Factors affecting chalcopyrite dissolution in NaCl-CuCl2 solutions

Transcription

1 Factors affecting chalcopyrite dissolution in NaCl-CuCl2 solutions M. Lundström, J. Aromaa, O. Forsen Helsinki University of Technology, Laboratory of Corrosion and Material Chemistry P.O. Box HUT, Finland. O. Hyvärinen, M.H. Barker Outokumpu Research Oy P.O. Box 60, Pori, Finland ABSTRACT In this study we aim to identify the factors affecting the rate of chalcopyrite leaching in the presence of cupric ions, in concentrated chloride solutions. Experiments were carried out near the boiling point of the solution (T = 90 C) at atmospheric pressure with [NaCl] = 280 g/l and [Cu 2+ ] = 1-40 g/l at ph = 2. Electrochemical methods such as cathodic polarization and cyclic voltammetry were used to investigate the solution reactions, specifically the cathodic reactions of copper complexes. To identify the nature and the rate-controlling steps of the reactions rotating disk electrodes were also used. The environment studied is similar to that of the Outokumpu HydroCopper TM process. HydroCopper TM is a novel process for leaching copper from chalcopyrite using cupric chloride solution. The dissolution rate of chalcopyrite depends on the formation of a reaction product layer on the mineral surface. Also thermodynamic and/or kinetic factors of cathodic reactions can affect the nature of the leaching. The results from this study suggest that in the typical redox-potential range of an HydroCopper TM type environment, the dissolved copper exists mainly as the complex [CuCl] +. The diffusion coefficient and the unit rate constants for the solution species were calculated. Also the exchange current density and rate constant for electron transfer was estimated. A simulation was made of the cathodic polarization curve and compared with the experimental results. The cathodic reaction rate did not appear to limit the rate of chalcopyrite dissolution in concentrated cupric chloride environment. 1

2 1 Introduction As sulfur dioxide emissions from the pyrometallurgical industry are highly undesirable, the development of alternative hydrometallurgical process options for sulfide minerals is of great importance from an environmental viewpoint. There is an incentive to develop economically beneficial processes, like HydroCopper TM, which turn the sulfur coming from the sulfide minerals into elemental form. Chalcopyrite, CuFeS2, is the most common copper mineral, available in large quantities and with widespread distribution across the globe. In the presence of cupric ions it is reported to dissolve according to reaction (1) forming elemental sulfur (Habashi, 1978; Hietala and Hyvärinen, 2003; Hyvärinen et al., 2002; Olper and Maccagni, 2003). CuFeS2(s) + 3Cu 2+ (aq) 4Cu + (aq) + Fe 2+ (aq) + 2S 0 (s) (1) In concentrated chloride solutions, copper ions readily form complexes. Thus the leaching reaction can also be suggested to progress in the presence of cupric and cuprous complexes (2) (Wilson and Fisher, 1981). CuFeS2(s) + 3[CuCl] + (aq) + 11Cl - (aq) = 4[CuCl3] 2- (aq) + FeCl2(aq) + 2S 0 (s) (2) Although a lot of hydrometallurgical studies in the field of chalcopyrite leaching has been carried out, the nature of leaching and rate limiting factors are not totally understood. One charasteristic behaviour of chalcopyrite mentioned is the formation of a a sulfur-rich layer on the mineral surface during leaching (Hackl et al., 1995; Munoz et al., 1979; Parker et al., 2003; Roman and Benner, 1973). Earlier (Lundström et al., 2005) it was shown that the process parameters can significantly affect the dissolution rate of chalcopyrite and the properties of the reaction product layers on the mineral. Increasing the temperature from 70 to 90 C doubled the corrosion current density of chalcopyrite. A critical cupric ion concentration of ca. 9 g/l was found, below which the changes in the cupric ion concentration did not affect significantly