Studies on copper electrorefining by factorial design methods a pilot plant experience

E. M. Alchtara
Caraiba Metais S.A.
Via do Cobre n "3 700, A.I. 0.- COPEC
Dias D - ~ v i l aBahia,
,
Brasil

ABSTRACT

This paper describes some aspects about the relationship between copper
electrorefining addition reagents and the hydrogen in the cathode. The conditions of
electrorefining were studied by factorial design methods. The advantages of this
technique are reduction in the number of tests, well organized experiments; improvement
in productivity and reliability of results. The dosage of the two main addition agents, glue
and thiourea, were varied simultaneously and compared with the hydrogen content in the
cathode. The influence of these factors, as well their interactions were analyzed by
statistical methods denominated ANOVA (Analysis of Variance). The increase of
thiourea quantities, within of'levels studied, influenced significantly at 95% confidence
in the hydrogen content in copper cathode.

Proceedings of
Copper 99-Cobre 99 International Conference
Volume II-Electrorefining and Electrowinning of Copper
Edited by J.E. Dutrizac J. Ji and V. Ramachandran
The Minerals, Metals & Materials Society, 1999

330

VOLUME I11

INTRODUCTION
The copper electrorefining is a very dynamic and complex process fiom the point
of view of numerous parallels reactions occurrence. Many times the investigation of its
behavior is seen as empirical work, especially when addition agents consumption in
tankhouse is discussed.
Although many efforts (1-4) have been done for organic additives control in the
industrial electrolytic copper refining and for the understanding of their influences on the
copper cathode quality by using overpotential measurements, subjectivity is still present
in procedures adopted nowadays, such as the visual inspection of the cathode deposit.
The overpotential has been used as a powerful response for monitoring the
electrorefining system and is defined as the difference between the applied potential on
the electrodes and the electrochemical equilibrium potential. In this way, the organic
additives, mainly glue and thiourea, have the strongest effect on the cathodic
overpotential in the normal industrial operating practices and are responsible for good
quality copper deposits.
The overall consumption of organic additives is influenced ,by temperature,
sulfuric acid, glue and thiourea concentration in the electrolyte and the presence of
impurities such as metal ions and slime. Of these variables, impurities and slime are
difficult to control and have great responsibility for the empirically operating adjustment
in each refineryindividually.
According to Chia & Sun '(4), there is a linear relationship between the cathode
overpotential, glue and thiourea concentration and hydrogen content in the copper
cathode. In this way, the present work has as purpose to study the glue and thiourea
influence on hydrogen content in the copper cathode using high current density in a pilot
plant. It used the factorial design of experiments (5) for providing a comprehensive
understanding of the effects and their inter-relationship. The advantages of this technique
are reduction in the number of tests, experiments being well organized, improvement in
productivity and reliability of results.
DEVELOPMENT
All experiments were carried out in an independent test section which has four
cells (42 anodes and 43 cathodes per cell), one centrifugal pump for electrolyte
circulation, one heat exchanger and one electrolyte storage tank. The operating conditions
for thc test section1are presented in Table I.

ELECTROREFINING AND ELECTROWINNING OF COPPER

Table I - Typical Pilot Plant Operating Conditions

Electrolyte
Temperature (" C)
Flow rate (Ymin)
Avitone
(ppm)
Copper
(dl)
Sulfuric acid (dl)
Nickel
(dl)
c1(mg/l)
As
(dl)
Sb
(mdl)
Bi
(mg/l)
Fe
(mdl)
Ca
(mdl)
Na
(mdl)
Anode Analysis (ppm)

Others
Current density (Alm2)
Anode spacing (cm)
Cathode life (day)

While all parameters during the experiments were maintained approximately

constant, glue and thiourea quantities were varied according the levels shown in Table II.
The additives control was done by means of monitoring of the tank level and adjustment
of the dosing pump, while the electrolyte bleed off was used for controlling the copper
and sulfuric acid concentration. At the end of a ten days cycle, three cathodes fiom
different cells were randomly selected, one sample of each cathode was drilled and' sent

for hydrogen analysis by LECO analyzer. All results obtained were statistically analyzed
by the well known Analysis of Variance technique (ANOVA) ( 5 ).
Table II - Factors and 'Respective Levels

+ level'

- level

Glue (g/t Cu)

120
60

Thiourea (g/t Cu)

80
40

RESULTS
The experiments were duplicated and their orders randomly chosen are identified
and organized with their respective hydrogen analysis in Table III.
Table IJI - Factorial Design Matrix
Identification
Order Glue
Thiourea
RI*
1
2"
2.4
2
lo
+
0.93
3
3"
+
0.46
4
6"
+
+
0.89
5
5"
1.16
6
7"
+
1.09
4"
+
0.94
7
8
8"
+
+
0.91
* part per million (ppm ) of hydrogen content in cathode.

