NPL Report CHEM 95

January 1979

NATIONAL P H Y SIC A L LABORATORY

VAPOUR-LIQUID EQUILIBRIA FOR THE SYSTEMS

WATER + J\JIEIIHANOL, WATER + ErHANOL, METHANOL
+ ErHANOL ANIJ WATER + ]\JIEI'HANOL + ErHANOL
AT 298.15 K DETERMINED BY A RAPID
TRANSPIRATION METHOD

by

D J Hall, C J Mash and R C Pemberton

Division of Chemical Standards

ABSTRACT

Isothermal vapour-liquid equilibria have been determined by means of

a rapid transpiration technique at 298.15 K for the ternary mixture water +

methanol + ethanol and for the binary mixtures water + methanol, water +

ethanol and methanol + ethanol. Values of the molar excess Gibbs energy

GE and the partial molar Gibbs energies G~l were computed and the results
tested for thermodynamic consistency.

The binary mixture data were correlated by the Wilson, Redlich-Kister,

NETL and UNIQUAC equations and the ternary systems data predicted from the

binary data. The Redlich-Kister equation "\\Tith adc1.ed ternary terms was

fitted to the binary and ternary data. The predicted values are compared

with the experimental results.

 CONTENTS

1• INTRODUCTION

2. EXPERIMENTAL 2

3. MATERIALS 5

4. RESULTS 6

S. DISCUSSION 9
6. CONCLUSIONS 11

APPENDIX A 12

TABLES

FIGURES

REFERENCES
 - 1 -

1. Introduction

Many industrial processes involve distillation, phase separation, or

analysis of aqueous solutions of widely varying compositions. The work

described in this paper was carried out to assess the usefulness of a

transpiration apparatus for the rapid study of the vapour-liquid equilibria

of multicomponent aqueous alcohol systems and to present suitable correlation

equations for the prediction of multicomponent vapour-liquid phase equilibria

from data for the constituent binary systems.

Measurements were made at 298.15 K on the binary mixtures water + methanol,

water + ethanol and methanol + ethanol and on the ternary mixture water +

methanol + ethanol.
 - 2 -

2. Experimental

Recently Arnikar, Rao and Bodhe(1) described a novel procedure which

considerably reduces the tedium involved in the measurement of the equilibrium

partial vapour pressures for multicomponent mixtures by the transpiration

teChnique(2). The transpiration apparatus used in the present work is shown

in figure 1 and was a modified form of that reported by Arnikar et al and

subsequently used by Rao, Bodhe and Gandhe(3) in a stu~ of a ternary liquid

s ys tern in which the component partial pressures were measured by gas chr8:i:.~"!;8

graphic analysis.

The liquid mixture (250 cm 3 ), which had been prepared by weight to within
+ 0.005 g, was poured into two glass spirals, which were then connected in

series and immersed in a water filled thermostat tank controlled at 298.15 +

0.01 K. The glass spirals were made from 1.4 em o.d., 1.0 em bore pyrex

glass tubing and contained approximately 60 em 3 of solution. The spirals had

a diameter of 7 em, pitch of 2.5 em and a length of 95 em. Pure dry helium
1
was passed at a rate of 0.167 cm3 sec- through the liquid as a stream of

small bubbles issuing from a fine capillary tube at the bottom of each spiral.

The helium gas stream was effectively saturated in the first spiral thereby

minimising any change in -the composition of the liquid in the second spiral.

I i was found -that this composi-tion varied by a mole fraction of < 0.0001

for the duration of a series of measurements. The li~~id phase mole fractions

were known to within:!: 0.0003. The saturated gas stream then passed through

a heated stainless steel vapour transfer tube to a pneumatically operated

gas sampling valve (Pye-Unicam Ltd, type 13440) mounted in a thernostatically

controlled oven at a temperature of 411.6 : 0.1 K. The vapour stream was

normally vented to the atmosphere through one of the ports inihe sampling

valve. Operation of the valve transferred a 5 cm 3 vapour sample, via a second

heated stainless steel transfer tube to the inlet port of a gas chromatograph

(Packard-Becker Ltd, model 417).

 - 3-

The gas chromatograph (g.~) was fitted with stainless steel columns

(length 4 Ill, 2.2 rnm bore) packed with 50-80 mesh porous polymer (Waters

Associates Inc, Porapak Q) held at 373 K and a thermal conductivity detector

which had tungsten-rhenium filaments (Gow-Mac Instrument Co, type W2X). The
3 -1
helium carrier gas flow rate was 0.334 cm sec • The detector was operated

at a filament current of 150 rna. and a temperature of 423 K. The two transfer

tubes were heated to approximately 408 K and the gas chromatograph column inlet

ports were controlled at 413 K. Before each series of measurements the g.c

colurrln packing was pre-treated by sequential injection of about 15 equilibrium

vapour samples over a period of 1500 s. Chromatographic analysis of subsequent

vapour samples required between 600 and 1200 s. depending on the components'in

the mixture. Between 10 and 20 samples were analysed for each mixture compos-

iiion siudied.

