Hydrometallurgy 67 (2002) 1 7 www.elsevier.

com/locate/hydromet

Recovery of copper, cobalt, and zinc from copper smelter and converter slags
Cu neyt Arslan a,*, Fatma Arslan b
Extractive Metallurgy Division, Metallurgical and Materials Engineering Department, Istanbul Technical University, 80626 Maslak, Istanbul, Turkey b Mineral and Coal Processing Division, Mining Engineering Department, Istanbul Technical University, 80626 Maslak, Istanbul, Turkey Received 4 February 2002; received in revised form 7 June 2002; accepted 9 June 2002
a

Abstract A study on the recovery of copper, cobalt, and zinc from copper smelter and converter slags by roasting with sulfuric acid has been conducted. Acid roasting of slags followed by hot water leaching was carried out to bring the metal values into solution. In the leaching experiments, the effects of roasting time, acid/slag ratio, roasting temperature, and application of thermal decomposition prior to leaching on the metals dissolution extents were investigated. A total of 88% of copper, 87% of cobalt, 93% of zinc, and 83% of iron were extracted in 2 h of roasting at 150 jC and 3:1 acid/slag ratio. Increasing acid roasting temperature and time did not have increasing effect on the Co and Zn dissolution extents, while significant improvements were observed in Cu dissolution. Application of thermal decomposition prior to leaching gave small decreases in metal extractions, but since there was no iron dissolution, it was favoured from the viewpoint of metal recoveries from the leachates due to the elimination of an iron removal step. D 2002 Elsevier Science B.V. All rights reserved.
Keywords: Copper smelter; Converter slags; Cobalt; Zinc

1. Introduction The flash smelting process has been used in copper industry for a number of years and has replaced most of the reverberatory applications, known as conventional copper smelting processes. In this process, due to high copper lockup in the slag, a separate processing system consisting of electric furnace slag cleaning, converter for copper matte blowing and fire refining is required (Traulsen et al., 1982; Jones and Deneys, 1998). Recycling copper converter slag back

* Corresponding author. E-mail address: arslanc@itu.edu.tr (C. Arslan).

to the flash furnace has been practised in some plants and it leads to an increase in the matte grade (Antonioni et al., 1982). In some plants, flash smelting furnace slag (1.6 2% Cu) and converter slag (5 7% Cu) are cleaned in an electric arc furnace, where coke is used as reductant and copper content of waste slag decreases to 0.5 0.8% (Kim, 1982; Suziki et al., 1982). Ausemelt top entry submerged lance technology is also used to recover Cu, Ni, and Co from slags (Hughes, 2000; Vernon and Burks, 1997). The flotation process plays an important role in the evaluation of slags and is utilized in some copper smelting processes (Bota et al., 1995; Edlund and Hussey, 1972).

0304-386X/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 3 8 6 X ( 0 2 ) 0 0 1 3 9 - 1

C. Arslan, F. Arslan / Hydrometallurgy 67 (2002) 17

There are also some hydrometallurgical methods given in the literature for the treatment of slags, such as leaching in nitrate, perchloride, chlorate, and sulfuric acids with and without pressure, ferric chloride, cyanide, ferric sulfate, and ammonia solutions (Lindblad and Dufresne, 1974; Jia et al., 1999; Shelley, 1975; dk, 1980). Prater et al., 1970; Kayadeniz and Sag Smelter slag and sulfur dioxide are the waste products of nonferrous pyrometallurgical processes. Dissolution behavior of Co, Cu, Fe, Ni, and Zn from smelter slag in aqueous sulfur dioxide has also been studied (Ahmed et al., 2000). Dissolution behavior of Cu is found to be fast, but subsequent precipitation lowers its dissolution. Studies on the recovery of Cu, Co, Ni, and Zn from copper converter slags by roasting with ferric sulfate, pyrite, ammonium sulfate, and sulfuric acid followed an and Tu by leaching have been reported (Altundog men, 1997; Tu men and Bailey, 1990; Sukla et al., 1986; Tu men, 1994). In the Black Sea Copper Works (Samsun, Turkey) about 35,000 tons of blister copper is produced annually along with approximately 100,000 tons of slag, which is disposed to a slag area and left there for 24 h for cooling (Sirkeci, 1998). Then, slag is treated by flotation after being crushed and ground. The flotation concentrate is fed to the flash smelting furnace. However, important metals such as Co and Zn are recycled continuously from slag to concentrate (and vice versa) and some of them are lost in tailings. Metal contents of smelter and converter slags, which materialize in copper smelting plants, are very high and should be recovered from an economic viewpoint. The aim of this research is to investigate the possibility of recovering Cu, Co, and Zn present in the slag through a pyrohydrometallurgical route as an alternative to flotation. A mixture of flash smelting and converter slags was subjected to the experimental study and treated by acid roasting followed by either direct hot water leaching or thermal decomposition and then hot water leaching techniques.

