SciELO - Scientific Electronic Library Online

Home Pagealphabetic serial listing  

Services on Demand



Related links


Brazilian Journal of Chemical Engineering

Print version ISSN 0104-6632

Braz. J. Chem. Eng. vol. 14 no. 3 São Paulo Sept. 1997 


B. Gabai1, N.A.A. dos Santos1, D.C.S. Azevêdo1, S. Brandani2 and
C.L. Cavalcante Jr.

1Universidade Federal do Ceará - GPSA - Grupo de Pesquisa em Separação por Adsorção
Departamento de Engenharia Química/CT - Campus do Pici - Bl. 710 - CEP - 60455-760 - Fortaleza - CE - Brazil
Phone: (55)(85) 288-9739/288-9613 - Fax: (55)(85) 288-9601/288-9752 -

2University of L'Aquila - Department of Chemical Engineering - 67100, L'Aquila, Italia


(Received: March 5, 1997; Accepted: August 5, 1997)


Abstract: Selective adsorbents have become frequently used in industrial processes. Recent studies have shown the possibility of using adsorption to separate copper refinery electrolyte contaminants, with better results than those obtained with conventional techniques. During copper electrorefinning, many impurities may be found as dissolved metals present in the anode slime which forms on the electrode surface, accumulated in the electrolyte or incorporated into the refined copper on the cathode by deposition. In this study, synthetic zeolites, chelating resins and activated carbons were tested as adsorbents to select the best adsorbent performance, as well as the best operating temperature for the process. The experimental method applied was the finite bath, which consists in bringing the adsorbent into contact with a finite volume of electrolyte while controlling the temperature. The concentration of metals in the liquid phase was continuously monitored by atomic absorption spectrophotometry (AAS).
Keywords: Electrorefining, adsorption, copper.




During the electrorefining of copper, many impurities are introduced to the electrolyte through the dissolution of an impure copper anode. Contaminants such as antimony, bismuth and arsenic may be found as dissolved metals in the anode slime which forms on the electrode surface, accumulated in the electrolyte or incorporated into the refined copper on the cathode by deposition. The removal of these impurities by conventional techniques involves the use of electrolysis processes in liberating tanks or bleeding the electrolyte in multi-stage electrowinning. In general, the mixture of deposited copper-arsenic-antimony-bismuth is recycled back into the smelter, to recover most of the copper from the slime. However, the use of these conventional methods often results in an insufficient reduction of contaminant concentrations. Besides, it presents disadvantages such as the need to reduce and control the copper concentration level in the electrolyte to facilitate the removal of impurities that may damage the quality of the final product (refined copper).

Recent studies have demonstrated the possibility of using adsorption on separation of heavy metals in several industrial processes, including copper refining, to remove electrolyte contaminants, with better results than those obtained with conventional techniques. The adsorbents mainly cited in the literature for metals removing processes are chelating resins and activated carbons.

This paper studies the possibility of removing antimony and bismuth from copper refinery electrolyte by selective adsorption. As adsorbents, activated carbon, synthetic zeolites, and chelating resins were tested to select the more adequate adsorbent to this separation, as well as the best operating temperature for the process.



Resistance to Electrolyte Acidity

Before initiating any of the experiments, resistance to the acidity of the electrolyte (pH @ 0) was checked for some of the adsorbents. Although zeolites are known to resist only to pHs as low as 2 (Ruthven, 1988), some samples of zeolites were tested as well. Pellets of Y, 4A, USY (dealuminized) zeolites, silicalite, some samples ofactivated carbon and chelating resins were chosen for this test. A synthetic electrolyte, composed of CuSO4 · 5H2O (Cu+2 @ 42 g/l) and H2SO4 (185 g/l), was prepared. Then, small samples of the adsorbents (@ 0.5 g each) were soaked in 100 ml of electrolyte and maintained for 24 hours at 60°C or 27°C (room temperature). Table 1 shows the results obtained in which only the activated carbons, the USY zeolite and the chelating resins samples remained intact.


Tested Adsorbents

Table 2 summarizes some physical properties of the adsorbents that were used in the finite bath experiments. Zeolite Ultra-Stable Y was provided by Degussa, activated carbon by Carbomafra S/A and chelating resins by Unitika and Rohm & Haas.