The out of balance signal from the detector bridge circuit was displayed

on a potentiometric recorder (Leeds and Northrup Ltd, model W). A gas chrom-

atograph integrator (Hewlett-Packard Ltd, model 3370A) was connected in parallel

with the recorder and significant deviations from the signal baseline were

recorded as peak areas in units of p.V. sec. Tho integrator had a binary coded

decimal output,the signals from which wore serialised and punched on to paper

tape in a form suitable for subsequent computer processing.

Transistor-transistor logic circuits were used to automate the repetitive

vapour sample injection and analysis operations so that the apparatus could be

run unattended for long periods. A computer was used io sari ihe gas chromaio-

graph peak area data, io calculate the mean areas of ihe peaks and iheir root

mean square standard deviaiions and ihen to calculate ihe mixture component

partial pressures Pi and activity coefficients y. and the molar excess Gibbs
l
E
energy G •

Measurements were made to test for vapour phase equilibration in the spirals

and the precision of the g.c. analyses under different experimental conditions.

The following experimental parameters were variedj the helium flow-raie through
 - 4-

the spirals (0.042 to 0.42 em 3 s-l); the volume of sample present in each of

the spirals; the g.e. oven temperature (353 to 413 k); the g.e. detector,

transfer line and sample valve temperatures (388 to 423 K) and the g.c.

det ector filament current (150 to 250 rna.). For liquid samples> 20 cm 3 in
3 -1
both spirals and helium flow-rates)0.042 (0.42 em s the g.c. peak areas

for the test mixture were constant, within the estimated experimental error.

0' for all peak areas A, was less than! 0.25 per eent of Ai at all x. ..~
1 1

studied. When the sample valve or transfer tubes were maintained at

temperatures >473 K an apparent decomposition of the methanol and ethanol

vapour occurred; also at a katharometer filament current of 250 rna. the g.c.

peaks for ethanol tailed to below the normal signal baseline.

The linearity of g.o. peak area \\Tith respect to the vapour pressure

for water, methanol and ethanol was investigated over the temperature range

278 to 303 K for fixed vapour sample volumes of 0.25 and 5.0 em3 • The

katharometer response was linear within our estimated experimental errors

for vapour sample volumes below 5.0 em 3 over the range of vapour pressures

studied.
 - 5 -

, lVlaterials
.).

&~alaR methanol and ethanol (James Burrough Ltd) were used without

further purification. Water analysis by g.l.c. with a Porapak Q column and

a thermal conductivity detector showed the methanol and ethanol to contain

0.03 and 0.06 per cent by mass of water respectively. The presence of this

water was taken into account in computing the composition of the binary and

ternary aqueous mixtures. The doubly distilled water used was prepared in

the laboratory. The helium (99.998 per cent purity) was supplied by British

Oxygen Company Ltd.

 - 6 -

4. Results

It has been shown by Halasz(4) that the katharometer detector has a

response proportional to the concentration of a component i in the carrier

gas. We have therefore assumed that the mass of component i present in unit

volume of the equilibrium vapour phase is, to a good approximation over the

pressure range of our experiments, proportional to its partial vapour pressure

Pi' Thus for a fixed volume of the vapour

p./p?
l l
= A./A?
l l
= .•. (1)

a
where for a species i, and p. are the partial and pure component
l

vapour pressures, x. and y. are the mole fraction and activity coeffioient
l l
o
of i in the mixture when A. is the observed g.c. peak area and A. is
l l

the peak area for x. = 1 under identical experimental conditions.

l

in the determination of the liquid phase activity coefficients Yi was better

than -+ 0.5 per cent.

E
The results y.
l
and molar Gibbs excess free energy G calculated

for the systems water + methanol, water + ethanol, methanol + ethanol and water

+ methanol + ethanol at 298.15 -+ 0.01 K are presented in tables 1 and 2 and

the pure component vapour pressures used in these calculations are shown in

table 3. In these and subsequent tables water, methanol and ethanol are denoted

by the subscripts 1, 2 and 3 respectively.

