Hydrometallurgy 103 (2010) 1-95

Contents lists available at ScienceDirect

Hydrometallurgy
journal homepage: www.elsevier.com/locate/hydromet

Hydrometallurgy 103 (2010) 2-95

The dissolution of chalcopyrite in chloride solutions Part 3. Mechanisms

Michael Nicola'*, Hajime Mikia, Lilian Velsquez-Yvenes b
a

Parker Centre, Faculty of Minerals and Energy, Murdoch University, Perth, WA 6150, Australia b Department of Metallurgical Engineering, Universidad Catlica del Norte, Antofagasta, Chile

Hydrometallurgy 103 (2010) 3-95 A R T I C L E I N F O A B S T R A C T

Hydrometallurgy 103 (2010) 4-95

Article history: Received 26 January 2010 Received in revised form 2 March 2010 Accepted 3 March 2010 Available online 12 March 2010 Keywords: Chalcopyrite Copper heap leaching Chloride leaching rate Nonoxidative dissolution H2 S C ov ell ite P yr ite Mechanism

In Parts 1 and 2 of this series, which describe the results of a study of the dissolution of chalcopyrite under conditions that could be expected in a heap leaching process for primary copper minerals, it was shown that enhanced leaching of chalcopyrite from several copper concentrates in dilute acidic chloride solutions can be achieved by controlling the potential in a "window" of 560-600 mV (SHE) in the presence of dissolved oxygen. It was also found that the rate is linear and essentially independent of the initial concentration of chloride and cupric ions under these conditions. Furthermore, the rate appears to be largely independent of the source of the mineral and is strongly dependent on the temperature (activation energy = 72 kJ mol -1). In this part, additional kinetic data on the effects of fine pyrite on the rate complemented by detailed mineralogical analysis of the residues will be used to demonstrate that sulfur forms a soluble intermediate such as H2S in the dissolution reaction. A summary of the results of a detailed study of the kinetics of the copper ion catalysed oxidation of H2S by dissolved oxygen is presented which provides further support for a mechanism for the dissolution of chalcopyrite under heap leach conditions which involves an initial step involving non-oxidative dissolution to form H 2S and either cupric ions or a covellite-like surface as the initial products. 2010 Elsevier B.V. All rights reserved.

1. Introduction
The mechanism for the dissolution of chalcopyrite in both sulfate and chloride media is still very controversial which is in part due to the often very different conditions under which experimental work has been conducted. According to Dutrizac (1978), the leaching behaviour of chalcopyrite in ferric chloride and ferric sulfate media is substantially different. Thus, this author believes that different mechanisms are operative for the two systems and, consequently, the data for one medium should not be used to support mechanistic interpretations in the other. In particular, many studies have been carried out in relation to the role of potential in the dissolution of chalcopyrite. However, according to Nicol and Lazaro (2003), the potential region studied in most of the electrochemical studies was above that in which the mineral actively dissolves and the relevance needs to be questioned. Several alternative mechanisms have been suggested involving oxidative, reductive/oxidative, non-oxidative and a combination of non-oxidative/oxidative reactions each of which are reviewed in the following sections.

The oxidative dissolution of chalcopyrite in acidic ferric or cupric solutions can be described by the normal mixed-potential electrochemical model proposed by Jones and Peters (1976), Miller et al. (1981) and others, where the following anodic reaction
CuFeS2 = Cu2+ + Fe2+ + S + 4e (1) is coupled to the cathodic

reactions
4Fe3+ + 4e = 4Fe2+ and/or 4Cu2+ + 4e = 4Cu+ (2)

* Corresponding author. E-mail address: M.Nicol@murdoch.edu.au (M. Nicol).

1.1. Oxidative dissolution

Contrary to conventional expectation that the rate of dissolution should increase with increasing potential at the chalcopyrite surface, the rate does not follow such a trend under ambient conditions and the mineral generally requires elevated temperatures in order to dissolve at acceptable rates. During the dissolution of chalcopyrite at elevated temperatures, elemental sulfur is formed on the surface of chalcopyrite and it has been suggested that this elemental sulfur layer inhibits leaching (Dutrizac, 1981). However, this is no longer accepted as the reason for "passivation", which is now thought to be due to the formation of a relatively thin copper-rich polysulfide layer. Under ambient conditions, this mechanism is appropriate at potentials of above about 0.6 V at which the anodic reaction is operative and is characterized by a rapid initial dissolution rate which

