THE ELECTRICAL RESISTIVITY OF

SILVER, COPPER, ALUMINIUM, AND ZINC AS

A FUNCTION OF PURITY IN THE RANGE A29S0 K
F. PAWLEK and D. ROGALLA
Technical University of Berlin, Berlin. Germany

Received 18 Ma-v 1965

S I N c E the war there has been a considerable increase zinc at room temperature changes to aluminium.
in the consumption of pure oxygen in various pro- copper, silver, zinc at low temperatures, and the
cessesin the steel industry and this has led to significant curves cross each other several times. From the theory
advances in the field of cryogenic engineering. Not we would expect resistance ratios between room
only the production of pure oxygen from liquid air on temperature and 4.2” K of the order of 106. This is
a very large scale but also the generation of lower not achieved in reality and the measured deviation of
temperatures, as required in the liquefaction of five particularly pure metals is also shown in Figure 1
hydrogen and helium, have been realized on large
installations. Therefore, the way is now open to
operate sizeable machines and equipment at low tem-
peratures. The electrical industry is interested in
exploiting the hundred or thousand-fold increase in
electrical conductivity of metals in the vicinity of
absolute zero for transformers and generators. In the
course of this development it seemed desirable to
determine the low temperature electrical resistivity of
metals which are good conductors and may be pur-
chased at economical prices in large quantities and of
sufficient purity. The metals, silver, copper, aluminium,
and zinc have been chosen for the determination of
electrical resistance between 4 and 298” K, and par-
ticular attention was paid to the influence of impurity
elements and strong magnetic fields as they may arise
in the windings of transformers and generators.
The theory concerning the temperature dependence
of electrical resistivity is due to Griineisenr and IO
Bloch.2 A paper by MacDonald3 contains formulae
and tables which permit the calculation of the ideal
electrical resistance. The following values of Debye
temperatures 6’ and resistivities prevailing at these IO
temperatures have been used:

/ Ag ~ Cu Al Zn
~~-- I I
O(“K) ‘228~ 330 408 j 300
W cm) , 1.20 1 1.93 4.05 6.29
1 5 IO 50 100

T- ("K)
The calculations are plotted on Figure 1 and listed in
Table 1. (All resistivities are given in r&cm, i.e. Figure 1. Ideal resistivity of silver, copper, aluminium, and
zinc, according to the Bloch-Grtineisen formula and re-
10-g Rem.) The shape of the theoretical curves
sistance measurement of commercially purest silver,
reflect the influence of the Debye temperature. The aluminium, copper, zinc, and zone-refined copper
well known sequence of silver, copper, aluminium, samples

14 CRYOGENICS FEBRUARY 1966

 103
Silver Aluminium Copper wire bor Double electrolytically Double electrolytically
99.9: (Cu 01 refined copper rd zone refined copper
(Ag 4) (A (Cu 0)

I0
,I’ If
P

L --
f’
/

-
+In orgon. Ih o1 400’ C
r t 1
1+ lnargon al 600° C,cooling rol e-= 50degC/h 1
Oln orgon of 400’ C, cooling rote c!XdegC/h
A In argon 01 600’ C, cooling rote in on organ
fIlled quartz tube cooled by air
I I I I I I
I00 0 50 100 0 50 IOC

Table 1. Ideal resistivities according to the Bloch-Griineisen Figure 2. Influence of heat treatment on the electrical
formula and
- measured resistivities resistivity of silver, aluminium, and copper

