SCJIL

CHEMICAL ANALYSIS

M. L. I AC11. S0 N
Professo r of Soils, Univers ity of Wiscons in
in
Agricul tural Experim ent Station, Madison , Wiscons

Pl\ENTICE-HALL, INC.
Engle wood Cliffs, N.J. · 1958
 ~1-1958, BY
PREN TICE- HALL. INC.
ENGLl\ \'OOD CLIFFS . N. J.

Al I RLSf R\'I D. NO f'.\R I OI,. THIS BOOK

Rllill IS
MAY BL RIPROD LICf() JN ANY FORM. HY MIMfO -
GRAPH OR ANY OTHI R Ml ;,;;s, WI
rt IOU I PFR-
PliBI ISHl-RS.
MISSICl' < IN WRITIN (, FROM ·1 HI

i .IBRARY OF ( ·oNtiRl· SS

CA I Al Oti C~Rl> NUMB ! R: -'~-(1932

A
PRINTE D IN TllE UNITED STATES OF AMERIC
82178
 Preface
follow sci-
As Sir Franci s Bacon wrote in his Of Studies, "Book s must
be measured
ences, not science books ." The maturity of any science may
raphs, and spe-
in the growth of integrative public ations -revie ws, monog
the older disci-
cialized textbooks. Soil Science is an infant compa red to
omy, physic s, chemis try, and medicine, but
plines of mathematics, astron
the field has been firmly established.
the trend toward integrative writing in
omy, and the monog raph series,
The annual review, Advan ces in Agron
oks both genera l and spe-
Agron omy, first appear ed in 1949 and textbo
ehensi ve treatm ent
cialized are available. The author hopes that the compr
Soil Chemi cal
of procedures and fundamental principles presented in
Analys is will foster progress in the science of soil chemistry.
as numerous
Methods applicable to the chemical analysis of soils are
of a chemical
and varied as the field of chemistry itself. The extraction
try, while the
constituent from soil is purely a procedure of soil chemis
constit uent is an analyti cal proces s, limited
determination of the extracted
of soil charac teristic s. Although
in range of methods only by consideration
filtrati on, centrif ugatio n, absorp -
the powerful techniques of fritted glass
sized, some proced ures are
tion and emission spectrophotometry are empha
included because they can be executed with simple equipm ent.
s are cited
The original publications of procedures and critical ·'!itudie
.,and im-
throughout Soil Chemical Analys is and the many modific'ati~ns
included: ·'fhe
provements developed by the author and his students are
the materials
free availability of information during the 20 years in which
many improv ements possib le. Great
for this book were accumulated made
proced ure and the author will
pains have been taken to eliminate flaws in
on to those inadve rtently
be grateful to readers who may call his attenti
missed ..
Dr. M. D.
The author wishes to extend grateful acknowledgments to
him to soil
Weldon of the University of Nebraska, who first introduced
nsin, who en-
science; to Professor E. Truog of the University of Wisco
Volk of Purdu e
couraged the writing of soil chemical methods; to Dr. N. J.
soils and soil
University, who fostered a broade ned outlook of the field of
of the Univer sity of Wisco nsin for his
chemistry; and to Dr. I. L. Baldwin
are extend ed to Dr. B. R.
warmth and encouragement. Special thanks
to Dr. J. L. White of Purdu e
Bertra mson of Washington State College,
tes who encour aged com-
University, and to the many other former associa
iii
iv PREFACE

pletion of this work. Appreciation is extended to Dr. W. R. Schmehl of

Colorado A. & M. College. to Dr. L. D. Swindale of the New Zealand
Soil Bureau, to Dr. A. W. Taylor of Rothamsted Experiment Station, and to
others who kindly read and supplied criticism of portions of the manuscript.

M. L. JACK SON
Madison, Wisconsin
 To the l\eader
s pro-
This book gives the most frequen tly used soil chemical analysi
researc h in soil chemis try, soil fertility ,
cedures , useful in instruc tion and
is essentia lly related to these lields,
and soil genesis. Becaus e plant growth
specialized
proced ures are given for plant inorgan ic constitu ents. More
exclude d in the interest of space
proced ures of these fields have had to be
s course will, later in re-
econom y. The student in a soil chemic al analysi
tion given. The teacher will
search, find a continu ing need of the informa
followi ng arc suggest ed
find time-sa ving discuss ions of princip les. The
weekly assignm ents for an underg raduate -gradua te course:

De I err11 ina I ions

Exercis e
number
I. Check desk. apparat us and reagents ; take soils samples
in field ( ~; 2-19)
or if weather does not permit. prepare samples issued by instruct or; in
case field samplin g to be introdu ced as soon as weather permits :
which
exercise on colorim etric pl-I indicato rs ( ~· 3-45. 3-46).
2. Determ ine soil pH with glass electrod e ( ~i 3-22): colorim
etric measure -
(~I 13-91) : thiocya nate test for acidity ( ~: 13-93) .
ment of soil pH
n of soils ( ~! 4-50); lime require ment of soils
3. Exchan geable hydroge
( ~ 4-64).
4. Neutral izing equival ence of limesto ne ( ~' 4-66).
5. Exchan geable metallic cations of soils ( ~. 5-1 or 18-24) .
6. Availab le phosph orus of soils (~ 7-69, 7-86, 7-99).
7. Organic matter of soils ( ~ 9-57).
8. Total nitrogen of soils ( ~i 8-13).
9. Ammon ia of soils ( ~' 8-33): nitrate of soils ( ~i 8-48).
10. Plant tissue analysis ( ~: 12-19) .
11.· Soluble salts of soils and waters (~I 10-25).
12. Rapid soil tests ( ~ 13-7 5).
13. Plant tissue test ( ~' 13-3).
14. Total phosph orus of soils ( •, 7-128) .
1

analysis
15. Total potassiu m and sodium of soils (~i 11-37. 11-177 ); silicate
of soils(~ 11-100 ).
16. Check in apparat us and reagent s; turn in reports; check
out.

v
 Contents
1
1. INTRODUCTION .
Y APPAR ATUS, 5 · LABOR A-
ANALY TICAi. Rf.AGr .NTS, 3 ·LABO RATOR
T 0UTl.I NE, 8 · LABOR ATORY 0RDFR LINES S, 8 ·
TORY REPOR
QUEST IONS. 9

") 10
--· SOIL SAMPLING .
MoNO l TTH MouNT JN<i OF Son,
Soll. PROFIL E SAMPL ING, 14
LISl!E[ ) EXPER IMENT AL
PROFIL ES. ! 8 . SAMPL ING Scm. OF EsTAB
SE! FCTION OF Son ExPFR l-
PLOTS . 22 · SAM Pl ING Son.s JN
MENT Fin.us , 25 . SAMPL ING FARM FIELDS , 29 . HANDL ING
THE LABOR ATORY , 30 . QUEST IONS. 36
Sou. SAMPL ES IN

3. HYDR OGEN ACTIVITY 38

DETERMINATION FOR SOILS

MEASU REMEN T OF Soll, PH. 41 . COLOR IMETR IC PH INDICA TORS

USEFU L JN Soll. ANALY SIS, 49 · Oxl!JA TION-R EDUCT ION INDI-
54 · QUEST IONS, 55
CATOR S USEFU L lN Soll ANALY SIS.

57
4. CATION EXCH ANGE DETERMINATIONS FOR
SOILS

EXCHA NGE CAPAC ITY DETER MINAT IONS, 59 · TOTAL

CATIO N
EXCHA NliEAB I E META LllC CATIO NS,
68 · GYPSU M OR SUI.FU R
71 . EXCHA NGEAB LE HYDRO GEN DE-
REQU! RFMEN T OF Sou.s,
73 · NEUTR ALIZIN G EQUlV AUoNC E OF AGRIC UL-
TERMI NATIO N,
0TURAL LIMES TONE, 77 · QUEST IONS, 80

r-::
a. EXCH ANGE ABLE METALLIC 82
CATI ON DETERMINATIONS FOR SOILS
NS WITH NH 4 0Ac, 84 ·
EXTRA CTION OF EXCHA NGEAB LE CATIO
OF CALCA REOUS Son.s, 88 . CALCI UM
EXCHA NGEAB LE CATIO NS
89 · MAGN ESIUM DETER MINAT ION, 97 · MAN-
DETER MINAT ION.
E DETER MINAT ION, J 02 · POTAS SIUM DETER MINAT ION,
GANES
106 · SODIU M DETER MINAT ION, 107 · QUEST IONS, I 09
vii
viii CONTENTS

6. POTASSIUM DETERMINATIONS FOR SOILS 111

POTASSIUM DETERMINA TION, 115 · COLORIMET RIC DETERMIN A-

TION OF POTASSIUM AS COBALT HYDROCARB ONATE, 119 • TITRI-
METRIC DETERMINA TION OF POTASSIUM , 122 • GRAVIMETR IC
DETERMINA TION OF POTASSIUM , 127 ·EXTRACTIO N OF EXCHANGE -
ABLE POTASSIUM FROM SOILS, I 28 • TOTAL POTASSIUM IN SoILS
OR MINERALS, 129 ·SOIL-POTA SSIUM EQUILIBRIU M, 130 • QUES-
TIONS, 133

7. PHOSPHORUS DETERMINATIONS FOR SOILS . 134

CHLOROSTA NNOUS-RED UCED MOLYBDOP HOSPHORIC BLUE COLOR

METHOD IN SULFURIC Ano SYSTEM (METHOD I), 141 . CHLORO-
STANNOUS- REDUCED MoLYBDOPH OSPHORIC BLUE COLOR METHOD
IN HYDROCHLO RIC Acm SYSTEM (METHOD II), 144 . MOLYB-
DENUM-RED UCED MoLYBDOPH OSPHORIC BLUE COLOR METHOD
IN SULFURIC Acrn SYSTEM (METHOD III), 146 • I, 2, 4-AMINO-
NAPHTHOL SULFONll ACID-REDU CED MOLYBDOP HOSPHORIC BLUE
COLOR METHOD IN PERCHLORI C Acm SYSTEM (METHOD IV),
148, • VANADOMO LYBDOPHOS PHORIC YELLOW COLOR METHOD
IN NITRIC ACio SYSTEM (METHOD V), 151 ·DILUTE Acm Sm.-
UBLE PHOSPHORU S OF SOILS, 154 ·WATER SOLUBI E AND PH 3
EXTRACTAB LE PHOSPHORU S IN RUNOFF WATERS, 157 ·FLUORIDE
EXTRACTAB LE PHOSPHORU S OF SOILS, 159 • HYDROXYL AND CAR-
BONATE EXTRACTAB LE PHOSPHORU S OF Son.s, 162 • DITHIONITE -
CITRATE EXTRACTAB LE PHOSPHORU S OF SOILS, 165 · ORGANIC
PHOSPHORU S OF So11.s, 169 • TOTAL ELEMENTA L PHOSPHORU S,
175 • PRECIPITAT ION AND IDENTIFICA TION OF ALUMINUM AND
IRON PHOSPHATE , 177 · PHOSPHATE EXCHANGE CAPACITY OF
Son.s, 178 • QuESTIONS , 181

8. NITROGEN DETERMINATIONS
183
FOR SOILS AND PLANT TISSUE

TOTAL NITROGEN DETERMINA TION, 183 · TOTAL AMMONIUM

AND NITRATE NITROGEN IN WATERS, 193 · EXCHANGEA BLE AM-
MONIUM DETERMINA TION, 193 ·NITRATE DETERMINA TION, 197
' NITRITE DETERMINA TION, 201 • NITRIFICAT ION RATE OF SolLS,
202 • QUESTIONS , 203

9. ORGANIC MATTER DETERMINATIONS FOR SOILS 205

ORGANIC CARBON DETERMINA TION AS co2

(DRY COMBUSTI ON),
208 • ORGANIC CARBON DETERMINA TION AS co~ (TITRATION ),
CONTENTS ix
Organic Matter Determinations for Soils (Cont.)
211 . OXIDATION MATTER BY CHROMIC ACID WITH EXTERNAL
HEAT APPLIED, 214 · 0XJDIZABLE MATTER BY CHROMIC ACID
WITH H 2 S04 HEAT OF DILUTION, 219 ·TOTAL ORGANIC MATTER
BY H 20 2 OXIDATION AND WEIGHT Loss, 222 ·QUESTIONS, 226

10. SOLUBLE SALT

ANALYSIS FOR SOILS AND WATERS 227
DIRECT SEMIQUANTITATIVE TEST FOR SOIL SALJNITY-CON-
DUCTOMETRICALLY ON SOIL PASTE, 229 · DETERMINATION OF
ELECfRIC CONDUCTANCE OF SOIL SoLUTIONS AND WATERS AS A
MEASURE OF SALT CONTENT, 234 · MEASUREMENT OF EFFECTS
OF Soll. SALINITY ON SEED GERMINATION, 251 · COMPOSITING
RUNOFF WATERS AND DETERMINATION OF TOTAL SUSPENDED
Souos OF RUNOFF WATERS, 253 . DETERMINATION OF TOTAL
DISSOLVED SOLIDS OF Sou. EXTRACT OR WATER, 256 . DETER-
MINATION OF lNDfVIDUAL CATIONS AND ANIONS OF SOLUBLE
SALTS IN SOILS AND WATERS, 256 ·DISSOLVED CARBONATE AND
BICARBONATE DETERMINATION, 260 · CHLORIDE DETERMINA-
TION, 261 · SULFATE DETERMINATION, 263 · GYPSUM DETER-
MINATION FOR Sou.s, 266 . CARBONATE CARBON DETERMINATION
FOR SOILS, 268 · QUESTIONS, 270

11. ELEMENTAL ANALYSIS OF MINERAL

COLLOIDS, SOILS, MINERALS, AND ROCKS 272
SEMIMICROCHEMICAL SYSTEM OF SILICATE ANALYSIS, 278 · Po-
TASIUM, SODIUM, CALCIUM, AND MAGNESIUM DETERMINATIONS,
285 · IRON AND TITANIUM DETERMINATIONS, 291 · SILICON DE-
TERMINATION, 294 · ALUMINUM DETERMINATION, 297 · CON-
VENTIONAL SYSTEM OF SILICATE ANALYSIS, 300 · SILICA DETER-
MINATION, 303 · CUPFERRON SEPARATION OF IRON AND TITANIUM,
307 · NH 4 0H SEPARATION AND ALUMINUM DETERMINATION,
309 · IRON DETERMINATION, 311 · TITANIUM DETERMINATION,
314 ·CALCIUM DETERMINATION, 315 ·MAGNESIUM DETERMINA-
7JON, 316 ·MANGANESE DETERMINATION, 318 ·POTASSIUM AND
SoDIUM DETERMINATIONS, 318 ·FERROUS IRON DETERMINATION
IN SILICATES, 320 . TOTAL SULFUR IN Son.s, 322 . MISCELLA-
NEOUS CoNSTITlJENTS, 323 · QUESTIONS, 324

12. PLANT TISSUE ANALYSIS-MINERAL

CONSTITUENTS 326
TOTAL ASH CONTENT OF PLANT TISSUE, 330 · WET OXIDATION
OF PLANT TISSUE, 331 · ELEMENTAL ANALYSIS OF RESIDUE
x CONTENTS

Plant Tissue Analysis-Mineral Constituents (Cont.)

FROM WET OXIDATION OF PLANT TISSUE, 335 ·SULFUR CONTENT
OF PLANT TISSUE, 336 ·CYANIDE IN PLANT TISSUE, 337 · FLUO-
RIDE JN PLANT TISSUE, 338 · QUESTIONS, 338

l:i. RAPID CHEMICAIA TESTS

OF SOILS AND PLANT TISSUE 339

NITRATE NrrnoGl'N IN SAP OJ- PLANT TISSUE, 342 . PHOSPHORUS

IN SAP OF f>LAN'J TISSUE, 347 · POTASSIUM IN THE SAP OF PLANT
TISSUF. 350 · QUICK SOIL AERATION TESTS. 354 · Soll. TESTING
SYSTEMS. 357 · QULSTlONS. 368

14. BORON DETERMINATIONS

FOR SOILS AND PLANT TISSUE 370

BORON DETERMINATION (WITH CURCUMIN), 372 · BORON DE-

TFRMINATION (WITH QUINALIZARIN), 375 · WATER SOLUBLE
BoRoN JN So11 s, 379 · AciD SOJ.UBLF BoRoN, 382 · ToTAL
BORON IN So11.s. 384 . TO'IAL BORON IN PLANTS, 385 . QUES-
TIONS, 387

1r}. IRON, MANGANESE, COPPER, ZINC,

MOLYBDENUM, AND COBALT DETERMINATIONS 388

IRON DETFRMINATION. 389 · EXCHANGLABI F FFRROUS IRON.

391 · COMBINED EXCHANGEABI F FERROUS AND FLRRIC IRON
AND DILUTE ACID Srn.u11LF IRON, 392 · Sm UBLF, Exc11ANGE-
ABLF. AND EASll Y REDUCIBLE MANGANESE OF SOILS, 393 · COP-
PER DETERMINATION. 396 · TOTAL COPPER AND ZINC EXTRAC-
TION FROM So11.s, 398 . EXTRACTABLE COPPER IN So11.s. 400 .
ZrNc DETJ:RMINATION, 402 · DrT111zoNE ExTRACTABI E So11.
ZINC, 406 · MOLYBDENUM DETERMINATION, 408 · TOTAL Mo-
LYllDENUM IN SOILS AND P1 ANTS. 411 · CoBAt T Dt·TERMINATION,
414 · QUJ'SflONS, 415

1(). POLAROGRAPHIC ANALYSIS

FOR SOILS AND PLANT TISSUE 416

COPPER AND ZINC DETERMINATION. 421 · MANGANESE DE-

TERMINATION, 426 · COPPER AND ZINC SEPARATION FROM RESI-
DUE OF WET 0X1DATION OF PLANT TISSUE, 427 · QUESTIONS, 428
CONTENTS xi

J7. ABSORPTION SPECTROPHOTOMETRY 429

CONCENTRATION DETERMINATION BY ABSORPTION SPECTRO-

PHOTOMETRY, 438 · ABSORPTION SPECTRUM DETERMINATION,
446 · QUESTIONS, 451

18. EMISSION SPECTROPHOTOMETRY . 452

DETERMINATIONS WITH FLAME EMISSION SPECTROPHOTOMETER,

455 · QUALITATIVE ANALYSIS WITH ARC OR SPARK EMISSION
SPECTROGRAPH, 465 · QUANTITATIVE ANALYSIS WITH THE ARC
OR SPARK EMISSION SPECTROGRAPH, 474 · QUESTIONS, 485
 List of Tables
1-1 Strength of "concentra ted" acids and bases . 3
2-1 Weights of representat ive furrow slice soil volume sometimes em-
ployed in expressing results of soil chemical analysis . 12
2-2 Relation of sieve size to minimum sample volume and weight. The
optimum sample size is on the order of 3 or 4 times the minimum 33
3-1 Selected list of colorimetri c pH indicators useful in soil analysis 52
3-2 Determined pH values of various solutions prepared with l gm
of solid mixed with I 00 ml of water 54
3-3 Three useful colorimetri c oxidation-r eduction indicators . 55
5-1 Equivalence as pp2m of I meq of exchangeab le cation per 100 gm
of soil and representat ive amounts in a fertile soil . 84
7-1 Summary of characteris tics of 5 spectropho tometric methods for
phosphorus described in text as methods l to V . 138
7-2 Appropriat e concentrati on ranges of phosphorus standard solutions
for various analytical methods 141
10-1 Approxima te amounts of salts in soils of various textures with a
given electrical resistance of soil paste in standard soil cup . 232
10-2 Approxima te salt content of water with given resistances in stand-
ard soil cup 233
10-3 Various units which have been employed for electrical conductanc e
of soil solutions and other waters . 236
I 0-4 Quantity of soil to be taken for each ml of saturation extract in
relation to the soil texture and moisture properties . 241
10-5 The salinity scale 244
10-6 Relationshi p of specific conductanc e in 1 : 2 soil: water extract to
that in the saturation extract for 2 soil textures . 250
I 0-7 Guide to quality of irrigation water . '257
12-1 Concentrat ion of some mineral elements in crop plant tissue 329
13-1 General character of several soil testing systems 361
14-1 Appropriat e concentrati on ranges of boron standard solutions for
curcumin and quinalizarin methods 374
17-1 Record form for absorption spectropho tometric measureme nt of
concentrati on . 444
18-1 Wave lengths and relative intensities of flame spectra determined
with the Beckman flame spectropho tometer . 456
18-2 Conditions with Beckman flame emissions spectropho tometer for
determinat ion of several cations excited in oxygen-acetylene flame 461
xiii
xiv LIST OF TABLES

18-3 Spectral lines employed for quantitat ive determin ation of a number
of elements by arc emission, together with the appropri ate internal
standard s 483
Periodic Table of the Elemen ts-Fron t Endpape r
Internati onal Atomic Weight s-Front Endpape r
Log Table-R ear Endpape r
 1

Introduction
The soil is a medium of great complexit y, and soil science has
not progresse d very jar.
-RICHAR DS 1

1-l. Soil Presents a Complex Analytical System. The analysis of soils

and plants presents an interestin g challenge to a prospecti ve soil scientist.
In his Fisher award address, H. H. Willard~ states,
The number of chemical reactions. the types of apparatus used. and the
varied technique s utilized by the analytical chemist today makes one who is
broadly trained in this field valuable in solving problems quite outside of
analytical chemistry .
This challenge is acute because a soil consists of an extraordi narily complex
chemical mixture of different mineral and organic substance s. The plant
presents a special type of soil extractio n. Soil chemical analysis deals (as
determin ations of soil elements or indirectly as reagents and equipme nt)
with over 60 of the naturally occurring chemical elements (Fig. 1-1). Ad-
ditional soil elements continual ly become of interest in plant nutrition, in
relation to essentiality or toxicity, or for physiological substitution. As
many as half of the remainin g 32 naturally occurring elements may be in-
volved· ultimately one way or another in soil and plant analysis. Appro-
priate chemical analysis systems have been develope d for quantitat ive
analysis of the mineral elements present. but the determin ation of mineral
structure requires an advanced system of mineralo chemical analysis. Like-
wise, the organic matter in soil or plants growing on or applied as residues
to soil is an analytica l field in itself. But the total organic matter of soil can

I Diagnosis and Improvem ent of Saline and Alkali Soils (Riverside, Calif.: U.S.
Salinity Lab., 1947). p. iii.
2 Anal. Chem., 23: 1726 (1951).
 INTRODUCTION
2

He 2
0.
Li Be
B C F Ne 8

ta~
K Ca Ti v
.. ~
Al
.. . . . ..
Si

Co Ni Cu Zn Ga Ge As Se Br
Cl A 8
16

Sn Sb I 11
Rb Sr Zr Rh Pd Ag
W Re Os Ir Pt Au Hg Pb Bi Rn 13
Cs Ba La • 2
Ra Ac

Ce
2
Th u
63

Fig. 1-1. Elements employed in soil chemistry

as determinants or in-
al to
are essenti
directly as reagents and equipment. Elements in boxes
plants. A complete chart appears inside the back flyleaf.

or by reactions with
readily be estima ted from the total organic carbo n
strong oxidizing solutior.s.
analyzed, the pro-
1-2. Because of the compl exity of the mater ial to be
r Presid ent Cham -
spective analys t should consid er the mand ate of forme
forth by Rich, 4
berlin:1 of the University of Wisconsin, as so aptly set
ve, and the best guide
To be efficient, the gather ing of data must be selecti
metho d of multip le workin R hypoth eses
to its collect ion is probab ly the
many years ago by Chamb erlin. The availab le inform ation is first
propos ed
g hypoth eses are formu lated; from
studied and analyz ed; severa l workin
are deduc ed the conseq uences which should follow if that
each of these
are then sough t and the old data re-
hypoth esis were correc t; furthe r data
ce which would either bear out or refute the hypoth esis
exami ned for eviden
under scrutin y.
a minim um of blind
When search for data is thus guided . there will be
clutter ing of the literat ure with
collect ion of irrelev ant facts and a minim um
undige sted factua l inform ation. . . .
of prope r foreth ought
These principles of research indicate the impor tance
tissue analys es.
befor~ launch ing into a set of soil or plant
tion betwe en the proce dures
1-3. The analyst should make a sharp distinc
soil or plant (whic h is a
of extracting a chemical consti tuent from the
and determ ining the con-
pheno menon purely of soil or plant chemi stry),
r solely of analyt ical
stituen t once extrac ted. The determ inatio n is a matte
ds, excep t as limite d
chemistry, with conse quent range in possible metho
ring substa nces that
by consid eratio n of conce ntratio n range and interfe
Newly develo ped instru ments contin ually broad en the
might be presen t.
:11. Geo/ .. 5:837 ( 1897).
4 Sci., 107:58 1 ( 1948).
 3
INTRO DUCT ION
ely be
field of soil or plant analysis. However, some proced ures can effectiv
the absence of more speciali zed equip-
execute d with simple equipm ent, in
when a
ment. For example, __p_o._~~sstum may. be determ ined by cobaltinitrite
le. Powerf ul techniq ues in-
flame emission spectro photom eter is not availab
fritted glass filtratio n,
clude absorpt ion and emission spectro photom etry,
of several constitu ents.
centrifugation, and systematic schemes of analysis
n, and
These avoid the time-consuming steps of evapor ation, paper filtratio
ure for a
heating to constan t weight. The analyst seeking a suitable proced
to learn the
given determ ination should consult first textbooks on analysis
differen t
range of choice and some of the advantages and disadvantages of
journal
procedures. Sometimes he will need to consult abstrac t journals and
articles for any recent advanc es on a particu lar determ ination .

ANAL YTICA L REAG ENTS

purest
1-4. Reagen ts are supplied commercially in different grades, the
the second, "C.P." and the
being "analyt ical reagent " or "reagen t grade,"
various grades has a distinct
third, "techni cal" or "U.S.P ."Each of the
chemist
purpos e and range of uses for which it is satisfactory. The alert
each speciali zed
will find that there seems to be a suitable reagent for
analytical function.
cture,
1-5. Concen trated Acids and Bases of Commerce. During manufa
bases of
it is custom ary to express the strength of concen trated acids and
ling
liquid form in terms of specific gravity because it is used for control
volu-
their strength and because specified weights can be approx imated
the laborat ory, their strengt h is best express ed on
metrically. However, in
the basis of chemical equival ence or normal ity (Table 1-1).

TABLE 1-1

Strength of "concen trated" acids and bases

·------ -----"·"-''- ---·--·----------
Concent ration Approxi mate
Normali ty Per cent by weight specific gravity
Reagent
11.6 37 to 38 1.19
HCI
35 to 36 97 to 100 1.84
H~S04 .
17.5 99.5 I. 13
HO Ac
16 70 to 71 1.42
HN0:1
9 to 11.6 60 to 70 1.51 to 1.67
HCI04
45 85 1.71
H~P04
15 28 to 29 0.90
NH40H
55 100 1.00
HOH

ed sub-
1-6. Distilled Water. Water freed to varying extent of dissolv
chemic al analysi s. The quality varies from
stances is essential for all
in a copper or tin still to double or triple
simple distilled water conden sed
4 INTRODUCTION
distilled water. A number of treatments are sometimes applied to the
water between distillations to effect oxidation of organic substances and to
suppress the passage of ions into the distillate. The substances employed
for oxidation include !5.~n0 4 or Br". Sometimes even for minor element
work, the_redistillation of ordinary distilled water in <1..J~,yr.ex glass" still
makes it sufficiently free from metallic ions for analytical purposes.
1-7. Sometimes condensation of high pressure steam coupled with some
filtration of the gas produces condensate of a quality suitable to he used as
distilled water. Steam condensate usually contains volatile oils and to some
extent suspension of solid particles carried from the steam lines. Successful
steam condensation as a substitute for distilled water was reported by
Margolis." A commercial high pressure steam condenser designed by
Truog is described by Stark.'1
1-8. Commercially available ionic columns are able to produce water of
sufficiently low ionic content for analytical work. Commercial sources in-
clude La Motte Chemical Products Co. (Towson, Baltimore, Md.), Wilken-
Anderson Co. (Chicago, Ill.), Sargent Co. (Chicago, Ill.), and Enley
Products, Inc. (254 Pearl St., New York 38, N.Y.).
1-9. Filter Paper. The several types of filter paper vary greatly in their
content of total ash and in their content of major and minor elements.
There is no substitute for experimental check on contamination; the ques-
tion of paper purity parallels that of reagent purity. Filter paper pulp often
is a useful expediter for filtrations through paper. Several types of filter
paper in a range of suitable porosities have been prepared by commercial
manufacturers. The leading brand names of filter paper. available from
the usual chemical supply houses, are Whatman, Munktells, Schleicher and
Schuell, and "E and D" (Lapine Co., Chicago, Ill.).
J-10. Chromic Acid Cleaning Solution. Chromic-sulfuric acid cleaning
solution is valuable for the final cleaning of glassware. Visible materials
and organic solvents should be rinsed out with water before using the
cleaning solution.
J-J I. The solution is made up by dissolution of 80 gm of K:!Cr:!0 7 or
Na:!Cr:!O, in 300 ml of water (with heating). The aqueous solution is
placed in a Pyrex container, and one liter of technical grade H:$0 1 is
added cautiously with stirring. Considerable red chromic oxide (Cr:!OJ
precipitates.
1-12. A cleaning solution which does not involve Cr~0, 1 crystallization
is made by dissolution of 5 gm of K"Cr"0 7 in a minimum of water and
addition of this solution to one liter of technical grade H~S0 1 •
1-13. Aqua Regia. The acid oxidant aqua regia is prepared by mixing

"Sci .. 115:552 ( 1952\.

6Modem Ho"p .. 43(3) (Sept. 1934).
INTRODUCTION 5
3 volumes of concentrated HCI with 1 volume of concentrated HN0 3 • Free
Cl2 is liberated.

LABORATORY APPARATUS
1-14. A. N. Whitehead states, "In science, the most important thing
that has happened is the advance in instrumental design." The soil chemist
is dependent on all manner of physical instruments, from analytical balance
to spectrophotometer, for the chemical analysis of soils. Fortunately for
efficiency in analysis, apparatus has been developed to a high degree, from
the resistant glass beakers to flame emission equipment.
1-15. Analytical Balances. The beginning student should review his
quantitative chemical text on the care and use of the analytical balance. He
should know the "method of swings" for calculating the rest point, the
method of checking for unequal balance arms, the method of calibration
of weights, and the theory and practice of balance weights. These matters
have been summarized concisely by MacNevin, 7 Swift, 8 and others.
1-16. Analytical balances differ in complexity from the simple balance,
which requires gram weights, fractionals, and a balance beam rider, to the
chainomatic balances, which require no fractionals or rider, to the key-
board balance, which requires only the heavier weights, and finally to the
Gram-atic balance, which has but one pan, has no weights to be handled by
the analyst, and automatically gives complete weighing to four places.
The analyst should have a chainomatic balance or better. He should also
have a torsion balance sensitive to 0.05 gm and a solution balance that will
weigh up to 10 kgm or more.
1-17. Chemical Glassware. The technology of chemical glassware has
been developed to a high state. Different glasses are available in varying
compositions and physical properties to meet the requirements of nearly any
of the common analyses. The analyst should review the composition of the
glass in relation to the analysis to be made. He should recall that Pyrex
glass is a borosilicate glass that contains arsenic; that it is suitable for most
laboratory operation, but unsuitable for boron determination, and a pos-
sible source of arsenic in an arsenic determination. On the other hand,
Cavalier or Coming 728 glassware is boron-free. Ordinary soft glass is
relatively low in boron and frequently can be cleaned sufficiently to con-
tain reagents for boron. It is also suited for some of the phosphorus re-
agents. However, soft glass is readily attacked by neutral or slightly alkaline
solutions, and thus is a ready source of sodium contamination. Care in
selection of the composition of the glass, coupled with adequate cleaning
and testing for contamination, are essential.

7The A nu/ytica/ Balance (Sandusky, Ohio: Handbook Publishers, Inc., I 951).

8Introductory Quantitative Analysis (Englewood Cliffs, N.J.: Prentice-Hall, Inc.,
1950).
6 INTRODUCTION

1-18. Volumetric glassware is now available with high standards of

calibration and utility. Standard-taper volumetric flasks are preferred.
Burcts arc availahle with calibration lines extending completely around for
elimination of parallax. Notice should be paid to whether the volumetric
pipette is calibrated to deliver (T.D.) or to contain (T.C.).
1-19. Filters Composed of Glass and Other Mineral Substances. Al-
though paper filters ( ~ 1-9) have been improved, purified, and toughened,
for many purposes there is no substitute for the mineral, or inorganic,
types of filter. The most commonly used is the fritted glass filkr, manu-
factured in various porosities and in various sizes, fitted in the bottom of
different types of crucibles and into various shapes of funnels. For semi-
micro work the 1iltcr tube (Fig. 5-2), which is a very ~mall funnel with
the top enclosed with porous glass and from which liquid is drawn through
the stem, is indispensable. In addition to glass filters there arc clay and
alundum filters of varicus porosities. Alundum powder, ''cclite,'' and other
mineral filter aids arc often useful.
1-20. Centrifu~e Washin~ in Lieu of Filtration. Alternative to the u~c of
filters is the washing of precipitates with the centrifuge. Elimination of fil-
tration hy the use of centrifuge tubes and centrifugation is especially ad-
vantageous for removal of precipitated clements which would other-
wise interfere with the determinations of ions in solution. The suspen-
sion may he simply diluted to volume in a centrifuge tube, clarified, and an
aliquot then taken of the supernatant liquid for analysis (for example,
~ 11--40).
1-21. For the analy~is of the precipitate wherein ions in solution inter-
fere, the centrifuge tube washing is not so advantageous. The filter tube is
ordinarily more rapid and convenient in this case.
1-22. Cutting of Large Glass Cylinders. There arc two techniques for
cutting large glass cylinders that seem to he more effective than the com-
mon hot-wire procedure. In one a large glass cylinder is surrounded by

Method J Method 2
-~
Wet paper strips
i!J' v /'--
___!~ ' "
I ,
Red hot rod
'-,Scratch
W1ngtop

Break
Burner occurs..._

Fig. 1-2. Cutting large glass cylinders and hottles.

INTRODU CTION 7
two wet paper strips (Fig. 1-2) placed on either side of a light scratch.
Heat is applied at the scratch with the wing top burner as the tube is
rotated. The cooling by the wet paper strips localizes the heat and gives a
clean cut of a rather large glass tube, 2 to 10 cm in diameter.
1-23. The second method, described by Dr. C. B. Tanner of the Uni-
versity of Wisconsin , is mainly applicable to large cylinders, including
bottles. The cylinder is filled with mineral oil to the depth at which the cut
is wanted (Fig. 1-2). Then a red hot iron rod is thrust into the oil causing
hot oil to flow out over the surface to produce a crack.
1-24. Mortars and Pestles. A mortar and pestle is a vital aid to the
analyst. The agate mortar and pestle of a suitable size (Fig. 1-3) is com-

Fig. 1-3. Agate mortars and pestles of suitable size are indispensab le to the a nalyst.
(Courtesy Eldot and Co .. 33 West 60th Street, New York.)

monJy used. Newer materials for mortars and pestles include mullite and
synthetic sapphire. The synthetic sapphire mortar has been recommended
for preparations for spectrographic analysis. Mortars made of this material
have been described 9 as being "hard as tungsten carbide." The porcelain
mortar is generally considered to introduce too much contamination for
most soil analysis work except for the preparation of bulk soil samples for
passing relatively coarse screens.
1-25. Sieve Openings versus Meshes per Inch. In stating the size of sieve
openings, one widespread custom is to specify the number of meshes per
9 The Laboratory, 18 (5): J31 (Pittsburgh: Fisher Scientific Company, 1949 ) .
8 INTRODUCTION
in
inch, without reference to the actual size of the opening. The opening
mm can be fairly well approx imated on the assump tion that the opening
is 0.63 of the mesh interval, whence,
. 16 (1-1)
mm per openmg = meshes per me . h

For example, a "l 00-mesh" screen has openings of 0.16 mm.

LABORA TORY REPORT OUTL INE

fol-
1-26. The report of each laboratory is to be developed around the
8.5 by 11 inches in size, and
lowing outline, to be submitted -on sheets
is not
typewritten (or neatly and legibly written in ink if a typewriter
available) .
I. Numbe r and title of experiment.
II. Objective.
III. Methods employed.
A. Princip les-Prin ciples studied are to be outlined; all chemica
l equations
involved are to be given.
ns is to
B. Appara tus-An y apparat us not already pictured in the directio
be sketched.
C. Proced ure-Gi ven by reference to directions, with careful
notes on any
departures from given procedure.
IV. Data-T he original laborato ry record is submitted in the data section.
indi-
A. Sample calcula tion-A sample calculation is given in outline form,
letters or words for the function s. All notation s
cating each step, with
one
are defined and all dimensions given. Finally, numerical data for
com-
sample calculation are substituted in the equation and the result
puted.
shown
B. Original as well as final comput ed results are tabulated and also
in graphic form.
and
C. Analysis of error-A n estimate is to be made of significant figures
the extent of inheren t error.
up. Answers
V. Conclu sions-T he significance of the experiment is summed
are to be written for question s supplied by the instruct or.
to the experi-
VI. Abstra cts-A brief abstract of one or more articles pertaining
ment may be assigned by the instruct or. The abstrac t will be submitted in
typewritten form, usually 0.2 to 0.5 page in length.

LABO RATO RY ORDERLINESS

uni-
1-27. Cleanliness and Orderliness in the Laboratory. These are
ory
versally conceded to be requisite for high efficiency of the laborat
and by his prospec tive
worker. Each student is judged by his associates
fol-
employer on the condition in which he maintains his laboratory. The
lowing general rules are presented as a guide:
-
1. Dirty glassware should be washed promptly, and stored away immedi
after it has drained . It should not be left on the drain board for more
ately
INTROD UCTION
9

than the few minutes required to complete the washing job, never left there
to dry.
2. All glass and other container s of solutions will bear a label of the
contents and the initials of the user.
3. Each student will discard his samples, clean the container s and return
them to the stock room.
1-28. General Laboratory Tables and Equipment. These need to be
cleaned periodica lly. Each user is responsible for leaving them clean. In
addition, cooperati ve arrangem ents will be made by all students for assign-
ment of cleaning special equipme nt to individua l students.
1-29. Everyone will profit by experienc e in a well-orde red laborator y
where high standard s of orderline ss and cleanline ss will be establish ed
for laborator ies placed in your charge in the future. It will pay good divi-
dends. Rememb er, your prospecti ve employer cannot see at a glance the
inner workings of your mind, nor the work you have done, but he can see
at a glance the condition of the space assigned to you. He can judge your
ability to organize by the way you organize your equipmen t.

QUESTIONS
1. Outline the functions of textbook treatment of analytical determinations.
2. Outline what should be done in making a selection among established
analytical procedures.
3. Explain how each element in the third period (or another period selected)
of the periodic chart is of concern in soil chemical analysis.
4. What are the normality and molarity of common acids and bases of com-
merce? What is the molarity of H 2 0 in water?
5. What are the chief impurities in distilled water made by condensation of
boiler steam?
6. Discuss several ways in which soil chemical analysis (and chemical anal-
ysis in general) has been aided by improved physical instruments; give several
examples.
7. Make an outline of the different kinds of filtering devices and their relative
advantages and disadvantages.
8. Describe two methods for cutting large glass cylinders.
9. Outline the kinds of mortars and pestles as to substances from which
made.
10. What is the approxim ate conversion to sieve size in mm from "meshes
per inch"?
11. What are the chief advantages to the research student of keeping a neat
laboratory ?
 Soil Sampling
The analysis can be no better than the sample.
-AN AXIOM

2-1. The sampling of soils is a challenging problem, worthy of detailed

consideration. The general problem of soil sampling has been summarized
by the Association of Official Agricultural Chemists 1 as follows:
In view of the variability of soils, it seems impossible to devise an entirely
satisfactory method for sampling. It is obvious that the details of the pro-
cedure should be determined by the purpose for which the sample is taken.
This passage brings into focus the twin problems of soil variability (sam-
pling "all out-of-doors") and the divergency of types of analyses to be per-
formed on samples. Soil sampling includes taking the soil material so as to
take into account the variability of soils, handling and processing the
sample, and final subsampling for the actual analytical determination.
2-2. Soil sampling must take into account variations of soil according to
profile depth and landscape area. These may be viewed in terms of natural
soil type units of soil development or in practical units of farm, field, or
plot size.
2-3. Importance of the Soil-Volume Concept. The soil body sampled
should be referred to as the sampling volume, rather than as the sampling
area. Farmers till the soil on a volume basis-the furrow slice. Chemical
analyses are often reported on that basis (Table 2-1 ) . Soil physicists refer
to percentage of moisture and air by volume. Soil geneticists classify the
soil three dimensionally (Fig. 2-1). Crop yields are produced by the entire
soil volume, the roots deriving moisture and nutrients from the entire depth

1 Official and Tentative Methods of Analysis, 6th ed. (Washington, D.C.: A.O.A.C.,
1945), p. 1.
10
SOIL SAMPLING 11

.,
·;;
N"'

LIMB CONCAVE L I M J
..
[ A TOPOSEQUENCE

- - - Length - - - - -

Fig. 2-1. Soil associations in a landscape viewed with reference to

three coordinates in space designated the X, Y, and Z landscape axes
extending from any designated point of origin, 0. The soil profile is
the landscape viewed as horizons developed anisotropically along the
vertical or Z-axis, the soil depth function, with depth units a. The con-
tour line follows, at any point, the direction of the X-axis, along which
the soil profile displays isotropism. The Y-axis is oriented with the
slope of the soil surface and perpendicular to the relief contour. The
angle, IJ, is the slope, conventionally expressed in per cent, 100 c/b. A
toposequence in a given landscape may have a characteristic length
and height (ht.) as a whole and of the convex and concave limbs. The
soil toposequence (~ 2-10) is the landscape viewed primarily in the
two dimensions, along the Y and Z-axes, and is of great importance
to all kinds of soil sampling. (Courtesy Dr. F. D. Hole.)

of the profile. Occurrence of gravel at a shallow depth has led to some of

the locally known "Poverty Ridges" because productive soil material was
too shallow. The deep loess in central Nebraska permits penetration of
alfalfa roots to a depth of 8 or t 2 meters and furnishes available water
throughout that depth at a rate of 1 cm of water to 6 cm of soil depth. Some
soils have most of their exchangeable potassium in the "plow layer"; the
deep loess referred to has about 500 to 1000 pp2m in each plow-layer
thickness throughout the 10 or more meters of depth.
2-4. Principal Sources of Variability of Soil Chemical Analysis. The
principal sources of variability in the chemical analyses of samples from
a given soil are:
12 SOIL SAMPLING

TABLE 2-1
Weights of representadve furrow slice soil volumes sometimes employed
in expressing results of soil chemical analyses
Weight of furrow
slice volume
Bulk Per hectare Per acre Factor
Soil density, ( 17 cm), (6.7 in.), from
texture gm/cm~ kgm pounds ppm
2,220,000 2,000,000 2.0
Silt or clay loam 1.3
2.5
Sands 1.6 2,770,000 2,500,000
550,000 500,000 0.5
Peats, mucks 0.32
--------------------------------
1. Variability among different samples drawn from the same volume
(sampling "error").
2. Variability introduced among subsamples of the same soil sample
(subsampling "error").
3. Variability from one chemical determination to another on the same
subsample (analytical "error").
Procedur es are well establish ed for decreasing the variability due to sub-
sampling and variability due to analytical procedur es. Thus, the main
source of variability in the analytical results stems from variability of the
soil samples. The standard error arising from sampling and sample treat-
ment has been shown 2 to be 3 to 6 times greater than those arising from
the subsampl ing and analytical procedur es, even when the soil sampling
procedur e in the field was much more refined than is common. Samples
taken at 6-inch intervals in 24 square feet of a uniform, virgin, red-brow n
earth soil area showed 3 coefficients of variation of 10 per cent for 6 metal-
lic cations (rarer elements ), and the highest concentr ation of each element
was twice the lowest among the 68 samples taken.
2-5. Composite Soil Sample Equivalent to an Average. A composit e soil
sample gives a mean analytical value represent ative of the soil sampling
volume from which the composit e sample was drawn. (Any analytical
value is a mean for the individual soil particles, hence the futility of an argu-
ment against obtaining mean-val ue analyses. ) Analyses 4 for carbon, nitro-
gen, phosphor us, and for soil pH values 5 made on composit e samples have
been found (as would be expected ) to be equivalen t to the mean of analyses
of individua l cores. The standard deviation from the mean can be calcu-
lated from separate core analyses; however, individua l cores are subject

2 Cline, Soil Sci., 58:275 (1944); Robinson and Lloyd, J. Agr. Sci., 7:144 (1915);
Waynick, Univ. Calif. (Berkeley), Pubs. Agr. Sci., 3:243 (1918).
~McKenzie, Australian J. Agri. Res., 6:699 (1955).
4 Robinson and Lloyd, U.S.D A. Cir. 139 (1939); Waynick and Sha.rp, Univ. Calif.
(Berkeley), Pubs. Ag•i. Sci.,4:120 (1919).
o Chapman ~t al., S.S.S.A. Proc., 5: 197 (1941 ).
SOIL SAMPLING 13
to large variations which are not significant to plant growth individually.
The extensive volume through which roots of one plant spread, together
with the interlacing of root systems of various crop plants growing in as-
sociation, severely limits the degree of dependence of the crop on chemical
properties of the soil in any given soil core. Disproportionate uptake oc-
curs, of course, from a soil volume of a few cm diameter that has a high
nutrient concentration, such as a fertilizer band, but this has been shown
by radiochemical tracer experiments to be less extensive than is sometimes
supposed. Growth restrictions in pots and small lysimeters further testify as
to the dependence of plants on rooting extensity. The root spread volume of
many field crop plants is a cylindrical soil volume of 1 to 2 meters radius
and of 1 to 10 meters depth. Crop plant spacing is ordinarily only 1 to 10
per cent of the land area over the root spread volume of each plant. Root
contact with the side of each cubic soil volume of 1 to 2 mm on an edge
can be observed in the field, illustrating the intensity and extensity of root
coverage of the soil. Composite sampling of soils restricts the samples to be
analyzed to those useful as a basis for soil management recommendations.
2-6. Requirements of Composite Sampling Procedure. The fundamental
requirements of valid composite sampling are:
l. Each core should be of the same volume and represent the same cross
section of the sampling volume.
2. The cores should be taken at random with respect to the sampling
volume, usually restricted (~ 2-35) to criss-cross the direction of cultural
operations and natural trends of change such as slope.
3. Enough cores should be taken to represent the whole sampling vol-
ume adequately.
4. There should be no chemical interactions of soil material composited
that are significant to the objective.
5. The soil unit selected for one composite sample should be homogenous
for the objective of the analysis, for example, division of a field into several
areas on the basis of observed or otherwise known heterogeneity permits
analysis of a sampling volume for 'each field area the farmer should treat
separately.
2-7. Number of Cores Required for a Composite Sample. The
A.O.A.C. 6 states that the sample should include " . . . a sufficient number
of [cores] to insure a composite sample that will be representative of the
tract sampled." Increasing the number of cores lowers the variability of the
sample characteristics. Standard statistical tables in books give the number
of replications (cores) to composite in terms of the variability of the analy-
ses of individual cores and the level of probability sought. In practice, com-
positing 20 to 30 cores usually narrows 7 the distribution curve down until
it has the same shape as the analytical distribution curve.

6 Official and Tentative Methods, 6th ed. (Washington, D.C.: A.U.A.C., 1945), p. 1.
7 Reed and Rigney, J. Am. Soc. Agron., 39:26 (1947).
14 SOIL SAMPLING

SOIL PROFILE SAMPLING

2-8. The generalizations as to the uniformity and differences in soils are
systematized by soil classification specialists and embodied in soil type
definitions. A soil chemist should always work in close cooperation with
soil classification specialists in sampling soil profiles.
2-9. Objectives. Chemical analysis of soil profiles generally are made for
information on the chemical processes of soil development and long range
soil fertility. The soil profiles to be sampled are selected:
1. To represent agriculturally important soils.
2. To represent soil development factors functionally.
3. To represent the sequence of mineral weathering functionally.
2-10. Selecting soil types to represent the soil development and chemical
weathering factors functionally simplifies the interpretation of results and
provides the maximum of useful information per determination on the proc-
esses responsible for the soil properties. It is important to recognize that the
variation of one factor, such as climate or parent material, inevitably in-
volves significant changes in other more or less "independent" factors, par-
ticularly the nature of the vegetation. Bray 8 studied profiles varying in
over-all degree of development, termed a maturity series. It is impossible
"to hold all other factors constant" while varying one. Rather, the change in
the soil type is observed as a function of the change in a developmental or
weathering factor, with the recognition that other factors also change as a
function of the factor under study. A sequence of soil types is selected to
isolate the effect of a single factor in so far as possible. Examples of such
sequences9 follow:
1. Soil types in toposequence (Fig. 2-1 ) , such as a developmental se-
quence from well-drained soils to planosols on a uniform loessial parent
material, or well-drained to poorly-drained soils in the tropical red and
black complex.
2. Soil types in climosequence, such as soils in successive great soil
groups on the same parent material; or soil types of a tropical region repre-
senting various proportions of wet and dry seasons, such as high mag-
nesium, manganous, ferruginous, and aluminous families of Latosols
(leaching and oxidation sequence); or soil types in a mountainous region
representing the same parent material at various altitudes (temperature and
rainfall sequence) .
3. Soil types in chronosequence, such as soils representing various peri-
ods of time since glacial, loessial, or volcanic deposition.
4. Soil types in biosequence, such as grassland and forest soils developed
on the same parent material.

s Am. Soil Survey Assoc. Bul., 15: 58 (1934).

D Namesof sequences proposed by Jenny, Soil Sci., 61 : 37S (1946) ; also, Jenny,
Factors of Soil Formation (New York: McGraw-Hill Book Company, Inc., 1941),
p. 6.
SOIL SAMPLING 15
5. Soil types in lithosequence, with closely similar environment but de-
veloped on different parent materials (a measure of specific mineral suscep-
tibility to weathering).
2-11. The catena (toposequence) was described in 1935 by Milne. 10 The
interrelationships of several Nebraska soil types were sketched 11 in 1935
as a toposequence. Soil sequences controlled primarily by drainage were
detailed by Bushnell. 12 Several other subdivisions of the toposequence
would be possible. 13 Hole 14 assigned index values to catena members.
2-12. Deep Profile Sampling Recommended. Sampling of the profile
of soil and subsoil to the full depth of geochemical weathering is recom-
mended in so far as possible. Interesting facts of mineral weathering and
soil formation are revealed by analyses of profiles to the full depth of geo-
chemical weathering, whether the deeper horizons are considered a part of
the soil profile proper or not. Knowledge thus gained of the course 111 of
mineral weathering as a function of depth (the "depth function") is useful
in the interpretation of soils, their formation, and potential fertility.

APPARATUS AND SUPPLIES

2-13. Apparatus needed for soil profile sampling includes spades for
excavation of the soil profile pit; soil augers, a small one for examination of
the soil and a large one (Fig. 2-2) for taking deeper samples; paper bags
for the samples; and a dark wax crayon for labeling.
2-14. The closed-cylinder type auger16 (Fig. 2-2a) has been found by
several state experiment stations to be suitable for dry soils, including
sands. Two tool steel cutting blades loosen the soil and feed it back into the
closed cylinder. Its disadvantages are that moist clay soils pack hard in the
cylinder and are difficult to remove, and the soil structure is not preserved
for examination. The standard 8- or 10-cm post-hole auger (Fig. 2-2b) is
often employed for sampling. Its disadvantages are that loose dry soils do
not remain in the · hopper as it is pulled from the hole, and, since it is
slightly tapered, some of the upper soil layers are continually mixed with
somewhat lower layers. The latter difficulty is not considered very serious
by Piper, 17 in recommending the 10-cm post-hole auger for taking soil
samples under Australian conditions.

10 Soil Research, 4(3): 183 (1935).

11 Jackson et al., S.S.S.A. Proc., 2:437 (1938).
12 Purdue Univ. Agr. Exp. Sta. Spec. Cir. No. 1 (1944); S.S.S.A. Proc., 10:335
(1946).
13 Winters, Soil Sci., 67: 131 (1949).
14S.S.S.A.Proc., 17:131 (1953).
11! Jackson et al., J. Phys. Colloid Chem., 52: 1237 (1948).
16 Available commercially from R. C. Jordan, Soil Testing Equipment Manufac·
turers, 4616 Olivewood Ave., Riverside, Calif.
17 Soil and Plant Analysis (New York: lnterscience Publishers, Inc., 1944), p. 1.
16 SOIL SAMPLING

c
\
Wood handle

100 to 150 cm

l15 to 23 cm

J
Scm
Cutting blade

~.jIO cm
(a) (b)

Fig. 2-2. Post-hole type soil sampling augers: (a), closed cylinder auger suitable for
dry soils (after Cole and Retzer, S.S.S.A. Proc., 1 :305, 1937); (b), conventional post-
hole auger.

PROCEDURE

2-15. Selection of the Profile-Sampling Site. The profile site for sampling
is chosen on the basis of vegetation, microclimate, degree of erosion, sur-
face drainage, proximity to trees, and any other factors which are pertinent
to identification of the profile with the soil type. Road cuts are not the best
sampling sites because they are likely to have an overburden and a deposi-
tion of limestone dust. To study the most extensive agriculturally important
profiles may necessitate the use of some cultivated sampling sites. The
degree of disturbance should be carefully noted in the profile description,
and the fact of disturbance recognized in the interpretation of the analyses.
If virgin and cultivated soil profiles are being compared, the virgin soil is
selected as the modal soil for the type, and the cultivated soil is taken as
near to it geographically as possible.
2-16. Location. The location of the site selected is carefully recorded
SOIL SAMPLING 17
on a detailed map, for example, on a detailed soil survey map or a county
road map. The state, island, county, detailed legal location (section, town,
and range) of the site, and directions for locating it from a nearby town
are recorded with the description. The soil map represents the landscape in
the dimensions along the X-axis and Y-axis (Fig. 2-1 ) . The soil type, or
phase, symbol denotes characteristics along the Z-axis. 18 Lines within the
XY quadrants mark the map into isotypic soil areas.
2-17. Replication. Soil profile sampling obviously must take into account
the range of variations in the soil type, which are most clearly defined if
resolved into those occurring specifically along either the X- or the Y-axis.
Different profiles along the X-axis tend to give the least variability and
narrowest range of confidence limits. Selection along the X-axis fits the
concept of selection to be "representative of the modal profile of the soil
type . . . ." 19 There is also true isotypic variation within one soil type along
the Y-axis; in fact, one soil type must necessarily grade into another along
the toposequence. Analyses of several profiles representing a developmental
sequence of soils, either modal profiles of a soil catena or a soil family,
provides a measure of systematic variability, and thus enhances the repli-
cation. Composite sampling is not employed.
2-18. At least three replicate soil profiles are sampled. They are sepa-
rated geographically as widely as possible, preferably by at least 50 miles
(I 00 kilometers) to represent normal isotypic variations resulting from
variations in parent materials. Analyses of at least three profiles are re-
quired before any generalizations are made for the soil type as such. Two
soil types may not be proved to differ if only one profile of each is analyzed.
2-19. Excavation and Description of the Profile. The profile pit is exca-
vated deep enough to reveal the principal features and to extend down to
the parent material. The pit should be oriented so that profile is uniformly
lighted. Before and during the taking of the samples, the profile horizons
are carefully described as to depth, color, morphology, texture, consistency,
and drainage. Detailed nomenclature and criteria for each feature are set
forth in the Soil Survey Manual. 20 A system of symbols may conveniently
be used to describe the various properties semi-quantitatively. For ex-
ample, a system for quantitative expression of colors is employed by the
U.S. Division of Soil Survey. 21 Nikiforoff22 described a system of symbols
for the other profile features. The parent material, age (from geological
data), vegetation, altitude, rainfall, temperature, and other factors such as

1s Different conventions were employed by Mattson and Wiklander, Soil Sci.,

49:154 (1940); Z corresponds to their "y" and Y to their "x."
111 Cline, Soil Sci., 59:3 (1945).
20 U.S.D.A. Handb. 18 (1951).
21 Jbid., p. 194 (color chips are obtainable from Munsell Color Co., 10 E. Franklin
St., Baltimore 2, Md.); also, Pendleton and Nickerson, Soil Sci., 71:35 (1951).
22 S.S.S.A. Proc., 1:307 (1937).
 SOIL SAMP LING
18
various
the wet/dr y ratio of seasons or frozen subsoil are recorded. The
a sketch
horizons are delineated and marked with pegs. A photograph and
t and
of the profile are usually made. Collaboration between the soil chemis
ists is always useful and usually mandat ory
the soil classification special
during the profile sampling process .
la-
2-20. Removal of Soil Samples from the Profile. The surface accumu
) is sample d separat ely and its average
tion of organic matter (A 0 horizon
area of
thickness estimated. This horizon may have to be collected over an
an adequa te sample volume , usu-
one or more square meters to provide
are next sample d, a volume of about
ally about 2 liters. Successive horizons
g volume is a cylindr ical
1 liter of each usually being taken. The samplin
er of the samplin g
core centered on the Z-axis of the profile. The diamet
l even from thin
volume is varied so as to provide adequate soil materia
cm and
horizons. If the horizons are uniform through a thickness up to 30
horizon
sharply delineated, the cylinder of soil is taken throughout the
y depth
thickness. The procedure of sampling every I 0 cm or other arbitrar
the
is avoided. For very thick horizons bounded by long transition zones,
vertical ly only through the most typical,
cylinder of soil taken extends
s being
central, I 0- or 15-cm portion of the horizon, the transition horizon
proced ures such as pH and organic
sampled separately. Less laborious
on all horizon s. The most extensi ve
matter determinations are carried out
out only on the 4 or 5
and laborious analyses may need to be carried
found satisfac tory in the
principal horizons. This procedure has been
23 He sug-
author' s laboratory and concurs with Piper's recommendation.
I .5 meters. Profile
gests sampling into the C horizon, to a depth of 1 to
extende d down
sampling for mineralogical studies almost always should be
into the D
into the parent soil material, preferably through the C and
horizon ( ~ 2-12). Monolith sampling should be considered ( ~ 2-22).
2-21. Labeling and Transporting the Profile Samples. Each horizon
labeled
sample is placed in a double-walled cloth or paper bag. The bag is
or red wax crayon is not satisfac tory), and
with a blue wax crayon (pencil
the same label, written in blue wax
a file card is dropped into it, bearing
soil is
crayon. If the samples are very wet, the bags are left open and the
The bags are then transpo rted to
placed to dry partially before packaging.
the laboratory.

MONO LITH MOUN TING OF SOIL PROFI LES

natu-
2-22. Mounted profile monoliths set in plastic for permanence and
e for instruc tion and study of profile
ral, moist appearance are valuabl
excava-
analyses. Frequently these monoliths are taken in the same profile
analysis. Method s of taking relative ly
tion as the profile samples taken for

2s Op. cit., p. 3.
SOIL SAMPLING 19

thin monoliths, in the interest of light weight and small space for display,
have been described by Harper, 24 Lyford, 25 and Berger and Muckenhirn. 26
The latter authors point out the need for a thickness of 2 ·to 3 cm to show
the natural structure of many soil types. They developed this effective pro-
cedure for cementing the soil material together in monoliths of this thick-
ness.

APPARATUS
2-23. Needed apparatus is a flat spade, knife, spatula, hand pick, mono-
lith sampling tool (Fig. 2-3a), cloth and wood board backing, and
hydraulic jack.

(a) (b) (c)

L-=4
~
14-20cm-j
~
'
:...~---

lOcm

Cutting
~ edge
"" 120cm
(d)
(e)
t:s cm:j

~1
~ .
..f 10 cm .

~.
·J.
-==N-.'.ie':
J
Sharpened

Fig. 2-3. Soil monolith sampling tools: (a), frame for profile
sample mounting (after Berger and Muckenhirn, S.S.S.A. Proc.,
10:368, 1946); (h, c, d), types of columns to be thrust into the soil
from above; (e), turf profile sampler.

24 Okla. Agr. Exp. Sta. Bui. 201 ( 1932).

25S.S.S.A. Proc., 4:355 (1940).
:ms.S.S.A. Proc., 10:368 (1946).
20 SOIL SAMPLING

REAGENTS

2-24. Two reagents are employed, which are mixed together in the
proper proportions just before using:
Solution A: 12 % vinylite resin in acetone.
Solution B: 12 % vinylite resin in methyl isobutyl ketone.
The vinylite resin employed is vinyl acetate-vinyl chloride copolymer, grade
VYHH in powder form (Bakelite Corporation, 230 North Michigan Ave.,
Chicago 1, Ill.) .

PROCEDURE

2-25. The Berger and Muckenhim procedure27 for removal of the mono-
lith from the soil varies somewhat according to soil texture and compact-
ness. For fine textured soils,
a portion of the exposure, about 18 inches [30 or 40 cm] in width. is
smoothed to a plane surface and the metal frame is pressed or driven into
the soil bank until it is flush with the soil. The back is then fastened to the
frame and the soil dug away around the sides. The soil in the frame is then
separated from the exposure by cutting between the bank and frame with a
large knife, starting at the top and working down. While cutting is in prog-
ress, slight pressure [is] exerted on the frame, pushing it toward the bank
to prevent slumping of the soil. If the soil tends to slump out of the frame,
a metal or plywood strip, just over 8 inches (20 cm] wide [is] inserted be-
hind the frame, from the top down, as the cutting progresses. After sever-
ance from the bank, the frame with the soil profile in it is tilted back and
the exposed face smoothed level with the edge of the frame.
The exposed face of the soil profile is then painted with vinylite solution
A. A board 9 X 50 inches [23 X 125 cm] is also moistened with the solu-
tion and, with moistened side down, placed and centered on the exposed
face of the profile. The board and frame are then held tightly together and
turned over so the profile rests on the board. The back is removed from the
frame and the soil is forced out by lifting the frame gently, while pressing
downward on the soil with a board small enough to fit inside the cutting
edges of the frame. Before the frame is entirely lifted away from the profile,
the latter is enclosed by placing ... wood strips around the sides. After the
frame has been removed, these strips are nailed to the board underneath and
the profile is ready for transportatio n to the laboratory.

2-26. For coarse textured soils, the exposed soil profile is first allowed
to dry to a low moisture content. Then
a liter or more of a solution consisting of two-thirds of solution A and
one-third of solution B is poured or painted over an area 8 inches [20 cm]
wide and 48 inches [120 cm] in [depth] ... A piece of gauze cloth about
16 X 60 inches (40 X 150 cm] is then pressed against the treated soil sur-
face and wetted with additional solution. The gauze provides some support

21 S.S.S.A. Proc., 10:368 (1946).

SOIL SAMPLING 21
and makes the profile easier to handle in the laboratory. After the solvent
has evaporated and the plastic has hardened, the metal ftame [is] pressed
into sandy soils around the area treated with plastic and gauze and the soil
removed as described above for fine-textured soils. Generally, however, and
especially with gravelly soils, a board about 9 X 50 inches [23 X 125 cm]
is placed against the profile and the loose ends of the cloth are tacked to
the board to prevent slumping of the soil layer during removal.
A profile about 2 inches [5 cm] thick is then separated from the soil bank
by cutting from the top downward with a knife or trowel. A plywood board
or sheet of metal is slipped behind the profile as cutting proceeds, although
slight pressure must be maintained throughout against the board to which
the gauze was tacked. When cutting is completed or nearly so, the profile,
held between the boards, is removed from the bank.
After the tacks are taken out, the surplus cloth is folded over the ex-
posed face of the profile and sewn together so as to hold the soil in place
during transportation to the laboratory. Later, before cementing the profile
to the board backing, the excess cloth is trimmed to the dimensions . . . of
the finished profile.
2-27. To complete the mount for display, wood strips are temporarily
tacked on a 23 x 125 x 2 cm (9 x 50 x 0.75 inch) wood board to form
a box 20 x 120 x 2.5 cm. Vinylite solution A is poured into the box to
cover the surface outlined by the strips.
These strips are needed to maintain straight edges during subsequent
trimming and plastic treatment of the profile. The soil profile is then placed
inside the rectangle enclosed by the strips, pressed down firmly, and the
plastic underneath allowed to harden.
The thickness of the profile is then [cut down] to about [2.5 cm] in the
case of silt and clay loam soils and about [1.5 cm] in the case of sands and
sandy loams. The excess soil is removed by prying and lifting with a sharp
pick or knife point so as to expose the natural structure. Where the edges
of the profile are irregular or broken, they are repaired with loose soil ma-
terial previously trimmed from a corresponding position in the profile. This
trimming and preparatory work is more easily and effectively done when
the soil is fairly moist; after it has been completed, the soil is allowed to
dry before being impregnated with the vinylite solution.
The vinylite solution is poured over the face of the profile until the soil
is nearly saturated, requiring usually about I to 2 liters. For most soils, a
mixture consisting of two-thirds of solution A and one-third of solution B
is satisfactory; for clay soils, however, a higher proportion of solution B is
preferable.
After the plastic has dried at room temperature for about 30 minutes, the
wood border strips are removed, and excess solution and soil material are
cleaned from the edges of the board. After standing for about 24 hours, the
profiles become dry and hard and may be set upright. If the face becomes
glossy in places, this glossiness may be removed by brushing the face lightly
with a paint brush moistened with pure isobutyl ketone.
The simulation of natural moist appearance and color is superior in the
above preparations to coatings with cellulose nitrate.2s

!!8 McClure and Converse, S.S.S.A. Proc., 4: 120 (1940).

22 SOIL SAMP LING

ALTERNATIVE PROCEDUR.ES
led
2-28. Miniature soil profiles, 5 cm wide and 30 cm long were assemb
the soil materia l trimme d from the full
by Berger and Muckenhirn from 29
ed
size profile. The respective material is placed to scale in a box as describ
or embedd ed in
for the full size profile, and set in the vinylite plastic,
au and
"Selectron" or "Castolite" resin by a process described by Bourbe
Berger. 30
2-29. A wide variety of tools and methods have been designed for mono-
pres-
lith samples (Fig. 2-3). Types (b) and (c) are thrust into the soil by
bar held down by two
sure from a hydraulic jack anchored against a cross
by means of
heavy screw-type augers. Type (d) may be thrust in by hand or
rs (c
the jack, but is suitable only for relatively friable soils. The cylinde
s
and d) provide soil sample cores suitable for preparation of thin section
has been used for examin ation of golf
by plastic embedding. Type (e)
dis-
greens, when a view of the profile to 15 cm is wanted without greatly
g machin e which takes
turbing the turf. Kelley et al. 31 describe a soil samplin
undisturbed soil cores 10 cm in diameter to a depth of 2 meters.
2-30. Hole:{:! has employed relief models to visualize further the relation
of the soil profile to the landscape.

SAMPLING SOIL OF ESTABLISHED EXPERIMENTAL PLOTS

in-
2-31. The purpose of soil sampling, to measure soil properties that
accomp lished by com-
fluence the production of crops, invariably is best
soil
posite sampling ( ~ 2-5) . When the set of replicate plots is all of one
A plot
type, phase, and subtype, each plot is composite sampled as a unit.
t field area that receive s a given treat-
is used here to designate the smalles
more than one soil type, phase, or
ment and is harvested as a unit. When
are present in
subtype occurs within one plot, and different proportions
different plots, it is difficult to obtain valid samples of soil.

APPARATUS AND SUPPLIES

2-32. Needed apparatus includes a soil-sampling tube n (Fig. 2-4);

3

pencil;
spade; paper bags, cardboard cartons, or sample bottles; dark wax
stony soils, a hand pick and soil auger
stakes; and string. For hard, dry, or
are needed.

29S.S.S .A.Proc ., 11:484 (1947).

ao S.S.S.A. Proc., 12:409 ( 1948).
a1 S.S.S.A. Proc., 12:85 (1948).
32 Agron. lour., 42:520 (1950).
33 Obtaina ble from the Oakfield Apparat us Compan
y, Oakfield, Wis., or Central
Scientific Compan y, 1700 Irving Park Road, Chicago , Ill.
SOIL SAMPLING 23

(a)
15cm

(b)

Plated case-hardened
iron or stainless steel ~f"'"'~'"'~~~~~~~

1.0 mm

Fig. 2-4. Details of construction of cut-away soil sampling

tube: (a), depth intervals grooved each 15 cm (6 inches), lengths
of 33 and 93 cm are convenient; (b), details of hardened cutting
edge, which should be hard enough to resist bending if thrust
into gravel, but not hard enough to be brittle.

2-33. The cut-away type soil sampling tube is remarkably effective for
composite sampling the plow layer and upper subsoil of moist, stone-free,
friable soils. Permanent grooves on the outside of the tube indicate the soil
depth sampled. The semicircular opening permits visual inspection of the
horizon for conditions of structure, mottling, mild compaction (Hoffer
chalk test, 34 11 13-7 4), contamination of surface with subsoil, moisture
penetration, and root distribution. These observations can be made rapidly
as the composite sampling progresses. The semicircular opening permits
the soil core to drop out into the paper bag or other sample container.
2-34. The King type of soil sampling tube~~ (closed cylinder with en-
larged portion back of cutting edge) has long been employed. The top
end is re-enforced for use with a soil tube driver that consists of a hammer
guided by a steel rod that fits into the tube. Threaded sections of this type
of tube permit extension to depths of 10 meters or more. This tube, when
ruggedly built, is suitable for sampling to great depths in fairly dry soils.
Wet soils tend to stick and clog the tube. A lighter weight cylindrical tube
similar to the King tube but with handle has been designed by Hankinson-
Hester. 36

34 Best used with extra-Jong cut-away soil sampler obtainable from Elano Corp.,
Xienia, Ohio, or Ken Standard Co., Evansville, Ind.
B5 Available from several supply houses, including R. C. Jordan, Soil Testing Equip-
ment Manufacturers, 4615 Olivewood Ave., Riverside, Calif.
86 Available from LaMotte Chemical Products Co., Baltimore 4, Md.
24 SOIL SAMPLI NG

PROCEDURE

2-35. Each plot is delineated by a stake at each corner, and sometimes

is outlined by a string. The operator proceeds across the plot in a zigzag
path (Fig. 2-5, A) taking a core with the cut-away core sampler (Fig. 2-4)
to pl~-layer depth every 2
A B c steps. Sampling of the 50-cm
(Effective) (Unsuitable) (Unsuitable)
border of the plot is avoided.
x x x x Criss-cro ssing the plot provides
x
x
x for chance sampling of fertilizer
x
x
ordinaril y applied parallel to
x
x x x the rows and the long direction
x of the plot. A composite sample
x
x
x x of 10 to 30 cores (usually 20
x
x
x to 25) is placed in a water-
x resistant paper bag that has
x x
x x x been plainly labeled with a
x
x dark wax crayon. A spade is
x x sometimes employed in lieu
x x x of a soil-sampling tube for
x
x x x x composite sampling of plots. A
x x
..
x=core-sam phng pos1t1on
spade or hand pick is best for
sampling stony or hard, dry
Fig. 2-5. A, Correct procedure, a zigzag soils, but is many times more
pattern across the plot gives a proper com-
posite of 10 to 30 cores for a single soil laborious than the sampling
sample. Valid replication is provided by tube for moist, friable soils. A
sampling a replicate plot similarly or resam- narrow, flat blade with parallel
pling the same plot in a similar pattern. B,
Unsuitable procedure , regular positioning of sides facilitates obtaining a
the cores in a plot, from which samples are soil-sample slice with uniform
likely to be biased by row applications; analysis
of these cores separately as replicates is an in-
size of cross section. However,
efficient practice. C, Unsuitable procedure , it is easily bent and cannot be
regular positioning of the cores with a distinct employed for the excavation.
bias toward the ends of the plot; too few cores
are represented to make an accurate composite
A hand trowel aids in manipu-
sample. lating a parallel-sided block of
soil for compositing from a slice
taken with the tile spade. Cline 37 indicates that in an hour 20 circular holes
40 cm in depth and diameter can be excavated, and a 10-cm slice cut from
the wall of each with the spade and sampled with the trowel. The time re-
quired, of course, depends on how hard the soil is. Twenty cores of a moist,
friable soil can be taken and composited in 5 or 10 minutes with the soil-
sampling tube.
2-36. It is convenient to leave the bags standing on the plot where taken
87 Soil Sci., 59:3 (1945).
SOIL SAMPLING 25
until the entire series of samples has been obtained. In this way, the sample-
numbering system is self-checking and no confusion as to tl~e sampling area
from which the sample was obtained can occur.
2-37. The samples are collected into suitable cardboard boxes and trans-
ported to the laboratory without drying. The particular determinations that
are to be made should be considered in deciding upon the procedure from
this point forward. Some determinations require field-moist samples so that
the equilibrium is not disturbed by drying, and others require air-dried
samples.
2-38. Replication. One composite sample is generally obtained from
each of several replicated experimental plots, thus providing the most suit-
able sample replication of each treatment. Replicate samples may be ob-
tained from each plot by repeated composite sampling; this procedure is
necessary for testing for differences between the different replicate plots in
the experiment, but is not done in routine plot sampling. A plan for replica-
tion is developed and mapped before sampling is initiated.
2-39. Sampling Soil Under a Single Plant. Soil sampling for correlation
with the analysis of a single plant is restricted to a composite sample from
the central portion of the root-spread volume (~I 2-5), which is more or Jess
a hemispherical body of soil surrounding the base of the plant. This root-
spread hemisphere represents the theoretical minimum part of a plot that a
single sample should represent.
SAMPLING SOILS IN SELECTION OF SOIL EXPERIMENT FIELDS
2-40. The popular directive given for selection of a soil experiment field
is "select a uniform soil area." Unfortunately, this is a very difficult if not
impossible requirement to fulfill completely. Except for some loessial, allu-
vial, and lacustrine soils under grassland vegetation, soils are generally
found to be lacking in uniformity over areas the size of a field experiment.
As nearly uniform soil as possible is located, and then soil sampling and
analysis are employed to assay the extent of soil variation that exists.
2-41. The Plot-Size Sampling Volume. In a field being examined for
adequate uniformity for fertility plots, the smallest unit generally employed
for variability measurement is a sampling volume having the size, shape,
and orientation the plot will have (Fig. 2-6). Composite sampling of such
a plot-sized soil volume performs the same integration as the roots of a
growing crop effects ( ~ 2-5) within each plot. The difference between
analyses of a series of such composite samples provides a basis for calcu-
lating the variance of soil properties among different plots. Once the
standard deviation of analytical characteristics is established for composite
samples from well distributed plot-size sampling volumes, it tends to apply
for the whole experimental field whether or not the actual fertility plots fall
exactly within the individual limits of the plot-size sampling volumes (Fig.
26 SOIL SAMPLING

j:f><;C--;;;·-·;- -_-;;•i<j \Sampling

volume= surface soil of plot
~----"-"---'-------' Composite sample= 2: 20 to 30 cores
I )CX1'i.-;- - ~)i;; -x-;~

.--~---!(~ ~- :\ ~---~---.::.> x _ _J

I )( ~-_?-~--
" 1f w •'t""'-;--- -----1
-~)( -1
Ll! _ _

FiJ?. 2-6. Chemical analyses of composite samples

drawn from plot-size volumes (dotted lines) permits
statistical analysis of significant soil variability between
future yield plots (solid lines) to be laid out later. even
though the latter may not fall within the original bound-
aries.

2-6). The significant "deviation from the mean" is the deviation of various
plot-size volumes from the mean of the field. The distribution of semimicro
soil variations and the characteristics of crop growth and management,
make the long, narrow plot, and therefore, a long, narrow sampling volume,
the most efficient for yield measurement. No gain in estimation of the vari-
ability of the plot-size sampling volume would be made by analysis of indi-
vidual cores. Individual cores taken at distances apart of plot dimensions
and analyzed separately characterize the field as a whole (as would one
composite ), but little or no significant information is provided about the
homogeneity or heterogeneity of a set of plots to be placed in the field.
Microvaria tions in soil are induced by individual trees, windrows or shocks
of grain, or deposits of animal excrements, but these are of little significance
because rooting of the individual plant ( ~ 2-5) and coverage of a single
long narrow plot average the soil properties over each plot.

APPARATUS AND SUPPLIES

2-42. Needed apparatus includes sampling tube (Fig. 2-4), spade, and
auger. Needed supplies include paper bags or cartons and dark wax pencil.

PROCEDURE

· 2-43. Preliminary Inspection. Within a prospective experimental field

area having suitable topography, the soil is examined to a meter depth or
more to establish whether the soil type is that wanted. This preliminary in-
spection is carried out in at least 5 places, well distributed in the prospective
experimental field.
SOIL SAMPLING 27
2-44. Preliminary Composite Sampling of Blocks. Between 10 and 30
cores (usually 20 to 25) are composited in a paper bag from the full plow-
layer depth, taken at random over an area having the size, shape, and
orientation of the prospective plot (Fig. 2-6). This procedure is repeated
for other plot-size areas to represent parts of different replicate blocks. One
sample per block is minimal, and, if tests of these with the field kit do not
eliminate the field from consideration, additional samples are taken to give
2 or more samples of plot-size volumes per replicate block for more detailed
analyses in the laboratory. This constitutes the preliminary sampling. Vari-
ability between blocks is calculated by analysis of variance. Between-block
variability should be no larger than necessary, and should not affect differ-
entially the treatments to be tested.
2-45. Detailed Sampling of Blocks. Additional detailed sampling of
plot-size volumes is carried out after the plots are tentatively laid out, to
provide an estimate of plot variability within blocks. This is the variability
that causes the most difficulty in testing differences due to treatments. If it
is large, the experimental site is unsuitable. Ordinarily the variability be-
tween long narrow plots is not excessive. This sampling can be used as a
basis for discarding a few plots if they fall outside the desired limits of
variability.
2-46. The chemical analyses employed as criteria for plot uniformity
(e.g., soil pH, available phosphorus, available potassium, organic matter)
are generally simple and few in number, and therefore can be applied to
relatively large numbers of samples. The labor for analysis is relatively
small compared to that involved in plot experiments. Thus, it is not un-
reasonable to analyze as many samples as plots to be included in the experi-
ment, although fewer samples suffice for many purposes.
2-47. Subsoil Sampling. Soil fertility plots are generally laid out on culti-
vated fields or grassland, seldom on virgin lands. Soils that have been
cropped are likely to exhibit their chemical variability most in the surface
plow-layer, because variations in fertilizer treatments, return of plant resi-
dues in manures, accumulation of organic matter from roots, and depletion
of organic matter through oxidation and erosion are concentrated in this
layer. Also, the underlying subsoil tends to become more uniform chemi-
cally throughout any given field of one soil type through the action of roots
in depleting available nutrients selectively in the originally more fertile
spots. On the other hand, differences in the subsoil texture and depth lead
to differences in the chemical properties of the subsoil in different portions
of the field that are not made uniform by cropping.
2-48. The greatest emphasis in soil sampling for plots is given to the
surface plot-layer of 15 to 20 cm ( 6 to 8 inches), after a preliminary in-
spection has been made for difference in the texture and chemical properties
 SOIL SAMPLING
28
subsurface 20 cm
of the entire profile. A few samples are taken from the
are taken of the
immediately beneath the plow-layer, and a few samples
taken of virgin
underlying profile by horizons. Comparative samples are
tion of the chemi cal status of the
profiles of the soil type for detailed evalua
soil of the experimental field.

SAMPLING FARM FIELDS

one soil sample
2-49. The size of a farm field area to be represented by
r is willing to give
is determined by the size of the area to which a farme
opera tions. Generally, in the
separate attention in his soil-management
es) are managed as a
smaller farms, fields of 5 or l 0 acres ( 2 to 4 hectar
analysis, recom-
single field unit, making this the logical unit for sampling,
on larger farms
mendations, and subsequent treatments. Larger field units
major natural soil
dictate larger soil sampling units, within the limitations of
relief, depth , or textur e of the soil.
sampling units delineated by changes in
by Jack of produ ctiveness,
On the other hand, small trouble spots delineated
may be sampled separately ( ~ 2-54) .

APPARATUS AND SUPPLIES

2-4) or spade.
2-50. Needed apparatus includes a sampling tube (Fig.
on smooth board,
a
Needed supplies include paper for a map to be mounted
paper bags or cartons, and dark wax pencil.

PROCEDURE
sampled is given
2-51. Field or Farm Diagram. The farm or field to be
the different fields,
a general inspection, and a diagram is prepared showing
upland or bottom
the drainage pattern, and the main kinds of soil, such as
and mann er of compo site sampling
land. A plan of the number of samples
design ated by letters , A, B, C,
is entered on this map, different fields being
a numb er. Detail s of soil
etc., and samples from each field by the letter and
retatio n of the
and crop management are needed for a satisfactory interp
e ( ~ 13-80 ). Sep-
analyses to be made in connection with soil testing servic
topography (up-
arate samples are taken to represent each distinct kind of
t (light colored
land separate from lowland), soil texture, soil organic conten
by crop growt h),
separate from dark colore d), fertility status (as indicated
tions not to be included in the
and management unit. Abnormal condi
ns, highw ays, and soil from
samples are soil near buildings, gates, field margi
bags are labeled
crop hills or rows that were fertilized. Water-resistant paper
with a dark wax pencil before sampling is begun.
to be sampled
2-52. The operator makes a traverse over each area
layer at intervals of
separately (Fig. 2-7), taking a core or slice of the plow-
SOIL SAMPLING 29

x=core-sampling position

Fig. 2-7. Farm field diagram showing layout

of four soil sampling units according to typo-
graphic position and size of the field significant
to management. Samples 2 and 3 are of the
nature of replicate samples in the main field
area. Each soil sample is a composite of from
IO to 30 soil cores.

15 to 20 steps and compositing them together in a bag. A thin slice of soil

(Fig. 2-8) is taken so that the sample size is not too large. Ten to 30 well
distributed cores or slices are composited for each sample. A sample that is
too large is thoroughly mixed in a pail or passed through a coarse screen,
and a representative portion is retained as the sample. One double handful
(250 to 500 ml volume) of soil is sufficient for simpler tests. The sample

Plow laye

Fig. 2--8. Taking a representative thin slice of soil

through the plow-layer by means of a spade. From
10 to 20 such slices are.composited for one sample.
 SOIL SAMPLING
30
ted
location previously marked on the bag is carefully checked and correla
with the field and area on the map.
2-53. Subsoil character. Seldom does farm field-sampling require
time of
samples from the subsoil below the plow layer. However, at the
possibl y samplin g) is
sampling the surface layers, an examination (and
nce of a highly
usually made of the subsoil at 1 or 2 locations. The occurre
le
fertile subsoil, well supplied with lime and having a high or low availab
eristics
water capacity is of great significance to crop growth. Such charact
for the soil type and, therefo re, may not neces-
are frequently well known
farm samplin g, simply being taken into ac-
sarily be re-examined during
of the analyse s of surface sample s.
count in the interpretation of the results
soil problem spots
2-54. Sampling Soil of a Local Problem Spot. Local
local
frequently lead to requests for analysis and diagnosis; examples are
ely
areas of high CaCOa content, local exposures of subsoil, local extrem
e soil
acid (containing S and H 2 SO 4 ) soils, local saline soil areas or alkalin
n, and local areas that grow chlorot ic
areas, local areas underlain by hardpa
iting 10 to 30 cores taken
plants. A surface soil sample is taken by compos
2 or more meters apart in representative areas. Each core extends through
the All or A 1 horizon. Subsoil samples are taken by profile hori;zon beneath
ited
the All or A 1 horizon to a depth of 1 meter by spade excavation, compos
local problem spots occur, each is sample d
from 3 locations. If several
within one area are noted, soil represe nting
separately. If visible variations
One or
each important kind of occurrence is composite sampled separately.
iately adjacen t soil profiles
more control samples are obtained from immed
oppor-
on which normal plant growth occurs. This precaution offers the
ies respons ible
tunity of greatly narrowing down the analysis of soil propert
for the local problem.

HANDLING SOIL SAMPLES IN THE LABORATORY

for
2-55. Handling soil samples in the laboratory involves procedures
drying, grinding, sieving, mixing, partitioning, weighing, and storing.

APPARATUS AND SUPPLIES

2-56. Handling soil samples in the laboratory requires a drying cabinet

6-mm,
with provision for air circulation; a drying oven, 100° to 110° C;
zed cloth, wrappi ng
4-mm, 2-mm, and smaller sieves; sheets of rubberi
and agate
paper, and glazed paper; a wooden rolling pin, rubber pestle,
paper
mortar and pestle (Fig. 1-3); a riffle or large spatula for quartering;
book,
bags or cartons for temporary storage; screw cap jars, labels, a log
standar d sample s; and a balance sensitive to 0.1 to
and a storage case for
0.5 per cent of sample weight.
SOIL SAMPLING 31

PROCEDURE

2-57. Drying. The soil sample is usually partially air-dried (until not
sticky) at a temperature of about 25° to 35° C, and relative humidity of 20
to 60 per cent. Many determinations are not significantly affected by com-
plete air drying for storage purposes. Analytical samples to be stored for
protracted periods of time are almost invariably air-dried because of
changes that occur in the chemical status of ions and organic matter of soil
samples stored moist. Owing to the large and rapid changes that take place
in the status of some ionic species on drying, many types of analyses must
be carried out on moist samples immediately after collection. Examples are
the determinations of exchangeable ferrous iron (Chapter 15) and to some
extent hydrogen ion activity (Chapter 3), exchangeable potassium (Chap-
ter 6), acid extractable phosphorus (Chapter 7), and nitrate nitrogen
(Chapter 8). The common weight basis for expressing the results is the
100° to 110° C oven-dry weight of soil material (usually the "fine earth,"
~ 2-60); this may be converted to the volume basis ("plow layer," etc.,
~ 2-3 ).
2-58. Sieving. The bulk soil sample for chemical analysis, in its natural
field moist conditions, is passed through a 6-mm ( 4 meshes per inch) sieve,
usually by rubbing with the fingers. Passing samples other than sandy soils
through the 6-mm sieve is much easier if the soil is field moist rather than
air-dried. If the soil is in a very friable condition, the -6-mm material
may approximate a -4-mm material without further sieving. In soil sam-
pling in the field, the stones and large gravels are ordinarily ignored, and a
few stones and gravel particles remaining on the 6-mm screen usually are
discarded, since they ordinarily are less than 1 per cent of the plow layer
(~ 2-60). Sieving effects some mixing of the soil; mixing is completed next
(~ 2-65), and the soil is partitioned (~ 2-66 or 2-67). An optimum
sample volume of -6-mm material is often about 1 liter (~ 2-63). If
further decrease of sample size is needed, usually the case, the process of
grinding ( ~ 2-61 ) , sieving, mixing, and partitioning are repeated. Suc-
cessively smaller sieves are employed until the required laboratory sample
is obtained for the particular analysis to be effected.
2-59. Soils in the right moisture condition to be mellow can be worked
through a 2-mm sieve (or 10- to 20-mesh per inch) by being rubbed over
the sieve surface with a rubber stopper. The common practice of passing
only a portion of the gross sample through the sieve and discarding the re-
mainder is likely to introduce positive bias in the sample by increasing the
concentration of most elements of soil fertility. That practice is justified
perhaps in preparing samples of primarily silty and clayey soils for rapid
soil testing, the assumption being that the unsifted aggregated material is
32 SOIL SAMPLIN G

the same as that which has passed the sieve. The entire partitioned sub-
sample is ordinarily made to pass a given sieve size before further parti-
tioning.
2-60. Soil scientists have generally employed as a basis for expressing
the analysis, the "fine earth," or soil material passing the 2-mm, round hole
sieve. The percentage of constituents is expressed in terms of this material
even if further sieving and segregation of coarse sand fractions is carried
out. A moderately stonyas field has 0.5 to 2 per cent of stones. In soils
exceeding this content of stones and gravel, the "fine earth" ( - 2-mm)
basis is biased by more than 2 per cent. For routine soil testing, the fraction
coarser than 2 mm is discarded. In research work, the appearance of sig-
nificant amounts (over 2 per cent) of gravel on the 2-mm sieve is a satis-
factory indication that the "fine earth" basis must be corrected for agricul-
tural soils in referring back to the plow-layer volume (~ 2-3). For research
purposes, the gravel retained on the 2-mm round hole sieve is examined for
the presence of concretions or granular secondary soil particles. The pri-
mary particles and concretions are preserved intact if of interest to the re-
search, but the tough granular secondary soil particles are further disaggre-
gated and passed through the sieve. This is sometimes accomplished best by
trituration in water. The fine particles are washed through the sieve into a
large beaker. The gravel and concretions remaining on the 2-mm sieve are
washed, dried, and weighed. The percentage by weight of "gravel and con-
cretions, + 2 mm" is calculated on the basis of the whole soil on an oven-
dry basis. This fraction is bottled and labeled for further examination if
significant to the analysis. Sieves finer than 2 mm are of course employed
for segregating the various sand fractions used in mineralochemical analy-
sis, but in that procedure the -2-mm sample is disaggregated chemically,
rather than mechanically, prior to sieving.
2-61. Grinding. The soil aggregates are broken up by grinding lightly
with a roller, rubber pestle in an agate mortar, or motorized grinder. Crush-
ing the primary sand and gravel particles is generally avoided. Clay soils
are best crushed for passing 2 mm before they reach complete air-dryness;
otherwise the crushing process is difficult.
2-62. Fine grinding of the mineral grains is permissible with samples
used for the determination of soil organic matter or total elemental analysis;
however, such samples are not suitable for some other analytical deter-
minations such as soil pH, exchangeable cations, easily soluble phosphorus,
or mineralochemical analysis. Fine grinding is carried out in an agate
mortar with agate pestle; a porcelain mortar should not be employed. Care
is exercised never to strike an agate mortar a sharp blow, as it is easily
chipped or broken. There is danger of breakage in heating it for drying

ss Stoniness is thoroughly treated by Nikiforoff, Soil Sci., 66: 347 ( 1948).

SOIL SAMPLING 33
purposes. The proper washing is with distilled water followed by drying in
air or wiping with surgical gauze or filter paper.
2-63. A simple guide to required fineness is that the sieve opening
should pass a spherical particle which will constitute 0.001 or less of the
minimum subsample volume (Table 2-2). Of course, the average particle
size passing the sieve will be much smaller than the maximum set by the
sieve opening. The optimum sample volume or weight will be 3 to 4 times
the minimum.
2-64. A few uncrushed sand particles may become an appreciable frac-
tion of a small analytical sample of sandy soils. It is difficult to get a repre-
sentative portion of the fine and coarse particles into each sample even by

TABLE 2-2
Relation of sieve size to minimum sample volume and weight, based on need in
the sample for at least 1000 particles of sieve-opening size. The optimum
sample size is on the order of 3 or 4 times the minimum
-----·---~. ------ -----------
Minimum sample
Sieve size, -------- --~---------

nominal opening Volume Weightt

----;----
mesh/in.
_____
mm _______ ,_____________
,____(approximate) cm gm
----··--
3

6 4 112 146
4 (i 34 44
2* ]() 4.2 5.3
1 20 0.52 0.68

0.25 60 8.2 13
0.16 100 2.1 2.7
0.1 140 0.52 0.68
----~--·---------·--·----·---·------------~---------~----

0 Standard of reference, round holes.

t Based on bulk density of soil equals 1.3 gm/cmB; for mineral or rock grains of 2.65 gm/cm3 (no
pore space) double these weights would be minimum.

careful partitioning (~ 2-66). Peech et al. 119 have suggested a partial solu-
tion of this problem by sieving out the coarse sand with a 1-mm sieve,
making the chemical analyses on the -1-mm fraction, and calculation
back to the basis of the - 2-mm soil by neglecting any contribution of the
1- to 2-mm fraction to the extractable constituents. The less than 2-mm
fraction is simply weighed, and its percentage of the - 2-mm soil is cal-
culated. Other procedures are use of a larger sample and taking a small
aliquot of solution for analysis; and fine-grinding of the -2-mm sample
prior to analysis when permissible.

39 U.S.D.A. Cir. 757 (1947).

34 SOIL SAMP LING
,
2-65. Mixing. The sample is mixed by a process of rolling or turning
are grasped , and
effected as follows: opposite corners of the cloth or paper
over
one is pulled diagonally across the sample slowly so that the soil rolls
corner of the
(not slides) toward the opposite corner. Then the opposite
is
cloth or paper is pulled back over the soil to roll it back. The process
opposite
repeated by grasping the other two opposite corners. Rolling on
if the
diagonals is repeated at least 5 times. Ten times may be necessary
volume of soil is large.
A
2-66. Partition of the sample. The sample is partitioned with a riffle.
d so that alterna te slots deliver to
riffle has a series of narrow slots, arrange
than
opposite sides. The procedure is somewhat faster and more reliable
quartering(~ 2-67).
r-
2-67. Alternative to riffling, the sample may be partitioned by "quarte
of the mixing sheet with
ing." The mixed soil material is coned in the center

,-- --:: -:- --- --,

I t : A
.J--Glaz ed
I paper

I I
I __P_le_~
Cored sam

~------
' D B B
~ --·""~'1
I----'-----! '·;·'
I 'I• su15ample

I I
IL ___ ! { c, I
_L __ ..!_ _ _ _ _ _ _J

Fig. 2-9. Paper quarteri ng technique for small

samples, especially for sand and silt grains that are
not to be ground. Four strips of paper arc woven
together and the cone formed at the center. Pulling
the strips apart results in accurate quarteri ng. (After
Pettijohn . J. Geo/., 39:432, 1931.)

l.
care to make it symmetrical with respect to fine and coarse soil materia
a flat metal
The cone is flattened and divided through the center with
atively.
spatula or metal sheet, one-half being moved to the side quantit
sepa-
Then each half is further divided into half, the four quarters being
quarter -
rated into separate piles or "quarte rs." For small samples, a paper
opposite
ing technique (Fig. 2-9) may be employed. Two diagonally
40

atively. The other two are either mixed by

"quarte rs" are discarded quantit
Manual of Sedi-
Micro-s ubsampl ing is discussed by Krumbe in and Pettijohn ,
40
n-Centu ry-Croft s, Inc., 1938), p. 549.
mentary Petrogra plly (New York: Appleto
SOIL SAMPLING 35

rolling if the entire half is to be retained, or resieved with a finer sieve if

further partitioning of the sample is to be made.
2-68. Weighing. The analytical sample is the soil material employed as
a whole in a single analytical determination. For example, the soil material
placed in the funnel for extraction of exchangeable cations. Weighing out
the analytical sample is fundamentally a continuation of the partition
process (~ 2-66). The minimum size of sample is best kept 1000 times
(~ 2-63) that of the sieve opening through which the sample was passed.
Schollenberger and Simon41 point out that the analytical sample of coarse-
textured soil samples, which tend to segregate, should be taken by riffling
or quartering down to about the desired weight, and then the entire portion
should be weighed accurately.
2-69. The sample is weighed on a torsion or analytical balance having
a sensitivity equal to 0.1 to 0.5 per cent of the sample weight. Thus for a
20-gm sample, a torsion balance sensitivity to 0.05 to 0.1 gm is adequate.
The sample, if fine enough to be nonsegregating, is handled on a spatula,
otherwise on a spoon. A smooth metal scoop and camel's hair brush are
essential accessories. A counterpoise weight for the scoop is highly
desirable.
2-70. Each duplicate analytical sample is taken in a different set; the
2 are not run side by side. The estimate of error obtained by duplication
then encompasses variations between sets. Many analytical procedures are
short compared to the time for sampling and preparation, and justify
duplication. When a routine has been established to check duplication un-
failingly in different sets, the analytical replication is decreased, reliance
being placed on comparisons of duplicate soil samples for replication. The
procedure then is to replicate only one analytical sample from the previous
set in each succeeding set, or to carry a single control sample in every set.
Finally, after complete reproducibility of procedure by a given analyst has
been established, replication of analysis of a single sample may be discon-
tinued. Replication is then still provided by means of replicate samples of
the original soil. In routine soil testing of farm fields, variability is meas-
ured by taking samples from different small areas in a given field rather
than outright replication by resampling of any one area.
2-71. Storage. Most soil samples are collected for a series of analyses
to be made immediately, and then are discarded. Storage of soil samples,
however, even for the period of analysis requires an orderly procedure. For
soil testing, a series of cardboard cups in a tray with dimensions similar to
the mass-handling testing apparatus is convenient. For longer analyses, the
most satisfactory procedure is to place the samples in screw-cap jars and
to array them on shelves in proper order. Great saving in time and improve-

41 Soil Sci., 59: 13 (1945).

36 SOIL SAMPLING

ment in accuracy is thus effected compared to the all-too-frequent practice

of letting the soil samples accumulate in paper bags in the laboratory, where
contamination may occur. A well-defined schedule for discarding samples
after use should be established. Few soil samples justify storage for pro-
tracted periods.
2-72. Moist Soil Storage. Storage of moist bulk soil samples for green-
house studies is sometimes advantageous, particularly with soils high in
organic colloids, which dry irreversibly. To maintain the moist samples in
tightly covered bulk storage cans, an effective procedure is to embed a
beaker of water in the soil surface. The vapor pressure of water is main-
tained sufficiently high to preserve the natural moistness of the soil. That
some chemical changes in the soil occur under these conditions must be
taken into account in the experimental interpretations.
2-73. Standard Soil Samples. Standard soil samples on which an ex-
haustive amount of analytical work has been done, such as those involved
in mineralochemical or minor element analysis, justify long-time or perma-
nent storage. An intensive research program can be accelerated consider-
ably by having representative reference samples at hand. A properly sieved
and mixed laboratory sample, usually of 1- to 5-liter volume is appropriate
for the soil chemical reference sample. The sample is placed in an air-tight
glass jar with a screw cap. A permanent laboratory number is assigned, a
well-designed label is carefully mounted on the jar, and a similar label is
placed inside it. A water-excluding label varnish is applied over the label
to insure permanence. Where great numbers of such samples are prepared,
a serial number alone on the label suffices. A log book of samples is kept
with copies of the information on the labels as well as additional details.
The samples are stored in a locked case or room fitted with shelves for dis-
playing them.

QUESTIONS
1. What are the two main problems of soil sampling'?
2. Explain the importance of the soil volume concept, as opposed to a soil
area.
3. Define the three coordinates of the landscape, depth function, topo-
sequence, slope.
4. State the principal sources of variability in soil chemical analysis.
5. How does the root spread of plants relate to composite sampling of soil?
6. List the requirements of a valid composite-sampling procedure.
7. How many soil cores should be composited to reduce the variability of
the composite sample to the variability of the analytical results?
8. Outline the considerations you would employ in sampling soil profiles for
chemical analyses.
9. Of what value are soil profile monoliths?
SOIL SAMPLING 37
10. How is replication ordinarily obtained when sampling established experi-
mental plots that are replicated?
11. Outline the sampling procedure used in experimental plot site selection.
12. Outline the procedure and precautions employed in sampling a farm field.
A problem spot.
13. Why must precautions be taken not to dry soil samples previous to some
types of analysis?
14. What precautions should be taken about grinding soil samples?
15. State the relationship of maximum particle size to subsample size.
 Hydrogen Activity
Iletern1ination for Soils
The active mass or activity as distinguished from the total
amount present.

3-1. Perhaps the most importan t chemical property of soil as a medium

for plant growth is its pH value or "hydrogen ion activity." So familiar is
this to soil scientists that the term "soil pH value" is often interpreted as an
entity without reference to its fundamental definition as interpreted from
the voltage obtained with an electrode (~ 3-9). Moreover, the activity in
soils of the twelve or more other ions that enter into plant nutrition is
highly dependent upon that of the hydrogen ion. The lime requirement of
soils (~ 4-64) depends upon the adjustment of the hydrogen ion activity
and the associated activities of metallic cations and anions, although it also
involves adjustment of the calcium and sometimes magnesium ion activities
as such. Activities of several cations other than hydrogen have been meas-
ured1 directly by a clay membrane method analogous to hydrogen ion ac-
tivity measurement with a glass electrode.
3-2. Electrical Potential Measurement with a Potentiometer. The elec-
trical potential of an electrochemical cell is measured with a potentiometer.
The line CD (Fig. 3-1 a) represents a uniform resistance wire of a po-
tentiometer. P and P' are points of contact with the terminals of the stand-
ard cell E. The adjustable resistance, R, at SC is varied until the drop in
potential from D to C is such that the potential drop between fixed points
P and P' is just equal to the standard cell voltage ( 1.0183 volts in this ex-
ample). The reading on galvanometer G is zero when this condition is
1 Marshall, J. Phys. Chem., 43: 1555 (1939); 48: 67 (1944); also McLean et al., Soil
Sci., 72:315 (1951).
38
HYDROGE N ACTIVITY DETERMIN ATION FOR SOILS 39

(a)

E= 1.0183 volts (standard cell)

Q r-
Clo p p'

(b)
D of water
Flow

E~
Fig. 3-1. Principle of the potentiometer: (a},
electric circuit; ( b), hydrologic analogy.

reached, indicating that the voltage from E exactly counterbalances any

tendency for the current to flow through the circuit P'GP. Current from
the working cell of course continues to flow through the path DP'PC.
3-3. The situation may be visualized by analogy to water pressure in the
hydrologic system (Fig. 3-1 b). No water will flow through the water
wheel if the water wheel is turned by an external power source (analogous
to the standard cell) at a rate which is just sufficient to offset the drop in
pressure from P' to P. Water will continue to flow through the pipe PP'.
3-4. The space between P and P' (Fig. 3-1 a) is divided off into a scale
of 1018.3 equal units, and the vol~age drop across each unit then becomes
one millivolt. Since the potential of the glass or hydrogen electrode is di-
rectly proportional to the pH, the space between P and P' may be calibrated
directly into a scale of pH units. It has been found that at 25°C each 59.16
millivolts (0.059 volts) corresponds to one pH unit (~ 3-9).
3-5. The voltage of an unknown cell may be compared to the standard
cell by the proportion of the two distances along the uniform resistance
wire required just to balance the voltages of the two cells. Thus an un-
known potential (as that from a glass electrode pH assembly) is substituted
for the standard cell E and, with the setting at SC and P remaining constant,
the contact point P' is moved along the slide wire CD until the galvanom-
eter reading is again zero. (The movement of P' along the wire corresponds
to the movement of the pH dial on the pH meter.) The spacing between·
the points P and P' when the balance against the unknown cell is obtained
then represents either (a) the millivolts potential of the test cell or (b) the
pH units, as read from the scales described. The details of operation of
40 HYDROG EN ACTIVIT Y DETERM INATION FOR SOILS
a given laboratory potentiometer should be obtained from the instructor or
manufacturer.
3-6. Because some electric current is drawn from the test cell in the
simple potentiometer during the time required to adjust the balance, it is
not suited for the glass electrode pH meter ( «,! 3-7). A vacuum tube circuit
2

is required to measure the glass electrode potential with practically no

current drawn from the glass electrode.
3-7. The Glass Electrode pH Meter. The glass electrode generally used
for pH measurement employs a glass membrane of special, chemically pure,
soft (rapidly soluble, low melting point) glass.a For example, the composi-
tion of Coming 015 electrode glass is 72 per cent Si0 2 , 6 per cent CaO,

/Fill

Hg, Hg2Cl2
or Hg, Hg2Cl 2
Ag, AgCI half cell
half cell

I<CI crystals
Glass .,.___+ Liquid junction
membrane Fiber
Test solution

Fig. 3-2. Glass electrode system for measuremen t of pH.

and 22 per cent Na 2 0. Across the glass membrane develops an electrical

potential that is proportional to the difference in pH between the two sides,
in accord with the concentration cell.: 4
(3-1)
(Sat.) (Sat.)

2 Daniels et al., Experimenta l Physical Chemistry, 4th ed. (New York:

McGraw-
Hill Book Company, Inc., 1949), p. 488.
a Macinnes and Dole, J. Am. Chem. Soc., 52:29 (1930).
4Dole, The Glass Electrode (New York: John Wiley & Sons, Inc., 1941); Dole,
Measuring pH with the Glass Electrode, Bui. 371 (Maywood, Ill.: Coleman Electric
Co., 1937); also Haugaard, J. Phys. Chem., 45: 148 (1941).
 41
HYDR OGEN ACTIV ITY DETER MINA TION FOR SOILS
calomel
in which ! is the glass membrane and 11 is a liquid junction. The left
bulb
half-cell (often replaced by an Ag, AgCl electrode) is within the glass
bulb.
(Fig. 3-2) and dips into the dilute acid of an 1 + that fills the glass
ely
The right calomel half-cell is the reference electrode that is separat
n of aH + that is to be measur ed. The voltage ,
placed in the outside solutio 2

E, of the glass electrode half-cell ideally is given by:

(3-2)

sion of
in which 0.0591 6 is the function RT /nF and the ln-to-log 10 conver
2.3026 , R = 8.313, F = 96,500 , n = I for H+, and T = 298°K (for
E develop s across the glass mem-
25 °C.) An "''asymmetric potential," Al"
the two
brane even when solutions of the same H+ ion activity are on
sides. For this reason the meter must be calibra ted with a standar d buffer.
it (a)
3-8. The glass electrode pH meter has the distinct advantages that
system under measur e-
does not expel dissolved gases such as C0 2 from the
colored solution s,
ment, (b) is adaptable to thick fluids, pastes, gels, and
t to quin-
(c) is not affected by oxidizing and reducing solutions in contras
c
hydrone or hydrogen electrodes, ( d) does not require H 2 gas or a catalyti
l
surface required by the H electrode, or the addition of auxilliary materia
for
as the quinhydrone electrode, ( e) has a relatively low salt error (except
and (f) after standar dizatio n, gives an accurac y
Li and Na above pH 9.5),
care, within 0.02 pH unit, and (g) is rapid,
within 0.1 pH unit and with
convenient, inexpensive, and adaptable for continuous recording.

MEASUREMENT OF SOIL pH
( Electrom etrically by means of the glass electrod e)
n:
3-9. Soil pH Defined. Soil pH is conventionally defined by the equatio

Soil pH = log --1- = - log 111aH + (3-3)

aH+

ed as
in which the activity of H + in the soil suspension, an+, is express
en ions include s
gm-ions per liter. The "effective" concentration of hydrog
and those
all sources such as those arising by dissociation of soluble acids
of
dissociated from soil particles. "Effective" is defined" in practical terms
elec-
the electrical potential (equation 3-2, ~ 3- 7) measured with the glass
vely by
trode applied to the soil system. Soil pH is measured almost exclusi
the glass electrode because of its advanta ges c,;
3-8) over the hydrog en
electrode or quinhydrone electrode.
soil
3-10. Factors Affecting the Measurement of Soil pH. Because the

5 Bates, Electrometric pH Measure ment (New

York: John Wiley & Sons, Inc.,
1954), p. 31.
42 HYDROGEN ACTIVITY DETERMINATION FOR SOILS
pH measure varies 6 widely with the method of preparation of a given soil,
the details of the preparation procedure must be carefully specified with
any soil pH data. In the preparation of the soil system the principal vari-
ables that affect the pH measurement are drying 7 of the soil sample in the
preparation, the soil water content used, the content of soluble salts, the
content of C0 2 as influenced by season 8 or drying, the amount of grinding
given the soil, and the field variation from core to core-w hich is best
handled by composite sampling (~ 2-6).
3-11. Measurement of the pH soil samples directly in the field moist
condition may be considered the most valid in terms of the existing soil-
biological environment. Measurement of air-dried soil samples is the most
convenient and generally used, and perhaps could be considered the stand-
ard procedure. There is reason to believe that certain soil chemical reac-
tions are hastened by the drying process and that dried samples are there-
fore more nearly at equilibrium. Whether dried or field moist samples were
employed for the soil pH determination should always be stated with the
tabulated data.
3-12. Effect of Water Content on Soil pH Measured. In general, the
more dilute the soil suspension, the higher the soil pH value found, whether
the soil is acid or alkaline. 9 The rise in soil pH with dilution, from the
sticky point to a soil : water ratio of 1 : 10 is usually of the order of 0.2 to
0.5 pH unit, but may be 1 or more pH units (Fig. 3-3) in certain neutral
and alkaline soils. 10 A rise was observed11 from pH 6.45 at 6.3 per cent
moisture in a calcareous soil to pH 8.60 at a 1 : 5 soil : water ratio. An
abrupt rise in pH occurs with alkaline soils (of pH 7 .5 and above) just be-
fore the moisture saturation percentage (~ 3-22) is reached. Thus dilution
to the saturation percentage, although convenient in soil preparation and
electrode insertion, involves some shift in pH for alkaline soils, though less
than for greater dilutions often employed.
3-13. Junction Potential. According to some workers, an appreciable
liquid junction potential, Ej, may occur at the liquid junction (Fig. 3-2),
particularly in a very concentrated colloidal electrolyte suspension. The
12

liquid junction potential is kept low (nearly zero) and as constant as pos-
sible by the use of a saturated KC! solution (or sometimes NH 4N0 3 when

Reed and Cummings, Soil Sci., 59: 97 ( 1945).

11
with
7McGeorge, Ariz. Agr. Exp. Sta. Tech. Bui. 57 (1935) noted a rise of pH
a rise was also observed in the author's laborator y with drying
drying calcareous soils;
unlimed acid soils.
B Baver, Soil Sci., 23:399 (1927).
D McGeorge, Ariz. Agr. Exp. Sta. Tech. Bui. 78
(1938).
19 Chapman et al., S.S.S.A. Proc., 5: 191 (1941 ).
11 Huberty and Haas, Soil Sci., 49:455 (1940).
12 Jenny et al., Sci., 112: 164 (1950); Coleman et al., S.S.S.A.
Proc., 15: 106 (1951).
In disputatio n: Marshall , S.S.S.A. Proc., 15:110; Sci., 113:43 (1951), 115:361
(1952); and Peech and McDevit , S.S.S.A. Proc., 15: 112 (1951).
 43
HYDR OGEN ACTIV ITY DETER MINA TION FOR SOILS

10.0

J:
a.

Aiken cl. lo.

S=sticky point
F=flow point
(Sat. %)

:2----''--- -1...l.:3_ _..___.....

3.0L--l:..J.0.-5--1'-:1_ __._ _1.... 1:4 _ _.....__-:-"'1:5

Soil:water ratio
ratio.
Fig. 3-3. Relation of observed soil pH values to soil : water
Chapma n et al., S.S.S.A. Proc., 5:191, 1941; dashed lines
(Solid lines from
from Huberty and Haas, Soil Sci., 49:455, 1940.)

equal
K would interfere). The anion and cation of these salts have nearly
con-
mobilities in solution. The junction potential does not remain exactly
because ions other than K and Cl are present
stant even in solution systems
Any
in the junction to contribute to the transpo rt of electrical charge.
the potenti al caused by H ions
liquid junction potential acts additively to
t

increas e in soil suspen-

acting on the glass electrode. An increase in Ej with
ed
sion concentration would correspond to a decrease in apparently measur
utes to the
soil pH. To the extent that a liquid junction potential contrib
nts
lower pH measured in thick soil suspensions, it detracts from the argume
e
for measurement of the soil pH in thick suspensions in order to simulat
field conditions.

APPARATUS
with
3-14. Needed apparatus includes a glass electrode pH meter
beakers , short stirring
calomel reference electrode and salt bridge, 50-ml
rods, spatula, and a distilled water wash bottle.

REAGENTS
and
3-15. Needed reagents are a standard buffer solution of pH 4.00
d, and
buffers of other pH values in the range of the soil pH values expecte
saturated KC! (about 40 gm per 100 ml) for the bridge.
 SOILS
44 HYD ROGE N ACTI VITY DETE RMIN ATIO N FOR
for calibration
3-16. Standard Buffers. The standard pH 4.00 buffer
mol. wt. 204.1 4). The re-
is 0.05 M potassium biphthalate (KHC 8H 40 4 ,

, and is sufficiently non-

agent grade has a purity of 99.9 per cent or better
soluti on of 0.3 M
hygroscopic to permit weighing without drying. A stock
liter of hot water.
is prepared by dissolving 61.2 gm of KHC 8H 4 0 4 in one
h. A dilutio n of
Three drops of toluene are added to discourage mold growt
a 0.05 M solution.
100 ml of the buffer with 500 ml of water results in
u of Standards
Buffers of various pH values that meet National Burea
ercially (Leeds
(Washington, D.C.) specifications can be obtained comm
The "certif ied buffer tablet s" of Cole-
and Northrup, Philadelphia, Pa.).
e an accura cy of ±0.02 pH at
man Electric (Maywood, Illinois) provid
distilled water. A 0.1 M KH-
30°C when one is dissolved in 100 ml of
of dilutio n witho ut an
phthate solution of pH 3.98 has a wide tolerance
per cent diluti on). Snell
appreciable change in pH (0.02 pH unit for 100
g 19.45 ml of
and Sneurn state that pH 8.0 may be obtained by mixin
0.2 M Na~HP0 4 with 0.55 ml of 0.1 M citric acid.

PROCEDURE
at the moisture
3-17. Preference is here given to pH determination
rcutine wetting of
saturation percentage (~ 3-22) because of the ease of
moisture content
various soils to "equipotential" moisture status. This soil
re films are thick
is the highest likely to occur in the field. The moistu
the glass electro de. Small changes in
enough to give good contact with
, so that only moderate care
dilution cause little change in pH ( ~ 3-12)
t of the soil for the pH
need be exercised in adjusting the water conten
y in the routine
measurement. This procedure has been used with facilit
(~ 3-25) is also
soil testing laboratory. The use of a 1 : 1 soil: water ratio
s other moistu re
common. The determination of the pH of a soil at variou
ered (~ 3-24)
contents, from the sticky point to a I : I 0 ratio, is consid
because of varying practi ces in different labora tories .
20 cores) is pre-
3-18. The Soil Sample. A composite soil sample (10 to
ng, and tested ( ~ 3-22) at once or
pared, with avoidance of severe grindi
14

nt conta ct with traces of ammo-

placed in a jar and sealed tightly to preve
nia and other laboratory fumes that would alter its pH.
The operational
3-19. Precautions in the Use of the Glass Electrode.
the manu facturer's in-
details for various instruments are obtained from
t to excess ively rapid
structions. The glass electrode is fragile and subjec
sive to replace.
deterioration if not properly cared for. It is also expen
These precautions help:
D. Van Nostra nd Com-
13 Colorimetric Metho ds of Analysis, Vol. l (New York:
pany, Inc., 1936), p. 677.
1 4 Baver, Soil Sci., 23: 399 ( 1927).
noted a 1.0 to 1.3 pH unit rise in soils of pH
4. 7 as a result of grinding from 2 mm to 0.16 mm ( l 00 meshes per inch).
HYDROG EN ACTIVIT Y DETERM INATION FOR SOILS 45

1. The electrode is not allowed to remain in the test solution or suspen-

sion longer than necessary, especially if more alkaline than pH 9.
2. Immediately after testing, the electrode is washed off with a strong
stream of distilled water from a wash bottle. If the system was alkaline, the
electrode should be dipped for a few seconds in the acid pH buffer or
dilute HCI to remove the film of CaC0 3 that sometimes forms.
3. For storage, after cleaning, the electrode is suspended in distilled
water, and the system is protected from evaporation. Drying out of the
electrode is avoided.
4. Failure of the glass electrode pH meter is indicated when, after stand-
ardization, it gives slow response to large pH changes. The glass electrode
is immersed in an alkaline buffer, then in the original acid standard buffer.
Readings of pH values a few tenths higher than the specified pH values of
the standard after as little as 60 seconds equilibration indicates "etching"
or an over-age glass membrane.
3-20. The temperatu re setting is adjusted, and a little KCl is passed
from the junction followed by flushing with distilled water. Then the stand-
ard buffer, phthalate of pH 4.00, or another buffer of pH near that of the
systems to be determined, is placed in the electrode vessel, and the glass
electrode and calomel half-cell with KCJ junction are immersed in the
buffer. The instrumen t's pH dial is set at the known pH value of the stand-
ard buffer. After a suitable elapse of "warm-up " time, the instrumen t is
balanced to eliminate the asymmetric potential. The buffer is removed, and
the electrodes are carefully flushed off with water.
3-21. The pH Test. A test solution 10 or soil suspension (~] 3-22) is now
placed in the electrode vessel, and the pH dial is turned until galvanometer
balance is reached. This test is repeated before each new sample is tested
until stability is assured, then after each fifth determinat ion. The pH meter
is turned off when the series of readings is completed.
3-22. Water-Saturation Percentage 16 Preparation. A volume of approxi-
mately 40 ml of -2-mm or -4-mm soil is placed in a 50-ml beaker (which
is about 0.7 filled with soil in lieu of weighing). Water is added to the soil
in increments from a wash bottle. It is more convenient to prepare a series
of several samples simultaneously. Successive increments of water are
added without stirring the soil until water just wets the entire soil mass,
then a few drops more are added slowly until the surface glistens slightly.
Stirring of the soil is avoided until it is fully wetted to prevent forming of
a puddled mass through which water moves only very slowly; most of the
wetting is thus accomplished through the undisturbe d pores. After this
moisture content has been attained, the soil is stirred with a glass rod, and

15 An instructive exercise on the glass electrode pH meter is to titrate 20 ml of

0.1 M HaP04 with 0.1 N NaOH delivered from a buret, the pH being determined at
frequent intervals as the titration proceeds.
rn Richards, ed., Diagnosis and Improvemen t of Saline and Alkali Soils, U.S.D.A.
Handb. 60 (1954),p. 17.
46 HYDROGEN ACTIVITY DETERMINATION FOR SOILS
drops of water are added until the soil is a "thin paste" that just barely
flows together to close around a hole left by the rod. The soil is now at the
"moisture saturation percentage," 17 an equipotential moisture content for
soils (~ 10-35). The surface of the water-saturated soil glistens, and air
has been excluded. The soil is at the "fl.ow point" or "liquid limit," or
slightly wetter than the "upper plastic limit." The U.S. Salinity Laboratory
allows the wetted soil to equilibrate for one hour before the pH measure-
ment is made. For most practical soil testing, the pH reading may be taken
right away.
3-23. The glass and calomel electrodes are inserted in the water-satu-
rated soil, and the pH measurement is made. The glass electrode is moved
about to insure removal of the water film around the electrode, and the pH
reading is again taken. When the reading is constant or nearly so (the C02
status of some soils gradually changes with time and therefore undue delay
should be avoided), the pH value is recorded. Whether field-moist or air-
dried samples (~ 3-11) were used is also recorded.

ALTERNATIVE PROCEDURES

3-24. Several alternative procedures, involving more concentrated or

more dilute suspensions (~ 3-12), are presented because each is widely
employed in different laboratories, and fairly extensive pH changes occur
with dilution. Stirring of water suspensions of soil is necessary18 to keep
the soil suspended during the pH measurement. Some methods involve the
addition of a strong electrolyte or buffer.
3-25. Soil : Water Ratio of 1 : 1.19 To a 20-gm sample of soil in a 50-
ml beaker is added a 20-ml volume of distilled water. The suspension is
stirred at regular intervals for about an hour. Then the pH is measured
with the glass electrode, the suspension being stirred well just before im-
mersing the electrode.
3-26. Soil: Water Ratio of 1 : 2.5.20 To a 10-gm sample of soil in a 50-
ml beaker is added a 25-ml volume of distilled water. The suspension is
stirred at regular intervals for 20 to 30 minutes. Equilibrium can be at-
tained in 5 minutes of continuous vigorous shaking with a reciprocating

17 Scofield, U.S.D.A. Cir. 232 (1932).

18 Bailey, Soil Sci., 55: 143 (1943); Miles and Reed, in Diagnostic Techniques
(Washington, D.C.: American Potash Institute, 1948), p. 93.
10 Biilman and Jensen, Trans. 2nd Comm. Intern. Soc. Soil Sci., B:236 (1927);
Purvis, J.A.O.A.C .. 23:219 (1940); Peech et al., U.S.D.A. Cir., 757:5 (1947); Reed
and Cummings, Soil Sci., 59: 103 ( 1945).
2osoil Reaction Committee, Intern. Soc. Sci., Soil Res., 2:241 (1930). Hester, in
Diagnostic Techniques (Washington, D.C.: American Potash Institute, 1948), p.
112, employed a I : 2 water ratio, and took 30 gm of soil in an effort to decrease the
possibility of soil heterogeneity.
HYDROGEN ACTIVITY DETERMINATION FOR SOILS 47
mechanical shaker. Then the pH is measured with the glass electrode, the
suspension being stirred well just before the electrode is immersed.
3-27. Soil : Water Ratio of 1 : 5.21 To a 20-gm soil sample in a flask,
100 ml of aerated distilled water is added and the soil is stirred for one
hour. The pH value is then determined within 60 seconds after the glass
electrode is immersed in the freshly shaken suspension.
3-28. Soil : Water Ratio of 1 : 10. To a l 0-gm soil sample in a flask,
100 ml of distilled water is added and the soil is agitated for one hour. The
pH value is then determined with the glass electrode. The pH of the 1 : 10
suspension is termed the "hydrolytic pH value" by McGeorge. 22 The hy-
drolytic pH value increases with degree of Na-saturation. The 1 : 10 meas-
urement has the advantage that all Na-saturated soils have the same pH
value regardless of exchange capacity (whereas the pH of Na soils are not
the same at the sticky point).
3-29. Soil: Water Ratio of 1 : 2 in 0.01 M CaCl2• 23 To mask the vari-
ability in salt content of soils, to maintain the soil in a flocculated condition,
and to decrease the junction potential effect (, 3-13), the soil pH meas-
urement is made in 0.01 M CaC1 2 solution. Twenty-five gm of soil is sus-
pended in 50 ml of 0.0 I N CaCl 2 solution and the suspension is stirred
thoroughly. The pH measurement is then made in the usual way. The soil
pH scale is shifted downward in this solution. Acid soils range from pH 4
to 5.0, pH 6 is the optimum pH of limed soils, and calcareous soils run
pH 6.9 to 7.1 in the CaCl 2 solution.
3-30. Soil: Water Ratio of 1: 2.5 with l N KCI Present. Because
leached soils showed little change in pH with dilution, the effect of dilution
was attributed 24 to the presence of small amounts of salts. To overcome
large relative changes in salt content, 1 N KCI was added. The observed
pH values were 1.5 units lower than those obtained in water suspension.
Varying soil : water ratios were employed from 1 : 2.5 to 1 : 25, and the
dilution had to be stated because the pH changed 0.6 to 1.0 unit with dilu-
tion even in the presence of 1 N KCI.
3-31. To a 10-gm soil sample in a 50-ml beaker, a 25-ml volume of
1 N KCI is added and the soil is stirred intermittently for one hour. The
pH is measured with a glass electrode.
3-32. Isohydric pH Value. 20 The isohydric pH value of a soil is that of
a buffer solution that shows no change in reaction on coming into contact

2 1 Division of Soils, C.S.l.R. Australia method, according to Piper, Soil and Plant
Analysis (New York: Jnterscience Publishers, Inc., 1944), p. 9.
22 Soil Sci., 59:271 (1945).
23 Schofield and Taylor, S.S.S.A. Proc., 19: 164 ( J 955 ).
24 Puri and Asghar, Soil Sci., 46: 249 (1938).
211 Puri and Sarap, Soil Sci., 46:49.(1938 ).
48 HYDROGE N ACTIVITY DETERMIN ATION FOR SOILS
with the soil and "which consequently brings about no change in the base
content of the soil." A series of buffers (HOAc-KOAc in the acid range;
H 8BOa, KCl, and KOH in the alkaline range) are equilibrated with soil
samples for a period of 2 hours. A 1 : 5 soil : buffer ratio is employed (no
difference is found up to 1 : 20). The isohydric pH can thus be obtained by
determining the pH of the series and interpolating the isohydric pH value.
3-33. Sticky Point Method. 211 Approximately 40-gm of soil is placed in a
50-ml beaker. (The beaker may be about 0.7 filled with soil in lieu of
weighing.) Water is added in increments from a wash bottle, the soil being
allowed to imbibe each increment before the next is added, until only about
5 or 10 ml of the soil remains unwetted. About one minute is allowed for
wetting to take place, while other soils in a series are being processed.
Then the soil is stirred and puddled with a spatula to make a stiff paste.
The "sticky point"27 is reached when the soil is just wet enough to stick
firmly to a spatula pressed to the soil mass and pulled directly away from
it. The soil remains "sticky" on the wet side of the sticky point, so this
moisture content should be approached from the dry side. If the soil is too
dry, additional drops of water are added, the soil is stirred, and the "sticky
point" test is repeated. A definite threshold force is required to break
("click") the spatula away from the soil when the sticky point is reached.
More soil may be added if the sticky point moistness is exceeded. Clay
soils, which usually exhibit considerable rigidity at the sticky point moisture
content, were moistened slightly more by Chapman, et al., to make them
soft enough to facilitate insertion of the glass electrode.
3-34. The soil paste is equilibrated for l 0 minutes. Then a slit is opened
in it with a spatula, the glass electrode and salt bridge are inserted and the
paste is formed to cover the active portions of the electrodes completely
and left 60 seconds for "conditioning" of the electrode. The pH value is
then read and checked for constancy.

AVERAGE SOIL pH VALUES

3-35. The basis for the use of average soil pH values lies in the observa-
tion that:
Soil pH, composite sample= Soil pH, cores (~r subcomposites) (3-4)
n
in which n is the number of separate pH values of cores also represented in
the composite sample. The right hand expression in equation 3-4 is the
average of soil pH values. Some opposition to the averaging of pH values

20 Chapman et al., S.S.S.A. Proc., S: 196 ( 1941).

27 Keen and Coutts, I. Agr. Sci., 18:740 (1928).
HYDROGE N ACTIVITY DETERMIN ATION FOR SOILS 49

exists because their being log au+. Actually many useful (and frequently
averaged) parameters have exponential relationships to oth~r variables.
(The average soil pH value, incidentally, is the geometric mean aH +.) The
relationship expressed by equation 3-4 is supported to an accuracy of 0.02
pH unit by published data28 and by the data of L. F. Marriott of this labo-
ratory with limed and unlimed Plainfield sand soil samples. As would
therefore be expected, averaging soil pH values gives a good estimate of
the meq of exchangeable hydrogen and lime requirement of an acid soil;
such estimate is very much better than average a 11 +. This conclusion is
equivalent to the often noted fact that the portion of the pH titration curve
involved in calculation of lime needs of a given soil approximates a straight
line. Chapman et al., placed two samples of soil, one acid and one alkaline,
on two sides of the glass electrode bulb. The pH value read was the arith-
metic mean of the two soil pH values. The soil pH value of a composite
sample or of an individual soil core is evidently the arithmetic mean pH
value of the various kinds of soil particles in a sample, such as of calcite or
feldspar (alkaline) and kaolinite or allophane (normally acid).

COLORIME TRIC pH INDICATORS USEFUL IN SOIL ANALYSIS

3-36. A colorimetric pH indicator is an organic dye, the color of which
is controlled by the hydrogen ion activity in solution. During analytical de-
terminations they provide a continuous check on solution pH. They are
useful in rapid testing of soil pH (~ 13-91) in the field and laboratory. A
set of the most useful and sensitive pH indicators (Table 3-1) may well be
prepared at one time, and stocked in the laboratory in a convenient dropper
bottle rack.
3-37. Color Range and Critical pH. The shift in pH indicator color re-
sults from a reaction of the dye with ions of the solution. Many indicator
dyes form weakly colored, little dissociated acids and highly colored,
strongly dissociated salts of metallic cations. Weak base forming indicators,
such as methyl red, are strongly colored in acid solutions. The full color
range of almost every colorimetric pH indicator is approximately the same,
namely ± one pH unit from midcolor for ±90 per cent of the color change
(Fig. 3-4), as determined by spectrophotometric color analysis 29 of in-
dicator titration curves. This fact arises from the relationship of color to
dissociation constant, K, of the indicator molecule. For example, a ten-
fold increase in [H +] (one pH unit decrease) changes the concentration
ratio of [salt]/[acid] by 10/100, or roughly 90 per cent toward a complete
change to acid color. The uniformity of behavior of pH indicators permits
28Chapman et al., S.S.S.A. Proc., 5: 191 ( 1941 ).
29Kolthoff, Colorimetric and Potentiometric Determination of pH (New York:
John Wiley & Sons, Inc., 1931), p. 26.
50 HYDROGEN ACTIVITY DETERMINATION FOR SOILS
pH

O.IM Na3P04 -------- -

0.11:![ NaOH_--------13-.------,,.------,

12 1 - - - - - - + - - - - + - t
O.IM Na2C03--------
Xpoint -of sharpest
Critical pH value ancJ
color change

Malachite green, alkaline range

Alizarin yellow R
10

Phenol red
tt•=oH-"neutrality, "·- 7 -h'~'---..,,,~--7fl"-+-I Brom thymol blue
conductivity water at 22"C
.Sat. Ca(HC03)2 }-- ---- Brom cresol purple
1 aim. C0 2J 6 t-:ir--~,,.__..,,,,,.~---ri
Methyl red
H2C03 in water,}-_ - - - Chlorphenol red
exposed to air
P·nitro phenol
5 HF-+----+-,~~~
O.IM NaH2P04------- -~"_,,._.+-Brom cresol green

0.IM KH-phthalate----4 .-.--..A::-7'"711~----!-Brom phenol blue

H 2C0 3 in water,}--' -- Methyl orange
1 aim. C02 2,6·dinitro phenol
0.002tj H2S04 }_,-3
( NH4)2S04 buffered
IN HOAc----------
O:itj KHSOr -- - -- - 2

O.ltj HCI - - - - - - --1 _ _,,,'---f--'.,...,_,.,,,,---t-Malachite green, acid range

Picric acid

Percentage of dissociation

Fig. 3-4. Critical pH values of selected colorimetric pH indicators,

shown on titration curves in relation to degree of dissociation and
pK values. Data adapted in part from Clark, Determination of Hydro-
gen Jons, Ed. 2, Williams and Wilkins, Baltimore, Md. (1923);
Snell and Snell, Colorimetric Methods of Analysis, Vol. l, p. 683,
Van Nostrand, New York (1936); Mellon, Colorimetry for Chemists,
p. 33, G. F. Smith Chem. Co., Columbus, 0. (1945); "Manual of
Methods for Pure Culture Study of Bacteria,'' Soc. A mer. Bact., Leaf.,
9:8 (1941); and from the data of the author (1946).

the identification (Fig. 3-5) of each indicator by a critical property, the pf

of most pronounced color change, termed the "critical pH value". 30 Th
critical pH value lies 0.1 to 0.3 pH unit from the pK value toward t.li
weaker color for 2-color indicators and 0.6 to 1.0 pH unit toward tt
colorless side for the single color indicators. The pH value of a solutio
can be estimated visually by the increment of indicator color change fro1

ao This is preferable to listing a pH range, which involves arbitrary assignment l

numbers to the two asymptotic ends.
HYDROGE N ACTIVITY DETERMIN ATION FOR SOILS 51

that at a critical pH, and thus dependence r-----------------1

: pH 4.0 I
on a color chart is eliminated for most l
I
!
BROM PHENOL BLUE
Y-Pu-V I
purposes. I I
I
I
: 0.04%Aq. J.Doe,1958!
REAGENTS L-----------------~
3-38. Aqueous Solutions of the Sulfon· Fig. 3-S. Specimen label for
colorimetric pH indicator bottle,
phthaleins (0.04 Per Cent). Aqueous with critical pH value (pH 4.0,
solutions of the sulfonphthaleins and simi- midcolor, Pu = purple) used as
larly acting dyes (No. 2, 3, 6-8, 11-14, central identification. The color
of changes are indicated for more
and I 8) are prepared by dissolution acid side (Y = yellow) to the Jess
0.4 gm of the dry indicator powder (acid acid side (V = violet).
dye) in 5 to 15 ml (Table 3-1) of 0.1 N
NaOH by trituration in an agate mortar. An hour or more may be required.
Although the process may be hastened by the addition of 10 ml 95 per cent
ethanol, an aqueous solution is preferred. The low surface tension of the
alcoholic solutions is generally a disadvantage, and the Na salts of sulfon-
phthalein type dyes are water soluble. The solution is washed into a 2-liter
beaker, diluted to 1 liter, and then HCl or NaOH is added dropwise to
bring the indicator to the midpoint or critical pH (~ 3-37). The color
change is more easily observed in a dilute solution, and therefore the color
is judged as a drop diluted with distilled water on a spot plate.
3-39. Caution. These indicator solutions will be contaminated with Na
and CI ions. Other anions and cations should be scrupulously kept out. For
work involving determination of Na or Cl especially prepared indicators
are used, for example, alcoholic solutions of the acidic dyes.
3-40. Aqueous Solutions of the Nitrophenols (0.25 Per Cent). Aqueous
solutions are prepared of the nitrophenols (mono, di, tri, or picric acid; Nos.
1, 4, and 10) by dissolution of 2.5 gm of the acid in 1000 ml of distilled
water, warming it on the steam plate. These indicators are ordinarily left
acid. However, for special work it may be desirable to adjust the pH to the
critical pH value by dropwise addition of dilute NaOH.
3-41. Alcoholic Solutions of the Phthaleins (0.2 to 1 Per Cent). Alco-
holic rather than water solutions of the phthaleins (phenolphthalein, thy-
molphthalein, Nos. 15 and 16) are required because, even with some NaOH,
re phthalein indicators are extremely insoluble in water, and tend to pre-
:tpitate out. The alcoholic solutions are prepared by dissolution of 2.0 to
p gm of the dry indicator powder (acid dye) in 1000 ml of 95 per cent
"1anol. For phenolphthalein, 1 per cent or 10 gm per 1000 ml is prepared.
)br thymolphthalein, 0.2 per cent or 2.0 gm per 1000 ml is satisfactory.
~· 3-42. Water Solution of the Sodium Salt Dyes. The water soluble sodium
~It forms of methyl orange (No. 5) and alizarin yellow R are available
.".Jo. 18); the solutions are prepared by simple solution of the salts in water
at the concentrations listed (Table 3-1).
 TABLE 3-1
Selected list of colorimetric pH indicators useful in soil analysis.* The following abbreviations designate colors:
B = blue; C =
colorless; G =
green; 0 =
orange; P pink; Pu = purple; R =
red; V =
violet; =
Y = yellow. The complete set includes 17 indicators, since Nos. 16 and 19 repeat Nos. 3 and 2

Color change 1 Indicator

Critical (center color NaOHf Ii concentration
pH is at critical Mot. 0.1 N to use
Common name Chemical name value+ pH value) wt. (ml) (per cent)
-------·---; -----
I. Picric acid 2.4,6=Trinitrophenol 0.4 C-Y-Y 299
2. Malachite green Phenyl di-p-dimethyl amino phenyl carbinol 1.0§ Y-YG-BG
** 0.25
373 10.7 0.04
3. Thymol blue Thymol sulfonphthalein 1.9 R-0-Y 466 8.6 0.04
4. Dinitrophenol 2,6=Dinitrophenol (or 2,4) 3.1 (2.7) C-Y-Y 184 0.25
5. Methyl orange Dimethylamineazobenz ene sodium sulfonate 3.7§ R-Y-0
**
6. Brom phenol blue Tetrabromo-phenol sulfonphthalein 4.0 Y-Pu-V
** ').J
669 6.0 0.04
----·--------------------·---- ·-

7. Brom cresol green Tetrabromo-m-cresol sulfonphthalein 4.6 Y-G-B 698 5.7 0.04
8. Chlor phenol red Dichloro-phenol sulfonphthalein 5.6§ Y-OP-V 423 9.5 0.04
v. 9. Methyl red Dimethylamino-azo-ben zene o-carboxylic acid 5.7§ R-0-Y 269 tt 0.125U
N I 0. P-nitrophenol ( m-) P-nitrophenol ( m-) 5.2 (6.7) C-Y-Y 139 ** 0.25
11. Brom cresol purple Dihromo-o-cresol sulfonphthalein 6.2 Y-Pu-V 540 7.4 0.04
12. Brom thymol blue ' Dibromo-thymol sulfonphthalein
'1'

______ ._ __ 6.9 Y-G-B 624 6.4 0.04

------------
13. Phenol red
--i--·-
Phenol sulfonphthalein 7.3 § Y-RO-V 354 I 1.3 I 0.04
14. m-Cresol purple m-Cresol sulfonphthalein 8.3 § Y-0-R
15. Phenolphthalein 382 10.5 0.04
Phenolphthalein
1·

8.3 C-P-P 318 tt 1.0

-------- ---------------- -·-- ------- --·---·-- '------· ------
16. Thymol blue (See No. 3) 8.9 Y-Pu-BV 466 8.6 I 0.04
17. Thymolphthalein Thymolphthalein 9.4 C-B-B 430 tt I 0.2
18. Alizarin yellow R Nitrobenzene-azo-salicy lic acid 10.3 Y-0-R 287 13.9 I o.o4
19. Malachite green (SeeNo.2) 11.5H BG-BG-C 373 10.1 1 o.o4
0 These indicators were selected on the basis of laboratory experience. from the more exten'iiYe lists of Clark and Luhs.
Cohen. and the Eastman Kodak Co.
f The critical pH value is that found with the glass dectrode at the point of sharpest color change of the indicator. A kno\\.·ledge of this characteristic value decrea.oSes
dependence on a color chart for indicator interpretation. The end colors are fully developed within I pH unit from the critical pH value and are rather well devel-
oped within 0.6 to 0.8 pH unit of it. Thus the color shift may be subdivided mentally into 3 or 4 color segments to obtain 0.2 pH unit precision.
t Required to convert 0.4 gm to Na-salt.
~ Color change is not sharp; critical pH value represents midpoint of color change.
0 0 Dissolved in water.

tt The acid form dissolved in 95 per cent ethanol.

U A brighter Pu-B-G color change is obtained with 0 .8 per cent methylene blue chloride added.
H Decolorizes above pH 11.5, irreversibly.
HYDROGE N ACTIVITY DETERMIN ATION FOR SOILS 53
3-43. Alcoholic Solutions of Methyl Red Mixed Indicato~. Methyl red-
methylene blue indicator (critical pH 5.6, G-C-Pu) is made by dissolution
of 1.2 gm of methyl red and 0.8 gm methylene blue chloride in 1000 ml of
95 per cent ethanol. This mixture is especially adapted to acid base titra-
tions in which only a small amount of C02 is likely to be present, as in the
case of the Kjeldahl method for nitrogen.
Brom cresol green-methyl red indicator (critical pH 4.53, R-G-B) is
made by dissolution of 5 gm of brom cresol green and 1.0 gm of methyl
red in 1000 ml of 95 per cent ethanol and neutralization to midcolor with
drops of 1 NNaOH (118-10).
3-44. Indicator Test Papers. Filter paper strips soaked in various of the
indicators (e.g., litmus paper) or their mixtures may be inserted in solution
or in contact with a paste to test pH. A complete set is sold as "Hydrion"
papers (Micro Essentials Laboratories, Brooklyn). Indicator paper strips
carrying a printed color chart are available (A. Daigger Co., Chicago IO).

PROCEDURE

3-45. Choice of pH Indicator for a Specific Use. For the colorimetric

determination of a soil pH, an indicator is chosen with a critical pH value
as close as possible to the pH to be measured. In some systems this is
unique, as for example, brom cresol purple for the pH 6.4 NH4 0H separa-
tion ( 11 11-136). From superficial consideration, pH 7 is often chosen as
the end point for acid-base neutralization reactions, but further study shows
that end points in specific titrations may be anywhere from pH 3 to 9, or
even outside this range. Standardization of NaOH solution by neutralization
of standard KH-phthalate requires titration to pH 8.3 (phenolphthalein).
If C0 2 is to be produced, as in standardization of HCI against Na 2 COa, and
is to be boiled out, an indicator of critical pH 4 to 5 (brom phenol blue or
methyl red) should be used, but if C02 is not to be boiled out, a critical pH
of 3 (methyl orange or dinitrophenol) is more appropriate. Use of a critical
pH of 7 to 8 is just as accurate finally after C02 is removed, but makes the
end point laborious to find (Fig. 3-6). Boiling only once permits find-
ing the end point with brom phenol blue, but boiling may need to be re-
peated 4 or more times with phenolphthalein (pH 8.3), or brom thymol
blue (pH 7.0).
3-46. An Exercise on the Use of pH Indicators. The various pH indi-
cators (Table 3-1) determine the pH value of each of a number of solu-
tions prepared by mixing 1 gm of various salts with 100 ml of water (Table
3-2). Some of the materials are not soluble enough to go entirely into solu-
tion. The pH values are also checked by means of the glass electrode
(11 3-21).
3-47. The pH value found in each solution is to be explained by means
of ionization and hydrolysis equations.
54 HYDROGEN ACTIVITY DETERMINATION FOR SOILS

14i--~~~~--.-~~~~~....-~~~~~....-~~--~ .....
13
12
11 pH rises when C02
10 is boiled out
9
:a s7
6
5
4
3
20 25 50 75 90 100
ml Of 0.11::! HCI added

Fig. 3-6. Use of indicator with a critical pH of 4.0 or below is advantageous

in a titration of a carbonate with acid.

TABLE 3-2
Detennined pH values of various solutions prepared with
1 gm of solid mixed with 100 ml of water
·~-------··- -- -----··-------
Name of appropriate pH value, found I pH value, by
Salt pH indicator by indicator glass electrode
- --------·---·-- --- ----------
1. KHS04 ----- - - - - - -
2. NaHCOs -----------
3. NaCl
4. Na2COs
5. CaCOa
--
6. CaS04 • 2H20
---------
7. Ca(H2P04)2 • H20
8. Ca2CHP04)2 • 4 H20
9. FeCla • 6 H20

OXIDATION-RED UCTION INDICATORS USEFUL IN

SOIL ANALYSIS
3-48. To complete the discussion of colorimetric indicators, the 3 most
valuable colorimetric oxidation-reduction indicators must be included, al-
though they are not clearly a part of hydrogen ion activity measurement.
Colorimetric oxidation-reduction indicators are available to detect rapid
shifts in the oxidation potential of solutions employed in analysis. The mid-
point of the oxidation-reduction indicator color change is analogous to the
critical pH change of the pH indicators. In the oxidation-reduction indica-
tors, the shift in color occurs at a specific oxidation potential (Table 3-3).
Either diphenylamine or "Ferroin" is employed in the chromic acid meth-
ods for organic matter (~ 9-54, 9-67); "Ferroin" is employed for sulfato-
cerate titrations ( ~ 5-43, 6-49) .
HYDROGEN ACTIVITY DETERMINATION FOR SOILS 55

3-49. When solutions are highly colored, the use of colorime tric indica-
tors is difficult. Potentiom etric measurem ent of the endpoint is then more
appropria te. A special "magic eye" potentiom etric endpoint indicator has
been offered (G. F. Smith Chemica l Co., Columbu s, Ohio).

TABLE 3-3
Three useful colorimetr ic oxidl.ltion-reduction indicators*
Oxidation
potential
Color change at color Preparation of
Indicator upon oxidation change* solution
----------
1. Methylene blue Blue to colorless -0.53 0.2 gm per I 00 ml of
(pH 2.86) H20
2. Diphenylamine Colorless to -0.76 1 gm in I 00 ml of
violet concentrated H 2S04
3. "Ferroin" ( Orthophenanth- Red to faint blue -1.06 1.487 gm with 0.695
roline ferrous complex) gm FeS04. 7 H20 in
100 ml of H20.

o Numerical values from Smith and Richter, Phenanthrol ine and Substituted
Phenanthrol ine Indi-
cators (Columbus, Ohio: 1944), G. F. Smith Chemical Co., pp. 28-29.

3-50. Starch Indicator for Iodimetr y. One gm of soluble starch is stirred

in 5 ml of cold water in a beaker. The resulting smooth suspensio n is
poured with constant stirring into 100 ml of boiling water containin g 3 gm
of boric acid preservative. This solution is cooled and transferre d to a
100-ml scalded dropper bottle provided with a 2- or 3-ml calibrate d drop-
per. The bottle is stored in a dark, cool place. A 2- or 3-ml portion of the
indicator is transferre d into the solution to be titrated (for example, stan-
nous chloride,~ 7-19) and as the 0.1 N 12 solution is titrated in, blue color
first appears in the droplets and disappea rs until at the end (slight excess
12 present) a deep blue color appears througho ut the solution.

QUESTI ONS
1. Why is the measurement of soil pH fundamental to the study of soil·
plant relationships?
2. Identify the contents of the glass electrode and the associated components
of its electrochemical cell.
3. Write the fundamental equation relating electrical potential of a cell con-
taining a glass electrode to activities of hydrogen in solution, defining the terms.
4. State the advantages of the glass electrode for measurement of soil pH.
5. What is the effect of drying soils on the pH value measured?
6. What is the effect observed on the pH of making the soil suspension more
dilute, and what is the relationship to a possible liquid junction potential?
56 HYDROGEN ACTIVITY DETERMINATION FOR SOILS
7. Why is it necessary to use a standard pH buffer in connection with the
glass electrode?
8. Explain why the average soil pH value rather than the mean activity of
hydrogen is often the most valuable datum. (Consider the use of soil pH value
as an estimate of liming need.)
9. Discuss the way in which colorimetric pH indicators respond to change
in solution pH, employing the dissociation constant concept.
I 0. What is the critical pH value of an indicator and how does it aid in find-
ing the pH of a test solution from the observed color?
11. What are the different general classes of methods of preparing pH indica-
tor solutions?
12. Discuss the basis of choice of an indicator pH range in relation to titra-
tions of different specific kinds of acids and hydroxides.
 4

Cation Exchange
Ileterminations for Soils
The cation exchanRe reaction is the second most important in
nature, surpassed in fundamental importance only by the
photosynthetic process of green plants.
-MARSHALL I

4-1. Cation exchange in soils 2 has long been of interest to soil chemists.
Cation exchange determinations for soils include the measurement of the
cation exchange capacity of soils and soil colloids, the measurement of the
total exchangeable metallic cations and percentage exchange saturation of
soils, the measurement of the percentage of saturation with alkali cations
and the gypsum or sulfur requirement of soils, and finally the measurement
of exchangeable hydrogen and the lime requirement of soils. Neutralizing
equivalence of limestone and other liming materials must also be considered
in raising the percentage saturation of soils by liming. The cation exchange
status with respect to groups of cations is the concern of the present chap-
ter, whereas the determination of the amounts and ratios of individual
metallic cation species is the subject of following chapters.
4-2. Washing Techniques. Cation exchange or removal of excess salt
from a soil or colloid is effected by continuous or repeated washings. Con-
tinuous washing may be done with a percolation device such as a funnel,
Gooch crucible, fritted glass crucible or leaching tube ( ~ 4-10). Repeated
washing may be done by repeatedly dispersing the sample in fresh portions
of liquid and removing the liquid with a centrifuge (~ 4-3) or a clay filter

tColloids in Agriculture (London: Edward Arnold & Co., 1935).

2Kelley, Cation E.tchange in Soil.I' (New York: Reinhold Publishing Corporation,
1948).
51
58 CATION EXCHANGE DETERMINATIONS FOR SOILS
(Norton Refractories, Worcester, Mass.). Washing can also be done by
changing the solution about a dialysis sack.
4-3. Centrifuge Washing of a Sample. In the centrifuge washing pro-
cedure, each washing of a sample consists of the following steps:
1. The washing solution (acid, salt, water, or acetone solution) is added
to the tube containing the sample in a volume ratio of at least 6 of the
solution to I of the bulk volume of the sample cake. The procedure de-
pends upon a cumulative dilution, which at the rate of 1 to 5 for each
washing effects a dilution to less than 0.1 per cent in 5 washings.
2. The tube is stoppered and shaken by hand or mechanically until the
sample is well dispersed. Samples that are resistant to dispersion by this
method are triturated with a rubber ball supported on a thick glass rod,
with short strokes to prevent splashing.
3. The stopper (or ball) and sides of the tube are washed down with
water or the organic solvent, whichever is being used in the washing solu-
tion. In so doing, the stopper is lifted slightly to break its seal, but is left
in place. Then a very fine stream of the solvent is delivered from a pressure
wash bottle onto the juncture of the tube and stopper. The stopper is
raised and lowered slightly and then removed, as a film of solvent sweeps
down the sides of the tube. Too thick a layer of solvent over the washing
solution is avoided, as it interferes with decantation later.
4. If the solvent is water or alcohol, the tube is placed for I to 5 minutes
in a hot water bath to hasten flocculation. Caution: Any open flames are
removed from the vicinity if an organic solvent is being employed.
5. The tube is centrifuged at 2100 to 2600 rpm (1800 rpm for tubes
more than half full in the International No. 2 centrifuge) for 5 minutes to
clear the suspension. The centrifuge is allowed to coast to a halt without
braking, to avoid disturbing the cake.
6. The clear supernatant liquid is decanted with caution so that the tube
is held in a position that permits observation of the entire clear liquid.
When the meniscus strikes the cake, a little solid may be dislodged, and if
so, the decantation should be stopped before the solid leaves the tube. The
tube is righted abruptly, and a second pouring is never attempted without
recentrifugation.
4-4. The excess salt is washed out with solvents such as water, alcohol,
acetone. The washing solvent must be free of electrolyte or part of the ex-
changeable cation will be displaced. Alcohol, for example, should be free
of organic acids. A former practice of neutralizing ethanol with NH 40H to
pH 7 (as measured by indicators or pH meter) is considered3 undesirable
because the pH measurement is not accurate in an alcoholic system (brom
thymol blue turns yellow in alcohol in the absence of free acid), and es-
pecially because some exchangeable cation would be introduced in the
washing solution. Distillation of alcohol after the addition of Ca(OH) 2 is a
satisfactory procedure of freeing it from acids and other electrolytes.
4-S. Dispersion During Washing. As the excess of soluble electrolyte is

a Peech et al., U.S.D.A. Cir. 757 (1947); Swindale and Fieldes, Soil Sci., 74:287
(1952).
CATION EXCHANGE DETERMINA TIONS FOR SOILS 59
removed, the colloidal material tends to disperse, more so in water than in
nearly anhydrous organic solvents. With monovalent ion saturations (par-
ticularly with Na) the tendency for dispersion is great even in anhydrous
organic solvents. With divalent ion saturations, the tendency toward dis-
persion is less. Dispersion results in the colloid remaining suspended in the
washing solvent after centrifugation, or passing through the filter (but not
through dialysis sacks) . Dispersion during washing is best prevented by one
or more of the following: {a) selecting a less dispersive cation, (b) select-
ing the solvent, particularly using as low a concentration of water in the
organic solvent as possible, ( c) keeping the volume of solvent at a mini-
mum in the third and successive washings, ( d) rewashing the small fraction
dispersed in small pointed tubes after reflocculation, and ( e) as a last re-
sort, removing the salts by simple dialysis in a sack, recognizing that some
hydrolysis of cation occurs. Dialysis may be effected with a Cellophane or
Visking sack made of commercial tubing, or with a Collodian sack.
CATION EXCHANGE CAPACITY DETERMINA TION
4-6. Soil mineral and organic colloidal particles have negative valence
charges that hold dissociable cations, and thus are "colloidal electrolytes"
as defined in 1912. 4 The cation exchange capacity determination involves
measuring the total quantity of negative charges per unit weight of the
material. The amount of cation exchange capacity measured varies some-
what according to the nature of the cation employed, concentration of the
salt, and the equilibrium pH ( ~ 4-8). This measurement should not be
thought of as highly exact, but rather as an equilibrium measurement un-
der the conditions chosen. This fact led to the statement, "Next to lime re-
quirement no other soil constant, perhaps, is so widely used and yet so
little understood as 'base' exchange capacity."11
4-7. The cation exchange capacity is usually measured by leaching the
soil or colloid with a 1 N salt solution b"11ffered at a neutral or slightly alka-
line pH value, and then washing out the excess salts with an electrolyte-free
solvent. The buffer function, of course, may be as separate washings, fol-
lowed by washings with a salt of the metallic cation employed for satura-
tion. For example, the buffer washing may be with a NaOAc solution, and
the saturating metallic cation may then be furnished from a 1 N CaCl2
solution ( ~ 4-20). Salts of a strongly dissociated acid are not satisfactory
unless a buffer washing precedes. Thus NH4 Cl was shown 6 not to saturate
acid soils completely because the reaction:
(4-1)
4 McBain, Sci., 109:291 (1949); Marshall and K.rinbill, J. Phys. Colloid Chem.,
46: 1077 (1942).
11 Puri and Uppal, Soil Sci., 47:245 (1939).
6 Parker, J. Am. Soc. Agron., 21: 1030 ( 1929).
60 CATION EXCHANGE DETERMINATIONS FOR SOILS
in which X is the soil exchange complex, is kept from completion by
the activity of the HCl produced. Pretreatment was given the soil with
Ba(OH)2 , or, better, with Ba(0Ac) 2 to avoid the "artificial" increase 7 in ex-
change capacity caused by Ba(OH) 2 (~ 4-52). Use of NaOAc for pre-
treatment (~ 4-20) or for saturation (~ 4-23) provides the same advan-
tage and eliminates difficulties of precipitation of carbonate or sulfate of
barium. Normal acetate salts of monovalent ions usually dissolve only
traces of organic matter from soils, in solutions having pH values in the
range from 7 to 8.3.
4-8. Because of the wide recommendation of liming soils to the pH
range near 7, determination of cation exchange capacities at this reference
pH has long been popular. Recent trends toward extraction of exchangeable
hydrogen (~ 4-50) at pH values of 8.1 to 8.3 has renewed interest in the
determination of the cation exchange capacity at the same reference pH.
There is some evidence that the use of a I N salt solution of any metallic
cation that forms a strong base, strongly buffered at any pH from 4 to 9,
tends to measure a constant amount 8 of exchange capacity except for spe-
cific fixation reactions ( ~ 4-9). This fact is attributable to the effect of the
large mass of metallic cations compared to hydrogen ions in the salt solu-
tion. Cations that form weak bases tend to give lower or higher exchange
capacity. For example, a lower cation exchange capacity was determined
by 1 N NH+OAc of pH 7 than with BaC1 2 -triethanolamine of pH 8.1 for
kaolinitic soils11 and soils with 2 : I types of clay. 10 With cations such as
Cu, Zn, Al, and Fe, a higher cation exchange capacity was measured 11 than
with cations of strong bases.
4-9. An extremely large number of "methods" for determining the
cation exchange capacity of colloidal electrolytes is provided by different
combinations of pretreatment, cation, salt, solvent, washing procedure, and
determinative technique for the ions ( ~ 4-22). Calcium 12 may be consid-
ered the standard because it is the most abundant exchangeable metallic
cation in the majority of near neutral soils. Potassiumrn has the advantage

7 Magistad, J. Am. Soc. Agron., 21: 1045 ( 1929).

RDeMumbrum and Jackson, S.S.S.A. Proc., 20:334 (1956); also pointed out by
Piper, Soil and Plant Analysis (New York: Interscience Publishers, Inc., 1944), p.
165.
fl Mehlich, Soil Sci., 60:289 (1945).
io Pratt and Holowaychuk, S.S.S.A. Proc., 18:365 (1954).
11 Lutz, S.S.S.A. Proc., 3:7 (1939); Bower and Truog, S.S.S.A. Proc., 5:86 (1941);
Sieling, J. Am. Soc. Agron., 33:24 (1941); Menzel and Jackson, S.S.S.A. Proc.,
15: 122 (1951); DeMumbrum and Jackson, Soil Sci., 81 :353 (1956).
12Kappan, Trans. 2nd Comm. Intern. Soc. Soil Sci. (Gronigen) B:179 (1927);
Truog and Drosdoff, Trans. 3rd Intern. Congr. Soil Sci., 1: 106 ( 1935); Bradfield
and Allison, Trans. Intern. Soc. Soil Sci., A:63 (1933); Jackson and Truog, S.S.S.A.
Proc., 4: 137 (1940).
13 Rendig, S.S.S.A. Proc., 12:449 (1948); Swindale and Fieldes, Soil Sci., 74:287
( 1952).
 CATION EXCHANGE DETERMINATIONS FOR SOILS 61
over Ca of greater sensitivity with the flame emission spectrophotometer;
also K has less tendency to cause soil dispersion than Na (~ 4-23). Fixa-
tion by vermiculite of potassium (~ 6-79) and ammonium (~ 4-25) ions
makes the exchange capacity lower with these ions than with nonfixing ions.