0304-386X/S - see front matter 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.hydromet.2010.03.003

decays to a slow rate at longer times. This has been shown to be the case for potentials within the so-called passive region in the anodic curve for the oxidation of chalcopyrite and is typical of the situation with ferric ions as the oxidant under ambient conditions (Hiroyoshi et al., 2004). The rate of the oxidative process is not significantly affected by the potential between 0.6 and 0.75 V but increases significantly in the trans-passive region above 0.75 V which is not accessible with common oxidants. 1.2. Reductive/oxidative dissolution In a study of oxidative leaching of chalcopyrite with dissolved oxygen and ferric ions in sulfate media, Kametani and Aoki (1985) and subsequently, Hiroyoshi et al. (2000, 2001) found that the rate was enhanced in the presence of high concentrations of ferrous ions in sulphuric acid solutions (5) containing cupric ions. However, they could not explain their results in terms of the typical oxidative model according to which the rate of copper extraction should decrease by (or be independent of) the addition of cupric and ferrous ions because these ions are the products of the anodic and cathodic reactions. In order to interpret the enhancement of chalcopyrite leaching by ferrous and cupric ions, these researchers proposed a two-step reaction model in which, (1) chalcopyrite is reduced by ferrous ions to Cu2S in the presence of cupric ions
CuFeS2 + 3Cu2+ + 3Fe2+ = 2Cu2S + 4Fe3+ (3)

Nicol and Lazaro (2003) this can simply be explained in terms of a nonoxidative process. Variations on this model have been made that include alternative reactions to explain, for example, catalysis of the reaction by silver ions (Hiroyoshi et al., 2002). These variations assume that during the reduction of chalcopyrite, hydrogen sulfide is released which reacts with silver ions to form silver sulphide thereby decreasing the concentration of hydrogen sulphide in the solution. This in turn should increase the critical potential of Cu 2S formation and broaden the potential range in which rapid copper extraction takes place. 1.3. Non-oxidative dissolution Dissolution of chalcopyrite in acidic solutions under non-oxidative conditions by a reaction such as
CuFeS2 + 4H+ = Cu2+ + Fe2+ + 2H2S

(2) the intermediate Cu2S is oxidised by ferric ions

2Cu2S + 8Fe3+ = 4Cu2+ + 8Fe2+ + 2S.

(4)

is not thermodynamically spontaneous ( K =2.8x10 19 at 35 C) under normal conditions. This process was suggested by Parker et al. (1981), Dutrizac (1990) and Ammou-Chokroum et al. (1981) and investigated in more detail by Nicol and Lazaro (2003). The lack of experimental study has been mainly due to the low equilibrium concentrations of the dissolved species in solution and to suggestions by authors such as Ammou-Chokroum (1981) and Hiroyoshi et al. (2001) that apparent non-oxidative dissolution can be attributed to oxidation of the mineral by residual dissolved oxygen. According to Nicol and Lazaro (2003), the latter is most unlikely given the low oxygen concentrations and the slow reduction kinetics of oxygen on chalcopyrite surface under ambient conditions. As a result of ring-disk studies, Nicol and Lazaro (2003), observed that dissolution of chalcopyrite can occur to some extent in the absence of any oxidising reagent with the formation of a detectable soluble sulfur species and they proposed a non-oxidative process based on reaction (5). It was suggested that the rate of reaction is governed by two steps:

The sum of these reactions gives the generally reported reaction for ferric leaching of chalcopyrite in sulfate media. On the basis of a thermodynamic analysis, the model predicts that the formation of intermediate Cu2S and therefore ferrous promoted chalcopyrite leaching will occur when the redox potential of the solution is below a calculated critical potential that is function ofthe ferrous and cupric ion concentrations. Cu 2S is more rapidly oxidised than chalcopyrite and this causes the enhanced copper extraction at low potentials in the presence of cupric and ferrous ions. On the other hand, according to Nicol and Lazaro (2003), this model fails to explain how step (3) can be achieved at potentials in which this process is unlikely to occur from a thermodynamic and particularly kinetic point of view given the over-potentials associated with these reactions on a chalcopyrite surface. Thus, there is no experimental evidence to support reduction of chalcopyrite to chalcocite at potentials above 0.5 V. In addition, although reaction (4) provides a means of oxidising Cu2S, the thermodynamics of this reaction are not appropriate given that oxidation of Cu2S proceeds through the intermediate formation of CuS and that the equilibrium potential for the oxidation of CuS is about 70 mV more positive than that for the oxidation of Cu2S.
(6) CuS = Cu2+ + S + 2e Eo = 0.63V Cu2S = 2Cu2+ + S + 4e Eo = 0.56V

a) rapid dissolution to assure equilibrium between soluble species at the b) rate-determining diffusion of the soluble species away from the mineral
surface. The maximum rate, in terms of H 2S production, of the non-oxidative dissolution reaction was shown to be Max Rate = km (4K)1/4[H+]in which km is the mass transport coefficient for transport of H 2Sfrom the surface of the mineral, K is the equilibrium constant for reaction (5) and [H+] is the surface proton concentration. The calculated rate of this non-oxidative process was shown to approximate that observed. 1.4. Non-oxidative/oxidative model Nicol and Lazaro (2003) extended the non-oxidative model to include the overall oxidative dissolution process. Thus, reaction (5) can be combined with (6) to give the overall reaction. Similar reactions can be written with copper(ll) ions as the oxidant in chloride solutions.
H2S + 2Fe3+ = S + 2Fe2+ + 2H+

surface of mineral and the solid.

