JOURNAL OF APPLIED ELECTROCHEMISTRY 9 (1979) 721-730

The electrodeposition of zinc from acidified zinc

sulphate solution
I. W. W A R K
CSIRO Division of Mineral Chemistry, P.O. Box 124, Port Melbourne, Vic. 3207, Australia

Received 31 October 1978

In the electrodeposition of zinc from acidified zinc sulphate solutions the loss in current efficiency (CE)
due to evolution of hydrogen at the cathode has often been attributed to the presence of metal im-
purities, and it has been thought that in a pure solution the CE would be 100%. In this paper it is argued
that, whatever the degree of purity achieved, hydrogen must always be produced simultaneously with
zinc. At the start of electrolysis the CE is determined solely by the zinc/acid ratio, but when impurities
are present the CE falls progressively as electrodeposition proceeds. Further, colloid silica which is
always present in plant solutions is not responsible for the CE falling below 100%, nor are the lead
anodes that are used in practice. Some effects of manganese, both at the anode and the cathode, are
reported. The deleterious influences of cobalt and manganese together are also discussed.

1. Introduction hydrogen [3], which strengthens the case for its

acceptance.
In t964 a relationship between the current effic- The value of k was almost independent of tem-
iency and the composition of the solution was perature, current density, and deposition period
formulated [ 1], for the electrolytic zinc process within the probable limits of operations for a com-
mercial plant. Current efficiency (CE)for the
zinc deposited
experimental conditions most used by us t was
hydrogen formed 97.3 -+ 0.3%; thus, about 2-5% of the.current pro-
1 zinc sulphate concentration duced hydrogen.
k sulphuric acid concentration" It has been widely thought that the results
from still more highly purified solutions might not
This was a semi-empirical or intuitive relationship conform to this equation, and indeed that current
based on experimental work with the highly efficiencies would then be substantially higher
purified solutions to be described later.* until finally, with spectroscopic purity, only zinc
G. C. Bratt [2] has provided a theoretical basis would be produced. Tainton [4] was the first to
for this relationship, which he calls 'Wark's Rule'. advance this view; it has never been explicitly
Bratt states that a similar relationship has been refuted.
found to hold for co-deposition of nickel and The main purpose of the present paper is to
* The research on which this relationship was based was demonstrate that, although experimental work
conducted during 1926-29 for the Electrolytic Zinc with more highly refined solutions is certainly
Company of Australasia under the general direction of
Sir David Orme Masson, with E. E. Jones as co-worker warranted, there is a considerable body of evi-
for a few months, and then with H. P. Matthews as dence from several hundred experiments that for
collaborator. Company policy precluded publication at the conditions mentioned above the current effic-
the time, but subsequently permission to publish certain
aspects has been granted. The present paper is based on iencies can never exceed the values shown in Table
the experimental work of Matthews and the author. 1. They may, however, approach them more
t These were: temperature 35~ C; current density 92 mA closely in operating plants as control measures
cm-2 ; duration of deposition 1 h; [ZnSO4 ] = 0.9 M,
[H2SQ ] = 0.6 M. The experimental technique was des- improve.
cribed earlier [1]. Table 1, which is taken from [ 1], indicates the

0021-891X/79/060721-10 $03.00/0 9 t979 Chapman and Hall Ltd. 721

722 I.W. WARK

Table 1. Current efficiency versus acidity

ZnSO JH=S0 4 CE calculated CE found

(molecular ratio) (k = 30)
1.5 M total sulphate 2.125 M total sulphate

5 99.3 99.2 98-9

2 98.3 98.2 98.0
1 96-8 96.4 96.5
0.5 93-7 93-0 93-3
0-2 85 -7 85 -2 85-8
0.09 73.2 (70)