T(-KI
I and Table I. The measurements were carried out on
2 mm diameter wires and potential drop measure-
coladared
.- ments were made with the aid of a very sensitive
Ag with 4.2 I .46 x IO-3 0.739 502
-c 36 p.p.m. 20.4 3.49 4.33 1 1.24 voltmeter. The chosen temperatures were the boiling
impurities 77 273 276 I.01 points of helium, hydrogen, and nitrogen, the sub-
(Ag 4) I95 1010 991 0.98
273 1 460 I 500 I .03 limation point of carbon dioxide, and the nielting
.-
Cu with 4.2 2.86 :x IO-’ 0.884 3 090 point of ice. While theory and experiment agree quite
< 57 p.p.m. 20.4 0.911 I .93 2.12 well at 77” K, the ratio of measured to ideal resistivity
impurities (Co 0) , 77 199 194 0.975
195 I 030 1010 0.980 at 4.2” K was found to be of the order of 103.
273
._
1 560 I 590 I .02 The Bloch-Griineisen theory is based on a perfect
CU with 2.86 x IO-’ 0.336 1 I 170 crystal lattice, while all practical materials contain
< 6 p.p.m. 2i.i 0.91 I I.11 I .22 impurity elements and physical lattice defects whose
impurities 77 199 192 0.964
(Cu 9.5) 195 1 030 996 0.966 influence on the resistivity becomes relatively large at
273 1 560 I 580 1.01
.- low temperatures as compared with the contribution
Al with 4.2 2.3 v IO-’ then made by thermal lattice vibrations. Foreign
-: 9 p.p.m. 20.4 0.664
impurities 77 232 atoms will scatter electrons only when they form part
(AI 99.999) 195 I 620 of the lattice. The solubility of impurities in metals is
273 2 520
._ generally temperature dependent and there exists,
Zn with 4.2 I.78 x IO-’ therefore, the possibility of precipitating some of the
< 41 p.p.m. 20.4 4.74
impurities 77 814 trace elements by suitable heat treatments and thereby
(Zn 4) 195 3 790 eliminating their contribution to the electrical re-
273 5660
- - sistivity. In general, physical defects may be removed

CRYOGENICS * FEBRUARY 1966 15

 Table 3. Impurity content and electrical resistivity of silver, copper, aluminium, and zinc with optimum heat treatment
T
!
Impuririrs (weight, p.p.m.) I p(nOcm) at T “K I
-- --
NO. S0lTlpll? A&t Cu/FeIMn1 Ni il I 20 Rem. EC, 1 4.2 1 20.4
-1. .-t - _. -~-- --
Ag 1 Silver c
&2 Electrolytically re-
fined silver
4 3 Chemically refined 322
silver
Ag4 99.999 % silver 376
-I_ .- - __
cu 01 Copper wire bar IIO.. 176
cu 0 Double electrolytic- 56 8 901
ally refined copper I I /
cu oooo Double electrolytic- IIO.. < 0.9 < om ’ <I2 I.11 2.42 771
ally refined and
electron beammeltec I
copper
cu 9.5 Double electrolytic- 3.1 0.6 I<O? 0.8 ~ / <6 0.336 I.11
ally and zone re
’ fined copper
.- -_ -~~ i I 1
Al 99.7 Aluminium 20110020/ !
Al 99.9 Pure aluminium 17 730 20
Al 99.99 Very pure aluminium 24 12
AlYY.999 Extra pure aluminiun 1 I .Y 2.4
__- -- / I I I
Zn 1 Zinc Unterharz 24.9
Zn 2 Zinc, Unterharz 25.8
Zn 3 Pure zinc, Unterharz 142
Zn 4 Very mre zinc, Unter 425 (3lOH
h&z’
Zn 5 ASARCO zinc 2 I25
Zn 7 Cominco electro-zinc 137
- - -

t Estimated value
 x

IO

-Double electrolytically
1 and zone refined copper (Cu 0000)
elecfrolyticolly refined copper
(Cu 0) (~57 w3.m.)
Double electrolyticolly ond zone
copper (Cu9.5)(<6p.p.m)

10-lI I I I I I I I
100 200 300 100 200 300
T- (“K) T--t (“K)

Figure 3. Influence of impurities on the resistivity of Figure 4. Influence of impurities on the resistivity of
silver. Annealing: 1 h in argon at 500” C. Cooling rate: copper. Annealing: 1 h in argon at 600” C. Cooling rate
~50 degC/h ~50 degC/h