2. Theoretical considerations Flash furnace and converter slags, usually fayalitic based, contain Cu, Co, and Zn in the form of silicates and ferrites, as determined from the X-ray diffraction

patterns (Arslan, 1982). Although copper in the slag is usually ferritic in nature, appearance of bornite (Cu5 FeS4), a metastable phase, may be explained by the matte inclusions in the slag. Moreover, metallic copper is also observed by the microscopic analysis of these slags. Standard free energy values (Barn et al., 1977) for the reactions, which probably occur between some noticeable compounds existing in the slag and either liquid or gaseous H2SO4, are plotted vs. temperature in Fig. 1a and b, respectively. Usually, the standard free energies of reactions between silicates (and ferrites) and H2SO4 are negative and decrease as the temperature rises. In contrast, DGTj values of some silicates, such as ZnO SiO 2 and 2ZnOSiO2 approach zero as the temperature increases. As seen from Fig. 1, thermodynamically speaking, the possibility of the reactions to occur between metal silicates and sulfuric acid is decreasing in the order of 2FeOSiO2, 2CoOSiO2, and 2ZnOSiO2 and between ferrites and silicates increasing in the order of ZnO Fe2O3, NiOFe2O3, CoOFe2O3, and CuOFe2O3. On the other hand, the standard free energy change of the reaction between bornite (Cu5FeS4) and H2SO4(l) rapidly goes toward negative values at high temperatures. In standard conditions, DGTj values of reactions between gaseous H2SO4 and silicates (and ferrites) are more negative than those of with H2SO4(l). Thermodynamic possibility of reactions to occur follows the same order in both cases. However, the possibility of reactions to occur with H2SO4(l) seems to be more likely since the vapor pressure of H2SO4 is not going to reach 1 atm until its boiling temperature (International Critical Tables, 1978). Since the vapor pressure of H2SO4 over 200 jC is high, the possibility of reactions with H2SO4(g) is high. On the other hand, the possibility of formation of water during the reactions with liquid or gaseous H2SO4 is considerable. The water formation will decrease the H2SO4 concentration; the vapor pressure of H2SO4 at lower H2SO4 concentrations will be lower at the same temperatures, and thus, the reactions with H2SO4(l) seem to be more likely up to 200 jC. Fig. 2 shows the stability regions of various metal sulfates, basic metal sulfates, and metal oxide phases in relation to roasting temperature and partial pressure of SO2 (Addemir, 1978). Considering the constant partial SO2 pressure, increasing roasting temperature causes the decomposition of metal sulfates to metal

C. Arslan, F. Arslan / Hydrometallurgy 67 (2002) 17

Fig. 1. DGTj temperature relationship of some silicates and ferrites with liquid and gaseous H2SO4.

C. Arslan, F. Arslan / Hydrometallurgy 67 (2002) 17

Fig. 2. Thermodynamic equilibrium conditions for sulfate decompositions in Fe Cu Zn Cd S O system, under 1 atm total and PO2 = 0.2 atm partial pressure (reproduced from Addemir (1978)).

oxides. Thermal decomposition of metal sulfates are in the order of Fe2(SO4)3, CuSO4, ZnSO4, CdSO4, and PbSO4.

quantities, where zinc, cobalt, and nickel were in the form of silicates and ferrites, while iron formed as a fayalite-type matrix. 3.2. Method

