R

'

hydrometallurgy
Hydrometallurgy 36 (1994) 271-283

ELSEVIER

Effect of some common impurities on mass transfer
coefficient and deposit quality during copper
electrowinning
T. Subbaiah, S.C. Das
Regional Research Laboratory, Council of Scientific & Industrial Research, Orissa 751013,
Bhubaneswar- 751 013, India
Received 24 August 1993; revised version accepted 14 January 1994

Abstract

The effects of impurities such as Fe2+, Fe 3+, Ni 2+, Coz+ , Mn2+ and LIX64N on copper
electrowinning were investigated. These impurities were found to affect the physicochemical properties of the electrolyte adversely, which ultimately caused a decrease in the limiting current density and mass transfer coefficient. Examination of the cathode copper
deposited from such electrolytes showed that changes in the physicochemical properties of
the copper electrolyte not only affected the mass transfer conditions during copper electrowinning but also effected the surface morphology and crystal orientation of the cathode
copper significantly. The presence of any of the metal ions studied promotes the growth of
pyramidal and planar structures, while restricting the growth of ridge-type structures.

I. Introduction

For the last 30 years there has been a marked increase in the production of
copper by electrowinning. The major drawback in the conventional electrowinning of copper is the low current density used which causes rather high capital
costs, due to the large tank house needed and the high energy requirements. Continuous efforts are being made to achieve higher current densities using higher
mass transfer rates, reduced equipment sizes and increased outputs [ 1-5 ]. The
energy consumption in a copper electrolytic cell depends on the physicochemical
properties of the electrolytes, such as density, viscosity and conductivity. Density
and viscosity affect the heat and mass transfer conditions in the cell and thereby
determine the energy consumption. The presence of metallic impurities not only
0304-386X/94/$07.00 © 1994 E; lsevier Science B.V. All rights reserved

SSDI 0304-386X (94) 00007-P

272

T. Subbaiah, S.C. Das / Hydrometallurgy 36 (1994) 271-283

affects the physicochemical properties of the electrolyte [ 6-13 ] but also the deposit quality [ 14,15 ].
More recently, processing of slags, complex sulphides and ocean nodules through
hydrometaUurgical routes has resulted in solutions containing a variety of metallic impurities [ 16-18 ] from which the copper is to be extracted. These leach liquors undergo purification steps, such as solvent extraction, before being treated
for electrowinning of copper. During the solvent extraction process, some impurities escape and report in the electrolyte. Although the concentrations of such
impurities are not high enough to affect mass transfer coefficients and deposit
quality substantially, there is a high possibility of their build-up in the closed
circuit (SX-EW), where they are likely to interfere with the electrowinning process. In addition, the electrolyte is also contaminated with entrained solvent.
In the present study, the change in the physicochemical properties of copper
electrolyte due to the presence of certain metallic impurities was determined. The
effect of such changes on the limiting current, mass transfer coefficient and cathode deposit during copper electrowinning was examined. Further, the effect of
entrained solvent on the limiting currents, iL, and mass transfer coefficients, K
values, during the process was investigated.

2. Experimental
All the reagents used in the present study were of reagent grade. The electrolytes were prepared using distilled water. In order to study the effect of various
metallic impurities, the respective sulphates were used. Desired quantities of these
impurities were added to the CuSO4-H2SO4 electrolyte from their stock solutions. 20% LIX64N stock solution was prepared using aliphatic, commercial grade
kerosene as the diluent. The desired quantity of the solvent was added to the
electrolyte.
The physicochemical properties, such as density, conductivity and viscosity
were measured using a constant temperature bath. Densities were measured by
means of a standard 50 ml pycnometer. Viscosities were measured with a standard U-tube viscometer. Conductivities were measured using a Phillips PR 9500
conductivity meter. The electrolytic cell used for limiting current measurements
was 13 cm long, 6 cm wide and 10 cm high. The cell was connected with a Watson-Marlow peristaltic pump for circulating electrolyte at the desired rate. The
anode was a lead-antimony sheet (7% Sb) 3 mm thick, 10 cm long and 5.8 cm
wide. The cathode was a copper sheet 10.0 cm long and 5.8 cm wide. Fresh cathodes were used for each limiting current measurement. The cathodes were carefully polished with fine (600 grade) emery paper to secure uniformity, washed
under running tap water, scrubbed with filter paper, soaked in acetone, rinsed
with distilled water and finally air dried. The electrode was then transferred into
the cell and connected to the circuit for limiting current measurement. A copper
rod 3 mm in diameter and 5 cm long, immersed in the same copper sulphate
sulphuric acid electrolyte as that used in the cell, was employed as the reference