APPARATUS

4-10. Needed apparatus for rapid semimicro determinations of cation

exchange capacity includes 100-ml centrifuge tubes (or smaller tubes for
small samples) and International No. 2 or other suitable centrifuge (in
lieu of centrifuge washing, the sample may be washed in the special small
carbon funnels, Fig. 4-1, or in fritted glass or asbestos crucibles, or in an

15 mm
15 cm

~------------------
:11 r r r r -r
I1 t'2 rows of 8 holes 26 mm dia
I1 '\ on 37 mm centers
~--~--~--c--~--~--
Filter
pad

Fig. 4-1. Special small carbon funnel and leaching rack for the semimicro filter
tube method of washing sample for exchange capacity. (After Swindale and Fieldes,
Soil Sci., 74:287, 1952; drawing courtesy of Dr. L. D. Swindale.)

ordinary funnel with filter paper), a 250-ml beaker, a 500-ml conical flask,
a 25-ml pipette, and a buret for the standard Versene for Ca determination
(or flame emission spectrophotometer suitable for Ca determination,
~ 18-21 ).
62 CATION EXCHAN GE DETERM INATION S FOR SOILS

REAGENTS

4-11. Needed reagents are 1 N NaOAc adjusted to pH 5, 82 gm of salt

and about 28 ml of glacial HOAc per liter; 30 per cent H 2 0 2 (if the or-
ganic matter is to be removed); neutral 1 N NaOAc, 82 gm per liter, with
pH adjusted to 7.0 with HOAc to neutralize the NaOH normally formed
by hydrolysis of this salt (neutral I N NH 4 0Ac is used in addition if the
flame emission spectrometer is to be employed for Ca); neutral 1 N CaC1 2 ,
approximately 109 gm of CaCl 2 • 6 H:!O or 73 gm of CaCl 2 • 2 H 20 per
liter of C0 2 -free distilled water, with pH adjusted to 7.0 with Ca(OH) 2 ; 80
per cent acetone (20 per cent by volume of water); and the following spe-
cial solutions for the Ca-Versene procedure.
4-12. Standard Ca Solution. A 0.5005-gm portion of pure dried CaCO~
is dissolved in a minimum of 0.2 N HCI. The solution is boiled to expel
C0 2 and is then diluted to 1 liter. The solution is 0.0100 N with respect
to Ca.
4-13. NH 4 Cl-NH 4 0H Buffer of pH 10. 14 This buffer is made up of
100 ml of 1 N NH 4 Cl and 500 ml of 1 N NH 4 0H.
4-14. Eriochrome Black T lndicator. 15 A solution of Eriochrome black
T indicator (Bersworth Chemical Co., Framingham, Mass., or Eastman,
Rochester, N.Y.) is prepared by dissolution of 0.5 gm of the indicator with
4.5 gm of hydroxyl amine hydrochloride in 100 ml of methanol ( ~ 11-5 8),
A 50-ml portion of 2 per cent NaCN solution is prepared.
4-15. Standard Versene Solution. A 2-gm portion of disodium Versenate
(disodium dihydrogen ethylenediamine tetra-acetic acid, from Bersworth or
Eastman) is dissolved in 900 ml of water. Then approximately 50 mgm of
MgC1 2 • 6 H 2 0 is added to the solution. The normality of Versene is then
determined by titration of a 25-ml portion of the standard Ca solution ac-
cording to the procedure ( ~ 4-21 ) .

PROCEDURE

4-16. The Sample. Enough of a sample is taken to give 0.04 to 0.2 meq
of cation exchange capacity, when the determination is with Ca. This re-
quires about 0.3 gm or less of mineral colloid and 0.5 to 5 gm of soil.
(Enough of a sample for 0.002 to 0.02 meq of K in a 40-ml final volume is

14 Schwarzenb ach, T. chimia, 2:59 {1948); Tech. Bui. 2, Sec. Ill (Framingha
m,
Mass: Bersworth Chemical Co., 1951 ), p. 5. Cheng and Bray, Soil Sci., 72:449 (1951)
used 5 ml of pH I 0 buffer consisting of 67 gm of NH 4Cl, 570 ml of concentrated
NH40H, and water to make 1 liter. Bowe.-, S.S.S.A. Proc., 19:40 (1955), to minimize
the complexing of Na of the NaOAc, used pH 9.5 (10 ml of NH4Cl-NH 40H
buffer, made up of 134 gm of NH 4 Cl in 1 liter of 2.5 NH40H), but the end point
is much less sharp.
15Tucker, J. Australian Inst. Agr. Sci., 21:100 (1955), proposed phthaleinco m-
plexone as a substitute.
CATION EXCHANGE DETERMINATION S FOR SOILS 63
satisfactory, 11 4-22. Also 0.004 to 0.04 meq of Mn may be employed,
11 4-29.)
4-17. Acidification of the Sample. The soil or colloid sample is placed
in a 100-ml centrifuge tube and stirred in 50 ml of 1 N NaOAc of pH 5.0
with a policeman-tipped rod. The soil suspension is digested 16 in a near-
boiling water bath for 30 minutes with intermittent stirring. The salts are
removed by centrifugation of the suspension and decantation of the clear
supernatant liquid (11 4-3). Two additional washings are given with
1 N NaOAc of pH 5, the 30-minute boiling water bath treatment being re-
peated if the sample is known to be calcareous.
4-18. Removal of Organic Matter (Optional). If the exchange capacity
of only the mineral portion of the soil is wanted, the organic matter is re-
moved by H 2 0 2 treatment. The acidified sample is transferred to a 250-ml
beaker with 10 ml of water (or as little as possible more). Then 5 ml of
30 per cent H 2 0 2 is added, the beaker is covered with a watch glass, and
the sample is heated cautiously to start the reaction. With beaker tongs, the
beaker is removed from the heat and the reaction is quieted if necessary by
a stream of water from a wash bottle. When the reaction is quiet at steam
plate temperature, additional 5-ml increments of 30 per cent H 2 0 2 are
added, one at a time, and the digestion on the steam plate is continued as
long as needed to bring about destruction of the organic matter. The evapo-
ration is not carried below about 5 to 10 ml. As many as 5 to 10 increment&
of H 2 0 2 may be needed for soils high in organic matter. The H 20 2 treat-
ment also removes any Mn0 2 present in this acidified soil system through
a rapid and specific reduction reaction bringing Mn to manganous.
4-19. Exchange Capacity of Soil Due to Organic Matter. The portion
of the cation exchange capacity arising from soil organic matter is meas-
ured17 by the determination of the exchange capacity before and after the
above treatment with H 2 0 2 • Each 1 per cent of Walkley-Black soil carbon
(11 9-57) was equivalent 18 to 4.9 meq of exchange capacity per 100 gm of
soil ( 490 meq of exchange per 100 gm of organic matter). No "uncover-
ing" by H 2 0 2 of blocked 19 exchange sites was noted. Selective destruction
of soil organic matter by heating at 350 to 400°C for 7 or 8 hours has also
been attempted, 20 but the exchange capacity of some mineral colloids is
decreased by such heating.
16 50 ml of this buffer per 5 gm of soil dissolved up to 50 per cent of the soil
carbonates including dolomite at steam plate temperature, in tests in this laboratory
by L. Gallardo; using the buffer is more convenient than acidifying to pH 3.5 with
HCI.
17Robinson, J. Agr. Res. 34:339 (1927); Alexander and Byers, U.S.D.A. Tech.
Bui. 317 (1932); McGeorge, Ariz. Agr. Exp. Sta. Tech. Bui. 30 (1930); Olson and
Bray, Soil Sci., 45 :483 (1938); and Mehlich, Soil Sci., 66:429 (1948).
18 Pratt, Soil Sci., 83: 85 (1957).
19Broadbent,Adv. Agron., 5:153 (1953).
20 Mitchell, J. Am. Soc. Agron., 24:256 ( 1932).
64 CATION EXCHANGE DETERMINATION S FOR SOILS
4-20. Exchange Saturation. The acidified soil sample (after transfer
from the beaker to centrifuge or leaching tube if the H 20 2 treatment has
been employed) is washed twice (~ 4-3) with neutral 1 N NaOAc to re-
move the remaining salts and to aid in making active some slowly active
vermiculite-like exchange charges (~ 4-25). Next, the sample is given 5
washings with I N CaC12 solution. (Four washings with N Ca ( OAc) 2 and 1
with CaCl 2 is also satisfactory.) The excess salt is removed by washings
(usually 5) with 80 per cent acetone (~ 4-4) until the excess CaCl 2 is re-
moved, as indicated by a negative AgN0 8 test for Cl in the last of the wash-
ings. Finally the Ca is replaced by means of 5 washings with a neutral
1 N NaOAc solution ( l N NH.,OAc is employed instead of NaOAc if the
flame emission spectrophotometer is to be employed for the Ca determina-
tion). The Ca is then determined directly in the solution employed for dis-
placement, by Versene titration, now to be described (or the Ca may be
determined by flame emission spectrophotometry, ~ 18-7).
4-21. Versene Titration of Ca for Exchange Capacity. The Ca solution
(about 250 ml) resulting from displacement in the determination of cation
exchange capacity is placed in a 500-ml conical flask. Next, 10 ml of the
NH4 Cl-NH 4 0H buffer solution (~I 4-13) is added to bring the solution to
pH 10 and then 10 drops of Eriochrome black T indicator solution and
lml of 2 per cent NaCN solution are added. A blank of NaOAc is similarly
prepared and is then titrated to a bright blue endpoint with standardized
(about 0.01 N) Versene solution containing Mg. The test sample is titrated
to the same color. The Mg present in the Versene forms a wine-red color
with the indicator after the titration is begun. The color shifts to blue as the
Mg is again complexed with Versene at the endpoint. The Ca complex with
Versene dissociates less than the Mg complex, and therefore all of the Ca
is titrated before the Mg is recomplexed by the Versene. The presence of
Mg in the Versene and the use of Eriochrome black T indicator make the
determination like that of Mg plus Ca (~ 11-59), but the titer does not
include the Mg because the Mg is the same form at the endpoint of the
titration as in Versene reagent. The cation exchange capacity for Ca is
calculated:

meq exchange capacity per 100 gm = ml Versene x N

100
x (4-2)
sample weight in gm

ALTERNATIVE PROCEDURES

4-22. Cation exchange capacity may also be determined by other ions

and with other determinative procedures. Determination with Na (~ 4-23)
or Ba (~ 4-32) can be effected without removal of CaC03 . In the deter-
CATION EXCHANGE DETERMINATIONS FOR SOILS 65
mination 21 with K ( ~ 4-16) in leaching tubes (Fig. 4-1 ) , 0.1 gm of soil is
leached successively with 40 ml of N KOAc, 40 ml of 95 per cent ethanol,
and 40 ml of NH 4 0Ac in which the exchanged K is determined by flame
emission spectrophotometry ( ~ 18-7). Also, the soil, after Ca or Mg satu-
ration and washing with .80 per cent ethanol, may be suspended in water
and the exchangeable cation titrated with Yersene. 22 The soil may be
leached with K 2 C08 and the filtrate titrated to determine 23 the alkali taken
up, by comparison to a blank. Also the soil can be leached with 0.2 N KC1
then with N (NH 4 )~C0:1 , with subsequent evaporation of the second filtrate
and titration of the residual K 2 C03 , thus avoiding the need to wash out the
excess electrolyte. 24 The cation exchange capacity can be approximated as
the sum of total exchangeable metallic cations (~ 4-36) and exchangeable
hydrogen(~ 4-50).
4-23. Determination with Sodium.:m Determination of cation exchange
capacity by means of NaOAc solution of pH 8.2 can be effected without
interference of soil CaC0:1 and is relatively constant from a pH of just
above 7 to over 9 in 1 N NaOAc solution. The Na ion also has the advan-
tages of being a prominent cation in some saline and alkali soils, and of not
being fixed as are NH 4 and K.
4-24. A 4-gm sample of medium and fine textured soil or a 6-gm sample
of coarse textured soil, of known moisture content, is transferred to a
50-ml round bottom, narrow neck centrifuge tube and 33 ml of N NaOAc
solution of pH 8.2 is added. The stoppered tube is shaken for 5 minutes in a
reciprocating shaker. After centrifugation, the clear supernatant liquid is
decanted as completely as possible and discarded. The sample is treated
with 3 additional 33-ml portions of N NaOAc solution for a total of 4
treatments. Then the sample is suspended in 33 ml of 95 per cent ethanol
and shaken for 5 minutes, centrifuged, and the clear supernatant liquid
discarded. The sample is washed with 2 additional 33-ml portions of eth-
anol. The ele~trical conductance of the supernatant liquid from the third
washing should be Jess than 40 micromhos per cm. (Bower, et al., show
that the quantity of Na released by hydrolysis just balances the quantity of
excess Na salt left in the solutions at about the third washing, and that
further washing results in an erroneous lowering of the measured exchange
capacity.) The adsorbed Na is replaced from the sample with three 33-ml
portions of N NH 4 0Ac, the decantate being collected in a 100-ml volu-
metric flask. This solution is made to volume and mixed, and the Na is

21 Swindale and Fieldes, Soil Sci., 74:287 (1952).

22 Perkins, Soil Sci., 74:443 (1952); also suggested removal of CaC03 from soil
wlth Versene.
23 Schofield. J. Agr. Sci., 23 :255 (1933 ).
24 Puri, Soil Sci., 37: 105 (1934).
2r. Bower et al., Soil Sci., 73: 251 (19 52).
66 CATION EXCHANGE DETERMINA TIONS FOR SOILS
determined by the usual flame emission ( 11 18-21 ) or other procedure
(115-88).
4-25. Determination with Ammonium. 26 The cation exchange capacity
can be determined with ammonium following extraction of exchangeable
cations ( 11 5-11 ) . The determination with ammonium ion may not displace
the last few per cent of the strongly held H + ions (weakly acidic) from
kaolinite or halloysite, and tends to become fixed to some extent by vermic-
ulite. 27 Neither is the determination of ammonium ion as rapid as that of
metallic cations by flame emission or Versene titration. Use of 80 per cent
ethanol solution of NH 4 0Ac for saturation and Ca(0Ac) 2 as the replacing
agent have been employed with calcareous soils. The ammonium exchange
capacity was employed by Peech, et al. for noncalcareous soils.
4-26. The reagents required for the determination with noncalcareous
soils include 95 per cent ethanol or methanol, acid-free U.S.P. grade, and
acidified to per cent KCl (too gm of KCl in 900 ml of water, the pH of
the solution being adjusted at 2.5 with HCJ, thymol blue indicator being
employed with a spot plate; the solution is diluted to 1 liter).
4-27. After the ammonium acetate leaching for the removal of ex-
changeable cations ( 11 5-11 ) , the upper parts of the apparatus are washed
with a stream of alcohol. Then the soil is leached with 250 ml of alcohol
added to the Buchner funnel in increments (500 ml through the delivery
tube for the carbon funnel apparatus). 28 The last of the alcohol is drawn
through by suction, but the soil is not dried lest ammonia be lost.
4-28. The soil is then leached with 450 ml of l 0 per cent KCI solution
of pH 2.5, and the leachate is collected in a suction flask. The leachate is
washed into a 500-ml volumetric flask, made to volume, and mixed. An
aliquot representing l to 2 meq of NH4 + is placed in a 800-ml Kjeldahl
flask and the ammonium is distilled ( 11 8-22) and back titrated with
N / 14 H 2 SO4 • As a blank, an equal portion of the to per cent KCl solution
is distilled.
4-29. Determination with Mn. The cation exchange capacity of small
samples (10 to 80 mgm) is advantageously measured by Mn++. Fortu-
nately, as shown by Bower and Truog, 29 the cation exchange capacity
measured by means of Mn++ is equivalent to that with Ca, which is the
standard cation because it constitutes the major percentage saturation of
agricultural soils. Manganese chloride or acetate solutions are invariably

26Kelley, J. Am. Soc. Agron., 21:1021 (1929); Schollenberger and Dreibelbis,

Soil Sci., 30: 161 (1930); Chapman and Kelley, Soil Sci., 30: 391 (1930); Peech et al.,
U.S.D.A. Cir. 757 (1947).
21 Bower, Soil Sci., 70:375 (1950).
28 Schollenberger and Simon, Soil Sci., 59: 13 (1945).
29 /nd. Eng. Chem., A.E., 12:411 (1940).
CATION EXCHANGE DETERMINATIONS FOR SOILS 67
acid, as a result of hydrolysis, in contrast to neutral solutions obtainable
with Ca. The Mn is readily determined in microchemical amounts by the
classical periodate method of Willard and Greathouse 30 as adapted to spec-
trophotometric measurements by Mehlig. 81
4-30. The quantity of sample (or aliquot) is taken to be equivalent to
0.004 to 0.04 meq of exchange charge (0.1 to 0.8 mgm of Mn) in 100 ml
of final HMn0 4 solution (with smaller samples, a smaller final volume is
employed). The soluble salts and CaC03 are removed ( ~ 4 -1 7) ; the or-
ganic matter is removed ( ~ 4-18) if the exchange capacity of only the min-
eral portion is to be determined; and the exchangeable hydrogen is
removed by a buffer washing (~ 4-20). The exchange charge is then satu-
rated with Mn+ + by the usual washing procedures ( ~ 4-2) with N MnC1 2
or Mn(0Ac) 2 solution. The excess salt is removed by the usual washings
with methanol or 80 per cent acetone solution in water, and the Mn++ is
displaced by means of washings with 1 N NH4 0Ac, and determined by the
usual procedure ( ~ 5-74). The cation exchange capacity is calculated:
final HMn04
meq of exchange _ ppm of Mn solution volume, ml
(4-3)
capacity per - 274.7 x sample represented in
l 00 gm of soil final solution, gm

4-31. Holtzinger et al., 32 replaced the Mn++ with N KCl, 4 33-ml wash-
ings being employed for a 4-gm soil sample. The KCl was made to 250 ml
with N KCI solution. To 20 ml of this 5 drops of 2 per cent gelatin were
added and the Mn was determined directly in the replacing solution with
a polarograph ( ~ 16-28) .
4-32. Determination with Ba.as Barium saturation is effected during ex-
traction of exchangeable cations from calcareous soils (~ 5-21) with
BaC12 buffered with triethanolamine. A final washing with BaCl 2 alone was
given by Mehlich to complete the saturation for cation exchange capacity
determination. Washing with Ba(0Ac) 2 has also been employed by
Mehlich, and has been combined with Ba determination in the NH 4 0Ac
extract by flame emission 34 (~ 18-21). Exchange capacity has been de-
termined by conductometric titration 31> of Ba-soil with MgS04 solution.

30J. Am. Chem. Soc., 39:2366 (1917).

31 Jnd. Eng. Chem., A.E .. 11 :274 (1939).
32 Soil Sci., 77: 137 (1954).
33 Bobko and Arkinasi, Z. Pflanz. Dung.. 6A:99 (1925); Burgess and Breazeale,
Ariz. Agr. Exp. Sta. Tech. Bul. 9 (1926); Mehlich, Soil Sci., 60:289 (1945); Hanna
and Reed, Soil Sci., 66:447 (1948); Pratt and Holowaychuk, S.S.S.A. Proc., 18:365
(1954).
:H Toth and Prince, Soil Sci., 67:439 (1949).
Sli Mortland and Mellor, S.S.S.A. Proc., 18:363 (1954).
68 CATION EXCHANGE DETERMINATIONS FOR SOILS

TOTAL EXCHANGEABLE METALLIC CATIONS

4-33. In the equilibrium pH method, 36 total exchangeable metallic ca-
tions of soil are estimated by replacement with HOAc, the resulting pH
change being carefully measured to determine the amount of H removed
from the solution by reference to the standard pH titration curve of acetic
acid. The reaction is made virtually complete by the great excess of HOAc,
even though the metallic cations in solution are in equilibrium with the ex-
change charges. The total of exchangeable metallic cations has been esti-
mated also by extraction with NH4 0Ac (~ 4-43) and with HCI (~ 4-41).

APPARATUS

4-34. Needed are 50-ml conical extraction flasks and stoppers, a torsion
balance, a 25-ml volumetric pipette, and a glass electrode pH meter.

REAGENTS

4-35. One N HOAc is made up by dilution of 57.2 ml of glacial acetic

acid to 1 liter; a pH value of 2.3 t should be obtained by slight adjustments
of the concentration if necessary. One N NH 4 0Ac is made up by diluting
57 .2 ml of glacial acetic acid to 800 ml and neutralizing to pH 7 .00 ± 0.01
with concentrated NH 4 0H and then diluting to t liter. Other reagents
needed are standard buffers for calibrating the pH meter, with pH values
known to ± 0.01 pH unit, preferably 1 in the range of pH 3 to 5 and t in
the range of pH 6 to 7.

PROCEDURE

4-36. Total Exchangeable Metallic Cations. To 2.50 gm of soil, placed

in a 50-ml conical flask, is added a 25-ml volume of 1 N HOAc. The sus-
pension is shaken intermittently for t hour, and then the pH of the suspen-
sion is determined accurately to 0.01 or 0.02 unit (~ 3-19). The pH of
the clear supernatant liquid is almost exactly the same as that of the sus-
pension in this system. The pH value of the original HOAc reagent is de-
termined at the same time. The meq of exchangeable metallic cations is
calculated from the curve (Fig. 4-2) of Brown or the following equation
derived therefrom:
meq exchangeable
metallic cations = (pH observed - 2.31) x 22 (4-4)
per 100 gm of soil
which is applicable from pH 2.31 to 2.8, that is, up to 10 meq of exchange-
able metallic cations per 100 gm of soil. The curve and equation are based
on meq per liter, but one liter of solution corresponds to 100 gm of soil.

se Brown, Soil Sci., 56:353 (1943).

CATION EXCHA NGE DETERM INATIO NS FOR SOILS 69

For amounts of exchangeable metallic cations from 10 to 20 meq, one-half

the amount of soil is taken ( 1.25 gm) and the meq obtained is multiplied
by 2. The total exchangeable metallic cations may be employed to derive
the cation exchange status ( ~ 4-42).

3.00 r--""T""" --.----r-. .-""T"""- .---.-..--. --.---.-..- -.---

2.90

"2.80
<
02.70
:x:
0 2.60
:x:
Q. 2.50
2.40
2.30 ~........~__.,-'---':---'-__.,-'--~-'-- -'--'---'--'---'
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
meq of total exchangeable metallic cations
per 100 gm of soil

Fig. 4-2. Relation of pH of HOAc solution to total exchange-

able metallic cations extracted from soil. (After Brown, Soil Sci.,
56:383, 1943.) The curve is for pure HOAc, while the points are
soil analysis values.

ALTERNATIV E PROCEDURES

4-37. Titration of Total Exchangeable Metallic Cations.37 The titration

of metallic oxides in the residue of the NH 4 0Ac extract of soil gives a
measure of the total exchangeable metallic cations. From the screened and
well-mixed soil sample, a 10.0-gm sample is weighed out on a torsion bal-
ance and transferred to a 250-ml conical flask. Then approximately 100 ml
of N NH 4 0Ac solution is poured into the flask, the flask is stoppered, and
the suspension is shaken intermittently for 10 minutes. The suspension is
then poured into the Gooch filter crucible in which the filter paper or as-
bestos has previously been wetted and seated firmly by suction. When the
suspension has been filtered, the flask is rinsed with more of the NH 4 0Ac
solution and the rinsings poured onto the filter. Additional increments of
NH4 0Ac solution are poured onto the soil until a total volume of approxi-
mately 250 ml of the extraction solution has been collected.
4-38. The NH 4 0Ac extract is transferred to a 400-ml beaker, and
evaporated to a small volume on a steam plate in the absence of chloride
fumes. Before the solution becomes gummy, it is transferred to a platinum
or silica dish, and the evaporation is continued to dryness over a low flame
or slowly in a muffle furnace to convert the acetates to oxides and carbon-
ates. The temperature is raised gradually to a full red heat (about 700 to
800°C) for 15 to 20 minutes, and then the dish and contents are allowed

87 Bray and Willhite, Ind. Eng. Chem., A.E., 1: 144 (1929).

70 CATION EXCHANGE DETERMINA TIONS FOR SOlLS
to cool slowly to room temperature, in a desiccator if necessary for protec-
tion from chloride fumes in the laboratory.
4-39. The residue is taken up in 50 ml (or other known quantity, suf-
ficient to be an excess) of standard 0.1 N HCI delivered from a pipette. The
acid is warmed gently to cause complete reaction with the metallic oxides
and carbonates and then is brought just to boiling to drive off C0 2 • Finally
the excess acid is back titrated with standard 0.1 N NaOH, brom cresol
purple being employed as the indicator. Heating to boiling sharpens the
end point.
4-40. The meq of exchangeable metallic cations per 100 gm of soil
equals the net ml of 0.1 N HCl consumed by the oxides and carbonates,
meq exchangeable
metallic cations =V - T (4-5)
per 100 gm of soil
in which V is the volume of 0.1 N HCl added and T is the back titer of
0.1 N NaOH. The metallic cations usually consist dominantly of Ca and Mg
with lesser amounts of K and Na. The presence of soluble chloride and
sulfate salts in the soil does not interfere, although the presence of car-
bQnates in the soil presents the usual difficulty ( ~ 5-18) of increasing the
quantities of metallic cations over those in truly exchangeable form.
4-41. Total Exchangeable Metallic. Cations by Dilute HCl Extraction.38
To 10 gm of acid or neutral soil (method not applicable to calcareous
soils) in a 250-ml conical flask is added a 100-ml volume of standard
0.1 N HCJ. The flask is stoppered and the suspension is shaken intermit-
tently for 1 or more hours. The suspension is allowed to stand until clear.
Then 25 ml is pipetted from the clear supernatant liquid and titrated with
standard 0.1 N NaOH with brom cresol purple as the indicator. Warming
the solution to drive off the C02 sharpens the end point. The calculation of
total exchangeable metallic cations is by equation 4-5. Iron and aluminum
extracted by the acid are reprecipitated and thus excluded from the ex-
changeable metallic cations, but some error occurs due to side reactions
with the soil. Schofield39 proposed the use of 0.05 N HCl in volume suffici-
ent that only 20 per cent of it is used up.

EXCHANGEABLE METALLIC CATION STATUS

4-42. The exchangeable metallic cation status 40 of a soil refers to the

Kappen, in Bodenaziditat (Berlin: J. Springer, 1929), p. 169.

88
According to Piper, op. cit., p. 189.
89
Metallic cations with hydroxyl ions form hydroxides or bases;
40 in older soil
literature the terms "base" hence "base status" was sometimes applied to the cation
instead of to the hydroxide. The hydroxide as a whole is the proton acceptor and,
therefore, the base. Compare base, the proton acceptor, to acid, the proton donor
<• 4-50).
CATION EXCHA NGE DETERM INATIO NS FOR SOILS 71

percentage exchange saturation with metallic cations:

Exchange status (percentage saturation) = : x 100 (4-6)

in which S (after Hissink) is the total exchangeable metallic cations

(~ 4-33) as meq per 100 gm of soil and T (after Hissink) is
the cation ex-
change capacity ( ~ 4-6) as meq per 100 gm of soil. The value T is also the
sum of S plus exchangeable hydrogen (~ 4-50). A more detailed evalua-
tion of the metallic cation status of soils includes the separate determination
of the percentage saturation with each metallic cation species (~ 5-5). The
percentage exchangeable hydrogen saturation, the complement of the per-
centage of metallic cation saturation, once was termed percentage "un-
saturation," but such usage is disfavored by the Soil Science Society of
America.
4-43. The percentage metallic cation saturation of acid soils is important
in soil-plant relationships, and can be raised by liming the soil (~ 4-64).
In alkaline soils, the determined exchange saturation with metallic cations
(~ 5-5) often exceeds I 00 per cent of the exchange capacity
as usually
determined. This situation is particularly common in soils
41 high in organic
matter. Calcium is released from positions not available for exchange ca-
pacity measurement by most methods. In calcareous soils, appreciable
amounts of Ca and Mg tend to dissolve from the carbonate form in the
extraction solution (~ 5-18) and thus are not distinguishable from ex-
changeable metallic cations.

GYPSUM OR SULFUR REQUIR EMENT OF SOILS

4-44. The development of physical impermeability in "alkali" soils
(~ 10-5) is correlated with the percentage Na saturation or
"degree of
alkalization." 4 ~ When the percentage Na saturation exceeds
15 per cent,
deterioration of the physical properties of the soil generally sets in. When
a soil containing an excessively high percentage saturation with Na is
shaken with a nearly saturated gypsum solution, Ca exchanges for Na and
the consequent loss of Ca in the solution is an approximate measure of the
gypsum requirement. The gypsum requirement is the equivalence of
CaS0 4 • 2 H 2 0 or of S that should be added to "reclaim" the soil by dis-
placing the net exchangeable Na (~ 4-48). Use of gypsum or sulfur, or
43

acidulation 44 is the counterpart of liming soils ( ~ 4-64).

41 Puri and Uppal, Soil Sci., 47:245 (1939).

42 Puri, Soils (New York: Reinhold Publishing Corporatio n, 1949), p. 83.
43 Harmer, S.S.S.A. Proc., 7:378 (1943).
44Sherma n and Harmer, J. Am. Soc. Agron., 33:1081 (1941); Aldrich, S.S.S.A.
Proc., 13: 191 (1949); Aldrich and Turrell, Soil Sci., 70: 83 ( 1950).
72 CATION EXCHANGE DETERMINATIONS FOR SOILS

APPARATUS

4-45. Needed apparatus is a 200-ml conical fl.ask and stopper, a 100-ml

pipet, and a buret and fl.ask for Versene titration.

REAGENTS

4-46. Needed reagents are an approximately saturated gypsum solution

of determined Ca concentration ( 5 gm of CaSO 4 • 2 H 2 0 shaken for an
hour in 1 liter of water) and Versene reagents ( ~ 4-1 1 ) . The gypsum
solution is filtered and titrated ( ~ 4-21 ) . It should have a Versene titration
of at least 28 meq per liter.

PROCEDURE45

4-47. A 5-gm sample of air-dried soil is added to a 200-ml conical fl.ask.·

Then 100 ml of the standardized gypsum solution (~ 4-46) is added by
means of a pipet. The suspension is shaken for 30 minutes and then a por-
tion is filtered and the Ca + Mg concentration is determined in an aliquot
by Versene titration (~ 4-21). Then:
Gypsum requirement, meq per 100 gm of soil = 2(S - T) ( 4-7)

in which Sis the Versene standardization blank of the gypsum solution as

meq per liter and Tis the Versene titration of the filtrate as meq per liter.
4-48. Each meq of exchangeable Na or "gypsum requirement"(~ 4-44)
is approximately equivalent to 1.7 tons of gypsum or 0.3 tons of sulfur per
acre-foot of soil to be reclaimed. Any gypsum already present in the soil is
taken into account in the extraction. In deciding the actual size of applica-
tion, the depth of soil to be reclaimed (more or less than 1 acre-foot) and
the fraction of exchangeable Na to be replaced (for example, 75 per cent
of it) must be considered. Also, exchangeable K does not contribute mate-
rially to the adverse effects noted for excessive Na saturation, yet this and
some other methods (~ 4-49) also include the degree of K saturation. For
this reason, the determined gypsum requirement may be somewhat exces-
sive for soils in which the saturation with K ( ~ 5-5) is over a few per cent.

ALTERNATIVE PROCEDURE

4-49. Puri46 gives a simple method of determining the total of exchange-

able Na and K by extracting the soil with (NH4) 2C03 solution, evaporating
the filtrate, and titrating the two alkalies (compare ~ 4-22).

4a U.S.D.A. Handb. 60 ( 1954) p. 49; Schoonover, Examination of soils for alkali

(Berkeley: U. Calif. Ext. Serv., mimeo., 1952); McGeorge and Breazeale, Ariz. Agr.
Exp. Sta. Tech. Bul. 122 (1951).
46 Op. cit., p. 83.
CATION EXCHANGE DETERMINATIONS FOR SOILS 73

EXCHANGEABLE HYDROGEN DETERMINATION

4-SO. The determination of exchangeable hydrogen in soils is in fact
an equilibrium determination of the proton (H + ) supplying power of the
colloidal material. The stronger the proton acceptor (the weaker the acid)
that is placed in equilibrium with the soil, the greater the quantity of pro-
tons removed from it. Acetate is generally used as a proton acceptor but
many others are available. The quantity of protons removed in the extrac-
tion solution is reported as the quantity of "exchangeable hydrogen." This
quantity in turn can be calculated into terms of "lime requirement"
(1! 4-64) of the soil.
4-51. Acid soils contain 47 a considerable quantity of active or exchange-
able aluminum, in the form of Al+++, Al(OH) ++and Al(OH) 2 +.When a
proton acceptor is equilibrated in a system containing active aluminum,
hydrolysis occurs with precipitation of aluminum hydroxide and simultane-
ous release of a proton from water:
(4-8)

The net effect is the release of a proton from the soil, which is equivalent
to the release of an "exchangeable" hydrogen ion. Salts of a strong acid
cannot be used to displace exchangeable hydrogen because they do not re-
move ("accept") the H+ from the solution.
4-52. Soils also contain tightly bound hydrogen associated to varying
degrees with oxygen. This hydrogen is of the nature of hydroxyls. As the
strength of the proton acceptor is increased in the equilibrium system,
greater and greater quantities of the hydrogen bound with varying strength
is released:
(4-9)

when X represents an organic or mineral colloidal electrolyte of soil. This

reaction is part of the so-called48 "build-up" of cation exchange capacity in
Ba(OH) 2 solution.
4-53. Therefore, the quantity of "exchangeable hydrogen" released is
a function of the equilibrium system in which it is measured. In general, the
practice is to determine exchangeable hydrogen at an equilibrium pH of 7
to 8.3 (pH 8.3 is 1/500,000 Nin hydroxyl), since this range of pH values
is a convenient reference point in the neutralization of acid soils in liming
practice ( 1l 4-8 ) .
47 Daikuhara, Bui. Imp. Centr. Agr. Exp. Sta. Japan, 2: 18 (1914); Kappan, op. cit.,
Paver and Marshall, J. Soc. Chem. Ind., 53:750 (1934); Mukherjee, Ind. J. Agr. Sci.,
12:105 (1942); Chatterjee and Paul, Ind. J. Agr. Sci., 12:113 (1942); Mukherjee
et al., I. Ind. Chem. Soc., 19:405 (1942).
4R Magistad,J. Am. Soc. Agron., 21: 1045 (1929).
74 CATION EXCHANG E DETERMIN ATIONS.FO R SOILS

APPARATUS
4-54. Needed apparatus consists of a carbon funnel with fl.ow rate regu-
lated with a rubber tube and screw clamp, fine glass wool, acid-washed
silica sand (coarse), and a 500-ml conical flask with stopper and delivery
tube for the extraction solution, a 500-ml receiving flask vented to permit
escape of air, and a 600-ml beaker.
REAGENTS
4-55. Needed reagents are 1 N Ba(0Ac) 2 extraction solution of pH 8.1,
adjusted with Ba(OH) 2 ; 1 per cent phenolphthalein indicator; and 0.05 N
NaOH or Ba(OH) 2 solution.
PROCEDURE41l
4-56. A small sphere of fine glass wool is tamped into the bottom of the
carbon filter funnel. It is then wetted with a little of the Ba(0Ac) 2 solution
and further compacted. Then a small portion of pure silica sand is spread
over the glass wool. The soil sample representing 0.2 to 1 meq of exchange-
able hydrogen is added to the funnel with care to prevent soil from sticking
to the sides of the tube. The soil is leveled off, and additional silica sand is
placed on top to a depth of about 5 mm. The flask containing 350 ml of the
Ba(0Ac) 2 extraction solution is inverted to deliver onto the soil. The
leaching rate is adjusted with a screw clamp so as not to exceed 10 to 20
drops per minute.
4-57. When all the extraction solution is passed through, its volume is
measured in a graduated cylinder and then it is transferred to a 600-ml
beaker. Approximately 10 drops of 1 per cent phenolphthalein solution are
added, and the extraction solution is then back titrated with 0.05 N NaOH
or Ba(OH) 2 to a faint pink color. A blank titration is made on 100 ml of
the extraction solution and calculated for the volume employed for extrac-
tion. The results are expressed in terms of meq of H+ per 100 gm of soil:
meq exchangeable H• per 100 gm soil= (T - B) x N x _____l_O_O_ _
sample wt., gm
(4-10)

in which T is the ml titration, B is ml blank, and N is the normality of

standard NaOH or Ba(OH) 2 • The results are also expressed as tons of
CaC03 per acre (~ 4-64).
ALTERNATIVE PROCEDURES
4-58. Exchangeable hydrogen has been determined by the change in
the pH of a 1 N NH4 0Ac solution (~ 4-61) or titration~ 0 or pH measure-

41JAfter Parker,/. Am. Soc. Agron., 21:1030 (1929), modified for extraction at
pH 8.1 following Mehlich, Soil Sci., 60:289 (1952).
r.o Schollenberger and Simon, Soil Sci., 59: 13 ( 1945).
CATION EXCHANGE DETERMINATION S FOR SOILS 75

ment 51 of the NH 40Ac extract (~ 5-11) with a standard hydroxide solu-

tion, although the end point is indistinct. Exchangeable hydrogen may be
approximated by subtracting the sum of exchangeable metallic cations
(~ 4-33) from the cation exchange capacity(~ 4-16).
4-59. Mehlich 52 modified the Ba(0Ac) 2 procedure slightly. A soil
sample having 0.5 to 1.0 meq of exchange capacity is placed in a crucible,
washed with 50 ml of 0.2 N Ba(OAc) 2 , then with 50 ml of water. The fil-
trate is made to 100-ml volume and an aliquot is back titrated with
0.05 N Ba(OH) 2 •
4-60. Instead of an acetate buffer, triethanolamine has been employed
by Mehlich53 for buffering BaCl2 • In a modification, Peech et al., 64 em-
ployed a solution like that described for extraction of exchangeable metallic
cations from calcareous soils ( ~ 5-20), except that the BaCl2 is made
0.5 N. A 10-gm soil sample is shaken with 25 ml of the buffer for 0.5 hour
and then filtered on a Gooch filter fitted with paper. The soil is completely
transferred with 25 ml more of the buffer, and then is further leached with
100 ml of 0.5 N BaC12 containing 2.5 ml of the original buffer solution per
liter. The extraction is made in not less than 1 hour. The combined filtrates
are titrated with standard 0.1 N HCl to an end point with brom cresol
green-methyl red indicator, the color being standardized against a blank
titration of similar volumes of the 2 extraction solutions.
4-61. Brown"r; determined exchangeable hydrogen of soils through
equilibration of 2.50 gm of soil in 25 ml of 1 N NH 4 0Ac (compare
~ 4-33). The suspension was shaken intermittently for 1 hour, and then the
pH values of the suspension and original solution were carefully determined
to 0.01 or 0.02 pH units (~ 3-19). The meq of exchangeable hydrogen
can be obtained from the curve (Fig. 4-3) or calculated from the following
equation derived from the curve:
meq exchangeable ff+ per 100 gm soil= (7.00 - pH observed) x 22 (4-11)
which is applicable from pH 7.00 to 6.65, that is up to 10 meq of exchange-
able H + per 100 gm soil. For amounts of exchangeable H + from 10 to 20
meq, one-half the amount of soil is taken ( 1.25 gm) and the meq obtained
is multiplied by 2.
4-62. Schofield56 employed the Ca salt of the p-nitrophenol buffer to
extract exchangeable hydrogen. This solution is strongly buffered in the
neutral pH range. The extraction solution was titrated to pH 4.5 with dilute

51 Maehl, Ind. Eng. Chem., A.E., 12:24 (1940).

s2 Soil Sci., 60:289 ( 1945).
53 S.S.S.A. Proc., 3: 162 ( 1939); Soil Sci., 66:429 (1948).
54 U.S.D.A. Cir. 757 (1947).
55 Soil Sci., 56:353 (1943).
56/. A~r, ~ci,~ ~l;252 (1933).
76 CATION EXCHAN GE DETERM INATION S FOR SOILS

0
~ 6.90
0
i6.80
z
0 6.70
:r:
0. 6.60
6.50
6 "4001,-....1.1_2L-....1.3_4L-...1.5_61-...l.7-8~J...9--.Jl0-l...1.l--.Jl2~1...1.3_1L..:4:=..il5
meq of total exchangeable
hydrogen per 100 gm of soil

Fig. 4-3. Relation of pH of NH 4 0Ac solution to exchangeabl e

hydrogen extracted from soil. (After Brown, Soil Sci., 56: 353,
1943.)

HCI, brom cresol green being used as an indicator, and the exchangeable
hydrogen was determined by calculating it in relation to a blank.
4-63. Woodruff57 employed a buffer mixture of acetate and p-nitrophe nol
that gives a linear change of pH in relation to exchangeable Hin soils in the
pH range 6 to 7 (compare to Fig. 4-3). The buffer consists of 8 gm of
p-nitrophenol, 40 gm of Ca(0Ac) 2 • Hp, and 0.625 gm of MgO per liter;
the pH of the solution is then adjusted to 7 .0 with HCl or MgO as required.
Five gm of soil, 5 ml of water, and 10 ml of the buffer solution are equi-
librated with stirring and allowed to stand for 30 minutes. Each change in
pH of 0.1 unit going from pH 7.0 to 6.0 equals 1 meq of exchangeable H
per 100 gm. If the pH drops below 6.0, indicating over 10 meq of H per
100 gm, the test is repeated with 2.5 gm of soil; each 0.1 pH unit drop
equals 2 meq of exchangeable H per 100 gm. Because of the linearity of
response, the pH meter dial can be directly calibrated to read as a lime-
meter.
THE LIME REQUIREMENT OF SOILS

4-64. Each meq of exchangeable hydrogen (~ 4-50) to be neutralized

per 100 gm of soil requires 1000 pounds of CaCOa neutralizing equivalence
(~ 4-66) per 2 million pounds of soil. For example, 1 ton of CaC0 11 per
acre plow layer (of 2 million pounds) would be needed to nt!utralize 2 meq
of exchangeable hydrogen present in each 100 gm of soil (Table 5-1 ) . The
CaC011 equivalence of the exchangeable hydrogen of soils is termed the
"lime requireme nt" of soils. Lime requireme nt is also judged from pH meas-
urements (~ 13-88) with consideration for soil texture, from solubility of
iron or release of sulfide ( ~ 13-93), and also through equilibration of soil

u7 Soil Sci., 66:53 (1948); also Graham, Mo. Agr. Exp. Sta. Cir. 345, p. 12 (1950).
CATION EXCHANGE DETERMINA TIONS FOR SOILS 77

with Ca(OH) 2 and C02 , 58 CaC03 , Ca(HC03 ) 2 , 119 and other methods. The
lime requirement should be contrasted with the gypsum requirement
(11 4-44).
4-65. Russell60 states that "the vagueness of the concept of lime require-
ment ... [makes it] a useful working concept that can have no exact mean-
ing." Besides the functional nature of the "exchangeable" hydrogen, de-
pending on the amount of Al(OH) 2 + present (11 4-51 ), the reference pH
(11 4-52) and so forth, the lime requirement concept is further complicated
by differences between plant responses. Consideration must be given to the
degree to which the exchangeable hydrogen should be neutralized. Differ-
ent crops have different optimum soil pH values (1113-89). Also, different
soil colloids have different calcium activities at a given pH. Organic soils
frequently require less liming than mineral soils to reach the point of no
crop response. Some soils require lower pH levels to avoid the appearance
of minor element deficiencies in plants. The actual amount of liming mate-
rial to be applied is also governed by the purity and reaction rate of the
liming material (11 4-73).

NEUTRALIZI NG EQUIVALENC E OF AGRICULTU RAL

LIMESTONE
4-66. When the percentage saturation (11 4-42) of a soil with metallic
cations is too low, agricultural limestones are applied. Only the neutralizing
equivalence of limestone, the "calcium carbonate equivalent," is considered
as a first approximation in "lime requirement" (11 4-64). More detailed
consideration takes into account the proportion of Ca and Mg in the liming
material. Carbonate rocks commonly employed in finely ground form for
liming soils include limestone in which the active material is calcite, CaC03 ,
dolomitic limestone, which contains dolomite, CaC03 • MgC0 3 , and marl,
in which the active material is calcite. Other materials used include burned
limestone containing CaO, slaked lime, containing Ca(OH) 2 with CaC03 ,
and limited amounts of other materials. Collectively, these materials are
termed "liming materials." The titration method given here includes the
determination of the CaC03 equivalence of all 9f these materials in ac-
cordance with the official A.O.A.C. method.

APPARATUS

. 4-67. Needed apparatus includes a torsion. balance; a 500-ml .c_onical

flask; two burets; a steam plate; and, for the estimation of percentage of

58Tr11og, J. Ind. Eng. Chem., 8:341 (1916); Bradfield and Allison, Trans. Intern.
Soc. Soil Sci., A:63 (1933 ); Truog, U.S.D.A. Yearbook; p. 563 (1938).
59 Patel and Truog, S.S.S.A. Proc., 16:41 (1952).
60 Soil Conditions and Plant Growth (London: Longmans Green & Co. Ltd. 1948),
p.105. .
78 CATION EXCH ANGE DETE RMIN ATIO NS FOR SOILS
and a 50-ml
MgC0 3 , a 250-ml beaker, a 200-ml volumetric fl.ask, a funnel,
pipet.

REAGENTS
(the latter
4-68. Needed reagents are standard 1 N HCl and 1 N NaOH
ate, then the titrime tric ratio is
is standardized with potassium acid phthal
and 1 per cent pheno l-
determined between the standard acid and alkali)
brom cresol
phthalein indicator. To determine the percentage of MgC0 3 ,

green indicator is needed.

PROCEDURE
0.5 gm of
4-69. A 1.00-gm sample of dried and ground limestone or
passed the 0.26-mm
burned, or slaked, or other liming material, that has
l conica l fl.ask.
(60 meshes per inch) sieve, is weighed out into a 500-m
is swirled to
Next, 25 ml of standard 1 N HCl is added. The suspension
a steam bath
mix thoroughly and then heated nearly to boiling, and held on
if necess ary to compl ete the reaction).
for 5 minutes (up to 45 minutes
solutio n is boiled for exactly 1
Finally, 100 ml of water is added and the
minute over a small burner, then cooled to room tempe rature.
indicator
4-70. Back Titration. Five drops of l per cent phenolphthalein
with standard
solution are added and the solution is then back titrated
solution is
1 N NaOH to a pink color, which persists for 15 seconds as the
for the pro-
shaken. This solution is saved if Ca or Mg is to be determined
portionality between CaC03 and MgC0 3 ( ~ 4-7 4).
is com-
4-71. The neutralizing equivalence as a percentage of CaC03
puted as follows:

% CaC03 equivalence = ( V - T) x N x ~2a~~JL x _1 ~O

5 (4-12 )
=net ml x -s

back titre,
in which Vis the ml of HCl initially added, Tis the ml of NaOH
N is normality, ands is the sample weight in gm, usually unity.
al is
4-72. Any iron oxide that is dissolved from the original materi
titratio n so that it does not enter into the
reprecipitated from the alkali
For compo unds that have a lower
neutralizing equivalence of the sample.
may exceed
molecular weight than CaC0 3 , the neutralizing equivalence
lence of 1 17 per
100 per cent. For example, MgC0 3 has a CaC03 equiva
The titration
cent, whereas CaO has a CaC03 equivalence of 179 per cent.
decom-
method includes carbonates, oxides, and hydroxides, and easily
posable silicates of alkaline earths.
al to
4-73. Use of CaC03 Equivalence. The quantity of a liming materi
CATION EXCHANGE DETERMINA TIONS FOR SOILS 79

be recommended depends on the CaC03 requirement of the soil (~ 4-64)

and the percentage CaC03 equivalence of the liming material:
Quantity liming _ Quantity CaC03 needed X 100
material recommended - % CaC03 equivalence, liming material
(4-13)
The quantity of liming material to be recommended may be increased over
this estimate in proportion to the estimated percentage of the CaC03
equivalence of the material which will not soon become effective, owing to
large particle size or other factors.
4-74. Estimation of Magnesium Carbonate. Estimation of the MgC08
equivalence of agricultural liming materials is often important. To do this,
the solution that has been back titrated with NaOH (~ 4-70) is acidified to
a pH of 4.5 as indicated by the brom cresol green indicator, and the acidic
solution is then transferred with a funnel to a 200-ml volumetric flask. The
conical flask is rinsed and the rinsing is added to the volumetric flask.
Finally the solution is made to volume and mixed well. The undissolved
portion is allowed to settle to the bottom of the flask, and a 50-ml aliquot
is pipetted from the clear supernatant liquid into a 250-ml beaker. Then Ca
in this solution is determined (~ 5-24).
4-75. The Ca so determined is calculated as the percentage of CaC03 in
the original sample, the fact that only a 0.25 aliquot was taken of the origi-
nal sample being taken into account in the calculation. Then:

% MgCOa -- (nt . 1ence - nt

C CO 3 eqmva . d) X 84.32
CaCO 3 d etermme
equivalence w a -;o 100~
(4-14)
For a more accurate determination of the MgC0 3 content, the Mg is de-
termined quantitatively(~ 5-45) as well as the Ca.

ALTERNATIVE PROCEDURES

4-76. The neutralizing equivalence of the carbonate form of the liming

material may be obtained by measuring the amount of C02 evolved, col-
lected and weighed in soda-lime, 61 or titrated (~ 10-121 ), or measured by
volume(~ 4-77).
4-77. The gas law corrections in gasometric62 determination of CaC03
equivalence can be eliminated by determining a standard CaC03 sample
at the same time and under the same conditions as the limestone equiva-
lence is determined. A weighed sample (approximately 1 gm) of pure

Methods of Analysis, 7th ed. (Washington, D. C.: A.O.A.C., 1950), p. 31.

61
Horton and Newsom, S.S.S.A. Proc., 17•:414 (1~53) after Pierce and Haenisch,
02
Quantitative Analysis, 4th ed. (New York: John Wiley & Sons, Inc., 1947).
80 CATIO N EXCHA NGE DETER MINAT IONS FOR SOILS

Graduated
cylinder 125-mf
flask
Vial

Fig. 4-4. Apparatu s for gasometric determin ation of CaCOa

equivalence of limestone by volume of C02 evolved. (After Hor-
ton and Newsom.)

CaCOa is placed in the 125-ml conical flask (Fig. 4-4) with a small vial
containing 3 ml of concentrated HCI. The 500-ml conical flask is filled
with water. All the connections are secured and the overflow tube is placed
in a graduate cylinder. Then the flask containing the sample is tilted to up-
set the vial of HCI. The volume of water that overflows into the graduate
is equivalent to the volume of C0 2 evolved. The volume of C0 2 evolved
by a limestone sample to be tested is determined in the same manner. The
CaC03 equivalence of the limestone is calculated:
volume co9 evolved
CaCOa equivalence = · · by 1 gm test sa~pl~-- x 100 (4-15)
volume C02 evolved
from 1 gm pure CaCO:i
The authors reported reproducible results with from 0.2 to 1 gm of CaC03
equivalence, and agreement with the titration method on pure calcite and
dolomite within 1 per cent. Since any change in the atmospheric pressure
results in a change in the volume of C0 2 evolved, a sample of pure CaC03
must be tested each time the method is employed.

QUESTIONS
1. Explain the principles employed in the centrifuge washing procedure for
cation exchange capacity determination.
2. Explain why leaching an acid soil with an unbuffered neutral salt solution
is an unsatisfactory method of effecting complete cation exchange.
3. How may the exchange capacity of the organic portion of a soil be
estimated?
4. What are some important advantages of the Ca ion for the determination
of exchange capacity? When may Na be employed advantageously in the deter-
mination of exchange capacity? When may K? When may Mn?
S. Why may the number of "methods" for cation exchange capacity be
very large?
CATIO N EXCHA NGE DETER MINAT IONS FOR SOILS 81

6. Explain why simple dialysis of soil or clay in water results in the satura-
tion of some cation exchange charges with hydrogen.
7. Explain how the total exchangeable metallic cations of a soil can be
measured by equilibration of a sample with a solution of a weak acid.
8. Describe the reactions that occur in the Bray and Willhite method for
total exchangeable metallic cations of soils.
9. Define "exchangeable metallic cation status" of soils.
10. What are the principles employed in the determination of the gypsum
requirem ent of soils?
ex-
11. What are the essential principles employed in the determination of
changeable hydroge n of soils?
n
12. How is the exchangeable hydrogen analysis interpreted for the estimatio
of lime requirement of a soil?
13. How may the CaC03 equivalence of a liming material be estimated?
14. Explain how the dissolved iron and aluminum ions are excluded from
acid-
the neutralizing equivalence of ground limestone as determined by the
hydroxide titration.
15. How may the percentage of MgC0 3 be estimated in connection with
limestone analysis?
 5
Exchangeable Metallic Cation
Determinations for Soils
Exchan geable . . . often termed "available"' constit uents

nt interest
5-1. The exchangeable metallic cation species of most freque
Mn++ , Na+, all of which are readily ex-
in soils are Ca++ , Mg++ , K+,
NH 0Ac extrac t of soil. The princip al
tracted and determined in the I N 4
NH + and H+ (~ 4-50) . Since
nonmetallic exchangeable cations are 4
the exchange-
NH 4 0Ac is employed (~I 5-7) as the extraction reagent for
is necess ary for ex-
able metallic cations, a separate extraction (~ 8-33)
, Fe+ + +, Cu++ , Zn++ ,
changeable NH 4 + determination. While AI+++
chemis try in
and Co++ are extracted by NH 4 0Ac to some extent, their
geable metal-
soils is more complex than that of the more abundant exchan
analyti cal determ ination s are given special considera-
lic cations, and their
tions in Chapters 11 and 15.
extracted
5-2. Three factors tend to raise the quantity of NH 4 0Ac
t truly exchan geable , namely (a) the
metallic cations over the amoun
( ~l 5-3), (b) some weathe ring release
presence of soluble salts in the soil
the course of prolon ged extrac -
of cations from silicate minerals during
and techni que, and ( c) the
tion, controlled by standardized extraction time
. These fac-
dissolution of some of the carbonates of Ca and Mg ( ~ 5-18)
ic cations only ap-
tors tend to make the analysis for exchangeable metall
to make it one of
proximate, yet the determination is sufficiently accurate
( ~ 5-6).
the most practical and valuable means of assessing soil fertility
on ex-
5-3. Exclusion of Soluble Salts. For the most accurate results
in the satura -
changeable cations, the quantities of soluble cations found
ed in the
tion extract (~ 10-78 ) are subtracted from the quantities remov
soluble salts may be remov ed by leaching
NH 4 0Ac solution. Or the water
82
EXCHANGEA BLE METALLIC CATION DETERMINA TIONS 83

the soil with 40 per cent ethanol until the leachate is free from CJ- and
S04 = prior to the NH 40Ac extraction. 1 Soils high in gypsum have been
preleached with water in 1 : 1 weight ratio to soil. 2 There is an equilibrium
between the exchangeable and soluble cations ( 11 10-18), and this equi-
librium shifts as the ratio of water to soil increases. The quantity and nature
of ions in solution therefore depends upon the soil : water ratio, particu-
larly if sulfates are present. For fertility evaluation (11 5-6), the soluble
salts may ordinarily be lumped with the exchangeable (available) cations.
For rather well leached soils that have not been recently fertilized, the con-
tent of water soluble salts may be ignored.
5-4. Report of Exchangeable Metallic Cation Species. Each exchange-
able cation species of soils is reported as meq of the cation per 100 gm of
soil on an oven-dry basis. The sum total of various metallic cation species
as meq per I 00 gm of soil, is found by summation (compare 11 4-33). The
percentage of metallic cation saturation is then computed by equation 4-6
( ~ 4-42). If the soil is known to contain carbonates of Ca and Mg, the
determination of these cations in the aqueous NH 40Ac extract may be
omitted. If the soil is known to contain enough gypsum ( ~ 10-11 3) not to
be completely soluble in the saturation extract, the determination of Ca
may be omitted. If the sum of exchangeable cations greatly exceeds the
cation exchange capacity, indicating the presence of free carbonates or
gypsum (but see 11 4-43), the determinations of Ca and Mg may be disre-
garded on completion of the analysis.
5-5. Meq Percentage of Metallic Cation Species. The amount of each
exchangeable metallic cation is also calculated as percentage of the total
cation exchange capacity ( ~ 4-6), for example, for Na:
. meq of Na per 100 gm of soil-
% Na saturat10n = ---Cation ------·----
exchange capacity
(5-1)

Also, each ion may be calculated as meq percentage of the total major
metallic cations, for example, for Ca:
meq of Ca per 100 gm of soil
meq % Ca = ---------- ·- -· -----····--- (5-2)
meq of (Ca+ Mg+ K +Na) per 100 gm of soil
For neutral soils, the two equations give· approximately the same result
(11 4-43). For soils which have over 15 per cent Na saturation (~ 10-5)
changing the ratio of percentage Na saturation to percentage Ca saturation
is important and the needed change is designated the gypsum requirement
of soil ( 11 4-44) .

1 Gill and Sherman, Pac. Sci., 6: 138 (1952).

2 Piper, Soil and Plant Analysis (New York: lnterscience Publishers, Inc., 1944),
p. 168.
84 EXCHA NGEAB LE METAL LIC CATION DETERM INATIO NS
5-6. rtow-Lay er Volume Basis. In soil fertility evaluation, the amounts
of exchangeable metallic cations are often expressed in terms of pp2m,
that is, pounds of the cation per acre plow-layer volume (~ 2-3) of the
soil. The relationship of meq to pp2m is given in Table 5-1. The quantity
of CaC03 equivalent to each meq of exchangeable H in soil is also given
in the table (discussed in ~ 4-64).