3. Experimental 3.1. Material A mixture of flash smelter and converter slags subjected to this experimental study was taken from the feed to the flotation unit in the Black Sea Copper Works, Turkey. Chemical analysis of the slag sample is: 2.64% Cu, 0.095% Co, 0.67% Zn, 47.2% Fe, 0.13% Pb, 0.004% Cd, 0.065% Ni, 8.5% Si, and 1.3% S. By examining the smelter and converter slag samples with optical microscopy and by X-ray diffraction analysis, it was found that copper exists as bornite and ferrites, and also in metallic form, though in small Samples were first crushed and ground before the tests so that 80% of them would pass below 0.1 mm. In the acid roasting experiments, a Heraeus Roka type tube furnace was used. A sample of 5 g was mixed with varying amounts of acid and placed into a quartz boat (10 cm length 1.5 cm width), which was then inserted into the tube furnace with a known temperature profile. Air was blown at a rate of 2.5 L min 1 to sweep the reaction gases out of the tube. After samples were cooled to room temperatures, leaching was carried out in 500 mL water at 70 jC and 400 450 min 1 for 1 h. Acid leaching experiments done after thermal decomposition were carried out in 100 mL solution containing 20 g/L H2SO4 at 70 jC and 400 450

C. Arslan, F. Arslan / Hydrometallurgy 67 (2002) 17

min 1 for 1 h. Two experimental routes were followed after the acid roasting step, namely, hot water leaching + filtration, and thermal decomposition + hot water leaching + filtration. Standard glassware, a heater + magnetic stirrer, and a contact thermometer were employed in leaching experiments. After leaching, the pulp was filtered and all chemical analyses were made using an Atomic Absorption Spectrophotometer.

4. Results and discussion The effects of roasting time, acid/slag ratio, and roasting temperature on the dissolution extents of metals were investigated in the leaching experiments. Increasing acid/slag ratio shifted metal dissolution extent roasting time curves to higher values except copper. A 3:1 acid/slag ratio was to be the optimum (Arslan, 1982). Fig. 3 shows the effect of roasting time on metal dissolutions at 150 jC and 3:1 acid/slag ratio. Under these conditions, 80% of Co, 83% of Zn, and 81% of Fe were dissolved. Increasing roasting time to 4 h increased Cu dissolution to 95%. Fig. 4 shows the effect of roasting temperature on metal dissolutions at 3:1 acid/slag ratio and 2 h of roasting time. In all cases, roasting temperature of 150 jC was chosen to be the optimum. Under these conditions 87% Cu, 87% Co, 93% Zn, and 84% Fe were dissolved.

Fig. 4. Effect of roasting temperature on metal dissolutions at 3:1 acid/slag ratio and for 2 h of roasting time.

Fig. 3. Effect of roasting time on metal dissolutions at 150 jC and 3:1 acid/slag ratio.

Copper flash smelting and converter slags contain considerable amounts of iron in the forms of silicate, magnetite, and metal oxy-ferrites as observed in the case of acid roasting experiments, and they were able to be transformed into the soluble compounds by sulfatizing as a result of roasting with concentrated acid. However, the main purpose of this study was to dissolve Cu, Co, and Zn while keeping Fe in the solid residue as an undissolved compound. Thus, the following experiments aimed at producing insoluble iron oxides by thermally decomposing iron sulfates, formed during acid roasting stage, at a temperature at which not all other metal sulfates are affected. For this purpose, slag sample roasted at 250 jC and 3:1 acid/ slag ratio was subjected to the thermal decomposition at 600, 650, and 700 jC for 0.5 4 h and by blowing 2.5 L/min air to the furnace and then dissolving in hot water followed by filtration. As seen from Fig. 2, increasing roasting temperatures caused an increase in the decomposition of metal sulfates to oxides. As a result of this, metal dissolutions dropped significantly by increasing thermal decomposition temperature. All metal decompositions at given temperatures were investigated, and 650 jC was found to give the best results in terms of Cu, Co, and Zn dissolutions, where Fe dissolution was the lowest (Arslan, 1982). Fig. 5 shows the change in the metal dissolution extents with time in the leaching experiments with 20 g/L H2SO4 at 70 jC after thermal decomposition at