the dissolution rate. The corrosion potential followed the Nernst equation, with a slope of 60 mv/decade. When the cupric concentration was above 9 g/l the increase in the cupric ion concentration was observed to increase the dissolution rate of chalcopyrite. Solution ph was shown to have a remarkable effect on the composition of the reaction product layers as well as on the dissolution rate. The change in the electrochemical behaviour of the system was observed between ph 2 and The decrease in the corrosion potential was ca. 20 mv/ph unit at ph < 2.5, whigh suggests that the reaction product layer formed reacts with protons. The current densities in the anodic polarization curves were higher when the ph was above 2 and the potentiostatic measurement and cyclic voltammograms indicated that the reaction mechanism changed somewhere between ph 2 and Quantitative calculations from the SEM analysis and visual observations suggested that at ph < 2.5 the reaction product layer on the chalcopyrite surface was sulfur-rich and chalcopyrite was rapidly passivated. At higher phs the reaction product layer was a two-phase layer consisting mainly of iron and oxygen (in the ratio 1:2 to 1:3). This was suggested to be a hydrated iron oxide or goethite type layer. The corrosion potential of chalcopyrite is in the range 450 mv to 600 mv vs. Ag/AgCl, depending on the process parameters (temperature, cupric ion concentration and ph) in concentrated chloride solution. The anodic reaction in the HydroCopper TM process is the dissolution of chalcopyrite. The dissolution was studied earlier (Lundström et al., 2005) and the results suggested that the anodic reaction is either under mixed or chemical control (with low cupric ion concentrations < 9 g/l, ph = 2

3 2, T = 85 C) or under diffusion control through a passivating product layer (with higher cupric ion concentrations 9 g/l, ph = 2, T = 85 C), diffusion coefficient, D, being cm 2 /s. To gain a deeper understanding about the leaching of chalcopyrite, it is essential to also study the nature of the solution and cathodic reactions. The present work focusses on the reactions of the leaching solution and the nature of the cathodic reactions in the leaching process. 2 Materials The concentration of NaCl in the solution was 280 g/l (4.8 M) and the cupric ion concentrations were varied between 1 and 40 g/l. The temperature was 90 C and the ph was adjusted to 2. The working electrode materials used were chalcopyrite, platinum and glassy carbon (GC). Chalcopyrite came from the Pyhäsalmi mine in Finland. The elemental composition of chalcopyrite samples was analysed several times with SEM/EDS. The composition varied, but the average value was Cu 30.0%, Fe 32.5%, S 36.3%, Si 0.5%, Al 0.4% and Mg 0.3% in percentage by weight. The analysed values were near to the theoretical composition of chalcopyrite (Cu 34.6%, Fe 30.4% and S 34.9%). The chalcopyrite electrode surface was polished between every measurement with wetted grade 800 waterproof abrasive paper on a polishing wheel. After polishing the chalcopyrite electrodes were rinsed with de-ionised water, then ethanol and then dried. Glassy carbon was a 5 mm disk in a PTFE sheath. The electrode was prepared at Outokumpu Research, using a Johnson Matthey Type 1 glassy carbon rod (Alfa Aesar, Germany). Since the chalcopyrite is a difficult material to make into electrodes, Pt and glassy carbon were used to get a clearer understanding of the cathodic reactions of the cupric chloride solution. The surfaces of platinum and glassy carbon are shown to be comparable to chalcopyrite for measuring cathodic reactions of solution species. A standard three-electrode electrochemical cell with a thermostated water jacket was employed for the electrochemical measurements. In all measurements (except RDE) the cell was stirred with a magnetic stirrer at 500 rpm. No purging of gases was done. The counter electrode was a platinum sheet