by annealing at higher temperatures. The vacancies of grain boundaries is more noticeable in purer copper.
produced by this can subsequently be eliminated by Annealing at 400” C is then noticeably inferior to
slow cooling. Hence each form of heat treatment gives 600” C. In even purer copper the recrystallization
rise to two opposing effects on resistivity and it is temperature, and therefore also the temperature of
necessary to decide by experiment the optimum con- collective recrystallization, drops below 400” C. It is
ditions for minimum resistance. Figure 2 shows the then possible to detect the effect of vacancies produced
influence of heat treatment on the electrical resistance by increasing the cooling rate. As a result of these
of silver, copper, and aluminium for various annealing experiments the following annealing temperatures
temperatures and cooling rates. The metals used for were chosen: 500” C for silver; 600” C for copper;
this were already quite pure, so that the impurity 400” C for aluminium; and 200” C for zinc. The
concentrations were insufficient to produce precipita- following materials were available:
tion. In particular, it has been shown that 400” C is
too low an annealing temperature for silver and the (1) Silwr. Degussa, Frankfurt/Main, supplied four
elimination of dislocations requires temperatures of different brands of varying purity with works analysis.
500” C and more in order to obtain minimum resis- (2) Copper. Wire bar copper was derived from a
tance. This is accompanied by the appearance of section of an actual bar of unknown origin. Double
large grains of about 1 mm. Many grain boundaries, electrolytic copper was supplied by Elmore Metal1
which represent the accumulation of dislocations, are AG, Schladern/Sieg. This material was used as
then eliminated. This difference is very small in the received and drawn into wire or alternatively re-
case of pure aluminium annealed at 300 and 400” C. melted either in an electron beam furnace or by ten
The difference is also small in the case of the copper passes of zone refining in a graphite boat. The mass
wire bar, which has a relatively high impurity con- spectrographic analysis of these samples was kindly
centration when comparing annealing treatments at provided by British Insulated Callender’s Cables
400 and 600” C with fast or slow cooling. The influence Limited, London.

CRYOGENICS . FEBRUARY 1966 17

6
Table 3. Atomic resistivity increments at room temperature
and 4.2” K

11 Impurirbs ‘-

at T = 273” K (I, T = 4.2’ K

Al I .9s
0.077
,“u 7,3
Cd 0.38
CU
Mll ;:y (0)
Pb 4.65
Pd 044
1.6
4.36 -Aluminium (Al 99,7)(<1860 p.p.m.1
-- --

0.2 0.1 (4
0.55 0.5 (4
6.9 54: 6.7 (e): (f)
9.3 3.7; 12.2 (P):(f)
I .4 I .4 (/I)
3.7 4.0: 3.0 (6): (0
I.1 (6) 1.8; 0.88 (/I); (i)
I.1 I .4 (4
5.5 4.4 (9 Very pure aluminium
3.1 2.9 (6) (Al 99.99)( 56 p.p.m)
3.1 2.8: 2.7 (j): (i)
I I.3 0.7 (k)
--