T. Subbaiah, S.C. Das / Hydrometallurgy 36 (1994) 271-283

273

electrode. Current potential curves for copper deposition were measured with a
high power potentiostat (Wenking Model HP 72 ) operated in the potentiostatic
mode. The scan rate was maintained at 5 m V / s controlled by a scan generator
(Wenking Model VSG 72 ). The current potential curves were recorded using an
X- Y recorder.
The electrolytic cell used for producing cathode copper for the characterisation
studies was a 250 ml glass container. The cathodes used were rectangular stainless
steel sheets 10 cm long and 5.8 cm wide. The anodes were of lead-antimony (Sb
7%) with the same area as the cathodes. The electrolyte contained 50 g/1 copper
and 150 g/l H2SO 4. In all the experiments one anode and one cathode were used
and kept face to face, 2.5 cm apart. The electrolysis was carried out by applying
DC voltage from a regulated power supply unit, at a current density of 250 A/
m 2. All these experiments were carried out at room temperature ( 30 _+ 1 ° C ). After
electrolysis, the cathodes were thoroughly washed with water, followed by acetone, and air dried.
The surface morphology of the deposits was examined by SEM. The crystallographic orientation of each deposit was determined by X-ray diffractometry.

3. Results and discussion

Copper electrowinning largely depends upon the physicochemical properties
of the electrolytes, since these properties of the electrolytes affect the energy consumption in the tank house. Hence, the electrowinning operations should be carried out under conditions of high electrical conductivity and high mass transfer
conditions to minimise the power consumption. The recent trends have been directed towards operation of the electrowinning processes at increased current
density. However, this is limited by the limiting current density, iL, expressed as:
iL--

nFDCb
~

(1)

The iL value can be increased if the mass transfer conditions in the electrowinning
cell can be improved according to:
K:

iL

nFCb

(2)

The limiting current density, iL, is affected by the solution properties, such as
viscosity and density, by inhibiting the diffusion of copper ions. This is shown by
the following equation [ 15 ]:

Fk2/3T 2/3V~ Cb
6.= ( 67~yi)2/3kgr12/3(rl/p)n

(3)

It has also been shown that the ~]2/3 term in the above equation contributes more

274

T. Subbaiah, S.C. Das / Hydrometallurgy 36 (1994) 271-283

than the term (q/p)n to inhibiting diffusion. This indicates that viscosity plays a
role in impeeding diffusion of a species.
In the present work, the solution properties of copper electrolyte containing
some cations and LIX64N were determined. The limiting current densities, iL, of
these electrolytes were measured, from which the K values were calculated. The
deposits obtained from these solutions were examined using X-ray diffraction
and scanning electron microscopy. Based on the data generated, attempts were
made to find a correlation between physicochemical properties of the copper
electrolyte, limiting current density, mass transfer coefficient and deposit
characteristics.

3.1. Effect of ferrous iron
The physicochemical properties, such as viscosity, density and conductivity, of
copper electrolyte containing ferrous ion in the range 0-11.95 g/1 were determined. The results are given in Table 1. Both the viscosity and density increased
with increase in ferrous iron in the electrolyte, but the conductivity dropped from
420 to 350 m t2-~ cm-1 as the ferrous ion concentration was raised to I 1.95 g/1.
The change in the iL values due to the increase in ferrous iron concentration is
shown in Fig. 1. It increased from ~ 1300 to ~ 1500 A / m 2 as the ferrous iron
concentration approached ~ 8 g/1. It then dropped to ~ 1300 A / m z when the
concentration of ferrous iron became ~ 12 g/1. The results indicate that Fe 2÷ up
to ~ 8 g/1 enhances the iL values and, thus, does not interfere in the diffusion of
copper ions, in spite of the increase in the viscosity and density. The impact on
the mass transfer coefficient, K, is given in Table 2. The K values increased up to
~ 8 g/l of ferrous iron, beyond this they dropped.