TABLE S-1

Equivalen ce as pp2m of 1 meq of exchangea ble cation per 100 gm of soil and
representa tive amounts in a fertile soil. If the furrow slice weighs 2 million
pounds, pp2m equals pounds per acre.

~~~ivalent of on~--~~j~r~ ()Q g~_of -~0~1_[_~-- Rep~~sentati~~for a ferti_le_s_o_il_ __

__catio~- __ . _ pp2~-- ______pp2m _J meq/100 gm
Ca 401 4010 10
Mg 243 729 3
Mn 549 11 0.02
K 782 234 0.3
Na 460 92 0.2
H 20 30 1.5
CaC0:1 1000

EXTRAC TION OF EXCHA NGEAB LE CATIO NS WITH NH 4 0Ac

5-7. The exchangeable cations are determined after extraction with
1 N NH10Ac. Prianishnikova employed this reagent for the extraction of
exchangeable K. The advantages 4 of this reagent are its effectiveness in
wetting the soil and replacing exchangeable cations, its ease of volatility
during analysis, and its suitability for use with flame emission. Although
some NH4 + is fixed (like K) in nonexchangeable form even under moist
conditions in soils containing vermiculite, accompanied by a decrease in
the cation exchange capacity ( ~ 4-25), the fixation of NH 4 + from NH 4 0Ac
apparently interferes little with the extraction of exchangeable metallic
cations. 5

APPARATUS

5-8. The apparatus needed for extraction of a soil includes a 5 .5-cm

Buchner funnel and 500-ml suction flask and a 250-ml conical flask. To
extract soil in runoff waters, a 11-cm Buchner funnel and 1-liter suction
flask are needed. Also needed are a torsion balance, a beaker of sufficient
volume to hold the NH 4 0Ac extraction solution, a cover glass with glass

3 Landw. Vers. Sta., 79-80:667 (1913).

4 Schollenbe rger and Simon, Soil Sci., 59: 13 ( 1945).
Ii Bower et al., Soil Sci., 73: 251 (1952).
EXCHA NGEAB LE METAL LIC CATIO N DETER MINAT IONS 85

hooks to support it, a glass rod with rubber policeman, a funnel, a 100-ml
volumetric flask, and a steam plate. ·

REAGENTS

5-9. Required reagents are 1 N NH4 0Ac (57 ml of glacial HOAc is

diluted to 800 ml with water, then neutralized with concentrated NH 4 0H
to pH 7 .O; the solution is then diluted to 1 liter); Whatman No. 42 filter
paper, 5.5-cm and 11-cm; 6 N HN03 , 30 per cent H 20 2 , and 6 N HCI.

PROCEDURE

5-10. The Soil Sample. Ideally, the soil sample should be in the field-
moist condition so that ionic equilibria are undisturbed by drying (most
important in relation to NH 4 +, K +, Mn 1 +, and Fe 1 1 ) • T~Qis.tw:e
content is determined by drying a separate sample at 100:.C. In practice, an
air-dried soil sample is frequently extracted. The sample weight in either
case is based on the "fine earth," particles passing a 2-mm sieve, and the
results expressed on the oven-dry basis. Samples rich in coarse and medium
sand are quartered or riffied down to the sample size (50 to 100 gm) for
analysis; the entire sample thus derived is weighed for analysis. Fine tex-
tured soils, on the other hand, may be transferred with spatula or spoon
while being weighed.
5-11. Extraction of Cations from Soil in a Buchner Funnel.' A 50-gm
1

air-dry soil sample (or equivalent weight of field-mo ist soil) is weighed out
into a 250-ml conical flask and 100 ml of 1 N NH 4 0Ac is added. The flask
is stoppered and shaken for several minutes and then allowed to stand
overnight. The contents of the flask is then transferred to a 5 .5-cm Buchner
funnel in which a moist Whatman No. 42 filter paper has been seated by
gentle suction. The soil is leached with an additional 400 ml of NH 40Ac, a
little at a time, so that the leaching process requires at least an hour. Two
reagent blanks are run on the same volume of NH.iOAc. The NH4 0Ac
leachate is evaporated ( ~ 5-14) or analyzed directly by flame emission
(~ 18-24). The exchangeable hydrogen may be estimate
d first from the
pH or titration of the NH4 0Ac solution (~ 4-61). The soil in the Buchner
funnel may be employed to determin e the cation exchang e capacity (~ 4-
25).
5-12. Extraction of Cations from Suspended Soil in Runoff. After
thorough agitation of the original runoff sample, 950 ml is quickly meas-
ured in a 1-Iiter conical flask bearing a 950-ml mark. Then approximately
75 gm of NH4 0Ac (crystals) is added (giving a 1 NNH4 0Ac solution).
The flask is stoppered and the solution is mixed by inversion every 5 min-
utes for 30 minutes (while other flasks are being prepared) . While the

6 Peech et al., U.S.D.A. Cir. 757, p. 8 ( 1947).

86 EXCHANG EABLE METALLIC CATION DETERMIN ATIONS
suspension is still well-mixed, a I -cm layer is poured into a 11-cm Buchner
funnel in which a Whatman No. 42 filter paper has been moistened and
sealed tightly by suction. The filtrate is received in a clean 1-liter suction
flask. (If the filtrate at first comes through cloudy, seldom the case, the
free liquid is suctioned through the Buchner, and the filtrate is washed back
into the original flask). When all of the NH4 0Ac has been suctioned
through, the flask and soil are washed once with about 25 ml of I N
NH 4 0Ac. The filtrate is transferred to a 1-liter beaker and the exchange-
able cations are determined(~ 5-14 ). Reagent blanks are run.
5-13. For rapid determination of exchangeable K only in runoff, a
special extraction and determination by flame emission ( ~ 18-26) is rec-
ommended.
5-14. Evaporation of NH 4 0Ac Extract. The beaker containing the 1 N
NH4 0Ac extract from the soil or suspended solids in runoff is placed on
the steam plate and evaporated to dryness. Then the sides of beaker are
washed down with a small amount of distilled water, and the solution
is evaporated to dryness again. If the residue is dark in color, indicating
organic matter remaining (usually the case), the beaker is fitted with a
closely-fitting cover glass, and then to the beaker are added 2 ml of 30 per
cent H~O~, and 2 ml of 6 N HNOw While tightly covered, the solution is
digested on a steam plate for 20 to 30 minutes. Then the cover glass is
raised on glass hooks and the solution is evaporated to dryness.
5-15. The residue is wetted with 10 ml of 6 N HCI, stirred with a police-
man to bring the acid into good contact, and then the solution is diluted
with 15 ml of water followed by immediate filtration through a retentive
(Whatman No. 42) filter paper to remove silica. The filtrate and washings
are collected in a l 00-ml volumetric flask, made up to volume at room
temperature, and designated Solution A, containing the exchangeable metal-
lic cations. (It is analyzed by ~ 5-24 and following.)

ALTERNATIVE PROCEDURES

5-16. A centrifuge extraction procedure 7 is rapid and convenient for

semimicrochemical methods of determination, and overcomes the diffi-
culty encountered in leaching soil samples of low permeability. A sample
weight 8 of 4 gm of medium- or fine-textured soil or 6 gm of coarse-textured
soil of known moisture content is weighed out and transferred to a 50-ml
round-bottom narrow-neck centrifuge tube. Then 33 ml of neutral 1 N
NH4 0Ac solution is added, the tube is stoppered, and the suspension is
shaken in a reciprocating shaker for 5 minutes. The stopper is then re-

7 Bower et al., Soil Sci., 73:251 (1952).

s Must be increased for soils of low metallic cation status to give determinable
concentrations in the extract volume.
EXCHA NGEAB LE METAL LIC CATION DETERM INATIO NS 87

moved, and the tube is centrifuged until the supernatant liquid is clear,
usually requiring 5 minutes in the International No. 2 centrifuge at a speed
of 1,500 to 2,000 rpm. The supernatant liquid is decanted as soon as pos-
sible into a 100-ml volumetric flask. The sample is then extracted 2 addi-
tional times with 33 ml of NH 40Ac each time, the supernatant liquid be-
ing decanted into the same volumetric flask. The solution in the flask is
made to volume and mixed, and the various cations extracted are deter-
mined by suitable semimicro procedures, such as the flame emission spec-
trophotometer ( ~ 18-7).
5-17. Extraction of Cations from Soil in a Carbon Funnel. 9 The end of
a small wad of fine Pyrex glass wool, or cotton is loosely stuffed into the
stem of a carbon funnel leaching tube (Fig. 5-1 ) with a glass rod. Then

75cm

Fig. 5-1. Leaching apparatus for exchangeable cations.

(After Schollenberger and Simon, Soil Sci., 59: 13, 1945.)

additional glass wool or cotton is arranged above to form a fiat pad with
a rubber stopper mounted on the end of a glass rod. A little water is run
through to make sure that the percolation rate will not be too slow because
of too tight packing of the glass wool. The glass wool is rearranged if
necessary. Next, the soil sample, consisting of up to 100 gm of air-dry

9 Schollenberger and Simon, Soil Sci., 59: 13 (1945).

88 EXCHANGEA BLE METALLIC CATION DETERMINA TIONS
sifted soil or the equivalent weight of field-moist soil is placed in the car-
bon funnel in several portions, each portion being packed slightly by jab-
bing with a spatula, to reduce shrinkage on wetting. Then a 1-cm layer of
glass wool is placed above the soil. The apparatus is next assembled (Fig.
5-1 ), the upper flask with 750 ml of the 1 N NH 4 0Ac being placed on a
shelf, and connected with a siphon to deliver through a tight-fitting stopper
into the leaching tube. Suction is applied at the collection flask to start the
siphon, and then removed. Finally, the screw clamp between the extrac-
tion tube and the receiving flask is adjusted to regulate the flow rate. The
time for leaching should be at least 4 hours, but no longer than overnight,
if the nature of the soil permits. The leaching should now proceed without
further attention ( ~ 5-14). Two blanks for each set of soil extractions are
run on the same volume of NH 4 0Ac, the same quantities of reagents be-
ing added throughout.
EXCHANGEA BLE CATIONS OF CALCAREOU S SOILS
(Extracted with BaCb-triethanola mine)
5-18. Extracting calcareous soils with NH 4 0Ac dissolves Ca and Mg
extensively from CaC0 3 (calcite) or Ca CO a · MgC0 3 (dolomite), beyond
their true activity for plant use. Kelley 10 concluded that exchangeable Ca
and Mg determinations for calcareous soils were meaningless, and pre-
ferred to estimate their sum as the exchange capacity less exchangeable K
and Na. Bower et al., 11 also disregarded the Ca and Mg extracted in
NH 4 0Ac if the soil was calcareous or gypsiferous. Peech 12 estimated ex-
changeable Ca in saturated soils as exchange capacity less exchangeable
K, Na, and Mg. There is still a demand for an extraction method for the
truly exchangeable Ca and Mg from calcareous soils. The various propos-
als include (a) use of 40 to 80 per cent alcoholic solutions 1 :1 of NH 4 0Ac,
(b) corrections for the solubility of CaCOa in the leachings, 14 or ( c) use
of alkaline solutions such as Na 2C0/ 5 or BaC1 2 -triethanolamine extrac-
tion solution of pH 8.1 (to be described), in which BaC0 3 coats the
surface of calcite and dolomite and makes them insoluble.

APPARATUS

5-19. Needed apparatus includes a crucible with a perforated bottom, a

crucible holder mounted on a suction flask, a 100-ml volumetric flask, a
25-ml pipet, a buret, and a 150-ml beaker.

Agr. Sci., 24:72 ( 1934).

10 J.
Soil Sci., 73:251 (1952).
11
12Jnd. Eng. Clzem.,A.E., 13:436 (1941).
13 Magistad and Burgess, Ariz. Agr. Exp. Sta. Tech. Bui. 20 (1928).
14 As reviewed by Kelley, Cation Exchange in Soils (New York: Reinhold Publish-
ing Corporation, 1948).
15 Puri, Soils (New York: Reinhold Publishing Corporation, 1949).
EXCHA NGEAB LE METAL LIC CATIO N DETER MINAT IONS 89

REAGENTS

5-20. Needed reagents include 0.2 N BaC12-triethanolamine of pH 8.1

(made up 16 of 25 ml of commercial triethanolamine, specific gravity 1.126
which is about 8 N, diluted to 250 ml with water and partially neutralized
with HCl to adjust the pH to 8.1 : this requires approximately 90 ml of 1 N
HCl: this solution is diluted to 500 ml with water and then mixed with 500
ml of 0.4 N BaC12 and the resulting solution is mixed and protected from
the C0 2 of the air). Also needed are filter paper discs, 0.1 N H 2 S04 , methyl
orange indicator, 4 per cent (NH4 ) 2C20 4 • H 20 and standard 0.025 N
KMn0 4 •

PROCEDURE! 7

5-21. A quantity of soil or a colloidal electrolyte equivalent to 0.5 to 1

meq of exchangeable metallic cations is weighed into a perforated crucible
fitted with a small, moistened filter paper disc. The sample is leached with
50 ml of 0.2 N BaC12 -triethanolamine of pH 8.1 and then is washed with
50 ml of H 2 0. The leachate is made to a volume of 100 ml. (To deter-
mine the exchange capacity with Ba, the soil is leached once more, with
BaC1 2 without triethanolamine, ~ 4-32).
5-22. Exchangeable Ca Determination. The exchangeable Ca is de-
termined on a 25-ml aliquot of the leachate made to 100 ml (~ 5-21). To
the aliquot, 25 ml of 0.1 N H 2S04 and 1 drop of methyl orange indicator
are added. Next 20 per cent NaOAc is added until the rose-orange color
disappears. The mixture is heated to about 70°C, and two 5-ml portions
of 4 per cent (NH 4 ) 2 C 2 0 4 • H 2 0 are added slowly with stirring. After
digestion for about an hour, the precipitate (BaS0 4 and CaC 2 0 4 ) is fil-
tered and washed with warm water. The CaC 2 0 4 is dissolved from the
precipitate in 50 ml of approximately 1 per cent H 2S0 4 , heated to 80°-
900C and titrated with 0.025 N KMn0 4 •
5-23. Determination of Mg. The Mg is determined on the filtrate from
the Ca determination, by means of the usual procedure ( ~ 5-45).

CALCI UM DETER MINAT ION

(As oxalate by titration with cerate or permanganate)
5-24. The exchangeable Ca obtained in the NH 40Ac extraction(~ 5-14)
may readily be determined by means of the oxalate-cerate titration pro-
18

16AfterP eechetal ., U.S.D.A. Cir. 757 (1947).

17 After Mehlich, Soil Sci., 60:289 (1945).
18Adapte d from Ellis, Ind. Eng. Chem., A.E., 18:426 (1938), and Katzman and
less con-
Jacobi, J. Biol. Chem., 118:539 (1937). Titration with permanganate is ton,
venient though still a standard procedure, Methods of Analysis, 7th ed. (Washing
D.C.: A.O.A.C ., 1950), p. 33.
90 EXCHANGEABLE METALLIC CATION DETERMINATIONS
cedure. Also, Ca from total elemental analysis of soils ( ~ 11-162), from
plant tissue (~ 12-39), from water analysis (~ 10-85), or from other
sources can be determined. Ions that may coprecipitate with Ca as the
oxalate include Sr, Fe++, Ti, Mn++, Li, Be, Mg, and Ba. For routine Ca
determinations for soils and plants, the interferences are almost invariably
negligible. When the coprecipitation of Mg is known to be a factor, rarely
the case, either a double precipitation (~ 11-166) or a high dilution com-
bined with the semimicro oxalate procedure (~ 5-38) for Ca is an effec-
tive remedy, as is also the use of the Versene procedure ( ~ 4-21 or 11-47)
or the flame emission spectrophotometric procedure ( ~ 18-7).

APPARATUS

5-25. Needed apparatus includes a 250-ml beaker, a buret containing

4 N NH 40H, a steam plate, and a fritted glass 19 crucible (medium po-
rosity) with a holder and suction flask.

REAGENTS

5-26. Needed reagents are 6 N HCl; glacial HOAc; concentrated and

4 N NH 4 0H; brom cresol green indicator; 10 per cent H 2 C 2 0 4 solution;
2 N H 2 S0 4 (55 ml of concentrated H 2 S0 4 slowly added to 900 ml of
water, then diluted to 1 liter); 0.025 M "Ferroin" indicator ( orthophenan-
throline-ferrons complex, from G. F. Smith Chemical Co., Columbus, 0.);
and the following special reagents.
5-27. Standard Oxalate. Standard Na 2 C 2 0 4 is dried in an oven at 100°C
and cooled. Then a 3.350-gm portion is weighed out and dissolved in ap-
proximately 600 ml of 2 N H 2 S04 • The solution is then made to a volume
of 1 liter in this solvent. This primary oxidation-reduction standard is
0.0500N.
5-28. Standard Cerate. Approximately 34 gm of (NH 4 ) 4 Ce(S0 4 ) 4 •
2 H 2 0, ammonium tetrasulfatocerate, is dissolved in 1 liter of 2 N H 2 S04
to obtain a solution which is approximately 0.05 N with respect to oxidation
reduction, Ce 4 + ~ CeH. (If permanganate is to be employed, a stock
solution of 1 N KMnO4 is prepared as directed in connection with the Mn
determination ( ~ 5-67), and an aliquot is diluted to 0.05 N and standard-
ized.) It is preferable to standardize the oxidant solution without further
dilution, rather than to attempt to dilute it to exactly 0.05 N.
5-29. Standard Ferrous Solution. A standard ferrous solution is pre-
pared by dissolution of 30 gm of Fe(NH4 ) 2 (S0 4 ) 2 • 6 Hp in 1 liter of
approximately 0.5 N H 2S04 (exact concentration of ferrous is not impor-

19 Marsden, J. Soc. Chem. Ind., 60:20 (1941). Found also by this author to be
eminently suitable, owing to low blank, rapidity, and convenience.
EXCHANGEABLE METALLIC CATION DETERMINATIONS 91
tant). This solution is approximately 0.05 N with respect to oxidation-
reduction. A 10-ml aliquot of the cerate solution is titrated with the ferrous
solution to the "Ferroin" end point just before each set of titrimetric de-
terminations, and the factor, R, for the cerate equivalent of the ferrous is
calculated (ml Ce/ ml Fe = R). Then, 20 ml of the standard cerate is
pipetted into 10 ml of standard oxalate in a beaker and the solution tem-
perature is brought to about 80°C and finally back to less than 50°C. The
excess cerate is then back titrated with ferrous solution to the "Ferroin"
end point. Then
ml oxalate x 0.05
N c.. = (ml cerate) - (ml ferro_u_s_x_R_)- {5-3)

The above procedure for standardization of the cerate is conventional for

sulfatocerate oxidimetry. 20

PROCEDURE

5-30. An aliquot of Solution A equivalent to 0.1 to 1 meq (2 to 20

mgm) of Ca (15 ml for silt loams, 25 ml for sands, 40 ml for runoff) is
transferred with a pipet to a 250-ml beaker and diluted to about 150 ml.
Then 5 ml of 6 N HCI is added to provide ample NH 4 Cl to prevent co-
precipitation of Mg with the CaC 2 0 4 • Mg, Fe, Al, and other ions remain
quantitatively in solution with the acid 21 precipitation employed, in the
presence22 of excess NH 4 Cl, NH 4 0Ac, and (NH 4 ) 2 C2 0 4 • Rarely, with
very unusual quantities of Mg present relative to Ca, is reprecipitation
necessary ( ~ 11-160).
5-31. Precipitation of CaC 20 4 • To the acid solution of Ca and Mg,
containing 5 ml or more of concentrated HCl in excess, 5 ml glacial HOAc
and 0.5 ml of brom cresol green indicator are added. Then concentrated
NH 4 0H is added with stirring, until the first greenish color appears; next
the solution is heated to boiling. Now 10 ml of 10 per cent oxalic acid
solution is added, which makes the solution strongly acid again (full yel-
low color). Finally, 4 N NH 40H is added dropwise from a buret with
rapid stirring until a strong green color appears, indicating pH 4 to 4.5.
Calcium oxalate precipitates slowly in the hot acid solution during the

20 Smith, Cerate Oxidimetry (Columbus, Ohio: G. F. Smith Chemical Co., 1942),

p. 81; Larson and Greenberg, J. Biol. Chem., 123:199 (1938); Reitemeier, Ind. Eng.
Chem., A.E., 15:393 (1943).
21 Chapman, Soil Sci., 26:479 (1928), employed a precipitation pH of about 4.0,
which suggests that brom phenol blue indicator (critical pH 4.0) may be employed
instead of brom cresol green.
22 That the use of ammonium and organic chelating agents permits a single pre-
cipitation separation of Ca from Mg is shown by Chapman, Soil Sci., 26:479 (1928);
Wright and Delaune, Ind. Eng. Chem., A.E., 18:426 (1946); and B. Chatterjee and
J. L. Huber working in the author's laboratory. See also semimicrochemical procedure
(11 5-38).
92 EXCHANGEABLE METALLIC CATION DETERMINATIONS
additions of NH 4 0H, with only traces of Ca left to precipitate as pH 4.5 is
approached. The solution is digested on the steam plate for 1 hour, which
suffices to increase the crystal size and leave a clear supernatant liquid.
Prolonged digestion is avoided, because Mg is likely to begin coprecipita-
tion.
5-32. Filtration of CaC 2 0 4 • The clear supernatant solution is decanted
through a fritted glass crucible (medium porosity) with suction. The
precipitate is suspended in the last of the liquid and poured into the filter.
Then the beaker is rinsed 3 times, thoroughly, including the sides but with
a minimum volume of water delivered as a fine jet, the rinsings being
poured through the filter. The filter is next washed 5 times with small
amounts of hot water. The filtrate is saved for magnesium. (In place of a
fritted glass filter crucible, filter paper, a Whatman No. 42, and a fluted
funnel may be used, in which case washing is continued until the washings
are free from chlorides, as shown by test of 3 ml of washings with 0.5 per
cent AgN0 3 solution, the tested solution being discarded. About 10 wash-
ings are usually required.)
5-33. Titration of CaC 2 0 4 • The fritted glass crucible and precipitate
are placed into the original beaker. Then 75 ml of 1 N H 2 S04 is added, as
well as sufficient water to cover the crucible. The solution is warmed to
90°C and promptly titrated with 0.05 N (NH 4 ) 4 Ce(S0 4 ) 4 (with ortho-
phenanthroline indicator) or KMn0 4 • The blank on fritted glass is negli-
gible. The oxidizing solution is freshly standardized against the primary
standard Na 2 C 2 0 4 and the normality of the oxidizing solution is recorded,
with no attempt to adjust it to exactly 0.05 N. For work involving 0.1 meq
of Ca or less, 0.015 N solutions are employed (~ 5-38). (In case a filter
paper was used, it is pierced, and the main mass of precipitate is washed
into the original beaker. Then 75 ml of H 2 S04 is poured over the paper to
dissolve the precipitate in the pores of the paper. The paper is then washed
several times with hot water. Finally the solution is diluted to 150 ml with
hot water, heated to 90°C and titrated with 0.05 N cerate or permanga-
nate. Near the end of the titration the paper is placed in the solution, and
the titration is continued until the end point is permanent for at least 30
seconds. The blank titration obtained with filter paper alone, which usu-
ally amounts to 0.1 to 0.2 ml, is subtracted from the main titre.)
Equations:
CaC2 0 4 + H 2 S04 ~ H 2 C 20 4 + CaS0 4 (5-4)
C2 0 4 - - + 2 Ce4+ ~ 2 Ce 3 + + 2 C0 2 (5-5)
or, molecularly,
H 2 C 20 4 +2 (NH 4 ) 4 Ce(S0 4 ) 4 ~ Ce 2 (S0 4 ) 3 + H 2S04
+ 4 (NH4 ) 2S04 + 2 C02
(5-6)
EXCHANG EABLE METALLIC CATION DETERMIN ATIONS 93

5-34. For soils, the results are reported as meq of Ca per 100 gm of
soil:
cerate )
meq of Ca per 100 gm of soil = (ml cerate) x ( normality
x (aliquot ) x _!Q2
factor s
(5-7)
The aliquot factor is 100/ml of Solution A taken. One ml of 0.05 N cerate
is equivalent to 1.0 mgm of Ca or 1.4 mgm of CaO.
5-35. For runoff, the pounds of CaCO:i per acre-inch of runoff (ppai)
is calculated for 40/100 aliquot of the 9 50 ml of runoff extracted as fol-
lows:
pounds CaC0 3 (ppai) = meq of Ca x 29.8 (5-8)

The factor 29.8 is derived as follows:

( meq wt. CuCO:i) (aliquot factor) (H,Q, ppai)

ALTERNATIVE PROCEDURES

5-36. The oxalate of the CaC 2 0 4 prec1p1tate may be evaluated alter-

natively by a gravimetric procedure if over 40 mgm of Ca is available. A
number of semimicro and microchemical procedures for Ca are available,
including the Versene method (~ 11-47), the picrolonate 2 :1 method, the
chloranilate 24 method, and the oxalate 25 methods. A semimicro oxalate
procedure (~ 5-38) is effective for amounts of Ca from 0.0 I to 0.25 meq
( 0.2 mgm to 5 mgm) and is readily adaptable outside this range.
5-37. Gravimetric Procedure. In place of the titration procedure, the
precipitate of calcium oxalate may also be ignited and weighed, if 2 meq
(56 mgm CaO) or more of Ca is being determined. In this case the precipi-
tate is washed free of chlorides with hot neutral I per cent (NH 4 ) 2 C 2 0 4
solution in place of hot water. The precipitate and filter paper are then
ignited in a porcelain crucible and the CaO is weighed.
5-38. Semimicro Oxalate Procedure. 2 u Ca is precipitated in a concen-

23 Snell and Snell, Colorimetric Methods of Analysis, Vol. I (New York: D. Van
Nostrand Company, Inc., 1941), p. 466; also 3rd ed., Vol. II (New York: D. Van
Nostrand Company, Inc., 1949), p. 602. Alten et al., Biochem. Z., 265: 85 ( 1933);
Bollinger, Australian .f. Exp. Biol. Med. Sci., 13:75 (1935).
24 Tyner, Anal. Chem., 20:76 (1948).
2r. Reitemeicr, T11d. Eng. Chem., A.E., 15:393 (1943); Peech et al., U.S.D.A.
Cir.
757, p. 18 (1947).
2 n Grateful acknowledgme nt is made to J. L. Huber and Dr. R. B. Corey for
assistance to the author during the adaptation of details of this procedure, and the
study of the range of its freedom from interference.
94 EXCHANGEA BLE METALLIC CA TlON DETERMINATJONS
trated solution of salts, under which conditions coprecipitation with Mg is
negligible up to a 20 : 1 Mg to Ca ratio with 1 mgm or less of Ca, up to a
ratio of 3 : I for 2 mgm of Ca, and up to a ratio of 1 : 1 for 5 mgm of Ca.
The method is amply sensitive to 1 mgm of Ca or less, so that a great
enough dilution of the Ca solution may nearly always be made to insure
freedom from Mg interference with a single precipitation. The evaluation of
the oxalate may be either titrimetric, or colorimetric by means of the excess
cerate color.
5-39. The reagents required are similar to those for the macro procedure
except that the standard reductor and oxidant solutions are made more
dilute. To make standard 0.015 N reductor solution, 1.005 gm of N a 2 C2 0 4
is dissolved in 1 liter of 2 N H 2 S04 in a volumetric flask. To make 0.0 I 5 N
cerate, 9 gm of (NH 4 ) 4 Ce(S0 4 ) • 2 H 2 0 is dissolved in 950 ml of 2 N
H 2 S0 4 in a 1.5-liter flask; the strength of this solution is determined in ac-
cordance with the oxidation-reduction titration procedure detailed later in
this paragraph; then the solution is diluted to give exactly 0.015 N. The
ferrous reference solution is prepared by dissolution of 5 gm of Fe (NH 4 ) 2
(S0 4 ) 2 • 6 H 2 0 in 1 liter of approximately 0.5 N H 2 S0 4 (exact concentra-
tion of ferrous is not important). A 10-ml aliquot of the cerate solution is
titrated with the ferrous solution to the "Ferroin" end point just before
each set of titrimetric determinations, .and the factor, R, for the cerate
equivalent of the ferrous is calculated (ml Ce/ ml Fe = R). Then, 20 ml
of the standard cerate is pipetted into 10 ml of standard oxalate in a beaker
and the solution temperature is brought to 80°C and finally back to less
than 50°C. The excess cerate is back titrated with ferrous solution. Then:
ml oxalate x 0.015.
N. = - -----···--· ---~---- (5-9)
ce (ml cerate) - (ml ferrous X R)
The precipitate washing solution consists of an ethanol-NH 4 0H solution:
concentrated NH4 0H is diluted with 9 volumes of water and then this dilute
solution is mixed with an equal volume of ethanol.
5-40. To the Ca solution, containing 0.2 to 5 mgm of Ca in a volume of
10 to 20 ml, are added 1 ml of 6 N HCl, 6 ml of glacial HO Ac, 3 drops of
brom cresol green indicator, and concentrated NH 4 0H dropwise from a
buret until the solution turns from yellow to faintly green. Next, 2 ml of 10
per cent H 2 C2 0 4 solution is added and the solution is heated to boiling.
Finally, 4 N NH 4 0H is added from a buret until the color changes to a full
green or bluish green. The CaC 2 0 4 ordinarily will have begun precipitation.
The beaker is placed on a steam plate for digestion of the precipitate and
is allowed to remain for about 30 minutes. If no precipitate is apparent
· after 10 minutes of digestion, indicating only a trace of Ca present, 10 ml
of 95 per cent ethanol is added and the precipitate and the solution are
digested for another 20 minutes.
EXCHAN GEABLE METALL IC CATION DETERM INATION S 95

5-41. When the digestion is completed, a Pyrex fine porosity filter tube
is placed in the beaker, arranged so that the supernatan t solution can be
drawn off by suction into a 300-ml conical flask (Fig. 5-2). The precipi-

(To suction (through water trap)

Glass tubing

Surgical
rubber
tubing
300 ml
conical Pyrex (fine)
flask filter tube

150 ml
beaker

Hot water
I bath

Mg in solution

Pig. 5-2. Filtration apparatus for semimicro oxalate pro-

cedure for Ca. (Sketch courtesy J. L. Huber and Dr. R. B.
Corey.)

tation beaker is kept in a hot water bath to increase the rate of filtration.
The supernatan t solution is drawn off almost completely and the first wash-
ing solution is added just as the last of the solution leaves the beaker, so
that (NH 4 ) 2 C 2 0 4 will not dry on the glass. (Dried (NH 4 ) 2 C 2 0 4 may not
redissolve completely in the washing solution and high results would be ob-
tained.) The precipitate is carefully washed 4 times with small portions of
the NH 4 0H-ethano l washing solution, the filter tube being allowed to draw
off all the washing solution each time.
5-42. The precipitate and filter tube contained in the original beaker,
are dried on a steam hot plate. When the precipitate and filter tube are dried
completely, approximately 30 ml of hot (90°C) H 2 S04 is added to dissolve
the precipitate including any adhering to the filter tube. After the precipi-
tate has dissolved, filtered air is forced back thropgh the stem of the filter
tube to remove the adhering solution. The filter tube is then raised from the
solution and about 5 drops of water are placed in it with the capillary
dropper and forced back through the fritted glass surface by filtered air to
complete the washing of the dissolved oxalate from the interior of the filter
96 EXCHANGEABLE METALLIC CATION DETERMJNA TIONS
tube. The exterior of the tube is flushed off with distilled water. (Fritted
glass crucibles are also very satisfactory for this filtration.)
5-43. A 10-ml aliquot of 0.015 N sulfatocerate solution is added to the
oxalate solution in H.!S0 4 • If the solution becomes colorless, more than
3 mgm of Ca is present, and so another 10-ml aliquot of the sulfatocerate
is added. A third 10-ml aliquot of sulfatocerate is added if the second be-
comes colorless. Not over 9 mgm of Ca should be determined ordinarily in
this procedure, and a smaller aliquot should be taken if more than 30 ml
of the sulfatocerate is required. The temperature of the solution is raised to
about 80°C to complete the oxidation reaction. The solution is then cooled
to below 50°C and excess cerate is back titrated with ferrous reference
solution to the end point with "Ferroin" indicator. A 10-ml portion of this
standard cerate solution is titrated with ferrous at the same time and the R-
value is rechecked at the time of the determinations.
5-44. Instead of back titration of the excess cerate, the yellow colored
solution may be evaluated colorimetrically. 27 It is cooled to room tempera-
ture, transferred to a 100-ml volumetric flask or smaller volumetric flask.
The filter and beaker are washed once with a 20-ml portion of 2 N H 2 S0 1
and finally with a fine jet wash bottle. The volume is adjusted exactly to
the mark with 2 N H 2S0 4 and the solution mixed. A portion is transferred
to a colorimeter tube for reading which may be taken immediately with a
420 mu light maximum. The color is stable overnight. The blank for I 00
per cent transmission setting is obtained with 2 N H 2 S0 4 • The mgm of Ca
in the sample is obtained from a standard curve. The standard curve is
prepared with increments of 0 to 10 ml of cerate similarly diluted. The

Ml of 0.015 N sulfatocerate added

10 20 30
0
Residual
Titer color
------
Residual
__ :U!.e_r:_____ _______ color
Residual
__ J~~--------------------------~,._c_ol_or~

0 3 6 9
Mgm of Ca determined

Fig. 5-3. Principle of colorimetry for semimicro oxalate

procedure for Cjl by residual sulfatocerate. (Sketch courtesy
J. L. Huber.)

27 After the procedure outlined by Weybrew et al., Anal. Chem., 20:759 (1948),
adapted in the author's laboratory with the generous assistance of Dr. B. Chatterjee
and J. L. Huber.
EXCHANGEABLE METALLIC CATION DETERMINATION S 97

curve may be checked by a series of standard Ca increments. 28 The reading

of Ca in the 0 to 3.0, 3.0 to 6.0, or 6.0 to 9.0 mgm of calcium range is il-
lustrated in Fig. 5-3.

MAGNESIUM DETERMINATION
(Titration of MgNH 4P0 4 and MnNH4P04 to NH 4 H2P0 4 , then Mn colorimetrically
and Mg by difference)
5-45. The exchangeable Mg obtained by the NH4 0Ac extraction may
be determined readily by the H 2 SO 4 titration of the phosphate precipi-
tated, 20 if 2 mgm or more is available. Alternatively, Mg can be determined
by the Versene titration ( ~l 11-54), by the micro 8-hydroxy quinoline pro-
cedure ( ~ 5-59), or the flame emission spectrophotometer ( ~ 18-24).

APPARATUS

5-46. Needed apparatus includes a 300-ml conical flask with a mechani-

cal shaking device or a' system of bubbling air through to stir the solution;
a fine porosity fritted glass filter, holder, and suction flask; a 1.5 N NH 4 0H
wash bottle; and a 250-ml beaker.

REAGENTS

5-47. Needed reagents are concentrated NH 4 0H, brom cresol purple

and brom cresol green indicators, IO per cent (NH 4 ) 2 HP0 4 , 1.5 N NH 4 0H,
95 per cent and absolute ethanol or methanol, and standard 0.0714 N
H 2 S0 4 and NaOH.

PROCEDURE

5-48. Precipitation of Mg and Mn. The filtrate from the calcium de-
termination containing over 2 mgm of Mg, is brought to pH 6.2 to 6.4
with NH 4 0H (brom cresol purple indicator) and heated on the steam plate
until evaporated to 150 ml. If a precipitate has formed (probably Fe
(0H) 3 ), it is filtered off. It is important that the pH not exceed 6.4, or
Mn0 2 may precipitate.
5-49. The solution is transferred to a 300-ml conical flask and then 20
ml of 10 per cent solution of ( NH 4 ) 2 HP0 1 is added slowly with stirring.
After thorough stirring, the solution is made strongly alkaline by slow
addition, with more stirring, of 30 ml of concentrated NH 4 0H. The solu-
tion is allowed to stand overnight. For small amounts of Mg, the solution
must be agitated by a mechanical reciprocal shaker or by aspiration to in-
sure complete precipitation.

2 8 Equivalent curves have been shown for the dilution of cerate, and partial oxida·
tion of it by standard Na2C 2 0 4 and standard increments of Ca as CaC204 in sys-
tematic studies by J. L. Huber in the author's laboratory.
29 Dean and Truog, Ind. Eng. Chem., A.E., 7:383 (1935).
98 EXCHANGEABLE METALLIC CATION DETERMINATIONS
5-50. Filtration. The solution is decanted through a retentive fritted
glass filter arranged with a suction pump (or filtered through a Whatman
No. 32 or 42 filter paper). The flask (and aspirator tube) are washed 3 or
4 times with small portions of 1.5 N NH4 0H solution. It is not necessary
to remove the precipitate sticking to the flask and aspirator tube. The
fritted glass filter and precipitate are then washed 3 times with the 1.5 N
NH 4 0H, the solution being allowed to drain from the precipitate com-
pletely between washings ( 10 or 12 washings are required if a filter paper
is used). The removal of the last traces of free ammonia from the flask,
aspirator tube, and filter are now accomplished by drying by means of two
rinsings with 95 per cent ethanol followed by 1 with absolute ethanol or
methanol and aspiration for a few minutes. The filtrate is retained until
complete recovery of the Mg is assured and then discarded.
5-51. Note on Drying Paper Filter. Free exposure to the air at room
temperature over night suffices for drying the filter paper in lieu of alco-
hol washing. Drying may be accomplished at a higher temperature in 2 or
3 hours, but this temperature must not exceed 45°C. Free NH 8 , if left in
the filter, will enter in the titration just as the MgNH 4 P0 4 and vitiate the
results.
5-52. Titration80 of Mg and Mn Phosphates. After all free ammonia is
removed from the precipitate and glassware, a carefully measured excess
of standard 31 0.0714 N H 2 S04 (usually 15 ml will suffice) is dispensed
from a buret into the original 300-ml conical flask (to which some of the
now washed and dried precipitate is adhering). The acid is diluted with
10 ml of water, and this solution warmed briefly on the steam plate to dis-
solve the precipitate. The flask is rotated to effect contact with the surface
of the flask and aspirator tube. This solution is then transferred quantita-
tively to a 250-ml beaker in which the fritted glass filter carrying the bulk
of the precipitate has been placed. Sufficient water is added to cover the
filter and 0.5 ml of 0.04 per cent brom cresol green indicator is added.
Warming on a steam plate is continued for 20 minutes with occasional
stirring to help dissolve the precipitate. (If a filter paper was used, the
paper and standard acid are added directly to the flask.) If the color shifts
from yellow to green, an additional measured amount of standard H 2 S04
is added to maintain an excess.
5-53. The excess H 2 S04 is then back titrated in the beaker containing
the fritted glass filter (or in the flask after addition of 100 ml of water in
case the filter paper has been used) with standard 0.0714 N Na OH to the
first distinct blue-green color. The titrated solution is saved for manganese.

Method proposed by Handy, J. Am. Chem. Soc., 22:31 (1900).

30
Use of 0.1 or 0.05 N H2S04 and NaOH may be substituted if desired according
31
to the amount of Mg present. The more dilute reagents are employed if necessary to
give a net titer of at leust 3 to 5 ml.
EXCHANGEABLE METALLIC CATION DETERMINATION S 99

A color standard for the end point is prepared by dissolving 1 gm of

KH 2 PO 4 in 140 ml of water and adding 10 drops of brom cresol green indi-
cator. The titration represents a conversion of MgNH 4 P01 and MnNH 4 P0 4
to NH 4 H 2 P0 4 •
5-54. The volume of standard H 2 SO4 required may be small, and there-
fore both standard solutions are very carefully standardized 32 and the buret
readings are estimated to within 0.01 ml. (With the use of filter paper, a
blank determination must be made, and applied to the titration value to
account for the appreciable amount of acid consumed by the filter paper,
often as much as 0.3 ml.)
5-55. Calculation of Results. The Mg determined is calculated:
meq of Mg per 100 gm soil = ml H 2 SO4 x normality H 2 SO 4 x 100 / s
(5-10)

wherein s = weight of soil represented. Each ml of 0.0714 N is equivalent

to 0.868 mgm of Mg.
MgNH 4 PO 4 + H 2 SO 4 ~ NH 4 H 2 PO4 + MgSO 4 (5-11)
(pH 4.6 m<l point)

The manganese present reacts in the same way as magnesium, and if ap-
preciable (usually not the case), the meq of Mn is subtracted from the titre.
5-56. Gravimetric Method. The MgNH 4 P0 4 and MnNH 4 P0 4 may be
ignited to the pyrophosphate according to the following equation, after
which, it is determined gravimetrically. 33
2 MgNH 4 P0 4 ~ Mg 2 P20 7 + 2 NH3 + H 2 0 (5-12)
(heat)

meq Mg per 100 gm soil= gm Mg 2 P 20 7 X 8.98 x l~Q (5-13)

Each mgm Mg 2 P 2 0 7 is equivalent to 0.218 mgm Mg. The gravimetric

method thus is most applicable when quantities of Mg exceed 10 or 20
mgm. If the content of Mn is appreciable, it is determined separately
(~ 5-71) and the total weight obtained on ignition is corrected to Mg
accordingly.

ALTERNATIVE PROCEDURES

5-57. Mg may be determined by a number of other satisfactory pro-

cedures; for semimicro quantities the alternative procedures are preferable
to the titration procedure given above. The MgNH 4 P0 4 precipitate can be
evaluated by a colorimetric determination of the phosphorus (the titrimetric
a2 The standard NaOH is titrated against a weighed portion of KH-phthalate, and
the acid-base ratio is determined of the solutions used.
33 This method is employed by A.0.A.C., Methods of Analysis, 7th ed (Washing-
ton, D.C.: A.O.A.C., 1950), p. 22.
100 EXCHA NGEAB LE METAL LIC CATION DETERM INATIO NS
procedure is essentially a titration of the first hydrogen of H 3 P0 4 and thus
the colorimetric procedure is essentially the same in principle, but is enor-
mously more sensitive.) The colorimetric evaluation of the precipitate has
been employed.:14 The evaluation can be made with either the blue or
yellow molybdophosphoric color (Chapter 7). A Beer's law relation be-
tween molybdophosphoric blue color and the Mg precipitated was ob-
tained35 in the range of 0.2 to l .2 mgm of Mg, and accurate recoveries
were obtained of Mg added to plant ash solutions.
5-58. The 8-hydroxy quinoline procedure is also a thoroughly satisfac-
tory semimicro colorimetric procedure for Mg, suitable for the range of
0.1 to 1.5 mgm of Mg as this procedure is detailed below. It requires a
somewhat higher manipulative skill than the titration procedure given
above. Other alternative procedures for Mg include the Versene titration
( ~ 4-21 and 11-54) and the ft a me emission spectrophotometer procedure
(~ 18-7).
5-59. Semimicro 8-hydroxy Quinoline Procedure for Mg.:m Needed ap-
paratus includes 60-ml pointed centrifuge tubes (Fig. 1 1-4) and centrifuge,
an air-jet stirrer, a hot water bath and cover glass, a 50-ml pipet prefer-
ably of the 3-stopcock (Lowy) type and smaller pi pets, an aspirator de-
canting device (Fig. 11-3), and an absorption spectrophotometer with a
660 mu light maximum.
5-60. Needed reagents are NH 4 Cl, A.R.; HOAc, glacial, C.P.; concen-
trated and 4 N NH 4 0H, and the following special reagents. For a three per
cent 8-hydroxy quinoline solution, approximately 3 gm of 8-hydroxy
quinoline is dissolved in 100 ml of 95 per cent ethanol. This reagent is
freshly prepared before each use. For the ammonium hydroxide-ethanol
wash solution, a water solution containing 10 per cent by volume of con-
centrated NHPH is mixed with an equal volume of 95 per cent ethanol.
For the ferric chloride-acetic acid reagent solution, a 25-gm portion of
FeC1:1 • 7 H 2 0 is dissolved in water, 25 ml of glacial HOAc is added, and
the solution is diluted to 1 liter. For the standard Mg solution, a 1.521-gm
portion of MgS0 4 • 7 H 2 0 is dissolved in water and diluted to approxi-
mately 900 ml. The Mg concentration in this solution is then determined
gravimetrically as the hydroxy quinolate, or as the pyrophosphate (~
5-56), and the volume is adjusted so that l ml of the solution contains
0.150 mgm of Mg. To obtain the standard Mg curve, aliquots of this solu-
tion (0.5, 1, 2, 4, 6, 8, 10, and 12 ml) are placed into 60-ml pointed
:l4 Reitemeier, Ind. Eng. Chem., A.E., 15:393 (1943).
ar. In the author's laboratory , by Dr. B. Chatterjee and J. L. Huber ( 1947).
3GThe details of the procedure given were adapted by Dr. R. B. Corey, working
in the author's laboratory ; previous reports on the method include Sideris, Ind. Eng.
Chem., A.E., 12:232 (1940); Gerber et al., Ind. Eng. Chem., A.E., 14:658 (1942);
in
Weeks and Todd, Ind. Eng. Chem., A.E., 15:297 (1943); and Dr. A. E. Peterson
this laboratory .
EXCHANGEABLE METALLIC CATION DETERMINATIONS 101
centrifuge tubes and diluted to approximately 40 ml. Then 1 gm of
NH 4 Cl is added and precipitation of the Mg and preparation and colori-
metric reading of the test solutions are carried out as described in the
procedure.
5-61. The Mg is determined directly on an aliquot of Solution A, with-
out prior removal of Ca in nearly all cases; Ca does not interfere except
when very unusually high in amount relative to Mg. In case of a shortage
of Solution A, the Mg may be determined on the filtrate from the Ca de-
termination as the oxalate, but the excess ammonium salts must first be
removed. To do this, 3 gm of NaOH pellets are added to the filtrate in a
conical flask and the solution is boiled to expel NH 3 • The solution is then
made slightly acid while hot, by the addition of concentrated HCI drop-
wise (indicators are already present in the solution). The Mg is then de-
termined on this solution instead of Solution A in the following paragraph.
5-62. An aliquot of Solution A containing 0.1 to 1.5 mgm of Mg is
placed in a graduated 60-ml pointed centrifuge tube (Fig. 11-4) and
diluted to a volume of 40 ml. Approximately 1 gm of NH 4 CI is dissolved
in this solution with agitation by the air-jet stirrer and then the tube is
placed in a hot water bath for 5 minutes. Three drops of brom cresol
purple indicator are added to the hot solution. Next, 4 N NH4 0H is added
dropwise while the solution is agitated with the air-jet stirrer, until the indi-
cator just turns purple. The tube is again placed in the hot water bath for 5
minutes, then cooled and diluted to exactly 60 ml. The solution is mixed
and centrifuged for 5 minutes at 2000 rpm to sediment the hydroxy oxides
of Fe, Al, and Ti.
5-63. An aliquot of the supernatant solution (or the standard Mg solu-
tion), containing from 0.1 to 1.5 mgm of Mg, is transferred to another
pointed centrifuge tube. For recovery of a maximum aliquot, a 50-ml Lowy
pipet is best. To this solution are added, with constant agitation with the
air-jet stirrer, 2 ml of concentrated NH 4 0H and 1 ml of 3 per cent
8-hydroxy quinoline solution. The tube is placed in a water bath that is
then covered with a watch glass so any precipitate that may rise on top of
the solution will not dry on the sides of the tube. The water bath is heated
to boiling, and then the tube is placed in a cool water bath for 2 hours,
after which it is centrifuged at 2000 rpm for 5 minutes. The supernatant
liquid is decanted with the device shown in Fig. 11-3. The precipitate is
washed twice, each washing consisting of the addition of approximately
40 ml of the NH 4 0H-ethanol washing solution, agitation with the air-jet
stirrer, centrifugation, and decantation. After the last decantation, exactly
2.5 ml of glacial HOAc is added to the precipitate in the bottom of the
tube, and the suspension is thoroughly mixed with the air-jet stirrer. When
the precipitate has dissolved completely, the solution is diluted in the tube
102 EXCH ANGE ABLE META LLIC CATIO N DETER MINA TIONS
n B,
to exactly 50 ml and mixed well with the air-jet stirrer. This is Solutio
used for the determination of magnesium.
,
5-64. To a 50-ml volumetric flask is added l 0 ml of FeC13 . solution
which is then diluted to approximately 30 ml with distilled water. Exactly
the
10 ml of Solution B containing the Mg is added. (It is important that
solution has
aliquot size not vary, as the concentration of HO Ac in the final
to the
an appreciable effect on the color intensity.) The solution is diluted
tube.
mark, mixed thoroughly, and a portion is transferred to a colorimeter
transm ission is read with a 660 mu light maxim um,
The percentage of light
Mg is obtaine d from the standar d curve. The color
and the concentration of
is stable over night.
MANGANESE DETERMINA TION37
(Colorimetric N a-paraperiodate oxidation in H 11P0 4 solution)
us
5-65. Manganese determination is needed in connection with numero
determ inative pro-
kinds of soil chemical analysis, but fortunately a simple
ap-
cedure is adequate for the many applications. A list of representative
Back-
plications is given at the beginning of the procedure (~ 5-73).
ground for absorption colorimetry is supplied in Chapter 17.
APPARATUS
a
5-66. Needed apparatus includes a 150-ml beaker and cover glass,
gas burner, and
glass funnel, a 100-ml volumetric flask, a steam plate, a
an absorption spectrophotometer with 540 mu light maximum.
REAGENTS
cent
5-67. Needed reagents are 30 per cent H 20 2 (Mn-free), 85 per
H;iP0 4 , concentrated HN011 , and the following special solutio ns:
al Co.,
1. Na 3 H 210 6 , Trisodi um paraper iodate (G. F. Smith Chemic
Columb us, 0.).
in a
2. Purified water diluent. To l liter of distilled or redistilled water
and 1 gm
2-liter Florenc e flask are added 100 ml of 85 per cent H 3 P0 4
plate
Na3 H 210u. The mixture is heated to boiling and digested on the steam
for I hour. It is then stopper ed with a foil-cov ered stopper .
mately
3. Standard KMn0 4 solution. A stock solution of approxi
is prepare d by dissolvi ng 31.6 gm of KMn0 4 crystals in I
1 N KMn0 4
in an amber bottle. It must be purified
liter of distilled water. It is stored
traces of
because of the develop ment of traces of Mn0 2 as a result of
er, etc., and through autooxi da-
reducing substances in the solvent, contain
tion of Mn++ impurit y by Mn0 4 - accordi ng to the equatio n:
3 Mn++ + 2 Mn0 4 - + 2 H 2 0 :=;: 5 Mn0 2 + 4H+ (5-4)
(precip itate)
87 Based on the classical method of Willard and
Greathouse, J. Am. Chem. Soc.,
39:2366 (1917), as adapted to photoele ctric colorim etry by Mehlig, Ind. Eng. Chern.,
A.E., 11 :274 ( 1939). Published improve ments and modifica tions are cited individ-
ually at appropriate places in the procedure.
 EXCHANGEABLE METALLIC CATION DETERMINATIONS 103
in which MnO~ is the generalized form of the precipitate. It may involve
various amounts of water and mixed valences of Mn. This reaction also
occurs in a KMn0 4 titration if insufficient excess of acid has been supplied,
when considerable Mn++ has been formed as the product of the reaction
and an excess of KMn0 4 is present. Purification may be hastened by boil-
ing and filtration; IQng standing for the reaction and sediment.ation of the
precipitate at room temperature is also effective.
5-68. The standard solution of 0.05 N KMn0 4 (N as oxidation-reduc-
tion) is prepared by dilution of 50 ml of the purified I N stock solution
to 1 liter. This solution is stored in an amber bottle fitted with a 2-hole
stopper bearing a delivery tube bent to deliver into a buret and an air
inlet tube through which air is forced to deliver the solution needed. This
solution is standardized by titration of primary standard Na~C~Oi. The
titration is conveniently carried out by pipetting 20 ml of 0.0500 N
Na~CP 4 solution in 2 N H~S0 4 (~ 5-27) into a 150-ml beaker, heating
the solution to 70°C, and titration of the 0.05 N KMn0 4 to the appear-
ance of a permanent faint pink color. A trace of MnCI~ may be added as
an accelerator to initiate the reaction. The exact normality of the KMn0.1
solution is calculated and the solution is labeled, a procedure that is prefer-
able to attempting a dilution to exactly 0.0500 N KMn0 4 •
5-69. A second dilution of 50.0 ml of the 0.05 N KMn0 4 to 500 ml in
a volumetric flask gives a 0.005 N KMn0 4 solution (N as oxidation-
reduction). The final colorimetric standard solution is prepared by a
further I : 20 duution, to 0.00025 N KMn0 1 as oxidation-reduction,
equivalent to 0.0001 N as Mn-f f (2.75 mgm Mn per liter). To make this
dilution to a permanently colored solution, 50 ml of the 0.005 N stock is
pipetted into a 1.5-liter beaker, then 100 ml of water, 100 ml of 85 per
cent RiP0 4 , and 2.0 gm of sodium paraperiodate arc added. The solution
is heated just to boiling, 725 ml additional water is added, and the solution
is digested on the steam plate for 0.5 hour. The hot solution is transferred
. to a I-liter volumetric flask, diluted to volume, and mixed. When cooled
to room temperature, the volume is adjusted exactly to the mark with the
purified water diluent, again mixed, and stored in a cleaned amber bottle
or in the flask in the dark. The color is stable for long periods, because an
excess of periodate is. present. Portions of the solution are read in the
colorimeter as needed with a 540 mu light maximum.ak Beer's law applies.
Any light maximum from 525 to 545 mu is satisfactory if used for both
standard and test sample; a 520-mu maximum has been employed but its
± 15 mm range encompasses much of the rapid rise in percentage trans-
mission below 522 mu.
5-70. As a test of the analytical technique for developing the HMn0 4

38 Mehlig, Ind. Eng. Chem. A .E., 11 :274 (1939); verified and refined by Cooper,
Anal. Chem., 25:411 (1953).
 ATIO NS
104 EXCH ANGE ABLE META LLIC CATI ON DETE RMIN
ts (2, 4,
color that is advisable for the beginning analyst a series of aliquo
and mixed
6, 8, 10, 12, and 14 ml) is taken of the freshly prepared
(55 ppm of
0.005 N (oxidation-reduction) KMn0 4 standard solution
s. These
Mn++ ). The aliquots are evaporated to dryness in 150-ml beaker
( ~ 5-71) . The percen tage transm ission
are treated as in the procedure
scale against concen tration on a linear
readings are plotted on a semilog
, both on linear scales ,~ 17-16 ).
scale (or L-values against concentration
derive d with the perma nent
A straight line should pass through the point
us para-
HMn0 4 color standard (2.75 ppm Mn++ ) prepared in the previo
graph.

PROCEDURE
ns be-
5-71. Preparation of the Mn Test Solution. Various test solutio
may be analyz ed by the manga nese
sides that carrying exchangeable Mn++
determ ination of exchan geable
determination now to be described. The
n of the
Mn+ + is carried out either on the solution resulting after titratio
t of Solutio n A (if
MgNH 1P0 4 and MnNH 4 P0/9 or on a separate aliquo
glass is emplo yed to
enough is available). If filter paper instead of fritted
g the solution
filter that precipitate, the filter paper is removed by passin
through a second filter paper and thorough washing of the filter.
analyzed,
5-72. An appropriate aliquot is taken of the test solution to be
test solution
to give 0.1 to 0.8 mgm (0.004 to 0.03 meq) of Mn++ in the
Increased
to be made up to a 100-ml final volume with the HMn0 4 color.
a use of a smalle r final volum e with the
sensitivity can be obtained by
ses in the volum e of each solven t being
HMn0 4 color, proportional decrea
added in the development of the HMn0 4 color( ~ 5-76) .
l final vol-
5-73. Some examples of proper aliquot fractions for a 100-m
concen tration s and sample
ume of HMn0 4 color and for assumed typical
sizes follow:
sample(~ 5-53):
I. For exchangeable Mn-t +, 5 ppm in soil from 25-gm
1.0 aliquot.
with 1-gm
2. For Mn++ exchange capacity of soil, 10 meq/l 00 gm,
sample (~I 4-29): 0.2 aliquot .
with 0.5-gm
3. For Mn++ exchange capacity of clay, 50 meq/1 00 gm,
sample ( ~ 4-29) : 0.1 aliquot .
1-gm ~mple
4. For total Mn++ in soil, 0.05 per cent MnO from
(~ 11-173 ): 0.5 aliquot .
5. For total Mn++ in plant tissue following wet-oxidation
, 50 ppm Mn
from 5-gm sample ( ~ 12-39 ): 1.0 aliquot .
to dryness,
5-74. The aliquot from any of these sources is evaporated
in a larger beaker if the origina l volume
preferably in a 150-ml beaker, but
Eng. Chem., A.E.,
ao Joint precipitation with magnesium after Dean and Truog, Ind.
7:383 (1935) .
EXCHANGEABLE METALLIC CATION DETERMINATION S 105

of solution to be evaporated is too large. To the residue, 3 ml of concen-

trated HN0 3 and 2 ml of 30 per cent H 20 2 are added; the beaker is cov-
ered and digested on the steam plate for 30 minutes. The sides of the
beaker are washed down with a fine stream of water from a wash bottle.
The solution is transferred to a 150-ml beaker if that size was not used for
the original evaporation. It is important that this transfer not be made be-
fore the HNO:i and H 2 0 2 treatment. The cover glass on the beaker is next
supported on glass hooks and the solution is evaporated to dryness to
complete the oxidation of the organic matter and the removal of the H 2 0 2 •
5-75. Treatment with HN0:1 and H 2 0 2 has been found to be highly
efficacious in insuring complete removal of reducing substances without
ignition in H 2 S04 • Some of the Mn tends to become insoluble as Mn02 in
HNO:i alone, but not in the presence of both HNOa and H 2 0 2 • lnterference
by chloride, bromide, iodide, thiocyanate, arsenite, and organic anions
such as oxalate is effectively avoided by this treatment.
5-76. Development of HMn0 4 Color. Ten ml of 85 per cent HaP0 4 is
added 40 to the beaker, a cover glass is placed on it, and the solution is
warmed over a burner to boiling, and then cooled to about 50°C to avoid
spattering when diluted with water. The cooled acid is diluted with 10 ml
of water, and the solution is mixed by rotating the beaker. Approximately
0.2 gm of sodium paraperiodate is added, the beaker is covered, and the
solution is heated on the steam plate until a purple color appears; then 75
ml of the purified water diluent is added and heating is continued for 40
minutes or until the purple color no longer increases. Approximately 0.1
gm additional periodate is added near the end of the period of digestion.
Equations:
5 ]7+ + 2 Mn+ f - 515+ + 2 Mn 7 + (5-15)
or, with trisodium paraperiodate,
15 NaaH 2 I0 11 + 2 Mn 3 (P0 4 ) 2 + 6 H:iP0 4 - 15 Nal0 3 + 6 HMnO,
,,,,, + 10 Na:{P0 4 + 6 H~O
(5-16)
while, with potassium metaperiodate,
15 KI0 4 + 2 Mna(P04 ) 2 + 9 H 20 - 15 KIOa + 6 HMn04 + 4 HaP0 1
(5-17)

40 Use of 85 per cent HaP0 4 is after Willard and Greathouse, J. Am. Chem. Soc.
39:2366 (1917), and Peech, Ind. En[?. Chem., A.E., 13:436 (1941 ). This acid has
the advantage that the HMn04 color development is Jess sensitive to H: 1P04 con-
centration than to that of H2S04, and that Fe++• is soluble and decolorized in it. The
H3 P0 4 also has the advantage over the H 2 S04 of nonprecipitation of Ca as CaS0 4
in case manganese is being run without prior removal of calcium, as pointed out by
Sherman et al., Soil Sci., 54:253 (1942).
106 EXCH ANGE ABLE META LLIC CATIO N DETER MINA TIONS
.
The paraperiodate appears to be a superior reagent to the metaperiodate
), and
The reason may lie in the requirement of acid (in an acid-rich system
the nonrequirement of water (from a water-p oor system ) in the former
(eq. 5-16).
5-77. The hot solution is transferred to a 100-ml volumetric flask
with
the purified water diluent. The flask is closed and the solutio n is mixed,
with
allowed to come to room temperature, and diluted to exactly J00 ml
is compar ed with the
the purified water diluent. A portion of this solution
colorim eter with
standard HMn04 solution, by means of a photoelectric
com-
540-mu light ( 525 to 545 mu, as explained for the standa rd). Visual
parison in Nessler tubes is also possible.
5-78. Calculation of Results. The HMn0 4 color obeys Beer's law with
n, is
540-mu light. The normality of Mn (as divalent) in the test solutio
(~ 17-35) . The exchan ge-
calculated from the percentage transmission
able manganese is calculated to meq Mn+ + per 100 gm soil:

meq Mn+ + per 100 gm of soil = N x 100 ml x 1OO

s
(5-18)

when N represents normality in the 100 ml of test sample, and s

is gm
sample weight represented in the aliquot. Also parts of Mn 1- + per 2 million
parts soil (pounds of Mn+ + per acre) is calculated:
pp2mM n++ = meqM n++ per lOOgmsoil x 549.3 (5-19)
ed as
In the analysis for total manganese of soil, the result is usually express
"% MnO":
70.93
% MnO = meq Mn++ per 100 gm x - - - - (5-20)
2000
= meq Mn++ per 100 gm x 0.03547
of
5-79. The exchangeable manganese is calculated to "equivalent ml"
Mn titratio n value.
standard H 2S04 and this is subtracted from the Mg plus
Thus the magnesium is obtained by difference.

POTASSIUM .DETERMINATION
be
5-80. The exchangeable K obtained in the NH4 0Ac extraction may
cobalti nitrite proced ure. To do this,
determined readily by means of the
the
an aliquot of Solution A that contains 0.5 to 6 mgm of K is taken for
a rapid determ ina-
determination of potassium (~ 6-20). Alternatively,
-
tion of exchangeable potassium can be made by flame emission spectro
photometry, either directly on Solution A, or through a rapid direct extrac-
tion of soil for Kand Na(~ 18-24) .
EXCHANGEABLE METALLIC CATION DETERMINATION S 107

SODIUM DETERMINATION41
(Precipitation as Mg uranyl acetate triple salt)
5-81. The exchangeable Na obtained in the NH4 0Ac extraction may
be determined readily by means of the magnesium uranyl acetate procedure
to be described, or alternatively, by flame emission spectrophotometry di-
rectly on Solution A or through a rapid direct extraction of soil for K and
Na(~ 18-24).
5-82. The triple salt formed by Na precipitation with magnesium and
uranyl acetates has the formula NaMg(U0 2 ) 3 (0Ac) 9 • 8 H 2 0. Forma-
tion of such a triple salt with other divalent cations such as Ni, Co, or
Zn is known, but decreasing sensitivity has been shown 42 with increasing
radius in the order listed; Mg forms the most insoluble of the triple salt
precipitates. The precipitate is determined by titration with standard NaOH
or alternatively by a gravimetric procedure.

APPARATUS

5-83. Needed apparatus includes 250-ml and 50-ml beakers, a 10-ml

pipet, a filter (Gooch type with sintered glass, asbestos, or porous porce-
lain filtration substance), and a glass dropper with a rubber bulb. For the
gravimetric procedure, a rubber policeman is needed.

REAGENTS

5-84. Needed reagents are standard 0.1 N NaOH and 0.1 N H 2 S04 , and
the following special reagents.
5-85. Magnesium Uranyl Acetate Reagent. Approximately 32 gm of
uranyl acetate, U0 2 (0Ac) 2 • 2 H:P, and 100 gm of Mg(0Ac) 2 • 4 H 2 0
are dissolved in 200 ml of water, the mixture being warmed if necessary to
obtain solution. The solution is cooled, and to it are added 20 ml of glacial
HOAc and 475 ml of 95 per cent ethanol. The solution is diluted to 1 liter
and mixed well. This solution is allowed to stand in a dark place for 48
hours and then is filtered into a Pyrex bottle. The filtered solution is stored
in a dark place.
5-86. Washing Solution. A few gm of precipitated sodium magnesium
uranyl acetate, made by precipitation according to the procedure, are placed
in a I-liter Pyrex bottle and then approximately 800 ml of 95 per cent
ethanol is added and the mixture is shaken intermittently for several hours.
Finally the suspension is allowed to settle for 24 hours. The almost clear
supernatant liquid is then filtered through a Whatman No. 44 filter and the
41 Precipitation method of Caley and Foulk, I. Am. Chem. Soc., 51:1664 (1929),
52:1349 and 4247 (1930); Piper,/. Agr. Sci., 22:676 (1932); after Peech et al.,
U.S.D.A. Cir. 757, p. 16 (1947). Titration in accordance with Dobbins and Byrd,
J. Am. Chem. Soc., 53:3288 (1931).
42 Rogers, Sci., 103: 420 ( 1946).
108 EXCHANG EABLE METALLIC CATION DETERMIN ATIONS
filtrate is transferred to a second Pyrex bottle with more of the crystalline
salt. The washing solution is prepared by filtrating this second stock supply
of alcohol saturated with the sodium salt just prior to use. Fresh prepara-
tions of ethanol may be made by addition to the first bottle, the solution be-
ing shaken and allowed to stand for 24 hours to precipitate organic sub-
stances prior to transfer to the second bottle, according to Piper. 43
5-87. Na Standards. Because conditions of precipitation may result in
some disparity from the theoretical composition of the precipitate, it is ad-
visable to take standard amounts of Na as NaCl in the range from 1 to 5
mgm to determine the titer or weight in relation to the quantity of Na taken.
This provides a calibration curve that is a reliable measure of the precipitate
composition under the conditions employed in the determination.
PROCEDURE

5-88. An aliquot of Solution A, containing 1 to 5 mgm of Na is placed

in a 50-ml (well weathered) Pyrex beaker and evaporated to dryness. The
residue is di~<.olvcd in 6 ml of water. It is not necessary to filter this solu-
tion if it happens to be slightly turbid either for volumetric or gravimetric
determination. Next, 15 ml of magnesium uranyl acetate reagent is added
and the solution is stirred for about 15 seconds, after which precipitate
formation normally has been started. The beaker is covered and allowed to
stand for 1 hour but for no longer than 2 hours.
5-89. The solution is filtered through a Gooch type crucible (sintered
glass, asbestos, or porous porcelain), and the beaker, filter, and precipitate
are washed once with 2 ml of magnesium uranyl acetate reagent with the
aid of a dropper and then 5 times with the 95 per cent ethanol saturated
with the triple salt.
5-90. Titrimetric Procedure for Na. The precipitate in the original
beaker is dissolved in a little water, and the solution is transferred to a
250-ml beaker. The crucible carrying the main mass of precipitate is placed
in the 250-ml beaker and water is added to a total volume of 100 ml. Then
5 drops of I per cent phenolphthalein indicator is added and an excess of
0.1 N NaOH is added as indicated by the formation of a persistent pink
color. The solution is heated to boiling, with further addition of NaOH if
the pink color begins to fade. When the solution just reaches boiling, it is
removed from the heat source and back titrated with 0.1 N H 2 S04 just to
the disappearance of the pink color.
Equation:
NaMg(U0 2 ) 3 (0Ac) 9 • 8 Hp+ 10 NaOH ~ 9 NaOAc + Na 2U 20 7
+ MgU04 + 13 H 20
(5-21)
43 Piper,/. Agr. Sci., 22:676 (1932).
EXCHANGEABLE METALLIC CATION DETERMINATIONS I 09
Each ml of 0.1 N NaOH is equivalent to 0.00023 gm of Na in the precipi-
tate. Ten times as much NaOH is consumed as the chemical equivalent of
the Na in the precipitate.

ALTERNATIVE PROCEDURES

5-91. Gravimetric Evaluation of Triple Salt Na Precipitate. The precipi-

tate is filtered on a sintered glass, asbestos, or porcelain Gooch crucible
filter. The beaker, crucible, and precipitate are washed with a sufficient
number of 2-ml portions of the magnesium uranyl acetate reagent ( deliv-
ered from a glass pipet) to complete the transfer of the precipitate en-
tirely to the crucible. Usually 3 to 5 washings are required. The last of the
transfer is aided by a little washing solution delivered from an extremely
fine jet wash bottle. Then the filter is washed 5 times with 2-ml portions of
95 per cent ethanol saturated with sodium magnesium uranyl acetate. Fi-
nally the filter is washed twice with small portions of anhydrous ether or
acetone. Air is drawn through the crucible for a few minutes to volatilize
the organic solvent. The crucible is then allowed to stand 10 to I 5 minutes
until the odor of the organic solvent is completely gone and then is weighed.
The difference in weight corresponds to the triple salt of sodium. A blank,
carried out on all reagents just as in the determination, is subtracted from
the above difference in weight.
weight of Na in gm= net weight of precipitate in gm x 0.01495
(5-22)
5-92. Colorimetric Evaluation of Na Triple Salt Precipitate. The evalua-
tion of a zinc uranyl acetate precipitate of Na can be done by the develop-
ment of color with sulfosalicylic acid. 4 "1 Amounts of Na in the range of
0.010 to 0.360 mgm can be determined by means of a sensitive centrifuge
washing procedure.

QUESTIONS
I . List the exchangeable cations of soils that can be determined in the
NH 4 0Ac extract?
2. Outline some of the factors that may raise the amount of metallic ca-
tions present in this extract over the quantities that are truly exchangeable in
the soil.
3. What is the equivalent in terms of pp2m of one meq of each of the fol-
lowing metallic cations: Ca, Mg, Mn, K, and Na?
4. Why is it advantageous to employ a field-moist soil sample for exchange-
able cations such as K +, NH 4 +,Mn++, and Fe++?
5. What is the purpose of allowing the sample to equilibrate with the
NH 4 0Ac for a considerable time, as long as overnight, prior to filtration?

44 Parks et al., Ind. Eng. Chem., A.£., 15: 528 ( 1943).

110 EXCHANGEABLE METALLIC CATION DETERMINATIONS
the
6. What is the purpose of having the beaker tightly covered during
H 2 0 2 and HNOa treatmen t of the NH 4 0Ac residue?
Mg
7. What principle is employed in the prevention of precipita tion of
with Ca as the oxalate?
8. What are the advantages of cerate over permang anate for the oxalate
titration?
9. List the advantages and disadvantages of the MgNH 4 P0 4 method for Mg?
10. State the advantages of the fritted glass filter over filter paper for recov-
be
ery of the MgNH 4 P0 4 precipitate. Why must the last traces of free NH:i
removed prior to titration?
de-
11. Why must the Fe and several other ions be removed prior to the
y quinolin e precipita tion? Why may
terminat ion of Mg by means of 8-hydrox
ed
moderat e amounts of Ca be present in the solution on which Mg is determin
in this way?
de-
12. Why must the last traces of organic matter be removed prior to the
terminat ion of Mn as the H MnO 4 purple color?
sys-
13. Why is the determin ation of K often considered separately from the
tematic determin ation of all of the exchangeable metallic cations?
of
14. What are the chief precautio ns to he followed in the determinations
Na, whether by precipita tion or flame emission spectrop hotomet ry?
 6

Potassium Determinations
for Soils
. exchange, fixation and release
-A THEME OF SOIL CHEMICAL RESEARCH 1

6-1. The determination of potassium has been the subject of so much

investigation that semimicro determination by gravimetric, titrimetric, and
absorption and emission (~I 18-7) spectrophotometric methods can now
be readily accomplished. The absorption spectrophotometric determination
of potassium is emphasized in this chapter, although titrimetric and gravi-
metric procedures are also given.
6-2. Potassium Is a Key Element in Soil Chemistry. Because the amount
of exchangeable potassium is critically low for crop production in exten-
sive soils throughout the more humid regions, this element is of key im-
portance in soil chemistry and soil fertility. 2 Its equilibrium reactions in
soils are somewhat more complex than can be predicted from the elemen-
tary principles of ionic exchange as exemplified by sodium and calcium
(Chapter 5). The analytical determination of potassium in solution, con-
sidered first, is carefully distinguished from the extraction of exchangeable
(~ 6-67) and other(~ 6-76) forms from soils.
6-3. Methods for Potassium Determination. Despite the prominence of

1 Reitemeier, U.S.D.A. Tech. Bui. 1049 (1951).

2 Cowie, Potash (London: Edward Arnold & Co., 1951); Eckstein et al., Potash
Deficiency Symptoms, 2nd ed. (Berlin: Verlagsgesellschaft fiir Ackerbau, 1937; Tur-
rentine, Potash in North America (New York: Reinhold Publishing Corporation,
1943).
111
112 POTASSIUM DETERMINATIONS FOR SOILS
the emission method for K ( ~ 18-7), several types of simple and inexpen-
sive potassium procedures are still important:
I. Precipitation in water as cobaltinitrite, (K,,NaL ,,)Co(N02)u · x H 2 0.
The term (K 11 Naa_ 11 ) refers to an isomorphous composition r~nge in the
crystals, in which K 1 •11 to K 1 • 7 is the most stable and reproducible compo-
sition ( ~ 6--7), and x stands for variable water of crystallization. The
method requires the prior removal of ammonium and barium ions, use of
a water medium of precipitation, and KC! standards for calibration. De-
termination is effected gravimetrically, titrimetrically for nitrite, or colori-
metrically for nitrite or cobalt.
2. Precipitation as chloroplatinate, K~PtCI,;. This method employs dif-
ferential solubility in alcohol for separations and requires prior removal of
ammonium. Procedures are gravimetric, or colorimetric. 3
3. Precipitation as K-dipicrylamine, K-hexanitrodiphenyla mine ( ~ 13-
48). This method requires the prior removal of ammonium and double
precipitation to remove Na interference. The procedure is colorimetric. 4
4. Precipitation as perchlorate, KCI0.1• This method employs differen-
tial solubility in alcohol for separation and is often used following a pre-
liminary separation as cobaltinitrite. It requires removal of ammonium and
·sulfate. The procedure is gravimetric.
5. Gravimetric determination of combined NaCl plus KC!, separate de-
termination of chloride, and computation of K or Na by simultaneous
equations ( ~ I 0-84).
6-4. Cobaltinitrite Method. Of the 5 methods, only the cobaltinitrite
procedure is given in detail herein. Of the precipitation methods, the
cobaltinitrite procedure is preferred for soils work because it is rapid, eco-
nomical, adaptable to semimicro technique, and the low solubility (~ 6-5)
of the precipitate in water makes for flexibility. The only faster and more
flexible method is the emission spectrophotometric method with the flame
photometer (~ 18-7). Because the cobaltinitrite of potassium (with so-
dium) is enormously more insoluble in water than other potassium pre-
cipitates, the cobaltinitrite has been employed5 for preconcentration of K
from salt solutions of such composition that the direct determination of K
is not feasible.
6-5. The cobaltinitrites of potassium and sodium present a striking dif-
ferential in the solubility in water. Were it not for the isomorphism (~ 6-
7) of Na in the K salt, the cobaltinitrite procedure would be ideal. None-
theless, it is an effective procedure. In considering variability in the pre-
cipitate composition, in relation to other methods, one must recall that

s Adams and St. John, Ind. Eng. Chem., A.E., 17:435 ( 1945).
4Kolthoff and Bendix, Ind. Eng. Chem., A.E., 11:94 (1939); Amdur, Ind. Eng.
Chem., A.E., 12:731 (1940); Lawton, S.S.S.A. Proc., 10: 126 (1946).
5 Washington, Chemical Analysis of Rocks (New York: John Wiley & Sons, Inc.,
1919); Piper, Soil and Plant Analysis (New York: lnterscience Publishers, Inc.,
1944); Wright, Soil Analysis, 2nd ed. (London: Thomas Murby &Co., 1939).
 POTASSIUM DETERMINATIONS FOR SOILS 113
selection of solvents and procedures for the K2 PtC1 6 and KCIO 4 deter-
minations are also empirically established. These must be calibrated against
quantities of K taken in standardization and give an over-all accuracy in
the same range as with the best cobaltinitrite procedures. Good concord-
ance of the chloroplatinate and cobaltinitrite procedures was found in a
careful experimental comparison. 6 The cobaltinitrite technique has an enor-
mous advantage for sensitivity to small amounts of K.
6-6. Cobaltinitrite Ion. Nitrite reacts with cobaltous ion to form a
cobaltic complex Co(N02 ) 6 - - - , the oxidation of Co++ to Co+++ be-
ing accomplished by the reduction of 1 nitrite ion. Sodium is commonly
used as the cation of this complex salt for a potassium precipitating re-
agent, but silver 7 and zinc 8 have been employed. The complex soluble salt
may be purified by recrystallization from organic solvents, and is avail-
able as a bright orange salt from chemical supply houses. In many pro-
cedures11 it is formed during the preparation of the precipitation reagent,
by mixing solutions of the nitrite and cobaltous salts and allowing time for
the reaction to be completed. The necessity of precipitation of the potas-
sium in the presence of the large excess of sodium salt (by-product) is not
a disadvantage to the method; many procedures which employ the purified
salt call for the addition of more sodium salts.
6-7. Isomorphism. Extensive studies have demonstrated the fact of iso-
morphism of Na and K in the cobaltinitrite precipitate crystal, between the
n-value of 3 and 1.5 in the formula (~ 6-3, I). The upper limit of Na con-
tent appears to be approximately at a l : l ratio to K, as would be ex-
pected from the extreme solubility of the sodium cobaltinitrite salt. One
Na can proxy for 1 K so long as there is a K ion immediately adjacent to
cover it in the crystal. The chief factors determining the composition of the
precipitate are (a) the concentration of sodium ion, including that from the
precipitation reagent, and its general magnitude in relation to that of potas-
sium (Fig. 6-1), and (b) the temperature of the solution from which it is
precipitated. Coincidence will be noted in Fig. 6-1 of curves representing
10 and 2 mgm of K precipitated under the same conditions by the author
and N. J. Volk, respectively, working independently. Yolk's 2 curves, ob-
tained with room temperature precipitation, illustrate the effect of a large
amount (5 per cent) of Na in the system in addition to that in the precipi-
tation reagent. Cooling the system to 3°C lowered the author's curve about
as much as Yolk's addition of 5 per cent extra Na at room temperature.

n Hibbard and Stout,J.A.O.A.C., 16:137 (1933).

7 Breth and Gaebler, J. Biol. Chem .. 87:81 (1930).
R Adams et al., Ind. Eng. Chem., A.E., 7:310 (1935).
9 An early procedure was described by Adie and Wood, J. Chem. Soc., 77: 1076
(1900).
114 POTASSIUM DETERMINATIONS FOR SOILS
20.0
19.2
18.4
17.6
16.8
16.0
15.2

c
14.4 K added {o---o Jackson (10 mgm K)
0 to reagent •---.. Volk (2 mgm K)
ti:::> 13.6
~ 12.8
Reagent {x----+<
Jackson (10 mgm K)
added to K - - - Volk (2 mgmK)
C:
.
2 12.0
~ 11.2
T
I Relation if precipitate were K 3Co ( N0 2) 8.x H 20
0
i 10.4 I
E I
9.6
8. 8.8
yAt room temperature
~

. I •

E 8.0
\ '-._
::!: 7.2 \, \ Relation if precipitate were K2NaCo< N0 2 )a.x H20

6.4
5.6
4.8
4.0 Relation if precipitate were Na2KCo(N0 2 )6 .xHgO
3.2
2.40
1.60
rdoption
0.80
0.0 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
% Na 3Co (N0 2)8 in final precipitation solution

Fig. 6-1. Relation of oxidation-reduction reaction of K

precipitate, (K.Na3-.)Co(N 02) 6 • xH 2 0, to concentration of
precipitation solution, at 3°C and with no extra sodium added
except as indicated. (Data from author's studies and Volk,
I. Am. Soc. Agron., 33:685, 1941.)

6-8. Lack of any inflexion of the curves for the hypothetical K2NaCo
crystal
(N0 2 ) 6 composition comes from the fact of isomorphism in the
and shows the fallacy of the
(established by X-ray diffraction studies 10 )
concen tration and other factors to
theory of adjusting the Na 8 Co(N0 2 6
11 )

10 Data obtained in detailed studies by the author (1943).

and many
11 Carried out by Kramer and Tisdall, /. Biol. Chem., 46:339 (1921) and
and Truog, /. Am. Soc. Agron., 36:537 (1934),
later workers, including Volk
Wilcox, Ind. Eng. Chem., A.E., 9: 136 (1937).
POTASSIUM DETERMINATION S FOR SOILS 115

approximate the K 2Na composition. The precipitate composition is sub-

ject to great variation in this range as shown by the steep rise in the com-
position curves with small changes in conditions as they cross the K2 NaCo
(N02 ) u composition line. The K3 Co (N0 2 ) 6 line lacks inflexion because
the solubility product of the precipitate leaves appreciable K in solution
with these low concentrations of Na+ and Co(N02 ) 6 - - - ions.
6-9. The pure K3 Co(N0 2 ) 6 • 1 · 5 H 2 0 salt can be prepared 12 with an
excess of K + and N0 2 - in the presence of low Co++ and Na+ concentra-
tions, as a method for determining cobalt.
6-10. Choice of Procedure. The adoption of the 10 per cent final con-
centration of precipitating reagent ( 20 per cent original concentration) is
based on the greater reproducibility 13 of precipitate composition, though
nonstoichiometric, than is obtainable under greater or lesser concentrations.
It greatly increases the sensitivity of the method with lower concentrations.
Addition of the K solution to the precipitating solution (rather than the
reverse) was adopted also for greater reproducibility. Adoption of 3 °C
precipitation is based on greater uniformity of composition over a wide
range of K concentrations. The I 0 per cent precipitating reagent concentra-
tion corresponds to 3.4 per cent Na in solution, which is sufficient to buffer
out considerable differences in the Na content of the test solution, and per-
mits greater reproducibility than higher Na concentrations.
6-11. Adding alcohol to the aqueous solvent increases the insolubility
of the precipitate and substitutes for low temperature in effect, but alcohol
is incompatible with formaldehyde in the precipitation system. The tem-
perature effect on the K content is linear from 5° to 50°C, being -0.5 per
cent per degree, but appears to be constant below 3 ° to 5 °C. The pH of
the supernatant liquid has no effect on the precipitate composition between
pH 3.2 and 5.6. Above this the Co begins to precipitate as the hydroxide;
below this the precipitate solubility increases rapidly. The cobaltinitrite
reagent brings the pH within this range even though the solvent for the K
may be considerably more acid in various procedures.

POTASSIUM DETERMINATION
(Precipitation as cobaltinitrite)
6-12. Determination of potassium is ordinarily quicker by the flame
emission spectrophotometry ( ~ 18-7) than by cobaltinitrite. In the ab-
sence of expensive emission equipment, cobaltinitrite is efficaceous for ex-

12 Kolthoff and Sandell, Textbook of Quantitative Inorganic Analysis (New York:

The Macmillan Company, 1938), p. 374.
13 Greater reproducibility led to the adoption of procedures, in which the K content
of the precipitate was lowered, by a number of workers, including Hibbard and
Stout, J.A.O.A.C., 16:137 (1933); Schueler and Thomas, Ind. Eng. Chem., A.E.,
5:163 (1933); Lohse, Ind. Eng. Chem., A.E., 7:272 (1935); Brown et al., Ind. Eng.
Chem., A.E., 10:652 (1938); and Volk, J. Am. Soc. Agron., 33:684 (1941).
116 POTASSIUM DETERMINATIONS FOR SOILS
changeable K ( ~ 5-14 or 6-67), K in soluble salts ( ~ 10-83), total K in
minerals (~ 6-75), and K in plant tissue (~ 12-39). The key principle
employed is the direct standardization in terms of standard KCl for the
conditions of the analysis. The existence of (K,Na) isomorphous relations
of the (K"Na3-n)Co(N0 2 ) 6 • xH 20 precipitate dictates the choice of the
most stable and reproducible conditions for its precipitation (~ 6-7) rather
than the conditions favoring a simulated stoichiometry. This principle is
widely applied in spectrochemistry; for example, the stoichiometry is not
known for the chromogen in the determination of phosphorus.

APPARATUS

6-13. Needed apparatus consists of beakers-150-ml for evaporation

and 50-ml for precipitation, g centrifuge and 25-ml pointed tubes, test
tubes or 50-ml conical flasks for cooling the K solution, a spot plate, a
policeman and glass rod, a 500-ml volumetric flask, 15 and 10-ml pipets
that deliver moderately rapidly, and a buret.

REAGENTS

6-14. Needed reagents include 10 per cent NaOH (low in K), am-
monium-free 1 N HCI (tested), Nessler's reagent ( ~ 8-43), phenolphtha-
lein indicator, and the following special reagents.
6-15. Precipitation Reagent. A 20 per cent solution of sodium co-
baltinitrite is prepared by dissolving 20 gm of reagent grade sodium co-
baltinitrite, Na 3 Co(N02 ) 11 (Mallinckrodt or Bakers) in 80 ml of cold
distilled water (5°C) and making to JOO ml volume at 5°C. After standing
24 to 48 hours, the solution is filtered through a retentive paper or cen-
trifuged in pointed centrifuge tubes to remove traces of insoluble matter in-
cluding potassium precipitate (filtration under reduced pressure is avoided
since a serious loss of nitrite results). The solution is tightly stoppered and
stored in the refrigerator at this temperature, and may safely be used for
1 to 3 weeks. It has been stored in tightly stopped full bottles without
change for 6 months.
6-16. Solvent for Potassium. The solvent for potassium is 0.17 N HOAc
containing 4 per cent formaldehyde, prepared by dilution of 10 ml of
glacial acetic acid and 100 ml of 40 per cent formaldehyde to 1 liter. The
formaldehyde prevents traces of ammonia from interfering through co-
precipitation with the potassium but does not eliminate the need for am-
monium removal in the procedure.
6-17. Standard KCI Solution (2 mgm K per 10 ml). Exactly 0.1905 gm
of dried KCI is dissolved in water. The solution is transferred to a 500-ml
volumetric flask, made to volume, and mixed. Each ml contains 0.2 mgm
ofK.
POT ASSlUM DETERMINATIONS FOR SOILS 117
6-18. Standardization. To obtain the standard curve and to establish the
technique, aliquots (3.75, 7.5, 15, 30, and 45 ml) of the.standard KCl
solution are measured out with a buret into separate 150-ml beakers. The
solutions are evaporated to dryness and each is carried, with blanks on
reagents, through the procedure (~ 6-20) and colorimetric evaluation
(~ 6-30).
6-19. In routine, 2 15-ml portions of the standard KC! solution are
evaporated (later giving 2 mgm K in 2/3 aliquot precipitated), and treated
as in the procedure ( ~ 6-20). These 2 standards are used to establish a
working calibration factor, which may vary slightly (but only slightly) ac-
cording to the amount of Na introduced in the various procedures.

PROCEDURE

6-20. Removal of Ammonium. The sides of the beaker in which the

potassium solution has evaporated arc washed down thoroughly to dissolve
any adhering ammonium. Two drops of l per cent phenolphthalein are
added and IO per cent NaOH then is added dropwise (usually 2 drops are
required) until the phenolphthalein turns distinctly red. The use of over
I ml of NaOH is avoided whenever possible, but the procedure accom-
modates more if necessary. The same quantities of NaOH are added to the
blanks and K standards. The solution is then evaporated to dryness on a
steam plate to drive off the last traces of ammonia. As evaporation pro-
ceeds, the solution color sh6uld remain pink. lf fading occurs, more phe-
nolphthalein and more drops of NaOH arc added as required. Ammonia
fumes must he kept away. When the evaporation is complete, 2 ml of am-
monium-free I N HCI is added and I small drop of this HCI solution of the
sample is added to a spot plate and tested for ammonium with Nessler's
reagent. H (If ammonia is present, rarely the case, the solution is again
evaporated to dryness, the above I 0 per cent Na OH treatment is repeated,
and the residue is taken up in 2 ml ammonium-free I N HCI as before.)
The HCI solution is dried out to extract of the K completely from the
siliceous residue and its conversion to KCI. The procedure is quickly con-
tinued until the solution is tightly stoppered.
6-21. Clarification of Solution. The beaker is cooled, and exactly 15 ml
of the solvent for potassium (0.17 N HOAc, containing 4 per cent formal-
dehyde) is added. The sides and bottom of the beaker are policed (using
no water), and the solution is immediately filtered through a dry filter paper
into a dry test tube or a 50-ml conical tlask, the latter then being stoppered
to prevent evaporation. Instead of filtration, this solution may be clarified by
centrifugation in a pointed 25-ml centrifuge tube just prior to precipitation.

14 No trace of the Nessler's solution, which is high in potassium, must be allowed

to come in contact with the test solutions.
118 POTASSIUM DETERMINA TIONS FOR SOILS
(A 7.5-ml volume of solvent for the potassium may be used, and 5 ml of
this solution may be used for precipitation.)
6-22. Potassium Cobaltinitrite Precipitation. The concentration of K in
the solution for analysis i!! kept in the range of 0.5 to 20 mgm K in 10
ml. The 15 ml of K solution in the tube or conical flask is cooled in an ice
bath to 3 °C (this will take about 20 minutes). Then 10 ml of the predpi-
tation reagent is pipetted into a dry 50-ml beaker and similarly cooled to
3°C in an ice bath. When both solutions have cooled to the proper tem-
perature, an aliquot of exactly 10 ml of the test solution is pipetted into the
beaker containing the precipitation reagent, the precipitation reagent being
stirred rapidly by rotation of the beaker during addition of the potassium
solution. The beaker is returned to the ice bath and kept at 4° to 6°C for
5 hours or preferably overnight.
6-23. Note. Various volumes of precipitation reagent and potassium
solvent may be used as long as the volume ratio remains 1 : I and the
potassium concentration does not exceed 2 mgm per ml. The K solution is
added to the precipitation reagent so that the concentration of sodium and
cobaltinitrite remain relatively high throughout the precipitation. The Na
concentration of the precipitation reagent is adequately high to buffer out
the effects of great variation in Na concentration in the K solution. Both
solutions are cooled before precipitation so that the temperature is constant
even for the precipitation of the last traces on standing. Cooling after the
initial mixing would result in a gradient of precipitate composition as cool-
ing took place.
6-24. Choice of Procedure for Evaluation of the Precipitate. Several
procedures are described below for evaluation of the cobaltinitrite precipi-
tate, and thus of potassium. The colorimetric method (as cobalt hydro-
catbonate) is the most rapid; the titrimetric method (with cerate or per-
manganate) is popular and equally as accurate; the gravimetric method is
most suitable for amounts of potassium of 5 mgm or above. The chief
variables in the cobaltinitrite determination are the conditions of precipita-
tion, which are satisfactorily controlled by the procedure given. A preci-
sion of 0.5 to 1 per cent is ordinarily obtained with any of the evaluation
procedures herewith described, and the precipitation procedure as de-
scribed is reproducible to I per cent.
6-25. Slightly different precipitate washing procedures are employed, ac-
cording to the procedure to be followed for evaluating the precipitate, and
therefore the washing procedure is detailed with each evaluation pro-
cedure below.

ALTERNATIVE PROCEDURE

6-26. For amounts of K less than 1 mgm, the prec1p1tation may ad-
vantageously be carried out in a 15-ml or 25-ml centrifuge tube. Washing
POTASSIUM DETERMINA TIONS FOR SOILS 119

by means of centrifugation follows, and the precipitate is evaluated by oxi-

dation in the tube. For quantities from 0.05 to 0.3 ll)gm, the centrifuge
tube technique is much to be preferred.
COLORIMETRIC DETERMINA TION OJ<' POTASSIUM AS COBALT
HYDROCARB ONATE
(Evaluation of potassium cobaltinitrite precipitate)
6-27. The precipitate of K as (K"Na:1 ,.)Co(N0) 0 • xH~O, obtained in
the previous section, is evaluated by dissolving the washed precipitate in
acid and determining the cobalt contained. The washing procedure for
the precipitate is given herewith. The standard curve is based on a series
of K standards derived from the same procedure. This colorimetric de-
termination is quantitative for cobalt in the range of 0.1 to 10 or more
mgm of Co. This is ordinarily too insensitive for the range of Co in soils
and plants. (Chapter 15 cites more sensitive methods.)

APPARATUS

6-28. Needed apparatus includes a colorimeter with 620-mu light maxi-

mum, colorimeter tubes with 40-ml calibration and stoppers to fit, fine
fritted glass filter crucibles, a suction flask and crucible holder, a 60° fun-
nel and rubber Gooch ring, a vacuum desiccator, a policeman, and a glass
rod.

REAGENTS

6-29. Needed reagents are 70 per cent ethanol, absolute methanol or

ethanol, 6 N HCl, 6 N KOH, 27 to 30 per cent U.S.P. grade H~O~ and 3
per cent H:P~ diluted from the concentrated reagent (made up fresh each
week), and saturated KHC0:1 (46 gm in 100 ml of water, shaken re-
peatedly and allowed to settle out).

PROCEDURE

6-30. Filtration and Washing of Precipitate. The cobaltinitrite precipi-

tate of K, formed at less than 5 °C as directed above, is filtered on a fine
porosity Pyrex fritted glass crucible with suction,rn and the 50-ml beaker
and precipitate are washed 3 times with 70 per cent ethanol. Then the
crucible is removed and the underside is washed with 70 per cent ethanol.
Finally the crucible is returned to the holder, the precipitate washed twice
more with 70 per cent ethanol and once with absolute methanol or ethanol.
The crucible is set in the 50-ml beaker and the apparatus and precipitate
are dried for 5 minutes at 100° to 110°C.

15 A centrifuge tube technique may be used to wash the precipitate, but that pro-
cedure is less rapid than the fritted glass filtration.
120 POTASSIUM DETERMINATION S FOR SOILS
6-31. Development of Cobalt Hydrocarbonate Color.lll The precipitate
adhering to the walls of the dried 50-ml precipitation beaker is dissolved
in 3 or 4 drops of 6 N HCl, a little heat being applied and the glass sur-
face being rubbed with the policeman. Then about 5 ml of hot water is
added to the beaker. Next the frittcd glass crucible is placed upright in the
beaker, and 2 ml 6 N HCI is pipetted into the crucible. The beaker and
crucible are placed on the steam plate for a few minutes until the precipi-
tate has dissolved.
6-32. A 60° funnel is arranged on a vacuum desiccator to deliver into
an Evelyn colorimeter tube placed upright inside. This tube bears a 40-ml
calibration mark and is colorimetrically standardized (Chapter 17). The
crucible bearing some of the solution of the precipitate is set in the rubber
crucible-holder in the funnel, suction is applied, and the cobalt solution in
the 50-ml beaker is washed through the filter into the colorimeter tube.
The transfer is completed with a few ml of water delivered from a fine-
stream, wash bottle. (For amounts of K greater than 8 mgm, the color is
made up in a volumetric flask.)
6-33. Now 1.5 ml 6 N KOH is added. The solution must remain slightly
acid as evidenced by absence of precipitation; if it is not slightly acid, a little
6 N HCI is added dropwise with stirring. Next, 0.5 ml of 3 per cent H~O~
is added, and the tube is stoppered and inverted to mix. (If a brown precipi-
tate begins to form, rarely the case, 2 to 5 drops 6 N HCI is added as re-
quired to clear the solution.) Finally 15 ml of saturated KHCO:: is added
and then water to make 40-ml total volume, followed by thorough mixing.
After the solution has stood for a few minutes for bubbles to rise, the color
is read in the colorimeter with 620-mu light maximum. The reagent blank
tube is employed for the I 00 per cent trammission setting of the colorime-
ter, automatically taking into account the potassium impurities in the re-
agents. The colorimeter readings are referred to the calibration curve
(based on working standards) to find the mgm K in the test sample.

ALTERNATIVE PROCEDURES

6-34. The cobalt of the potassium precipitate has been colorimetrically

determined 17 by the nitroso-R salt, a method that is well adapted to ex-
tremely small amounts of K below the range of the cobalt hydrocarbonate
method. The nitrite has been extracted with 0.1 N NaOH and determined
by diazotization. 1 k
6-35. The cobaltinitrite precipitate has also been evaluated by oxidation
in chromic acid, and the residual mixed color has been measured with 425-

16 Colorimetric procedure is adapted from that of Blanehetiere and Pirlot, Compt.

rend. soc. biol., 101:858 ( 1929).
17 Sideris, Ind. EnJ?. Chem .. A.E .. 9: 145 ( 1937), 14:821 ( 1942).
18 Taylor, J. Biol. Chem., 87: 27 ( 1930).
POTASSIUM DETERMINATION S FOR SOILS 121

mu light. 19 The residual dichromate color was measured later with 400-mu
light. 20 It should be noted that any method based on residual color, in the
lower range of the constituent, measures a small decrease in strong color,
which is less accurate than measuring the first increment of color.
6-36. The cobaltinitrite precipitate can be estimated turbidimetrically
for soil or tissue testing (Chapter 13), but this is difficult to make quantita-
tive because of the variability in the number of precipitation nucleii. 21 A
satisfactory procedure was devised 22 in which solid Na 3 Co(N0 2 ) 6 was em-
ployed for the precipitation nucleii.
6-37. Centrifuge Washing of the Precipitate. In lieu of filtration on a
fritted glass crucible, the precipitate may be washed by means of the cen-
trifuge. The precipitate and supernatant liquid, in the tube ( ~ 6-26) or
transferred from the 50-ml beaker with a rubber policeman and 70 per
cent ethanol dispensed from a wash bottle, are covered by a layer of 70
per cent ethanol. The precipitate is thrown down by centrifugation at 2000
rpm in a No. 2 International or other suitable centrifuge until clear, usu-
ally requiring 5 to 10 minutes or more. The supernatant liquid is removed
by gentle suction (10-20 cm of Hg) through a glass needle with tip bent
at a 90° angle. Then the precipitate is stirred with a fine jet of 70 per cent
ethanol (completed with a pointed rod if necessary) and the washing vol-
ume made to 5 to 10 ml. The precipitate is thrown down by centrifugation,
and the supernatant liquid is removed as before. Washing is repeated once
with 70 per cent ethanol and then once with absolute methanol or ethanol.
The tube is placed on its side to allow the alcohol to flow away from the
precipitate, dried in an oven at 100°-l 10°C for 30 minutes, and cooled.
The precipitate is then dissolved in 2 ml of 6 N HCl added to the centrifuge
tube, and the color is developed by the cobalt hydrocarbonate procedure.
The volume of colored solution is kept below 10 or 20 ml for small amounts
ofK.
6-38. The precipitate has been washed by means of a centrifuge by
many workers, several of whom have shown that the washings may be
limited to 1 if the procedure is so standardized that traces of residual pre-
cipitating reagent are represented in the calibration curve. Iced water and
other washing solutions such as 0.0 l N HNO/a are satisfactory because
the precipitate is relatively insoluble even in aqueous solvents, and the pro-
cedures are calibrated for any slight loss, even though in some cases the
stoichiometric composition of the precipitate was in the end approximated
by compensating factors. A 1 per cent solution of Al 2 (S0 4 ) 3 has been

rn Wander, Ind. Eng. Chem., A.E., 14:471 (1942).

20 Parks et al., Ind. Eng. Chem., A.E., 15: 530 ( 1943).
21 Volk, Proc. Soil Sci. Soc. Fla., 3 :99 ( 1941).
22 Garman, Soil Sci., 56: 101 ( 1943).
2a Wilcox, Ind. Eng. Chem., A.E., 9: 136 ( 1937).
122 POTASSIUM DETERMINATIONS FOR SOILS
employed24 as a washing solution to help keep the precipitate packed
down, thus facilitating inversion of the centrifuge tube for complete drain-
age.
TITRIMETRIC DETERMINATION OF POTASSIUM
(Evaluation of potassium cobaltinitrite precipitate by cerate or permanganate titration)
6-39. As an alternative to the cobalt hydrocarbonate procedure, the
precipitate of K as (KnNaa_JCo(N0 2 ) 11 • xH 2 0 (~ 6-22) is dissolved in
2 N H 2 S0 4 in the presence of an excess of cerate or permanganate, and
the excess of oxidant is determined by back titration. Because the washing
procedure varies according to the type of evaluation, it is detailed herewith.
APPARATUS

6-40. Needed apparatus includes a Gooch crucible suitable for holding

asbestos, a crucible holder, a suction flask, a glass rod fitted with a stopper
for tamping asbestos, a 100°-l I 0°C oven, pipets, and burets.
REAGENTS

6-41. Needed reagents are 70 per cent ethanol, absolute methanol or

ethanol, 2 N H 2 S04 , and the following special reagents.
6-42. Standard Oxidant. Approximately 50 gm of (NH,) 4 Ce(S0 4 ) 4 •
2 H 2 0 is dissolved in 1 liter of 2 N H 2 S0 4 to obtain a solution that is ap-
proximately 0.05 N with respect to oxidation-reduction, Ce 4 -1 -~ ce:i ' .
(If permanganate is to be employed, a stock solution of 1 N KMn0 4 is
prepared as described under manganese determination, ~ 5-67, and an
aliquot is diluted to 0.05 N and standardized.)
6-43. Oxidation-Reduction Indicator. Orthophenanthroline-ferrous com-
plex, 0.025 M, is obtainable as "Ferroin" from G. F. Smith Chemical Co.,
Columbus, Ohio.
6-44. Standard Oxalate. Standard Na 2 C 2 0 4 is dried in an oven at l00°C
and cooled. Then a 3.350-gm portion is weighed out and dissolved in
2 N H 2 S04 • The solution is made to a volume of I liter in a volumetric flask
with this solvent. This primary oxidation-reduction standcird is 0.0500 N.
6-45. Ferrous Reference Solution for Cerate Titration. Approximately
30 gm of Fe (NH 4 ) 2 (SO 4 ) 2 • 6 Hp is dissolved in approximately 1 liter
of 0.5 N H 2 S0 4 to obtain a solution of approximately 0.075 N with respect
to oxidation-reduction, Fe t-+ -~Fe+++. The volumetric ratio (R) of
cerate to iron is determined by titration at room temperature of 10 ml of
the cerate with the ferrous reference solution to the end point, 1 drop of
Ferroin being used as an internal indicator (red to faint blue color change).
Apparatus has been designed 25 to protect the stock solution from oxida-
tion, although daily determination of R is no disadvantage.

24 Thrum, Ind. Eng. Chem., A.E., 5:99 (1933).

2u Schollenberger, Ind. Eng_ Chem, A.E._ 7: 199 (1935).
POTASSIUM DETERMINATIONS FOR SOILS 123
6-46. Asbestos for Filtering Precipitates. The commercial grade asbestos
is ground in a water suspension with a rotary stirrer (drink mixer or
blender) then digested in aqua regia for 10 minutes at boiling temperature,
and finally washed with water on a Buchner funnel. The asbestos improves
with use and is collected and repeatedly used for the potassium filtration.

PROCEDURE

6-47. Filtration and Washing of Precipitate. An asbestos pad is made

by pouring a suspension of fairly fine purified asbestos through a Gooch
crucible, tamping with a stopper attached to a glass rod, and finally pouring
through an additional suspension of still finer asbestos 26 to give a total pad
thickness of about l.O to 1.5 mm. (A fritted glass filter is not satisfactory
with the titration procedure because some of the cobaltinitrite precipitate in
the pores of fritted glass does not react with the oxidant before it is lost by
side reactions with the acid and water, ~ 6-50. This contrasts with oxalate,
~ 5-33.)
6-48. The potassium precipitate (~ 6-22) is suspended by stirring and
poured into the center of the pad. The beaker is washed 3 times with small
portions of 70 per cent ethanol and once with absolute ethanol. The Gooch
(washed externally with ethanol) is dried with the beaker in the oven at
100°C for 10 to 30 minutes (while additional samples are filtered).
6-49. Titration Procedure (cerate). The Gooch is removed, and the pad
and precipitate are transferred to the original 50-ml beaker with a stirring
rod. Traces of precipitate adhering to the Gooch are washed into the beaker
with a fine stream of water. To the 50-ml beaker containing the suspension
of precipitate, approximately 1 ml of 2 N H 2S0/ 7 is added, followed im-
mediately by an excess of cerate ( 15 ml of 0.07 5 N suffices for 6 mgm K),
and the solution is rapidly heated to boiling while constantly being stirred.
The reaction is complete after 30 seconds of boiling. (If the yellow color
becomes rather faint, indicating insufficient excess of cerate, a measured
additional quantity of cerate is quickly added. A large excess of cerate
does no harm, so the procedure should be adjusted so that a sufficient ex-
cess is always added with the first pipetting.) The beaker with the solution
is set aside to cool, while cerate is added to other samples. The excess
cerate is back titrated with the ferrous ammonium sulfate reference solu-
tion to an end point with orthophenanthroline indicator.
6-50. The nitrite is oxidized to nitrate by the combined action of cerate

26 Asbestos gives a much lower titration blank than paper, according to Curtis and
Finkelstein, Ind. Eng. Chem., A .E., 5: 318 (1933), but its titration blank must be
carefully taken into account. Sized talc particles have been successfully employed
as a filter medium by Hibbard and Stout, I.A .0 .A .C., 16: 13 7 (193 3).
21 It is important that the filter system be acidified prior to the addition of cerate,
because eerie oxide otherwise precipitates at the interface of the acid cerate solution
and the nearly neutral precipitate phase.
124 POTASSIUM DETERMIN ATIONS FOR SOILS
( Ce 4 +) and cobalt ( co:i + ) , each of which is reduced by one valence
charge. 28 The cobalt is stabilized as trivalent by sixfold nitrite coordina-
tion but is stable only as divalent in the absence of 'nitrite excess. The
Co+++ ion is a strong enough oxidizing agent to liberate 0 2 from water:
Co+1 t + e~Co++ (6-1)
2 H 20 ~ 4H+ + 0 2 + 4e (6-2)

On addition:
4 Co+++ + 2 H 20 -~ 4 Co++ + 4H+ + 02 ( 6-3)

This takes place in the absence of a more ready source of electrons, such
as nitrite.
6--51. Strong acidification retards reactions 6-2 and 6-3 and liberates
HN0 2 for stoichiometric reaction with the Cot-++. The essential reactions
are as follows:
6N0 2 - + 60--~6NO:i- + 12e (6-4)
Co+-t + + e-~ Co++ (6-5)
11 CeH + 11 e ~ 11 Ce+++ (6-6)
On addition and doubling:
2 (K. Na)3Co(N02) u + 22 H4Ce(S04)4 + 12 H20 ~ 6 (K, Na)NOa
+ 2 Co(NO:i)2 + 2 HN0:1
+ 11 Ce2(S04):i + 55 H2S04
(6-7)

6--52. The K determined cannot be calculated from equation 6-7 be-

cause the K content of the precipitate, represented by (K, Na) :i• must be
evaluated by means of potassium standards, under precipitation conditions
that hold the K/Na ratio in the precipitate constant ( ~] 6-7). Then the
computation of the mgm K is as follows:
. mgm K (std) (6-8)
mgm K determined = --------- ---- x net ml cerate
net ml cerate (std) (unknown)

when mgm K (std) equals the number of mgm K precipitated in standard

(usually 2 mgm K), net ml cerate is ml cerate pipetted into the determina-
tion minus ml Fe back titration time R, and R, the ratio Ce volume : Fe
voH:tme, is determined in a separate titration, in which IO ml of the cerate
is titrated to the Ferroin end point with the ferrous reference soluti.on.

AlTERNA TIVE PROCEDURES

6--53. Besides the sulfatocerate procedure for evaluating the potassium
cobaltinitrite precipitate, detailed above, there are several other volumetric
28 Intraction of cobalt with nitrite in stoichiometric proportions was noted by
Drushel, Am. J. Sci., 24:433 (1907), through removal of Co from the titration by
NaOH treatment.
POTASSIUM DETERMINATION S FOR SOILS 125

procedures that may be employed, the chief ones of which utilize nitrato-
or perchloratocerate, permanganate, or chromate. The cerate procedures
have advantages over permanganate. They have great stability even in very
dilute solutions; no precipitate is formed in acid solutions and thus they
are usable in large excess in routine constant volume, which is particu-
larly convenient for use in centrifuge tubes with micro quantities of K.
They give a very sharp end point with Fcrroin and Nitrofcrroin indicators.
The sulfatocerate in either 1 or 2 N H 2 S04 solution, with oxidation poten-
tial of -- 1.44 volts, has the slight disadvantage of requiring the ferrous
reference solution, and a temperature over 50'C for reaction with Na 2 C 4 0 4 ,
and back-titration at less than 50CC. It is a slightly less active oxidizer than
permanganate, which has an oxidation potential of -1.5 volts in 1 N H 2 S0 4 •
Nitratocerate with an oxidation potential of - 1.61 volts in 1 N HN0:1 and
perchloratocerate with an oxidation potential of --1.71 volts in 1 N HC10 4 ,
are stronger oxidizers than permanganate. Oxalate can be titrated with
these cerates at room temperature. The permanganate is slightly cheaper,
and more commonly called for in older procedures, but has the decided dis-
advantage of instability in storage and use.
6-54. Permanganate Titration of Cobaltinitrite. On the basis of an esti-
mate of the amount of potassium present in the precipitate to be titrated,
approximately 2~ml excess of 0.05 N KMn0 1 is dispensed from a buret
into the 50-ml precipitation beaker (I mgm K is equivalent to approxi-
mately 3 ml of 0.05 N KMn0 1 ). To the beaker is then added 20 ml of
2 N H~S0 4 • The crucible containing the precipitate is now washed on the
,outside with a fin\: stream of water. The asbestos pad is removed from the
•crucible with a glass rod, and transferred to the acid KMnO" and the solu-
tion is stirred vigorously as it is heated rapidly. If the solution begins to
,turn colorless, more KMn0 1 is added at once to prevent loss of N0 2 • Heat-
ing is continued just to boiling. The crucible, containing traces of precipi-
tate, is placed in the KMnO, solution and stirred continuously for about
1 minute, which usually suffices to complete the reaction as evidenced by
no further change in the permanganate color. (If a larger titration volume
is required than will go into the 50-ml precipitation beaker, the titration
may be carried out in a larger beaker and a portion of the solution contain-
ing excess KMn0 4 can be used to wash the last traces of precipitate from
the 50-ml beaker.)
6-55. Next a slight excess of 0.05 N sodium oxalate is added from a
buret and stirring is continued until the excess of permanganate is con-
sumed, and a water clear solution is obtained. The excess of oxalate is
then back titrated with the 0.05 N KMnOj to a faint pink color. The vol-
ume of 0.05 N Na 2 C~0 4 used is subtracted from the total volume of 0.05
N KMn0 4 , which has been dispensed from the buret, to obtain the net
iKMn0 4 used in oxidizing the precipitate.
126 POTASSIUM DETERMIN ATIONS FOR SOILS
6-56. The essential reactions with permanganate are similar to those
with cerate:
6 N0 2 --+ 6 o-- ~ 6 NO:i- + 12e (6-9)
Co-f + 1 + e-~ Co++ (6-10)
Mn 7 I +Se~ Mn1 + (6-1 J)

On addition:
5 (K, NahCo(N02)H +I I HMn04 + 11H 2 S04~15 (K, Na)NO~ + 5 Co(NO:ih +
5 HNOa + 11 MnS04 + 14 H20
(6-12)

Also:
5 Na2C204 + 8 H2S04 + 2 KMn04 ~ 5 Na2S04 + K2S04 + 2 MnS04 +
10 C02 + 8 H20 (6-13)

6-57. The quantity of K is calculated by equation 6-8, the word "per-

manganate" being substituted for the word "cerate," when
Net ml of permanganat e= ml of permanganate added - ml of Na 2C 2 0 4 added X R

. d .
( permanganate volume ) d etermme
. h' h R
lTI W IC = ratJO. ---()xaJate-v olume-- Jn a separate

standardization in which 10 ml of hot Na 2 C 2 0 4 solution is titrated with

KMn0 4 solution. Ordinarily the strength of the KMn0 4 solution will not
be exactly equal to that of the oxalate because the former is continuously
changing. The quantity of K cannot be calculated from equation 6-13 be-
cause the content of K, represented by (K, Na)a (~I 6-7), must be evalu-
ated under the conditions of the precipitation.
6-58. Various procedures have been evolved for insuring good contact
of the precipitate with the oxidant to prevent escape of unoxidized nitrogen
oxides. One is to dissolve the precipitate in NaOH with precipitation of the
cobalt as a hydroxide. 2 !! This is a satisfactory routine, which gives the
same titer value as the direct acidification procedure above.
6-59. Centrifuge Washing Technique. When the volumetric evaluation
of the cobaltinitrite precipitation is combined with the centrifuge method
of washing the precipitate, the same washing procedure and solvents are
employed as described in connection with Gooch filtration and colorimetric
evaluation ( ~ 6-30). The volumetric cerate procedure is the same, except
that a more dilute cerate solution is used in the tube if the amount of K is
less than 0.5 mgm. After the tube and precipitate have been oven-dried, the
excess cerate is pipetted directly into the tube, the precipitate is stirred with
a glass rod, and the suspension is heated by partly immersing the tube in a
hot water bath to oxidize the cobaltinitrite. The excess cerate is back

:iu Klein and Mendel, Ind. Eng. Chem., A.E., 12:687 (1940).
POTASSIUM DETERMINA TJONS FOR SOILS 127

titrated with the ferrous solution, usually being more conveniently done
after the solution is transferred to a 50-ml beaker.
6-60. Permanganate is much less suitable than cerate as an oxidant for
use in excess in the centrifuge tube method.

GRAVIMETRIC DETERMINATION OF POTASSIUM

(Evaluation of potassium cobaltinitrite precipitate of over 5 mgm of K)
6-61. The weight of the potassium cobaltinitrite precipitate as
(K 11 Na 3 _ 11)Co(NO~)n · xH~O has been found to be as reproducible as
the colorimetric or volumetric evaluation, and therefore the gravimetric
method is highly satisfactory for macro amounts of K. The only precaution
is that the concentration of K must not exceed 2 mgm per ml (as stated in
~ 6-23). Any volume of the 2 solutions may be employed, so that this
precaution places no upper limit on the amount of K that may be deter-
mined.

APPARATUS

6-62. Needed apparatus includes precipitation beakers, a fritted glass

filter crucible, a holder, a suction flask, a policeman and glass rod, an oven
at I 00° to 110°C, and an analytical balance.

REAGENTS

6-63. Needed reagents include precipitation reagents as given in ~1 6-15,

and washing solutions-70 per cent ethanol and absolute methanol or
ethanol.

PROCEDURE

6-64. The potassium is prepared and precipitated (~ 6-20 and following)

by adding the cooled K solution containing 5 mgm of K or more to the
cooled 20 per cent solution of cobaltinitrite precipitating reagent in a 1 : 1
volume ratio of the 2 solutions. After standing over night, the precipitate
is filtered on a fritted glass filter that has been dried at 100° to 110°C and
weighed. The precipitation beaker and precipitates are washed 3 times with
70 per cent ethanol and once with absolute methanol or ethanol.
6-65. The crucible and precipitate are then dried at 100° to l 10°C for
30 minutes, cooled, and weighed. Then:
mgm K = Gravimetric factor x weight of precipitate, mgm ( 6-14)
Kprecipitated (std), mgm . . .
mgm K = --;-~-- - .---~------~ -- x weight of prec1p1tate ' mgm
weight precipitate (std), mgm
(6-15)
The gravimetric factor has been found to be 0.135 mgm K per mgm pre-
cipitate for the operating conditions specified, but should be redetermined
128 POTASSIUM DETERMINATION S FOR SOILS
for the actual operating conditions employed. A standard consisting of 10
to 20 mgm of K is suitable.
6-66. Precipitation and washing in a centrifuge tube is poorly suited for
the gravimetric determination of potassium because of the large quantity
of precipitate usually involved.

EXTRACTION OF EXCHANGEABLE POTASSIUM FROM SOILS

(Equilibrium extraction method in I N NH10Ac)
6-67. Although the exchangeable K has been extracted by the 1 N
NH 4 0Ac in the system for exchangeable cations (~I 5-7), a separate rapid
procedure is given here because of the frequent interest in exchangeable
potassium, separately from other exchangeable metallic cations. The potas-
sium obtained from this extraction may be determined by the cobaltinitrite
procedure (or another). A slightly different extraction is ordinarily em-
ployed for determination of exchangeable potassium from soil or runoff
by flame emission (~I 18-24).

APPARATUS

6-68. Needed apparatus includes 500-ml conical extraction flasks and

stoppers, a rotary shaker for flasks, a Buchner funnel attached to a suction
flask (or a vacuum desiccator that holds a 400-ml beaker), 400-ml beakers,
cover glasses, and glass hooks.

REAGENTS

6-69. Needed reagents are 1 N NH~OAc (prepared by dilution of 57 ml

of glacial HOAc to 800 ml, neutralization with concentrated NH 4 0H to
pH 7, and dilution to I liter), and 30 per cent H~O~.

PROCEDURE

6-70. A sample, cons1stmg of 30.0 gm of field-moist soil (preferably

freshly taken) that has passed a 5-mm sieve, is transferred to a 500-ml
conical flask. Then 300 ml of 1 N NH.10Ac solution is added. The flask
is stoppered and shaken for 30 minutes on a rotary shaker, or vigorously
by hand intermittently every 5 minutes. A separate 20-gm moisture sample
is weighed out and dried in an oven at I 00°C, then reweighed.
6-71. The suspension is then filtered through a Buchner funnel on which
a moistened filter paper has been tightly sealed by suction, the filtrate being
collected in a suction flask or a 400-ml beaker in a vacuum desiccator. The ·
extraction flask is rinsed with two 25-ml portions of N NH 4 0Ac, the wash-
ings being poured through the filter. The filtrate is evaporated to dryness on
a steam hot plate overnight. (For greater speed, the solution may be boiled
down gently to about 10 to 20 ml, but should not be allowed to go to dry-
ness over a burner, because carbon is formed, which resists oxidation later.)
POTASSIUM DETERMINATIONS FOR SOILS 129
6-72. After the solution has evaporated to dryness, the sides of the
beaker are washed down and the evaporation to dryness is repeated. The
second evaporation gives any mechanically held NH 4 0Ac a chance to es-
cape. Next, 5 ml of 30 per cent H 20 2 is added, the beaker is covered with
a watch glass, and the solution is digested for 30 minutes to oxidize residual
organic matter. The refluxing that takes place in the closed beaker washes
down the sides of the beaker, and makes the digestion much more effective
than simple evaporation with H 2 0 2 • Finally the cover glass is lifted onto
glass hooks, and the solution is evaporated to dryness. The H 2 0 2 treatment
is repeated if necessary to obtain a white residue indicating absence of
organic matter. The beaker is protected from deposition of NH 4 Cl. The
residue is then ready for the potassium determination ( ~ 6-20).
6-73. Calculation of Results. The mgm of K obtained in the potassium
determination is converted to meq of K per 100 gm of soils:
. mgm of K x 100
meq of K per I 00 gm of soil = 39 . h f (6-16)
.1 x we1g to
sample, gm
"Weight of sample, gm" is identical to the net oven-dry weight of the 20-
gm sample dried for a moisture determination, provided a 10/15 (or a
. 5/7.5) aliquot of potassium solution was taken for precipitation (~ 6-22).
The range of 0.05 to 0.3 meq per 100 gm ( 50 to 250 pp2m) is usual for
unfertilized soils in the humid region ( ~ 5-6). For pp2m of K in the soil:

pp2m of K = ----~~m_K 2?_~_2Q -- (6-17)

weight of sample, gm
ALTERNATIVE PROCEDURE

6-74. An ordinary filter may be employed instead of the Buchner funnel,

but it is advantageous then to use 40 gm of soil and exactly 400 ml of ex-
traction solution. After the shaking period, the extraction solution is filtered
through a dry filter into a dry receiving flask, and the rinsing of the flask
is omitted. If the first portion of the extraction solution comes through
cloudy, it is poured back through the filter. Finally, exactly 300 ml of the
extraction solution (representing 30 gm of soil) is evaporated to dryness as
in the regular procedure given.
•
TOTAL POTASSIUM IN SOILS OR MINERALS
6-75. Total K in soils or minerals is best determined by decomposition
of the sam pie by means of HF ( ~ 11-3 1 or 11-176) followed by deter-
mination by the emission spectrophotometric method (~ 18-27) or by
th~ cobaltinitrite precipitation ( ~ 6-22). The total Na can be determined
on the same sample as the total K when the HF decomposition is employed.
The total K of soils may also be extracted by a Na 2C03 fusion (~ 11-104)
130 POTASSIUM DETERMIN ATIONS FOR SOILS
followed by disintegration of the melt with the addition of 30 ml of HCl
(no HC10 4 , ~ 11-118) and removal of sesquioxides (~ 11-180). Blanks
are carried on all reagents. Two K standards which will yield 2 mgm of K
in the final aliquot are analyzed with the addition of all reagents. (The 5
gm of Na 2 C0 8 containing 0.002 per cent K carries 1 mgm of K into the
sample, which must be taken into account by the blanks and standards.)
The J. Lawrence Smith method for total K and Na involves fusion in
NH 4 Cl-CaC0 3 (~ 11-181).

SOIL-POTASSIUM EQUILIBR IUM

6-76. As the exchangeable potassium of soils is removed by leaching
or by crops, some of the reserve (or "total") potassium weathers from
feldspars and micas and becomes exchangeable. The exchangeable potas-
sium tends by equilibration to return to the original Jevel. 30 The addition of
soluble potassium salts to soils causes a reversal of this reaction. That is,
some of the potassium is "fixed" in nonexchangeable for~, more so when
the soil is dried. 31 The equilibrium nature of this reaction is vividly il-
lustrated by the release of potassium as a result of drying a soil that had
been cropped severely to remove the readily exchangeable potassium.:i 2
Also, exhaustive cropping with oats resulted in a recovery of potassium that
had been fixed in nonexchangeable form. In regions where soils have rela-
tively high K supplying power, soil testing laboratories frequently dry the
soil sample for a week or more at 95°F with low humidity to permit the
release of K to exchangeable form before the testing is done. Either fixa-
tion or release of potassium occurs 33 as a result of freezing and thawing,
depending on the equilibrium level of K.
6-77. Soil, NH 4 -saturated and dried at 500°C, released 34 much more
K than extractions with acids, bases, or salts without the heating. The "log
K released" fell off linearly with successive NH 4 -drying treatments.
6-78. Addition of colloidal alumina or use of a Na 2C03 boiling treat-
ment35 increased the quantity of K fixed by soil on drying, while removal
of colloidal materials by H 2 0 2 and H 2 S-HC1 treatments decreased the
quantity of K fixed. Addition of K2 Si0a to montmorillonite or illite greatly
increased:l6 the quantity of K fixed compared to that fixed with KOH
added in equivalent amounts.
6-79. These findings and many others, clearly reveal the equilibrium

:io Bray and DeTurk, S.S.S.A. Proc., 3: 101 (1939).

:n Volk, Soil Sci., 37:267 (1934).
32Attoe,S.S.S .A.Proc., 13:112 (1949).
33 Fine et al .. S.S.S.A. Proc .. 5: 183 (1941 ).
34 Kolterman and Truog, S.S.S.A. Proc., 17:347 ( 1953 ).
8oVolk,Soi/S ci., 45:263 (1938).
36 Mortland and Gieseking, Soil Sci., 71: 381 (1951).
POTASSIUM DETERMINATIONS FOR SOILS 131
nature of the reactions by which potassium is interchanged from exchange-
able to nonexchangeable form. The equilibrium reaction with the layer
silicate vermiculite37 involves closure of the interlayer spaces by K ions,
after which the K is not readily exchanged. Extending this concept, fixation
of potassium by 2 : 1 layer silicate clays was attributed 88 to a frequency
distribution of layer change density and K fixation in interlayer spaces in-
volving high charge density. Potassium in interlayers surrounded by low
charge density remains exchangeable. This mechanism accords with the
evidence 39 that mixed layer structures are formed by K fixation. The
equilibrium and rate functions of K fixation have been summarized. 40
Phosphate added to soils can also fix potassium by precipitation of potas-
sium aluminum or iron phosphate salts. 41
6-80. NH 4 0Ac Standard for Exchangeable K. Use of N NH 4 0Ac ex-
traction ( ~ 6-70) was advocated 4 ~ as the standard for extraction of the
exchangeable form of soil K because the level of exchangeable K of soils
rose rapidly within an hour after replacement with H or Ca but did not
after NH 4 0Ac replacement. Since NH 4 + holds highly charged layers to-
gether (~ 8-35) just as K, the release of nonexchangeable K to exchange-
able form is retarded during NH 4 0Ac extraction.
6-81. K-fixation Capacity of Soils. The following procedure for potas-
sium fixation capacity of soils (wetting-and-drying basis) was suggested by
N. J. Volk: to 10 gm of soil, 10 mgm of K is added in the form of IO ml
of KC! solution. The soil is dried and wetted 10 times on a steam plate.
The K is then extracted in N NH 4 0Ac (~ 6-70) and determined. A second
10-gm sample of soil is similarly treated with 10 mgm of K, and the soil
is immediately extracted with N NH 4 0Ac and the K is determined. The
difference, expressed as meq of K fixed per 100 gm of soil, is the K fixation
capacity of the soil.
6-82. Nonreplacement of fixed K by NH 4 from vermiculite was demon-
strated by Barshad. 4 :1 Both of these ions are released by boiling in NaOH.
The release may be attributed to the fact that the Na-vermiculite has
expanded spacings from which the K may migrate out into solution. Deter-
mination of the K-fixation capacity of soils (wet-fixation basis) was sug-
gested as follows: of 2 samples, 1 is saturated with K, then both with am-
monium. Then both are subjected to NaOH distillation of NH 4 , and the
difference (expressed as meq of K fixed per 100 gm of soil) in NH 4 held is

37 Barshad, Soil Sci., 72:361 (1951).

:is Jackson et al., S.S.S.A. Proc., 16:3 ( 1952).
:111 Dyal and Hendricks, S.S.S.A. Proc., 16:45 (1952).
40 Wiklander, Potassium-Symposium (Berne: Int. Pot. Inst., 1954), p. 109.
41 Cole and Jackson, S.S.S.A. Proc., 15:84 (1951); Haseman et al., S.S.S.A. Proc.,
15:76 (1951).
42 Merwin and Peech, S.S.S.A. Proc., 15: 125 (1951 ).
4a Soil Sci., 72:361 (1951).
132 POTASSIUM DETERMINATIONS FOR SOILS
the K-fixation capacity. Wet fixation and dry fixation were measured by K
saturation, 0.5 N Mg ( OAc) 2 being used as a replacing agent. 44
6-83. Acid-Soluble K of Soils. Extensive efforts have been made to de-
velop a method of soil potassium extraction that is not exchangeable to be-
gin with but is taken up by plants in the course of exhaustive cropping. 45
This form is sometimes called the "moderately available form." Some of
these extractions involve various concentrations of acids and electrodi-
alysis ( ~ 6--85). Much more K is extracted40 by 0.5 N HCl than is ex-
changeable with N NH 4 0Ac, and such extraction after drying of the H-
saturated soil (cycle repeated 6 times) releases 47 an amount of K that was
related to the K recovered by exhaustive cropping.
6-84. Extraction of soil with l N HNOa for l 0 minutes 48 removed an
amount of K that seemed to be correlated with removal by cropping. A
1 : 10 ratio of soil to the 1 N NH0:1 gave an amount of K extracted that
was insensitive to dilution but was sensitive to time of boiling. 49 A 10-
minute boiling time was adopted. The quantity of K released by this acid
treatment was correlated significantly with the quantity removed by ex-
haustive cropping of several soils. A correlation was found 50 of the K thus
extracted and the microcline content of the silt fraction of 12 Indiana soils.
A correlation coefficient of 0.9 was obtained51 between alfalfa uptake of K
from some Iowa soils and the K released from 2 gm of soil in 25 ml of
N NH Ox on boiling 10 minutes after standing 15 minutes.
6-85. Electrodialysis Release of K. Electrodialysis for 30 days was used" 2
to measure the supplying capacity of Hawaiian soils. Release of K by elec-
trodialysis could be correlated with the K removal by exhaustive crop-
ping.r.:i The K removed by electrodialysis was related neither to the ex-
changeable K nor to the total K, but the total quantity of K removed when
the rate leveled off and became constant was related to past K fertilization
and potassium fertility level. 54 From 10 to 20 days of electrodialysis was
required before the rate became constant.

44 Mare), Verslag. Landhouwk. Onderzoek. No. 61, 8, S-gravenhage (1955).

45 Drake and Scarseth. S.S.S.A. Proc., 4:201 (1940); Olsen and Shaw, J. Am. Soc.
Agron., 35:1 (1942); Bear et al., Soil Sci., 58:139 (1944); Chandler et al., J. Am.
Soc. Agron., 37:709 (1945); Evans and Attoe, Soil Sci., 66:323 (1948); Gholston
and Hoover, S.S.S.A. Proc., 13: 116 ( 1949); Legg and Beacher, S.S.S.A. Proc., 16:210
(1952); Williams and Jenny, S.S.S.A. Proc., 16:216 ( 1952 ).
4G Attoe and Truog, S.S.S.A. Proc., 10: 81 ( 1946).
47 Evans and Simon, S.S.S.A. Proc., 14: 126 ( 1950).
48 DeTurk et al., Soil Sci., 55: I ( 1943).
40 Rouse and Bertramson, S.S.S.A. Proc., 14: 113 ( 1950).
r.o Phillippe and White, S.S.S.A. Proc., 16:371 (1952).
r.1 Pratt, Soil Sci., 72:107 (1951).
52 Ayres et al.. S.S.S.A. Proc., II: 175 ( 1947).
l'i3 Reitemeier et al., S.S.S.A. Proc., 12: 158 (1948).
54 Pearson, Soil Sci., 74:301 (1952).
POTASSIUM DETERMINATIONS FOR SOILS 133

QUESTIONS
1. What are the principal methods for the determination of potassium of
soils and plants?
2. Why is the flame emission method preferable to precipitation methods
for K, and why are precipitation methods still widely employed?
3. Outline the principles of the cobaltinitrite method for the determination
of potassium, including the role of Na.
4. How docs the water solubility of potassium cobaltinitrite compare to that
of KCI0 4 and K~PtClu?
5. Why must ammonium ions be removed before potassium is precipitated?
6. List several procedures available for evaluating the potassium cobaltini-
trite precipitate, and give the principle involved in each.
7. What factor is involved in the gravimetric determination of potassium
cobaltinitrite that is not involved in titrimetric determination, and how is this
additional factor taken care of analytically?
8. What reagent is most commonly employed for the extraction of exchange-
able K?
9. What are the chief methods for the extraction of total K from soil or clay
for its determination?
I 0. Why is there a difference between wet fixation and dry K fixation capacity
of soil?
I I. What is the purpose of applying stronger extraction procedures such as
with 0.5 N HCI, I N HNO;i and electrodialysis in the extraction of K from soils?
 7
Phosphorus Determinations
for Soils
Availability, fixation, fractionation

7-1. Phosphorus determinations for soils have received extensive atten-

tion from soil chemists because of the importance of phosphorus in soil
fertility .1 The chemistry of phosphorus in solution as well as in soil is
rather complex, and methods of analytical determination have required ex-
acting attention to every controllable detail. The extraction of phosphorus
from soils has received extensive study along 2 lines: (a) the extraction of
more active or available phosphate with high specific surface, and (b)
fractionation of the total of each chemical class or species of phosphate.
The similarity of arsenic reactions to those of phosphorus and the fact that
arsenic is added to soils through vegetation sprays and in studies of phos-
phorus reactions make the differentiation of arsenic from phosphorus an
important consideration in the determination of phosphorus.
7-2. Methods for Phosphorus Determination. Phosphorus determination
for soils was greatly expedited by the development of sensitive colorimetric
methods, although less sensitive and more laborious precipitation-titri-
metric2 and gravimetric 3 methods have long been employed. The Osmond 4
1 Pierre and Norman, Soil and Fertilizer Phosphorus (New York: Academic Press,
Inc., 1953); Pierre, ed., Summary of Phosporus Research in the United States (Belts-
ville, Md.: Natl. Soil and Fert. Res. Com., U.S.D.A. Soils Div., 1950); Waggaman,
Phosphoric Acid, Phosphate, and Phosphatic Fertilizers, 2nd ed. (New York: Rein-
hold Publishing Corporation, 1952); DeMolon and Marquis, Le phosphere et la Vie
(Paris: Presses Univ. de France, 1949).
2 Methods of Analysis, 7th ed. (Washington, D.C.: A.O.A.C., 1950), p. 36.
3 Washington, Chemical Analysis of Rocks, 3rd ed. (New York: John Wiley & Sons,
lnc., 1919), p. 216.
4 Bui. soc. chim. biol., 47:745 (1887).

134
PHOSPHORUS DETERMINATIONS FOR SOILS 135
method, involving the molybdophosphoric blue color produced by selec-
tive reduction of the heteropoly molybdophosphoric acid, has been exten-
sively adapted. Four widely used modifications:
I. Chlorostannous-reduced molybdophosphoric blue color method, in a
sulfuric acid system.
II. Chlorostannous-reduced molybdophosphoric blue color method, in a
hydrochloric acid system.
III. Molybdenum-reduced molybdophosphoric blue color method, in a
sulfuric acid system.
IV. 1, 2, 4-aminonaphtholsulfonic-reduced molybdophosphoric blue
method, in a perchloric or sulfuric acid system.

are described herein. Also included is method V, which is based on the

yellow color of the unreduced vanadomolybdophosphoric heteropoly com-
plex. Each of these methods offers its respective advantages of sensitivity,
tolerance to various levels of interfering substances, speed, or stability of
color. The choice of method for a given phosphorus analysis depends on
the amount of phosphorus available for the determination, the acid system
involved in the scheme of analysis, and the concentration of interfering
substances present in the solution to be analyzed.
7-3. Basic Principles. The heteropoly complexes are thought to be
formed by coordination of molybdate ions, with phosphorus as the central
coordinating atom," the oxygen of the molybdate radicals being substi-
tuted for that of PO 4 :
(7-1)

Ions besides (P 5 t ) , which act as the central coordinating atom to form

12-fold heteropoly acids with molybdate, include arsenic (As·~+ ) , silicon
(Si 4 +), germanium (GeH), and under some conditions molybdenum
(Mo 6 +) and boron (BH ). Also, wolframate 11 (also termed "tungstate")
ions are coordinated about P as a central atom in a fashion analogous to
molybdate ions, though with somewhat less avidity. The heteropoly com-
plexes, before reduction, give a yellow hue to their water solution. With
high P concentrations, a yellow precipitate is formed ( eq. 7-1). In solu-
tions of low enough concentration to be suitable for determination by re-
duction to form the blue color, the yellow color is so faint that it is not
noticed. At about 100 times this concentration, the yellow color is suitable
for spectrophotometry as for molybdosilicic acid color ( ~ 11-72). The
vanadomolybdophosphoric yellow color (method V) is greatly intensified
by the vanadium component.

"Boltz et al., Anal. Chem., 21:563 (1949); Kraus, Z. Krist., 100:394 (1939);
Keggin, Proc. Roy. Soc. (London), A 144:75 (1934), Nature, 131:908 (1933);
Hastings and Frediana, Anal. Chem., 20:382 (1948).
6 International Union of Chemistry, Sci. American, 181, 5 :30 (Nov. 1!149).
136 PHOSPHORUS DETERMINATIONS FOR SOILS
7-4. A characteristic blue color (the "molybdenum blue reaction" 7 ) is
produced when either molybdate or its heteropoly complexes are partially
reduced. Some of the molybdenum ions are reduced from 6+ to a lower
valence, probably 3 + and/or 5 +, involving unpaired electrons from which
spectrophotometric resonance (blue color) would be expected. The spec-
trophotometric absorption curves 8 for the blue color of molybdenum blue
in the presence or absence of heteropoly complexes of phosphorus shows
2 wave lengths for characteristic light absorption, at 660 mu and 830 mu.
Expressed as an extinction coefficient, E, in which E = log (Tblank/Ttest),
the sensitivity at 660 is one-third that in the infrared at 830 mu. Tblank
and Ttest refer to the percentage transmission of the blank and test solu-
tion, respectively.
7-5. Optimum Concentration of Reagents. The optimum concentration
of acid, molybdate, and reductant is that which will give the maximum of
color per unit of P present (in accord with Beer's law), and the minimum

260--~~--------.~--------------.-----r~..--.

240
220
200
180 ~ability plateau
bD
:g 160 I I
m
... 140 I
I
Wet-ashed corn leaf
~E 120 I
I

~ 100 ,1 I I
8
80
I !I I I
11lm' In j1v lrva
II i;--r-+----'---~
60
40 II I
o ppm P I
20
'V
0 2 4 6 8 l.0 l.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
HzS0 4 normality
Fig. 7-1. Effect of acid concentration on molybdophosphoric
heteropoly blue color intensity, with acid concentrations employed
in methods I, II, Ill, IV, and IVa. The curves were made with
1,2,4-aminonaphtholsulfonic acid reduction, with 0.4 per cent
ammonium molybdate as in method IVa. (Adapted in part from
Cotton, Ind. Eng. Chem., A.E., 17:736, 1945, and do not apply
in detail to the other methods with different molybdate concen-
trations.)

7 Berzelius, Pogg. Ann., 6:380 (1826); McAlpine and Soule, Qualitative Chemical
Analysis (New York: P. Van Nostrand Company, Inc., 1933 ), p. 239.
s Boltz and Mellon, Anal. Chem., 19:874 (1947).
PHOSPHORUS DETERMINATION S FOR SOILS 137

of fading. 9 The range of acidity over which the color is not affected by the
acid concentration 10 may be termed the "acid-stability plateau" (Fig. 7-1).
This plateau narrows as the phosphorus concentration increases and widens
as it decreases, the low acidity boundary extending through 0.4 N H 2 S0 4 •
The acid-stability plateau shifts position somewhat with different re-
ductants, different molybdate concentrations, and in different acid systems.
At low acid concentration, molybdate itself, as well as all of the heteropoly
complexes, is easily reduced to form a blue color. The molybdate color
(no P present) falls off above 0.35 to 0.6 N acid, depending upon the
molybdate and reductant concentrations. The presence of phosphorus in-
creases the molybdomolybdic blue color in the low (0.2 N) acidity range,
but disproportionately to the phosphorus present. This disproportionate
color development from molybdate and phosphorus extends to higher
acidity progressively with higher phosphorus concentrations. At high acid
concentration, the normality varying with the molybdate : acid ratio, the
reduction of the molybdophosphoric complex itself falls off to zero.
7-6. Molybdoarsenic acid blue color is usually excluded from the phos-
phorus analysis (~ 7-64) by reduction to arsenious acid prior to the
addition of ammonium molybdate to form the heteropoly complex. The
complex does not form with the arsenious radical. Also, the P can be precip-
itated11 with Al(OH):1 and the precipitate treated with HF, HBr, HCI, and
H 2 S04 to volatilize As, Ge, and Si, which may be coprecipitated, leaving
the P for analysis. Usually, the molybdomolybdic and molybdosilicic com-
plexes are separated by adjusting the acidity. In a 0.35 N H~S0 4 solution,
the molybdate is not reduced, but the heteropoly molybdosilicic and phos-
phoric acid complexes are still reduced. In method I, with 0.39 N acid,
the molybdosilicic acid (with Si below 200 ppm) is not reduced. Molybdo-
silicic and molybdogermanic acid are reduced (as well as molybdophos-
phoric and molybdoarsenic acid) at acidity greater than 0.4 N under
some conditions, 12 but are not a problem under suitable conditions. The
acid strength is usually determined by the amount of acid added with
the molybdate reagent. The test solution is previously brought to pH 2. 7 to
3.0, 2, 4-dinitrophenol being used as the indicator. (The indicator 2, 6-
dinitrophenol is preferred to 2, 4- except for its unavailability.)
7-7. Increasing the molybdate concentration increases the required con-
centration of acid to prevent molybdate reduction without altering the
phosphorus concentration. That is, the acid-stability plateau (Fig. 7-1) is
shifted to the right by increasing the molybdate concentration. Thus it
extends 13 only from 1.7 to 2.1 NH 2 S04 with a 0.75 per cent molybdate
11 Dickman and Bray, Ind. Eng. Chem., A.E., 12:666 ( 1940).
10 Cotton, Ind. Eng. Chem., A.E., 17:736 ( 1945).
11 Levine et al., Sci., 119:327 (1954).
12 Boltz and Mellon, Anal. Chem., 19:873 (1947).
13 Fontaine, Ind. Eng. Chem., A.E., 14:77 (1942).
 TABLE 7-1
as methods I to V
Summary of the characteristics of 5 spectrophotometric methods for phosphorus described in text
I
Molybdopho sphoric blue color methods,
reductant consisting of
--
I Vanado-
Chlorostann ous
molybdo-
acid
I II I, 2, 4 amino phosphoric
I I yellow
In In Reduced naphthol
H2S04
system*
I HCI
system*
molybdate
system*
II sulfonict
acid*
color
method
III
I
! IV (IVa) v
Item compared I II
,-

0.4§ 0.8 0.9 2.2 6

p c nc at 50% transmission,! ppm P . .
%T ransmission, l ppm P, 660 mu . . . 18 36 50 75 -
0---2.5 0-4 0---10 0---20
::;:; Ran :e of conformity to Beer's Law. ppm P 0---1.0
0.8-20
0.02-1 0.05-2 0.1-5 0.2-10
00
Wor dng range, ppm P 15 min. 5 min.
Tim : to develop color . 5 min. 5 min. 30 min.
20°-25°C 25°C 100°c 25°C 25°C
Tern !)erature of color developmen t Changes** Infinite
Stab lity 15 min. 20 min. 24 hrs.
0.39 0.7 0.44 0.9 (1.3) 0.2-1.6
Fina I concentratio n of acid, N . . •,
0.5
0.1 0.3 0.0506 0.4
Fina I concentratio n of molybdate, % None I None
Effe :t of excess molybdate-a cid reagent Slight Slight Slight
decrease decrease decrease
Slight Slight Slight Increase -
Effe, :t of excess reductant
increase increase increase I
I I

0 After Woods and Mellon, Ind. Eng. Chem., A.E., 13:760 ( 1941 ).
t l, 2, 4-aminonapht holsulfonic acid is found to be more effective than 2, 5, 7-; l, 4, 8-; 2, 6, 8-.
t For 1 cm cell at 700 mu for methods I to IVa; Evelyn tube at 440 mu for method V.
1941 ).
§Calculated at 660 mu, from Fig. 1 of Woods and Mellon, Ind. Eng. Chem., A.E., 13:760 (
00 Measurement must be made at a definite time.
PHOSPHORUS DETERMINATIONS FOR SOILS 139
concentration, compared to the wide range from 0.6 to 1.6 N with 0.4
per cent molybdate (Fig. 7-1 ). The plateau extends further to the left
than in Fig. 7-1 under the conditions of method I with 0.1 per ceµt
molybdate (zero color with no phosphorus, at 0.39 N acid). The sensi-
tivity, final concentration of acid and molybdate, and other characteristics
of methods I to V are listed in Table 7-1.
7-8. Many different reductant reagents can be used to develop the
heteropoly blue color. Each has certain advantages, but it is doubtful if
any one reductant is in every respect better than all others. 14 The calibra-
tion curve for each method fits only 1 fairly definite amount of reducing
agent. A great excess cannot be employed because the reduction must; be
selective for the heteropoly complexes, and the excess molybdate reagent
must not be reduced. There is usually a small plateau over which slight varia-
tion in the amount of reducing agent is permissible. The plateau depends
upon the slight difference in reducibility between the heteropoly complex
and the uncomplexed molybdate. The plateau is so small that the amount
of reducing agent must be controlled closely.
7-9. Choice of Phosphorus Method. Sensitivity and freedom from inter-
fering ions arc the most important factors in the choice of phosphorus
method for any given application. For example, method I has much the
greatest sensitivity (Table 7-1 ) and is thus the most satisfactory for phos-
phorus extracted from infertile soils with relatively low soil : extractant
ratios. Method II is suited for systems high in chlorides. Method III is
relatively sensitive, eliminates the effect of arsenic, and has been used by
the U.S.D.A. in regional cooperative studies.rn Method IV, though less
sensitive, permits the determination in the presence of 200 ppm of ferric
iron, is little sensitive to moderate variations in acidity, and eliminates the
effect of arsenate. Method V, though much less sensitive, provides the ad-
vantages of color stability, freedom from the reduction step, greater free-
dom from interfering ions and contamination from glassware and lower
dilution required for a convenient size of soil or plant sample analyzed. The
proper choice of analytical method is indicated with each type of soil phos-
phorus extraction procedure to be described.

14 Reducing agents that have been employed include: chlorostannous acid (Osmond,
Chem. News, 56: 160, 1887), in methods I and II; gallic acid (Passerini, Gazz. chim.
ital., 41:182, 1911); hydroiodic acid (Wu, J. Biol. Chem., 43:218, 1920); hydroqui-
none (Bell and Doisy, J. Biol. Chem .. 44:55, 1920) and (Official Methods of Analysis,
6th ed., Washington, D.C.: A.O.A.C .. 1945, p. 127); sodium thiosulfate (Losana,
Giorn. chim. ind. applicata, 4:60, 1922); 1, 2, 4-aminonaphtholsulfonic acid (Fiske
and Subbarow, J. Biol. Chem., 66:375, 1925), employed in method IV; benzidine
(Feigl, Z. anal. Chem., 74:386, 1928); molybdate reduced with a metal, usually Mo
itself (Zinzadze, Ann. agron., n.s., 1: 321, 1931), employed in method III; p-meth-
ylaminophenol (Lingren, Analyst, 58:755, 1933); and hydrazine sulfate (Hague and
Bright, J. Research Nat. Bur. Standards, 26:505, 1941).
1s Peech et al., U.S.D.A. Cir. 757 ( 1947).
140 PHOSPHORUS DETERM INATION S FOR SOILS
7-10. Precautions Against Phosphorus Contamination. The glassware to
be used must be free of contamination with phosphorus (or arsenic, which
gjjes the same test unless reduced with NaHS0 3 ). Since Pyrex glass con-
ta'fns 0. 7 per cent arsenic oxide, new glassware must be thoroughly
weathered before use by treatment with warm sulfuric acid-dichromate
solution for at least 24 hours. Washing soaps and powders, if used, must
0

be completely removed by cleitaing in strong acid, as they frequently con-

tain phosphorus. As the last step in cleaning, the glassware is dipped or
rinsed in 6 N HCl after apparently clean, then filled with tap water, and
finally rinsed 3 times with distilled water. The reagents and filter paper
seWifted should be as free of phosphorus as possible. Dust, saliva, perspira-
tio!, and tobacco ashes carry appreciable amounts of phosphorus and
therefore should be excluded.
7-11. Primary Phosphate Standard, 50 ppm of P. Potassium dihydrogen
phosphate (KH 2 P0 4 , Sorenson's ~tandard, or recrystallized at pH 4.5) is
dried at 40°C, and 0.2195 gm is dissolved in about 400 ml of distilled
water in a 1000-ml volumetric flask. Then 25 ml of 7 N H 2 S0 4 is added,
and the solution made to 1000-ml volume and mixed, giving 50 ppm of P.
Thus preserved with H 2 S0 4 , the solution keeps indefinitely, 16 but it should
be stored in a weathered soft-glass bottle (rather than one of Pyrex), to
minimize contamination with arsenic. This solution is diluted directly for
the yellow colored solutions of the vanadomolybdophosphoric acid (method
V, Table 7-2).
7-12. Secondary Standards, 2 and 20 ppm of P. For the 2 ppm standard,
20 ml of the 50 ppm stock solution is diluted to 500 ml. For the 20 ppm
standard, 200 ml (pipetted or measured in a 200-ml volumetric flask) is
diluted to 500 ml. These more dilute stock solutions, especially the 2 ppm
solution, do not keep well, even with toluol1 7 added, and must be made up
fairly frequently. They are used in making the blue-colored solutions
(methods I, II, III, and IV, Table 7-2).
7-13. Working Standards and Blank. All reagents including the extrac-
tion solution and sample processing chemicals must be included in each of
the standard solutions and in the blank employed for any of the calibration
curves (Table 7-2). The influence of the extraneous ions and the im-
purities thus are taken into account. A slight color in the blank does not
interfere, since the photometer is set with the blank reading, G = 100
(L = 0), and the remaining solutions are read relative to this blank ( ~
17-39). The blank transmission percentage usually will be more than 95
per cent of a similar but uncolored solution. The "center setting" of the
Evelyn colorimeter should be between 76 and 90 if the reagents are rea-
sonably free from phosphorus and arsenic. If more than this slight trace of
16 Bertramson, Soil Sci., 53: 135 ( 1942).
17 Holman and Pollard, J. Soc. Chem. Ind., 56T:339 (1937).
PHOSPHORUS DETERMINATIONS FOR SOILS 141
color is noted in the blank, the source of contamination should be located
(~ 7-10).
TABLE 7-2
Appropriate concentration ranges of phosphorus standard solutions
for various analytical methods

Concentration obtained when stock solution is

diluted to 50 ml final volume
------
Volume of 2 ppm stock 20 ppm stock 150 pp_:_stock
stock*
solution taken I III IV
---·······-·-- I II ·--- ------ %
ml ppm ppm ppm ppm ppm Transmission
- - - · - - --·-·----- ----- --··---
0 (Blank) 0 0 0 0 0 100.0
0.50 0.02 0.2 0.2
1.00 0.04 0.04 0.4 0.4 1.0
2.50 0.10 0.10 1.0 1.0 2.5
5.00 0.20 0.20 2.0 2.0 5.0
7.50 0.30 3.0 7.5
I 0.00 0.40 4.0 10.0
12.5 0.50 5.0 12.5
15.0 0.60 0.60 6.0 15.0
20.0 0.80 8.0 20.0
25.0 1.0 10.0

0 Stock solution of KH"PO, containing: for methods I and II, 2 ugm of P per ml ( 2 ppm of P);

for methods Ill and IV, 20 ugm of P per ml (20 ppm of P); for method V, 50 ugm of P per ml
(50 ppm of P).

CHLO ROST ANNO US-REDUCED MOL YBDOPHOSPHORIC BLUE

COLOR METHOD, IN SULFURIC ACID SYSTEM 18
(METHOD I)
(Includes arsenate)
7-14. Method I has the highest sensitivity per unit of phosphorus present
(Table 7-1 ) , providing a working range from 0.02 to 0.6 ppm of P. It
provides for noninterference of Si in solution up to 200 ppm, Fe++ up to
100 ppm, Fe+ ·t + up to 2 ppm, Ti up to 20 ppm, Ca and Mg up to 500 or
more ppm, NOa up to 100 ppm, F up to 5 ppm, Cl up to 250 ppm, S04 up
to 1000 ppm-but it includes arsenate in chemical equivalence to P. Excess
of nitrate or chloride prevents maximum color development and hastens
color fading, and therefore is held to a minimum. Great excess of sulfate
18 Adapted from Deniges, Compt. Rend., 171: 802 ( 1920), Truog and Meyer, hid.
Eng. Chem., A.E., 1:136 (1929); Woods and Mellon, Ind. Eng. Chem., A. E., 13:760
(1941); and Bertramson, Soil Sci., 53: 135 (1942).
142 PHOSPHORUS DETERMIN ATIONS FOR SOlLS
intensifies the color. Preparation of the standard curve in the presence of the
same amount of salts as will be present in the test sample permits deter-
minations with Fe, Cl, S0 4 , and NO~ in much larger concentrations than the
stated limits. Determination of P can be carried out with several hundred
ppm of Fe if reduced(~ 7-132).
APPARATUS

7-15. Apparatus needed consists of a special 2-ml pipet for sulfo-

molybdic acid; a special storage container for chlorostannous acid; 50-ml,
500-ml, and 1000-ml volumetric flasks; 125-ml conical flasks; a colorim-
eter with 660-mu light maximum or visual comparator tubes; pipets; and
a buret.
REAGENTS

7-16. Needed reagents consist of 2 N H 2S0 4 , 4 N Na 2C0 3 , 2, 4-dinitro-

phenol indicator, standard phosphorus solution (Table 7-2), and the fol-
lowing special reagents.
7-17. Sulfomolybdic Acid Solution, 2.5 Per Cent. Exactly 25.0 gm of
c.p. ammonium molybdate, (NH 4 )nMo 7 0 24 • 4 H 2 0, is dissolved in 200 ml
of distilled water and warmed to 60°C. The solution is filtered to remove
sediment, if necessary. Then 275 ml of phosphorus-free and arsenic-free
concentrated sulfuric acid (35 to 36 N) is diluted to 750 ml with distilled
water. After both solutions have cooled, the ammonium molybdate solu-
tion is added slowly, with stirring, to the sulfuric acid solution. After the
combined solution has cooled to room temperature, it is diluted with water
to exactly 1000 ml. This is a 9.7 to 9.9 N sulfuric acid solution, contain-
ing 2.5 gm of ammonium molybdate per 100 ml. It is stored in an acid-
weathered amber glass bottle, 19 with a 2-ml rapid-delivering pipet at-
tached. Thus stored, protected from light, the solution keeps indefinitely. 20
7-18. Chlorostannous Acid Reductant. Approximately 25 gm of reagent
grade SnC12 • 2 H 20 is dissolved in 50 ml of concentrated HCl, with warm-
ing if necessary to dissolve. This solution is diluted (with rapid stirring) to
approximately 500 ml with recently boiled distilled water, giving about
0.2 M Sn++. The molar concentration of stannous in the solution is best
determined (~ 7-19) by titration of a 5-ml aliqµot with 0.1 N standard
iodine solution, although with reagents of known high purity, this stannous
solution may be diluted with 1.2 N HCI (~ 7-20) to 1 liter, giving 0.1 M
Sn++.
7-19. To standardize the stannous solution, an iodine solution is pre-
pared by dissolving 12. 7 gm reagent grade 12 and 40 gm of iodate-free KI
in 25 ml of water. This mixture is stirred until complete solution results
and then is diluted to 1 liter and stored in an amber glass-stoppered bottle.
19 Smith et al., Can. J. Research, B 17: 178 ( 1939).
20 Holman and Pollard, J. Soc. Chem. Ind., T56:339 ( 1937).
PHOSPHORUS DETERMINATIONS FOR SOILS 143
The iodine solution is standardized against a 0.1 N arsenious solution, with
starch as an indicator. The 0.1 N arsenious solution is prepared by weigh-
ing 2.4725 gm of pure, dry arsenious oxide, dissolving it in 20 ml of 1 N
NaOH, then adding 1 N H 2 S04 until the solution is neutral or very slightly
acid to litmus paper, and finally diluting to 500 ml in a volumetric flask.
A 25-ml aliquot is pipetted into a 250-ml conical flask, to which are then
added 50 ml of water, 1 gm of NaHCOa, and a few ml of starch solution
( ~ 3-50). The iodine solution is titrated into the flask until a blue color
results. The iodine solution is then titrated into a 25-ml aliquot of the
chlorostannous acid solution in the presence of starch indicator to a blue
color. An atmosphere of C02 may be maintained over the titration for most
accurate results, as Sn++ is rapidly oxidized in air.
7-20. If the purity of the chlorostannous solution is less than 90 per cent,

and stored in an amber glass bottle under a 1-cm layer

i
a fresh solution should be prepared or a new reagent obtained because the
stannic ion will seriously affect the test. 21 The chlorost.annous solution is
next diluted with 1.2 N HCI to 0.1 M SnCI., · 2 H90, about 1 liter),
mmeral 011, pro-
tected from oxygen by passing the entering air through o traps contain-
ing alkaline pyrogallol, or by means of a CO., generator. The original air
in the bottle is removed by bubbling through \~atftral g;s or C02 • The de-
livery of the chlorostannous acid solution is arranged ·by a siphon or bottom
exit in the bottle, through a dropper tip calibrated to deliver almost ex-
actly 0.15 ml in 3 drops, 22 and protected by a rubber c!l.p (policeman)
when not in use.
7-21. Precaution. A ml or so of the solution in the dropper is run out
and discarded before use after it has stood exposed to the atmosphere for
more than half an hour. To assure preservation of the SnC1 2 solution, it is
desirable to check it occasionally against the standard iodine solution,
which keeps indefinitely in a cool, dark place.

PROCEDURE

7-22. Preparation of Phosphorus Solution. The phosphorus solution

(prepared according to Table 7-2 for the phosphorus standards, from
the extract from a soil~amle, or from stock solution resulting from a
fusion, fertilizer, or pla issue analysis) should have a phosphorus con-
centration in the final 0-ml dilution to be made, of 0.05 to 0.6 ppm P.
This corresponds to 2.5 to 30 ugm of P per 50-ml volume. An aliquot is
placed in a 50-ml volumetric flask. If its pH is appreciably at variance from
pH 3, the reaction is adjusted to pH 3, using 2,4-dinitrophenol indicator23
21 Pierce and Haenisch, Quantitative Analysis (New York: John Wiley & Sons, Inc.,
1940), p. 199.
22 0.25 ml per 50 ml is recommended by Woods and Mellon, Ind. Eng. Chem., A .E.,
13:761 (1941). M~r..e than that leads to turbidity of the colored solution on standing.
23 Zinzadze, /nd. Eng. Chem., A.E., 7:228 (1935).
144 PHOSPHORUS DETERMINA TIONS FOR SOILS
( 0.25 per cent in H 20). To do this, a few drops of indicator are added to
the solution and if it gives a yellow color, 2 N H 2 S04 is added dropwise
until colorless. If the indicator gives a colorless solution, indicating a solu-
tion pH below 3, then 4 N Na 2C03 is added dropwise just until a yellow
color appears and finally 2 N H 2 S04 until yellow becomes faint.
7-23. Development of the Molybdophosphoric Blue Color. The test
solution aliquot (of pH 3) is placed in a 50-ml volumetric flask. Then 2
ml of the sulfomolybdic acid solution is added with the pipet. The solu-
tion is made to volume, mixed, and poured into a drained 125-ml conical
flask. The working temperature is kept at 25° ± 5°C, and the volume
ratios are adhered to strictly, to maintain a 0.4 N H 2 S0 4 solution, 0.1 per
cent in ammonium molybdate. Next, 3 drops ( 0.15 ml) of the chloro-
stannous acid reductant solution is added, and the solution is mixed
thoroughly. The full color intensity develops in 3 to 4 minutes and begins
to fade in 10 or 12 minutes. It is read photometrically with a 660-mu light
maximum, within;1hat interval. The blue solution cannot be diluted if too
strongly colored;r... e color is developed in a new aliquot if further dilu-
tion is needed.
7-24. Calibration Curve. The transmission percentage readings of the
standards are plotted ( ~ 11'...'..16). The photo response follows Beer's law
up to 0.4 ppm, or somewhat more with some types of colorimeters. The
transmission percentage readings of the test samples are referred directly
to this calibration curve to obtain the ppm of P in solution in the test
sample. Precautions: For a satisfactory phosphorus procedure, maintain
constant conditions in the blank, standard, and test solutions. Contamina-
tion must be scrupulously avoided (~ 7-10). The standard curve should
be repeated as necessary to obtain all the data points on the straight line
calibration curve and a consistent position of the curve.
7-25. Instead of a photoelectric colorimeter and calibration curve, the
standard colors (0.1 and 0.25 ppm P) may be visually compared to those
in the test solutions. In this case, the development of the standard colors is
delayed until several test solutions are prepared; then the color is developed
in the series of solutions simultaneously," l!nd is compared by means of
comparator tubes or a visual colorimeter (~ 17-42).
CHLOROSTA NNOUS-REDU CED MOL¥:JJDOPHOSPHORIC BLUE
COLOR METHOD, IN HYDROCHLO RIC ACID SYSTEM24
(METHOD II)
(Includes arsenate)
7-26. Method II has approximately one-half the sensitivity of the sul-
furic acid system (method I), and a working range from 0.05 to l ppm of
.· ~·
~--,;;~/'
24·Afte?J)ickman and Bray, Ind. Eng. Chem., A.E., 12:665 (1940), and Woods arid
Mellon, Ind. Eng. Chem., A.E., 13:760 (1941).
PHOSPHORUS DETERMINATION S FOR SOILS 145
P (Table t.-..1). Concentrations of 2000 ppm or more of CI, 15 ppm of
Fe+++, 1000 ppm of Al, 200 ppm of Ca or Mg, 600 ppm oJ SO 4 , 50 ppm
of Cl0 4 , or 25 ppm of N0 3 do not interfere. The method is thus eminently
suited for application in the Na 2 CO:i-HCI system (~ 7-133) for total ele-
mental analysis of soils. Sulfates or other anions (except chloride) in gross
amounts do interfere, just as chloride does in the sulfuric acid system. De-
termination of P can be carried out with several hundred ppm of Fe if
reduced (~ 7-32).

APPARATUS

7-27. Apparatus needed consists of a special storage container for the

chlorostannous acid (~ 7-20), 50-ml and 1000-ml volumetric flasks, 125-
ml conical flasks, a colorimeter with 660-mu light maximum or visual com-
parator tubes, pipets, and a buret.

REAGENTS

7-28. Needed reagents consist of 4 N NH 40H, 4 N HCl, 2, 4-dinitro-

phenol indicator, standard phosphorus solution (Table 7-2), and the fol-
lowing special reagents.
7-29. Chloromolybdic Acid Reagent, 1.5 Per Cent. Exactly 15.0 gm of
c.p. ammonium molybdate, (NH 4 )flMo 70 24 • 4 H 2 0, is dissolved in about
300 ml of distilled water warmed to about 50°C, and the solution is filtered
to remove sediment, if necessary. The molybdate solution is cooled and 350
ml of 10.0 N HCI is added slowly with rapid stirring. When this solution
~as cooled again to room temperature, it is diluted with distilled water to
~· 1000 ml in a volumetric flask, mixed thoroughly, and stored in an
amber, glass-stoppered bottle. This is a l.5 per cent ammonium molybdate
mlution in 3.5 N HCI; it should be replaced every 2 months.
!f"i 7-30. Chloro~tan~ous Acid Reductant. The same reagent is. employed
a"or the sulfunc acid system (Method I) except that 0.25 ml 1s used per
50 ml of colored solution.

PROCEDURE

7-31. Development of the Molybdophosphoric Blue Color. An aliquot

of the phosphorus-containing test solution. is pipetted into a 50-ml volu-
metric flask, adjusted to pH 3 with 4 N NH 4 0H or 4 N HC1, 2,4-dinitro-
phenol being used as an indicator (becomes yellow as pH 3 is approached
from the acid side). Then 10 ml of the chloromolybdic acid solution is
added by means of a pipet. The solution is diluted to the 50-ml volume
mark, mixed and poured into a drained 125-ml conical flask. The tem-
perature throughout the determination is maintained at 25° ± 5°C. Next,
5 drops (0.25 ml) of the chlorostannous acid reductant solution is added
fe the solution in the flask and mixed thoroughly. The color intensity is
146 PHOSPHORUS DETERMINATIONS FOR SOILS
nearly constant between 4 and 20 minutes, and is read phBtometrically
after 5 minutes with a 660-mu light. For greater rapidity, the solution may
of course be mixed in 25 x 200-mm test tubes calibrated for the appro-
priate volume.

ALTERNATIVE PROCEDURES

7-32. Reduction of Large Amounts of Ferric Iron. If greater than 15

ppm Fe+++ in the final dilution are encountered, an aliquot of the acid
test solution is passed through a small Jones reductor. A convenient size
is a 15-cm column of amalgamated zinc supported on 5 mm of glass wool
in a 50-ml buret. The solution is rinsed from the reductor with three small
portions of 0.25 N HCI, and the combined effluent is adjusted to pH 3 by
the addition of 4 N NH 4 0H, with care to avoid exceeding that pH.-
otherwise a precipitate forms. If this happens, the turbid solution is dis-
carded and a new aliquot is reduced. When over 30 ppm of iron is pres-
ent, the reduced solution should be used for a blank reading, since it has
an appreciable colpr. The color is developed as in the previous paragraph.
7-33. Elimination of Fluoride Interference25 in the Molybdenum Blue
Reaction. Fluoride ions, if present in greater than about 5 ppm, 20 produce
a negative interference with the molybdenum blue reaction used for the
determination of phosphorus. The interference of F ion is eliminated by
the addition of boric acid,
4 F- + H 3B03 + 3 H+ ~ (BF 4 ) - + 3 HP (7-2)
( fluorohoratP. VPry s1ight1y ionizt'd)

Neither the boric acid ions or the fluoroborate interfere. An aliquot of·1he
test solution, containing less than 0.15 mole of fluoride ion is pipetted into
a 50-ml volumetric flask, and 15 ml of 0.8 M H:1B0 3 ( 50 gm of H 3B03 per
liter) is added. The color is then developed as usual (~ 7-31 ). A less
convenient method 27 is the removal of F by evaporation with HCI0 4 , the
excess HCI0 4 being neutralized before the determination of phosphorus.

MOLYBDENUM-REDUCED MOLYBDOPHOSPHORIC BLUE

COLOR METHOD, IN SULFURIC ACID SYSTEM28
(METHOD nn
(With reduction to exclude arsenate)
7-34. Method III is about one-third as sensitive as Method I, and re-
quires heating for 30 minutes for color development (Table 7-1). The

25 Kurtz, Ind. Eng. Chem., A.E., 14:855 (1942).

26Woods and Mellon, Ind. Eng. Chem., A.E., 13:762 (1941 ).
27 Robinson, Ind. Eng. Chem., A.E., 13:465 (1941 ).
28After Zinzadze, Zts. Pflanzenerdhr. Dung. Bodenk., 16A:l29 (1930), Ann.
Agron., n.s., 1:321 (1931), Ind. Eng. Chem., A.E., 7:227 (1935), as modified by
Gerritz, J.A.O.A.C., 23:321 (1940), and Peech et al., U.S.D.A. Cir. 757, p. 3 (1947).
PHOSPHORUS DETERMINATION S FOR SOILS 147
color, onccf developed, is stable for 24 hours. Over 2, 25, JOO, and 25 ppm
respectively of iron, arsenate, arsenite, and silicate interfere. 2 ~

APPARATUS
7-35. Needed apparatus includes an 800-ml Kjeldahl flask and heating
rack, a 50-ml volumetric flask, a 125-ml conical flask, pipets, a buret, a
I 00°C water bath, a colorimeter with 660-mu light maximum, and tubes.

REAGENTS
7-36. Needed reagents consist of 99.5 to 100 per cent pure and phos-
phorus-free Mo0 3 , 99 .5 to I 00 per cent pure Mo metal powder (passed
an 80-mu or 200 meshes per inch sieve), concentrated and 2 N H 2 SO 4 , 0.1
N KMn0 4 , 4 N Na 2 CO:i• 0.25 per cent 2,4-dinitrophenol, standard phos-
phorus solution (Table 7-2), and the following special reagents.
7-37. Sulfomolybdic Acid "Molybdenum Blue"ao (Reagent A). Approxi-
mately 19.5 gm of Mo0 3 is placed in an 800-ml Kjeldahl flask and 500
ml of 36 N H 2 S0 4 added. The mixture is heated gently with occasional
mixing until solution is complete, and then is cooled to J 50°C. Next, 1.25
gm of finely powdered Mo metal is added. The temperature is kept at
140° to l 50°C, and the solution is mixed vigorously until all of the metal
is dissolved except possibly a few large particles. The solution is cooled
and a 5-ml aliquot (a previously wetted pipet is used and the aliquot is
washed out to transfer the proper aliquot of this viscous solution) is placed
in a 125-ml conical flask. Next, 20 ml of distillea water is added and the
solution is titrated with 0.1 N KMnO 4 to a pink that persists for 1 minute.
The reagent should be 0.1 1 N; if less than 0.109, a calculated amount of
the Mo powder is added and dissolved by reheating in the Kjeldahl flask
at l 50°C. Evolution of fumes and loss of acid concentration is avoided.
The reagent is cooled, then transferred to a dark Pyrex, gla~toppered
bottle without dilution. This solution keeps indefinitely.
7-38. Dilute Sulfomolybdic Acid "Molybdenum Blue":n (Reagent B).
One volume of reagent A is diluted with 3 volumes of water and cooled.
Since this reagent deteriorates rapidly upon standing, it is freshly prepared
as needed.
7-39. Sodium Bisulfite-H 2S04 Solution, 8 Per Cent. Approximately 40
gm of NaHS0 3 (meta, powder) is dissolved in 500 ml of 1.0 N H 2 S0 4 •
This reagent is freshly prepared each week.

w Woods and Mellon, Ind. Eng. Chem., A.E., 13:763 (1941).

:io Peech et al., U.S.D.A. Cir. 757 (1947).
a1 The concentration of molybdate expressed as (NH 4 ) 6 Mo7024 • 4 H20 is l.28
per cent; the hexavalent molybdate, similarly, expressed is 1.19 per cent. Thus, a lower
molybdate concentration is employed in method III than in methods I and II.
148 PHOSPHORUS DETERMINATIONS FOR SOILS
PROCEDURE
~·
7-40. Preparation of the Phosphorus Solution. The appropriate volume
of the standard phosphorus solution (Table 7-2), the blank, and an ex-
tract from the soil sample, fertilizer, or plant tissue analysis, is taken to
give 0.2 to 3 ppm of P (final concentration) in a 50-ml volumetric flask.
This corresponds to 10 to 150 ugm P. If the pH is appreciably at variance
from pH 3, the reaction is adjusted to pH 3, using 2,4-dinitrophenol indi-
cator (0.25 per cent in H 2 0), with 2 N H 2S04 added dropwise just to
colorless or 4 N Na 2 C0 3 added dropwise just to a faint yellow.
7-41. Reduction of Arsenic to Arsenious and Ferric to Ferrous. The
phosphorus solution (of pH 3) is diluted to a 35-ml mark on the flask and
then 4 ml of the sodium bisulfite-H 2 S04 solution is added.:i 2 The solution
is mixed with a swirling motion and heated in a water bath at 100°C for
40 minutes. This treatment reduces As 2 0r, to As 2 0a, the latter form of
which does not give the heteropoly blue reaction, but does not reduce the
P 2 0 5 • Ferric iron is, within limits, reduced to ferrous. Arsenate should be
limited33 to 25 ppm and iron to 2 ppm.
7-42. Development of Molybdophosphoric Blue Color. To each flask
containing a 39-ml total volume of phosphorus-containing solution, at
I 00°C, 2 ml of the dilute molybdenum blue (reagent B) is added by means
of a pipet, the stream being directed to the center of the flask, not being
allowed to run down the sides. The solution is quickly mixed by swirling,
and the heating is continued for 25 minutes. Next the solution is cooled in
cold running water, diluted to 50 ml, and mixed. The color is stable for as
long as 24 hours and is measured at 660 mu. A calibration curve is pre-
pared in the usual way, with transmission percentage plotted on a log scale
against the P concentration on a linear scale. If the color in a test sample
is too strong, the determination is repeated, since dilution of the final solu-
tion is not permissible.

1, 2, 4-AMINONAPHTHOLSULFONIC ACID-REDUCED
MOLYBDOPHOSPHORIC BLUE COLOR METHOD, IN
PERCHLORIC ACID SYSTEM34
(METHOD IV)
(With a modification to exclude arsenate)
7-43. Method IV is unique in that the perchloric acid is added separately
to the test solution near the end of the formulation, the reductant and

112zinzadze, Ind. Eng. Chem., A.E., 7:221 (1935), added 5 ml of NaHS0:1 follow-
ing separate addition of 5 ml of I N H 2 S04 .
SB Woods and Mellon, Ind. Eng. Chem. A.E., 13:763 (1941).
34After Sherman, Ind. Eng. Chem., A.E., 14:182 (1942), as modified from King,
Biochem. J., 26:292 (1932), the reductant of Fiske and Subbarow, J. Biol. Chem.,
66: 375 (1925), being employed.
PHOSPHORUS DETERMINAT IONS FOR SOILS 149
sodium sulfite for arsenate elimination having been mixed previously. Fi-
nally the molybdate is added to develop the color. This permits separate
adjustment of the final acidity in solutions containing an appreciable resid-
ual acidity.The total acidity is 0.9 N in the final solution, which is 22 per
cent greater than that used by King (0.7 N), the latter being too low to be
in the acid-stability plateau ( ~ 7-5) for amounts of phosphorus above 2
ppm.
7-44. Method IV is about one-sixth as sensitive as method I, ranging