Finite Bath

The finite bath immersion method (Azevêdo, 1993), which is widely used to screen adsorbents and temperatures in liquid phase sorption processes, was chosen for the adsorption studies. This method, with a properly designed adsorption cell, illustrated in Figure 1, allows experiments by adsorbent immersion in a finite volume of liquid (containing the sorbate) and satisfies the following conditions:

1) Operational ease in set up, control over the experimental starting conditions and sample collection;
2) Thermostatization of the cell to assure isothermality of the system;
3) Lowest variation in system volume during the experiment;
4) Minimal mechanical damage to the particles to maintain their volume constant over the entire experiment.

The experimental system is sketched in Figure 2 and consists basically of a glass cell immersed in a thermostatic bath. A properly design shaker is used to assure good mixing of the solid/liquid system. Samples are taken with syringes during the experiments without disturbing the overall conditions of the system.


Table 1: Results of the Test of Adsorbent Resistance to Electrolyte Acidity

Adsorbent Behaviour
Activated Carbon Remained undamaged.
UR-3300S Resin Remained undamaged.
Duolite C-467 Resin Remained undamaged.
Zeolite Y Structure softened, degenerating easily when shaken by a glass stick.
Zeolite 4A Solubilized as soon as there was contact
Zeolite USY Resisted, apparently undamaged. However, it’s possible that the structure
may have been weakened.
Silicalite Solubilized.


Table 2: Some Physical Properties of Adsorbents Used

Adsorbent Manufacturer Typical Shape Main Dimension
Zeolite Ultra-Stable Y Degussa Cylindric pellets rp @ 1; L @ 5
Activated Carbon Carbomafra S/A Random rp @ 0.6
Duolite C-467 Resin Rohm & Haas Spherical rp @ 0.28
UR-3300S Unitika Spherical rp @ 0.3


Figure 1: Detail of the Glass Cell (ports on top are for extracting samples with the syringe, and introducing the components and the shaker).


Figure 2: Finite Bath System.


The experimental procedure consists in bringing the adsorbent into contact with a limited volume of electrolyte in the glass cell, under controlled temperature, and continuously monitoring the variations in metals concentrations. The analyses of contaminant concentrations were done by atomic absorption spectrophotometry (AAS) (Interlab/Varian Mod. AA 1475), with hollow cathode lamps of Sb and Bi and air-acetylene flame. In regard to the antimony analysis, copper and nickel appear as interferences in the air-acetylene flame, decreasing the absorption signal, specially in reducing flame. This effect, however, may be eliminated without problems with an oxidizing flame. There is no risk of interference with the other electrolyte components in the analysis for bismuth.

The synthetic electrolyte prepared for the experiments was composed of a solution of CuSO4 · 5H2O and H2SO4 (in concentrations given above) with the addition of dissolved Sb+3 (@ 296g/l) and Bi+3 (@ 142,5g/l) salts.

Before each experimental series, the adsorbent was previously activated. Activated carbon and zeolites were dried in an oven at 150°C to remove moisture occasionally present into the pores. The chelating resins, however, were activated by soaking in deionized water for at least 24 hours.

The experiment began with the addition of the electrolyte to the cell. Then, the adsorbent was dropped into the electrolyte and immediately agitation was initiated to maximize the contact between the liquid and the adsorbent. At predetermined intervals, samples were taken from the electrolyte (@ 3.5 ml) with a syringe, to observe the decrease in concentrations (adsorption kinetics) of antimony and bismuth in the electrolyte during the process. A typical experiment lasted about 3 to 4 hours.

For the equilibrium experiments, the cell with adsorbent and electrolyte was left still for 24 hours (or more) in a thermostatic bath at a constant temperature. Then, a sample of the electrolyte was taken for concentration analysis. For the experiments with activated carbon or zeolite, the electrolyte with the adsorbent was maintained in balloons after the tests. A few samples were extracted several days later, for equilibrium concentration analysis.

After each run with resins, the used resin was separated from the electrolyte and rinsed with deionized water. Then, elution was done with Hcl 6M, followed by a new rinse with deionized water. The resin was maintained in this water to be reused.