3.2. Effect of ferric iron
The physicochemical properties of copper electrolyte due to the presence of
ferric iron in the range 0-15.8 g/1 are given in Table 1. As in the case of ferrous
iron, both the viscosity and density increased with increasing ferric iron concentration; however, at the same time, the conductivity diminished. The plot of iL
against Fe 3+ concentration is shown in Fig. 1. The iL value increased up to ~ 1550
A / m 2, after this it dropped and remained constant beyond ~ 7 g/l. As in the case
of ferrous iron, the iL values in this case also showed a tendency to increase in
spite of the increase in the viscosity and density. The mass transfer coefficients
at various ferric iron concentrations are given in Table 2. This shows a similar
trend as observed in the case of ferrous iron.

3.3. Effect of manganese
The density and viscosity of copper electrolyte increased as the manganous ion
was added to it. The detailed results are given in Table 1. The conductivity of
copper electrolyte dropped from 420 m Q - l / c m to 370 m ~ - ~/cm in the presence

T. Subbaiah, S. C. Das /Hydrometallurgy 36 (1994) 271-283

275

Table 1
Physicochemical properties of copper electrolyte containing dissolved impurities 30 ° C
Soln.

Conc.

Conc.

Metal ion

Density

No.

o f Cu
( g d m -3 )

of H2SO4
( g d m -3 )

conc.
( g d m -3 )

( g c m -3 )

38.5
38.5
38.5
38.5
38.5
41.6
41.6
41.6
41.6
41.6
41.6
38.5
38.5
38.5
38.5
38.5
40.9
40.9
40.9
40.9
40.9
40.9
39.3
39.3
39.3
39.3
39.3

143.2
143.2
143.2
143.2
143.2
163.5
163.5
163.5
163.5
163.5
163.5
143.6
143.6
143.6
143.6
143.6
152.2
152.2
152.2
152.2
152.2
152.2
143.3
143.3
143.3
143.3
143.3

0.985 Fe z+
2.760 Fe 2+
8.350 Fe 2+
11.95 Fe 2+
0.779 Fe 3+
1.550 Fe 3+
3.160 Fe 3+
7.400 Fe 3+
0.900 M n 2+
4.580 M n 2+
9.020 M n 2+
19.90 M n 2+
19.90 M n 2+
0.839 Ni 2÷
1.590Ni 2+
4.270 Ni 2+
7.880 Ni 2÷
19.60 Ni 2+
0.860 Co 2+
4.290 Co 2+
8.300 Co 2+
17.30 Co 2+

1.163
1.165
1.168
1.179
1.186
1.177
1.180
1.183
1.187
1.191
1.162
1.165
1.173
1.183
1.204
1.204
1.177
1.180
1.181
1.187
1.195
1.211
1.163
1.168
1.175
1.183
1.199

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

Absolute
viscosity

Conductivity
(m.Q -1 c m - : )

( k g / m s)
0.162
0.164
0.168
0.176
0.179
0.169
0.171
0.172
0.174
0.185
0.162
0.163
0.164
0.171
0.187
0.187
0.167
0.170
0.170
0.172
0.181
0.194
0.162
0.164
0.167
0.171
0.183

420
400
390
370
350
440
435
435
430
415
420
390
385
375
370
370
415
415
410
400
390
370
400
395
390
375
350

of 19.9 g/1 of Mn 2+. The iL values at various manganous ion concentrations are
plotted in Fig. 2. In spite of the decrease in the viscosity and density the i L value
showed an increase up to ~ 5 g/1 of manganous ion. The K values at various
manganous ion concentrations are given in Table 3. The trend is similar to that
observed in the case of i L values.
From the data on the physicochemical properties and i L values due to presence
of ferrous and ferric iron and manganous ion in the copper electrolyte, it can be
seen that Eq. (3) is not obeyed up to certain concentrations of the metal ions.