Al 0.8 I .22 I I
6.7
(c)
I .4
(0
, L- ’ pure aluminium
Extra
0.8 I .02

zn
I_ 0.4 0.5 (Al 99.999)(9 p.p.m.) I c

(0.36)
4.2

lo-tl I00
CU
Fe (:.A, WI)
Ni (4.1)
Pb 20
Sn (II) 200 300
- T- (‘lo
Figure 5. Influence of impurities on the resistivity of alu-
minium. Annealing: aluminium 99.7,99.9, and 99.99, 1 h in
argon at 300” C. Cooling rate ~50 degC/h
REFERENCES IN TABLE 3
(3) Aluminium. Commercial aluminium, pure alum-
(a) LANDOLT-BORNSTEIN. ‘Znhlenwerre wxi Funktionen: II Band. inium, and very pure aluminium were supplied, with
6 Teil, Elektrische Eigensch~ften I.’ p. 22 (Springer, Berlin, works analysis, by Vereinigte Aluminiumwerke AG,
1959) Bonn. The extra pure sample was received, with works
(6) PAWLEK, F., and REICHEL, K. Mefall. 12, 1 (1958) analysis, from Aluminium-Htitte Rheinfelden GmbH,
(c) VASSEL, K. R. Aluminirrm 33, 781 (1957)
(d) MACDONALD, D. K. C.. and PEARSON, W. B. Acta metal/. 3.
Rheinfelden.
459 (1955) (4) Zinc. Unterharzer Berg- und Htittenwerke
(e) JACOBS, L. S., and SCHMITT, R. W. Phys. Rev. 113,459 (1959) GmbH, Zinkhiitte Harlingerode, supplied four types
(f) CHRISTENSON. E. L.. and DOMENICALI. C. A. J. oval.
II Phvs.
_ of zinc with works analysis. Very pure zinc, with
32, 2450 (1961) analysis, which originated from the American Smelt-
(g) FRANCK, J. P., MANCHESTER, F. D., and MARTIN, D. L.
Proc. roy. Sot. A263, 494 (1961) ing and Refining Co. (ASARCO) was made available
(h) MACDONALD, D. K. C., and PEARSON, W. B. Acta metal/. 3. by ‘Berzelius’ Metallhiitten-GmbH. Electrolytic zinc
403 (1955) from the Consolidated Mining and Smelting Co. of
(i) KROPSCHOTT, R. H., GARBER, M., and BLATT, F. J. Phys. Canada Ltd (Cominco) was also available.
Rev. Letf. 2, 91 (1959)
(j) MACDONALD, D. K. C., and PEARSON, W. B. P~oc. ray.
The analytical information and the results of
Sot. A219, 373 (1953) measurements are shown in Table 2 and plotted in
(k) LINDE, J. O., BACKLUND, N., and HUMBLE, S. G. Proc. Figures 3-6. From these arise the following points.
Tenth in,. Congr. Refrker., Copenhqen 1, 66 (1960) (1) Siloer (Figure 3). The electrical behaviour of the
(I) CARON, M. ‘ConductibilitP docu-
Plecfrique . . .’ Service de various brands is similar down to about 50” K, but
mentation et d’information technique de l’aeronautique,
Paris 1957 II, II 58, S.4 below this temperature the resistivity curves split up
(m) DESALVO, A., GONDI, P.. LEVI, F. A., and ZIGNANI, F. Met. and at 4.2” K the resistances differ by a factor of 10.
Iral. 55, 511 (1963) Because of the lack of sensitivity of the spectro-
graphic method of impurity analysis it is not possible
(The values for zinc have been calculated for an impurity to explain why two samples of apparently similar
concentration of 10 p.p.m. (at.) ) impurity content differ in resistivity by a factor of 2.

18 CRYOGENICS . FEBRUARY 1966

 reached so that the given impurity concentrations can
only be approximate values. Nevertheless, the indus-
trially refined purest zinc shows, at 4.2” K, a resistivity
which is only five times as large as that of double
electrolytically refined copper.
Taking the whole set of results, we conclude that,
given. optimum heat treatment of the metals, their
impurity concentrations exert a strong influence on
the resistivity-temperature relationship below 60” K.
It was of interest to see whether the magnitude of this
effect could be estimated from the analytical informa-
Zinc, Unterharz (Zn 2)(2970 p.p.m. tion and the known values of atomic resistance
I I I increments for the various impurity elements. Table 3
c,Unterhorz (Znlj(6730 p.p.m.1 shows the atomic resistance increment of the various
I
impurity elements of the four metals for room tem-
--. ASARCO zinc (Zn 5)(29 p.pm.) perature and 4.2” K as determined in the past by
into elecIrozinc (Zn 7)(40 p.p.m.) experiment. The first point to note is that Mathiessen’s
‘Pure zinc,Unterhorz (Zn 3)(43 p.p.m.1 rule is not obeyed at low temperatures, and this has
I--) Pure zinc,Unterharr (Zn4)(41 p.p.m.) 1 already been pointed out by Griineisen and Goens’and
Justi. An attempt was then made to calculate the
resistance increment due to the impurities. The
measured resistivities for each of the purest metals
have been plotted in Figure 1. Table 4 shows the
differences between ideal and measured resistivity
and this is then compared with the atomic resistance
increment calculated in accordance with the published
figures. Besiring in mind the considerable uncertainties
--
IOU 200 inherent in the chemical analysis, there appears to
300
T- (OKI be a satisfactory agreement in order of magnitude.
In addition, we conclude that the influence of physical
Figure 6. Influence ofimpurities on the resistivity of
zinc. Annealing: 1 h in argon at 200” C. Cooling rate lattice defects is already very small in silver and
~50 degC/h copper, but a certain improvement in aluminium by
special heat treatment seems possible. It seems
(2) Copper (Figure 4). Differences in the various appropriate to mention that the processing of con-
samples are already apparent below 70” K. It is not ductor materials for transformers and generators
surprising that ordinary wire bar copper, with its must inevitably involve deformation and, therefore,
relatively large impurity content, shows a rather the properties of single crystals are of little significance
high resistance at 4.2” K. It is remarkable that the for practical applications.
sample which was re-melted in an electron beam In these applications the conductors are used in
furnace, in order to increase the purity further, was current-carrying windings for producing magnetic
found to be worse than the starting material. This fields. At low temperatures the presence of the
effect is the result of the reduction in oxygen content magnetic field will increase the electrical resistivity.
in the electrolytic copper which occurs during electron
beam melting. Similar effects have previously been Table 4. Comparison of calculated and measured electrical
reported by Sekula,4 and Dolecek and Schu1tz.s The resistivities of four pure metals
zone-refined copper curve is shown for purposes of
P(dl)Cln
comparison; apart from this the metals investigated NO. TPKJ
can all be purchased on a large scale. I I Ideal I Measured I PmeasuredlPideal Pchemlcdt
I-----