Rz*
2.3
1.63
0.83
1.19
0.83
1.77
1.35
0.49

R3*
1.9
2.2
0.91
1.47
1.0
0.7
1.06
0.55

The test of significance, of .glue and thiourea as well their .interactions,are shown
in table [Link] Ratio (FbbJ.at95%confidence from Snedcor's statistical table for
1 x 23 degree of freedom is 4;28. This number was used' for significance test and was
compared with the calculated F .ratio (F,*.) from ANOVA7s table for each factor
specifically. 'The glue and gluelthiourea :interaction did not influence significantly the
hydrogen. content in copper cathode p,,~,< Fbb), while thiourea had influenced
significantly at 95% confidence [Link] study (FcaIc> Ftab).

ELECTROREFINING AND ELECTROWINNING OF COPPER

Table IV
Source of
variation
Glue
Thiourea
Interaction
Residual
SQT

Sum of
Square
0.1
1.9
0.5
4.6
7.1

Analysis of Variance

Degree of
Freedom
1
1
1
20
23

Mean of
Square
0.1
1.9
0.5
0.23

FCalc. Significance
at 95%
0.3
No
8.5
Yes
2.3
No

To assure that the results provided by ANOVA technique can be considered true,
it is necessary that experimental error must be independent, normally distributed and
present constant variance.
The residues were calculated fiom the difference between real and mean values
and figure 1 shows them plotted according with the order of execution of the
experiments. As can been seen, there is no specific conformation and the error
independence was not violated.

-1 ,o
-1-5
0

10

15

20

25

Order

Figure 1 - Residue versus The Execution Order of Experiments

The residues were also plotted versus the levels of the main variables, glue and
thiourea, (figures 2 and 3), as well as versus the mean of results (figure 4). The variance
was practically constant between the levels of factors and the residues did not present any
special conformation when it was plotted versus the mean of results.

Residue
in

'

"

in -b

..

em--

-wee.-

Residue
1 . 6 0 0 ,

Residue

ELECTROREFINING AND ELECTROWINNING OF COPPER

t-

- &e4
-1 .O

r = 0.999
Mean = 0.0
Standard dev. = 0.47
n=24

e
-0,5

0.0

03

1to

1.5

2.0

Residue
Figure 5 - Normal Probability

DISCUSSION
Independent of the reaction mechanisms occurring, the hydrogen content in the
cathode obtained in this present work was 1.6 ppm when 40 .g/t of thiourea was used. On
the other hand, the hydrogen concentration was around 0.9 ppm when thiourea was
increased to 80 g/t. According to Chia & Sun (4) the optimum overpotential range for
industrial purpose is below - 60 mV and at this value, hydrogen content can be
maintained below 1 ppm. The increase in thiourea consumption when high current
density is used can be explained 'by the increase of the anode dissolution rate. According
to Kolevatova et a1 (6), the thiourea consumption up to 83 % is concentrated into the
anode slime and the remainder is incorporated in the cathode. The impurities contained
in the anode are oxidized easily by oxygen h m sulfate ion decomposition on ,anode
surface ( 7 ), according with the reactions (1 a) and ( 1 b). In this way, the presence of
AS'"), CU* ( 8 )( 9 )
metallic ions in their .high oxidation states such as ~ i ~Fe+3+,, sbCV),
(10) have an oxidizing effect on thiourea

followed by the recovery of ~

0ion ~

Although the hydrogen content in cathode has not been influenced by changing
the glue quantities within the levels of 60 and 120 g/t, it is interesting to observe that
almost all the operating problems, such as bad current distribution, poor electrolyte flow
and distribution, poor electrodes verticality, etc., normally tend to be overcome by
additional increments in the glue addition. This practice may lead to an increase in

cathodic overpotential and simultaneous hydrogen gas evolution may occur when the
current density limit is reached (1 1) (12). This phenomenon is characterized by a porous
and spongy deposit and many times low adherent deposits on cathode and starting sheets
surface. Figure 6 shows a typical metallographic structure of a bad copper deposit
obtained in presence of high cathodic overpotential with simultaneous hydrogen gas
evolution.