Assuming that the rate of reaction (6) is rapid compared to the rate of diffusion of H2S from the surface, the equilibrium at the surface will be perturbed by the removal of H2S by oxidation and the dissolution reaction can continue. The above authors applied Fick's second law at

Furthermore, a mechanism in which iron(ll) acts as a reducing agent and simultaneously iron(lll) is an oxidant is unlikely. Another interesting result is that the rate of dissolution of chalcopyrite in sulfuric acid solution under nitrogen atmosphere was higher in the absence of either cupric and or ferrous ions than in their presence. This result was explained in terms of the effect of residual oxygen by these authors. Similar results were obtained by Ammou-Chokroum et al. (1981), but according to

steady-state for the rate of reaction (6), with some assumptions based on a reanalysis of the data of Tekin et al. (1999) (cited in Nicol and Lazaro, 2003). According to the data of Tekin et al. (1999), the rate of reaction (6) decreases with increasing acidity. The following expression was derived for the flux of copper ions from the surface of the chalcopyrite
J =[H2S]s (feD)0-5cotli[(feD)-5/fem (7)

Cu(II) + H2S = CuS + 2H+

in which J is the flux, [H2S]s is the concentration of H2S in equilibrium with the mineral surface, k is the pseudo 1st-order rate constant in 0.1 M acid, D is the diffusion coefficient and km the mass transfer coefficient for Fe3+ at the surface. This model predicts that the rate of dissolution of chalcopyrite should not increase significantly with addition of ferric ions above about 0.1 M; that increasing concentrations of acid should not increase the rate; and that addition of ferrous ions should reduce the rate of dissolution. The results of the simulated rates of dissolution were found to be comparable with some reported studies of the kinetics of dissolution in acid solutions. lt should be apparent from this brief review of the mechanisms of the dissolution of chalcopyrite in both sulfate and chloride media that there is very little agreement on the nature of the rate-determining step and the mechanisms involved in this important process.

The calculated equilibrium concentrations of the soluble species in a saturated solution of CuS in 0.2 M HCl containing increasing concentrations of chloride (added as NaCl) show that the equilibrium does not allow for measurable concentrations of copper(II) ions (less than 4x10 -9M). However, experimental observations show that addition of sulfide ions at low concentrations (less than 1 mM) to acidic chloride solutions containing copper(II) ions does not result in the immediate precipitation of black CuS, but rather a white turbidity appears after some minutes which is due to the precipitation of elemental sulfur. Thus, dilute solutions of copper(II) ions in the presence of low concentrations of H2S are meta-stable and therefore it is possible that elemental sulfur may not be observed on the surface of partially dissolved chalcopyrite grains. These observations are in accord with those made by Luther (2002) who found that copper(II) ion reduction by sulfide to copper(I) occurs prior to precipitation as CuS (covellite is now accepted as a copper(I) sulfide i.e. Cu2S2) in sea water solutions. In the chloride system, CuS is also not thermodynamically stable in the presence of excess copper(II) ions as shown by the data presented in Fig. 2 which summarizes the thermodynamics of the various possible couples involved in this system for initial concentration ratios of Cu(II)/Cu(I) = 100 and Fe(III)/Fe(II) = 0.001. These ratios were chosen such that these two couples and the couple for the oxidation of CuS (i.e. CuS = Cu(I)+S + e) are all in equilibrium as shown for a chloride concentration of 15 g/L and a total copper ion concentration

2. Experimental

2.1.

Materials

As described in previous paper of this series, chalcopyrite concentrates were obtained by flotation from operations in Chile and the USA. Details of the chemical analysis and mineralogical composition of the materials have been given in Parts 1 and 2.

2.2.