Diffusion Model

Ruthven (1984) shows the solution for the uptake curve of a sorbate in a finite bath system:


where Pn is given by the non-zero roots of:




This model was employed for evaluation of the effective diffusivities of antimony and bismuth in the chelating resins.



Activated Carbon

Experimental runs with activated carbon, were performed at temperature of 60°C and 27°C (room temperature). The results showed a very smooth or almost negligible decrease in the contaminants in the electrolyte while under agitation during 4 hours. However, after left still for a few days, there was a noticeable, although only partial, reduction in the level of contaminants. As expected, increased values of solid/liquid ratio decreased the level of contaminants. But even with higher values of amount adsorbed (g metal /g adsorbent), the adsorption kinetics of the activated carbon was too slow, usually taking several days to achieve equilibrium. As previously observed by Reed and Nonavinakere (1992), there was no adsorbate competition between the two adsorbing metals. However, some selectivity favoring Sb over Bi was observed. Using fixed bed adsorption, Toyabe et al. (1987) also observed high Sb sorption preference, with somewhat higher selectivity values. Table 3 shows a summary of experimental results for activated carbon adsorption equilibrium.


USY Zeolite

No reference was found in the literature to the removal of copper electrolyte contaminants using zeolites. However, the removal of heavy metals from industrial wastes using zeolites has been reported (Zamzow and Murphy., 1992; Rupp et al., 1995). Based on these studies, room temperature was initially chosen for the experiments. The results showed that even after several days of the experiment, there was no significant variation in the Sb and Bi concentrations in the electrolyte and the amount of adsorbed metals in the zeolite was still very low (Table 4).

We may infer as well that, due to their chemical structure, zeolites are strongly affected by the very low pH of the electrolyte, and their physical resistance to high acidity hinders any possibility of usage with copper electrolyte at process conditions.


Chelating Resins

Chelating resins have been reported as viable adsorbents for heavy metal removal (Oda et al.,1986; Shibata et al., 1987; Abe and Takasawa, 1987; Sasaki et al., 1991; Dreisinger et al., 1994; Dreisinger and Scholey, 1995; Yang et al., 1995), even for copper electrolyte conditions. In this study, Duolite C-467 and Unitika UR-3300S were tested at temperatures of 45°C and 55°C. Both resins showed high levels of contaminant removal, as may be seen in Table 5.

The sorption kinetics was essentially the same for both resins, as may be observed in Figures 3 and 4. Increasing temperature accelerated the overall adsorption process, as expected (Figures 58), which indicates a diffusion-controlled process. Applying the prposed diffusion model (equation 1), we could estimate values for the diffusivities by matching the theoretical curves against the observed experimental data. Some curves are shown in Figures 911

The estimated values for the effective diffusion coefficients are shown in Table 6. These again confirm that the two resins have very similar behaviors and may be applied in practice for the removal of Sb and Bi. The diffusivity values for Sb are slightly higher than those for Bi at the same temperature.


Table 3: Experimental Results for the Adsorption of Antimonyand Bismuth with Activated Carbon (Equilibrium)

Adsorbent Activated Carbon
T (°C) 27 60
Mass (g) for 100 ml of electrolyte 0.3 - 20 10 - 20
Initial Concentration Sb 261 - 304 225
(mg/l) Bi 128 - 141 124
Final Concentration Sb 149 - 207 130-170
(mg/l) Bi 70 - 109 92 - 125
Capacity Sb 60 0.4 - 0.6 (*)
(1000 ´ g/g Ads.) Bi 22 0.1 - 0.3
Time to reach Sb (days) 48 hours
90% equilib. (min) Bi (days) 48 hours


Table 4: Experimental Results for the Adsorption of Antimony and Bismuth with Zeolites (Equilibrium)

Adsorbent Zeolite
T (°C) 27
Mass (g) for 100 ml of electrolyte 3.5
Initial Concentration Sb 285
(mg/l) Bi 131
Final Concentration Sb 203
(mg/l) Bi 105
Capacity Sb 2.2
(1000´ g/g Ads.) Bi 0.8
Time to reach Sb (days)
90% equilib. (min) Bi (days)


Table 5: Experimental Results for the Adsorption of Antimony and Bismuth with Chelating Resins (Equilibrium)

