Minerals Engineering 21 (2008) 1126–1130

Contents lists available at ScienceDirect

Minerals Engineering
journal homepage: www.elsevier.com/locate/mineng

Extractive separation of copper and nickel from copper bleed stream

by solvent extraction route
Archana Agrawal *, M.K. Manoj, S. Kumari, D. Bagchi, V. Kumar, B.D. Pandey
Metal Extraction and Processing Division, National Metallurgical Laboratory, Jamshedpur 831007, India

a r t i c l e i n f o a b s t r a c t

Article history: Bleed stream from electro refining step of copper smelter was processed to recover the metals as high
Received 22 October 2007 value products such as copper and nickel powders or salts. The process consists of partial decopperisation
Accepted 27 February 2008 of the bleed stream followed by crystallization of a mixed salt of copper and nickel sulphate, leaching of
Available online 18 April 2008
the mixed salt, removal of iron, solvent extraction for the separation of copper and nickel and winning the
solution to produce metal powders. After partial decopperisation of copper bleed stream, the resultant
solution was subjected to crystallization which produces composite crystals with the chemical composi-
Keywords:
tion of 8.4–12.5% Cu, 13.7–14.38% Ni and 1–2 ppm of Fe as impurity. This mixed salt was leached with
Copper
Nickel
water and was treated for iron precipitation. The purified solution was subjected to solvent extraction
Solvent extraction using two solvents namely LIX 84 or Cyanex 272. A 20% LIX 84 in kerosene extracted 99.9% copper
Electrorefining and 0.059% nickel at a pH of 2.5, similarly a 5% Cyanex 272 in kerosene extracted 98.06% copper and
Bleed stream 0.51% nickel at a pH of 4.85. LIX 84 was used for metal separation in the mixer-settler unit and a flow
Hydro metallurgy sheet was developed using this solvent. The pure solutions of copper after stripping it from the loaded
Powder metallurgy organic and nickel left in the raffinate were further electrolysed to produce pure copper (99.9%) and
nickel (99.8%) powders, alternatively pure sulphate salts could also be crystallized. Since it is well known
that Cyanex 272 is used for the separation of cobalt and nickel, a few experiments were performed on the
extraction of copper by using Cyanex 272. A complete study is yet to be done to develop a flow sheet by
using this solvent.
Ó 2008 Elsevier Ltd. All rights reserved.

1. Introduction
Copper bleed electrolyte is generated during the purification of
impure anode copper in the electro-refining plant. Due to recycling
of the electrolyte the impurities viz. arsenic, iron, bismuth, anti-
mony, chloride etc. get enriched in the solution. When the concen-
tration of the impurities reaches their upper limit of tolerance it
becomes deleterious for the cathode copper production. In order
to control the purity of the cathodes during the continuous opera-
tion, a part of the solution is discarded as bleed stream periodically
which contains 40–50 g/L Cu, 170–200 g/L H2SO4, 20 g/L Ni and
1–2 g/L Fe, and traces of other impurities like As, Sb, Bi, Cl etc.
In the existing practice the bleed electrolyte is purified by
removing copper, arsenic and antimony by electrolytic stripping
in the liberator cell. The purified liquor is then concentrated to
crystallize out nickel sulphate and the solution left in the form of
black acid can be recycled back to the system. But the associated
drawbacks with this process are liberation of the arsine gas, high
operation cost, poor grade NiSO4, and loss of acid (Habashi and
Ismail, 1975).
* Corresponding author. Tel.: +91 657 2271709; fax: +91 657 2270527.
E-mail address: archana_nml03@yahoo.com (A. Agrawal).
Hence research in this area has been done on such bleed
streams/electrolytes containing copper and other metals which
could be treated by various processes such as hydrogen reduction,
solvent extraction, cementation and electrowinning to produce
copper as its sulphate salt which could be recycled back to the
electrolysis tank or pure copper powder for powder metallurgical
(P/M) applications. Of these processes the solvent extraction is of-
ten in vogue in the copper industry and more than 1.5 million tons
of copper are produced by this process annually from the lean
grade ores/waste streams. Copper was recovered from motor scrap
and other lean grade ores/waste streams by ammonia leaching,
solvent extraction with LIX 84, and electrowinning of metal from
the stripped solution (Kirk-Othmer Encyclopedia, 1995; Majima
et al., 1993). The extraction of copper and nickel has attracted
much interest and usually oximes are used for the extraction of
these metals, although b-diketones have also been used in some
studies (Ritcey, 1983; BhaskaraSarma et al., 1987; Nathsarma
and BhaskaraSarma, 1993; BhaskaraSarma and Nathsarma, 1996).
PT-5050 is one of the Acorga products having the advantages of
strong extraction, good stripping and high selectivity for copper
(Nigo et al., 1993; Kordosky, 1992). Recovery of copper from HCl
solution and its transfer into H2SO4 solution were reported using
extractant mixtures such as trioctylamine-LIX 54 (Barakat and