3.4. Effect of nickel and cobalt ions
The plots of i L values at various nickel and cobalt concentrations are shown in
Fig. 3. In both cases, the iL values decreased linearly. The linear relationship between i L and Ni or Co concentrations derived are as follows:

T. Subbaiah, S. C Das /Hydrometallurgy 36 (1994) 2 71-283

276
2000]

Fe 2+

E

A Fe 3"1"

~_______~

'

1200
O

D
U

800

ol
c

.l
.d

400

0

0

I

J

1

I

4

8

12

16

Concentration

of

20

impurities in g / L

Fig. 1. The effect of Fe 2÷ and Fe 3+ on limiting current density.
Table 2
Effect of iron on mass transfer coefficient
Ferrous iron

Ferric iron

Fe 2+
(g/l)

Mass transfer
coefficient
K (ram/s)

Fe 3+ conc.
(g/l)

Mass transfer
coefficient
K (ram/s)

0
0.985
2.76
8.35
11.95

1.160
1.160
1.253
1.290
1.124

0
0.779
1.55
3.16
7.60
11.80

1.160
1.177
1.245
1.245
1.075
1.075

Circulation rate = 0.5 l/min; temperature = 30 ° C.

iL =

1530.2- 17.9Ni

and

iL --

1530.2-24.3Co

(4)
(5)

It can also be seen from the Fig. 3 that the decrease in iL values was comparatively more in the case of cobalt ions than for nickel. The changes in the physicochemical properties of the copper electrolyte due to the above ions are summarized in Table I. The viscosity and density increased, whereas the conductivity

T. Subbaiah, S.C. Das / Hydrometallurgy 36 (1994) 2 71-283

¢"E

277

2000
1800
1600
1400(

u
t"t

1200

E;

1000
t~O0
600
._1

400
200
0
0

I

1

1

I

I

i

i

i

I

I

I

I

I

2

4

6

8

10

12

14

16

18

20

22

24

26

Concentration

of

Manganese

in

28

g/L

Fig. 2. The effect o f M n on limiting current density.
Table 3
Effect o f manganese on mass transfer coefficient
Manganese conc.
(g/l)

Mass transfer coefficient
K (mm/s)

0
0.900
4.580
9.020
19.900

1.1600
1.253
1.382
1.198
1.069

Circulation rate = 0.5 1/min; temperature = 30 ° C.

decreased with rising nickel or cobalt ion concentrations in the electrolyte. In
these cases, the iL value decreased with increasing viscosity and density of the
electrolyte, which is in agreement with Eq. (3). The change in the K values at
various nickel and cobalt ion concentrations are given in Table 4. The K values
decreased from 1.174 mm/s to 0.905 mm/s on addition of 19.6 g/1 of Ni, whereas
it decreased to 0.965 m / s when 17.3 g/1 of Co was added to the copper electrolyte.

3.5. Effect of LIX64N
The effect of LIX64N on deposit quality during copper electrowinning has been
studied by Hopkin and co-workers [ 19 ] who indicated that presence of the solvent in the electrolyte caused 'organic burn' on the cathode. Harvey et al. [20 ]
reported that good cathode copper could be achieved at a current density of 643.3
A / m E, even from an electrolyte containing dissolved solvent (40% LIX64N). No

T. Subbaiah, S. C. Das / Hydrometallurgy 36 (1994) 271-283

278
200~

A Ni z÷
Co z+

~E160(
\

:•

]20C

C3

80(
o't

._~

coo:
_J

2
Concentration

i

8

I

,2

20

of impurities in g/L

Fig. 3. The effect of nickel and cobalt on limiting current density.

Table 4
Effect of nickel and cobalt on mass transfer coefficient
Nickel ion

Cobalt ion

Ni 2+ conc.
(g/l)

Mass transfer
coefficient
K (mm/s)

Co 2+ conc.
(g/l)

Mass transfer
coefficient
K (mm/s)

0
0.839
1.59
4.27
19.60

1.174
1.174
1.158
1.125
0.91

0
0.860
4.29
8.30
17.30

1.174
1.125
1.075
1.042
0.965

Circulation r a t e = 0 . 5 l/min; temperature= 30°C.

information is available on the change in iL and K values during copper electrowinning due to the presence of LIX64N. In the present work, the iL and Kvalues
were determined for a copper electrolyte containing LIX64N in the range of 0200 mg/1. Table 5 gives the K values at different LIX64N concentrations. It is
observed that the mass transfer coefficient increased with increasing concentration of dissolved LIX64. Fig. 4 shows the plot of iL against LIX64N concentration. A similar trend was observed to that of the mass transfer coefficient.

T. Subbaiah, S.C. Das / Hydrometallurgy 36 (1994) 271-283

279

Table 5
Effect of LIX64N on mass transfer coefficient
LIX64Nconc. dissolved in
electrolyte (g/I)

Mass transfer coefficient
K (mm/s)

-

1.169
1.235
1.342
1.366

0.02
0.100
0.200

Circulation rate = 0.5 1/min; temperature 30 ° C.