Figure 6 - Metallographic Structure of Bad Copper Cathode Deposits (X80)

CONCLUSIONS
The increase of thiourea quantities, within the studied levels, influenced
significantly at 95% confidence the decrease of the hydrogen content in copper cathode.
The increase of thiourea quantities when high current densities are used may occur
because the anode dissolution rate increases and by presence of the metallic ions oxidized
on the anode surface. The ideal glue and thiourea quantities that should be used in the test
section are 60 g/t ( more economic value for glue consumption), and 80 g/t, respectively.

ELECTROREFINING AND ELECTROWINNING OF COPPER

AKNOWLEDGEMENT
The author wishes to thank Engineer Ivan Moraes for his comments and great
help in the organization of this paper, Technical Assistant Isa Pedro and Chemist Jose
Maria for all technical support and hydrogen analysis, respectively.

REFERENCES
S. E. Afifi,, A. A. Elsayed and A . E. Elsheriet, Additive Behavior in Copper
Electrorefining, Journal of Metals, February, 1987,38-41.
C. T. Wang and T. J. O'Keffe, " The Influence of Additives and their Interactions
on Copper Electrorefining", Proceedings of the International Svm~osiumon
Electrochemistrv in Mineral and Metal Processing, P. R. Richardson, S.
Srinavasan and P. Woods, Eds, Vol. 84, 1984,655-670.
C. Maere. and R. Winand, "Study of the Influence of Additives in Copper
Electrorefining, Simulating Industrial Conditions", Proceedin= of Co~uer95 International Conference, W. C. Cooper, D. B. Dresinger, J. E. Drutizac, H. Hein
and G. Ugarte , Eds, The Metallurgical Society of CIM, Vol 111, 1995, pp 267286.
E. H. Chia and Y. Y. Sun, Organic Additives: A source of Hydrogen in Copper
Cathodes, Journal ofMetals, April, 1987,42 - 45.
D. C. Montgomery, Designing and Analvsis of Experiment, John Wiley & Sons,
Inc, 3" Ed, New York, USA,1.991 b..
V.S. Kolevatova, V.I. Korobkov, and A. I. Levin, The Role of Thiourea in
Electrorefining of Copper. Tsvetnve Metally, UDC no 669.337:546.86,27.
L. I. Antropv , Theoretical Electrochemistry, Mir Publishers, Moscow, Russian,
1977,467 - 468.
K. I. Baikenov, V. B. Syzdykbaeva and E. A . Buketov , Effect of Thiourea and
Nickel on Antimony solubility in Copper Electrolyte, Tsuetnve Metally, UDC no
669.34:621.039.85,38 - 40.

9.

F. Noguchi, H. Itoh and T. Nakamura, "Effect of Impurities on the Quality of

electrorefined Cathode Copper; Behavior of Antimony in the Anode, Proceedings
of Copper 95 - International Conference, W. C. Cooper, D. B. Dresinger, J. E.
Drutizac, H. Hein and G. Ugarte ,Eds, The Metallurgical Society of CIM, Vol Ill,
1995,337-348.

10.

E. Zhang, "Evaluation of Thiourea Consumption for Gold Extraction from

Complex and Refractory Gold Ores, G. W. Warren, Eds., The Minerals Metals &
Materials Society, USA, 1995,443-458.

1 1.

H. Chagas, Introduc&. a Ennenharia Eletroquimica- Processos Eletrometalunicos,

AssociqSio ~rasileirade Metais, Silo Paulo, Brasil, 1985,113-114.

12.

D. A. Uceda, J. D. Scott and T. J. O'Keefe, "Electrochemical Evaluation of Gas

Sparging Effects on Copper Deposition in the Presence of Additives",
tProceedin~sof Copper 91 - International Conference, W. C. Cooper, D. J. Kemp,
G. E. Lagos and K. G. Tan, Eds, The Metallurgical Society of CIM, Vol. ID,
1991,pp 303-3177.