degree of reliability of the Wark Rule and empha- to the normal salt by the acid condition of the
sizes that the current efficiency is dependent pri- electrolytic cell.
marily on the zinc sulphate/sulphuric acid ratio. It is pertinent to enquire whether these small
amounts of metallic impurities or perhaps some
2. Experimental unidentified impurity could have been responsible
for the failure to obtain current efficiencies of
The solutions used, possibly the purest ever ob- 100% in the results reported in Table 1.
tained in quantity, were prepared by E. E. Jones It would appear that any suspect impurity
who followed in general the purification steps then would, like cobalt, be'deposited progressively" on
used in the commercial plant. Electrolytic zinc the zinc cathode. Its concentration in the solution
was dissolved in C.P. contact sulphuric acid. The would therefore decrease with time. As a conse-
solution was heated with excess zinc dust for self quence, if the electrolyte (with adjustment for
purification; basic zinc sulphates form during this concentration changes) were used with a fresh
treatment and these and/or zinc hydroxide, in cathode for a second and a third time, increases
separating out, remove iron, arsenic, etc. The solu- in current efficiency could be expected. No such
tion was f'fltered hot. Addition of silver sulphate, effects were ever observed in the work under
followed by filtration, removed chloride. Cobalt discussion. The bulk solutions were used re-
was removed by the arsenic purification process, peatedly without improvement or deterioration,
namely treatment of the hot solution with zinc as Table 3 shows in two instances (tests 87 and
dust in the presence of copper sulphate and 88, and tests 89 and 90).
sodium arsenite, followed by filtration. The
Table 2. Stock solutions o f zinc sulphate
Filtrate was zinc dust purified to remove possible
traces of arsenic, copper and cadmium. The final
Component First batch Second batch
'highly purified solution' was obtained by filtra- (rag 1-1) (mg 1-~)
tion of this solution after cooling to 25 ~ C and air
agitation; a considerable amount of basic zinc Zn 9.77 x 10~ 1.012 x lOs
sulphate is removed during this treatment, and 0.5 0.6
Mn 0.6
iron in solution is reduced to a very low concen-
Co nil < 0.1
tration. Cu < 0.02 < 0-05
Cd possible trace < 0-2
3. Results and discussion Ni nil < 0-2
Fe nil < 0-t
The assays, taken from the earlier paper [1], are C1 trace < 0-5
As < 0.02 < 0.1
reproduced in Table 2; they show, in addition Sb < 0.02 < 0-05
to minor traces of metallic elements, a significant Pb < 0.2 < 0.6
concentration of silica, to which reference is made A1 - < 0-5
later. They also indicate a low concentration of SiO~ 15 15
SO4 1.430 • l0 s 1489 X l0 s
dissolved basic sulphate which would be converted
THE ELECTRODEPOSITION OF ZINC FROM ACIDIFIED ZINC SULPHATE SOLUTION 723

Table 3. Influence o f current density on electrodeposition o f zinc (ZnS04 = 0.9 M," H2SO 4 = 0.6 M; 35 ~ C; duration
1 h]

Test Cathode Anode Nature o f Solution Cathode Current Current

number potential* potential* cathode current efficiency efficiency
(V) (V) density . (%) corrected for
(mA cm-2) re-solution
(%)

87 1.058-1.066 1-860-1.859 standard fresh 92 97.6 97.7

88 1.056-1.064 1.856--1.853 from 87 from 87 86 97-6 97.7
28 1-053 - standard fresh 31.8 97.4 97.7
31 1.050 1.814 standard fresh 31-8 97.4 97.7
89 1-048 1.754 standard fresh 9.5 96.9 97.9
90 1.050 1.758 from 89 from 89 9-3 96.1 97.2
95 1.043 1-691 from 88 fresh 2.9 90-5 (94.4)?
96 1-046 1-648 from 95 fresh 0.95 87.0 97.8
open 4.4 mg
0.987 1.45-1.37 from 96 ,from 96 circuit loss
inlh

* Against saturated calomel electrode at room temperature (21 ~ C -+ 30 C).