Total vapour pressure and liquid phase activity coefficient data for water

+ ethanol have been reported at 283.15 to 303.15 K(5), 323.15 and 333.15 K(6)

and 303.15, 323.15, 343.15 and 363.15 K(7). Partial vapour pressure data at

298.15 K have also been published (2,8). The Pi (x ) data from the present
1
work and the transpiration experiments of Dobson(2) and Shaw and Butler(8) are

shown in figure 2. In figure 3 our values are compared with the T'PSl1 I I

of D'Avila and Silva(5) from static isoteniscope measurements and the corrp.lat.~ -

by Larkin and Pemberton(9) of the experimental vapour pressures (303.15 to

363.15 K)(7) and excess enthalpies ~ (298.15 to 383.15 K)(10) for this ",~T~'+om
 - 7-

The partial pressure data reported here agree well with those of Dobson and

Shaw and But 1 er. Deviations from ideality are positive fOT all
1
has a maximum value of 747 ~ 6 J mol- at x z 0.45 which is to be compared
3
-1
with 746 J mol at x ~ 0.44 quoted by Larkin and Pemberton at 298.15 K.
3
Total vapour pressures for water + methanol have been reported at 308.15,

323.15, 338.15 K(11) and at several temperatures in the range 313.15 to

47 3. 15 K e1 2) • , I pressure d e t ermlna
Parcia . t lons
. f or t his system at 335. 65,

323.15(13), 332.15 and 313.15(14), 313015(1 5 ) and 298.15 K(13,16) have also

been published. Our measurements are in good agreement with those of Butler,

Thomson and ~~cLennan

(16) as is shown In figures 4 and 5. Deviations from

ideality are positive for all x and our work at 298.15 K shows a maximum
2
E + -1
in G of 320 - 6 J mol at x2·~ 0.46.

Total vapour pressures for thes,ystem metllanol + ethanol have been measured

at various temperatures in the ranges 273.15 to 373.15 K(18) and 373.15 to

433.15 K(19). The only partial pressure data for this system appear to be

those of Morris, Munn and Anderson(20) at 313.15 K. In the present study

deviations from ideality were negative for all x at 298.15 K with a minimum
2
1
in GE of -22! 6 J mol- • Total pressure measurements(21) at 343.143 K for

the range x
2
= 0.00143 to 0.38441 using a precise static method(7) have also

shown deviations from ideality to be small and negative with E -1

G = -9.2 J mol

at x 2 -- 0.38441(1 7 ).

The three binary mixture data sets were tested for thermodynamic cons is-
1 22 23'
tency by applying the area and symmetrical area tests of Herington\ 1 ).

The symmetrical area test confirms that the data for water + ethanol are con-

sistent, \vhile data for water + methanol ShOlJ'T minor inconsistencies in the

range xl from 0 to 0.2. The methanol + ethanol system is almost ideal and

the data 1J'lere regarded as acceptable even though the tests suggested -thermo-

dynamic inconsistency.

There are several isobaric vapour-liquid equilibrium studies of the water

 - 8 -

+ methanol + ethanol system in the literature

(24 ,2-
),
26) • The most extenslve
.

measurements are those of Delzenne(26) which include a study of the binary

system methanol + ethanol from 339 to 350 K. These data show positive devia-
E 1
tioncfrom ideality with G = 31 J mol- at xi = 0.47 at 344 K whereas the

present study has indicated that small negative deviations occur at 298.15 K.
E
Our experimental G (x.) data set for the ternary system water + methanol
1

+ ethanol is shown in table 2 and represented on a contour diagram in figure 6.

The errors introduced in calculating aE for the binary systems neglecting

vapour phase imperfections were less than 3 J mol-i. Values for the pure

component second virial coefficients and molar volumes are shown ln table 3.
-1
The estimated experimental uncertainty in was ± 6 J mol .

The effect on the liquid-vapour equilibrium of the heliwn used as the

transpiration gas has been estimated. Gupta, Lesslie and King(2 7) have shown

that the second mixed virial coefficient :E.. for heliu.'ll + ethanol is only
lJ
32 -+ 6 em 3 mol -1 at 29 8 .15 K. B.. for the systems helium + methanol
Values of
lJ
and helium + water are expected to be of similar magnitude and, as the second

virial coefficient for helium is 11 cm3 mol- 1 at 298.15 K(28), it can be shown

that, at our experimental hcliu.m partial pressure of about 113 kPa, the helium

vapour phase interactions are negligibly small. However, even if ideality of

the vapour phase is assumed the vapour pressure of a volatile liquid will be

enhanced under the partial pressureaf an inert gas(2 9 ). For our experimental

conditions the partial pressures of water, methanol and ethanol at x. = 1

1

would be increased by only 3, 32 and 21 Pa respectively through this effect.