0.4 to 2.4 ppm (Table 7-1). The chief advantage of method IV is that
200 parts per million of ferric iron do not interfere with the development
of the blue color, and it is thus eminently suited to the determination of
total phosphorus (~ 7-134) of soils following perchloric acid digestion of
the sample.x" The iron gives the solution a greenish cast, but that effect is
eliminated by the light filter. Ti and V do not interfere, nor does Mg from
the magnesium nitrate ashing procedure. Silica and nitrate are eliminated
by HC10 4 predigestion.

APPARATUS

7-45. Needed apparatus includes a 5-ml pipet, 2 burets, 50-ml volu-

metric flasks, a colorimeter with 660-mu light maximum, and tubes. Addi-
tional apparatus needed for the purification of the reagents includes a 2-
liter conical flask and 32-cm suction filter apparatus.

REAGENTS

7-46. Required reagents include 60 per cent and 2 N HC10 4 , 2 N

Na 2 C0: 1, 0.25 per cent 2,4-dinitrophenol indicator, and standard phos-
phorus solutions (Table 7-2). To exclude arsenates, NaHSO:i crystals are
needed as well. To purify the reagents, concentrated HCl, NaHSO:l'
Na 2 SO:i• and 95 per cent ethanol are needed. The following special reagents
are prepared freshly every 2 weeks.
7-47. Ammonium Molybdate, 5 Per Cent. Five gm of ammonium molyb-
date are dissolved in 100 ml of distilled water, and allowed to stand over
night. The solution is then filtered into an amber glass bottle.
7-48. 1, 2, 4-Aminonaphtholsulfonic Acid Reductant. The commercial
reagent I-amino, 2-naphthol, 4-sulfonic acid (Eastman Kodak Co.) is re-
crystallized as follows: ~ 0 to 1000 ml of water in a 2-liter conical flask,
heated to 90°C, in which there have been dissolved 150 gm of NaHS0 3 and
10 gm of crystalline Na 2 S0a, 15 gm of the crude 1,2,4-aminonaphthol-
sulfonic acid is added. The flask is stoppered and shaken until all but the
"amorphous" impurity is dissolved. Next the hot solution is filtered through
a large filter paper (about 32 cm), and the filtrate is then cooled promptly

35 Bray and Kurtz, Soil Sci., 59: 39 (1945).

36FiskeandSubb arow,J.Bio. Chem., 66:375 (1925).
150 PHOSPHORUS DETERMINAT IONS FOR SOILS
under a water tap. Then I 0 ml of concentrated HCl is added to the solu-
tion. Finally the solution is filtered with suction and washed with about 300
ml of water followed by 95 per cent ethanol until the washings are color-
less. The purified aminonaphtholsulfonic acid is dried in air with the least
possible exposure to light, then powdered and transferred to a brown bottle.
7-49. The 1,2,4-aminonaphtholsulfonic acid may be synthesized:17 from
B-naphthol and the crystalline product so prepared, while still wet on the
filter, may be washed further with 95 per cent ethanol as long as any color is
extracted.
7-50. The reductant reagent38 is prepared by mixing 0.125 gm of the
purified 1,2,4-aminonaphtholsulfonic acid crystals with 44 ml of 15 per
cent NaHS0 3 solution in a dark glass stopped bottle. A 20 per cent solu-
tion Na:iS0:1 is then added drop by drop until the solution is clear, usually
requiring 5 to 7 ml.

PROCEDURE

7-51. Development of the Molybdophosphoric Blue Color. An aliquot

of phosphorus solution (less than 25 ml containing 20 to 120 ugm of P,
0.4 to 2.4 ppm in the 50 ml) is pipetted into a 50-ml volumetric llask. The
pH, if appreciably at variance from pH 3, is adjusted to pH 3 with 2 N
Na 2 CO:i or 2 N HC10 4 , 2,4-dinitrophenol being used as an indicator. (In
certain routines, the amount of HC10 4 in the aliquot, if accurately known,
may be subtracted from that to be added and the pH adjustment dis-
pensed with.) Then 5 ml of 60 per cent HClO, is added and the solution
volume is adjusted to approximately 40 ml. The solution is thoroughly
mixed after each successive reagent addition. The temperature is maintained
at 25° ± 4°C, throughout the color development. Next, 1.6 ml of the
1,2,4-aminonaphtholsulfonic acid reagent is added. Exactly 15 minutes
before the colorimetric reading, 4.0 ml of the ammonium molybdate solu-
tion is added, the volume quickly adjusted to 50 ml with distilled water,
and the solution mixed. At the end of 15 minutes, transmission percentage
is read with a 660-mu light maximum. The photo response is linear to 2.4
ppm.
7-52. A convenient routine 39 is to prepare as many as 20 or 30 samples
at once through the addition of the aminonaphtholsulfonic acid. Then,
when the colorimeter is ready, the ammonium molybdate is added at in-
tervals of 2 minutes; and the samples are made to volume and mixed. When
7 samples are prepared ( 14 minutes), the first is read. Subsequently addi-
tional samples are prepared, and readings are taken alternately, so that the
15-minute reading interval is maintained between color development and

B7Folin,J. Biol. Chem., 51:386 (1922).

as Sherman, Ind. Eng. Chem., A.E., 14: 182 ( 1942).
811 Cole, Ph.D. Thesis, Department of Soils, University of Wisconsin ( 1950).
 PHOSPHORUS DETERMINATION S FOR SOILS 151
reading. Carefully matched tubes (or the same tube) are used to avoid
variance. A standard solution and blank are included with .each run to re-
check the calibration curve.
7-53. Modification to Exclude Arsenate. 40 To eliminate arsenates, which
may be present in amounts equal to phosphorus in soils sprayed with
arsenic to control insects, an aliquot of the test solution is placed in the
volumetric tlask, and the perchloric acid is added as in the regular de-
termination. Then 0.8 gm of crystalline NaHS0:1 is added by means of a
scoop of appropriate size. The neck of the tlask is washed down with water
and the total volume brought to 40 ml. The flask is agitated a little to
hasten solution of the solid, and the solution is allowed to stand on the
bench for 3 hours, after which the color is developed (,i 7-5 I). Heat is
unnecessary 41 for the reduction of arsenate in the presence of perchloric
acid; in fact the solution must not be heated because it affects the reading.

ALTERNATIVE PROCEDURE

7-54. Modification for Mid-Range of Acid-Stability Plateau in Sulfuric

Acid System 4 ~ (Method IVa). Although method IV (above) utilizes 0.9 N
HC10 4 , which is well within the acid-stability plateau, phosphorus of
plants has been determined at 43 the mid-range of acid-stability (~ 7-5).
In acid digestions for the oxidation of organic matter of tissue, the residual
acidity may be utilized as a part of the total acidity of the formulation,
without detailed adjustments for losses by volatilization.
7-55. To eliminate the effect of arsenate from the sample or contamina-
tion from glassware, approximately 0.03 gm of crystalline NaHSOa is added
to a 5-ml aliquot of plant digest solution; the solution is heated to 80°C
for at least 30 minutes and cooled. The total acidity is adjusted to give
1.3 N H~S0 4 in the final 50-ml volume, after which the determination is
carried out according to method IV44 (~ 7-5 I). The color is read at 25
minutes; it varies slowly after this.

V ANADOMOLYBDO PHOSPHORIC YELLOW COLOR METHOD,

IN NITRIC ACID SYSTEM
(METHOD V)
7-56. The exact nature of the yellow chrnmogen of the vanadomolybdo-
phosphoric system (method V) is not known, but the color is attributed to

40Sherman, /nd. Eng. Chem., A.E., 14:182 (1942).

41 Sherman, Ind. Eng. Chem., A.E .. 14:182 (1942).
42Woods and Mellon, Ind. Eng. Chem .. A.E.. 13:762 (1941).
4aCotton, Ind. Eng. Chem., A.E.. 17:734 (1945).
44 Woods and Mellon employed the equivalent of 5 ml of 5 per cent ammonium
molybdate per 50 ml, but stated that the amount was not critical; Cotton, according
to King, Biochem. J., 26:292 (1932), employed 3.3 ml, and Sherman (method IV)
employed 4 ml of 5 per cent per 50 ml final volume.
152 PHOSPHORUS DETERMINATIONS FOR SOILS
substitution 4 ~
of oxyvanadium and oxymolybdenum radicals for the 0 of
P0 4 to give a heteropoly compound that is chromogenic. The method has
long been used in steel analysis 40 and biological materials. 47 The advan-
tages of method V are extreme simplicity, lower sensitivity ( 1 to 20 ppm
of P, or about one-tenth the sensitivity of the blue-color methods, Table
7-1, permitting applications on a more macro scale), stability of color,
freedom from interferences with a wide range of ionic species in concentra-
tions up to 1000 ppm, and adaptability to HN0:1, HCI, H~S0 4 , or HCI0 4
systems. Ions that do not interfere in concentrations up to 1000 ppm are:
Al, Fe+++, Be, Mg, Ca, Ba, Sr, Li, Na, K, NH 4 , Cd, Mn, Pb, Hg+, or
Hg++, Sn++, Zn, Cu and Ni only by a change in hue, Ag, U, Zr, OAc, As
(ite), Br, CO:l' Cl0 4 , cyanide, pyrophosphate, molybdate, tetraborate,
selenate, iodate, silicate, nitrate, nitrite, sulfate, sulfite, benzoate, citrate,
oxalate, lactate, tartrate, formate, and salicylate. Positive interference is
caused by silica and arsenic with heating; negative interferences are caused
by As (ate), F, Th, and Bi; blue color is caused by Fe++ (but without in-
terference up to I 00 ppm), sulfide, thiosulfate, thiocyanate, or excess
molybdate. Chloride over 75 ppm interferes in HN0:1 system.
APPARATUS

7-57. Required apparatus includes a buret, 50-ml and 1000-ml volu-

metric flasks, and a colorimeter with facilities for light maxima at 400 to
490 mu (variable according to sensitivity needed,~ 7-61).
REAGENTS

7-58. Needed reagents include standard phosphorus solutions (Table

7-2), and the following special combined HNO~-vanadate-molybdate 48
reagent: Solution A is prepared by dissolution of 25 gm of ammonium
molybdate in 400 ml of water. Then solution B is prepared by dissolving
1.25 gm of ammonium metavanadate in 300 ml of boiling water. Solution B
is cooled and then 250 ml of concentrated HN0:1 is added and the solu-
tion is again cooled to room temperature. Finally solution A is poured into
solution B and the mixture is diluted to 1 liter.
PROCEDURE

7-59. Preparation of the Test Solution. An aliquot of the phosphorus

solution (standard, or test sample from plant tissue, soil, fertilizer, or rock
analysis) that contains 0.05 to 1.0 mgm of P, in acid not exceeding the
equivalent of 0.2 N in the final 50 ml, is placed in a 50-ml volumetric
flask in a volume not exceeding 35 ml.

411 After Kitson and Mellon, Ind. Eng. Chem., A.E.. 16:379 (1944).
4H Mission, Chem. Ztg., 32:633 ( 1908).
4 7 Koenig and Johnson, Ind. Eng. Chem., A.E., 14: 155 (1942).
48 After Barton, Anal. Chem., 20: 1068 ( 1948).
PHOSPHORUS DETERMINATIONS FOR SOILS 153
7-60. Acidity. The acid concentration in the determination is not critical,
but the final concentration of 0.5 N is recommended. It must be above
0.2 N (to eliminate the yellow nitric acid color), but not over 1.6 N above
which there is less color produced. Color development is slowed up in the
higher portion of this range; and to obtain full color development in 5
minutes, the acidity should be less than I N. The acidity obtained from the
combined reagent in the following procedure is approximately 0.8 N. Al-
though this system is in HNOa, equivalent systems in HCl, H 2 S04 , or
HC10 4 are satisfactory.
7-61. Vanadomolybdophosphoric Yellow Color. To the phosphorus so-
lution in a 50-ml volumetric flask, 10 ml of the vanadomolybdate reagent
is added, and the solution is diluted to 50 ml with distilled water and mixed
well. The color develops rapidly but is usually read after 10 minutes to as-
sure full strength. A blank must be prepared and read with the samples be-
cause the blank vanadate color itself is noticeable even when phosphorus
impurities have been carefully excluded. Variation in room temperature has a
negligible effect on the color intensity. The color is read in the colorimeter
with a light maximum from 400 to 490 mu, according to the sensitivity
needed. The sensitivity varies I 0-fold with wave lengths from 400 to 490
mu, but ferric ion causes interference with the lower wave lengths, par-
ticularly at 400 mu. The 4 70 filter is generally employed. Concentration
ranges for different light maxima are:
Range, ppm P Light maximum, mu
0.75-5.5 400
2.0--15 440
4-17 470
7-20 490
When ferric ion is low enough not to interfere, a convenient practice is to
plot a family of calibration curves of 1 series of solution concentrations
with a number of light maxima. A suitable selection of light maximum can
be then made with any given test solution according to the transmission
percentage found. This permits a wide latitude in the concentrations in one
series of determinations.

ALTERNATIVE PROCEDURES

7-62. A procedure 49 employing 2 reagent solutions has been widely em-

ployed. A 0.25 per cent ammonium metavanadate solution is prepared in
dilute HNOa by dissolving 2.5 gm of NH4 VOa in 500 ml of boiling water,
cooling the solution somewhat and adding 20 ml of concentrated HNOa-
The solution is cooled, diluted to 1 liter, and stored in an amber glass
bottle. The vanadate concentration is moderately critical (within 10 per
cent); the effect of an excess is a slight vanadate reagent color at a similar

49 Kitson and Mellon, Ind. Eng. Chem., A.E., 16:379 (1944).

154 PHOSPHORUS DETERMINATIONS FOR SOILS
wave length of absorption. A 5 per cent solution in water of ammonium
molybdate is prepared by dissolving 25 gm of (NH4 )nMoP 24 • 4 H 20 in
400 ml of distilled water warmed to 50°C. The solution is cooled and
filtered if cloudy, and diluted to 500 ml, mixed, and stored in an amber
glass bottle. The molybdate concentration is not critical but should be held
constant for a given series of determinations.
7-63. To the phosphorus solution (with acid equivalenf•0 to 0.5 N fig-
ured on the basis of a final volume of 50 ml, but contained in a solution not
to exceed 35 ml) 5 ml of 0.25 per cent ammonium vanadate solution is
added. The solution is mixed, 5 ml of 5 per cent ammonium molybdate
solution is added, and the mixing is repeated. The solution is made to vol-
ume and mixed well. The yellow color is read as before.
7-64. Phosphorus Determination After Removal of Arsenic by Bromide
Distillation. Gross amounts of arsenic are removed from the solution in
which the phosphorus is to be determined 51 rather than being reduced to
suppress the arsenic-derived color as applicable to small amounts of As
(~ 7-6). To remove the As as AsBra, a IO-ml aliquot of the clear slightly
acid test solution is placed in a 70-ml Kjeldahl flask, 10 ml of 48 per cent
HBr is added, and the contents of the flask is brought just to dryness with
moderate heating. Then the neck of the flask is washed down with IO ml of
1.5 N HN0 3 , and the contents is boiled gently until colorless. The cooled
solution is transferred to a 50-ml volumetric flask and brought to about
35 ml, and the vanadomolybdophosphoric yellow color is developed and
read (~ 7-61).

DILUTE ACID SOLUBLE PHOSPHORUS OF SOILS

7-65. A number of dilute acid extraction methods have been proposed
for measuring a fraction of the soil phosphorus ("available" phosphorus),
that could be correlated with field response of crops. The degree of correla-
tion varies according to the nature of the soil. In general, the phosphorus
extractable from acid soils by dilute solutions of strong acids can be cor-
related with crop yield response to phosphate fertilizers. The correlation
is lower with neutral soils, but little or no correlation is obtained for soils
that are alkaline, calcareous, or that have been fertilized with phosphate
rock. The strongly dissociated acids that have been employed by different
investigators52 include 0.5 N H 2S0 4 (~ 7-74), 0.2 N HN0 3 , 53 and 0.002
50 Obtained by the addition of 5 ml of 5 N HN03 less the equivalent of residual
acidity if over 0.1 N. ·
»l Rubins and Dean, Soil Sci., 63:392 ( 1947), found that the usual molybdophos-
phoric blue color methods are not satisfactory following this removal of arsenic used
in large amounts as a replacing reagent. Residual impurities, particularly Sb, present in
most commercially obtainable arsenates, apparently interfered.
52 Reviewed in detail by Nelson et al., "'Soil and Fertilizer Phosphorus," Agronomy
(New York: Academic Press, 1953), Vol. 4, ch. 6. p. 153.
53Fraps, Tex. Agr. Exp. Sta. Bui. 126 (1909).
 PHOSPHO RUS DETERM INATION S FOR SOILS 155
N H 2 S0 4 54corresponding to a range from pH 0.7 to pH 3. Solutions of
weakly dissociated acids (~ 7-73) have been used for extraction of phos-
phorus in efforts to correlate with crop yield responses to phosphate fer-
tiliza.tion. The procedure for pH 3 extractable phosphorus is given in this
section. A higher correlation of soil test with crop response is obtained with
dilute acid-fluoride extraction (~ 7-86) or NaHCOa extraction (~ 7-99)
when all soils are considered, including acid, neutral, alkaline, and calcare-
ous soils.
7-66. Effect of Drying the Soil. Freshly taken field-moist samples are
recommended for measurement of the chemical conditions existing in the
field. Storage in the moist condition may decrease the dilute acid ex-
tractable phosphorus. At the same time air-drying the sample, especially
if the soil is acid, may increase the extractable phosphorus test found by
10 to 30 per cent. Oven-drying the sample, especially if acid, may increase
the extractable phosphorus found by over 100 per cent. With limed and
highly fertilized soils, T. M. Yu and the author found the dilute acid ex-
tractable phosphorus was decreased by drying.

APPARATUS

7-67. Needed apparatus includes a 2-mm (10 meshes per inch) sieve, a
drying oven, moisture dishes, an analytical balance, 500-ml conical extrac-
tion flasks fitted with rubber stoppers, a rotary shaker, I 0-cm filter funnels
and phosphorus-free filter paper, and a 50-ml graduated cylinder.

REAGENTS

7-68. Extraction Reagent. A 1>tock solution of exactly 0.1 N H!.!SO 4 is

prepared by titration against standard alkali. A convenient volume of this
is diluted to 0.002 N and buffered by addition of 3 gm of ( NH 4 ) 2 SO 4 or
K2 S04 per liter to produce a pH of 3 in the final solution.

PROCEDURE

7-69. Extraction of Soil. 55 The freshly taken field-moist soil sample is

passed through a screen having 2-mm openings (or 10 meshes per inch),
mixed, bottled tightly, and the moisture content determined by oven-drying.
Then a sample is weighed out equivalent to 1.000 gm oven-dry soil and
placed in 200 ml of the buffered 0.002 N H!.!S0 4 extraction solution in a
500-ml conical flask. The suspension is shaken on the rotary shaker for
exactly one-half hour. The suspension is immediately filtered through a
retentive, P-free paper (the first of the filtrate being discarded, the clear
filtrate saved). Then 50-ml of the clear filtrate is taken for determination of

54 Truog, J. Am. Soc. Agron., 22:874 ( 1930).

5f>After Truog, J. Am. Soc. Agron., 22:874 (1930), modified for application to
freshly taken field-moist soils samples.
156 PHOSPHORUS DETERMINATION S FOR SOILS
the molybdophosphoric blue color ( ~i 7-22). For soils extremely high in pH
3 extractable phosphorus (over 100 ppm), a 10-ml aliquot of this solution
is diluted to 50 ml prior to development of the color.
7-70. Interpretations. The concentration of Pin the test solution is read
as ppm from the standard curve. Then:
. . total ml extraction solution used
ppm Pin soil =ppm P m so1ution x · . ···-··--·····--~-
gm soil extracted
(pH ,3 extractable)
= ppm P in solution x 200 (7-3)
when 1 gm of soil and 200 ml of extraction solution are employed. For
slightly acid to acid soils tested, under 5 ppm is very low in terms of crop
response to phosphate fertilizers, 5 to 15 ppm is low, 15 to 25 ppm is
medium, and over 25 ppm is adequate to high.

ALTERNATIVE PROCEDURES

7-71. Air-dry soil samples may be tested for pH 3 extractable phos-

phorus, but a change from the field analysis should be recognized.
7-72. A modification of the Truog 1 : 200 soil to extractant ratio was
employed by Peech et al. 56 It consisted of doubling the soil suspension con-
centration to a 1 : 100 ratio. The phosphorus was determined by method
III (~ 7-40). Peech et al. also suggest extraction of coarse textured soils
after they have been through a 1-mm sieve, in order to decrease the sam-
pling error. The results were corrected to the 2-mm basis by assuming no
P to be extractable from the 1 to 2 fraction.
7-73. The list of weak acids used for the extraction of soil phosphorus
includes saturated carbonic acidr. 7 (usually employed with calcareous
soils), dilute acetic acid~ 8 of pH 4.8 buffered with NaOAc, and 1 per cent
citric acid. 50 Stelly and Pierre 110 employed the whole range of pH values
from acid to alkaline to characterize the soil inorganic phosphorus. Cal-
cium phosphate of phosphate rock dissolves in extraction solutions of pH
3 or lower. 61 Even neutral ammonium citrate solution dissolves 5 to 34
per cent of the phosphate of phosphate rock, but 2 per cent citric acid dis-
solves 2 to 3 times more. 62 Water soluble and neutral ammonium citrate

GU U.S.D.A. Cir. 757 (1947), p. 3.

0 70aubeny, Roy. Soc. (London) Phil. Trans. 179 (1845); McGeorge, Ariz. Agr.
Exp. Sta. Tech. Bui. 82 (1939); Ensminger and Larson, Soil Sci., 58:253 ( 1944);
Smith, J. Am. Soc. Agron .. 40: 1045 (1948).
r.s Morgan, Conn. Agr. Exp. Sta. Bui. 392 (1937); Spurway and Lawton, Mich. Agr.
Exp. Sta. Bui. 132 ( 1949); Lunt et al., Conn. Agr. Exp. Sta. Bui. 541 (1950).
r.o Dyer, J. Chem. Soc .. T65: 115 ( 1894); Wiley, Principles and Practices of Agri-
cultural Analysis, 3rd. ed., Vol. 1 (Easton, Pa.: Chemical Publishing Co., 1926),
p. 429.
so S.S.S.A. Proc., 7. 139 (1943).
61 Fraps and Fudge, J. Am. Soc. Agron., 37: 532 (1945).
62 Jacob et al., J.A.O.A.C., 30:529 (1947), 19:449 (1936).
 PHOSPHORUS DETERMINAT IONS FOR SOILS 157
soluble phosphate is defined 63 as the available phosphate of fertilizers. The
use of citrate with Na 2 S2 0 4 reduction is presented below (~ 7-105).
7-74. Extraction of Calcium Phosphate of Soil in 0.5 N H 2S04 • In the
soil phosphorus fractionation system (~ 7-106), the 1-gm soil sample,
from which aluminum phosphate(~ 7-92) and iron phosphate (~ 7-104)
have been extracted, is washed twice with 25-ml portions of saturated
NaCl solution. It is then extracted with 50 ml of 0.5 N H~S0 4 on a rotary
shaker for I hour 114 and the suspension is centrifuged. The supernatant
solution is decanted into a clean, dry flask and the soil sample is saved for
further extraction with citrate-Na~S~0 4 (~ 7-109). To test for noninter-
ference of Fe in the P determination, 2 identical aliquots (usually J to 5 ml
each) are taken from the decanted solution. To I aliquot, enough standard
P solution ( ~ 7- I 2) is added to give 0.2 ppm final concentration of added
P. Then Pis determined in both aliquots (,[ 7-22). The complete recovery
of the 0.2 ppm of P (difference between the 2 determinations) establishes
noninterference (usually the case) by Fe present. Low recovery, indicating
Fe interference, can be eliminated by adding chlorostannous acid (~
7-120) or using a Jones reductor (~ 7-132).

WATER SOLUBLE AND pH 3 EXTRACTAB LE PHOSPHORUS

IN RUNOFF WATERS
7-75. The water soluble and "loosely bound" (,i 7-91) phosphorus in
runoff is obtained in the clear solution after flocculation of the soil with
NaCl solution. The concentration of P is determined by means of method
I ( ~ 7-22). The pH 3 extractable phosphorus is then determined with a
200 to I extractant : solids ratio. The sum of the water soluble phosphorus
and the pH 3 extractable phosphorus is considered to be the available
phosphorus in runoff.

APPARATUS
7-76. Needed are a graduated cylinder cut to contain 60-ml; a 50-ml
and other graduated cylinders, a 250-ml conical flask, a centrifuge and
100-ml centrifuge tube, and apparatus for the colorimetric determinations
(,I 7-15).

REAGENTS
7-77. Jn addition to the reagents regularly used in the determination of
available phosphorus (,I 7-16), 5 N NaCl is needed. It is prepared by dis-
solving 293 gm of NaCl in water and making the solution volume to 1
liter.

na Methods of A na/ysis. 7th ed. (Washington, D.C.: A.0.A.C., 1950), p. I 0.

64 Dean,/. Agr. Sci. 28:234 (1938).
158 PHOSPHORUS DETERMINATIONS FOR SOILS

PROCEDURE

7-78. The Sample. The amount of runoff is calculated which will give
between 100 and 200 mgm of solids, based on a separate determination
of the solids content ( ~ I 0-66). The first 60-ml increment of well-mixed
runoff is poured into a 100-ml centrifuge tube, and 10 ml of 5 N NaCl is
added and mixed. The tube is placed in a hot water bath until the suspen-
sion is flocculated, and the solids are then thrown down by centrifugation
for 5 minutes at 2000 rpm. Of the clear supernatant liquid, 50 ml is de-
canted into a graduated cylinder and the rest is discarded. The soluble phos-
phorus as ppm Pin solution is determined (~i 7-22) with a separate cali-
bration curve made up with 1 N NaCl present in the standards, since the
slope of the curve is affected by chloride.
7-79. If less than 100 mgm of solids is obtained in the first 60-ml of
runoff, 1 or more additional increments are taken and flocculated as be-
fore. The solids are thrown down as before, in the same centrifuge tube,
the clear supernatant solutions being discarded.
7-80. Extraction of pH 3 Extractable Phosphorus. The pH 3 extract-
able phosphorus is extracted with a 200 : I ratio of the extractant volume
to solids weight exactly as for soils (~ 7-69). A variable volume of extract-
ing solution is employed, 20 ml per 100 mgm of solids. The extraction is
carried out in the 100-ml centrifuge tube if the extractant volume is under
75 ml, otherwise in a 250-ml conical flask. If less than 50 ml of extractant
is employed, the solution is diluted with extraction solution to 50 ml after
extraction, and the dilution factor represented is then ( 50 / ml extractant
used in the determination). The ppm P in the pH 3 extraction solution is
determined by reference to the calibration curve.
7-81. Calculations. The soluble phosphorus is calculated, as pounds per
acre inch (ppai) of runoff:

ppa1. = ppm p.m soluhon

. 7 x 0.227
x .~ (7-4)
( soluhk P) (curve) 6
The ppai is multiplied by the correction factor ( ~ 10-69) . The parts of
soluble P per 2 million solids ( pp2ms) is calculated:

2ms = ppai (solubl~_~) x 2 000 000 (7-5)

pp - ppai (solids) ' '
Similarly, for the pH 3 extractable P:
pp2ms = ppm P in solution x 200 x 2 x dilution factor (7-6)
(extractable P ) (curve)

in which the dilution factor is 1.0 if 50 ml of extractant was analyzed

(otherwise as defined in ~ 7-80). Finally, the pounds of extractable P per
acre inch of runoff is calculated:
PHOSPHORUS DETERMINATION S FOR SOILS 159
ppai = ppai x J>p2ms ( extract~~~--p)
( flX!ractahle P) ( •olid•) 2,000,000 (7-7)

FLUORIDE EXTRACT ABLE PHOSPHORUS OF SOILS

7-82. The F- ion has the special property of complexing AI+++ and
Fe+++ ions in acid solution, with consequent release of phosphorus held
in the soil by these trivalent ions. 6" The reaction in acid solution may be
represented as follows:
3 NH 4 F + 3 HF+ AlP0 4 ~ H:iP0 4 + (NH 4 )aAlF0 (7-8)
3 NH 4 F + 3 HF+ FeP0 4 ~ H:iP0 4 + (NH 4 ):iFeF0 (7-9)
The formula AlP0 4 represents various hydrated and hydroxyl phosphates
of aluminum, including any adsorbed or precipitated surface layers on
oxides and aluminosilicates. The formula FeP04 similarly represents vari-
ous hydrated and hydroxyl phosphates of iron, including adsorbed or pre-
cipitated surface layers on iron oxides. The reaction with FeP0 4 (eq. 7-9)
does not go to completion for concretionary or iron-oxide coated iron phos-
phate (~[ 7-105). Also the fluoroferrate is unstable in neutral or alkaline
system, so phosphorus from AlP0 4 can be fractionated by 0.5 N NH 4 F
extraction (~ 7-91 ). The (NH 4 )aAIF6 may precipitate with large excess of
fluoride.
7-83. The dilute acid-fluoride procedure1111 given here for extraction of
available phosphorus of soils with 0.03 N NH 4 F in 0.025 N HCI (Bray and
Kurtz, No. t solution) has been employed widely and has been found to
give results that are highly correlated with crop response to phosphate
fertilization. Inclusion of acid results in the dissolution of the more active
calcium phosphates and prevents the precipitation (as calcium phosphate)
of phosphorus dissolved by equations 7-8 and 7-9. The method removes, in
the words of Bray and Kurtz, "proportional parts of [or] . . . the more
readily soluble portion of each form of available soil phosphorus." Also,
"the term 'available forms' [of soil phosphorus] is restricted to those which
are of most immediate significance to crop growth and whose variations in
amount are responsible for variation in crop growth and response to added
phosphates."

APPARATUS

7-84. Needed apparatus includes a 50-ml extraction bottle or tube fitted

with stopper; 20-ml, 2-ml and 1-ml pi pets (preferably automatic); a
filter tube or funnel and 7-cm filter paper (Whatman No. 42), and a col-
orimeter with 660-mu light maximum and comparator tubes.

ur, Swenson et al., Soil Sci., 67:3 (1949); Turner and Rice. Soil Sci., 74: 141 (1952).
1111 Bray and Kurtz, Soil Sci., 59: 39 ( 1945).
160 PHOSPHORUS DETERMINATIONS FOR SOILS

REAGENTS

7-85. Needed reagents include the 0.03 N NH 4 F in 0.025 N HCl ex-

traction solution ( 1.11 gm of solid NH 4 F and 4.16 ml of 6 N HCI per
liter) and ammonium molybdate, in HCI as specified for method II (~
7-28). The extraction solution may be stored in a glass container for a year
without appreciable deterioration. Also needed is a stannous chloride stock
solution prepared by dissolution of 10 gm of reagent grade SnC1 2 • 2 H~O
in 25 ml of concentrated HCI. Kept in a dark, tightly stoppered bottle, this
solution need be replaced only at 2-month intervals. For each 4-hours work,
a freshly diluted stannous chloride solution is prepared by pipetting I ml
of the stock solution into 330 ml of freshly boiled, cooled, distilled water.

PROCEDUREli7

7-86. Dilute Fluoride-Dilute Acid Soluble P. A 2.85-gm sample of

crushed, sieved soil is weighed out (or measured, about 2.5 ml) into a 50-
ml extraction bottle or tube. Then 20 ml of the extraction solution (0.03
N NH 4 F in 0.025 N HCI) is added from a pipet (preferably arranged as
an automatic dispenser), and the bottle is stoppered and shaken for 1
minute. 68 The suspension is immediately filtered on a moist Whatman No.
42 filter paper held in a filter tube or funnel. The filtrate should be clear. If
not, the solution is quickly poured back through the same filter. A 2-ml
aliquottm of the clear filtrate is transferred to a colorimeter test tube by
means of an automatic pipet, after I ml has previously been discarded
to rinse the pipet. Then 5 ml of distilled water is added. The determina-
tion is next carried out essentially in accord with method II (~ 7-31). First
2 ml of ammonium molybdate ( chloromolybdic acid) reagent (~ 7-85) is
added and the solution is mixed, conveniently accomplished in routine by
an air jet stirrer. Finally, 1 ml of freshly diluted stannous chloride reagent
(~ 7-85) is added with immediate mixing. After 5 to 6 minutes and before
15 to 20 minutes the color is read in a colorimeter. The P standards are
made up in the range of 0.1 to 1 ppm of P through the same steps as in
this procedure, including 2 ml of the extraction solution in each I 0 ml final
volume since the F- content of the final solution is slightly above the 5
ppm limitrn of noninterference (~ 7-33) and would be expected to shift
the curve slightly. A reagent blank is made with each series of determina-
tions and is employed for the 100 per cent transmission setting.
7-87. The concentration of dilute fluoride-dilute acid soluble soil phos-
phorus is calculated:

117Adapted from Bray and Kurtz. Soil Sci., 59: 39 ( 1945).

us More dilute soil : extractant ratios also have been employed ( ~i 7-88).
au For unusually high or low tests another aliquot size is employed to give proper
range of P concentration; the aliquot volume plus the distilled water must total 7 ml.
70 Woods and Mellon, Ind. Eng. Chem., A.E., 13:762 (1941 ).
PHOSPHORUS DETERMINATIONS FOR SOILS 161

ppm of P in soil = ppm of P in solution x1 x 20

2 °
2.85
= ppm of Pin solution x 35 (7-10)
As a general guide to crop response, below 3 ppm is very low, 3 to 7 ppm
is low, 7 to 20 is medium, and above 20 ppm is adequate to high. Different
crops shift the scale some because of different root characteristics.
ALTERNATIVE PROCEDURES

7-88. A soil extractant ratio of 3 : 20 is substituted, for the 2.85 : 20

( 1 : 7) ratio of Bray and Kurtz, by the Nebraska Soil Testing Laboratory. 71
A 5-minute shaking time is also employed. Soil extractant ratios of 1 : 10
and 1 : 50 72 also have been employed, and the more dilute extraction is
thought to raise the correlation with yield response on some kinds of soils.
7-89. 0.1 N HCI and 0.03 N NH 4 F. To include more of the soil apatite
(for example, that in near neutral or calcareous soils or that remaining
from phosphate rock additions) in the extraction, a higher concentration of
HCl is included (Bray and Kurtz, No. 2, acid-soluble and "adsorbed" phos-
phorus). The procedure is the same (~ 7-86) except that the 0.1 N HCl
and a 40-second shaking period are employed. Some workers have em-
ployed 0.2 N HCl instead of 0.1 N HCl and considered that all of the
apatite was thereby dissolved during the 40-second shaking period.
7-90. For a "quick test," the soil is allowed to settle with either No. 1
or No. 2 solution instead of being filtered, and the color is developed in the
clear supernatant by the addition of 0.30 ml of concentrated molybdate and
stirring with a tin rod. The color range of this undiluted solution is usually
too intense for use in a colorimeter.
7-91. Extraction of Aluminum Phosphate of Soil in Neutral 0.5 N
NH 1F. In neutral or alkaline solutions, the fluoride complex of Al (~ 7-82)
forms, but that of iron does not to any appreciable extent. Neither is cal-
cium phosphate appreciably dissolved in this reagent (~ 7-106). There-
fore, the phosphate extracted by neutral 0.5 N NH 4 F7a is mainly AIP0 4 ,
and the method is therefore an important step in the soil phosphorus frac-
tionation system (~ 7-106). A I-gm sample of soil is placed in a 100-ml
centrifuge tube and extracted with 50 ml of 1 N NH 4 Cl for 30 minutes on
a rotary shaker to remove water soluble and loosely bound phosphorus and
exchangeable calcium. The suspension is centrifuged, and the supernatant
solution is discarded. Water soluble P could be determined on this solu-
tion, but the amount is generally very low.
11 Information courtesy Dr. R. A. Olson, Agron. Dept., University of Nebraska,
Lincoln.
12 Information courtesy Dr. Floyd Smith, Agron. Dept., Kansas State College,
Manhattan. Kansas.
73 Reagent of Bray and Dickman, S.S.S.A. Proc., 6:312 (1942) and Bray and Kurtz,
Soil Sci., 59:39 (1945), extractant No. 3, adapted to a somewhat different procedure
162 PHOSPHORUS DETERMINATIONS FOR SOILS
7-92. To the NH 4 -soil in the centrifuge tube, 50 ml of neutral 0.5 N
NH 4 F ( 18.5 gm solid NH 4 F per liter of water adjusted to pH 7 and stored
in a wax-lined bottle) is added and the suspension is shaken for 1 hour.
The suspension is then centrifuged and an aliquot (usually 10 ml) is taken
for analysis. The soil sample is saved for the iron phosphate extraction in
0.1 N NaOH (~ 7-104). The extract may be slightly colored with organic
matter, but a slight color usually does not interfere with readings at 660 mu.
With strong coloration, an aliquot may be prepared without the reductant
and used as a blank (the reagent blank in this case is handled separately
and the ppm P therein is subtracted from the analysis). In extreme cases,
the organic matter may be flocculated in HCI and filtered off, with or with-
out the aid of aqua regia-purified charcoal. Boric acid is added (~[ 7-33) to
remove fluoride interference, and the total volume is made to 50 ml during
the P analysis. Then:
ppm P m . s01·1 = ppm p·m so Iut10n . x - 1-.-50
- - -1
(neutral NH.,F <·xtractnblc) a tquot, m
x ,'11_l_~_!r-~_~tio!l solution (7-11)
gm soil extracted
The aluminum phosphate removed includes much that is only slowly avail-
able to crops, and therefore the results are not as highly correlated with
yield responses to phosphate fertilization as the dilute acid-fluoride extrac-
tion (~ 7-86). It is correlated with phosphate released by liming acid soils
and the resultant low response to phosphate fertilizer on such soils.
7-93. Acid-0.5 N NH 4 F Extraction. A separate 1-gm soil sample is
shaken in 50 ml of 0.1 N HCI for 30 minutes. Then 1 gm of solid NH 4 F is
added and the shaking is continued for an additional 60 minutes (Bray and
Kurtz, No. 4). The suspension is filtered on Whatman No. 2 paper, and
the P is determined on a I 0-ml aliquot in the presence of H:iBOa ( ~ 7-33).
Calcium phosphate and some of the iron phosphate (~ 7-82) is included
with the aluminum phosphate by this extraction.

HYDROXYL AND CARBONATE EXTRACTABLE PHOSPHORUS

OF SOILS
7-94. Two general types of alkaline extraction of soil phosphorus have
been employed: (a) moderately alkaline solutions such as sodium bicarbon-
ate74 or potassium carbonate (~ 7-103) have been employed to extract a
portion of the soil phosphorus that could be correlated with the response of
crops to the addition of phosphate fertilizers, and (b) strongly alkaline
solutions such as NaOH have been employed for the extraction of phos-
phorus from aluminum and iron phosphates (~ 7-104) and organic forms
of soil phosphorus ( ~ 7-112) . The organic phosphorus extracted by al-

74 Olsen et al., U.S.D.A. Cir. 939 ( 1954).

PHOSPHORUS DETERMINATIONS FOR SOILS 163
kaline solutions is flocculated and excluded from the analysis of the inor-
ganic phosphorus brought into solution.
7-95. The 0.5 N NaHC0 3 method employs a solution of pH 8.5, de-
signed to control the ionic activity of Ca, through the solubility product of
CaC0 3 , during the extraction of calcareous soils. As the carbonate activity
in the soil is raised by this solution, the calcium activity is decreased. Thus
some phosphate from the surface of calcium phosphates is extracted
through tho: solubility product of calcium phosphate. As Ca activity de-
creases, pf.osphate activity increases. The importance of buffering car-
bonate during extraction is illustrated by the 2 trends produced by carbonic
acid in calcareous soils; (a) a trend toward increased solubility of calcium
phosphate as expected with an acid, and (b) a trend toward decreased
solubility of calcium phosphate owing to the increased calcium activity as
CaC0,1 is dissolved by the carbonic acid. The reagent also extracts some
phosphate ( ~ 7-143) from the surface of aluminum and iron phosphates,
which are more abundant (~ 7-106) in acid and neutral soils. By repres-
sion of the Al and Fe activities (the Al, 75 by aluminate complex formation
and Fe, 76 by precipitation as the oxide) the phosphate activity is increased.
By the solubility product principle, 77 the activity of phosphate ions must
rise as aA 1 and a~, .. decrease in the presence of AIP0 4 and FeP0 4 • Also, the
carbonate ion added in the reagent, by the solubility product of CaCOa,
maintains the Ca activity low enough in all soils (acid, neutral, or alkaline)
to prevent reprecipitation of the liberated phosphate as calcium phosphate.
The secondary precipitation of calcium phosphate during the extraction is
a problem with dilute acid extractions (~ 7-73) including C02 extractions.
The procedure of Olsen et al., with NaHC0 3 can be likened to the brief
extraction with dilute acid-dilute fluoride (~ 7-86), since the most reactive
(high specific surface) phosphate is extracted from phosphates of alumi-
num, iron, and calcium.

APPARATUS

7-96. Needed apparatus consists of a weathered 250- or 300-ml soft

glass extraction bottle or conical flask fitted with a stopper, an end-over-
end shaker, a filter funnel and Whatman No. 40 filter paper, a 50-ml volu-
metric flask, and a l 0-ml pi pet.

REAGENTS

7-97. Needed reagents include the extraction solution, activated char-

coal, and the colorimetric phosphorus reagents. The extraction solution is
0.5 M NaHCO:i ( 42.0 gm per liter) adjusted to pH 8.5 with NaOH. The

7r. Cole and Jackson, S.S.S.A. Proc., 15:84 (1951 ).

711 Chang and Jackson, S.S.S.A. Proc., 21: 265 ( 1957).
77 Kittrick and Jackson, J. Soil Sci., 7:81 (1956).
164 PHOSPHORUS DETERMINATIONS FOR SOILS
pH of the solution tends to increase on standing exposed to the atmosphere,
but this change can be prevented by the addition of a layer of mineral oil.
To remove phosphorus from the activated charcoal (Darco G-60), it is
preleached with NaHCO:i and then is washed with water and dried.
7-98. The Dickman and Bray ch\oromolybdic acid reagent (~ 7-29) is
modified by the addition of 60 ml more of I 0 N HCl per liter (for a total
of 410 ml) to neutralize the NaHCOa in the 10-ml aliquot. The 0.1 M
chlorostannous acid (~ 7-30) is employed.

PROCEDURE

7-99. In the procedure of Olsen et al., a 5-gm sample of soil is sus-

pended in 100 ml of the NaHCOa extraction solution of pH 8.5 along with
1 teaspoon of carbon black. The suspension is shaken for a period of 30
minutes. The solution is filtered through a Whatman No. 40 or other
suitable filter paper. lf the filtrate is not clear, it is returned to the soil and
more carbon black is added to the flask followed by a quick shaking and a
second filtration.
7-100. A 10-ml aliquot of clear filtrate is pipetted into a 50-m\ volu-
metric flask. The P is determined by method II (~I 7-31) except that the
modified molybdate reagent ( ~ 7-98) containing extra HCI is employed.
Ten ml of the acid molybdate is added down the side of the flask contain-
ing the aliquot, and the flask is allowed to stand quietly Jest too rapid evo-
lution of C0 2 result in some loss of the solution. After C0 2 evolution has
subsided (following a final swirling of the solution), the neck of the flask
is washed down, and the solution is diluted to 40-ml volume. Then 0.25
ml ( 5 drops) of the 0. I M chlorostannous acid solution is added, followed
by immediate shaking, dilution to volume, and mixing. Failure to observe
these precautions with respect to the chlorostannous acid addition causes
erratic results.
8Z-1JH. he concentration of P found is above the range of the method,
a~li less than l 0 ml is taken, and additional extraction solution is
·~llllJ';iru/ make up a total of l 0 ml of NaHC0:1 solution, in order to main-
n:· the proper acidity during color development. The standard curve is
prepared with the same quantity of NaHCOa included.
7-102. The quantity of P extracted is calculated as ppm of the soil:

ppm of P m . x -50 x -100

. soi.1 = ppm o.fP'm sol ution (7-12)
10 5
As a general guide to crop response, 5 ppm of P ( 25 pounds P ~O" per
acre) is low, and response to phosphate is likely, 5 to 10 ppm of P ( 25
to 50 pounds P 20" per acre) is medium and response is probable, and over
10 ppm of P ( 50 pounds P 2 0 5 per acre) is adequate and response is un-
PHOSPHORUS DETERMINATIONS FOR SOILS 165
likely. Values in the range of 18 to 25 ppm of P ( 90 to 125 pounds of
P2 0 5 per acre) are characteristic of fertile soils.

ALTERNATIVE PROCEDURES

7-103. A more alkaline carbonate solution, 1 per cent K 2C03 , was

used 78 to measure the available phosphorus in calcareous soils. With acid
and neutral soils, this method causes the dissolution of so much of the
strong alkali soluble phosphate (~ 7-104) that a poor correlation was ob-
tained with yield response to phosphate fertilization on such soils.rn A 1.5
per cent Na 2 C03 solution was employed 80 to measure the quantity of phos-
phate rock converted to soil phosphate in acid soils. A 0.01 N Na 2 CO:i and
H 3 B0 3 buffer, adjusted to the pH of the soil, was employed 81 for available
phosphorus. An 8-hydroxy quinoline reagent was employed 82 to repress the
activity of Fe and Al in dilute acid extracts of the soil.
7-104. Extraction of Iron Phosphate of Soil in 0.1 N NaOH. In the soil
phosphorus fractionation system (~l 7-106), the I-gm soil sample extracted
with 0.5 N NH 4 F (~ 7-92) is washed twice with 25-ml portions of satu-
rated NaCl solution. It is then extracted with 50 ml of 0. I N NaOH on a
rotary shaker at room temperature for I 7 hours. 83 The soil suspension is
centrifuged for ·at 2400 rpm (and recentrifuged, if necessary) to
obtain a clear supernatan . The solution is decanted into another
centrifuge tube and the soil samp e . ved for the extraction with 0.5 N
H 2 S04 (~I 7-74). Two ml of 2 N H 2 S0 4 ts. to the decantate and then
1 or more drops of concentrated H 2 S0 4 until the organic colloids begin to
flocculate. Then the suspension i~-f,~otrifl!ge~ an aliquot (usually I to
5 ml) of the supernatant liquid is taken for analysis·{~ 7-22).
DITHIONITE-CITRATE EXTRACTABLE PHOSPHORUS OF SOIU!
~
7-105. The residual soil phosp snot extr.acted by succe~·ve fluoride
(~ 7-91 ), alkali (~I 7-104), and a (~ 7-74) treatments- ~round
40 per cent of the soil phosphorus was sometimes termed 84 "in ohible
phosphorus" and attributed to os ·ants in the structure 85 of layer sili-
cate clays. All or nearly all of this residual, so-called "insoluble" phos-
phorus is extracted 86 when the iron oxides are removed by an effective re-

78 Das, Soil Sci., 30:33 (1930).

19 Gardner and Kelley, Soil Sci., 50:91 (1940).
80 Joos and Black, S.S.S.A. Proc .. 15: 69 ( 1951 ) .
HI Rhoades, Nebr. Agri. Exp. Sta. Bui. 113 (1939).
82 Ghani, Ind. J. Agr. Sci., 13 :562 ( 1943 ).
s;; After Williams, J. Agr. Sci .. 40:233 ( 1950), who used 40 ml per gm of soil.
84 Bray and Kurtz, Soil Sci., 59: 39 ( 1945); Allaway and Rhoades, Soil Sci., 72: 119
(1951).
8 5 Marshall, J. Soc. Chem. Ind. (London), 54:393 ( 1935).
su Bauwin and Tyner, S.S.S.A. Proc., 21 :250 ( 1957); Lancaster, Ph.D. thesis, Univ.
Wis. (1954).
166 PHOSPHO RUS DETERM INATION S FOR SOILS

duction-chelation procedure (~ 7-109). Since a reduction-chelation treat-

ment is rather specific for iron oxide removal, the phosphate released by
such treatment is attributable to forms occluded in the iron oxides rather
than to tetrahedral P, and may aptly be termed "reductant soluble" P. Iron
oxides, unlike aluminum hydroxides, are not dissolved by NaOH or neutral
NH 4 F solutions and therefore coatings of iron oxides would be expected to
remain on the surfaces of iron oxide aggregates or concretions during the
process of extraction with these reagents, thus preventing the dissolution of
phosphate occluded in the interior. The phosphate of freshly precipitated
aluminum or iron phosphate is largely extractable by NaOH. Also, phos-
phate accumulated in soil as residues from fertilizers is all or largely ex-
tracted by NaOH-acid treatment. 87 The occluded phosphate thus must be
so insoluble as to be highly unavailable to plants.
7-106. Soil Phosphorus Fractionation System. Soil phosphorus long ago
was classified88 into the categories:
I. Inorganic phosphorus in near neutral soils, probably calcium
phosphate.
2. Inorganic phosphorus in acid soils, presumably in combinatio n with
aluminum and iron.
3. Organic phosphorus compounds .
Fractionation was carried out 811 by 0.2 N NaOAc, 0.25 N NaOH, and 0.5
N H 2 S0 4 treatments; and later 90 by 0.2 N HOAc, cool 0.25 N NaOH and
2 N H 2 S0 4 treatments. Also,. functional acid and alkali treatments were ap-
plied.111. A soil phosphorus fractionation procedure92 (flow sheet, Fig. 7-2)
has been developed which measures discretely the total of each chemical
form, aluminum phosphate(~ 7-91 ), iron phosphate (~ 7-104), calcium
phosphate (~ 7-74 ), and reductant soluble phosphate (~ 7-109). It is
now known 9 ~ that calcium, aluminum, and iron phosphates, and reductant
soluble phosphates occur in varying proportions in most neutral, alkaline,
and acid soils. Controls with the different chemical species showed that the
NH 4 F and NaOH treatments must precede the H 2S04 treatment since the
latter removes considerable aluminum and iron phosphate as well as all of
the calcium phosphate. Knowledge of the distribution of soil phosphorus
among chemical forms should prove useful in the development of soil
chemistry, even though the objective of such fractionation is completely
distinct from the measurement of the amount of available phosphate (~
87 Dean, J. Agr. Sci., 28:234 ( 1938).
88 Russell, Soil Conditions and Plant Growth (New York: Longmans Green & Co.
Inc., 1932), p. 238.
HU Dean, I. Agr. Sci., 28:234 ( 1938).
110 Ghani, Ind. J. Agr. Sci., 13: 29 ( 1943).
111 Stelley and Pierre, S.S.S.A. Proc., 7: 139 ( 1943).
112 Chang and Jackson, Soil Sci., 84: 13 3-144 ( 19 57).
113 Chang and Jackson, I. Soil Sci., ( 1958), presented before
the North Central
Branch, American Society of Agronomy, Lafayette, Ind., Aug. 1956.
 (2)
(1)
1.0-gm sample
1.0-gm sample

Na2COa
fusion
112-hr shaking centrifuge

Fusion
Solution Soil cake

(discard) 50 ml 0.51'.!
neutral NH 4 F
1-hr shaking centrifuge

SolutionF
Solution A

Sat. NaCl wash twice

50 ml 0. lti NaOH
17-hr shaking, centrifuge

Sat. NaCl wash twice

0.5 ~ H 2 S0 4
Flocculate centrifuge

Residue Solution B Solution C

(discard) Na-citrate
Na-dithionite
15-min water bath
centrifuge

Solution Soil

Sat. NaCl wash twice

Centrifuge after
each washing

Combined Soil
washings
50 ml 0.5N
neutral NH 4 F
1-hr shaking
centrifuge

Solution D Soil

(discard)

P from P in occluded Total P

P from P from
Fe-phos Ca-phos Fe-phos min & org
Al-phos
Determined by Determined by Determined by
Determined by Determined by Determinedby method I
method I method I method I method II
method II
s is not
Fig. 7-2. Flow sheet for soil phosphoru s fractionati on system. Organic phosphoru
included in this fractionati on, but is determined separately (see Fig. 7-3 ).
167
168 PHOSPHORUS DETERMINATION S FOR SOILS
7-95), involving any chemical form that has high activity owing to its high
specific surface.

APPARATUS

7-107. Needed apparatus includes a centrifuge and 100-ml tubes, a

steam plate, a water bath, a burner, a 15-ml centrifuge tube, a 150-ml
conical flask, a 50-ml volumetric flask, and a colorimeter with 660- and
490-mu light maxima.

REAGENTS

7-108. Needed reagents include 0.3 M sodium citrate (88 gm of tribasic

sodium citrate, Na 8 CuH,,0 7 • 2 H 20 per liter) 1 M NaHCOa (84 gm per
liter), sodium dithionite (Na 2 S2 0 4 , from Amend Drug and Chemical Co.,
New York, or Eastman Kodak Co., Rochester, N.Y.; also called sodium
"hydrosulfite"), 30 per cent P-free H 2 0 2 (obtained by 3 successive treat-
ments with kaolinite 111 ), 0.5 M FeCl 3 ( 135 gm of FeCl 3 • 6 Hp dissolved
in 1 liter of water), and saturated NaCl ( 400 gm shaken in a liter of H 2 0).

PROCEDURE

7-109. Dithionite-Citrate Extraction. 95 In the soil phosphorus frac-

tionation system ( ~ 7-106), the 1-gm soil sample, just previously extracted
in 0.5 N H 2 S04 to remove calcium phosphate (~] 7-74), is washed twice
with 25-ml portions of saturated NaCl solution. It is then suspended in
40 ml of 0.3 M sodium citrate solution and 5 ml of M NaHCOa, and the
suspension is heated in a water bath at 80°C. Then 1.0 gm of Na 2 S2 0 4 is
added with rapid stirring. The suspension is kept at 80°C for 15 minutes
and centrifuged. The supernatant solution is collected in a 100-ml volu-
metric flask. The soil is washed twice with 25-ml portions of saturated NaCl,
the 2 washings being combined with the extract in the I 00-ml volumetric
flask. The solution in the flask is made to volume and an aliquot (usually 1
to 5 ml) is taken for the determining P ( ~l 7-110). Another aliquot is taken
for Fe determination (~ 7-111) if desired. For soils high in iron oxides, the
residue is extracted with neutral NH 4 F (~ 7-91) to remove occluded alumi-
num phosphate or with NaOH (~ 7-104) to remove occluded aluminum-
iron phosphate (barrandite-like).
7-110. Phosphorus Determination. A suitable aliquot (usually 1 to 5
ml) of the dithionite citrate extract (~ 7-109) is placed in a 150-ml conical

94 Chang and Jackson, Sci., 124: 1209 (I 956); this procedure gives nearly full
strength H 2 02 and little loss in volume. Serious losses of concentration occur with
distillation under reduced pressure, Baumann, /. Biol. Chem., 59: 667 ( 1924), and
Dickman and DeTurk, Soil Sci., 45:29 (1938).
11 ~Aguilera and Jackson, S.S.S.A. Proc., 17:359 (1953). 18:223, 350 (1954); pro-
cedure for removal of free iron oxides from soils and clays. Modified for inclusion of
NaHCOa buffer by Mehra and Jackson.
PHOSPHORUS DETERMINATIONS FOR SOILS 169

flask. About 10 ml of distilled water and 5 to 10 ml of the 30 per cent

P-free H 2 0 4 (5 ml for 1- to 2-ml aliquot and 10 ml for 3- to 5-ml aliquot)
are added, and the solution is heated cautiously on a burner. Vigorous
splashing must not occur. The burner is moved under or away from the
flask as needed. One drop of 0.5 M FeCla (or 10 ml of 100 ppm Fe solu-
tion) is added to moderate the oxidation. The cessation of foaming (gas
evolution) and the beginning of ordinary boiling indicate the completion
of oxidation. Complete drying must be avoided before the oxidation is
complete, otherwise the very concentrated H:.P 2 and the high temperature
will ignite the organic matter, leaving some carbon particles. Small amounts
of distilled water are added as necessary. After completion of oxidation,
the solution is boiled for an additional I or 2 minutes, and then dried on a
steam plate. About 10 ml of 2 N NaOH is added. The solution is boiled
for 1 to 2 minutes and digested on a steam plate for 5 minutes. The sus-
pension is poured into a 15-ml centrifuge tube and centrifuged to throw
down the iron oxide precipitate. The supernatant liquid is decanted into a
50-ml volumetric flask. The original 150-ml flask is washed with 10 ml of
water into the same 15-ml centrifuge tube. The precipitate is resuspended
by shaking and again thrown down by centrifugation. The supernatant solu-
tion is poured into the same 50-ml volumetric flask. The washing is re-
peated once more and the supernatant solution placed in the same flask.
The combined supernatant solutions are made to volume, and the phos-
phorus is determined by method I ( ~! 7-22).
7-111. Iron Determination.He An aliquot of the dithionite-citrate extract
(~ 7-109), which will make a final solution containing 0.5 to 3 ppm Fe,
is placed in a 50-ml volumetric flask. Water is added to make 35 ml, then
l drop of 30 per cent H~O~, 5 ml of 6 N HCI, and 5 ml of 20 per cent
KSCN solution are added. The solution is made to volume with water. The
color is measured with a 490-mu light maximum.

ORGANIC PHOSPHORUS OF SOILS

7-112. Organic phosphorus of soils has been extracted and determined
by several methods, but these can be classified into two general types; (a)
alkaline extraction in NaOH or NH 10H after acid pretreatment, and (b)
dilute acid extraction after oxidation of organic matter by H~O~ or ignition.
In either type of method, the phosphorus is determined colorimetrically,
and the difference between the inorganic phosphorus in the extract before
and after oxidation of the organic matter represents organic phosphorus.
In the colorimetric determination of phosphorus, only the inorganic ortho-
phosphate yields the blue or yellow color. In method (a) the oxidation of
organic matter is carried out after extraction, in method (b), before ex-

96 Aguilera and Jackson, S.S.S.A. Proc., 17 :359 ( 1953), 18:223 and 350 ( 1954).
 (1) (2) (3)

1.0 gm sample l.O·gm sample l.O·gm sample

10 ml cone HCI I I
130% H 20 2 150 ml 0.5~
150 ml 0.5N acid NH F !acid NH•F
I - •
r1 !h7~akirig, r1h7!t;a'ki~,,
IB
10 ml cone HCI centrifuge I I centrifuge

~ ['°'":oo A ~So-1-ut
.....io_n_B...,,

50ml water I

Centrifuge
I orgf min
Determined by
I I mikral
Determined by
I
Soil Solution method II method II

I
30 ml 0.5~ NaOH I
I
l hr room temp
I I
centrifuge L J
Difference
-----r---- -
Soil Solution

60 ml 0.5~ NaOH

8 hr warm 90°C
centrifuge

Soil Solution

Combined
solution

Aliquot A Aliquot B

72% HC104
digestion
p p
org & min mineral
Determined· by Determined by
method II method II

Difference

Organic P

Fig. 7-3. Flow sheet for determination of soil organic phosphorus. Dotted lines
indicate an alternative procedure.
170
PHOSPHORUS DETERMINATIONS FOR SOILS 171
traction (Fig. 7-3). Release of phosphorus from mineral form (either be-
fore or after extraction, the latter from suspended mineral colloids) by the
oxidation procedure gives a ( + ) inference with the organic phosphorus de-
termination. Release of orthophosphate from organic matter by hydrolysis
before the initial inorganic phosphorus determination gives a ( - ) inter-
ference with the organic phosphorus determination. The HCl-NaOH ex-
traction, method97 (a), is given in the procedure. The dilute acid extraction,
method (b), after H 20 2 oxidation or ignition, is considered in the alterna-
tive procedures (~ 7-125). The first NaOH extraction is given at room
temperature to minimize the hydrolysis of organic phosphorus.

APPARATUS

7-113. Needed apparatus includes a 100-ml centrifuge tube and cen-

trifuge; 250-ml and 50-ml volumetric flasks; a 50-ml beaker; 15-, 10-, and
5-ml pipets; an oven at 90°C; a suitable fume hood for HC10 4 fumes; an
electric plate; a rubber policeman; and a colorimeter and tubes, with 660-
mu light maximum.
REAGENTS

7-114. Needed reagents include concentrated HCl, 0.5 N NaOH, 72

per cent HC10 4 , 6 N NH 4 0H, 0.5 N HCI, 0.5 per cent p-nitrophenol indi-
cator in water, and the following special reagents.
7-115. Chloromolybdic Acid. Twenty-four gm of ammonium molybdate
is dissolved in 300 ml of water. Then 560 ml of 10 N HC1 is added slowly
with stirring. The solution is then diluted to 1 liter with water.
7-116. Chlorostannous Acid. Forty gm of SnC12 • 2H 2 0 is dissolved in
100 ml of concentrated HCl, and 40 ml of water is added. The solution is
protected from oxidation (~ 7-20).
PROCEDURE

7-117. Extraction. In the procedure of Mehta et al., 1 gm of soil is

placed in a 100-ml centrifuge tube and 10 ml of concentrated HCl is added.
The suspension is heated on a steam plate for 10 minutes (final solution
temperature about 70°C), then removed, and an additional 10 ml of con-
centrated HCl is added and mixed. This suspension is allowed to stand
at room temperature for I hour. Then 50 ml of water is added and mixed.
The suspension is centrifuged, and the clear supernatant liquid is poured
into a 250-ml volumetric flask containing about 50 ml of water.
7-118. Thirty ml of 0.5 N NaOH is added to the tube, the soil is stirred,
and the suspension is allowed to stand at room temperature for 1 hour.
The suspension is then centrifuged and the supernatant liquid is poured
into the volumetric flask containing the acid extract. Then 60 ml of 0.5 N
97 Mehta et al., S.S.S.A. Proc., 18:443 (1954).
172 PHOSPHORUS DETERMINA TIONS FOR SOILS
NaOH is added to the tube, the soil is stirred, and the tube is covered with
an inverted 50-ml beaker, followed by warming in an oven at 90°C for 8
hours. The tube is cooled, the suspension centrifuged, and the supernatant
liquid is poured into the 250-ml flask containing the previous extracts. The
combined extracts are diluted to volume with water and mixed thoroughly.
7-119. Total Phosphorus Extracted. The flask containing the mixed ex-
tracts is shaken thoroughly to suspend the flocculated material, and im-
mediately a 15-ml aliquot is pipetted into a 50-ml beaker. To the aliquot is
added 1 ml of 72 per cent HC10 4 , followed by evaporation to a residue of
HC10 4 on a steam plate. The temperature is raised (with care to avoid
spattering due to a rapid temperature change) until fumes of HC10 4 appear
(use a suitable fume hood for HC10 4 fumes) after which the beaker is
covered with a watch glass to reduce further loss of the acid. The digestion
is continued until the color of the solution no longer changes. When the
temperature of the digest approaches the boiling point, the organic matter
is oxidized rapidly. After oxidation, extracts low in iron usually become
colorless; those high in iron retain a light yellow color. Care is taken not to
dry the solution completely. The beaker is cooled and the contents are
transferred quantitatively to a 50-ml volumetric flask with the aid of a
rubber policeman to insure complete transfer of the silica. The solution is
diluted to volume with water and mixed thoroughly. An aliquot, usually
l 0 ml of the clear supernatant liquid, is pipetted into a test tube graduated
at 35 ml.
7-120. Determination. The phosphorus solution in the test tube grndu-
ated at 35 ml is neutralized with 6 N NH,10H to yellow with p-nitrophenol
and then with 0.5 N HCI just to colorless. Then water is added to 35-ml
volume, followed by addition of 5 ml of chloromolybdic acid and thorough
mixing. Finally 3 drops of chlorostannous acid solution is added, followed
by immediate mixing. A blank is carried through the same steps. The per-
centage transmittance is read with 660-mu light after 10 minutes. If more
than 75 ppm of Fe is present, the quantity of SnCJ 2 is increased to give
noninterference. The standards are made up with the same reagents and
procedure, in the range from 0.05 to 0.6 ppm of P. The total phosphorus
extracted is calculated as ppm of Pin the soil.
7-121. Inorganic Phosphorus Extracted. After the aliquot of the com-
bined soil extracts has been taken for total extracted phosphorus, the sus-
pended material is allowed to flocculate and settle out. Then an aliquot of
the clear, colorless (~ 7-122) supernatant liquid (usually 5 or 10 ml) is
pipetted into a test tube graduated at 35 ml. The inorganic phosphorus is
then determined (~ 7-120) and expressed as ppm of Pin the soil.
7-122. Compensation for Brown Coloration. To compensate98 for brown
color, if present (a problem with peat soils), a separate aliquot is taken
9R Dyer and Wrenshall, Can. J. Research, 168:97 ( 1938).
PHOSPHORUS DETERMINATION S FOR SOILS 173

of the supernatant solution and treated as in the determination (~ 7-120)

except for omission of the SnC1 2 reagent. This solution is then employed for
the I 00 per cent transmission setting for the test sample. (The test sample
may be used for the I 00 per cent transmission setting after the addition of
all reagents except SnC1 2 • The rest point of the galvanometer is immedi-
ately recorded; then the colorimeter is reset at this rest point 10 minutes
after addition of the SnC12 and the test sample is then read.) The optical
density of both the sample P and the Pin the reagents (blank) are repre-
sented in the test reading. The standard phosphorus curve is made with
solutions containing the proper aliquots of all reagents employed in the
determination, including a reagent blank (zero P added). The transmis-
sion percentage is read for each solution including the reagent blank against
a colorless solution including all of the solvents and solutes except SnC1 2 •
The standard curve does not pass through I 00 per cent transmission at
zero P added, to the extent of P impurities in the reagents. The transmission
percentage of the test sample, when referred to the standard curve made up
in this way, excludes the P from the blank and gives the net ppm of P
attributable to the determination.
7-123. Calculation. The organic phosphorus is calculated as follows:
ppm of P = ppm of P - ppm of P (7-13)
(organic) (total extracted) ( inor!(anic)

ALTERNATIVE PROCEDURES

7-124. Soil organic phosphorus has also been extracted with cold 0.1 N
NH 10H,uu with hot 0.5 N NH 4 0H, 100 and with hot 4 per cent NH 4 0H. 101
The NH 4 0H extractions are less complete than the NaOH extraction.
7-125. Organic Phosphorus Estimation by Means of Hydrogen Perox·
ide. 10 ~ One gm of soil having passed a 0.5-mm sieve is placed in a tube with
a 50-ml graduation. (The carbonates should be removed by acidification
prior to application of the H 2 0 2 method to calcareous soils.) To the soil
is added the phosphorus-free H 2 0 2 equivalent to 15 ml of phosphorus-
free10a 30 per cent H 2 0 2 • The suspension is thoroughly mixed and placed
on a steam bath for 0.5 hour. Then to the solution are added 15 ml of
water, 10 ml of 0.5 N HCI, and water to give a 50-ml total volume. The
tube is stopped and shaken for 30 minutes. Then 1 gm of solid NH 4 F is
added and the shaking is continued for 1 hour. Finally the solution is fil-
tered on a small Buchner funnel. An aliquot of 5 or 10 ml of this solution

99 Schollenberger, Soil Sci., 6:365 (1918).

100 Pearson, Ind. Eng. Chem., A.E., 12: 198 (1940).
101 Dyer and Wrenshall, Soil Sci., 51 : 159 (1941).
102 Bray and Kurtz, Soil Sci., 59:39 (1945); as adapted to HCI by Dickman and
Bray, Ind. Eng. Chem .. A.E.. 12:665 (1940) instead of H 2 S04 used by Dickman and
De Turk, Soil Sci., 45: 29 ( 1938).
1oa Chang and Jackscon, Sci., 124: 1209 ( 1956).
174 PHOSP HORUS DETER MINAT IONS FOR SOILS
is placed in a 250-ml beaker, 15 ml of 0.8 M H 3 B0 3 is added, and this
solution is evaporated to dryness. Approximately 10 ml of 0.1 N HCl is
added and the solution is again evaporated to dryness. The residue is taken
up in 10 ml of 0.1 N HCl and the P is determined ( ~ 7-31) on this aliquot
in the presence of the H:iB0 3 therein. The organic phosphorus is taken as
the difference between the phosphorus removed by this procedure and that
by acid-0.5 N NH 4 F without H 2 0 2 (~ 7-93) removed from a duplicate
sample. Disadvantages of the method are that the H 2 0 2 may not release all
of the organic phosphorus and may change the status of the inorganic
phosphorus and thus introduce an error in the difference attributed to or-
ganic phosphorus.
7-126. Ignition Method. Extraction of the phosphorus with acid before
and after ignition has been employed to estimate organic phosphorus. The
increase due to ignition is attributable to organic phosphorus. Complicat-
104

ing factors are the release to acid soluble form of some phospho rus from
mineral form by ignition 105 and the possible hydrolysis of some phospho rus
from organic forms to orthophosphate during the acid extractio n.
7-127. Chemical Characterization of Soil Organic Phosphorus Com-
pounds. The determination of the chemical nature of the various soil organic
phosphorus compounds is also important. The 3 general classes of soil
organic phosphorus compounds include phospholipids, nucleic acids, and
inositol (including phytin and related compou nds). Phospholipids are in-
dicated by the presence of organic phosphorus in the ether and alcohol
extracts of soils and are confirmed by the isolation of choline. The pres-
106

ence of nucleic acids is indicated 107 by the identification, in the hydrolysates

of soil phosphorus preparations, of constituent parts of nucleic acid such
as pentose sugar, cytosine, adenine, guanine, uracil, xanthine, and hypoxan-
thine, as well as H;iP0 4 • Support is also found in the known susceptibility
of nucleic acids to dephosphorylation and the !ability observed of the
nucleic acid portion of soil organic phosphorus compared to the stability
108 of the or-
of inositol-like soil organic phosphorus compounds. Analysis
ganic phosphorus of several groups of soils shows that soil organic phos- _
phorus is abundant in inositol-like compounds. Of these, phytin or inositol
hexaphosphate (precipitated by ferric iron) is the most abundant, and
inositol triphosphate (precipitated by Ca) is next most abundant. How-

104 Legg and Black, S.S.S.A. Proc .. 19: 139 (1955).

10:; Fraps, J. Ind. Eng. Chem .. 3: 335 (1911).
(1913).
106Aso, Coll. Agr. Tokyo Imp. Univ. Bui., 6:277 (1905); Shorey, Bui. 88
"Soil and Fertilizer Phosphor us," Agronom y (New York:
107 Black and Goring,
Academic Press, 1953), Vol. 5, ch. 5, p. 123.
108 Bower, Soil Sci., 59:277 (1945); Yoshi ta, Soil Sci
.. 50: 81 (I 940); Dyer et al.,
Sci., 9 I: 319 (1940); Dyer and Wrensha ll, Soil Sci .. 51: 159 (194 I ) ; Young, Biochem.
( 1915 ).
]., 28: 1435 (1934); Anderson, J. Biol. Chem., 18:441 (1914), 20:463, 475
PHOSPHORU S DETERMINA TIONS FOR SOILS 175

ever, anion exchange chromatography and P 32 tracer techniques 109 show

that the two-thirds or more of the organic phosphorus-containing material
from soil that behaves chemically as phytin does not behave chromato-
graphically as inositol hexaphosphate. The preparations were extracted in
0.5 N NaOH, treated with alkaline hypobromite, precipitated with Fe+++
and Ca, and washed thoroughly in N HCl, the standard procedure for
separation of phytin.

TOTAL ELEMENTAL PHOSPHORU S

7-128. The total content of the element phosphorus in silicates and other
solids can be extracted and determined by several methods: (a) by a
Na 2 CO:i fusion, (b) by perchloric acid digestion followed by colorimetric
or titrimetric determination, ( c) by heating of the soil sample with Mg
( N0:1 ) 2 , followed by titrimetric determinations as the yellow ammonium
molybdophosphate; 110 or (d) by digestion in the HF followed by gravi-
metric determination as the Mg 2 P 2 0 7 • 111 Methods a, b, and d are satisfac-
tory for most soils and rocks, but method c apparently does not extract112
all of the soil phosphorus. The perchloric acid method (~ 7-134) has been
widely used for total P in soils, although Jess recovery is sometimes ob-
tained11 :1 than with the Na 2 C0 3 or HF methods.

APPARATUS

7-129. Needed apparatus consists of a platinum crucible and cover, a

filter funnel and paper, a 50-ml volumetric flask, and pipets.

REAGENTS

7-130. Reagents required are analytical grade Na 2C0 3 , 1 N H 2 S0 4 , and

the reagents for method l (~I 7-14).

PROCEDURE

7-131. Total P of Soils by Na2 C0 3 Fusion. A 0.1- to 1-gm sample is

fused in Na 2 C0 3 (~I 11-104 to 11-106) followed by disintegration of the
melt in distilled water, and filtration to remove the curd containing the iron
and much of the silica.11 4 The filtrate is diluted to a volume appropriate to
give 5 to 30 ugm of P in an aliquot to be taken. The aliquot is placed in a
50-ml volumetric flask, neutralized with 1 N H 2 S04 to pH 3 and brought
to 50 ml by method I (~ 7-22). This method works well for silicious soils

109 Clark and Smith, Soil Sci., 72:353 (1951 ).

110 Methods of Analysis, 7th ed. (Washington, D.C.: A.O.A.C., 1950), p. 37.
111 Washington, Chemical Analysis of Rocks (New York: John Wiley & Sons, Inc.,
1919); Robinson, U.S.D.A. Cir. 139 (1930).
112 Sherman, Ind. Eng. Chem., A.E., 14: 182 (1942).
113 Muir, Analyst, 77:313 (1952).
114 Pearson et al., J. Am. Soc. Agron., 32:685 (1940).
176 PHOSPHO RUS DETERM INATION S FOR SOILS

and rocks, but if the silica does not exceed the quantities of F plus Ca
present (as with highly calcareous soils, rock phosphate, etc.), some of the
Pis retained in the curd and the determination is not successful (~ 7-132).
ALTERNATIVE PROCEDURES

7-132. An alternative Na 2C0 8 procedure, employed (1945) by S. C.

Chang and the author, is to take up Na 2C0 3 fusion cake (~ 11-105) in 30
ml of 9 N H 2 S0 4 , with care to avoid loss by effervescence. The crucible and
cover are boiled in a little 2 N H 2S04 in a separate small beaker and this
solution is added to the main solution. The solution and suspended ma-
terial are then transferred to a 200-ml volumetric flask, made to volume,
and mixed, giving approximately 1 N H:.?S0 4 • To eliminate ferric iron, if
present in interfering amounts (a test is given in ~I 7-74), an aliquot of the
clear supernatan t liquid remaining after the silica settles out is passed
through a small amalgamated Zn Jones reductor ( ~ 11-144), made up in
a 50-ml buret, followed by determination by method I (~I 7-14). Full
recoveries of phosphorus were obtained at a concentration of 0.25 ppm
of P in the presence of 100 to 1000 ppm of iron, after reduction in the
Jones reductor in from 0.25 to 2 N H 2 S0 4 •
7-133. Use of HCI to decompose the melt and method II (~ 7-26) per-
mits up to 15 ppm of ferric before interference occurs, and the upper limit
can be extended by the use of a Jones reductor (~ 7-32 or 7-120). Hydra-
zine sulfate has been employed 115 as the reducing agent and ferric iron did
not cause interference up to 150 ppm.
7-134. HCl0 4 Method for Total P. 116 A 2-gm sample of soil (5 gm of
soil low in phosphoru s) that has passed an 0.5-mm sieve is weighed and
transferred to a 300-ml conical flask. A smaller sample of a rock material
high in phosphate is employed. If the sample is high in organic matter, 20
ml of HN0 8 is added to effect preliminary oxidation on a steam plate. For
ordinary soil or rock, fairly low in organic matter, the HNO.~ treatment is
omitted. Then 30 ml of 60 per cent HCl0 4 is added and the digestion is
carried out at 130°C in a special digestion apparatus (~I 12-23) designed
to remove HCI0 4 fumes. A funnel can also be used to reflux the HC10 4
during the digestion in the flask.
7-135. The HC10 4 digestion of the sample is carried out until the solu-
tion appears.colorless, with a slight increase of temperature if necessary.
Usually about 40 minutes of digestion suffices. As the digestion is com-
pleted, dense white fumes of HC10 4 appear and the silica becomes white.
Additional HC10 4 may be employed to wash down any dark particles that
stick to the sides of the flask.
115Sheldon and Harper, Iowa State Coll. J. Sci., 40, 4:403 (1941).
116 Digestion in HCI04 was used by Volk and Jones, S.S.S.A. Proc., 2: 197 (1938),
with a titrimetric procedure; Sherman, Ind. Eng. Chem., A .E., 14: 182 (1942), with
method IV; and Bray and Kurtz, Soil Sci., 59:39 (1945), with method III.
PHOSPHOR US DETERMIN ATIONS FOR SOILS 177

7-136. When the digestion is completed, the flask is removed. When it

has cooled sufficiently to avoid spattering, 50 ml of distilled wa~er is added,
and the solution is transferred through a filter to a 200-ml volumetric flask.
The residue is washed to bring the volume of solution to the mark. An
aliquot is taken for analysis by the vanadomolybdophosphoric (~ 7-59) or
the perchloric acid ·methods ( ~ 7-51 ) . The silica has been removed by the
HC10 4 treatment, and ferric iron does not generally interfere.

PRECIPITA TION AND IDENTIFIC ATION OF ALUMINU M

AND IRON PHOSPHAT E
7-137. It is often desirable to prepare precipitates of amorphous or
crystalline phosphates of aluminum or iron for study in the laboratory or
greenhouse. Characterization of the solid phase~• precipitated can be ef-
fected by X-ray diffraction, elemental, thermal, and infrared analyses. The
precipitates may be compared in behavior to soil phosphates by the several
extraction methods given above ( ~ 7-106).

APPARATUS

7-138. Needed apparatus includes 1-liter beakers, pipets. burets, a

steam plate, a centrifuge and tubes, and apparatus for characterization of
the precipitate~.

REAGENTS

7-139. Needed reagents consist of 1 M AlC1 3 , 1 M FeC1:1, 1 M NaH 2P0 4 ,

0.1 N Na Cl, and acetone.

PROCEDURE

7-140. Aluminum Phosphates. To precipitate aluminum phosphate, 500

ml of distilled water is placed in a 1-liter beaker and 30 ml of I M AlCla
is added. Then, with stirring, 90 ml of 1 M NaH 2 P0 4 is added. The beaker
is covered with a cover glass and placed on the steam plate. After a period
of heating, a precipitate of aluminum phosphate begins to form. Digestion
for about 40 hours makes the precipitated crystalline. To determine the
identity of the solid phase, the suspension is stirred, an aliquot is removed,
and the precipitate is thrown down by centrifugation, washed twice with
0.1 N NaCl to remove saloid-bound phosphate, then with water until free
of chlorides, and finally with acetone. The precipitate gives an analytical
composition of AIP0 4 • 2 H 2 0 or Al(OH) 2 H 2 P0 4 , and an X-ray diffrac-
tion pattern of the mineral variscite. To follow the process of precipitation
and crystallization, the pH values of the unmixed and mixed solutions are
measured at the beginning (about pH 2.2) a net at 8-hour intervals (a slight
decrease occurs with time). Also at 8-hour intervals, a portion of the pre-
cipitate is removed, washed, and X-rayed. Usually a precipitate having the
 178 PHOSPHORUS DETERMINATION S FOR SOILS
analytical composition AlP0 4 • I 1,;2 H 2 0 and a distinctive X-ray diffrac-
tion pattern appears first, then gradually disappears as variscite becomes
predominant. Removal of fine colloidal material from the sample by size
separation aids in the separation of pure crystalline variscite in the coarser
fraction. Use of low pH values hastens the growth of crystals (greater
solubility of the phosphate) and prevents the precipitation of hydroxides or
oxides of aluminum.
7-141. Iron Phosphates. To 540 ml of water in a 1-liter beaker, 10 ml
of 1 M FeCl 3 is added and then, with rapid stirring, 30 ml of 1 M
NaH 2 P0 4 • A precipitate forms rapidly and becomes crystalline with diges-
tion on the steam plate for about 24 hours. The composition is FeP04 •
2 H 2 0, or Fe(OH) 2 H 2 P0 4 , and the crystalline diffraction pattern is that
of the mineral strengite. The precipitate is washed as for the aluminum
phosphates and is then ready for X-ray diffraction analysis and other tests.

ALTERNATIVE PROCEDURES

7-142. Inclusion of a K or NH 4 ions in the mixture results in forma-

tion of crystalline products with these cations contained in the crystals; for
example, crystals of minerals such as minyulite and taranakite. There is an
extensive series of relatively insoluble aluminum and iron phosphates, 117
the compositions of which include the ions K, NH 4 , Na, H, and Ca in
varying proportions depending on the ions present and their source, con-
centrations, acidity, and the temperature of the precipitation solution. Col-
loids of Fe and Al oxides, hydroxides, and silicates, when placed in phos-
phate solutions at steam plate temperatures, result in crystalline iron and
aluminum phosphate precipitates. 118

PHOSPHATE EXCHANGE CAPACITY OF SOILS

7-143. Of the plant nutrient anions, only phosphate shows much anion
exchange in soils. Sulfate may exchange some, but chlorides and nitrates
exchange little or none. Anion exchange, therefore, must be thought of as
having a different general character from cation exchange, which occurs
more or less the same for all of the common cations. Phosphate "ex-
change," "adsorption," or "precipitation" involves change of the phosphate
from solution to solid phase (~ 7-95), and when the phase change is ex-
amined in detail, any distinction in the meaning of the 3 terms as applied
to the phosphate soil system is difficult and probably useless. The fraction
11 7 Haseman et al., Soil Sci., 70:257 (1950), S.S.S.A. Proc., 15:76 (1951); Cole
and Jackson, S.S.S.A. Proc., 15:84 (1951), J. Phys. Colloid Chem., 54:128 (1950);
Larsen, Am. Mineral., 25:315 (1940); McConnell, Am. Mineral., 25:719 (1940),
Bui. Geo/. Soc. Am., 54:707 (1945), J. Geo/., 58:16 (1950).
118 Kittrick and Jackson, Sci., 120:508 (1954); S.S.S.A. Proc., 19:292 and 455
(1955),Soi/Sci., 79:415 (1955);J.Soi/Sci., 7:81 (1956).
PHOSPHORUS DETERMINATIONS FOR SOILS 179
of fixed phosphate that is available for "exchange" can be calculated from
the specific surface (particle size) of the phosphate phase. 119 -The amounts
of phosphate present on surfaces can be measured by the p:i 2 exchange. 120
The reaction rates increase with temperature and therefore chemical bond-
ing rather than physical (the usual adsorption) bonding is responsible. 121
The phase change appears to be controlled by the formation of chemical
bonds with Fe, Al, and Ca ions that are on mineral surfaces or in solution.
All of these cations are known to form insoluble phosphate salts, and the
"exchange" is of the nature of a chemical double decomposition reaction.
Authors agree unanimously that measurement of phosphate exchange ca-
pacity (phosphate fixation capacity) is dependent upon the conditions such
as time, reference pH, and pretreatment, and that these conditions must be
stated with the report of phosphate exchange capacity.
7-144. To eliminate precipitation of phosphate by divalent cations, the
soil was prewashed 122 with 0.5 N NaOAc of pH 5. 7, saturated with the 0.5
M salt solution of the anion to be tested, and ad1usted to pH 5.7. To Ca soil
(from which the free CaCOa has been removed), the cation exchange
equivalent of dilute H 3 P0 4 (on the basis of the first hydrogen) was added 123
to measure the phosphate reaction. The iron and aluminum extracted from
acid soils in 0.5 M citric acid was shown 124 to be related to the phosphate
exchange capacity as measured by J. S. Hosking 125 with 0.5 M KH 2 P0 4
plus HaP0 4 at pH 3.4.

APPARATUS

7-145. Needed apparatus includes 400- and 250-ml beakers, a 500-ml

volumetric flask, a 250-ml conical flask, HC10 4 digestion apparatus (~
12-23), a filter funnel and Whatman No. 40 or 41 and 42 filter papers, a
25-ml pipet, and the apparatus for colorimetric phosphorus determina-
tion.

REAGENTS

7-146. Needed reagents are 0.5 M citric acid (91 gm per liter), 1 and
20 per cent NaCl, 60 per cent HCI0 4 , 0.03 M KH2 P04 , approximately N
NaOH and N HCI, 0.1 per cent methyl orange indicator (~ 3-42), and
the reagents for colorimetric phosphorus determination.

11 9 Kittrick and Jackson, J. Soil Sci., 7:81 (1956).

120 Olsen and Watonabe, S.S.S.A. Proc., 21: 144 (1957).
121 Hemwall,Soi/Sci., 83:101 (1957).
122 Rubins and Dean, Soil Sci., 63: 389 ( 1947).
12:! Mehlich, Soil Sci., 66:429 (1948).
124 Bass and Sieling, Soil Sci., 69:269 ( 1950).
125 Piper, Soil and Plant Analysis (New York: Interscience Publishers, Inc., 1944),
p. 191.
180 PHOS PHOR US DETE RMIN ATIO NS FOR SOILS

PROCEDURE
exchange
7-147 . In the procedure of Bass and Sieling for the phosphate
is weighe d into a 400-m l beaker,
capacity of acid soils, a 20-gm soil sample
digeste d in a boiling water bath
mixed with 75 ml of 0.5 M citric acid, and
is then filtered throug h
for 1 hour with periodic stirring. The suspension
volum etric flask, follow ed
Whatman No. 40 or 41 filter paper into a 500-ml
room tempera-
by washing with hot water. When the filtrate has cooled to
pipette d into a
ture, it is made to volume, mixed, and a 25-ml aliquot is
s, then 10
250-ml conical flask. The aliquot is evaporated nearly to drynes
a steam plate
ml of concentrated HNO:i is added. The mixture is heated on
added. Heating
for about 20 minutes and then 5 ml of 60 per cent HC10 4 is
te with a digesti on manifo ld to collect the
is continued on an electric hotpla
g is contin ued until heavy white
HC10 4 fumes in water ( ~ 12-23 ). Heatin
20 minute s to insure complete
fumes of HC10 4 appear , then an additional
is sullicie ntly cooled
oxidation of the organic matter and iron. The sample
of distille d water.
to avoid spattering, and then is diluted with 20 ml
0.05 to
7-148 . The digested solution (organic matter free), containing
transferred to
1.0 millimol of Fe plus Al or either element separately, is
of 20 per cent
a 250-ml beaker. Then I 00 ml of 0.03 M KH~PO 4 and I 0 ml
added, and the
NaCl are added. Two drops of methyl orange indicator are
solutio n is then titrated with approxi-
solution is heated to boiling. The
to yellow is observ ed. Dilute HCl is
mately N NaOH until a color change
slight orange color, giving a pH
then used to back titrate the solution to a
then digeste d at near boiling
of approximately 3.4 to 3.5. The sample is
digesti on, the hot solu-
temperature on a steam plate for 30 minutes. After
the filtrate is dis-
tion is filtered through Whatm an No. 42 filter paper, and
hot 1 per cent
carded. The precipitation beaker is washed thoroughly with
the precipi-
NaCl adjusted to pH 3.4, and this washing liquid is poured over
from the filter
tate on the filter. The solution is allowed to drain completely
are washe d 2 additio nal times with hot 1 per
and the filter and precipitate
direct the wash solutio n around the top
cent NaCl solution. Care is taken to
itate as little as possib le. When the
of the filter so as to disturb the precip
h the filter (the precip itate
third washing has completely drained throug
a hot 0.2 N
must not be allowed to dry), the precipitate is dissolved with
in the colorim etric
acid solution, the acid being the one to be employed
d II, HCI0 4 for
determination (H 2 S04 for method I, HCI for metho
from dissolu -
method IV, or HN0:1 for method V). The solution resulting
The solu-
tion of the precipitate is washed into a 500-ml volumetric flask.
e, mixed , and an aliquo t is taken for colori-
tion is cooled, made to volum
metric analysis of the phosphorus.
lent to
7-149 . The numbe r of millimols of phosphorus found is equiva
t in the origina l citric acid
the numbe r of millimols of Fe plus Al presen
PHOSPHORUS DETERMINATIONS FOR SOILS 181

extract of the soil and represents the phosphate exchange (fixing) capacity
of the soil. It is calculated in terms of millimols (mgm atoms) of P per
100 gm of soil.

ALTERNATIVE PROCEDURES

7-150. In the Bass and Sieling procedure for measurement of phosphate

exchange capacity by direct addition of phosphate, 5 gm of soil is weighed
into a 50-ml centrifuge tube and enough 0.5 M KH 2 P0 4 (brought to pH
3.4 with HaPOJ is added to bring the total volume to 40 ml. The mixture
is stirred thoroughly and then digested on a water bath for 4 hours. At the
end of this time the centrifuge tube is removed and centrifuged for I 5
minutes to clarify the solution. The supernatant liquid is poured off and
40 ml of hot I per cent NaCl is added to each sample. The tubes are
thoroughly shaken, heated for 3 minutes in a water bath, and then centri-
fuged again. The supernatant liquid is poured off and the second washing
with 1 per cent NaCl made. Following this, 3 washings are made with 70
per cent ethanol containing 2 per cent NaCl. Samples are shaken thor-
oughly each time and centrifuged without being heated. All the liquid is
drained out of the centrifuge tubes after each centrifuging.
7-151. To displace the phosphate, the sample is digested on the hot
water bath for I hour with 0.5 M citric acid. The suspension is filtered into
a 250-ml volumetric flask, and the solution is cooled and made to volume.
A 25-ml aliquot is taken for oxidation with HNOx and HCIO, as in the
procedure ( ~! 7-14 7). When the oxidation has reached the stage of dense
HCI0 4 fumes, the solution is cooled, diluted, and filtered into a 250-ml
flask. An aliquot is taken for the colorimetric phosphorus determination.
The phosphate extracted is calculated as millimols (mgm atoms) of P per
100 gm of soil, the phosphate exchange capacity.

QUESTIONS
I . Distinguish between the analytical determination of phosphorus as op-
posed to the principles and problems involved in the extraction of various forms
of soil phosphorus.
2. What is the relation of the phosphorus atom to the molybdate radicals in
the heteropoly complex ion?
3. What chemical principle is employed in the elimination of the interfer-
ence by silica in the phosphorus determination?
4. What is the "acid-stability plateau" with reference to the heteropoly blue
color methods for phosphorus?
5. What factors dictate the choice of phosphorus method?
6. By what means is the interference of fluoride eliminated in the chloro-
stannous-HCI method?
7. State the advantages of the vanadomolybdophosphoric yellow color phos-
phorus method.
182 PHOSPHORUS DETERMINATIONS FOR SOILS
8. What forms of soil phosphorus are extracted by dilute acid?
9. What chemical form of phosphorus is extracted by neutral fluoride
solution?
l 0. Why is NaHCOa of pH 8.5 effective in the measurement of available
phosphorus of calcareous soils? Of acid and neutral soils?
11. Why is citrate-dithionite an effective cxtractant of acid and alkali insolu-
ble soil phosphate?
12. What chemical principles are applied in the extraction of the total organic
phosphorus of soils?
13. List 4 different methods for the release of the total phosphorus from
soils.
14. Why do the phosphates of Fe and Al quickly become crystalline when
formed at pH 2 to 3 but not when precipitated at higher pH values?
15. Explain the equivalence of citric acid soluble Fe and Al and the phos-
phate exchange capacity of acid soils.
 8
Nitro~en Oeterminations
for Soils and Plant Tissue
NitroRen-css cntial link in protein

8-l. Most of the nitrogen in soils is in organic form. Relatively small

amounts ordinarily occur in ammonium and nitrates, the available forms.
Two general types of analytical procedures are widely used-the Kjeldahl
conversion of nitrogen to (NH,1 ) 2 SO 4 and the Dumas conversion to nitro-
gen gas. The Kjeldahl method is employed in macro, micro, and ultra-
micro procedures. The macro procedure is given here.
8-2. Microbiological conversion of organic nitrogen to ionic forms is an
important aspect of the nitrogen chemistry of soils. The Kjeldahl method
includes both organic and ammonium forms, and with modifications (~
8-14) includes the nitrate form. Nitrate nitrogen should be included in the
total nitrogen of soils that contain appreciable amounts. For example, 100
ppm N in nitrate form is 0.01 per cent of the soil and 10 per cent of the
total nitrogen in the soil containing 0.10 per cent nitrogen (expected in a
soil containing 2 per cent organic matter). Procedures for the separate de-
termination of ammonium and nitrate are also detailed herein.
8-3. A general approximation of total soil nitrogen content can be made
from organic matter content (0.M. x 0.05 = N) and ignition loss at
400°C (loss x 0.022 to 0.03 = N). These approximate factors are sub-
ject to great change with widely different soils.

TOTAL NITROGEN DETERMINATION

(Modified Kjeldahl method)
8-4. Although the 9riginal Kjeldahl procedure has been modified many
times, the determination of total nitrogen is still not as simple as it is often
183
184 NITROGEN DETERMIN ATION-SO ILS, PLANT TISSUE
thought to be. 1 It is subject to many difficulties, any of which may lead to
low results. For example, an hour of digestion may be necessary after the
digest turns clear, to release all of the nitrogen, the most important con-
siderations being the catalyst selected and the digestion temperature. If the
temperature is too low (below 360°C), the release is slow or incomplete,
and if too high (over 410°C) , some loss of NH3 from the mixture re -
sults.2 The soil should be more finely ground 3 than the sample size would
ordinarily require in order to assure complete oxidation of the organic
matter within the small aggregates, especially in the case of heavy clay
soils. Soaking4 clay soils in water also aids in achieving complete oxidation.
APPARATUS
8-5. Needed apparatus consists of the Kjeldahl digestion manifold (~
8-6), a Kjeldahl NH3 distillation rack, 800-ml or 500-ml Kjeldahl flasks,
a 25-ml pipet, 500-ml conical receiver flasks with 8-mm diameter re-
ceiver tubes long enough to reach from the bottom of the conical flask to
the distillation condenser, burets, and analytical balance.
8-6. Kjeldahl Digestion Rack with Fume Aspirator. The special type of

Fig. 8-1. Kjeldahl digestion apparatus for disposal of acid fumes through water
pump and sewer, successfully operated at the University of Wisconsin for over
20 years.
1 Bal and Meter, Anal. Chem., 23 : 1632 (1951), show wide variation between
laboratories in both the Kjeldahl and Dumas determination of total nitrogen of
petroleum products, which are somewhat more difficult materials to analyze than
soils and plants.
2 Lake et al., Anal. Chem., 23:1634 (1951) .
s Walkley, J. Agr. Sci., 25:598 (1935); Prince, Soil Sci., 59:48 (1945).
4 Bal, J. Agr. Sci., 15:454 (1925); Ashton, J. Agr. Sci., 26:239 (1936).
NITROGE N DETERM INATION -SOILS, PLANT TISSUE 185

H 2 SO4 fume disposal by evacuation5 through a large lead water pump

(Fig. 8-1) is m~re!erred to the less satisfactory H 2S0 4 fume chimney
type that is commo~featured commercially. The shut-off and regulator
valves for the water line, which flushes condensed acid out of the lead
manifold, is seen in the upper right of Fig. 8-1. A similar set of valves
regulates the flow through the lead suction pump at the end of the manifold.
Details of the manifold and pump are shown in Fig. 8-2.
,,,, .. iron pipe nipple
~ (soldered in)
Wipe 1oint

21/i'

Four 1/8' holes

slanted downward ~
41/,'' o.c. ~ ['h'' inside d1a
%" outside
to improve 1et action
8" long
!"inside d1a
All lead pipe extra strong

To 2" soil pipe and trap placed 30" below.

A 3" drain 1s required to carry water from
2 pumps and manifolds,

Fig. 8-2. Pump and manifold for Kjeldahl digestion apparatus. (Lead fabricated
by Crown Metal Products Co., 117-119 Washington Street, Milwaukee, Wis.)

REAGENTS

8-7. Needed reagents include concentrated H:!S0 4 , granulated zinc or

N
pumice, reductant for mercuric compounds, standard-14 H:!S0 4 , and the
special reagents listed in the following paragraphs. If Na 2 S or K2 S is to
be employed for reduction of mercuric compounds, 40 gm of the com-
mercial grade is dissolved in I liter of water. lf Na 2 S20a · .5 H 20 is to be
employed, it is dissolved in the NaOH (~ 8-9).
8-8. Digestion Accelerators. Approximately 20 gm of CuS0 4 • 5 H 2 0
(previously ground and dried in an oven at 110°C), 3 gm HgO, and l gm
of Se powder are mixed 11 by grinding in a mortar. Caution: The Se powder
5 Kjeldahl fume disposal through the sewer has been successfully employed
by
Prof. H. A. Schuette of the University of Wisconsin by an apparatus after which the
one described here was designed. Tyner, Anal. Chem., 20:273 (1948), described a
slightly different apparatus that successfully achieved the same results .
. 6 Mixture proponions after Poe and Nalder, Ind. Eng. Chem., A.E.,
7: 189 ( 1935),
and Lauro, Ind. Eng. Chem., A.E., 3:401 (1931 ), who found a greatly accelerated
digestion rate over copper alone. Patel and Sreenivasan, Ana/. Chem., 20: 63 (1948}
noted the value of Hg in helping prevent loss of N, which may occur with Se alone as
a catalyst.
186 NITROGEN DETERMIN ATION-SO ILS, PLANT TISSUE
must not be dried at l 10°C since this element is volatile and toxic. One
part of this mixture is thoroughly mixed with 20 parts of anhydrous Na 2S04
and the mixture labeled "Na 2S04 -plus-catalyst." The purpose of the Na 2S04
salt is to raise the boiling temperature of the H 2S0 4 digestion; K2 S04 works
as well as Na 2S04 and may be substituted. Alternatives to the dry mixture
of digestion accelerators is the separate addition of each component; for
example, 5 ml of 1 N CuS0 4 solution has been used instead of the crystals.
8-9. 40 per cent NaO.ff, for NH 3 Distillation. First 50 per cent NaOH
solution is made up and impurities are allowed to settle out. To do this,
3 kgm of NaOH (technical, low N) is dissolved in 3 liters of distilled
water in a heavy-walled Pyrex flask or bottle. The solution is allowed to
stand several days for the Na 2C03 to settle out. Then the clear supernatant
liquid is siphoned (caution) into 1.5 kgm of water; the solution is then
mixed and placed in the 40 per cent NaOH bottle (Fig. 8-3) with a glass

Fig. 8-3. Arrangement for delivery of 40 per cent (I ON) NaOH

and protection from C0 2 contamination.

delivery tube that is too short to act as a siphon, and an Ascarite tube to
prevent absorption of C02 • If Na 2S20 3 is to be employed to precipitate
mercuric compounds, 360 gm of Na2S20 3 • 5 H 20 is dissolved in the 1.5
kgm of water prior to addition of the 50 per cent NaOH.
8-10. Mixed Indicator Solution. Brom cresol green (0.5 per cent) and
methyl red (0.1 per cent) mixed indicator7 is prepared by dissolving 0.5
7 Indicator proportions after Ma and Zuazaga, Ind. Eng. Chem., A.E., 14:280
· -'1942); but brom cresol green concentration 5-fold greater, was found to be more
efficacious in the author's laboratories.
NITROGEN .DETERMINATION-SOILS , PLANT TISSUE 187
gm of brom cresol green and 0.1 gm methyl red in 100 ml of 95 per cent
ethanol, and adjusting the solution to the bluish purple midcolor at pH
4.5, with dilute NaOH or HCl. This indicator is pink at pH 4.2 or lower
and bluish green as the pH rises to pH 4.9 and above.
8-11. Boric Acid. Approximately 40 gm of H 8B08 is dissolved in 1 liter
of distilled water containing 5 ml of the mixed indicator. This boric acid
stock solution is adjusted by dilute H 2 S04 or HCl titration until the bluish
color of the indicator weakens toward pink. The equilibrium constant of
H 3 B03 (6.4 x 10- 10 ) shows that pH 8.6 is reached when 20 per cent of
the first H has been neutralized with NH 3 , which corresponds to 48 mgm of
N per 25 ml. The solution is 0.65 M H 3 B03 and thus its pH is 4. 7 when all
of the H 3 B03 has been reformed in the acid titration. Therefore the pH
4.5 midcolor of the indicator permits a sharp end point on titration with a
strong acid.
8-12. The reagents are tested for nitrogen by digestion of a filter paper
blank, all steps of the procedure being carried out as in the determinations.
The blank determination is subtracted from each determination.

PROCEDURE

8-13. For Soils or Plant Tissue. A soil sample of 5.00 gm (1.000 gm of

muck or peat, or 20.00 gm of sandy soil) that has been ground to pass a
0.15 mm (I 00 mesh per in.) screen, or 0.500 gm of dried plant tissue that
has been ground to pass a 0.4 mm ( 40 mesh per in.) screen is wrapped in a
11-cm qualitative filter paper and dropped as a package into a 800-ml
Kjeldahl digestion flask. Then 20 ± 1 gm 8 of the Na 2S0 4 -plus-catalyst di-
gestion mix is added.
8-14. Nitrates may be included, (~8-2, 8-15) but usually are not, since
they are often negligible in relation to the total nitrogen. The procedure
for plant tissue or sandy soil samples is to add 35 ml of concentrated
H 2 SO4 and to mix the contents by swirling the flask, with care not to throw
the sample onto the sides. The digestion is then commenced ( ~ 8-18). The
procedure for medium and fine textured soils is to add 50 ml of water9 and
to allow the sample to soak for 30 minutes. Then 35 ml of concentrated
H 2 SO4 are added and digestion follows ( ~ 8-18).

8 This quantity of Na2 S04 controls the boiling point of the digestion as shown by
Lake et al., Anal. Chem., 23:1634 (1951), between 360° and 410°C, which is the
critical range for efficacious digestion. Since the boiling temperature is affected by
altitude and the rate of H 2 S0 4 discharge during digestion, some control work on
completeness of digestion and recovery is suggested by each laboratory. The 2~gm
quantity, adopted by Lake et al., and used in this procedure, is double the conven-
tional 10 gm.
o Bal modification, J. Agr. Sci., 15:454 (1925), which raises the N recovery con-
siderably, especially in clayey soils.
188 NITROGEN DETERMINATION-SOILS, PLANT TISSUE
8-15. If soil nitrates 10 are to be included in the determination, 35 ml of
concentrated H 2 S0 4 containing 1 gm of commercial salicylic acid is added
to the sample in the flask ( ~ 8-13). The flask is swirled until the acid is
thoroughly mixed with the soil, and the mixture is allowed to stand for 30
minutes for the nitrates to react with the salicylic acid. Then 5 gm of
Na 2 S2 0:1 • 5 H 2 0 (or 2 gm of zinc dust-gra nulated zinc will not do) and
50 ml of H 2 0 11 are added and the mixture is heated slowly and with care
at first to avoid frothing over. When this danger is past, the digestion is
continued as usual ( ~ 8-18) .
8-16. Procedure for Runoff Suspensions. The runoff suspension is
shaken thoroughly and a 250-ml aliquot ( l 00 ml if over 5 gm of solids
occurs in 100 ml) is quickly measured out in a calibrated (~ 10-64)
beaker. The aliquot is transferred to an 800-ml Kjeldahl flask, a few glass
beads and 20 ± 1 gm of the Na 2 S04 -plus-catalyst are added, and finally
35 ml of concentrated H 2S0 4 • The suspension is mixed cautiously by swirl-
ing the flask, and heated gradually to evaporate the water. The heat is then
increased and digestion in the H 2 SO 4 is effected ( ~ 8-18).
8-17. The above procedure for runoff does not include the nitrates.
Separate determination of nitrates by the phenoldisulfonic acid (~ 8-59)
is simpler, and permits addition of the nitrate equivalent to the Kjeldahl
nitrogen excluding nitrates. The salicylic acid method (~ 8-15) works only
in concentrated H 2 S0 4 and will not recover nitrates from aqueous solution.
8-18. Digestion in H 2 S0 4 • Digestion is effected on the Kjeldahl diges-
tion rack with low flame for the first 10 to 30 minutes, until frothing stops,
and then gradually more strongly until the sample is completely charred.
The heat is gradually raised until the acid reaches a boil, and condensation
of acid reaches approximately one-third the way up the neck of the diges-
tion flask.12 The flame is not allowed to touch the flask above the part
occupied by liquid; otherwise there may be a loss of NH:i in consequence
of decomposition of (NH 4 ) 2 S0 4 • Heating at an excessive rate may be a
disadvantage because of undue volatilization of acid before the organic
matter is all oxidized. Some NH:1 may be lost if the acid is largely volatil-
ized, because the temperature may rise above 410°C. The flask is rotated at
intervals and heating is continued until the organic matter is destroyed,

10 Nitrates are ordinarily lost by volatilization of HNO:i

during the H 2 S0 4 diges-
tion. In this procedure, the nitrate is combined with salicylate in concentrated H 2 S0 4
•

According to the work of Dr. J. C. Kaudy and the author at this laboratory ( 1948),
nitrates in runoff in aqueous H2S04 are virtually impossible to reduce quantitatively by
e
means of granulated zinc, Davada's alloy, or iron powder, although one quantitativ
result was reported by Ashton, J. Agr. Sci., 26:239 (1936), with 2 gm of colloidal
iron reduced from the oxide. The ferrous sulfate formed from iron powder has the
advantage over zinc sulfate of solubility in the H 2 S0 4 •
11 Combinati on of the Bal modification and the salicylate procedure
, after Ashton,
J. Agr. Sci., 26:239 (1936). .
12 Lake et al., Anal. Chem., 23:1634 (1951).
NITROGEN DETERMINATION-SOILS, PLANT TISSUE 189
best judged by timing the digestion for 1 ± 0.25 hour after the solution has
cleared (light yellow or gray color). According to Lake et al., the best as-
surance of complete digestion is careful regulation of the dig~stion tempera-
ture so that it exceeds 360°C but does not reach 410°C. Extra digestion
time may be substituted for higher temperature only to a limited extent.
8-19. At the end of the digestion, the heating is stopped, but the
fume exhaustion is continued until fuming stops. When the flasks are cooled
just to the point where crystals start to form (not cooled completely, as the
salts redissolve only slowly), 300 ml of NH 3-free water is added as the
solution is cautiously mixed. This solution is further cooled (heat of dilu-
tion).
8-20. If large quantities of sand are present, particularly from runoff,
bumping during distillation is sometimes severe. This can be avoided by
washing the acid solution into another flask, the sand being left in the
original flask.
8-21. Several pieces of granulated zinc or a teaspoon of pumice is added,
followed by 25 ml of K2 S (reducing agent for mercuric salts; may be r,.e-
placed by Na 2 S or Na 2 S2 0a, the latter being most effectively added in the
40 per cent NaOH as in the next paragraph). The solution is mixed, and
is then ready for determination of the ammonium content.
8-22. Distillation of NHa into Boric Acid. 13 Approximately 14 25 ml of
4 per cent boric acid is pipetted into a 500-ml conical flask, and 4 drops
of brom cresol green-methyl red indicator solution are added. A glass re-
ceiver tube is attached to the still and placed in the flask so that its end is
below the surface of the boric acid in the flask. The cooling water is then
started flowing in the. condenser. The contents of the Kjeldahl flask are
mixed by rotation, and the flask is placed on the distillation stand and
checked for a good fit with the condenser connection. Then, with the
Kjeldahl flask held at a 45 ° angle, about 125 ml (or 100 ml, if bumping
is a problem) of 40 per cent NaOH are poured so that it runs down the
neck to the bottom of the flask without mixing. The burner is then lighted,
the flask is attached to the still, and the solution is mixed thoroughly by
swirling. Immediately after this mixing, the flask is set to rest on the still
support and is heated to avoid the danger of "sucking back. " 15 The flame

13 Wrinkler modification, Scales and Harrison, Ind. Eng. Chem., 12:350 (1920),
J.A.0.A.C., 8:455 (1925).
14 Neither the volume nor the strength of the boric acid need be known exactly,
because the NH4-borate formed is titrated back to HaBOa in the titration. Twenty-five
ml of boric acid will absorb 48 mgm of nitrogen as NH:i. which is approximately
equivalent to 0.9 per cent N in soil ( 5 gm sample), or 4.8 per cent N in peat (1 gm
sample), or 9.6 per cent Nin vegetation (0.5 gm sample), or about 34 meq per 100
gm exchange capacity (10 gm sample).
15 If sucking back occurs, it is only necessary to wash all of the boric acid into the
digestion flask and redistill the NHa into a fresh lot of boric acid.
190 NITROGEN DETERMINATION-SOILS, P~NT TISSUE
receiver
is increased gradually. About 150 ml is distilled over and then the
flask and tube are disconnected to preven t sucking back.
bly
8-23. The boric acid is back titrated with a standar d acid, prefera
_!'!_ HnS0 4 or HCl in routine soil analysis (0.0788 N acid for runoff). At the
14 - to 0.05
end point the blue color just disappears. One drop in excess (0.02
ml) turns the solution pink.
ted
8-24. For soil or plant tissue, the percentage of nitrogen is calcula
111

as follows:

% Nin soil or plant tissue= (T - B) x N x !·~­ (8-1)

s
in which
T = sample titration , ml standar d acid
B = blank titration , ml standar d acid
N = normali ty of standar d acid
s = sample weight, gm
calcu-
For runoff (~ 8-16), pounds N per acre inch of runoff (ppai) is
lated for a 250-ml sample of suspension analyze d as follows:

Nin runoff (ppai) = (T - B) x 1.0 (for 0.0788 N H~S0 4 )

= ( T - B) x 0.906 (for~ H~S01 )

= (T-8 ) x 1.27 (for0.JONH~S0 4 )
(8-2)

in which
T = sample titration , ml standar d acid
B = blank titration , ml standar d acid
ppai = pounds N per acre inch runoff
Then the nitrogen in the soil solids present in the runoff is calculated:
N in runoff = pp~i nitrogen in runoff x 2 000 OOO (8-3)
(pp2ms) ppai solids in runoff ' '
runoff.
in which, pp2ms = parts N per 2 million parts of soil solids in
correct ion factor
These values, for greatest precision, are multiplied by the
(~ 10-69, eq. 10-21 ).

rn Approxi mations :
% N X 6.25 = % crude protein in plant tissue
% N X 20 = % organic matter in soil
% N X 20,000 = pounds nitrogen per acre in soil
NITROGEN DETERMINATl ON-SOILS, PLANT TISSUE 191

ALTERNATIVE PROCEDURES

8-25. A number of suggestions have been made for insuring complete

liberation of the organic nitrogen in the form of ammonium sulfate. Greater
completeness is claimed 17 through the addition of HC10 4 near the end of
the digestion followed by warming below boiling, but Joss of nitrogen by
this treatment will occur if the solution is brought to boiling. Another pro-
cedure calls for the addition at the end of the digestion (after cooling
&lightly) of KMn0 4 crystals in pinches until the solution remains green or
purple. Loss of nitrogen occurs if the solution is heated after this treatment.
It is believed that the elevated temperature empJ.oyed in the procedure given
is the most satisfactory method.
8-26. Acid-Base Titration of Distilled Ammonia. An alternative pro-
cedure is to collect the ammonia in dilute (about 0.1 N) H~S0 4 or HCl, the
excess acid being back titrated with standard NaOH. A common error is
made in thinking that this dilute acid need be of standardized strength,
whereas only its volume need be known exactly. The ammonia is distilled
over (~[ 8-22) into exactly 25 ml of approximately 0.1 N H~S0 4 or HCl,
containing 3 drops of methyl red-methylene blue indicator. Twenty-five ml
of ~ acid is equivalent to 0.5 per cent N in 5 gm of soil; or 18 meq of
14
exchange capacity per I 00 gm, when a 10 gm sample is employed. The
excess acid is back titrated with standard _""! NaOH (0.0788 N NaOH
14
may be used for runoff) to an end point of about pH 6.0 with methyl red-
methylene blue indicator (0.6 gm methyl red and 0.4 gm of methylene blue
in 500 ml of 95 per cent ethanol). The color change is from lavender to
colorless to green. For soils or plant tissue, the percentage of nitrogen is
calculated as follows:

% N in soil or plant tissue = (S - T) x N x I. 4 (8-4)

s
in which
S = standardization titration, ml standard NaOH for 25 ml H~S0 4 used
for receiving the distillation of the blank
T = titration of sample. ml standard NaOH for 25 ml H~S0 4 receiving the
sample
N = normality of standard alkali
s = sample weight, gm
For runoff, lbs. N per acre inch runoff (ppai) is calculated:
Nin runoff = (S - T) x 1.0 (for 0.788 N NaOH) (8-5)
(ppai)

17 Pepkowitz et al., Ind. Eng. Chem., A.E., 14:856 ( 1942).

192 NITROGEN DETERMINATION-SOILS, PLANT TISSUE
N
= (S - T) x 0.906 (for 14Na0H)
= (S - T) x 1.27 (for 0.10 N NaOH)
8-27. Titration of Ammonium in Digest. Instead of being distilled the
ammonium, as NH:1, may be titrated directly, 18 under suitable conditions,
in the Kjeldahl flask after the usual H 2 S0 4 digestion. The principle is based
on the fact that the excess H 2 S04 can be neutralized with strong NaOH
while the ammonium, previously displaced from the mercuric complex by
the addition of NaBr, which complexes the mercuric ions as HgBr 4 - - , is
left as (NH 4 ) 2 S04 as follows:

This reaction is controlled by methyl red indicator, and the addition of

10 N NaOH is stopped at the point at which the indicator turns from pink
to yellow (after the solution is boiled to remove C0 2 ). Then 30 ml of 18
per cent formaldehyde is added to complex the ammonium as hexameth-
ylenetetramine (slight reversal of indicator color to pink is ignored) and
standard 0.1 N NaOH is added to yellow and then further added to pink
with 8 drops of .l per cent phenolphthalein indicator. The total titer of
standard NaOH between methyl red and phenolphthalein is equivalent to
the ammonium present:

(NH 4 ) 2 S04 + 2 NaOH-----+

(to pH 8.3)
Na 2 S0 4 + 2 H 20 + 2 NH 2
(8-7)
(8-8)

The procedure is facilitated by digestion in 500 ml, round bottom or slightly

flattened bottom, Pyrex flasks with 29 / 42 standard-taper openings that fit
20-cm (eight-inch) detachable necks. 19
8-28. As would be expected, some materials in the sample may interfere
with titration in the presence of the entire digest. Precipitates of iron and
aluminum tend to obscure the end points, and phosphorus tends to undergo
conversion that gives a positive error. Zirconyl chloride, however, precipi-
tates the phosphate in nonreactive form. Silica impurities in the NaOH also
interfere by buffering the solution near the phenolphthalein end point if
more than 15 ml of concentrated H 2S04 must be neutralized.

18 Marcali and Rieman, Ind. Eng. Chem., A.E., 18:709 (1946), Anal. Chem.,
20:381 (1948).
19 Available from Fisher Scientific Co., Pittsburgh 19, Pa.
NITROGEN DETERMINATION-SOILS, PLANT TISSUE 193

TOTAL AMMONIUM AND NITRATE NITROGEN IN WATERS

8-29. To determine the ammonium and nitrate nitrogen in drainage
waters, the nitrate is reduced to ammonium with H 2 in alkaline solution, and
the total ammonium is volatilized as NH 3 into boric acid.
PROCEDURE

8-30. A 250 ml (or other) volume of filtered runoff or drainage water

is placed in an 800-ml Kjeldahl flask, and the flask is fitted to the Kjeldahl
distillation rack. Boric acid solution is placed in the receiver flask as in the
total nitrogen procedure(~ 8-22). Then, 2 gm of Davarda's alloy (Cu, 50
per cent; Al, 45 per cent; Zn 5 per cent) and IO ml of 40 per cent NaOH
are added. The flask is attached to the still and the heating is started while
the solution is simultaneously mixed by swirling the flask.
8-31. Distillation is continued until approximately 225 ml of the distil-
late has been collected. It is necessary to carry the distillation nearly to dry-
ness to obtain complete reduction of all nitrate.
8-32. Quantitative tests with standard nitrate samples are suggested as
a measure of the percentage recovery being obtained, since full recovery is
frequently difficult to obtain.

EXCHANGEABLE AMMONIUM DETERMINATION

8-33. Because the ammonium ion is subject to oxidation to nitrite and
nitrate in soils stored warm and moist, its determination should quickly
follow the taking of the soil samples. The samples may be extracted moist,
a separate determination of moisture being made, or they may be spread
out in a thin layer and dried rapidly in an oven at 50°C.
8-34. Care must be exercised to prevent hydrolysis of the organic com-
pounds of the soil to ammonium. This precludes a satisfactory direct dis-
placement of ammonium by distillation as ammonia through treatment of
the soil with hot alkali. Use of an acidified salt solution for displacement
purposes appears to be a satisfactory precaution. Exchangeable ammo-.
nium ion has been determined by aeration of the soil at room temperature
in the presence of a solution containing 4 per cent K~CO:i and 20 per cent
KC! in a suitable apparatus, 20 but the method is more tedious than extrac-
tion.
8-35. The ammonium ion is held in exchangeable form in soils just as
are the exchangeable metallic cations (Chapter 5). It must be extracted by
some other exchangeable cation. Ammonium ion undergoes equilibrium
fixation in the 2 : 1 layer silicates, particularly in the highly charged ver-
miculite interlayer spaces, in exactly the same way as K +, by closure of the
20 Mathews,/. Agr. Sci., 10:72 (1920).
194 NITROGEN DETERMIN ATION-SO ILS, PLANT TISSUE
interlayer space. The ammonium ion thus fixed undergoes only slow ex-
change and is reluctant to nitrify. 21 The sodium ion has been selected for
the replacing ion because it is among the best for replacing ammonium and
potassium from the slow exchange positions. 22
8-36. The amount of exchangeable NH4 + frequently falls in the range
of 0.01 to 0.1 meq per 100 gm of soil, which range corresponds to 1.4 to
14 ppm or 2.8 to 28 pp2m. The upper limit of this range corresponds to 1
per cent of an exchange capacity of 10 meq per 100 gm of soil, and there-
fore, the exchangeable NH4 + usually may be neglected in the calculation
of percentage cation saturation (~ 5-4). The possibility of an exchange-
able ammonium content of 10 or more times the upper limit of this range
in some circumstances should not be overlooked. Equivalent to the value
0.1 meq per 100 gm soil, are the values 28 pp2m of NH 4 + and 78 pp2m
of K +, and thus the percentage saturation with K + will generally exceed
that of NH 4 + in most soils.

APPARATUS

8-37. Needed apparatus includes a torsion balance, a 500-ml conical

flask, a 11-cm Buchner funnel and suction flask, 11-cm Whatman No. 42
filter paper, a 800-ml Kjeldahl flask, and ammonia distillation apparatus.

REAGENTS

8-38. Needed reagents include 10 per cent (1.7 N) NaCl, acidified to

pH 2.5 with HCl, 23 standard N or!!_ H 2S0 4 , 40 per cent ( 10 N) NaOH,
14 56
N-free ( ~ 8-9), methyl red-brom cresol green mixed indicator ( ~ 8-10),
and 4 per cent boric acid solution ( ~ 8-11 ) .

PROCEDURE

8-39. Extraction of Ammonium. A 100-gm, freshly taken soil sample

is weighed out and placed in a 500-ml conical flask. (A separate moisture
determination is made.) Then 200 ml of the acidified NaCl solution is
added. The suspension is shaken thoroughly at first and then intermittently
for 0.5 hour. Then it is poured onto the Buchner funnel on which a 11-cm
Whatman No. 42 filter paper has been moistened and seated firmly by
suction. Finally 250 ml more of the acidified NaCl solution is passed
through the soil in increments, the first increment being used to rinse out
the conical flask.
8-40. The entire leachate from 100 gm of soil is generally needed for

21 Bower, S.S.S.A. Proc., 16: 119 (1951).

22 Barshad, Soil Sci., 72:361 (1951).
28 Peech et al., U.S.D.A. Cir. 747, p. 9 (1947).
NITROGEN DETERMINATION-SOILS, PLANT TISSUE 195
the distillation determination of ammonium because of the small quanti-
ties of ammonium usually present in soil ( ~ 8-36). If much less than 10
ppm of N is expected, the Nessler procedure (~ 8-43) may be elected for
greater sensitivity. Since 25 ml of 4 per cent boric acid, used to collect the
ammonia, will hold the equivalent of 48 mgm of N, it will collect the NH4
equivalent of up to 960 pp2m of N in the soil.
8-41. Detennination of Ammonium. The NaCl leachate, containing 1 to
48 mgm of N, is transferred to a 800-ml Kjeldahl flask. Caution: the still
should be scrupulously clean, conveniently insured by distilling 50 ml of
distilled water through it prior to the ammonia distillation. A reagent
blank is run. Also, the pH of the boric acid should be carefully adjusted
to the end point. Next, 80 ml of 40 per cent NaOH is added carefully down
the side of the flask so as to collect at the bottom of the flask. The am-
monia is then distilled into 25 ml of 4 per cent boric acid by means of
Kjeldahl still ( ,, 8-22). The boric acid is back titrated with standard ~
H 2 SO4 (or ~ , if the amount of ammonia is expected to be large).
8-42. The meq of ammonium is the product: (ml of standard acid)
x (normality). The nitrogen in ammonium form, expressed as pp2m
(pounds per acre) may be calculated:

N (pp2m) = (T - B) x 5 00
s
(for ~ acid)
(8-9)
= (T - B) x 2000
s
(for ~ acid)

when
N = nitrogen, in ammonium form, pp2m

T = sample titration, ml standard acid, ~ or ~

B = blank titration, ml standard acid, ~ or ~

s = sample weight, gm

ALTERNATIVE PROCEDURES

8-43. Nessler Method for Ammonium. 24 The ammonium may be de-

termined by means of the Nessler reagent. This reagent is prepared25 by
dissolving 45.5 gm of mercuric iodide and 35.0 gm of KI in a few ml of
water:
(8-10)

24 Peech et al., U.S.D.A. Cir. 757, p. 10 (1947).

211 Vanselow, Ind. Enf(. Chem., A.E., 12:516 (1940); Peech, Soil Sci., 59:27 (1945).
196 NITROGEN DETERMINATION-SOILS, PLANT TISSUE
The solution. is washed into a I-liter volumetric flask. Then, 112 gm of
KOH is added and the volume is brought to about 800 ml. The solution
is mixed well, cooled, and diluted to 1 liter with water. The solution is al-
lowed to stand for a few days, and the clear supernatant liquid (Nessler's
reagent) is decanted off into an amber colored bottle for use.
8-44. Other reagents needed are 10 per cent sodium tartrate (100 gm
of Na 2 C4 H 4 0 6 • 2 H 2 0 per liter of solution), and standard NH4 Cl solution
( 1.337 gm of NH 4 Cl in 1000 ml of water containing 1 ml of chloroform
as a preservative). A dilute NH4 + standard, containing 0.001 meq of
NH 4 + per ml, is prepared from 20 ml of the first standard solution diluted
to 500 ml.
8-45. Aliquots of 3 to 30 ml of the dilute standard are placed in a series
of 100-ml volumetric flasks together with 2 ml of I 0 per cent tartrate solu-
tion and the same amount of acidified NaCl solution (~ 8-38) as will be
employed in the determinations on soil extract:
2 K2Hgl 4 + 3 KOH + NH:i ~ Hg20(NH)) + 7 KI + 2 H 2 0
(orang<') ( 8-11 )
Water is added to make about 93 ml total volume. Then 5 ml of the Nessler
reagent is added with rapid mixing. The solution is brought to volume,
mixed, and read at the end of 25 minutes in a colorimeter with 410-mu
light maximum. Transmission percentage is plotted on a log scale against
concentration on a linear scale to obtain the calibration curve.
8-46. To determine the NH 4 + in the extract, the leachate is washed into
a 500-ml volumetric flask and made to volume, and the solution is mixed.
Then a 50-ml aliquot is taken for Nesslerization as for the standards. A
greater or lesser volume of aliquot may be taken if the NH 4 + present is
out of range of the curve. For the 50-ml aliquot:
NH 4 + ( meq per 100 gm soil) = I 0 x ( meq NH 4 + from curve)
(8-12)
N (pp2m) = 2800 x (meq NH4 +from curve)
(8-13)
8-47. Equilibrium Extraction. An equilibrium extraction of exchange-
able ammonium is satisfactory and more rapid. The 100-gm soil sample
is placed in a liter flask. Exactly 500 ml of 10 per cent solution acidified
to pH 2.5 is added, and the suspension is shaken for 30 minutes in a
mechanical shaker. The solution is filtered on an 11-cm Buchner funnel
into a dry suction flask. Then 400 ml of the filtrate is added to a Kjeldahl
flask. A small piece of paraffin is added to prevent foaming. An excess
(3 to 4 gm) of MgO is added, 20 and the ammonia is distilled into boric
26Harper, Soil Sci., 18:409 (1924): McLean and Robinson, J. Agri. Sci., 14:548
(1924): Prince, Soil Sci., 59:47 (1945). The regular NaOH distillation (~ 8-22) also
may be used.
NITROGEN DETERMINATION-SOILS, PLANT TISSUE 197
acid ( ~ 8-22), followed by back titration. A correction factor of 1.25 for
the aliquot taken is applied in the calculation ( ~! 8-42).

NITRATE DETERMINATION
(Colorimetrically with nitrophenoldisulfonic acid27)
8-48. Several common chemical reactions are available for the de-
termination of soil nitrates, the most important of which is the nitro-
phenoldisulfonic-yellow color method. Reduction of nitrate with H.? gen-
erated by iron filings in H!1S0 4 has been employed to include nitrate with
total nitrogen (~ 8-15); Davarda's alloy in alkaline solution has been
similarly used ( ~ 8-30). The diphenylamine-blue color method is as used
in the qualitative test for nitrates in the sap of green plants (~ 13-12). The
alpha naphthylamine-pink color method may be employed if the nitrate is
reduced to nitrite (~i 13-22). The brucine-blue or -yellow color methods
have been used in qualitative tests for nitrates in soils. .
8-49. The phenoldisulfonic acid method for nitrates depends upon the
nitration of position 6 of 2, 4-phenoldisulfonic acid in fuming H~S0 4 :
C 0 H:1 OH(HS0a) 2 + HNOa ~ C0 H 2 OH(HS0a) 2 N0 2 + H 2 0
(8-14)
The nitrate solution is dried out previous to determination since the reac-
tion must be effected in the virtual absence of water. The product behaves
as a nitrophenolic type indicator with C-Y-Y reaction ( ~ 3-40), that is,
is colorless in acid and yellow when neutralized or in alkaline solution. A
hydroxide such as KOH or NH 4 0H is therefore employed to shift the pH
to the yellow range for the colorimetric determination.
APPARATUS

8-50. Needed apparatus includes a torsion balance, a 500-ml extraction

bottle and tight-fitting rubber stopper, an 18-cm filter paper and funnel,
8-cm evaporating dishes, 3 by 70 mm glass stirring rods, a 3-ml pipet
with its tip cut off to deliver rapidly, volumetric pipet, colorimeter tubes
with 40-ml calibration marks (or Nessler tubes), and a colorimeter with
420-mu light maximum.
REAGENTS

8-51. Needed reagents (all tested to be nitrate-free) include 6 N NH 4 0H,

Ca(OH) 2 , MgC0 3 , activated charcoal (G Elf or Darco G 60), approxi-
mately 1 N CuSO 4 ( 125 gm of CuSO 4 • 5 H 2 0 per liter), and the follow-
ing special reagents.
8-52. Phenol 2, 4-Disulphonic Acid. Twenty-five gm of pure phenol
(crystal white in color) is dissolved in 150 ml of concentrated H 2SO4 • Then
27 Adapted for photometric determination from Harper, Ind. Eng. Chem., 16: 180
(1924) and Prince, Soil Sci., 59:47 (1945).
198 NITROGEN DETERMINA TION-SOILS, PLANT TISSUE
75 ml of fuming H~S0 4 is added. This solution is mixed and heated by
placing the flask in boiling water for 2 hours. The resulting phenoldisul-
fonic acid, C 11 H:PH ( HS0:1 ) ~· solution is stored in a brown bottle. Caution:
this reagent is highly corrosive.
8-53. Standard Nitrate Solution. Exactly 0. 7221 gm of pure dry KNO:i
is dissolved in water and the solution is diluted to exactly 1 liter, giving
0.1 mgm N per ml, or 100 ppm stock solution. This stock solution is then
diluted, 20 ml to 200 ml in a volumetric flask. This latter solution contains
0.01 mgm N per ml, or I 0 ppm. Aliquots (2, 5, I 0, and 15 ml) of the
I 0 ppm N standard nitrate solution are placed in separate 8-cm porcelain
evaporating dishes and evaporated to dryness on the steam bath in an
atmosphere free from HNO:i fumes. Color development follows ( ~ 8-61 ) .
8-54. Ag~S0 1 Solution, to Remove Chlorides. Six gm of Ag~S0 4 is dis-
solved in 1 liter of H~O. This gives a 0.6 per cent solution.
8-55. Nitrate Extraction Solution. This is prepared by mixing 200 ml of
1 N CuS0 1 solution and I liter ot' 0.6 per cent Ag~S0 4 solution (~i 8-54)
and dilution to I 0 liters with H~O. The Ag~SO, is equivalent to 338 ppm
Cl- present in 50 gm soil, or 0.03 per cent Cl . If less than I 0 ppm of Cl -
is present in soil, the Ag~S0 4 may be omitted from the extraction solution.
If more than 0.03 per cent CJ- is present, as in saline soils, 2.25 gm of
powdered Ag~S0 4 salt for each I per cent Cl- present is mixed with the
soil prior to extraction; or, the Cl in the soil is determined and a CJ-
correction factor is established by addition of CJ- to a standard nitrate
series.

PROCEDURE

8-56. Soil Sampling and Preparation. Composite soil samples are ob-
tained (~ 2-7) freshly from the field or pot. The soil is mixed thoroughly
by passing it through a 6-mm sieve. Clayey soils that have dried and con-
tain hard granules are pulverized to pass a 2-mm sieve, to facilitate com-
plete wetting of the sample by the extractant in the time allowed.
8-57. Rapid changes in the nitrate and ammonia contents of soil samples
occur after removal of the samples from field or pot, because of the in-
creased aeration and rise in temperature. It is, therefore, desirable that the
extraction of nitrates and ammonium follow the collection of the samples
closely. If this is not possible, nitrification and ammonification in the
samples is retarded by adding 3 ml of toluene per kgm of soil and sealing
and refrigerating the samples. Retardation by this means is only moder-
ately satisfactory. If the elapsed time is to be greater than a day or two, the
samples are dried at a temperature not exceeding 55'C.
8-58. Extraction of Nitrate from Soil. Fifty gm of soil (25 gm of peat)
is weighed out and placed in a 500-ml, wide-mouthed bottle, and 250 ml
of extraction solution is added. (At the same time, a 25-gm sample is
NITROGEN DETERMINATION -SOILS, PLANT TISSUE 199

weighed out for moisture determination.) The suspension is shaken for

10 minutes and then 0.4 gm Ca (OH) 2 is added. This is foJJowed by 5 minutes
further shaking and the addition of 1 gm of MgC03' These 2 reagents pre-
cipitate the copper and silver and clarify the suspension. The suspension
is filtered on a dry filter paper, and the first 20 ml of filtrate discarded. A
10-ml portion of the clear filtrate~ 8 (25 ml if the soil nitrate nitrogen con-
tent may be less than 10 ppm) is pipetted into an 8-cm evaporating dish
and evaporated to dryness in an atmosphere free of HN0:1 fumes. Color
development follows (~I 8-61 ) .
8-59. Extraction of Nitrate from Evaporated Runoff. Samples of run-
off on which nitrate is to be determined are made alkaline by the addition
of 0.25 gm of CaCOa and evaporated to dryness immediately after coJlec-
tion. (Denitrification in runoff suspensions is not prevented by the addition
of 1 ml of toluol.) To the evaporated residue from runoff, 100 ml of the
extraction solution is added, and the solids arc suspended by thorough agi-
tation for 10 minutes with a rotary stirrer or an end-over-end shaker. Then
0.2 gm Ca (OH)~ is added, and the shaking is continued for 5 minutes.
Finally, 0.5 gm of MgC0:1 is added, and the suspension is shaken, allowed
to stand for a few minutes, and filtered on dry filter paper. The first 10 ml
are discarded to rinse the apparatus and about 40 ml is collected. If this
filtrate is colorcd~ 0 with organic matter, I gm of nitrate-free carbon black
or activated charcoal ( G Elf or Darco G 60) is added to the filtrate, fol-
lowed by shaking and filtration.
8-60. A 20-ml aliquot of the clear filtrate is placed in an 8-cm evapo-
rating dish, evaporated to dryness (with protection from HN0:1 fumes),
and analyzed (~I 8-61 ) .
8-61. Development of the Nitrophcnoldisulfonic Color. The 8-cm evapo-
rating dishes arc allowed to cool, and 3 ml of phenoldisulfonic acid is added
rapidly directly in the center of each. The dish is rotated to effect contact
with all of the residual salt, and the reagent is allowed to act for I 0 min-
utes. Then 15 ml of cold water are added, and the solution is stirred with

~H For certain acid soils that give a colored extract. the soil is allowed to settle
before the addition of the Ca (OH h and M gCO:i· Then about 150 ml of the super-
natant liquid is decanted off. To this arc added 0.2 gm of Ca(OH )~ and 0.5 gm of
MgCOR. It is then shaken for 5 minutes and filtered on a dry filter. The first 20 ml of
filtrate may be discarded.
For certain (usually alkaline) soils that give highly colored soil extracts that can-
not be decolorized by this treatment of the decanted supernatant liquid, 1 gm of
carbon black or activated charcoal (Darco G. 60) is added to 100 ml of the super-
natant liquid. and the suspension is shaken 15 or 20 minutes before the addition of
the Ca(OH h and MgC0: 1 to the solution. If the soil is calcareous, 5 ml of 1 N
copper sulfate is added to the soil extract with the carbon black or charcoal to insure
enough copper hydroxides to remove the colloidal carbon completely on filtration.
w Although organic matter that may color the filtrate can be removed by treatment
with 30 per cent H:iO:i, the treatment tends to result in an off color in the final
nitrate solution, and therefore is not recommended.
200 NITROGEN DETERMINATION -SOILS, PLANT TISSUE
a glass rod until all the residue is in solution. After the dishes are cool,
6 N NH 4 0H is added slowly until the solution is distinctly alkaline as in-
dicated by the development of a yellow color, then 3 ml more is added.
This solution is then diluted to volume with water. The standard series is
diluted to 100 ml and contains 0.2, 0.5, I, and I .5 ppm of nitrate nitrogen.
Soils extract is usually diluted to l 00 ml. Runoff nitrate is conveniently
diluted to 40 ml in calibrated tubes.
8-62. The transmission percentage of the nitrate solutions is read in a
colorimeter with 420-mu light maximum. Alternately, the color may be
evaluated by visual comparison to the standard solution by means of Nes-
sler tubes ( ~ 8-65).
8-63. Preparation of Standard Colorimetric Curve. A calibration curve
is plotted from the standard nitrates on semilogarithmic paper, the log
scale being employed for the transmission percentage readings. This curve
is usually not exactly linear, and thus it is best to refer to the graph to de-
termine nitrate concentration in the test sample.
8-64. Calculation of Results. The results are reported in parts of N
(nitrate form) per million parts of oven-dry soil. The concentration of the
nitrate test solution as ppm N is obtained from the standard curve. Then
the calculation is as follows:
ppm N in soil = ppm N in test solution x Aliquot dilution x Soil dilution
(nitrate form ) ( from curve)
(8-15)
ml final ml extraction
color volume solution
= ppm N in test solution x ·----·---·· ····-
ml aliquot
x --·---·- ··-·~-
gm O.D. soil
evaporated extracted

= ppm Nin test solution x 1OO x 250 ±.!11l_tlP.

10 50 - ml H 2 0
The dilution of the extracting solution by the soil moisture present in the
original moist sample is taken into account in equation ( 8-15) as ( +) and
( - ) ml H~O. In this way the results are based on the oven-dry weight of
· the soil.
8-65. For visual comparison in Nessler tubes:
. . ml standard solution -
ppm Nm test solution = -·· ------
ml test solution matched by standard
x ppm Nin standard solution (8-16)
Then computations are continued (eq. 8-15).
8-66. For runoff, the nitrate content is expressed as parts N per million
NITROGEN DETERMINATION -SOILS, PLANT TISSUE 201
parts of original runoff suspension and also per million parts solids in the
runoff. The quantity of N removed per unit area of field from which the
runoff was collected is calculated:
ppm Nin ppmN
original from
runoff working
suspension curve
ml final ml extracting
analytical solution solution
x ------. --- x ml runoff (8-17)
ml extracting so-
lution analyzed evaporated
The runoff from each tank is sometimes evaporated separately. The com-
posite nitrate analysis is calculated by proportion of each runoff suspension
needed to form a composite. Then:
ppai =ppm Nin composite runoff x 0.227 (8-18)
( nitrate-N)

This value is next corrected by factors for dilution of runoff by rain and
for the volume of the runoff occupied by solids (~ 10-68, 10-69). Finally,
the nitrate concentration on a solid basis is:

---- nitrate-N in runoff x 2 OOO OOO

pp 2 ms -_ ppai, (8-19)
ppai, solids in runoff ' '

NITRITE DETERMINATION
8-67. Nitrites accumulate in soils that are above a critical pH value of
7.7 ± 0.1. 30 Thus, in somewhat alkaline soils that have been fertilized
heavily with ammonium fertilizers, nitrites may accumulate instead of
nitrates and in similar amounts,:u up to almost 100 ppm. Otherwise, the
accumulation of nitrites is generally so neglible that they can scarcely be
detected.
8-68. No formal procedure is given here for nitrite determination. In
principle, the nitrite is extracted from soils by the same water extraction
employed for nitrate. In practice it is difficult to obtain a clear, colorless
extract without oxidation of some of the nitrite. The standard procedure for
its chemical determination is that employing sulfanilic acid and alpha-
naphthylamine. 82 The nitrite standard is prepared from AgN0 3 and NaN0 2 ,
from which AgN0 2 is prepared.

ao Martin et al., S.S.S.A. Proc., 7:223 ( 1942).

a1 Chapman and Liebig,S.S.S.A. Proc., 16:276 (1952).
:i2 Prince, Soil Sci., 59: 50 ( 1945); Fraps and Sterges, Tex. Agr. Exp. Sta. Bui. 439
(1931); Rider and Mellon, Ind. Eng. Chem., A.E., 18:96 (1946). An alternative pro-
cedure has also been proposed by Shinn, Ind. Eng. Chem., A.E., 13:33 (1941).
202 NITROGEN DETERMINATION-SOILS, PLANT TISSUE

NITRIFICATION RA TE OF SOILS
8-69. The nitrification rate of a soil is a measure of the rate of release
of available nitrogen by the organic matter in the soil. The nitrogen in
nitrate form is released by microbiological activity (a) from fresh organic
residues from crops and (b) from soil organic matter. Fresh organic matter,
if succulent and with a relatively low C : N ratio, releases nitrate more rapidly
than does dry carbonaceous material that may depress nitrate release tem-
porarily. Nitrate release from soil organic matter may be less rapid but
continues over a longer period, with as much as 10 per cent of the total N
being converted to the nitrate form. This determination assumes that the
NH4 + formed in the decomposition of organic matter is converted to
N0 3 - , so that the N0 3 - formed is a measure of both processes. 3 :i The
actual rate of nitrification may be higher under conditions of nearly com-
plete nitrate removal by a crop than under conditions of accumlation as
nitrate. The measurement of nitrification in the field is not meaningful un-
less the amount of nitrate used by the crop and lost by leaching is accounted
for. When incubated under optimum conditions, the accumulation of nitro-
gen in nitrate form may go to 100 pp2m in 6 weeks and exceed 200 pp2m
in 4 months.

APPARATUS

8-70. Sampling tube or spade, 6-mm (4 meshes per inch) sieve, a sample
container, a knife or scissors, a torsion balance with capacity of at least 3
kgm, a 2-liter crock, a mixing table, a paper towel for the crock cover,
string or rubber bands.

REAGENTS

8-71. Reagents needed include only those for N0 3 - determination (~

8-51 ).

PROCEDURE

8-72. Field Sampling. The soil samples are obtained by one of two
methods: (a) For bare soil, random cores are taken well distributed over
the area ( ~ 2-7), to a depth of the plow-layer, and (b) to obtain a repre-
sentative sample of the tops and roots of vegetation such as grass, clover,
or small grain, the vegetation is tramped down flat on the ground, a slant-
ing hole is dug to the plow-layer depth with a sharp flat spade, and a thin
slice ( 1 cm) of the vegetation and soil is taken to obtain a representative
sample. Several random samples are composited.

aa Russel et al., Soil Sci., 19:381 (1925).

NITROGEN DETERMINATION-SOILS, PLANT TISSUE 203
8-73. Mixing and Potting. As soon as possible after the field sampling,
the samples are mixed thoroughly by being passed through a- 6-mm sieve.
Any plant material present is cut up into small pieces so that a uniform
mixture can be made. The soil (in 4 replications) is filled into a weighed
crock to within an inch of the top, and the crock and contents are weighed.
The soil moisture percentage is determined on a 25-gm sample taken at
the time of filling the crock. Sufficient water is added to the soil in the crock
to bring it to approximately 0.7 or 0.8 of field-moisture capacity. At the
time of potting, soil nitrate and total N are determined.
8-74. Incubation and Sampling. The crock is covered with a paper towel
secured by a rubber band or string, and stored at a constant temperature of
28° to 30°C. Water is added to the soil in the crocks twice weekly to main-
tain the moisture level, the surface crust being pulverized by hand before
watering. Weekly mixing of the soil provides aeration, which is an im-
portant factor affecting the rate at which nitrification takes place. At the
end of the first week, the soil is removed, broken up, and mixed thoroughly.
Care is taken not to lose any soil or to contaminate it. The crock is jarred
to settle the soil, and water is added to the required weight. If the soil is
watered 2 or 3 days before sampling, it can be mixed more easily and a
more uniform sample can be taken for the N03 - determinations. At the
end of the second week, the above mixing procedure is repeated, and 50
gm of soil is removed for NOa- determination (~ 8-58) and 25 gm for
moisture determination. The new constant weight is recorded.

QUESTIONS
1. What forms of nitrogen in the soil are determined by the Kjeldahl
method?
2. What is the role of the HgO in the catalyst mixture?
3. How is the complete liberation of nitrogen insured without undue dan-
ger of nitrogen loss in the Kjeldahl digestion process?
4. How is the digestion temperature controlled?
5. What are the advantages of the disposal of fumes in water to the sewer
as compared to disposal in air by means of a fume chimney?
6. How is the Kjeldahl procedure for soils and plant tissue modified to in-
clude nitrates in the determination?
7. Why is the NaOH added so that it runs down to form a layer on the
bottom of the flask just prior to the distillation?
8. In the exchangeable ammonium or nitrate determinations, what precau-
tions must be observed in handling the sample prior to the extraction process?
9. Why is it not satisfactory to distill off the NH 3 directly by treatment of
the soil with hot alkali?
10. How is NH 4 + held in the soil, and how can it be extracted?
11. What is the usual range of NH 4 + concentration in soils?
 TISSUE.
204 NITROGEN DETERMINATION-SOILS, PLANT
in the extraction of soil
12. What is the purpose of each of the following
nitrates: CuS04 , Ag 2S04 , Ca(O 2H) , and MgC 0 8 ?
dryness in the nitrate test,
13. Why is the aliquot of the extract evaporated to
residue?
and what is the action of phenoldisulfonic acid on the
procedure?
14. What is the role of the NH 4 0H in the nitrate
amount of vegetation in
15. Why is it important to include a representative
determ inatio ns?
soil samples to be used for nitrification rate
 9
Organic Matter
Deter1ninatio11s for Soils
Organic matter . . . a distinction of soil from rock

9-1. Carbon occurs in soils in 4 forms of mineral and organic matter:

1. Carbonate mineral forms, chiefly CaCO:i and MgCOR · CaCOa (~ 10-
121); but highly active and important small amounts also occur as CO~,
and HCOa-- and COa-- ions of more soluble salts(~ 10-87).
2. Highly condensed, nearly elemental organic carbon (charcoal,
graphite, coal).
3. Altered and rather resistant organic residues of plants, animals, and
microorganisms, sometimes termed "humus" or "humate," but not, as these
latter terms tend to suggest, a single compound.
4. Little altered organic residues of plants and animals, and living and
dead microorganisms, subject to rather rapid decomposition in soils.

9-2. The total carbon of soils obviously includes all 4 forms. Total
organic carbon includes the latter 3, the mineral form being eliminated
by treatment with dilute reducing acid prior to the organic carbon deter-
mination. The most reproducible soil organic carbon determination is one
that includes all 3 forms of organic carbon without any attempt at fraction-
ation. The chemically active organic matter that is related to soil genesis
and fertility includes forms 3 and 4. Some effort is sometimes made, there-
fore, to exclude 2, the carbon in highly condensed form, from soil organic
matter determinations. Also, a distinction may be made between the older,
more resistant organic matter which is the humus proper, 3, and the
freshly added organic residues, 4, which are subject to rapid decomposition
and release of their nutrient elements to the current crop ( ~ 9-73). Al-
though fractionation, or distinction between forms of soil organic matter,
205
206 ORGANIC MATTER DETERMINATIONS FOR SOILS
involves loss in reproducibility of measurement, when perfected, it may
promote more significant interpretations of the analyses.
9-3. Total Organic Carbon of Soils. Organic carbon determination is
carried out (after removal of carbonate) by (a) dry combustion in a fur~
nace (~ 9-9), or (b) by chromic acid oxidation (~I 9-21), followed by
measurement of the C0 2 evolved. The C0 2 evolved may be measured by
volume, weight, or titration. The organic carbon content of soil may be
reported directly as percentage of C; or calculated as organic matter by
multiplication by a factor. The conventional carbon to organic matter
factor of long standing is 1.724, based on the assumption that soil organic
matter is 58 per cent C. This factor is sometimes referred to as the "Van
Bemmelen factor," although it is of older origin. Russel and Engel 1 state
that ". . . organic matter calculated by multiplying carbon content by the
conventional factor I .724 agrees closely with the direct determination by
means of hydrogen peroxide, individual discrepancies being within the
range of experimental error. . . ." Several other studies on the factor have
been reported. 2 The factor for a conversion of the carbon content of many
surface soils to organic content has been found: 1 to be 1.9, and the factor for
many subsoils is about 2.5. The variation in the carbon to organic matter
ratio makes it desirable to report the organic content rather than the carbon
content for comparisons between different horizons or between dissimilar
soils. Comparisons of organic carbon content as such may be satisfactory
for comparisons between similar soil horizons.
9-4. Oxidizable Soil Or~anic Matter Determination. The oxidizable
matter of soils is determined by (a) chromic acid oxidation with heat ap-
plied(~ 9-33), or (b) by chromic acid oxidation with spontaneous heat-
ing (~ 9-57). It can be carried out without prior carbonate removal be-
cause the CO~ evolved is not measured. The mcq of chromic acid reduced
may be measured by titration or by the green color formed. The meq may
be related to the total carbon or the total organic matter by calibration with
other methods. Oxidizable matter determination by chromic acid methods
is the most rapid and popular type of analysis and has the advantage of
moderately satisfactory discrimination of humus from highly condensed
forms including graphite and charcoal. The determination of oxidizable
matter is often termed the determination of "organic carbon"; however,
since the highly condensed forms of organic carbon are excluded, the term
"oxidizablc matter" should be given preference. The total organic carbon
thus would be reserved for those methods ( ~ 9-3 ) that do in fact measure
all forms of organic carbon without a correlation factor. Oxidizable matter

1 Proc. /st Intern. Congr. Soil Sci., 4:343 ( 1928).

~Leighty and Shorey, Soil Sci., 30:257 ( 1930); Lunt, Soil Sci., 32:27 (1931 ); and
Wilson and Staker,]. Am. Soc. A!?ron., 24:477 ( 1932).
B Broadbent, Adv. Agron., 5: 176 ( 1953).
ORGANIC MATTER DETERMINATIONS FOR SOILS 207
can be reported conveniently as the meq of oxidizablc matter per 100 gm
of soil. Conversion factors from oxidizable matter to carbon and organic
matter are often employed, and greatly facilitate comparison with the con-
tents of these constituents in various soils. The use of carbon to organic
matter factors other than conventional 1. 724 factor ( ~ 9-3) would not in-
fluence the net results of the chromic acid methods, because it would merely
change the assumed percentage recovery of organic matter, leaving the
over-all factor for conversion of titer to organic carbon or organic matter
the same.
9-5. Total Soil Organic Matter by Weight Loss. The total organic
matter content of soils is determined by (a) oxidation with H~O~ ( ~i 9-74),
(b) ignition at moderate temperatures ( ~ 9-85), or ( c) ignition after
decomposition of silicates with HF-HCI ( ~ 9-86). Distinction between the
determination, by these methods, of organic matter and organic carbon
(~[ 9-3) is based on the fact that the total organic matter weight is meas-
ured by the former, while only that of carbon is represented by the latter.
Serious errors enter into the various weight-Joss methods for measuring the
organic matter weight because other constituents-such as salts including
CO~ of carbonates, H~O, and OH (~[ 9-85 )-also influence the weight
change. Also, the organic matter may not be completely oxidized.
9-6. Other Estimates of Soil Organic Matter. Soil organic matter has
been shown to correlate (a) with total nitrogen content, (b) with climate,
and ( c) with clay content. Multiplication of the total nitrogen content of
soil by 20 roughly approximates the organic matter content. This assumes
5 per cent N in the organic matter of a C : N ratio of 1 1.6 since the or-
ganic matter is conventionally assumed to contain 58 per cent carbon
(~ 9-3). The estimation of soil organic matter from the nitrogen content
may be as accurate as from the carbon content.·1 Likewise, a correlation of
organic matter with the climate" and with the clay content of the soils of a
given region has been suggested by some investigators.
9-7. Determination of the specific compounds, organic structure, and
functional groups of organic matter in soils is a soil chemical analysis field
in itself. Organic exchange capacity (~ 4-19), car boxy I, phenolic, acetyl,
methoxy, and other functional properties can be measured. Specific molecu-
lar species carrying phosphate, such as phytin and inositol monophosphate,
have been isolated. Solubility classes include fats and waxes, carbohydrates,
lignins, and proteins. Alcohol, acetyl bromide, alkali, and pyrophosphate
dissolution techniques---<:oupled with acid and other types of precipitation
-give information on undecomposed plant residues, black pigment, brown
pigment, and other humus fractions.

4 Read and Ridgell, Soil Sci., 13: I ( 1922) .

. 5 Jenny, Soil Sci., 27: 169 ( 1929).
ORGANIC MATTER DETERMINATIONS FOR SOILS 209

Fig. 9-1. Fisher Induction · Carbon Apparatus. The sample in a combustion boat
is pfaced in a combustion tube at 9 and oxygen is passed into 1 and out at 2, sweeping
the C02 along. Other features are explained in Fig. 9- 2 and the text.

,....-,
I
-1 r-1 . r-1 -,I
I I I I
A I
I
B tI Combustion tube
c I
I D ·1 t
I .I I . I
I and coil
I I I
I ,, I
~'2'~e~
(1)
Solenoid
. valve
,I
I
I
I l~l
I
I
I
I
I
I
I
l Platinum
oxidizer (2)
I I . I I I
'' t t t t 't
I I I I I
I I I I I
I I I I I
L-..J L--' L-..J L_J L-..J

·A-Magnesium perchlorate
B-Caroxite
C-Manganese dioxide
D-Magnesium perchlorate
E·Flow meter

Fig. 9-2 • .Schematic diagram of gas train . of Fisher In4uction Carbon Apparatus.
Letters and numbers also refer to Fig. 9- 1.

(Complete apparatus is available from Fisher Scientific Co., Pitts,burgh,

'P a.) If carbonates are .to l?.e rem<;>ved by S0 2 , 50-ml beakers and capi1lary
aspirator tubes are ·required.

REAGENTS

9,-12. Needed reagents include tank oxyg~n; the three absorbants (~

9- 10) , including granular magnesiinn. perchlorate, a spedaf grade of man-
210 ORGANIC MATTER DETERMINATIONS FOR SOILS
ganese dioxide, and "Caroxite" (C0 2 absorbing chemical with color change
to indicate exhaustion); electrolytic iron powder; granulated tin, passing
0.5 mm sieve (30 meshes per inch); and alundum, passing 0.2-mm sieve
( 90 mesh per inch). (All reagents are available from Fisher Scientific Co.,
Pittsburgh, Pa.) A mixture is prepared, 50 per cent each by weight, of the
granular tin and electrolytic iron powder; 0.5 gm of the mixture is required
for each determination. If carbonates are to be removed by so2, tank so2
is required.
PROCEDUREll

9-13. The soil sample is air-dried, ground to pass a 0.2-mm sieve (80
meshes per inch), and thoroughly mixe-d. S:lmples containing the chloride
equivalent of I per cent NaCl or more are preleached to prevent positive
interference. Such interference is not prevented by the inclusion of Ag 2 S0.1-
H2S04-saturated pumice or silver wool in the train. The preleaching can
be combined with the carbonate removal procedure that follows.
9-14. Removal of Carbonate. If carbonate is present, either it must be
removed prior to the ignition, or its amount must be determined separately
( ~ 10-121 ) and its equivalent deducted from the co2 found by ignition.
Carbonate can be removed readily without oxidizing the organic matter by
treating the soil with sulfurous acid.
9-15. A soil sample equivalent to 2-gm or more of carbonate-free mate-
rial is placed in a 50-ml beaker, 15 ml of water is added, and then S02
gas is bubbled through the suspension with a capillary tube until all car-
bonates are destroyed. The reaction may be hastened by warming the sus-
pension slightly. The reaction is allowed to proceed over night. Then the
soil is dried at 100°C and thoroughly mixed, and the sample for the de-
termination is weighed out. In the presence of high amounts of Mn0 2 , the
addition of FeC12 aids in the prevention of oxidation of any organic matter.
9-16. Combustion. The alundum boat is lined with granular alundum
until it is half filled. Then approximately 0.5 gm of powdered electrolytic
iron is spread evenly over the alundum. The soil sample (0.2727 gm for
samples containing over 5 per cent carbon; 1.364 gm for samples contain~
ing less than 5 per cent carbon) is thoroughly mixed with approximately
0.5 gm of a 50-50 mixture of granular tin and electrolytic iron powder
combustion accelerator (~ 9-12). The mixture is transferred quantitatively
to the bed of electrolytic iron in the boat. The sample with the accelerator
is then covered with an additional 0.5 gm (approximate) of electrolytic
iron powder. The boat and sleeve are positioned in the induction carbon
apparatus with the special tool provided.
9-17. The apparatus is fired according to the directions supplied with

11 Results with this procedure have been published by the author, S.S.S.A. Proc.,
16:370 (1952).
ORGANIC MATTER DETERMINATIONS FOR SOILS 211
the instrument, and the co2 is collected and weighed in the weighing tube.
Each 10 mgm of C02 collected equals 1 per cent C for a sample weighing
0.2727 gm; each 50 mgm of C0 2 collected equals 1 per cent C for a sample
weighing 1.364 gm.
9-18. Each ignition cycle with a Fisher induction carbon apparatus re-
quires 2 minutes. With rapid weighing conveniences such as the "Gram-
atic" balance, the complete carbon determination for a soil sample can be
effected within 3 minutes.
ALTERNATIVE PROCEDURES

9-19. The C0 2 may also be collected in standard alkali and titrated

(~ 9-28).
9-20. The Lindberg high frequency combustion apparatus has been
employed 12 for the carbon in steel, and could undoubtedly be adapted for
carbon in soils. The C0 2 was measured gasometrically in a buret, the walls
of which were treated with a silicone to hasten volumetric equilibrium.

ORGANIC CARBON DETERMINATION AS co2

(Wet oxidation by chromic acid and determination by titration)
9-21. The soil organic matter may be wet oxidized by chromic acid and
the C0 2 collected for determination as in the dry combustion method. The
wet-oxidation method to be described differs from the chromic acid meth-
ods (,I 9-49, 9-65) for the determination of oxidizable matter in soils; the
C0 2 is measured in the present method, whereas the amount of chromic
ion reduced is measured in the methods for oxidizable matter. Though the
results may be shown to be highly correlated with each other, the deter-
minations are quite different in principle.
9-22. The carbonate carbon must be excluded with the wet-oxidation
method just as for the dry combustion method, by separate determination
( ~ 10-12 I ) or expulsion by means of sulfurous acid ( ~ 9-14).
9-23. Although gravimetric and titrimetric determination of the C0 2
are both appropriate for either the dry combustion or wet-oxidation pro-
cedures, the titrimetric determination is employed with the procedure given
here.
APPARATUS

9-24. Needed apparatus includes a small gas burner and wind shield, a
25-ml pipet, a buret, and a special apparatus (Fig. 9-3). The sample
flask is a 200-ml conical flask fitted with a 3-hole stopper. The condenser
tip is drawn out and a hole is blown in the side to permit free passage of
the gases without interference from condensation at the tip. The U-tube
has glass beads on the bottom, a layer of glass wool on each side, and 20-
mesh pumice filling each arm. The pumice in the left arm is saturated with
12 Pepkowitz and Chebiniak, Anal. Chem., 24:889 (1952).
212 ORGANIC MATTER DETERMJNATIONS FOR SOILS

Pumice with Ag1S04

---1+-Pumice with cone H 2 S04

Trap

Fig. 9-3. Apparatus for organic matter determination by wet-oxidation pro-

cedure and titrimetric determination of the C0 2 evolved. (After Heck, Soil Sci.,
28:225, 1929.)

concentrated Ag 2 S0 4 , and in the right arm with freshly boiled, concentrated

H 2S04 • The brom thymol blue trap is a side arm test tube. The bead tower
is constructed of a 2 x 25 cm test tube, the bottom of which is drawn out
and the sides pinched to hold the beads away from the bottom opening.
The beads are 4 to 6 mm in diameter. A little glycerol is applied to the out-
side of the bead tower to facilitate raising and lowering it. The soda lime
(or Ascarite) tower admits C0 2-free air to the apparatus.
REAGENTS
9-25. Required reagents include standard 0.5 N HCI, I per cent alco-
holic phenolphthalein indicator, 0.04 per cent brom thymol blue indicator
(1 ml of this in 10 ml of water in the trap, Fig. 9-3), C0 2 -free water, ap-
proximately 0.5 N NaOH (20 gm per liter), approximately 1 M BaC12
ORGANIC MATTER DETERMINATIONS FOR SOILS 213
(244 gm of BaC1 2 • 2 H 20 per liter of C02-free water), and the following
special digestion acids. ·
9-26. Chromic Acid. Approximately 340 gm of chromic oxide, Cr0 3 , is
dissolved in 400 ml of hot water and made to a volume of 1 liter with 85
per cent H 3 P0 4 •
9-27. Sulfuric Phosphoric Acid. Equal volumes of concentrated H 2 SO 4
and 85 per cent H 3 P0 4 are mixed together and stored in a glass-stoppered
bottle.

PROCEDURE13

9-28. A I-gm sample of 0.2-mm soil is weighed out and transferred to

the sample flask, after pretreatment with sulfurous acid (~ 9-14) if car-
bonates are present. (If the organic matter content is between 4 and 20 per
cent, a 0.5 gm sample is employed. A 0.2-gm sample of muck and peat is
employed.) The sample flask is then attached to the apparatus. To free the
train of C0 2 , a flow of C02 -free air is sent through for 1 or 2 minutes
longer than required for the brom thymol blue in the trap to turn from yel-
low to blue. Then 25 ml of approximately 0.5 N NaOH is pipetted into the
C0 2 absorption flask followed by 50 ml of C02 -free water. The condenser
cooling water is next started. While the air flow continues, the bead tower
is lowered into the NaOH until the solution reaches the top of the beads,
and then raised so that no more NaOH solution enters. The air flow is
regulated to about 3 to 5 bubbles per second.
9-29. Next, 10 ml of the chromic acid solution is added through the
separatory funnel, followed by 50 ml of the H 2 S04-H3 P0 4 mixture. A small
flame with a wind shield is placed under the sample flask, and the suspen-
sion is brought to boiling as rapidly as is possible without danger of froth-
ing over. The boiling is continued for 5 minutes longer than required for
the brom thymol blue in the trap to change from yellow to blue.
9-30. Back Titration. The clamp on the suction side of the C0 2 absorp-
tion flask is closed and the flask is detached. The NaOH is washed from the
bead tower and entrance tube with C02-free water. Next, 15 ml of 1 M
BaC12 solution is added to precipitate the carbonate. The excess NaOH is
then back titrated with standard 0.5 N HCl, 3 drops of 1 per cent phe-
nolphthalein indicator being employed. A standardization blank is made to
measure the NaOH concentration.
9-31. Calculation of Results. Since the carbonate is all precipitated as
BaC03 the net titration is as follows:
NaOH + HCl ~NaCl + H 20 (9-1)

13 Heck, Soil Sci., 28: 225 (1929); Friedemann and Kendall, J. Biol. Chem.,
82:45 (1929); White and Holben, Ind. Eng. Chem., 17:83 (1925). Robinson,
U.S.D.A. Cir. 139 (1939), employed this titration procedure in combination with
ignition to liberate the C0 2 •
214 ORGANIC MATTER DETERMINATIONS FOR SOILS
The C0 2 evolved may be calculated to its equivalent value as follows:
meq of C0 2 = (S - T) x N (9-2)
in which Sis the standardization blank titration of 25 ml of NaOH, Tis the
back titration, and N refers to the normality of the standard HCl.

% organic c = meq co2 x 0 ·6 (9-3)

s
in which s is the sample weight in gm and the factor 0.6 is the meq weight
(12/2000) x 100.

ALTERNATIVE PROCEDURE

9-32.- The C0 2 may be passed through a Mg(Cl0J 2 bulb to remove

water vapor and then may be collected in a gravimetric absorption bulb
and weighed instead of being titrated.
OXIDATION MATTER BY CHROMIC ACID WITH EXTERNAL
HEAT APPLIED
( Schollenberger)
9-33. To avoid the laboriousness of conventional carbon determination
methods involving ignition, collection of co2, and weighing, various rapid
approximate methods of determining organic matter in soils have been de-
veloped. These methods involve the oxidation of the organic matter by an
oxidizing agent added to the soil in excess, and the subsequent titration or
colorimetric determination of the excess oxidizing agent.
9-34. In the Schollenberger14 method, oxidizable matter in soil is oxi-
dized by chromic acid in the presence of excess H 2 S04 , with external heat
applied, the excess of standard chromic acid being back titrated with fer-
rous solution. This is a rapid method for approximating the organic matter
content, since a linear relationship between the titer and organic content
can be established for a given procedure. Moreover, the arbitrary measure-
ment of "oxidizable" material, expressed as meq per gm soil, has a direct
interpretation of its own, without regard to total organic content, as a meas-
ure of the active organic content of soil. By these methods, carbon from
carbonate does not interfere, and also the various forms of elemental car-
bon (graphite, charcoal, coal, etc.) are only attacked in part and thus are
mainly excluded from the measurement. In this regard, Allison 15 gave an
example that indicated discrimination between organic matter of soil and
the carbon in the form of cinders and coal by his modification of the Schol-
lenberger procedure. The Walkley-Black method which involves less heat-
ing has been shown (~ 9-58) to give almost complete discrimination .

. 14 Schollenb,erger, Soil Sci., 24: 65 ( 192 7).

15 Soil Sci., 40:311 (1935).
ORGANIC MATTER DETERMINATION S FOR SOILS 215

9.;..JS. The following consideration of the effect of chloride, manganese

dioxide, and ferrous iron applied equally to the Schollenberger and the
Walkley-Black (~ 9-57) chromic acid procedures.
9-36. Effect of Chloride. The reaction of Cr 2 0 7 - - with CI-:
Cr 2 0 1 ·· · + 6Cl· + 14H+ ~ 2Cr+++ + 3 Cl2 + 7 H 20
(9-4)
is quantitative, thus permitting an accurate use of an approximate chloride
factorrn of 1/12 (from eq. 9-6 below, C/4Cl- = 12/4 x 35.5) as follows:
Soil C content in % = Uncorrected soil C content in %
Soil Cl content in %
- ---· . -·~--~--···----- (9-5)
12
The correction factor has been found to be valid up to a Cl : C ratio of
5 : 1.
9-37. Alternatively, the chloride may be leached out17 on an asbestos
filter and the asbestos and sample together placed in the determination.
The oxidation of Cl - can be prevented by the use of Ag 2 SO4 in the diges-
tion mixture. HgO and HgS0 4 have also been found effective.
9-38. Effect of Higher Oxides of Manganese. Although the presence of
active higher oxides of manganese causes an error in the chromate titer of
the soil, inactive higher manganese oxides such as occur in ores cause no
interference. Ordinarily the content of higher oxides of manganese in soils
will cause no interference. However, any possible interference of highly
reactive (as freshly precipitated higher oxides of manganese) can be nulli-
fied by preliminary treatment of the soil with FeS0 4 • Such active oxides
react rapidly in the cold with acidified N FeS0 4 , and their amount can be
determined by preliminary titration. 1 R Two ml of H 3 P0 4 , 5 ml of water, and
1 ml of diphenylamine indicator are added to the soil followed by sufficient
N FeSO 4 to give an excess as judged by the color of the indicator ( 5 ml
usually suffices). The mixture is allowed to stand with occasional agitation
for 10 minutes, and then the excess FeS0 4 is back titrated with K 2Cr 2 0 7 •
The amount of FeS0 4 oxidized by the Mn0 2 , as determined by this titra-
tion, is then added to the fresh soil sample together with 2 ml of H 3 P0 4 •
The mixture is allowed to stand about 5 minutes, whereupon most of the
Mn0 2 will have been dissolved and what remains may be neglected. The
K 2 Cr 2 0 4 solution is then added, and the digestion of soil organic matter is
carried out by the usual procedure.
9-39. The use of H 2 0 2 by Degtjareff 19 in his modification would cause

16 Walkley, J. Agr. Sci., 25: 398 ( 1935); Soil Sci., 63: 257 ( 1947).
11 Schollenberger, Soil Sci., 59: 53 ( 1945). ·
1s Walkley, Soil Sci., 63:257 (1947). ·
19 Soil Sci., 29:239 (1930).
216 ORGANIC MATTER DETERMINATIONS FOR SOILS
the decomposition of the Mn0 2 but would still involve the quantitative
error.
9-40. Effect of Ferrous Iron. Ferrous iron in soils, if present, leads to
high results for the chromic acid titer of soil organic matter. 20 However, soil
samples that have been air-dried for 1 or 2 days contain insignificant
amounts of soluble ferrous compounds, even though the ferrous had been
high in the fresh sample. Thus no interference with the organic matter de-
termination is caused by the ferrous iron.
9-41. Use of an iron or steel mortar is avoided because of the introduc-
tion of reducing material in the form of metallic iron.
9-42. The Chromic Oxidation Equivalent of Soil Organic Matter. The
reactions of dichromic acid with soil organic matter may be represented,
separately with organic carbon and organic hydrogen, as follows:
4 Cr11 + + 3C 0 --
(acid)
4 Cr++ + + 3 C4 + (9-6)

2 H 2 Crp 7 + 3 C + 6 H 2S04 - - 2 Cr 2 (S0 4 )a + 3 C02 + 8 H 20

0

(9-7)
(9-8)
The H and 0 content of the organic matter is ordinarily not considered in
the stoichiometric relationships. However, in methane, CH 4 , the organic H
would require ( eq. 9-8) as much di chromic acid as would the carbon. As
the carbon chain length increases, the H drops to the ratio 2H : C; in
phenolic compounds, to H : C. !so-linked carbon compounds have still
less hydrogen present. The oxygen in groups attached to the organic matter
complex, R, lowers the dichromate titer that would be required by the car-
bon:
RCOOH - - RH
(acid)
+ CO~- (9-9)

Thus the presence of organic oxygen tends to counterbalance the effect of

organic hydrogen in the titration. Also, some of the carbon may not be
oxidized.
9-43. These various compensating factors apparently summate to give
74 to 100 per cent of a 4-electron change equivalent for each carbon. A re-
port21 that the C02 evolved by wet oxidation with chromic acid corre-
spond closely with the quantity of reduction of chromic acid, suggest a near
balance of organic oxygen and hydrogen effects. External heat was applied,
so the procedure is similar to that of Schollenberger. An oxidation factor 22
of 90 per cent approximates the results of the dry combustion method; 84

20 Walkley, Soil Sci., 63:257 (1947).

21 Aldrich et al., Soil Sci., 59: 299 (1945).
22 Schollenberger, Soil Sci., 59:53 (1945).
ORGANIC MATTER DETERMINA TIONS FOR SOILS 217

to 91 per cent oxidation factors were found for a number of soils in this
laboratory by the procedure described here ( ~ 9-49).

APPARATUS

9-44. Needed apparatus includes an electric hot plate with adjustable

temperature; 250-ml and 400-ml beakers; 30 x 200 mm Pyrex test tubes;
a 2-liter Pyrex beaker (bath); a 1-liter shallow Pyrex tray (bath); a ther-
mometer, 0° to 250°C; a torsion balance, ±0.05 gm; a capillary tube stirrer
and compressed air supply; stirring rods with ends flattened; volumetric
pipets; and a 50-ml buret.

REAGENTS

9-45. Needed reagents include 85 per cent H 3P0 4 , both for the baths
and as a reagent, and the following special reagents.
9-46. Standard 0.4 N Chromic Acid Solution. Exactly 19.61 gm
K2 Cr 20 7 (oven-dry) is dissolved in about 50 ml of water and then the
solution is diluted to one liter with concentrated H 2 S0 4 •
9-47. Ferrous Ammonium Sulfate Solution, 0.2 N. Exactly 78.44 gm of
Fe(NH 4 ) 2 (S0 4 ) 2 • 6 H 2 0 is dissolved in 300 ml of water containing 20 ml
of concentrated H 2S04 , and the solution is diluted to 1 liter with water. This
solution is made freshly, or titrated against the standard chromic acid each
day.
9-48. Orthophenanthroline Indicator. 0.025 M solution ("Ferroin" from
G. F. Smith Chemical Co., Columbus, Ohio).

PROCEDURE21l

9-49. The Soil Sample. The soil sample is ground to pass a 0.2-mm (80
meshes per inch) sieve and 0.25 gm of mineral soil (0.05 gm of peat, 1.00