or wire, the reference electrode was Ag/AgCl (REF201, Radiometer Analytical, France) placed in a sintered glass tube containing a gel of agar powder, potassium chloride and distilled water. The reference electrode junction was positioned in an external beaker and connected to the cell via a salt bridge and a Luggin capillary. The measurements were carried out using two electrochemical workstations: (i) a PAR 273 Potentiostat/Galvanostat controlled by EG&G PAR s Model 352 Corrosion Analysis Software 1.00 and (ii) a Potentiostat/Galvanostat 2000 working together with a 5050 frequency response analyser (FRA) and 1731 Intelligent/Arbitrary function synthesizer (both NF Corporation, Japan) controlled by in-house software. 3 Procedures The RDE technique was used to study the mass transport as the Levich equation (3) predicts the variation in the transport-limited current as a function of the electrode rotation rate jlim = 0.62 zfd 2/3 ν -1/6 ω 1/2 c (3) where jlim is the limiting current density (A/cm 2 ), z is the number of electrons involved in the reaction, F is Faraday s constant (96485 C/mol), D is the diffusion coefficient (cm 2 /s), ν is the kinematic viscosity (cm 2 /s), ω is the rotation rate (1/s (=2πf)) and c is the concentration (mol/cm 3 ). From the Levich equation the limiting current when plotted as a function of the square root of rotation rate should yield a straight line. If the plot gives a linear response, the system can be 3

4 assumed to be diffusion controlled and the diffusion coefficient can be calculated. On the grounds of equation (3) it can be concluded that (i) jlim is proportional to ω 1/2 when the concentration is constant (Levich plot). On the other hand it follows also that (ii) at constant ω, jlim is proportional to the bulk concentration c. These can be applied to the study of the reaction controlling mechanisms. The corresponding unit rate constant, kt (cm/s), characterising the rate of transport of a solute species to the rotating disc electrode is given by the equation (4) (Gregory and Riddiford, 1956). kt = 0.62D 2/3 ν -1/6 ω 1/2 (4) According to Gregory et al. (Gregory and Riddiford, 1956) a discrepancy in the Levich equation (3) lies in the This coefficient is based on the I( ) value of A better estimation for the value of D is suggested to be achieved by (5) (Gregory and Riddiford, 1956). I( ) = (D/ν) 0.36 (5) for values of (D/ν) in the range cm 2 /s. As the value of D 2/3 / I( ) remains constant, the correlation factor for the Levich equation can be taken into account. The equation (5) is called Gregory-Riddiford equation (MacHardy and Janssen, 2004). If a Levich plot is non-linear at low rotation rates, it might be due to kinetic limitations. A Koutecky-Levich analysis is applicable to first order reactions and used to study the combined effect of diffusion current and kinetic current (Lyons, 2002). This relationship can be described by the Koutecky-Levich equation (6) (Bard and Faulkner, 2001). 1/I = 1/Ik + 1/Id (6) Where I is the total current, Ik is the charge transfer current and Id is the diffusion current. If equation (6) is valid, the plot of I -1 versus ω -1 gives a straight line and the Ik can be calculated from the y-axis intercept. The intercept, near the value of zero, indicates a neglible effect of the charge transfer process on the dissolution rate (Ohno, 1990). Cyclic voltammetry (CV) is a widely practiced electrochemical method and although CV is not ideal for quantitative evaluation of system properties, it is however, a powerful method, because of the ease to interpret qualitative and semi-quantitative behaviour (Bard and Faulkner, 2001). It can also provide rapid and reasonably accurate determinations for reaction parameters, such as rate constants (Nicholson, 1965). The peak height can be presented as a function of diffusion coefficient, concentration and scan rate (7) (Bard and Faulkner, 2001; Sundholm, 1987). (zf) A (RT) 3/2 