The presence of helium as an additional component in the systems studied was

therefore ignored.
 - 9 -

5. Discussion

The Wilson (36), NRTL (37) and UNIQUAC(3 8 ) equations and the t,1TO- and three-

coefficient Redlich-Kister(39) equaiions (see equations 4 to 13 in Appendix A)

were fitted by a leasi squares meihod, with minimisation on the residuals in

(GE/RT), to the data for the bina~J systems. The coefficients determined for

these equations and the root mean square deviations o(GE/RT) in the molar

Gibbs energy of the fits are shown in table 4. ~(GE/RT) is uefined in equation 2.
1

\~ ((GE/RT) _(GE/RT) )2/~Y2 ••• (2)

L!- / calculated experimental J
E
where n is the nTh"1lber of experimental data points. The values of cf (G /RT)

recorded in table 4 show that the three-coefficient Redlich-Kister equation

gives the best fit for all three binary systems. If the NRTL equation is eon-

sidered as having only two coefficients (ec being set as 0.3 in our fitting
E
program) then the 6(G /RT) values for the two-parameter equations are of

similar magnitude.

Equations 4 to 13 and the coefficients in table 4 for the three binary

systems were used to predict the values of the activity coefficients y and
i
E
the molar Gibbs excess free energies G , as a function of x., for the ternary
l

system water + methanol + ethanol at our experimental ternary compositions.

E
The resi~Qals 6(G /RT) are recorded in table 5. The form of the NRTL equation

appears to predict the ternary data best while the three-coefficient Redlich-

Kister equation produced the largest residual. It is not possible to give a

firm recor,~endation as to which is the best two-coefficient equation as all

four residuals are of similar magnitude and about three times our estimated

overall lli~certainty in GE •
The binary and ternary data were also fitted to a truncated orthogonal
( 11 ~)
Redlich-Kister equation ,u with nine binary and three ternary coefficients

(see 8.luation 3 and table 6).

 - 10 -

GE/ID' l a .. x.x. + rb ~J
.. x.x. (x.-x.)
~J ~ J ~ J ~ J

+ r: c .. x.x.
~J ~
J
(x~
~ lJ
2
- 8/3x.x. + x.)
J
+ A. 'k x.x.~ + B. 'k x.x.~ (x. + ~ - 2x.)
~J ~ J ~J ~ J J ~

+ Cijk XiXj~ (~ - x j ) ... (3)

The value of d(GE/RT) for this fit is about three times smaller than

the predictions from binary data alone but uses about twice the number of

data points and fitted coefficients.

The perspective view of the water + methanol + ethanol

surface shown in figure 7 was computed from equation 3 and the coefficients
E
in table 6. The values of G were calculated at x' J x. corresponding to
~ J
the intersections on a triangular mesh superimposed on the triangular

Xi' x ' ~ base. The perspective view was produced by projecting the lines
j
E
joining adjacent points of -me G mesh thereby represent ing the GE(x. ,x .)
l J
surface as a series of planar triangles.
 - 11 -

6. Conclusions

The transpiration Lechnique described is capable of producing vapour-

liqUid equilibrium data for the binary and multicomponent systems studied

with a precision in the activity coefficients of better than ~ 0.5 per cent
over the composition range studied when vapour phase non-ideality can be

neglected. The limitations of the method are principally those of the

availability of a suitable gas chromatographic column for the analysis of

the vapour samples.

 - 12 -

Appendix A

Wilson Equation

where A.. - 1.
II

NRTL EqJJation

N
N
j=1
I T ••
Jl Jl J
G.. x.
GE/RT = L x.
l N
i=l
L GkiXk
k=1

where

T •• (g .. - g .. )/RT (6 )
Jl Jl II

G• •
Jl
= exp (- a .. 't . . )
Jl Jl
where ot·· 01.. .
Jl lJ

InUQUAC Equation

N ¢i z N e.
GE/RT = I
i=1
x. In - +- L q.x. In .2.
l X.
l
2
i=1 l l ¢ l.
(8)

- L
N

i=l
q.x. In
l l [J, OjT ji]

where T· •
Jl
= exp - [( u .. - u .. ) IRT]
Jl II

q.N. q.x.
l l l l
e.l - =
L q.N. J J L q.x.
J j J J

r.N. r.x.
l l l l
¢ l. = = " (11)
L r.N. J J
I r.x.
J J
j j
 - 13 -

Redlich-Kister Equation

"Ii:
G':'./RT = x.x. [a .. + b .. (x.-x.) + c .. (x.-x.)2] •• (12)
lJ l J lJ lJ l J lJ l J

= G~ .IRT .. (13)
lJ
 E
TABLE 1. Molar excess Gibbs free energies G , vapour-phase mole fractions