0892-6875/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.mineng.2008.02.016
 A. Agrawal et al. / Minerals Engineering 21 (2008) 1126–1130 1127

Mahmoud, 1999), and Alamine336-LIX 54 (Kyuchukov and Mihay-
lov, 1991, 1993). Recently, some work by Kumar et al. (1989) was
done at NML for separating copper and zinc from sulphate solution
in the presence of gangue metals like iron and manganese using
LIX 84.
A process has been developed at NML for the recovery of copper
(Agrawal et al., 2007a) and nickel (Agrawal et al., 2007b) from cop-
per bleed stream from an electro-refining unit. The process steps
consists of partial decopperisation of the bleed stream, concentra-
tion and crystallization of a mixed salt, leaching of the mixed crys-
tal, iron purification, solvent extraction of the purified solution
using LIX 84 or Cyanex 272, and electrolysis of the copper and nick-
el solutions to produce pure copper and nickel powders. This paper
deals with the solvent extraction separation of Cu and Ni using LIX
84 and Cyanex 272.
2. Experimental
Synthetic copper bleed solution containing 38.42 g/L Cu,
19.37 g/L Ni, 0.757 g/L Fe, 0.2 g/L Bi, 171 g/L H2SO4 was prepared,
similar to that obtained from ICC, Ghatsila (which contains:
39.86 g/L Cu, 9.577 g/L Ni, 194.04 g/L H2SO4, 0.0007 g/L As, 0.26 g/
L Fe, 0.055 g/L Sb and 0.11 g/L Bi. 20% LIX 84 or 5% Cyanex 272 in
kerosene were diluted for solvent extraction studies. Since the cop-
per concentration was very high in copper bleed stream, a partial
decopperisation of the bleed stream was carried out in a lead lined
tank of 1L capacity to bring down the copper level similar to the
concentration of nickel present. Subsequently, this solution was
evaporated to crystallize out mixed sulphate salt of copper and
nickel in two stages (Kumari et al., 2006). The first stage salt con-
tained 21.7% Cu and 2.9% Ni. The second stage salt had 8.4% Cu and
15.6% Ni that was washed and leached with water to give an aque-
ous solution with a chemical composition: 8.4 g/L Cu, 15.55 g/L Ni,
15.68 g/L H2SO4, 0.1 g/L Bi, and 0.52 g/L Fe. Presence of iron is del-
eterious for solvent extraction and in electrolysis by decreasing the
cathode current efficiency during electrowinning leading to high-
energy consumption and low production rates. Hence iron removal
from the electrolyte is considered essential. Thus the mixed crystal
leach solution containing iron was removed from the solution with
the drop wise addition of NaOH solutions with a constant stirring
to increase the pH of solution from 1.12 to 3.42 and aerated to pre-
cipitate the iron as ferric hydroxide thereby bringing down the iron
content from 0.52 to 0.0041 g/L. The iron free solution containing
8.39 g/L Cu, 13.67 g/L Ni was subjected to solvent extraction using

Copper bleed solution (g/L) Cu-

38.42, Ni-19.37, Fe-0.757, Bi -0.2 &
H2SO4-171

Decopperisation

Decopperised solution (g/L)