It is well known in the industry that higher [ 1] show conclusively. It appears, therefore, that
concentrations of impurities can be tolerated with cobalt could not have been responsible for the
higher current densities. In fact the Tainton version failure of electrodeposition to yield 100% current
of the process was partly based on this fact. There- efficiencies.
fore, had the value of less than 100% CE of our
'pure' solutions been due to the presence of un- 3.1. E f f e c t o f manganese
identified metal impurities a substantial increase
in current density should have raised the current Current plant-purification procedures do not
efficiency. While there was some slight improve- remove manganese from the feed solutions, which
ment with current density, it is significant (see contain several grams of manganese per litre.
Table 3) that a maximum value o f CE had been Plant operators are well aware that simultaneously
reached at a relatively low current density. with oxygen evolution manganese ions may be
There was some corrosion of cathode zinc oxidized at the anode, either to permanganate and
deposits wh6n they were left on open circuit for other manganates or to a more or less adherent
an hour in the electrolyte. The cathodic reactions layer of oxide on the lead anodes. There is also a
associated with the corrosion could be hydrogen possibility of alternating oxidation at the anode
evolution or the reduction of an impurity (prob- and reduction at the cathode with loss of electrical
ably dissolve d oxygen). If the current efficiencies efficiency.
(Table 3) are Corrected for the dissolution of zinc With up to 20 g-i Mn it was found that, as for
in open circuit, there appears to be little effect of cobalt, the current efficiency is not initially dimin-
current density on current efficiency. The magni- ished. However, the efficiency deteriorates sub-
tude o f the corrosion is of the order to be stantially over a period of hours: simultaneously
expected for the diffusion-controlled reduction of the permanganate colour may develop, and small
dissolved oxygen. particles of oxide, becoming detached from the
Still more conclusive evidence came from tests anode, become suspended in the solution and may
[1] when cobalt sulphate was added deliberately settle on the cathode. An attempt was made to
to the stock dlectrolyte. Cobalt did have a delayed determine the cause o f the loss, which might have
I
disastrous effect on current efficiency, but there been due to (a) deleterious changes in the ionic
was no increased hydrogen evolution at the com- state of the manganese similar to those for cobalt
i
mencement of electrolysis, as the curves of paper [1], (b) deterioration of the cathode due to
 t~
4~

Table 4. Influence o f additions o f manganese f r o m various sources or o f MnO 2 (tests o f 6 h duration)

Mn form CE test Initial Mn Final Mn Mean H2SO 4 Mean ZnSO 4 CE A n o d e potential Mean Cathode current
and source number (g 1-1) (g 1-1) (M) (M) (%) (v) temperature density
Cc) (mA cm-2)

no Mn 125 nil nil 0.61 0.88 97-0 (1.815)* 34-95 29.3

MnSO4, Hopkin 126 20-0 19.0 0.61 0.89 93.6 1.82 34-95 29-6
and Williams
M n S Q , Merck 127 20.0 19.6 0.61 0-89 94.2 1-83 34-95 29.1
M n S Q , from 135 18.7 18.2 0.61 0.88 93.4 1-80 35.05 29.7
Merck KMnO 4
added 0.6 g 128 nil nil 0-61 0.89 93.1 1-78 35-0 30.0
MnO2 from
Hopkin and
Williams ?
added 2-0 g 131, 2, 3 nil nil 0.61 0.89 91-7 1.78 35-0 29.6
MnO~ from
Merck KMnO4 $

* The usual value for these anodes is rather lower [cf. 1.78 V in test 128 (without MnSO4)]
t This MnO 2 was collected from tests 114 and 126
Mean of three consecutive tests over 1--4 h.

>
 t-

0
r
9

Table 5. Influence o f manganese on electrodeposition o f zinc (1 h test#

Z
9
Mn CE CE Cathode Anode Mean Cathode Cathode Solution Final Final N
concentration test (%) potential* potential* temperature current Mn acidity Z
(g 1-1) number (V) (V) density concentration at 35 ~ C
(mA cm -=) (gl -l)
9
0 107 97.5 1.046 1.804 35,0 31-8 fresh fresh 0-63 M N
m
1087 96.8 1.042 1.795 34.85 31.2 from 107 from 107 0-68 M >

5 110 97.0 1.046 1.63-1-89 35.1 30.2 fresh fresh 4-56

1117 97.1 1.050 1.63-1-79 35.0 29.4 from 110 from 110 4.41 0"65 M
filtered t~

2O 112 97-4 1-054 1-66-1-76 35.0 29.1 fresh fresh 19,1 0-63 M N
113"~ 96.9 1-054 1-69-1-80 35.0 29.6 from 112 from 112 18,8 0-64 M z
filtered r~3

* Against saturated calomel electrode at room temperature.