Adsorbents Resin UR-3300S Resin Duolite C-467
T (°C) 45 55 45 55
Mass (g) for 100 ml of electrolyte 4.3 - 10 4.3 - 10
Initial Concentration Sb 227-232 227-237 216 216-233
(mg/l) Bi 127-139 130 118 118-138
Final Concentration Sb 24-32.6 21.6-36 18.7-26 17.1-30
(mg/l) Bi 7.8-10.2 6-9 6.5-12 5.6-16.5
Capacity Sb 3-5 1-5 2-5 1-5
(1000g/g Ads.) Bi 2-3 1-3 1-3 1-3
Time to reach Sb 30 20 - 15-30
90% equilib. (min) Bi 25 15 - 10-20



Figure 3: Adsorption Kinetics of Sb on resins UR-3300S and Duolite C-467 at 55°C.


Figure 4: Adsorption Kinetics of Bi on resins UR-3300S and Duolite C-467 at 55°C.


Figure 5: Adsorption Kinetics of Sb on resin UR-3300S (m = 10.5g) at 45°C and 55°C



Figure 6: Adsorption Kinetics of Bi on resin UR-3300S (m = 10.5g) at 45°C and 55°C.


Figure 7: Adsorption Kinetics of Sb on resin UR-3300S (m = 6.0g) at 45°C and 55°C.


Figure 8: Adsorption Kinetics of Bi on resin UR-3300S (m = 6.0g) at 45°C and 55°C.


Figure 9: Theoretical diffusion model (full line) versus experimental results (squares) for sorption kinetics of Sb on UR-3300S at 55°C.


Figure 10: Theoretical diffusion model (full line) versus experimental results (squares) for sorption kinetics of Bi on UR-3300S at 55°C.


Figure 11: Theoretical diffusion model (full line) versus experimental results (squares) for sorption kinetics of Sb on Duolite C-467 at 55°C.


Table 6: Estimated Effective Diffusivity Coefficients for Chelating Resins

De (cm2/min)
Resin UR-3300S Duolite C-467
45°C Sb 6.48 ´ 10–7
Bi 3.60 ´ 10–7
55°C Sb 1.69 ´ 10–6 1.57 ´ 10–6
Bi 7.50 ´ 10–7 1.04 ´ 10–6



Several adsorbents were tested for Sb and Bi removal from a copper electrolyte using the finite bath experimental method.

Activated carbons are selective for antimony and bismuth with reasonable adsorption capacities (g metal/g adsorbent). However, the adsorption kinetics for these samples of activated carbon was too slow (15 or more days to reach the equilibrium). So, it may be concluded that these adsorbents are not recommended for industrial applications of this process.

Several zeolites did not physically resist the process conditions, as was expected. For USY zeolite, the results indicated very little selectivity towards the contaminants, with very slow kinetics. So zeolites are not adequate adsorbents for this application either.

Chelating resins definitely presented the best results, as compared to zeolites and activated carbon. They reduced the contamination level considerably and in a short time (about 3 hours under shaking and 24 hours or fewer at rest).

The sorption rate in both resins increased as temperature increased. Both resins, UR-3300S Unitika and Duolite C-467 Rohm & Haas, achieved similar results under all the conditions that were studied. The effective difusion coefficients were very similar for both resins and on the order of 10–6 to 10–7 cm2/min in the temperature range of 45 to 55°C.



C0 , C¥ Fluid phase concentration of sorbate at t = 0 and t = ¥ . (For negligible uptake C0 » C¥), mg/l

De Effective diffusivity, cm2/min

m, (mt , m¥ ) Mass adsorbed (at time t and as t ® ¥ ), mg

pn Nonzero roots of equation 2

rp Particle radius

t Time, min

T Temperature, °C

L (C0 C¥ )/C0 = fraction of sorbate taken up by adsorbent



The support of Caraíba Metais S/A, Departamento de Solos/Universidade Federal do Ceará and ASTEF (Associação Técnico Científica Engenheiro Paulo de Frontin) is gratefully acknowledged.