Cu-18..96, Ni- 20.4, H2SO4 - 206

Evaporation/
Crystallisation
Mother liquor for Mixed crystal of
recycling to refinery Cu & Ni sulphate

Leaching of mixed crystal

& iron removal

Filtration

Solution of Cu & Ni

Solvent extraction
Raffinate Ni
LIX 84 or cyanex 272
Regenerated
solvent
Electrowinning
Loaded organic (Cu)

Ni powder Ni scrubbing and Cu stripping

Spent
Electrowinning electrolyte

Cu powder

Fig. 1. Copper bleed solution treatment following decopperisation-crystallization-solvent extraction-electrowinning.

1128 A. Agrawal et al. / Minerals Engineering 21 (2008) 1126–1130

LIX 84 or Cyanex 272 to separate copper from nickel. Finally copper 3

and nickel sulphate thus separated were subjected to electrolysis
to give P/M grade copper and nickel powders. Various parameters Slope = 1.5133
2.5
were optimized on bench scale and continuous operation for the
separation of copper from nickel and are discussed briefly in the 2
following text (see text below). The flow sheet shown in Fig. 1 de-

Log D
picts the steps followed. 1.5
Analyses of the samples for low concentration of metallic con-
tents were carried out by an atomic absorption spectrometer. Solu-
1
tions containing high metals (Cu and Ni) were analyzed by
conventional method (Lewis and Ott, 1970; Jeffery et al., 1989).
0.5
All the chemicals used for the chemical analysis were Analar grade
reagents.
0
0 0.5 1 1.5 2 2.5 3
3. Results and discussion pH