1/2 1/2 I p D 1/2 0 c0υ (7) Where Ip is the peak height (A), A is the electrode surface area (cm 2 ), D0 is the diffusion coefficient of the reactive species (cm 2 /s), c0 is the bulk concentration (mol/cm 3 ) and υ is the sweep rate (V/s). According to the equation (7), if Ip is linearly dependent of the square root of the sweep rate, the diffusion coefficient can be calculated. This equation (7) at T = 25 C is also called the Randles- Sevčik equation, applicable for a reversible system, for which the heterogeneous electron transfer reaction is fast. The difference between anodic peak potential (Epa) and cathodic peak potential (Epc) is represented by Ep and for totally reversible has a value of 59 mv (T = 25 C) or 72 mv (T = 90 C) for one 4

5 electron transfer (z=1) (Sundholm, 1987). When the electron transfer reaction is quasireversible, the rate of electron transfer is too slow to keep the redox couple in equilibrium as the potential is changed. Thus Ep increases with the scan rate (υ) and can be used to estimate the rate constant for electron transfer, k 0 (method of Nicholson (Bard and Faulkner, 2001; Nicholson, 1965)). The exchange current (i0) is proportional and can be calculated from k 0. The Tafel method can also be used to determine the value of (i0) from which the value of k 0 can be calculated (8) (Bard and Faulkner, 2001). i0 = FAk 0 C (8) 4 Results It is clear that the anodic reaction is the leaching of chalcopyrite. The leaching behaviour of chalcopyrite in the HydroCopper TM environment was reported earlier (Lundström et al., 2005). According to the thermodynamics, the cathodic reaction is the reduction of [CuCl] + (aq) (Lundström et al., 2005). Platinum is a strong adsorber and an excellent electrocatalyst for hydrogen gas evolution. The conventional voltammetric response of a polycrystalline platinum/aqueous solution interface is usually divided into three sections: (i) adsorption and desorption of hydrogen, < 0.4 V vs. NHE, (ii) double layer charging, 0.4 to 0.8 V vs. NHE and (iii) oxygen adsorption and oxide formation, 0.8 to 1.5 V vs NHE. The shape, number and size of the peaks caused by adsorption/desorption processes depend on the crystal faces, electrode pre-treatment, impurities present and the electrolyte. (Bard and Faulkner, 2001; Casey, 1999; Sundholm, 1987) Cyclic voltammetry (not shown) was carried out both with platinum and glassy carbon in a concentrated sodium chloride solution ([NaCl] = 280 g/l) without cupric ions. Glassy carbon did not appear to be very reactive in a concentrated sodium chloride solution. The current densities (from 1.5V to +1.2 V vs. Ag/AgCl) were in the range of -1 ma/cm 2 to 1 ma/cm 2. At lower and higher potentials gas evolution was observed. Also, the magnitudes of the current densities of the reactions of platinum (from 1.0 V to +1.1 V vs. Ag/AgCl) were very low (from -6 to 4 ma/cm 2 ) compared to those measured in a concentrated cupric chloride solution, similar to HydroCopper TM environment (Figure 1). At lower and higher potentials gas evolution was observed. The electrochemical response of platinum suggested characteristic adsorption and desorption processes. As the current densities measured in sodium chloride solution were small compared to those measured in cupric chloride solution (Figure 1), both platinum and glassy carbon can be considered as suitable electrode materials for the study of cathodic reactions in cupric chloride solutions. 