These tests were second 1 h tests with cathode and solution from the preceding test.

r~
9
t~

9
Z

to
726 I.W. WARK

Table 6. Effect o f manganese; deterioration with time

CE CE Anode Mean Duration Solution Cathode

test (%) potential temperature (h)
number (V)

128 93-1 1.783 35.0 6 fre~ ~e~

129 91.9 1.785 35.0 1 ~om 128 ~e~
130 96.5 1-787 35.05 1 from 129 ~e~
filtered

adhering oxide particles, or (c) alternate oxidation the 6 h periods of Table 4 current efficiencies
to manganate or manganese (III) at the anode and were already seriously reduced. The electrolyte
reduction to manganese (II) ions at the cathode. from test 128 of Table 4 containing only sus-
To determine the significance of the last- pended MnO2, when used with a fresh cathode in
mentioned possibility the technique of Buttinelli test 129, gave a current efficiency in a 1 h test of
and Iacchelli [5] could be employed, namely to only 91 "9%. Comparing this with the results of
measure the amount of hydrogen produced at the Table 5 it is apparent that any inroads on CE are
cathode. If this, added to the zinc produced, falls made by the oxidation products of manganese
short of 100% in current efficiency then changes rather than dissolved manganous ion.
in the ionic condition of the manganese must be The influence of suspended MnO2 is further
held responsible. Another useful technique would confirmed by the fact that after filtration the
be to use a two-component electrolysis cell [1], solution from test 129 recovered to give a current
determining the manganate concentrations in both efficiency of 96.5% (see Table 6).
anolyte and catholyte compartments. Deposition There is little or no permanent deterioration in
of manganese metal with the zinc must be ruled the cathodically deposited zinc from prior contact
out because of the very low manganese content with manganese oxide, as is shown in Table 7.
in zinc cathodes. In plant practice glue is usually added to mini-
Care was taken to ensure that troublesome mize the loss in current efficiency due to undesir-
impurities were not introduced with the manga- able impurities. Its marked influence for cobalt-
nese. Three different samples, one starting from bearing solutions has been reported [1] : glue
permanganate, were prepared. Table 4 shows that delayed but did not prevent a fast-accelerating
over a period of 6 h they behaved similarly. It also deterioration in CE. A similar situation arises for
proves that manganese dioxide in suspension, even manganese, as is shown in Table 8: the lower than
in the absence of dissolved manganese, caused normal current efficiencies were due to use of a
serious losses. higher acid concentration. Because of the manga-
In Table 5 the results for two different con- nese dioxide carried by the electrolyte the current
centrations of manganese are compared with the efficiencies cannot be duplicated with the same
quite normal results for pure solutions (tests 107 precision as in 'pure' solutions or in solutions con-
and 108). For these tests of short duration the taining cobalt as the only added impurity.
manganese had little or no effect, whereas with Cathode potentials (routine measurements
Table 7. Cathode or solution effect

CE Mean Cathode Solution CE Anode Mean

test Mn (%) potential acid
number (g1-1) (V) at 35 ~ C

119 18.5 seven th hour's use 94-3 - 0-64 M

120 18-6 fresh from 119 94.0 1.83 0.64 M
121 nil from 119 fresh 97"1 1.84 0"61 M
'pure'
THE ELECTRODEPOSITION OF ZINC FROM ACIDIFIED ZINC SULPHATE SOLUTION 727

Table 8. Effect o f glue in restraining manganese in electrodeposition o f zinc

Test Glue Mn CE Cathode Anode Duration Deposit

number (dissolved) (%) potential potential o f test
(g 1-~) (V) (V)* (h)
Left Right

291 nil nil 96-5 1,038 1.85-1.83 1.85-1-83 24 a small number of nodules
had begun to form; shining
surface.
295 5 mg 1-1 nil 96.3 1.047 1.84-1.82 1-84-1-82 24 typical smooth deposit with
+ 1 mg h -1 a few nodules; duller than
test 291.
296 1ill 2.0-1-93 75.1 1.033 1.73-1.75 1.82-1.83 15 marked edge growths and
nodules; dull and even; few
holes.
288t 5 mg 1-1 2.0-1.63 88.0 1-044 2.39-1.84 2-46-1-97 24 very highly vegetated at the
+ 1 m g h -1 edge and even in the centre;
would give very low melting
efficiency.

* The cathode was placed centrally between two anodes, referred to here as left and right.
? The high anode potential of test 288 was due to the anodes having stood out of action for some time prior to the test.