3.1. Solvent extraction with 20% LIX 84 Fig. 3. Plot of log D vs pH with LIX 84.

Preliminary experiments (Kumar et al., 1989) showed that 20%

LIX 84 diluted with kerosene oil is suitable for the extraction of
copper from the concentrated aqueous feed solutions. Hence all
the experiments were conducted with 20% LIX 84 diluted with ker-
osene. Effect of parameters for maximum metal recovery was stud-
ied and are given below.
Aqueous solution containing 8.39 g/L Cu and 13.67 g/L Ni was
shaken with 20% LIX 84 at O:A ratio of 2:1 for 10 min at different
equilibrium pH varied from 1 to 5 as shown in Fig. 2. At equilib-
rium pH of 2.5 copper extractions was 99.9% whereas nickel
extraction was only 0.55%. Plot of log D for copper vs equilibrium
pH (Fig. 3) gives a slope of 1.51 which indicates the release of
2H+ leading to the extraction of copper as CuA2 (HA: active organic
hydroxime species) in the organic phase as shown below
Cu2þ þ ½2HA0 ! ½CuA2 þ
0 þ 2H
þ
ð1Þ
The loading capacity of 20% LIX 84 was determined by repeated
contact method and was found to be 7.94 g/L Cu at a pH of 2.73 in
three contacts. The copper loaded organic was stirred with a strip
solution containing 31.77 g/L Cu and 180 g/L H2SO4. Stripping was
done at different O:A ratios and at a ratio of 4:1 the stripped solu-
tion was found to contain 57.19 g/L Cu. Data on stripping of loaded
copper from LIX 84 are presented in Table 1.
3.2. Simulation of counter current extraction of copper with LIX 84
Effect of aqueous to organic ratio was studied to carry out the
shake out experiments. It can be seen from Fig. 4 that with the de-
100
90
80
70
60 Cu
50 Ni
40
30
20
10
0
0 1 2 3 4 5
Fig. 2. Copper and nickel extraction at different pH on LIX 84, aqueous solution (g/
L): 8.39 Cu, 13.67 Ni, 20% LIX 84 in kerosene, A:O 1:2, stirring time 10 min.
Table 1
Effect of O:A ratio on stripping of copper from loaded copper from LIX 84 (Loaded
organic (20% LIX 84) –7.94 g/L Cu; strip solution – 31.77 g/LCu and 180 g/L H2SO4;
stirring time – 10 min)
Stripping stage O:A ratio [Cu]org (g/L) [Cu]aq (g/L) Stripping (%)
1 4:1 1.59 57.19 80.0
2 2:1 1.87 43.85 76.4
3 1.33:1 0.31 41.94 96.1
4 1:1 0.29 39.41 96.3
crease in A:O ratio from 2.5 to 0.25, the percent extraction of cop-
per increases from 46% to 99.9%. McCabe Thiele plot for the Cu
extraction shows that at an A:O ratio of 1:1 complete copper is ta-
ken up by 20% LIX 84 in two stages leaving behind only 0.0004 g/L
copper in the raffinate. A few experiments were then carried out in
mixer settler unit where extraction was done with 25% LIX 84. The
mixer settler unit consists of a mixer of 620 mL capacity and a set-
tler of 860 mL capacity. A diaphragm pump was used for pumping
the aqueous and the organic phases at a flow rate of 3.17 and
5.04 L/h, respectively. The stirrer was a pump-mix type with a mo-
tor power of 1/20 hp. The stirring speed was maintained at
800 rpm. Since the volume of the aqueous and organic was 1 L
each, extraction was performed in four stages to extract all the
copper from the aqueous feed. Some nickel is also co-extracted
with the copper by the organic phase, so the loaded organic was
scrubbed with dilute acid (1.76 g/L H2SO4) to remove Ni in two
stages and finally copper was stripped with 10% H2SO4 in two
stages. The composition of raffinate, scrubbed solution and
stripped solution showed the performance of the circuit (Fig. 5).
The stripped solution rich in copper could be used for producing
copper powder. Thus an attempt was made to produce pure copper
powder from the stripped copper solution obtained in running the
mixer settler unit. The composition of solution used for copper
powder preparation was 37.75 g/L Cu and 0.57 g/L Ni and at differ-
ent current densities from 600 to 1000 A/m2, copper powder of the
purity varying from 99.5% to 99.98% was produced (Agrawal et al.,
2007a). Nickel powder of 99.8% purity was also produced by
electrowinning of raffinate on copper extraction (Agrawal et al.,
2007b). These pure metal powders could be used for powder
metallurgical applications.
3.3. Solvent extraction using Cyanex 272
Initially the extraction of copper from the aqueous feed solution
with 8.26 g/L Cu and 13.67 g/L Ni was carried out at different pH
using 20% Cyanex 272 and 4% isodecanol in kerosene. Aqueous
 A. Agrawal et al. / Minerals Engineering 21 (2008) 1126–1130 1129

A/O
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75 3
12 120

10 100

8 80

% Cu extraction
[Cu]org, g/L

6 60

4 40

2 McCabe thiele plot 20

% Cu extraction
0 0
0 1 2 3 4 5 6 7 8 9
[Cu]aq, g/L

Fig. 4. Effect of A:O on % Cu extraction and McCabe Thiele plot for the extraction of Cu by LIX 84.

25% LIX 84
(1.27g/Lcu, 0.3g/LNi
Flow rate-5.04L/h
Ni powder
NaOH flow
rate-0.26L/h
Aqueous feed(g/L)
Cu - 9.53, Ni - 12.81 Extraction Raffinate(g/L)
Flow rate - 3.17 L/h Cu-0.001,Ni 11.37
pH - 3.06
4 stages
Flow rate 3.43, pH 2.7
Organic (g/L)
Cu-7.27, Ni-0.61

Scrub solution Scrubbed solution(g/L)

H2SO4-1.76g/L Scrubbing
Cu – 0.036 ,Ni - 0.25
Flow rate – 4.52L/h 2 stages
Flow rate –4.52L/h

Organic (g/L)
Cu-7.23, Ni-0.39
Stripping solution Regenerated organic
Cu- 30.248g/L, Stripping Cu 0.57g/L, Ni 0.09g/L
Ni- 0.203g/L 2 stages
H2SO4-179.34g/L Stripped solution(g/L)
Flow rate –4.52L/h Cu – 37.75, Ni - 0.57
Flow rate –4.52 L/h
Electrowinning

Spent electrolyte Cu powder

Fig. 5. Separation of copper and nickel in mixer settler unit using LIX 84.