Abe, S. and Takasawa, Y., Prevention of Floating Slimes Precipitation in Copper Electrorefining. The Electrorefining and Winning of Copper, Hoffmann et al. (Eds.), TMS-AIME, Warrendale, PA, USA, pp. 87-98 (1987).         [ Links ]

Azevêdo, D.C.S., Estudo Cinético e Termodinâmico de Adsorção para o Sistema Etanol-Água Sobre Zeolita 3A. Msc. Thesis, Universidade Federal de São Carlos-São Paulo, Brazil (1993).         [ Links ]

Dreisinger, D.B., Leong, B.J.Y. and Grewal, I., Recent Advances in the Solvent Extraction and Ion Exchange of As, Sb, and Bi from Copper Refinery Electrolytes. 24thannual hydrometallurgical Meeting, Harris et al. (Eds.), Canada, pp. 71-89 (1994).         [ Links ]

Dreisinger, D.B. and Scholey, B.J.Y., Ion Exchange Removal of Antimony and Bismuth from Copper Refinery Electrolytes. Electrorefining and Hydrometallurgy of Copper, Proceedings of Copper ‘95, Volume III, Cooper et al. (Eds.), The Metallurgical Society of CIM, pp. 305-314 (1995).         [ Links ]

McLaughlin, H. S., Regenerate Activated Carbon Using Organic Solvents. Chemical Engineering Progress, July, pp. 45-53 (1995).         [ Links ]

Oda, Y., Hashiuchi, M., Kato, T. and Yonekawa, H., Tamano Chelating Resin Adsorption Process. Metallurgical Review of MMIJ, 3 (2), Nov., Symposium Proceedings, pp. 155-163 (1986).         [ Links ]

Reed, B.E. and Nonavinakere, S.K., Metal Adsorption by Activated Carbon: Effect of Complexing Ligands, Competing Adsorbates, Ionic Strength, and Background Electrolyte. Separation Science and Technology, Hassan et al. (Eds.), USA, 27 (14), pp. 1985-2000 (1992).         [ Links ]

Rupp, M.T.C, Russo, C., Sousa-Aguiar, E.F. and Imenes, E.S., Avaliação dos Parâmetros Dinâmicos na Remoção de Metais Pesados Através da Utilização da Zeólita Mineral Clinoptilolita. Paper, XXIII ENEMP, Maringá, PR, Brazil, (14) (1995).         [ Links ]

Ruthven, D. M., Principles of Adsorption and Adsorption Processes, Wiley, New York, NY, USA (1984).         [ Links ]

Ruthven, D.M., Zeolites as Selective Adsorbents. Chemical Engineering Progress, Feb, pp. 42-50 (1988).         [ Links ]

Sasaki, Y., Kawai, S., Takasawa, Y. and Furuga, S., Development of Antimony Removal Process for Copper Electrolyte. Hydrometallurgy and Electrometallurgy of Copper. Proceedings of Copper ‘91, Volume III, Cooper et al. (Eds.), Elmsford, N.Y., USA, pp. 245-254 (1991).         [ Links ]

Shibata, T., Hashiuchi, M. and Kato, T., Tamano Refinery’s New Processes for Removing Impurities from Electrolyte. The Electrorefining and Winning of Copper, Hoffmann et al. (Eds.), TMS-AIME, Warrendale, PA, USA, pp. 99-116 (1987         [ Links ]).

Tarasenko, Yu. A., Bagreev, A.A., Reznik, G.V. and Strelko, V.V., Reductive Sorption Methods for Extraction of Noble Metals from Solution. Hydrometallurgy ‘94, Jul., Cambridge, England, pp. 517-526 (1994).         [ Links ]

Toyabe, K., Segawa, C. and Sato, H., Impurity Control of Electrolyte at Sumitomo Niihama Copper Refinery. The Electrorefining and Winning of Copper, Hoffmann et al. (Eds.), TMS-AIME, Warrendale, PA, USA, pp. 117-128 (1987).         [ Links ]

Yang, J.H.K., Burban, J.H. and Cussler, E.L., Copper Selective Adsorption with a Microemulsion-Based Resin. AIChE Journal, 41(5), May, pp. 1165-1170 (1995).         [ Links ]

Zamzow, M.J. and Murphy, J.E., Removal of Metal Cations from Water Using Zeolites. Separation Science and Tecnology, Hassan et al. (Eds.), USA, 27, pp. 1969-1984 (1992).         [ Links ]



* To whom correspondence should be addressed.