gm of soil having less than 1 per cent organic matter) is placed into a 250-
ml beaker ( or 30 x 200 mm test tube).
9-50. The Runoff Sample. If the beaker in which the runoff was dried
( ~ 10-71 ) contains more than 1 gm of solids, they are transferred from
the beaker, ground to pass the 0.2-mm (80 meshes per inch) sieve, and a
0.25 gm sample (more if the organic matter content is less than 2 per cent)
is weighed out into a 250-ml beaker or 30 x 200 ml test tube. Eight test
tubes can be digested simultaneously in the H 3 P0 4 bath in a 2-liter beaker.
9-51. If the beaker containing the dried runoff (~ 10-72) contains less
than 1 gm, the entire residue is analyzed directly in the 250-ml beaker, the
shallow Pyrex tray with H 3P04 being employed for the heating bath.

2a The procedures given are modified from Schollenberger, Soil Sci., 24:65 (1927),
59:53 (1945); Allison, Soil Sci., 40:311 (1935); Wilde and Patzer, J. Am. Soc.
Agron., 32:551 (1940). Dr. H. F. Massey studied the heating rate and bath pro-
cedures with soils of standard organic carbon contents.
218 ORGANIC MATTER DETERMINATION S FOR SOILS
9-52. Oxidation of Organic Matter. From a pipet, 20 ml of 0.4 N
chromic acid solution ( 50 ml for peats) is added to the soil sample in the
250-ml beaker or test tube, and a similar quantity is taken for the standardi-
zation blank. The vessel with mixture is placed· in the HaP0 4 bath and
heated on an electric hot plate at such a rate that a temperature of 155 °C
is reached in 20 to 25 minutes. 24 The contents of the tube or beaker are
mixed every 5 minutes during the heating period. The temperature is held
at 155° to I 60°C for an additional 5 minutes. The thermometer is kept in
the blank, which is simultaneously heated, to follow the solution tempera-
ture.
9-53. The vessels with samples and blank are then removed from the
bath, are allowed to drain in air for 30 seconds, and are then placed in a
water bath at room temperature for 2 minutes. The thermometer is re-
moved with care not to break it by thermal shock.
9-54. Back Titration. The chromic acid solution, now cooled to room
temperature, is diluted with water to 75 to 200 ml, either in the tube or
250-ml beaker. Then 5 ml of 85 per cent H;1P0 4 and 4 drops of ortho-
phenanthroline indicator are added. The solution is back titrated with the
0.2 N ferrous ammonium sulfate until the solution color turns from green
to red at the end point. An air-jet stirrer is used with the tubes. The color
at the start is dark brownish, and then shifts sharply from blue to red at
the end point. The blank is similarly titrated. More chromic acid should be
added to fresh samples if the amount added proves to be inadequate; not
over one-half of the chromic acid should be consumed by oxidation of the
organic matter.
9-55. Calculation of Results for Soils. The percentage of organic matter
in soil is estimated as follows:

% OM (in soil).= 20(1 - I) x 0.92 (9-10)

S = Standardization blank titration, ml of approximately 0.2 N ferrous

solution
T = sample titration, ml of approximately 0.2 N ferrous solution

0.92 is an empirical factor derived as a product:

(0.4 N) x ~ x I. 724 x lOO = 0.92

4000 0.90 0.25
when 0.4 N is the normality of the chromic acid and 0.25 is the sample
weight in gm. In deriving the factor, it is assumed a 4 valence change of
carbon occurs, 58 per cent carbon occurs in soil organic matter, only carbon

24 Heating the chromic acid in an electric oven has been described by Purvis and
Higson, Ind. Eng. Chem., A.E., 11: 19 (1939).
ORGANIC MATTER DETERMIN ATIONS FOR SOILS 219

is oxidized, and only 90 per cent of the total soil organic matter is oxidized,
the assumed figures being somewhat arbitrary.
9-56. Calculation of Results for Runoff. The percentage of organic
matter in the runoff solids is calculated as for soils, as percentages of or-
ganic matter, the appropriate sample weight being inserted in equation
9-10. Then the pounds of organic matter per acre inch of runoff is calcu-
lated from this percentage and the pounds of solids per acre inch ( ~ l 0-
73 ).

OXIDIZABLE MATTER BY CHROMIC ACID WITH H2 S0 4 HEAT

OF DILUTION
(Walkley-Blac k)
9-57. The chromic acid method based on spontaneous heating by dilu-
tion of H 2 S04 (Walkley-Black method) involves essentially the same pro-
cedure as that of Schollenberger (~! 9-49) except that the heating is less
than that externally supplied. For this reason, somewhat Jess of the total
organic matter is oxidized, and this is held by some workers to be an ad-
vantage, since the less active organic matter is not measured.
9-58. Exclusion of Elementary Carbon. The lesser heating largely dif-
ferentiates soil humus matter from extraneous sources of organic carbon
such as graphite and charcoal, a distinct advantage. While dry combustion
methods include all of the elementary carbon, the Walkley-Black method
excludes 90 to 95 per cent of it. 2 '' Oxidation of flake graphite stopped at
the graphite oxide stage as indicated by its typical swelling reaction when
warmed while moistened with water. The method of Tiurin, 2 ij involving
greater oxidation intensity, gave 84 per cent recovery of charcoal, ac-
cording to Walkley.
9-59. The effects of chloride, manganese dioxide, and ferrous iron on
the chromic acid titration methods are discussed in ~l 9-36 to 9-41.
APPARATUS
9-60. Needed apparatus includes 500-ml conical flasks, pipets of 10-
ml and 20-ml volume and preferably self-adjusting to zero, a buret for the
ferrous solution, and an analytical balance. A storage vessel to prevent
oxidation of ferrous iron is a convenience and may be arranged to keep
hydrogen or C0 2 over the solution. For convenience, the chromate solution
is employed as the standard of reference and the ferrous concentration is
determined only in ratio to the chromic acid.
REAGENTS
9-61. Required reagents include 85 per cent HaP0 4 , solid NaF, concen-
trated H 2 S0 4 not less than 96 per cent, and the following special reagents.
25 Walkley, Soil Sci., 63: 251 (1947).
26 Pedology, 26:36 (1931 ); Trans. Dokuchaev Soil Inst. (U.S.S.R.), 10:27 (1934).
220 ORGANI C MATTER DETERM INATION S FOR SOILS
9-62. Standard 1 N K 2 Cr2 0 7 • Exactly 49.04 gm of K 2Cr 20 7 is dissolved
in water, and the solution is diluted to 1 liter.
9-63. Diphenylamine Indicator. Approximately 0.5 gm of reagent grade
diphenylamine is dissolved in 20 ml of water and 100 ml of concentrated
H 2 S04 •
9-64. Ferrous Solution. The ferrous concentration may be made stand-
ard and preserved or made approximate and the chromate used as the
standard of reference. A 0.5 N solution of ferrous is prepared by dis-
solution of 196.1 gm of Fe (NH 4 ) 2 (SO 4 ) 2 • 6 H 2 0 in 800 ml of water con-
taining 20 ml of concentrated H 2S0 4 and dilution to 1 liter. Walkley
27

employed 278.0 gm of FeS0 4 • 7 H 2 0 per liter with 15 ml of concentrated

H 2 S04 , a 1 N solution.

PROCEDURE28

9-65. Pretreatme nt to eliminate readily oxidizable Mn0 2 (~ 9-38) is

given if needed.
9-66. Oxidation of Organic Matter. The 0.5-gm soil sample (0.05 gm
for peat; 2.00 gm for soils less than 1 per cent organic matter), having
passed a 0.2-mm ( 80 meshes per inch) nonferrous sieve, is placed in a
500-ml conical fl.ask. Next, exactly 10 ml of 1 N K2 Cr 2 0 7 solution is
pipetted onto the soil, and the 2 are mixed by swirling the flask. Then 20
ml of concentrated H 2 S0 4 is added and mixed by gentle rotation for 1 min-
ute, to insure complete contact of the reagent with the soil, with care to
avoid throwing soil up onto the sides of the flask out of contact with the
reagent. The mixture is allowed to stand 20 to 30 minutes. A standardiza-
tion blank (without soil) is run in the same way.
9-67. Back Titration. The solution is diluted to 200 ml with water, and
10 ml 85 per cent H 3 P0 4 , 0.2 gm NaF, and 30 drops of diphenylamine in-
dicator are added. The solution is back titrated with ferrous ammonium
sulfate solution delivered from a buret. The color is dull green with
chromous ion at the beginning, then shifts to a turbid blue as the titration
proceeds. At the end point, this color sharply shifts to a brilliant green,
giving a I -drop end point. If over 8 of the 10 ml of chromic acid has been
consumed, the determination is repeated with a smaller soil sample.
9-68. Calculation of Results. The results are calculated by the equation:
T
% OM = 10(1 - S) x 1.34 (9-11)

S = standardization blank titration, ml ferrous solution

T = sample titration, ml ferrous solution

21 Walkley, Soil Sci .. 63:251 (1947).

28After Walkley and Black, Soil Sci., 37:29 (1934); Walkley, J. Agr. Sci., 25:598
(l935);Soi1 Sci., 63:251 (1947).
ORGANIC MATTER DETERMINATIONS FOR SOILS 221
The factor 1.34 is derived as follows:

(1.0 N) x _g_ x t.n x IOO = 1.34

4000 0.77 0.5
in which 0.5 is the sample weight, 1. 72, the factor for organic matter from
carbon, and 12/4000, the meq weight of carbon. The 77 per cent recovery
factor found by Walkley 29 is used; a recovery factor of 74 per cent had been
suggested by Smith and Weldon 30 for Nebraska soils. Good correlation of
the Walkley-Black method with the dry combustion method was obtained
for Florida soils when the 77 per cent recovery factor was employed. 31
9-69. If the results are expressed in terms of readily oxidizable organic
matter, the recovery factor ( 1/0.77) is omitted, thus giving:

% OM (readily oxidizable) = 10(1 - ~) x 1.03 (9-12)

The readily oxidizable organic matter may be calculated by the equation:

T 100
(9-13)
meq per 100 gm= 10(1 - -) x -
s s
wherein s is the sample weight.
9-70. The correction for chloride (~ 9-36 and eq. 9-5) is applied if
CJ- is present.

ALTERNATIVE PROCEDURES

9-71. Colorimetric Chromic Acid Procedures. The quantity of chromic

acid reduced by the soil organic matter has been measured colorimetrically
instead of titrated. 32 The soil is filtered ofP 3 after dilution with water, and
the green chromous color is measured with a photoelectric colorimeter and
a light maximum of 645 mu. Under these conditions the amount of chro-
mate added need not be measured accurately. The change in the orange
chromic color may be measured instead, but that proc'.!dure is not as satis-
factory from several standpoints. The color changes may be approximated
visually 34 as a rapid test of soil organic matter (orange color-low, green
color-high).
9-72. Permanganate Use in the Titration. The Walkley-Black procedure
was slightly modified311 by addition of an excess of ferrous solution to the
chromic acid after the digestion, then back titrating the ferrous with stand-

29 Walkley, Soil Sci., 63:251 (1947).

ao S.S.S.A. Proc., 5: 177 (1941).
31 Dyal and Drosdotf, Proc. Soil Sci. Soc. Fla .. 3:91 (1941).
32 Graham, Soil Sci., 65: 181 ( 1948).
33 Carolan, Soil Sci., 66:241 (1948).
34 Wilde, S.S.S.A. Proc., 7:393 (1943).
35 Smith and Weldon, S.S.S.A. Proc., 5: 177 (1941).
222 ORGANIC MATTER DETERMIN ATIONS FOR SOILS
ard KMn04 • So long as the same volume of ferrous and dichromate are
added to the blank and to the deterqiination, no solution need to be stand-
ardized except the KMn0 4 • This method was further extended 36 to smaller
samples consisting of 0.1 to 0.3 gm of soil.
9-73. Readily Oxidizable Organic Matter Fractions of Soils. Oxidation
of the more readily available organic matter ( ~ 9-2) by means of a more
dilute chromic acid solution or more dilute H 2 S04 has been considered as
a means of measuring the readily oxidizable organic matter that is possibly
related to current crop production. However, the content of fresh organic
matter may not be correlated with its supplying power for available nitro-
gen, if it has a high carbon : nitrogen ratio. Thus the content of readily
available organic matter might be correlated with an adverse effect on crop
production. Even the measurements of ammonium ion released through the
action of mild oxidants is a difficult method to calibrate to measure the
effect of fresh organic matter on nitrogen supplying power.

TOTAL ORGANIC MATTER BY H 20 2 OXIDATIO N AND

WEIGHT Loss:17
9-74. Decomposition of organic matter by treatment with 30 per cent
H 20 2 is effected at temperatures below 110°C, and total organic matter is
estimated gravimetrically by the weight loss. This low temperature permits
the retention of the hydroxyl and strongly sorbed water in the mineral col-
loids. Thus the determinatio n of organic matter is more accurate than can
be achieved by ignition temperatures high enough to oxidize the organic
matter thermally, because H 2 0 and OH are also lost by the latter procedure
(~ 9-85).
9-75. Elemental carbon and resistant, paraffin-like organic material is
not attacked by the H:P 2 (~ 9-2). To some extent structural (nonhumi-
fied) organic matter is not attacked, especially if the H 2 0 2 is more dilute
than 6 per cent. Up to 16 per cent of the original carbon content has been
found 38 in the residues from the H 20 2 method. In this respect, the H 20 2
method is comparable to the chromic acid methods for oxidizable matter.
9-76. The free Mn0 2 present in soil is changed by the procedure to
Mn++ and finally to Mn:p 4 • The CaCOa is decomposed with acid, the soil
being brought to pH 5.8 or less to increase the effectiveness of the organic
matter oxidation with H 2 0 2 • The carbonates are formed again at the end of
the procedure.

Chesnin, Agron. lour., 42: 385 ( 1950).

36
After procedure of Robinson, J. Agr. Res., 34:339 (1927), as modified by J.C.
37
Russel and described by Judd and Weldon, J. Am. Soc. Agron., 31:217 (1939).
ss Alexander and Byers, U.S.D.A. Tech. Bui. 317 (1932).
ORGANIC MATTER DETERMINATION S FOR SOILS 223

APPARATUS

9-77. Needed apparatus consists of tall-form 250-ml Pyrex beakers and

closely fitting cover glasses; a gas or electric hot plate; 100-ml centrifuge
tubes and centrifuge; 75-ml platinum or rhotanium dishes; an analytical
balance and counterpoised scoop; a 110°C oven; and a series of 50-ml
glass-stopped weighing bottles, numbered in ascending order of weights,
arranged so the odd numbered bottle of each pair is slightly lighter. The
even numbered bottles are dried and weighed to the nearest mgm.
REAGENTS

9-78. Needed reagents consist of 0.1 N HCl; 30 per cent H 2 0 2 , reagent

grade, residue free, in wax-lined bottles; 10 per cent (NH 4 ) 2 C0:1 pre-
pared by dissolving I 0 gm of reagent grade salt in l 00 ml of water. The
latter solution is filtered if not entirely clear; 10 ml should not leave a
weighable residue after evaporation to dryness and heating in an oven at
l 10°C.

PROCEDURE

9-79. The Samples. A system of weighing is employed that minimizes

systematic errors, number of calculations, and time. Two successive 2-gm
samples of air-dry soil having been passed through a 0. I 5 mm (I 00 mesh
per inch) screen are weighed out on a counterpoised scoop and transferred
to 2 paired weighing bottles that have stood open to weight-equilibrium in
air. (For soils containing over 5 per cent organic matter, a 1-gm sample is
employed. A 0.2-gm sample of peat or other organic material is employed.)
The bottles are tightly stoppered as filled. When the series of samples has
been weighed out, the pair of bottles containing samples of I soil is placed
on the balance, the odd numbered on the right pan, and the other on the
left. The difference in their weights is then determined to the fourth decimal.
The even-numbered bottle is set aside for the time and the sample from
the odd-numbered bottle is transferred to a 250-ml tall-form beaker, a little
water being employed to rinse out traces of soil. The weighing bottle is set
aside.
9-80. Acidification of the Sample and Removal of Carbonates. Ten ml
of water is added to the soil and then 4 ml of 0.1 N HCI. If the soil is
calcareous, 4 additional ml of 0.1 N HCl is added for each 1 per cent of
CaC03 present. The suspension is stirred to facilitate reaction, and finally
is warmed on the hot plate for an hour. A drop of salt-free brom cresol
green indicator is added, which should remain yellow or green, indicating
acidity greater than pH 5.8 (BCG is blue at pH 5.8 and above). If the in-
dicator turns blue, additional 0.1 N HC1 is added and digestion is continued
until the pH remains below 5.8. The sample is then slightly acid and car-
bonate-free, ready for organic matter oxidation.
224 ORGANIC MATTER DETERMINA TIONS FOR SOILS
9-81. Oxidation of Organic Matter. To the 250-ml tall-form beaker con-
taining the slightly acid soil suspension, 10 ml of 30 per cent H 20 2 is added,
the beaker is covered with a tightly fitting watch glass, and the suspension is
digested on a gas or electric hot plate that is carefully regulated to avoid
frothing over of the sample. From time to time the suspension is mixed
by rotation of the beaker. After all the peroxide is decomposed, and the
solution has evaporated to a volume of approximately 5 ml (beaker un-
covered briefly if necessary), 5 ml additional 30 per cent H 2 0 2 is added to
rinse down the sides of the beaker. The digestion is continued until all
peroxide is decomposed. Oxidation is usually complete at this point, but
further additions of H 2 0 2 may be required for complete oxidation of the
organic matter of some soils.
9-82. After digestion is complete, the cover glass is scrubbed with a
policeman and rinsed into the beaker. The contents of the beaker are trans-
ferred to a 100-ml centrifuge tube into which 5 ml of the 10 per cent
(NH 4 ) 2 C03 solution has been previously placed, a policeman being used
to complete the transfer. The beaker is covered and set aside to be used
again, to receive the soluble salts from the sample. Finally the suspension
in the tube is mixed with a strong jet of water, and the tube is set aside for
the soil to flocculate. The (NH4 ) 2 C03 serves also to reconvert the CaC12
and MgC1 2 formed by the HCl treatment back to CaC0 3 ; each 5 ml is
equivalent to 25 per cent CaC03 in the soil. After flocculation has oc-
curred, the suspension is centrifuged until the supernatant liquid is entirely
clear. This clear liquid is decanted into the original beaker in which the
soil was digested, and saved for the procedure in the next paragraph. Two
additional washings are given the residue by resuspending with a water jet,
adding 5 ml of 10 per cent (NH 4 ) 2 C03 , flocculating, centrifuging, and
decanting it into the beaker. Finally it is resuspended with a jet of water,
and quantitatively rinsed into the respective original weighing bottle with
the aid of a policeman. With care and a fine water jet, the transfer can be
made with 1 filling of the weighing bottle. The weighing bottle and contents
are placed in the oven, the liquid evaporated to dryness, and the bottle and
cover finally placed beside the companion bottle (even-numbered) contain-
ing the moiswre sample. Drying at 110°C is continued for 8 hours or over
night. The pair of weighing b0ttles i:s cooled in the same desiccator and the
difference in their weights is determined ( even--numbered bottll.e on left
,pan .as before) to the nearest 0.1 mgm. The odd-numloered IDottle is re-
,m<i>ved and the other bottle and conten1'.s weighed to the nearest 1 mgm.
'ifhe net oven-dry weight is determined by !Mllbtracting the weight of the
weighing bottle.
9-83. To determine the weight of the soluble salts removed from the
sample, a series of platinum or rhotanium dishes is weighed, cooled in a
.des,iccator, and weighed again to the nearest 0.1 mgm. The1;1 the suyer-
ORGANIC MATTER DETERMINA TIONS FOR SOILS
natant solution from the beakers is transferred to the dishes, evaporated to
dryness, and ignited at 550°C (medium red) in a muftle furnace for a few
minutes. The soluble salts are thus dried, excess NH4 Cl is removed (formed
from the HCl treatment), and the solublized soil organic matter volatized.
The dishes are cooled in a desiccator and weighed, and the weight of the
ignited residue is calculated.
9-84. Calculation of Results. The weight of organic matter is calculated
as follows:

Weight of _ Final weight difference Initial weight difference Ignited residue

organic matter - in weigh bottles - in weigh bottles - in dish
(9-14)

. (weight of organic matter) x 100 (9-15)

% Orgamc matter= - - ··---··-·-- ------·---
(oven dry sample weight)

ALTERNATIVE PROCEDURES

9-85. Ignition at Low Temperature. A procedure for the oxidation of

organic matter at moderate temperature is described by Mitchell. 39 Dis-
crimination of organic matter loss from weight loss of H 2 0 and OH is
based on selection of temperature ranges: the H 2 0 is driven off at l 10°C;
the organic matter is oxidized by heating at 350° to 400°C for 7 to 8
hours; and the mineral matter is assumed to be unchanged at these tem-
peratures. Unfortunately, the discrimination between organic and mineral
matter is far from complete, particularly for soils containing amorphous
materials.
9-86. HF-HCI Pretreatment. A proposal was made by Rather 40 to de-
compose the hydrous silicates prior to heating to oxidize the organic mat-
ter, and thus to eliminate the heating weight loss due to water and hydroxyl.
The silicate is decomposed by HF-HCI treatment and the residue is then
weighed. The organic matter is driven off by heat, and the weight loss in
the absence of H 2 0 and OH effects is noted. Complete preservation of the
organic matter while the silicate is being decomposed presents a problem,
and loss of some organic matter during the process of washing out the
salts presents another. The method, while requiring much time, is worthy
of con3ideration for some purposes. Alexander and Byers 41 give a modified
Rather procedure, but consider it long and tedious. Gottlieb and Hen-
dricks42 employed the Rather procedure for extraction of organic matter
for comparison to that extracted with NaOH.

3ft J. Am. Soc. Agron., 24:256 (1932).

40Ark. Exp. Sta. Bui. 140 (1917).
41U.S.D.A. Tech. Bui. 317 (1932).
42S.S.S.A.Proc., 10:117 (1946).
226 ORGANIC MATTER DETERMINATIONS FOR SOILS

QUESTIONS
1. List the 4 general forms in which carbon may occur in soils.
2. State representat ive ratios of carbon to organic matter of soils.
3. What is the ratio of organic matter to nitrogen in the more resistant
stabilized portion of the organic colloids of soils?
4. How is carbonate excluded from the determinat ion of organic carbon of
soils?
5. Why is the determinat ion of soil organic carbon as CO~ by means of the
dry combustion method or the wet oxidation method a fairly reproducib le
determinati on?
6. What methods are available for the determinat ion of CO~ evolved from
the soil organic carbon?
7. Why is it more proper to speak of the chromic acid methods as de-
termination s of "oxidizable matter" rather than as methods for organic carbon
determinati on? Jn what sense are they estimates of total organic carbon?
8. What are the percentages of recovery of total organic carbon for (a) the
chromic acid method in which external heat is applied, and ( b) the chromic
acid method in which no external heat is applied?
9. Why is it possible to employ a ferrous solution of a concentrati on not
known exactly in the back titration of the chromic acid?
10. Why may the hydrogen peroxide method be termed a direct determina-
tion of soil organic matter as distinguished from the CO~ measureme nt methods?
11. What principles are applied in the ignition loss method for organic mat-
ter to distinguish the organic from H~O and OH constituent s? To what extent
is it successful?
 10
Soluble Salt Analysis
for Soils and Waters
The salt of the earth
-FROM AN OLD PROVERB

10-1. All fertile soils have at least small amounts of soluble salts in them.
The exchangeable cations equilibrate with the H~C0:1 dissolved in soil
moisture, yielding soluble carbonates and bicarbonat es of the metallic ca-
tions and leaving the correspond ing amount of hydrogen ion on the ex-
change. Traces to 100 ppm or more of nitrogen occur as nitrate salts in
soils. Natural waters from rivers, lakes, and wells contain varying amounts
of dissolved salts. Runoff waters from soil carry soluble salts as well as
suspended solids. Lysimeter leachates contain dissolved salts. The ac-
cumulatio n of soil salts in larger amounts is mainly through influx of
seepage, runoff, and irrigation waters, followed by concentrat ion by evapo-
transpirati on. Varying amounts arise from the processes of nitrification,
sulfofication, acidification, and fertilization.
10-2. Soil Salinity. When a soil contains an excess of soluble salts, it is
termed a saline soil. Sometimes it is called a "white alkali" soil because of
the white saline crust that sometimes appears on drying. Occurrenc e of soil
salinity problems fall into 2 main classes:
1. Natural occurrence of excess salts in soil, in the absence of adequate
drainage, usually in semiarid or arid regions, 1 but also through marine
waters or sediments even in humid and tropical areas.

IThe salt problem in irrigation agriculture is reviewed by Richards, ed., Diagnosis

and Improvemen t of Saline and Alkali Soils, U.S.D.A. Agr. Handb. 60 (1954); and by
Hayward and Magistad, U.S.D.A. Misc. Publ. 607 ( 1946).
227
228 SOLUB LE SALT ANALYSIS FOR SOILS AND WATE RS
2. Occurrence of excess salts in soil as a result of fertilization, trouble-
This
some in highly fertilized greenhouse soils2 and in fertilizer bands.
soil regions , and include s the
situation occurs frequently, even in humid
problem of salt index of fertilize rs.
s
10-3. Soluble salt analysis for soils and waters generally concern
ger-
whether enough salt is present to cause interference with normal seed
of the
mination, plant growth, or plant intake of water. The determination
species in the soil salts is also import ant in the
actual amoun t of each ionic
interfer ence with plant functio n. Be-
interpretation of the extent of their
tration and compo sition of salts in
sides the determination of both concen
salt analysi s include s the
the soil solution and natural waters, soluble
m)
analysis of salt crusts and salt deposits such as Ca SO 4 • 2 H 2 0 (gypsu
and Ca CO a (calcite ) in the soil profile.
over
10-4. Water soluble salts occurring in soils in total amounts of
Ca++,
0.1 per cent usually consist principally of the four cations Na+, K +,
to Cl and S0_ sometim es to NOa and
and Mg++, linked mainly 1 - - ,

HCOa -- . Usually 98 per cent of the

C0:1 - - , and to a limited extent to
ions. Soil salinity problem s3
soluble salts of saline soils consist of these
Mg+ -t,
frequently arise from Na+, CI-, S0 4 - - but seldom from Ca+',
high HC0: con-
or C0:1 - -- • Chlorosis is sometimes associated with 1 -

borate
centration coupled with high pH (~! 10-5). Toxicity to plants of
in saline
ion even in small amounts is important in irrigation waters and
are also toxic to plants if present in soil
soils. The ions Mn++ and Al+++
as soluble salts at more than quite low concen trations .
or its
10-5. Soil Alkalinity. Associated with the occurrence of salinity
in general is evidenc ed
aftermath is the occurrence of soil alkalin ity, which
red (pH 8.3 to
by sufficiently high soil pH to turn phenolphthalein pink or
age ( 15
10 or 11 ) . This range is frequently associated with a high percent
Na in soils. 4 The per-
to 85 per cent or more) of exchan ge saturat ion with
tion
centage of exchange saturation with Na is called the degree of alkaliza
to water (under 1 cm/hr) fre-
( ~ 4-44). Poor permeability of the soil
15 per cent of Na saturat ion
quently develops in association with over
persists as a detrime nt
when the salinity is low. The poor permeability often
and even
to productivity after the alkalinity has dropped to near neutral
alkaline
after the percentage Na saturation has been decreased. Saline and
meaning
soils taken collectively are often termed alkali soils, although this
5

accepta ble in a chemic al sense. A dis-

of the term "alkali " is not entirely

2 A problem in 20 per cent of the greenho

use soils examine d by Merkle and
Dunkle, J. Am. Soc. Af?ron., 36: 10 ( 1940).
3 Scofield, Soils and Men, U.S.D.A . Yearboo k,
p. 712 ( 1938).
4 Fireman and Reeve, S.S.S.A. Proc., 13: 494 ( 1949).
5 Kelley, Alkali Soils (New York: Reinhol
d Publishing Corpora tion, 1951);
Richard s, op. cit., p. 5.
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 229

persive effect on the soil, somewhat like that of Na saturation, is thought6

to result from Mg saturation in excess of 30 per cent.

DIRECT SEMIQUAN TITATIVE TEST FOR SOIL SALINITY -

CONDUCT OMETRICA LLY ON SOIL PASTE
10-6. A semiquantitative estimate of the salt content of a soil, by the
electrical resistance of the soil paste, may be made by means of the salt
bridge. The soil paste method gives a fair semiquantitative approximation
except for soils low in salts but high in exchangeable Na. 7 Appreciable vari-
ations occur 8 in resistance found with different soils having the same per-
centage of salts. Soils at the saturation moisture percentage gave conduct-
ances 0.5 to 4.8 times the conductance of the corresponding extracted
solutions. 9 Nonetheless, the approximate measurement of the salt content
by soil paste resistance is simple and rapid, and has been found to be ex-
tremely useful in field work, such as soil surveys, 10 where wide variance
in soil salt content occurs in short distances. The soil paste-salt bridge tech-
nique has been indispensable for the survey and reclamation of soils of arid
regions.

APPARATUS

10-7. Needed apparatus includes a thermometer and salt bridge. A

standard soil paste "salt bridge" apparatus, consisting of an A.C. poten-
tiometer with standard Bureau of Soils hard rubber soil resistance cup is
suitable. Also, a standard conductance meter (~ 10-21) and separate soil
resistance cup (Fig. 10-1) are suitable. The ohm resistance range needed
is from about 10 to 10,000. The principles of resistance or conductance
measurements are presented in ~ 10-22.

REAGENTS

10-8. Needed are distilled water, 1 per cent phenolphthalein indicator

solution, and carefully collected and representative soils samples, the
soluble salt content of which are to be determined.

PROCEDURE

10-9. The soil sample is sieved through 6-mm screen (0.25 inch) and
thoroughly mixed on a rubberized sheet. Pebbles and large root fragments
are discarded. The xnbber resistance cup is filled about half full of distilled

6,Qill and Sherman,.Pac . Sci., 6:137 (1952); Smith et al., Soil Sci., 68:451 (1949).
7 Richards, U.S.D.A. Agr. Handb. 60, p. 16 ( 1954).
8 Magistad et al., Soil Sci., 59: 70 ( 1945).
9 Reitemeier and Wilcox, Soil Sci., 61 :281 (1946).
10 Davis and Bryan, U.S.D.A. Bur. Soils Bui. 61 ( 1910); Kellogg, U.S.D.A. Misc
.
.Pub. 274,p.123 (1937);U.S.D .A. Aj;r. Handb. 18,p.343 (1951).
230 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS

Fig. 10-1. Standard Bureau of Soils soil resistance cup. (Photo courtesy Industrial
Instruments, Inc., Jersey City 5, N.J .)

water (or somewhat less if the soil is rather moist, and somewhat more if
the soil is dry and of very fine texture) . The soil is added with stirring
until the paste is wet enough to glisten on the surface, but thick enough
so that no free water stands on the surface. The cup is tapped to remove
air bubbles. The top of the soil is struck off to leave the cup just level full,
and the cup (clean externally) is placed between the spring clips of the salt
bridge. The soil-water mixture may be made in a beaker and transferred to
the cup. The mixture is equilibrated for a period of 20 minutes.
10-10. The knife switch of the "salt bridge" is set in the "out" posi-
tion (the "in" position places an extra resistance of 100 ohms in series
with the soil cup). Then the coil resistance knob is turned to 10, the phone
placed to the ear and the button depressed. An electronic eye replaces the
buzzer in some bridges. If a buzzing is heard, the dial resistance knob is
turned back and forth until the point of minimum sound is located. If such
a position cannot be found, the coil resistance knob is turned to 100 and
the dial resistance knob again turned back and forth to the position where
the sound is the faintest. An adjustment of the soil resistance to the 1000
factor ma be necessary to get a minimum with the knob. When the mini-
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 231

mum point sound is found, the dial reading is observed and multiplied by
the setting of the coil resistance factor 10, 100, or 1000 as the case may be.
The product is the ohms resistance in the cup. For example, if the dial read-
ing is 1.45 and the coil resistance factor is 100, the cup resistance is 145
ohms.
10-11. Soils containing large amounts of soluble salts may have a re-
sistance below the range of the dial. To obtain the resistance of these soils,
the knife switch is thrown to the "in" position to throw 100 ohms in series
with the soil cup, and the dial is adjusted to minimum sound. From the
product obtained as in the preceding paragraph, 100 is subtracted. The
100 ohms resistance controlled by the knife switch is also used to test the
bridge, and when in the "in" position should balance the 100 ohm bridge
setting when the cup contacts are short circuited with a buss bar.
10-12. Temperature Correction. The temperature of the soil suspension
is measured and corrected to 15.8 °C ( 60°F), which is the conventional
temperature for soil paste resistance comparisons. Each degree centigrade
away from 15.8°C represents a 2.49 per cent resistance change 11 (1.38
per cent per degree F away from 60°F). The resistance (R) corrections
may be made by fairly tedious calculations from tables, 12 or more con-
veniently (and with fewer arithmetical steps), by the formula,
(10-1)

in which RT refers to observed resistance at T 0 C. The approximation is to

within ± 2 per cent of table values from 0° to 33°C (32° to 92°F) and
within± 5 per cent up to 40°C.
10-13. The specific conductance of the soil paste is 0.25 / R wherein
0.25 is the constant for the Bureau of Soils cup, and R is the resistance of
the soil paste in ohms. The ratio of electrical conductance of the saturation
extract (mmho/cm) to the electrical conductance of the soil paste in the
standard cup was found 13 to be 2. 71 for some muck soils and a silty clay
loam.
10-14. Interpretation of Resistance of Soil Paste in Terms of Salt Con-
tent. The soil paste resistance values corrected to 15.8°C (60°F) are inter-
preted in terms of salt content as percentage by weight of the soil, by means
of the chloride and sulfate column of Table 10-1. These data were based
on average saline soils encountered in the field, for which the soluble salt
contents were determined by extraction and weighing ( ~ 10-17). The data

11 Derived graphically from table values of Whitney and Means, U.S.D.A. Div. of
Soils, Bui. 8 (1897).
12 Whitney and Means, U.S.D.A. Div. of Soils, Bui. 8 (1897); Davis and Bryan,
U.S.D.A. Bur. Soils Bui. 61 (1910); or Kellogg, U.S.D.A. Misc. Pub. 274, p. 123
(1937); U.S.D.A. Agr. Handb. 18, p. 343 (1951).
13 Campbell and Richards, Agron. lour., 42:582 (1950).
232 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS
of
of Davis and Bryan (Tables 10-1 and 10-2) agree well with those
they are some-
Whitson and King. 14 Salt contents as parts per 100,000 (as
a
times expressed) may be obtained by multiplying by 1000. A soil having
for the
salt content of less than 0.1 per cent is considered nonsaline except
cent
more sandy soils (~ 10-35) . A higher percentage ranging up to 0.3 per
nonsali ne clay soils ( ~ 10-42) . The salt conten t of
is the upper limit for
re-
water or soil solutions may be estimated by the determination of the
sistance when the soil cup is filled with it (Table 10-2).

TABLE 10-1
electrical
Approximate amounts of salts in soils of various textures with given
resistance of soil paste in standard soil cup*

Percenta ge salt content of soils containi ng salts consisting

predomi nantly of chloride s in part of sodium carbona te (turn
and sulfates phenolp hthalein pink)

Resistance of I
soil paste at
15.8°C (60°F), Clay I Clay
in ohms Sand I Loam loam Clay Sand I Loam loam Clay
----
3.0 . .. .. . ... ... . .. . ..
18 3.0
19 2.4 2.6 3.0 . .. ... ... ... . ..
20 2.2 2.4 2.8 3.0 . .. ... . .. ...
1.7 1.9 2.2 2.9 3.0 3.0 ...
25 1.5
1.3 1.4 1.6 2.1 2.2 2.2 3.0
30 1.2
I.I 1.2 1.3 1.6 1.9 1.9 2.6
35 1.0
0.94 1.0 1.1 1.4 1.7 1.7 2.3
40 0.86
0.71 0.77 0.86 1.3 1.4 1.4 1.9
50 0.67
0.58 0.63 0.70 0.87 1.2 1.2 1.6
60 0.55
0.50 0.53 0.59 0.74 0.98 1.0 1.4
70 0.48
0.44 0.47 0.51 0.64 0.86 0.90 1.2
80 0.42
0.39 0.41 0.45 0.56 0.77 0.82 1.1
90 0.37
0.35 0.37 0.39 0.51 0.69 0.75 0.97
100 0.33
0.28 0.29 0.32 0.43 0.57 0.64 0.79
120 0.27
0.23 0.24 0.26 0.38 0.49 0.55 0.66
140 0.22
0.20 0.21 0.22 0.34 0.43 0.49 0.56
160 0.20
... . .. 0.31 0.38 0.44 0.49
180 ... ...
0.43
200 ... ... ... ... 0.29 0.34 0.40

... . .. . .. 0.21 0.28 0.34 0.33

240 ...
300 . .. ... ... ... ... 0.22 0.28 0.29
380 ... ... ... . .. ... . .. 0.20 0.20

61 (1910).
•Adapted from data of Davis and Bryan, U.S.D.A. Bur. Soils Bui.

14 Wis. Agr. Exp. Sta. Bui. 85 (1901).

 TABLE 10-2
Approximate salt content of water with given resistances
in standard soil cup•
--------·---·---·
Salt content of water containing salts consisting
-··--·---····-----
in part of sodium car-
Resistance of water at predominantly of chlor- bonate (turn phenol-
15.8°C (60°F), in ohms ides and sulfates phthalein pink)
------· .... ---
p.p.t. t p.p.t. t
--·----- -------
30 7.5 7.5
35 6.7 6.7
40 6.0 6.0
45 5.3 5.3
50 4.6 4.6
55 4.0 4.3
60 3.6 4.0
65 3.1 3.8
70 2.7 3.6
75 2.3 3.4
80 2.1 3.2
85 2.0 3.0
90 1.95 2.9
95 1.88 2.8
100 1.81 2.6
110 1.70 2.5
120 1.60 2.3
130 1.50 2.1
140 1.41 2.0
150 1.32 1.87
160 1.24 1.76
170 1.16 1.65
180 1.09 1.54
190 1.02 1.44
200 0.96 1.38
220 0.87 1.22
240 0.79 1.10
260 0.71 1.00
280 0.65 0.90
300 0.59 0.83
340 0.50 0.71
380 0.44 0.60
400 0.41 0.55
450 0.35 0.46
500 0.31 0.38
550 0.28 0.32
600 0.25 0.27
700 0.22 0.23
800 0.20 0.20
1000 0.18 0.18
1500 0.16 0.16
2500 0.15 0.15
------
•Adapted from data of Davis and Bryan, U.S.D.A. Bur. Soils Bui. 61 ( 1910 ).
f Parts per thousand parts water.
233
234 SOLUBL E SALT ANALY SIS FOR SOILS AND WATER S
10-15. Davis and Bryan checked for the presence of Na 2C0 3 by addi-
tion of drops of 1 per cent phenolphthalein solution to the soil or solution.
If a pink color was obtained, indicating pH 8.3 or above, the salts were in
part sodium carbonate, and the corresponding columns in Tables 10-1 or
I 0-2 were used. In practice, the estimate of salt content based on the sul-
fate and chloride columns (Table 10-1) is better 15 than from the carbonat e
columns for all soils, even of alkali soils above pH 8.3.
10-16. Care of the Salt Bridge. The salt bridge is carefully protected
from mechanical shock and kept clean. Particularly, the contact metals are
kept clean, bright, and free from grease. After use, the soil cup is washed,
rinsed, and dried.

ALTERNATIVE PROCEDURES

10-17. Standardization. It is possible to standardize soil paste resistance

in terms of the salts occurring under the particular field conditions. The salt
content of I 0 or 12 representative soils is determined by the conductance
of the saturation extract ( ~ I 0-27) or gravimetrically ( ~ I 0-7 5). The ac-
curacy of the soil resistance method seldom justifies this refinement. For
greater accuracy, the salt content is determined on the saturation extract of
all samples(~ 10-27).

DETERMINATION OF ELECTRICAL CONDUCTANCE OF SOIL

SOLUTIONS AND WATERS AS A MEASURE
OF SALT CONTENT

10-18. A fairly quantitative estimate of the salt content of solutions

extracted from soils or of natural waters can be made from their electrical
conductance. Extracts of soils, particularly those made with high water to
soil ratios, are a less accurate measure of the solute content of the soil it-
self because more salts may be removed than are ever present in the soil
at field moisture contents. Also, the ionic species extracted may be different
from those present in the soil solution. For example, the amount of calcium
and sulfate from a gypsum-bearing soil is about 5 times as much in a 1 : 5
extract as in a 1 : 1 extract. Ii CaC0:1 is present, HC0:1 - and Na+ in-
crease with dilution, the latter being displaced by Ca t + dissolved from
CaCOw rn The content of Na+ may be twice as great in the 1 : 5 extract a3
the 1 : 1 extract; likewise, the dissolved Ca+ +, SO 4 - - , and Na+ are
greater in the 1 : 1 extract than in the soil saturation extract. To some ex-
tent, hydrolysis of the exchangeable sodium increases with the extent of

15Fireman and Reeve, S.S.S.A. Proc., 13:495, Fig. 2 (1949); Richards, op. cit.,
p. 16.
16 Magistad et al., Soil Sci., 59:65 ( 1945).
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 235
dilution. Chloride and nitrate concentrations decrease more than propor-
tionately by dilution of the suspension, 17 an effect attributed to "negative
adsorption" (Donnan distribution) or to bound water of the colloids not
acting as a solvent. Increased quantities of sulfate extracted with more
dilute suspensions in one soil was attributed to anion replacement by
hydroxyl.
10-19. Because the soil : water ratio influences the amount and composi-
tion of salts extracted, the soil : water ratio employed must be specified
with the analyses. Extraction of soil at natural field moisture contents (~
I 0-45) gives the most accurate measure of soluble soil salts. The soil
moisture saturation percentage (~ 10-27) is related to soil moisture con-
stants, and field moisture content, provides a simply interpreted soil salinity
scale based on electrical conductance of the extract at this moisture content
(~ 10-35), and is the highest soil moisture content at which the soil itself
may be employed to regulate the soil : water ratio for extraction.
10-20. Definitions. Electrical resistance is defined by the equation.

E=IR (10-2)

in which E is the electrical potential in volts, I is the current in amperes and

R is the resistance in ohms. Electrical conductance, C, or conductivity of a
solution in mhos is the reciprocal of resistance, R, in ohms,

c-!..
-R (10-3)

For convenience, conductance is often expressed as millimhos ( 1000 times

mhos). Specific conductance, L, of a solution is the conductance that
would be measured at 25°C between electrodes 1 cm 2 in cross section and
placed one cm apart, and may be visualized as the conductance across a
cm:i, or mhos/cm. Specific conductance may be measured with a cell of
various dimensions by means of a cell constant (~ 10-32). Because the
numbers obtained expressing specific conductance of soil solutions are gen-
erally small, it has been found convenient to express specific conductance as
millimhos/cm 1 H ( 1000 times mhos/cm) and this unit has been adopted
widely. Other units such asrn 10r. x mhos/cm, and 20 micromhos/cm (10 8 x
mhos/cm) have been employed (Table 10-3).

17 Reitemeier, Soil Sci .. 61: 195 (1946).

18 Fireman and Reeve, S.S.S.A. Proc., 13:495 (1949).
1ll Magistad et al .. Soil Sci., 59:65 (1945).
20 Employed by Wilcox, U.S.D.A. Cir. 784 (1948), the Rubidoux Laboratory, and
also several agencies, such as U.S. Geo!. Surv., Quality of Water Branch; U.S. Fish
and Wild Life, and Reclamation Research Laboratory, Denver.
236 SOLUB LE SALT ANALYSIS FOR SOILS AND WATE RS

TABLE 10-3

Various units which have been employed for electrical conduct ance
of soil solutions and other waters
Exampl e, L of moisture saturatio n
Relative size Factor for extract of a moderat ely
of unit calculat ion saline soil

I X mhos/cm 0.006 mho/cm

10:1 10~ X mhos 6 millimh os/cm (adopte d)
10-r. 105 X mhos 600 mhos X 10°/cm*
10-6 J06 x mhos 6000 microm hos/cm
-· - - · - - · · - - - - - - - - - - - - - - - - -l - ---------
value reported was mhos X
•Some writers have stated "mhos X 10-5/cm" when the numerica
10'•/cm.

APPARATUS
al
10-21. Needed apparatus consists of an AC "salt bridge" or electric
10-4)
resistance bridge'2 1 (Fig. I 0-2 and I 0-3), conductance cell (Fig.
funnel
with platinum-blackened electrodes, a Buchner or spociaP vacuum
2

tor, a suction pump (a centrifu ge and tubes

with a vacuum flask or desicca
of the soil extract , large test tubes, bottles or
may be used) for separation
beakers for collection of the extract , and a thermo meter.
ce or
10-22. Principles of the Salt Bridge. In a salt bridge (a resistan
R and a variabl e resistan ce
conductance bridge) 2 fixed resistances R 1 and '2

tance cell having re-

Rv are connected in a branched circuit with the conduc
An AC
sistance R.x (Fig. 10-5). An AC potential is applied at C and D.
polariz a-
potential is employed to prevent electrolysis of the solution and
a 1000-
tion of the electrodes in the conductance cell at R,.. Ordinarily
bridges operate on a 60-cycl e power line
cycle source is used, but some
capacit ance effects become import ant and
source. As the frequency rises,
R • The
are compensated for with the variable condenser in parallel with 2
is no current passing in the
variable resistance R,, is adjusted until there
m sound or shift in
phone circuit from A to B, as indicated by a minimu
and the voltage
the electric eye. Then A and B are at the same potential,
be-
drop JRJ, (eq. 10-2), between BD must equal l'R 2 , the voltage drop
tween AD:
(10-4)

Also:
{10-5)

1940), available
21 Bouyoucos and Mick, Mich. Agr. Expt. Sta. Tech. Bui. 172 (
from Wood and Metal Product s Co., Bloomfi eld Hills, Mich.
'2'2 Richard s, Agron. lour., 41 :446 ( 1949), availabl
e from Instrum ent Develop ment
and Manufa cturing Corp., Box 191, East Pasadena, Calif.
 Fig. 10-2. Conductance bridge suitable for the measurement of soluble salt con-
tent of soil solutions (as well as gypsum block resistances). (Available from In-
dustrial Instruments, Inc., Jersey City 5, N.J.)

237
 measurement of electrical
Fig. 10-3. Conductance bridge and cell suitable for y saline soil extracts).
cm (nonsa line to strongl
conductance from 0.1 to 10 millimhos/ City 5, N.J.)
(Available from the Industrial Instruments, Inc., Jersey

(b)
(al
ns: (a), large size, requiring
Fig. 10-4. Conductance cells suitable for soils solutio
requiri ng only a few ml of solution. (Both
about 35 ml of solution; (b), small size, Industrial Instruments, Inc.,
availab le with a cell consta nt range of 0.1 to 2.0, from
Jersey City 5, N.J.)

238
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 239

Potent1ometric
rheostat

Fig. 10-5. Schematic diagram of salt bridge.

The dial of R,. is calibrated in terms of resistance
or conductance of R,. the conductance cell, as ex-
plained in text.

Then division of equation I 0-4 by I 0-5 gives:

/Ra: l'R 2
{10-6)
IR,, l'R 1
Whence:
R = R2 R (10-7)
"' R1 v
Since R 1 and R 2 are fixed, the dial on R, can be calibrated to read R,.. the
resistance of the test sample. Or, the dial on R,, can be calibrated to read
1IR,,., that is, directly in conductance of the solution.
REAGENTS

10-23. Needed reagents consist of 0.02 M KCl (1.4912 gm of KCl per

liter of solution), C0 2 -free distilled water, and filter paper.
10-24. For purposes of the standardization work, salts such as CaS04 •
2 H 20, CaC03 , NaCl, Na 2 C0 3 , and 3-12-12 fertilizer (commercial or
 ERS
240 SOLU BLE SALT ANALYSIS FOR SOILS AND WAT
an equal amount of
compounded of NH 4 NOa, KC!, and Ca(H 2 P0 4 ) 2 with
and saline soils.
gypsum) may be added to a slightly acid nonsaline soil

PROCEDURE
23 soil samples
10-25 . The Soil Sample. For complete salinity analysis,
feet) . Some times the surface
are taken to a depth of 120 to 180 cm ( 4 to 6
imes this situati on is re-
has greater salinity than the subsoil, and somet
each succee ding 30-cm
versed. The surface 5 cm, the next 5 to 30 cm, and
produ ces good crops,
layer are sampled separately. If 1 portion of the field
to the troubl esome
this is sampled separately and analyzed for comparison
sampl e of appro ximat ely 1-liter volum e is placed in a
area (~ 2-54) . Each
the outside of the
strong paper bag with label inside on a tag, as well as on
to the laboratory.
bag. Wet samples are partially air-dried prior to shipment
on locati on, soil descri ption, area rep-
The label is supplemented with data
ity, and qualit y of water used
resented; depth to water table; source, quant
for irrigation; and crop variety grown and condi tion.
air-dry, but not
10-26 . The soil sample may be either field moist or
or fertilizer (~
oven-dry. Salts such as CaS0 4 • 2 H 2 0, CaCO~, NaCl,
rdization work.
10-24 ) may be added to various soil samples for standa
soils can more
The soil is mixed by passage through a 2-mm sieve. Clay
Stones and coarse
easily be sieved before they are completely air-dry.
roots are discarded.
tion moisture
10-27 . Moisture Saturation Extract of Soil. The soil satura
nt of water held in the puddled
content is defined 24 as the maximum amou
made in the soil mass.
soil without free water collection in a depression
ds on the soil texture
The quantity of soil sample to be extracted depen
be emplo yed. The
(Tabl e 10-4) and the volume of the conductance cell to
volum e is placed
approximate weight of soil to provide the needed filtrate
or two-th irds of the water is added down the side
in a beaker. The first half
ries. The soil is not
of the beake r so that it passes through the large capilla
through puddled
disturbed during this process because water movement
are added until the soil mass is fully
soil is very slow. Increments of water
add increm ents of water to several
wetted by capillarity. It is convenient to
allowe d for full imbib ition
successive samples to be analyzed, time being
sampl e. The soil is then
of one increment before more is added to each
to give the final ad-
stirred with a spatula, and more water or soil is added
the soil barely
justment of water content. The water content is right when
re slides off the
flows together into a hole made with spatula, the mixtu
water does not
spatula, and the soil surface is wet enough to glisten. Free
few minutes. If free
collect in the depressions on the surface on standing a

.
2s Magistad and Christiansen, U.S.D.A. Cir. 707 (1944)
24 Scofield, U.S.D.A. Cir. 232 (1932) .
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 241

TABLE 10-4
Quantity of soil to be taken for each ml of saturation extract in relation
to soil texture and moisture properties
Sample weight in
Field Moisture gm, oven-dry
Soil Wilting capacity saturation equivalent, per
texture percentage percentage percentage ml filtrate
--·---------
Sand 1 2 4 50.0
Sandy loam 4 8 16 12.5
Silt loam IO 20 40 5.0
Clay 25 50 JOO 2.0
Peat 35 70 140 1.4
---·--..-·-----------·------
Relative* 2 4
-----~------·- -----------·--------·---------·---------·---·----
0 Relativ<· values for 3 soil moisture conditions pointed out by Dr. L. A. Richards ( ](•cture at the

University of Wisconsin, I f.J49 ).

water stands on the surface, too much water has been added, and a little
more soil is added to blot up the excess. With a little practice, the char-
acteristic moisture saturation percentage can be reproduced.
10-28. For many purposes the percentage moisture at saturation need
not be determined. It may be determined by oven-drying a sample, or it
may be estimated ( ~ l 0-44).
10-29. Equilibration Time. For conductance measurements, the satu-
rated soil is equilibrated for I 0 minutes if gypsum is absent or for 2 hours
if gypsum is present. If the extract is to be analyzed chemically for ionic
composition (~ 10-78), the moisture-saturated soil is equilibrated for 6
hours to permit ionic-exchange equilibrium to be attained ( ~ 10-18), but
no longer because of changes in composition that result from bacterial
activity.
10-30. Filtration. The soil is placed on a suitable size of Buchner fun-
nel with tightly seated filter paper, and the "saturation extract" filtrate is
removed by suction. The soil saturation extract also may be obtained by the
pressure membrane (~ I 0-47) or by centrifugation. 25 A porous tube de-
vice has been employed for sampling soil solutions during water-spreading
operations. 26
10-31. Coloration of the extract by dissolved organic matter does not
appreciably affect the conductance or chemical analysis, so may generally
be ignored. Turbidity, on the other hand, may lead to an appreciable error
in the chemical analysis. Turbid solutions may clear on standing, or may
be cleared up by passage through a Pasteur-Chamberland filter (~ 10-53).
10-32. Determination of Cell Constant. The cell constant, k, of a con-

25 Chesnin and Johnson, Soil Sci., 69:497 ( 1950).

26 Krone et al., Soil Sci., 73: 211 (1952).
 242 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS
ductance cell is determined by measurement of the electrical conductance,
C, of a standard KCI solution, and use of the equation:

L
k=- (10-8)
c
in which
L = known specific electrical conductance of standard solution, usually
0.02 M KC!, mmhos/cm
C = conductance of the standard solution measured in the given cell,
mmhos
The specific conductance, L, of the 0.0200 M KC! is 2.39 mmhos/cm at
I 8°C, and 2.768 mmhos/cm at 25°C. Various standard conductance solu-
tions are available.2 7 The measured conductance, C, of a test solution (~
I 0-33), in millimhos, multiplied by the cell constant gives the specific con-
ductance, Lrnmho/ «m' of the test solution:
L = kC (10-9)

10-33. Determination of Solution Conductance. The conductance cell is

ordinarily stored immersed in distilled water. The cell is usually rinsed
twice with the test solution, but if insufficient solution is at hand, the cell
may be rinsed with and dried from acetone to prevent dilution of the ex-
tract with water. The cell is filled with the test solution (~I I 0-30, I 0-79)
to immerse the electrodes completely. The bridge is balanced and the read-
ing is recorded as resistance or conductance (specify units, ~ 10-20). The
specific conductance as L 11 ..111101 em is calculated ( eq. I 0-9). The "Solu-
bridge" with its conventional cell reads directly in specific conductance, 28
no cell constant thus being needed.
10-34. The temperature of the solution is taken into account in the
calculation of the result. Electrical conductance of a solution increases ap-
proximately 2 per cent per degree C. Temperature corrections may be
avoided by use of a temperature bath at 25°C, or approximately by use of
a reference solution at the same temperature as the test solution.
10-35. The Soil Salinity Scale. Soluble salts decrease the availability of
the soil water by contributing osmotic pressure 211 to the integrated soil
moisture stress, and the latter decreases yields if it exceeds low values.an
As the soil moisture content is changed from the wilting percentage to field

27 Handbook of Chemistry and Physics (Cleveland, Ohio: Chemical Rubber Pub.

Co., 1930-1957); Jones and Bradshaw, J. Am. Chem. Soc. 55: 1780 (1933).
28 In some dials (Fig. 10-3) "mhos X 10-0 " means "mhos X 10°/cm" (~ 10-20).
2u Magistad, Bot. Rev., 11: 181 (1945).
aowadleigh, Soil Sci., 61:225 (1946).
SOLUBL E SALT ANALYSIS FOR SOILS AND WATER S 243

capacity or saturation percentage, the salts present in the soil solution are
diluted. Conversely, the salts become more concentr ated in the soil solu-
tion as the soil moisture is used up and the wilting percentage is approached.
Greater salt damage to crops is often observed in hot summers than in
cooler summers, presumably because the wilting percentage is approach ed
more frequently. Since the wilting percentage is smaller for sandy soils
than for finer textured soils, a given absolute amount of salts per unit weight
of a sandy soil creates a greater concentration of salts in its soil solution at
the wilting percentage than that same amount of salts would create in a
finer textured soil ( ~ 10-42). For all soil textures, the concentration of
salts in the soil moisture saturation extract is approximately one-fourth that
in the soil solution at the wilting percentage and one-half that present in
the soil solution at the field capacity, owing to the fundamental relation of
the saturation moisture percentage to soil moisture constants (Table 10-4,
above). The saturation extract is thus an "equipot ential" soil moisture
content for various soils. The specific electrical conductance of the satura-
tion extract, which is linearly related to osmotic pressure as well as concen-
tration of salts in solution (~I 10-3 7), can be interpreted directly in terms
of plant growth, by means of the salinity scale~ 1 (Table 10-5). Although
different plants vary in their tolerance to the presence of soluble salts,
32

the salinity scale is found applicable to plants classified into relatively few
groups. Workers of long experience with saline soils tend to prefer electrical
conductance units to units of concentration of salts in solution. The rela-
tive conductance units of the salinity scale can be interpreted as readily as
the relative numbers of the soil pH scale.
10-36. Irrigation waters should range from 0.1 to 0.75 mmhos per cm
or below. High salinity hazard is incurred in the use of irrigation water
having conductance much above this range ( ~ 10-80).
10-37. Calculation of Specific Electrical Conductance to Salt Concen-
tration in Solution. A linear relationship existsa:i between the specific elec-
trical conductance in a water extract of soils or irrigation water and the
concentration of salts as found by analysis (~ 10-78) and expressed as
meq of anions (or cations) per liter of solution,
meq of salt per liter = equiv. per million = 12.5 Lmru1ioi cm ( 10-10)

The factor for the single salt solutions of NaCl, CaCl~, MgCl~, Na~S0 4 ,
CaS0 4 , MgS0 4 , and NaHC0:1, according to data published in the Inter-
national Critical Tables,:H ranges from 8 to 20.

:11 Scofield, U.S.D.A. Cir. 232 (1932).

32Ali and Powers, Plant Physiol., 13:767 (1938); Ayers et al., J. Am. Soc. Agron.,
35 :796 (1943); for boron, Wilcox, U.S.D.A. Cir. 784 ( 1948).
aa Fireman and Reeve, S.S.S.A. Proc., 13: 494 ( 1949).
34 Richards, U.S.D.A. Agr. Handb. 60, p. 10--12 (1954).
 ,.

TABLE 10-5
The salinity scale*
- - - - - - - - - - - - - - Specific conductance of the saturation extract of soil, millimhos per cm. -------------~
0 2 4 8 16
Very slightly Moderately Strongly Very strongly
Nonsaline saline saline saline saline
Salinity effects Yields of very Yield of many Only tolerant Only a few very
mostly sensitive crops crops crops yield tolerant crops
negligible. may be restricted. satisfactorily.
.,..
N
.,.. restricted. Alfalfa, cotton, Bare spots
yield satis-
factorily. Only
sugar beets, appear because salt tolerant
cereals, and of injury to grasses, herba-
grain sorghums germination. ceous plants,
adapted. shrubs, and
trees grow.
0 0.1 0.3 0.5 1.0
~------------------- Percentage of salts in moisture saturation extract - - - - - - - - - - - - - - - - - - - - '
•Adapted from descriptions of Scofield, Reports of Participating Agencies, Part III, Sec. 6, U.S. National Resources Planning Board, June ( 1942 ), pp. 263-
334; Richards, ed., Diagnosis and Improvement of Saline and Alkali Soils, U.S.D.A. Agr. Handb. 60, p. 9 ( 1954 ); and Campbell and Richards, Agron. four., 42:
582 ( 1950 ).

,.
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 245

10-38. The effects of salts on plants is more closely related 35 to the

equivalents of salts per million parts of solution (~ 10-37) than gravi-
metric weight units per million. However, the latter can be estimated 36 for
waters in the alkaline soil regions:

ppm of salts= 640 Lmmbo/ cm (10-11)

% salts in solution = 0.064 Lmmbol em (10-12)

% salts in soil = 0 064 L x % water in soil at ext':ilcti_~I!

• mm ho/ <'m 100
(10-13)

Equivalents per million multiplied by the gm-equivalent weight gives the

parts per million. Equation 10-11 involves the assumption of an average
gm-equivalent weight of 51 for the various salts present. The conversion
factor to percentage salts for highly fertilized soils of the humid region is
approximately 0.1 ( ~ 10-51 ) , or somewhat higher than 0.064, and sig-
nifies that the average equivalent weight is greater than 51.
10-39. Osmotic Pressure of Solutions. Specific conductance can be con-
verted to osmotic pressure:

Osmotic pressure of solution, atm. = 0.36 Lmmho/ .. m ( 10-14)

The factor 0.36 applies well for NaCl and for solutions extracted from
alkali and saline soils 37 and should be applicable to irrigation waters in arid
regions. The factor is 0.3 for common MX 2 and M 2 X salts and 0.28 for
MgS0 4 , and thus a factor of about 0.3 would be expected for highly fer-
tilized soils of the humid region.
10-40. The salt concentrations of the soil extract or water may be
checked by gravimetric determination through evaporation ( ~ 10-7 5).
The anion and cation species in the solution may be determined(~ 10-78).
10-41. The salt index of a fertilizer may be estimated by electrical
conductance:
Specific conductance of
solution when 1 gm offertilizer is (10-15)
. suspended in 1 liter of water 100
Salt mdex = ---·------ ---------- ·--·------ X
. Specific conductance of 0.1 % NaN0 3
solution

35 Magistad, Bot. Rev., 11: 181 (1945).

sa Richards, U.S.D.A. Agr. Handb. 60, p. 16 (1954); and Magistad et al., Soil Sci.,
59: 70 (1945), used the factor 700.
87 These factors are derived from freezing point depression data reported in the
International Critical Tables, 4:254-260, based on O.P. =
12.06 a T - 0.021 (i:i T) 2 ,
as reported by Richards, op. cit., p. IO, 17; Bouyoucos and McCool, J. Agr. Res.,
15:331 (1918); Campbell, Soil Sci., 73:221 (1952).
246 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS
10-42. Toxic Limit Percentage of Salt in Soil. The salt content of soil
expressed as percentage or ppm of the dry soil is not simply related to the
salt toxicity to plants grown there ( ~ 10-35). For example, in 2 soils, 1
a loamy sand and the other a clay, both with same salt content on the dry
soil weight basis, the salt concentration in the soil solution at the wilting
percentage will be approximately 10 times as high in the sandy soil, because
its wilting percentage is about 0.1 as great as that of the clay soil. Thus no
1 definite percentage of salts in soils can be given at which toxicity begins
in all soils. If the moisture saturation percentage of the clay soil happened
to be 100 per cent, the concentration of salts in the soil solution would be
equal to that in the dry soil. Supposing L = 10 mmho/crn in the saturation
extract, from equation 10-12:
% salts in soil = L x 0.064 = 10 x 0.064 = 0.64
(:::: that in solution)

This value of 0.64 per cent is plotted as point A (Fig. 10-6) at a saturation
extract conductance of l 0 mmho/crn. A line drawn through point A from
the origin is the locus of salt percentages for all values of specific conduct-
ance for soils having 100 per cent moisture saturation percentage. For a
second soil, having a saturation moisture percentage of 80 per cent, the
concentration of salts in the dry soil would be 80/100 of that in the satura-
tion extract from it, and for L = 10 mmho/cm in the saturation extract,
from equation 10-13:
. . L x 0.064 x % water at sat.
% salts m sod= -----·--··-~--·-----------··-··--·--
100
= lg_x_0.0§4 x 80 = 0.51
100
which is plotted at point B (Fig. 10-6). The various lines show the rela-
tionship of salt content (of soils having the stated saturation moisture per-
centages) to specific conductance of the saturation extract. Taking Scofield's
value of 4 mmhos/cm as the upper limit of salinity that is harmless to
plants (Table 10-5), the toxic limit percentage of soil salts is 0.35 per
cent for a peat soil (Fig. 10-6), about 0.1 per cent for a silt loam, and
about 0.05 per cent for a coarse loamy sand. Salt contents in soils in ex-
cess of these percentages would be harmful to some plants. Double these
values may be taken as the toxic limit for more resistant plants (limit of
moderate salinity, Table 10-5).
10-43. A fertile soil may contain 0.02 to 0.05 per cent (200 to 500 ppm)
of soluble salts. A content of 0.1 per cent salts in a silt loam corresponds to
1000 ppm in the soil or 4000 ppm in the soil solution at the field moisture
capacity of 25 per cent. For a silt loam soil, 1500 ppm in the soil is often
found to be the maximum salt content for growth of soft-stemmed plants,
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 247

0 4
Specific conductance of the saturation
extract from soil, millimhos per cm

Fi~. 10-6. Relationship of percentage of soluble salts to spe-

cific conductance of the saturation extract of soils of varying
texture. The moderately saline limit at 4 mmho/cm falls at widely
varying percentages of salts, depending on the soil texture. (After
Richards, op. cit., p. 17 .)

and 2500 ppm of the soil is the maximum for growth of woody plants such
as rose bushes.
ALTF.RNATIVE PROCEDURES

10-44. Determination of Saturation Moisture Percentage. It is often de-

sirable to determine the saturation moisture percentage. To do this, a small
portion of the soil paste is removed, and the moisture content is determine,d
by oven-drying. Alternatively, Wilcox 38 has published a formula for the
calculation of the saturation moisture percentage from the weight of a
known volume of saturated soil paste, as follows:
. . (2.65V - W) (10-16)
Saturat10n moisture percentage = 37 .74 --w--V---

in which V is the volume in ml of the saturated paste, and W is its weight

in gm. The density of water is taken as unity and that of the soil particles as
2.65. The formula is suitable for mineral soils but not for organic soils; the
estimate was found to be 4 to 6 per cent low for those mineral soils that
swell considerably.
10-45. Displacement of the Soil Solution. Water or other displacing
liquid appears to replace the sorbed films of the soil solution, pushing it
ahead through the column. The displacement method has been employed to
study the effect of fertilizers on the concentrations of the soil solution 39 and
as Soil Sci., 72:233 (1951 ).
39 White and Ross, S.S.S.A. Proc., I: 181 (1937).
248 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS
other soil conditions, 40 and to study the solution phase of mixed fertilizers. 41
Various displacing liquids 42 used include water, alcohol, acetone, and oil.
The soil with the desired moisture content is passed rapidly through a 2-
or 3-mm screen. Quartz gravel or glass beads are placed over the opening
in the glass leaching tube or metal cylinder 43 (inverted 1-liter glass bottle
with gently sloping neck and with bottom cut off, ~ 1-23). A 5-mm layer
of glass wool is next laid over the gravel or beads. The soil is then packed
to three-fourths fill the tube, with a wooden rod. Sandy soils or peats are
packed as firmly as possible at all moisture contents, as there is little danger
of making them impervious to the displacing liquid. The finer classes of
soil are packed more lightly and used at a moisture content somewhat be-
low the field capacity to insure adequate rate of solution displacement. A
silt loam is best used at a moisture content of about 20 per cent, and when
properly packed has an apparent specific gravity of 1.50 to l .60.
10-46. The leaching tube is suspended from a ring stand and 300 to 500
ml of water containing 0.5 per cent of KCNS is added. This solution dis-
places or pushes the soil solution ahead of it without rapid mixing of the 2
solutions, making it possible to get a portion of the actual soil solution
(usually 20 to 50 per cent) before any of the displacing solution comes
through. After about 20 per cent of the soil solution has come through,
the remainder is caught in successive 25-ml portions and a few drops are
tested on a spot plate with 4 per cent FeCl:1 solution. The first appear-
ance of a pink or red color in this test indicates CNS- ions of the dis-
placing solution are coming through and collection of the solution is
stopped.
10-47. Soil Solution Obtained with a Pressure Membrane.H Extraction
of the soil solution by means of the pressure membrane is slower and re-
quires a larger soil sample than do water extracts, but has the advantage
that the soil solution is obtained without dissolution of additional materials
from the soil. The method serves well for more exacting studies and as a
means to check the extent that the results are being affected by water ex-
traction procedures. 45 In the event that air-dried soil is to be examined,
the soil first is sprayed with a fine mist of water while being rapidly turned
in a can or on a rubberized mixing sheet. The soil is equilibrated with the
water by storage at high humidity and constant temperature for a period of
2 weeks, with occasional mixing. The field-moist, or moistened soil after

40 Burd and Martin, J. A gr. Sci .. 13: 265 (1923).

41 Rader, Anal. Chem., 19:229 (1947).
42 Parker, Soil Sci .. 12:209 ( 1921 ).
48 Anderson et al .. U .S.D.A. Tech. Bui. 813 ( 1942); White and Ross, J. A gr. Res ..
59:81 (1939).
44 Reitemeier and Richards, Soil Sci., 57: 119 ( 1944).
45 Reitemeier, Soil Sci., 61: 195 ( 1946).
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 249

storage, is packed by hand into the pressure membrane cylinder46 to a

depth of 5 to 10 or more cm. The cylinder is then closed, and the extrac-
tion process carried out at 15 kgm per cm 2 ( 15 atmospheres, or 225 lbs. per
sq. inch).
10-48. The 1 : 1 Soil: Water Extraction of Soluble Salts. A soil sample
consisting of 50 to 1500 gm of dry soil or its equivalent of field-moist soil
is added to an equal weight of CO~-free water (the water in the soil being
included in this total) in a bottle of suitable size. (A small sample is em-
ployed if only conductance measurements are to be made.) The bottle is
stoppered and shaken for a period of 2 hours, preferably in a rotary shaker.
The suspension is then filtered on a large Buchner funnel fitted with a care-
fully sealed medium porosity filter paper. The filter paper is sealed by wet-
ting it with a little distilled water and pulling it down with suction. The
excess water pulled through is discarded before the filtration is begun. The
first portion of the filtrate may be cloudy, and is either discarded or poured
back through the filter. The filter is covered with a glass during the filtra-
tion process to retard evaporation. The conductance is measured ( ~ 10-3 3)
and then the concentration of salts in the 1 : 1 extract solution is auto-
matically the concentration in the soil, oven-dry basis.
10-49. As a simple alternative 47 a 50-gm soil sample is placed in a Col-
lodian bag suspended in 50 ml of water in an extraction flask. Gentle agita-
tion gives chloride equilibrium into the outer solution within 24 hours.
l0-50. The 1: 2 Soil: Water Extraction of Soluble Salts. The 1 : 2
soil : water extraction has been employed extensively by research workers
in the humid soil region, particularly in connection with highly fertilized
greenhouse soils. 48 One hundred gm of dry soil or its equivalent of field-
moist soil (more if ionic analyses are to be made) is placed in 200 ml of
water in a 500-ml conical flask. The flask is stoppered and the suspension
is shaken for a period of 2 hours or over night. The solution is filtered, and
the conductance is measured (~ 10-33). The concentration of soluble
salts in the 1 : 2 extract, multiplied by 2 is the concentration in the soil,
oven-dry basis.
l0-51. The factor(~ 10-38, eq. 10-12) calculated for the 1 : 2 extract
data from greenhouse soils49 was 0.073 (instead of 0.064) after 2 weeks
equilibration of fertilizer salts with soil before extraction, but increased to
0.094 after 4 weeks equilibration; and this value was also the "best fit" for

46 Available commercially from the Instrument Developmen t and Manufactur

ing
Corporation , P.O. Box 191, East Pasadena, California.
47 Hester, Sci., 107:99 (1948); Pierre and Parker, Soil Sci., 23: 13 (1927).
48 Merkle and Dunkle, J. Am. Soc. Agron., 36: to (l 944); Sweet and
Peech, Farm
Research, 11:4 (April 1945), and Industrial Instruments, Litho. circular, Jersey City,
N.J.
49 Dunkle and Merkle, S.S.S.A. Proc., 8: 185 (1944).
 AND WAT ERS
250 SOL UBL E SAL T ANA LYS IS FOR SOILS
autho r's labor atory gave a
a large numb er of collected soils. Data in the
50
d with 4-10 -10 fertilizer,
factor of 0.12 for the filtrate of a soil equilibrate
filtrate. It appears that a
based on the gravimetric weight of solutes in the
in solution is applicable to
factor of abou t 0.1 for Lmmh or cm to % salts
h calcium and sulfate are in
1 : 2 extracts from highly fertilized soils in whic
fairly high proportion.
ce of solutions derived
10-5 2. The interpretation of specific conductan
subject to variation with the
from 1 : 2 or other fixed soil : water ratios is
literature reveals attem pts
wilting percentage or soil texture ( ~ 10-3 5). The
ence to soil texture. A 1 : 2
to set a limit for the 1 : 2 extract without refer
of the soil solution at wilt-
extract of a sand involves much more dilution
relation between the spe-
ing than does a 1 : 2 extract of a silt loam. The
extracts may be calcu-
cific conductance of the 1 : 2 and the saturation
lated (Tab le l 0-6) by the equation:
200 . - (10- 17)
L•At. ext. = L, : 2 X -·-- ·--- --.. :--- ·---;-··--- --
% water m soil at saturation
x dissolved ( ~l 10-1 8) in
except for additional Ca SO, · 2 H 2 0 and CaCO
of cond uctance for I : 2 soil
the more dilute extraction. The uppe r limit
for best growth of toma to
extract appeared to be 0.75 to 1.25 mmh os/cm
ty due to flooding of soils
seedlings, 51 1.0 mmh o/cm for correction of salini
freedom from repression of
by sea water, 52 and 2 to 4 mmh o/cm for
53 Thes e values corre spon d
germination and growth in greenhouse soils.
ty scale (Tab le 10-5 ) for
(Tab le 10-6 ) to satisfactory values on the salini
be excessive on coarse tex-
good growth on fine textured soils, but would

TABL E 10-6
: water extrac t to that
Relati onshi p of specific condu ctance in I : 2 soil es
in the satura tion extrac t for 2 soil textur
---- -,-- ---- ---- ----of---· ·--- ---
the I : 2 extrac t Specific condu ctance the satura tion
Specific condu ctance of
(obse rved) extrac t (calcu lated)
-----
For clay loam, high
---
·-----
~-~---
------

Expressed as For silt loam organ ic matte r (sat.

Expressed as
millim hos/c m mhos X 105/cm (sat. % = 40) % = 100)

75 3.75 1.5
0.75 2.0
1.0 100 5.0
200 10.0 4.0
2.0 6.0
3.0 300 15.0
400 20.0 8.0
4.0
------
( 1944) .
50 Merkle and Dunk le, J. Am. Soc. Agron ., 36: 10 .
Farm Resea rch, 11 :4 (Apri l 1945)
l'il Sweet and Peech.
ii!! Jndus trial Instru ments , Litho.
circul ar, Jersey City, N.J.
5a Dunk le and Merkl e, S.S.S. A. Proc.,
8: 185 ( 1944) .
 S AND WATERS 251
SOLUBLE SAL T ANALYSIS FOR SOIL
greenhouse soils undoubtedly
tured soils. The 1 : 2 extracts from fertilized
um that did not actually occur in
involved extraction of considerable gyps
es found are higher than would
the soil solution. Thu s the conductance valu
indeed striking, nonetheless, how
be found in the saturation extract. It is
saline soil studies of the other-
similar the conductance ranges are for the
ns.
wise contrasting alkaline and humid soil regio
n of Soluble Salts.M Field-moist
10-5 3. The 1: 5 Soil : Water Extractio
to 140 gm of oven-dry soil, and
or air-dry soil is weighed out equivalent
of C0 2-free water (less the cal-
transferred to a 1-liter bottle. The n 700 ml
ple) is added. The bottle is stop-
culated amount of water in the soil sam
2 hours, and then allowed to settle
pered and placed in the rotary shaker for
liquid is filtered. A Pasteur-Cham-
for 0.5 hour, after which the supernatant
re cleanness of the filter tubes,
berland filter may be employed. To insu
collected filtrate shows a very low
distilled water is passed through until the
disconnected and drained. The
electrical conductance. The filters are then
is decanted into the Pasteur-
supernatant liquid in the extraction flask
ted under 10 or 15 pounds per
Chamberland filter cup, and filtration is effec
ml of filtrate is discarded and then
square inch pressure. The first 30 to 50
clay filter tubes are cleaned and
250 or more is collected for analysis. The
y for use again. The concentration
rinsed under pressure, and air-dried, read
iplied by 5 is the concentration
of the soluble salts in the 1 : 5 extract mult
in the soil, oven-dry basis.
soil at extraction) of equation
10-5 4. The factor (0.064 x % water in
or, for L in mhos, 320. Joseph
10-1 3 for a 1 : 5 extract becomes 0.320,
on the specific resistance of 50
and Martinr.n used the factor 250 based
per cent NaCl and 50 per cent
ohms for a 1 per cent salt solution (50
ivalent to 0.07 5 in equations
Na!.!SO 4 ); Piper511 used the factor 37 5 (equ
10-1 1). The range in these values
10-1 2 and 10-1 3, and to 750 in equation
analysis.
is some measure of the uncertainties of the
SOIL SALINITY
MEASUREMENT OF EFFECTS OF
ON SEED GERMIN ATI ON
sensitive to soil salinity during
10-5 5. Many crop plants are especially
y occurs even with plants that
the seed-germination stage. This sensitivit
s of growth, examples being
are relatively salt tolerant during later stage
of salinity during germination are
sugar beets and alfalfa. 117 Adverse effects
of Soils, University of Nebr aska (193 6),
Dilution was employed at the Division to
Institute, Adelaide, Australia according
54
and the Waite Agricultural Research York: Interscience Publishers, Inc., 1944).
Piper, Soil and Plant Anal ysis (New by
an's pressure extract has been reported
Com paris on of the 1 : 5 extract to Lipm
Burgess, Soil Sci., 14: 191 (192 2).
uaJ. Agr. Sci., 13:52 (192 3).
r.s Piper, op. cit., p. 32.
57 Ayers and Hayward, S.S.S
.A. Proc., 13:224 (194 9).
252 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS
largely responsible for the bare spots contrasting to good growth with only
a few cm transition in the field. The literature on this subject has been ex-
tensively reviewed. 58

APPARATUS

10-56. Needed apparatus includes a constant temperature room, 20-cm

culture dishes, a 4-mm sieve, rubberized cloth or Cellophane, a mixing
spatula, moist soil containers, and an electrical conductance bridge to test
the final salinity of samples.

REAGENTS

10-57. Needed reagents include NaCl and other salts, including fertilizer
salts. Barley and other seeds and saline and nonsaline soils are also needed.

PROCEDURE

10-58. The soil is passed through a 4-mm sieve, and the salt content is
adjusted by the addition of with NaCl (or other salt to be tested) so as to
give in successive lots, 0, 0.1, 0.15, 0.2, 0.25, and 0.3 per cent salt on a dry
soil basis. The soil-moisture content is then brought into the moisture range
from wilting percentage to field capacity as follows: the soil sample is
spread out in a thin layer on rubberized sheet or waterproof Cellophane," 11
and the calculated amount of water is sprayed on the soil in small incre-
ments, after each of which the soil is mixed with a spatula with care to
avoid puddling. If wetted lumps appear, the soil is placed on the 4-mm
sieve and gently shaken to pass only the fine soil, the lumps being retained
on the sieve. When the calculated amount of water has been added, the
whole sample of soil is mixed and placed in a container, which then is
tightly closed. The soil is stored in a constant temperature room at 70°C for
about two weeks for equilibration.'w Occasional rotation of the container to
mix the soil speeds up the equilibration of moisture and salts throughout
the soil mass.
10-59. After equilibration, 1.4-kgm soil samples are weighed and placed
in 20-cm culture dishes. A definite number of seeds is planted, 20 in the case
of barley. The dishes are covered and maintained at constant temperature
to prevent distillation and condensation of moisture. The number of emerg-
ing seedlings is counted each day. The results are expressed as the percent-
age of emergence at various time intervals.
10-60. The electrical conductance of the saturation extract is determined
(~ 10-27) from subsamples of the moistened soil (~ 10-58) taken at the
time of planting. From this and the measured soil moisture content of the

r.RUhvits, Am. J. Bot., 33:278 (1946).

GilTechnique described to the writer by Dr. G. M. Volk ( 1946).
oo Ayers and Hayward, S.S.S.A. Proc., 13 :224 (1949).
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 253
soil, the osmotic pressure of the soil solution may be calculated (~ 10-39).
1f the relation of moisture stress to soil moisture content is known, the total
moisture stress may be calculated. 61
10-61. The procedure may be adapted to the determination of the influ-
ence of different single salts, cation exchange status, germination tempera-
tures, and the effect of different fertilizer salts 62 on germination. Critical
percentages of NaCl (dry soil basis), above which germination was greatly
retarded were found by Ayers and Hayward to be 0.1 per cent for barley,
and 0.04 to 0.05 per cent for corn and beans. Expressed as percentage
of NaCl in solution,u:i the critical percentage for hemp was 0.25 per cent
and 0.5 for clover, wheat, rye, beans, and peas, which agree well with the
above values for the dry soil basis.

COMPOSIT ING RUNOFF WATERS AND DETERMIN ATION OF

TOTAL SUSPENDE D SOLIDS OF RUNOFF WATERS
10-62. The analysis of runoff waters differs somewhat from that of well,
lake, and river waters ( ~ 10-81 ) , because they contain rather high amounts
of suspended solids in addition to the dissolved solids ordinarily determined
in waters. The determination of total suspended solids of runoff waters is
carried out (~ 10-7 I) to provide a base for the calculation of the chemical
constituents in terms of soil material lost by erosion, as well as to determine
the amount of physical erosion that has taken place. In addition, a few
special procedures are given here for getting the runoff waters in condition
for the respective chemical determinations given elsewhere in the text in
the respective sections where the analysis of soil for the same constituent
is given.
10-63. Of interest in the interpretation of runoff analyses in terms of soil
fertility is the "enrichment ratio":
. . Concentration in soil material in runoff
Ennchment ratio=----- -- ---------------------- ----------- ·-----
(of a soil constitiwnt) Concentration in soil from which runoff originated
(10-18)
APPARATUS
10-64. Needed are an electric stirring apparatus with collapsible paddle
(Fisher Scientific Co., Pittsburgh, Pa.) and a set of graduated beakers each
of which holds, when level full, the size of aliquot wanted for a particular
determination. The beaker volume is adjusted with melted paraffin at the
bottom.

u1 Wadleigh, Soil Sci., 61 : 225 (l 946).

6~ Lundstrom, Research report 134, Division of Fertilizers and Agricultural Lime,
Bur. Plant Ind., Soils Agric. Eng .. U.S.D.A. (1948) as cited by Ayers and Hayward,
S.S.S.A. Proc., 13:224 (1949).
1m Harris, Soil Alkali, Its Origin, Nature, and Treatment (New York: John Wiley
& Sons, Inc., 1920).
254 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS

REAGENTS

10-65. Reagents needed are toluene and precipitated CaC03 (for treat-
ment of the nitrate samples).

PROCEDUREli4

10-66. Compositing of Runoff Water Samples. Each suspension is thor-

oughly mixed and the volume of each is taken in proportion to the fraction
of runoff from a given plot it represents. A few samples are readily com-
posited by agitation of the bottles by hand, but if many samples are to be
composited, an electric stirring motor is employed. The proper amount of
each suspension to give a total 3 liters of composited suspension volume is
measured in a graduated cylinder with caution to keep the suspension of
uniform concentration during the process. The composited suspension is
placed in a 4-liter ( 1-gallon) jug, and 1 ml of toluene is added to suppress
bacterial activity.
10-67. Nitrate Samples. Since nitrates are subject to denitrification in
solution, a 1-liter sample of the runoff water from each tank is treated with
0.25 gm of CaCO:i and dried immediately after collection. When separate
samples are dried from each of several tanks that are to be reported as 1
composite sample, data are submitted with the samples showing the per-
centage from each tank for the composite sample. Nitrates are determined
according to ~I 8-59.
10-68. Correction Factor. The amount of dilution caused by precipita-
tion that falls in the tanks is calculated from the exposed area of the tank,
and the "corrected runoff volume" in the tank is calculated. The analytical
results for the diluted runoff are corrected by a "dilution factor" calculated
as follows:
Dilution factor = lJ_n~orrected ru.noff_v()It1111e me<lsured ( 10_ 19 )
Corrected runoff volume calculated
10-69. The runoff volume measured also includes the volume of the
solids present. For runoff containing more than 1.5 gm of solids per I 00
gm, a correction is made for the volume of solids:

% water in runoff = 100 - ~. ~!11 solid e.~~_!Q2_~11:1-~_noff)

2.65
(10-20)
The total correction factor to be applied to analyses of runoff ( ~ 10-73) is
then calculated:

64 The writer is indebted to Dr. H.F. Massey, J. C. Kaudy, and J. Dana-Bashian at

the University of Wisconsin for their participation in the runoff analysis through
which these procedures were developed. Also, helpful suggestions were made by 0. E.
Hays of the U.S.D.A., and Dr. H. Kohnke of Purdue University.
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 255
. f t
C orrec tion Dilution factor x 100
ac or = (10-21)
% water in runoff

10-70. Subsampling for Analysis. In subsampling of a composite runoff

water for analysis, the suspension is first thoroughly mixed. Since it is easier
to measure out a constant volume, a calibrated (~ 10-64) beaker is used
that holds just the desired quantity of suspension for a given analysis. This
allows the sample to be poured quickly without settling, the excess merely
spilling over the top.
10-71. Total Solid Content of Runoff. Two separate 100-ml samples are
measured out and poured into the same tared 250-ml beaker, the solids be-
ing rinsed out of the measuring beaker with distilled water. The 250-ml
beaker is placed on the steam plate and brought to dryness at I 00°C. (The
process may be hastened by boiling the liquid down to low volume on an
electric hot plate.) The sample is not allowed to bake on the hot plate after
drying, as organic matter may be decomposed. The beaker is allowed to
cool and then is weighed.
10-72. When the beaker contains less than 1 gm of solids, it is weighed
on an.analytical balance to the nearest mgm. Otherwise, a torsion balance
sensitive to 0.05 gm is satisfactory. At least 0.25 gm of solids is required
for the organic matter determination (~I 9-50), and it may sometimes be
necessary to dry down more than 200 ml of runoff to obtain this amount of
solids. If additional volume is necessary, the sample is removed from the
hot plate just prior to dryness, 2 additional 100-ml aliquots are added, and
the beaker is again placed on the hot plate.
10-73. The concentration of solids as gm per I 00 ml of runoff is calcu-
lated, and from this the pounds solids per acre inch of runoff (ppai) is
calculated:
Solids (ppai) = gm solids per 100 ml x 2270 x correction factor ( 10-22)
The correction factor is obtained from equation I 0-21. The factor 2270 is
derived for the pounds of water per acre inch as follows:

43560 s . ft. er acre x 62.4 ~ounds per Cl1:_!!..: = 226512 poun~s of water
q P 12 mches per foot per acre mch
(10-23)

The conversion factor for percentage to ppai is therefore 2270, and for
ppm to ppai is 0.227.
10-74. The chemical analyses usually made on runoff waters include:
organic matter (~! 9-50), combined organic and ammonium nitrogen
(~ 8-16), nitrate nitrogen (~ 8-59), soluble and extractable phosphorus
(~ 7-75), and exchangeable potassium(~ 18-26).
256 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS

DETERMINATION OF TOTAL DISSOLVED SOLIDS

OF SOIL EXTRACT OR WATER
10-75. The soil extract or water is evaporated to dryness and the quan-
tity of dissolved solids is determined gravimetrically. The method is a direct
measure of total salinity, and serves as a check on the results by the more
rapid electrical conductance methods ( ~ 10-18) and on the chemical an-
alysis for individual ions (~ 10-78). Two procedures have been widely em-
ployed, 1 involving drying at 110°C and the other at 550° to 600°C for 5
minutes. The object of the latter is to remove organic matter and measure
only salts; prolonged heating is avoided because of losses of salts, particu-
larly MgCl 2 •

APPARATUS

10-76. Needed apparatus includes platinum or porcelain evaporating

dishes, pipets, a l l0°C oven, a muffle furnace, and an analytical balance.

PROCEDUREllfi

10-77. An evaporating dish, preferably of platinum, is carefully cleaned,

ignited, and weighed to 4 places. Then 50 ml of the clear extract or water
is measured out by means of a pipet. The solution is evaporated to dry-
ness at I 10°C (the platinum dish being set on a cloth-wrapped ring, not on
metal or porcelain) then is cooled and weighed. The total dissolved solids
are reported as ppm in solution.
DETERMINATION OF INDIVIDUAL CATIO NS AND ANIONS
OF SOLUBLE SALTS IN SOILS AND WATERS
10-78. Analyses are often made of soil extracts (~ 10-18, 5-3), of
lysimetern 6 percolates, and of waters from rivers, lakes, wells, and runoff.
The determinations of the individual cations of the soluble salts and waters
include those of Nat, K +, Ca++, Mg++ (~ 10-82 to l 0-86). The
soil : water ratio used in the extraction of soils affects the ionic composition
of the extract ( ~ 10-18). Procedures for the dissolved anions C03 - - ,
HCO:i - , CJ-, SO 4 - - , gypsum, and carbonate carbon are given in follow-
ing sections. Other anions in solutions are determined by procedures given
in appropriate places: NOa- (~ 8-48), B0:1 - - - (~ 14-7), P0 4 - - -
(~ 7-14) andSiO:i-- (~ 11-72).
10-79. Water Samples.n 7 Because representative waters have been ana-
lyzed, an inquiry from the state chemist or engineer, or Agricultural Experi-

i;;; Official and Tentative Methods of A na/ysis, 6th ed. (Washington, D.C.:
A.O.A.C., 1945).
6!; Lysimeter technique has been reviewed by Kohnke et al .. U.S.D.A. Misc. Pub.
372 (1940) and Harrold and Dreibelbis, U.S.D.A. Tech. Bui. 1050 (1951).
67 Magistad and Christiansen, U.S.D.A. Cir. 707 (1944 ).
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 257
ment Station should be made prior to water sampling from common
sources. If sufficient information is not available, 4 liters t 1 gallon) of
water sample is collected for analysis. The container i5 thoroughly washed
and then rinsed several times with the water to be sampled. In sampling a
stream, water is taken from a rapidly flowing part. The kind of water
(spring, stream, lake, well) should be specified, together with location,
depth, temperature, odor, color, use to be put to, users opinion of its qual-
ity, and other pertinent information. Collection of rain water (particularly
for sulfur studies) by means of a frost-proof rain and snow gauge has been
described.° 8
10-80. Interpretation of Quality of Irrigation Waters. The quality of
irrigation waters is dependent on the total salt content, on the nature of the
salts (particularly of Na and B) present in solution, and the proportions of
meq of Na to meq of other metallic cations, and of Ca to Mg to bi-
ca bonates. The total salt content as parts per million can be estimated
rather accurately from the specific conductance (~ 10-33). The ppm
multiplied by 0.00136 gives the tons of salt per acre foot of water. Also
ppm x 0.0586 equals grains of salt per U.S. gallon. The quality of irrigation
water can be evaluated in terms of the conductance and the sodium adsorp-
tion ratio (Fig. 10-7). A guide to the quality of water in terms of specific
conductance, sodium percentage, boron content, 69 and residual Na 2 C0x 70
is given in Table 10-7. The sodium percentage is the percentage of Na meq
of the total meq of cations by analysis. The range of concentration of boron
is given for the more sensitive crops such as lemons, grapefruit, and navy
beans (low number of the range) to the more tolerant crops such as car-
rots, cabbage, and alfalfa (higher number in the range). The "residual
Na 2 COa" of water is defined as follows:

TABLE 10-7
Guide to the quality of irrigation water

Specific Residual Quality of

conductance, Sodium Boron, Na2COa irrigation
mmho/cm percentage ppm meq/I water
<0.75 <65 0.3-1 <<1.25 Excellent to good
0.75-2.0 50-65 0.7-2 <1.25 Good to permissible
2.0-3.0 92 1-3 1.25-2.5 Doubtful to unsuitable
>3.0 >92 1.2-3.8 >2.5 Unsuitable
·--··--··--~

as Fried and Jackson, Sci., 106: 19 (1947).

mi Wilcox, U.S.D.A. Cir. 784 (1948) (condensed and rearranged from this source).
70 Wilcox et al., Soil Sci., 77:259 (1954).
258 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS

0.1 0.2 0.3 0.4 0.7 1.0 2 3 4 5.0

30

28 Cl·S4

26 C2-S4

24 C3..s4

22
C4·S4

.......
z"'
g 16
E
II
.Q 14
eC: Cl·S2
0 12
~
~ C2-S2
~ 10
E
:::i

~ 8

6
Cl SI
4 C2-Sl
C3·Sl
2

.10 0.25 0.75 2.250

Conductivity MM HOS/cm (EC x 108) at 25°C

2 3 4
Low Medium High Very high
Salinity hazard

Fig. 10-7. Guide to quality of irrigation waters. (After Richards, ed.,

U.S.D.A. Agr. Handb. 60, 1954, p. 80.)

when all quantities are expressed as meq/liter. The salt content of river
water tends to increase with distance from the source, especially if the
water is repeatedly used for irrigation en route.
10-81. The following are representative analyses 71 of river water for
conductance and individual ionic species.

71 Wilcox, U.S.D.A. Tech. Bui. 962, Table 2 (1948).

SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 259
North Platte R. Gila R.
Scottsbluff, Nebr., Amhm:st-Hyden Dam
1945 Ariz., 1933
L mmho/cm ........................ . 0.28 1.33
Per cent Na ......................... . 22. 60.
B, ppm ............................. . O.o3 0.20
Silica, Si0 2 , ppm ..................... . 15. 15.
Cations: meq/I.
Ca ............................... . 1.61 3.39
Mg .............................. . 0.57 1.59
Na ............................... . 0.63 7.75
K ................................ . 0.o7 0.25

Total ........................... . 2.88 12.98

Anions: meq/I.
CO a ............................. . 0.0 0.30
HC03 ............................ . 1.59 2.75
S04 .............................. . 1.12 2.29
CI ............................... . 0.09 7.07
NO:i .............................. . 0.03 0.01

Total ........................... . 2.83 12.42

10-82. Sodium Determination. An aliquot for Na analysis is trans-

ferred to a small beaker and the Na is determined by emission spectro-
photometry (~I 18-16). Determination of Na can also be made by the
sodium magnesium uranyl acetate procedure. An aliquot of extract equiva-
lent to I to 5 mgm of Na is transferred to a small Pyrex beaker, neutralized
with HOAc, and evaporated to a volume of I ml, with care not to evapo-
rate to dryness with consequent dehydration of silica. The Na is precipi-
tated from this solution ( ~ 5-81 ) . The change in specific gravity caused by
uranyl acetate precipitation of Na has been adapted to rapid testing.i 2
10-83. Potassium Determination. An aliquot for K analysis is trans-
ferred to a small beaker and determined by emission spectrophotometry
( ~ 18-16). Determination of K can also be made by precipitation as the
cobaltinitrite. An aliquot containing 0.5 to 5 mgm of K is transferred to
a Pyrex beaker, ammonia is removed, and the precipitation is effected
( ~ 6-20). The K can be estimated in rapid testing by the turbidimetric co-
baltinitrate procedures ( ~ 13-56).
10-84. Combined Sodium and Potassium Determination. A simple ap-
proximation methodrn for the estimation of combined Na and K is provided
by their combined weight as the chlorides and the chloride content (similar
to the titration of the combined alkali carbonates, ~ 4-49). Bray's alco-
holic ( NH 4 ) 2 C0a solution is prepared by dissolution of I 00 gm of pow-
dered ( NH 4 ) 2 C0 3 in 190 ml of water and addition of 210 ml of concen-

n Bower, J. Am. Soc. Awon., 40: 1100 ( 1948).

n Division of Soils, University of Nebraska (1935).
260 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS
trated NH 40H. This solution is then diluted with an equal volume of 95
per cent ethanol as needed. The filtrate from the precipitation of sulfate as
BaS0 4 (~ 10-102) is evaporated to dryness in a porcelain dish and 5 ml
of water and 20 ml of Bray's reagent are then added. The mixture is
stirred occasionally for a half hour or more (or over night) and then is
filtered through a dry paper. The filter is washed with Bray's reagent and
then discarded. The filtrate is evaporated to dryness in a weighed platinum
or rhodium dish and ignited to faint redness (540°C) to expel ammonium
salts. The residue is taken up in water and a few drops of HCI and brought
to dryness again, then dried in an oven at 110°C. The residue is weighed as
combined NaCl and KCI. The residue is dissolved in water and made to
volume in a volumetric flask. The Cl is determined by titration (~ 10-94)
of a suitable aliquot diluted to 50 ml. Then:
weight Na= 3.004 (chloride) - 1.428 (residue.) (10-25)
weight K = 2.428 (residue) - 4.004 (chloride) ( 10-26)
in which (chloride) and (residue) refer to the weights of Cl and combined
NaCl and KCI.
10-85. Calcium Determination. An aliquot for Ca analysis is trans-
ferred to a small beaker and the Ca is determined by emission spectro-
photometry (~ 18-16). Determination of Ca (0.2 to 5 mgm) can also be
made by procedures employing Versene (~I 11-47) or oxalate(~ 5-38).
10-86. Magnesium Determination. An aliquot for Mg analysis is trans-
ferred to a small beaker and determined by emission spectrophotometry
(~ 18-16). Determination of Mg (0.3 to 2 mgm) can also be made by pro-
cedures employing Versene (~ 11-59) or hydroxy quinoline (~ 5-59).

DISSOLVED CARBONATE AND BICARBONATE

DETERMINATION
("Total alkalinity" of waters)
10-87. The titration of dissolved carbonate (C0 3 - - ) and bicarbonate
(HCO:i - ) in waters, including soil solutions, as conventionally carried out,
is often termed "total alkalinity" because it includes the usually small
amounts of phosphates, borates, and silicates as well as the two species of
carbo"nate ions. The alkalinity other than bicarbonates and carbonates in
waters and soil solutions are generally small enough to be negligible, and
thus the total alkalinity is virtually equivalent to the 2 dissolved carbonates.
Solid carbonate is determined separately as carbonate carbon ( ~ I 0-121).

APPARATUS

10-88. Needed apparatus includes a 150-ml beaker, a 50-ml pipet, and

a buret for standard 0.05 N H 2 S04 •
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 261

REAGENTS

10-89. Needed reagents are standardized 0.05 N H~S0 4 ,- phenolphthal-

ein and methyl orange pH indicators (alternatively, brom phenol blue pH
indicator may be substituted for methyl orange).

PROCEDURE74

10-90. To 50 ml of water sample in a 150-ml beaker, 0.15 ml of phenol-

phthalein indicator is added and, if a pink color develops, normal carbonate
(C0:1 - - ) is present. The C0: 1---- is titrated with 0.05 N H~S0 4 , a drop be-
ing added every 2 or 3 seconds until the pink color disappears. The buret
reading in ml multiplied by 2 gives the C0:1- - as meq per liter, or equiva-
lents per million parts of solution.
10-91. Bicarbonate Titration. To the solution resulting from the COa - -
titration, or to the original solution if no pink color resulted, 0.1 ml of
methyl orange or brom phenol blue indicator is added, and the titration is
continued (without a refilling of the buret) to the first change in the methyl
orange color or to the midcolor of brom phenol blue. The total buret read-
ing is recorded. The solution is reserved for the chloride determination.
10-92. If C0:1 - was absent, the buret reading in ml is numerically
equal to the meq of HCOa per liter, or equivalents per million parts of
solution. If C0:1- - was present, the CO:i - - titer in ml multiplied by 2 is
subtracted from the total titer volume and the difference in ml is the meq of
HC0:1- per liter or the equivalents per million parts of solution. Blanks are
run on the reagents and subtracted from the determinations.
10-93. If small amounts of C0 3 - -- are determined in the presence of
large amounts of HC0: 1 - , a more accurate evaluation of CO a···- ··· can be ob-
tained by the use of the Hirsch carbonate equilibria slide rule. 70

CHLORIDE DETERMINATION
10-94. As AgNOa solution is titrated into a chloride solution in the
presence of Cr0 4 , only momentary formation of red Ag~Cr0 4 occurs so
long as some chloride persists in the solution. When the Cl in solution is
exhausted through precipitation as AgCl, the red precipitate of Ag~Cr0 4
sharply signals the end point. Alternatively, an Ag-AgCl electrode also
sharply registers the change to an excess of Ag ion. Thus the AgNOa titra-
tion of chloride is the standard method, either end point measurement be-
ing satisfactory. The electrode method may be employed in the presence of
soil and thus eliminates the need of extraction, if only chloride is to be
determined.

74 Magistad et al., Soil Sci., 59: 65 (1945).

75 Ind. Eng. Chem., A.E.. 14:943 (1942).
262 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS

APPARATUS

10-95. Needed apparatus includes a white casserole or other dish, a

buret, and pipets.

REAGENTS

10-96. Needed reagents are 0.1 N H 2 S04 and 0.1 N Na 2 COa, phenol-
phthalein indicator, standard NaCl (2.923 gm per liter for 0.05 N; 1.648
gm per liter for I mgm Cl per ml), and the following special reagents.
10-97. Standard Silver Nitrate Solution. Exactly 8.494 gm of AgNOa is
dissolved in water and diluted to I liter to obtain an 0.05 N solution. (Al-
ternatively 4. 791 gm per liter gives a solution each ml of which is equivalent
to 1 mgm of Cl.) The AgNO:i solution concentration is usually checked by
titration against a standard solution of NaCl.
10-98. Potassium Chromate Indicator. Approximately 5 gm of K2 Cr0 4
is dissolved in 80 ml of water, and then saturated AgNO:i solution is added
dropwise with stirring until a permanent red precipitate is produced. The
solution is filtered, and the filtrate is diluted to 100 ml.

PROCEDURE7fl

10-99. The solution employed for the bicarbonate titration (or a fresh
solution containing 5 to 25 mgm of chloride in a volume of 25 to 100 ml,
obtained by dilution or evaporation; the pH is adjusted to 8.2, just colorless
to phenolphthalein indicator with 0.1 N H 2 SO 4 or 0.1 N N a 2 C0,1 solution)
is employed for the chloride titration. One ml of the chromate indicator is
added and the solution is titrated with the standard AgN0:1 solution to the
appearance of the first permanent red coloration due to precipitation of
Ag2 Cr0 4 • A blank is titrated consisting of the same volume of chloride-free
water. For a 50-ml aliquot of water, the net ml titer, after subtraction of the
blank, is equal to the equivalent of Cl per million parts of solution (or
equal to the mgm of Cl titrated with the alternative AgN0:1 standard
solution).

ALTERNATIVE PROCEDURES

10-100. More than 25 mgm of Cl may be determined satisfactorily by

precipitation as AgCl and weighing. 77 In order to avoid a large excess of the
AgN0 3 precipitation reagent, a preliminary titration of the Cl is made. The
pH of the solution is adjusted as in the regular procedure and a slight excess
of AgNOa is added. The solution is heated to boiling, protected from light,

76 Magistad et al., Soil Sci., 59: 73 ( 1945), which is only slightly modified from
Methods of Analysis, 6th ed. (Washington, D.C.: A.0.A.C .• 1945), p. 632.
77 This procedure is employed for standardization of HCI after neutralization with
the CaC03, Methods of Analysis, 6th ed. (Washington, D.C.: A.0.A.C., 1945), p. 25.
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 263
and allowed to stand until the precipitate is granular. The precipitate is
filtered on a Gooch crucible (previously weighed at 140° to 150°). The
precipitate is washed with hot water and the filtrate is tested to verify an
excess of AgNOa· The AgCl is dried at 140° to I 50°C and weighed.
10-101. The silver-silver chloride electrode may be substituted 78 for the
chromate indicator in the AgN0;1 titration of Cl. The cell consisted of the
following:
Pt, Quinhydrone, KN0 8 11 KN0 3 11 Agel, Ag (10-27)
(Sat.) (Sat., in (In the Cl-
agar) solution)

in which 11 represents a liquid junction and AgCl, Ag the silver-silver

chloride electrode. Standardized dilute AgN0 3 is buretted into the soil sus-
pension and the equivalence point is indicated when a rapid shift occurs in
the potential of the cell.

SULFA TE DETERMINATION
10-102. Sulfate in extracts and waters is conventionally determined
gravimetrically as BaSQ1• The sulfate solution is made 0.1 to 0.3 N in HCl
and boiled to remove carbonates, and then BaCl 2 is added to cause the
precipitation ( ~ 11-188). Because precipitation, digestion, filtration, wash-
ing, ignition, and weighing constitute a time-consuming procedure, rapid
titrimetric and turbidimetric procedures have been developed. The titri-
metric procedure presented here is based on the Versene chelation of the
excess Ba remaining after BaS0 4 precipitation. The Versene indicator for
Mg is employed and therefore Mg is introduced into the system, to be
chelated at the end point after the Ba has been completely chelated. 79 Be-
cause an excess of barium chloride is employed, the sulphate is precipitated
quantitatively, even in very low concentration. It is not necessary to re-
move the precipitate prior to the titration of the excess Ba, and therefore
the procedure is rapid.
10-103. The sulfate range to which the method is applicable is from 5 to
200 ppm. Interferences occur with Cu, Mn, Co, and Ni, but the concentra-
tions of these ions are generally not sufficiently high to interfere with sulfate
determination in most soils and natural waters. Modifications (~ 11-56)
provide for elimination of their interference. xo

APPARATUS

10-104. Required apparatus includes conical titration flasks, pipets,

and burets.

7H Best, Trans. 4th Intern. Congr. Soil Sci., 3: 162 ( 1950).

111 Biedermann and Schwarzenbach, Chimia, 2: 56 ( 1948).
80 Diehl et al., J. Am. Water Works Assoc., 42:40 (1950), also provide for their
elimination.
264 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS

REAGENTS

10-105. Required reagents include 0.02 N MgC1 2 , BaCl 2 , and Versene

solutions that are standardized against a standard Ca solution (below);
Eriochrome black T, 0.5 gm of the indicator with 4.5 gm of hydroxyl
amine in 100 ml of methanol (~I 11-55); and reagents for dissolved car-
bonate and bicarbonate (~I I 0-89) except that standard HCl is employed
instead of standard H~S0 4 •
10-106. A buffer solution is required consisting of 8.25 gm of NH 1Cl
and 113 ml of concentrated reagent NH.,OH in I liter, a concentration that
should give pH I 0.0 when I 0 ml is added to a 50-ml aliquot of water
sample.
10-107. A standard 0.02 N Ca solution is required for the standardiza-
tion of the Versene, Ba, and Mg solutions. It is prepared by dissolution of
1.001 gm of dried CaCO:; in a minimum of I : 50 HCI; the solution is
boiled to expel the CO~ and then diluted to I liter.

PROCEDURESl

10-108. The combined Ca and Mg are determined by titration with

Versene (~I 11-54). On a second aliquot, the "total alkalinity" is deter-
mined by titration with HCl (not the usual H~S0 4 solution) (~I 10-90). To
a third aliquot, the standard HCl equivalent to the alkalinity or slightly
more is added and the sample is boiled to destroy carbonates. Then a
known amount (enough to exceed the sulfate) of the standard Ba Cl~ solu-
tion is added, and the solution mixture is allowed to boil for a few seconds.
Then it is cooled, I 0 ml of buffer and 5 drops of Erioehrome black T in-
dicator are added. Finally the solution is titrated with the standard Versene
solution to the first end point. The first end point cannot be used as final
because the accuracy is poor even when the standard solution containing
some magnesium ion is employ.ed. A small, known volume of standard Mg
solution is added, and a second end point is obtained by titration with ·
additional Versene solution. The titration is stopped at the same end point
as for the original determination of Ca plus Mg.
10-109. Calculation of Results. The meq of sulfate ion may be calcu-
lated as follows:
meq of S0 4 = (B +Ba+ Mg - T) (10-28)

in which B is the blank, meq of Ca + Mg in the original water; Ba is the

meq of Ba added, Mg is the meq of Mg added, and Tis the meq of Versene
added in the total titration of the sample with Ba and Mg added.

81 Slightly modified from Munger et al., Anal. Chem., 22: 1455 ( 1950).
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 265

AlTERNA TIVE PROCEDURES

10-110. Besides the gravimetric BaS04 (~ 10-102) and Versene titri-

metric methods for sulfate, a wide variety of other methods are available.
Two 82 are the benzidine hydrochloride method and titration with barium
chloride in the presence of disodium tetrahydroxy quinone (THO) indica-
tor. The latter is limited to fairly high concentrations of sulfate because
BaS04 precipitates slowly when the sulfate concentration is low. Another
titrimetric method is based on precipitation of sulfate and chromate with
barium. Barium chromate is added in acid solution and the excess barium is
then precipitated as chromate by addition of NaOH to pH 8.3; the excess
chromate, recovered in the filtrate, is titrated with thiosulfate. This titer
must 83 be standardized against standard sulfate solutions. A micro method
has been described 84 for sulfate in the range of 1 to 300 ugm of S. Sulphate
has been determined by conductance or oscillametric 85 titration with BaCl~;
this is an alternative to titration with BaCI~ in the presence of tetrahydroxy
quinone indicator.
10-111. Turbidimetric Sulfate Determination. The determination of sul-
fate through the turbidity developed on precipitation as BaS0 4 has long
been an attractive and much used method. The estimation has been re-
fined from relatively semiquantitative procedures to satisfactory quantita-
tive procedures. Establishment of relative freedom from interference was
made by Sheen et al. 811 The procedure here given is that of Chesnin and
Yien. 87 To a 20.0-gm air-dry soil sample contained in a 250-ml conical
flask is added a 100-ml volume of Morgan's extraction solution ( 100 gm
of NaOAc and 30 ml of 99.5 per cent HOAc dissolved and mixed in 500
ml of water, and the volume made to l liter). The suspension is shaken for
one-half hour and then is filtered through Whatman's No. 42 filter paper
(or centrifuged until clear). A 10- or 20-ml aliquot is transferred to a
25-ml volumetric flask. The precipitation is then carried out with I gm
(cup measure) of sized BaCl~ crystals (agate mortar ground to pass 0.5
mm and to be caught on 0.25 sieve) added to the aliquot in the flask fol-
lowed by 1-minute of shaking. Then:
1. if the sulfate is below 20 ppm, one ml of 0.25 per cent gum acatia
solution is added.
2. if the sulfate is between 20 and 40 ppm, 2 ml of gum acatia solution
is added.

82 Standard Methods for Examination of Water and Sewage, 9th ed. (New York:
Am. Pub. Health Assoc., 1946).
8X Cantino, Soil Sci., 61 :361 ( 1946).
84 Johnson and Nishita, Anal. Chem., 24:736 ( 1952).
Hr. Milner, Anal. Chem., 24: 1247 ( 1952).
su Ind. Eng. Chem., A.E., 7:262 ( 1935).
87 S.S.S.A. Proc., 15: 149 (1951 ).
266 SOLUBLE SALT ANALYS IS FOR SOILS AND WATERS
The suspension is made to volume and shaken for 1 minute. Turbidity
readings (blue filter in colorimete r) are taken between 5 and 30 minutes
after the precipitation, and the sulfate is then determined by reference to a
standard sulfate curve.
10-112. Exchange Column Purification of Sulfate Solutions. According
to Samuelson, 88 Bahrdt 811 published a method for the rapid estimation of
sulfate in natural waters. The water to be analyzed was softened in a labo-
ratory column containing sodium zeolite. To the effluent from the column,
a known amount of barium chloride was added, and the excess barium was
back titrated by means of potassium palmitate. The Ca and Mg were taken
up so effectively that they could not be detected in the effluent water, and
therefore their interference with the palmitate method was completely re-
moved. Present day organic resin exchangers for cations may be employed
for the Bahrdt separation ; his sulfate determinat ion by difference with Ba
may be completed with Versene titration or with flame emission instead of
palmitate.

GYPSUM DETERMINATION FOR SOILS

10-113. The content of gypsum, CaS0 4 • 2 H~O, in soil is commonly
estimated by the separate determinat ions of Ca (~ 5-38) and S0 1 (~i JO-
I 02) in a water extract that is made at a sufticiently dilute soil : water ratio
to permit dissolution of all the gypsum. A 1 : 5 soil : water ratio dissolves
gypsum to the extent of 1.3 per cent of the soil H I 0- I 8). Extraction of
the soil solution or of the saturation extract (~ I 0-27) removes CaS0 4
only to the extent that it normally contribute s to the osmotic pressure of
the soil solution in the field. The determinat ion of total gypsum in soil is,
however, an important aspect of soil soluble salt analysis. Reitemeier
110

found that 3 factors besides the solution of gypsum may influence the
amount of Ca and S0 4 extracted from soil: (a) solution of Ca from sources
other than gypsum, for example from CaCOx; (b) exchange reactions by
which the dissolved Ca replaces some ions, such as Na and Mg; and ( c) the
solution of S04 from sources other than gypsum. These errors preclude a
highly accurate determinat ion of gypsum.
10-114. Because the separate determinat ion of Ca and S04 ions is a
fairly lengthy procedure, a rapid conductan ce method for gypsum determi-
nation was developed by Bower and Huss, 01 which is given below. Those
authors found good concordan ce of the conductan ce method with the de-

88 Samuelson. Ion Exchangers in Analytical Chemistry (New York:

John Wiley &
Sons, Inc .. 1953), p. 2.
Ml Bahrdt, z. anal. Chem .. 70: I 09 ( 1927).
UOJnd. Eng. Chem., A.E .. 15:393 (1943).
Ill Soil Sci., 66: 199 (1948).
 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 267
termination of the separate ions, whether either Ca or S04 was present in
excess.
APPARATUS

10-115. Needed apparatus consists of 250-ml extraction flasks and

stoppers; 10-ml and 20-ml pipets; filter paper of medium porosity and a
funnel; a centrifuge and conical tube of 50-ml capacity; and a conductivity
cell and resistance or conductance bridge, such as described in ~ 10-21.
REAGENTS

10-116. Needed reagents consist of water for the soil extraction and
reagent grade acetone.
PROCEDURE!l2

10-117. A suitable weight of air-dry soil that has passed a 2-mm round
hole sieve ( 10 gm of soil for each 50 ml of water extract if the gypsum con-
tent is not over 1.3 per cent) is placed in a 250-ml extraction flask and
distilled water in sufficient volume is added to dissolve the gypsum present.
Fifty ml of water will dissolve approximately 0.1 gm of CaS0 4 or 0.13 gm
of CaS0 1 • 2 H~O or approximately 1.5 meq of the salt. If the gypsum
content is found to approach 1.3 per cent or 15 meq per I 00 gm of soil
(in a I 0-gm per 50-ml extraction), the determination should be repeated
with a more dilute extract. Air-dry soil is used rather than oven-dry soil
because oven drying converts the gypsum to CaS0 4 • 0.5 H 2 0, which
has a higher solubility in water for an indefinite period following solution.Ila
10-118. The bottle is stoppered and shaken by hand 6 times at 15-
minute intervals or agitated for 30 minutes in a mechanical shaker. The sus-
pension is filtered through a paper of medium porosity and a 20-ml aliquot
of the filtered extract containing 0.1 to 0.6 meq of CaS0 4 is placed in a
50-ml conical centrifuge tube. To the tube is then added 20 ml of acetone,
the contents of the tube are mixed, and the suspension is allowed to stand
until the precipitate flocculates, usually 5 to 10 minutes. The suspension is
clarified by centrifugation at 1000 times gravity (2000 rpm with a 24-cm
radius) for 3 minutes. The supernatant liquid is decanted away and the tube
is inverted to drain on filter paper for 5 minutes. The precipitate is then dis-
persed in the fresh 10-ml portion of acetone delivered from a pipet so as
to wash down the walls of the tube. The centrifugation, decantation, and
drainage on a filter paper is repeated as before. Finally exactly 40 ml of
distilled water is added to the tube, which is stoppered and shaken until
the precipitate is completely dissolved. Electrical conductance of the solu-
tion is measured by the usual procedure (~ 10-33). The conductance is
corrected to 25 °C (it increases 2 per cent per degree centigrade).
112 Essentials of the procedure are from Bower and Huss, Soil Sci .. 66: 199 ( 1948).
Ila Reitemeier and Ayers,!. Am. Chem. Soc., 69:2759 ( 1947).
 ERS
268 SOLUBLE SALT ANALYSIS FOR SOILS AND WAT
10-11 9. The gypsum content of the solution is found
by reference to a
Huss from the Interna-
graph of the following data given by Bower and
tional Critical Tables.114

CaS0 4 concen tration Electrical conduc tance at 25°C

meq/li ter millim hos/cm
1
0.121
2
0.226
0.500
5
0.900
to
1.584
20
2.205
30.5

n:
A close approximation of the graph is provided by the relatio
(10-2 9)
meq of CaS0 4 per liter= Lmmho ! .. m X 12.5
as in equation I 0-10.
by reference to
10-12 0. The meq of CaS0 4 • 2 H 2 0 in the soil is found
tage of gypsum
the soil : water ratio employed in the extraction. The percen
t:
in the soil may be calculated from the meq of gypsum presen
% CaSO • 2 H 2 0 in soil = meq per 100 gm of soil x
0.0861 ( 10-30 )
4

SOILS
CARB ONA TE CARB ON DETE RMIN ATIO N FOR
("Inorg anic" carbon of soils)
organic carbon
10-12 1. The carbonate carbon of soils, as opposed to
e earth comp ounds such
(1[ 9-1), occurs as various sparingly soluble alkalin
pedog enesis ) and
as CaC011 (calcite, the chief carbonate resulting from
t mater ials. Rare
dolomite, CaC0 3 • MgCOa, which occurs in some paren
Remo val of car-
occurrence of pedogenic dolomite has been reported.
115

the determ inatio n

bonate carbon has been mentioned in connection with
nates are some-
of organic carbon ( ~ 9-14) . Since the alkaline earth carbo
with the determ inatio n of excha ngeable cations
what soluble, they interfere
(~) 5-2) and also affect the composition
of the extracted soil solution (1[
ined by titration
10-18 ). Soluble carbonates and bicarbonates are determ
occur s as a part of the soil atmos-
in solution ( ~ 10-87 ). Gaseous C0 2
in terms of percen tage of mineral
phere, but the amount is inappreciable
high as 20 per cent of the
carbonate in soils even if the content is as
soil gases.
as for limestone,
10-12 2. Carbonate carbon can be estimated for soils,
n is quantitatively
from neutralizing equivalence (~ 4-66) . Carbonate carbo
sampl e with acid.
determined by evolution of C0 2 by treatment of the soil
Vol. I, pp. 231, 236.
04
an and Thiel, Bui. Geol.
Sherm an, M. S. Thesis, Univ. of Minn. (1937) ; Sherm
9G
Soc. Am., 50: 1535 (1939) .
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 269

The CO~ evolved is collected in standard hydroxide solution followed by

back titration of the excess hydroxide. The procedure is similar lo that for
organic carbon ( ~ 9-21), but involves essential differences. The C0 2
evolved may also be determined gasiometrically. The different mineral
species of sparingly soluble carbonates are sensitively determined by
X-ray diffraction analysis and polarizing microscope. Allocations between
Ca and Mg forms can also be made by elemental analysis ( ~ 4-7 4).

APPARATUS

10-123. Needed apparatus is the same as for organic carbon by wet

oxidation and determination as co~ (~ 9-24).

REAGENTS

10-124. Needed reagents are alcoholic phenolphthalein indicator, 0.04

per cent brom thymol blue indicator (I ml of this to I 0 ml of water in the
trap, Fig. 9-3), CO:!-free water, approximately 0.5 N NaOH ( 20 gm per
liter), approximately I M BaCI" (244 gm of BaCl:! · 2 H 2 0 per liter of C0 2 -
free water and adjusted to neutrality) and approximately 1 N HCI con-
taining 5 gm of SnCI:! · 2 Hp per I 00 ml.

PROCEDURE

10-125. The soil sample is first passed through a 0.2-mm screen. A

sample containing considerably less than 0.625 gm of CaCOa equivalent
( 10 gm of soil containing 5 per cent CaC0:1 or less or 0.25 gm of lime-
stone is taken) is weighed out and transferred to the sample flask. The
sample flask is then attached to the apparatus. To free the train of CO:!, a
flow of CO:!-free air is sent through for I or 2 minutes longer than required
for the brom thymol blue in the trap to turn from yellow to blue. Then 25
ml of 0.5 N NaOH is pipetted into the CO:! absorption flask followed by
50 ml of CO:!-free water. The condenser cooling water is started, while the
flow of CO:!-free air is continued. The bead tower is lowered to dip into the
NaOH until the solution reaches the top of the beads and then the bead
tower is raised so that no more NaOH solution enters. The air flow is
regulated to about 3 to 5 bubbles per second. Next, 50 ml of 1 N HCl
(with 5 per cent SnCJ/H) is added through the separatory funnel, slowly
at first if the sample is high in carbonate. A small flame with a wind shield
is placed under the sample flask and the suspension is slowly brought to
boiling. The boiling is continued for 5 minutes longer than required for
the brom thymol blue in the trap to change from yellow to blue.
10-126. The C0 2 evolved is collected in the absorption flask (Fig. 9-3)

or. Inclusion of SnCl2 is after Methods of Analysis, 7th ed. (Washington, D.C.:
A.O.A.C., 1950), p. 30.
270 SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS
and determined by back titration with standard HCl after addition of
BaCl2 ( ~ 9-30) . The carbonate carbon is calculated as follows:
meq of C0 2 = (S - T) x N (10-31)
in which S is the standardization blank titration, T is the back titration
and N refers to the normality of the standard HCl. Also:
5 (10-32)
% CaCO.,, = meq CO.,- x s

in which sis the sample weight in gm, and the factor 5 is derived from the
meq weight of CaCO:i (I 00/2000) x 100.

ALTERNATIVE PROCEDURES

10-127. The C0 2 may be passed through a Mg(Cl0 4 ) 2 bulb to remove

water vapor and then be collected in a gravimetric absorption bulb and
weighed instead of being titrated. In the absence of MgC0:1 (as magnesite
or dolomite), the C0 2 can be evolved without heating ( ~ 4-77) and de-
termined gasiometrically.

QUESTIONS
1. List the principal ions that are likely to be present as soluble salts in soils.
2. Describe the procedure for direct qualitative determination of soil salin-
ity conductometrically on a soil paste. What difficulties are involved?
3. What is the reason for conversion of the conductance readings to specific
conductance?
4. Define: (a) specific conductance, (b) cell constant. (c) salt bridge,
( d) millimhos/ cm.
5. Why must tables generally be employed to convert resistance readings to
salt content, while a proportionality factor can be used for conversion of con-
ductance readings?
6. By what mechanism is the electric current carried through the solution
portion of conductance circuit?
7. State the approximate relation between the magnitude of the wilting per-
centage, field capacity percentage, and saturation percentage of soil moisture.
8. What is the main influence of an excess of soluble salts in soils on growth
of plants?
9. Why is the specific conductance of the soil saturation extract a more
valid measurement of soil salinity than the specific conductance of an extract
made with a constant soil : water ratio?
10. State the upper limit of the specific conductance (mmho/cm) of the
soil saturation extract:
(a) Above which even salt tolerant crops do not thrive.
(b) Above which only a few highly salt tolerant plants survive.
(c) Above which crops tolerant of salts thrive but those nontolerant do not.
SOLUBLE SALT ANALYSIS FOR SOILS AND WATERS 271
11. What is the effect of making the extraction with more and more dilute
soil : water ratios on salt concentration calculated to the soil basis? Consider the
several possible cases.
12. List the numerical factors for conversion of Lrnmho/ <'m to:
(a) meq of salt per liter
( b) equivalents per million
( c) ppm in solution
( d) osmotic pressure of solution
( e) per cent in soil
13. What evidence is there that the soil solution may be displaced from a soil
by means of a solution infiltrated through a column of soil?
14. Explain the reason for sharp differentiation between bare spots and tall
growth of a crop in a field affected by salinity.
15. Define the "enrichment ratio" with reference to runoff waters.
16. State the principle involved in the analytical determination in soluble salt
solutions from soils of (a) bicarbonate, (b) carbonate, (c) sulfate, and (d)
chloride.
 11
Elemental Analysis of Mineral
(~olloids, Soils, Minerals,
and Rocks
The elements and their combination

11-1. Elemental analysis of soils and rocks is the determination of the

total amount of a mineral element present in the sample. Historically ele-
mental analysis was developed for the analysis of rocks and minerals. Later
it was used to evaluate soil fertility directly from the quantities of the ele-
ments present, but was supplanted largely by the concept of "available"
nutrients--ex changeable, extractable, and equilibrium-released ions, except
for the minor elements Cu, Zn, and Mo (Chapter 15). Elemental analysis
of mineral colloids now is employed mainly as a means to study their crystal
chemistry and their allocation to the mineral species present. Such applica-
tion of soil chemical analysis is of great importance to the fundamental in-
terpretations of the chemical processes of soil development and as a back-
ground to soil fertility interpretations.
11-2. The elements of soils and rocks are usually combined as constitu-
ents of 1 or more minerals. Because ionic oxygen is the major anion in
silicates, the elemental constituents are generally reported as the percent-
ages of oxides, so that the total approaches I 00 per cent of the ignited
sample weight ( ~ 11-3). There is also a logical basis for reporting the
analyses as the percentage of the element. The equivalent percentages of
0, F, and OH can readily be included so that the analyses tend to total
100 per cent as before. The basis for reporting has no bearing on either the
choice or execution of the procedures.
272
 MINER AL COLLO IDS, SOILS, MINER ALS, AND ROCKS 273
11-3. Ignited Weight Basis. The ignited weight basis is in some respects
the most satisfactory for reporting the results of elemental analysis because
the organic matter and combined water are not determined as an integral
part of the elemental analysis system. Thus most of the constituents are
determined in or calculated to the ignited state, and the determinations thus
add up to 100 per cent within the accuracy and completeness of the ele-
mental determinations.
11-4. Oven-Dry Weight Basis. Strictly speaking, the H 2 0, -OH and
other constituents are integral parts of the sample and should be determined
as part of the elemental analysis. The weight loss at 100°C is usually re-
ported as adsorbed water ( ~ 11-5) or H 2 0 ( - ) . The loss between 100°C
and 800° to 1000°C is reported as organic matter and H!!O( +). Fre-
quently the losses are calculated on the 100°C weight basis, adsorbed H 0
2
thus being excluded from the 100 per cent total. Fluoride, sulfide, MgCl!!,
and other salts, if present in the sample, constitute a portion of the volatile
matter losses at 800° to 1000°C. Except for losses of such constituents,
interconversion is readily made between the 100° and the ignited basis of
reporting elemental analyses.
11-5. The analytical sample, unless derived from suspension(~ 11-12),
is often weighed out in the air-dry condition. A separate sample is taken
for the determination of adsorbed water. There is no precise oven tempera-
ture or period of drying at which all of the hygroscopic water and none of
the water of constitution have been driven off. The common procedure for
bulk materials consists of drying a separate 3-gm sample in a dish for
which a tightly fitting cover is available, at 100° to 1 I 0°C for I 0 to 16
hours for removal of hygroscopic moisture; then cooling it in a desiccator
for 0.5 hour and weighing. The weight after such drying is the oven-dry
basis for the analysis, the basis employed for most common soil analyses.
11-6. Methods of Bringing the Sample into Solution. The usual 2 pro-
cedures for bringing the elements of sample into solution are (a) fusion
in Na 2 CO:i, or (b) some system of heating in acids such as with HF, HCI0 ,
4
H 3 P0 4 , and H 2 S04 • These 2 procedures are used in the analytical systems
detailed in this chapter. A few minerals such as chromite and zircon do not
become soluble through these treatments, and another type of fusion, such
as in Na 2 0 2 or K 2 S2 0 7 (~ 11-198 ) is necessary.
11-7. Systems of Elemental Analysis. Systems of elemental analysis may
be divided into 4 categories:
1. Rapid micro and semimicro absorption spectrophotometric ( colori-
metric) systems
2. Micro and semimicro emission spectrophotometric (flame, arc,
spark) systems
3. Semimicro titrimetric systems
4. Macro gravimetric systems
 S
274 MINE RAL COLLOIDS, SOILS, MINERALS, AND ROCK
of analysis
11-8. The 3 main factors that dictate the choice of method
tional gravi-
are accuracy, speed, and simplicity. For macro samples, conven
consid ered to be more accurate 2
metric and titrimetric systems are usually
1
For semimicro
than the semimicrochemical spectrophotometric methods.
te and have the ad-
samples, spectrophotometric metho ds are more accura
ophotometry
vantages of greater simplicity and rapidity. Absorption spectr
ty of anal-
(Chap ter 17) has progressed sufficiently far to permit the majori
(Chap ter
yses to be made by that means.x Emission spectrophotometry
most accura te for the alkali
18) with flame excitation is the easiest and
earth metals . Analys es
metals and is usually satisfactory for the alkaline
al proper ties
of difficultly separable elements having nearly the same chemic
ter 18). Micro
requires the specificity of emission spectrophotometry (Chap
ed to correlate
samples, such as individual sand grains that are to be analyz
proper ties, or soil specim ens to be
with their respective individual optical
the sensiti vity of the spectro-
employed for criminal case work also require
graph ( ~l 18-51 ) .
semimicro-
11-9. Because of the simplicity and great saving in time, a
tal analysis
chemical spectrophotometric and titrimetric system of elemen
Na. A com-
is given first for the elements Si, Al, Fe, Ti, Ca, Mg, K, and
of system s conven tionall y employed
bination of several of the general types
ntiona l) system , the more
is given as a second system. In the second (conve
gravim ctrically,
abund ant constituents, such as Si and Al, are determined
as Fe (in-
whereas the constituents present in intermediate amounts, such
ined titrimetri-
cluding ferrous iron separa tely), Ca, Mg, and K, are determ
uents, such as Ti, Mn, and P, are de-
cally, and the least abund ant constit
determ ination s of ferrous iron, S,
termined colorimetrically. The separate
F, Zr, and Li are considered near the end of this chapte r.

tal analysis are avail-

I Compre hensive treatises of the classical method s of elemen
, Applied Inorgan ic Analysi s (with special referen ce to
able; Hillebr and and Lundell
rocks) (New York: John Wiley & Sons, Inc.,
the analysis of metals, mineral s, and
. Silicate Analvsi s, 2nd
1929); Hillebr and, U.S. Geo!. Surv. Bui. 700 (1919) ; Groves
Chemis che Analyse Der
ed. (Londo n: George Allan & Unwin Ltd., 1951); Jakob, Birkhau ser, 1952);
lien (Basel. Switzer land: Verlag
Gestein e Und Silikati schen Minera
3rd ed. (New York: John
Washington, Manua l of the Chemic al Analysi s of Rocks,
Inorganic Analysi s with
Wiley & Sons, Inc., 1919); Mellor, A Treatise on Quantit ative
Analysi s of Clays, Silicate s and Related Minera ls (Londo n:
Special Referen ce tu the
Charles Griffin & Compa ny, Ltd., 1913).
~ Even with the greates t care, the accurac y
of the conven tional method s is not as
high as often is suppose d, as indicate d by compar ative studies by Schlecht, Anal.
Chem., 23: 1568 (1951) ; Fairbai rn, U.S. Geo!. Surv .. Cir. 980 ( 1951); Fairbai rn and
Schaire r, Am. Mineral, 37:744 (1952) .
3 Colorim etric systems have been present
ed by Guthrie and Miller, MineraloRical
Mag., 23:405 (1933) ; Hecht, Mikroc him. Acta, 2: 188 ( 1937), abstr. Analyst , 63:209 ;
Materials, Proc. Swed.
Hedin, Colorim etric Method s For Rapid Analysi s of Silicate
Concre te Res. Inst., Stockho lm (1947) ; Shapiro and Branno ck, U.S.
Cemen t and
Corey and Jackson , Anal. Chem., 25: 624 (1953).
Geo!. Surv. Cir. 165 ( 1952);
 MINERA L COLLOIDS, SOILS, MINERAL S, AND ROCKS 275
11-10. Methods of Checking Accuracy. Methods of checking the accu-
racy of silicate analysis consist of 1 or more of the following:
1. Analysis of pure chemicals (not satisfactory for the final check)
2. Addition of standard amounts to samples (test of increase in the
recovery)
3. Analysis of standard samples, such as those provided by the National
Bureau of Standards, and calculation of the standard deviation.

11-11. Preparation of the Soil or Rock Sample. A bulk soil or rock

sample is sieved (,! 2-58), mixed, air-dried, and quartered to a minimum
of 1 kgm ( ~ 2-63). In case of a rock or mineral, a hard steel mortar is
used for coarse crushing. The - 2 mm soil ( ~ 2-60) is ground in an agate
mortar 4 ( ,, 2-62) and partitioned ( ~ 2-66) until the final 5-gm portion is
passed in its entirety through a 150-u ( 100 meshes per inch) bolting cloth.
11-12. Preparation of the Mineral Colloid Sample. Minerals frequently
require elaborate individual separations and purification or preconcentra-
tion in preparatio n for elemental analysis. Usually, from the mineral colloid
suspension an aliquot is taken that contains the correct sample weight
(known from drying a separate aliquot). The particles are flocculated with a
minimum of HCl and thrown down by centrifugation. The sample is then
washed by means of the centrifuge procedure (,I 4-3) to remove all solu-
ble electrolytes and exchangeable metallic cations, five times with 0.05 N
HCl and twice each with 70 per cent ethanol, absolute methanol, and finally
benzene (to remove oil originating from the supercentrifuge). The suspen-
sion is either transferred with methanol and water to the platinum crucible
and dried for analysis, or (if prepared as a bulk sample) is transferred to a
porcelain dish with benzene and dried.
11-13. Care and Use of Platinum Utensils." Platinum utensils are virtu-
ally indispensable for accurate elemental analyses, but must be used with
care because of their expensiveness. Platinum costs upwards of $3 per gm.
An ordinary fusion crucible costs $100 or more. Platinum metal is re-
markably resistant to a variety of chemicals although it is soluble to an
appreciable extent ( 0.5 to 1 mgm per Na 2 COa fusion, for example) in the
majority of the reagents that may "properly" be used in platinum utensils.
Platinum has a melting point of about l 770°C but is appreciably volatile
at I000°C and above (about 0.1 mgm/cm 2 /hr at 1000°C and 1 mgm/-
cm~ /hr at 1200°C). Corrections for volatility are made by heating the
empty crucible for a similar time and weighing. Platinum density is 21.3 7.

4 Robinson, Soil Sci., 59:7 (1945).

G For additional information on the properties and care of platinum utensils the
reader is referred to Kolthoff and Sandell, Textbook of Quantitative Inorganic
Analysis (New York: The Macmillan Company, Inc., 1945), p. 179; and Hillebrand
and Lundell, Applied Inorganic Analysis (New York: John Wiley & Sons, Inc., 1929),
pp. 18-20, 272-273, 295.
 S
276 MINE RAL COLL OIDS, SOILS, MINE RALS , AND ROCK
usually with a
Being too soft when pure, platinum is alloyed for stiffening,
utensils should
fraction of a per cent of iridium. The useful life of platinum
ng specific
be guarded by (a) keeping them chemically clean, (b) avoidi
ng them so
reagents in which platinum is soluble or reactive, and ( c) handli
as to keep them in good mechanical condition.
ctory life,
11-14 . Cleaning and Burnishing Platinum Utensils. For satisfa
and in proper mecha nical condit ion. Clean-
platinum utensils are kept clean
ents as necess ary to make the utensil
ing should .include chemical treatm
and burnis hing, and finally washin g
appear clean, mechanical straightening
d singly. The initial chemic al
in 6 N HCI. The crucibles should be cleane
cleaning procedure consists of:
I. Soakin g in hot water and scrubb ing; if not clean, then
the crucibl e
2. Boiling in 6 N HCl for a few minutes. If still not clean,
is dried, and
poured into
3. Fused with potassium pyrosu lfate, K~S 2 0i, the melt being
waste sand. and the residue dissolv ed from the crucibl e in warm 6 N
dry
case except for
HCI. If certain types of impuri ties persist, not usually the
that have been improp erly used, the crucibl es are further cleaned
crucibles
by
a little HF to
4. Fusion with Na~COa. or, in some cases, digestion with
which 3 drops of H~S0 4 have been added.
form or a
11-15 . The cleaned crucible is straightened with a wood
mm rounde d sea
smooth maple rod. It is then burnished with 0.1 to 0.5
are never used
sand held on a soft cloth on the finger. (Grou nd powders
smooth s the
because of angularity and resulting abrasiveness.) Burnishing
The scale and
surface and retards the development of scale and cracks.
be remov ed promp tly becaus e once started, they
microscopic cracks should
not continued
develop rapidly with further use of the utensil. Burnishing is
e there is always some abrasio n of the plat-
longer than necessary becaus
crucib le is warme d for a few minutes in
inum. Following burnishing, the
and dried in an oven, being grasped
6 N HCI, rinsed with distilled water
only with platinum tipped or pure nickel tongs.
Soluble or
11-16 . Avoidance of Specific Reagents in Which Platinum is
ally, platinu m is readily at-
Reactive. Although remarkably inert chemic
reactio ns with impro per
tacked, and may easily be alloyed and cracked by
ions are as
reagents added or products formed. Salient precautionary condit
follows:
not be di-
1. Chlorin e attacks platinu m. Thus platinu m utensils must
chlorin e is liberate d. Similar ly, manga nate
gested in aqua regia, from which
from HCI and, therefo re, the manga nate formed in the
liberate s Cl 2
additio n of ethano l or mechan ically
Na 2 C0 3 fusion should be reduce d by halo-
the platinu m before the additio n of HCI. Two other
separa ted from a-
platinu m at ordina ry temper
gens, free bromin e and iodine, also attack
ts should not be mixed with nitrate s in acid
tures. Salts of these elemen
solutio ns in platinu m.
 MINERA L COLLOIDS, SOILS, MINERAL S, AND ROCKS 277
2. Ferric chloride in the presence of HCI attacks platinum, with the
partial reduction of the iron to ferrous and the simultaneous appearance of
platinum in solution. Such solutions should not be evaporated in platinum.
3. Hydroxides, oxides, peroxides, nitrites, and cyanides of the alkalies
strongly attack platinum. Use of crucibles of other metals is necessary for
fusions involving these fluxes. BaO also attacks platinum.
4. Reducing conditions, found in a luminous flame involving incomplete
combustion or produced by contact with carbonaceo us material in the ab-
sence of a good circulation of oxygen, are conducive to formation of
platinum carbide. Silica and borate react with platinum under reducing
conditions. However, silica and precipitated iron held in filter paper may
be heated with a low, nonlumino us flame with an abundant circulation of
air without damage to the crucible if the paper and contents, placed in the
crucible, are first oven-dried.
5. Reducing conditions may result in alloying the platinum with any
easily reducible metal that is present. To be excluded from platinum are
compounds of the following easily reducible elements, as well as the ele-
ments themselves: Ag, Pb, Hg, Bi, Sn, Sb, Se, and As. Free phosphorus ,
sulfur, and sulfides combine with platinum, and should, therefore, be ex-
cluded.
11-17. Precautions in Handling Platinum Utensils. Platinum utensils are
easily damaged mechanically unless proper care is exercised. Most im-
portant precautions are as follows:
l. The utensils are handled in such a manner as to prevent deformatio n.
Storing in cotton packing avoids denting by collision.
2. Triangles of platinum, silica-tubed wire, or pipe clay should be em-
ployed. Triangles of nichrome arc fairly satisfactory for crucibles if kept
scrupulously clean and dry, but are not satisfactory under large platinum
dishes. Nichrome triangles have frequently been employed with a false
sense of security. Iron triangles are never employed.
3. Platinum-ti pped or pure nickel tongs are employed (never brass,
nickel-plated, or iron tongs).
4. Before the burner is lighted, the utensils are put in place, and the
mounting is checked for position with respect to the burner. Then the
burner should be directed away, lighted, the flame adjusted to nonluminous,
and the burner position fixed so that the platinum is well out of the green
cone, which contains incompletely burned (reducing) gases.
5. Before the crucible is grasped after heating, it is allowed to cool
momentarily. Stiffening of the platinum occurs as red heat is lost.
6. A porcelain plate, beaker, or asbestos pad is employed to set the
platinum utensil on, never a desk top or ring stand.
11-18. As a matter of habit, the exterior of the platinum utensils should
be kept chemically clean, because it frequently must be extracted by im-
mersion in the solution, and because contaminants on the exterior may
cause injury to the utensil. A good analyst always has his platinum well-
shaped, bright, clean, and protected from damage and contamination. So
used, the utensil will last for decades.
278 MINERA L COLLOI DS, SOILS, MINERA LS, AND ROCKS

SEMIMI CROCH EMICAL SYSTEM OF SILICA TE ANALYSIS

11-19. Semimicro (0.1-gm) samples of mineral colloid, soil, mineral, or
rock are analyzed for their content of the elements Si, Al, Fe, Ti, Ca, Mg,
K and Na by rapid spectrophotometric techniques ( ~ 11-7). Spectro-
photometric determinations of Mn and of P (Chapter 7) can also be made
with slight adaptations of the procedures. The principal problems in a rapid
semimicrochemical system of analysis center around the chemical separa-
tions and flow sheet ( ~ 11-22), rather than the spectrophotometric de-
terminations of the individual constituents. This is true because procedures
for the individual elements have become available in the field of general
analytical chemistry. The separations of the constituents and the prepara-
tion of solutions free from interfering substances is the special contribution
of the method. 11 The chemical determinations of the elements have been
refined and adapted to fit the conditions resulting from the method employed
for bringing the sample into solution and making the separations.
11-20. Extension of the Semimicrochemical System. The same elements
from other sources, such as in solutions derived from differential solubility
studies, may also be determined by the same procedures. Each element must
be obtained in the proper concentration range in a solution free from inter-
fering substances, and the salt concentration and pH values of the solution
must be adjusted to approximately the same as in the procedure given in
the semimicrochemical system.
11-21. Accuracy. The accuracy of the individual determinations in the
semimicrochemical system is on the order of 1 to 3 per cent of the total
amount of the constituent present. This accuracy of individual constituents
permits an over-all accuracy of ±2 per cent, which is satisfactory for a
majority of interpretations in which elemental analyses are applied. Al-
though the precision is less than is ordinarily considered attainable with the
conventional longer methods, the conventional methods themselves, even
with the greatest care, are much less accurate than is sometimes supposed
(~ 11-8).
11-22. Flow Sheets. In general plan or flow sheet, the analytical system
makes use of 2 0.1-gm samples, 1 of which is decomposed by heating with
HF and HC10 4 (Fig. 11-1) and the other by fusion with Na 2 COa (Fig.
11-2). These reagents release the elements concerned from nearly all soil
minerals ( ~ 11-6). Two samples are necessary because it would be ex-
tremely difficult to determine Si and the alkali metals in the same sample.
Separations are speeded by the use of volumetric pointed centrifuge tubes
and centrifugation. The point in the flow sheet at which each element is

6Corey and Jackson, Anal. Chem., 25:624 (1953); special thanks are due to
Ors. L. D. Whittig and L. D. Swindale for assistance with several of the improvements
included.
 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 279

0.100 gm sample

'----- ----- -1 Evaporate

HF.HCI04
to dryness
Salts of K. Na, Fe, Ti, Al, Ca, Mg, Mn

HCI. boil
Dilute to 100 ml

---- ---- ---- -,

~-~- r---L- -,
I 20 ml I
pH 4. 7 buffer
NH,Cl, NH40H
L--i--.. .J
Tiron
·sr2 water
Dilute to 60 ml I
Centrifuge I
Read@ 565 mu
Solution B
I
25 ml aliquot
Na:i-"204 HN03 I
Read @4oomu Evaporate to dryness I
CaCl2, MgCI 2
I
I
25 ml
0.4~
I150
HCI ml aliquot
,----- ------ -- --,
IAl(OH)3, Fe(OH)3,FeP 04,Ti(OH)4. Mn02I Read@@ @ IHN03
L - - - - - - -r.-- - -- ----' I Evaporate to
, ______ _rHCI
____ , and ~ on flame dryness
AJ3"': Fe 3"'; Ti 4+, photometer IHCI
I I Mn 2+
L------T N;oH__ ...J IKOH
r....J.---- --....,
- - __ j::·~e~..:.? ~ - --, 1 pH 12 solution
L.,.--- ---J
1

rJ£e!:!"~!!! __ ----, r-l--,

I Murexide
I
I Fe(OH)3, Ti(OH)4, Mn02
1 _____ _____ L-r.-J
L .JI I AI02- I Titrate @ with
Versene
HCI I
IHN03 andH202 r--1---- ---...., I Bromine water
I to dryness
.l~~~~':'.:...,
I Dilute to 100 ml I
I L....,-------~
H3P04 1 Aliquot of above
II Penodate
. pH 10 solution
LT _____ ...JI
1
II pH 4.2 buffer I Eriochrome
~

J_ Read K
r - ------, r
..1 Aluminon
--------, I black T and N
I Dilute to 25 ml I I Dilute to 50 ml I
LT------~ L,------~ I (.;). on flame

Read 9 540 mu .Read @ 520 mu

Titrate ~ with
Versene
photometer

Fig. 11-1. Flow sheet for sample decomposed with HF in semimicrochemical

system.

determined is indicated in a circle. Since 2 samples are employed, some of

the elements may be determined in either sample. The procedures that are
generally employed are shown in solid lines. Alternative procedures, which
are usually employed when only I sample is analyzed (for fewer than 8
elements ), or when the flame emission spectrophotometer is not available
for Ca and/or Mg, are indicated by broken lines on the flow sheets.
280 MINE RAL COLLOIDS, SOILS, MINERALS, AND ROCKS

.0.100 gm sample
Na2C03 fusion, HCI04
dehydrate to fumes
Dilute to 60 ml
Centrifuge

NH40H, Br2 water

Dilute to 60 ml

Centrifuge
---- ---- ..,
I
Fe(OH)3, FeP04, Ti(OH)4,A l(OH)3, Mn02
HCI

Read@ 400 mu Fe 3+, Ti 4 +, AJ 8 +, Mn2 +

Centrifuge

---lHCI
r--L- ----- -,I
to 100 ml_
r ____
._ __Dilute 1 _J
Aliquot of above
pH 4.2 buffer
Aluminon
r---- ----- -L-1
I Aliquot of above I
I Evaporate to dryness I Read@ 520 mu

----~ I
I
I
I Aliquot of above
f pH 4. 7 buffer
rlT~o.'.'._ _____ ,
HN03 + H202
to dryness 1 Dilute to 50 ml I
'-y--- ---...J
H 3P0 4 Read @ 565 mu
Per iodate rl--- ----- ,
Dilute to 25 ml LT _______
1 Same solution 1
_J

Read S 540 mu I Na2S204

Read@ 400 mu

semimicrochemical
Fig. 11-2. Flow sheet for sample decomposed with Na2C0:1 in
system.

d in
11-23. The products obtained by the various procedures are enclose
listed be-
rectangular boxes. Reagents added and operations performed are
double
side the lines joining the boxes. The precipitates, indicated by
their boxes, are written to the left and the solutio ns to
vertical lines above
con-
the right. The method of separation is shown under the horizontal line
the flow sheet. The wave lengths refer to the
necting each box to the rest of
light maxima employed for the respect ive determ ination s.
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 281
ll-24. The sample decomposed by HF is used for the determination of
K, Na, Ca, and Mg by the flame emission method. It is also· used to de-
termine Fe and Ti. Also, Ca and Mg can be determined by Versene if de-
sired. Likewise it is possible to determine Al with this solution. The Na 2 C0:1
fused sample is used for the determination of Si and Al. Also, the total of
Mg plus Ca plus Mn can be determined by Versene if desired. It is also
possible to determine Fe and Ti on this sample fused in Na 2 CO:i. Analysis
of 8 materials ( 4 in duplicate) for 8 elements can be accomplished in a
few days, or, under ideal conditions, in 1 day.
11-25. The order of procedures adopted for greatest efficacy is as
follows:
I. Decomposition of sample with HF ( ~i 11-31), resulting in Solution
A (Fig. 11-1)
2. Fusion of second sample in Na2 COa (~I 11-35), resulting in Solutions
C and D (Fig. 11-2)
3. Determination of K, Na, Ca, and Mg by flame emission, on Solution
B (~: 11-37). [Alternatively, Kand Na may be determined on Solution A
( ~ 11-45) and Ca by flame emission or Versene ( ~; 11-44) on Solution B;
in this event, the total of Ca plus Mg is determined with Versene on Solu-
tion B (~: 11-54) or Solution E (~! 11-60), derived by the NH 4 0H-Br
separation ( ~i 11-94) carried out on Solution C.]
4. Determination of Fe and Ti on Solution A ( ~ 11-62)
5. Determination of Si on Solution D after Solution C is removed follow-
ing HCI0 4 dehydration (~ 11-72)
6. Determination of Al on Solution F (Fig. 11-2) resulting from the
NH 10H-Br (~ 11-94) and NaOH (~ 11-96) separations on Solution C
11-26. Blanks. Blanks are carried throughout for control of impurities
of reagents and contamination from glassware.

APPARATUS FOR SEMIMICROCHEMICAL SYSTEM

11-27. For analysis of 8 materials for 8 elements and for 1 blank, the
following apparatus is required: 9 30-ml or larger platinum crucibles and
covers; 8 500-ml nickel or platinum beakers; 1 I 50-ml Pyrex beakers; 9
125-ml Pyrex conical flasks; 18 50-ml volumetric flasks, 1 set of 9 to be
used exclusively for the determination of Fe, Ti, and Al, and the other set
for Si; 17 I 00-ml volumetric flasks, 1 set of 9 to be used exclusively for
Solution A and the other 8 for Solution F; 8 500-ml volumetric flasks; l
50-ml Lowy pipet (with 3-way stopcock); pipets delivering 1, 2, 3, 4,
5, and 10 ml of solution; a suction-decantation apparatus (Fig. 11-3); a
photoelectric colorimeter with 400-, 520-, and 565-mu light maxima; 18
matched colorimeter tubes, 1 set of 9 tubes to be used exclusively for the
determination of Fe, Ti, and Al, and the other set for Si; an electric hot
plate, 200° to 225°C and sand bath; and a flame emission spectropho-
tometer.
 Rubber tubing

Glass
tubing

Fig. 11-3. Suction apparatus for removal of

supernatant liquid.

38mm
2mm

25mm

60ml

50

40 75 mm

~·~01 34mm

170mm
I I
l
5___..~29mm--J
~-

mm ,_1_,1_~~-1

}5 /
\\mm' 1
'J./ 34 mm
70mm
r~ l
!+-31 mm-J

Fig. 11-4. Special 60-ml pointed centrifuge tube

and special rubber cushion enclosed in brass
cylinder for support during centrifugation. (After
Corey and Jackson, Anal. Chem., 25:624, 1953.)

282
 MINERAL COLLOIDS , SOILS, MINERALS , AND ROCKS 283
11-28. An International size 2 centrifuge is employed, with 16 graduated
60-ml pointed centrifuge tubes (Fig. 11-4) and special cushiohs (obtained
from lnternation Equipment Co., Boston, Mass.). The precipitates are
stirred with an air-jet stirrer consisting of a 20-cm length of 3-mm outside
diameter glass tubing pulled out to a fine point on one end (filtered air is
forced through this tube to mix solutions in the pointed tubes and also to
dislodge precipitates from the bottom of the pointed tubes). A 2-liter
Pyrex beaker is used as a hot water bath for sets of 8 pointed centrifuge
tubes.
11-29. A crucible radiator is made from a 100-ml crucible of nickel,
iron, or porcelain over which a silica-covered nichrome triangle is fitted to
hold a platinum crucible.

REAGENTS

11-30. The acids and other reagents are analytical reagent grade. Re-
agents needed for the HF decomposition include 48 per cent HF, 60 per
cent HCI0 4 , and 6 N HCI; and for the Na 2 COx fusion, anhydrous Na 2 COa
powder. The reagents are listed separately for the determination of each
element.

PROCEDURE

11-31. Decomposition of Sample with HF. A 0.1000-gm, finely ground

sample that has been dried at I 00°C for 2 hours (or on which a separate
moisture sample has been run) is placed in a 30-ml platinum crucible. The
sample is wetted with a few drops of water, then 0.5 ml of HCl0 4 and 5 ml
of 48 per cent HF are added. 7 The crucible, with the lid covering nine-
tenths of the top, is placed on a sand bath at a temperature of 200° to
225°C, and the acids are evaporated to dryness. The solution must not
boil vigorously or spattering may occur. The HCl0 4 should drive off vir-
tually all of the F, because appreciable F interferes with the Fe determina-
tion. Organic matter, which sometimes condenses onto the cover and upper
sides of the crucible, is often not attacked completely by the HCI0 4 • If a
dark color is present on the cover and sides, indicating organic matter, a
Meker burner is directed on the sides of the crucible exterior just long
enough to dispel the organic matter. A faint red heat is usually all that is
necessary.
11-32. The crucible is removed from the sand bath and cooled, 5 ml of
6 N HCl is added, and the suspension is diluted to two-thirds of the vol-

7 Volatilization of Ti in the absence of H 2 S0 4 (~ 11-117) was shown to be almost

negligible by data of Hoffman given by Hillebrand and Lundell, Applied Inorganic
Analysis (New York: John Wiley & Sons, Inc., 1929), p. 724, and verified by Dr. L.
D. Whittig in this laboratory. Use of HzS04 is abandoned because of interference with
dissolution of high Ti samples and in the Na and K determination with the flame
photometer.
284 MINER AL COLLO IDS, SOILS, MINER ALS, AND ROCKS
ume of the crucible with water. The crucible is then covered and placed in
the radiator described with the apparat us ( ~ 11-29). The radiator crucible
is then heated with a relatively low flame so that the solution boils gently
without bumping. After 5 minutes of boiling, the residue should be com-
pletely dissolved 8 (~I 11-34).
11-33. If the sample is high in Al 2 0:1 or coarse-grained particles, diffi-
culty is sometimes encountered in dissolution of the residue. If no addi-
tional sample is available, the solution is evapora ted to dryness and the
HF-HC l0 4 treatme nt is repeated (~ 11-31). The residue is once more
treated with 6 N HCl, water, and boiling. This process is repeated until all
the residue is dissolved. With some unusual types of sample, a different
type of dissolution must be considered (~I 11-6).
11-34. When the residue is completely dissolved, the solution in the
crucible is transferred to a 100-ml volumetric flask, cooled, and diluted to
a volume of 100 ml. (In high Ti samples, some opalescence tends to de-
velop by Ti precipitation, but when the solution in the volumetric flask is
heated for several minutes in a hot water bath, the Ti goes into solution.)
This is Solution A, used for the determination of K, Na, Ca, and Mg by
flame emission (~ 11-37) and Fe and Ti (~ 11-62). The K and Na de-
terminations are run as soon as possible to minimize contamination from
the glassware. If necessary, this solution may also be used for the deter-
mination of Ca(~ 11-44) and Ca plus Mg with Yersene (~ 11-54). Also,
Al may be determined, but the F must have been completely removed by
fuming inHCl0 4 (~ ll-31)to preven titsinte rferenc e (~ 11-86).
11-35. Decomposition of Sample by Na 2C0:1 Fusion. A 0.1000-gm,
finely ground sample that has been dried at I 00°C for 2 hours (or for
which a separate moisture sample has been run) is weighed in a 30-ml
platinum crucible. Then approximately 0.75 gm of Na 2 CO:i is added, and
the sample and the carbona te are thoroughly mixed by rotation of the
crucible with the fingers. Approximately 0.25 gm of Na 2 CO:i is then added
on top of the mixture. The crucible is placed in a slanting position on a
silica-covered triangle with the lid covering about seven-tenths of the top.
The low flame of a Meker burner is placed so as to heat one side of the
crucible, and the heat 9 is gradually increased until the melt is liquefied. If
the material is finely ground, the fusion is usually complete in 1 to 2 min-
utes, at which time the flux appears quiet with small curds of precipitate
floating in the otherwise clear liquid, and the evolution of gas has ceased.
The crucible is then grasped with nickel tongs and swirled to spread the

8 Traces of organic matter not removed ( ~ 11-31),

appearing as dark specks in
the solution, sometimes are simply ignored.
11 Shell, Anal. Chem., 26: 591 ( 1954), recomme nded
fusion in the electric furnace
occurs if
at 1200°C to avoid transfer of iron from the sample to the Pt metal, which
very good oxidizing conditions are not maintained.
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 285
flux in a thin layer over the sides of the lower half of the crucible so that
any silica adhering to the sides is fused. Heating is continued under full
Meker flame for 2 or 3 minutes, after which the burner is removed and the
crucible is swirled to spread the melt in a thin, easily dissolved layer.
ll-36. The Na 2C03 fused sample is used for the determination of Si
( ~ 11-72) and Al (~I 11-86). Ca plus Mg may be determined with Ver-
sene ( ~ 1 1-54). Likewise, Fe and Ti may also be determined in this
sample (~I 11-62), but their determinations are faster and more accurate
in the HF decomposed sample.

POTASSIUM, SODIUM, CALCIUM, AND MAGNESIUM

DETERMINATIONS
(Rapid semimicrochemical system)
11-37. The flame emission spectrophotometric determinations of K, Na,
Ca, and Mg are far more rapid and better adapted to the sample size em-
ployed in this semimicrochemical system than are titrimetric, colorimetric,
or gravimetric determinations. If flame emission techniques are not avail-
able, suitable alternative procedures (~I 11-44) are adaptable to the semi-
microchemical system. The NH 4 0H separation is given to remove Fe, Al,
and P, which interfere with the flame emission determination of Ca and
Mg. If titrimetric procedures ( ~ 11-46) are to be carried out for Ca and
Mg, then the Br2 separation (~I 11-41) is included with the NH 4 0H sepa-
ration to remove Mn, while K and Na are determined on Solution A (Fig.
11-1 and ~I 11--45).

APPARATUS

11-38. Apparatus needed for the NH 4 0H or NH 4 0H-Br separation in-

cludes 60-ml pointed centrifuge tubes and centrifuge, an air-jet stirrer, a
hot water bath, a 125-ml conical flask, and a steam plate or electric hot
plate with sand bath. A Beckman or other suitable flame emission spec-
trophotometer is needed for the determination ( ~ 18-7).

REAGENTS

ll-39. Reagents needed for the NH 4 0H separation include 4 N NH 4 0H,

solid NH 4 Cl, and concentrated HNOa. For the NH 4 0H-Br separation,
saturated Br2 water is also needed. For the determination, 0.4 N HCI and
standard solutions of K, Na, Ca, and Mg in base electrolyte concentrations
comparable to those of the test sample ( ~ I 8-15) are needed.

PROCEDURE

11-40. NH 4 0H Separation. A 50-ml aliquot of Solution A (~ 11-34) is

placed in a 60-ml pointed centrifuge tube. Approximately 1 gm of NH 4 Cl
is dissolved in this solution by agitation with the air-jet stirrer, 3 drops of
286 MINERA L COLLOI DS, SOILS, MINERA LS, AND RUCK.::S
brom cresol purple indicator are added, the solution is warmed in a boiling
water bath, and 4 N NH.PH is dispensed from a buret until the full purple-
violet color is reached, and then 2 ml of additional NH 4 0H is added.
11-41. NH 4 0H-Br Separation. To cause the precipitation of the Mn as
Mn0 2 with the R 20a precipitate, 2 ml of saturated Br2 water is added to
oxidize the Mn. The Br2 also decolorizes the indicator. If both Ca and Mg
are to be determined by Harne emission instead of by Versene, this para-
graph may be omitted.
11-42. The tube is placed in a hot water bath for 5 minutes to flocculate
the R 2 0:i (and Mn0 2 ) precipitate. lt is then cooled, and the volume is ad-
justed to exactly 60 ml. The suspension is mixed thoroughly with the air-
jet stirrer, then centrifuged 5 minutes at 1800 rpm. The K, Na, Ca and Mg
are in the supernat ant liquid (Solution B).
11-43. A 25-ml aliquot of the supernat ant liquid is pipetted into a
125-ml conical flask. To destroy most of the ammonium salts, the solution
is evaporat ed to 10 ml, 10 ml of concentr ated HNO:i is added, and the
solution is evaporat ed to dryness on a steam plate or an electric plate with
sand bath. The residue is taken up in 25 ml of 0.4 N HCI, and the K, Na,
Ca, and Mg are determined by flame emission(~ 18-27). The Solution A
was diluted by a factor of 5/6, hence, for K, Na, Ca, or Mg:
0.120
--- ···--- ···---
% element = mgm of element per 100 ml x weight
<from curve) of sample, gm
(11-1)

ALTERNATIVE PROCEDURES

11-44. Calcium is most expeditiously determined by the Harne emission

spectrop hotomete r ( ~ 18-16) applied to Solution B obtained after the
NH 4 0H separatio n (~ 11-40) even when Mg cannot be because of equip-
ment limitations. In this case, Ca plus Mg is determined with Versene (~
11-59) followed by subtracti on of Ca determined by flame emission, to
obtain Mg by difference. The Mg may also be determined by 8-hydroxy
quinoline ( ~ 5-59). Also, Ca may be determined in Solution B by Versene
( ~ 11-46) or by the semimicro oxalate-cerate procedure (~I 5-3 8).
11-45. The K and Na are most expeditiously determin ed (~ 18-27) by
a flame emission spectroph otometer , even if Ca and/or Mg cannot be (~
11-44) because of equipme nt limitations. In the complete absence of flame
emission equipment, the K and Na must be determined by conventional
methods ( ~ 11-176) rather than by this semimicrochemical system. Ap-
proximately 20 ml of Solution A ( ~ 11-34) is transferr ed to a 50-ml beaker
or other suitable vessel and then K and Na are determined by flame emis-
sion ( ~ 18-27). Since Solution A is not diluted or otherwise changed, the
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 287
concentration reading from the flame emission curve gives the mgm of K
or Na in the entire 100 ml of Solution A:

% Kor Na = mgm of Kor Na per 100 ml x - -0.100

----
wt. sample, gm
(11-2)
11-46. Versene Method. The Versene (disodium dihydrogen ethylene-
diamine tetra-acetic acid) titration procedure of Schwarzenbach and Bieder-
rnann10 is employed alternatively for the Ca and Ca plus Mg determina-
tions because its speed and simplicity are next to those of the flame emission
method. Versene exhibits very strong complexing powers with various
polyvalent cations. The logarithms of the equilibrium constants of some of
the metal Versene complexes for the reaction M + + Ver 4 - ~ M-
11

Ver"14 are as follows: Fe+ -J +, 25. I; Cu++, 18.4; Ni++, 18.4; Pb++, 18.2;
Cd++, 16.5; Zn++, 16.2; Co++, 16. l; La+++, 15.4; Fe++, 14.2; Mn++,
13.5; Ca++, 10.59; Mg++, 8.69; Sr++, 8.63; Ba++, 7.76; Li+, 2.79; and
Na+, 1.66.
11-47. Calcium by Versene Titration. Any metal whose Versene com-
plex is less dissociated than that of Ca (whose equilibrium constant is
higher) will be preferentially complexed before Ca and thus will be in-
cluded in the Ca titration. Interfering 11 ions include Hg ( ous), Be, Cd, Sn
( ous or ic), Cu, Zn, Fe ( ic) and Mn ·( ous). The interfering ions that would
be encountered in this system are Fe and Mn. The Fe is removed by the
NH 4 0H separation, but the Mn normally passes through. The addition of
bromine water, in the NH 4 0H-Br separations given, serves to remove Mn
with the Fe. The removal of Mn is important for the analysis of some
Latosols and many rocks. If more than 0.5 mgm of Mn is present, it makes
the end point less distinct. The Mn can also be removed as carbamate (~
11-57), as Mn0 2 by KMn0 4 , 12 or as MnS by the addition of Na 2S to the
solution and filtration, prior to titration of the Ca.
11-48. Standard Ca Solution. A 0.5005-gm portion of pure dried CaCOa
is dissolved with a minimum of 0.2 N HCI. The solution is boiled to expel
the C02 and is then diluted to 1 liter. The solution is 0.0100 N with respect
to Ca.
11-49. Standard Versene Solution. A 2.0-gm portion of Versene (di-
sodium dihydrogen ethylenediamine tetra-acetic acid, from Bersworth
Chemical Co., Framingham, Mass., or Eastman Kodak) is dissolved in 900
ml of water. The normality of this solution is obtained by titration of a 25-
ml portion of the standard Ca solution according to the procedure (~ 11-

10 Helv. Chim. Acta, 31: 678 ( 1948).

11 Cheng and Bray, Soil Sci., 72:449 (1951).
12 Ingram and Bean, Anal. Chem., 25: 1217 (1953 ).
288 MINERA L COLLOID S, SOILS, MINERA LS, AND ROCKS
52), and the solution is then diluted so that the normality is exactly 0.0100.
A check standardization is then made.
11-50. Murexide Indicator. A 0.2-gm portion of Murexide (Eastman
Kodak) is mixed with 40 gm of powdered K~S0 4 • The Murexide indicator
is best added as the powder, since it is extremely susceptible to oxidation
when used as a solution. The susceptibility of Murexide to oxidation also
makes the Versene determination of Ca difficult on the samples derived
by fusion (~ 11-35) because of the presence of perchlorate. The sharpness
of the end point is greatly reduced by large concentrations of salts, which
is one reason the ammonium salts from the NH 4 0H-Br separation are de-
stroyed(~ 11-51) before the Ca titration. Another reason for the destruc-
tion of the ammonium salts is to facilitate the solution pH being raised to
12 for the titration.
11-51. Removal of Ammonium Salts. A 50-ml aliquot (or less if neces-
sary to contain less than 5 mgm of Ca) of Solution B, the supernatan t liquid
in the centrifuge tube after the NH 4 0H-Br separation (~ 11-41 ), is trans-
ferred to a 125-ml conical flask and 10 ml of HNO:; is added. The solution