Yi' liquid-phase activity coefficients Yi' and total pressure P,

for the systems H 0(I) + CH OH(2), CH 0H(2) + C H OH(3) and

2 3 , , 3 2 S
H 0(I) + C2H OH(3) at 298.15 K.
2 5

x x E
Y1 Y2 ...E... ln Y ln Y lny G
1 2 1 2 3
kPa Jmol-1

0.8796 0.1204 0.483 0.517 5.856 0.014 0.395 148

0.7961 0.2039 0.347 0.653 7.476 0.027 0-346 228
0.7081 0.2919 0.2705 0.7295 8.804 0.059 0.261 292
0.6019 0.3981 0.2105 0.7895 10.051 0.104 0.162 315
0.5169 0.4831 0.174 0.826 10.975 0.154 0.102 319
0.,1651 0·5349 0.156 0.845 11·505 0.194 0.070 316
0.4129 0.5871 0.1355 0.8645 12.071 0.223 0.048 298
0.3019 0.6981 0.096 0.904 13.283 0.290 0.015 243
0.1977 0.8023 0.061 0.939 14.445 0-346 -0.003 164
0.1478 0.8522 0.045 0.955 15.080 0-366 -0.003 128
0.0841 0.161 8.564 -0.030 -0.005 -18
0.1353 0.247 9.046 -0.025 0 - 8
0.1902 0.331 9.555 -0.020 -0.002 - 5
0.2919 0.467 10.418 -0.016 -0.005 -20
0.3585 0·543 11.036 -0.014 -0.002 -16
0.4414 0.628 11. 765 -0.013 -0.004 -20
0·5372 0.715 12.639 -0.008 -0.010 -22
0.6392 0·792 13·557 -0.009 -0.007 -21
0.7489 0.865 14.582 -0.006 -0.009 -17
0.8535 0.928 15·484 -0.007 -0.024 -24
0.9165 0.961 16.122 -0.003 -0.046 -16
0.9450 0.677 4.444 0.004 1.199 173
0.875~ 0·503 5.608 0.016 1.044 357
0.7858 0.421 6.397 0.079 0.786 571
0.6059 0.352 6.988 0.248 0.377 741
0.4504 0.2995 7·355 0.434 0.174 722
0.2994 0.2315 7·574 0.614 0.054 546
0.2158 0.1815 7.695 0·714 0.020 418
0.1604 0.1415 7.825 0·780 0.016 341
0.1210 0.1105 7.798 0.810 0.002 246
0.0635 0.061 7.879 0.864 0.003 143
TABLE 2. Molar excess Gibbs free energies GE I vapour-phase mole fractions
liquid-phase activity coefficients Y.7
1.
and total pressure P,
for the system H 0(1) + MeOH(2) + EtOH(3) at 298.15 K.
2

x2 E
Xi Y1 Y2 .J2... In Y1 In Y2 In Y G
3
kPa Jmol-1

0.0955 0.0992 0.017 0.183 8.611 0·793 --0.065 0.004 180

0.1030 0.1974 0.072 0.335 9.448 0·740 -0.056 0.017 191
0.0940 0.7 047 0.035 0.847 13.863 0.492 -0.017 0.032 100
0.0966 0.8061 0.032 0·915 14.160 0.418 -0.012 0.044 86
0.1513 0.2243 0.100 0.370 9.623 o. croo -0.065 0.035 281
0.1445 0.2943 0.088 0.459 10.157 0.665 -0.068 0.042 247
0.1461 0.3496 0.082 0.528 10·577 0.624 -0.059 0.039 224
0.1422 0.4280 0.073 0.614 11.300 0.601 -0.045 0.045 212
0.1457
0.1432
0.1476
0.1935
0.4946
0.6299
0.7032
0.1065
0.068
0.057
0.053
0.144
0.677
0.798
0.855
0.191
11.844
12.976
13.630
8.543
0.556
0.489
0.438
0.695
-0.044
-0.031
-0.022
-0.099
0.061
0.054
0.064
0.030
201
156
145
359
I
0.1862 0.3062 0.108 0.478 10.166 0.620 -0.066 0.053 302
0.1956 0.5019 0.088 0.692 11.731 0·505 -0.046 0.080 248
0.1930 0.7015 0.068 0.866 13·501 0.403 -0.017 0.087 186
0.2118 0.1577 0.147 0.273 8.826 0.657 -0.103 0.032 354
0.2144
0.2989
0.2948
0.6332
0.1524
0.2044
0.079
0.191
0.177
0.815
0.267
0.346
12.845
8.663
9.104
0.406
0.558
0·548
-0.024
-0.111
-0.095
0.119
0.083
0.095
223
485
470
I!
Ii