C. Arslan, F. Arslan / Hydrometallurgy 67 (2002) 17

650 jC. Only Cu dissolution extent increased enormously after 2 h of leaching time reaching 37%. Since Co dissolution was almost negligible at all times, it was not shown in the figure. Fig. 6 shows the metal dissolution extent thermal decomposition time (at 650 jC for 2 h) relation. Increasing decomposition time decreased the metal dissolutions. For instance, Cu dissolution decreased from 92% at 0.5 h to 78% at 2 h and was in sulfate form. The reason for the decrease in copper dissolution was not related to the formation of copper oxides, since the dissolution of copper was not increased much (0.4%) in acidic solutions at the given thermal decomposition conditions. It may come from the quick formation of ferrites instead of being in the forms of free oxides. The quick formation of ferrites may be due to the very high iron contents. As a result of thermal decomposition experiments, although there are small decreases in metal dissolutions compared to the ones observed in the case of acid roasting followed by water leaching, it was found that the iron dissolution was almost negligible, which is favored in the next step, recovery of metals from sulfate solutions. In order to increase the metal ion contents of the leachates, solutions could be recycled in the leach circuit until they reach the desirable metal contents. After filtration, copper and zinc could be recovered by H2S precipitation and Co could be precipitated by reduction with H2 gas in autoclaves which is similar to

Fig. 6. Thermal decomposition time vs. total metal dissolutions (in acid and water) at 650 jC.

the process used in Outokumpu Oy (Finland) for recovering Cu, Co, Ni, and Zn from leach solutions produced by sulfatizing roasting of pyrite cinders followed by leaching (Outokumpu News, 1980).

5. Conclusions It was found that roasting the mixture of smelter and converter slags with sulfuric acid caused the sulfation of slag constituents, which is in agreement with theoretical considerations. Results obtained from roasting the slag with H2SO4 and leaching the product with hot water revealed that the highest metal recoveries observed at 150 jC roasting temperature and increasing roasting temperature caused decreased sulfation of Co and Zn compounds in the slag. A total of 88% of copper, 87% of Co, 93 % of Zn, and 83% of Fe were extracted in 2 h of roasting at 150 jC and 3:1 acid/slag ratio, and increasing the roasting time to 4 h increased Cu extraction to 95%. Increasing roasting temperature was favourable for the production of Cu sulfate compounds and 100% of Cu dissolution was observed as a result of roasting at 250 jC and 3:1 acid/ slag ratio for 3 h. Application of thermal decomposition after acid roasting resulted in lower metal dissolutions. However, decreasing Fe dissolution almost down to zero after 2 h of thermal decomposition at 650 jC is favourable from the viewpoint of recovering nonferrous metals from solutions, since there will be no additional payments for removal of iron in the

Fig. 5. Thermal decomposition time vs. metal dissolutions (in acid) at 650 jC.

C. Arslan, F. Arslan / Hydrometallurgy 67 (2002) 17

solutions. Lower Co and Zn dissolution extents could be reasonable owing to their low amounts in the slag and having higher charges for iron removal. Therefore, a small decrease in metal dissolution extents could be acceptable and 79% Cu, 66%v Co, and 41% Zn dissolution extents were achieved at these conditions. As a concluding remark, it can be said that this study has proven the recovery of Cu, Co, and Zn in the slag by the application of acid roasting followed by hot water leaching with acceptable metal extraction extents.

Acknowledgements Authors would like to dedicate this paper to the memory of their beloved teacher and adviser, Prof. Dr. Fuat Yavuz BOR.