is evaporated to dryness on a sand bath. If all of the visible ammonium
salts are not destroyed, 5 ml of aqua regia is added, and the evaporation is
repeated. The cooled residue is dissolved with 2 ml of 6 N HCl, with heat-
ing if necessary to get a clear solution, and the solution is diluted to ap-
proximately 50 ml.
11-52. Titration of Calcium. A 5-ml portion of 10 per cent KOH is
added, and 0.3 gm of Murexide indicator powder is dissolved in the solu-
tion. The amount of Ca in the solution is then determined by titration with
0.0100 N Versene solution (~ 11-49) to a purple end point. Murexide,
also known as ammonium purpurate, forms an orange-red complex with
Ca ions at pH 12, which is converted to a red-violet color when the Ca ions
are completely complexed by Versene. The color change is gradual, mak-
ing it imperative to compare the end point color with a standard. The end
point can be reached from either direction, and back titration with standard
Ca solution is possible and often helpful:
0.0481 (11-3)
% Ca = ml Versene x -··-; -- ------
weight of sample, gm

11-53. The solution resulting from the Ca titration with Versene (Mu-
rexide) may be used ( ~ 11-59) for the titration of Mg plus Ca with Versene
(Eriochrom e black T indicator) instead of the same titration on Solution E
(~ 11-60).
11-54. Magnesium Through Versene Titration. Total Mg is most easily
determined with the flame emission spectrophotometer on Solution B at the
same time as the K, Na, and Ca determinations (~ 11-37). In the absence
 MINERAL COLLOIDS , SOILS, MINERALS , AND ROCKS 289
of suitable equipment, the total Mg plus Ca is determined by Yersene
titration 13 with Eriochrome black T indicator, following the' NH 40H-Br
separation; the total Ca, determined separately, is subtracted, thus giving
total Mg.
11-55. The Eriochrome black T dye gives a wine-red colored complex
with Mg, but changes to blue when all the Mg ions have been removed by
the Versene; the titration is much more satisfactory than that with Murexide
(~ 11-52). The formation of the Mg complex is optimum at pH 10 in the
presence of NH4 Cl and NH 4 0H (~I 4-13). The Ca is complexed by Ver-
sene before the Mg, and therefore the titration value gives total Mg plus Ca.
11-56. Ions which would interfere 14 (~11-46) include: Cu, Co, and Ni,
which form a red complex with the indicator (not decomposed by the Ver-
sene titration); Fe, which precipitates the dye (if the latter is added before
the solution is made alkaline) and obscures the end point if present in large
amounts; and Mn, which, if present as MnO~, causes fading of the dye and
an indistinct end point. In the semimicrosystem, Fe and Mn are removed by
the NH 4 0H-Br separation, and Cu, Co, and Ni are complexed by NaCN.
11-57. In other applications, the interference of Fe, Mn, and Cu in the
Versene titration procedure for Ca plus Mg can also be removed by the
carbamate separation.ir. The sample solution in a volume of 50 ml is
brought to pH between 1 and 4 and then the solution is made 0.05 to 0.1
per cent with respect to carbamate (~I 15-45). The solution is shaken, and
the brownish complexes of Mn, Fe, and Cu with carbamate are formed.
Then I 0 to 20 ml of isomyl alcohol is added and the flask is stoppered,
shaken for about 10 seconds, and then allowed to stand until the alcohol
separates from the aqueous phase. The alcoholic upper layer carrying the
Mn, Fe, and Cu is removed by suction through a small glass tube connected
to an evacuation flask (or can be removed by means of a separatory funnel,
with less convenience). The isoamyl alcohol extraction is repeated until
the aqueous solution becomes colorless. The Ca and Mg remain in the
aqueous solution.
11-58. Reagents needed for the Versene titration (~ 11-59) include con-
centrated and 4 N NH 4 0H; NH.1Cl salt; thymolphthalein and alizarin yellow
indicators; 0.0100 N Versene (~I 11-49); Eriochrome black T indicator
(Eastman Kodak), prepared as 0.5 gm of the indicator with 4.5 gm of
hydroxylamine hydrochloride dissolved in 100 ml of methanol; and a 2
per cent NaCN solution. The hydroxylamine keeps any traces of Mn in

13 Gysling and Schwarzenbac h, Helv. Chim. Acta, 32: 1484 ( 1949); Schwarzenbac
h
and Gysling, Helv. Chim. Acta, 32: 1314 ( 1949).
14 Cheng and Bray, Soil Sci., 72:449 (1951); Conners, J. Am. Water Works Assoc.,
42:33 (1950); Diehl et al., J. Am. Water Works Assoc., 42:40 (1950); and Betz and
Noll, J. Am. Water Works Assoc., 42:49 (1950).
15 Cheng et al., Soil Sci., 15 :37 (1953).
290 MINERA L COLLOI DS, SOILS, MINERA LS, AND ROCKS
divalent form and the NaCN removes interference of traces of Cu, Co, or
Ni present. 1 H
11-59. Mg plus Ca Titration with Versene. The Versene titration of Ca
plus Mg may be carried out either on the solution left after Versene titra-
tion of Ca with Murexide (~ 11-53) or on Solution E (~ 11-60) resulting
from the NH 4 0H-Br separation on the Na 2 COa fused sample. The latter
supplies nearly twice as large aliquot of the sample and is thus advantageous
if the amount of Mg is small. The Murexide indicator remaining after the
Ca titration with Versene on Solution B is destroyed by the addition of a
few drops of bromine water (an excess is avoided) and the solution is then
acidified to bring the Mg(OH) 2 into solution. A solution containing 1 gm
of NH 4 Cl is added and then sufficient NH 4 0H to bring the final solution to
pH 10 (above blue to thymolphthalein and yellow-orange to alizarin yellow
indicator s). Next, 5 drops of the Eriochrome black T indicator solution
and 1 ml of 2 per cent NaCN solution are added. The solution is then
titrated with 0.0100 N Versene solution to a bright blue end point. This
titration is a measure of the total Mg plus Ca in the solution. To obtain the
percentage of Mg in the sample, the Ca equivalent is subtracted:
0.024
meq Mg+ Ca per gm sample= ml Versene x ----- ----
( Eriochrome ) Wt. sample, gm
(11-4)

From~ 11-52:
0.024
meq Ca per gm sample = ml Versene x - sample, gm
( Murexid•·) wt.
(11-5)
Then:
% M = (meqMg +Ca_ meqCap er) x 1. 216
g per gm sample gm sample
(11-6)
The meq of Mn (in the absence of Br2 in the NH 40H separation) is de-
termined separately ( ~ 1 1-172) and subtracted as was Ca.
11-60. To titrate Mg plus Ca on Solution E resulting from the NH 40H-
Br separation (~ 11-94), a 50-ml aliquot is placed in a 125-ml conical
flask, and 5 ml of concentrated NH 4 0H 17 is added to give pH 10. Then 5
drops of the Eriochrome black T indicator solution and l ml of 2 per cent
16 Dr. C. V. Cole recommen ded making the Eriochrom e black T solution I per cent
with respect to KCN to repress the minor element interferenc e (possibly introduced
that
as impurities in the indicator) . Diskant, Anal. Chem .. 24: 1856 ( 1952), notes
e black T indicator is stable when dissolved in di- or triethanola mine, but
Eriochrom
states that no other solvent or salts should be added.
1 7 This addition of NH.10H together with the NH 4 CI in Solution E are equivalent
to an NH40H-N H4CJ buffer solution of pH 10 (~I 4--13) ordinarily employed with
the Versene titration with Eriochrom e black T of Mg, Mg plus Ca, or Ca alone
(~ 4--21 ).
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 291

NaCN solution are added and the solution is titrated with 0.0100 N Ver-
sene as before (~ 11-59) from wine-red to a bright blue end point. Nearly
twice as large an aliquot of Mg and Ca is obtained in Solution E as in Solu-
tion B.
11-61. The meq of Mg plus Ca is calculated:
0.0144
meq Mg+ Ca per gm sample= ml Versene x --····-··-·-··-
(Eriochrome) Wt. sample, gm
{11-7)
The meq of Ca per gm sample is calculated ( ~ 11-52) and then the per-
centage of Mg ( eq. I 1-6). The meq of Mn (in the absence of Br2 in the
NH 4 0H separation) is determined separately (~ 11-172), and subtracted
as was Ca.

IRON AND TITANIUM DETERMINATIONS

(Rapid semimicrochemical system)

11-62. The Tiron method'H for the determination of Fe and Ti was

adopted in this system because of its simplicity, speed, and freedom from
interfering ions. The yellow color of Tiron ( disodium 1,2-dihydroxyben-
zene 3,5-disulfonate) with Ti is stable over a pH range of from 4.3 to 9.6,
but the purple color with Fe++ varies in intensity and hue from pH 4. 7
j

to 7 or above and therefore the pH is buffered at 4.7 for most satisfactory

results. At this pH value, the Fe complex shows a light absorption maxi-
mum at 560 mu and the Ti complex shows a maximum at 380 mu. The Fe
color is read first and then is destroyed by dithionite so as not to interfere
with the reading of the Ti color. The color of dithionite interferes with the
Ti color unless Ti is read with a 400- to 420-mu light maximum. The
colors of both the Fe and Ti complexes increase with increased concentra-
tion of reagent so that a large excess of reagent is required. The color, once
formed, increases slowly in intensity for 18 hours, after which it is stable
almost indefinitely. The increase, however, is very slow, so that readings
may be made over a considerable period of time, and 5 to 30 minutes was
adopted.
11-63. Few ions interfere with the color formed by Tiron with Fe and
Ti. The only other ions forming colored solutions with Tiron are VO++
(purple), MoO 4 - - , OsO 4 - - , U0 2 + + (yellow) and Cu++ (greenish yel-
low in alkaline solution). Ag and Au Cl- are reduced to the metals by the
reagent. Ions which consume the reagent are Al+ + +, Ca++, Ce 4 +, Mg++,
Fe 1 1 , SnH, Th 4 +, ZrO + +, and W0 4 - - • Tolerance to these ions is in-
creased and interference is overcome by the addition of a large quantity of
reagents. The only anions that interfere seriously are fluorides, which de-

IH Yoe and Armstrong, Anal. Chem., 19: 100 ( 1947).

292 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS
crease the color intensity when present in concentrations greater than 10
ppm, and perchlorates, which tend to increase the color of the Fe complex.
APPARATUS
11-64. Needed apparatus includes an absorption spectrophotometer with
565- and 400- (or 420-) mu light maxima; absorption tubes; 50-ml volu-
metric flasks; and 5-ml, 10-ml, and other pipets.
REAGENTS
11-65. Needed reagents include 3 per cent H 2 0 2 , sodium dithionite
(Na 2 S2 0 4 , from Amend Drug and Chemical Co., New York), Tiron re-
agent solution prepared fresh daily by dissolution of 4 gm of Tiron (La-
Motte Chemical Products Co., Towson, Md.) in 75 ml of distilled water
and dilution of the solution to 100 ml, buffer solution of pH 4. 7 prepared
by mixing equal volumes of 1 M HOAc (60 ml of glacial HOAc per liter)
and 1 M NaOAc (82 gm of anhydrous NaOAc per liter) and adjustment of
the pH to 4.7 (glass electrode) with NaOH or HOAc, and standard Fe
and Ti solutions prepared as follows.
11-66. Standard Fe Solution. A 0.0500-gm portion of pure iron wire is
dissolved in 10 ml of 0.6 N HCI and 1 ml of concentrated HN0:1, and the
volume is adjusted to 1 liter, so that the solution contains 50 ugm of Fe
per ml. Aliquots (2, 4, 6, and 8 ml) of this solution are taken for the
standard curve, and the color is developed as described in the procedure,
the standard solutions being substituted for the Solution A aliquot. The per-
centage light transmission for this series of solutions is plotted against ugm
of Fe per 50 ml on semilogarithmic paper. The curve follows Beer's law.
11-67. Standard Ti Solution. A standard Ti solution is prepared by
fusion of 0.1668 gm of standard ignited Ti0 2 in K 2Sp 7 (~ 11-113), tak-
ing up the melt with 10 ml of 6 N HCl. The solution is then made to a vol-
ume of I liter by dilution with 50 ml of 6 N HCI and then with water (rapid
stirring). The concentration is 100 mgm of elemental Ti per liter, in 0.4
N HCl. After this solution has been thoroughly mixed, 10 ml is pipetted
into a 100-ml volumetric flask and is diluted to volume with 0.4 N HCl
and mixed, giving 10 ppm of Ti.
11-68. The alternative use of TiCla for the Ti standard, which is often
recommended, gives difficulties in standardization, attributed to coprecipi-
tationlll of S0 4 with Ti(OH) 4 •
rn In tests by Dr. L. D. Whittig, TiCI :i (which characteristically hydrolyzes to
some extent to yield a precipitate) was diluted with I N H2S04, the Ti in solution
oxidized with H202. and the solution filtered. An aliquot of the clear filtrate was used
for precipitation of Ti with NH 4 0H. The precipitate was washed free from salts with
NH 4Cl, dried, and ignited over a Meker burner for 1 hour. On the basis of this
gravimetric standardization, the curve made with fresh portions of the solution gave
a low recovery of Ti in National Bureau of Standards samples analyzed by the pro-
cedure. It was concluded therefore that sulfate precipitated with the Ti and persisted
during the ignition, giving a fictitiously high Ti standardization.
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 293
11-69. Aliquots (2, 4, 6, 8, and 10 ml) of the 10 ppm Ti standard
(~ 11-67) are taken for the standard curve, giving 20 to 100 ugm of Ti.
The color is developed as described in the procedure ( ~ 11-70), the stand-
ard solutions being substituted for the Solution A aliquot. The percentage of
light transmission for this series of solutions is plotted against the ugm of
Ti per 50 ml on semilogarithmic paper. The curve follows Beer's law.

PROCEDURE

11-70. A drop of 3 per cent H 2 0 2 is added to the remaining portion of

Solution A to oxidize the Fe and Ti. To develop the color, first, a 10-ml
portion of pH 4. 7 buffer solution is pipetted into a clean 50-ml volumetric
flask. Next, the volume is adjusted to approximately 30 ml with water, and
5.0 ml of 4 per cent Tiron reagent solution is pipetted in. An aliquot that
contains from 50 to 250 ugm of Fe and I 0 to 100 ugm of Ti is now trans-
ferred from Solution A to the 50-ml volumetric flask. (A 5-ml aliquot is
convenient for samples containing up to 6 per cent Fe and 1.2 per cent
Ti.) The solution is diluted to the mark with distilled water and mixed, and
then approximately 20 ml of this solution is transferred to a colorimeter
tube. After 5 minutes, the percentage light transmission of the blue-violet
color of the Fe complex is read with a 565-mu light maximum. The readings
are completed within 30 minutes. Approximately 3 mgm (a bit on the end
of a spatula 1 mm wide) of Na 2 S~0 4 is next added to this same solution in
the colorimeter tube to destroy the color due to the ferric complex by re-
ducing it to ferrous. The colorimeter tube is stoppered with a rubber
stopper, and the content of the tube is mixed by gentle inversion of the
tube 4 or 5 times. The percentage light transmission of the yellow Ti com-
plex is obtained within 10 minutes after the addition of the Na 2S20 4 with
the 400-mu light maximum. The concentrations of Fe and Ti (ugm/50 ml)
are then obtained from the standard curves:
ugm Fe or Ti
n1
-10
F e or T"I = -per
- -50-ml
-- X --- 0.01
--- {11-8)
ml aliquot wt. sample, gm
taken
ALTERNATIVE PROCEDURES

11-71. Iron may also be determined by the orthophenanthroline pro-

cedure ( ~ 15-13), or by the SCN color method. Cheng et al. 20 describe
the determination of Fe in soil or limestone by a Versene titration at pH
2 to 3 with salicylic acid as an indicator. The color change at the end point
is from purple-red to colorless or light yellow, and the procedure is similar
to the Versene titration of Ca ( ~ l 1-4 7). The Tiron procedure for Fe and
Ti is much more convenient for use with the semimicrochemical system.

20 Anal. Chem., 24: 1640 (1952), 25:347 (1953).

294 MINERAL COLLOIDS , SOILS, MINERALS , AND ROCKS
Titanium may also be determined as the peroxided yellow color (~ 11-
155).
SILICON DETERMIN ATION
(Rapid semimicrochemical system)
11-72. The silicon from the Na2 C0 3 fusion(~ 11-35) is dehydrated with
HC10 4 , 21 washed free of metallic cations with 6 N HCl, and then is brought
into solution in NaOH and determined colorimetrically by the molybdo-
silicic yellow color method22 (~ 7-3). Perchloric acid dihydrate boils at
203°C and at this temperature it is a very powerful dehydrating agent. By
this method, the silica dehydration takes much less time than the hydro-
chloric acid evaporation, and the Si02 obtained is much purer, practically
no Fe being coprecipitated. Titanium, however, if present in fairly large
amounts will tend to be coprecipitated, as will Mn0 2 • This does not inter-
fere with the subsequent silicon determination. Only l dehydration is re-
quired to recover all but a very minute amount of the Si present. In addi-
tion, all the metal perchlorates with the exception of KC10 4 are freely
soluble in water or dilute acid. Removal of Ca, Mg, Fe, and Ti is necessary
to prevent coprecipitation with the Si. Phosphorus interferes if left in the
solution with Si, but is removed in the present system in the washings em-
ployed to remove the metallic cations.
11-73. The rate of yellow color development with molybdosilicic acid
increases with the concentratio n of Si, high concentrations attaining maxi-
mum color development in a few minutes and very low concentrations in
about an hour. After the maximum color is obtained there is a slow fading.
The rate of change in intensity of color is sufficiently low after 30 minutes
that 5 minutes does not produce a significant change, and a reproducible
curve can be obtained.
11-74. The molybdosilicic yellow color is very sensitive to pH and
somewhat sensitive to salt concentration. However the large amount of
molybdate added serves as an excellent buffer so that a 0.25-ml error in
measuremen t of the 6 N HCl does not produce a significant change of pH
or color intensity.

APPARATUS

11-75. Needed apparatus includes an absorption spectrophoto meter with

400-mu light maximum and tubes; 60-ml pointed centrifuge tubes and
centrifuge; an air-jet stirrer; a 3-way, 50-ml pipet with stopcock (Lowy);
1-liter, 500-ml, and 50-ml volumetric flasks; 250 to 500-ml Ni (G. T.
Walker, Minneapolis, Minn.) or Pt beaker; an electric hot plate; a rubber
policeman, pipets, and a buret.

21 Willard and Cake, J. Am. Chem. Soc., 42:2208 ( 1920).

22 Dienert and Waldenbulcke, Compt. rend., 176: 1478 ( 1923); and Hedin, op cit.
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 295

REAGENTS

11-76. Needed reagents include NaOH (pellets and 5 per cent solution),
60 per cent HC10 4 , 6 N and 1.2 N HCl, (NH 4 ) 6Mo 70 24 • 4 H 20 (150
gm in 1 liter of solution in distilled water, filtered if cloudy), and the fol-
lowing standard Si solution.
11-77. Standard Si Solution. Clear quartz crystals are digested for an
hour in concentrated HCl to remove surface impurities, then washed and
ground in an agate mortar to a fine powder that will pass a 0.15-mm sieve.
The powder is ignited briefly in a crucible, then cooled, placed in a vial,
and tightly stoppered. A 0.1070-gm sample is placed in a platinum crucible
and fused with Na 2 C03 (~ 11-35). The melt is dissolved in water and
diluted to a volume of 1 liter in a volumetric flask. This solution contains
50 ugm of elemental Si per ml. Aliquots (1, 2, 3, 4, 5, 6, 7, 8, 9, and 10
ml) of this solution are taken for the standard curve, and the color is de-
veloped (~ 11-78), the aliquots of standard solution being substituted for
the Solution D aliquot. The percentage light transmission for each of these
solutions is plotted against ugm of Si per 50 ml on semilogarithmic paper.

PROCEDURE

11-78. When the crucible from the Na 2C0 3 fusion(~ 11-35) has cooled,
the cover is placed on the crucible, and 8 ml of 60 per cent HC104 is added
dropwise under the slightly raised lid. When effervescence has ceased, the
lid and sides of the crucible are washed down with a minimum of water, and
the crucible, with the lid covering three-fourths of the top, is placed in a
sand bath on an electric hot plate, and the suspension is evaporated (vigor-
ous boiling must be prevented or loss of sample may result) to fumes of
HC10 4 • When dense fumes appear, the crucible is covered, and the suspen-
sion is boiled gently for I 0 minutes at a temperature a little above 200~C.
11-79. When the crucible has cooled, approximately 5 ml of distilled
water is added, and the suspension is carefully mixed and heated almost to
boiling to dissolve the salts that have solidified on cooling. The suspension
is then transferred to a 60-ml pointed centrifuge tube, and the crucible is
rinsed with a wash bottle, the washings being added to the tube. All of the
silica does not have to be removed from the crucible at this time. Approxi-
mately 2 ml of 6 N HCl is added, and the suspension is diluted to exactly
60 ml with water, thoroughly mixed with the air-jet stirrer, and centrifuged
at 1800 rpm for 5 minutes to throw down the silica. The suspension ad-
hering to the air-jet stirrer is washed back into the crucible.
11-80. A 50-ml aliquot is pipetted with a Lowy pipet from the super-
natant liquid in the centrifuge tube and transferred to another 60-ml pointed
centrifuge tube. This is Solution C, which is used for the determination of
Al ( ~ 11-93) (and also Mg plus Ca, Fe, and Ti, if so desired). Although
296 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS
the pipetting operation is more easily carried out by means of a Lowy
pipet, an ordinary 50-ml pipet may be used if care is exercised. The
suction apparatus shown in Fig. 11-3 may be used as a source of suction.
11-81. To wash the silica in the tube, about 50 ml of 1.2 N HCl is
added and stirred, and the suspension is centrifuged at 1800 rpm for 5
minutes. The supernatant liquid is then decanted by suction and subse-
quently discarded, since an aliquot for analysis of the solutes has already
been taken.
11-82. The silica is then washed from the tube into a 250- to 500-ml
nickel or platinum beaker with a stream from the wash bottle, and the
silica adhering to the sides of the crucible in which the dehydration was
carried out is also transferred. The silica adheres rather tightly to the
crucible so that it must be loosened with a policeman and washed out with
a stream from the wash bottle. The final washing of both the crucible and
centrifuge tube is made with warm 5 per cent NaOH to make sure all of
the SiO~ is removed. Approximately 2.5 gm of NaOH pellets is added to
the suspension in the nickel or platinum beaker, and the volume is ad-
justed to approximately I 00 ml. The solution is then boiled for 5 minutes
to dissolve the SiO~. When cool, this solution is transferred to a 500-ml
volumetric flask and diluted to the mark with distilled water. This is Solu-
tion D, used for the determination of Si. A slight turbidity is ignored (~i 11-
72).
11-83. A I 0-ml portion of ammonium molybdate solution is placed in
a 50-ml volumetric flask, and the volume is adjusted to about 30 ml with
distilled water. Then 5 ml of 6 N HCl is added, and the flask is swirled to
dissolve the white precipitate that forms. Finally, a Solution D aliquot con-
taining 50 to 500 ugm of Si is added, and the solution is diluted to the mark
with distilled water. A 5-ml aliquot of Solution D is desirable for samples
containing from 5 to 50 per cent elemental Si. The solution is mixed well,
transferred to a colorimeter tube, and allowed to stand for 30 minutes be-
fore the percentage light transmission is read with a 400-mu light maximum.
Reference to the standard curve gives the ugm of Si per 50 ml:

% Si = IJj~_~_i_p~r-~-~~ x ----- 9._Q~-- (11-9)

ml in aliquot wt. sample, gm

ALTERNATIVE PROCEDURES

11-84. For Si determination in systems other than the microchemical,

the pH of the solution in which the Si is to be determined should be ad-
justed to approximately pH 3 to insure that the buffer capacity of the sys-
tem is not exceeded as a result of excess acid or alkali in the test solution.
To be determined spectrophotometrically, the Si must be in true solution.
Digestion of organic compounds and colloidal aluminosilicates in HCIO~
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 297

usually converts the Si to forms soluble in NaOH. Silica gel is appreciably

soluble even in water if the solution is sufficiently dilute.
11-85. Artificial standards for Si, consisting of chromate colored solu-
tions, have been proposed. 23 When greater sensitivity is required than is
provided by the molybdosilicic yellow color, it is provided by the reduced
molybdosilicic blue color methods. 24

ALUMINUM DETERMINATION
(Rapid semimicrochemical system)
11-86. The aluminon method for Al was adopted in this system because
of its sensitivity and reliability as standardized by Smith et al. 2 ~ The red
color of aluminon (aurin tricarboxylic acid) with Al develops slowly for
a chelated compound; approximately 20 minutes are required for full
color development with the free acid at pH 4.2 at room temperature. At
lower pH values, the rate is still slower, and at higher pH values an in-
tense red color of the dye itself develops although the rate of fading is also
high. The solution is buffered at pH 4.2 controlled to ± 0. I pH unit. Con-
trolling the pH is better than raising it to 7 .0 to 7 .2 in an effort to cause the
excess reagent to fade selectively. Under the conditions selected, no heat or
stabilizers such as gum arabic are necessary. Because aluminon is highly
colored, the concentration of this reagent is a critical factor influencing
the color intensity of the final solution. Also, hot solutions have a lower
color intensity, but ordinary fluctuations in room temperature do not pro-
duce a significant change.
11-87. Interference by both cations and anions is extensive with Al
determination, and for this reason the Al is separated in practically pure
form (~I 11-88) prior to the determination. Ions that form red colors
deeper than that of Al include Be, La, Ce, Zn, Tl, Y, Nd, Er. Also, Fe
forms 2 n a color approximately one-third as intense as Al at 520 mu. White
precipitates are formed with Bi, Pb, Sb, Sn, Hg ( ic), V, Ti, and Si. Inter-
fering anions include fluosilicate, fluoride, tartrate, citrate, oxalate, malate,
and borate.
11-88. The Al is separated from other ions through the NH 4 0H sepa-
ration followed by the NaOH separation.

APPARATUS

11-89. Needed apparatus includes an air-jet stirrer; 60-ml pointed cen-

23Swank and Mellon, Ind. Eng. Chem., A.E., 6:348 ( 1934).

24Kahler, Ind. Eng. Chem., A.E., 13:536 (1941 ); Carlson and Banks, Anal. Chem.,
24:472 (1952).
25Anal. Chem .. 21:1334 (1949).
20 Dependent in part on the amount of Al present; Roberson, J. Sci. Food Agr.,
2:59 (1950), employed thioglycollic acid or hydroxylamine hydrochloride to prevent
the formation of color with iron.
298 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS
trifuge tubes and centrifuge; a buret for NH 4 0H; a hot water bath, a 50-ml
pipet (preferably with 3-way stopcock); a 125-ml conical flask; 1-liter,
500-ml, 100-ml, and 50-ml volumetric flasks; a buret for Al standard;
pipets; a colorimeter with 520-mu light maximum and tubes; and pH meter.

REAGENTS

11-90. Needed reagents include 6 N HCl, 4 N and concentrated NH 4 0H,

1 per cent NH 4 Cl solution, saturated Br~ water, NaOH (25 gm per 100
ml of distilled water, made up freshly for each set of determinations and
used while still hot), buffer solution of pH 4.2 ( 60 ml of HO Ac diluted to
900 ml with distilled water plus I 00 ml of 10 per cent NaOH, with final
adjustment to pH 4.2 as measured with the glass electrode), and aluminon
and standard aluminum solutions, made up as follows.
11-91. Aluminon Reagent. Exactly 0.200 gm of aluminon (aurin tri-
carboxylic acid,~ 7 Eastman, Rochester, N.Y.) is dissolved in 100 ml of the
4.2 buffer solution, and then the solution is diluted to 500 ml with dis-
tilled water.
11-92. Standard Al Solution. Exactly 0.500 gm of electrolytically pre-
pared metallic Al sheet or wire free from a surface coating of aluminum
oxide is dissolved in 15 ml of 6 N HCI. This solution is then diluted to 1
liter. A dilute standard solution is prepared by dilution of 10 ml of the
first solution to 1 liter, giving 5 ugm of Al per ml of solution. Aliquots ( 1,
2, 3, 4, 5, 6, 7, and 8 ml) of the dilute standard Al solution are taken for
the standard curve, and the color is developed as described in the pro-
cedure, the aliquots of standard solution being substituted for the Solution
F aliquot. The percentage light transmission for each of these solutions is
plotted against ugm of Al per 50 ml on semilogarithmic paper.

PROCEDURE

11-93. Aluminum is readied for the aluminon determination through the

following separations. Calcium and Mg are also obtained for the Versene
titration ( ~ 11-60).
11-94. The NH 4 0H-Br Separation. The 50-ml portion of Solution C
(the supernatant solution from the silica centrifugation, ~l 11-80) has al-
ready been placed in a 60-ml pointed centrifuge tube. Next 3 drops of
brom cresol purple indicator are added. The solution is heated in a boiling
water bath and then concentrated NH 4 0H is added until the first traces of
precipitate start to form, the solution being agitated constantly with the
air-jet stirrer. (Enough HCl is present to provide adequate NH 4 Cl for keep-
ing Mg in solution.) About 4 N NH 4 0H is then added dropwise until the
indicator turns from yellow to purple, followed by 2 ml additional NH 4 0H.
Then 2 ml of saturated Br2 water is added to oxidize Mn and cause pre-

21 The commercial preparation is as satisfactory as that synthesized, Smith et al.,

Anal. Chem .. 21: 1334 ( 1949).
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 299

cipitation of Mn0 2 • The Br2 also decolorizes the indicator. If the Ca and
Mg both are to be determined by flame emission ( ~ 11-3 7), the Br2 may
be omitted. The tube is placed in a hot water bath for 5 minutes and
cooled, and the volume is adjusted to exactly 60 ml. The suspension is
mixed thoroughly with the air-jet stirrer, then centrifuged for 5 minutes at
1800 rpm. The supernatant solution contains Ca and Mg, whereas the
precipitate contains the oxides of Al, Fe, Ti, and Mn. A 50-ml aliquot of
the supernatant liquid is pipetted into a 125-ml conical flask, care being
taken that the precipitate is not disturbed. This is Solution E, which is used
for the Versene titration of Mg plus Ca(~ 11-60).
11-95. Approximately 50 ml of a 1 per cent NH1CI solution is then
added to the tube, and the precipitate is thoroughly mixed with the air-jet
stirrer. The suspension is again centrifuged at I 800 rpm for 5 minutes, after
which the supernatant liquid is drawn off with the suction apparatus (Fig.
I 1-3) and discarded.
11-96. The NaOH Separation. The precipitate is dissolved by the addi-
tion of 3 ml of hot 6 N HCl, the suspension being stirred and heated in a
hot water bath to effect solution. (If the Ti content is over about 5 per
cent, the Ti precipitate may be in a form that is insoluble in HCl at this
stage, but this is of no consequence if Ti is run on Solution A, ~ 11-70.)
Next, 10 ml of hot 25 per cent NaOH solution is added with stirring, and
the suspension is placed in a hot water bath for 5 minutes and then allowed
to cool. It is then diluted to exactly 50 ml with distilled water and mixed
with the air-jet stirrer. The stirrer is allowed to drain completely, but the
adhering solution is not washed back into the tube. The suspension is
centrifuged for 5 minutes at 1800 rpm.
11-97. An aliquot containing 100 to 700 ugm of Al (5 ml for sample
containing 1 to 8 per cent Al) is placed in a 100-ml volumetric flask, and
3 ml of 6 N HCl is added for each 5 ml of aliquot taken. The solution is
adjusted to pH 4.2 (glass electrode), made to volume with distilled water,
and mixed. This is Solution F, used for the determination of Al(~ 11-99).
11-98. If the Fe content is over 10 per cent, occlusion of Al on the
Fe 2 0H precipitate may be excessive. If this problem arises, the entire super-
natant liquid from the NaOH separation is decanted into a 200-ml volu-
metric flask instead of an aliquot of it being taken. The precipitate is dis-
solved in 3 ml of hot 6 N HCl and the NaOH separation is repeated, the
supernatant liquid being decanted into the same 200-ml volumetric flask.
The NaOH separation is carried out still a third time in lieu of washing the
Fe 20:i precipitate. The 3 decantates combined in the 200-ml volumetric
flask are diluted to volume, an aliquot that contains 100 to 700 ugm of Al
(20 ml for samples containing 1 to 8 per cent Al) is placed in a 100-ml
flask. Then 0.45 ml of 6 N HCl is added for each ml of aliquot taken, the so-
lution is adjusted to pH 4.2 (glass electrode), and made to volume. This (as
in~ 11-97) is Solution F, used for the determination of Al (~ 11-99).
300 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS
11-99. Development of Color with Aluminon. A 10-ml portion of pH
4.2 buffer solution is placed in a 50-ml volumetric flask, and water is added
to adjust the volume to about 30 ml. Exactly 10.00 ml of 0.04 per cent
aluminon reagent is added, and the flask is swirled to mix the solution.
Finally, a 5-ml aliquot of Solution F (containing 5 to 35 ugm of Al) is
added, the solution is immediately diluted to the mark, and the contents are
well mixed. After 25 minutes some of the solution is transferred to a
colorimeter tube, and the percentage light transmission is read with a 520-
mu light maximum. Reference to the standard curve gives the ugm of Al per
50 ml of colored solution:
0.12
% Al = ugm Al per 50 ml· x --- ·········-----····- (11-10)
ml in aliquot wt. sample, gm
CONVENTIONAL SYSTEM OF SILICATE ANALYSIS
11-100. Conventional gravimetric, titrimetric, and colorimetric proce-
dures may be employed for Si, Al, Fe, Fe(ous), Ti, Ca, Mg, Mn, K, Na, P,
and S ( ~ 11-7). The accuracy of the gravimetric procedures, though not
as high ( ~ 11-8) as sometimes believed, is thought in most cases to be
somewhat higher than the procedures employed in the semimicrochemical
system given above, when adequate amounts of sample material is avail-
able. Preparation of the sample is considered in~ 11-11.
11-101. The Flow Sheet. The flow sheet for the conventional system is
shown in Fig. 11-5. Decomposition of I sample by Na 2 CO:i (~ 11-104) is
employed for the analysis of the elements, Si, Al, Fe, Ti, Ca, Mg, and Mn.
Separate samples are fused (~ 11-104) for K (~ 6-75), for P (~I 7-131),
and for S (~ 11-188) determinations. Decomposition of the separate
samples by HF is accomplished for Kand Na (~I 11-176) and for Fe(ous)
(~ 11-182). Dehydration in HC10 4 is employed for silica, the cupferron
separation is employed for separation of Fe and Ti from Al and P, and the
NH 4 0H separation is employed for separation of AI and P from Ca, Mg
and Mn.
APPARATUS
11-102. Apparatus needed for the Na 2 CO:i fusion consists of a 30-ml
platinum crucible and cover, an analytical balance, a 10-cm glass rod, a
camel's hair brush, a silica-tube triangle, a Meker burner, nickel or plati-
num tipped tongs, an 8-cm platinum dish or 400-ml Pyrex beaker with
watch glass to fit, a 50-ml beaker, a rubber policeman, and a 15-cm glass
rod with flattened end.
REAGENTS
11-103. Reagents needed for the Na 2 CO:i fusion and decomposition of
the melt are anhydrous Na 2 CO:p 70 per cent HCI0 4 ; and concentrated and
6 NHCl.
 Sample of mineral, soil or rock
Quartered to 25 gm Ground to pass 0.15 mm sieve

!OO'C
Oven·dry
weight

Titrate

0
Read @620mu
Titrate Q
~
Molybdote BaCI 2
SnCI 2

Molybdophosphoric
Blue color
Weigh
Read0660 mu

Fe, Al, Ti, Ca, Mg, Mn, P04 ----,

Ignite (excess Na, K. CI, S04)
weigh I
Cupfcrron separation 5°C
in It;! HCl
I
Insoluble I In absence
Ignition I of cupferron
Soluble I separation
Traces of I
Fe, Ti, Mg I
_ _ _ _ _ _ _ _ __J
I

Soluble
r:----'-1
~--- ____ --J Fe203, Fe.P041
I KzS207 fusion LA~~:_'._f~zj
Acid oxalate
I Jones reductor
I KMnO 4 titration Insoluble
Soluble

Fe, Ti
® Al 20 3, AIP0 4
Mg, Mn
excess salts

KMn04 HNOa
titration Versene titration

@ ~
Ti yellow Mg, Mn
color

ReadQY 420mu @ by difference

Read 9 540 mu

@ by difference

Fig. 11-5. Flow sheet for conventional system of silicate analysis.

301
302 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS

PROCEDURE

11-104. Na2 C0 3 Fusion of Silicates.28 A 1.000-gm sample of mineral

colloid, soil, or other silicate material, passed through a 0.16-mm ( 100
meshes per inch) sieve, is weighed out. Determinations of hygroscopic
moisture and ignited weight (~ 11-4) are made separately. This sample
is placed on 4.0 gm of anhydrous Na 2 C0:1 in a 30-ml platinum crucible
(care of platinum ware, ~ 11-13), and mixed in with a glass rod. Finally
1.0 gm additional Na 2 C03 is placed on top of the sample, and the rod is
brushed off on the fresh flux. The crucible is placed on a silica tube (or
nichrome) triangle in a slightly inclined position, and the cover is adjusted
to leave the crucible about 0.1 open to prevent reducing conditions. A low
flame of a Meker burner is placed under 1 side of the crucible, opposite to
the side left one-tenth open by the cover, and heating is continued for 5
or 10 minutes so that the side of the bottom of the crucible takes on a dull
redness. This low heating removes moisture from the sample gradually and
thus prevents spattering. Now the heat is gradually increased so that after
5 or 10 minutes more the mass completely fuses and forms a liquid melt.
The cover is kept adjusted so that the crucible is partly open but the open-
ing is not enveloped with the flame, as that would prevent access of oxygen
to the contents of the crucible.
11-105. After about 5 minutes of intense heating, bubbles of gas cease
to come off, which indicates that the fusion is complete. To insure com-
plete fusion of any material sticking on the upper sides of the crucible, the
crucible is inclined so that the main fused mass covers this material, and
while in this position, heat is applied until the main mass again fuses. In
order to assure complete oxidation of all material, the cover is removed
and the melt is heated 1 to 2 minutes more with a full flame. Then the flame
is removed, and after the top of the crucible ceases to glow, the crucible is
grasped with the crucible tongs and given a rotary motion to spread the con-
tents over the sides of the lower half of the crucible, thus expediting subse-
quent removal and solution of the melt.
11-106. Removal of the Melt. After the crucible and contents have
cooled, sufficient water is added just to cover the melt. Now gentle heat is
applied to the crucible by a small flame around the sides, causing the metal
to expand rapidly and thus break away from the cake inside. If the fusion
has been properly made and the crucible is clean and not badly dented, the
cake should loo,sen sufficiently to be removed, in which case it is tipped into
a platinum dish or 400-ml beaker. Should the cake resist becoming free in
the manner indicated so that it cannot be readily removed, the crucible with

28 For an extensive discussion of the various fluxes available, interfering substances,

and alternative pr,ocedures, the reader is referred to Hillebrand and Lundell, Applied
Inorganic Analysis (New York: John Wiley & Sons, Inc., 1929), pp. 700--713.
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 303

contents is placed directly in the dish or beaker, and removal is brought

about by digestion in water. If some material has spattered onto the cover,
this is transferred to the beaker by means of a policeman and a little water.
11-107. Disintegration of Melt. The melt is disintegrated in 50 to 100
ml of water, with heating and the aid of a glass rod with a flattened end, and
frequent stirring, the beaker being kept covered with a watch glass. For
the usual silica determination and complete analysis, the solution covering
the melt is acidified with I 0 20 ml of concentrated HCI and 10 ml of 70 per
cent HCl0 4 , and digested on the steam hot plate, with stirring from time to
time. The crucible and cover are extracted with a little 6 N HCI in a sepa-
rate small beaker, this solution being brought to a boil and then transferred
to the main solution. The silica is determined by volatilization with HF
(~ 11-110).

Al TERNA TIVE PROCEDURES

11-108. Samples that are high in oxides of manganese and iron tend to
cause damage to platinum crucibles by alloying Mn and Fe with the plati-
num. To avoid these difficulties, Dr. G. D. Sherman pretreated the sample
(prior to fusion) with aqua regia in a covered 400-ml beaker to dissolve as
much of the sample as possible. The dissolved portion of the sample is de-
canted into a second 400-ml beaker and the residue is washed thoroughly
with small portions of 6 N HCI. The residue is transferred to the platinum
crucible and then dried in an oven. The residue is weighed as an estimate of
quartz. The residue is then fused in Na 2 C0:3 by the procedure, and the ele-
mental analysis is made on the combined solutions resulting from the aqua
regia and fusion treatments.
11-109. If the iron oxide content is low, but the manganese oxide con-
tent is high, damage to the platinum crucibles is prevented by pretreatment
of the disintegrating melt with 2 ml of 50 per cent ethanol solution in 6 N
HCI and allowing the manganese to be reduced prior to carrying out the
procedure of removal of the cake from the crucible.

SILICA DETERMINATIONa 0
(by HF volatilization and weight Joss)
11-110. The silica is dehydrated in a boiling solution of HCl0 4 and the
other constituents are separated from the silica by filtration. The boiling
HC10 4 method 31 of silica dehydration can be effected in a few minutes with-

29 Or 30 ml of concentrated HCJ if the HCI0 4 is to be omitted for the alternative

HCI dehydration of silica(~ 11-118).
ao For more detailed discussion of silica determination, the analyst is referred to
Hillebrand and Lundell, Applied Inorganic Analysis (New York: John Wiley & Sons,
Inc., 1929); Robinson, Soil Sci., 59:7 (1945), U.S.D.A. Cir. 139 (1930); and Lenher
and Truog, J. Am. Chem. Soc., 38: 1050 (1916).
31 Willard and Cake, J. Am. Chem. Soc., 42:2208 ( 1920).
304 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS
out the necessity of the time-consuming baking of the silica necessary with
the conventional HCl dehydration method ( ~ 11-118). The use of per-
chloric acid permits elevation of the temperature to 200°C, whereas if
HClO 4 is not present, "a temperature of 130°C must not be exceeded, as
high temperatures lead to formation of basic salts with resolution of sil-
ica. "32

APPARATUS
11-111. Needed apparatus includes a 30-ml platinum crucible and
cover, glass hooks to support a 400-ml beaker cover glass, a funnel, a
100°C drying oven, a Meker burner, a silica-tube triangle, a sand bath on
electric hot plate, and a fume hood to carry off HCI0 1 fumes, preferably
through a water aspirator pump (~I 12-23).

REAGENTS
11-112. Needed reagents include 70 per cent HCl0 1; 0.5 N HCI; reten-
tive, ashless filter paper; 48 per cent HF; 2 N H 2 SO 4 ; and K~S~0 7 crystals
(~11-113).
11-113. Preparation of Potassium Pyrosulfate. A platinum dish (or cru-
cible) is filled nearly full of reagent grade KHS0 4 crystals, with care to
avoid getting any crystals on the exterior or near the upper edge of the dish
or crucible. The dish is placed on a silica-tube triangle, and the contents are
warmed gently with a low flame from a Meker burner adjusted for a fully
oxidizing flame. The water is gradually driven off, the heat being increased
somewhat after most of the water has been expelled. Precautions: Should
any of the salts spill or spatter onto the exterior of the dish, the heat is re-
moved and the dish is cooled, washed off, and dried, after which the heat-
ing is begun again. The reaction is:
2 KHSO, ~ K~S 2 0 7 + H 20 j (ll-11)
(lwat)

The slow heating is continued for 5 minutes after active bubbling ceases. A
slight bubbling (fine, nonspattering bubbles) will continue, accompanied
by evolution of a small amount of smoke ( S0:1 ), according to the following
reaction:
K.S.,0 7 ~ K.,SO,'
~ ~ (heat) -
+ S0.'1 j (11-12)

The objective is to obtain completion of reaction 11-11 with a minimum of

reaction 11-12. This is easily accomplished if excessively rapid heating is
avoided.

32 Smith, Perchloric Acid, 4th ed., Vol. I (Columbus, Ohio: G. F. Smith Chemical
Co., 1940), p. 23.
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 305

PROCEDURE

11-114. Dehydration of Silica with HCI0 4 • The solution (~ 11-107)

resulting from the Na 2COa fusion and HC10 4 and HCl dissolution of the
cake is covered with a watch glass supported on glass hooks and evaporated
to dense fumes (fume hood!), the evaporation being completed on an
electric or gas hot plate. The HC10 4 solution is gently boiled (fuming) for
15 or 20 minutes to dehydrate the silica, then cooled and diluted with 20
to 30 ml of hot water. All salts are soluble, and the silica is filtered off
through a retentive ashless paper and washed with 0.5 N HCI until free of
salts, about 10 or 12 washings being required. The silica contains fewer im-
purities than when separated by the HCl dehydration method. The filtrate
is saved for the R~0:1 separation (~ 11-122 or 11-132). A second silica
dehydration from HCl0 4 usually is considered unnecessary because only a
few tenths of a mgm of SiO~ is generally recovered:rn by evaporation of the
filtrate and fuming for 5 minutes in HCI0 4 • The paper and silica from the
first (and second) filtration is transferred to a platinum crucible, and the
crucible and contents are placed in an oven at I 00°C until completely dry.
11-115. Ignition of Silica. After the silica has been dried completely at
I 00°C, the crucible containing the filter papers is placed on a silica-tube
triangle and one side is heated very gently at first with a Meker burner hav-
ing a small flame. It is necessary that the paper be burned slowly in order
to prevent occlusion of carbon by silica (probably as silicon carbide). Also,
the SiO~ precipitate is light and fluffy and must be protected from Joss by
air currents. If the paper ignites, the flame is removed until the paper
ceases to burn. Sparking indicates incomplete washing out of HCl0 4 • With
the crucible tilted at a 45° angle, the flame is applied with just sufficient
heat, first to I side of the crucible and then the other, so that the crucible
shows dull red only on I side.
11-116. After 13 to 20 minutes slow heating, when the ash has turned
white on the under side, the crucible cover is placed to cover half the
crucible, and the heat is increased somewhat on that side so that the carbon
glows. To hasten oxidation of the last traces of carbon, it may be necessary
to cool the crucible and contents and break up the crumbs and granules
with a glass rod. When the carbon is completely burned and the ash is per-
fectly white, the crucible is blasted with the strongest flame of a Meker
burner for about I 0 minutes:i 4 during which time the crucible is about three-
fourths covered and the flame is applied so that the crucible opening is not
enveloped. The crucible is cooled for I 0 minutes in a desiccator and

33 As compared to several mgm from the second HCI dehydration, Hillebrand and
Lundell, op. cit., p. 723.
34 This blasting is necessary to volatilize traces (I to 3 mgm) of NaCl, which
usually is present, according to Robinson, Soil Sci., 59: 8 ( 1945).
306 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS
weighed, reheated for 5 minutes and reweighed. This process is repeated
until the crucible and contents are brought to constant weight. The final
weight is recorded.
11-117. Volatilization of Silica. The weight of silica is determined by
volatilization weight loss:
Si0 2 + 4 HF-- SiF4 + 2 H 2 0 (11-13)
(gas)

The silica is moistened with 3 ml of 2 N H 2 S0 4 , and then about 5 ml of 48

per cent HF is added slowly. The liquid is evaporated on a sand bath on the
electric hot plate until the crucible contents are almost dry and S0:1 begins
to be evolved. The crucible is cooled, 1 to 2 ml more HF is added and the
crucible is rotated in an inclined position so as to dissolve any Si0 2 that
may be stuck to the sides. The acid is evaporated to complete dryness on
the sand bath. Then the uncovered crucible, in an inclined position, is
gradually heated with a flame. The heating is finished with a fairly strong
flame of a Meker burner for a few minutes. The crucible and contents are
weighed, and this weight is subtracted from the previous weight to get the
loss in weight of pure Si0 2 • Volatilization of Ti0 2 :
Ti0 2 + 4 HF - - TiF4 + 2 H 20 (11-14)
(gas)

is preventeda 0 when H 2 S0 4 is present. The residue in the crucible is then

fused with 1 gm of K2 S20 7 (or Na 2COa). The crucible is placed in a clean
beaker and the cake is dissolved in 25 ml of N HCI, with heat. This solution
is added to the filtrate from the silica separation ( ~ I 1-1 14).

ALTERNATIVE PROCEDURES

11-118. Dehydration of Silica with HCI. To the beaker containing the

melt and suspended silica ( ,, 1 1-107) is added 30 ml of concentrated HCl,
and the beaker is placed on the hot plate with the cover glass supported on
glass hooks. The solution is evaporated to dryness, during the process of
which lumps and crusts are broken up with a glass rod with flattened end to
expedite evaporation. The residue is treated with I 0 ml of concentrated
HCl, and the solution is covered and heated on a hot plate for 10 minutes
or until the iron is in solution (but no longer), which is indicated by the
absence of brown specks. Immediately the solution is diluted with about
15 ml of hot water, stirred up well and without delay filtered on an ashless
filter paper (Whatman No. 40) while hot. With the aid of a policeman and
a fine stream wash bottle containing hot 0.5 N HCl, all but traces of the
silica are transferred onto the filter. The volume of acid used for the trans~

R5 McAlpine and Soule, Qualitative Chemical Analysis (New York: D. Van Nos-
trand Company, Inc., 1933), p. 385. The loss is small without H 2 S04 (~ 11-31 ).
MINERAL COLLOIDS, SOlLS, MINERALS, AND ROCKS 307
fer is kept sufficiently small to permit a complete transfer of the solution
and silica with I filling of the funnel. The filter and contents· are washed
until salts are all removed, usually requiring about 15 good washings.
11-119. In order to recover silica that is always left in solution after the
first HCI dehydration, the filtrate is transferred to the original dish or
beaker and evaporated to dryness again, and then the silica is dehydrated
in an oven at 100°C for 2 hours.
11-120. Then 10 ml of concentrated:rn HCI is added, the beaker is
covered and heated on hot plate for 5 to 10 minutes, and then immediately
15 ml of hot water is added, followed by prompt filtration on a second ash-
less retentive filter paper ( Whatman No. 42). The filter is thoroughly
washed with hot 0.5 N HCl, 12 good washings usually being sufficient. The
filtrate is saved for R.!0. 1 separation. The papers and silica from the first and
second filtrations are transferred to a platinum crucible, and the crucible
and contents are placed in an oven at I 00°C until the papers are com-
pletely dry. Ignition and volatilization of the silica are carried out as given
in the procedure.
11-121. The dehydration of silica from HCI is sometimes expedited by
addition of methanol to the encrusted salts and evaporation, this process
then being repeated twice for each dehydration step.

CUPFERRON SEPARATION OF IRON AND TITANIUM

(Precipitation of Fe, Ti, V, Zr, Sn, in IN HCI solution; remaining
in solution are Al. Co. Ni, Mn, Ca, Mg, P0 4 and Pt)

11-122. Cupferron (nitroso-phenyl-hydroxylamine-ammonium) forms

insoluble "chelate" coordination compounds:n with iron (titanium, etc.:i 8 )
in acid solution, forming Cp: 1Fe in which Cp refers to a cupferron anion
joined to Fe by one primary and one secondary valence, as follows: ~ 9

:w Concentrated HCI used for maximum insolubility, after Lenher, Merrill, and
Baldwin, .I. Am. Chem. Soc .. 39:2630 ( 1917); Robinson, Soil Sci., 59:8 ( 1945).
:J7 AcL:ording to Diehl. Chem. Rev .. 21: 39 ( 1937), the term "chelate" was proposed
by Morgan . .I. Chem. Soc., 117: 1856 (1920), to designate those cyclic structures that
arise from the union of metallic atoms with organic and inorganic molecules; it is
derived from the Greek word "L:hela," referring to the great claw of the lobster and
other crustaceans, and is applicable to these ring systems because of the caliper-like
character of the associating molecule. Smith (see footnote 39) states "The formation
of these rings may involve either primary or secondary valence. 'Chelate' rings may
be defined to cover all 3 types, that is, rings formed by 2 primary valences, by I
primary and 1 secondary valence, or by 2 secondary valences. Cupferron is repre-
sentative of the bidentate classification of chelate rings in which there is one acidic
group, and one coordinating group."
:m For further details, the analyst is referred to Hillebrand and Lundell, op. cit.;
and also Lundell and Hoffman, Outlines of Methods of Chemical Analysis (New
York: John Wiley & Sons, Inc., 1938).
311 Smith, Cupferron and Neocupferrott, their Preparation, Properties, and Analyti-
cal Applications (Columbus, Ohio: G. F. Smith Chemical Co., 1938).
308 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS

CH-N-OH
6 5 I
N
II (11-15)
0
Tautomeric forms of the cupferron molecule form the corresponding 2
chelate ferric iron precipitates:

( 11-16)

in each of which Fe is 6-fold coordinated with oxygen.

11-123. The cupferron separation is often omitted, and the filtrate from
the silica dehydration is taken directly to the ammonia separation. How-
ever, this necessitates the estimation of Al 2 0:i + P 2 0 0 by difference.
APPARATUS
11-124. Needed apparatus consists of an ice box or bath, a rubber
policeman, a filter funnel, a platinum crucible and cover, a I 00°C drying
oven, a Meker burner, and a silica-tube triangle.
REAGENTS
11-125. Reagents needed consist of ashless filter paper (Whatman No.
41), concentrated H 2 SO 4 and HNO:i• K 2 S2 0j and the following special re-
agents.
11-126. 6 Per Cent Cupferron. Six gm of cupferron reagent 40 is shaken
in I 00 ml of ice water until dissolved. Approximately 10 ml of this reagent
is required per determination of 100 mgm of Fe 2 0:i· This solution is stored
in a refrigerator, since the cupferron decomposes in warm solution.
11-127. Filter Paper Pulp. One ashless (Whatman No. 41) tilter paper
(per determination) is shredded and suspended in water, and the suspen-
sion is shaken vigorously to pulp the paper.
PROCEDURE

11-128. Precipitation. To the ice cold (4°C) solution resulting from the
filtration of the silica determination, combined with the residue from silica
volatilization fused in K 2 S2 0 7 or Na 2 CO:i, containing Fe, Ti, etc., sufficient
concentrated HCI is added to give a 1 N solution of HCI plus HCI0 4 • A
little filter paper pulp (equivalent to I filter disc) is added to help aggregate
the precipitate later. Finally the ice cold cupferron reagent is added drop-
wise ( 10 ml per 100 mgm of Fe 2 0x) until the precipitate aggregates into
large clumps and the supernatant liquid is the color of the reagent, but not
reddish. The aggregates are allowed to settle for 30 seconds, and then, to
test for completeness of precipitation, a few drops additional cupferron
40 Ammonium salt, as obtained from the G. F. Smith Chemical Co., Columbus, 0.
 MINERAL COLLOIDS , SOILS, MINERALS , AND ROCKS 309
reagent are added. Formation of white flashes with no further brown pre-
cipitate indicates complete precipitation.
11-129. Filtration. After the aggregates settle a few seconds, the solution
is filtered through a porous (Whatman No. 41) filter paper, the sides of the
beaker being policed well. The filtrate has a cloudy tan color that is due
entirely to the products of decomposition of the reagent. (Presence of iron
may be tested by further addition of cupferron, which should give only a
temporarily white precipitate.) The precipitate is washed 5 times with cold
1 N HCl containing a little cupferron, and then 3 times in addition with
ice cold water to remove most of the chlorides. The filter is then dried at
100°C.
11-130. Ignition. The precipitate, dried at 100°C, is carefully charred
(as in ~ 11-115) and finally ignited strongly to constant weight. The pre-
cipitate from clays or soils is mainly Fe~0:1 and Ti0 2 . The precipitate is
fusedinK~S~0 7 (~111-148) for the iron determination.
11-131. The cool filtrate from the cupferron separation is treated with
1 ml of concentrated H~SO 4 and I 0 ml of concentrated HN0 3 to remove
the excess cupferron. (The mixture would be slightly explosive if the filtrate
were heated before the acids were added, but is quiet if the acids are added
to the cool filtrate obtained after the normal precipitation procedure.) Then
the solution is placed on the steam hot plate and evaporated to dryness to
evolve free Cl~ and to oxidize and decolorize the cupferron. The mixture is
finally heated on the electric hot plate to fumes to remove HC10 4 • 3 or more
HNO:i treatments will be required. When the cupferron is destroyed, the
residue is ready for the NH 4 0H separation.

NH 4 0H SEPARATION AND ALUMINUM DETERMINATION

(Following cupferron separation41 of Fe and Ti. separation of Al
and P0 4 from Ca, Mg, and Mn.)
ll-132. The NH 4 0H separation is dependent on the insolubility of the
hydroxide and phosphate of Al (and Fe and Ti in the 'absence of the cup-
ferron separation) at or near the neutral point, and the solubility in the
presence of considerable NH 4 Cl, of the hydroxides of Ca, Mg, and Mn++.
ll-133. In the event the P0 4 exceeds the equivalents of Al (and Fe and
Ti) present, as will always be the case in the analysis of phosphate mate-
rials, it will combine with the divalent bases, causing their precipitation,
thus invalidating the separation. In this case, an excess and known amount
of Al (or Fe) is added to effect the separation.
APPARATUS
11-134. Needed apparatus consists of a crude buret for dispensing 4 N
NH 4 0H, a platinum crucible and cover, a Meker burner, and a silica-tube
triangle.
41 In the absence of the cupferron separation, Fe and Ti precipitate with Al and
P04.
310 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS

REAGENTS
11-135. Needed reagents consist of concentrated HCl, 4 N NH40H,
thymol blue and brom cresol purple indicators, filter paper (Whatman No.
41 ), and 0.5 per cent NH 4N0a washing solution (3 ml of concentrated
HN0 3 per liter, to which NH4 0H is added to pH 6.4, brom cresol purple).

PROCEDURE

11-136. Precipitation of R 2 0a. Five ml of concentrated HCl is added to

the H 2S0 4 -HN0 3-treated residue from the cupferron separation, and the
solution is then made to a volume of 250 ml (or, in the absence of the cup-
ferron separation, the HCl is added to the silica filtrate combined with the
K2 S20 7 solution of the residues from the silica volatilization, in a volume
of about 400 ml). Next, 4 N NH 4 0H is added dropwise from a buret until
a drop of thymol blue indicator dropped into the solution turns from red to
yellow, indicating a pH in the range of 2 to 3. Then 1 ml of brom cresol
purple (BCP) indicator is added, the solution is heated to boiling, and
more 4 N NH4 0H is added dropwise until pH 6.4 is reached from the acid
side. (A glass electrode pH meter or methyl red may be used in lieu of
BCP.)
11-137. In the presence of much iron, the indicator color may not be
clearly visible. In this event, small drops of the suspension are tested on a
spot plate, or a drop of BCP indicator is added from time to time and the
color is observed when it first strikes the solution. The pH is not allowed
to rise above the final value of 6.4 because Mn may be precipitated and will
not redissolve if the pH is again dropped to 6.4. 1f the pH value inad-
vertently exceeds 6.4, the reprecipitation later (~I 11-139) is mandatory;
otherwise it is sometimes omitted. The solution is finally brought just to the
boiling point and then is set aside for the precipitate to settle, followed by
prompt filtration.
11-138. Filtration. The bulk of the supernatant solution is decanted
through a porous, ashless, Whatman No. 41 filter paper, the filter and pre-
cipitate being washed once or twice with rinsings from the beaker with hot
0.5 per cent NH 4N0 3 solution adjusted to pH 6.4. It is not necessary to
police the beaker after the first precipitation if the R 2 0 3 is to be reprecipi-
tated. The filtrate is saved for the determination of Ca ( ~ 11-162), Mg,
and Mn ( ~ 11-1 72) .
11-139. Reprecipitation of R2 0 3 • In order to remove traces of man-
ganese, calcium, magnesium, and other salts that contaminate the R 2 0a
precipitate, a reprecipitation is necessary. The filter paper and contents are
placed into the original beaker, and about 20 ml of 6 N HCl is added. The
solution is heated to boiling to dissolve the precipitate, and the filter paper
is disintegrated with a glass rod. The solution is diluted with water to about
250 ml and heated to boiling. Then 1 ml BCP indicator is added and the
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 311
solution brought just to pH 6.4 by adding 4 N NH 4 0H, the same technique
being employed as in the first precipitation. The precipitate is allowed to
settle, and the solution is poured through the filter while hot. Immediately
the precipitate is washed 5 times with a hot solution of 0.5 per cent
NH4N0 3 solution. The filtrates are next combined and the solution re-
adjusted to pH 6.4, followed by digestion on a hot plate. If any more pre-
cipitate forms, it is caught on a third filter paper and washed I 0 times.
Filtrates from the ammonia separation are combined and saved for the cal-
cium ( ~ 11-162), magnesium, and manganese determinations ( ~ 11-172).
11-140. Ignition. The filter paper and precipitate are placed in a weighed
platinum crucible, oven-dried at 100°C and then charred carefully (as in
~ 11-115), and finally ignited to constant weight. The ignition is completed
by heating with a Meker burner for 5 to 10 minutes. Good oxidizing con-
ditions are rigorously maintained during ignition.
11-141. Content of Al 20 3 • The net weight of the contents of the cru-
cible, consisting of Al 20:{ and P 20r, (and Fe 20 3 and Ti0 2 in the absence of
the cupferron separation) is calculated. The content of P 2 0 5 is subtracted
(and also the content of Fe 2 0a and Ti0 2 is subtracted in the absence of
the cupferron separation) from the total, and Al2 0 3 determined by dif-
ference.
IRON DETERMINATION
(Following cupferron or NH40H separation)
11-142. The Fe is determined in H 2 S0 4 solution titrimetrically with
cerate (or permanganate) after reduction in the Jones reductor (or in HCI
after reduction in Walden Ag-reductor). The iron and titanium may also
be determined with Tiron ( ~ 11-62) or the iron by orthophenanthroline
( ~ 15-6) after the K2 S2 0 7 fusion given here.

APPARATUS

11-143. Apparatus needed consists of 600-ml and 400-ml beakers, stir-

ring rods, a suction flask, and a Jones reductor (prepared as follows) :
11-144. Preparation of a Jones Reductor. Amalgamated zinc is prepared
by treatment for about 30 seconds of approximately 300 gm of zinc gran-
ules (I mm) with enough I N HCl to cover. Then 200 ml of approximately
0.2 N HgCl 2 solution is added followed by vigorous stirring until no more
H 2 is evolved. The supernatant solution is decanted, and the amalgamated
zinc is thoroughly washed with distilled water and dried on a watch glass
at 100°C. The reductor column is prepared from a glass column 2 cm in
diameter, with a length for the zinc of 20 cm, fitted with a glass stopcock
on 1 end and a funnel on the other. A 1-cm depth of glass wool is placed
in the bottom of the column, followed by 20 cm of the amalgamated zinc,
and then 1 cm of glass wool.
312 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS
11-145. Precautions in the Use of a Jones Reductor.
I. The zinc column is kept covered with distilled water when not in use,
and with the I N H 2 S0 4 solution (never stronger than 2.5 N) when in use.
The objective is to avoid getting air into the column since this may cause
low results. The amalgamated zinc corrodes and cakes if exposed at length
to air.
2. Acids stronger than 10 per cent H 2 S0 4 (2.5 N) are avoided. H~S0 4
is preferred to HCI, although the latter may be employed.
3. Copper and other easily reducible metals, which may be reduced to
the free metal in the reductor, are avoided.
4. Ammonium solutions render the column worthless.
5. A reductor that is newly made or has not been used for several days
is treated by the passage of 100 ml of I N H 2 S0 4 followed by 100 ml of
distilled water. prior to use. This serves to remove any possible reducible
substances and insures efficacy of operation.
11-146. Blank Determination on the Jones Reductor. At the beginning
of a series of determinations, a blank is run on the reductor. To do this,
approximately 75 ml of l N H 2 S0 4 is passed through the reductor at the
rate of about 25 to 50 ml per minute, the stopcock being closed before the
level of the liquid drops sufficiently to expose the zinc at the top. This is
followed by 3 25-ml portions of distilled water. Each portion is added just
before the previous portion sinks below the top surface of the zinc column.
At no time is the zinc exposed to the air. When the last addition of water
has reached the upper glass wool pad, and the top surface of the zinc is
still covered, the stopcock is closed and the solution in the receiving flask
transferred to a 400-ml beaker, and the flask is rinsed 3 times with water.
The combined solution and rinsings are titrated with 0.05 N (NH 4 ) 4 Ce-
(S04) 4 (~ 11-154) and orthophcnanthroline indicator (or with 0.05 N
KMn0 4 ). If the reductor is in good condition, the blank determination
should not exceed 0.1 to 0.15 ml. This blank titration value is subtracted
from the values obtained in the regular determinations.

REAGENTS

11-147. Needed reagents consist of K 2 S2 0 7 (~ 11-113), NaF, 2 N and

1 N H 2S0 4 , 0.05 N (NH 4 ) 4 Ce(S0 4 ) 4 solution in 1 N H 2 S0 4 (~ 6-42),
and orthophenanthroline indicator (or 0.05 N KMn0 4 instead of cerate and
the indicator).

PROCEDURE

11-148. Pyrosulfate Fusion of the Oxides. Five gm of K 2S20 7 is placed

in the crucible containing the weighed precipitate of oxides of Fe and Ti
(and Al and PO 4 in the absence of the cupferron separation). The mixture
is fused for 1 or 2 minutes (under the fume hood) with a low flame just
sufficient to bring the bottom of the crucible to faint redness and to cause
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 313
a moderate evolution of S0 3 fumes. The crucible is then allowed to cool
until the melt is partially crystallized.
11-149. Then approximately 0.05 gm of NaF is added and heating is
continued until all the oxides are in solution, which is indicated by the ab-
sence of solid brown or white particles in the liquid fusion. Fuming of the
pyrosulfate to volatilize HF is necessary to prevent interference in the Ti
determination. It usually takes 5 to 10 minutes for the fusion, but heating
is not continued any longer than necessary since the crucible is attacked to
some extent. After cooling, the crucible is placed (clean externally) into
40 ml of 2 N H 2 S04 and heated until the melt dissolves. (The 2 N H 2 S0 4
is sufficiently strong to maintain Ti0 2 in solution.) The crucible is then
removed and washed off.
11-150. The solution should be clear at this point if the fusion has been
complete. However, if dark particles remain, they are filtered off on an ash-
less Whatman No. 42 filter paper, the filter then being dried at I 00°C and
ashed in a Pt crucible. The contents of the crucible are fused with 1 gm
K2 S2 0 7 and 0.05 gm of NaF (~ 11-148). The melt is cooled and dissolved
in 2 N H 2 S0 4 , and this second solution is combined with the solution re-
sulting from the first fusion.
11-151. Reduction and Titration of Iron. The H 2 S0 4 solution of Fe and
Ti (and Al and P0 4 in absence of cupferron separation) is now ready for
reduction of the iron. The solution is stirred and then passed through the
reductor (freshly washed, ~ 11-145,5) at the rate of about 25 ml per
minute. This is followed with 2 25-ml portions of 1 N H 2S0 4 and 2 25-ml
portions of distilled water so as to wash out all the test solution. The top
surface of the zinc is always left covered with water. The solution in the
flask is transferred to a 600-ml beaker, but the total volume is held down
to 400 ml to avoid the need for evaporation for the Ti determination.
11-152. The appearance of a purplish tinge in the solution indicates the
presence of at least 0.10 mgm of reduced Ti per ml. Less Ti than this is
reoxidizcd during the regular operations. To reoxidize quantities of 0.10
mgm or more, about 25 ml of distilled water is added (to carry in small
amounts of oxygen), and the solution in the beaker stirred vigorously for
about 6 minutes. If, at the end of this time, the color has not entirely dis-
appeared, 25 ml more distilled water is added, and stirring is continued
until the color is no longer apparent when the beaker is held over a white
background, and then is continued about 2 minutes longer. This procedure
does not cause reoxidation of iron. 42
11-153. The efficacy of the reductor in reducing all of the Fe to ferrous
may be tested with a few drops of KSCN added to a ml of the solution. De-
velopment of a pink or red color indicates incomplete reduction of ferric
iron, but the persistence of any ferric iron is almost never found. This test
42 Truog and Pearson, Ind. Eng. Chem., A.E., 10:631 (1938).
314 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS
also permits verification that Fe has remained reduced while the Ti was
being oxidized. Chromium, V, and Mo are also reduced by the Jones re-
ductor, but like Ti, are oxidized by air unless protected by collection out
of contact with air.
11-154. The solution is titrated with 0.05 N (NH 4 ) 4 Ce(S0 4 ) 4 to blue
with orthophenanthroline (or with 0.05 N KMn0 1 to the first appearance
of a pink color). The solution is saved for the determination of Ti. The
blank titration is subtracted; then 1 ml 0.05 N solution = 0.003992 gm
Fe 2 0 3 , or:

% Fe = ml x N x -- 5·??~-­ (11-17)
wt. sample, gm
TITANIUM DETERMINATION
(Following cupferron or NH 4 0H separation, spectrophotometrically
after peroxidation in 2 N H2 S0 4 )
ll-155. The yellow-colored compound formed when Ti is peroxidized
in 2 N H 2 S0 4 has long been employed for the Ti determination. 4 :1 The
color is thought possibly to involve a complex of Tin+ with S0 4 - --. Inter-
fering ions include V, Cr, W, Mo, and F. Also, yellow-colored ferric salts
must be compensated for; Hillebrand 4 '1 gives a correction equal to 0.2 mgm
of Ti0 2 peroxidized for each 100 mgm of FeP:i in I 00 ml of 1.2 N H~SO,.
At least 1.2 N H 2 S0 4 is necessary to prevent hydrolysis of the Ti(S0 4 ) 2
to insoluble basic sulfate. Also alkali sulfate causes bleaching of the color
except in the presence of a large excess of H 2 S0 4 • A concentration of 2 N
H 2 S04 is therefore employed.

APPARATUS

ll-156. Needed apparatus includes an absorption spectrophotometer

with 420-mu light maximum and tubes, and 500-ml and 100-ml volumetric
flasks.

REAGENTS

11-157. Needed reagents include concentrated and 2 N H 2 S0 4 , 30 per

cent H 2 0 2 , and a standard Ti solution prepared as follows:
ll-158. Standard Ti Solution. A standard Ti solution is prepared by
fusion of 0.250 gm of ignited Ti0 2 in 3 gm of K 2 S2 0 7 ( ~ 11-11 3) in a
platinum crucible. The fusion is carried out under the fume hood, the heat-
ing being just sufficient to cause moderate evolution of SOa fumes. Heat-
ing is continued long enough to obtain all of the Ti0 2 in solution, as indi-
cated by a perfectly clear, quiet, fused mass. Too rapid heating, however,

43 Weller, Ber. 15 :2592 ( 1882); Snell and Snell, Colorimetric Methods of Analysis,
Vol. I (New York: D. Van Nostrand Company, Inc., 1936).
44 U.S. Geo!. Surv. Bui. 700: 160 ( 1919).
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 315
will convert the K2 S2 0 7 to K2 S0 4 , which is not readily fusible. The crucible
(clean externally) is placed in a beaker containing 7 5 ml 2 N H2 SO4 and
the mixture is warmed gently until solution is complete. The solution is
transferred to a 500-ml volumetric flask and diluted to volume with 2 N
H 2 S04 • This solution contains 0.5 mgm of Ti0 2 per ml.
11-159. Aliquots (1, 2, 5, and 10 ml) of the Ti standard solution are
pipetted into respective l 00-ml volumetric flasks and diluted to 75 ml with
a 2 N H 2 S0 4 • Next, 5 ml of 30 per cent H 2 0 2 is added and the solutions
are then mixed. The yellow color develops instantly. When the color has
developed, each solution is made to volume with 2 N H 2 S04 and mixed,
giving 5, I 0, 25, and 50 ugm of Ti0 2 per ml (or 3, 6, 15, and 30 ugm of
Ti) in the flasks. This series is read in the colorimeter, 2 N H 2 S04 being
used as the blank. The concentration of Ti or Ti0 2 as ugm per ml is plotted
on a linear scale against percentage transmission on a log scale. The curv~
is linear, and after having once been established, is only rechecked occa-
sionally, with the 6 ugm of Ti per ml standard.

PROCEDURE

11-160. To the solution left after completion of the Jones reductor titra-
tion of the Fe (~I I I- I 54), enough concentrated H 2 SO4 (usually about 20
ml) is added to make the solution approximately 2 N in a final volume of
500 ml, the amount of acid already present being taken into account. If
the volume is less than 500, the solution is transferred to a 500-ml vol-
umetric flask, but is not yet made to volume.
l l-161. Development of the Ti Yellow Color. To develop the color,
approximately 5 ml of 30 per cent H 2 0 2 is added per l 00 ml of the Ti
solution at hand and the solution mixed. If the color is distinctly yellow (in
range of standards), sufficient Ti concentration is present for the determina-
tion and the solution is made to volume. If too dilute, the solution is
evaporated as needed to bring the Ti concentration into range. The solu-
tion is then diluted to an exact volume and mixed, the percentage trans-
mission is read in the colorimeter with a 420-mu light maximum, and the
reading is referred to the calibration curve to obtain the ugm of Ti per ml.
Visual comparison may also be made to the standard by Nessler tubes. A
mineral colloid or soil containing 0.5 per cent Ti0 2 (common) gives 10
ugm of Ti0 2 or 6 ugm of Ti per ml in a 500-ml dilution volume:

% Ti = ugm per ml x ~ytic:m volume, ml

(11-18)
10 4 wt. sample, gm
CALCIUM DETERMINATION
(Total, in conventional silicate analysis system)
11-162. The total calcium obtained in the filtrate from the NH 4 0H
separation (~ 11-138) may be determined by the oxalate procedure that
316 MINER AL COLLO IDS, SOILS, MINER ALS, AND ROCKS
follows. Through suitable standardization, total calcium may also be de-
termined with Versene ( ~ 11-46) or by the flame emission spectrophotom-
eter (~ 11-37).

APPARATUS

11-163. Needed apparatu s includes a 400-ml beaker, a steam hot plate,

and fritted glass crucible (medium porosity ) with holder and suction flask.

REAGENTS

11-164. Needed reagents are brom phenol blue indicator, 6 N HCl,

4 N NH 4 0H, and reagents for precipitation of calcium (~! 5-26).

PROCEDURE

11-165. The filtrates from the NH 4 0H separation are brought to a vol-

ume of approximately 250 ml by evaporation, but the solution is not al-
lowed to go to dryness :15 The solution is transferred to a 400-ml beaker, 1
ml of brom phenol blue indicator is added, and finally a little 6 N HCI
is added if necessary to make the solution slightly acid. The calcium is next
precipitated ( ~ 5-31 ) in a 400-ml beaker and reprecipitated ( ~ 11-166 )
if the amounts of calcium and particularly magnesium are large, as will be
the case in the analysis of calcareous soils or dolomitic limestone. The fil-
trate is saved for the determination of Mg and Mn. The calcium oxalate
is determined usually by titration (~ 5-33 or 5-38) or, rarely, gravimetri-
cally (~! 5-37).
11-166. Reprecipitation of Calcium. Reprecipitation is necessary when
the amount of magnesium is large. The precipitate of CaC~0 4 (~! 11-165 )
is redissolved with about 15 ml hot 6 N HC1 and washed out of the filter
with hot water, diluted to 150 ml with water and reprecipitated as in the
first precipitation. The second filtrate from calcium is united with the first
for the determination of Mg plus Mn ( ~ 11-167 ).
MAGNESIUM DETERMINATION
(Total, in conventional silicate analysis system)
11-167. The procedure for Mg actually determines Mg plus Mn, but the
latter is usually so small in amount as to be within the experimental error
of the Mg determination. This is checked on, however, by a Mn determina-
tion(~ 11-172 ). Total Mg plus Mn may be determined
by titration with

4r. If the solution should happen to go to dryness at this point,

some dark colore<l
15 ml of
material becomes insoluble. If this happens, 5 ml of concentr ated HNO~ and
with a watch
water are added and the solution is stirred. The beaker is then covered The
glass supported on glass hooks, and the solution is evaporat ed to dryness.
now be white or only slightly yellow. If not, the residue is taken up in
residue should
These treat-
about 3 ml of HNO:i and 10 ml of HCI and evaporat ed to dryness again.
soluble. The
ments destroy coloring matter and render other constituents easily
residue is taken up in 5 ml 6 N HC1 and the solution diluted to 250 ml.
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 317
Versene (below) or as MgNH 4 P0 4 and MnNH 4 P0 4 (~ 5-48). Through
suitable standardization, total Mg may be determined by flame emission
spectrophotometry ( ~ 11-3 7).

APPARATUS

11-168. Needed apparatus includes a cover glass and glass hooks to

raise it from the beaker, a steam hot plate, and apparatus for the titrimetric
determination of Mg ( ~ I 1-54).

REAGENTS

11-169. Needed reagents consist of concentrated HN0 3 , concentrated

HCI, 4 N NH 4 0H, brom thymol blue indicator, and the reagents for Mg
determination ( ~ 1 1-58).

PROCEDURE

11-170. Removal of Excess Ammonium Salts. Approximately 20 ml of

concentrated HNOa is added to the filtrate from the calcium determination
(or combined filtrates in case of reprecipitation), and the beaker is covered
with a cover glass supported on glass hooks. The solution is evaporated
to dryness in order to decompose the excess of ammonium salts, yielding
volatile products.
Equations:
NH 4 CI + HN0 3 - NH4 NOa + HCI (11-19)
2HC1+2HNO:i- 2N0 2 +Cl 2 +2H 2 0 (11-20)
(brown gas)

N 0 + 2H 2 0 (11-21)
( colorfess gas )

3 Cl 2 + 2 NH 4 C l - 8 HCl + N2 (I 1-22)

H 2 Cz0 4 + Cl 2 - 2 HCl + 2 C0 2 ( 11-23)

The treatment is repeated by adding I 0 ml concentrated HN0 3 and 30 ml
concentrated HCl (aqua regia) and evaporating to dryness again, and is
repeated still again if considerable ammonium salts remain, which will be
the case if more than the usual quantity of acid and NH 4 0H were used in
the NH 4 0H separation.
11-171. The residue is dissolved in about 3 ml of concentrated HCl,
diluted to 100 ml and neutralized with 4 N NH 4 0H using brom thymol
blue indicator. Then Mg (plus Mn) is determined titrimetrically with
Versene (~ 11-59). Also, Mg may be precipitated as the phosphate (~
5-48), and determined titrimetrically ( ~ 5-52). In either case, the meq
of Mn ( ~ 1 1-61 ) is subtracted from the meq of Mg plus Mn titrated, and
the Mg is derived by difference.
318 MINERA L COLLOIDS, SOILS, MINERA LS, AND ROCKS

MANGA NESE DETERM INATIO N

(Total, in conventional or rapid semimicrochemical system)
11-172. The total manganese may be determined in an aliquot of the
filtrate from the NH 4 OH separation ( ~ 11-139, 11-40, or 11-94) or from
the Ca determination ( ~ 11-165) , and also in the precipitate from the
NH 4 0H-Brse paration (~ 11-41) orNaOH separatio n (~ 11-96).
11-173. Manganese may also be determined in the solution resulting
from the titration of Mg plus Mn (~ 11-171) in the conyentional system,
or in the solution remaining after Versene titration of Ca plus Mg plus Mn
with Versene in the semimicrochemical system ( ~ 11-59). Also, following
a gravimetric determination as Mg 2 P 2 0 7 and Mn2P 2 0 7 (~ 5-56), the ig-
nited ash may be fused in K2 S2 0 7 and Mn determined in a H:1P0 4 solution
of the melt.
11-174. The aqua regia treatment is given for removal of organic matter
and excess ammonium salts as detailed in the Mg determination (~ 11-
170) if Mn is to be determined on an aliquot of the filtrate from either the
NH 40H separation or the Ca determination in the conventional system.
This treatment is also given if Mn is to be determined on the solution re-
maining after titration of Ca plus Mg and Mn with Versene in the semimi-
crochemical system.
11-175. The Mn solution is freed of traces of organic matter, ammo-
nium, and halides by the treatment detailed in ~I 5-74; it is obtained in
solution in H 3P0 4 and then is ready for the usual periodate determina-
tion of Mn as detailed in~ 5-76.

POTASSIUM AND SODIUM DETERM INATIO NS

(Total, conventional system by decomposition of sample in HF)
11-176. Potassium and sodium cannot be determined on the sample
fused in Na 2 COa earlier in this system because appreciable potassium has
been introduced in the reagents during the course of the analysis, and of
course sodium has been introduced in the Na 2 C0 3 • A separate sample is
decomposed in HF, and the K and Na are subsequently determined by
means of titrimetric, spectrochemical, or gravimetric methods. The follow-
ing procedure details the method of decomposition of the sample in HF;
it is similar to that given in the semimicrochemical system of analysis (~
11-31), but it is adapted to macro samples.

APPARATUS

11-177. Needed apparatus includes a 30-ml platinum crucible and

cover; an electric hot plate with sand bath; a 100-ml volumetric flask; and
150-ml, 100-ml, and 50-ml beakers.
MINERAL COLLOIDS, SOILS, MINERALS , AND ROCKS 319

REAGENTS

11-178. Needed reagents are 18 N H 2S0 4 , 48 per cent HF solution,

concentrated HN0 3 , 60 per cent HC10 4 , and 6 N HCI. In the conventional
procedures for K and Na, 6 N NaOH and 6 N NH 40H and brom cresol
purple indicator are required in addition.
PROCEDURE

11-179. Decomposition of Sample in HF. A sample, of mineral colloid,

soil, silicate rock, or mineral, having passed the 0.16-mm sieve ( 100
meshes per inch) and weighing from 0.1 to 1.0 gm, is placed in a 30-ml
platinum crucible. The sample is moistened with a few drops of 18 N
H 2 S0 4 , then I ml of HC10 4 and 5 ml of 48 per cent HF are added (the
H 2 S0 4 is omitted for a sample containing over 2 per cent Ti). The crucible
is then about 0.7 covered and is placed on a sand bath on an electric hot
plate regulated at 200° to 225°C, after which the acids are evaporated to
dryness. The heating rate is regulated so that H 2 S0 4 fumes are evolved, but
so that the solution does not boil. Additional portions of the H 2 S04 , HC10 4
and HF acids are added so that a total of 3 treatments and evaporations
have been given. At last, a few additional drops of H 2S0 4 are added and
the mixture is fumed to drive off fluorides. Then the crucible is cooled and
5 ml of 6 N HCl is added. The suspension is diluted to 20 ml with water.
The crucible is then placed in a radiator (on a triangle within a large
crucible). The solution is brought to very gentle boiling with a low flame
applied to the outer crucible. The residue usually goes into solution in
about 5 minutes.
11-180. Analysis of the Solution. Analysis of the solution in the crucible
may be made for K and Na, and in addition, for Fe, Ti, Al, Ca, Mg, and
Mn if desired. The method of transfer depends on how many elements are
to be determined and the determinative method to be chosen, outlined as
follows:
1. For K and Na by flame emission spectrophoto meter, the solution is
transferred to a 100-ml volumetric flask and the determination is made
directly in the HCl solution (~ 11-45). The determinations are run as
promptly as possible to minimize contaminatio n from glassware.
2. For K determination by the cobaltinitrite precipitation method (suit-
able for range of 0.5 to 6 mgm of K in 5 to 10 ml, the solution in the
crucible (or a known fraction of it by a prior transfer to a 100-ml volu-
metric flask) is transferred to a 150-ml beaker, diluted to 75 ml, heated to
boiling and brought to pH 6.4 ( brom cresol purple) by means of 6 N
NaOH. The solution is transferred to a JOO-ml (or other volume to give the
proper range) volumetric flask, brought to volume at room temperature,
mixed, and filtered through a dry Whatman No. 40 filter paper, with no
washings and with precaution against evaporation. An aliquot of 5 or 10
ml, containing 0.5 to 6 mgm of K, is taken for precipitation of the K ( ~
6-22).
320 MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS
3. For Na determination by the magnesium uranyl acetate procedure
(suitable for the range of 1 to 5 mgm of Na), the solution in the crucible
(or a known fraction of it taken by dilution in a volumetric flask) is trans-
ferred to a 150-ml beaker, diluted to 75 ml. brought to boiling and neutral-
ized to pH 6.4 with 6 N NH 4 0H (brom cresol purple), transferred to a
100-ml volumetric flask, made to volume at room temperature, and well
mixed. It is then filtered through a dry Whatman No. 40 filter paper with-
out washing and with precaution against evaporation. An aliquot containing
1 to 5 mgm of Na is transferred to a well-weathered 100-ml beaker. Then
10 ml of aqua regia (3 HCI to I NHOa) is added and the solution is
evaporated to dryness to destroy ammonium salts. If much ammonium salt
appears to remain, the residue is taken up to 2 ml of aqua regia and the
evaporation is repeated. The Na is then determined(~ 5-88).

AlTERNA TIVE PROCEDURE

11-181. Total potassium and sodium of minerals may be determined by

the classical NH 4 Cl-CaCOa method of J. Lawrence Smith, 4 n but it is not
easy to obtain these reagents free from the alkali metals; also the method
requires a relatively large sample. The sample is intimately ground into a
mixture of NH 4 Cl and CaCO:i and heated. Finally the mixture is leached,
and the filtrate is subjected to precipitation with (NH 4 ) 2 CO:i, and the filtrate
from this is evaporated to dryness, after which NH 1Cl is driven off by
cautious heating and the NaCl and KC! weighed. The K may then be de-
termined separately by precipitation as chloroplatinate or cobaltinitrite, the
NaCl being determined by difference from the combined weight of the
mixed chlorides (~ 10-84).

FERROUS IRON DETERMINATION IN SILICATES

11-182. The determination of ferrous iron in silicates is essential for the
allocation of elemental analysis to the mineral constituents. After the de-
termination of ferrous iron, the ferric iron is determined by difference from
the total iron(~ 11-62 or 11-142).

APPARATUS

11-183. Needed apparatus includes a 30-ml or larger platinum crucible,

a burner and windshield, a 400-ml beaker, platinum or nickel tongs, and a
small buret. Also needed is a radiator consisting of a porcelain crucible into
which the platinum crucible is suspended on a triangle.

REAGENTS

11-184. Needed reagents are 18 N H 2 S04 , 48 per cent HF, saturated

H 3 B0a solution, standard 0.025 N (NH 4 ) 4 Ce(S0 4 ) 4 • 2 H 2 0 in 1 N H 2S0 4
(or standard K2 Cr 20 7 , or KMn0 4 solution). If the cerate solution is to be

46 Washington, Chemical Analy~·is of Rucks, 3rd ed. (New York: John Wiley &
Sons, Inc., 1919 ); Smith, Am. J. Arts, J,3:269 (1881).
MINERAL COLLOIDS, SOILS, MINERALS, AND ROCKS 321

employed, 85 per cent H 3 P0 4 and diphenylamine sulfonate indicator solu-

tion are needed.

PROCEDURE47

11-185. A 0.1 to 0.5 gm of sample, ground to pass a 0.15-mm sieve is

placed in a 30-ml platinum crucible and 1 to 2 ml of water is added to
moisten the powder. Then 10 ml of 18 N H 2 SO4 is added to the crucible,
the cover is placed completely over the crucible, and the solution is heated
nearly to boiling with a burner, heat being applied to the radiator described
with the apparatus. The crucible cover is then moved slightly to the side,
and 5 ml of HF is added. The cover is replaced and the liquid is quickly
heated to boiling. Gentle boiling is continued for 10 minutes, the wind-
shield being employed to maintain steady slow heat.
11-186. While a sample is being decomposed by the HF treatment, a
400-ml beaker is half-filled with cold water and 10 ml of 18 N H 2 S0 4 and
I 0 ml of the saturated H:iBOa solution are added. (The boric acid is em-
ployed for complexing F as weakly dissociated HBF 4 , necessary to prevent
formation of ferric fluoride complex and acceleration of ferrous oxidation
in the presence of air.) By means of tongs, the crucible and cover are placed
in the solution in the beaker. The solution is stirred to suspend the solid
matter in the solution and then the ferrous iron is titrated with the 0.025 N
standard oxidizing solution, 2 drops of "Ferroin" being added as indicator
for the cerate titration (or 6 drops of diphenylamine disulfonate solution
along with I 0 ml of 85 per cent HaP0 4 if K2 Cr:i0 7 is to be employed for
the titration). If organic material is present in the sample, the end point
may fade and erratic results will be obtained. After the titration the bottom
of the beaker is examined for any dark particles that would indicate in-
complete decomposition of the sample. If such dark particles are found, the
determination is repeated with a more finely ground sample, but excessively
fine grinding is avoided because ferrous iron may be oxidized as a result
of the grinding.

ALTERNATIVE PROCEDURE

ll-187. Maintenance of an atmosphere of C02 over the crucible has

been 4 H employed as a precautionary measure for prevention of oxidation of
iron by air. However, the water vapor evolved on boiling the sample ordi-
narily protects against contact with oxygen and permits a satisfactory de-
termination of ferrous iron.

47 Procedure employed at this laboratory is essentially as described by Kolthoff

and Sandell, Textbook of Quantitative Inorganic Analysis, rev. ed. (New York: The
Macmillan Company, 1945), p. 746.
48Sarvar,J. Am. Chem. Soc., 49:1472 (1927); Schollenberger, !. Am. Chem. Soc.,
53:96 (1931).
322 MINER AL COLLO IDS, SOILS, MINER ALS, AND ROCKS
TOTAL SULFUR IN SOILS49
(Separate fusion in Na 2 C0 3 and NaNOa, and precipitation as BaS04)
11-188 . Total sulfur determination for soils should be thought of much
as the total nitrogen determination for soils. Sulfate may accumulate in soils
as gypsum, CaS0 4 • 2 H 20, and its determination in this form has been
considered ( ~ 10-102 ) in connection with soluble salts in soils. Much of
the sulfur is in organic form and slowly becomes available for crop use. The
sulfur requirement of crops is relatively high, being 0.5 of the P requirement
of cereals and equal to the P requirement of legumes.
APPARATUS

11-189. Needed apparatus includes a 25-ml platinum crucible, electric

furnace, filter funnel, and Meker burner.
REAGENTS

11-190. Needed reagents are anhydrous Na 2 C0:1> NaN0:1, 6 N HCI,

10 per cent BaC12 solution, HF, H~S01 and filter paper.
PROCEDURE

11-191. Fusion. One to 2 gm of the finely ground and well-mixed soil is

fused in a 25-ml platinum crucible with 5 times the weight of Na~C0: 1 and
0.2 or 0.3 gm of NaNOa. The fusion is best done in an electric furnace. If
gas is used for the ignition, the melt is likely to be contaminated with sul-
fate. To avoid the latter, the crucible is supported in platinum foil with a
hole cut in it to let the crucible in for about half its height. This foil, in
turn, is supported by a sheet of thin asbestos. When the flame is played on
the lower part of the crucible, the upper part of the crucible is protected
from the products of combustion. After fusion, the melt is thoroughly dis-
integrated in water on the steam bath, preferably by digestion over night.
The solution is then filtered, and the silica washed. The filtrate is then made
up to 150 to 175 ml if not already of this volume. Enough HCl is added to
neutralize the Na 2 CO:i and to make the solution about 0.3 N HCl in excess.
11-192. Precaution. Occasionally, because of a long digestion, or too
large an excess of acid, the silica remaining in the filtrate gels. One of the
2 following procedures may be used before the BaC1 2 is added: (a) The
silica may be removed by dehydration and filtration, or (b) the silica may
be filtered from the solution neutralized with NH 40H.
11-193. Precipitation of BaS0 4 • The solution is brought to boiling, and
I 0 ml of I 0 per cent BaC1 2 is slowly added to precipitate the sulfate. The
solution is allowed to stand until cool, then is passed through a small fine-
porosity filter paper, and the precipitate is washed.
11-194. Ignition of BaS0 4 • The paper is ignited at a low tempera ture

49 Robinson, Soil Sci., 59: 11 (1945).

 S 323
MINE RAL COLLOIDS, SOILS, MINERALS, AND ROCK
a few drops
and the precipitate weighed. The ignited BaS04 is treated with
of HF and H 2S04 , cautiously ignited, and weighed again.
blanks are
11-19 5. As the reagents used invariably contain some sulfur,
carried along in the same manner as the determinations.

x . 13 ·7 (11-2 4)
% S =gm BaSO 4 weight of sample, gm

MISCELLANEOUS CONSTITUENTS
interest
11-19 6. In the elemental analysis of soils, minerals, and rocks,
fluorin e, barium , chrom ium,
occasionally arises in the elements zirconium,
vanadium, wolfram, chlorine, and lithium.
11-19 7. Zirconium Detennination. Because of the resista
nce of the
yed as a refer-
mineral zircon (Zr0 2 ) to weathering, Zr is sometimes emplo
the determ ina-
ence element in soil weathering. It has been proposed that
50

conten t
tion of total zirconium is a fairly satisfactory measure of the zircon 51
ed. Dr.
of soil, although weathering of zircon grains has been observ
highly podzo lized horizo ns of certain soils con-
L. D. Swindale found that
of show-
tain only 25 to 80 per cent as much Zr as the parent rock, instead
for a referen ce elemen t. Soil Zr is ef-
ing Zr enrichment as should occur
cence. The sample is freed of
fectively determined by X-ray fluores
52

to pass a 0.1-mm sieve (140

organic matter with H 2 0 2 , dried, and ground
fluorescence
meshes per inch). The Zr is excited by W X-radiation and its
intensi ty of ZrK
spectrum is obtained by a rock salt analyzer crystal. The
for quanti-
radiation appearing at about 15.9 degrees of 2-theta is employed
100 ppm,
tative estimation. The Zr0 2 content in soils and rocks is about
but ranges from about 10 to 800 or more ppm.
15 to 25
11-19 8. Zirconium has also been determined by fusion of
53

pyrosu lfate in a transp arent quartz cru-

mgm of heavy mineral residue in
(as the phosp hate), ignited , and
cible. The Zr was finally precipitated
being necess ary. A Na 0 fusion
weighed as ZrSi0 4 , a 5-place weighing 2 2
the determi-
has also been employed. 54 A series of papers has appeared on
nation of zirconium spectrophotometrically. 55

68 (1945).
50 Haseman and Marshall, Mo. Agr. Exp. Sta. Res. Bui. 387, p.
r.1 Carroll, J. Sed. Petrol., 23: 106 (1953 }.
" 2 Fluorescence analysis of Zr in metals,
Mortimore and Romans, J. Opt. Soc. Am.,
42:673 (1952) ; fluorescence analysis of Zr in minerals, Carl and Campbell, Fluor-
Soc. Testing Materials
escent X-ray Spectrographic Analysi s (Atlantic City, N.J.: Am.
Meeting , 1953).
r.:i Hasem an and Marshall, Mo. Agr. Exp. Sta. Res.
Bul. 387, p. 68 (1945) ; precipita-
tion was also employed by Willard and Hahn, Anal. Chem., 21:293 (1949).
54 Petretic, Anal. Chem., 23: 1183 (1951).
r,r, Green, Anal. Chem., 20:370 (1948)
; Petretic, Anal. Chem., 23:1183 (1951) ;
Kiefer and Boltz, Anal. Chem., 24:542 (1952) ; Wengert, Anal. Chem., 24:1449
berg and Papucci, Anal.
(1952) ; Oesper et al., Anal. Chem., 24:1492 (1952) ; Klingen
Chem., 24: 1861 (1952).
324 MINERA L COLLOIDS, SOILS, MINERALS, AND ROCKS
11-199. Fluoride Determination. In connection with the determination
of fluoride, attention is called to the method of Rowley et al. 56 in which
marked improvements are claimed over the classical distillation method of
Willard and Winter. 57 Jeffries 58 determined F in limestone. A number of
methods are offered for rapid determination of fluoride once it is in solu-
tion. 5H
11-200. Other Determinations. Barium determination" 0 is important be-
cause Ba occasionally contributes to infertility of soil. The determination
of chromium and vanadium, 01 and of wolfram 112 (tungsten) in soils is some-
times of interest. The chlorine in silicate rocks can be determined. rm Lithium
determination° 4 is sometimes of interest in connection with rock and min-
eral analysis because it substitutes structurally for Mg. The separation and
determination of alkali metals has been studied for silicates.u 5 Separation
of the alkali metals derived from insoluble silicates has been accomplished
by means of ion exchange chromatography. rin

QUESTIO NS
I. When the sample weighed out in an air-dry condition for elemental analy-
sis, how is the oven-dry basis obtained?
2. What are the limitations to interconver sion between the oven-dry basis
and the ignited weight basis for reporting the elemental analysis?
3. List 4 types of system for elemental analysis of soils, and briefly state the
advantages and disadvantag es of each.
4. Outline the steps in cleaning and burnishing platinum utensils. and make
a list of the most important commonly used chemicals that attack platinum
utensils.
5. What elements are determined on the sample decompose d by HF in each
of the 2 systems for silicate analysis? What are the disadvantag es of the deter-
mination of Mg and Al on this sample in the semimicroc hemical system?
6. What elements are determined on the sample decompose d by Na 2 COa in

r.ts Anal. Chem., 25: 1061 ( 1953).

~.1 Ind. Eng. Chem., A.E., 5:7 (1953); employed with rock material by
Olsen, Sci.,
112:620 (1950).
C.8 Soil Sci., 71 :287 (1951 ).
Ml Hill and Reynolds, Anal. Chem., 22:448 (1950); Thrum, Anal. Chem.,
22:918
(1950); Willard and Horton, Anal. Chem., 22: 1190, 1194 ( 1950); Horton et al.,
Anal. Chem., 24:548 (1952); Willard and Horton, Anal. Chem., 24:862 (1952);
Price and Walker, Anal. Chem., 24: 1593 (1952); Bum~tead and Wells, Anal. Chem.,
24:1595 (1952); Miller and Phillips, Anal. Chem., 25:172 (1953); Powell and Saylor,
Anal. Chem., 25:960 (1953).
uo Robinson et al., U.S.D.A. Tech. Bui. 1013 ( 1950).
Ill Groves, Silicate Analysis (London: George Allen & Unwin, Ltd.,
1951 ).
Ii:! Ward, U.S. Geol. Surv., Cir. 119 ( 1951).
G~ Kuroda and Sandell, Anal. Chem., 22: 1144 (1950).
G4 White et al., Anal. Chem., 23 :478 (1951 ).
65 Elving and Chao, Anal. Chem., 21:507 (1949).
66 Sweet, et al., Anal. Chem., 24:952 ( 1952).
 325
MINER AL COLLOIDS, SOILS, MINERALS, AND ROCKS
nation of Fe
each of the 2 systems? What are the disadvantages of the determi
and Ti on this sample in the semimi crochem ical system?
following
7. State the principles utilized for determi nation of each of the
s in the semimi crochem ical silicate analysis system: K, Na, Ca, Mg, Fe,
element
Ti, Si, Al.
its separa-
8. What reagent is employ ed for the dehydra tion of silica prior to
tion in an acid solution ?
of alumi-
9. State qualitatively the reactions employ ed for the purification
num prior to its determi nation with aurin tricarbo xylic acid.
sample with
10. What advantages may be gained by the pretrea tment of a
aqua regia prior to the sodium carbona te fusion?
reagent over
11. What are the advantages of HC10 4 as a silica dehydra ting
the HCI system of dehydra tion of silica?
12. List the steps by which SiO~ is determi ned gravimetrically.
cupferr on.
13. List the elements precipit ated and those kept in solution by the
precaut ions are required to get all of the P
By the NH 4 0H separation. What
and no Ca into the latter precipit ate?
r. Why
14. List the steps by which iron is determi ned with a Jones reducto
does Ti not interfer e with this determi nation of Fe?
nation of
15. What oxidizing reagents are commo nly used for the determi
calcium as the oxalate?
nations ?
I 6. In what form is magnesium precipit ated for gravimetric determi
In what form is it weighed?
a manner
I 7. In what valence form does manganese react chemically in
similar to magnesium?
convent ional
18. Why is total K or Na determi ned on a separate sample in the
system of silicate analysis ?
e the essen-
19. Why is ferrous iron determi ned on a separate sample? Describ
tial features of the procedu re.
gravime tric
20. Into what chemical combin ation is sulfate convert ed for its
determi nation?
21. How can zirconium of soils be determi ned?
 12
Plant Tissue Analysis
-Mineral Constituents
The analysis of the soil hy means of the plant
-HALL 1 (ROTHAMS TED)

12-1. The extraction of mineral nutrients from soil by growing crops is

a unique type for soil chemical analysis. Bradfield 2 states, "The founda-
tions of agricultural chemistry . . . had to wait the development of
methods for determining the composition of plants . . . . " Plant tissue
analysis aids in the characterization of soil chemical properties in terms of
soil fertility and mineral nutrition of plants. Mineral is employed for pres-
ent purposes to cover the mineral elements in plants. A wet-oxidation pro-
cedure is given which results in the conversion of the elements Na, K, Ca,
Mg, Mn, Fe, P, Cu, Zn and others to proper form for analytical determina-
tion. Attention is also given to total ash, dry ashing, and sulfur, cyanide,
and fluorine contents.
12-2. Chemical testing of plant tissue sap for mineral nutrient levels,
while similar in objective to mineral analysis in plant ash, is entirely dif-
ferent in procedure and is therefore considered separately, in the next chap-
ter. The extraction and determination of total nitrogen of plants is described
in Chapter 8; of total boron, in Chapter 14; and of total molybdenum, in
~ 15-104.
12-3. Preparation of Plant Tissue Sample. Freshly taken plant tissue is
dried in air or in an oven at 60° to 80°C, in either case protected carefully
from fumes that would lead to contamination. Rapid drying by forced

1 /. Agr. Sci., I :65 ( 1905).

2 Moulton, ed., Liebig and after Liebig (Washington, D.C.: Am. Assoc. Adv. Sci.,
1942), p. 49.
326
 327
PLAN T TISSU E ANAL YSIS- MINE RAL CONS TITUE NTS
bags is
ventilation over thin layers of tissue contained in mesh trays or
to avoid the growth of molds or to mini-
provided. Speed in drying helps
mize loss of weight by enzymatic action in the tissue.
, so
12-4. All but the smaller sizes of plants and seeds must be ground
analysi s. Contam ination with
that suitable sample sizes may be obtained for
sam-
the major elements is ordinarily negligible from grinding mills. Large
ground through a 30-
ples such as corn stover or large forage samples are
hp, manufa ctured
cm diameter hamme r mill (for example, Bell No. 10, 2
imately
by C. S. Bell Co., Hillsboro, Ohio), and quartered down to approx
effected
I-liter bulk volume. Furthe r grinding to pass 0.3- to 0.5-mm sieve is
lphia,
by mortar and pestle, Wiley mill (Arthu r H. Thomas Co., Philade
d), or
Pa.), C and N mill (Christy and Norris, Ltd., Chelmsford, Englan
similar apparatus.
by
12-5. Considerable minor .element contamination is likely to occur
g mill. For this reason careful study
abrasion of materials from the grindin
ined.
is given to this problem in relation to each element that is to be determ
appreci able contam ination " of
Mortar grinding by hand resulted:i in "no
The minor elemen t
Fe, Zn, Cu, B, Co, Mn, Mo, Ca, Na, Mg, P, S, or K.
mills and still
contamination was appreciable with the above mentioned
Menzel
worse with ball mills operate d with various types of balls. Dr. R. G.
stainles s steel s~een for the Wiley mill accepta ble
found a specially made
and zinc determ ination s, but contam ination
for grinding samples for copper
with iron was high.
bri-
12-6. For analysis of plant sap, and for radiochemical assay of
by means of the Carver
quetted plant tissue, the plant tissue is pressed
Summit,
Labora tory Press (Fred S. Carver, lnc., One Chatham Road,
N.J.) with suitable attachments.
tissue
12-7. Plant Tissue Sample Size for Analysis. The size of plant
s is determ ined by (a) the size necessa ry to be
sample needed for analysi
by fineness of grindin g (~I 12-4), ( b) the con-
representative as determined
and ( c) the sensitiv ity of the method
centration of the element in the plant,
m
of determination to be employed. From the first consideration, the minimu
gm, and usually not less
size of plant tissue sample is ordinarily about 0.2
kept as
than a 1-gm sample is taken. The plant tissue sample for analysis is
t to use
small as practicable to conserve time and reagents. It is more efficien
interest ,
a single sample for the determination of the several elements of
t. De-
rather than the prepara tion of a separate sample for each elemen
or radical s that arc lost during the normal prepara -
terminations of elements
separat e sample s, for exampl e
tion of tissue for mineral analysis require
determinations of sulfur, cyanide, and fluoride.
The
12-8. Concentration of Mineral Elements in Crop Plant Tissue.

3 Hood et al., Ind. Eng. Chem., A.E., 16:202 ( 1944).

328 PLANT TISSUE ANALYS IS-MINE RAL CONSTIT UENTS
concentrat ion of each element in a given type of plant tissue varies greatly
with the time of sampling and type of environment. 4 The typical (~ 12-11)
concentrat ion range of mineral elements of several kinds of crop plant
tissue (Table 12-1) are expressed as percentage or ppm of the dried tissue.
12-9. Another useful method of expression of concentrat ion is as per
100 gm of tissue. Concentration of a mineral element multiplied by the
crop yield gives the content of element per crop yield unit, representative
values for which are given by Romaine 5 for field, fruit, and vegetable crops.
12-10. The sensitivity range of the method to be employed is ascertained
and the size of sample is selected accordingly. For example, for P only 0.2
to 1 gm is employed. For P, K, Ca, and Mg from 2 to 5 gm is taken. For
Cu and Zn, 2 to 5 gm is sufficient for polarograp hic determinat ion. Up to
10, 20, or more gm may be prepared satisfactorily by the wet oxidation
procedure given, when such a large sample is required.
12-11. Interpretation of Plant Tissue Analyses. The subject of inter-
pretation of plant tissue analyses has been summarize d well by Ulrich.
8

Occurrenc e of a characteristic content of each element in a given plant

species was early recognized. 7 But the characteris tic content was found to
vary according to the activity of the respective element in the soil, and
according to age of the tissue, cul5ural conditions, and climate. The plant
content of a nutrient has been correlated with soil content,~ with yield, and
9

with various nutrition values and quality factors such as carotene.

12-12. Highest yields are associated with the correct balance of the
nutrient elements in the tissue, coupled with sufficient magnitude of the
content of all nutrient elements. 10 This sufficient magnitude has been termed
sufficient "intensity of nutrition," but may be interpreted equally well as
sufficient activity of the respective nutrients in soil. There is a minimum
plant content of a nutrient for any growth at all, a low range over which
yields increase rapidly with little increase (with some elements there may
be a decrease 11 ) in content, an intermediate range over which both content
and yield increase fairly rapidly ("poverty adjustmen t" 12 ), and a high

4Beeson, U.S.D.A. Misc. Pub. 369 (1941); Lindsey et al., Mass. Agr. Exp. Sta.
Bui. (1919); Lucas et al .. Purdue Univ. Agr. Exp. Stu. Bui. 468 (1942).
"Better Crops, 24, 3:6 (1940).
6 Diagnostic Techniques for Soils and Crops (Washingto n, D.C.: American Potash
Institute, 1948), p. 157.
7 Hall, J. Agr. Sci., I :65 ( 1905).
8 Attoe, J. Am. Soc. Agron., 38: 194 (1946); Seay et al., S.S.S.A.
Proc., 14:245
(1950).
!l Lundegardh , tr. Mitchell, Leaf Analysis (London: Hilger & Watts,
Ltd., 1951 ).
10 Lagatu and Maume, Ann. eco/e nat. agr. Montpel/ier , 22:257 (1934);
Thomas,
Plant Physio/., 12:571 (1937) and Soil Sci., 59:353 (1945); Shear et al., A.S.H.S.
Proc., 47:239 (1946).
11 Steenbjerg, Plant and Soil, 3: 97 (1951).
12 Macy, Plant Physiol., 11: 749 (1936).
 TABLE 12-1
60° or 80°C basis
Concentration* of some mineral elements in crop plant tissue,
ppm
Percent age
Zn B Co
p K Ca Mg s Fe Mn Cu

Legume s
0.2 -0.4 0.2 130-10 00 10-120 4-15 14-110 4-30 -
Alfalfa 0.1 -0.5 0.5-4.5 0.5 -4.5
1.1 -2.l 0.4 -0.7 0.2 100-13 00 25-540 6-20 24-70 36 -
Clover, red 0.2 -0.3 1.1-3.4
0.7 -2.1 0.1 -0.4 - 100-10 00 50-420 - - - -
Lesped eza 0.1 -0.4 0.6-2.0
0.8-2.4 0.2 -0.3 0.2 -0.3 0.1 -0.2 60-570 20-280 4-12 27-80 2-29 -
Soybea n, seed 0.5 -1.l
Grasses
0.3-0.8 0.02 -0.l 0.07-0. 2 0.08-0. 3 14-350 7-38 6-41 21 2 -
Barley, grain 0.15-0. 6
- - - ·-
Barley, straw I
0.04-0. 6 1.1-2.0 -
0.1 -0.4 0.2
- 0.08-0. 2
0.1
-
60-430
7
30-220 7-14 20-90 - -
Blue grass (Ky.) 0.16-0. 4
· 0.2 -0.8
1.3-2.9
0.2-0.9 0.006-0 .05 0.1 -0.3 0.04-0. 3 25-50 5-19 4-17 20 - -
Corn, grain 160-19 0 50-270 2-9 5-80 - 0.01
0.04-0. 4 0.3-1.9 0.1 -0.8 0.1 -0.5 0.08-0. 3
Corn, stover
- 8-17f - - -
\H
N Corn, leaves 0.05-0. 2 0.2-1.0 0.1 -0.9 - - -
20-80 4-51 20 - -
ID
0.05 -0.2 0.06-0. 3 0.07-0. 3 7-350
Oats, grain 0.15-0. 5 0.3-0.7
0.6-3.5 0.15 -0.7 0.06-0. 5 0.09-0. 5 60-370 4-1660 3-54 4-200 - -
Oats, straw 0.02-0. 4
Vegetables
O.o7 -0.3 0.1 -0.3 0.06-0. 6 80-210 14-36 7-16 23-56 - 0.01
Beans, seed 0.3 -0.8 1.1-1.6
0.7-4.1 0.09 -0.4 0.01-0. 5 0.12-0. 2 I 70-280 20-100 6-27 25-69 - -
Beets, roots 0.1 -0.6 0 0'.)_0 LI. 0 R -1 Q , 11-~00 '\_LI.LIO ~-?R - ~7 00'7
01 -OR 1 _Q 0 LI. _l R
Cahhae e 4-94 1-26 11-14 2-16 0.06
0.02 -0.1 0.05-0. 2 0.06-0. 4 7-360
Potato, tuber ! 0.1 -0.5 1.0-4.0
Fruits
0.1 -0.2 1 -1.5 1+ 0.2 - 40-350 20-170 - 14-55
Apple, JeaH 0.03-0. 1 20-40 1-22 5-7 3-9 3-80
0.02-0. 1 0.5-1.4 0.02 -0.1 0.02-0. 06
Apple, fruit
0.4-2.8 0.9 -3.0 0.4 -1.2 - I 19-740 45-280 5-200 - 10-160
Cherry , leaf§ 0.1 -0.7 4-30** 6-40 12-110
Peach, lean I 0.1 -0.5 0.8-2.4 1.1 -2.7 0.4 -1.4 - I 40-540 17-220

. Misc. Pub. 369 ( 1941 ), except as indicated .

0 Analyses rounded from compilat ion of Beeson, U.S.D.A
t Pack et al., Soil Sci., 75: 433 ( 1945).
p. 35 ( 1949 ).
t Walrath , Vermont State Hort. Soc. Proc., 53rd meeting, ), except as indicated .
§ Kentwor thy, Proc. Am. Soc. Hort. Sci., 55:41 ( 1950
• 0 Zubriski, Ph.D. Thesis, Universi
ty of Wiscons in, p. 85 ( 1950 ).
330 PLANT TISSUE ANALYSIS -MINERAL CONSTITU ENTS
range over which yields increase little while the content increases rapidly
("luxury consumption "). The critical percentage 13 of a nutrient element in
a plant is that "below which the yield is progressively reduced by deficiency
of that nutrient and above which only moderate yield response is expected
to further application of that nutrient," thus corresponding to the transition
between the intermediate and high range.

TOTAL ASH CONTENT OF PLANT TISSUE

12-13. Total ash content of plant tissue is often determined as a meas-
urement distinct from ashing for the determination of individual mineral
elements. Volatilization of constituents during dry ashing is less of a prob-
lem for total ash than for elemental analysis of the ash, but volatilization
error may be appreciable even for total ash as shown by Liebenthal1 4 for Si.

APPARATUS

12-14. Needed apparatus consists of a porcelain or platinum crucible, a

Meker burner, a 100°C oven, a muffle furnace with temperature regulator,
a steam plate, and an analytical balance.

REAGENTS

12-15. Required reagents include pure olive oil and distilled water.

PROCEDURE tr.

12-16. A 3- to 5-gm (weighed to the nearest mgm) sample of the plant

tissue, collected so as to be as free as possible of dust, is placed in a weighed
porcelain or platinum crucible, dried at 100°C for I 0 hours or overnight,
and weighed. This weight is the base.
12-17. Approximately 2 ml of pure olive oil is added to the tissue and
the crucible is heated slowly until smoking ceases over a Meker burner
(hood) at very low flame to prevent ignition. The crucible is then placed
in a (preferably cool) muffle furnace and heated to about 525°C for 45
minutes, or until the ash is nearly white.
12-18. The crucible is removed and cooled, and the ash is moistened
with water from a fine-jet wash bottle to dissolve occluding salts. The ash
is then dried on a steam plate and reheated in the muffle at about 525°C to
constant weight (I hour usually suffices). The crucible is cooled in a desic-
cator, weighed to the nearest mgm, then reheated, cooled, and weighed to
check constancy of weight.

13Jackson et al., S.S.S.A. Proc., 12:282 (1948); Terminology Committee, S.S.S.A.

Proc., 15:429 (1951).
14 Sci., 114:636 (1951 ).
rn The essential features of this procedure are in accord with Methods of Analysis,
A.O.A.C., Ed. 6 (1945).
PLANT TISSUE ANAL YSIS-M INERA L CONST ITUEN TS 331

WET OXIDATION OF PLANT TISSUE

12-19. Oxidation of the organic matter of plant tissue and release of the
mineral elements such as Ca, Mg, and P may be effected through either wet
oxidation by means of oxidizing acids such as the HN0 3-H 2S04-HC10 4
ternary acid mixture employed in the procedure to be described herein or
by dry ashing (considered in the alternative procedure, ~ 12-32). Potas-
sium and Na may be extracted directly from plant tissue (without oxidation
of the organic matter) by means of appropriate salt solutions (~ 18-32).
12-20. Wet oxidation with HC10 4 avoids the loss of K through volatil-
ization16 and gives a clear solution of all constituents 17 except Si, which is
quantitatively dehydrated and precipitated and is removable by filtration or
centrifugation. The resultant solution is ideal for analysis of both the major
and minor elements. Parks et al. 18 removed the Si by HF treatment after
HNOa- HCl0 4 digestion. It has been reported 19 that H:1P0 4 can be lost dur-
ing HCl0 4 digestion if excessive temperature (much over 200°C) is used.
In the proposed procedure, recovery of phosphorus is complete up to a
temperature of 230°C. Danger of explosion with perchloric acid is com-
pletely overcome by (a) predigestion with concentrated HN0:1, ( b) in-
clusion20 of H 2 S0 4 in the HC10 4 -HNOa solution used to complete the oxi-
dation, and (c) exhaustion of the HC10 4 fumes through a water pump
according to the arrangements in Fig. 12-1 to prevent their accumulation
in the hood ventilation system.
12-21. In the HC10 4 procedure presented, the proportion of acid to
tissue weight is carefully regulated for maximum economy of digestion and
evaporation times as well as for economy of reagents.
12-22. The procedure employed departs from the usual practice of slow
heating of the digestion mix containing HNOa, H 2 S0 4 , and HC10 4 from
room temperature. The mix is brought to a temperature of 180° to 200°C
rapidly. The power of oxidation of the HC10 4 appears to increase more
rapidly than the rate of evaporation, and consequently the efficiency of
oxidation of the organic constituents is increased at higher temperatures.
Smith21 states that HC10 4 on heating yields anhydrous HCI0 4 , which
further dissociates into nascent chlorine and oxygen. At a temperature of

rn St. John and Midgley, Ind. Eng. Chem., A.E., 14:301 (1942).
17 Gieseking et al., Ind. Eng. Chem., A.E., 7:185 (1935); Toth et al.. Soil Sci.,
66:459 (1948).
18 Anal. Chem., 15:527 (1943).
D. Van
rn Snell and Snell, Colorimetric Methods of Analysis, Vol. 1 (New York:
Nostrand Company , Inc., 1936), p. 497.
1944),
20 Piper, Soil and Plant Analysis (New York: Interscience, Publishers, Inc., of explo-
p. 272, emphasiz es that the presence of H 2 S04 greatly decreases the danger
c acid and ammoniu m perchlora te at the end of the
sive decomposition of perchlori
digestion.
21 Perchloric Ac;id, 4th ed., Vol. l (Columb us, Ohio: G. Frederick Smith
Chemical
Co., 1940).
332 PLANT TISSUE ANALYS IS-MINE RAL CONSTIT UENTS

12.8 cm

Evacuation tube

nn n
Reflux tube

Fig. 12-1. Apparatus for wet oxidation of organic tissue in HNOa-H 2 S04-HCI04
mixture. Special reflux tuhes and manifold provide for evacuation of HCI04 fumes
through an NaOH trap and their discharge into the sewer line. (Apparatus developed
by author and associates at the University of Wisconsin; available from Erway
Glassblowing, Oregon, Wis.)

203°C, the constant boiling point of the system HC10 4 • H 2 0, the de-
composition of the anhydrous HC10 4 appears to be complete. These nas-
cent oxidants account for the increased oxidation efficiency at the elevated
temperature employed in the digestion procedure.
APPARATUS

12-23. Needed apparatus includes 500-ml conical digestion flasks

( 125-ml, 250-ml, or I -liter flasks may be used), a special digestion mani-
fold shown in Fig. 12-1, an electric hot plate with covered element and
thermostat for heating the digestion flasks (such as the Lindberg, Wilkins-
Anderson, Chicago, Ill.), a sand tray to cover the electric hot plate, and
burets for dispensing the digestion acids. The Kjeldahl digestion apparatus
(Fig. 8-1) with disposition of fumes in water may also be employed for
HC10 4 digestion, particularly for large samples.
REAGENTS

12-24. Needed reagents include concentrated HN0 3 and a ternary solu-

tion of three acids prepared by mixing 100 ml of concentrated HN0:1, 10
ml of concentrated H 2 S04 , and 40 ml of 60 per cent HC10 4 (any quantity
in volume ratio of 10 : 1 : 4) and then allowing to cool before use.
PROCEDURE22

12-25. Precaution. Predigestion of plant tissue in HN0 3 prior to addi-

22 This procedure was developed during the period from 1944 to 1950, by the
author and several of his associates, including Mrs. Pauline Frink of Purdue Univer-
sity, Dr. B. Chatterjee, J. L. Huber, and Dr. L. D. Whittig. Drs. J. A. Kittrick and A.
Kaufman also participated in the design of the aspirator; Dr. J. C. Kaudy aided in
the testing of complete recovery of phosphorus after the wet oxidation procedure;
and Dr. R. G. Menzel assisted in testing the procedure in connection with the recovery
of Cu and Zn.
PLANT TISSUE ANALYSIS-M INERAL CONSTITUENTS 333

tion of HC10 4 is highly important to preclude danger of explosion and fire.

Sixty or more per cent HC10 4 is never added directly to plant tissue without
predigestion in HNO:i· Piper 23 employed a higher proportion of HN0 3 to
HCl0 4 than does the present procedure, and omitted the predigestion with
HN0 3 when H 2S0 4 was included in the digest, except for oily seeds. Ex-
perience in this laboratory indicates the efficacy as well as safety of a HN0 3
predigestion followed by the combined ternary acid mixture. The digestion
procedure thus designed can be carried out rapidly at an elevated tempera-
ture with complete safety and with no exceptions for various types of tissues.
12-26. Predigestion of Small Tissue Samples of 2 gm or less in HN0 3 •
To 2 gm or less of dried and powdered plant tissue in a 500-ml conical
flask, 5 ml of concentrated HN0:1 is added for each gm of plant tissue. The
flask is swirled to moisten the entire mass of tissue and then is placed on a
steam plate for 30 minutes and then on the electric hot plate at 180° to
200°C as measured in a flask of glycerol standing on the hot plate. The
aspirator system (Fig. 12-1 ) is connected to the flask to exhaust the ox-
ides of nitrogen and to condense unreacted acid fumes. The suspension is
boiled until taken nearly to dryness. This predigestion with HN0 3 requires
only about 45 minutes.
12-27. Predigestion of Large Tissue Samples Over 2 gm in HNOa• The
predigestion of large samples is essentially the same as outlined in the previ-
ous paragraph with the following exceptions: (a) glass beads (8 to 10)
are added to alleviate bumping, (b) with extra large samples (20 gm or
more) a 1000-ml conical flask is employed, and (c) the digestion is not
placed on the hot plate, until the preliminary action has subsided, as other-
wise the sample may froth over the top of the flask. The acid is brought into
contact with the tissue by thorough stirring and, if the spontaneous heating
becomes excessive, it is slowed down by immersion of the flask in a water
bath at 25°C. Predigestion of samples as large as 20 gm requires only about
an hour.
12-28. Blank. Blank digestions (in duplicate) are run on the reagents,
added in the same amounts as employed in the determinations. All steps
are carried out parallel to the sample.
12-29. Digestion of samples in HN0 3 -H 2S0 4-HCI0 4 • The digestion
flask and contents are cooled slightly. Then an appropriate amount of the
ternary mixture of acids (HNO:i-H2 S04 -HCI0 4 ) is added, consisting of 5
ml for each gm of tissue up to 2 gm of tissue and 4 ml per additional gm
of tissue. Digestion is carried out at 180° to 200°C until dense white fumes
of H 2S04 and HC104 are evolved. A brown or greenish scum of Mn0 2 may
appear while HC10 4 is present, but this redissolves in the concentrated
H 2 S0 4 at the end of the digestion. The digestion is continued at 180° to
200°C until the a<;id liquid is largely volatilized.

23 Pip~r, Of?· cit,

334 PLANT TISSUE ANALYS IS-MINE RAL CONSTITUENTS
12-30. The digestion mix rarely shows charring near the end of the di-
gestion. Even so, there usually is sufficient HC10 4 remaining to oxidize the
charred material completely in virtually all cases. If the acid liquid turns
brown with carmelized organic matter as the volume becomes low, 5 ml
more of the ternary mixture of acids is added and digestion is continued as
before.
12-31. The digestion process is continued until a clear solution remains
after the acids are largely volatilized. The digestion is stopped when the
residues in the flask are clear and white and only slightly moist with H 2 S04 •
The HCl0 4 , at this point, has been largely removed. The residue is now
ready for analysis for elemental mineral constituents.

ALTERNATIVE PROCEDURES

12-32. Dry Ashing. Dry ashing (ignition) has been a popular and very
satisfactory alternative to wet oxidation of plant tissue for the release of
mineral elements. However, significant amounts of potassium may be vola-
tilized at the usual ignition temperature of 550° to 600°C. Both P and K
may be Jost at ignition temperatures over 600°C. No phosphorus is lost by
volatilization in dry ashing if the ash is alkaline but some may be Jost if the
ash is acid. 24 For this reason, ashing for phosphorus is generally carried out
in the presence of an alcoholic solution of Mg(NO:i) 2 (tJ 12-34) or
Mg ( OAc) 2 the latter sometimes being preferred~~ because the former
causes deftagration. If one of these salts is added, the metallic cations
cannot be determined on the same ash as the phosphorus (tJ 12-34). Ad-
dition of Na 2C0 3 to the tissue serves to retain S and Cl, which are other-
wise volatilized. Occasionally charring occurs with large samples with
resultant incomplete ashing. Also, significant errors may occur in the deter-
mination of phosphorus and the trace elements due to a portion becoming
occluded or insoluble in the ash. Dry ashing is often more time consuming
than wet oxidation. Total ash content of plant tissue is occasionally de-
sired (tJ 12-16). A technique is available 26 for ashing plant tissue in micro-
section for microchemical analysis.
12-33. M. Peech of Cornell University (personal communication) has
devised the following simple procedure for complete ashing of plant mate-
rial at low enough temperature to prevent Joss of volatile mineral constitu-
ents such as K. The sample of plant material is ashed at 400° to 450°C in
a muffie furnace for several hours or overnight. The sample is then cooled
and treated with an excess of 1 N HN0 3 , evaporated to dryness on a hot
plate, and placed back in a muffie furnace at 400°C ± 10° for about 10

24 Piper, op. cit., p. 259.

25 /bid., p. 268.
26 Struckmeyer, Am. J. Bot., 30:477 (1943).
PLANT TISSUE ANALYSIS-M INERAL CONSTITUEN TS 335

minutes. The perfectly clean white ash is then cooled and taken up in ap-
propriate acid.
12-34. Dry Ashing for Phosphoms.27 A 0.5-gm sample of finely ground,
well-mixed, dried plant material is placed in an evaporating dish of ap-
proximately 100-ml capacity. Five ml of 0.5 N Mg(N0 3 ) 2 or Mg(0Ac) 2
and 10 ml of distilled water are added. This mixture is evaporated to dry-
ness on a steam bath and then the dish is allowed to dry. The dish and
contents are placed in a muffle and ignited at about 600°C until the residue
is uniformly gray in color ( 30 minutes is usually sufficient time). The dish
is cooled, and then 10 ml of approximately 2 N H 2 S04 is added. The dish
is rotated to bring the acid in contact with the entire ash. Then 15 ml of
distilled water is added and the dish is placed on the steam bath to evapo-
rate the suspension to a volume of Jess than 5 ml. The dish is removed
from the steam bath and 20 ml of distilled water is added. When the dish
has cooled, the sides are rubbed down with a rubber policeman and the
contents are filtered into a 100-ml volumetric flask. The filter is washed and
then the solution is made to volume and mixed. Aliquots of this solution
are employed for the determination of phosphorus (~ 7-22 or 7-31).
ELEMENTAL ANALYSIS OF RESIDUE FROM WET OXIDATION
OF PLANT TISSUE
12-35. The analytical system for the residue from wet oxidation of plant
tissue must be designed (a) to keep the silica from rehydrating and thus
dispersing, (b) to keep K from precipitation as KCI0 4 , and (c) to dissolve
all of the Ca, which tends to precipitate as CaS0 4 • 2 H~O. The digestion
is carried far enough in the procedure to decrease the HC10 4 sufficiently to
prevent KCl0 4 precipitation in the take-up solution. Use of HCI as the
solvent converts the Ca to soluble form as the chloride salt and the HCl is
used at sufficient concentration to prevent the rehydration of silica.
12-36. Stock Solution of Sample (Solution A). The digestion flask con-
taining the residue from the wet oxidation of plant tissue is cooled and 5 ml
of concentrated HCl is added. The flask is swirled and policed and then the
solution is poured into a 25- or 50-ml calibrated centrifuge tube. Five addi-
tional ml of concentrated HCI is added to the flask, and the flask is rotated to
bring the HCJ into contact with all the inside surfaces. This solution is de-
canted into the same centrifuge tube. Two additional rinsings of the flask are
given with small portions of 6 N HCI, the rinsings being transferred to the
tube each time. Water is not used in the transfer.
12-37. The solution in the centrifuge tube is made to the volumetric
mark with 6 N HCI; the solution is then mixed with an air jet stirring rod,
washing being omitted, and the tube is centrifuged for 5 minutes or until
clear. The clear supernatant HCJ solution is immediately decanted into a
27 Bertramson, Plant Physiol., 17:447 (1942).
336 PLANT TISSUE ANALY SIS-MIN ERAL CONSTI TUENTS
dry flask, stoppered, and labeled Solution A. Washing of the silica is
omitted.
12-38. The analyses of the elements present in the residue from wet
oxidation of plant tissue are done by the same procedures as employed for
the usual soil analyses. Interfering elements present are similar to and
usually less extensive than those commonly encountered in soil analysis.
12-39. Aliquots of Solution A are taken for analysis. Either dilution or
evaporation of an aliquot to dryness permits the control of HCl concen-
tration, for example, prior to development of the vanadomolydophosphoric
yellow color(~ 7-56) in HNO:i. A secondary dilution is usually necessary
for phosphorus determination prior to taking the final aliquot for the devel-
opment of the blue color (~ 7-26). Potassium, Na, Ca, Mg, and other
metallic cations are readily determined by flame emission spectrophotom-
etry (~I 18-7). Potassium can be determined by the cobaltinitrite pro-
cedure (~I 6-12), and the others by usual procedu res-Ca (~I 5-30), Mg
(~ 5-48), and Mn(~ 5-71). Sodium is best determined by
direct extrac-
tion (~I 12-40) rather than in the wet oxidation digest, to insure freedom
as directed in ~ 11-70,
from contamination. Iron is readily determined
15-13 or 7-1 11. Copper and zinc are readily determined polarogra phically
( ~ 16-39) or by absorption spectrophotometry ( ~ 15-47, 15-69).

ALTERNATIVE PROCEDURES

12-40. A determination of total K or Na in plant tissue or manure with-

out the oxidation of the organic matter can be effected by direct extraction
with NH 4 0Ac (~ 18-32). Moreover, plant tissue samples may be formed
into pellets and ignited directly in the emission spectrograph ( ~ 18-5 8) for
determination of the elements present without any form of preparation ex-
cept drying and grinding. The determination of an individual element such
as phosphorus is often wanted on plant material; either wet oxidation as
given above or dry ashing ( ~ 12-34) is satisfactory.

SULFUR CONTENT OF PLANT TISSUE

12-41. Sulfur in organic combination is volatilized 28 during the dry
ashing process, and H 2 S0 4 has been introduced in the wet oxidation pro-
cedure given ( ~ 12-29). Thus a separate preparati on must be employed for
the sulfur determination in plant tissue. Sulfur contents obtained by
29

peroxide fusion methods or wet oxidation methods without H 2 SO 4 were 2 to

100 times higher than obtained by the old dry ashing method. The organic
sulfur content of plant tissue has been reported:m to range from 0.1 to 0. 7

28 Hart and Peterson, Wis. Agr. Exp. Sta. Res. Bui. 14 (1911 ).
29 Halverson, J. Am. Chem. Soc., 41: 1494 (1919).
ao Thomas et al., Soil Sci., 70:9 ( 1950).
PLANT TISSUE ANALYSIS-MINERAL CONSTITUENTS 337
per cent though total sulfur by the bomb procedure ranges on up to nearly
3 per cent.
12-42. Total sulfur analyses are usually made by ignition of the plant
sample in a bomb and determination by precipitation with BaS0 4 , but care
must be employed not to air-dry samples that contain appreciable contents
of volatile sulfur compounds. A bomb procedure is also given by Piper. 31
Johnson and Nishita 32 prepared plant samples by the A.0.A.C. procedure
of dry ashing in the presence of Mg(NO:i) 2 dissolved in 95 per cent
ethanol for the determination of micro quantities of sulfur by their special
procedure.

CYANIDE IN PLANT TISSUE

12-43. The determination of cyanide and cyanide-producing substances
in plants such as Sudan grass and sorghums is of interest to the soil chemist,
for when the soil phosphorus activity is too low in the presence of adequate
nitrogen, compounds that hydrolyze to form cyanide may accumulate, ac-
cording to Boyd et a1.a:i A brief outline of the procedure of Boyd et al.
(developed by Boyd and Truog) is given herewith, because of the im-
portance and efficacy of the procedure, although CN- is outside of the
usual determination of mineral constituents of plants. An 8-gm sample of
fresh, green tissue (air-dried tissue may be employed also) of Sudan grass
or sorghum, previously cut into 8-mm lengths with a razor blade or knife
and mixed thoroughly, is placed in a 800-ml long neck Kjeldahl flask and
then 250 ml of distilled water and 5 ml of chloroform are added. Steam is
passed through this suspension of tissue to steam-distill the HCN through a
condenser into a 100-ml test tube containing 5 ml of 2 per cent KOH. The
delivery end of the condenser is kept below the surface of the KOH solu-
tion. Approximately 60 ml of distillate is collected and the distillation is
then stopped. A 5-ml aliquot of the well-mixed distillate is pipetted into a
colorimeter tube. To this 5-ml aliquot is added 5-ml of an alkaline picrate
solution (50 gm of Na 2 C0 3 and 5 gm of picric acid in 1000 ml of distilled
water, brought nearly to boiling to obtain dissolution), the contents of the
tube are mixed and digested in a boiling water bath for 5 minutes. The
color is determined with a colorimeter with a 520-mu light maximum or by
visual comparison to a series of standards (0.241 gm of KCN in 1000 ml
of water gives 0.1 mgm of HCN per ml). The range of the method is from
5 to 500 ugm of HCN per determination.
12-44. The results are expressed as mgm of HCN per 100 gm of oven-
dry tissue. Under 25 mgm is considered safe, whereas from 75 to 100 mgm
is considered dangerous.

a1 Piper, op. cit., p. 299.

32 Anal. Chem., 24:736 (1952).
sa /.Am. Soc. Agron., 30:569 ( 1938).
 AL CONSTITUENTS
338 PLA NT TISSUE AN ALY SIS -MI NER

FLU ORI DE IN PLA NT TISSUE

t tissue is often wanted. Attention
12-4 5. Fluoride determination in plan
34 in which marked improvements
is called to the method of Rowley et al.
35 based on the Willard and Winter
311
are claimed over the A.O.A.C. method fluorine in plan t
rt determinations of
distillation. Remmert et a1.,:i 7 also repo
materials.
QUESTIONS
are
tissue samples very soon after they
1. Why is it necessary to dry plan t
take n? e
are available for grinding plan t tissu
2. Wha t types of mills and mor tars of grind ing meth od
how is the choice
dow n to suitable size for analysis and
cont amin ation of the samp le?
related to possible
dete rmin e the size of plan t tissue sample emp loye d for analy-
3. Wha t facto rs
sis of the mine rals pres ent?
ral elements variable in one type of
4. To wha t exte nt is the cont ent of mine term s
how is this varia tion relat ed to the
tissue of a given plan t species and " and "crit ical perc enta ge"?
tion,
"pov erty adju stme nt," "lux ury cons ump ts
esse ntial featu res of the dete rmin ation of total ash of plan
5. D~scribe the
and give the reasons for each main step.
oxid ation meth od for plan t tissue.
6. State the chie f adva ntag e of the wet
of H 2S0 4 with the HC1 0 4 and HN 03
7. Wha t is the purp ose of the inclusion
of plan t tissue?
in conn ectio n with the wet oxid ation p
8. Why is it impo rtan t to exha ust the HCJ 0.1 fumes thro ugh a wate r pum
usua l fume hood vent ilatio n syste m?
the
and into a sewer rath er than thro ugh prio r
9. Wha t is the purp ose of the pred igestion of plan t tissue with HNO:i
to the addition of HC1 0 4 ? ng
of virtually all of the HC1 0 4 by fumi
10. Wha t is the purp ose of removal are the quan tities of acid s
stion, and why
the H 2S0 4 near the end of the dige
samp le?
adde d acco rdin g to the weight of the e for
11. Why is the temp eratu re held
dow n to 400° C in the Peec h proc edur
emp loye d for facil itatio n
ation system is
dry ashing, and wha t step of a wet oxid
of completeness of ashing?
d to plan t tissue prio r to dry ashing?
12. Why is Mg( OAc h sometimes adde
ed in the residue from wet oxid ation
13. Why may the elem ents be dete rmin d for the usual soil analyses?
ods as emp loye
of plan t tissue by the same meth

Anal. Chem., 25: 1061 (1953). ), p. 398.

ed. (Washington, D.C.: A.0.A.C., 1950
:14
M Meth ods of Anal ysis, 7th
36/n d. Eng. Chem., A.E. , 5:7 (1933).
87 Anal . Chem ., 25:450 (195
3).
 15
Rapid Chemical Tests
of Soils and Plant Tissue
Diagnos is
-SCARS ETH

13-1. Rapid chemical testing of soils and plant tissue is analogous to

clinical testing and diagnosis in medicine. More elaborate analytical pro-
cedures of soil chemical analysis are simplified to be suitable for ready use
by Jess specialized personnel in county agent offices and commercial farm
service laboratories. The "diagnosis" 1 of a particular soil condition, made
under the direction of a skilled agricultural agent and crop production
supervisor, has become as indispensable and as widely required by farmers
as that of the general practitioner in medicine.
13-2. Testing in the Field. Plant sap testing is generally carried out in
the field or greenhouse, frequently with concurrent soil testing. Field tests
make possible a coordinated diagnosis directly in the field by means of
both plant tissue and soil tests. Reagents and apparatus for simple soil and
plant tissue tests are packed into portable kits (~ 13-78) for ready trans-
port to the field. Although field tests are generally less accurate than labora-
tory tests, they are usually adequate in view of the variability of samples
obtained by cursory sampling. Furthermore, testing soils and plant tissue in
the field dramatizes the diagnosis for .students and farmers by providing im-
mediate answers to nutrition problems. In this way, interest is created in
the soil testing service offered in central laboratories. Then systematic soil
testing ( ~ 13-7 5) in spring in preparation for preplanting applications can
be developed in the normal pattern. Rapid soil tests are given below for
1 Kitchen, ed., Diagnostic Techniques of Soils a11d
Crops (Washington, D.C.:
American Potash Institute, 1948).
339
340 RAPID TESTS OF SOILS AND PLANT TISSUE
aeration and porosity ( ~ 13-71), for soil pH ( ~ 13-87) and lime require-
ment, and for readily extractable soil phosphorus ( ~ 13-94), potassium,
magnesium, calcium, nitrate, and ammonium. In saline soils, rapid testing
of chloride, sulfate, and sodium are often important ( ~ 13-109).
13-3. The basis for rapid chemical testing of plant tissue is the colori-
metric determination of the levels of nitrate, phosphorus, and potassium,
(sometimes magnesium and calcium, 2 manganese, 3 zinc, 4 and other ele-
ments), is the sap of fresh plant tissue. Plant tissue tests, though only semi-
quantitative, are highly useful as guides to interpretation of the relative
supply of nutrients actually being taken up by the plant. The underlying
assumption is that an adequate supply of the element is indicated by an
abundance in the plant sap(~ 13-4).
13-4. Interpretation of the Plant Tissue Tests. The presence of an
abundant supply of one nutrient element in the sap may lead to a mistaken
interpretation in plants when the plant growth is held back by a severe
shortage of another. Use of the tests for all 3 major nutrients-nitrogen,
phosphorus, and potassium-is nearly always more indicative than use of
only 1. Two typical situations diagnosed by plant tissue tests are shown in
Fig. 13-1. A marked response to nitrogen applied as a side-dressing was
correctly predicted from the tissue tests (Fig. 13-1, A) coupled with fairly
abundant growth. The need for more than 1 element (Fig. 13-1, C) can
also be predicted by proper interpretation of the tissue tests and the ob-
served generally retarded growth. The general appearance, size, and color
of the plant are an aid in the interpretations. Leaf symptomsfi give an ac-
curate measure of severe deficiencies, but the tissue tests measure declining
supplies before the visual symptoms are evident.
13-5. Tests for available phosphorus(' 13-94) and potassium in the soil
often corroborate tissue tests for these elements but do not always do so.
Factors of depth and function of rooting, controlled by soil aeration and
moisture supply; root pruning through cultivation; and unusually cool tem-
peratures may greatly affect the growth aside from the levels of mineral
nutrient supplies in the soil. The plant roots stay in contact with the soil for
long equilibrium times compared to the soil extraction times employed in
soil testing. Plant roots extend to subsoil layers where available nutrients
may be more abundant than in the surface soil where the soil tests are most
often made. In addition, different plants have widely different capabilities
for utilization of any 1 nutrient from a given soil, whereas soil testing ex-
tractants are designed for general crops and are not generally standardized

2 Mikkelsen and Toth (magnesium), Agron. lour., 41:379 (1941); Cheng and
Bray (calcium and magnesium), Better Crops, 36, I: 13 ( 1952).
a Cook and Lawton, S.S.S.A. Proc., 8: 327 ( 1944).
4 Shaw, Soil Sci., 74:479 (1952).
11 Hambridge, ed., Hunger Signs in Crops (Washington, D.C.: Natl. Fert. Assoc.,
1941 ).
 dlJ
RAPID TESTS OF SOILS AND PLANT TISSUE 341

Adequate[

Doubtful N P K

Deficient
A. Before nitrogen B. After side·dressing
side·dressing severe with nitrogen; the
deficiency of nitrogen nitrogen deficiency
was indicated. was corrected.
Plants growing rapidly on fairly fertile soil.

::::I lh [18 ] Adequate

Doubtful

Deficient l cd_Jj
C. Severe deficiency of D. Phosphate shortage
Deficient

nitrogen and doubtful appeared when nitrogen

potassium supply; and potassium fertilizers
high relative supply were used.
of phosphorus.
Plants having retarded growth on a less fertile soil.

Fig. 13-1. Plant tissue tests indicative of single (A,B) and multiple
(C.D) deficiencies of nutrients. Two situations could be separated by
observation of relative over-all growth rate, and fertilizer applications
could be adjusted accordingly without the losses involved with in-
adequate fertilizer applications.

with respect to these different plant capabilities. 6 All of these factors are to
some extent measured by the tissue tests, since they reflect the relative
supplies of nutrients actually obtained by the plants. Like factory produc-
tion, plant production depends on the raw materials (nutrients) actually
conveyed to the site of manufacture (living cells), and no quantity of ma-
terials stored beyond the conveyor lines (roots) influences current produc-
tion ( ~ I 3-81 ) .
13-6. The laboratory analysis of plant tissue (Chapters 8, 12, 14, and
15 ) , sometimes termed "foliar diagnosis," is similar in aim to rapid plant
tissue testing in the field. The 2 systems differ in that the former determines
the total content of nitrogen, phosphorus, potassium, boron, etc., but the
latter determines only the currently unassimilated forms. The rapidity and
on-the-spot results of field testing of tissue make the system indispensable
in crop nutritional diagnosis. The total percentage of each nutrient ele-

6 Scarseth, Better Crops, 27, 5: 11 (1943).

342 RAPID TESTS OF SOILS AND PLANT TISSUE
ment in plants from elemental analysis usually can be correlated with the
unassimilated form found by tissue testing, and with soil supply. 7
13-7. Concentration Scale. The concentrations of unassimilated nutrient
elements in the sap of plants may be interpreted in a purely relative way
(high, medium, low) for each plant species, or they may be interpreted in
terms of the usual numerical concentration units, such as parts per million
of sap or of tissue. In either case, about 5 categories of concentration are
enough to permit ready comparison; for example, very high, high, medium,
low, and very low. Some tissue testing concentration scales are simplified
more; for example, adequate and not adequate. Hoffer, in the Purdue Uni-
versity bulletin that pioneered rapid testing of plant tissue, showed the rela-
tive abundance of potassium in growing corn stalks by an inverse correla-
tion with the colorimetric test for the accumulation of iron in the nodes
( ~ 13-62). Thus the relative test for potassium was expressed in terms of
the relative test for a physiologically related element; no absolute concen-
tration standards were needed.
13-8. Type of Plant Tissue to Test. The type of plant tissue to test varies
with the plant species and growth stage, as well as the preference of the
operator. Hoffer employed split corn stalks; leaf, stem, and nodal slices
have also been used. Juicy tissue is susceptible to mechanical macerations
followed by filtration, with tests on the clear sap or extract. Carbon black 11
has been employed as a filtering aid in leaf extractions.

NITRATE NITROGEN IN SAP OF PLANT TISSUE 10

13-9. Nitrates accumulate in the sap of grass tissue roughly in proportion
to the supply of soil nitrogen in relation to the rate of its utilization. The
rate of utilization depends mainly on (a) the stage of growth, (b) how
favorable the weather conditions are for growth, and ( c) the relative sup-
ply of other nutrients. If other nutrients are in short supply, the nitrate tests
may be high though the soil supply of available nitrogen may be only mod-
erate or low (Fig. 13-1 ). The nitrate content of the sap varies in different
parts of a plant, and it is necessary to standardize on a favorable part for
the test on each species. The sensitivity of the test can be varied by choice
of different parts of the plant. ,
13-10. When diphenylamine in H 2 S0 4 is dropped on a slice of tissue,
nitrate in the sap reacts to give a blue color, the intensity of which varies
with the nitrate concentration. The advantages of the method are sim-
plicity and rapidity. The disadvantages are corrosiveness of the solution

7 Lundegardh, tr. Mitchell, Leaf Analysis (London: Hilger & Watts, Ltd., 1951).
s Brendler, Sci., 114:61 (1951 ).
9 Daniel and Turk, Mich. Agr. Exp. Sta. Quart. Bui. 32, 2: 199 (1949).
10 Grateful acknowledgement is extended to Dr. L. F. Marriott for his generous
assistance with the sections on plant tissue and soil testing.
RAPID TESTS OF SOILS AND PLANT TISSUE 343
and the possibility that a few substances other than nitrate may give the
test. The blue color develops in the presence of copper, lead, chromium, and
iron, but these substances seldom cause interference in plant tissue testing.

APPARATUS

13-11. Apparatus needed includes a sharp knife, a porcelain spot plate,

and a glass-stoppered dropper bottle.

REAGENTS

13-12. Needed reagents include distilled water for washing the spot
plate and the diphenylamine reagent, which is prepared as follows: 1 gm
of diphenylamine powder is dissolved in 100 ml of concentrated H~S0 4 •
The solution is stored in a glass-stoppered bottle and the portion for use is
placed in a glass-stoppered dropper bottle. The solution becomes discolored
with time or with contact with the rubber bulb, and is replaced with fresh
solution when this occurs. Caution: This reagent is corrosive.

PROCEDURE

13-13. The test is carried out somewhat differently with each crop. Pro-
cedures for corn, oats, soybeans, and grass are given as examples.
13-14. Stalk Test for Nitrate in Corn. The accumulation of nitrate is
usually greatest at the base of the stalk and decreases toward the top. Dur-
ing the period of ear development, the nitrate utilization at the ear is so
great that often a high test can be obtained just below the ear shank
though a low or deficiency test can be obtained above this point. These
facts should be kept in mind in the interpretation of the test results. When
destruction of the stalk is not desirable, the test can be made on a thin
vertical slice of nodal tissue cut with a clean sharp knife. A drop or 2 of the
diphenylamine solution is placed on a freshly cut surface. The immediate
development of a dark blue color indicates an abundant supply of nitrate
present. A lack of blue color (brown color arising from tissue char by the
acid is ignored) indicates no accumulation of nitrate, which usually is due
to a nitrogen deficiency condition. Variation of intensity of the blue color
ranges between these extremes, depending on the amount of nitrate present.
Since the nitrate varies with height in the stalk, tests can be made at dif-
ferent nodes to determine the highest level at which it is present, as well as
of the relative amount present. The base of the leaf midrib can also be used
as the test site, the conducting tissues being cut open and the test solution
being applied as indicated above. When destruction of a stalk is not a criti-
cal factor, a freshly split stalk provides a very satisfactory means of check-
ing the amount of nitrate present from base to tassel. The test should be
made on at least 6 or 8 representative stalks within the area checked so the
proper interpretation of the test for a particular soil treatment can be made.
 Fig. 13-2. N-P-K plant test kit by which nitrogen, phosphorus, and potassium
status of plants can be determined in a few minutes and at low ~ost. Special filter
paper color strips (lower) onto which the plant sap is squeezed eliminate filtration,
glassware washing. (Available from the Urbana Laboratories, 406 N. Lincoln Ave.,
Urbana, Ill; Photos courtesy J. N. Bray.)

345
346 RAPID TESTS OF SOILS AND PLANT TISSUE

separately with a portion of the BaS04 • Then all components including the
remaining BaSO4 and citric acid are mixed together thoroughly. Extreme
care is exercised to have the room, table tops, and equipment free of nitrate
and nitrite. The powder is stored in a bottle painted black since light af-
fects the alpha-naphthylamine. A sharp knife, filter paper strips, and a small
pair of pliers are needed for the test.
13-21. The test may be standardized against known amounts of nitrate.
To do this a standard solution is prepared by dissolution of 0.072 gm of
KN0 8 in 100 ml of distilled water. This solution contains 100 ppm of N in
nitrate form. Nitrate-free filter paper is moistened with this solution or
with a solution diluted by a known factor. A thin layer of the white powder
is then applied, and the paper is folded over and squeezed. The intensity of
the pink color after 1 minute measures the concentration of nitrate placed
on the paper.
13-22. Stalk Test for Nitrate in Com. The same principles govern the
portion of the stalk to be tested by the white powder method as for the
diphenylamine method (~ 13-14). The powder test may be made in 1 of
2 ways, either on a filter paper or directly on the stalk. When the filter
paper is used, a diagonal cut is made into the conductive tissues of the
stalk, just deep enough to reach the plant sap, then downward a short dis-
tance so that the filter paper strip can be inserted into the cut. The paper
is pressed against the fresh cut until it is wet with the sap. A small amount
of the powder is sprinkled on the wet paper with a toothpick and then the
paper is folded and pressed together so that the powder becomes wet. The
sap is allowed to act on the powder for 1 minute so that the pink color
develops fully. The intensity of the pink color is a measure of the nitrate
level in the plant. When the powder remains colorless, a deficiency in plant
nitrate is indicated. When the color is pink, adequate nitrates are present.
When the color is dark pink to red, the plant has an excess of nitrates.
13-23. The test may also be made by sprinkling a small amount of the
powder directly into the cut in the stalk and pressing it with the cut sliver
so that the powder is moistened by the sap. Or, the stalk is split through the
entire length for an over-all test. The color is read in the same manner as
when the filter paper is used. The test may also be made at the base of the
midrib of a leaf. A piece of tissue to be tested may be removed so that ad-
ditional sap does not diffuse and bring more nitrate to the area tested.
13-24. Stem Test for Nitrate in Oats. Since it is often difficult to expose
enough sap of an oat straw or leaf to wet the powder directly, it is most
practicable to press a strip of filter paper against a node with a pair of
pliers until the paper is wet with the sap. The paper often becomes stained
green, but by using the side away from the node or by doubling the paper
so that the chlorophyll is filtered out by the first layer of paper, a clean
RAPID TESTS OF SOILS AND PLANT TISSUE 347
surface is available for application of the powder, and the pink color is not
masked by the green. The color is read as in the stalk test.
13-25. Stem Test for Nitrate in Soybeans. In making the test, the stem
is cut diagonally at the point desired and a small amount of the white
powder is applied to the freshly cut surface. The powder must be wet with
the plant sap. The color resulting after about a minute is read as it was for
corn. The test may also be made by the use of filter paper strips as indi-
cated for corn, but the different distribution of nitrate in soybeans (~ 13-
16) must be kept in mind.
13-26. Test for Nitrate in Grass. One measure of finely chopped plant
material, 1 measure of the white powder, and 14 measures of distilled
water are shaken together for 25 to 30 seconds and allowed to settle. The
color is interpreted as for corn.

PHOSPHORUS IN SAP OF PLANT TISSUE

13-27. Phosphorus is absorbed from the soil by plants chiefly in the
inorganic or phosphate form. Within the plant it is changed rapidly into
organic form and utilized to produce new growth, particularly the forma-
tion of new cells. When the phosphorus supply is abundant, the plant is
able to accumulate a small reserve of inorganic phosphates in the plant
juices, and this gives the blue coloration in the test indicating "abundant"
or "adequate" supplies. A high phosphorus test may be found in plants
growing slowly because of limited supplies of nitrogen and/or potassium,
the phosphorus supply being relatively more abundant. Application of the
limiting element or elements will increase the growth, but phosphorus may
then become the limiting factor (Fig. 13-1)
13-28. A low phosphorus test can be corroborated in conjunction with
the nitrate test. Phosphorus-deficient plants usually show a high test for
nitrate and a dark green color because the most limiting growth factor is
phosphorus. Phosphorus-deficient plants may also show a purple coloration
of the main stalk arising from high sugar accumulation because insufficient
phosphorus is present to permit full conversion of the sugar formed in the
leaves into growth and yield production.
13-29. When the molybdatic acid solution is dropped on filter paper wet
with the plant sap (N-P-K test kit, Fig. 13-2) and this is followed by con-
tact with a tin rod, a blue color resulting indicates that a reserve of phos-
phate is present. The advantages of this method are simplicity (little ap-
paratus) and rapidity.

APPARATUS

13-30. Needed apparatus includes a sharp knife, filter paper strips, and
pliers.
348 RAPID TESTS OF SOILS AND PLANT TISSUE
REAGENTS

13-31. Needed reagents includes the acid molybdate solution ("P-K

developer") and the tin rod supplied in the kit (Fig. 13-2) .
PROCEDURE
13-32. The same general portion of the plant must be sampled each time
for comparative work. Plants and the portions of each that may be used
are as follows: corn-sta lk near tassel, stalk or midrib or leaf just below
the lower ear node; oats-up per stem or node; soybean s-upper part of
the stem or petiole of leaf in upper part of the plant.
13-33. The surface of the tissue being tested is sliced so that a freshly
cut surface is exposed. A spot on the filter paper strip is moistened with the
sap by placing the paper in contact with the cut surface. In some cases it
may be necessary to squeeze the paper and stem (or leaf) together gently
to accomplish the wetting. A small drop of the molybdate-acid solution is
applied to the wet spot, care being taken to avoid too much of the solution.
The brightened end of the tin rod is immediately pressed against the spot
for about 10 seconds. The color resulting is read as follows:
No color-pl ant very deficient in phosphorus
Slight blue-pla nt is deficient in phosphorus
Medium blue-pla nt is only slightly deficient in phosphorus
Dark blue-pla nt is adequately supplied with phosphorus
The test is made on 6 or 8 typical stalks in the area so that the representa-
tive level of phosphorus may be determined.
ALTERNATIVE PROCEDURES

13-34. The reaction in the Purdue test12 is similar to that described in

~ 13-33, with the exception that this test is carried out in a solution.
The
most soluble phosphates are extracted with ammoniu m molybda te-acid
solution, and the color is developed by the addition of the stannous oxalate
powder. Needed apparatus consists of a sharp knife or razor blade, %
teaspoon measure, rigid-toothed comb, wood block, and glass vial 11 mm in
diameter and 76 mm high with a 10-ml graduation.
13-35. Needed reagents consist of stannous oxalate powder and 2 con-
centrations of molybdate reagent prepared as follows.
13-36. Molybdate Reagent. Eight gm of ammonium molybdate is dis-
solved in 200 ml of distilled water. To this solution, a mixture of 126 ml
of concentrated hydrochloric acid and 7 4 ml of distilled water are added
slowly with constant stirring. A small portion of this concentrated molyb-
date reagent is diluted with 4 volumes of distilled water just before use
for the phosphate test. The diluted reagent may become unsuited for use

12 Thornton et al., Purdue Univ. Agr. Exp. Sta. Bui. 204 ( 1939).
RAPID TESTS OF SOILS AND PLANT TISSUE 349

after standing a few weeks and is checked by running a standard determina-

tion before it is used on plant material ( ~ 13-42).
13-37. Test for Phosphorus in Com. Leaf blades appear to be the best
part of the corn plant to test during the early stages of growth, and are satis-
factory in the later stages. Stem tissue may also be used when the plants
are larger. Since tissue should be taken from at least 6 or 8 stalks to get
a representative sample from an area, there is less mutilation of the plants
when leaf tissue is used. The leaf just below the lower ear node has been
widely used, throughout the season. If stem tissue is used, the portion
selected for the test is the section just below the growing tip or below the
tassel in older stalks. Tissue from several typical stalks should be com-
bined as a representative sample for an area.
13-38. The tissue must be cut into very small uniform sections to obtain
comparable results since the sap of the ruptured cells is essentially the solu-
tion tested. Uniformity of the sections also aids in the measurement of uni-
form quantities of tissue. To accomplish this, leaf blades with the midrib
removed are superimposed one upon the other on a wood block. Then,
with a rigid-toothed comb as a guide, they are cut lengthwise and crosswise
with a razor blade or thin sharp knife. A similar method may be used for
the stem tissue.
13-39. Next, 1;4 teaspoon of the finely cut tissue is placed in a fiat-
bottomed glass vial. The vial is filled to the 10-ml mark with the diluted
molybdate reagent, stoppered, and shaken vigorously for I minute. For
comparable results, this time must be observed carefully, since this extrac-
tion removes only the most readily soluble phosphates. Then a Heck of
stannous oxalate reagent (about the size of a pin head) is added, the solu-
tion is again shaken, and the color is observed. The resulting color may
be read as follows:
Colorless or yellow-very deficient phosphorus supply
Green or bluish green-deficient phosphorus supply
Light blue-medium phosphorus supply
Medium hlue--adequate phosphorus supply
Dark blue-abundant phosphorus supply
Caution: Arsenic gives the same test as available phosphorus, and therefore
the test as described is not applicable for plants that have been recently
sprayed or dusted with arsenic compounds.
13-40. Test for Phosphorus in Oats and Other Grasses. Similarly located
leaves from each of several plants are combined and finely chopped as in-
dicated for corn. The leaf used should be from the upper part of the plant.
One-fourth teaspoon of tissue is taken for the test, and the phosphate is
determined by the same procedure as for corn.
13-41. Test for Phosphorus in Soybeans. The terminal leaflet from the
third node below the growing point is taken from each of t 5 to 20 repre-
350 RAPID TESTS OF SOILS AND PLANT TISSUE

sentative plants. The leaflets are superimposed one upon another and cut
as with corn. Then 1,1,i teaspoon of tissue is used for the test and the phos-
phate determined by the same procedure as for corn.
13-42. Phosphorus Standards. A solution is made up by dissolving 0.02
gm of dicalcium phosphate (CaHP0 4 • 2 H 2 0) or 0.016 gm of potassium
dihydrogen phosphate (KH 2 P0 4 ) in 100 ml of 0.75 N HCl. Then the pro-
cedure is followed in the usual way, the volumes of standard phosphorus
solutions, being used in place of the plant tissue, as follows:

Simulated phosphorus Volume of standard

level phosphorus solution
Very deficient phosphoru s supply 0.1 ml
Deficient phosphoru s supply 0.2ml
Medium phosphoru s supply 0.6 ml
Adequate phosphoru s supply 1.5 ml
Abundant phosphoru s supply 3.0 ml

POTASSIUM IN THE SAP OF PLANT TISSUE

13-43. Nearly all of the potassium in plant tissue remains in soluble or
extractable form in the plant sap. If the potassium supply is limited, its
concentration in the sap declines greatly as growth continues, giving con-
centration differences that are easily tested. The potassium, when in de-
ficient supply, tends to be translocated in grasses from older tissue to
younger growing parts. The older grass leaves, near the base of the stem,
then turn brown and die along the margins. Severe potassium deficiency
causes corn ears to be poorly filled at the tip and may cause stunting of the
plants. In legumes, potassium becomes deficient first in the younger leaves
and white or brown necrotic spots appear between the veins. In soybeans,
for example, the lower leaves are nearly always higher in potassium than
the upper leaves.
13-44. The test for potassium in the sap of plant tissue is carried out
with filter paper test strips, 13 or in a suspension of finely cut tissue. Sap
samples may be collected on filter papers in the field, dried and the tests
performed later on a routine basis in the laboratory 14 by means of the
dipicrylamine reagent. 15 Test papers may be prepared with the dipicryla-
mine dried on test spots. 16

APPARATUS

13-45. Needed apparatus consists of a sharp knife and pliers.

13 Hoffer, Better Crops, 29, 4:9 (1945).

14 Richer, Better Crops, 31, I :26 ( 1947).
15 Poluektoff, Mikrochem ie, 14:265 (1933-193 4).
16 Melsted, Better Crops, 34, 1:26 (1950).
RAPID TESTS OF SOILS AND PLANT TISSUE 351

REAGENTS

13-46. Needed reagents consist of the prepared test papers from the
N-P-K test kit (Fig. 13-2) and the acid molybdate P-K developer of the
kit, or the prepared reagents described below.
13-47. In lieu of the P-K developer, the potassium test can be com-
pleted with 0.5 N HCI.
13-48. The test papers may be prepared according to the Melsted 17
procedure. Solution A is prepared by dissolution of 0.60 gm of dipicryla-
mine (hexanitrodiphenylamine) and 0.60 gm of Na 2 C0 3 in 16 ml of dis-
tilled water, the mixture being stirred and brought to a boil to hasten
solution. The mixture is then cooled and filtered through a small filter
paper, and the filter is washed with distilled water. The filtrate is made up
to a volume of 25 ml. For convenience, the mixture may be filtered and
washed directly into a 25-ml graduate.
13-49. Solution B is prepared by dilution of 8 ml of Solution A to 25 ml
with distilled water in a 25-ml graduate.
13-50. Solution C is prepared by dilution of 10 ml of Solution B to 15
ml with distilled water in a 25-ml graduate.
13-51. Preparation of the Test Papers. Whatman No. 1 filter paper is
cut into approximately 2 x 7 cm strips. At 1 end of the filter paper, a
small drop of Solution A is placed to form a test paper spot. Then about
1 cm from this spot, a small drop of Solution B is placed. Finally, about 1
cm from the second spot, a small drop of Solution C is placed. The papers
are allowed to dry in the open air or in a drying oven for 3 to 5 minutes at
85 °C. When dry, the test spots should vary in color from a deep orange
color for the first (Solution A) spot to a light orange color for the last
(Solution C) spot. It is recommended that the prepared papers be used
within a year to insure good results. Calibration:
Spot on paper Sensitivity, ppm of Kin solution
Solution A 750-1000
Solution B 2000 or more
Solution C 3000 or more

PROCEDURE

13-52. The same general portion of the plant must be sampled each time
for comparative results. The portion suggested for use in this test is indi-
cated for some of the crop plants as follows: corn-the base of the midrib
of the leaf at ear level or the stalk at about the same point; oats-a node
near the middle or upper part of the plant; soybeans-the enlarged base of
the petiole on a leaf from the top of the plant; grass-leaves.

17 Better Crops, 34, 1:26 (1950).

352 RAPI D TESTS OF SOILS AND PLAN T TISSUE
of the leaf.
13-53 . For com, the test paper is placed along the midrib
the other under-
One jaw of the pliers is placed on I of the test spots and
until the plant sap
neath the midrib. The jaws are squeezed together gently
other 2 spots on
wets the spot. The procedure is repeated for each of the
d to react with the
different portions of the same midrib. The sap is allowe
more. Then the spots are wet with one
test spots for about 30 seconds or
(or 0.5 N HC1). The test is positive
or more drops of the P-K developer
persis ts on the test spot area. The
if the reddish-orange or brownish color
w color. The test may be read
test is negative if the spot turns a lemon-yello
as follows:
deficient in potas-
I. If all spots turn lemon yellow, the plants are very
sium and show definite potash deficiency sympt oms.
hut the others are
2. If the end spot ( 1000 ppm) tests a solid orange
deficie nt in potass ium and would prohab ly
yellow, the plants are usually
respond to potass ium treatm ent.
is usually ade-
3. If the first 2 spots test a solid orange, the potassium
grain. soybea ns, pastur e, etc.
quate for field crops, such as corn, small
is in abund ant supply .
4. If all spots test orange, the potassium
g at the concen-
13-54 . The test was developed to give a definite readin
ium in solution
trations indicated. Therefore, as the concentration of potass
Likewise, a strong
approaches the indicated levels, weak tests are obtained.
a positive test. Thus
test indicates a stronger concentration than needed for
of the color of the
some range in interpretation is provided by the intensity
test spots.
in the area so
13-55 . The test should be made on 6 or 8 typical stalks
ined.
that the representative level of potassium may be determ

ALTERNATIVE PROCEDURE
itation with
13-56 . The Purdu e test 18 for potassium involves precip
ity. Neede d appar atus consis ts of a
cobaltinitrite and estimation of turbid
on measu re, a rigid tooth comb, a
sharp knife or razor blade, a h teaspo
1
76 mm high with a 10-ml
wood block, glass vials 1 I mm in diameter and
d cobaltinitritern re-
graduation. Needed reagents consist of the concentrate
o(N0 2 )n and 30 gm of
agent prepared by dissolution of 5 gm of Na:;C
l HOA c is added,
NaN0 2 in 80 ml of distilled water. Then 5 ml of glacia
to stand for several
iind the volume made to 100 ml. The solution is allowed
reagent is -prepare:d