4.1 Cathodic reactions Cathodic polarization curves were measured with a platinum RDE. The measurements were carried out at 90 C and ph 2, with 280 g/l of NaCl and the cupric ion concentrations were 1, 10, 20, 30 or 40 g/l. The characteristic curve from the cathodic polarization measurements is similar to that of platinum shown in Figure 1. The curve begins from the corrosion potential, reaching a limiting current (i), which increases both with increasing cupric ion concentration and increasing electrode rotation rate. On the grounds of the experiments done and the thermodynamic considerations (Lundström et al., 2005), it is suggested that at the corrosion potential, copper is present mainly as the cupric complex [CuCl] +, and when the electrode is polarizised in the cathodic direction this reduces to the complex [CuCl3] 2-. At approximately E = V vs. Ag/AgCl (Figure 1) a small peak (ii) is observed with lower, but not with higher rotation rates. The peak around E = V is dependent on the initial cupric ion concentration, after which that the current increases strongly (iii). 5

6 Independent of whether or not the peak was present, precipitated copper was always observed on the platinum electrode surface after the measurement. According to the Pourbaix diagram drawn with the HSC Chemistry package, ([Cu 2+ ] = 26.6 g/l, [NaCl] = 250 g/l, T = 90 C) the precipitation of copper begins at approximately E = -0.2 V vs. Ag/AgCl. It can also be concluded that decreasing the cupric ion concentration decreases the reduction potential of copper. Thus it is suggested that the peak (ii) is the reduction of the complex [CuCl3] 2- to solid copper at approximately E = V vs. Ag/AgCl. The reduction potential depends on the cupric ion concentration. Copper deposition is followed by hydrogen evolution (iii). CURRENT DENSITY (ma/cm 2 ) (iii) (ii) (i) Glassy carbon Platinum CuFeS 2 CuFeS 2, [Cu 2+ ] = 0 g/l POTENTIAL (mv vs. Ag/AgCl) Figure 1. Cathodic polarization curves for glassy carbon, platinum and chalcopyrite RDEs (100 rpm). [NaCl] = 280 g/l, [Cu 2+ ] = 20 g/l (for three upper curves) and [Cu 2+ ] = 0 g/l (for the lowest curve), sweep rate = 0.7 mv/s, ph = 2 and T = 90 C. 4.2 The effect of mass transfer In a concentrated cupric chloride solution, the shape of the polarization curve in the potential range 200 mv to -200 mv vs. Ag/AgCl (Figure 1) shows a limiting current density suggesting that the cathodic reaction ([CuCl] + reduction to [CuCl3] 2- ) is controlled by mass transfer in this potential range. The limiting current densities at E = 0 mv vs. Ag/AgCl with cupric ion concentrations from 10 to 40 g/l were taken from the cathodic polarisation curves and plotted as Levich plots (Figure 2). The response was linear indicating a diffusion controlled process at the potential investigated. Plots of [Cu 2+ ] vs. jlim, when E = 0 V vs. Ag/AgCl, gave a linear response for all the rotation rates studied: 100, 900 and 2500 rpm (10.5, 94.2 and Hz). Diffusion coefficients calculated from the Levich plots and [Cu 2+ ] vs. jlim curves are presented in Table 1, the mean value being 8.33± cm 2 /s. Using Gregory-Riddiford equation (Gregory and Riddiford, 1956) the diffusion coefficient of Cu 2+ gave higher values than with the Levich equation. The D values used in calculations were those shown in the Table 1. The kinematic viscosity was estimated to be cm 2 /s (Lobo and Quaresma, 1989). The diffusion coefficients calculated with the Gregory-Riddiford equation are also included in Table 1. The mean diffusion coefficient was 8.79± cm 2 /s, being approximately 5.5% bigger than that calculated with the Levich method. 6

7 LIMITING CURRENT DENSITY (ma/cm 2 ) Cu 2+ = 40 g/l Cu 2+ = 30 g/l Cu 2+ = 20 g/l Cu 2+ = 10 g/l SQUARE ROOT OF ANGULAR SPEED (1/s) Figure 2. Levich plot with cupric ion concentrations 10, 20, 30 and 40 g/l. [NaCl] = 280 g/l, ph = 2, T = 90 C. Limiting current density taken at 0 mv vs. Ag/AgCl. Table 1. Diffusion coefficient of Cu 2+ ion or complex calculated from the experimental data. E = 0 V vs. Ag/AgCl. [NaCl] = 280 g/l, ph = 2 and T = 90 C. METHOD [Cu 2+ ] (g/l) Rotating speed (rpm) Levich D (cm 2 /s) Gregory-Riddiford D (cm 2 /s) Levich plot Levich plot Levich plot Levich plot j lim -[Cu 2+ ] j lim -[Cu 2+ ] j lim -[Cu 2+ ] The corresponding unit rate constants (kt) characterise the rate of transport of the solution species to the rotating disk. In this case the species are assumed to be cupric ions or complexes. The values of kt calculated are presented in Table 2, which shows that kt is independent of the cupric ion concentration in the concentration range studied and varies as a function of the rotating speed. The average unit rate constants are suggested to be , , , and cm/s with 100, 400, 900, 1600 and 2500 rpm, respectively. 7