Fig. 1. (a) Zinc deposit from highly purified solution: rising hydrogen bubbles have caused the striations; (b) fine-
grained deposit from highly purified solution to which glue has been added; (c) front and (d) reverse views of zinc
from cobalt-bearing solution; the striations of the reverse side are due to use of a roughened aluminium anode; note
that corrosion has already commenced in the typical 'cobalt holes'; (e) typical deposit from a manganese-bearing
solution (X 7-2)
728 I.W. WARK

against the saturated calomel electrode) are re- series except when manganese was present. (Lead
ported for this series of tests because they show anodes tend to become passive when not in con-
that glue, which improved current efficiency, may tinuous use, sometimes to the extent o f several
slightly reduce energy efficiency by raising the volts of overvoltage. They return gradually to
cell voltage. The final column reports on the con- normal after a few days, and in the meantime in
dition of the deposited zinc which was markedly pure solution have negligible effect on current
dependent on the solution composition and the efficiency.) In the presence of manganese the
particular impurities present. All our deposits were establishment of relatively stable conditions is
examined microscopically, and Fig. 1 shows slow. The solution may become clouded with sus-
characteristic photographs for some experimental pensions of varying degrees of fineness; it may
conditions. Each element has a characteristic develop the permanganate colour; sizeable chips
effect on the morphology of the deposit, but it is of oxide may break away from the anode. These
not always possible to identify it by examining chips may contain some cobalt oxide; at any rate
deposits formed in the complex solutions of prac- the two elements together can be much more
tice. Antimony in sufficient amounts gives rise to harmful than either alone. Once again the current
a cauliflower effect, selenium to a smooth deposit efficiency of the first few minutes is unaffected.
to which hydrogen bubbles adhere and glide Table 9 records the results from a typical series
slowly upwards. of tests of 5 h duration, indicating how toxic
cobalt can be with anodes from solutions con-
3.2. Anode effects in the presence o f taining both cobalt and manganese, and also how
manganese and cobalt important is voluntary or involuntary disturbance
of the MnO2 deposit.
When cobalt and manganese were both introduced The high CE (96-7%) of the 5 h test 285, with
the results could not be quantitatively predicted a pure solution but with anodes aged in an elec-
from the results for each separately. Dominating trolyte containing both cobalt and manganese,
factors were the prior history of the lead anodes illustrates that the current efficiency of zinc depo-
and the condition of the anode scale. Routine sition is at first unaffected by the condition of the
measurements of the anode potential, accurate to anodes. Later as the impurities passed into solution
1 mV, were found to vary little during the progress or suspension, a marked loss would have occurred.
of electrodeposition for any test within the whole The anodes themselves require several days to
Table 9. Influence of anode condition on current efficiency and nature of deposit (tests of 5 h duration)

Test Anodes Solution Mn elimination, CE Zinc deposit

number first 2 h (%)
(g 1-1 h-l)

285 with heavy scale from standard zinc and - 96.7 no visible
Co-Mn solution acid concentrations, 'Co holes'
Co and Mn-free
286 aged in manganese-free standard 89-8 minor cobalt
cobalt-bearing solution plus 100 mg F 1 Co, re-solution
Mn-free holes
281 aged and scaled in as for 286 0-035 52.0 heavy
Co-Mn solution plus 2 g 1-1 Mn re-solution
282 as for 281 as for 286 0-17 62-3 heavy
plus 5 g 1-1 Mn re-solution
283 from 281, as for 286 - 53.2 heavy
scale scraped re-~olution
284 from 282, as for 286 - 70.2 heavy
heavy scale re-solution
THE ELECTRODEPOSITION OF ZINC FROM ACIDIFIED ZINC SULPHATE SOLUTION 729