days. Shortly before tests are to be perform~d a dilute
prepared to a solu-
by the addition of 5 ml of the concentrated reagent just
and the pH is a:d-
tion of 15 gm of NaN0 2 in 100 ml of distilled water,
1939).
Thornt on et al., Purdue Univ. Agr. Exp. Sta. Bui. 204 (
18
erably and this test is
Co conten t of comme rcial prepar ations varies consid
rn The tration is an_ im-
's Analyz ed" produc t. The cobalti nitrite concen
based on the "Baker
oortan t factor in determining the sensitiv ity of the test.
RAPID TESTS OF SOILS AND PLANT TISSUE 353
justed to 5.0 with HOAc. Other reagents needed are anhydrous isopropyl
alcohol and 95 per cent ethanol. When ethanol for use as a- reagent is
difficult to obtain, a mixture of 60 parts anhydrous methanol, 40 parts
isopropyl alcohol, and 5 parts of distilled may be substituted. If this mix-
ture turns turbid, it is filtered. Denatured alcohol is not satisfactory.
13-57. Tests for Potassium in Corn. The base of the leaf, near the ear
node or middle of the stalk is most appropriate for the potassium test. The
tissue is first cut into very small uniform sections. To accomplish this, leaf
blades are superimposed upon one another on a wood block and cut length-
wise and crosswise with the blade, a rigid tooth comb being used as a guide.
13-58. Next, t/2 teaspoon of finely cut tissue is measured into a glass
vial. Then 10 ml of the dilute cobaltinitrite reagent is added and the vial is
shaken vigorously for 1 minute. Finally, 5 ml of the 95 per cent ethanol is
added and the suspension is mixed. The yellow coloration of the solution is
approximately constant and is disregarded. The degree of turbidity formed
is an indication of the potassium content:

Only a trace of turbidity Deficient potassium supply

Medium turbidity Doubtful potassium supply
Very high turbidity Adequate potassium supply

13-59. Potassium Standards. A standard potassium solution is made by

dissolution of I gm of KCl in 100 ml of 15 per cent NaN0 2 solution. Then
the procedure is followed in the usual way, with the following volumes of
this solution being employed instead of the plant tissue:

Volume of standard
Simulated potassium level potassium solution
Deficient potassium supply 0.2 ml
Doubtful potassium supply 0.4 ml
Adequate potassium supply 1.0 ml

13-60. Potassium may be tested in small grains such as wheat, oats,

barley, and rye by obtaining tissue at the base of the leaf in the middle of
the main stem. The tissue from alfalfa, clover, and soybeans is taken from
the main stem.
13-61. In a modification for peach tree leaves, 20 1 teaspoon of finely cut
green tissue taken in mid season, Hi teaspoon of DARCO G60 activated
carbon, and IO ml of 15 per cent NaN0 2 having pH adjusted to 5.0, are
placed in the vial and the suspension is shaken for 1 minute, then filtered.
Next, 2.5 ml of the filtrate is placed in a clean vial. The temperature is held
0

near 20°C (68° ± 3°F) during the precipitation to follow. First, 5 ml of

the dilute cobaltinitrite solution ( ~ 13-56) is thoroughly mixed with the

20 Garrard, Better Crops, 34, 6: 17 ( 1950).

354 RAPID TESTS OF SOILS AND PLAN T TISSUE
the side
solution in the vial. Then 2.5 ml of 95 per cent ethanol is run down
for I
of the tube to obtain a layer on top. This layer is allowed to stand
with a
minute to start precipitation. Finally, the alcohol is mixed in slowly
allowed before the estimat ion of
rotary motion, and 3 minutes are then
test is approx imately 1
turbidity(~ 13-58) . The equivalent for a medium
to 1.5 per cent K in the tissue.
is
13-62. Iron accumulates in the nodes of corn stalks when the plant
be in-
potassium deficient ( ~ 13-7). Iron concentration was shown to
21

nodal, and interno dal tissue. The test

versely related to total K in the leaf,
a few drops of 10 per cent
is made by splitting the stalk and applying
two of
aqueous KSCN solution to the nodal area, followed by a drop or
deficien cy.
6 N HCI. A red color (detecting iron accumulation) denotes K

QUICK SOIL AERA TION TESTS

the
13-63. Oxygen activity in soil is difficult to measure directly, hut
it, and these ions may be determ ined
valence of soluble iron ions reflects
oxygen
easily. 22 The proportion of ferric to ferrous reflects the intensity of
supply.
al
13-64. The direct determination of oxygen activity or oxidation potenti
adapt-
with the platinum electrode is a somewhat longer test, but it is also
with portabl e appara tus. The same dy-
able to rapid field determination
is measur ed by Hoffer test
namic intensity factor of soil oxygen activity
for ferric and ferrous iron, given here.
ries,
13-65. Functions of Soil Oxygen. A few plants, such as rice, cranber
s
weeping willows, and cypress, are able to live with extremely low activitie
above-
of soil oxygen because they can utilize oxygen absorbed in the
it to the roots. 2 =1 In contras t, a high ac-
ground portions and translocate
agricul-
tivity of soil oxygen is required for the satisfactory growth of most
alfalfa. Roots of these plants
tural crops including corn, oats, wheat, and
resultan t liberati on of energy
require free soil oxygen for respiration and
the
used in the uptake of water and nutrient ions. 24 Oxygen activity affects
of the balance
utilization of ammonium 2 " by roots and bears on the effect
of
between the activities of the various nutrients in the media. Uptake
26

in the
applied nitrogen, phosphorus, and potassium is more efficacious
of soil oxygen . 27 Decomposition of organic
presence of adequate activity

21 Hoffer, Purdue Agr. Exp. Sta. Bui. 298 (1930);

Hoffer and Trost, /. Am. Soc.
Agron., 15, 8:323 (1923). •
22 Hoffer, Better Crops, 29, I: 6 (1945); 29, 2: 19
(1945).
23 Cannon , Carnegie Inst. Wash. Pub. 368
(1925).
24 Hoaglan d and Broyer, Plant Physiol., 11: 471
( 1936).
25 Amon, Soil Sci., 44:91 (1937).
20 Shive, Soil Sci., 51 : 445 ( 1941 ) .
27 Lawton, S.S.S.A. Proc., 10:263 ( 1946).
RAPID TESTS OF SOILS AND PLANT TISSUE 355

residues in soils and release of mineral constituents for crop use requires
an abundance of soil oxygen.
APPARATUS

13-66. Needed apparatus consists of a spade for the excavation of the

soil and 3 glass-stoppered dropper bottles.
REAGENTS

13-67. Needed reagents include filter paper, 7 to 10 cm in diameter;

1.2 N HCl (concentrated HCl diluted with 4 parts of water); and the fol-
lowing special reagents.
13-68. Potassium Thiocyanate. Ten gm of KCNS is dissolved in 100 ml
of distilled water.
13-69. Potassium Ferricyanide. One-half gm of K 3 Fe(CN)t; is dissolved
in 100 ml of distilled water.
13-70. Each of the 3 solutions is placed in a separate dropper bottle,
to be ready for quick service.
PROCEOURE2H

13-71. The soil tests for ferric and ferrous iron must be made within 20
or 30 seconds on samples from freshly exposed soil surfaces. The aeration
and attendant chemical reactions change very rapidly. A hole 30- to 50-cm
deep is spaded in the area to be tested. The test is carried out at 10-, 20-,
30-, and possibly 40-cm depths. The test on a sample from one depth is
carried out on a freshly taken sample before sampling at the next depth.
13-72. To make the test, a filter paper is creased along the diameter
(Fig. 13-3, I). Two soil samples (each consisting of a pinch of soil) are
placed on opposite ends of the paper. Two drops of the HCI are added to
each of the soil samples. The paper is then folded over and squeezed tightly
so that the liquid comes through the paper. To the wet areas on the outside
of the paper l drop of the KCNS solution is added to the left wet area, and
one drop of the K3 Fe(CN) 6 solution is added to the right wet area (Fig.
13-3, II). Appearance of a red area from the KCNS treated area indicates
ferric iron (good oxygen supply). Appearance of a blue color from the
K3 Fe ( CN) 6 treated area indicates ferrous iron (poor oxygen supply). A
faint red color, or none at all, generally is found when a blue color appears.
The results are recorded at once. If both ferric and ferrous iron tests are
obtained, the oxygen deficiency is usually not severe.
13-73. The test is then repeated for other soil depths. The importance
of making the test very quickly can be demonstrated by permitting addi-
tional soil samples to be exposed to the air for a few minutes, particularly
in sunlight, and then repeating the test. A negative ferric iron test soon

28 Hoffer, Better Crops, 29, 1: 6 ( 1945).

356 RAPID TESTS OF SbILS AND PLANT TISSUE

II. Wet areas showing through are treated

(a) with KCNS and (b) with K 3 Fe(CN)6
(Filter paper) solutions.

I. Drops of HCI are placed on soil at

(a) and (b), and the paper folded.

a a
III. Red color at (a) IV. Faint red or no color at (a)
indicates ferric iron. and blue color at (b) indicates
(GOOD AERATION) ferrous iron.
(POOR AERATION)

Fig. 13-3. Chemical tests for ferric and ferrous iron in soils, indicative of
oxygen supply. (After Hoffer, Better Crops with Plant Food, 29, I :6, 1945.)

becomes a positive one and shows the rapidity of oxidation once the soil
is exposed. Interesting information on the oxygen distribution in soils can
be noted by making the test for ferric iron in the channels formed by pene-
trating roots of previous crops such as alfalfa or ragweed. Also, the depth
of aeration over tile drains as compared to other areas is of interest. Com-
paction at the plow sole can be noted by tests at that depth and above and
below it. The deficiency of oxygen is most likely to occur during years of
adequate rainfall, except in soils that have been adequately maintained by
the growth of deep-rooted crops.
ALTERNATIVE PROCEDURE

13-74. Hoffer Chalk Test. Hoffer has also proposed a test for soil po-
rosity that gives an indication of oxygen supply. A suspension is prepared of
10 ml of dry powdered precipitated CaC03 in 50 ml of water. This sus-
pension is shaken well just before each use. A core sample of soil is taken
by means of the open faced sampling tube(~ 2-33). The face of the core
RAPID TESTS OF SOILS AND PLANT TISSUE 357

is shaved off with a knife with care not to compress the surface, and then
drops of the CaC03 suspension are placed at different depths along the
core. If the CaC03 particles penetrate into the soil and disappear from the
surface, coarse porosity and satisfactory aeration is indicated. If, on the
other hand, the white particles remain on the surface of the core, fine
porosity and poor oxygen supply are indicated.

SOIL TESTING SYSTEMS

13-75. Soil testing refers to chemical tests 29 that can be made rapidly
and with low cost per determination, as contrasted to conventional methods
of soil chemical analysis, which are more accurate but more time consum-
ing and expensive. Soil testing covers both rapid analysis in the field (~
13-2) and in the laboratory. Testing for soil aeration, being entirely a field
test (not adaptable to laboratory samples) and closely allied to plant tissue
testing for field diagnosis, has been given above ( ~ 13-71 ) .
13-76. Function of Soil Testing. Soil testing serves as a basis for advice
to individual farmers about particular fields ( ~ 13-1 ) , as contrasted to the
conventional more elaborate methods that serve not only for calibration of
the rapid soil tests but also for the determination of general principles
about soil types on the basis of which recommendations of general ap-
plicability are formulated. Soil tests are often designed for general use (~
13-5) without special calibrations for either particular crops or particular
soils. Dr. C. E. Kellogg, in personal conversations with the author, has
emphasized the need for standardization of soil tests according to soil
types as well as according to crops. It is believed that soil-testing practice
must eventually be refined to this point.
13-77. Mass Handling in Soil Testing. Because of the greater rapidity of
rapid microchemical soil testing procedures than the conventional "long
methods," they are generally used for soil testing on a mass scale in central
soil-testing laboratories. Convenient mass-handling techniques have been
described in the literature. 30 Multiple apparatus for shaking, filtration,
pipetting, etc. are aids to mass handling. Time and motion studies assist in
the achievement of high efficiency. A system of rapid microchemical soil
tests designed for a precision approaching that of conventional "long
methods" has been described. 31
13-78. Soil-testing.Kits. A number of commercial soil-testing kits:i~ are

. 20 Comparable to the spot tests of chemistry, Fiegl, tr. J. W. Matthews, Spot Tests
(New York: Nordeman Publishing Company, Inc., 1937); B.D.H., Reagents for Spot
Tests (London: British Drug Houses, 1936).
ao Constable and Miles, J. Am. Soc. Agron.;.33 :623 ( 1941).
:n Peech, Ind. Eng. Chem., A.E., 13:436 (1941 ); Peech and English, Soil Sci.,
57: 167 ( 1944); Peech et al., U.S.D.A. Cir. 757 (1947).
32 A comparison of various systems was ma<:Ie by Anderson .. and Noble, U.S.D.A.
Misc. Pub. 259 (1947).
358 RAPID TESTS OF SOILS AND PLANT TISSUE

available, complete with reagents and apparat us for making certain soil
tests in the field. As examples, the Purdue kit is supplied by Purdue Uni-
versity, Lafayette, Ind.; the Hellige-Truog kit is supplied by Hellige, Inc.,
3718 Northern Blvd., Long Island City I, New York; the Morgan type kit
is supplied by LaMott e Chemical Products, Towson, Md.; and the Spurway
type kit is supplied by the Edwards Laborat ory, Cleveland 11, Ohio.
13-79. Attenda nt Factors in Interpre tation of Soil Tests. Soil tests in
themselves, and apart from knowledge of the field conditions and other
attendan t circumstances, do not provide certain knowledge of proper soil
management recommendations (~ 13-5). The photometer reading cannot
be transmitted to the farmer without the intermediary of human judgement
-balanc ing the test against attendant circumstances. What are the most
importa nt attendan t circumstances? The interpreter of the soil test must
know the soil type and the climatic zone in which it is located. He must
consider the topography of the soil, the type of drainage, the depth and
aeration of the soil, and whether the subsoil is a potential source of nutri-
ents. :i:i The history of liming and fertilization must be known, since this
provides a clue to the yield level to be expected. The yield history permits
comparison to yields on similar soils and aids in establishing the amount
and analysis of fertilizer to be used. For the immediate recommendation,
the crop to be grown in the following year must be known. In this respect,
the different rooting habits and the differing abilities of various crop plants
to extract nutrient elements from the soil must be considered. With all
34

these factors in mind, a practical soil treatment may be formulated, assum-

ing average moisture and temperature conditions.
13-80. Sampling Soils for Soil Chemical Testing.:ir. The procedure for
sampling a farm field for soil testing (details in ~ 2-49) consists of com-
positing thin slices of soil through the plow layer from 10 to 20 places dis-
tributed throughout each uniform portion of the field. One such composite
sample is taken for each 5 to 10 acres. Upland, middle, and bottom lands
are sampled separately. Unusual soil spots such as those near gates, under
stacks of manure, from dead furrows, or along field boundaries arc
avoided. The following information should be supplied with the sample:
1. Sample number . . . . . . . . . . . . . . . . . . . . Date ....... ....... ..... .
2. Field number . . . . . . . . . . . . . . . . . . . . . County ....... ....... ... .
3. Crop to be grown ....... ....... ....... ....... ....... ....... .
4. Will the field be manured for next crop? ....... ....... ....... ... .
5. Yield of crop last year ....... ....... ....... ....... ....... .... .

33 Truog, S.S.S.A. Proc., 1: 135 (1937).

34 Drake et al .. S.S.S.A. Proc., 4:201 (1940).
S5 Reed, Better Crops, 37, 8:13 (1953).
RAPID TESTS OF SOILS ·AND PLANT TISSUE 359
6. Field is upland ......... , bottomland ......... , marsh
7. When field was last limed . . . . . . . . . . . Amount per acre .......... .
8. Kind and amount of commercial fertilizer used ................... .
9. Name and address .......................................... .

13-81. Extractants for Soil Testing. Soil extractants to be employed for

testing have been selected generally to simulate plant feeding (Morgan,:rn
Spurway37 ) or to extract all or a proportional part of the available form or
forms of the element being tested (Bray, 38 Truog~"). Although the results
from the 2 types of extractants may differ greatly, as in the phosphorus
test, they may be so related that a factor may be employed to convert from
one system to the other. 40 In any case, the results of a test must be cor-
related with crop response to applications of the element.
13-82. The extractant must not contain ions that would interfere with
the analysis to be made. Use of a single extractant for several elements de-
creases the time required for several tests. Since no 1 extractant is suitable
for all soils or all elements, various solutions have been used. These ex-
tractants fall generally into 4 types, as follows:
1. The strong mineral acid type, e.g., 0.3 N HCl (Truog as given by
Nelson et al. 41 )
2. The highly buffered weakly ionized acid type, e.g., 10 per cent NaOAc
in 3 per cent HOAc buffered at pH 4.8 (Morgan).
3. The fairly concentrated solution of a neutral salt, e.g., 3 N NaNO:i
(Bray) .42
4. The solutions with a solvent action of the same order as water, e.g.,
0.025 N HOAc (Spurway).

Many different extractants are being employed within any region due to de-
velopment by various independent workers. Some standardization would
be desirable.

13-83. Nutrients elements are extracted from both sandy and fine-
textured soils for testing by means of minitaure electrodialysis cells in the
New Jersey Experiment Station testing laboratory. 43

36 Lunt et al., The Morgan Soil Testing System, Conn. Agr. Exp. Sta. Bui. 541
(1950).
:H Spurway and Lawton, Mich. Agr. Exp. Sta. Tech. Bui. 132 (4th Rev.) (1949).
as Bray, Soil Sci., 66:83-89 (1948).
39 Truog, S.S.S.A. Proc., I: 135 ( 1937).
40 Peech, S.S.S.A. Proc., 10:245 ( 1946).
41 Nelson et al., Soil Testing in the United States (Beltsville, Md.: Natl. Soil and
Fert. Res. Committee, U.S.D.A. Soils Div., 1951).
42 Bray, Diagnostic Technics for Soils and Crops, Ch. 2 (Washington, D.C.:
American Potash Institute, 1948).
43 Purvis, Better Crops, 37, 3: 19 (1953 ).
 NT TISSUE
360 RAP ID TESTS OF SOILS AND PLA

PROCEDURE
still individualistic with each
13-8 4. Because soil-testing procedures are
in respective bulletins, no par-
experiment station, appropriately detailed
n here. As summarized in Table
ticular system of procedures is to be give
considerably in the extractants,
13-1 , the various soil testing systems vary
extraction; but many of the dif-
soil to extractant ratios, and methods of
ytical determination. The present
ferent systems are similar in type of anal
ytical determinations that are
objective is to illustrate representative anal
employed by various laboratories.
13-8 5. The determinations considered
are soil pH, phosphorus, potas-
um, and magnesium. Soil testing
sium, nitrogen and organic matter, calci
asingly important, but thus far the
for the minor elements is becoming incre
determinations; for example, for
methods are mainly the usual analytical
0). Tests for copper, zinc, and
manganese (~ 15-2 7) and boron (~ 14-4
sively beyond the determination
molybdenum have not been developed inten
of the total quantities (Cha pter 15).
or color developed in soil tests
13-8 6. The measurement of the turbidity
photometers (~ 17-2 1) or by
is generally by photoelectric absorption
permanent plastic 41 standards is
printed color charts. The preparation of
and mobility of the soil testing
an aid to permanence, inexpensiveness,
equipment.
13-8 7. The Soil pH Test. Perhaps the
most important of all rapid soil
soil pH indicates much about the
tests is the soil pH test(~ 13-9 0). The
the availability of P, Ca, Mg,
interpretation of the other soil tests, since
pH. An acid soil test, coupled
Fe, Mn, and B are greatly affected by soil
n and the soil type, permits a
with a knowledge of the crop to be grow
nd limestone needed ( ~ 13-8 8).
recommendation of the quantity of grou
agement.
Soil areas testing alkaline need different man
e, inclu ding rotations with lime-loving
13-8 8. For intensive agricultur
pH value to the range of 6.5 or
legumes, the soil is limed to bring the
slightly higher. The following is a guide:
pH value to 6.5
Tons of groun d limestone per acre to bring
(appr oxim ate-guide )
Soil pH value ---------
--------------
1---=-~-~--c
Sand s and sandy loam s Silt loams, clay loams, muck s

6 9
4.0 7
4.5 5
4 5
5.0 3
5.5 2
1 2
6.0 0
6.5 0 --~~--
--
------
------
------
------

n., 40:94 0 (194 8): S.S.S.A Proc., 15:15 2

, 44Ly nd and Turk , J. Am. Soc. Agro
(195 1).
 TABLE 13-1
General character of several soil testing systems

Wisconsin Indiana I Michigan§ Connecticut** Illinoistt CornellU Missouri§§

(Truog) (Ohlrogge) (Spurway) (Morgan) (Bray) (Peech) (Graham)
Wt. of soil 3* P-2.5 2.5 5 P-1 10 P-1
used, gm K--5.0 K-5 K-5
Soil : extrac- 3 : 7* P-1: 2 1:4 1:2 P-1: 7 1:5 P-1: 7
tant ratio K-1: 4 K-1: 2 K-1 :2
Extractant 0.3 NHCl* P-(NH4)2Mo04 0.025 NHOAc 10% NaAcin P-0.03 N NH4F Morgan Bray
+ 0.75 N HCl (active) 3% HOAc in 0.1 N HCl
K-Na3Co(N02)6 0.13 N HCl K-3 NNaN0 3
+ HOAc + NaN02 (reserve)
pH of 0.5 P-5.04 3.3 (active) 4.8 P-1.0 4.8 P-1.0
extractant 0.9 (reserve) K-7.0 K-7.0
.., K test Turbidimetric Sodium cobaltinitrite
°'
P04 te~t Molybdenum f blue color from formation of molybdophosphoric acid in acid solution with subse-
quent reduction with stannous ion or organic reducing solution.
Ca test Oxalatet - Oxalate Oxalate Oxalate Stearate Oleate
Mg test I Titan - Titan Thiazole Titan Titan Thiazole
II - yellowt yellow yellow yellow yellow yellow
--
Cl test I AgN03t - AgN03 AgNOa AgNOa
(special (active)
extractant)
S04 test I BaCl2t - BaCb BaCl2 BaCl2
0 Nelson et al., Soil Testing in the United States (Beltsville, Md.: Natl. Soil and Fert. Res. Committee, U.S.D.A. Soils Div., 1951).
t Hellige, Long lsland City, N.Y., Bui. 690 ( 1936 ).
I Ohlrogge, Purdue Univ. Agr. Exp. Sta. Bui. 584 ( 1952 ).
§ Spurway and Lawton, Mich. Agr. Exp. Sta. Tech. Bui. 132 (4th Rev.) (1949).
00 Lunt et al., Conn. Agr. Exp. Sta. Bui. 541 ( 1950 ).

tt Bray, Ill. Agr. Exp. Sta. Mimeo. Circ. AG 878 ( 1940); AG 1028 ( 1942) for change in P test.
U Peech and English, Soil Sci., 57: 167-195 ( 1944 ) .
§§Graham, Mo. Agr. Exp. Sta. Circ. 345 ( 1950 ).
362 RAPID TESTS OF SOILS AND PLAN T TISSUE
de-
The amoun t of change of soil pH with quantity of limestone applied
y and pH titratio n
pends on the soil type, particularly its exchange capacit
in re-
curve, but these factors are usually worked out in each laboratory
a soil
lation to the important local soil types. The lime requirement of
more elabora te quantit ative method s (~
can also be determined by the
4-64).
13-4.
13-89. The pH preferences of various crops are outlined in Fig.
exampl e, soils
Some conditions require a limited application of lime. For
extensi vely,
below pH 5.0 to be planted to turfs that are to be watered
lime-
should receive only a light dressing of 500 pounds of finely ground
the
stone per acre at seeding. The lime in the water will gradually bring
high values) and therefo re heavy liming is
soil pH values up (often to too
(those showin g mangan ese toxic-
avoided. For excessively acid potato soils
ground
ity), a limited amount ( 500 to 1000 pounds per acre) of a finely
limestone is applied.
be by
13-90. The determination of soil pH for soil testing purposes may
r of
the usual glass electrode techniques (~ 3-17) or by any of a numbe
ors. The glass electro de is used in nearly
systems of colorimetric pH indicat
on a
all 40 soil testing laboratories for the determination of soil pH even
e the determ ina-
mass production scale, because of its accuracy and becaus
than by
tion can be made in the laboratory as rapidly as, or more rapidly
ment
colorimetric methods. The equilibrium pH methods for lime require
pH measur ement, and take into accoun t
( ~ 4-61 ) are as rapid as a simple
the buffer capacity of the soil.
and
13-91. Colorimetric indicators are most useful for field testing kits
give approx imate but
are also used in some soil testing laboratories. They
uses 3 separat e
satisfactory results when properly used. The Purdue system
blue, to
indicators, brom cresol green, chlor phenol red, and brom thymol
, 47 use
cover the usual range of soil pH. Other systems, such as the Morgan
and
a mixture of the indicators brom cresol green, brom cresol purple,
to cover the pH range of 4 to 8.
cresol red (0.025 per cent each in water)
cresol green, 0.10 per cent
A similar mixed indicator (0.05 per cent brom
volume
brom cresol purple, and 0.02 per cent cresol red in 40 per cent by
ory. Three drops
of ethanol in water) has been used in the author's laborat
soil is
of the indicator are placed in a white spot plate. Just enough air-dry
or.
then sprinkled onto the indicator to remain rather wet with the indicat
for a minute , the soil is moved with the spatula
After stirring with a spatula
follow-
and the color is read against the white spotplate background. The
ing colors denote the soil pH:

Nelson et al., op. cit.

46
Sta. Bui. 541
Conn. Agr. Exp. Sta. Bui. 392 (1937); Lunt et al., Conn. Agr. Exp.
47
(1950).
 pH RANGE
CROPS 4.0 4.5 5.0 6.0 6.5 7.0
5.5 7.5 8.0
FIELD AND FORAGE
ALFALFA
CLOVER, SWEET
•
BARLEY •
CLOVER, RED
BLUE GRASS, KY.
••
TIMOTHY
ORCHARD GRASS
••
••
RAPE
CLOVER, WHITE

•••
WHEAT
CLOVER, MAMMOTH
CORN, FIELD
SOYBEAN
OAT
CLOVER, ALSIKE
CLOVER, CRIMSON
VETCH, HAIRY
MILLET
BUCKWHEAT '
RYE
RED TOP
•• '
BENT GRASSES •
TOBACCO
POTATOES, WHITE
•• •
•
FESCUE GRASSES
•
VEGETABLES
ASPARAGUS
•
••
SPINACH
LETTUCE
CELERY
RADISH
ONION

••
BEET, GARDEN
CAUL! FLOWER
BROCCOLI
CABBAGE • •
BRUSSEL SPROUTS
•
••
PEA, GARDEN
CUCUMBER
MUSKMELON •
••
RHUBARB
CARROT
BEANS, LIMA
BEANS, GARDEN • •
••
CORN, SWEET
ENDIVE

•••
PARSNIP
PUMPKIN
SQUASH

••
PEPPERS, SWEET
RUTABAGA

••
TURNIP
TOMATO
EGGPLANT • •
• •
POTATOES. SWEET

FRUITS
- --
CURRANT
•
•• •
QUINCE
PEAR
GOOSEBERRY

••
APPLE
GRAPE
PEACH
RASPBERRY, RED

••
RASPBERRY,BLACKCAP
STRAWBERRY
BLUEBERRY
CRANBERRY
•
•MINIMUM pH. • MAXIMUM pH LIMIT
- DESIRABLE pH RANGE. FOR DISEASE CONTROL.
Fig. 13-4. Desirable soil pH range for a number of crops. (Courtesy National
Plant Food Institute, Washington, D.C., arranged by Morgan, Fert. Rev., 12, 2:7,
1937.)
363
364 RAPID TESTS OF SOILS AND PLAN T TISSUE
Soil pH Indicator Color
Soil Acidity
4.0 Yellow
Very strongly acid Greenis h yellow
Strongly acid 4.5
5.0 Yellowish green
Acid Light green
Moderately acid 5.5
6.0 Bluish green
Slightly acid Greenish blue
Very slightly acid 6.5
7.0 Dark blue
Neutral Purple
Alkaline 7.5

reagent
The soil may also be covered over with neutral, ignited, and ground
bring out the
grade precipitated BaS0 4 powder, so as to mask the soil and
ponen t indica tor and ground and purified
indicator color. A special 3-com
pH testing is sold by Hellige, Inc., 3718
barite mineral for colorimetric
Northern Blvd., Long Island City, N.Y.
13-92 . A colorimetric test48 for soil acidity with a range of:
Tndicator color
Very strongly acid ...... .... Bright red
Slightly acid ...... ...... ... Yellow
Neutral ...... ...... ...... . Green

thymol blue,
employs an indicator composed of 0.08 per cent Na-brom
0.04 per cent Na-methyl red, and 0.02 per cent methyl orange .
thiocyanate
13-93 . Other methods for measuring soil acidity include the
cent KSCN
test 49 and the sulfide test. 50 Jn the thiocyanate test, a 4 per
e of field-
solution in 95 per cent ethanol is shaken with an equal volum
acid mineral
moist (fairly dry) soil in a glass vial. On standing 10 minutes,
acidity. Soils
soils develop a pink to red color depending on the degree of
ss solutio n. In the sulfide test, 9 ml of
of pH 6.5 or above give a colorle
l flask, 0.8 ml of a mixtur e of BaC1 2 •
soil is placed in a 300-ml conica
tely ground togeth er in the propor tion
2 H 2 0-ZnS (both neutral and intima
t) is added, follow ed by 100 ml
10 of BaC1 2 • 2 H 2 0 to l of ZnS by weigh
then a moiste ned paper con-
of water. The suspension is boiled for 1 minute
the liberation
taining Pb(OA c) 2 , is placed over the flask. Acid soil causes
degree of blackening
of H 2 S which blackens the Pb(0A c) 2 paper. The
with PbS is proportional to the soil acidity.
lly in-
13-94 . Soil Phosphorus Testing. The soil phosphorus test genera
n with sub-
volves the formation of molybdophosphoric acid in acid solutio
n) to give
sequent reduction with stannous ion (or organic reducing solutio
given as an
molybdenum blue (~ 7-4). The Purdu e51 phosphorus test is
example of this test.
demons trated for
48Yagi, H., Agr. and Hort. Research Sta., 25: 39 (1950) ; kindly
the author by S. Aomine .
49 Linsley and Bauer, Ill. Agr. Exp. Sta. Cir.
346 (l 941).
50 Truog, Wis. Agr. Exp. Sta. Bui. 312 ( 1920).
lil Ohlrogg e, Purdue Agr. Exp. Sta. Bui. 584
(1952) .
RAPID TESTS OF SOILS AND PLANT TISSUE 365

13-95. A glass vial graduated at 10 ml is filled to the mark with molyb-

date reagent (~ 13-36). One-half level teaspoon of soil (14 teaspoon for
greenhouse and garden soils) is added, and the vial is shaken vigorously for
1 minute. The solution is filtered into a funnel tube. To 5 ml of the filtrate
is added about 1 cubic mm of dry powdered stannous chloride or stannous.
oxalate. The solution is mixed by rotation of the funnel tube, and the color
produced is observed by comparison with the Phosphate Test Chart. More
of the stannous chloride or oxalate is added to make certain that enough has
been used. If the color becomes more intense, the latter reading is used. The
range of the test is from a dark blue (abundant phosphorus) to colorless or
a light yellow (low phosphorus supply). The test does not give reliable re-
sults on alkaline soils.
13-96. The most common tests for soil phosphorus involve the dilute
acid extractable phosphorus, which is chiefly in calcium phosphate form.
Other tests involve the extraction with solutions containing fluoride ( ~
7-82), in order to include some phosphorus forms other than the acid ex-
tractable. Testing for soil phosphorus is made difficult because in soils that
are not acid, the acid soluble form may not be available. In other soils the
fluoride extractable and easily oxidized organic forms of soil phosphorus
may make substantial contributions to the phosphorus available to crops in
a single season.
13-97. Soil Potassium Testing. The turbidimetric potassium cobalti-
nitrite test is used extensively. The results are affected by temperature,
time of precipitation, particle size, and crystal form. A recent modification
of this test is designed to reduce the effect of temperature. 52
13-98. The Purdue potassium test is described as an example of this
type of method. A glass vial graduated at 10 ml is filled to the mark with
the sodium cobaltinitrite reagent (~ 13-56). One level teaspoon of air-dried
soil (lh teaspoonful for mucks, greenhouse and garden soils) is added, and
the vial is shaken vigorously for 1 minute. The solution is filtered into a
funnel tube. To 5 ml of the filtrate is added 2.5 ml of anhydrous isopropyl
alcohol with the pipet, and the 2 solutions are mixed. After 3 minutes, the
turbidity of the solution is compared with the standard potassium chart.
The range of the test is from a clear solution (deficient supply) to a very
turbid, dense precipitate (abundant supply).
13-99. Soil potassium tests ordinarily aim to measure the exchangeable
potassium. Maintenance of a high level of exchangeable potassium is highly
desirable for assurance of high yields. Acid extraction measures some po-
tassium in addition to the exchangeable (~ 6-83). The soil potassium that
becomes available to a crop in a given year includes not only the exchange-
able but a certain amount of the potassium released by current weathering.

52 Olson, S.S.S.A. Proc., 17:20 (1953 ).

366 RAPID TESTS OF SOILS AND PLAN T TISSUE

In some soils and for some crops the potassium released

by weathering is
are supplied
rather important. Substantial amounts of available potassium
variables may
by some subsoils, whereas by others little is supplied. These
need consideration in the interpr etation of the soil potass ium test.
cted efforts
13-10 0. Soil Nitrogen and Organic Matter Testing. Protra
of nitroge n release (~ 8-69) through
have been made to measure the rate
consid erable attenti on has been given
biological nitrification.r.a In addition,
of organi c matter based on am-
in various countries to the partial oxidation
n. The design of a useful test
monia released in terms of available nitroge
e of the effect of climat ic
for available nitrogen is very difficult becaus
effect is greate r on nitro-
factors on the release of soil nitrogen. The climatic
Also, chemi-
gen release than on the availability of such elements as P or K.
from fresh organic
cal tests do not correlate well with biotic N release
tories there-
matter residues in soils because of C : N effects. Many labora
rapid chrom ic
fore simply determine the organic matter content with the
instead of at-
acid method (~ 9-57) applied on a mass production basis,
of the N release d from the oxidiz able organic
tempting a direct estimation
soil aids ma-
fraction. A knowledge of the total organic matter content of
ement recom menda tions. The immediately
terially the making of soil manag
taken into accoun t in additio n to any
preceding cropping history must be
of the chemical tests.
n may
13-10 1. Rapid tests for nitrate, ammonium, and nitrite nitroge
than the other tests, except
also be made. They are generally of less value
ive. The level of these
that abnormally high or low results may be indicat
soil, croppi ng, and cli-
3 forms of nitrogen in the soil is dependent on the
matic conditions prevailing immediately prior to the test.
diphenyl-
13-10 2. Nitrates may be determined rapidly with brucine or
of soil extract
amine. In the Morgan system 54 with diphenylamine, I drop
of diphen yl-
is placed on a spot plate, and 4 drops of the reagent (0.05 gm
s the mix-
amine in 25 ml concentrated H 2 S04 ) are added. After 2 minute
ty of the resulti ng blue color is determ ined with
ture is stirred and the intensi
comparison to the color chart.
Lawtonr, 5
13-10 3. The active ammonium test as given by Spurway and
drops of soil
involves the addition of 1 drop of Nessler's solution to 3
color develo ps rapidly . If a brown
extract (Table 13-1) on a test plate. The
the test repeat ed.
precipitate forms, the soil extract is diluted and
le of nitrite nitrogen
13-10 4. The Morgan method is given as an examp
placed on a spot
determination. Ten drops of soil extract (Table 13-1 ) are
is dissolv ed by
plate. One drop of nitrite reagent ( 1 gm of sulphanilic acid

Russel et al., Soil Sci., 19: 381 (1925) .

lill
Exp. Sta. Bui. 541
Lunt et al., The Morgan Soil Testing System, Conn. Agr.
54
(1950) .
55 Miclt. Agr. Exp Sia. Tech. Bui. 132 (4th Rev.)
( 1949).
RAPID TESTS OF SOILS AND PLANT TISSUE 367
heating in 100 ml of a saturated solution of NH 4 Cl, then 1.5 gm of phenol
is added followed by thorough mixing), 1 drop of 6 N HCl, ·and 4 drops
of 15 per cent NaOH are added. The mixture is stirred, and let stand
J minute. The resulting color ranges from yellowish orange for very high
nitrite test, to a trace of yellowish tint for very low test.
13-105. Soil Calcium Testing. Calcium in soil testing is usually esti-
mated turbidimetrically as the oxalate. This procedure is subject to errors
due to different pH levels, presence of organic matter, and interfering ions.
Calcium may also be determined as the stearate in the presence of citrate
(Cornell) r.n or as the oleate (Missouri). fi 7 In the Michigan" 8 system, 2
drops of ammonium oxalate ( 5 gm in 100 ml of distilled water) are added
to I ml of soil extract in a small tube and mixed by shaking. A white pre-
cipitate indicates calcium, and the amount is determined by comparison to
a turbidity chart.
13-106. The soil calcium test is generally high if the soil pH is properly
adjusted. Ordinarily, therefore, the soil pH test obviates the need for test-
ing soil calcium.
13-107. Soil Magnesium Testing. Magnesium in soil testing is usually
determined by the titan yellow or thiazole yellow procedure. This involves
the precipitation of Mg(OH ) 2 , which absorbs the dye to form a magnesium
lake. In the Michigan" 11 system, I drop of titan yellow solution (0.15 gm of
titan yellow dye dissolved in a mixture of 90 ml of either ethyl or isopropyl
alcohol, and 10 ml distilled water) is added to I ml of soil extract in a small
tube and mixed by shaking. To this mixture is added 1 drop of 5 per cent
NaOH solution and this is followed by thorough shaking. The magnesium
test colors range from light orange to peach red, with a yellow color indi-
cating no magnesium. In other procedures (Peech, Graham), a protective
colloid (starch) is introduced to disperse the precipitate and intensify the
color. Calcium and aluminum are added in a compensating solution to
produce their maximum intensity of color, since their presence in varying
amounts causes erratic results.
13-108. Soil magnesium levels are usually satisfactory if dolomitic lime-
stone has been employed for adjustment of the soil pH. However, if calcic
limestones are prevalent as a source of liming material in a locality, soil
magnesium testing is essential, independently of soil pH testing.
13-109. Chloride, Sulfate and Sodium Testing. In saline and alkali soils
the testing of soil chloride, sulfate, and sodium becomes important. The
determination of soil chloride and sulfate is especially important also in
connection with saline soils in commercial greenhouses where heavy fertili-

511 Peech and English, Soil Sci., 57: 167 ( 1944).

,-,; Graham, Mo. Agr. Exp. Sta. Cir. 345 ( 1950).
r.H Spurway and Lawton, Mich. Agr. Exp. Sta. Tech. Bui. 132 ( 1949).
59 Spurway and Lawton, Mich. Agr. Exp. Sta. Tech. Bui. 132 (1949).
368 RAPID TESTS OF SOILS AND PLANT TISSUE
zation has been practiced. Tests for these constituents have been given
(chloride, ~ 10-94; sulfate,~ ·10-102; and sodium,~ 5-81).
13-110. Biotic Soil Tests. Since the crop relation to soil is ultimately a
biotic one, the most basic calibration of chemical soil tests must be with
biotic tests. To make possible testing on an extensive scale, small-scale
biotic systems have been devised as contrasted with the field plot test. The
most noted of the small-scale biotic tests are the Mitscherlich110 pot test and
the Neubauer seedling cup. 61 The pot method for soil cultures has been
described in detail.u 2 Yield equations derived from field tests or pot tests
can be projected into the quantity of nutrient elements evidently "available"
from the untreated soil itself, and data are thereby provided for calibra-
tion of chemical soil tests. In general, biotic methods themselves do not ful-
fill the definition of soil testing ( ~ 13-7 5) because of not being rapid and
low in cost. They are even more costly than conventional long methods of
soil chemical analysis, and thus must be employed strategically in calibra-
tion of soil tests.

QUESTIONS
1. Discuss the analogy in function between the soil testing laboratory and
clinical testing and diagnosis in medicine.
2. What is the underlying assumption that forms the basis for rapid chemical
testing of the sap of plant tissue as a means of diagnosing adequacy of plant
nutrient supply?
3. Explain how a mistaken interpretation of plant sap tests might be made
of abundance of 1 nutrient element in the soil supply when the growth of the
plant is held back by a severe shortage of another nutrient element.
4. Explain what is indicated when the diphenylamine test gives a strong blue
color on a slice of nodal corn tissue. Is the blue diphenylamine test obtained in
such tissue as clover and asparagus? Explain.
5. Explain the principle and advantages of the pink powder test for nitrate
in the sap of plants.
6. Summarize briefly the different procedures for testing for phosphorus in
plant sap.
7. Summarize briefly the different procedures for testing for potassium in
the sap of plants.
8. Why is it possible to determine the instantaneous oxygen activity of a soil
by field colorimetric tests for ferric and ferrous iron in the HCI extract of the
soil?
9. Why may the rapid imbibition of suspended chalk particles in a soil
column surface be used as a measure of good soil aeration?

60 Hoover and Norman, Soil Sci., 53:329 ( 1942).

61 Thornton, Purdue Univ. Agr. Exp. Sta. Bui. 399 (1935); McGeorge, Soil Sci.,
58:389 (1944).
62 Maclntire and Winterberg, Soil Sci., 62:33 (1946).
RAPID TESTS OF SOILS AND PLANT TISSUE 369
10. Why do soil tests not automatically provide specific recommenda tions for
soil management without the intermediary of human judgment?
11. What 2 general types of soil pH tests are employed? Why is it that soil
testing for calcium and magnesium is seldom practiced in routine?
12. What difficulties are presented in the extraction of "available" soil phos-
phorus?
13. What difficulties are presented in the extraction of "available" soil potas-
sium?
14. Why is the determination of "available" soil nitrogen especially difficult?
15. Under what circumstances may analysis for chlorides, sulfates, and
sodium be of interest in soil testing?
16. Discuss the relationship of biotic tests to soil chemical tests.
 14
Boron Determinations
for Soils and Plant Tissue
. . . plants will not make growth without boron any more
than . . . without phosphorus or potassium which they re-
quire in considerable amounts.
-TRUOG 1

14-1. Of the 4 elements essential to plants that occur in soils principally

as anions (N, P, S, and B), the one plants need in the smallest quantities
is boron. (However, as will be pointed out in Chapter 15, molybdenum in
soils is most available in anion form, and this element is needed by plants
in still smaller quantity.) Like soil nitrogen and sulfur compounds, soil
boron compounds of highest availability are in water soluble form. Hot
water soluble soil boron was found 2 to be correlated with the boron con-
tent of red beet leaves, and such extraction appeared to be a better measure
of available soil boron than acid soluble boron or total boron. Analyses for
water soluble, acid soluble, and total boron for a large number of repre-
sentative soils are published. 3
14-2. The amount of water soluble boron in cultivated soils appears to
be influenced more by the soil pH than any other factor, but is influenced
also by the soil organic matter content, mineral colloid content, age of the
soil, and by the type of irrigation water that has been employed. Significant
positive correlation of water soluble boron was found 4 with soil pH values

1 Statement before Wisconsin Canner's School, March 13, 1940.

2 Berger and Truog, J. Am. Soc. Agron., 32:297 ( 1940); Berger, Adv. Agron.,
1:321 (1949).
3 Whetstone et al., U.S.D.A. Tech. Bui. 797 ( 1942).
4 Berger and Truog, S.S.S.A. Proc., JO: 113 ( 1946).

370
BORON DETERMINA TIONS FOR SOILS AND PLANTS 371
from pH 4. 7 to 6. 7, and negative correlation between pH 7. I and 8.1.
Although over-liming is known° to accentuate boron deficiency in accord
with these findings, it has been shown6 that the fixation of boron in water
insoluble form is caused by the rise in pH independent of the presence of
high activity of calcium ion. Acidification releases the boron fixed in the
presence of high -OH activity.
14-3. Increasing the organic matter content in soil increases the water
soluble boron, although a soil of higher organic matter content fixed more
as the pH was raised. n This indicates that the organic matter is in active
equilibrium with water soluble boron. Since a soil of pH value of near
neutrality holds boron better than an acid soil, the low level of water soluble
boron in old, acid, leached soils is explained. Fixation at pH values of 8
to 10 did not exceed 15 to 40 per cent of the water soluble boron, thus
indicating that boron seldom will become deficient in alkaline soils, and
may even become toxic in alkaline soils.
14-4. The condensation of the borate radical into long chains in the
presence of calcium, increasing hydroxyl activity, and low moisture has
been pointed out 7 as a possible mechanism for making boron less available
under some circumstances. Marked decrease in boron availability under
conditions of low soil moisture has been observed extensively. The be-
havior of boron in soils has been studied indirectly in simplified chemical
systems involving pure chemicals 8 and soil organic matter 0 by observa-
tion of precipitation and solubility. Entry of boron into the structure of
calcium silicates and calcium aluminosilicates was indicated. The magnitude
of boron retention by humus systems and the chemical reactions between
boron and dihydroxy organic compounds suggests 10 that soluble soil boron
is fixed by diols such as of uronic acid of soil organic matter.
14-5. There are rather narrow limits between adequate amounts and
toxic amounts of boron in soils. Boron toxicity has been a problem in con-
nection with soils irrigated with waters 11 containing over 1 to 3 ppm of B,
and the value for excellent irrigation water should be below 0.3 to 0.6 ppm
(~ I 0-80). Boron toxicity occurred in soils fertilized with potash carrying
a high boron content. 12 Toxicity resulted from use of boron in manure to
repress fly larvae. rn
14-6. Methcds for Boron Determination. Boron in soils and plants can
be determined titrimetrically (,I 14-52) in macro and semimicro amounts,
5 Naftel, J. Am. Soc. Agron., 29:761 ( 1937).
u Olson and Berger, S.S.S.A. Proc., 11 :216 ( 1947).
7 Colwell and Cummings, Soil Sci., 57: 37 ( 1944 ).
8 Parks and Shaw, S.S.S.A. Proc., 6: 219 (1941 ).
9 Drake et al., J. Am. Soc. Agron., 33:454 (1941).
10 Parks and White, S.S.S.A. Proc., 16:298 ( 1952).
11 Wilcox, U.S.D.A. Cir. 784 (1948).
12 Conners and Fergus, Ind. Agr. Exp. Sta. Bui. 239 ( 1920).
13CookandWils on,J.Agr.Res., 13:451 (1918).
372 BORON DETERMINATIONS FOR SOILS AND PLANTS
and colorimetrically or spectroscopically (Chapter 18) in micro amounts.
Biological assay of available soil boron also has been perfected. 14 Colori-
metric methods have a distinct advantage for the determination of the micro
amounts of boron of the order of 0.5 to 7 ugm, which are frequently de-
rived from convenient sample weights of soils or plants. Both the curcumin
( ~ 14-7) and the quinalizarin ( ~ 14-17) colorimetric methods are adapted
for either visual or photometric evaluation of the color. Because the chemi-
cal separations are easier, the colorimetric determinations are generally
more convenient than the titrimetric method, even when the amount of
boron is not a factor. The methods based on anthraquinones, 15 such as the
quinalizarin and caramine 16 methods, all have their colors developed in
concentrated H 2 S0 4 and therefore temperature change causes great varia-
tion in colorimetric readings.

BORON DETERMINATION
( Colorimetrically with curcumin)
14-7. The curcumin procedure for boron has the advantages over the
anthraquinone procedures of having a less corrosive solvent than concen-
trated H 2 S04 , nonsensitivity to small changes of temperature in the solution
to be read, and a sharp spectral separation between the reagent color (ab-
sorption slight at 500 mu and higher), and the boron-dependent color
( rosecyanine, 550 mu maximum absorption). The boron-dependent color
develops from curcumin, extracted from its crude vegetable source, tu-
meric. 17 The formation of rosecyanine takes place in proportion to the B
present, but requires the presence of 0.2 gm of oxalic acid. This amount
of oxalic acid is not soluble in the aqueous HCl-H 2 C 2 0 4 solution ordinarily
recommended except as a supersaturated solution or a suspension. In the
improved 18 procedure detailed here, the combined curcumin-H 2 C2 0 4 re-
agent in 95 per cent ethanol provides the required oxalic acid and the
advantages of a single combined reagent. Colored lakes develop with cur-
cumin in alkaline solution in the presence of Be, Al, Fe, or Mg, and there-
fore an acid medium must be maintained during color development. The
presence of H 2 C 2 0 4 maintains sufficient acidity for the color development
if the solution containing the boron is slightly acid, thus eliminating the
need for addition of concentrated HCl in the procedure.
14-8. Nitrate interferes in the acid solution if there is over 20 ugm of
N as nitrate in the 1-ml aliquot taken for analysis. Nitrates can be elimi-
nated from the test solution by evaporation of an aliquot (to which suffi~

14 Colwell, Soil Sci., 62:43 (1946).

rn Ellis et al., Anal. Chem., 21: 1345 (1949).
111 Hatcher and Wilcox, Anal. Chem., 22:567 (1950).
17 Naftel, Ind. Eng. Chem., A.E., 11 :407 (1939).
18Dibleetal.,Anal. Chem., 26:418 (1954).
BORON DETERMINATIONS FOR SOILS AND PLANTS 373
cient saturated water solution of Ca(OH) 2 has been added to make it
alkaline 19 ) and gentle ignition; the ash is taken up in 0.05· N HCI. Ac-
cording to Feigl, 2° Fe, Mo, Ti, and Zr interfere also, if present in unusual
amounts (over about 300 ppm) in the boron solution to be analyzed.

APPARATUS

14-9. Needed are a photoelectric colorimeter with 540- or 550-mu light

maximum for most sensitive range (0.2 to 1.5 ugm of B) and 580- or 600-
mu light maximum for a less sensitive range ( 1 to 7 ugm of B); standard-
ized colorimeter tubes; a boron-free glass beaker, 250-ml; pipets, 1-ml,
4-ml, 25-ml, and 50-ml; volumetric flasks, I-liter, 500-ml, series of 50-ml;
and a drying oven or bath, regulated at 55 ± 3 °C.

REAGENTS

14-10. Ethanol, 95 per cent, A.R., is needed, redistilled if necessary to

make it boron-free.
14-11. Curcumin-Oxalic Acid Reagent. 21 0.04 gm of finely ground cur-
,cumin (Eastman Kodak No. 1179) and 5.0 gm of H 2 C2 0 4 • 2 H 2 0 are
dissolved in 100 ml of boron-free 95 per cent ethanol. This solution is
stored in an amber bottle in a cool dark place. Because of the slow decom-
position of curcumin, this reagent is made up fresh every few days. This
may be extended to a week or more by storage in a refrigerator without
undue exposure at room temperature.
14-12. Boron Standard Solution. A stock boric acid solution is prepared
by dissolving 0.572 gm of dried reagent grade H:1B0: 1 in I liter of distilled
water, giving a 100 ugm of B per ml standard. A dilute stock solution is
prepared by pipetting 50 ml of the first stock solution into a 500-ml volu-
metric flask and diluting it to volume with distilled water, giving a I 0 ugm
of B per ml standard. The dilute stock solution is employed for making up
a series of standards, in 50-ml volumetric flasks, for the calibration curves,
in accordance with Table 14-1.

PROCEDURE22

14-13. A I-ml aliquot of slightly acid aqueous solution, containing 0.2

to 5 ugm of boron is transferred to a 250-ml boron-free glass beaker. (The
same 1-ml pipet is employed to transfer standard and test solutions, and
therefore no special volumetric calibration is needed; however, the 1-ml
constant volume is critical.) Then 4 ml of the curcumin-oxalic acid solu-

19 Gooch, Am. Chem. J., 9:23 (1887).

20 Tr. J. W. Matthews, Spot Tests (New York: Nordemann Publishing Company,
Inc., 1937),p.211.
21 Dible et al., Anal. Chem., 26:418 (1954).
22 Dible et al., Anal. Chem., 26:418 (1954).
374 BORON DETERMINATIONS FOR SOILS AND PLANTS

TABLE 14-1
Appropriate concentration ranges of boron standard solutions
for curcumin and quinalizarin methods
Concentration range and steps
I------ --~----·-····-- --·
Quinalizarin
Curcumin method method
I Volume of stock* Concentration
obtained when 580- to
solution diluted
Solution to 50 ml diluted to 50 mil 540-mu 600-mu Color-
number (ml) (ugm of B/ml) light light Visual imeter
Blank 0.0 0.0 0.0 0.0 0.0 0.0
1 1.0 0.2 0.2 0.2
2 2.0 Q4 0.4 0.4

3 2.5 0.5 0.5 0.5 0.5

4 3.0 0.6 0.6 0.6
5 4.0 0.8 0.8 0.8

6 5.0 1.0 1.0 1.0 1.0 1.0

7 7.5 1.5 1.5 1.5 1.5 1.5
8 10.0 2.0 2.0 2.0 2.0
9 12.5 2.5 2.5 2.5 2.5
10 15.0 3.0 3.0 3.0
11 17.5 3.5 3.5

12 20.0 4.0 4.0 4.0

13 22.5 4.5 4.5
14 25.0 5.0 5.0 5.0
15 27.5 5.5 5.5
16 I 30.0 6.0 6.0 6.0
__1_1__1_ _ _ .. 3~:~___________ 7.0 7.0 7.0
. --·-- -··-···-·-··--·-----·----
0 Stock solution containing 10 ugm of B per ml.
f The ug1n of B pt.•r detcnnination, since 1 ml is takrn. The ug1n of B ppr m1 of final solution is
1/11 these values, but all concentrations specifil'd in procP<lures lwrein refer to the concentration in
the 1-ml volume taken for analysis, not the final <.'otlCcntration.

tion is added and the 2 solutions are mixed by rotating the beaker. Finally,
the solution is evaporated to dryness in an oven (or in a bath) regulated 2 H
at 55 ± 3 °C, and the residue is baked at this temperature for 15 minutes
to insure dryness. The colored substance, rosecyaninc, is developed dur-
ing the evaporation and drying.
14-14. The beaker containing the dried residue is cooled to room tem-
perature. Then 25 ml of 95 per cent ethanol is added, the residue is tritu-
rated to extract the color, and the solution is filtered through a Whatman
No. 2 filter paper directly into the colorimeter tube. (Slight boron con-
tamination from the paper is not a factor here since the color-forming step
is past.)
14-15. The color is read with a 540-mu light maximum within 2 hours,

23 Naftel, Ind. Eng. Chem., A .E., 11 :407 (1939); Hafford, Ph.D. thesis, Univ. Wis.
(1942).
BORON DETERMINATIONS FOR SOILS AND PLANTS 375

since rosecyanine gradually hydrolyzes to curcumin; this is noticeable after

2 hours. If the percentage transmission is too low (less than 25 to 30 per
cent), indicating over 1.5 ugm of B, the solution is immediately reread with
a 580- or 600-mu light maximum. Reference is made. to the calibration
curve corresponding to the light maximum employed for the determination,
and the ugm of B contained in the 1-ml aliquot is thus determined.

ALTERNATIVE PROCEDURES

14-16. When an insufficient amount of boron is obtained in a 1-ml

aliquot, the solution can be concentrated by evaporation after being made
alkaline with Ca(OH) 2 solution, 24 and then dissolved in a small volume of
HCl of sufficient strength to give a slightly acid solution.

BORON DETERMINATION
(Colorimetrically with quinalizarin)
14-17. The quinalizarin procedure is suitable for the determination of
0.2 to 8 ugm of Bin the final aliquot. Careful regulation of the H 2 S0 4 con-
centration in the final solution to 89 per cent has been considered neces-
sary for reproducible color development. This is accomplished by obtain-
ing the B in 1 ml of aqueous solution and subsequently adding 10 ml of
the reagent in 98 per cent H 2 S04 • The intensity of coloration of quinalizarin
is sensitive not only to H 2 S0 4 concentration but also to temperature
changes. The blue coloration of the quinalizarin-boric acid complex in-
creases with decreasing temperature and increasing acid concentration, up
to about 90 per cent. This is attributed 25 to increasing the distortion of the
molecule rather than increasing the completeness of the reaction. The color
may be read visually (0.2 to 2.5 ugm of B) or by means of a photoelectric
colorimeter (0.5 to 8 ugm of B). Sensitivity of the quinalizarin method to
H 2 S04 concentration is greatly decreased 26 by increasing the quinalizarin
concentration to 45 mgm per liter (instead of the 25 mgm generally em-
ployed27). Ordinary concentrated H 2 SO4 ( 96 ± 1 per cent) can then be
substituted for the 98 per cent H 2 S04 without appreciably decreasing the
sensitivity of the method; and 2.5 ugm of F per ml of test solution does
not interfere with the boron determination.

APPARATUS

14-18. Needed apparatus consists of stoppered 20 x 100 mm shell vials

for visual comparison or colorimeter tubes and photoelectric colorimeter

24 Gooch, Am. Chem. J., 9:23 (1887).

25 Berger and Truog, Soil Sci., 57: 30 ( 1944).
26 MacDougall and Biggs, Anal. Chem., 24:566 ( 1952).
21 Olson and Berger, S.S.S.A. Proc., 11: 216 ( 1947).
376 BORON DETERMINATIONS FOR SOILS AND PLANTS
equipped with 620-mu light maximum, a 1-ml pipet, and a dispenser for
the quinalizarin-H 2 S0 4 solution (Fig. 14-1).

Mercury

Reagent

Fig. 14-1. Apparatus for dispensing quinalizarin-

H2S04 solution with protection from moisture of air. The
volume of the reagent is exactly reproducible, though not
necessarily exactly 10 ml. (After MacDougall and Biggs,
Anal. Chem., 24:566, 1952.)

14-19. The glassware must be boron-free, such as Kavalier or Corning

728. Ordinary soft glass may be used for most of the reagents. Borosilicate
glassware such as Pyrex is not used. All glassware is weathered in strong
HCl prior to use.
BORON DETERMINATIONS FOR SOILS AND PLANTS 377
14-20. Standardization of Colorimeter Tubes. In the standardization of
colorimeter tubes, a large number of clean, unscratched (preferably new)
tubes made of boron-free glass is taken for preliminary testing. To each
tube, 10 to 15 ml of the quinalizarin-sulfuric acid solution (11 14-27) is
added. The tubes and contents are allowed to stand at room temperature
for at least 1 hour to insure uniform temperatures and absence of air bub-
bles. The outside of each tube is wiped clean with a cloth dampened slightly
with distilled water, and then polished with a clean, dry, lint-free towel.
With a 620-mu light maximum in the colorimeter, the first tube is inserted
in it and the galvanometer is adjusted to give a reading of between 60 and
100 per cent transmission. The first tube is rotated to give a reading not
subject to small changes in the orientation of the tube, and this orientation
is marked on the tube. Each successive tube is placed in the colorimeter
and rotated to a position insensitive to rotation, the galvanometer is read
to the nearest 0.25 per cent transmission, and the results are recorded. Be-
tween each successive tube reading the colorimeter is adjusted to read ex-
actly the same as with the first tube observed.
14-21. A working set of tubes that agree within 1 or 2 per cent trans-
mission is selected. This selected set of tubes is rechecked ( 11 14-20), and
the tube representing the mode of the readings is selected as the primary
standard. The other tubes are then reread against the primary standard
tube, and the deviations from the standard in percentage transmission are
recorded respectively for each tube. The deviation for each tube is applied
as a correction each time it is used for a determination.
14-22. The high refractive index of the concentrated sulfuric acid solu-
tion magnifies the effect of scratches on the tubes. Care therefore must be
taken not to scratch the tubes in any way. In cleaning the tubes, the sulfuric
acid is poured out and the tubes are rinsed out with distilled water. A brush
is not used in the cleaning process. Occasional restandardization of the
tubes (1114-21) is made as a precaution.

REAGENTS

14-23. Needed reagents consist of brom phenol blue indicator, standard

boron solution ( 11 14-12) and the following 2 reagents.
14-24. Sulfuric Acid, 98 ± 0.5 Per Cent. A 454-gm (1-pound) bottle of
Merck reagent grade fuming H 2S04 is added to 978 gm of Grasselli 95 per
cent (concentrated) H 2 SO4 contained in a soft glass bottle. The solutions
are thoroughly mixed by rotation of the bottle. The concentration is de-
termined by weighing about 2 gm of the acid into a covered weighing bottle,
pouring the acid into a 400-ml beaker containing about 200 ml of boiled
and cooled water, the bottle and cover being quickly immersed. The acid is
378 BORON DETERMINATION S FOR SOILS AND PLANTS
titrated with 1 N NaOH with brom phenol blue as an indicator. Both hydro-
gens are replaced:

(14-1)

in which sis the gm of H 2 S0 4 taken.

14-25. The quantities of the 2 acids taken may with practice be selected
so that the mixture comes out 98 per cent each time. Proportionately more
fuming or 95 per cent H 2 S0 4 is added as needed to bring the concentration
to 98 ± 0.5, and the titration is then repeated. If the acid concentration is
much outside this range, the quinalizarin color development is not a maxi-
mum (~ 14-17).
14-26. Quinalizarin-Sulfuric Acid Solution. For visual measurement, 5
mgm of quinalizarin is dissolved per liter of 98 per cent H 2 S0 4 (prepared
above). The reagent is stored in a boron-free glass bottle under a dry
atmosphere. Incoming air is passed through an anhydrous CaC1 2 tube (Fig.
14-1).
14-27. For photoelectric colorimeter measurement, 25 mgm of quin-
alizarin is dissolved per liter of 98 per cent H 2 S0 4 , and stored as in the pre-
ceding paragraph.

PROCEDURE28

14-28. Visual Comparison. One ml of the standard or test solution

(Table 14-1) is placed in the 20 x 100 mm soft glass vial, with care to
measure exactly the 1 ml with the same pipet for all standards and test
samples. Then exactly 10 ml of the 98 per cent sulfuric acid solution con-
taining 5 mgm of quinalizarin per liter is added. The vial is stoppered im-
mediately and the contents mixed by whirling gently, with care not to allow
any of the acid solution to come in contact with the rubber stopper.
14-29. The concentration in the test sample is determined by direct
comparison to the series of tubes containing 0.0 to 2.5 ugm of B (Table
14-1). In making the comparisons, the stoppers are removed so that the
solutions can be viewed from the top.
14-30. The acid solutions are somewhat hygroscopic, and therefore be-
come diluted by moisture in the atmosphere when the tubes are opened for
comparisons. New color standards must be made up from time to time de-
pending on how frequently the tubes are opened.
14-31. Colorimeter Comparison. Exactly 1 ml of the standard or test
solution (Table 14-1) is transferred to a tube by means of the same 1-ml

28 Berger and Truog, Ind. Eng. Chem .. A .E.. 11: 540 (1939); McHargue and Hodg-
kiss, J.A.O.A.C., 25:311 (1942); Berger and Truog, Soil Sci., 57:25 (1944); Olson
and Berger. S.S.S.A. Proc .. 11:216 (1947). Dr. R. V. Olson of Kansas State College
generously assisted with the preparation of this section.
BORON DETERMINATIONS FOR SOILS AND PLANTS 379
pipet. Then exactly 10 ml of the 98 per cent sulfuric acid solution con-
taining 25 mgm of quinalizarin per liter is added to each colorimeter tube.
The tube is stoppered immediately and the contents is mixed by whirling
the tube gently. Care is taken not to allow any of the acid solution to come
in contact with the rubber stopper. The tubes are allowed to come to room
temperature by standing for at least 2 hours. The outsides of the tubes are
cleaned with a cloth moistened in distilled water followed by polishing
with a dry lint-free towel.
14-32. The blank (0.0 ugm of B) is placed in the colorimeter with
620-mu light maximum, and the colorimeter is adjusted to give 100 per
cent light transmission. The percentage transmission is read for all of the
other tubes to the nearest 0.25 per cent transmission. Then the readings
are corrected with the corresponding tube correction ( ~ 14-21 ) . Duplicates
should agree within 0.5 per cent transmission.
14-33. The Standard Curve. Since the room temperature variation
causes a shift in the standard curve, 1 of the standard boron solutions
(containing 1.0 to 1.5 ugm of B) is saved and designated the "standard
tube" for correcting the curve to the original temperature for each set of
determinations. The "standard tube" is placed in the colorimeter at the be-
ginning of each set of determinations of test solutions and the percentage
transmission is adjusted to read the same as was obtained in making up the
standard curve, thus correcting the entire curve to the temperature of the
test solutions. The percentage transmission readings are plotted against
ugm of B. The color is stable except for the effect of moisture absorbed
from the air.
14-34. The calibration curve obtained varies slightly with changes in the
amounts of sulfuric acid and quinalizarin in the solutions so that it is nec-
essary to make a new curve each time a new 98 per cent sulfuric acid-
quinalizarin solution is prepared.

WATER SOLUBLE BORON IN SOILS

14-35. The range of water soluble boron in mineral soils of the humid
region is generally from 0.2 to 1.5 ppm, and ranges on up to 2 or more
ppm in muck soils and down to 0.2 in fairly fertile sandy soils. Soils of
the semiarid regions may fall in the same ranges but occasionally contain
I 0 to 40 ppm or more. The water soluble boron is considered to be the
form immediately "available" to plants. The boron extracted in a 1 : 2
soil : water ratio is given as the common procedure; that extracted by leach-
ing is given in the alternative procedure. The analytical determination may
be either by the curcumin ( ~ I 4-13) or quinalizarin ( ~ 14-28) procedure.
14-36. The level of water soluble boron may vary greatly over short
distances in the field. Sampling at the time a growing crop such as alfalfa
is on the field permits observations of areas of acute boron deficiency.
380 BORON DETERMINATIONS FOR SOILS AND PLANTS
Separate soil samples are drawn from suspected deficient areas and com-
pared to areas on which normal growth occurs.
14-37. Nitrates are generally below interfering levels of 40 pp2m in soil
samples taken in the spring before nitrification has begun or taken under a
growing crop. Fallowed or incubated soils may develop sufficient amounts
of nitrate to cause interference and nitrate must be removed ( ~ 14-43).
APPARATUS

14-38. Needed in addition to apparatus for the colorimetric determina-

tion are a torsion balance, 125-ml boiling flasks (boron-free glass) fitted
with a I-hole stopper and condenser, 100-ml centrifuge tubes and centri-
fuge, a 20-ml pipet, 9-cm porcelain evaporating dishes, small glass fun-
nels, and 9-cm filter paper (boron-free, E and D or Whatman No. 2).
REAGENTS

14-39. Reagents needed include distilled water and 1 N CaCl 2 for ex-
traction for the curcumin determinative procedure, unless soil nitrates ex-
ceed 20 ppm. In the latter case, additional reagents needed include approxi-
mately 0.1 N HC1 and saturated Ca(OH) 2 (protected from C0 2 ). In
preparation for the quinalizarin determinative procedure, 0.36 N H 2 S04 is
needed instead of 0.1 N HCI.
PROCEDURE29

14-40. Equilibrium Extraction. To a 20-gm sample of air-dried and

sieved soil placed in a 125-ml boiling flask, a 40-ml volume of distilled
water is added, and a reflux condenser is attached. (Five or 10 gm or other
weight of soil sample may be taken, so long as the soil : water ratio is kept
at l : 2.) The suspension is boiled for 5 minutes and allowed to cool, and
the condenser is disconnected.
14-41. A blank is determined by placing 40 ml of water in the extraction
flask and boiling for 5 minutes as in the soil extraction. This solution is
carried through all steps in the subsequent procedure.
14-42. The soil suspension is transferred to a centrifuge tube, 2 or 4
drops (not to exceed 5 drops) of a I N solution of CaC1 2 is added, and
the suspension is centrifuged for 5 to 10 minutes. In lieu of centrifuging,
the CaC1 2 -treated suspension may be filtered.
14-43. For the curcumin determinative procedure, 1 ml of the clear
supernatant liquid is taken directly for the test ( ~ 14-13) unless soil nitrates
exceed 20 ppm, in which case the solution must be made alkaline, evap-
orated to dryness, and ignited as in the quinalizarin procedure.
14-44. For the quinalizarin determinative procedure, a 20-ml aliquot of

2 9 Essential features are of the procedure of Berger and Truog, Soil Sci., 57:32
(1944).
BORON DETERMINATIONS FOR SOILS AND PLANTS 381
the clear supernatant solution is transferred by means of a pipet to a
porcelain evaporating dish, 2 ml of saturated Ca(OH) 2 solution is added,
and the entire solution is evaporated to dryness on a steam hot plate. It is
essential that the extract be alkaline before evaporation so that, if the soil
is extremely acid or has had acid treatments, the H:1B03 will not volatilize.
If the acid treatments have been such as to exceed the 2 ml of Ca (OH) 2
solution, more of the latter is added to insure alkalinity.
14-45. The residue is ignited gently to destroy nitrates and all organic
matter. It is then cooled and 5 ml of approximately 0.36 N H~S0 4 is em-
ployed if the quinalizarin procedure is to be employed ( ~ 14-28). Five ml
of 0.1 N HCl is added if the curcumin procedure is to be employed (~ 14-
13). The residue is triturated thoroughly with a policeman to dissolve all
soluble matter. The solution is filtered through a 9-cm filter paper on a
small funnel. A 1-ml aliquot of the clear filtrate is taken for color develop-
ment.
14-46. Calculation of Results. The ppm of water soluble B in soil is
calculated from the ugm of B in solution, minus the blank, by means of the
following equation:
ppm of water soluble B in soil = ugm B in 1 ml of solution tested
extractant
x 40 ml
. ·-·-·-······-···--
20 gm soil
1
x
ml extractant represented in test
(14-2)
ALTERNATIVE PROCEDURES

14-47. Extraction with a more dilute soil: water ratio, such as 1 : 5, 30

causes an increase in the B extracted, in contrast to the behavior of simple
water soluble anions such as nitrate ( ~ 14-1). An equilibrium apparently
exists between the soil boron and the water solution (~ 14-2, 14-49).
14-48. In a continuous leaching procedure, 31 the soil sample is placed in
a soxlet thimble supported in the neck of a Kjeldahl type flask made of
boron-free glass. Boiling the solution in the flask drives steam into a con-
denser in the top of the flask that drops the water into the thimble. Leach-
ing for 6 hours results in extraction of as much boron as 12 or 24 hours,
indicating completeness of extraction.
14-49. The boron extracted by continuous leaching, L, is related to the
equilibrium extracted boron, E ( ~ 14-40) ,
E = 0.24L + 0.01 (14-3)
A correlation coefficient of 0.94 was noted. The quantity L is not particu-

30 Haas, Soil Sci., 58: 123 (1944).

31 Mcclung, S.S.S.A. Proc., 15:268 (1951).
382 BORON DETERMINATIONS FOR SOILS AND PLANTS
larly dependent on the quantity of soil in ratio to solvent, and decreases by
an amount equal to the boron removed by cropping (r = 0.89). As long
as 3 months may be required for re-establishment of the equilibrium level
of boron after it has been lowered by cropping (as measured by the con-
tinuous leaching method). The continuous leaching method may offer a
more absolute measure of the quantity of boron available to plants than an
equilibrium extraction, although the latter can be equally well calibrated
for crop responses to boron.
14-50. Water soluble boron has also been extracted:1 2 with a 1 : 1.5
soil : water ratio and boiling for 15 minutes; followed by filtration and 3
successive similar extractions with a I : I soil : water ratio. Approximately
60 per cent of the boron extracted was obtained in the first of the 4 equi-
librium extractions, compared to equilibrium extractable boron of 24 per
cent of that extracted by continuous leaching ( ~ 14-49).
ACID SOLUBLE BORON:i:i
14-51. Treatment of the soil with 85 per cent HaP0 4 extracts the boron
present in the organic colloids and precipitated forms, but excludes that in
tourmaline ( 3 .5 per cent B) and other resistant borosilicate minerals.
Other acids, such as HCl and H~SO.P are less satisfactory. The acid soluble
boron averages 17 ppm for 203 soils and is a small fraction of the total for
stony mineral soils, approximately half for many fine-textured mineral
soils, and a large fraction of the total for soils that are high in mineral and
organic colloids.
14-52. The titration (~ 14-58) of H:1B0a requires 0.1 to 2 mgm of B. It
is based on the formation of a complex between boric acid and mannite that
permits titration of the first hydrogen of boric acid with a standard hydrox-
ide solution. Interfering substances include other weak acids, 4-valent
germanium, and hexavalent tellurium,: 14 all of which are removed by steam
distillation of H:iBO:i as the methyl ester (~l 14-56) or are ordinarily rare
in the preparations from soils or plants.

APPARATUS

14-53. Needed are a round-bottom 500-ml extraction flask with short

neck (all glassware should be boron-free), 250-ml beaker, a platinum dish,
alcohol "steam" distillation apparatus, a buret, and a pH meter.

REAGENTS

14-54. Reagents needed include 85 per cent HaP0 4 , boron-free anhy-

drous methanol, 2 N and 0.1 N HCI, 0.5 N, 0.2 N and 0.02 N NaOH

32 Whetstone et al .. U.S.D.A. Tech. Bui. 797 ( 1942).

aa Whetstone et al .. U.S.D.A. Tech. Bui. 797 (1942).
34 Eaton and Wilcox, U.S.D.A. Tech. Bui. 696 ( 1939).
BORON DETERMINATIONS FOR SOILS AND PLANTS 383
(carbonate-free, the latter standardized), saturated Ca (OH) 2 , neutral man-
nite, and brom thymol blue indicator.
PROCEDURE:iil

14-55. Extraction. A 50-gm sample of air-dry soil that has passed a

2-mm sieve, is placed in a 500-ml short necked, round bottom, flask. Then
7 5 to 100 ml of 85 per cent H 8 PO 4 is added, the amount of acid depend-
ing upon the composition of the soil. Sandy soils require less than soils high
in bases and carbonate. The soil and acid are thoroughly mixed and heated
on a steam bath overnight. Any lumps are disintegrated by means of a
stirring rod, and the mixture is cooled. A blank is run on the reagents.
14-56. Distillation. Fifty to 100 ml of anhydrous methanol is added to
the cooled digest, the flask is then connected to the distillation set up, and
the mixture is agitated. The anhydrous alcohol in the still reservoir should
be boiling when the digestion flask is connected to prevent the soil mixture
from stopping up the inlet tube. The methyl borate is "steam-distilled"
(with alcohol vapors from the reservoir) until about 500 ml of distillate has
been collected. During the major portion of the distillation, the volume of
the mixture in the digestion flask is kept constant by regulating the burners.
Toward the end the volume may be somewhat decreased.
14-57. Ignition. The distillate is made just alkaline to brom thymol blue
with 0.2 N Na OH solution, and then 2 ml of saturated Ca (OH)~ is added.
The methanol is then distilled off and recovered. The aqueous residue is
transferred to a platinum dish, evaporated to dryness, and gently ignited to
destroy the organic matter.
14-58. Titration. The residue is taken up in water and transferred to a
250-ml beaker or conical flask, the dish being rinsed with a few drops of
2 N HCI. The solution is diluted to 150 ml, 2 or 3 drops of 1 per cent
brom thymol blue indicator solution are added, and the solution is acidified
with 2 N HCI. The solution is then boiled to expel C0 2 and more acid is
added as necessary to maintain slight acidity. The flask is then cooled by
immersion in cold water. The solution is titrated, brom thymol blue indi-
cator or a glass electrode pH meter being employed. To perform the titra-
tion, the solution is adjusted to a definite pH that is near neutrality with
0.5 N Na OH (carbonate-free), 0.1 N HCI, and finally with 0.02 N NaOH.
Then 5 gm of neutral mannite is added, and the solution is titrated with
standard 0.02 N NaOH to exactly the initial pH. The NaOH is standardized
against known amounts of HaBOa in the same manner.
ALTERNATIVE PROCEDURE

14-59. Colorimetric Determination. The residue from evaporation of the

distillate may be dissolved and a small aliquot used for colorimetric de-

35 Whetstone et al., U .S.D.A. Tech. Bui. 797 (1942).

384 BORON DETERMINATION S FOR SOILS AND PLANTS
termination of boron by means of the curcumin ( ~ 14-13) or quinalizarin
(~ 14-28).

TOTAL BORON IN SOILS

14-60. The total boron content of 118 soils averaged 36 30 ppm. The
concentration in igneous rocks averaged 37 10 ppm, and in sea water, 38 4.5
ppm. The total boron in soils ranged from 4 to 98 ppm, and fine-textured
humid soils analyzed about 30 to 60 ppm. Sandy soils often have as low as
2 to 6 ppm. Of the total, that portion that is strong acid insoluble is at-
tributable to tourmaline ( ~ 14-51 ) .

APPARATUS

14-61. Needed apparatus includes an analytical balance, a platinum

crucible, a 250-ml beaker and glass cover (boron-free glassware through-
out), a 500-ml volumetric flask, 100-ml centrifuge tubes and centrifuge, a
600-ml beaker, a rubber policeman, a platinum dish, a funnel, 9-cm boron-
frce filter paper, and a 500-ml graduated cylinder.

REAGENTS

14-62. Needed reagents include anhydrous Na~C0 3 , approximately 4 N

H 2 S0 4 (50 ml of concentrated H 2 S04 to 400 ml of water), 0.1 N HC1
( curcumin procedure), and boron-free ethanol or methanol (redistilled
in boron-free glassware, if necessary). For the quinalizarin procedure, ap-
proximately 0.36 N H 2 S04 is employed instead of 0.1 N HCI.

PROCEDURE:m

14-63. Fusion. A 0.5-gm sample of a soil that has passed a 0.16-mm

( 100 meshes per inch) sieve is placed in a platinum crucible. Then 3 gm of
anhydrous Na 2 C0:1 is placed in the crucible and mixed with the sample
with a glass rod. The mixture is fused over a Meker burner until the re-
action is complete, and then the crucible is cooled and placed in a 250-ml
beaker containing about SO-ml of distilled water. The beaker is covered
with a glass, and 4 N H 2S0 4 is added from time to time (about 14 ml of
4 N H 2 S04 is required) until the melt has disintegrated and the solution
has a pH of 6.5 or below (tested with brom thymol blue employed as an
external indicator). The melt is triturated from time to time with a rubber
policeman to hasten the disintegration.
14-64. With each set of determinations a duplicate set of blanks is run

36 Whetstone, et al., U.S.D.A. Tech. Bui. 797 ( 1942).

:n Clarke and Washington, U.S. Geo!. Surv., Prof. Paper 127 (1924).
38Moberg and Harding, Sci., 77:510 ( 1933 ).
39Fusion and separation after Berger and Truog, Soil Sci., 57: 25 (1944), as slightly
modified by Dible et al., Anal. Chem., 26:418 ( 1954).
BORON DETERMINATIONS FOR SOILS AND PLANTS 385
by the addition of all reagents except the soil, the same procedure being
carried out as with the soil. The blank is employed for the l 00 per cent
transmission setting to find the ugm of B found for the soil.
14-65. The resulting suspension is transferred to a 500-ml volumetric
flask, the beaker and crucible being washed several times with distilled
water and the washings being added to the flask. The total volume of the
solution should not exceed 150 ml. Then methanol or ethanol is added to
the flask to make a volume of nearly 500 ml and the contents are mixed
thoroughly. Flecks of Na 2C03 are added with mixing until the solution is
slightly alkaline. Then the solution is brought to full volume with alcohol.
During this time the excess of Na:!S0 4 is thrown out of solution. The sus-
pension is centrifuged (or filtered) until the supernatant liquid is clear.
14-66. A 400-ml aliquot of the clear supernatant solution is placed in a
600-ml beaker and 100 ml of distilled water is added to prevent subsequent
precipitation. The solution is evaporated to a small volume, then trans-
ferred to a platinum dish and finally evaporated to dryness and ignited
carefully. The dish and residue are cooled and then 5 ml of 0.1 N HCl (5
ml of 0.36 N H:!S0 4 for the quinalizarin procedure, ~ 14-28) is added and
the residue triturated thoroughly with a policeman. This solution is then
filtered and a 1-ml aliquot is taken for the boron determination with cur-
cumin (~I 14-13).
14-67. Calculation of Results. The ppm of total boron in soil is deter-
mined as follows:
ppm B in soil = ugm B per ml in solution tested x 12.5 ( 14-4)
when the factor 12.5 represents:
5 ml final solution 500 ml volume
0.5 gm s01I. x 400 ml evaporated

TOTAL BORON IN PLANTS

14-68. The range of boron contents of plants is wide, but 4 to 10 ppm
can be taken as typical for cereals such as oats, 20 to 50 ppm for legumes
such as alfalfa, and 20 to 100 for crops such as turnips and beets. The
content in plants that are deficient in boron varies widely with the plant
part and the species. A content of 8 ppm has been shown 40 to be the con-
tent in the apical portion of boron-deficient alfalfa plants, whereas the
lower part of the plant contained 30 ppm. This result emphasizes the need
for care in systematic sampling of plant parts for boron analysis, since
boron is not very mobile in the plant. A content of 28 to 30 ppm has been
reported to the critical content in red clover. 41 In the following procedure

40 Dible and Berger, S.S.S.A. Proc., 16:60 (1952).

41 Tucker and Smith, S.S.S.A. Proc., 16:252 (1952).
386 BORON DETERMIN ATIONS FOR SOILS AND PLANTS
the plant is ground and ashed, the boron in the ash is extracted with dilute
acid, and finally the boric acid is separated by a filtration step without
distillation.
APPARATUS

14-69. Needed apparatus for the ashing and extraction consists of a

9-cm porcelain evaporating dish, a flame or muffle furnace, 10-ml and 1-ml
pipet, a rubber policeman, a small centrifuge tube and centrifuge or a
funnel, and 9-cm filter paper. A Christy mill is employed for the tissue
grinding.
REAGENTS

14-70. Reagents needed for the extraction and preparation of plant

tissue for the analysis include 0.1 N HCl (0.36 N H 2S0 4 for quinalizarin
procedure) and solid Ca(OH) 2 for use with seed tissue.
PROCEDURE42

14-71. Ignition. A 0.50-gm sample of oven-dry and finely ground plant

tissue is placed in a porcelain evaporating dish. Vegetative tissue ordinarily
contains enough bases to prevent Joss of B on ignition; 4 a for seeds, particu-
larly oily seeds, a little base such as Ca(OH) 2 may be added. The tissue
is ignited gently to a white or gray ash, over an open flame or in a muffle
furnace at 550°C.
14-72. Extraction. The dish and contents are cooled and then 10.0 ml
of 0.1 N HCI (or approximately 0.36 N H 2 SO4 for the quinalizarin pro-
cedure) is pipetted into the dish, and the residue is triturated with a police-
man. After the suspension is filtered or centrifuged until clear, it is ready
for the boron determination. An aliquot of 1.0 ml is taken for either the
curcumin or quinalizarin procedure ( ~ 14-13 or 14-28).
14-73. With each set of determinations, 2 reagent duplicate blanks are
carried throughout the procedure, and the ugm found in the blank is sub-
tracted from that found in the sample.
14-74. Calculation of Results. The content of boron in the plant tissue
is obtained as follows:
ppm Bin plant tissue= ugm Bin 1 ml solution tested x 20 (14-5)

in which the factor 20 arises from 10 ml solution/0.5 gm sample.

ALTERNATIVE PROCEDURES

14-75. A plant tissue sample of 10 to 15 gm may be treated 44 with 75 ml

42 Dible et al., Anal. Chem., 26:418 (1954).

43Berger and Truog, Ind. Eng. Chem .. A. E., 11:540 (1939), Soil Sci., 57:25
(1944); McHargue and Hodgkiss, J.A.O.A.C., 24:518 (1941 ).
44 Whetstone et al., U.S.D.A. Tech. Bui. 797 (1942).
BORON DETERMINATIONS FOR SOILS AND PLANTS 387
of HaP0 4 and the HaBOa immediately distilled from the mixture as for acid
soluble boron (~ 14-55). Also, the plant tissue may be ashed followed by
determination by titration. 45
14-76. Boron has been determined 4 ff on fresh 'strips of plant tissue with-
out the necessity for ashing. The direct determination was preferred be-
cause of noted losses of boron on ignition of organic materials at tempera-
tures as low as 100°C, possibly as volatile esters in organic constituents of
the plants, and also because of the boron contamination noted even in the
double acid washed filter paper. The procedure was to cut 5-cm strips from
leaves and to determine the color by addition of the curcumin directly to
the strips. Results were reported as ugm of B per 10 cm~ of leaf tissue.

QUESTIONS
1. What 4 elements essential to plants normally occur in soils as anions?
2. Which of these anions occurs in soils largely in water soluble form?
3. Of which of these anions can the water soluble form he correlated very
closely with plant response?
4. State the relationship of boron availability in a given soil to hydroxyl ion
activity adjusted with NaOH.
5. Because of the reversion of boron compounds to water insoluble forms in
alkaline pH ranges, is boron likely to he deficient in alkaline soils?
6. List the chief methods hy which boron has heen determined commonly.
7. Name the 2 most commonly used colorimetric methods for boron.
8. What is the principle utilized in obtaining 2 concentration ranges in the
curcumin method for boron determination?
9. What are the effects of sulfuric acid concentration and temperature on the
blue color intensity of the quinalizarin-boric acid molecule?
10. What is the purpose of the addition of Ca(OH) 9 before evaporation of
the boron solution derived from acid extraction of plan( ash or from the fusion
analysis?
1 I. State the quantitative relation between equilibrium-extracted water sol-
uble boron from soils ( 1 : 2 soil : water ratio) to the water extractable boron
obtained by continuous leaching.
12. State the general concentration ranges of boron in soils in (a) water
soluble form, (h) acid soluble form, (c) total.
13. How can the content of the mineral tourmaline of soils he estimated by
boron analysis?
14. What is the normal boron content as ppm of tissue of different classes of
plants?

4r. Wilcox, Ind. Eng. Chem., A.E., 12:341 (1940).

46 Winsor, Anal. Chem., 20: 176 (1948).
 15
Iron, Manganese, Copper, Zinc,
Molybdenum, and Cobalt
]]eterminations
More minor elements are found to be e.1·se11tia/ to plants as
methods are refined.
-A TREND IN MINOR ELEMENT CHEMISTRY

15-1. The elements needed by plants only in trace amounts- minor

elements in amounts but not in importanc e-are iron, manganese, copper,
zinc, molybdenum, cobalt (considered here), and boron (considered in
Chapter 14). Sometimes magnesium, sodium 1 (Chapters 5 and I I ), and
sulfur (Chapters l 0 and l I ) are also considered to be minor elements.
Cobalt is unlikely to be deficient enough in soils to affect plant growth, but
it is important in animal feeds, which are supplied with it by plant uptake
from soils. For this reason the soil chemist is sometimes called on for soil
and plant analyses for cobalt ( 1) 15-115).
15-2. The 6 elements, Fe, Mn, Cu, Zn, Mo, and Co, when added to soils
in various moderately soluble compounds, tend to revert largely to forms
of much lower solubility and availability than common exchangeable ca-
tions such as calcium and magnesium (Chapter 5). The extent and rapidity
of this conversion depends upon the soil pH. Increasing soil acidity favors
higher activity of all of them except Mo, which becomes less active, possibly
by precipitation as the extremely insoluble MoS 2 (molybdenite) and MoO:i·
Increasing soil pH, particularly above pH 7, favors their conversion to
oxides, hydroxides, and silicates, in which form the activity of all but Mo

1 Volk, J. Am. Soc. Agron., 37:821 (1945).

388
IRON DETERMINATIONS 389
decreases toward the point of deficiency in plants. Molybdenum appears
to go over to anionic molybdate form of sufficient activity t() be adequately
available or even toxic.
15-3. The chemistry of soil availability of these 6 elements largely re-
mains to be worked out. The water soluble and exchangeable forms are, of
course, assumed to be readily available. But other soil forms are of suffi-
cient activity to be important in the nutrition of crop plants. Thus dilute
acid soluble and easily reducible iron and manganese are determined. Total
and strong acid soluble copper, zinc, and cobalt appear to be related to
availability of these elements in soils. Total molybdenum has been de-
termined as a measure of the adequacy of supply of this element in soils.
15-4. The chief methods for the determination of these elements are
spectrochemical and polarographic. Of the spectrochemical methods, the
absorption spectrophotometric methods are considered in this chapter,
whereas the emission spectrophotometric and spectrographic methods are
considered in Chapter 18. The polarographic method is considered in Chap-
ter 16.
15-5. First the chemical determination is described for an element and
then consideration is given various methods for extraction of the element
from soil. Procedures for extraction from plant tissue are given in Chapter
12. A systematic procedure for 12 nutritionally important elements in
plant tissue including the 6 involved in this chapter has been described. 2

IRON DETERMINATION
( Colorimetrically as o-phenanthroline red ferrous complex~)

15-6. Procedures are given here for iron (in soils) that is exchangeable,
readily soluble in dilute acids, or easily reducible. Iron determination by
orthophenanthroline, as described here, is a sensitive and fitting method for
the small amounts likely to be extracted for these analyses or from plants
( ~ 12-19). That extracted from soils is readily determined with SCN ( ~
13-71, 13-93). The larger amounts of total elemental Fe are readily de-
termined by reduction with amalgamated zinc ( ~ 11-142) or, when Ti is
to be determined also, with Tiron (~I 11-62).
15-7. The iron is determined by reduction to ferrous with hydroxylamine
hydrochloride, and formation of a ferrous complex of orthophenanthroline,
a chelate ring compound of intense red color. A slightly acidic reaction
must be used, since some cations will interfere in the alkaline ranges because
of precipitation of their hydroxides. No common soil anions interfere ex-
cept orthophosphate through precipitation of Ca and Fe. Twenty ppm of

2 Parks eta/., Ind. Eng. Chem .. A.E .. 15:527 (1943).

asaywell and Cunningham, Ind. Eng. Chem., A.E., 9:67 (1937); Fortune and
Mellon, Ind. Eng. Chem., A.E., 10:60 (1938).
390 IRON DETER MINAT IONS

P 20 5 from (NH 4 ) 2HP0 4 caused an error 4 of 1.4 per cent, but this quantity
is not commonly encountered with the amount of iron determined in soil
or plant extractions.
APPARATUS

15-8. Needed apparatus includes a colorimeter with 490-mu light maxi-

mum, colorimeter tubes with a 20-ml calibration mark, pipets for taking
aliquots, 1-liter and 100-ml volumetric flasks, and a pH meter or an indi-
cator spot plate.
REAGENTS

15-9. Needed reagents include dilute HCl and NH 4 0H, 2,4-dinitro-

phenal indicator, and the following special reagents.
15-10. Orthophenanthroline, 1.5 Per Cent in 95 Per Cent Ethanol. To
100 ml of 95 per cent ethanol is added 1.5 gm of the white crystalline
orthophenanthroline, and the solution is stirred until the crystals are dis-
solved. (This reagent should not be confused with the oxidation reduction
indicator "Ferroin ," which is the ferrous complex of orthophenanthroline,
sometimes termed "orthophenathroline" in laboratory practice.)
15-11. Hydroxylamine Hydrochloride, 10 Per Cent. Ten gm of hy-
droxylamine hydrochloride is dissolved in 100 ml of distilled water.
15-12. Standard Iron Solution. A piece of pure iron wire weighing ap-
proximately 0.10 gm, or 0.7023 gm of Fe(NH 4 ) 2 (S0 4 ) 2 • 6H 2 0, is
weighed out. The iron wire is dissolved in 20 ml of 0.6 N HCI with warm-
ing if necessary; or the salt is solved in water and 20 ml of 0.6 N HC1 added.
The solution in either case is transferred to a liter volumetric flask and
diluted to volume with water. This solution contains 100 ppm Fe. From
this solution, a second solution is prepared containing 10 ppm Fe by dilu-
tion of 10 ml to 100 with water. Aliquots (0.5, 1, 2, 3, 4, 6 ml) are then
taken for the Fe calibration curve ( ~ 15-13), giving 5 to 60 ugm of iron.

PROCEDURE

15-13. An aliquot of iron in dilute HCI solution (standard or test solu-

tion) containing from 5 to 60 ugm of Fe is placed in the colorimeter tube
graduated at 20 ml and diluted to about 5 or 6 ml. The pH of the test solu-
tion is adjusted to pH 1.5 to 2.7 with dilute HCl or NH 4 0H (the standards
fall in a suitable range 5 of pH t.08 to 3.1 without adjustm ent). A pH
meter can be applied to a separate aliquot, or a test can be made with
2,4-dinitrophenol (yellow above pH 2. 7) on an indicator spot plate of a

4 Hummel and Willard, Ind. Eng. Chem .. A.E., 10: 13

( 1938). These authors em-
4.5 for re-
ployed I per cent hydroquinone solution buffered in NaOAc-H OAc of pH
of the iron, in a solution of pH 3 to 4 to avoid precipita tion of calcium
duction
phosphate with occlusion of iron.
5 Range limits determin ed by I. Hashimo to in the author's
laborator y (1957).
IRON DETERMINATIONS 391

small drop carried on a glass rod. Next, 2 ml of 10 per cent hydroxylamine

hydrochloride and 1 ml of 1.5 per cent orthophenanthroline solution are
added. The solution is made up to volume with distilled water, and read in
the colorimeter with light of 490 mu maximum. The color reaches maxi-
mum strength immediately and is stable for months.
15-14. The concentration of iron in solution is read from the calibration
curve drawn from the readings of the standard iron solutions.

EXCHANGEABLE FERROUS IRON°

15-15. Exchangeable ferrous iron is not completely extracted from soils
by neutral ammonium acetate in the usual procedure for exchangeable ca-
tions because it oxidizes during the time for ordinary extraction and the
ferric ions are largely precipitated. A separate extraction of the soil sample
is therefore carried out for the determination of exchangeable ferrous iron.

APPARATUS

15-16. Needed apparatus includes a 11-cm Buchner funnel, a 1000-ml

suction flask and rubber adapter for the funnel, 11-cm Whatman No. 5
filter paper, an aspirator pump, and a 500-ml conical extraction flask. The
filter funnel and paper are set in readiness before the extraction is begun.

REAGENTS

15-17. Needed reagents include neutral, 1 N NH 4 0Ac solution, 400 ml

per extraction.

PROCEDURE

15-18. From a freshly taken soil sample, field-moisture condition un-

changed (moisture determined later on separate sample), a 25-gm sample
is quickly weighed out, placed in a 500-ml conical flask, and 250 ml of
neutral 1 N NH 4 0Ac solution added. The suspension is shaken vigorously
for 20 to 30 seconds and filtered quickly on the previously prepared Buch-
ner funnel fitted with Whatman No. 5 filter paper. Three successive 50-ml
portions of NH 4 0Ac solution are employed for further extraction of the
soil. To avoid oxidation of ferrous iron, the entire extraction is completed
in 5 minutes or less. 7 Walkley 8 noted that the ferrous iron was rather com-
pletely oxidized in soils during the operation of air-drying for l or 2 days,
even those which had been highly reduced (smelled strongly of H 2 S). This
observation was made in connection with the interference of ferrous iron

11 Procedure for exchangeable ferrous iron kindly suppled by Dr. G. D. Sherman,

Hawaiian Agr. Exp. Station, Honolulu, Hawaii, by personal communication.
7 Oxidation of the ferrous iron can amount to 20 per cent in 15 minutes, and to
80 per cent in 60 minutes.
s Soil Sci., 63: 261 ( 1947).
392 IRON DETERMINATIONS
. The ferrous
with chromic acid titration of soil organic matter ( ~ 9-40)
immediately
iron determination must be carried out on the field-moist soil
following the collection of the samples.
is freed of
15-19 . The filtrate, containing exchangeable ferrous iron,
tion of the iron of no conseq uence here)
NH4 0Ac by evaporation (oxida
of organi c matter arc remov ed by
on a steam hot plate. The last traces
regia, and then the solutio n is
treatm ent of the residue with 10 ml of aqua
into solutio n in 1 ml of 1 N HCl
again brought to dryness. The iron is taken
determination
followed by water to give the correct dilution for the iron
(~ I 5-13) .

ALTERNATIVE PROCEDURES
, and
15-20 . Sherman added 20 ml of concentrated HCl to the filtrate
ml of 8 N HN0: to oxi-
evapo rated it to a small volume, then added a few 1

clear solutio n of small

dize the organic matter, and evaporated to a water
plate, on which
volume. This procedure is suitable for use on an electric hot
the evaporation to dryness is likely to cause spattering.
a 1-minute
15-21 . Paddic k11 tested alkaline soils for available iron by
approximate
extraction with thioglycolic acid. Morgan 10 described a rapid
qualita tive test for ferrou s iron in soils
soil test for ferrous iron. A rapid
is given in ~ 13-71.
IC IRON
COMBINED EXCHANGEABLE FERROUS AND FERR
AND DILUTE ACID SOLU BLE IRON
Fe++ and
15-22 . Exchangeable iron exists in soils as (a) ferrous,
can be extrac ted by a
(b) [Fe(O H)n](:1 - 111 +. Although the ferrous iron
ly extrac ted
rapid procedure with NH 4 0Ac, the ferric iron is only partial
e some of
even though replaced from the exchange charges by NR1 +, becaus
are someti mes
it precipitates. Rather large amounts of iron and aluminum
neutra l NH 0Ac. Extrac tion with NH.10Ac at
extracted from acid soils by 4
iron in solutio n, and has been applied
a pH value of 3 keeps all of the ferric
ous clays. 11 However, in
successfully in ferric exchange studies with alumin
extracted. This
soils and ferruginous clays, some nonexchangeable iron is
its plant availab ility.
dilute acid soluble iron may be of significance to

APPARATUS
a 1000-ml
15-23 . Needed appara tus includes a 11-cm Buchn er funnel,
the funnel , an aspira tor pump, and
suction flask and rubber adapte r for
Whatm an No. 5 filter paper.

9 S.S.S.A . Proc., 13: 197 (1949).

10 Conn. Agr. Exp. Sta. Bui. 372:457 (1935).
11 Bower and Truog, S.S.S.A . Proc., 5:86 (1941
).
MANGANESE DETERMINATIONS 393

REAGENTS

15-24. Needed reagents include 1 N NH 40Ac, of pH 3.0, made from

approximately neutral NH 4 0Ac by the addition of concentrated HCI.

PROCEDURE

15-25. A 25-gm sample of soil is extracted as for exchangeable ferrous

iron (~ 15-18) except that the NH 4 0Ac of pH 3 is employed. The same
speed of extraction is maintained to avoid undue contact of the acid ex-
tractant with the soil.

ALTERNATIVE PROCEDURE

15-26. Centrifuge tubes and centrifuge may advantageously be substi-

tuted for the Buchner funnel filtration for studies of ferric iron as an ex-
changeable cation in clays.

SOLUBLE, EXCHANGEABLE, AND EASILY REDUCIBLE

MANGANESE OF SOILS
15-27. The extraction of water soluble and easily reducible manganese
are the chief concern here, but the exchangeable Mn++ is also extracted
in the system described. The analytical determination of Mn as MnO 4 -
with periodate has been outlined in connection with the exchangeable metal-
lic cations ( ~ 5-74). Extraction of the total Mn in soils ( ~ 1 1-172) and
plants ( ~ 12-19) has been given.
15-28. There is an equilibrium of manganous and manganic forms of
manganese in soils, and this is related to availability of manganese to plants;
the exchangeable (divalent) form was considered to be the best index of
plant availability. 12 Easily reducible manganese oxides are distinguished
by the test from those in which the Mn is rather inert and unavailable.
Easily reducible manganic oxides may involve mixtures of Mn valences
from 2 to 4. The equilibrium system may be represented:
Water sol. Mn•+~ Exchangeable Mn"~ Colloidal hydrated Mn01,., 2 ~Inert Mn02
( 15-1)

15..:.29. Strongly acid soils that contain 25 ppm or less of easily reducible
Mn are likely to become manganese deficient when limed. 13 Individual soils
vary greatly as to the limits in ppm of Mn under which the manganese will
become deficient. 14 Some acid Kentucky soils with only 3 to 10 ppm of
easily reducible manganese showed 15 over-liming injury (manganese de-

12 Piper, J. Agr. Sci., 21 :762 (1931 ).

ia Leeper, Proc. Roy. Soc. Victoria, 47(II) :225 (1935).
14 Leeper, Soil Sci., 63: 79 (194 7).
15 Sherman et al., Soil Sci., 54: 253 ( 1942).
394 MANGA NESE DETERM INATIO NS

ficiency), but clay soils having over 10 ppm of easily reducible manganese
did not show over-liming injury. Acid organic soils frequently are low in
total manganese owing to continual leaching of the manganese under con-
ditions favoring its conversion to Mn+ -1 • Manganese deficiency, although
not present in the acid condition, is likely to develop when this type of
soil is limed.
15-30. Acid soils that contain less than 10 to 25 ppm of easily reducible
manganese will generally not supply plants with sufficient manganese for
normal growth. Highly acid soils which are low in manganese sometimes
show manganese toxicity if large amounts are applied, because the process
of oxidation of added manganese to manganic form is slow. Productive neu-
tral or alkaline soils usually contain 100 or more ppm of easily reducible
manganese. Alkaline soiJsH1 should have at least 3 ppm of exchangeable
Mn++ in addition to I 00 ppm of easily reducible manganese to be free of
deficiency. The mechanisms of retention of Mn by soil colloids have been
examined. 1 7
15-31. The total Mn in soils ranges 18 generally from 10 to 2000 or more
ppm, of which 0.3 to less than 0.1 is easily reducible. In Hawaiian soils,rn
the total Mn is sometimes 1 to 4 per cent.
15-32. The Sample. The soil sample should be freshly taken from the
field, because air-drying (or heating) of the sample may increase the ex-
changeable Mn by a large amount ( 4 ppm raised to 80 ppm ). Drying
20

releases Mn f 1 from the Mn0 1 M complex as water is removed. A sepa-

21

rate portion of the sample is taken for moisture determination.

APPARATUS
15-33. Needed apparatu s consists of 500-ml conical extraction flask, a
shaking machine, a Buchner funnel, a suction flask, an asbestos and Gooch
crucible and holder, a 400-ml beaker, and a steam hot plate.

REAGENTS
15-34. Needed reagents consist of distilled water, filter paper to fit
Buchner funnel, l N NH.10Ac of pH 7, 30 per cent H~0 2 , concentrated
HN0:1, and 1 N NH.,OAc of pH 7 to which 0.2 gm of hydroquinone has
been added to each 100 ml.

lG Sherman and Harmer, S.S.S.A. Proc., 7: 398 ( 1943).

17 Hemstock and Low, Soil Sci., 76: 331 ( 1953).
18Leeper, Soi/Sci., 63:79 (1947).
19 Fujimoto and Sherman, Soil Sci., 66: 131 ( 1948).
20Sherma n and Harmer, S.S.S.A. Proc., 7:398 (1943).
21 Fujimoto and Sherman, S.S.S.A. Proc., 10: 107 ( 1946).
MANGANESE DETERMINATIONS 395

EXTRACTION PROCEDURES22

15-35. Water Soluble Manganese. To 25 gm of soil in a 500-ml conical

flask, 250 ml of distilled water is added and the flask is stoppered tightly.
The suspension is shaken by a machine for 30 minutes. Then the mixture
is filtered through a Buchner funnel and, if the filtrate is not clear, re-
filtered through a layer of acid-washed asbestos in a Gooch crucible. The
filtrate is evaporated to dryness, the organic matter is destroyed, and the
manganese is determined as described in ~] 5-74. The soil and asbestos
filter are returned to the 500-ml conical flask.
15-36. Exchangeable Manganese. The determination of exchangeable
manganese has been described (~ 5-71 ) in connection with the NH 4 0Ac
procedure for exchangeable cations. However, if easily reducible manganese
is to be determined, the exchangeable Mn is generally determined also on
the same sample.
15-37. To the 25-gm soil sample from which the water soluble Mn has
been removed, 250 ml of N NHpAc of pH 7 is added. The flask is stop-
pered and shaken in the machine for 30 minutes and then allowed to stand
at least 6 hours with frequent shaking. The soil mixture is next filtered
through a Buchner funnel, and the filtrate is evaporated to dryness on the
steam hot plate. The excess ammonium acetate is destroyed by moistening
the residue with water and reevaporating it to dryness. Then the residue is
treated with 30 per cent H 2 0" and HNOa, and the Mn is determined (~
5-74). The use of H 2 S04 is avoided at this point because of the possibility
of formation of CaS0 4 with soils high in calcium. The soil sample is re-
turned to the original flask for the determination of easily reducible man-
ganic compounds.
15-38. Easily Reducible Manganese. To the original 25 gm of soil, 250
ml of N NH 10Ac solution containing 0.2 per cent of hydroquinone is
added. The suspension is shaken at frequent intervals for 6 hours to in-
sure completeness of equilibrium, and then is filtered on a Buchner funnel.
To the filtrate, I 0 ml of concentrated HN0: 1 is added, the filtrate is evap-
orated to dryness on the steam hot plate, and the Mn is determined (~
5-74). It is essential that all of the hydroquinone be destroyed.
ALTERNATIVE PROCEDURES

15-39. Extraction of manganic manganese with an 0.05 per cent water

solution of quinol with I hour of contact was suggested by Jones and
Leeper 2 a as being more selective than quinol in NH 4 0Ac for the extrac-

22 Extraction procedures of Leeper, Proc. Roy. Soc. Victoria. 47(11) :225 (1935),
as modified and described by Sherman et al., Soil Sci., 54:253 ( 1942), and Sherman
and Harmer, S.S.S.A. Proc., 7:398 ( 1943).
2H Sci., 11 :463 (1950).
396 COPP ER DETE RMIN ATIO NS
0Ac-q uinol
tion of easily reducible manganese. The solution l N NH4
nese from hausman-
(0.05 per cent quinol) extracted considerable manga
did not correc t the
nite (having a formula MnOui~), a compound that
n extrac ted little
manganese deficiency of oats. The water-quinol solutio
much manga nic
from this compound, but it did extract 0.2 to 0.6 times as
ammo nium
manganese from a number of higher oxides of manganese as the
acetate-quinol solution extracted.
with 0.2 N
15-40 . Extraction of a quantity of available manganese
and plants
HO Ac was suggested by McCool. 24 A rapid test for Mn in soils
is available ( ~ 13-3) .
COPPER DETE RMIN ATION 20
( Colorim etricall y as carbam ate)
amate")
15-41 . The reaction of sodium diethyldithiocarbamate ("carb
which has
with copper gives the copper salt of diethyldithiocarbamic acid,
most sensi-
a golden brown color. Formation of this compound is one of the
pH over the
tive methods for copper determination and is unaffected by
as 0.01 ugm of Cu per ml may be detected,
range of 5.7 to 9.2. As little
determ ination withou t concen tration in an
and the range for quantitative
25 ml of solutio n, or 1 to 10 ugm in 10
organic solvent is 10 to 70 ugm in
ml of isoamyl acetate extractant.
Ni and Co
15-42 . Interfering substances are few in the procedure given.
ts of the order of 10 ugm are
give high positive interferences when amoun
do not interfe re with the
present with 10 ugm of Cu, and thus ordinarily
and accom panies Cu
Cu determination on soils and plants. Bi interferes
) if that is emplo yed;
through the dithizone preliminary separation (~ 16-34
Bi-carbam-
however, it would seldom interfere with soil and plant analysis.
e, and the
ate is not decolorized by a KCN solution as is Cu-carbamat
26
in terms of
presence of Bi can therefore be established and corrected for
n. A me-
its Cu color equivalent after a KCN treatment of the final solutio
itate Fe++ -1
dium alkaline with NH 40H is used in this procedure to precip
to the soluble
and Al+++ , which would interfere, and to convert the copper
ion form. Interfe rence by traces of Fe+++ and
cupric-ammonia complex
the presen ce of NH Cl and other condit ions of
Zn remaining is prevented by 4

the procedure.
APPARATUS
eter with
15-43 . Needed apparatus consists of a photoelectric colorim
a 125-m l conica l flask;
440-mu light maximum and colorimeter tubes;
Boyce Thomp son Inst. Contrib., 6: 147 (1934) .
24
method has been
After Callan and Hender son, Analyst . 54:650 (1929) . The
25
, Soil Sci ..
Sherma n and McHar gue, J.A.O.A .C., 25:510 (1942) , Holmes
employed by this laborat ory and
A. Kittrick employ ed the method in
56:359 (1943) and others. J.
worked out many details of the procedu re given.
26 Hibbar d, Hilgardia, 13: 1 (1940) .
COPPER DETERMINA TIONS 397
volumetric flasks, I-liter, 500-ml, 200-ml, set of 25-ml; a centrifuge and
centrifuge tubes with 15-ml volume calibration, a buret, and pipets to
take necessary aliquots. For the most sensitive procedure, a 150-ml separa-
tory funnel is required. In addition to usual cleaning, all glassware is
rinsed twice with 0.5 N HCl, twice with tap water to remove the HCl, and
finally twice with redistilled water. Care is used to avoid contamination
from stoppers, tubing, and other sources.

REAGENTS

15-44. Reagents needed consist of redistilled water (Pyrex still; stored

in Pyrex bottle), litmus paper, concentrated NH 1 0H, 0.5 N HCI (415 ml
of concentrated HCI diluted to I 0 liters with distilled water), and 25 per
cent NH 4 Cl solution (25 gm per 100 ml). Isoamyl acetate and 15 per cent
citric acid in water are required for the procedure of highest sensitivity.
15-45. Carbamate Reagent, 2 Per Cent. Two gm of sodium diethyl-
dithiocarbamate (obtained from Eastman Kodak Co., Rochester, N.Y.) is
dissolved in about I 00 ml of redistilled H~O, and the solution is filtered
through a coarse quantitative paper into a 200-ml volumetric flask. The
solution is diluted to volume with redistilled water and mixed. Stored in a
brown bottle in a cool, dark place, this reagent keeps for several months.
15-46. Standard Copper Solution. Exactly 0.500 gm of pure metallic
copper (preferably Bureau of Standards copper) is dissolved in 15 ml of
3 N HNO:i at room temperature in a covered 125-ml conical flask. When
the solution has cooled, 1 ml of concentrated H~S0 4 is added and the solu-
tion is evaporated cautiously to SOa fumes. The solution is cooled again,
diluted with I 0 to 15 ml of redistilled water, and again evaporated to S03
fumes. Finally, when cooled, the solution is transferred to a I -liter volu-
metric flask and made up to volume with redistilled water. This stock solu-
tion contains 0.50 mgm of copper per ml. A more dilute standard solution
containing 10 ugm per ml is prepared by diluting 10 ml of this stock solu-
tion to 500 ml. A series of aliquots (1,2,3,4,5, and 6 ml) is taken of the
10 ugm per ml standard solution for the standard curve, and the color is
developed as described in the procedure.

PROCEDURE

15-47. A 10 ml or smaller volume of solution of Cu (25 ugm or more)

in HCl, obtained from extraction from soil(~ 15-64) or plants(~ 12-39),
usually without dithizone purification (~ 16-34), is placed in a volumetric
centrifuge tube. Then 5 ml of 25 per cent NH 4 Cl solution is added, fol-
lowed by concentrated NH 4 0H from a buret with stirring (air jet held
under the solution surface) until the solution is neutral to litmus and then
3 ml in excess. The solution is diluted to I 5 ml with redistilled water, mixed,
and centrifuged at about 2000 rpm for 5 minutes.
398 COPPE R DETER MINA TIONS

15-48. From the supernatant liquid a suitable aliquot containing from

I 0 to 60 ugm of Cu is transferred to a 25-ml volumetric flask,
5 ml con-
n is diluted to about 22 ml with
centrated NH 40H is added, and the solutio
carbam ate is added from a pipet
redistilled water. Then I ml of I per cent
with redistil led water and
or buret, and the solution is diluted to volume
mixed.
the
15-49. After 15 minutes the color is read in the colorimeter with
1 hour, after which
440-mu light maximum. The color is stable for at least
ugm
it slowly decreases in intensity. Reference to the standard curve gives
Cu per ml of solution (ppm in solution).
ALTERNATIVE PROCEDURES

15-50. Isoamyl Acetate Extraction 27 for Greate r Sensitivity. To a 150-ml

of
separatory funnel, 5 ml of 15 per cent citric acid solution and an aliquot
added. Then concen -
the copper solution containing 1 to I 0 ugm of Cu are
and then 2 drops
trated NH 4 0H is added dropwise to neutral with litmus
added
in excess are added. Next 1 ml of the I per cent carbamate solution is
Finally , I0
and the solution is made to approximately 50 ml with water.
and the sus-
ml of isoamyl acetate is added, the funnel is tightly stoppered,
ls of I
pension is vigorously shaken for 1 minute each of 4 times at interva
of 20
or 2 minutes. The 2 liquids are allowed to separate during a period
iso-
to 30 minutes, and the aqueous layer is slowly drained off. The colored
uged
amyl acetate layer is then drained into a dry colorimeter tube (centrif
and the color is determ ined with a 440-mu light
if necessary to clear)
maximum.
, in
15-51. Refinements of the carbam ate method are availabJe. H Copper
2

from interfer ence, can be determ ined 29

low amounts and with great freedom
Chemi-
by 2,9-dim ethyl-l , 10-phenanthroline ("Neoc uproin e," G. F. Smith
light maxim um. The copper is re-
cal Co., Columbus, 0.) with a 457-mu
color is extract ed with n-hexyl
duced with hydroxylamine sulfate, and the
line,
alcohol. A similar reagent, 2, 9-dimethyl-4, 7-diphenyl-1, 10-phenanthro
is also specific for Cu. 30
TOTA L COPPE R AND ZINC EXTR ACTIO N FROM SOILS
31 gen-
15-52. The total copper in temperate and tropical soils ranges
t soils and
erally from 5 to 40 ppm but falls to I or 2 ppm in copper deficien

27 Details of this procedu re were kindly supplied

by J. A. Asleson from his M.S.
Thesis, Univ. of Wis. (1947). Chem., 25:1274
28Cheng and Bray, Anal. Chem., 25:655 (1953); Chilton, Anal.
(1953).
2ll Smith and McCurdy, Anal. Chem., 24:371
(1952); Gabler, Anal. Chem., 26:
577 (1954).
30 Smith and Wilkins, Anal. Chem., 25:510 (1953).
JordbrForsk.,
a1 Holmes, Soil Sci., 56: 359 (1943); Stenberg and Eckman, Nord.
4-6:689 (1948); Vermaa t and Vander Bie, Plant and Soil, 2:257 (1950).
COPPER DETERMINATIONS 399
may rise to l 00 ppm or more in some soils. Total zinc of soils falls in the
same range. Decomposition of soils with HF (with a few dr6ps of H 2 S0 4 )
was found by R. G. Menzel to yield up to 50 per cent more copper and
zinc than decomposition of the soils in HC10 4 (~ 15-57). Much of the
copper in some soils may be in organic combinations 32 and be completely
released by HC10 4 , whereas in other soils much of the copper may be in
silicates that resist HCl0 4 decomposition.

APPARATUS

15-53. Needed apparatus consists of a platinum crucible, a small agate

mortar, a 250-ml Pyrex beaker, an electric hot plate and sand bath, a rub-
ber policeman, and an analytical balance.

REAGENTS

15-54. Needed reagents consist of redistilled water, 48 per cent HF,

concentrated H 2 SO 4 , concentrated HNO:i> and a ternary acid mixture
consisting of 10 ml of concentrated HN0 3 , 1 ml of concentrated H 2 S0 4 ,
and 4 ml of 60 per cent HCI0 4 •

PROCEDURE33

15-55. One gm of air-dry soil sample, finely ground in an agate mortar,

is weighed and transferred to a platinum crucible. The soil is moistened
with water and 3 drops of concentrated H 2 S04 • Then two successive 5-ml
portions of HF are evaporated from the sample on the sand bath at l 80°C
to volatilize the Si02 • The crucible with the sample is placed in a 250-ml
beaker with l 0 ml of concentrated HN0 8 , and then covered with redis-
tilled water. The residue of the sample is loosened by heating the solution,
and is washed into the beaker with redistilled water, with the aid of a
rubber policeman. The solution is evaporated to dryness and the organic
matter oxidation is completed by digestion with l 0 ml of ternary acid mix-
ture at 200°C. No brown color from organic matter must be left if a good
dithizone separation is to be obtained. The residue, including a small
amount of H 2 S0 4 that moistens the residue after the ternary liquid diges-
tion, is taken up in l 0 ml of redistilled water. Not enough Pt is dissolved
at the low temperatures employed to interfere with the Cu carbamate de-
termination. (Pt in solution would interfere, and interfering amounts are
dissolved from a Pt crucible by dry ashing=14 at 450°C or above or in the
Na 2 C0 3 fusion. 35

a2 Lucas, Soil Sci., 5 I : 46 I (I 948).

33 After Menzel and Jackson, Anal. Chem., 23: 1861 (1951).
34 Bailey and McHargue, Plant Physiol .. 20:79 ( 1945).
35 J. A. Asleson, private communication (Aug. 1952).
400 COPPE R DETER MINA TIONS
of a
15-56. The total copper and zinc in soil is determined by means
for
usual semimicroanalytical procedure either colorimetrically (~I 15-41
zinc) or polarog raphica lly after dithizo ne separat ion
copper, ~ 15-69 for
(~ 16-34) .

EXTRACTABLE COPPER IN SOILS

varies
15-57. The soil copper extractable in various reagents (~ 15-65)
small frac-
from nearly all of it (I to 100 ppm of soil) in HCI0 4 to a very
(and
tion of it in neutral NH 4 0Ac. Release of nearly all of the soil copper
Tests of
zinc, cobalt, lead, etc.) by HCl0 4 digestion has been reportcd.:w
ing in
the insoluble residue generally showed only traces of copper remain
HF release , ~I 15-52) . Copper in minera ls that would
the residue (compa re
HC10 might be conside red of no signific ance to
not be decomposed by 4

soils.
e of
15-58. The HCI0 4 extractable Cu in soils was used:i; as a measur
soils. The copper content of plants
the "potentially available" Cu in organic
able
increased from 2 to 5 ppm with increasing soil content of HCI0 4 extract
in plants increas ed only to
Cu from 1 to 8 ppm; the content of copper
increas ed on up to I 00
around 7 ppm as the extractable copper content
ppm.
15-59. Uptake of copper (and zinc) by plants from montmorilloni
te was
but was less
proportional to the degree of saturation from 0 to 0.1 per cent
soil by other
than proportional above this degree.a~ That copper is held in
ii•
than the usual cation exchange forces has been shown by several studies.:
:"'
Even extremely insoluble compounds show some availability of copper
15-60. The HC10 4 treatme nt brings about oxidati on of the organic
soluble
matter and extraction of the metallic cations from most minerals as
ation and precipi tation of free Si0 2 •
perchlorates, with concurrent dehydr
that when the decomp osition is
J. A. Kittrick of this laboratory found
S0 the HC10 can
carried out in the presence of a small quantity of H 2 4 , 4
baking
be boiled off and the higher-boiling H 2 S0 4 keeps the residue from
dissolv ed
as a crust on the digestion vessel, and holds the cations as readily
sulfates.

APPARATUS
wide-
15-61. The HCI0 4 digestion of soil has been carried out in a

36 Holmes, Soil Sci., 59: 77 (I 945).

37 Lundbla d et al., Plant and Soil, I: 277 (1949).
38 Epstein and Stout, Soil Sci., 72:47 ( 1951 ).
66: 119 (1948);
:i9Bower and Truog, S.S.S.A. Proc., 5:86 (1941); Lucas, Soil Sci.,
, Trans. 4th Intern. Congr. Soil Sci., 1: 125 ( 1950), S.S.S.A. Proc.,
Menzel and Jackson S.S.S.A. Proc.,
15:122 (1951); DeMum brum and Jackson , Soil Sci., 81:353 (1956),
20:334 (1956).
40 Steenbje rg, Physiologia Plantaru m, 4: 677 (1951).
 COPPER DETERMINA TIONS 401
mouthed conical flask 41 covered with a special conical cover glass arranged
to drop the refluxed HCI0 4 at the center of the bottom of the flask. J. A.
Kittrick found that the HCI0 4 digestion of soil could conveniently be
carried out in Pyrex digestion tubes placed in a 4-liter beaker bath contain-
ing H 3 P0 4 , as illustrated in Fig. 15-1. The HC10 4 fumes are carried off by

Large funnel

4 Iller beaker

. D1gest1on tubes

Cone. H 3 PO,,
I inch deep

..*~~~~r-- Wire gauze with

asbestos center

-Gas burner

Fig. J 5-1. Digestion bath for de-

composition of soil in HCI04 for release
of copper. (After Dr. J. A. Kittrick.)

means of a glass funnel collector through tubing attached to an aspirator

pump.
REAGENTS

15-62. The needed reagents employed are redistilled water and an acid
mixture consisting of 100 ml of 60 per cent HCI0 4 mixed with 10 ml of
concentrated H 2S04 • This mixture is termed the 10 : I digestion mix.
PROCEDURE

15-63. To a 2.000 gm soil sample (up to 5 gm for sands and other soils
low in Cu) placed in the digestion tube, 10 ml of the 10 : 1 digestion mix-

41 Holmes, Soil Sci., 59: 77 ( 1945).

 ZINC DETE RMIN ATIO NS
402
d in the HaP0 4 bath until
ture is added. Several such tubes are slowly heate
S0 is left. If the samples
the HCl0 4 is boiled off and only l ml or so of H 2 4
ml of the 10 : 1 mixture
do not digest to a clear white residue, a few more
is added and the digestion is continued.
from the bath and
15-64 . When the tubes have cooled, they are removed
redistilled water. The in-
the digested sample is diluted to about I 5 ml with
proce dure at the first of
soluble residue is removed durin g the separ ation
ined colorimetrically (~
the analytical determ inatio n. The Cu is determ
( ~ I 6-34 ). The Zn
15-47) or polarographically after dithizone separ ation
by the usual Zn pro-
released by this same proce dure can be determined
cedures ( ~ 15-69 or 16-34 ) .

ALTERNATIVE PROCEDURES
ppm of Cu from
15-65 . Steenbjerg and Boken42 extra cted 0.1 to 0.3
ng Jess than this amou nt of
fertile soils with HCI of pH 2.0. Soils yieldi
coppe r were usually coppe r deficient.
coppe r by extra ction
15-66 . Antipov-Karataev 4 ~ deter mined available
the available coppe r was
with 0.5 N HNO:{• and noted coppe r deficiency if
total was less than I 8
less than 50 per cent of the total present, or if the
ppm.
complex, CuP ~0 7 - -
15-67 . Eriks son 44 suggested the use of the stable
that all coppe r sorbed on
for the extraction of available copper. He states
as well as that organi-
clay minerals should be removed by this treatm ent
nder 1" extra cted soil with
cally boun d by oxygen and amine groups. Wikla
ted ion excha nge resin, the soil being
an equal weight of hydro gen-s atura
ed on the cloth filter. The resin was
washed away while the resin was retain
the Cu and Zn. Small amou nts of
then extra cted with 2 N HCI to release
ively remov ed by this extraction
these elements added to soils were effect
even after the soils had been dried.
on in this labor atory
15-68 . Many workers, including Dr. J. A. Asles
Soil coppe r extra cted by
( 1951 ) have extra cted soil Cu in N NH 4 0Ac.
1 ugm amou nts either by
Asper gillus niger can also be assayed n in 0.1 to
4

mycelium weights or spore coloration.

ZINC DETERMINATION
( Colori metric ally as dithizo nate)
zone) is soluble in
15-69 . The acid form of diphe nylth iocar bazon e (dithi
water containing a slight
CCI 4 whereas the ammo nium salt is soluble in

42 Tids. Planteavl, 52:375 (1948 ).

4a Pedology, 1947:6 52 (1947 ).
44 Ann. Roy. Agr. Coll. Sweden, 16:72 ( 1949).
45 Ann. Roy. Agr. Coll. Sweden, 16:670 ( 1949).
4fl Dole, Soil Sci., 73: 135 (1952 ).
ZINC DETERMINA TIONS 403

excess of NH 4 0H. Dithizone forms complexes with Zn, Cu, Co, and Ni,
which can be extracted from a water solution into CC1 4 at pH values be-
tween 8 and 10. 47 In the present procedure, Cu, Ni, Co, and Pb are held
in carbamate complex form in the aqueous layer whereas the Zn is sepa-
rated into the CC1 4 layer at pH 8.8.
15-70. The interference by Ni is not wholly removed, but is satisfactorily
eliminated for soil and plant analyses. The range of zinc that can be de-
termined is from 1 to 25 ugm.
15-71. The analytical procedure given here can be adapted to total zinc
of soils ( ~ 15-52) or plants ( ~ 12-39) and to the various types of ex-
tractable zinc from soils ( ~ 15-8 3) .

APPARATUS

15-72. Needed apparatus consist of 125-ml pear-shaped separatory fun-

nels, 4-liter separatory funnels; a mechanical vertical shaker (Fig. 15-2),
a colorimeter with 535- and 620-light maxima; colorimeter tubes; volu-
metric flasks, 1-liter and 25-ml; and pipets including 1 of 5-ml volume.

REAGENTS

15-73. Needed reagents include zinc-free distilled water prepared by re-

distillation in a Pyrex still or passage through an ion exchange column;
l N HCl, prepared by distillation of approximately 6 N HCl in a Pyrex
still, and dilution of the condensate to 1 N; 1 N NH4 0H prepared by distil-
lation of concentrated NH 4 0H into zinc-free water in a Pyrex container
packed in ice (or by collection of anhydrous NHa in a Pyrex container of
zinc-free water); ACS purity CC1 4 (or redistilled S-free CCl." stored in the
dark); thymol blue indicator; and the following special reagents.
15-74. Carbamate Solution, 0.2 Per Cent. A 0.2 per cent solution is
prepared by dissolving 0.2 gm of sodium diethyldithiocarbamate (Eastman
Kodak Co.) in 100-ml of zinc-free water. This solution keeps satisfactorily
in a brown bottle, if kept in a cool, dark place.
15-75. Dithizone Solution in CCl4 , 0.01 Per Cent. Into a 4-liter separa-
tory funnel, 0.2 gm of diphenylthiocarbazone (Eastman Kodak Co.) and 1
liter of CC1 4 are placed, and the solid brought into solution by frequent
agitation for about 15 minutes. To this solution, 2 liters of zinc-free 0.02
N NH 4 0H is added and the mixture is shaken to transfer the dithizone to
the aqueous phase. The CC1 4 (light green color) is discarded and the
aqueous phase is rinsed with several 100-ml portions of CC1 4 • Then 500 ml
of CC14 and 50 ml of zinc-free 1 N HCl are added. The mixture is shaken
to transfer the dithizone to the CC1 4 layer, and then the CC1 4 -dithizone

47 The formula for dithizone is given in ~ 16-34. Further details of the condi-
tion for formation of complexes with various elements is given by Welcher, Organic
Analytical Reagents, Vol. 3 (New York: D. Van Nostrand Company, Inc., 1947).
404 ZINC DETERMINATIONS

Fig. lS-2. Vertical shaker for separatory funnels. (Photo courtesy Dr. R. S~ Holmes.
Details of construction are given by Holmes and Mullins, Soil Sci., 69:233, 1950.)

phase is diluted to 2 liters. The solution is stored in a glass-stoppered Py-

rex bottle, in a refrigerator.
15-76. Ammonium Citrate, 0.4 M. To 90 gm of ammonium citrate,
enough water is added to make l liter. To this solution, enough zinc-free
concentrated NH4 0H is added to bring the pH to 8.5. The zinc impurities
are removed by extraction with portions of dithizone in CC14 in a large
separatory funnel until the latter reagent no longer changes color, and then
ZINC DETERMINATIONS 405
with portions of CCl 4 until the citrate solution is free from dithizone color.
15-77. Standard Zinc Solution. Exactly 0.1 gm of pure zinc is dissolved
in 50 ml of 0.02 N H 2 SO., and diluted to 1 liter to give a I 00 ugm/ml
concentration of Zn. A secondary dilution of 10 ml to 500 ml gives a 2
ugm/ml working standard.
PROCEDURE4~

15-78. A solution containing 5 to 20 ugm of zinc in 30 to 40 ml of

0.02 N HCI is placed in a 125-ml separatory funnel. Fifty ml of ammonium
citrate buffer and 3 ml of carbamate are added and the solution pH is ad-
justed to 8.5 to 8.8 with redistilled NHpH or HCI (thymol blue indi-
cator). Exactly 10 ml of dithizone reagent in CCl 4 is added. The mixture is
shaken for 5 minutes. The CCl 4 phase is transferred to another separatory
funnel, 25 ml of 0.01 N NH 4 0H is added to the CCl 1 phase, and the mix-
ture is shaken again for 3 minutes to extract the excess dithizone into the
aqueous phase.
15-79. A 5-ml aliquot of the organic phase is taken with a pipet and
diluted with CCl 4 to 25 ml. The solution thus obtained is mixed and trans-
ferred to a colorimeter tube, and the light transmission is measured at a
535- (or 540-) mu light maximum. The amount of zinc present is deter-
mined by reference to a standard curve prepared in an identical manner
with known amounts of zinc. All equipment may be satisfactorily cleaned
with either CCl 1 or water if it is restricted to this analysis.
Al TERNATIVE PROCEDURES

15-80. R. G. Menzel found that the distribution of excess dithizone

between aqueous and CCl 1 layers is markedly affected by the solution pH
and other conditions, and that to measure its color was better than to at-
tempt to hold its amount constant. The test solution is read at 540-mu
against air as I 00 per cent transmission (L;, 40 ). Then the optical density
(Ln~o) of the excess dithizone is determined at 620-mu (at which wave
length zinc dithizonate docs not have absorption) against air as 100 per
cent transmission. The zinc dithizone optical density (L,,.J is found, for the
Evelyn Colorimeter, by the equation:
(15-2)
Since the excess dithizone has 0.345 times as much optical density at 540

4 8 Cowlings and Miller, Ind. Eng. Chem .. A .E., 13: 145 (1941); Holmes, Soil Sci.,
59: 77 ( 1945); Sandell, Colorimetric Determination of Traces of JHetals, 2nd Ed.
(New York: lnterscience Publishers, Inc., 1950), p. 628; Shirley, J.A.0.A.C., 31:285
(1948); J.A.0.A.C., 32:276 (1949); Shaw and Dean, Soil Sci., 73:343 (1952); Dr.
R. G. Menzel, personal communication.
406 ZINC DETERMIN ATIONS

mu as it has at 620 mu, the optical density of zinc dithizonate alone is

obtained by the equation. The constant term arises from the fact that the
CCl 4 blank would read over 100 per cent transmission against the air read-
ing as 100 per cent.
15-81. Some off color of the dithizone extraction may be obtained occa-
sionally with soil or plant extracts, but ordinarily has little effect on the
final reading. It can be eliminated by returning the zinc from CCl 4 to fresh
ammonium citrate-carbamate reagent through the addition of HCI to 0.02
N concentration, and re-extraction of the Zn with fresh dithizone reagent
by repetition of the procedure.
15-82. More elaborate means of chemical separation of zinc have been
employed by several groups of investigators. 4 n A field method is available. 50

DITHIZONE EXTRACTABLE SOIL ZINC

15-83. The method of Shaw and Dean° 1 given here for extraction of zinc
from soils in dithizone was shown to be fairly promising for extraction of a
fraction of the soil zinc that can be correlated with that which is relatively
available to plants. The interpretation of the results depends in part upon
a correlation with the soil pH. For soils below pH 6.5, zinc deficiency seems
to be correlated with 0.5 ppm or less of zinc extracted. For zinc deficient
soils above pH 6.5, the dithizone extractable zinc may extend to 2.5 ppm.
A dithizone extractable zinc content of above 0.5 ppm seems to be cor-
related with nondeficiency of zinc for soils of pH between 5 and 6.5. Dithi-
zone extractable zinc may extend to 5 and 17 ppm in acid soils high in
available zinc.
APPARATUS
15-84. Apparatus needed consists of 125-ml pear-shaped separatory
funnels, mechanical vertical shaker (Fig. 15-2), 60-ml conical centrifuge
tubes and centrifuge, and a 10-ml pipet.
REAGENTS
15-85. Reagents needed include redistilled water, HCI, CCl 4 , and di-
thizone in CC14 ( ~ 15-7 5) and the following.
15-86. NH 4 0Ac Buffer, 1 M. Seventy-seven gm of c.p. NH 4 0Ac is dis-
solved in enough redistilled water to make a final volume of 1 liter. The pH
is adjusted to 7 with zinc-free NH 4 0H. The buffer is purified in a large
separatory funnel with dithizone and CC1 4 , the organic phase being dis-
carded until it no longer changes color. The dithizone that has become
dissolved in the aqueous phase is removed by repeated extraction with CCl 4 •
49Parks et al., Ind. Eng. Chem., A.E., 15:528 (1943); Holmes, Soil Sci., 59:77
(1945).
;,o Reichen and Lakin, U.S. Geo!. Surv. Cir. 41 ( 1949).
r.t Soil Sci., 73:346 (1952).
ZINC DETERMINATIONS 407
PROCEDUREG2

15-87. Soil Samples. Since the quantities of zinc to be determined are

very small, the soil samples are collected and prepared with the usual care
recommended for minor element work. The soils are air-dried, crushed with
a wooden pestle, and passed through a zinc-free I -mm sieve ( 18 meshes per
inch). The sieve is made of iron wire or any zinc-free wire commercially
available. The samples are stored in cardboard ice cream containers.
Cleansing tissues generally have been sufficiently free of zinc to be useful
for such operations as wiping out weighing pans.
15-88. Extraction. To extract the dithizone extractable zinc from soils,
25 ml of 1 N NH 4 0Ac and 25 ml of dithizone-CCl 4 solution are pipetted
into a 125-ml pear-shaped separatory funnel. Then 2.5 gm of soil is added.
The separatory funnel is firmly stoppered and shaken on a mechanical ver-
tical shaker (Fig. 15-2) for 1 hour. The separatory funnel is then placed
on the stand and the soil-CCl 4 suspension is drained into a 60-ml conical
centrifuge tube. While the suspension is centrifuged to develop a continuous
CC1 4 phase, the original separatory funnel is rinsed with water.
15-89. A I 0-ml aliquot of the CC1 4 solution containing the zinc is re-
moved by means of a pipet, care being exercised to locate the tip of the
pipet in the CCl 4 phase. Air is expelled from the tip as it passes through
the aqueous and soil layers and the tip is kept away from the walls of the
tube where soil and water may be contacted. A mark on the pipet greatly
facilitates its placement at the proper depth in the centrifuge tube. The
pipet is withdrawn and the adhering soil and water are washed from the
tip with a wash bottle, the volume is adjusted to the mark, and the tip is
rinsed again. The 10-ml aliquot is added to the cleaned separatory funnel
from which it originally came, and 50 ml of 0.02 N HCI is added. The mix-
ture is agitated for 3 minutes on the shaker to cause the extraction of the
zinc from the CCl 4 phase into the HCl phase. The CCl 1 phase, which con-
tains interfering elements and some oxidized dithizone, is discarded. The
aqueous phase is rinsed twice with CCl 4 with agitation by hand. The zinc
so extracted may be determined by the standard methods, colorimetrically
as the dithizonate ( ~ 15-69) or polarographically ( ~ 16-34).

ALTERNATIVE PROCEDURES

15-90. Many types of extraction of soil zinc have been studied. As with
copper, the HC10 4 extractable zinc represents the majority of the zinc pres-
ent0a and may be obtained by the same procedure as HCl0 4 extractable
copper ( ~ 15-57).
15-91. Wear and Sommer•i 4 extracted the available zinc from soils by

r.~
Shaw and Dean, Soil Sci., 73: 342 ( 1952).
Holmes, Soil Sci .. 59:77 ( 1945).
r,;i
»4S.S.S.A. Proc., 12:143 (1948).
408 MOLYBDE NUM DETERMIN ATIONS

means of 0.1 N HCl and claimed good correlations between this type of
extractable zinc and zinc deficiency on a number of Alabama soils.
15-92. Hibbard"" recommended extraction of soil zinc with 1 N KCl
acidified to pH 3.2 with HOAc. Lyman and Dean" 11 used NH 4 0Ac of pH
4.6, and noted that pineapples were likely to show zinc deficiency on soils
that yielded less than l ppm of Zn in this reagent. Bergh"' recommended
0.1 N MgS0 4 of such acidity that the final pH of the extracted solution was
the same as the original soil pH. Soils yielding less than 5 ppm of zinc in
this solution, were considered likely to be zinc deficient. Thorne et al. 08
used KCl-HOAc at pH 3.2.
MOLYBDENUM DETERMINATION
( Colorimetrical ly by thiocyanate orange-red color)
15--93. Determination of Mo in soils and plants has been carried out by
various improvements in the method based on the thiocyanate-colored com-
plex, developed through reduction with acetone,'rn ascorbic acid,no or
chlorostannous acid. 111 Spectographic and polarographic methods are also
known. 112 Alkaline separation of Mo from many interfering ions as em-
ployed by Robinson and intensification and stabilization of the color have
made possible the elimination of need for extraction of the chromogen with
organic solvents for the range of I to 7 5 ugm of Mo. The method is suit-
able for determination of Mo extracted from soils or plants. Any rhenium
(Re) is included with Mo by the method, but ion exchange separation of
these two elements has been reported. u:i

APPARATUS

15-94. Needed apparatus includes a colorimeter with 470-mu light maxi-

mum (the light absorption maximum is 470 mu, although from 420- to
525-mu light maxima have been used by various workers with resulting
much lower sensitivity), colorimeter tubes, volumetric flasks (I-liter, 25-
ml), and pipets.
REAGENTS

15-95. Needed reagents include 10 per cent potassium thiocyanate (10

gm of KSCN dissolved in 100 ml of water), acetone (reagent grade), HCl
(concentrated, specific gravity 1.18 to 1.19), and the Mo standard.

r.5 Hilgardia, 13: 1 (1940).

r.11 Soil Sci., 54: 315 ( 1942).
<>7 K. Norske Vidensk. Selskabs, Skrifter, 1945, No. 3 ( 1948) .
.-.8 Soil Sci., 54:463 (1942).
"~Ellis and Olson, Anal. Chem., 22:328 (1950).
110 Robinson, Soil Sci., 66: 317 ( 1948).
Ill Barshad, Anal. Chem., 21: 1148 ( 1949).
6 :! Nichols and Rogers, lnd. Eng. Chem., A.E .. 16: 137 ( 1944)
63 Fisher and Meloche, Anal. Chem., 24: 1100 ( 1952).
MOLYBDENU M DETERMINA TIONS 409

15-96. Molybdenum Standard. Exactly 0.1840 gm of ammonium molyb-

date, (NH 4 )uMoP 24 • 4 H 2 0, is dissolved in 900 ml of water and the
solution is diluted to exactly 1 liter, giving 100 ppm of Mo. A 5-ppm Mo
standard is prepared by dilution of 50 ml to 1000. Aliquots of this solu-
tion (0.5, 1, 2, 5, and 10 ml) are taken for the standard curve, the color
being developed as for the test solution in the procedure. The above gives
the optimum range of 2.5 to 50 ugm in 25-ml volume of colored solution,
or 0.1 to 2.0 ppm Mo. A 1-ppm Mo standard is employed to extend the
range down to 1 ugm of Mo, and a 15-ppm Mo standard is employed for
75 ugm of Mo (3.0 ppm Mo) to cover the maximum range of 1 to 75 ugm
of Mo.

PROCEDURE64

15-97. Preparation of Test Solution. The test or standard solution is

prepared to contain 2.5 to 50 ugm (1 to 75 ugm permissible) of Mo in 10
ml of water, to 15 HCl, or H 2S04 solution. The solution should contain no
actively oxidizing acids, such as HN0 3 or HCl0 4 , and if present they are
removed by evaporation. In the extraction procedure ( ~ 15-104) , the Mo
is oxidized to hexavalent state and iron is removed by the Na 2 C03-water
slurry separation. If some other test solution contains more than the equiva-
lent of l 00 ppm of Fe in the 25-ml final volume, the iron is precipitated
and removed by the NH 4 0H separation, since more than this amount
causes a fine precipitate in the final Mo test solution.
15-98. Development of the Amber Color. The standard or test solution
of Mo (~ 15-97) is placed in a 25-ml volumetric flask. Enough HCl
must be present or be added to make the final 25-ml of solution 1.2 to 4 N
with respect to nonoxidizing acids (3 ml of concentrated HCI is added for
water solutions); the HCl has already been added in the solutions taken from
the silica separation procedure (~ 15-112). Then 1.5 ml of 10 per cent
KSCN solution and 8 ml of acetone are added and the solution is made to
volume and mixed. The solution in the flask is digested in a 60° to 70°C
water bath for an hour, or longer if necessary for disappearance of the red
Fe(SCN) 3 color. (The color of the standards develops in 20 minutes, but

64 Adapted here to increased sensitivity by use of 470-mu filter and smaller vol-
umes, from Ellis and Olson, Anal. Chem., 22:328 (1950). These authors obtained
a more intense color by acetone reduction than by chlorostannous reduction. Their
calibration curve with 420-mu light is nonlinear; their reported range of 5 to l 000
ugm of Mo resulted in reported transmissions of 96 to I per cent respectively, which
is considerably beyond the optimum range. The range of l to 75 ugm of Mo em-
ployed here gives transmissions of 97 to l 8 per cent, but the percentage transmission
can be decreased enough for the determination of 0.15 to 15 ugm of Mo by extraction
of the color in organic solvents (~] 15-101) and by use of a 5-cm absorption cell, as
reported by Parks et al., Ind. Eng. Chem., A.E., 15:528 (1943). Appreciation is ex-
pressed to Drs. Roscoe Ellis, H. G. Raj, and H. H. Hull for help with testing at the
University of Wisconsin the modifications of the original procedure incorporated here.
410 MOLYBDENUM DETERMINATION S
interfering red Fe(SCN) 3 color of soil extracts may not be gone for 1 to 3
or more hours. The color remains satisfactory if heating is continued as
long as 2 or 4 hours.) The solution is then cooled to room temperature and
mixed. It is examined for traces of turbidity. If they are present, a portion
of the solution is centrifuged (or filtered through Whatman No. 42 filter
paper). The color is read at once in a colorimeter with 4 70-mu light maxi-
mum.
15-99. The ugm of Mo present is obtained by means of a standard
curve. The Mo content of the reagents is taken into account through blank
determinations carried through all of the steps of the preparatory pro-
cedures.

ALTERNATIVE PROCEDURES

15-100. Intensification of the thiocyanate color with chlorostannous re-

duction has been reported as resulting from the presence of ferric 65 and
nitrate ions, but these additions are apparently unnecessary when acetone
reduction is employed. In 1 procedure, 66 3 or 4 drops of 0.01 N FeCI~ are
added per determination. The color is developed with freshly prepared
chlorostannous acid (SnCl 2 • 2 H 20 in 1.2 N HCl) in sufficient quantity
to cause the loss of ferric thiocyanate color and maximum development of
molybdic thiocyanate color, but a large excess of reductant is not per-
missible. Tests of increments of chlorostannous acid are suggested for the
particular system to be employed. The intensification by nitrate is at-
tributed to more efficient conversion of the Mo to hexavalent state prior to
color development. In another procedure, 67 1 mgm of Fe is added as a
ferric solution to each determination for color intensification. In another, 68
3 drops of 0.01 N FeCla, 1 ml of 5 N NaN0:1, and 6 ml of 10 per cent
SnCl 2 • 2 H 2 0 in 1.2 N HCl are added to each Mo determination.
15-101. Extraction of Colored Mo Solution in Organic Solvents. As a
means of increasing the sensitivity of the method, and so decreasing the
required size of sample when the Mo content is low, the Mo color de-
veloped with SnCl 2 has been extracted in organic solvents. A solution
heavier than water, 69 which is a considerable advantage, consists of CCl 4
and isoamyl alcohol mixed in equal volumes. Silicone stopcock grease was
employed on the separatory funnels and the color read at 470 mu. The Mo
thiocyanate color has been extracted with peroxide-free isopropyl ether. 70

65 Nichols and Rogers, Ind. Eng. Chem., A.E., 16: 137 (1944).
66 Barshad, Anal. Chem., 21: 1148 ( 1949).
67 Fujimoto a'ld Sherman, Agron. lour., 43: 425 ( 1951).
ss Sarthou, Ph.D. Thesis, Univ. Wis. (1951 ).
so Arkley and Johnson, Anal. Chem., 26:572 (1954).
70 Reichen and Ward, U.S. Geol. Surv. Cir. 124 (1951); Purvis and Peterson,
Soil Sci., 81 :223 (1956).
MOLYBDENUM DETERMINATIONS 411

To detect peroxides, 5 ml of isopropyl alcohol was shaken vigorously with

5 ml of acidified aqueous solution of KI. If the Kl solution does not remain
practically colorless, enough peroxides are indicated to make the ether un-
satisfactory.
15-102. Fujimoto and Sherman 71 placed a 50-ml aliquot of the Mo test
solution in a 125-ml separatory funnel and then added 1 ml of l 0 per cent
ammonium citrate, 3 ml of 10 per cent NH 4 SCN, and 3 ml of 10 per cent
SnC12 • 2 H 2 0 in 2 N HCl, followed by shaking. The colored complex then
was extracted in 6 ml of normal butyl alcohol by vigorous shaking for 1
minute. Since butyl alcohol is partially soluble in water, the initial 6 ml
decreased to about 4 ml. The aqueous layer was drained into a beaker and
kept for further extraction. The butyl alcohol phase was slowly run into a
l 0-ml calibrated centrifuge tube. The aqueous portion was returned to the
separatory funnel, 4 ml of fresh butyl alcohol added, and the extraction re-
peated. For the third extraction, only 2 ml of butyl alcohol was used. The
combined alcohol extract was diluted to 10.0 ml with butyl alcohol and
shaken vigorously. The small amount of aqueous solution, which un-
avoidably drained with the alcohol, disappeared when the solution was di-
luted to 10 ml with alcohol and shaken. The occasional turbidity of the
alcohol solution was removed by a few minutes of centrifugation. The clear
supernatant solution was transferred to the comparator tube and allowed
to stand for 10 minutes and then the percentage transmission was measured
spectrophotometrically. The standard solutions were evaporated and sub-
jected to fusion and other treatments similar to the samples.
15-103. Molybdenum of plants 72 and soilsrn has been determined by the
green-colored molybdenum-dithiol complex. Determination of Cu and Mo
in the one dithiol system has an advantage because of the interrelationship
of these 2 elements in animal nutrition. The lengthy separations involved
has led to preference74 for the thiocyanate method for Mo in some labora-
tories.

TOTAL MOLYBDENUM IN SOILS AND PLANTS

(Na 2 CO:i fusion and alkaline separation)
15-104. The total molybdenum content of soils for parts of North and
South America and Asia range from 1 to I 0 ppm, but extend in some soils
to 20 and even to 30 or more ppm. Liming a soil containing 32 ppm of Mo
resulted 75 in the production of legume vegetation containing over 10 ppm

71 Agron. lour., 43:425 ( 1951 ).

72 Piper and Beckwith, J. Soc. Chem. Ind., 67:374 (1948); Proc. Specialists Con/.
in Agr. Australia, 1949: I 44 ( 1951); Dick and Bingley, Australian J. Exp. Biol. Med.
Sci., 25: 193 (1947).
n Williams, J. Sci. Food Agr., 6: 104 (1955).
74 Purvis and Peterson, Soil Sci., 8 I: 223 (1956).
75 Robinson et al., Soil Sci., 72:267 (1951).
412 MOLYBDE NUM DETERMIN ATIONS

of Mo, the content at and above which toxicity to cattle begins. The
Mo content of various plants ranged from 0.1 to 47 ppm. However, plants
in Hawaii ranged 76 mainly below 1 ppm in spite of unusually high soil
content of Mo ( 11 15-113). Soil Mo becomes tightly associated 77 with free
iron oxides and then has an extremely low availability. Barshad 78 reported
as much as 200 ppm of Mo in some alfalfa in California, and Cunningham
and Hogan 79 reported an increase from 3 ppm Mo in pasture plants to 100
ppm Mo, as a result of the application of 150 gm of ammonium molybdate
or 82 gm of Mo per hectare ( 0.15 and 0.08 pounds per acre, respectively)
of acid peat.
APPARATUS

15-105. Needed apparatus consists of a platinum dish (plants) or 35-ml

crucible (soils), an electric furnace or Meker burner, 100-ml and 25-ml
graduated cylinders, a 250-ml beaker, a centrifuge and 25-ml volumetric
centrifuge tube, a filter funnel and Whatman No. 40 paper, a 100-ml volu-
metric flask, and pipets.
REAGENTS

15-106. Needed reagents are anhydrous Na 2C03 (Mo-free or low in Mo,

such as Baker's, Phillipsburg, N.J.), 3 per cent of ethanol in water, and
concentrated and 3 N HCl.
PROCEDURE

15-107. Soil or the ash of plant tissue is fused in Na 2 C03 to effect solu-
tion of molybdate as the sodium salt and to separate much of the interfer-
ing substances as insolubles in the water extract. The platinum ware is
cleaned by alternate fusions in Na 2 C03 and KHS0 4 followed by digestion
in dilute HCI until the Mo blank is reproducible and fairly low.
15-108. Ashing and Fusion of Plant Tissue. 80 One to 10 gm of dried
plant tissue is ashed in a platinum dish in an electric furnace at 500°C. The
ash is then fused in 2 gm of anhydrous Na 2 C03 (of low Mo content as de-
termined in preliminary tests) at 1000°C in an electric furnace or over a
Meker burner. Care is taken to bring all of the ash into contact with the
flux for a complete fusion. The melt is treated as for soils ( 11 15-111 ) .
15-109. Fusion of Soils. 81 A 2-gm sample of soil, ground to pass 0.25
mm openings ( 60 meshes per inch) of silk bolting cloth, is transferred to
a 35-ml platinum crucible containing about one gm of anhydrous Na2 C03 •

76 Fujimoto and Sherman, Agron. lour., 43:424 (1951).

77 Robinson and Edgington, Soil Sci., 77:237 (1954).
78 Soil Sci., 66: 187 (1948).
79 New Zealand I. Sci. Tech., A 31:(1)39 (1949).
80 Robinson and Edgington, Soil Sci., 77: 23 7 (1954).
81 Robinson et al., Soil Sci., 72:267 (1951).
MOLYBDENUM DETERMINATIONS 413
Six gm more of Na 2 C0a is mixed with the soil by stirring with a glass rod,
the latter being brushed off afterwards, to return adhering par~icles to the
crucible. The heating is carried only to about 500°C to burn off the soil
organic matter during the first 5 or 10 minutes, in an oxidizing atmosphere,
to avoid the possibility of Mo loss, ~ 2 before the charge is actually fused.
15-110. The crucible is partially covered and the temperature is gradu-
ally raised to 950° to 1050°C in an electric furnace or with the full heat of
a Meker burner, for 15 to 20 minutes or until fusion is complete, the flux
being swirled to mix at intervals after it becomes molten. When bubbles
cease to rise and the melt is quiet, the crucible is rotated to spread the melt
on the sides as it sets into a cake. (Precautions in handling platinum are
described in ~ 11-1 3).
15-111. Extraction of Mo from Cake. When the crucible or dish, con-
taining the Na~C0: 1 cake from soil or plant-ash fusion, has cooled, it is in-
verted over a 250-ml beaker and the cake is dropped from the platinum
container as the latter is gently rolled between the thumb and fingers. If the
cake docs not become detached, the crucible is placed in the beaker. The
cake is disintegrated in I 00 ml of water (containing 2 per cent by volume
of ethanol to reduce the manganate": 1 and avoid attack of the platinum ware
if the crucible has been placed in the beaker). The disintegration of the
cake is hastened by trituration with a rubber policeman on a glass rod and
warming on a steam hot plate for a period of several hours. The volume
or the resulting slurry suspension, in which the particles have been thor-
oughly disintegrated, is measured in a graduated cylinder and filtered
through a dry Whatman No. 40 filter paper into a 250-ml beaker. The
volume of the undiluted filtrate is measured and all (or an aliquot) is taken
for the Mo determination. (In lieu of filtering, the suspension may be clari-
fied by centrifugation.) The Mo remains soluble as sodium molybdate
while Fe, Ti, Ca, Mg, and many other ions remain in the precipitate. 84 The
filtrate should be perfectly clear.
15-112. Removal of Silica. The clear filtrate solution is acidified with
HCI (2 ml of concentrated HCl per gm Na 2 C0,1 contained) and is evapo-
rated to dryness. The residue is taken up in exactly (pipet) 20 ml of 3
N HCI (in lieu of HCI addition in the color development, ~I 15-98). The
suspension is briefly warmed on a steam plate while being triturated thor-
oughly. It is then transferred to a 25-ml centrifuge tube and centrifuged to
throw down the silica and undissolved salt. As large an aliquot as possible
(but not over 15 ml) of the clear supernatant solution is taken for the Mo

82 Robinson. Soil Sci .. 66:318 (1948); Fujimoto and Sherman, Agron. l<ll!r., 43:
425 ( 1951).
s:i Fujimoto and Sherman, Agron. lour., 43: 425 ( 1951 ) .
~4 Hillebrand and Lundell, Applied Jnor!(anic Analysis (New York: John Wiley
and Sons, Inc., 1929); Sandell, o,n. cit., p. 462; Robinson and Edgington, Soil Sci., 72:
?.67 (1951).
414 COBAL T DETERM INATIO NS

analysis ( ~ 15-97). If the amount of Mo is too low, the use of more sample
is preferable to substitution of washing of a filter in either of the two separa-
tions.
ALTERNATIVE PROCEDURES
15-113. With plant ash low in silica, silica does not need to be removed
from the filtrate of the second separation 80 (~ 15-112) . Direct HCl extrac-
tion of Mo from the plant ash obtained in a porcelain dish has been em-
ployed. H6 A procedure for extraction of Mo from plants by wet oxidation in
HNO:l' HC10 4 , and H 2 S04 has been employed.~ 7 But for thiocyanate color
stability, the strongly oxidizing acids must be removed. Strong acids have
been employed to extract the Mo from the Na 2 COa fusion of soils, but this
sacrifices the alkaline slurry separation obtained after the Na 2 CO:i fusion
( ~ 15-111 ) . Reportedx~ Mo contents of Hawaiian soils ranged from 9 to
74 ppm with strong acid extraction, while only 1.8 to 18.6 ppm were
found 8 n for the same soils when the alkaline separation (~ 15-1 I I) was
employed, and the lower values were confirmed by spectrographic analysis.
Field methods have been reported for Mo in plantsm1 and rocks and soils Jt
1

that are suitable for biogeochemical prospecting.

15-114. One proposed 1n method for available soil Mo involves extrac-
tion of the soil with acid ammonium oxalate solution buffered at pH 3.3.
COBAL T DETERM INATIO N
15-115. Two micro methods have found wide use for the determination
of cobalt in the small quantities available for analysis from convenient sizes
of plant, animal, and soil samples. Both are based on a chromogen de-
veloped by cobalt with the "o-nitros o-R" type salts, 11 :i with 550-mu light
maximum 114 giving a Beer's Law relationship. This type of salt was first
prepared, and later made readily available through an easy process of syn-
thesis, by Baudisch and coworkers. 9 r. Determination of cobalt by o-nitro-

85 Robinson et al., Soil Sci., 72:269 (1951 ).

86 Barshad, Anal. Chem., 21 :1149 (1949).
87 Parks et al., Ind. Eng. Chem., A .E., 15: 527 ( 1943); Piper and Beckwith, Proc.
Specialists Con/. in Agr. Australia (1951).
88 Fujimoto and Sherman, Agron. lour., 43 :425 ( 1951);
Bartrand, Compt. rend.,
211 :406 (1940).
su Robinson and Alexander , Soil Sci., 75 :287 (1953 ).
90 Reichen and Ward, U.S. Geo!. Surv. Cir. 124 (1951 ).
91 Ward, Anal. Chem., 23:788 (1951).
92 Grigg, New Zealand J. Sci. Tech., A34:405 (1953), Analyst,
78:470 (1953);
Purvis and Peterson, Soil Sci., 81 :223 (1956).
93 Sandell, op. cit., p. 287.
94 Cooper and Mattern, Anal. Chem., 24: 574 (1952); Claassen
and Westerveld,
Rec. trav. chim., 67:720 (1948); Haywood and Wood, J. Soc. Chem. Ind .. 62:37
(1943).
95Ber., 45:1164 (1912), 48:1660 (1915); Sci., 92:336 (1940); J. Am. Chem. Soc.,
63 :622 (1941).
COBALT DETERMINATIONS 415
sophenol has been described by Cronheim, 96 and by o-nitrosocresol by
Ellis and Thompson. 97 The latter method has given more consistent re-
sults and, in spite of the dithizone separation involved, is essentially a
simpler method, according to K. C. Beeson. 98 It has been employed for
the determination of total Co in soils by Holmes. 99 A colorimetric field
method for cobalt has been described by Almond and Bloom. 100 No details
of the micro procedures are given here.
15-116. Determination of larger quantities of cobalt (0.1 to 10 mgm
Co) has been described in connection with potassium determination by the
cobaltinitrite procedure (~ 6-27). It can also be determined as KaCo
(N0 2 ) 6 • 1.5 H 20 (~ 6-61 ).

QUESTIONS
1. State why the extraction and availability of the elements Fe, Mn. Cu. Zn,
Mo, and Co cannot be treated in the same way as that of metallic cations that
are strong base formers?
2. What is the relation of iron valence to o-phenanthroline color formation?
How can the orthophenanthroline compound be employed as an oxidation-
reduction indicator?
3. Why must the extraction of ferrous iron from soils be carried out very
rapidly?
4. Why cannot exchangeable ferric iron of soils be extracted in neutral
NH 4 0Ac?
5. Why is a reducing agent employed in the extraction of available man-
ganese from soils?
6. Why may manganese deficiency be induced in a soil, which did not pre-
viously show deficiency, by the application of ground limestone?
7. Why is the color intensified by isoamyl acetate extraction of the copper
carbamate?
8. What principle is employed in the separation of zinc as dithizonate from
elements that would interfere with its determination as the dithizone compound?
9. List several types of extractions employed for fractions of soil copper and
of soil zinc that may be correlated with availability.
10. Discuss the valence relations of molybdenum determination as the thio-
cyanate.
11. Into what chemical form is the molybdenum converted during the alkaline
slurry separation?
12. Why is cobalt determination in plant materials and soils important to the
soil chemist?

96 Jnd. Eng. Chem., A.E., 14:445 (1942).

97 Ind. Eng. Chem., A.E., 17:254 (1945).
98 Private communication.
011 Soil Sci., 59:77 (1945).
100 U.S. Geol. Surv. Cir. 125 ( 1951).
 16
Polarographic Analysis
for Soils and Plant Tissue
Voltammet ric 1 determinati on of substances which arc reducible
or oxidizable at the dropping mercury electrode

16-1. The minor nutrient elements, Cu, Zn, Fe, Mn, Co, and Mo, from
soils and plants are readily determined voltammetrically ( "polarogra phi-
cally'') with a dropping mercury electrode. Though highly effective colori-
metric methods are available for most of these elements (Chapter 15), the
polarograp hic method of analysis is the most effective for several of them,
especially Zn. It has been employed for Mn obtained in cation exchange
capacity measurem ent ( ~ 4-31 ) . Procedure s are given ( ~ I 6-34) for
dithizone separation of Cu and Zn from solutions derived from soils (~
15-52, 15-83) or plant tissue ( ~! J 2-39), and for determinat ion of Cu, Zn,
and Mn.
16-2. Organic compound s that are reduced or oxidized at the dropping
mercury electrode at characteris tic potentials can be determined . Possibili-
ties exist for many applications to soils and plant systems. For example,
the voltage recorded~ on a polarograp h attached between a plant stem and
a solution bathing the roots was equal to the ionization potential (EY2) of
the ion supplied in the solution. Also, the polarogram of a soil varies with
pH and exchangea ble cation.:{
16-3. Basic Principles. Polarograp hic analysis is based on the fact that
both qualitative and quantitativ e analyses can be made from the voltage-

1 Kolthoff and Jordan, Ana/. Chem., 25: 1833 (1953).

2 Breazeale and l\lcGeorge, Soil Sci., 75 :443 ( 1953).
:i Puri. Soils, their Physics and Chemistry (New York: Reinhold Publishing Corpo-
ration, 1949), p. 141.
416
POLAROGRAPHIC ANALYSIS FOR SOILS AND PLANTS 417
current curve obtained when an electro reducible or electro oxidizable ele-
ment or compound in solution is electrolyzed with an increasing voltage on
the electrodes (EMF). The analysis is thus essentially a voltammetric de-
termination. The voltage measured is the polarization voltage, recorded
graphically, hence the name polarograph. 4
16-4. The voltage is applied between a small, readily polarized electrode
(dropping Hg) and a large, nonpolarized electrode (and calomel cell-KCI
junction) separated by an indifferent supporting electrolyte (Fig. 16-1).

Dropping Hg
electrode

Saturated KC! Supporting

Calomel cell electrolyte

Fig. 16-1. Schematic arrangement of apparatus for polarographic analysis.

(After Menzel and Jackson, Anal. Chem., 23: 1861, 1951.)

The technique is distinguishable from amperometric, conductometric, elec-

trophoretic and electrodialysis determinations by its dependence on the oxi-
dation or reduction potential of the constituent. It has similarities to the

4 Heyrovsky, Chem. Listy, 16:256 (1922); automatic recording, Heyrovsky and

Shikata, Rec. trav. chim., 44:496 ( 1925); Heyrovsky, Po/arographie (Ann Arbor.
Mich.: J. W. Edwards, 1947); Kolthoff and Lingane, Polarography, 2nd ed. (New
York: Interscience Publishers, Inc., 1952); Muller, The Polarographic Method of
Analysis, 2nd ed. (Easton, Pa.: Chemical Education Publishing Co., 1951 ).
418 POLARO GRAPHIC ANALYSIS FOR SOILS AND PLANTS
graded cathode potential methodll of separation of metals by electrodeposi-
tion.
16-5. The characteristic current-voltage curve obtained is called the
polarogram (Fig. 16-2). At low potentials, the curve shows only a small

Discharge (reduction) of
7
second cation begins ------------

Diffusion current plateau

6

4
Wave-height
c or Characteristic voltage
or
~ 3 diffusion current, half-wave potential
:::J
(.) proportional to
(voltage at inflection point)
concentration
2

0
,RM"d"" l i
--=-------=~1dual current

0 -0.2 -0.8
Voltage applied to dropping Hg electrode

Fig. J6-2. General character of the polarogram or voltam-

metric curve of a single reducible substance. An ionic, atomic,
or molecular species or organic functional group can be identi-
fied qualitatively by the characteristic voltage, and determined
quantitative ly by the wave height.

"residual current" and only a slight rise in current with increasing EMF.
The EMF has not yet become sufficiently high to cause the main electrode
reaction to occur. Only a small current, due to cathodic capacitance effects
or possibly to minute traces of other materials, is flowing.
16-6. Diffusion Current. As the EMF increases further there is a sudden
sharp rise in the current flow, followed by a leveling off at a new and higher
current (Fig. 16-2). The rise in current indicates that an EMF has been
reached that is great enough to produce the electrode reaction with the con-
sequent increase in current passage. As soon as the electrode reaction be-

r. Diehl, Electrochemical Analysis with Graded Cathode Potential (Columbus,

Ohio: G. F. Smith Chemical Co., 1948).
POLAROGRAPHIC ANALYSIS FOR SOILS AND PLANTS 419
gins, the solute involved in the reaction becomes exhausted in the vicinity
of the small electrode, producing a condition of concentration polarization
at the electrode.
16-7. The supply of solute at the electrode is renewed only by kinetic
diffusion from the body of the solution, since movement of solute to the
electrode and the influence of the potential gradient between the electrodes
has been reduced to a negligible quantity by the addition of an excess of a
"supporting" electrolyte that does not enter into the electrode reaction.
This means that the resistance of the solution is low and thus provides a
very low potential gradient, E =IR.
16-8. The solute diffusion rate is proportional to the difference in con-
centration of solute at the electrode and in the body of the solution. Since
the concentration at the electrode is zero, the rate of diffusion is directly
proportional to the concentration in the body of the solution. When the
"diffusion current" has risen to the value corresponding to the rate of ar-
rival of solute at the electrode by diffusion, it levels off and remains constant
even though the EMF continues to rise. Since the rise in current is pro-
portional to the rate of diffusion of solute, it is also proportional to the
concentration of solute in the body of the solution, and this concentration
may be determined by measuring the height of the current rise or "step"
(wave height) through the application of the necessary instrumental fac-
tors.
16-9. The Ilkovic equation for diffusion current (it1) is
id = 605 n CD' 1'm'''t' 1' (16-1)
in which n is the number of electrons involved in the electrode reaction, C
is the concentration of the reduced or oxidized ion, D is the diffusion co-
efficient, m is mass of Hg flowing per second from the capillary, and t is
the number of seconds required for a drop to form. It will be seen that for
a given capillary and reacting substance, the current will be proportional
to the concentration of the reduced or oxidized substance. The usual range
of concentration is from 10-:1 to I0- 5 molar.
16-10. Sometimes a substance being determined causes a maximum in
the diffusion current just after its decomposition potential is reached and
then drops back to its normal current. The production of this maximum
may result from adsorption on the dropping mercury. The production of
maxima may be suppressed by the addition of a suitable organic reagent
such as gelatin or fuchsin. Gelatin also helps to stabilize the viscosity of the
solution through a series of test samples.
16-11. The voltage may shift from ( +) to (-) within the course of a
single wave step (Fig. 16-2).
16-12. The Characteristic Voltage or Half-Wave Potential. The mid-
point or inflection point of the current rise (Fig. 16-2) occurs at an EMF
420 POLARO GRAPHIC ANALYSIS FOR SOILS AND PLANTS
that is the constant for and characteristic of a particular solute under closely
prescribed conditions of environment, such as supporting electrolytes, ref-
erence electrode, or Hg drop size. The EMF corresponding to the midpoint
of the step is known as the "half-wave potential" and serves for qualitative
identification of the solute being oxidized or reduced. Each solute species
tends to exhibit its own characteristic oxidation or reduction potential, and
is not reduced until that potential is reached.
16-13. Supporting Electrolyte. A method for a specific ion species con-
sists of a supporting electrolyte solution with suitable complexing agents to
permit each solute species to be reduced or oxidized at a potential that is
well separated (about 0.2 volt) from the potentials at which other species
are reduced or oxidized. Dissolved oxygen must be removed from the po-
larographic sample before analysis, since, if present in the supporting elec-
trolyte, it would be reduced at a more negative potential than the elements
being determined, and thus would interfere. The supporting electrolyte
solution composition can be rather heterogeneous as to other solutes pres-
ent. But the accuracy falls off somewhat with increase in solute concentra-
tion much over 0.1 M, because of the increase in "residual current" not
attributable to the reduction of ions of interest (Fig. 16-2). Sometimes the
separation cannot be made by selection of supporting electrolyte composi-
tion, in which case the separation must be made by conventional chemical
methods prior to the instrumental analysis.
16-14. The Dropping Mercury Electrode. The dropping mercury elec-
trode consists of a very fine glass capillary through which mercury is pass-
ing at such a rate that it will produce falling drops at a constant rate (every
2 to 5 seconds). This mercury drop has proved to be the most suitable small,
readily polarized electrode. No electrode of more general scope has been
developed, although many other small metallic electrodes have been tested.
The dropping mercury electrode is peculiarly suited to polarography be-
cause of the following properties: (a) it gives complete polarization
rapidly, (b) it presents a continuously renewed electrode surface, and ( c)
it gives a high hydrogen "over-voltage."
16-15. At any fixed EMF, the current through the dropping mercury
electrode oscillates regularly over a small range because the growth and de-
tachment of mercury drops causes consequent variation in the electrode
area. Thus the recorded polarographic step is oscillatory in character rather
than even. The various methods of measuring step height automatically
compensate for these oscillations. The amplitude of the oscillations can be
varied by capacitive damping.
16-16. The rate of flow of the Hg is controlled by the capillary size to-
gether with the height of the Hg column. The Hg height is arranged to be
variable and is often scale mounted.
POLAROGRAPHIC ANALYSIS FOR SOILS AND PLANTS 421

COPPER AND ZINC DETERMINATION

( Polarographically in sulfite-ammonia supporting electrolyte)
16-17. Copper and zinc derived from soils or plants (~ 16-40) can be
determined in a solution 0.10 molar in NH4 0H and 0.25 molar in Na 2S03
by means of a single polarogram. 6 In this supporting electrolyte, the re-
duction reactions and half-wave potentials are:
Cu(NH3 ) 2 + + e- = Cu(Hg) + 2 NH 3 (H 20) (16-2)
E! = -0.50 volts