8 Table 2. Unit rate constants (kt, cm/s) of cupric ions/complexes calculated from D values (Levich plot) for different cupric concentrations with varying rotation rate. [NaCl] = 280 g/l, ph = 2 and T = 90 C. [Cu 2+ ] = 40 g/l [Cu 2+ ] = 30 g/l [Cu 2+ ] = 20 g/l [Cu 2+ ] = 10 g/l Average Rotating speed (rpm) k T (cm/s) k T (cm/s) k T (cm/s) k T (cm/s) k T (cm/s) When the diffusion coefficient of the cathodic reaction is compared to the value of cm 2 /s measured earlier for the anodic leaching reaction (Lundström et al., 2005), it can be concluded that they are of a different order, the cathodic value being significantly higher. This suggests that mass transfer in the solution does not control chalcopyrite dissolution in concentrated cupric chloride solutions. However, the real HydroCopper TM process does not operate at such low potentials (-200 mv to 200 mv vs. Ag/AgCl) in which the mass transfer is observed to control the cathodic reaction (Figure 1). Thus it is necessary to study the potential range around the open circuit potential of the concentrated cupric chloride solution. 4.3 The effect of charge transfer In the previous section it was shown that mass transfer of the solution species is not the rate limiting step in the leaching of chalcopyrite in cupric chloride solutions. However, Figures 1 and 3 show that in the potential range near the open circuit potential of the studied solution, the cathodic reactions are not totally diffusion controlled. This potential range is also of the greater interest, when considering the operation of the real HydroCopper TM process. Thus it is important to find out the effect of charge transfer on the dissolution process in the cathodic reaction. Cyclic voltammograms were carried out with a glassy carbon electrode to study the cathodic reaction in cupric chloride solutions near the corrosion potential of chalcopyrite in the same solution. Each measurement consisted of ten cycles in the potential range E = 200 to 900 mv vs. Ag/AgCl. The 6 th sweeps, after which the system had achieved a steady state, are presented in Figure 3. The peak currents of the anodic and cathodic reaction in the solution were of the same magnitude and similar in shape and area for the scan range studied ( mv/s). The average ratio of the anodic and cathodic peak currents was 1.0±0.1, this was independent of the scan rate. Since the anodic and cathodic peak currents were equal, the oxidized and reduced forms of the ion or complex must have similar diffusion coefficients. Both the anodic and cathodic peak currents varied linearly with the square root of the scan rate for scan rates between 5 and 200 mv/s (correlation coefficient >0.99 in all cases). D0, calculated from the slope of the Randles-Sevčik plot, was approximately cm 2 /s being twice as much as the value presented in Table 1. That can be explained in Figure 3, which shows that the reaction does not represent a totally reversible system, as stated for Randles-Sevčik equation (7), as the value of ΔEp is not constant. That causes an error in the calculated D0 values. 8