200 O,

N E 1600
\
f

:"

12~

g
Q

~

80

U
C

•-

4~

E
-d

0

I

0

i

o.~

0-05
0!1
0.15
Concentration of LIX 6/.,N in g / L

0.2

Fig. 4. The effect of LIX64N on limiting current density.

Table 6
Crystallographic orientation
Impurities

0
Fe z+
Fe 3+
Ni 2÷
Mn 2÷

Impurity
conc.
(mg/l)

500
500
150
500

Crystallographic orientation
(Peak heights, cm)

(111)

(200)

(220)

16.1
18.1
17.7
16.2
17.9

7.3
8.7
8.1
7.9
8.2

13.85
10.0
7.2
11.8
11.4

280

T. Subbaiah, S.C. Das / Hydrometallurgy 36 (1994) 271-283

Fig. 5. SEM photomicrograph of electrodepositedcopper ( × 600).

Fig. 6. (a) SEM photomicrograph of electrodeposited copper (× 600). Nickel content= 2000 ppm.
(b) SEM photomicrograph of electrodepositedcopper ( × 600 ). Iron (Ili) content = 500 ppm.

3. 6. Effect o f metallic impurities on the cathode deposit
In order to characterise the cathode copper deposited in the presence of different metal ions, copper was electrowon from a static bath. The electrodeposits
were subjected to X-ray diffraction and scanning electron microscopy.

3.7. Crystallographic orientation
The effects of various metallic ions on the crystallographic orientations of the
copper deposits were examined using X-ray diffraction. O f the five copper lines
which were scanned, only three--( I l l ), (200) and (220)Mshowed some changes
in their respective peak hights and thus only these three lines are considered here.
The results are summarised in Table 6. In all the cases, the crystalloplane (111 )

T. Subbaiah, S.C. Das /Hydrometallurgy 36 (1994) 271-283

281

Fig. 7. SEM photomicrograph of electrodeposited copper ( × 600 ). Iron (II) content = 500 ppm.

Fig. 8. SEM photomicrograph of electrodeposited copper ( × 600). Manganese (II) content = 1500
ppm.

remained preferred. The peak hights for the planes ( I I I ) and (200) due to the
presence of any of the metallic ions studied increased whereas that of (220) decreased. This shows that the presence of any of these metallic ions promotes the
growth of a pyramidal as well as a planar structure and restricts the growth of a
ridge-type structure. The preferred planes for cathode copper from pure electrolyte and from electrolyte containing Fe 2+, Mn 2÷ and Ni 2÷ are in the order:

(II1)> (220)> (200)
but from electrolytes containing Fe 3 +, the order is:

(111)> (200)>220).

282

T. Subbaiah, S.C. Das / Hydrometallurgy 36 (1994) 271-283

3.8. Surface morphology

The surface morphology of the copper deposits obtained in the presence of different metallic ions was examined by SEM. Some typical photomicrographs are
presented in Figs. 5-8. The photomicrograph of the cathode from pure electrolyte
is shown in Fig. 5a. The crystallite size and shape are greatly influenced by the
presence of the metallic impurities in the copper electrolyte. The deposits produced in the presence of Ni 2+ (Fig. 5b) and Fe 3÷ (Fig. 6) seem to be even and
compact. The smallest crystallites are observed in the presence of Ni 2÷ (Fig. 5 ),
whereas in the case of Fe 3+ (Fig. 6) the crystallites are a little larger than those
of Ni 2÷. In the case of Fe 2÷ and Mn 2+, however, the growth of larger crystallites,
which are more crystalline, is observed (Figs. 7 and 8). The presence of Mn 2÷
caused well developed pyramidical growth.
From these studies it is clear that the changes in the physicochemical properties
of the copper electrolyte due to the presence of the metallic impurities not only
affect the iL and K values but also greatly influence the characteristics of the
deposit.

4. Conclusions
The effects of some common impurities such as Fe 2+, Fe 3+, Ni 2+, Co 2+, Mn 2+
and LIX64N during copper electrowinning were studied. The changes in the values of physicochemical properties, such as density, viscosity and conductivity of
copper electrolyte, due to the presence of such impurities were measured. It was
found from the results that both the density and viscosity of copper electrolyte
increased due to presence of the impurity whereas the conductivity decreased.
The effects were comparatively larger when the concentrations of the impurities
were higher. The decrease in the conductivity values of the electrolytes would
adversely affect the power consumption in a copper tank house. Such changes in
the physicochemical properties caused decreases in the limiting current density,
ic, and the mass transfer coefficient, K, during copper electrowinning.