Zn(NH3 )~ ++ + 2e- = Zn(Hg) + 4 NH3 (H2 0) (16-3)
E 1 = -1.23 volts

The diffusion current is proportional to the concentration of Cu or Zn up

to about 10 millimoles per liter. As little as 0.05 millimoles per liter of Cu
can be determined within ± I 0 per cent. The accuracy is higher with Zn
or with larger amounts of Cu.
16-18. Ferric ion interferes with the polarographic determination of Cu
and Zn because it is reduced at a more positive potential. Therefore, Cu,
Zn, Co, and Ni are extracted from plant tissue digests by the dithizone
separation procedure ( ~ 16-34). Cobalt is reduced at nearly the same po-
tential as Zn; however, the content7 of Co in plants is usually less than I
ppm compared to 20 to 40 ppm of Zn commonly present. Thus Co rarely
influences the Zn determination by more than 2 per cent and can be
neglected except with plants of low Zn content. When the Co content is
sufficiently great to cause significant error, it can be determined colori-
metrically ( ~ I 5-115), and the equivalent Co wave height then can be
deducted~ from the Zn plus Co wave height. The wave for Ni is at -0.9
volt, well separated from the Cu and Zn waves, and Ni was determined
polarographically along with Cu and Zn in analysis of iron pyrites samples.
A polarographic accuracy of better than ± I 0 per cent for Ni determination
in plants would be difficult to attain consistently, as Ni often occurs in
lesser concentrations than Cu.

APPARATUS

16-19. Needed apparatus includes a recording polarograph, dropping

mercury electrode, saturated calomel half-cell, 2 x 7 cm sample vials, and
50-ml conical flasks. In addition to the usual cleaning, all glassware is
rinsed twice with 0.5 N HCI and twice with redistilled water.

"Menzel and Jackson. Anal. Chem., 23: 1861 ( 1951 ).

7 Beeson, U.S.D.A. Misc. Pub. 369 ( 1941 ).
B Cooper and Mattern, Anal. Chem., 24:574 ( 1952).
 PLANTS
422 POLA ROG RAP HIC ANALYSIS FOR SOILS AND
(or voltammetric9 )
16-20. Several kinds of commercial polarographic
ic film or by pen on
apparatus record the diffusion current on photograph
this, an electric motor
chart paper (examples, Fig. 16-3 and 16-4 ). To do

ing Polarograph . (Photo courtesy

Fig. 16-3. Sargent's Model XII photographic record
E. H. Sargent & Co., Chicago, lll.)

sly. The applied volt-

drives a voltage divider and the recorder synchronou
or negat ive value through
age range can be set to extend from a positive
a total range of up· to
zero to a value of the opposite sign (Fig. 16-2 ) , with
gh the dropp ing mer-
about 4.5 volts. The current ( ~ 16-6 ) passing throu
ter, the sensitivity of
cury electrode is measured by a sensitive galvanome
operated polarograph
which is adjusted with a variable shunt. A manually
le for an analysis in
(example, Fig. 16-5 ) is less expensive and is suitab
the step shape is approxi-
which only 1 substance is to be determined and
copyrighted names such as
Voltammetric ( ~ 16-3) appara tus is sold under
9
(Leeds & North rup Co.,
ro-Ch emogr aph"
"Polar ograp h" (E. H. Sargent & Co.), "Elect
and "Elecd ropod e" (Fishe r Scientific Co.).
POLAROGRAPHIC ANALYSIS FOR SOILS AND PLANTS 423

Fig. 16-4. Leeds and Northrup pen-recording Electro-Chemograph. (Photo courtesy

Leeds & Northrup Co., Philadelphia, Pa.)

mately a pure form. The latter is characteristic of reversible reactions in-

volved with simple ions in fairly high concentrations. The step height can
be measured as the difference between 2 points, 1 on the residual current
plateau and 1 on the diffusion current plateau (Fig. 16-2).
424 POLARO GRAPH IC ANALYSIS FOR SOILS AND PLANTS

Fig. 16-5. Fisher's manually operated

Elecdropode. (Photo courtesy Fisher Sci-
entific Co. , Pittsburgh , Pa.)

REAGENTS

16-21. Needed reagents include purified distilled water (distilled water

run through an exchange resin column or redistilled in Pyrex glass ap-
paratus) , 0.1 per cent gelatin ( 0.1 gm of gelatin is dissolved with gentle
heating in 100 ml of redistilled water), and the supporting electrolyte solu-
tion.
16-22. Supporting Electrolyte Solution. For supporting electrolyte
enough for 1 set of 12 determinations, 2.1 gm of Na 2S0 3 is dissolved in 66
ml of 0.1 N NH40H obtained by dilution of distilled NH 40H with redis-
tilled water. Because NH3 is volatile and S03 - - is oxidizable, the solu-
tion is freshly prepared each day.

PROCEDURE

16-23. To the dry sample residue from the dithizone separation proce-
dure (~ 16-41) exactly 5 ml of supporting electrolyte solution and 1 drop
of gelatin solution are added. For samples known to be very low in copper,
as little as 2 ml of supporting electrolyte may be added. One hour, with
POLAROGRAPHIC ANALYSIS FOR SOILS AND PLANTS 425
occasional gentle swirling of the flask, is enough to dissolve the copper and
zinc. The solution is then poured into a dry sample vial.
16-24. The mercury column above the capillary is previously adjusted
to a height that gives a drop time of about 4 seconds in the supporting
electrolyte with no applied potential. If the capillary is clean, the drop time
remains constant for a given height. The sample vial is placed in position
with the dropping mercury as 1 electrode and a KCl-agar bridge leading to
a calomel half-cell as the other electrode (Fig. 16-1 ). Recording of the
polarogram is begun with an applied potential of -0.2 volt on the dropping
mercury electrode. The Cu half wave potential is approximately -0.5 volt.
When the potential reaches -0.8 volt, it may be desirable to reduce the
sensitivity in order to record the Zn wave. The Zn half-wave potential is
approximately - 1.2 volt. The recording of the polarogram is stopped when
the potential reaches - 1.5 volts. The temperature of the solution is meas-
ured.
16-25. On the polarogram (photographic print, developed and dried, or
pen-recorded polarogram), the diffusion current wave height is measured
(Fig. 16-2). If the temperature of the polarographic solution was not
25°C, a correction of 2 per cent per degree difference is made, the cor-
rection being added if the temperature was lower and subtracted if it was
higher than 25°C. The relation between the diffusion current wave height
and the concentration of copper or zinc is determined by the use of stand-
ard samples.

ALTERNATIVE PROCEDURES

16-26. The sulfite-chloride supporting electrolyte (~] 16-32) can also be

used for copper and zinc determination. The half-wave potentials for re-
duction of copper and zinc are -0.3 and -1.0 volts, respectively. However,
the more negative reduction potential obtained for copper in the sulfite-
ammonia supporting electrolyte allows a more accurate estimate of low
concentrations of copper.
16-27. The polarograph has been applied to zinc analysis of soils and
plants after precipitation of iron, 10 for zinc after concentration by dithizone
extraction 11 (~ 16-34), for copper, 12 and for several elements simultane-
ously after the separation of silica. 13 Different supporting electrolytes have
been used by each investigator. Trace elements have also been preconcen-
trated by ion exchange. 14

10 Cummings and Reed, S.S.S.A. Proc., 5: 167 ( 1941 ).

11 Stout et al., Coll. Czech. Chem. Comm., 10:129 (1938); Walkley, Australian J.
Exp. Biol. Med. Sci., 20: 139 (1942).
12 Cranston and Thompson, Ind. Eng. Chem., A.E., 18:323 (1946).
13 Zak, Biedermanns Zentr., B. Tiererniihr., 14:301 ( 1942).
14 Riches, Nature, 158:96 (1946).
426 POLAROGRA PHlC ANALYSIS FOR SOILS AND PLANTS

MANGANESE DETERMINA TION

(Polarographicall y in sulfite-chloride supporting electrolyte)
16-28. Manganese can be determined in a solution 0.25 molar in
Na~S0 3 and 0.10 molar in NaCl. ln this supporting electrolyte the reduc-
tion reaction and half-wave potential are:
Mn+++ 2e- =Mn (Hg) (16-4)
E 1 = - 1.5 volts

Standard samples with manganese concentrations ranging from 0.01 to 0.1

millimoles per liter are determined within ± 5 per cent in the presence of
similar concentrations of copper and zinc.
16-29. Manganese is not extracted with Cu and Zn by the dithizone-
CC14 separation procedure (~ 16-34). Manganese is more abundant in plant
material than copper and zinc and its wave occurs after those of copper
and zinc at a suitable potential of -1.5 volts in the sulfite-ammonia sup-
porting electrolyte (~ 16-22). Manganese, copper, and zinc of soils and
plants could be determined on a single polarogram if the separation pro-
cedure were modified to include manganese and still exclude iron. A dis-
advantage for Mn determination arises because the diffusion current for
Mn in the sulfite-ammonia supporting electrolyte is abnormally low, 1.69
microamperes per millimole of concentration. The low diffusion current
may arise from the repression of manganese solubility. The solubility prod-
uct of Mn (OH)~ is exceeded and the ammonia concentration may be too
low to form a soluble manganous amine ion. In a supporting electrolyte
consisting of sodium sulfite alone or sulfite and chloride ( ~; 16-32), the
diffusion current for millimolar Mn is about 4 microampcres and gives a
satisfactory determination (~ 16-33).

APPARATUS

16-30. Needed apparatus includes a recording polarograph, a dropping

mercury electrode, a saturated calomel half-cell, 2 x 7 cm sample vials,
and 50-ml conical flasks. In addition to the usual cleaning, all glassware is
rinsed twice with 0.5 N HCl and twice with redistilled water.

REAGENTS

16-31. Needed reagents include redistilled water (distilled water is re-

distilled in the Pyrex glass apparatus), 0.1 per cent gelatin ( 0.1 gm of
gelatin is dissolved with gentle heating in 100 ml of redistilled water), and
the supporting electrolyte solution.
16-32. Supporting Electrolyte Solution. 2.1 gm of Na 2 S03 and 0.39 gm
of NaCl are dissolved in 66 ml of redistilled water. The solution is freshly
prepared each day.
 POLAROGRA PHIC ANALYSIS FOR SOILS AND PLANTS 427
PROCEDURE

16-33. The procedure for polarographic determination of manganese is

the same as that for copper and zinc, with appropriate substitution of re-
agents and voltages in recording the polarogram.
COPPER AND ZINC SEPARATION FROM RESIDUE OF
WET OXIDATION OF PLANT TISSUE 15
(Dithizone separation)
16-34. Diphenylthiocarbazone (dithizone) forms complex with most

l
transition metals according to the general reaction:

n S= C N=N-C>
1
+ Mn+ --+ S=C ;N=N-~ M + n H +
'N-N-C> 'N-N-0
H H H n
(16-5)
These complexes are to be quantitatively extracted from water solution into
CC1 4 • The pH range for stable complex formation varies somewhat with
the different metals. Thus, Cu, Zn, Co, and Ni are extracted at pH values
from 8 to 10. Ferrous ion is extracted at pH values from 6 to 7 and Mn is
reported to be extracted at pH 1 1. 16
APPARATUS
16-35. Needed apparatus includes a I -liter separatory funnel, 125-ml
separatory funnels, and 500-ml and 50-ml conical flasks.
REAGENTS
16-36. Needed reagents include thymol blue indicator, dithizone (di-
phenylthiocarbazone, Eastman Kodak Co., Rochester, N. Y.), redistilled
water (distilled water is redistilled in Pyrex glass apparatus), 6 N HCl (re-
distilled in Pyrex apparatus), 4 N NH 40H (NH 3 absorbed in redistilled
water cooled in an ice salt bath), and the following.
16-37. Ammonium Citrate Buffer. This buffer is made by adding 50 ml
of 10 per cent citric acid solution to 200 ml of 4 N NH 40H.
16-38. Distilled CCl4 • Technical grade or used CC1 4 is purified by wash-
ing successively with 20 per cent H 2 S0 4 , 20 per cent NaOH, and distilled
water. The washed product is distilled over Na2 C0 3 in Pyrex glass ap-
paratus.
PROCEDURE
16-39. Three hundred ml of ammonium citrate buffer, 10 ml of redis-
tilled CC1 4 , and 0.1 gm of dithizone are shaken together vigorously for 1
minute in a 1-Iiter separatory funnel. The buffer solution becomes red, ow-
15Menzel and Jackson, Anal. Chem., 23: 1861 ( 1951).
An extensive discussion of dithizone reactions will be found in Welcher, Organic
1fl
Analytical Reagents, Vol. 3 (New York: D. Van Nostrand Company, Inc., 1947).
 TS
428 POLAROGRAPHIC ANALYSIS FOR SOILS AND PLAN
Copp er and zinc
ing to the solubility of ammonium dithizonate in water.
ng the CC1 4 layer out of
impurities from the reagents are removed by drawi
more by shakin g for 1
the funnel. The buffer solution is extracted once
a clear green color,
minute with 10 ml of redistilled CC1 4 , which then has
hlorid e.
due entirely to the solubility of dithizone in carbon tetrac
ml of purifie d buffer soluti on contai ning dithizone,
16-40 . Twenty-five
tory funnel. The
and 5 ml of redistilled CC1 4 are placed in a 125-ml separa
, 15-57 , 15-83 ) is
plant tissue digest (~ 12-39 ) or soil extract(~ 15-52
the flask with redis-
transferred to the separatory funnel with 2 washings of
l is shake n for 1 minut e to bring most of
tilled water. The separatory funne
layer. If the pH of the aqueous phase is
the copper and zinc into the CC1 4
by thymo l blue indica tor on a spot
not between 9 and 10, as indicated
or distill ed NH 0H. The funnel is
plate, it is adjusted with distilled HCI 4

tetrac hlorid e phase is withdrawn

again shaken for l minute and the carbon
collec ts betwe en the layers
into a 50-ml conical flask. None of the silica that
ngs with redisti lled CC1 4
must enter the stopcock bore. Two 2-ml washi
indica ted by the clear
usually suffice to remove all the copper and zinc, as
green color of the CC1 4 layer in the last washing.
to dryness. The
16-41 . The carbon tetrachloride extract is evaporated
l flask with 2
dithizone is oxidized by digesting the residue in the conica
c hot plate for
ml of ternary acid mixture (~ 12-24 ) at 300°C on an electri
briefly and cau-
2 hours. Finally the sides of the conical flask are heated
of sulfuric acid.
tiously above a Meker burne r to remove the last traces
graphic analysis (~
The Cu and Zn in the residue is ready for the polaro
16-23 ).
QUES TION S
polarograph.
1. State the basis for qualitative analysis by means of the
the polarograph.
2. State the basis for quantitative analysis by means of
(b) residual curren t,
3. Define the following: (a) supporting electrolyte,
( c) diffusion curren t plateau.
with the polaro graph
4. Draw a typical voltammetric curve such as obtained
pure form: label the diagra m as to residual
when the step shape approximates indica te the wave
u; and
current, half-wave potential, diffusion curren t platea
height.
that are available
5. List the 3 general types of voltammetric instruments
ating types rather than comm ercial makes are wante d.)
commercially. (Oper
the supporting elec-
6. What is the purpose of the addition of gelatin to
trolyte?
NH 4 0H-N a 2S0 3 sup-
7. List the half-wave potentials for Cu and Zn in the
portin g electrolyte.
ined in a single
8. Under what conditions can Cu, Zn, and Mn be determ
solution and record ed on a single polaro gram?
ne separation of
9. Outline the typical procedure employed for the dithizo
Cu, Zn, etc. from iron. Why is the separa tion necess ary?
 17
Absorption Spectrophotometry
. . . a .vystematic property of the material containinR the con-
stituent, such as color . . . liRht ahsorptive capacity . . .
-MELLON 1

17-1. Absorption spectra have been observed by man since earliest

times, through visual color in rainbows, stones, waters, plants, and animals.
White light changes to colored light when some wave lengths have been ab-
sorbed, and substances from which such light is passing are said to be
colored. Precisely speaking, the substance shows "light absorptive capac-
ity,'' and the actual color is dependent on the quality of light incident on
the material. Certain wave lengths are preferentially absorbed, but the color
seen is that characteristic of the wave lengths not absorbed. Thus, the hue
of color seen will vary according to the spectrum of the incident light.
17-2. Basic Principles. Energy of atoms and interatomic bonds in crys-
tals, in solvated ions in solution, in radicals, or in molecules is subject to
changes in level by absorption of light energy. Light absorption of a general
character merely heats up the absorbing material without coloration. But
usually a given absorbing mechanism favors absorption of light in specific
short wave length bands; and the resulting variation in light intensity as a
function of wave length is an absorption spectrum. Absorption spectro-
photometry began by visual estimates of intensity and hue of light trans-
mitted by a sample. Quantitative measurement of color began with com-
parison of standard and test solutions in tubes of the same size and shape.
Flat bottoms and parallel sides incorporated in Nessler tubes permitted ac-
curate comparisons of different thickness of solution reciprocally with dif-

1 Colorimetry for Chemists (Columbus, Ohio: G. F. Smith Chemical Co., 1945),

p. iii.
429
430 ABSOR PTION SPECT ROPHO TOMET RY

ferent concentrations. Addition of lenses to bring the view of the two tubes
together gave the visual colorimeter. 2
17-3. Colorimetry concerns the absorption of visible light, usually by a
solution. Absorpt ion spectrophotometry 3 (also absorption spectroscopy )
4

concerns radiation absorption by solid or solution in the visible, ultravio let,

and infrared wave length ranges of electromagnetic radiation. Absorption
spectrophotometry performs 2 distinct functions: (a) radiation absorption
is measured as a function of concentration at a given narrow wave length
band(~ 17-19) and (b) the radiation absorption is measure
d as a function
of wave length (narrow band increme nts), for a given constan t quantity of
sample, to obtain the absorption spectrum (~ 17-44). Emissio n spectro-
photometry (also emission spectroscopy) refers to measure ment of the
quality of spectrum and quantity of radiation emitted by the sample con-
stituents (Chapte r 18). Taken together, the emission and absorpti on tech-
niques are the basis of spectrochemical methods. There is no sharp
boundary between the 2 techniques, because radiation absorption effects
accompany emission. Also, fluorescent emission sometimes occurs when
radiation is absorbed in solutions or in crystal s-the basis of fiuorescence
analysis.
17-4. Spectral Sensitivity of Human Vision. Human vision detects elec-
tromagnetic wave lengths extending from about 400 mu to about 730 mu,
the range of "visible" radiation. The spectral colors defined by human
vision have wave length range in mu as follows:
violet-4 00 to 450 green-5 00 to 570 orange- 590 to 620
blue-45 0 to 500 yellow- 570 to 590 red-620 to 730

Eye sensitivity in the extremes of the visible range vary considerably in

different individuals. The human eye has its maximum sensitivity to green
at 540 mu (520 to 600), responding to as little as 4 or 5 quanta of light.
5

17-5. Photocell Sensitivity to Wave Length. Each type of photocell has

its wave length band of maximum sensitivity (Fig. 17-1). The photronic
cell (also termed rectifier, barrier layer, or photogalvanic cell) has a sensi-
tivity similar to human vision, and is suitable for measurement of the wave
lengths emitted by the incandescent wolfram (W) filaments used in ordi-
nary light bulbs. The photronic cell gives sufficient current to be registered
directly on the galvanometer without amplification. Because of this, and
D. Van
2 Snell and Snell, Colorimetric Methods of Analysis, Vol. I (New York:
Nostrand Company , Inc., 1936).
23:2
3Hiskey and Young, Anal. Chem., 23:1196 (1951); Mellon, Anal. Chem.,
Mat. Bui.,
(1951); Rosenbaum, Anal. Chem., 23:12 (1951); Stillman, Am. Soc. Test.
125: 17 (194 3 ) .
4 Mellon, Analytica l Absorpti on Spectroscopy (New York: John Wiley
& Sons, Inc.,
1950).
5 Hecht, Am. Scientist, 32: 159 ( 1944).
ABSORPTION SPECTROPHO TOMETRY 431

Photoemission type KH vacuum cell

7

6
~

.!!?
·c
.,::i 5
>
...
:;;

!c: 4
0
a3
0
"'
.t:J
<
2

500 600 800

Wavelength-mµ

Fig. 17-1. Relative response of photoelectric cells and the hu-

man eye compared to energy distribution of wolfram (W) filament
emission. (After Withrow, et al., Ind. Eng. Chem., A. E., 8:214,
1936.)

since no cathode heating power is required, maximum simplicity in instru-

ment design is possible. Typical potassium hydride (KH) vacuum tube
photocells are most sensitive in the shorter wave lengths of the violet and
ultraviolet spectrum. The Cs~O type vacuum tube cell has maximum
sensitivity for red and infrared wave lengths. Choice of photocell type is
influenced by the absorption spectrum of the colored constituent to be de-
termined (~ 17-6).
17-6. Sensitivity Dependent on Wave Length of Incident Light. The
color or wave length band of incident light used for a colorimetric de-
termination is selected to coincide with that most absorbed by the test solu-
tion. The greatest sensitivity is thus obtained because the determination of
concentration is based on the change in light intensity that results from
light passage through the colored solution. Little or no change occurs in
intensity of wave lengths other than those specifically absorbed.
17-7. Example. Consider that 90 per cent of the incident light is of wave
lengths that are not absorbed, and that only 10 per cent of the incident light
is of wave lengths that are absorbed. Further, Jet it be supposed that the
concentration of the colored solution is such as to absorb 25 per cent of the
incident light of the wave lengths appropriate to be absorbed. Then the trans-
mitted light intensity is 90 per cent plus o/<i X 10, or 97.5 per cent. Little
sensitivity of measurement could be expected in this case, since only 2.5 per
 TRY
432 ABSORPTION SPE CTR OPH OTO ME
rbed. Next, let it be supposed that a
cent of the incident light would be abso
light which removes 8/9 of the light
light filter is interposed in the incident
tion, and transmits all of the light of
of wave length not absorbed by the solu
abso rptio n by the solution. The incident in-
wave length appr opri ate for
thro ugh instr ume nt settings. The n
tensity is raised by a facto r of ( 100/ 20)
the transmitted light is:

-100 x 10 + -100 ( -3 x 10) = 87.5%

20 20 4
ased to 12.5 per cent, a 5-fold in-
The relative absorption has been incre
the wave lengths not absorbed by the
crease in sensitivity. Removal of all of
n of 25 per cent of the incident light and
solution would result in absorptio
g a 10-fold increase in sensitivity.
transmission of only 75 per cent givin
itivit y is increased by the use of inci-
This example points out how the sens
ths most efficiently absorbed by the
dent light rich in precisely the wave leng
solution.
The light absorption band of
17-8 . Selection of Wave Length Band.
gh to permit the incident light wave
some colored solutions is narrow enou
daries (Fig. 17- 2). The light ab-
length band to be kept within both boun

Light transmission ///

'' \
maximum of filter or slit
I
\ I
\ I
\ I
\ .
. absorption
I light
\ I maximum of KMn 0 4

v
/
\
\
\ I
\ I
\ I
\
\ I
580
Wavelength·mµ

the incident beam

Fig. 17-2 . The light maximum of should coincide
imet ric meas urem ent
selected for a color ed solution being
with the absorption maximum of the color
meas ured .
ABSORPTION SPECTROPHOTOMETRY 433
sorption by other colored solutions cuts off the portion of the spectrum
below a given wave length (Fig. 17-3). For the latter, the incident light

Light absorption /
curve of blank--../
/ //
I I
-
. r:
~
I I
-.,u
c:
"O
/
I light transmission
maximum of I
I
.!: / fi lier or slit I
~
I I
.2l I
~ I I Light absorption
;;;
c: I J-
0
c I / curve of solution
0
;;; I I
"'E I
"'c I
~., I I
00 I I
~
.,uc I I
I I
~ /
/
400 500
Wavelength-mµ

Fig. I 7-3. Light maximum of the incident beam

selected to coincide with the absorption maximum of the
colored solution, but at maximum transmission of the
blank.

wave length band is kept below the steep rise in the absorption curve of the
test solution, but above the strongly absorbed wave lengths of t~e blank.
The appropriate light absorption wave length bands for most colorimetric
procedures are available in the literature but can be determined or re-
checked by the analyst (~ 17-44). Some examples are 535 mu for HMn0 4
(~ 5-69), 620 mu for cobalt hydrocarbonate green(~ 6-31), and 660 mu
for molybdophosphoric blue (~ 7-4). It is frequently possible to obtain
more than one range in sensitivity of a colorimetric method by selection
of a succession of incident light maxima which progressively encroaches on
the rising portion of the absorption curve of the solution, as for vanado-
molybdophosphoric yellow (~ 7-61). Nonlinearity often accompanies the
less sensitive calibration curves in such a series.
17-9. Obtaining the Wave Length Band. The appropriate light wave
length band is obtained either (a) by passage of the light through light
filters or (b) by slits to cut off a segment of a light spectrum arrayed by
 OMETRY
434 ABSORPTION SPECTROPHOT
are g~~n­
grating (~ 17- 47) . Light filters
means of a prism (~ 17- 46) or specti':al
2 or more glass plates of selected
erally made by superpositioning r yellO\v
. 17- 4), much as coloring ove
transmission characteristics (Fig

Standard
combinations
A

Special
"'monochromatic"
combinations

700mµ

series

700m µ
Wavelength in millimicrons
transmis-
acteristics. A, percentage light
Fig. 17-4 . Glass light filter chare with the Evelyn colorimeter, each rela-
labl
sion of a num ber of filters avai ll and lam p
sion and weighted for photoce
tive to the max imu m transmis band filter series. C, D, percentage trans-
characteristics. B, special narr ow glasses utilized in light filter construction.
ent
mission of a num ber of com pon
Philadelphia, Pa.)
(Fro m Bulletin 460, Rubicon Co.,
Since the
ow paint with blue gives green.
crayon with blue or mixing yell ctrum in
t filters advance through the spe
spectral bands provided by ligh ed an
r instrument is sometimes term
discrete increments, a light filte 20 mu
tometer. Narrow wave bands of
"abridged" absorption spectropho & Lomb,
rference in 1 colorimeter (Baush
or less have been obtained by inte th.with
wave filters of 7- to 15-mu wid
Rochester, N.Y .). Special narrow
ABSORPTION SPECTROPHOTOMETRY 435
high percentage transmissions are available (Baird Associates, Cambridge,
Mass.).
17-10. Percentage of Light Transmission. The fraction of the incident
light of intensity, / 0 , which is transmitted by the solution is termed trans-
mission, T, defined:

T= ~ (17-1)
lo
in which I is the light intensity after passage through the solution. Per-
centage transmission is 100 I/ lo- In any spectrophotometer in which the
nature of the photoelectric circuit is such that the current, G, flowing
through the galvanometer is directly proportional to the light, /, striking
the photocell, G /G 0 , may be substituted for I/ lw G 0 is the galvanometer
reading of the blank solution. It is apparent that the transmittance or T
scale of such absorption spectrophotometers is simply the usual linear
galvanometer scale.
17-11. When the percentage of light transmitted or absorbed is plotted
against the concentration, a curve results (Fig. 17-5). Evidently, the per-

100

80 •

\
c:
0
'Vi
</)

E
</)
60
c:
~
Q)
OD
40

·~
"'
i:
Q)
u
Q;
n.
20
·~
00 2 3 4 5
Concentration

Fig. 17-5. Percentage light transmission

plotted against concentration on linear graph
paper. illustrating nonlinearity of the relation-
ship.

centage of light transmitted or absorbed is not proportional to concen-

tration. A quantitative insight into the relationship of light absorption to
concentration (Beer's law) is desirable.
17-12. Beer's Law. Beer's law states that equal successive increments,
de, of concentration of a true solution containing a colored constituent
 RY
436 ABS ORP TIO N SPE CTR OPH OTO MET
ent light. (Lambert's law states
absorb equal fractions, -diI I, of the incid
thickness" of the solution.) The
the same proposition for "increments of
tion increases, thus the negative
light intensity decreases as the concentra
sign:
-di (17- 2)
--- = kdc
I
changes of intensity correspond-
Summing up by integration all the small
ing to increments of concentration,

f kdc = .! ~/!_ (I 7-3)

becomes,
which on performance of the integration,
(17- 4)
kc= -In I+ A
tion, that is, when c =0 and
in which A is a constant. At zero concentra
I= / 0 ,
(17- 5)
A = ln 1 0

17-4 ),
Thus, in general, substituting ( 17-5 ) in (
{17- 6)
kc = -In I + Zn 10

Or, in the form of Beer's law,

I /0 1 I0
c= k lnT = -k-1 log 10 T (17- 7)

ity, D, of a solution (or solid)

17-1 3. Optical Density. The optical dens
is defined as:

D = log 10 1= log 10 - } {17- 8)

concentration:
and, from equation 17-7 , is proportional to
D = k 1c (17- 9)

when Beer's law applies (Fig. 17-6 ).

sometimes desirable to calcu-
17-1 4. In reviewing older literature, it is
depths or volumes in Nessler
late optical density from different solution
ler tubes:
tubes. The equation used in matching Ness
ml (std ) cone (test ) (17- 10)
- - - - ---·-- ---·-· ·--
ml (test ) cone (std )

resolves itself into:

__ml (st~ = k c (17- 11)
2
ml (test )
ABSORPT ION SPECTRO PHOTOM ETRY 437

1.0

0.8
c:i~
_i. c.J
·v; 0 0.6
c: ,;;;
Q) 0
"O-;
-N 0.4
.~ Ii
c.c:i
o~ 0.2

2 3 4 5
Concentration

Fig. 17-6. Optical density plotted

against concentratio n, both on linear
scales. A straight line indicates con-
formity to Beer\ law.

in which c is the concentrat ion of the test solution and k 2 is a proportion al-
ity constant (reciproca l to constant standard concentrat ion). Then from
equations 17-9 and 17-1 1, D is proportion al to the ratio E11 ~~d) .
ml (test)
17-15. Extinction Coefficient. Optical density per unit of concentrat ion
and per unit thickness is the extinction coefficient. Color producing capacity
of a given chromogen (color producing constituen t) is important in the
evaluation of sensitivity of methods. It is quantitatively expressed in the
extinction coefficient, k:
I I
k= log 10 ° (17-12)
ex I,,
in which x is cm thickness of solution through which the light, /,,, passes,
and c is the concentrat ion in gm per liter. Optical density per mole per
liter per cm of thickness is the molar extinction coefficient, E:

L~ --
r.. MI og10 -10
Mk -- -- (17-13)
ex Iii]
in which M is the molar weight in gm.
17-16. Linearity of Calibratio n Curves. When the percentage transmis-
sion is plotted on a logarithmic scale against concentrat ion (normality ,
mgm per liter, ppm, etc.) on a linear scale (Fig. 17-7), points lying on a
straight line indicate conformity to Beer's law ( eq. 17-7). The point ( +)
in Figs. 17-5, 17-6, and 17-7 are the same point. Such a semilogarithmic
plot substitutes for mathematical or instrumental conversion to optical
density in testing for adherence to Beer's law.
438 ABSORPTION SPECTROPHOTOM ETRY
100 '-
90 .........
80
~
70
~I..
60
50
"-..
40

"' ~
""" ~ ~
~ ~
'

4
0 2 3 4 5
Concentration (linear scale)

Fig. 17-7. Percentage of light transmission plotted against concentration on semi-

logarithmic paper. A straight line indicates conformity to Beer's law.

17-17. Plotting to obtain linear calibration curves is a distinct advantage.

Data points that are at variance with the trend and the general reproduci-
bility of a colorimetric method can be judged. It is to be noted, however,
that nonlinearity of the calibration curve does not hinder graphic determina-
tion of concentrations of test solutions or use of a colorimetric method that
does not obey Beer's law (~ 17-20).
17-18. If the structure of the colored ions or of the colored nonelectro-
lytes in the dissolved state does not change with a change in concentration,
Beer's law will usually hold over a wide range of concentrations. However,
deviations from Beer's law are likely to occur in solutions whose color is
dependent upon dissociation phenomena, complex formation, and suspen-
sions of precipitates. The color may then depend upon the concentration
of the reactants, temperature of preparation, time of standing, and the
presence of other electrolytes in addition to the concentration of the light
absorbing substance under analysis.
CONCENTRATION DETERMINATION BY ABSORPTION
SPECTROPHOTOM ETRY
17-19. The concentration of a sample component that gives color is
measured by the height of a given light absorption peak of the sample
ABSORPTION SPECTROPHOTOMETRY 439

relative to that of a standard. The wave length position of the characteristic

absorption peak must be known ( ~ 17-8) or determined ( ~ 17-44) before
the simpler routine analytical determination of concentration can be made.
But once an absorption peak suitable for the analysis has been determined
for a given colored compound, it seldom has to be repeated, even though
many modifications of analytical procedure may be made concerning pre-
concentration, purification, color development, and other steps in the
colorimetric method.
17-20. Colorimetric procedures are most efficacious when the color (a) is
highly specific for the element to be determined, (b) is intense per unit of
concentration, ( c) is stable, ( d) is developed rapidly at room temperature,
( e) is developed in the same solvent as the reagents and does not require
extraction with an organic solvent, and (f) is not affected by excess of
reagents or changes of pH. Knowledge of the origin of the colored con-
stituent is helpful in procedure design. The color of solutions frequently
arises from oxidation, reduction, formation of a complex ion, or coupling
with a large molecule. Color of crystals has been correlatedH with struc-
ture and the presence of mixed valences of the same element.

APPARATUS

17-21. Needed apparatus consists of a photoelectric absorption spectro-

photometer with spectral control either with light filters or with prisms or
gratings, and cells or tubes of appropriate dimensions and light trans-
mission character to hold the test and standard solutions.
17-22. Several kinds of commercial absorption spectrophotometers are
available. Representative of the glass filter instruments is the Evelyn 7 photo-
electric colorimeter (Fig. 17-8) which is used extensively in soils labora-
tories. The popularity of the instrument is attributable to its simplicity and
accuracy. The glass filters (~ 17-9) give a perfectly reproducible incident
light of sufficiently narrow wave band ( 20- to 60-mu width) for high
sensitivity. The more generally used macro section of the instrument re-
quires 6 to 10 ml of solution and is 5- to 20-fold more sensitive than visual
colorimeters. A micro section, requiring 0.15 to 1.1 ml, provides a 40-
fold increase in sensitivity. Great stability of light source is attained by
use of a 6-volt storage battery to operate a small bulb about the size of a
pocket flashlight bulb, at much less than its rated voltage. A photronic
cell (~ 17-5) is used. Light filtration is effected prior to incidence on the
solution (Fig. 17-9), heating of the solution thus being avoided. The in-
strument is designed to use standardized test tubes approximately 2 cm
in diameter for holding the colored solutions.

6 Weyl, J. Phys. Colloid Chem., 55: 512 ( 1951).

7 Evelyn, J. Biol. Chem., 115:63 (1936).
 Fig. 17-8. Evelyn photoelectric colorimeter, a prec1S1on light filter absorption
spectrophotometer. The galvanometer reads percentage transmission on a linear sea.le
directly from current output of photronic cell without amplification. (Courtesy Min-
neapolis-Honeywell Regulator Co., Rubicon Instruments , Philadelphia, Pa.)
D D'
Test tube Diaphragm
(D, D ')

Photronic
cell

6 v. bulb

6 v.
storage
battery

Fig. 17-9. Diagram of Evelyn colorimeter, illustrating typical components of the

light filter type of absorption spectrophotometer. Special features include light filtra-
tion prior to incidence on the test solution, low energy light source, test tube com-
parator cell, and direct reading photronic cell, requiring no amplification.
440
ABSORPTION SPECTR OPHOTO METRY 441

Fig. 17-10. Coleman Junior absorption spectrophotometer, used for colorimetric

determinations in many soil testing laboratories. A replica grating gives a continuous
series of wavelength bands from 400 to 700 mu (Fig. 17- 14). (Photo courtesy Cole-
man Instrument , lnc., Maywood, Ill.)

17-23. Grating or prism type spectrophotometers (~ 17-46, 17-47) are

also used for concentration determination (example, Fig. 17-10). Care
is required in reproducibly setting the wave length band for the determina-
tion of a given constituent.
17-24. Double-beam absorption specttophotometers provide for com-
parison of the light coming simultaneously through the standard and test
sample. They are advantageous when the light source intensity is variable.
The voltage corresponding to the difference in light intensity between the
standard and test sample is balanced out (galvanometer null point) and
translated by scale calibration of the potentiometer into percentage trans-
mission or optical density. Galvanometer null-point reading may also be
built into single-beam instruments, and virtually always is used in vacuum
photocell instruments.
442 ABSORPTION SPECTROPHOTOM ETRY

REAGENTS

17-25. Standard Solutions. A series of standard concentrations of the

constituent to be determined is prepared, and the color is formed according
to the appropriate analytical procedure. All of the reagents included for
soil extraction should be included in the standard series.
17-26. A blank solution is prepared exactly the same way as in the pro-
cedure, including extraction solution and all other reagents and conditions.
A slight color may appear because of impurities in the reagents.
17-27. A series of dichromate (440-mu maximum) or permanganate
( 535- or 540-mu maximum) solutions of appropriate color intensity range
may be prepared by dilution of a standard solution. The permanganate is
diluted as in the manganese procedure ( ~ 5-68) and the dichromate with
distilled water. The proper concentration range is determined to give 4 to
6 well-distributed transmission readings between 20 and 95 per cent, be-
ginning with 0.05 N solutions (oxidation-reduction normality).
PROCEDURE

17-28. Operation of Evelyn Colorimeter. A suitable light filter(~ 17-9)

is placed in position in the slot at the left of the coarse adjustment knob.
The filter holder is pushed down (a) until a catch and notch are felt to
become engaged in order to bring the lower filter in position, or (b) until
it will go no further, to bring the upper filter into position. Next, the bake-
lite diaphragm tube is pulled up until the first encircling line appears above
the panel, and then it is rotated until the number 10 faces the operator and
an engaging pin is felt to arrest the tube. Caution: the diaphragm tube in
the photometer must be kept carefully in place as indicated by its arrest
by the engaging pin. Even slight displacement will greatly change the ob-
served readings. The galvanometer lamp switch is turned on, and then the
galvanometer is adjusted to give a zero reading by turning the lever on top
of the box and by sliding the frosted glass galvanometer scale. The colori-
meter lamp is turned on and allowed to warm up for 5 minutes.
17-29. Operation of the Beckman Absorption Spectrophotometer as
a Colorimeter. The Beckman (Fig. 17-11) light source and instrument
switches are turned on and the instrument is allowed to warrn up for a
period of 10 to 20 minutes. During this time the dark current adjustment
is kept at zero, with the light shutter closed. When it becomes unchanging,
warm-up is complete. In the meantime, the wave length selection(~ 17-8)
is made and the slit-width is adjusted to the appropriate value, usually as
narrow as possible within limits of sensitivity of the method. A sensitivity
control gives a 10-fold range in sensitivity.
17-30. Calibration of Test Tube Absorption Cells. The absorption cells
must be accurately paired optically, either by the manufacturer or by the
analyst. Ordinarily the analyst makes his own selections and pairing of test
ABSORPTION SPECTROPHOTOMETRY 443
tubes if this type of absorption cell is employed. To do this, the test tubes
that fit the instrument are cleaned in warm chromic acid cleaning solu-
tion, but cleaning with a brush is avoided because of the likelihood of pro-
ducing scratches. The tubes are half filled with distilled water or the solu-
tion to be analyzed (the latter is particularly important with solutions of
high refractive index, ~ 14-22). The same light wave length band is used
as will be used in the determination. With the tube in place in the instru-
ment, the settings are adjusted so that the reading of the first tube falls at
100 per cent transmission. The tube orientation is adjusted for the most
stable reading and the orientation is marked. Successive tubes are placed in
the instrument and a set selected which give readings within ± 0.5 per cent
of the average. Alternatively, 2 accurately paired tubes ( 1 for reference
and 1 for test solution) may be selected to give the same light transmission
at all orientations. With solutions of high refractive index ( ~ 14-21 ) , each
tube of a set is numbered and its deviation from the mean is recorded as a
correction.
17-31. It will be noted that the transmission reading of the Evelyn col-
orimeter is lower without a tube (air-reading) than with the tube in place
in it. The reason is that the cylindrical tube acts as a condenser lens and
increases the amount of light that passes from the light source through the
diaphragms to the photocell. The air reading or "center setting" is noted and
kept constant by rotating the rheostats as required. This is equivalent to
maintaining the standard tube reading at 100. Changes in filament and cir-
cuit temperatures at first cause a drift from the standard adjustment of
100, which is easily corrected in this manner.
17-32. The Blank Setting. The blank (~I 17-26) is placed in position for
light transmission and the percentage transmission is adjusted to 100
per cent by means of the light intensity controls ("sensitivity" on the Beck-
man). When the light transmission cannot be brought to 100 per cent in
the usual range, a "brighter" setting may be switched in (spreading the
0 to I 0 per cent section of the scale to 0 to 100 per cent for the Beckman).
If the blank setting still cannot be brought to 100, the attained transmis-
sion percentage with the blank can simply be recorded (~ 17-43). The
"center setting" ( ~ 17-31 ) of the Evelyn colorimeter is recorded.
17-33. Standard Solutions. Each standard solution tube or cell is pol-
ished externally and inserted (after check ()f both zero point and 100 per
cent transmission reading if much time has elapsed after the original ad-
justments) and the percentage transmission is observed to the nearest
0.25 per cent, recorded (Table 17-1 ) , and plotted as a calibration curve
(~ 17-16and 17-35).
17-34. Test Solutions. The same procedure is followed for the test solu-
tions as for the standard solutions. Greater accuracy 8 is assured if the read-
s Ayres, Anal. Chem., 21 :652 (1949).
444 ABSORPTION SPECTROP HOTOMET RY
ings of the test solutions are kept between 20 and 80 per cent transmission
(a 3 7 per cent transmission gives the lowest error). The concentrations are
maintained in the optimum range by suitable aliquot and dilution volumes.
The total volume of test solution is recorded. (The solution volume in the
Evelyn tubes may be decreased to 6 ml if the "6" mark on the Bakelite
diaphragm is turned toward the operator and raised to engage the pin. Tube
selection and calibration curve must be repeated,~ 17-33.)

TABLE 17-1
Record form for absorption spectrophotom etric measurement of concentration

-----,--~~~~::
Solution I diluted to
No. _ _ml
-·-···-·.---- ---- -----· - - -
ml
0 (blank) 100.0 0.0 0.0 0.0
I ........ .
2 ........ .
3 ........ .
4 ........ .
5 ........ .
6 ........ .
7 ........ .
8 ........ .
9 ........ .
]{) ........ .
0 Takt-•n of' staudanl or from calibration curve of test sample.

17-35. Interpretation of Results. Concentration is determined from per-

centage transmission readings (Table 17-1) through a calibration curve
(Fig. 17-7). The concentration may be expressed as normality, gm per
liter, ppm, or other convenient units. Quantity of constituent determined
is the product: Concentration x Volume.
17-36. Concentration can also be calculated (~ 17-13) from optical
density, D, by proportion:
D, test sample
Cone test sample= --- ---- -- x Cone ' standard {17-14)
' D, standard
Since the blank was set at 100 per cent transmission, from equation 17-7,
optical density, D, is given:
100
D = log · -- = log 100 - log G
G
=2 - log G (17-15)

The value of D is often replaced by L, giving L = 2 - log G. Tables of

optical densities (L = 2 - log G) for various percentage transmission
values are available. Values may also be calculated from ordinary log
ABSORPTI ON SPECTROP HOTOMET RY 445

tables. Some instruments have optical density scales, as well as percentage

transmission scales, so that optical density values can be read directly and
plotted. The optical density is converted to concentratio n by equation
17-14.
17-37. Before leaving the Evelyn colorimeter at the end of work, the
following are checked:
I. The lamp switches are turned off.
2. A light filter is always left in position for use (a protection against ex-
cessive illumination reaching the photocell when the operation is resumed).
3. The coarse rheostat is turned to the left (to lower the intensity to
guard against excess illumination if a less dense filter is next used).
17-38. Excess drift of the Evelyn galvanometer usually indicates either
that the battery needs recharging or that it has been over-charged. Over-
charging can be corrected by shorting the battery for a few seconds and
then allowing the lamp to run for 20 minutes. Unsteadiness of the rheostat
settings usually can be corrected by wiping the rheostat coils with a soft
cloth and smearing a light dressing of Vaseline along the region of contact.
17-39. Polychromogen Systems. Polychromogen (more than 1 color
source) systems are of 3 types. The simplest is monochromatic, with the
reagent blank carrying a certain amount of color of the same hue as that
derived from the test sample. Justification for setting the colored blank at
100 per cent transmission to eliminate the effect of the blank comes from
the fact that optical densities are additive, that is:
D (test) =D - D (blank) (17-16)

in which D (or L, ~! 17-36) is the optical density of the test sample if read
against a colorless blank not containing any analytical impurities. Or, in
terms of galvanometer readings,

D( ) ' Go l Go (17-17)
test = iog G - og G(blank)

in which G values correspond to the reading of each colored solution

against a clear solution blank. This simplifies to:
G (blank)
D (test) = log (]--- - (17-18)

which is the mathematical equivalent of reading the test sample against the
colored analytical blank; and the latter, simpler procedure is thus estab-
lished as fundamental.
17-40. The second case is polychromatic, in which the reagent blank has
an invariable color different from that of the principal chromogen in the
test sample. This effect is cancelled by maintaining the blank color con-
stant in the test sample as a "background " color, and minimizing its effect
446 ABSORPTION SPECTROPHOTOMETRY
by using a filter system (~ 17-8) that is designed to maximize absorption
by the principal chromogen and minimize absorption by the extraneous
chromogen of the blank.
17-41. The third case is polychromatic, in which the reagent color varies
with the amount of the test element and its light absorption cannot be com-
pletely separated by selection of the incident light wave length maximum.
This type of system is analyzed by direct use of a calibration curve, which
is inevitably nonlinear. Bicolor system analysis, for example, by measure-
ments at 2 wave length maxima, has also been effectively used ( ~ 15-80)
with suitable equations or nomographs. 9
ALTERNATIVE PROCEDURES

17-42. Nessler tubes may be used to compare the intensity of the color
of the test sample to that of the standard. To do this, 20 ml of the standard
solution is placed in 1 Nessler tube, and then increments of the test solu-
tion are added to the second Nessler tube until the color densities appear
equal when viewed from top to bottom. The 2 tubes are held adjacent to
one another in one hand, and viewed toward a white background (creased
filter paper) by indirect outdoor light. The crease is oriented to divide the
center field of view of each tube. The volume of the test sample re-
quired is recorded and then the process is repeated 4 or 5 times, the read-
ings being averaged. Then:
, . ml standard
Cone test sample =-= - --- - - x Cone standard
ml test sample (av.)
(17-19)
17-43. When a blank reading of 100 cannot be obtained by the full
setting of rheostats (~I 17-32), it is possible to set the blank at 50 (or 80)
and to multiply the galvanometer readings by 2 (or 1.25) before applica-
tion of the optical density (L = 2 - log G) table (~ 17-36). Since with
the blank at 100, L = log ( l 00/G), with the blank at 50 (or 80),
50 100 80 100
log - = log ·---- or log -- (17-20)
G 2G G 1.25G
Thus, the 2G (or 1.25G) value for the test solution is entered in the optical
density (L-G) table as though it was the G value observed.

ABSORPTION SPECTRUM DETERMINATION

17-44. The absorption spectrum of the specimen is yielded by deter-
mination of the amount of radiation absorption by a solution or solid as a
function of the wave length of incident light (by narrow band increments)
9 Kitson, Anal. Chem., 22:664 (1950); Knudson et al., Ind. Eng. Chem., A.E.,
12:715 (1940); Kozelka and Klochesky, Ind. Eng. Chem., A.E., 13:484 (1941);
Venning et al.,1. Biol. Chem., 120:225 ( 1937).
ABSORPTION SPECTROPHOTOMETRY 447
- ------ -·· ---
over a wide range of wave lengths. The components (bonds, molecules,
ions) of the specimen are identified qualitatively by the positions of the
spectral absorption peaks, in the visible, ultraviolet, and infrared regions.
The quantity of each component is estimated by the relative height of the
characteristic absorption peaks ( ~ 17-19).
APPARATUS

17-45. Needed apparatus includes a prism or grating absorption spectro-

photometer equipped with absorption cells of transmittancy suitable for
the spectral regions to be investigated. Glass cells are used for the visible

Fig. 17-11. Beckman quartz prism absorption spectrophotometer, model DU. Per-
centage transmission and optical density scales read with galvanometer at null. (Photo
courtesy Arthur Thomas Co., Philadelphia, Pa., and Scientific Instruments Division,
Beckman Instruments, Inc., Fullerton, Calif.)

region, quartz cells for ultraviolet, and KBr or NaCl "sandwich" cells for
the infrared region. Automatically recording infrared absorption spectro-
photometers are manufactured by Baird (Cambridge, Mass.), Perkin-Elmer
(Norwalk, Conn.), and Beckman (Fullerton, Calif.) .
17-46. The Beckman10 DU absorption spectrophotometer (Fig. 17- 11)
is suitable for near ultraviolet, visible, and near infrared spectra up to 2000
mu, and is used extensively in soils laboratories. The usual 4-compartment
solution holder has a 1-cm optical length of Corex or quartz, but 2-, 5-, and
10-cm cells are available and are used, with an interchangeable cell hous-
ing, when needed to increase the sensitivity of colorimetric methods. This
instrument well illustrates the principles of the prism type instrument (Fig.
17-12). A 6-volt storage battery and 25-watt W-filament lamp are em-
ployed as the light source for absorption down to 320 mu, and a stabilized
110-volt A.C. hydrogen discharge lamp is used for ultraviolet absorption
below 320 mu. An image of the light source A is focused by the condensing
mirror B and diagonal mirror C, on the entrance slit at D. The entrance
slit is the lower of 2 slits placed vertically over each other. The light fall-
1ocary and Beckman, J. Opt. Soc. Am., 31 :682 (1941); variability treated by
Castor, Anal. Chem., 23: 1229 (1951).
448 ABSORPTION SPECTROPHOTOM ETRY

Fig. 17-12. Diagram of Bl!ckman spectrophotometer. (Courtesy Scientific Instru-

ments Division. Beckman Instruments, Inc., Fullerton, Calif.)

ing on the collimating mirror E is rendered parallel and reflected toward

the quartz prism F. The back surface of the prism is aluminized so that
light refracted at the first surface is reflected back through the prism, under-
going further refraction as it emerges from the prism (Littrow prism prin-
ciple). The collimating mirror E focuses the spectrum in the plane of the
exit (upper) slit at D. Variation of the prism position by means of the wave
length selector knob setting determines the wave length band which passes
out of the slit, through the absorption cell G to the sealed, desiccated
phototube compartment at H. One vacuum tube photocell (~ 17-5) is the
cesium oxide type (for 600 to 1000 mu) and the other is a blue-ultraviolet
sensitive type (for 200 to 625 mu). A sliding rod brings either phototube
into position and simultaneously switches the electrical connections. A
switch coupled to the phototube aperture shutter permits the dark current
of the phototube to be checked at any time without changing any control
setting or removing the cells. The 200- to 2000-mu wave length scale is 1
meter long, and light wave length bands as narrow as 0.1 to I mu can be
defined by the slits. As with glass prisms generally, the extent of angular
dispersion of light is a function of the wave length. Attachments are avail-
able for emission spectrophotometry (~I 18-12), the flame unit replacing
the light source at A.
17-47. The Coleman replica grating absorption spectrophotometers are
also widely used in soil analysis. The Universal model (Fig. 17-13) gives
light wave length maxima from 320 to 800 mu (Fig. 17-10). As is true
of spectral gratings generally, the extent of angular dispersion is the same
for all wave lengths. Cells of 0.5-, 1-, 2-, 3.5-, 4-, and 5-cm optical length
are available. A stabilized power source that operates on a 110-volt A.C.
supply is employed for energizing the lamp run at 8 volts. A photronic cell
(~ 17-5) is used for registration, on direct reading scales, of transmission
percentage or optical density (Fig. 17-14).
17-48. A rapid-scanning absorption spectrophotometer that projects the
ABSORPTION SPECTROPHOTOMETRY 449

Fig. 17-13. Coleman replica grating absorption spectrophotometer, Universal model.

(Photo courtesy Coleman Instruments, Inc., Maywood, Ill.)

absorption spectrogram on an oscilloscope (60 times per second) is of-

fered by the American Optical Co., Buffalo, New York. Permanent records
are made with an oscilloscope camera.

REAGENTS

17-49. Absorption spectra can be made with any standard colored solu-
tion such as KMn0 4 (~ 5-68) or K2Cr 20 7 •

PROCEDURE

17-50. The light absorption curve of KMn0 4 or K2 Cr 20 7 is prepared

either with a prism or grating instrument (or approximated with a series ot
light filters) to determine which incident light color gives the greatest sensi-
450 ABSORPTION SPECTROPHOTOM ETRY

a spectrum, con-
taining all colors
of visible light.
This spectrum is /
passed over a nar-
row slit and con-
trolled by
i
a knob and dial -._
marked to show
wavelength (color)
of the beam ...
which emerges from
the slit. This
monochromatic beam
~ I
passes through ~~~ c~1.1-.---~-­
sample, which a~ _
sorbs part of I
the light. I
The unatsorbed _..........:
(transmitted) I
light ...
falls on a /
photo-cell
which actuates .
a galvanometer
reading the amount of
light absorbed or
transmitted by the
sample.

Fig. 17-14. Operating principles of

the Coleman replica grating ahsorp-
tion spectrophotometer. (After Waco
Catalyst, 9: 10, 1952, Wilkins-Ander-
son Co., Chicago, Ill.)

tivity. For advance work, the infrared absorptionll of a specimen such as

clay may be determined.
17-51. The specimen is first checked for proper concentration range to
give at least 20 per cent but not much over 30 per cent transmission at
the greatest absorption maximum or transmission minimum. To do this
the procedure given in the following paragraphs is run through rapidly and
approximately. Then the procedure is repeated in detail after the instru-
ment sensitivity, slit width, and specimen concentration have been ad-
justed. The appropriate photocell and light source are employed for the
spectral region in which measurements are being made.
11 Hunt et al., Anal. Chem., 22: 1478 ( 1950).
ABSORP TION SPECTR OPHOT OMETR Y 451

17-52. The transmission percentage is observed at a wave length maxi-

mum near one end of the spectrum against the blank as 100 per cent
transmission. The wave length band is advanced by an increment and the
transmission percentage determined again, the blank being reset at 100
per cent transmission at this wave length setting. Increments are taken
throttgh the entire spectral region of interest. The size of increment take11
is chosen according to the refinement desired.
17-53. Interpretation of Results. Absorption spectra of solutions are frl!·
quently made to select a suitable wave length maximum for the determina-
tion of concentration (~ 17-19). These are most effectively recorded and
interpreted as a plot of transmission percentage or optical density against
wave length.
17-54. Characterization of solutions and solids is effected by the posi-
tions of the absorption maxima, and components are determined quanti-
tatively by the relative heights of the absorption maxima compared to those
of the standards. Separations of components previous to absorption meas-
urement may aid the characterization. In the characterization of organic
and mineral colloids, the interpretation is based on comparison of the ab-
sorption spectrophotometric pattern with that of standard colloidal sub-
stances. Methods for this type of analysis have many details yet to be
worked out, but enough research work has been carried out to show
12

the promise of the methods for distinguishing molecular and crystalline

species not fully differentiated by other techniques. For example, the ab-
sorption maximum for water is at a longer wave length than that for
hydroxyl, and these 2 can thus be distinguished in crystalline materials.

QUESTI ONS
I. What procedure is followed to determine the correct light wave length
band for a given colorimetric procedure?
2. State Beer's law.
3. Describe the absorption spectroph otometric procedure followed to deter-
mine whether Beer's law applies to a solution.
4. Does nonlinearity of the optical density-concentration graph necessarily
prove nonobeyance of Beer's law for a given solution under all circumstances?
5. Describe the procedure for measurem ent of a colored constituen t con-
centration, when it is initially known that Beer's law is not obeyed.
6. Given a ml of test solution that matched a known volume and concentra -
tion of standard solution in Nessler tubes, derive the equation for expressing the
data as optical density.
7. How is the absorption spectrum determined?
8. In what ways are absorption spectra useful?

12 Hunt et al., Anal Chem., 22: 1478 ( 1950); Adler et al., Preliminary Reports
Reference Clay Minerals, Am. Pet. Inst. Res. Proj. 49, No. 8 ( 1951); Keller and
Pickett, Am. J. Sci., 248:264 (1950); Gore, Anal. Chem., 23:7 (1951); Launer, Am.
Mineral, 37:764 ( 1952).
 18
Emission Spectrophotometry
. . . no doubt that this [method of analysis] is one of the most
powerful now available for investigating the natural universe.
-HARRISON, LORD, AND LooFBOURow1

18-l. Emission spectra have been observed by man from earliest times
through visual detection of colors in fire. Flame tests used in qualitative
analysis utilize easily excited atomic spectra. For example, the bright
yellow flame of sodium is familiar to the analyst. Use of a colored glass
to screen out the sodium flame color permits a search for the reddish flame
of potassium. Instruments have been developed for precise determinations
of many elements through their emission spectra excited in various ways.
18-2. Basic Principles. When atoms, ions, or their groupings are sub-
jected to some form of energy excitation, as in a flame or in an electric arc
or spark, an emission spectrum results. Energy is first absorbed (~ 17-2)
by electron shifts to positions more distant from the atomic nucleus. As the
electrons regain or partially regain their stable or reference state, the pre-
viously absorbed energy is re-emitted as electromagnetic radiations, the
wave lengths of which correspond to the quantity of energy involved in the
respective electron shifts, in accordance with quantum theory. Thus:
(18-1)

in which l:o,E is the change in energy level, between initial E 1 and final E 2
states, h is Planck's constant, and f is the wave frequency of light. One
electron shift, yielding a given quantity of energy, results in the produc-
tion of a given emission maximum or line in the spectrum. For each ele-
ment there is a tremendous number of possible shifts and corresponding
spectral maxima or lines. Band spectra generally originate from molecules

1 Practical Spectroscopy (Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1948), p. 1.

4S2
EMISSION SPECTROP HOTOMET RY 453

or incandescen t polyatomic gases and vapors that are cool enough not to
be totally dissociated. For example, band spectra are produced by 0- H,
H 2 , and C-H bonds.
18-3. When excitation of an element has occurred, the subsequent emis-
sion of energy does not necessarily have to occur by shift directly to the
stable state, but may first occur by shift to some intermediate level fol-
lowed by a secondary shift to the stable state, giving 2 emission maxima.
The greater the level of excitation, the greater the number of electron
shifts and energy levels involved, and the more complex the spectrum pro-
duced. If the excitation is at an extremely high level, the emission becomes
a continuous spectrum, because the number of emission phenomena is so
great as to leave no finite gaps.
18-4. Emission spectrophoto metry or emission spectroscopy~ ( ~! 17-3)
is chemical analysis by measuremen t of spectral line intensity. Emission
spectra may best be visualized as a plot of intensity of radiation as a func-
tion of wave length. The spectrum useful for analysis extends from 2000 to
I 00,000 A (Fig. 18-1). Instruments used for emission spectrophoto metric
analysis consist of 3 discrete component systems: (a) sample excitation
system (the source), (b) radiation dispersion system (prism, grating),
and ( c) radiation measuremen t system (photograph y, photocell). The con-
ventional spectrograph employs high temperature excitation in an electric
arc or spark and records a spectrum through a wide range of wave lengths
on a photographic plate or film. The flame emission spectrophoto meter em-
ploys a relatively low temperature excitation and measures with a photo-
cell the emission intensity in a selected wave length range, correspondin g
to a given element.
18-5. Emission spectrophoto metry is an absolute qualitative method, in
as much as each particular element has its own spectrum, specific for that
atomic species. As a qualitative technique it is far more specific than most
precipitation tests for a single ionic species. Emission spectrophoto metry
is a comparative or empirical quantitative method, the quantity of emission
being compared, for a given instrumental set-up, with a known quantity
of the element to be determined. It is thus analogous to colorimetric
methods of absorption spectrophoto metry (Chapter 17).
18-6. One of the advantages of emission spectrophoto metric analysis lies
in the economy of time when large numbers of soil or plant analyses are to
be made on a routine basis. (If only a few or one determination is to be
made, the conventional chemical methods are faster.) The sensitivity of the

2 Brode, Chemical Spt!ctroscopy, 2nd ed. (New York: John Wiley & Sons, Inc.,
1947); Sawyer, Experimental Spectroscopy (Englewood Cliffs. N .J.: Prentice-Hall,
Inc., 1946); Candler, in Practical Spectroscopy (London: Hilger & Watts, 1949);
Thompson, A Course of Chemical Spectroscopy (Oxford: Clarendon Press, 1938);
Gerlach and Schweitzer, Foundations and Methods of Chemical Analysis by the Emis-
sion Spectrum (London: Hilger & Watts, Ltd., 1929).
Frequency Wave Wavelength
in sec- 1 number Characteristics of radiation cm µ. AU
incm-1

T----T--__l_ _ _
Natural Laboratory Prism Detectors
origin sources materials
10 7 10 1

1017 _____ ._
Inner High

1016 _____ _

iois_ ____ _
Outer
electrons
Arcand
spark;
CaF 2
LiF
,I
in
atoms
flame; t
and d1s~~:rge Quajrtz

lPhTI
molecules Glass

1014 _____ _

--Rock_____ -
salt KBr
J03
Molecular
t i
Vibrations
iota _____ _

10 2
Thermal j
- - - raciiatiO;;-- -only- -- -- - -
Radiometer
10 6
gratings
1012 _____ _ are suitable
Molecular
rotations

10

10 11 - - - - - -
__J___ _
I Microwave
----r--
Microwave
I generators receivers
I
t
I i
Fig. 18-1. Emission spectral regions and instrumental conditions for excitation
and detection. (From Harrison et al., Practical Spectroscopy, Englewood Cliffs,
N.J.: Prentice-Hall, Inc., 1948, p. 3.)

'454
EMISSION SPECTR OPHOTO METRY 455

spectral method to both macro and micro amounts of 60 to 70 elements

makes it ideally suited to refined and comprehensive analysis of soils in re-
lation to the growing list of soil elements that are important to the growth
of plants.
DETERM INATIO NS WITH FLAME EMISSION
SPECTR OPHOT OMETE R
18-7. In 1910, Klemperer8 began flame emission spectrophotometry by
matching oxyacetylene flame colors produced when various solutions were
sprayed into the flame, a split field spectroscopic eye-piece being used for
comparison of standard and test solutions. The method was made more
quantitative by the combination of the flame excitation with the prism
system for the light dispersion, photographic recording, and spectral line
intensity determination by a photocell densitometer. 4 Determinations of
K, Na, Ca, Fe, Mn, Sr, Tl, Cu, Ag, and later many other elements (~] 18-
34) were effected.
18-8. Spraying of the sample solution into a flame is used in various
forms in modern commercial instruments. A wide variety of elements can
be determined (Table 18-1 ) . The elements K, Na, Ca, and Mg, extracted
from soil, rocks, runoff, plants, or residues can easily and quickly be de-
termined by a flame emission spectrophotometer. Elements such as Ba and
Cu derived from cation exchange capacity measurements can also be de-
termined. The appropriate concentration range, which varies with the ele-
ment and with the instrument, must be considered in a particular procedure
(~ 18-16).
18-9. Accuracy. Most manufacturers of flame emission spectropho-
tometers claim an accuracy of 1 to 3 per cent in the analysis of K, Na, Ca,
Mg, and other elements, in solutions of chemically pure salts in suitable
concentrations. Variations in spray and flame production such as those
caused by variation in solution viscosity or pressure can be diminished
greatly by the use of an internal standard (~] 18-33). Error arises also by
the nonreproducibility or incomplete isolation of the characteristic spectral
line of the element to be determined.
18-10. Interference in the determination of an element may be caused by
(a) the enhancement of background intensity by extraneous cations, or ( b)
decrease in cation emission intensity by certain anions through forma-
tion of high melting point or high boiling point compounds in the flame.
5

Such interferences cause some loss of accuracy with unpurified soil and

According to Mitchell, Spectrochim. Acta, 4:62 (1950).

3
Lundegard h, Die quantitative Spectra/analyse de Elemente (Jena: Fischer, Ed. 1,
4
1929; Ed. 2, 1934), Z. Phys., 66: 109 (1930); Leaf Analysis, tr. Mitchell (London:
Hilger & Watts, Ltd., 1951 ), p. 115.
5 Margoshes and Vallee, Anal. Chem., 28: 180 (1956).
 TABLE 18-1
Wave lengths and relative intensities of flame spectra, determined with the Beck.man
flame spectrophotometert
Wave Wave Wave
Ele- length, Relative* Ele- length, Relative* Ele- length. Relative*
ment mu
----a;------- 238.9- ------=-- intensity ment
c_ _,___ _P_d
__ _
mu
363.5
intensity
-----
1.5
----~-
ment
Gd
y
mu
461.4
intensity

Hg 253.6 0.15 Pb 364 0.1 464.4

Au 267.6 1.5 Pb 368.5 0.2 y 467.5
Pb 283.3 0.1 Rh 369.1 B 473 5
Mg 285.2 10 Rh 369.6 y 486.0
In 303.9 - Mg 370.8 IO B 495 10
Cu 324.8 50 Fe 372.0 7 Mn 5IO 10
In 325.6 - Ru 372.7 3 Ba 520 20
Cd 326.1 0.15 Fe 373.6 IO B 521 15
Sn 326.2 0.2 Tl 377.6 80 Tl 535.0
.,.. Cu 327.4 20 Mg 383 8 Mn 541 30
v. Ag 328.1 7 Fe 386.0 7 B 548 20
"' Na 330.2 7 Ga 403.3 Ba 550 20
Rh 332.3 - to.In 403.4 100 Ca 556 200
Sn 333 0.2 K 404.6 7 Mn 561 70
Ag 338.3 50 Pb 405.8 0.3 La 563
Rh 339.7 - In 410.2 Na 589.3 10,000
Ni 341.5 3 Ga 417.2 Ca 603.5 100
Rh 343.5 - Rb 420.2 3 Ca 626 300
B 345 0.3 Ca 422.7 IOO Ca 650 IOO
Ni 349.3 2 La 438.4 Li 670.8 2.000
Co 350.2 20 La 443.3 La 714
Ni 352.5 6 Rh 449.2 La 745
Co 352.7 20 Gd 451.4 K 767 2,000
Rh 352.8 - Sn 452.5 0.1 La 798
Cr 357.9 70 B 454 2 Ba 850 50
Rh 358.3 - Cs 455.5 0.7 Cs 852 1.000
Rh 359.6 - Sr 460.7 150 La 860
0 Relative intensity of an element at a given wave length is equal to 100 divided by the number of parts per n1illion of the element necessary to give a photo-
metric response equal to 0.5 per cent of the flame background.
t lnstrnction manual 193-B, Table 2, courtesy Beckman lnstn1ments Inc., Fullerton, Calif.
EMISSION SPECTROPHOTOMETRY 457

plant extracts, which is not noted with chemically pure solutions of the
element to be determined. They can best be eliminated by chemical separa-
tions since they are not susceptible to correction by the internal standard
method. In practice, a variability of 3 to 8 per cent ot some elements deter-
mined is expected and is usually acceptable in rapid analysis for fertility
diagndsis.

APPARATUS

18-11. A flame em1ss1on spectrophotometer and suitable flasks and

beakers to hold the sample are required.
18-12. The Beckman model DU (Fig. 18-2) is an efficient and popular

with
Fig. 18-2. Beckman quartz prism single-beam spectrophotometer equipped scales
flame emission source, model DU. Percentage transmission and optical density cour-
read these properties of the spectral line with the galvanometer at null. (Photo nts
tesy Arthur H . Thomas Co., Philadelphia, Pa., by permission of Scientific Instrume
Division, Beckman Instruments, Inc., Fullerton , Calif.)

single-beam flame emission spectrophotometer. The spectrophotometer por-

tion of this instrument is also used for absorption spectrophotometry (~
17-46) . A photomultiplier increases the sensitivity. A special "bucking
circuit"6 has been developed to give greatly increased sensitivity (0-2 ppm
of Mg and 0-5 ppm of Mn) . The flame attachment to this instrument con-
n.
fl E. C. Boycks and V. V. Meloche, Chemistry Department, University of Wisconsi
458 EMISSION SPECTROPHOTOMETRY
sists of 2 parts, the oxygen and fuel (acetylene or hydrogen) regulation
system, and the burner and sample introduction system. Coarse pressure
adjustments are made at the respective supply tanks and the finer adjust-
ments are made at the pressure regulation system of the flame attachment.
The burner and sample sprayer ("atomizer" ) functions are combined into
a single small assembly. The sample is introduced through a capillary tube,
the oxygen from a chamber concentric with the capillary, and the fuel from
another chamber concentric with both the capillary and the oxygen cham-
ber. The oxygen serves both for combustion of the fuel and for dispersion
of the sample solution through a venturi effect. The fuel, oxygen, and
sample spray come together and burn immediately above the burner.
Special cooling of the burner is not necessary. Constancy of sample and
standard solution characteristics and operating conditions are highly essen-
tial, as with any single beam instrument.
18-13. The Perkin-Elmer Model 52C can be operated as a double beam
or single beam flame emission spectrophotometer (Fig. 18- 3). In the
double beam type of operation, an internal standard element, commonly

Ffa, 18-3. Perkin-Elmer prism double-beam flame spectrophotometer, model S2C.

(Photo courtesy Perkin-Elmer Corp., Norwalk, Conn.)
EMISSION SPECTROPHO TOMETRY 459

lithium, is placed in both the standard and test samples (~ 18-33). Spectral
dispersion is by a prism system. The elements K, Na, and Ca, and, with
modifications, Mg can be determined. The sample is introduced from a
dropping funnel into a sprayer. The spray is fed through a chamber that
has a provision for elimination of any condensed liquid and finally into the
bottom of a Meker type burner.
18-14. MultichanneF flame emission spectrophotometers are feasible, as
used by Lundegardh ( ~ 18-7) or better, through combination with a bat-
tery of photocell recorders (~ 18-55) set to record several wave bands
simultaneously.

REAGENTS

18-15. A series of stock solutions containing, respectively, 1000 ppm of

K, Na, Ca and Mg are prepared as follows: for K, 1.907 gm of KCI is dis-
solved in 1 liter of water; for Na, 2.541 gm of NaCl is dissolved in 1 liter
of water; for Ca, 2.500 gm of clear calcite (CaCO;i) is dissolved in 10 ml
of N HCl, and the solution is boiled to expel C0 2 and then diluted with
water to 1 liter; and Mg, 1.000 gm of metallic Mg foil is dissolved in 10
ml of N HCl, and the solution is diluted with water to 1 liter. The blank,
containing all salts that will be present in the analytical determination, ex-
cept the element to be determined, is called the base electrolyte solution.
The base electrolyte solution must be exactly the same in the standard and
the test samples, to provide the same interferences and viscosity effects.
Dilutions are made to give 40 ppm of K, 40 ppm of Na, 50 ppm of Ca, and
60 ppm of Mg in 1 N NHpAc, 0.4 N HCl, or other base electrolyte solu-
tion. More dilute standard solutions are made if the bucking circuit (~
18-12) is employed to increase the sensitivity. A 10 to 20 ppm standard
is used for the 100 per cent transmission setting when the content of cation
in the test solution is less than this (for example, if K or Na is less than
2 per cent in the total elemental analysis). Serial dilutions with base elec-
trolyte solution are made of each of the standards. The appropriate kind
of extraction solution for soil or ash is suggested in individual procedures.
The use of Li as an internal standard, is considered below (~ 18-33).

PROCEDURE!!

18-16. An inexperienced person must always obtain instructions from

an experienced operator of the flame emission spectrophotometer. He
should also carefully study the bulletin for a particular instrument, fur-
nished by the manufacturer.
18-17. The regular (almost weekly) replacement of the desiccant in the

7 Vallee and Margoshes, Anal. Chem., 28: 175 (1956).

s Appreciation is extended to Dr. L. D. Swindale for assistance with the writing
of portiQQS of this procedure.
460 EMISSION SPECTROPHOTOMETRY
Beckman flame emission spectrophotometer is highly important. The instru-
ment is allowed to warm up (~ 17-29) for a period of 10 to 15 minutes
prior to putting it in operation. The selector switch is set to 0.1. It is im-
portant to select the correct photoelectric cell and load resistor (suggestions
in Table 18-2) and to allow them to warm up at the same time as !he rest
of the instrument. The determination of K9 or other elements requiring the
red-sensitive photoelectric cell (those elements whose major flame emis-
sion line or band occurs between 600 and 1000 mu, Table 18-1) is car-
ried out first because this cell takes a considerable time to stabilize, whereas
the blue-sensitive cell reaches its operative condition in a few seconds and
no delay is experienced when the changeover is made. During the warm-up
period, the various standard solutions and sample solutions are poured into
5-ml beakers and a suitable record sheet is drawn up. When a steady dark
current is obtained the instrument is ready for the analysis.
18-18. Acetylene or hydrogen fuel gas is employed, burned with tank
oxygen. Acetylene is cheaper and is completely satisfactory for Na, K, and
Ca determinations. For Mg, acetylene gives a bent calibration curve instead
of the desired straight line obtainable with hydrogen. Acetylene gives more
sensitivity at low Mg concentrations, and less sensitivity at high Mg con-
centrations than hydrogen. Also, acetylene gives a hotter flame and less in-
terference of Ca with Mg and of P with Ca.
18-19. The burner is lighted in accordance with the manufacturers in-
structions. The oxygen is first adjusted to about half the operating pressure
at the pressure regulator on the emission spectrophotometer attachment.
Following this, the fuel gas is turned up slightly, at which time the burner is
lighted from the bottom. The pressure of oxygen and of fuel are then ad-
justed to give a hot oxidizing flame. Optimum operating conditions vary
from burner to burner and some experimentation is necessary to obtain
proper settings for the lowest flame background with the highest setting of
the sensitivity control and the smallest slit width. A reducing flame and
deposition of carbon on the burner should be avoided.
18-20. The most concentrated standard is put into the instrument with
the proper wave length selected and the transmission dial set at 100. The
sensitivity control and the slit width are varied to balance the galvanometer
needle at null-point. Standards of intermediate concentrations are then put
into the burner and their transmission percentage is obtained. From these
readings the calibration curve is drawn. Finally, solutions derived from
soils or plants are run through the spraying procedure, the percentage trans-
mission is observed, and the concentration is determined by reference to
the calibration curve. The most concentrated standard is run through from

9 The photomultiplier attachment, used at full sensitivity, makes possible the

advantage of K determination with the blue-sensitive cell. Rich, Agron. lour.,
48:430 (1956).
EMISSION SPECTR OPHOTO METRY 461

TABLE 18-2

Condition s with Beckman flame emission spectropho tometer for determina tion
of several cations excited in oxygen-acetylene flame

K Na Ca Mg Ba Cu

Working range, ppm 0-100 0-50 0-200 0-200 200-500 5-50

Selector switch 0.1 0.1 0.1 0.1 0.1 0.1

Phototube color Red Blue Blue Blue Red Blue

sensitivity

Phototube load 3 2 2 2 3 2
resistor position

Photomult iplier Full Full Full Full

sensitivity or 3 or 3

Zero suppressio n

Wave length,* mu 770 590 554 285 873 325

Slit width, mm 0.1 o.oi 0.02 0.06 0.25 0.06

Oxygen pressure, 10 10 10 8 8 7
p.s.i.

7 7 7 8.5 8.5 8.5

Acetylene pressure,
p.s.i.
the
0 Selection for midrange of the line or band is made by dialing for the 1naximum response near
wave length specified.

time to time to insure that the operating conditions are constant. If a change
in operating conditions is indicated by a change in the transmission per-
centage of the standard solution, the sensitivity or slit width must be altered
slightly until the original operating conditions are restored. If the test ele-
ment concentration exceeds that of the most concentrated standard, the
test solution may be diluted with the base elecrolyte solution ( ~ 18-15).
Greatest accuracy is generally obtained if the test samples are kept in the
range of 20 to 100 per cent of the standard curve range.
18-21. Cation Exchange Capacity by the Flame Emission Method. The
flame emission technique is a rapid and sensitive method foi: the determina-
tion of the cation exchange capacity of soil or clay (~ 4-6). Measurement
with K, Na, Ca, Ba, Cu, Mn and other ions all have been adapted to the
flame emission determination. The use of Ca or K is recommended ( ~ 4-9).
The saving in time over titrimetric, colorimetric, or gravimetric determina-
tions is considerable. In each case, the cation exchange determination con-
sists ( ~ 4--16) of saturation of the exchange charges with 1 of the above
cations 1 remov~l of the excess soluble salt by leaching with alcohol, and
462 EMISSION SPECTROPHOTOMETRY
then displacement of the exchangeable cation by means of a solution of a
second saturating cation. Ammonium acetate is a satisfactory displacement
solution because flame emission determinations can be carried out .directly
in this solution ( ~ 18-16) . Other solutions can be used for displacement,
but interferences and viscosity effects may be considerable.
18-22. Low sensitivity of the instrument (Table 18-2) to Ba and inter-
ference by Ca in the Ba determination detracts from the exchange capacity
determinations with Ba. Sufficiently concentrated solutions can usually be
obtained by using appropriate soil-to-solution ratios. One procedure 10 with
Ba consisted of leaching 2.5 gm of soil on a 7-cm conical funnel consecu-
tively (in 10-ml increments) with 50 ml of 0.1 N HCl (to remove Ca in-
terference with the flame emission method, since 10 ppm of Ca responded
as 400 ppm of Ba), 50 ml of BaC12 solution buffered to pH 8 .1 with tri-
ethanolamine, 50 ml of BaC1 2 solution, 100 ml of distilled water, and 90
ml of neutral 1 N NH 4 0Ac. The NH 4 0Ac extract was caught in 100-ml
volumetric flask containing sufficient LiCl solution to make 25 ppm Li in
the final volume. The Ba was determined by means of the Perkin-Elmer
model 52C flame emission spectrophotometer.
18-23. Dr. L. E. DeMumbrum in this laboratory determined exchanged
Cu (Table 18-2) with the Beckman instrument (~ 18-16), obtaining a
precision of 1 per cent. When the hydrogen burner was used instead of
acetylene, concentrations below 5 ppm could be obtained with 1 per cent
precision.
18-24. Exchangeable K, Na, Ca, and Mg of Soils. The NH4 0Ac extract
of many soils contains low enough concentrations of interfering ions such as
phosphate, Al, and Fe to obviate the need for chemical separations. The
standards are made up in 1 N NH 4 0Ac as the base electrolyte and the ex-
changeable cations of soil are determined ( ~ 18-16) directly in the extract.
A higher soil : extractant ratio and smaller increments of extractant than
usual (~ 5-11) are generally employed in the NH4 0Ac leaching in order
to give sufficiently high concentrations of cations for direct excitation in the
solution without preconcentration. Preconcentration may also be readily ac-
complished. The exchangeable cations have also been determined 11 directly
in a 0.05 N HCl extract. Systems of chemical purification of the extract to
remove interfering ions have been described. 12
18-25. The effect of variability of surface tension on the feeding rate of
the capillary which supplies the fog to the burner (Perkin-Elmer flame
emission spectrophotometer operated without an internal standard) was

IO Pratt and Holowaychuk, S.S.S.A. Proc., 18:365 (1954), adapted to flame emis-
sion from Mehlich, Soil Sci., 66:429 (1948).
11 Rich, S.S.S.A. Proc., 16:51 (1952).
1 2 Fieldes et al., Soil Sci., 72:219 (1951); Toth and Prince, Soil Sci., 67:439 (1949).
EMISSION SPECTRO PHOTOM ETRY 463

overcome 18 by the use of an extraction solution for K and Na that was

0.2 N with respect to Mg(0Ac) 2 as well as 2 N with respect to NH 40Ac.
From 15 to 30 gm of soil was extracted with 50 ml of extraction solution
for 0.5 hour. Filtration was effected in a small Buchner funnel followed by
leaching with 3 successive 15-ml portions of the extraction solution. Fi-
nally' the suction flask was rinsed with 5 ml of the extraction solution, and
the solution was brought to 100-ml volume in a graduated cylinder. The
pH of this solution after passage through soils of pH values between 4.9
and 7.7 did not cause the solution pH to vary beyond 6.6 to 7.2 a pH range
within which the photomete r readings checked to within 1 per cent.
Elimination of the interference of large amounts of Ca with the Na de-
termination with a filter instrumen t was effected 14 by inclusion of 0.5 gm
of (NH 4 ) 2 C 2 0 4 in with the soil extracted. This interference does not occur
with a prism instrument.
18-26. Exchangeable K in Runoff Waters. By means of a graduated
beaker ( ~ l 0-64), 100 ml of well-mixed runoff suspension is measured out
and transferred to a 500-ml conical flask. Then 100 ml of extraction solu-
tion which is 4 N with respect to NH 4 0Ac and 0.4 N with respect to
Mg(0Ac) 2 is added. The mixture is shaken for 0.5 of an hour and 25 to
35 ml of solution is filtered into a 50-ml beaker. The exchangeable K is de-
termined ( ~ 18-16) with the flame emission spectrophotometer. A 0 to 5
ppm K standard is usually employed; occasionally a 0 to 10 ppm K stand-
ard is required. Then:
ppai = ppm K x 2 x 0.227 x correction factor (18-2)

in which ppai is the pounds of K per acre inch of runoff, and the correction
factor is given by equation 10-20. Also:

pp2ms = ___p_pai of I{ __ x 2 000 000 (18-3)

ppai of solids ' '

in which pp2ms refers to the parts of K per 2 million of solids in runoff.

18-27. Total K, Na, Ca, and Mg of Minerals, Clays, Soils, or Rocks.
The K, Na, Ca, and Mg are determined in the order listed, with the Beck-
man flame emission spectrophotometer. The sample is decomposed in HF
(~ 11-31 ), and ions which interfere are removed by the NH 4 0H separa-
tion ( ~ 11-40) . The resulting Solution B is treated to remove ammonium
salts (~ 11-43) and taken up in 0.4 N HCl for the determination by flame
emission ( ~ 18-16) . The HCl solution is advantageous for the determina-
tion of Mg, and is excellent for the other cations also. Phosphate and Al,

13 Attoe and Truog, S.S.S.A. Proc., 11 :221 (1947). Myers et al., S.S.S.A.
Proc.,
12: 127 ( 1948) destroyed the NH4 0Ac and determined the Na and Kin 0.1 N HN0 3 •
14 Seay et al., Soil Sci., 71: 83 (1951).
464 EMISSION SPECTROP HOTOMET RY

which interfere with Ca determination at 554 mu, are removed by the

NH 4 0H separation. A calibration curve is prepared for each element in
0.4 N HCl base electrolyte solution ( ~ 18-15). The concentration of each
element is expressed as mgm per 100 ml of solution (~ 11-43, eq. 11-1).
The 40 ppm standards for Kand Na are 4 mgm of Kor Na per 100 Il)l.
18-28. It is advisable to determine the cations as soon as possible after
they are brought into solution in order to keep contamination with Na
from glassware as low as possible. Blanks should be run with every set of
determinations.
18-29. No ions that interfere with K or Na determinations arc likely to
be present in Solution A (~i 11-45) derived from soils or clays. These 2
ions are therefore determined ( ~! 18-16) on Solution A when Ca and Mg
are determined by methods other than flame emission (~ 11-44).
18-30. For total K, Na, and Ca, a 1-gm mineral sample was decom-
posed1'' in 15 ml of HC10 4 and 10 ml of 47 per cent HF evaporated in a Pt
dish (2 or 3 evaporations were sometimes necessary). Treatment was given
with HCl0 1 alone at the end to expel F from CaF 2 . Then the K, Na, and Ca
were determined with a Perkin-Elmer model 52 instrument.
18-31. Total K of clays can also be determined by decomposition of the
sample by Na~CO:i fusion, take-up of the melt in HCI (~ 6-75), removal
of the bulk of silica and R~O:i with the NH4 0H separation, and determina-
tion by flame emission after dilution of the solution with equal volume of
4 N NH 4 0Ac-0.4 N Mg(OAc)~ to approximate the composition of the ex-
traction solution for exchangeable cations ( ~[ I 8-25).
18-32. K, Na, Ca, and Mg of Plants and Organic Residues. The K and
Na can be extracted directly from plant tissue with 2 N NH 4 0Ac-0.2 N
Mg(OAc)~ (~] 18-25) with the same recoveries of these 2 elements found
as by ashing the plant tissue. 111 Extraction· of the K and Na of organic
residues gives that organically bound and the exchangeable forms of K
and Na but appropriately excludes that in the extraneous mineral particles.
A 0.5-gm sample of ground and dried organic material is extracted for 1
hour with 100 ml of 2 N NH 4 0Ac-0.2 N Mg(0Ac) 2 solution, the sus-
pension is filtered, and the concentrations of K and Na are then determined
( ~ 18-16). The Ca and Mg of plant tissue or organic residues are rendered
soluble by oxidation of the organic matter (~ 12-25) dissolved in 0.4 N
HCl, and determined (~I 18-16). The Cu of plant tissue has been de-
termined similarly .17

ir. Knight <'t al., Anal. Chem., 23: 1704 (1951).

16Attoe, S.S.S.A. Proc., 12:131 (1948). Myers et al., S.S.S.A. Proc., 12:127
(1948), ashed plant tissue at 475° to 500°C and took up the residue in 0.1 N HCI
for Kand Na determination by flame emission.
17 Massey, Anal. Chem., 29:365 (1957).
 465
EMISSI ON SPECT ROPHO TOMET RY

ALTERNATIVE PROCEDURES

18-33. Use of Internal Standar d. Lithium is employed for an internal

standard for the determination of K, Na, Ca, and Mg with the Perkin-E lmer
model 52C flame emission spectrop hotomet er ( ~ 18-13). Two optical
systems pick up the emission simultaneously 1 having a wave length setting
for Li at 671 mu, the other for a critical wave length maximum for the
test element (for example, 589 mu for Na or 767 mu for K). The Li emis-
sion is balance d against that of the test element, thus measuring the ratio of
Li concent ration to that of the test element. Variations of emission intensity
caused by changes in solution acidity and viscosity, flame tempera ture, and
pressure changes are thus cancelled out. With K, from 25 to 200 ppm of
18

Li is employed for the 0 to 20 and 0 to 100 ppm K standard curve and 0 to

10 ppm Na standard curve. A stock solution containing 1000 ppm of Li
is prepare d by dissolution of 9.93 gm of oven-dry LiN0 3 per liter of dis-
tilled water. This stock is further diluted ( 1 to 5 for 200 ppm Li) in the
standards. With Ca, a 100 ppm Li concent ration has been employed for
19

0 to 480 ppm Ca standard curve. With Mg, a 25 ppm Li standard is em-

ployed20 with the 0 to 1000 ppm Mg standard curve.
18-34. The Lundeg ardh type (~ 18-7) of flame emission spectrograph
has been employed21 for the determination of a number of elements in
soils and plants. In a modification, 22 5 gm of soil is extracte d with 200 ml
of N NH 4 0Ac, leached through in the usual way (~ 5-11) to remove ex-
changeable cations. The leachate is made up to 250 ml, and a 50-ml ali-
quot is diluted to 250 ml (or more if the soil contains over 8 meq of Ca
per 100 gm) for Ca determination. The remaind er of the leachate is evap-
orated to dryness and the organic matter is removed by means of H 20 2
•

The solution is made up to 50 ml, and then all other cations are determi ned
on that solution. The following concentrations in the standard stock solu-
tion, expressed as moles per liter, were employed for exchangeable cation
determinations for soils: Ca, 0.25; Mg, 5.0; K, 5.0; Na, 10.0; Li, 0.25; Sr,
0.1 O; Mn, 0.25; and Fe+++, 5 .0. This stock solution is diluted for each
photogr aphic plate by factors of 1, 2, 5, 10, and 20.

QUALI TATIV E ANALY SIS WITH ARC OR SPARK EMISSI ON

SPECT ROGRA PH
18-35. Qualitative analysis by the emission technique may best be
thought of in two distinct categories: (a) search for a specific element, and

18 Toth, Soil Sci., 67:439 (1949).

19 Jones and Hoover, S.S.S.A. Proc., 14:96 (1950).
20Tucker and Smith, S.S.S.A. Proc., 16:252 (1952).
21 Mitchell, Comm. Bur. Soil Sci. Tech. Com. 44, Harpenden, Herts., England
(1948).
22 Ells and Marshall, S.S.S.A. Proc., 4: 131 ( 1940).
466 EMISSION SPECTROPHO TOMETRY
(b) identification of spectral lines originating from an unknown element.
Theoretically, all elements can be identified by emission spectra; however,
some of the nonmetals are difficult to excite because of high ionization po-
tentials. Related to qualitative analysis is the comparison of samples as to
possible identity of source, through comparison of the spectrographic pat-
terns in a comparator. This technique applies the "fingerprint" matching
approach; identification of the different spectral lines with elements is not
required, nor need the amounts be determined.

APPARATUS'

18-36. Needed apparatus includes a spectrograph (~ 18-3 7), equipment

for shaping the electrodes, an agate mortar for grinding the soil sample, a
master spectrogram, and a comparator-densitometer (~ 18-48). Equip-
ment may be obtained from: Applied Research Laboratories, Glendale,
Calif.; Jarrel-Ash Co., Boston, Mass.; Baird Associates, Cambridge, Mass.;
and Bausch & Lomb Optical Co., Rochester, N.Y.
18-37. Emission spectrophotometers that use a high energy source for
excitation are generally known as "spectrographs." A spectrograph consists
of 3 principal components, the A.C. or D.C. arc or A.C. spark excitation
source (several types may be combined in a "multisource" unit), grating
or prism for spectral dispersion (~ 18-38), and recording components
(usually photographic, but sometimes by photocell, ~ 18-55).
18-38. In the grating spectrograph, a concave grating, entrance slit, and
recording device (photographic film or photocell) are placed on the circum-
ference of the Rowland circle (Fig. 18-4). The diameter of the circle is

Fig. 18-4. Principle of the Rowland Circle and the optics of the grating spectro-
graph. R. the Rowland Circle; S, entrance slit; G, diffraction grating; P, principal
focus, white light; A, B, spectral distribution of images of slit, dispersed linearly
about the circumference of the circle. Left diagram, Paschen mounting in which
recorders are placed at A and B. Middle and right diagrams, Eagle mounting, in
which movement of the grating brings different portions of the spectrum into posi-
tion near the entrance slit. (From Better Analysis, No. 2, p. 3, Cambridge, Mass.:
Baird Associates, Inc., 1950.)
~
-i

spectrograph (center), popular for soil and

Fig. 18-5. Spectrograp hic laboratory equipped with the ARL 1.5-meter Paschen-mo unted grating
c recording is on 35-mm film, covering about 2500 A in one exposure and 16 exposures in one 38-cm length of film.
plant analysis. Photographi
comparator- densitomete r. (Photo courtesy Applied Research Laboratorie s, Glendale, Calif.)
Right, multisource excitation unit. Left, projection
468 EMISSION SPECTROPHO TOMETRY
the radius of curvature of the grating. Grating systems give an angular
dispersion that is approximately the same for all wave lengths, in contrast
to prism systems (~ 18-39) which give angular dispersion, the degree of
which is a function of the wave length. The grating type of instrument is
superior to most prism type instruments with respect to spectral range,
linear aperture, resolving power, relative dispersion, and the fact that the
second order is usable. The gratings effect dispersion of emitted radiation
of wave lengths of about 1800 to 21,000 A (one-half, one-third, etc. of
this for higher order spectra). Either the position of the photographic plate
(Paschen mounting, Fig. 18-5) or the position of the grating (Eagle

Fig. 111-6. Baird 3-meter grating spectrograph equipped with Eagle mounting
(stands against wall, needing no access to back) and photographic plate recording.
(Photo courtesy Baird Associates, Cambridge, Mass.)

mounting, Fig. 18-6) must be varied for different wave length ranges of
radiation. The greater the number of grating lines per cm, the greater the
linear dispersion.

Kind and Grating Dispersion, A per mm

size lines per cm 1st order 2nd order Mounting
ARL I.Sm 9600 6.9 3.5 Paschen
ARL2m 9600 5.2 3.4 Paschen
ARL2m 14400 2.6 1.7 l>aschen
Baird, 3m 5900 5.5 Eagle

The Paschen mounting provides for equal dispersion of a given wave length
interval regardless of its position in the ~pectrogram, thus facilitating com-
parison to a master film in the comparator-der.sitometer. In the Eagle
mounting, both the grating and the photographic plates are rotated slightly
EMISSION SPECTR OPHOTO METRY 469

so that they remain on the Rowland circle, and the focus remains sharp.
The Eagle mounting gives a more compact instrument (Fig. 18-4), but
involves a different linear dispersion for different grating positions.
18-39. In the prism spectrograph (Fig. 18-7) the spectral radiation is
disper~ed by the Littrow-type quartz prism (one-half of a 90° prism
with
the cut surface silvered, Fig. 18-8). The optical system will be recognized
as being of the same general arrangem ent as the quartz spectroph otometer
(Fig. 17-12), reemphasizing the parallelism between absorption and emis-
sion spectrophotometry. Normally, 3 photographic plates, each 10 inches
long and of different emulsion characteristics, are used to cover the spectral

Fig. 18-7. Bausch & Lomb, Littrow prism spectrograph and illuminating unit. The
usable spectrum, extending from 2100 to 8000 A with full resolving power, is recorded
photographically. (Photo courtesy Bausch & Lomb Optical Co., Rochester , N.Y.)

Photographic plate

Fig. 18-8. Optical system of Bausch & Lomb, Littrow prism

spectrograph. (Photo courtesy Bausch & Lomb Optical Co.,
Rochester, N.Y.)
470 EMISSION SPECTROPHO TOMETRY
range. Glass prisms can also be designed for use at higher wave lengths at
some sacrifice of shorter wave lengths. Prism instruments tend to be better
than grating instruments in freedom from astigmatism and spurious lines.
Prisms also give a higher percentage collection of the total emitted radiation.
18-40. A simple and inexpensive emission spectral instrument is 9ffered
in the Spectranal (Todd Scientific Co., Springfield, Pa.). Metals are ex-
cited in either solid or solution form and their lines are inspected visually
in a Bausch & Lomb prism spectroscope.

REAGENTS AND SUPPLIES

18-41. Needed reagents and supplies include electrodes (Fig. 18-9),

Scale
One inch

A B c D E

Fig. 18-9. Graphite electrodes for arc excitation of test elements in samples.
A, upper. DC negative; B. lower, for iron spectra; C. sample crater, empty; D, crater
for sulfide precipitates; E, sample crater filled, with excess graphite filed off. (From
Vanselow and Liebig, Spectrochemical Methods, Berkeley, Calif.: University of Cali-
fornia, 1948, p. 39.)

usually graphite rods of high purity to avoid or minimize the introduction

of extraneous lines; spectrographic grade 35-mm photographic film or glass
plates; and photographic processing chemicals. Cloth gauze is employed for
sieving the sample to avoid contamination with metals.

PROCEDURE

18-42. The aim in the qualitative procedure is to detect as many ele-

ments as possible in the sample. Some difficulty is encountered in detecting
a representative portion of both the more volatile elements in the first part
of the exposure and the less volatile elements that tend to delay volatiliza-
tion until the last part of the burn. Volatility of the sample is an especially
important consideration in the qualitative analysis of soil or plant extracts.
Spectrograph "buffers," which control the rate of volatilization, are gen-
erally added. Thus Na 2S0 4 or NaCl may be added to a slowly volatile sili-
EMISSION SPECTROPHOTOMETRY 47l
of
cate to speed up volatilization, or Si0 2 to the chloride or acetate extract
zation. Also, 3 or more separat e
cations to slow down the rate of volatili
differen t arcing time to rep-
spectrograms of the sample may be taken with
ation, a
resent elements of various volatilities in the sample. In one modific
renews the sample for high
rotating disc of carbon or silver continuously
23
flame does for low tem-
temperature excitation much as spraying into the
perature excitation ( ~ 18-7).
dried
18-43. Mounting the Sample. The soil sample, after having been
salt in some
and ground to pass a fine bolting cloth (ground with buffer
cratere d
cases), is placed in a high purity carbon electrode which has been
24

in a
to a sufficient depth (Fig. 18-9, C). A liquid sample is evaporated
solution
dish and powdered with a buffer, then mounted. Small amounts of
a crater directly . The crater contain ing the sample
may be evaporated in
or 20 second s in a small gas blowpip e flame at be-
may be heated for 15
e if its presenc e proves objectio nable. Pow-
low red heat to remove moistur
soil, rock, glass, paint, and minera ls have
dered samples of plant tissue,
ively by direct placem ent in the carbon
been successfully analyzed qualitat
arc crater without buffer or other special arrangements.
be
18-44. Adjustment of Slit Width. The film background intensity must
e, so as not to mask any of the
controlled by adjustment of grating apertur
spectro-
lines of interest by over-exposure. With an ARL 1.5-m grating
to a minimu m (24 micron s)
graph, the slit width is adjusted at the start
g of 4 on a scale of 1 to 10.
and the grating apertures are set at an openin
(of over
A narrow slit width is somewhat more desirable than wider ones
spectru m lines may
60 microns) because with wider slit widths, prominent
at times overlap minor lines.
18-45. Arcing the Sample. A high current D.C. or A.C. arc is usually
of the
employed because it gives high sensitivity. Complete volatilization
the spec-
sample is easily achieved and all elements are thus represented in
6 mm).
trogram. The electrodes are spaced with the standard spacer ( 4 to
m is produc ed for referen ce purpose s (iron sample,
First the iron spectru
opened
Fig. 18-9, B), by a 5-second spark. Then the grating aperture is
is energiz ed by a 6-ampe re A.C. arc for 10 to 30
to I 0 and the first sample
Liebig suggest a 10-seco nd first exposu re fol-
seconds. Vanselow and 25

of the sample with the apertur e grating at

lowed by a second spectrogram
with apertur e grating set at 2 for 80
5 for 40 seconds, and a third run
e
seconds. Additional spectrograms are run at narrower grating apertur

23 Meloche and Shapiro, Anal. Chem .. 26:347 (1954).

re. Comme rcial
24 After preburn ing empty, for the same time as in the procedu
. Electrod es may be
electrod es can be obtained sufficiently pure for most purposes
rated HNOa and
purified by Soxlet extractio n with 20 per cent HCI and the concent
the procedu re of Vanselo w and Liebig, Spectrochem-
finally with water accordin g to
p. 5.
ical Methods (Berkeley: University of Californ ia, mimco., 1948),
25 Ibid., p. 11.
472 EMISSION SPECTROP HOTOMET RY
openings and longer exposure times if necessary to obtain complete vola-
tilization of the sample. Lastly, an iron reference spectrum is again re-
corded.
18-46. For identification work with soil, paint, rock, glass, and minerals,
ordinarily 1 arcing time can be selected as representative of a given material,
thus permitting a satisfactory test with 1 spectrogram per sample. This is
particularly an advantage when a large number of samples are to be checked
qualitatively.
18-47. The film or plate is processed according to standard procedure,
usually with a 3-minute development, 1-minute washing, and 2-minute fix-
ing time, followed by washing in tap water, rinsing in distilled water, spong-
ing off with fine textured synthetic sponges, and a 2-minute forced hot-air
drying.
18-48. Interpretation of Results. The photographic film or plate bearing
the spectrogram is mounted in a comparator (Fig. 18-10) and the spectral
lines are projected in juxtaposition with those of a master spectrogram.
After 1 or 2 lines for some known element in the sample have been lo-
cated on the sample spectrogram (for example, the iron triplet at about
3100 A), it is orientated and locked with lines adjacent to the same lines
on the master spectrogram. Unknown lines then can be identified.
18-49. The standard reference in which all known lines are listed is the
MIT wave length tables 26 which give 100,000 lines from 2000 to 100,000
A. Most elements are readily identified by their persistent lines (Raies
Ultimes or letzten Linien). These lines are observable if an element is pres-
ent in any amount. If the persistent lines are absent for some element, the
assumption can be made that the element is not present. If the presence of
some element is in question, the problem is simply one of searching for
the persistent lines. To identify each of various lines, the wave length is
measured and referred to tables of elements arranged by wave lengths 27 for
a list of possible elements represented. The more probable ones are then
looked up in the second set of tables, which gives all emission wave lengths
for each element. A series of lines in the spectrum corresponding to a given
element serves to identify it positively. The identification of all of the sig-
nificant lines present in the spectrogram is usually possible in this way.
18-50. If a grating spectrograph and Paschen mounting (Fig. 18-4) are
employed, the spectrogram is easily projected with the test sample lines in
juxtaposition to the standard spectrogram because the dispersion of the
lines is always the same for a given wave length. If a prism instrument or
Eagle-mounted grating instrument is employed, the dispersion must be

Harrison, MIT Wavelength Tables (New York: John Wiley & Sons, Inc., 1950).
26
Given in Harrison et al., op. cit.; Brode, op. cit.; and Ahrens, Wavelength Tables
27
of Sensitive Lines (Cambridge. Mass.: Addison-Wesley Publishing Company, Inc.,
1951).
MISSION SPECTROPHOTOMETRY

The spectro gram of the

Fig. 18-10. ARL projection compa rator-d ensitom eter.the lines projected in juxta-
set to track with a master spectro gram and
test sample is
used to measure light -absorp-
positio n for qualitative work. The densitometer can be
work. (Photo courtesy Applied Resear ch
tion of line and backgr ound for quantit ative
Labora tories, Glenda le, Calif.)

the same disper-

calculated or a standard spectrogram employed that has
rement of the
sion of the spectral lines as the test sample. Exact measu
ic spectrogram
position of the spectral lines with any type of photograph
scope, suitably
may also be made by means of a mechanical traverse micro
calibrated.
is of soil
18-51 . Soil Analysis in Crime Detection. The detailed analys
ion. The soil chemi st is called
often becomes of interest in crime detect
soil at the scene with that
upon for interpretation of possible identity of
474 EMISSION SPECTROPHOTOMETRY
found on clothing of a suspect. Spectrographic methods provide a method
for the analysis of as many as 70 elements on a few mgm of soil sample.
General similarity of soil samples is not sufficient for proof of identity, how-
ever, because many soils are more or less alike in composition. The soil
chemist can do service to justice in pointing out this relationship, to pre-
vent the mistaken claim to identity of similar samples. The presedce of
similar amounts of the same elements in both samples presents evidence
of probable identity; discrepancies between the two samples present evi-
dence that identity is unlikely.

QUANTITA TIVE ANALYSIS WITH THE ARC OR SPARK

EMISSION SPECTROG RAPH
18-52. In devoting one complete issue of Soil Science to the arc and
spark emission spectrograph, editor Bear28 states ". . . the emission spec-
trograph merits much more consideration than it has yet received from
biological scientists." Many soils laboratories now employ the technique.
The objective here is to outline some of the broad principles of quantita-
tive emission analysis. Because of the importance of routine application for
gaining economy with the emission spectrochemical method, the use of
cooperative facilities for the work of several departments and agencies is
important. For organizations of size insufficient to develop such a labora-
tory, the use of commercial service laboratory 2 n facilities helps.
18-53. Quantitative emission spectrochemical analysis has several ad-
vantages:w over the wet chemical methods. The size of sample may be
small, although this is usually a minor advantage. The identification of any
element determined is absolutely positive, an important consideration for
the minor elements. Furthermore, several minor elements can be deter-
mined simultaneously by fusion and extraction:ll or direct arcing of soil,
plant tissue, or fertilizer.:1 2 The determination of Zr, important in the
chemistry of soil development, can be accomplished.:i:i Use of the method
will undoubtedly lead to determination of elements not now considered in
the analysis of soils and plants. Another advantage is that chemical isola-
tion of a specific element is not required prior to analysis. It is necessary
only to preconcentrate the element in question to the extent necessary for
its determination. Sulfide and 8-hydroxyquinoline precipitations, which
bring down whole suites of elements, are highly satisfactory for the pre-
concentration of many trace elements, even though large amounts of more

28Bear, Soil Sci., 83: 1 (1957).

20One example, Chicago Spectro Service Laboratory, Inc., 2454 W. 38th St.,
Chicago 32, Ill.
30 Vanselow and Liebig~ op. cit., p. 28.
:n Wark, Anal. Chem., 26:203 (1954).
32 O'Connor, Ind. Eng. Chem., A.E., 13:597 (1941 ).
33 Horton, Anal. Chem., 25: 1331 (1953).
EMISSION SPECTR OPHOTO METRY 475

common elements come down also. In fact, macro quantities of an ele-

ment may be added to act as a carrier for recovery of some trace elements.
The use of solid samples has the advantage over the solution system, such
as is used with the flame photometer, that the entire ash or soil need not be
brought into solution. Experience indicates that bringing of the entire
ash of plants into solution, including all trace elements, is not as easily
accomplished as is commonly thought. Quantitative spectrochemical analy-
sis has been discussed at length. :H
18-54. Accuracy. The accuracy of a spectrographic determination de-
pends on: (a) instrument precision (capacity to give identical results un-
der identical operating conditions); (b) uniformity between standard and
test sample as to matrix composition, bonding of the elements present, and
degree to which the spectroscopic buffer (~ 18-42) is successful in its
function; ( c) amount of element available for analysis (the nearer to the
optimum quantity of test element, the greater the accuracy) ; and ( d) the
particular element to be determined, since some elements inherently give a
greater degree of accuracy than others. A report by Mathis:m shows that an
accuracy of 92 to 98 per cent is readily attained for a number of cations de-
termined with spark excitation on solutions, but that accuracy with boron
is only about 85 to 90 per cent with this method. With a D.C. arc, an ac-
curacy of 85 to 96 per cent was obtained with plant ash and with solutions.
The Quantometer ( ~ 18-55), in eliminating the photographic problems,
enhances precision by about one-third over the limits found with photo-
graphic methods. With the best available refinements of the procedure,
including appropriate matrix for standards, spectroscopic buffers, precon-
centration sufficient to bring the concentration to the optimum, using
elements that lend themselves to greatest accuracy, and with mathematical
treatment of the data, the error in the high temperature excitation methods
is held to about 3 per cent. In practice, keeping the error within 10 per cent
is considered satisfactory.

APPARATUS
18-55. Needed apparatus includes an arc or spark emission spectrograph
of grating or prism type (~ 18-37) with a recorder. Equipment for shap-
ing the electrodes is needed unless preshapoo electrodes are purchased. For
photographic recorders, processing accessories and a comparator-densi-
tometer (~ 18-48) are needed. Automatic photocell and chart recorder
equipment (Fig. 18-11 ) are highly efficient for routine work. Photomulti-

34 Ahrens, Spectrochemical Analysis (Cambridge, Mass.: Addison-Wesley Pub-

lishing Company, Inc., 1950), pp. 76-120; Mitchell, Comm. Bur. Soil Sci. Tech.
Com. 44, pp. 11-44 (1948); Harvey, Spectrochemical Procedures (Glendale, Calif.:
Applied Research Laboratori es, 1950), pp. 184-293.
ar.1.A.O.A.C., 34:604 (1951).
 Fig. 18-11. ARL Quantometer. Excitation unit at right, grating spectrograph in center, and detector unit at left. Photocells ~re aligned for direct
and simultaneous measurement and recording of intensity of critical spectral lines. (Photo courtesy Applied Research Laboratories, Glendale,
Calif.)
EMISSION SPECTR OPHOTO METRY 477
plier cells are situated in adjustable positions, each placed to intercept the
radiation at wave length maxima specific for an element to be determined.
The intensities of spectral lines are graphed automatically. Such an instru-
ment may be compared to a flame emission spectrophotometer ( ~ 18-7) in
mode of operation except that it employs the higher temperature of exci-
tation. of the conventional spectrograph. The usual photographic film ex-
posure, its processing, and the densitometric measurement of intensity are
thus supplanted.

REAGENTS AND SUPPLIES

18-56. Needed reagents and supplies include electrodes, usually graphite

rods or discs; spectrographic grade 35-mm photographic film or glass
plates, according to the design of the instrument; chemicals for internal
standards; spectrographic buffers; and reagents for preconcentration of
sample constituents. Copper, aluminum, and silver rods or discs are other
possible electrode materials. The electrodes should be of highest purity to
avoid or minimize the introduction of extraneous lines. A purification pro-
cedure was suggested in connection with qualitative emission spectrochemi-
cal analysis ( ~ 18-43) . One procedure ( ~ 18-5 8) calls for coating the
graphite electrode craters with DuPont methacrylate clear solution RC-901
diluted 5-fold with acetone, to prevent soaking-in of the sample solution.

PROCEDURE

18-57. The quantitative em1ss1on spectrochemical procedure involves

the following steps: (a) qualitative exploration ( ~ 18-42) for establish-
ment of instrumental conditions suitable for the elements to be deter-
mined; (b) design of convenient and effective form of sample, buffer,
matrix, crater shape or disc-wetting conditions; (c) preconcentration of
elements, as necessary; ( d) calibration of photographic emulsion response
characteristic if a Quantometer is not available; ( e) choice of internal
standards; and (f) preparation of standard curves for the elements to be
determined.
18-58. Fonn of Sample. The sample may be of ground soil or the whole-
ash of plants, and thus be mounted in solid form. Plant tissue in pellet
form has been employed. Fo_r the total elemental analysis of soil, Mitchenas
air-dried the soil, ground it to pass 0.16 mm openings ( 100 meshes per
inch), ignited it at 450°C to destroy the organic matter, and placed it in
the electrode craters ready for arcing. Solutions resulting from extractions
may be dried in an evaporating dish and mixed with the buffer (frequently
Na2S04 ) and internal standard by means of an agate pestle. 37 The plant

86 Comm. Bur. Soil Sci., Tech. Com. 44 (1948).

37 Vanselow and Liebig, op. cit.
478 EMISSION SPECTROPHOTOMETRY
ash solution has been evaporated in the crater filled with graphite powder38
resulting from the crater drilling. The solution may be mixed with the
buffer and internal standard in solution, 89 which gives thorough mixing,
and then evaporated in the graphite electrode crater which has previously
been sealed with methacrylate. For quantitative determination of cobalt, a
9-ampere D.C. arc with narrow electrodes and exposure until 1. to 2
seconds after the green color of iron internal standard spectrum disappears
has been employed successfully. Alternatively, the solution may be used
directly for excitation by means of a rotating graphite, copper, or silver
disc ( ~ 18-42), one side of which dips in the solution while simultaneously
the other 'side is sparked with the upper electrode. In the latter procedure,
a wetting agent is often added to facilitate uniform adherence of the solu-
tion to the electrode. Proper adjustment of the cycle must be made to
avoid excessive evaporation of the solution by over-heating of the disc.
18-59. Preconcentration of the Test Element. Increasing the concentra-
tion of an element per unit weight of the matrix in which it is to be excited
is termed preconcentration (~ 18-53). One or more elements present in
macro amounts are removed to increase the relative concentration of the
test element. It is frequently necessary to preconcentrate some of the minor
elements before spectrographic determination. A detailed procedure for
preconcentration of soil cobalt with dithizone has been outlined by Carri-
gan and Erwin. 40 It illustrates in some detail the principles of preconcen-
tration of a minor element. For total soil cobalt, l to 5 gm of soil sample,
previously ignited for 5 to 6 hours at 450°C to destroy organic matter and
ground to pass fine bolting cloth, is decomposed with 20 ml of 50 per cent
HF and 4 ml of 72 per cent HCI0 4 and the filtered acid insoluble residue
is decomposed by a Na 2 CO:i fusion and HCl extraction. The resulting acid
solutions each are extracted in the presence of 5 ml of 10 per cent sodium
citrate buffer and 0.5 ml excess of I : 3 NH 4 0H with 10 ml of 0.05 per
cent dithizone in CC1 4 and then with three 5-ml portions of 0.01 per cent
dithizone. To the combined dithizone extracts are added 2 ml of an Fe
internal standard solution and the whole is evaporated to dryness in a 50-
ml beaker. The residue is treated with 0.75 ml of concentrated H 2 S04 and
6 drops of 72 per cent HCI0 4 , and the mixture is digested on a hot plate
until the organic matter is destroyed. The H 2 S04 is then fumed off by a
flame playing across the beaker. The residue is dissolved, with heating, in
2 ml of 1 : 20 HN0:1 containing 2.5 mgm of Na 2 S04 per ml. Then the solu-
tion is evaporated to 1 ml and transferred to a glass cone with a capillary
tipped pipet (no washing is required because of the internal standard)
for further evaporation to 0.2 ml, an aluminum heating block being em-

as Mathis, Anal. Chem., 25:943 (1953).

39 Carrigan and Erwin, S.S.S.A. Proc., 15: 145 ( 1951).
40 S.S.S.A. Proc., 15: 145 (1951).
EMISSION SPECT ROPHO TOMET RY 479

ployed. Finally, drops of this solution are transferred to the plastic-sealed

electrode craters and evaporated to dryness. Cobalt extracts of soil with N
and 6 N HCl, and with HOAc and NH 4 0Ac were similarly preconcentrated.
18-60. The dilute acid extractable trace elements Co, Ni, Mo, Cu, and
Zn were preconcentrated by Mitchell 41 by 8-hydroxyquinoline. To 20 gm
of soil.is added 800 ml of 2.5 per cent HOAc (0.5 N, pH 2.5). The suspen-
sion is shaken over night and filtered. The cake is washed with distilled
water and the combined filtrates are then evaporated to dryness. The resi-
due is treated with H 20 2 to remove organic matter, taken up in 50 ml of
4 N HCI and filtered through 9-cm Whatman No. 41 filter paper. Enough
iron and aluminum in HCl solution are added to give the equivalent of 2
to 5 mgm of Fe 20:1 and 30 mgm of Al 20 3 total considering the amounts
already present, and the solution is diluted to 150 ml. Then 15 ml of a 5
per cent solution of 8-hydroxyquinoline in 2 N HOAc is added, followed
by 7 N NH 4 0H dropwise until the color changes from yellow to emerald
green at pH 1.8 to 1.9. Finally 50 ml of 2 N NHpH is added, and the
solution is stirred and allowed to stand over night at room temperature.
The precipitate is filtered on a 9-cm Whatman No. 540 filter paper and
washed with cold water. After it is partially dried, the precipitate is ignited
with the paper in a muffle furnace at 450°C. The precipitate is weighed,
and ordinarily amounts to 30 to 50 mgm. If below 30 mgm, it is made up to
40 mgm by the addition of pure Al 2 0 3 powder. The ash is thoroughly ground
in a small agate mortar, and is then ready for emission spectrochemical
analysis of the trace elements and the colorimetric determination of iron
for application of the variable internal standard method. Recovery of
42

Pb, Sn, Cr, V, and Cd is incomplete by the 8-hydroxyquinoline procedure,

but Mitchell and Scott 43 found that the addition of tannic acid and thional-
ide with the 8-hydroxyquinoline gave complete recovery of Co, Ni, Mo,
Cr, V, Bi, Ge, Sn, Pb, Ti, Zn, Cd, and probably Ga, Th, and Ag. Addition
of CdCl 2 was substituted for the addition of iron and thus the Zn deter-
mination made easier, since Fe interferes with the Zn determination. The
Cd serves as an internal standard for Zn.
18-61. Calibration of Photographic Emulsion Response Curve. One of
the greatest sources of error in the quantitative spectrochemical analysis
arises from difficulties inherent in the photographic process. Use of the
Quantometer (~ 18-55) eliminates this source of error. In order to obtain
reproducible photographic results, factors such as the age of the film, the
age of the developer, the developing temperature, and the agitation during
development must be rigidly controlled. Kodak spectrum analysis No. 1
film is usually used for quantitative analysis. The film is processed as fol-

41Comm. Bur. Soil Sci. Tech. Com. 44 (I 948).

42Davidson and Mitchell ,/. Soc. Chem. Ind., 59:213 (1940).
•a/. Soc. Chem. Ind., 66:330 ( 1947).
480 EMISSION SPECTROPHOTOMETRY
lows: 3 minutes for development in D-19 (Eastman-Kodak), 10 seconds of
acid wash, 1.5 minutes in fixer F-5 (Eastman), and 2 minutes of drying in
an infrared dryer.
18-62. The photographic darkening in a spectral line is read with a
densitometer. White light is suitable for measurement of the black emulsion
darkening, which consists of metallic silver grains. Optical density,. D, is
related ( ~ 17-13), to light transmission by:
1 100
(18-4)
D = log 10 T = log 10 % transmission

in which T is the ratio of transmitted to incident light intensity. The inci-

dent light intensity is, for convenience, taken as the intensity of light pass-
ing through the background or relatively clear portions of the film near the
lines to be measured. The densitometer readings, D, are linearly related to
concentration for low density lines (D below 1.25) and are related to the
log of concentration for high density lines.
18-63. The first operation in quantitative work is the calibration of the
photographic emulsion in the region of the spectrum used. A proper and
accurate emulsion calibration is important. In the Churchill44 2-step filter
method of calibration of photographic film, a quartz optical filter, one-half
of which is clear and the other half aluminized to decrease the amount of
light passing through it, is placed in front of the slit of the spectrograph.
The light transmission percentage value of the aluminized portion relative
to the clear portions is calibrated (the calibration is ordinarily furnished
with the filter) as a function of wave length. To do this, an iron spectro-
gram is made, a portion of the light being passed through each portion of
the 2-step filter. The 2 percentage transmissions for each pair of 30 iron
lines in the desired spectral region ( 100 A band width) are measured by a
densitometer and then plotted, one against the other, on linear graph
paper. This is known as the preliminary emulsion calibration curve. The
wave lengths of the analysis and the internal standard lines are usually
chosen to fall within the linear portions of the emulsion response curve
(2600 to 3300 A for spectrum analysis film No. 1). Churchill 45 gives direc-
tions for conversion of the preliminary to the final calibration curve and
summarizes graphical calibration and interpretation of emission spectra.
18-64. Internal Standards. The use of internal standards in spectro-
graphic analysis, pioneered by Gerlach, 46 has obviated many of the diffi-
culties earlier experienced with attempts to compare directly the line densi-

44 In Boltz, Modern Instrumental Analysis (Englewood Cliffs, N.J.: Prentice-Hall,

Inc., 1952), p. 201. Further discussions of the calibration of photographic emulsion
are given by Nachtrieb, Spectrochemical Analysis (New York: McGraw-Hill Book
Company, Inc., 1950), pp. 102-141; and Sawyer, Experimental Spectroscopy (Engli:--
wood Cliffs, N.J.: Prentice-Hall, Inc., 1946), 192-214, 254-257.
45 Op. cit.
46 z.anorg. Chem., 142:383 (1925),
EMISSION SPECTROP HOTOMET RY 481

ties of standard and test spectrograms. Even though the conditions of ex-
citation have been maintained as closely as possible alike, great differences
in intensity occur for identical samples. The internal standard is used by
placing a definite concentration of a selected element in with the test
sample and exciting it simultaneously with the test elements. The intensity
of 1 or more lines the standard emits is compared with the lines of the test
element. Any variations in excitation conditions are then represented
equally for both standard and test lines.
18-65. For the internal standard, an element is chosen that emits a
spectral line with approximately the same excitation energy as the test
element, and is located at a spectral position not too far distant from that
of the test element. The internal standard line of course must not fall on or
too near lines of other elements that may be present in the sample, and
must be of an element that does not occur in appreciable or unpredictable
amounts in the test sample. The internal standard line must not exhibit
self absorption.
18-66. A good preliminary choice of an internal standard can usually be
made in the basis of its similarity to the analysis element with respect to
atomic weight, electron configuration and periodic group. Ideally, the
members of the homologous pair, as the internal standard and the analysis
element are often called, will excite similarly under differing environmental
conditions. They are unaffected by enhancement (intensity increase of one
member independent of the other) or degradation (loss of intensity of one
member independent of the other) caused either by changes in the condi-
tions of excitation or in the matrix composition. In practice, this situation
is never completely attained. In the choice of an internal standard, it is first
desirable to match the internal standard, and the analysis element as re-
gards a similar rate of volatilization (Fig. 18-12).
18-67. Because of the minute quantities of internal standard required
per sample for minor element work, the standard substances have been
diluted 47 with Na 2S0 4 buffer salt in large enough quantities to permit
weighing of each component, and the resultant powders are thoroughly
mixed by grinding in an agate mortar, drying at 400°C, and regrinding. One
part of the Na 2 S04 mixture to 4 parts of solid sample has been employed.
The internal standards and buffer may be added as solutions, when the
sample is in solution, and all evaporated together.
18-68. The internal standard elements found to be appropriate for a
number of plant constituents are listed in Table 18-3. As an internal stand-
ard, Mg from the Mg (NO a) 2 used in plant tissue ashing or Ge in plants of
high ash content has been employed 48 for the line width technique of spec-
trochemical analysis.

47 Vanselow and Liebig, op. cit., p. 18.

48 O'Connor and Heinzelman, Anal. Chem., 24:1667 (1952).
482 EMISSION SPECTROPHOTOMETRY

Time in minutes
0 2 3 4

Hg-
As-
Cd-
Rb-
Pb-
TJ-
Cs-
Zn-
Bi------
Na------
K
In--------
Ag--------
Ge------
Li------
Sb------
Ga--------
Sn-----------
Au--------------
Cu---------
Mo------------------~
p
Mn-----------
Mg----------------
Ba---------------
Sr --------·-
Ca -- - - - - - - - - - - - - -
Ni --
Fe
Co - --
Cr----- - - - - - - - - - - - - -
Pd
v
Si
Al
Be
Yb
Ti
Yt
Er
Pt
La
Ce
Th
Nd
Zr

Fig. 18-12. Rate of vaporization of various elements during

the arcing cycle, which provides a basis of selection of internal
standard and test sample homologous pairs. (From Vanselow
and Liebig. Spectroc/1e111ical Methods, Berkeley, Calif.: Univer-
sity of California, 1948, p. 41.)

18-69. Standard Curves of the Constituent to Be Determined. Because

the amount of excitation of a constituent per unit concentration is strongly
affected by the nature of the matrix or the bulk composition of the sample,
the standards must be prepared with the same matrix as the test sample. As
an added means of making the matrix uniform a spectrographic buffer is
added as already explained.
18-70. A suitable range of concentrations of the test elements and in-
EMISSION SPECTROPHOTOMETRY 483

TABLE 18-3
Spectral lines employed for quandtative determination of a number of elements
by arc emission, together with the appropriate internal standards'"
---------
· Internal
• Elements ______ .... (A.)
Wave lengths standards
----- .· - -
Ag 3208.7 3382.9 Ge
Al 3961.5 3082.2 2575.l Be, Pd
As 2780.2 2456.5 Tl
Au 2676.0 3122.8 Ge, Pd
B 2497.7 2496.8 Pd, Ge
Ba 4554.0 Pd
Bi 3067.7 2898.0 Tl
Cb 4058.9 3094.2 Be
Cd 3261.l 3466.2 Tl
Co 3453.5 3405.1 Pd
Cr 4254.3 4274.8 2986.5 Pd
Cs 4555.4 4593.2 Tl
Cu 3247.5 3274.0 Ge, Pd
Fe 3020.6 3440.6 3021.1 Pd
Ga 2943.6 2874.2 2944.2 Ge
Hg 2536.5 Tl
In 4511.3 3256.1 2932.6 Ge
La 4086.7 4123.2 Pd
Li 3232.6 2741.3 Ge, Tl
Mn 4030.8 4034.5 2576.1 Pd
Mo 3132.6 3170.3 3208.8 Ge, Pd
Ni 3414.8 3458.5 3101.6 Pd
Pb 3683.5 2833.1 2614.2 Tl
Pt 3064.7 2998.0 Pd
Rb 4201.9 Tl
Sb 2598.1 Ge
Sn 3175.0 3262.3 3034.1 Ge
Sr 4607.3 Ge
Ta 3311.2 3318.8 Be
Ti 3653.5 3234.5 Pd, Be
v 4379.2 4384.7 4390.0 Pd
Zn 3345.0 3282.3 Tl
Zr 3392.0 3438.2 3273.1 Be
---- ~-~--·---·· ·-------- --
° From Vanselow and Liebig, Spcctrochemical Methods, (Berkeley: U nivcrsity of California,
mimeo., 1948 ), p. 35.

ternal standards is made up with chemically pure substances which, in ag-

gregate, approximate the bulk composition of the test samples. If more than
3 elements are to be determined, it is sometimes advantageous to make a
separate series of standards for each set of 3 elements. Selection of the
elements to be grouped together in 1 standard preparation is made to pro-
vide a clear separation of the spectral lines to be measured (diagnostic
lines are given in Table 18-3), both of test elements and of the internal
standard. The spectrographic buffer is included in the formulation. Solid or
liquid standards are then arced or sparked and the concentrations are
further adjusted as necessary to give the proper range of spectral radiation
intensity. The line densities are then converted to radiation intensities and
484 EMISSION SPECTROPHOTOMETRY
the intensity ratio of internal standard to test element is then plotted against
the concentration to obtain the working curve.
18-71. Interpretatio n of Results. The photographic spectrogram is
placed in the comparator-densitometer (Fig. 18-10), and the character-
istic lines (Table 18-3) of the elements to be determined are located. as in
qualitative analysis (~ 18-57) by reference to the master spectrogram or
reference spectrogram (frequently iron) recorded with that of the sample.
Each worker usually accumulates a set of standard 49 spectrograms for the
different elements and marks a series of index lines for comparisons with
test samples. The Applied Research Laboratories supplies a library of
standard film spectrograms for use with their grating spectrographs.
18-72. Then each spectral line density is measured as percentage trans-
mission or as optical density (~ 18-62) with the densitometer. In the
densitometer, light passes from a very narrow slit through the spectral
analysis line on the emulsion to a photocell and amplifier system. This
scanning device is incorporated into the light projection system (Fig. 18-
10) so that the density measurement can be observed as the deflection on
a sensitive galvanometer. The percentage transmission of each line is re-
ferred to the photographic calibration curve (~ 18-63) and the relative
spectral line intensities are obtained. The ratio of intensity of the specific
analytical line for each element to be determined to that of the internal
standard or reference line is calculated. Then the quantity of the constituent
may be calculated by proportion:
. Intensity ratio, test .
Quantity, test element= ----:··-··--- .--·--·--··--- x Quantity, standard
Intensity ratio, standard
(18-5)

In practice, the intensity ratios for a series of standard samples of the test
element are plotted against the quantities of the standard taken, and the
quantities in the test samples are then read from the curve. Graphic calcu-
lators have been described. 50
18-73. In the line-width method 01 of interpretation of the spectrogram,
the width of the standard line on the microphotometer tracing is measured
at the intensity (peak height) of the element the concentration of which is
to be determined.

4D For wave lengths of specific elements, the following references are recommended
:
Harrison, op. cit.; Ahrens, Wavelength Tables of Sensitive Lines (Cambridge, Mass.:
Addison-Wesley Publishing Company, Inc., 1951) and Spectrochemical Analysis
(Cambridge, Mass.: Addison-Wesley Publishing Company, Inc., 1950); Brode, op.
cit., pp. 400-658; Dingle. Practical Applications of Spectrum Analysis (London:
Chapman & Hall, Ltd., 1950), pp. 86-122.
uO Vanselow and Liebig,/. Opt. Soc. Am .. 34:219 (1944); Oplinger, Anal. Chem.
24:807 (1952); Frederickson, Anal. Chem., 24:2019 (1952).
0 1 O'Connor and Heinzelman, Anal. Chem., 24: 1667 (1952).
EMISSION SPECTROP HOTOMET RY 485

QUESTION S
1. Explain how em1ss1on spectra originate, particularly with reference to
the atomic mechanism.
2. Distinguish the fundamental principle of emission spectrophoto metry
from !hat of absorption spectrophoto metry.
3. Compare emission and absorption spectrophoto metry as to the common
features involved in both their principles and procedures.
4. Distinguish the flame from arc and spark methods of excitation, particu-
larly as to applicability to soil chemical analysis.
5. Draw a sketch of the general optical relations in emission spectropho-
tometry.
6. List several elements that are readily determined by flame excitation. Why
are they more readily determined than other elements?
7. Contrast prisms and gratings as to the properties of the spectral array
produced.
8. What types of information can be obtained in qualitative emission spec-
trochemical analysis?
9. List the advantages of quantitative emission spectrochemi cal analysis over
other types of chemical analysis.
l 0. Explain the procedure by which an internal standard is generally used in
quantitative emission spectrophoto metry.
11. What is the purpose of spectroscopic buffers?
12. What should be the composition of the matrix of the standards?
13. Explain how preconcentra tion of a minor element for emission spectro-
chemical analysis may be easier than its chemical separation and determination
by conventional wet chemical methods.
14. What order of accuracy can be achieved by emission spectrophoto metric
analysis of Na? Of K? Of Ca? Of Mg? State the methods of excitation and ob-
servation considered.
INDEX
 Index

A Ammonium salts:
removal for calcium, 288
Absorption spectrophotometr y, 429, 450 potassium, 117
absorption spectrum determined by, removal of excess, 317, 392
446 Ammonium separation, 285
colorimetry in relation to, 430 after cupferron separation, 309
concentration determination by, 438- before cupferron, 309
446 with bromine, 286, 298
exercise in, 442-444, 449 Anion exchange capacity for phosphate,
instrumentation, 440-442 178
photo cell sensitivity for, 430 Aqua regia, 4
polychromogenic systems of, 445-446 Arc and spark em1ss10n spectrophotom-
record form, 444 etry, 465, 474
solids characterized by, 451 accuracy of, 475
test solution procedure, 443 arcing a sample for, 471
tube calibration, 377, 442 automatic recording, 475-476
Alkalization, degree of, 228 buffers for, 470
Aluminon, 298, 300 electrodes for, 470
Aluminum determination: identity of sample source, 466
colorimetric, 297 instrumentation, 466-470, 475-477
gravimetric, 311 internal standard, 480
standard solution, 298
photographic calibration, 4 79
Aluminum phosphate. 159. 161, 177, 179
preconcentration for. 474, 478
Ammonium:
qualitative analysis, '465
amount in soil, I 94
direct titration in digest, 192 quantitative analysis, 474, 484
distillation into boric acid, I 89. 195 volatilization rate of elements, 482
exchangeable, I 93 Arsenic:
extraction from soil, 191, 194, 196 bromide distillation of, 154
fixation of, 61, 193 in glass, 5
in soil testing, 366 in phosphorus determination, 135, 140
in waters, 188, 193 Asbestos filters, 23
Nessler method for, 195 Ash, total of plants, 330
titration of in H2S04, 195 Ashing:
titration in digest, 192 dry, 330, 334, 335
total in waters, 19 3 use of Mg(OAch in dry ashing, 334
Ammonium acetate: wet-oxidation, 331-334
destruction of, 86 Aspirator disposal of fumes, 184, 332
for exchangeable cations, 85, 87 H2S04 from Kjeldahl method. 184
ferrous iron, 391 HCl04 from plant digestion, 332
hydrogen, 74- 7 5 Aurin tricarboxylic acid, 298
489
 INDEX
490
B Calcium determin ation (Cont.)
reprecipi tation for, 316
Balances, analytical, 5 semimicr o oxalate method, 93
Barium: total in silicates, 285, 315
exchange capacity for, 67 Versene method, 64, 287
exchangeable cations in calcareou s Calcium phosphat e, 157, 179
soils, 88 Carbama te, for copper, 396
relation to infertility of soils, 324 with dithizone for zinc, 403
use in precipitation of sulfate, 264, with Versene, 289
322 Carbon:
Base exchange (see Cation exchange ) differentiation of mineral from organic,
Beer's law, derived, 436 206
Bicarbon ate: dry combusti on of, 208
determin ation, 260 exclusion if elementa ry form, 219
extractio n of phosphor us, 164 four classes in soils, 205
Biotic soil tests, 368 induction furnace for, 208
Boric acid: inorganic form, 205, 268
distillation of, 383 wet-oxidation, 211
distillation of NH 3 into, 189 (see also Carbonat e)
receiving solution for NH 3 , 187 Carbonat e:
standard solution of, 373 removal from soil, 210
Boron: total inorganic in soil, 268-270
acid soluble, 382 (see also Carbon and Calcium carbon-
as essential element to plants, 370 ate)
as fluoride complexer, 146, 320 Carbonat e determin ation, 260, 268
availability affected by pH, 370-371 carbonat e carbon, 268
extractio n from soils, 381 choice of indicator for titration, 53
loss from tissue, 386, 387 dissolved, 260
relation to irrigation water, 257 titration with acid, 53, 54
total in plants, 385 total alkalinity, 260
soils, 384 with bicarbona te determin ation, 260
toxicity, 257, 371 Carbon dioxide:
water soluble in soils, 379 determin ation of organic matter by,
Boron determin ation: 208, 211
concentr ation ranges of various meth- removal from strong NaOH, 186
ods, 374 Cation exchange capacity, 57, 59
curcumin method, 3 72 colloidal electrolytes in, 59
methods outlined, 3 71 definition, 59
quinaliza rin method, 375 different cations for, 60, 64
standard solution for, 373-374 flame emission method for, 62, 461
of organic matter, 63
pH dependence of, 60
c Versene method for, 62, 64
Calcium: Cell constant, conducta nce, 241
exchangeable in calcareous soils, 89 Chemical glassware, 5
exchange capacity for, 82 arsenic in Pyrex, 5
in plants, 336, 340, 464 boron-fre e glassware, 5
soils, 285, 315 cleaning for phosphor us, 140
in soil testing, 367 cutting of, 6
Calcium carbonate : glass filters, sintered, 6
effect on exchangeable cations, 82, 88 sodium from, 5
soluble salt analysis, 234 volumetric, 6
equivalence in limestone, 77 Chloride determin ation, 261
Calcium determin ation, 89 chromate indicator for, 262
flame photomet ric method for, 461-464 silver electrode for, 263
in presence of barium salt, 89 silver nitrate standard, 262 ·
in waters, 260 Chloride in soil testing, 367
oxalate method, 89 Chlorine, total in rocks, 324
 491
INDEX
Chrom ic acid: Conduc tance, electrical: (Cont.)
carbon equival ence of, 216 temper ature effect on, 231, 242
cleaning solution , 4 units for expression, 236
effect of chlorid e on, 21 S Copper :
ferrous iron on, 216 extracti on from plant residue, 427
mangan ese on, 215 from soil with HCl04, 400
mention of colorim etric method , 221 other soil extracta nts, 402
method for soil organic matter, 214, solubility affected by pH, 388
219 total of soils, 398
perman ganate use with, 221 Copper determi nation, 396
amyl acetate extract of color, 398
Chromi um, in rocks, mention ed, 324
ane electrod e, 38 carbam ate method , 397
Clay membr
polarog raphica lly, 421
Cleanin g solution, chromi c acid, 4 standar d, 397
Cleanliness in the laborat ory:
Crime detection and soil analysis, 473
ability to organize judged by, 9
Cupfer ron iron separat ion from alu-
need for coopera tion, 9 minum , 307-309
requisite for efficiency, 8 Curcum in, for boron, 372-373
Cobalt, 388, 414 Cyanid e:
determi nation, reviewed, 414 in plant tissue, 337
in polarog raphy, 421 in Versene titration s, 64, 289, 290
in potassiu m determi nation, 112
preconc entratio n for emission spectro- D
photom etry, 478
Colloid al electrolyte, 59 Digestion acceler ators for nitroge n di-
Colorim eter tube standar dization , 377 gestion, 185
Colorim etric oxidati on-redu ction indica- Diphen ylamine :
tors, 54, 55 as oxidati on-redu ction indicato r, SS,
starch indicato r for, iodimet ry, 55 220
Colorim etric pH indicators, 49, 52 reagent for nitrate, 343
critical pH of, 49, 50 Dipicry lamine reagent for potassium,
exercise on use of, 53 350-351
labeling, 51 Droppi ng mercur y electrode, 416, 420
papers, 53
pK value of, 49 E
prepara tion of solutions of, 51
table of, 52 Elements:
titration curves of, 50 chart of, back endpap er
Compo site soil sampling: essential to plants, 2
equival ent to an average, 12 Elemen tal analysis, 272
in selection of experim ental field, 26, accurac y of, 274, 275, 278
27 alumin um, 297, 311
number of cores require d, 13 bringing sample into solution, 273
of established experim ental plots, 13 calcium , 285, 315
pattern of cores in a plot, 24 conven tional system of, 300
require ments of, 13 ftow sheets for, 279, 280, 301
root spread integra tion effect, 13 fluo.ride, 324
Concen trated acids and bases strengt h of, iron, ferric, 291, 309, 311
3 ferrous, 320
Conduc tance, electrical: magnesium, 28S, 316
cell constan t, 241 mangan ese, 279, 280, 318
definition, 23S periodi c chart of, 2, endpap er
quantit ative test on saturati on extract, potassium, 285, 318
234,24 0 semimicrochemical system, 278
relation to temper ature, 23 I, 242 silicon, 294, 306
salt bridge for, 229, 236-239 sodium , 285, 318
semiqu antitati ve test on soil paste, 229, sulfur, 322
231 systems of, 273
standar d solution for, 239 titanium , 291, 314
 INDEX
492
Elemental analysis (Cont. ) Fluoride complex: (Cont. )
total in soil, 272 of aluminilm, 159, 161
weight basis for, 273 of iron, 159
zirconium, 323 with boric acid, 146
Emission spectrophotometry, 452 Furrow slice, weight of, 12 (see also
spectra, 452-454, 456, 482, 483 Plow layer; Volume basis)
(see also Flame emission spectropho-
tometr y) G
Exchangeable cations. determination of,
68,82 Germanium in phosphorus determina-
extraction with NH 4 0Ac, 84 tion, 135, 137
equivalence as pp2m of I meq, 84 Glass cutting, 6
in fertile soil, 84, 85 Glass electrode, 40
calcareous soils, 88 advantages of, 41
runoff, 85 asymmetric potential of, 41
meq percentage, 83 precautions with, 44
meq per 100 gm, 83 standa rd buffers for, 44
metallic cation status, 70 use for soils, 44
total of exchangeagble metallic, 68-70 Glassware, 5
Exchangeable hydrogen: Gypsum determination, 266 (see also
barium ethanolamine extraction, 75 Sulfate; Calcium; Salt concentra-
determination of, 73 tion)
NH 4 0Ac method, 74 Gypsum requirement, determination of,
other buffers for, 76 71
relation to aluminum, 73 relation to Na saturation, 71, 72
relation to lime requirement, 76
Exchange capacity (see Cation exchange
capacity; Anion exchange capac- H
ity)
Hydrofluoric acid:
F complexing (see Fluoride complex)
decomposition of silicate. 283, 319, 320
Ferrou s iron determination: for organic matter determination, 225
boric acid protection in, 321 volatilization of silicon, 306
exchangeable form, 391 Hydrogen, exchangeable in soils, 73 (see
in silicates, 320 also Exchangeable hydrogen)
rapid oxidation of, 391 Hydrogen ion activity:
Filter paper, 4 definition, 41
Flame emission spectrophotometry, 455 glass electrode for, 40
accuracy of, 455 (also see Soil pH value)
base electrolyte solution, 459 Hydroxyl in phosphate extraction, 162
cation exchange capacity with, 461
elements in plants by, 464 I
exchangeable cations by, 462
instrumentation, 457-459 Indicators, colorimetric pH, 49-54
internal standards for, 465 brom cresol green-methyl red, 53, 186
relative intensity, 456 in soil testing, 362, 364
runoff analysis by, 463 Instruments, physical for chemical analy-
spectra of the elements, 456 sis, 5
standards, 459 Internal standards:
total elements in silicates by, 463 emission, 465, 480
Flame photometer (see Flame emission homologous pairs for, 482, 483
spectrophotometry)
Iodine, titration of, 142-143
Fluoride: Ionic activity, 38, 163
extraction of soil phosphorus, 160-162
total in plants, 338 Ionic species:
soils, 324 as affected by soil dilution, 234
Fluorid e complex: determination of ions, 256
interference with aluminum, 284 in soluble soil salts, 228
INDEX 493

Iron: Luxury consumpti on of plant nutrients,

exchangeable, 391-393 330
in soil acidity testing, 364 Lysimeter percolates, 227 (see also Water
in soil aeration test, 355 analysis)
relation to potassium deficiency in
plants, 342, 354 M
soluoility as affected by pH, 388
Iron determina tion, 169, 191 Magnesium:
dithionite extracted, 169 extraction of exchangeable, 82, 84
ferrous in silicates, 320 in plants, 335
in plants, 3 89 in soil testing, 367
orthophen anthroline method, 389 total in soils, 285, 3 16
reductor-ti tration method, 311-314 Magnesium determination, 97
separation from titanium, 293, 313 as 8-hydroxy Quilolate, 100
standard, 292, 390 colorimetrically as phosphate, 99
thiocyanate method, 169, 364 flame photometr ic method for, 461-
tiron method, 291 464
total in silicates, 291, 309, 311 gravimetrically as pyrophosphate, 99
Iron oxides, free oxides removed by in silicates, 285, 316
dithionite-citrate, 168 in waters, 260
Iron phosphate, 159, 165, 178, 180 titration as phosphate, 98
Irrigation water quality: Versene method, 288, 290
conductan ce determina tion of, 242, 258 Manganese:
guide to, 257-258 easily reducible, 393, 395
relation to residual Na2C0:1, 257 exchangeable, 393, 395
representative analyses, rivers, 259 influence on organic matter determina-
tion, 215
solubility affected by pH, 388, 393-394
J testing in plants, 340
total in plants, 335
Junction potential, 42 soils, 318
water soluble, 393, 395
K Manganese determination, 102
development of HMn0 4 color, 105
Kilograms per hectare plow layer, 12 exchange capacity for, 66
exercise in colorimetry, 442
in silicates, 318
L polarograp hic determina tion of, 67,
426
Laborator y report outline, 8 (see Report standard solution for, 103
of analysis) Melt, removal of Na2 CO:i type, 302
Light wave band: Microvariations of soil, 26
glass filters for, 433-434 Mineral elements in plants:
gratings for, 434 concentration, 329
means of obtaining, 433 luxury consumption, 330
prisms for, 434 poverty adjustment, 328
selection for colorimetry, 432-433 total ash, 330
Lime requireme nts of soils, 76, 84 Mineral elements in soils (see Elemental
relation to crop species, 361, 362, 363 analysis; Exchangeable cations)
Limestone, neutralizing equivalence of, Moisture saturation percentage :
77 determination, 240, 247
Liming soils: use for soil pH, 45
materials, 77 use for soil salinity analysis, 240
requireme nt, 76, 84, 360 Molybden um:
relation to crop, 360, 362, 363 in phosphoru s determina tion, 135-139
to pH, 360, 362 solubility affected by pH, 388
to texture, 360 total in plant tissue, 412
Log table, inside back co'Yer in soil, 411
 IND EX
494
Moly bdenu m determ inatio n, 4-08 Organic matte r determ inatio n (Cont .)
stand ard, 409 HF-H CI pretre atmen t, 225
other methods, 414 ignition at low tempe rature , 225
thiocyanate, 408 in runoff, 217
thiocyanate extracted, 410 peroxide method for, 222
Morta rs, 7 readily oxidizable. 214, 219, 222
relation to ignition loss, 207
N removal of inorganic carbo n from, 210,
223
Nessler reagent, 195 total carbo n metho d for, 206, 208, 211
Neutralizing equivalence: wet-oxidation method for, 211
HCl metho d for, 78 Organic phosphorus, 169
magnesium carbo nate equivalence, 79 chemical chara cteriz ation, 174
of agricu ltural limestone, 77 determ inatio n by H 2 02, 170, 173
of other liming materials, 77 extraction, 171
rapid gasometric method, 79 Ortho phena nthro line:
use of CaCO a equivalence, 78 indicator, 217
Nickel, in polaro graph y, 421 iron determ inatio n with, 389
Nitrat e determ inatio n, 197 Osmotic pressure of salt solution, 245
extrac tion from runoff, 199 Oxygen, functions in soil, 354
from soil, 198
in plant sap, 342-347 p
pheno l disulphonic acid for, 197
Perchloric acid:
stand ard, 198
dehyd ration of silica, 294, 303, 305
Nitrification rate determ inatio n, 202
in Cu and Zn extraction, 400
field sampling for, 202
plant ashing with, 331
Nitrit e determ inatio n, discussed, 201
Periodic chart of elements, back endpa per
included with nitrat e in sap, 344
elements essential to plants in relatio n
Nitro gen: to, 2
ammo nium form, 193
Phosp hate:
estim ation of organic matte r from, 183
acid soluble, 154
nitrat e form, 197
nitrite form, 201 alumi num phosp hate extraction, 161
precipitation, 177
Nitro gen determ inatio n, 183
bicarb onate extractable, 162
Duma s metho d mentioned, 183
calcium phosp hate extraction, 157
estimation from ignition, 183
dithionite extractable, 165
fume disposal from, 184
exchange capacity for, 178
Kjeldahl metho d, 183
fluoride extractable, 159
plant tissue analysis for, 187
fracti onatio n of inorganic, 166-167
runoff analysis for, 188
hydroxyl extractable, 162
Nutri ent balan ce in plants, 328
iron phosphate extraction, 165
precipitation, 177
0 water soluble in runoff, 157
Optic al density, 436, 480 Phosp horus :
extinction coefficient in relatio n to, 43 7 organic forms in soils (see Organic
phosp horus )
from Nessler tube readings, 436
testing in plant sap, 347-350
linearity of calibr ation curves from, soils, 364
437
total in phosp hate rock, 176
Organ ic matte r: plants, 334-336
cation exchange capacity of, 63 soil by HCl0 4 , 176
chara cteriz ation in polaro graph y, 416 by Na2COa, 175
in soil testing, 366
(see also Phosp hate; Phosp horus de-
Organ ic matte r determ inatio n, 205 termi nation )
chromic acid method, 216, 219
classes of methods for, 206 Phosp horus determ inatio n, 134, 151
dry comb ustion method for, 208 arsenic relations, 137, 148, f51, 154
basic principles, 135
exclusion of eleme ntary carbo n from,
219 chloro stanno us reage nt, 142, 171
INDEX 495

Phosphor us determin ation (Cont.) Polarogra phic analysis (Cont.)

chlorosta nnous reduced, 141-146 diffusion current, 418
choice of methods, 134, 139 dropping mercury electrode, 420
concentra tion ranges of methods, 140- half wave potential. 418
141 supporting electrolyte, 420
fluoride interference overcome, 146 voltamm etric analysis, 416
fractiona tion of inorganic, 166-167 wave height, 418, 419
iron interferen ce overcome , 146, 176 Potassium :
molybde num reduced, 146 acid soluble. 132
naphtholsulfonic reduced, 148, 151 electrodialysis extractable, 132
pH values for solutions, 45, 49, 53 equilibrium release, 130
precaution against contamin ation, 140 exchange capacity for, 61
reagent concentrations, 136, 138 exchangeable form, 128
spectroph otometric methods outlined, fixation capacity of soils, 131
138 key element in soil fertility, 111
standard solution, 140, 350 testing in plant sap, 350-354
student exercise in, 54 soil, 365
test solution preparati on, 143 total in silicates, 129, 285, 318
vanadom olybdoph osphoric yellow, 151 Potassium determin ation, I 06, 111, 115
Plant sap testing, 339 centrifug e washing of precipitate, 122
boron in leaf strips, 387 cobalt hydrocar bonate color, 120
calcium, 340 cobaltinitrite isomorphism with Na,
concentr ation scale, 342 113-115
corn,343 , 346, 349, 352, 353, 354 combined with Na, 259
diagnosis in field, 339-341 flame photome ter method for, 461-464
iron, 354 gravimetric method for, 127
magnesium, 340 in waters, 259
manganese, 340 permang anate evaluation of, 125
nitrate nitrogen. 342-347 removal of ammoniu m from, 117
oats, 344, 346, 349, 353 standards , 116, 353
phosphor us, 347-350 total in silicates, 285, 318
potassium, 350-354 Potentiom eter principle, 3 8
relations to tissue analysis, 341 Pounds per acre plow layer, 12
soybeans, 344, 347, 349 Poverty adjustme nt of plant nutrient con-
type of tissue, 342 tent, 328
zinc, 340 Pyrosulfa te:
Plant tissue analysis, 326 preparati on of K2S201, 304
concentr ation of mineral elements, 327- fusion in, 306, 312
329
direct extractio n for Na and K, 336, Q
464 Quick tests, 339, 354, 357
grinding tissue for, 327 (see also Soil testing; Plant sap testing;
interpreta tion, 328 soil aeration)
relation to plant sap testing, 326, 339 Quinalizarin, 375, 376
sample size for analysis, 327
sample preparati on for, 326 R
stock solution for, 335
Plant tissue testing (see Plant sap test- Reagents :
ing) analytical reagent grades, 3
Platinum utensils: concentra ted acids and bases, 3
burnishing, 276 Report of analysis:
care of, 275 laborator y report outline, 8
cleaning, 276 pH value to include moisture content,
handling precautions, 277 42
specific reagents to avoid, 276 soil testing, 358
Plow layer, weight of, 12 water analysis, 256-257
basis of analysis, 10, 84 Resistance, electrical, 23 5
Polarogr aphic analysis, 416 of soil paste, 229
basic principle , 417 relation to conductance, 235
 INDEX
496
Resistance, electrical (Cont.) Silicon: (Cont. )
tempera ture correction for, 231, 242 relation to molybdo phospho ric color,
Runoff analysis: 135, 137
compositing, 253, 254 standard , 295
correcti on factor, 254 volatilization by HF, 306
enrichm ent ratio, 253 (see also Silica)
nitrate, 199 Sodium:
sample preparat ion. 254 degree of soil saturatio n with, 83, 228
nitrogen, total and ammoni um, 188 direct extraction from organic residues,
organic matter, 217 336,464
phospho rus, 157 exchangeable, I 07
potassium, 463 in soil testing, 367
pounds per acre inch, 255 percentage saturatio n with, 83, 228
total solids content. 253, 255 relation to alkali soils, 228
Sodium adsorpti on ratio, 258
s Sodium carbona te fusion of soil:
for boron, 384
Salinity (see Irrigation water; Soil sa- phosphorus, 17 5
linity) potassium, 129, 318
Salinity scale: molybdenum, 412
for soils, 244 in silicate analysis, 284, 302
waters, 258 Sodium determin ation:
Salt bridge, electrical, 229, 236 as magnesium uranyl acetate, 107
care of, 234 combined with K, 259
Salt concent ration: exchange capacity with, 65
from electrical conductance, 243 flame photometric method for, 461-464
in soil, 245 in waters, 259
ionic species, 256 in silicates, 285, 318
total dissolved solids in water. 256 (see also Sodium )
water sampling, 256 Sodium Hydroxide:
Salt index of fertilizers, 245 iron phospha te dissolution with, 165
Saturati on extract of soils, 45, 240 organic matter dissolution in, 171
conduct ance linear to salt concentra- protection from C02, 186
tion, 243 separati on of aluminum, 299
equilibr ation time, 241 Soil acidity (see Exchangeable hydro-
equipotential relation to moisture, 243 gen; Lime requirem ent; Soil pH
quantity per unit weight of soil, 241 value)
relation to I: 2 extract of soil, 250 Soil, a complex analytical system:
salinity scale for, 244 broad training involved, 1
water content at, 241 extractio n and determin ation distin-
Semimicrochemical silicate analysis, 278 guished, 2
accuracy, 278 selective gathering of analytical data, 2
apparatu s for-, 281-283
elements involved, 1
flow sheets for, 279-280
(see also Element al analysis )
use of powerful techniques, 3
Sesquioxide separati on (see Ammon ium Soil aeration test:
separati on; Sodium hydroxide sep- based on iron valence, 354-356
aration) on permeability to chalk, 356
Sieve openings, relation to meshes, 7 Soil alkalinity, 228
Silica: relation to irrigatio n water, 258
to sodium percentage, 228
dehydra tion in HCl, 306
HCl04, 305 Soil profile sampling:
ignition of, 305 catena, 11, 15
Silicate analysis, 272 (see also Element al description, 17
analysis) functionality, 14
Silicon: monolit h samples, 18, 20, 22
colorimetric determin ation, 294 objectives, 14
gravimetric determination, 303, 306 replication, 17
INDEX
Soil profile sampling: (Cont.) Soil pH value (Cont.)
sequences, 14, 15 effect of salts on, 4 7
site, 16 water content on, 42, 43, 46
Soil salinity: factors affecting, 41
definition, 227, 242-247 isohydric, 47
effect on seed germination , 251 junction potential, 42
occurrence in arid regions, 227 measuremen t of, 45
humid regions, 227, 228 relation to liming, 360, 362
salinity scale for, 244 sticky point method, 48
toxic limits for plants, 246 water saturation percentage method,
Soil sample: 45
size in relation to sieve size, 33 (see also Hydrogen ion activity)
weighing, 36 Soil testing:
Soil sample preparation, 30 comparison of methods, 357-359
drying, 31 definition of, 337
for available phosphorus, 158 diagonsis, 339
exchangeabl e cations, 85 extractants for, 359
pH, 44 interpretatio n of. 340, 358
salt analysis, 240 in the field, 339
silicate analysis, 27 5 kits for, 357-358
zinc, 407 nitrate, 193, 197, 366 (see Nitrate de-
termination )
from colloidal suspension, 255, 275
nitrogen, mentioned, 366
grinding, 33
mixing, 34 phosphorus, 364
pH value, 360
partitioning , 34
sieving, 31 potassium, 365
sampling soils for, 358
standard samples, 36. 275
storage, 35 Soil testing systems, 357
diversity of in different localities, 360
Soil sampling:
relation to biotic tests, 368
field diagram for, 29
tabular comparison of, 36 l
for nitrification rate, 202
soil testing, 358 Soluble salt analysis, 227
local problem spots, 30 conductome trically in soils, 229, 234
on a farm, 29 exclusion from exchangeab le cations,
relation to rooting, 10, 13 82
under a plant, 25 soil sampling for, 240
(see also Composite soil sampling; Soil
water sampling for, 256
profile sampling) Soluble salts:
Soil sampling depth: relation to resistance of soil paste, 232
deep sampling recommend ed, 15, 18 of water, 233
"poverty ridges" in relation to, 11 Spark emission, 465, 474 (see Arc and
volume basis important in, 10 spark emission spectrophot ometry)
weights of furrow slice, 12 Spectral sensitivity:
Soil sampling tools: of human vision, 430
augers, 16 photocells, 430-431
frames for monoliths, 19 relation to wave length, 431
sampling tubes, 23 Spectrograp h, 453, 465 (see also Arc and
Soil solution: spark emission spectrophot om-
changes of composition on dilution, etry)
234 Spectrophot ometry:
displacemen t of, 247 absorption, 429
pressure membrane extraction of, 248 emission, 452
soil dilution 1 : I, 249 Spectrum:
1:2, 249 grating production of, 448-450, 466
1:5,251 of human vision, 430
Soil pH value, 38 prism production of, 447-448, 469
averaging, 48 Starch indicator, for iodimetry, 55
colorimetric method for, 360, 362, 364 in chlorostann ous standardizat ion, 142
 INDEX
Subsoil sampling, 27
(see also Soil sampling depth)
w
Sulfate: Washing techniques:
in soil salinity, 228, 243, 263, 266 centrifuge washing, 58
in soil testing, 367 colloid dispersion during, 58
Sulfate determination, 263 in cation exchange, 57
exchange column purification, 266 Water:
gypsum, 266 distilled, 3
in waters, with Versene, 263 exchange columned, 4
turbidimetric, 265 steam condensation, 4
Sulfur: Water analysis, 227, 242, 256
total in plants, 336 Weight basis:
soils, 322 ignited basis, 273
Sulfur requirement (see Gypsum require- oven-dry basis, 273
ment) Wet-oxidation:
Supporting electrolyte solution, 417, 424, apparatus for plants, 332
425,426 soils, 212
fume disposal for, 186, 332
T methods for organic matter, 21 l, 214,
219
Titanium determinatio n: need for H 2S0 4 with HCJ04, 332-333
in silicates, 291, 3 I 4 of plant tissue, 331-334
peroxide method, 315 soil organic matter, 211, 214, 219,
standard, 292, 314 222
Total analysis (see Elemental analysis) Wolfram:
Trace elements, 370, 388 in phosphorus determination, 135
Transmission percentage: rocks, mentioned, 324
Beer's law in relation to, 435
concentratio n from, 438, 443-444 x
of light, 435
optical density in relation to, 436 X-ray fluorescence in zirconium determi-
plotting for linearity, 438 nation, 323
Tungsten (see Wolfram)
z
u Zinc:
Uranyl acetate, I 07 extraction from plant residue, 427
from soil with HCI04, 407
extractable from soil with dithizone,
v 406
other extractants, 407
Vanadium: sampling soil for, 407
in iron determination, 314 solubility affected by soil pH, 388
phosphorus determination, 151 total in soils, 398
rocks, mentioned, 324 Zinc determination, 402
Voltammetric analysis, 416, 422 dithizone method, 405
Volume basis, 84 polarographically, 42 I
for soil analysis, 2 standard, 404
plot-sized volume, 25 Zirconium, total in soils, 323, 474

c
 Table of logarithms

8 P.P.
N 0 I 2 3 .4 5 6 7 9 I 2 3 4 5
-- -- -- _.__ - - - - -- -- - - 8 12 17 2I
10 0000 oo86 0128 0170 0212 0253 0294 0334 0374
0043 4
II 0414 0453 0492 0531 0'569 o607 0645 0682 0719 0755 4 8 I I IS IQ.
12 0792 0828 0864 0899 0934 0969 1004 1038 1072 II06 3 7 IO I4 I7
13 xr39 u73 1206 1239 1271 1303 1335 1367 1399 1430 3 6 IO 13 16
14 1461 1492 1523 1553 1584 1614 1644 1673 1703 1732 3 6 9 12 15

15 1761 1790 1818 1847 1875 1903 1931 1959 1987 2014 3 6 8 II 14
16 2041 2068 2095 2122 2148 2175 2201 2227 2253 2279 3 5 8 II 13
2304 2330 2355 238o .2405 2430 2455 248o 2504 2529 2 5 7 IO 12
17 12
18 2553 2577 2601 2625 2648 2672 2695 2718 2742 2765 2 5 7 9
19 2788 2810 2833 2856 2878 2900 2923 2945 2967 2989 2 4 i 9 II

20 3010 3032 3054 3075 3096 3u8 3139 3160 3181 3201 2 4 6 8 II
2I 3222 3243 3263 3284 3304 3324 3345 3365 3385 3404 2 4 6 8 10
22 3424 3444 3464 3483 3502 3522 3541 3560 3579 3598 2 4 6 8 IO
23 3617 3636 3655 3674 3692 37II 3729 3747 3766 3784 2 4 5 7 9
24 3802 3820 3838 3856 3874 3892 3909 3927 3945 3962 2 4 5 7 9
25 3979 3997 4014 4031 4048 4o65 4082 4099 4116 4133 2 3 5 7 9
26 4150 4166 4183 4200 4216 4232 4249 4265 4281 4298 2 3 5 7 8
27 4314 4330 4346 4362 4378 4393 4409 4425 4440 4456 2 3 5 6 8
28 4472 4487 4502 4518 4533 4548 4564 4579 4594 4609 2 3 5 6 8
29 4624 4639 4654 4669 4683 4698 4713 4728 4742 4757 l 3 4 6 7
4771 4786 48oo 4814 4829 4843 4857 4871 4886 4900 3 l 4 6 7
30 6
31 4914 4928 4942 4955 4969 4983 4997 sou 5024 5038 3 I 4 7
32 5051 5o65 5079 5092 5105 5II9 5132 5145 5159 5172
I 3 4 5 7
5185 5198 5:zx1 5224 5237 5250 5263 5276 528g 5302
l 3 4 5 6
33 6
34 5315 5328 5340 5353 5366 5378 5391 5403 5416 5428
1 3 4 5
5441 5453 5465. 5478 5490 5502 5514 5527 5539 5551 I 2 4 :s 6
~g 5563 5575 5587 5599 5611 5623 5635 5647 5658 5670 I 2 4 5 6
37 5682 5694 5705 5717 5729 5740 5752 5763 5775 5786 1 2 3 5 6
38 5798 58oq 5821 5832 5&43 5855 5866 5877 5888 5899 1 2 3 5 6
39 5911 5922 5933 5944 5955 5966 5977 5988 5999 6o10 1 2 3 4 6
40
41
6o21
6128
6031
6138
6042
6149
6os3
6160
6263
6064
6170
6274
6o75
6180
6284
6085
6191
6294
6096
6201
6304
6107 6n7 1 2
6212 6222 I
6314 6325 I 2
. 3
3
3
4 5
4 5
4 5
42 6232 6243 6253
43 6335 6345 6355 6365 6375 6385 6395 6405 6415 6425 I 2 3 4 5
44 6435 6444 6454 6464 6474 6484 6493 6503 6513 6522 I 2 3 4 5

6532 6542 6551 6561 6571 658Q 6590 6599 66og 6618 I 2 3 4 5
45 2
46 6628 6637 6646 6656 6665 6675 6684 6693 6702 6712 I 3 4 5
6721 6730 6739 6749 6758 6767 6776 6785 6794 6803 I 2 3 4 5
:149 6812
6902
6821
69II
6830
6920
6839
6928
6848
693.7
6857
6946
6866
6955
6875
6964
6884
6972
6893
6981
I
I
2
2
3
3
4
4
4
4
so 6990 6998 7007 7016 7024 703§ 7042 7050 7059 7o67 I 2 3 3 4
SI 7076 7o84 7093 7101 7no 7II 7126 713~ 7143 7152 I 2 3 3 4
52 'l16o 7168 7177 7185 7193 7202 7210 721 7226 7235 I 2 2 3 4
53 7243 7251 7259 7267 7275 7284 7292 7300 :;308 7316 I 2 2 3 4
54 7324 7332 7340 7348 73s£> 7364 7372 738o 7388 7396 i: :I 2 3 4
to the Base 10
P.P.
N 0 I 2 3 4 5 6 7 8 9 I 2 3 4 s
--- -- -- -- -- -- -- -- - --
7466 7474 I 2 2 3 4
55. 7404 74u 7419 7427 7435 7443 7451 7459 2 2 3 4
56 7482 7490 7497 7505 7513 7520 7518 7536 7543 7551 I

7559 7566 7574 7582 7589 7597- 7604 7612 7619 7627 l 2 2 3 4

H! 6o
7634
7.709
7782
7642
7716
7649
7723
7789 7796
7657 7664 7672 'l-679
7731 7738 7745 7752
7803 7810 7818 7825
7686 7694 7701
7760 7767 7774
7832
I
l
2 3 4

7839 7846 I l 2 3 4
l
2 3 4
l

61 7853 7860 7868 7875 7882 7889 7896 7903 7910 7917 l l 2 3 4
62 7924 7931 7938 7945 7952 7959 7966 7973 7980 7987 l l 2 3 3
63 7993 8ooo 8007 8014 8021 8028 8o35 8041 8048 8055 l l 2 3 3
. .b4 8o62 8o69 8075 8o82 8089 8096 8102 8109 8u6 8u2 l I 2 3 3

65 8129 8136 8142 8149 8156 8162 8169 8176 8182 8189 I l 2 3 3
66 8195 8202 8209 8215 8222 8228 8235 8241 8248 8254 l I 2 3 3
67 8261 8267 8274 8280 8287 8293 8299 8306 8312 8319 . I I 2 3 3
68 8325 8331 8338 8344 8351 8357 8363 8370 8376 8382 I I 2 3 3
6o 8388 8395 8401 8407 8414 8420 8426 8432 8439 8445 I l 2 3 3
8451 8457 8463 8470 8476 8482 8488 8494 8500 8506 I l 2 2 3
70 8561 8567 I I 2 2 3
71 8513 8519 8525 8531 8537 8543 8549 8555
72 8573 8579 8585 8591 8597 8603 8609 8615 8621 8627 I I 2 2 3
8633 8639 8645 8651 8657 8663 8669 8675 8681 8686 1 1 2 2 3
73
' 74 8692 8698 8704 8710 8716 8722 8727 8733 8739 8745 l l :z 2 3
75 8751 8756 8762 8768 8774 8779 8785 8791 8797 88o2 I I 2 2 3
76 88o8 8814 8820 8825 8831 8837 8842 8848 8854 8859 I I 2 2 3
77 8865 8871 8876 8882 8887 8893 8899 8904 8910 8915 I I 2 2 3
78 8921 5927 8932 8938 8943 8949 8954 8960 8965 6971 I I 2 2 3
79 8976 8982 8987 8993 8998 9004 9009 9015 9020 9025 1 I 2 2 3
8o 9031 9036 9042 9047 9053 9058 9063 9069 9074 9079 I 1 2 2 3
81 9085 9090 9096 9101 9106 9u2 9u7 9122 9128 9133 I I 2 2 3
82 9138 9143 9149 9154 9159 9165 9170 9175 9180 9186 l l 2 2 3
83 9191 9196 9201 9206 9212 9217 9222 9227 9232 9238 1 I 2 2 3
84 9243 9248 9253 9::158 9263 9269 9274 9279 9284 9289 I I 2 2 3
85 9294 9299 9304 9309 9315 9320 9325 9330 9335 9340 I I 2 2 3
86 9345 9350 9355 9360 9365 9370 9375 9380 9385 9390 1 I 2 2 3
94i5 9440 00 II II 2 2
2 2
87 9395 9400 9405 9410 9415 9420 9425 9430
88 9445 9450 9455 9460 9465 9469 9474 9479 94 4 9489
So 9494 9499 9504 9509 95 13 9518 9523 9528 9533 9538 0 l I 2 2
9562 9566 957:r 9576 9586
9581 0 1 l :z 2
90 9542 9547 9552 9557 0 I I 2 2
91 9590 9595 96oo 9605 9609 9614 9619 9624 ¢289633
9652 ,9661 9666 9671 ¢75968o 0 I I 2 2
92 9638 9643 9647 9657 0 I I 2 2
93 9685 9689 9694 9699 9703 9708 9713 9717 9727
9722
9759 9763 9768
9773 0 1 1 2 2
94 9731 9736 9741 9745 9750 9754
9782 9786 9791 9795 98oo 98o5 98oc) 9814 9818 0 I l 2 2
9777
;I 9823 9827 9832 9836
9881
9841
9886
9845
9890
9850
9894
9854 9859
9899 9903 ~
0
0
I I 2
l I I
2
2
9868 9872 9877
~ 9912
9956
9917
9961
9921
9965
9926
9969
9930
9974
9934
9978
99i9
99 3
9943 9948 9952
9987 9991 99¢.
0
0
I
I
l
I
2
2
2
2
99