9 100 Current Density (ma/cm 2 ) mv/s 100 mv/s 50 mv/s 20 mv/s 10 mv/s Potential (mv vs. Ag/AgCl) Figure 3. Cyclic voltammetry as a function of scan rate (200, 100, 50, 20 and 10 mv/s) for stationary glassy carbon electrode (6 th sweeps, Ecorr 900 mv 200 mv). Solution had [Cu 2+ ] = 20 g/l, [NaCl] = 280 g/l, ph = 2 and T = 90 C. The peak-to-peak potential separation, ΔEp, for Glassy carbon in cupric chloride solution ([Cu 2+ ] = 20 g/l and [NaCl] = 280 g/l) varied from 100 to 200 mv between 10 and 200 mv/s. This behavior is characteristic of A quasi-reversible system. This means that near the open circuit potential the reaction kinetics are determined not only by diffusion but also by electron transfer. The standard rate constant for electron transfer, k 0, was calculated from ΔEp by Nicholson analysis (Nicholson, 1965) but did not appear to give reliable k 0 and i0 values for the temperatures used. The approximation of the exchange current density (j0) value was done by the Tafel method from the cathodic polarization curves measured with a platinum working electrode. The exchange current density was independent of the cupric ion concentrations ([Cu 2+ ] = g/l) and gave an average value of 4.6 ma/cm 2 (corresponding to an exchange current (i0) value of 0.83 ma and k 0 value of cm/s). Also the magnitude of k 0 supports the assumption, that the system studied is quasireversible near the corrosion potential, while 0.3 υ 1/2 k υ 1/2 cm/s (Bard and Faulkner, 2001). It is clear that the j0 value for the cathodic reaction (4.6 ma/cm 2 ) is larger than that of the anodic reaction (<1.2 ma/cm 2 ) (Lundström et al., 2005). 5 Simulation of polarization curves If the system has no mass-transfer effects and the surface concentrations do not differ appreciably from the bulk values, the current measured can be expressed by the Butler-Volmer equation. It is mentioned to be a good approximation for i values, which are smaller than 10% of the limiting current (Bard and Faulkner, 2001). Under the conditions i = ilim the current follows the Levich equation (3). In a system under mixed control, it can be studied with the Koutecky-Levich equation (6). Levich plots at E = 0 mv vs. Ag/AgCl indicated that the system studied is diffusion controlled at relatively high over potentials (η). The value of the diffusion coefficient at the area of limiting current density (Figure 1, (i)) was calculated to be 8.33± cm 2 /s. On the other hand, cyclic voltammograms in Figure 3 showed that at potentials near to the open circuit potential, the system 9

10 is controlled also by the electron transfer. As the system is quasireversible at small cathodic overpotentials and diffusion controlled at higher cathodic overpotentials, the polarization curve can be simulated by using the Koutecky-Levich equation (6). Based on the experimental data ( [Cu 2+ ] = 20 g/l, [NaCl] = 280 g/l, T = 90 C) the equilibrium potential was assumed to be 600 mv vs. Ag/AgCl in all the simulated curves. Ik was calculated from the Butler-Volmer equation (α is assumed to be 0.5 and i0 estimated with Tafel method) and Id from the Levich equation. Figure 4 shows that the experimentally measured and simulated polarisation curves are in good agreement. Some deviation near the equilibrium potential can be caused by the estimated α value, which has an effect of the curve shape. Also i0 value, which is estimated with Tafel method, affects on the same area. Equilibrium potential is dependent on the cupric ion concentration and does not stay constant in the simulated range, which causes also some scatter CURRENT DENSITY (ma/cm 2 ) Simulated, [Cu 2+ ] = 40 g/l Simulated, [Cu 2+ ] = 30 g/l Simulated, [Cu 2+ ] = 20 g/l 1 Simulated, [Cu 2+ ] = 10 g/l Simulated, [Cu 2+ ] = 1 g/l Experimental, [Cu 2+ ] = 20 g/l Experimental, [Cu 2+ ] = 10 g/l POTENTIAL (mv vs. Ag/AgCl) Figure 4. Experimental ([Cu 2+ ] = 10 and 20 g/l) and simulated ([Cu 2+ ] = 1-40 g/l) polarisation curves with [NaCl] = 280 g/l, [Cu 2+ ] = 20 g/l, T = 90 C, ph = 2, ω = 1600 rpm, i0 = A, D = cm 2 /s and α = Conclusions The cathodic reactions under the conditions studied are suggested to be (i) the reduction of [CuCl] + to the complex [CuCl3] 2-, (ii) the reduction of [CuCl3] 2- to solid copper at E V vs. Ag/AgCl and (iii) hydrogen evolution. From the cathodic polarization curves, the diffusion coefficient for [CuCl] + calculated with the Levich method was 8.33± cm 2 /s and the value corrected