solution was stirred with 20% Cyanex 272 at O:A ratio of 1/1 for
10 min. Plot of equilibrium pH vs % extraction as shown in Fig. 6
shows 98.05% extraction of Cu and 0.51% Ni at a pH of 4.85. Plot
of log DCu at different equilibrium pH (Fig. 7) shows a straight line
with a slope of 1.75 indicating the extraction of the complex CuA2
in the organic phase. Similar observations were obtained by Van de
Voorde et al. (2006). Since the neutral form of the extractant exists
as monomers in the organic phase and the acidic form exists as
dimmers. Thus in the acidic as well as in the neutral form, the
extractant takes part in the extraction process of copper by the fol-
lowing equation.
Cu2þ  þ
aq þ Rorg þ 2ðHRÞ2 ðorgÞ $ MR 2  3HRorg þ Haq ð2Þ
As 20% Cyanex 272 formed a viscous organic phase while deter-
mining the loading capacity, the maximum loading of the extract-
ant was determined using 5% Cyanex 272 by repeated contact
1130 A. Agrawal et al. / Minerals Engineering 21 (2008) 1126–1130

served that almost all nickel and some copper was removed from
100 loaded organic during scrubbing resulting in a Ni free copper solu-
tion. This pure copper solution can further be electrolyzed to pro-
duce pure copper powder or precipitated to produce various
80 copper salts/oxides.
% Extraction

60 %Cu 4. Conclusion
%Ni
The studies were carried out for the treatment of copper bleed
40
electrolyte generated during the purification of impure anode cop-
per in the electro refining plant. Of the different process routes
20 investigated, partial decopperisation, crystallization, solvent
extraction electrowinning route was found suitable to produce
copper and nickel powder of required purity. LIX 84 and Cyanex
0 272 were used as extractants for copper and nickel separation.
3.5 4.5 5.5 6.5
25% LIX 84 was found suitable for complete extraction and separa-
Equilibrium pH tion of copper and nickel in the mixer settler unit. These solutions
after separation could be used for metal powder by electrolysis or
Fig. 6. Effect of equilibrium pH on Cu and Ni extraction with Cyanex 272. Aq.Feed:
8.26 g/L Cu, 13.67 g/L Ni, Solvent 20% Cyanex 272 and 1% isodecanol in kerosene, any other metal salt such as oxides/sulphates.
A:O: 1:1.
Acknowledgement

Authors are thankful to Director, National Metallurgical Labora-

3 tory, for giving permission to publish the paper.

2.5 References

Agrawal, Archana., Kumari, Sarita., Bagchi, D., Kumar, V., Pandey, B.D., 2007a.
2
slope 1.75 Recovery of copper powder from copper bleed electrolyte of an Indian copper
smelter by electrolysis. Minerals Engineering 20, 95–97.
Log D