0.3029 0-3037 0.160 0.487 9.829 0.492 -0.073 0.115 427

0.2974
0.2968
0.3399
003437
0.5007
0·5517
0.1536
0·5032
0.122
0.115
0.208
0.134
0.715
0.766
0.273
0.735
11.444
11.881
8.642
11.381
0.393
0.369
0.513
0.339
-0.037
-0.026
-0.099
-0.019
0.161
0.169
0.118
0.213
325
299
542
346
I
0-3946 0.3055 0.188 0·510 9.651 0.372 -0.051 0.212 483
0.4107 0.4380 0.163 0.689 10.727 0.294 -0.004 0.286 402
0.4959 0.1529 0.260 0.295 8.245 0.312 -0.062 0.281 604 u
0.4997 0.1973 0.248 0.375 8.647 0.302 -0.032 0.314 594
0.4952 0.2988 0.216 0·542 9.420 0.258 +0.008 0.344 498
0.4968 0-3510 0.203 0.618 9.841 0.236 +0.023 0.386 456
0·5448 0.3064 0.225 0.580 9.388 0.201 +0.047 0.450 473
0.6167
0.6210
0.6925
0.6946
0.1524
0.2272
0.1101
0.1540
0.294
0.266
0.331
0.314
0.336
0.489
0.277
0.383
7.942
8.575
7·525
7.828
0.178
0.148
0.127
0.110
+0.032
+0.086
+0.110
+0.140
0.481
0.560
0.641
0.687
560
488
561
SOO
I
0.6981 0.2013 0.297 0.499 8.169 0.093 +0.178 0.742 435
0.7989 0.1021 0.384 0.326 6.861 0.040 +0.258 0.936 375
 o
'.I:ABLE 3. Vapour pressures Pi' second virial coefficients B..
11
o
and molar volumes Vi' H20(1), CH 0H(2) and
3
C H OH(3) at 298.15 K
2 S

0
p. -B .. VO
..L II 1
kPa 3
em mol
-1
em 3 mol-1
18(e)
! Water 3.170(a)
I
1265° i

II
16.949(b) d 41(g)
Methanol 2075
7.8n(b) f 59(g)
Ethanol 2413

(a) from reference 29. (d) from reference 31. (g) from reference 34.

(b) from reference 30. (e) from reference 33.

(c) fron reference 27. (f) estimated from data in references 27 and 32.
TABLE 4. Coefficients of the Wilson, NRTL, UNIQUAC and Redlich-Kister equations

fitted to e-E/RT for the systems H 0(1) +

2
CH30H~2)1 CH 0H(2) + G H 0H(3)
2 5
3
and H 0(1) + C H OH(3) at 298.15 K. Tne value of a used in the
2 2 s ij
NRTL equation was 0.3. The root mean square deviations O(CE/RT) =
E
[2: (G7RJ1\-ca1 c. -(C jID'l
dbs.
)2jn]; where n is the number of data points.

Wilson NRTL
I UNIQUAC
Redlich-Kister
2 coefficient 3 coeffic ient

H O(1) + CH OH(2)
2 3
A
12 ,T 12 1.1357 0.019'1 -0.1376

A21 "21 0.4546 -0.3405 -0. 136t)

a 12 0.5092 0·516 5
b
12 0.11t)7 0.11t) 9
c
I -0.056 9
12
0(GEjRT) 0.0031 0.0029 0.0039 0.0023 0.001 5
Ii
CH Gr(2) + C H OH(3) i
3 2 5
I
1\23"23 1-3052 0.2863 -0.3465

1\32"32 0.7674 -0.2973 0.4334

I
a
23 -0.0370
I -0.030 6
b
23 ,
-0.0155
II -0.016 7
c ! -0.041 3
23 !
o(GEjRT) 0.0025 0.0026 0.0024 0.0024 I
I
,
0.001 9

H20(1) + C2H OH(3)

5
Ii , o 7850 1 5326 o 0316
0.2547 . -0.0150 0.3349
a
31 1.1848 1.1 97 2

b -0.2793 75
31
c
31 I 18
o (GE/RT) I 0.0074 0.0058 0.0064 0.0044
I 0.0039
TABLE 5 Prediction of (GE/RT) at 298.15 K for ternary mixtures of H 0(1) +
2
CH 0H(2) + C H 0H(3) using the Wilson, NRTL, UNIQUAC and Redlich-
3 2 5
Kister equations and the binary coefficients from Table 4.