with the Gregory-Riddiford method was 8.79± cm 2 /s. When compared to the value of the anodic diffusion coefficient published earlier ( cm 2 /s), it was concluded that the mass transfer of the cathodic reaction does not limit the chalcopyrite dissolution in a HydroCopper TM type environment. The exchange current density was estimated to be 4.6 ma/cm 2. The magnitude of the exchange current density compared to the corrosion current densities of chalcopyrite published earlier (<1.2 ma/cm 2 ) support the statement that not either the charge transfer of the cathodic reaction is controlling chalcopyrite leaching. Simulated cathodic polarization curves, based on the calculated values, were in good agreement with the experimental curves. However, even small changes in the value of α or i0 affect remarkably on the shape of the curve near the equilibrium potential and thus the kinetics in this area must be studied further. In 10

11 general, it can be concluded that the cathodic reactions do not limit the chalcopyrite leaching in a HydroCopper TM process solution. 7 Reference Bard, A.J. and Faulkner, L.R., Electrochemical methods Fundamentals and Applications. John Wiley & Sons. inc., 240,335,347 pp. Casey, D.P., Anomalous Redox Behaviour of Platinum Electrodes in Aqueous Media. M.Sc Thesis Thesis, National University of Ireland, Cork, 148 pp. Gregory, D.P. and Riddiford, A.C., Transport to the Surface of a Rotating Disc. Journal of the Chemical Society: Habashi, F., Chalcopyrite its Chemistry and Metallurgy. McGraw-Hill, Chatham, 165 pp. Hackl, R.P., Dreisinger, D.P., Peters, E. and King, J.A., Passivation of chalcopyrite during oxidative leaching in sulfate media. Hydrometallurgy, 39(1): Hietala, K. and Hyvärinen, O., HydroCopperTM - A New Technology for Copper Production, Alta 2003 Copper Conference, Perth, Australia, pp Hyvärinen, O., Hämäläinen, M. and Leimala, R., Outokumpu HydroCopperTM Process A Novel Concept in Copper Production. In: E. Peek and G. van Weert (Editors), Chloride Metallurgy 2002, 32nd Annual Hydrometallurgy Meeting. MetSoc, Montreal, Quebec, Canada, pp Lobo, V.M.M. and Quaresma, J.L., Handbook of Electrolyte Solutions Part B. Elsevier, Coimbra, Portugal, pp. 377, 1603, Lundström, M., Aromaa, J., Forsén, O., Hyvärinen, O. and Barker, M.H., Leaching of chalcopyrite in cupric chloride solution. Hydrometallurgy, 77: Lyons, M.E.G., Mediated Electron Transfer at Redox Active Monolayers. Part 3: Bimolecular Outer-Sphere, First Order Koutecky-Levich and Adduct Formation Mechanisms. Sensors, 2: MacHardy, S.J. and Janssen, L.J.J., The diffusion coefficient of Cu(II) ions in sulfuric acidaqueous and methanesulfonic acid-methanol solutions. Journal of Applied Electrochemistry, 34: Munoz, P.B., Miller, J.D. and Wadsworth, M.E., Reaction Mechanism for the Acid Ferric Sulfate Leaching of Chalcopyrite. Metallurgical Transactions B, 10B(June): Nicholson, R.S., Theory and Application of Cyclic Voltammetry for Measurement of Electrode Reaction Kinetics. Analytical Chemistry, 37: Ohno, H., Nishihara, H, Aramaki, K, The protection of iron against corrosion with polymer films prepared by cathodic polymerization of halogenated xylenes. Corrosion Science, 30(6/7): Olper, M. and Maccagni, M., The modified Ecuprex process: A promising hydrometallurgy approach for chalcopyrite-bearing copper concentrates. In: P.A. Riveros, D.G. Dixon, D.B. Dreisinger and J.H. Menacho (Editors), Copper 2003-Cobre Met Soc, Santiago, Chile, pp Parker, A., Klauber, C., Kougianos, A., Watling, H.R. and van Broswijk, W., An X-ray photoelectron spectroscopy study of the mechanism of oxidative dissolution of chalcopyrite. Hydrometallurgy, 71(1-2): Roman, R.J. and Benner, B.R., The dissolution of copper concentrates. Minerals Science and Engineering, 5(1): Sundholm, G., Sähkökemia 502. Otakustantamo, Hämeenlinna, 240 pp. Wilson, J.P. and Fisher, W.W., Cupric Chloride Leaching of Chalcopyrite. Journal of Metals, 33(2):