Agrawal, Archana, Bagchi, D., Kumari, S., Kumar, V., Pandey, B.D., 2007b. Recovery of
1.5
nickel powder from copper bleed electrolyte of an Indian copper smelter by
electrolysis. Powder Technology 177, 133–139.
Barakat, M.A., Mahmoud, M.H.H., 1999. Separation of metal values from brass alloy
1
scrap. In: 29th Annual Hydrometallurgical Meeting, Quebec, Canada, August,
pp. 329–342.
Bhaskara Sarma, P.V.R., Srinivasa Rao, K., Nathsarma, K.C., RoyChoudhury, G., 1987.
0.5
Processing of nickel- and cobalt-containing leach liquors obtained from
different raw materials. Hydrometallurgy 19 (1), 83–93.
BhaskaraSarma, P.V.R., Nathsarma, K.C., 1996. Extraction of nickel from ammoniacal
0
solutions using LIX 87QN. Hydrometallurgy 42 (1), 83–91.
3.5 4 4.5 5 Habashi, F., Ismail, M.I., 1975. Health hazards and pollution in the metallurgical
Equilibrium pH industry due to phosphine and arsine. CIM Bulletin 68 (760), 99–103.
Jeffery, G.H., Bassett, J., Mendhan, J., Denney, R.C., 1989. Vogel’s Textbook of
Fig. 7. Plot of log D vs equilibrium pH with Cyanex 272. Quantitative Chemical Analysis, fifth ed. Longman Publishers (pp. 393–394).
Kirk-Othmer Encyclopedia, vol. 15/31, John Wiley, New York, 1995, p. 3.
Kordosky, G.A., 1992. Copper solvent extraction: the state of the art. Journal of
Metals 44 (5), 40–45.
method at O:A ratio 1:1. The maximum loading of 4.19 g/L Cu and Kumari, Sarita, Agrawal, A., Bagchi, D., Kumar, V., Pandey, B.D., 2006. Synthesis of
2.23 ppm Ni was obtained in three contacts. The loaded organic copper metal/salts from copper bleed solution of a copper plant. Mineral
Processing & Extractive Metallurgical Review 27, 159–175.
was scrubbed with 0.9–3.6 g/L H2SO4 to remove the co extracted Kumar, V., Pandey, B.D., Bagchi, D., Akerkar, D.D., 1989. Scope of using LIX 84 for
and entrapped nickel and then stripped with 0.5–10% H2SO4 (Table separation of copper and zinc from complex sulphate solutions. In: Proceedings
2). It was observed that total copper present in loaded organic can of the International Conference on Base Metals Technology, Jamshedpur, India,
8–10 February, pp. 495–500.
be stripped even with 0.5% H2SO4. Since the electrolysis of copper Kyuchukov, G., Mihaylov, Y., 1991. A novel method for recovery of copper from
takes place at a very high acid concentration therefore stripping of hydrochloric acid solution. Hydrometallurgy 27 (3), 361–369.
copper from the scrubbed loaded organic was done with 10% acid. Kyuchukov, G., Mishonov, I., 1993. A new extractant mixture for recovery of copper
from hydrochloric etching solution. Solvent Extraction Ion Exchange 11 (4),
The A:O ratio in both scrubbing and stripping was 1:1. It was ob-
555–567.
Lewis, C.L., Ott, W.L., 1970. Analytical Chemistry of Nickel, first ed. Pergamon Press.
Majima, H., Hirato, T., Awakura, Y., 1993. Dissolution of copper with aqueous cupric
amine solution I. Studies on the selective recovery of copper and iron from
Table 2 motor scrap. Journal of the Mining and Materials Processing Institute of Japan
Scrubbing and stripping of loaded copper from 5% Cyanex 272 109 (3), 191–194.
Nathsarma, K.C., BhaskaraSarma, P.V.R., 1993. Processing of ammoniacal solutions
Sl. No Conc. of H2SO4 Composition Conc of H2SO4 Composition of containing copper nickel and cobalt for metal separation. Hydrometallurgy 33
for scrubbing of scrubbed for stripping stripped solution (1–2), 197–210.
(g/L) solution (ppm) (%) (g/L) Nigo, S., Hirato, T., Awakura, Y., Iwai, M., 1993. Recovery of copper from motor scrap
Cu Ni Cu Ni utilizing ammonia leaching technique – II. In: Studies on the selective recovery
of copper and iron from motor scrap. Journal of the Mining and Materials
1 0.9 0.572 2.16 0.5 4.193 LD Processing Institute of Japan 109 (5), 337–340.
2 1.8 1.271 1.18 2.0 4.193 LD Ritcey, G., 1983. In: Lo, T.C., Baird, M.H.I., Hanson, C.J. (Eds.), Handbook of Solvent
3 2.7 1.779 2.23 4.0 4.193 LD Extraction. Wiley, NewYork, pp. 673–687.
4 3.6 2.288 1.39 10.0 4.193 LD Van de Voorde, I., Pinoy, L., Courtijn, E., Verpoort, F., 2006. Equilibrium studies of
Nickel(II), Copper(II) and Cobalt(II) Extraction with Aloxime 800 D2 EHPA and
LD: lower than detectable limit. Cyanex reagents. Solvent Extraction and Ion Exchange 4 (6), 893–914.