Acknowledgements
The authors are indebted to Prof. H.S. Ray, Director, and Dr. R.P. Das, Deputy Director, Regional Research Laboratory, Bhubaneswar, Orissa, India, for their
keen interest in this work. They also are grateful to Prof. Ray for his kind permission to publish this paper.

Nomenclature
Cb

=

concentration of copper in the bulk electrolyte, m o l / m 3

T. Subbaiah, S.C. Das /Hydrometallurgy 36 (I 994) 2 71-283

D
F
iL

K
k
m, p
n
ri

T

r/
P

=
=
=
=
=
=
=
=
=
=
=
=
=
=

283

diffusion coefficient o f the electrolyte, mE/s
Faraday constant
limiting current density, A / m E
mass transfer coefficient, m m / s
Boltzmann constant
constant for a given electrode geometry
constants
number o f electrons taking part in the reaction
radius o f the diffusion species
temperature, °C
velocity constant o f the electrolyte parallel to the electrode surface at
a distance greater compared to diffusion layer thickness
absolute viscosity of the solution, k g / m s
density o f the solution, k g / m 3
diffusion layer thickness, m

References
[ 1]
[2]
[3]
[4]
[5 ]
[6]
[7 ]
[8]
[9]
[ 10]
[ 11 ]
[ 12]
[ 13]
[ 14 ]
[15 ]
[ 16]
[ 17]
[ 18 ]
[ 19]
[20]

Balberyszski, T. and Andersen, A.K., Proc. Aust. Inst. Min. Met. 244 (1972): 11.
Flett, D.S., Chem. Ind., 16 (1972): 933.
Eliasen, P.D. and Edmunds, E., Jr., CIM Bull., 67 ( 1974): 82.
Harvey, W.W., Randlett, M.R. and Bangerskis, K.L., Kennecott Copper Corp. Tech. Note, 108
(1974).
Jagannadha Raju, G.J.V., Venkateswarlu, P. and Sarveswara Rao, S., Ind. J. Technol., 18 (1980):
229.
Kocharev, V.I., Kondrat'eva, N.M. and Levin, A.I., Tsvetn. Metall., 7 ( 1966): 29.
Pomosov, A.V. and Prishvitsyna, G.N., lzv. Vyssh. Uchebn. Zaved, 7 (1964): 45.
Gorbachev, S.V. and Vasenin, R.M., Zh. Fiz. Khim., 28 ( 1954): 135, 1795, 1922, 1928, 2156.
Stender, V.V., J. Appl. Chem. USSR, 19 (1946): 231.
Classens, P., Ph.D. Thesis, Univ. Lauvain (1967).
Classens, P., Feneauh, C1. and Breakpot, R., Bull. Soc. Chem. Belg., 77 (1968): 213.
Price, D.C. and Davenport, W.G., Metall. Trans., 12B ( 1981 ): 639.
Kern, E.F. and Chang, M.Y., Trans. Am. Electrochem. Soc., 41 ( 1923): 181.
Ettel, V.A., Gendron, A.S. and Tilak, V.B., 102nd Annu. AIME Meet. (Chicago, 1973).
Anderson, T.N., Wright, C.N. and Richards, K.J., Int. Symp. on Hydrometallurgy (Chicago,
1973).
Das, R.P., Anand, S., Sarveswara Rao, K. and Jena, P.K., Trans. Inst. Min. Met., 96 ( 1987):
152.
Das, R.P., Anand, S., Das, S.C. and Jena, P.K., Hydrometallurgy, 16 ( 1986): 16.
Anand, S., Sarveswara Rao, K. and Das, R.P., Trans. Ind. Inst. Met., 39 ( 1986): 51.
Hopkins, W.R., Egyett, G. and Scuffham, J.B., In: D.J.I. Evans and R.S. Shoemaker (Editors),
Int. Syrup. Hydrometallurgy. AIME, New York, (1973) p. 127.
Harvey, W.W., Randlett, M.R. and Bangerskis, K.I., Trans. Inst. Min. Metal, Sect, C, 84 ( 1975 ):
210.