References
Addemir, O., 1978. Metallurgical evaluation of complex sulfide ores by sulfatizing roasting and examining on Lahanos Ore. Doctoral Thesis, Istanbul Technical University, (in Turkish). Ahmed, I.B., Gbor, P.K., Ji, C.Q., 2000. Aqueous sulfur dioxide leaching of Cu, Ni, Co, Zn, and Fe from smelter slag in the absence of oxygen. Canadian Journal of Chemical Engineering 78 (4), 694 703. an, H.S., Tu Altundog men, F., 1997. Metal recovery from copper converter slag by roasting with ferric sulfate. Hydrometallurgy 44 (1 2), 261 267. Antonioni, T.N., Diaz, C.M., Garvern, H.C., Landolt, C.A., 1981. Control of the Inco oxygen flash smelting process. In: George, D.B., Taylor, J.C. (Eds.), Copper SmeltingAn Update. AIME, New York, pp. 17 31. Arslan, C., 1982. Evaluation of Black Sea Copper Works smelter slags via acid roasting method, MS Thesis, Istanbul Technical University, (in Turkish). ., Knacke, O., Kubaschewski, O., 1977. Thermochemical Barn, I Properties of Inorganic Substances. Springer Verlag, Berlin. Supplement. Bota, G., Gican, I., Poputa, G., 1995. The beneficiation of the nonferrous metals from secondary resources at regia autonoma a Cuprului DevaRomaina. In: Krstev, B., Golomeov, B. (Eds.), Proceedings of 6th Balkan Mineral Processing Symposium, 18 21 Sep. 1995. The Ministry of Science in Government of Republic of Macedonia, Ohrid, Macedonia, pp. 426 431. Edlund, V.E., Hussey, S.J., 1972. Recovery of Copper from Con-

verter Slags by Flotation. United States Bureau of Mines Report of Investigation, No. 7562, 12 pp. Hughes, S., 2000. Applying Ausmelt technology to recover Cu, Ni and Co from slags. Journal of Metals 52 (8), 30 33. International Critical Tables of Numerical Data Physics, 1978. Chemistry and Technology, vol. VIII. McGraw-Hill, New York and London. Jia, C.Q., Xiao, J.Z., Orr, R.G., 1999. Behavior of metals in discard nickel smelting slag upon reacting with sulfuric acid. Journal of Environmental Science and Health. Part A, Toxic/ Hazardous Substances and Environmental Engineering 34 (5), 1013 1034. Jones, R.T., Deneys, A.C., 1998. Using a direct current arc furnace to recover cobalt from slags. Journal of Metals 50 (10), 57 61. dk, U., 1980. The cyanide leaching of copper in Kayadeniz, I., Sag Ku re, and Ergani slags. Chimica Acta Turcica 8, 299 310. Kim, T.J., 1981. Flash smelting in Korea. In: George, D.B., Taylor, J.C. (Eds.), Copper SmeltingAn Update. AIME, New York, pp. 33 39. Lindblad, K.O., Dufresne, R.E., 1974. Acid leach of copper reverberatory slaga new approach. Journal of Metals, 29 31. Outokumpu News, 1980. Vol. 17, No. 2, Outokumpu Oy, Finland. Prater, J.D., Queneau, P.B., Hudson, T.J., 1970. The sulfation of copper iron sulfides with concentrated sulfuric acid. Journal of Metals, 23 27. Shelley, T.R., 1975. Possible methods for recovering copper from waste copper smelting slags by leaching. IMM Transactions, C 84 (820), C1 C4. Sukla, L.B., Panda, S.C., Jean, P.K., 1986. Recovery of cobalt, nickel, and copper from converter slag through roasting with ammonium sulphate and sulfuric acid. Hydrometallurgy (16), 153 165. Suziki, T., Yanagida, T., Goto, M., Echigoya, T., Kikumoto, N., 1981. Recent operation of Mitsubishi continuous copper smelting and conProcess at Naoshima. In: George, D.B., Taylor, J.C. (Eds.), Copper SmeltingAn Update. AIME, New York, pp. 51 76. Traulsen, H.R., Taylor, J.C., George, D.B., 1981. Copper smeltingan overview. In: George, D.B., Taylor, J.C. (Eds.), Copper SmeltingAn Update. AIME, New York, pp. 1 15. Tu men, F., 1994. Metal recovery from secondary copper slag by roasting with ammonium sulphate. Turkish Journal of Engineering and Environmental Science 18, 1 5. Tu men, F., Bailey, N.T., 1990. Recovery of metal values from copper smelter slags by roasting with pyrite. Hydrometallurgy 25, 317 328. Turkish Copper Inventory, 1998. Sirkeci, A. (Ed.), A Collective Study. Istanbul Mine Exporters Union, Istanbul. Vernon, P.N., Burks, S.F., 1997. The application of Ausmelt technology to base metal smelting, now and in the future. Journal of the South African Institute of Mining and Metallurgy 97 (3), 89 98.