Wilson JlJRTL UNIQUAC Redlich-Kister

2 coefficient 1 3 coefficient ;

I
d (G
E
/m) x 163 i
I
6.75 II 5.99 6.56 6.53
I 7.27
TABLE 6 Coefficients and the root mean square deviation o"(GE/RT)

fa:!.' the orthogonal Recilich-Kister equation 0) used' to

represent the ternary and binary data for the system H 0(1) +
2
CH 0H(2) + C2H OH(3) at 298.15 K.
3 S

i j k a ..x10-3 b. ,xia-3 c ..x' 10- 3 A.lJ'kx1O-3 B _'Ci0-3

C.lJ"kxrGJ.
-3 a( GIY~)
1J lJ l J - ' ijk x 10-

1 2 508.788 132.244 -54.8048

2 3 ~35.9548- -18,0391 0.787681
3 1 1190.71 -276.329 -45.9307 !

1 2 3 I 1-'249.520 -72.2536 -3.29347

2·90
Legends for fiKQreS

FIGURE 1. Transpiration apparatus

A, pressure controlled heliwn (99.998 per cent purity) inlet; B~ drying

tower; C ' C2~ working and reference g.c colwnns; D~ katharometer detector;
1
E? g.c gas sampling valve (Fye-Unicam Ltd, type 13440); F, gas sampling valve
pnewnatic activator; H, gas chromatograph (Packard-Becker Ltd? model 417);

I, glass cone and socket joint; L , L , heated vapour transfer lines;

1 2
M1 , M , mass flow controllers; P, pressure controller.
2

FIGURE 2. Partial vapour pressures for the system H 0(1) + C H OH(3) at

2 2 S

298.15 K.

A comparison of published partial vapour pressure data with the present

work. 0, present work, 6, Dobson(2), 0, Shaw and Butler(8).

FIGURE 3. Partial molar Gibbs free energies for the system H O( 1) + C H 0H(3)
2 2 5
at 298.15 K.

0, present work, • , dfAvila and Silva(5) , 6, Larkin and Pemberton(9)

from a correlation of experimental ~ (298.15 - 383.15 K) and vapour pressures

(303.15 to 363.15 K).

FIGURE 4. Partial vapour pressures for the system H 0(1) + CH30H(~) at
2
298.15 K.

0, present work, o, Butler, Thomson and MacLennan(16).

FIGURE 5. Partial molar Gibbs free energies for the system H 0(1) + CH 0H(2)
2 3
at 298.15 K.

0, present work, 0, Larkin and Pemberton(9), 0, Butler, Thomson and

MacLennan(16) •

FIGURE 6. Molar excess Gibbs free energy (x.) for the ternary mixture
l
E
H 0(1) + CH 0H(?) + C H OH(3) at 298.15 K. Computer interpolated G contours
2 3 2 S
1
at 0.05 kJ mol- intervals and the compositions x. of the experimental
~

mixtures, o , are shown. For values of x. refer to tables 1 and 2.

l

FIGURE 7. Computer generated projection of the GE(x.) surface for the

l

ternary system H 0(1) + CH 0H(2) + C H 0H(3) at 298.15 K. The maximum excursion

2 3 2 5
E 1
of G is shown in J mol- for each of the constituent binary mixtures.
 >----- -{~._}-!-_._.~~-_._ . _~-~.~
A B ~ ~-~- - - ~- ~ - - '- ~- -'-'l'- -'- -
M1 1/
--r--
p[2J M2 1/
--r--