(3) Aluminium (Figure 5). Here the resistance h4 4.2 OGO146 0.734 0.733 3.0
ranking corresponds to the degree of purity and the --
20.4 3.49 4.33 0.84

value for the purest aluminium, which was produced Al 4.2 0+0023 I.11 1.11 0.40
99.999 20.4 0654 2.38 1.72
by a special electrolytic process,6 is almost as low as
that for double electrolytically refined copper.
(4) Zinc (Figure 6). The curve for the ultra-pure
zinc (Zn 4) (41 p.p.m.) indicates a measurement error Zn 4 4.2

at 20” K, and the broken line is likely to be a better

representation of the true resistivity curve. Here t Calculated from the atomic resistivity increment and the impurity con-
again, the sensitivity limit of analytical methods was centration.

CRYOGENICS . FEBRUARY 1966 19

Magnetoresistance increments in a field of 18 kG, 225 for Ag 4 at 20.4” K and H = 0. The calculation
as measured by Ltithi,g and our results for zero-field is based on the following prices (in DM/kg): alum-
are collected in Table 5 for the four purest metal inium, 6; copper, 7.8; silver, 190. Therefore extra
samples. Under these conditions super-pure alum- pure aluminium with a total impurity content of less
inium (Al 99.999) gives the lowest resistivity, with than 10 p.p.m. seems to be the most suitable conductor
copper and silver lying above this value. Zinc shows material for technical low temperature applications.
an extremely large magnetoresistance which rules it We have not been able to measure resistivities in
out for practical applications. the presence of alternating currents.
The relative prices for 1 m conductors of equal
resistivity are 1 for Al 99.999; 3.4 for Cu 0; and We would like to thank the Low Temperature
Laboratory of the Physikalisch-Technischen Bunde-
Table 5. Electrical resistivity of the purest commercial samples sanstalt for carrying out these measurements. We are
in a transverse magnetic field? indebted to the Stifterverband NE-Metalle Dusseldorf
for a grant which made this investigation possible,
and finally we would like to think the various indus-
TT trial concerns which provided samples with analytical
information.

44 4.2 ~ 0.734 8.8

20.4 , 4.33 , IS REFERENCES

cu 0 4.2 ~ 0.884 / 1. GRONEISEN, E. Anrr. Phys. Lpz. 16, 530 (1933)

20.4 I 1.93 IO 2. BLOCH, F. Z. P/y. 52, 555 (1928); 59, 208 (1930)
-1 I 3. MACDONALD, D. K. C. ‘Electrical Conductivity of Metals
Al 99.999 ’ 4.2 I.1 1 2.7 and Alloys at Low Temperatures.’ Handbuch der Physik
20.4 2.38 ~ 5.7 Vol. XIV, p. 127 (Springer, Berlin, 1956)
.___ ~- 4. SEKULA, S. T. Phys. Rev. Lett. 3, 416 (1959)
Zn 4 4.2 3.81 650 5. DOLECEK, R. L., and SCHULTZ, D. J. Acta Metall. 8. 664
I
20.4 (20) 260 (1960)
I 6. SCHM~, H., and KOCH, W. Ermerall. 11, 427 (1958)
7. GRUNEISEN, E., and GOENS, E. Z. Phys. 44, 615 (1927)
t Calculated from the magnetoresistance measurements by 8. JUSTI, E. Z. Mefall. 51, 1 (1960)
LtithP and our measurements in zero magnetic fields. 9. L~~THI, B. He/u. phys. octn 33, 161 (1960)

20 CRYOGENICS - FEBRUARY 1966