attain a steady condition in manganese-bearing While 15 mg 1-1 SiO2 had no more influence on
solutions, a fact noted also by Buttinelli and the current efficiencies of pure solutions than
Iacchelli [5]. 2 mg 1-1 , silica has a marked effect in reducing the
While discussing the influence of the condition ravages of cobalt and manganese. Table 10 com-
of lead and lead-coated anodes it is pertinent to pares the effects o f sodium silicate, zinc silicate
ask whether the current efficiencies lower than and glue on a solution containing 20 mg 1-1 Co
100% could have been caused by their use. It was and 2 g 1-1 Mn that had previously been rendered
found, however, that substitution for them of toxic by prolonged electrolysis.
platinum black anodes left the current efficiencies The zinc salt seems, on this limited series of
practically unaffected. tests, to be a better restraining agent than the
sodium, but neither is as effective as glue. A con-
3.3. E f f e c t o f silica clusion to be drawn is that it is unlikely that the
15 mg 1-1 SiO2 in our pure solutions could have
Table 2 showed that the 'pure' solutions used con- been responsible for their falling short of 100% in
tained 15 mg 1-1 SiO2, presumably in colloid form current efficiency.
derived from the ore and the glass vessels used. In One other possibility remains. It will be seen
commercial plants the concentration may reach that during the purification sequence our two
100 mg 1-1 . It is well known that some colloids stock solutions were treated with zinc dust. It
exert a restraining influence on such impurities as might therefore be suspected that they had carried
cobalt (see for example [6, 7] ). Thus, the silica forward some degree of the cobalt 'devilment'
in these solutions may have raised (but surely not factor ( [ 1 ] , p. 897) into the electrolytic cell. How-
lowered) current efficiency. A comparison was ever, test 52 of Table 11, for which the electrolyte
made between the stock solution and another pre- had not been pre-treated in this manner, gave pre-
pared from highly purified zinc sulphate crystals cisely the same result as test 34, which had.
and sulphuric acid, containing only 2 mg 1-1 SiO2. Despite the passage of 50 years since the pro-
A higher than usual acid/zinc ratio led to lower cess was introduced, it was still so little under-
current efficiencies than normal, but in agreement stood in 1969 that a disastrous situation could
with Table 1. arise at Risdon [8], where there were periods of

Table 10. Effect o f silica on electrolysis o f Co-Mn solutions

Test Addition agent and amount Duration CE

number (h) (%)

341 nil 5 78
340 glue, 10 mg 1-1 7 94.1
342 SiO2, 20 mg V1 as sodium salt 5 85-1
345 SiO~, 100 mg 1-1 as sodium salt 5 83-6
344 SiO~, 1000 mg 1-1 as sodium salt 5 89.7
351 SiO2, 100 mg F 1 as zinc salt 5 88.3
346 SiO2, 1000 mg 1-1 as zinc salt 5 92.0

Table 11. Influence o f dissolved silica on current efficiency (ZnSO 4 = 0.25 M; H2SO 4 =
1.25 M; duration 1 h)

Test Silica content Current Solution

number (mg 1-x ) efficiency
(%)

34 15 84-8 'pure' (bulk solution)

46 15 85.5 'pure' (bulk solution)
52 2 84-8 from ZnSO4 crystals
730 I.W. WARK

'catastrophic unexplained collapses' in current help of Dr D.F.A. Koch, Chief of the Division of
efficiency. At times, more zinc was dissolving than Mineral Chemistry, in the preparation of this
was being deposited: the trouble was traced to paper.
nickel and chromium, and efficiencies then im-
proved. A continuation of the type of work References
reported here seems desirable for such elements as
nickel and chromium, and a reversion to X-ray [1] I.W. Wark, Proceedings of the First Australian
analysis of the deposited zinc [1] as a means of Conference on Electrochemistry, (Edited by
J. A. Friend and F. Gutman), Sydney and
early detection of obtruding impurities. Hobart (1963), Pergamon Press, London,
p. 889.
4. Conclusions [2] G.C. Bratt, in Tasmania Conference 1977,
Australasian Institute of Mining and Metallurgy,
Melbourne (1977) p. 277.
Summarizing, it is concluded that at the com- [3] V.L. Kheffets and L. S. Reyshakhnit, Uch. Zap.
mencement of electroylsis the current efficiencies Leningr. Gos. Univ., Ser. Khim. Nauk 272(18)
(1959) 40.
with respect to hydrogen and zinc deposition are [4] U.C. Tainton, Trans. Amer. Electrochem. Soc.
determined solely by the acid/zinc ratio. With im- 41 (1922) 389.
purities present there is an increasing rate of [5] D. Buttinelli and A. Iacchelli, Ind. Min. (1976)
421.
hydrogen evolution as the deposition progresses, [6] T. Ejima, K. Shimakage, T. Muto and T. Suda,
and this may be a result of hydrogen deposition on Nippon Kinzoku Gakkaishi 38(8) (1974) 761.
exposed foreign-metal aggregates. [7] R.C. Kerby, H. E. Jackson, T. J. O'Keefe and
Yar-Ming-Wang,Met. Trans. AIME 8B (1977)
661.
Acknowledgement [8] A.R. Ault, J. H. Bain, D. J. Palmer and J. B.
Pullen, in South Australia Conference 1975,
Australasian Institute of Mining and Metallurgy,
The author acknowledges with thanks the valuable Melbourne (1975) p. 225.