Vent Vent
"Tl
G)
-'
V L1
rt
I Lz

~.-

,....",A,},)I, <,",I~I~

E
H

S1 S2 C1 C2
;,ssssss.~

Water thermostat Air thermostat Air thermostat

 7·
i
i
I
I
I
6
1
I

5~
I
II

4~ II
I

I
I
!
]
0
n.. 3r- I

-
~

.-
n..
II
I
I
I
II
i
2~
I
II
I
I

+ I
I
...-1 ..__ J .l -J- Il-----L..I__~. __.L_L . I
0'2 0'4 0·6 0·8 1
X1
FIG. 2.
3·0

i
i
2·5 -i

,
2·0 - -j

1- 0

0-5

o J-----_l_______L -l -l ~~L ..J

o 0·2 0'4 0·6 0'8 1

X1
FIG.3.
 15

10 ~
I

a
0-

-
~

Sf-
I
i

o _1 L .. _.L ..__ ..L__ .__. .J --.-.-L L

o 0·4 0·6 0·8 1
X1
FIG. 4.
1·5

1·0

1
0-5 ~

0
DO
1

j
I

l
0
I
0·2
I
0·4
I
0·6 0·8 1
X1
FIG.5.
 o

FIG.5.
 747 J mol- 1

320 J mol- 1

1
1
o
"/ .....-.....- - - - - - - - - - - - - - - - - 1:
- - /
o ... "1
1: ...
3

FIG. 7.
References

1. ARNlKAR, H.J.; RAG, T.S.; BGDHE, A.A~ J. Chern. Educ. 1970, 47, 826.

2. DOBSON, H.J.E. J. Chern. Soo. 1925, 127, 2866.

3. RAO, T.S.; BODHE, A.A.; GANDEE, B.R. J. Chromatog. 1971, 59, 151.

4. HALASZ, I. Anal. Chern. 1964, 36, 1428.

5. D'AVILA, S.G.; SILVA, R.S.F. J. Chern. Eng. Data, 1970, 15, 421.

6. JONES, C.A.; SHOENBORN, E.M.; COLBURN, A.P. Ind. Eng. Chern. 1943, 35, 666.

7. PEMBERTON, R.C.; MASH, C.J. J. Chern. Thermodynamics 1978, 10, 867.

8. SHAW, R.; BUTLER, J.A.V. Froc. Roy. Soc. (London), 1930, A129, 519.

9. LARKIN, J.A.; PEMBERTON, R.C. Nat. Phys. Lab. Report Chern. 43, 1976.

10. LARKIN, J.A. J. Chern. Thermodynamics 1975, 7, 137.

11. :MeGLASHAN, M.L.; TdILLIAlWSON, A.G. J. Chern. Eng. Data 1976, 21, 196.

12. BREDIG, G.; BAYER, R. Z. Fhys. Chern. 1927, A130, 1.

13. DULITSKAYA, K.A. Zhur. ObSDCh. Khim. 1945, 15, 9.

14. WREWSKY, M.S. Z8it. physik. Chern. 1912, 81, 1.

15. FERGUSON, J.B.; FUN1iELL, W.S. J. Phys. Chern. 1929, 33, 1.

16. BUTLER, J.A.V.; THOMSON, D.W.; MacLENNAN, W.H. J. Chern. Soc. 1933, 674.

17. MASH, C.J.; PEMBERTON, R.C. Unpublished work.

18. SCHMIDT, G.C. Zeit. physik. Chern. 1926, 121, 221; 1921, 99, 71.

19. BUTCHER, K.L.; ROBINSON, W.I. J. Appl. Chern. 1966, 16, 289.

20. MORRIS, A.C.; MUNN, L.T.; A~IDERSON, G. Can. J. Res. 1942, 20B, 207.

21. MASH, C .J.; ~rnERTON, R.C. Unpublished work.

22. HERINGTON, E.F.G. Nature, Lond. 1947, 160, 610.

23. lIERI NGTON , E.F.G. J. Appl. Chern. 1968, 18, 285.

24. GRISWOLD, J.j DINWIDDIE, J.A. Ind. Eng. Chern. 1942, 34, 1188.

25. CARLSON, C.S.j SMITH, F.V.; MORRELL~ C.E. Ind. Eng. Chern. 1954, 46, 350.

26. DELZE~[lli, A.O. Chern. Eng. Data Sere 1958, 3, 224.

27. GUPTA, S.K.; LESSLIE, R.D.; KING, A.D. J. Phys. Chern. 1973 1 77 1 20110

28. DYMOND, J .H. i SMITH, E.B. liThe virial coefficients of gase~l. Clarendon

Press: Oxford, 1969.

29. SMITH, N.O. J. Chern. Educ. 1963, 40, 317.

30. AMBROSE, D.; LAWRENSON, I.J. J. Chern. Thermodynamics 1972, 4, 755.

31. AMBROSE, D.; SPRAKE, C.H.S.; TOWNSEND, R. J. Chern. Thermodynamics

1975, 7, 185.
32. KUDCHADKER, A.P.; EUBANK, P.T. J. Chern. Eng. Data 1970, 15, 7.

33. KRETSCHMER, C.B.; WIEBE, R. J. Amer. Chern. Soc. 1954, 76, 2579.
34. KELL, G.S. J. Chern. Eng. data 1967, 12, 66.
35. HALES, J.L.; ELLENDER, J.H. J. Chern. Thermodynamics 1976, 8, 1177.
36. WILSON, G.M. J. Amer. Chern. Soc. 1964, 86, 127.
37. RENON, H.; PRAUSNlTZ, J.M. A. I. Ch. E.J. 1968, 14, 135.

38. ABRAMS, D.S.; PRAUSNITZ, J.M. A. I. Ch. E. J. 1975, 21, 116.

39. REDLICH, 0.; KISTER, A.T. Ind. Eng. Chern. 1948, 50, 345.
40. COUNSELL, J.F.; LEES, E.B. Unpublished work.