Leaching experiments

Leaching experiments were carried out in mechanically agitated instrumented reactors constructed at Murdoch University using the method described in Part 1 of this series. The rate of the copper-catalysed oxidation of hydrogen sulfide by dissolved oxygen was followed by measuring the rate of consumption of dissolved oxygen using an oxygen probe with the apparatus shown in Fig. 1. An aerated solution (4 mL) with known concentrations of HCl, copper(ll) and chloride ions was placed into the cell which was stirred by a small magnetic bar. The dissolved oxygen probe holder was designed to act like a plunger which fitted the cylindrical glass cell. A small groove machined along the length of the probe allowed air to be expelled as the probe was lowered into the cell. The cell was immersed into a temperature controlled water bath maintained at 35 C. After temperature equilibration, 40 uL of 0.01 M Na2S solution was injected using a thin Teflon tube and micropipette inserted into the groove. The dissolved oxygen concentration was recorded as a function of time using data acquisition and Labview software. All reagents used were analytical grade. The Na 2S solution was analyzed by titration with standard AgNO3 solution. In several comparative experiments, a copper(l) solution was injected into the reactor instead of a Na2S solution. The copper(l) solution was prepared from reagent grade CuCl and HCl and was stored under a nitrogen atmosphere in the presence of a piece of copper wire to minimize oxidation to copper(ll).

3. Results and discussion

3.1. Thermodynamic aspects Stirrer It is well known that copper(ll) ions are not stable in the presence of sulfide ions even in strongly acidic solutions due to precipitation by the reaction
(8)

of 1 g/L. Data for the stability constants for the copper(I) and (II) chloro complexes was taken from the NTIS compilation (Martell and Smith, 2004). As a result of the increased stability of copper(I) in chloride solutions, copper(II) ions can oxidise both CuS and H 2S to elemental sulfur. Obviously iron(III) can fulfill the same role but the important distinction between the Cu(II)/Cu(I) and Fe(III)/Fe(II) couples is that copper(I) ions are rapidly oxidised by dissolved oxygen whereas the corresponding oxidation of iron(II) ions is very slow at ambient temperatures. It is also important to note that the concentration of the oxidant is significantly greater by several orders of magnitude for copper(II) than for iron(III) at the potentials (less than 0.6 V) at which enhanced rates of leaching of chalcopyrite are observed. The potentials in Fig. 2 confirm that CuS is not stable in solutions containing an excess of copper(II) ions even at low concentrations of chloride and that elemental sulfur should be the product of oxidation of H 2S. Thus, even if CuS is the first product of reaction with H 2S, it will be oxidised by copper(II) ions, albeit at a slower rate than could be expected for the homogeneous oxidation of H2S by copper(II) ions. At potentials below about 0.57 V in a solution of 15 g/L chloride, CuS is stable and this is in accord with the observations of low rates of dissolution of chalcopyrite at potentials below the "window" and of a CuS-like surface layer which was observed under these conditions as outlined in Part 1. Note that there are no thermodynamic restrictions to the oxidation of chalcopyrite in chloride solutions with either copper(II) or iron(III) as oxidants. 3.2. Mineralogy of leach residues The deportment of sulfur in the dissolution of chalcopyrite is an important indicator of the mechanism. Detailed mineralogical studies were carried out on feed materials and the residues from a selected number of dissolution experiments. The details of the feed materials and the dissolution experiments have been given in Part 1 of this series. In most cases, mineral liberation analysis (MLA) was used to investigate the mineral particles after dissolution. Thus, Fig. 3 shows typical particles of chalcopyrite and elemental sulfur after controlled potential dissolution of Concentrate 1 (83% chalcopyrite with small amounts of bornite and pyrite see Part 1 for full description) in 0.2 M HCl with 0.5 g/L copper(ll) at 35 C. A characteristic of dissolution at low potentials in chloride solutions is that elemental sulfur does not form on the surface of the chalcopyrite particles as is conventionally assumed in oxidative leaching of sulfide minerals. As shown in these figures, very little of the elemental sulfur formed during dissolution in chloride solutions is associated with chalcopy-rite particles, but occurs largely as isolated spherical globules with an occasional association with chalcopyrite. lnterestingly, under conditions of low acidity, the sulfur forms concentric but isolated rims around chalcopyrite particles as shown in the inset in Fig. 3. These observations support those made by Dutrizac (1990) who, on the basis of microscopic examination of leach residues, suggested that sulfur was formed by oxidation of hydrogen sulfide by ferric ions adjacent to but not on the dissolving mineral surface. ln terms of the non-oxidative model, the actual location of the formation of sulfur will depend on the kinetics of the oxidation of hydrogen sulfide relative to that of its formation by dissolution of the mineral and mass transport from the surface. Thus, high temperatures and strongly oxidising conditions will tend to favour formation of sulfur close to or on the chalcopyrite surface whereas low temperatures and weakly oxidising conditions will result in oxidation of H2S away from the mineral surface. As will be discussed in the next section, addition of fine pyrite can result in enhanced dissolution of chalcopyrite and Fig. 4 shows the formation of elemental sulfur on fine pyrite particles but not on the coarser pyrite particles. ln fact, in this case, no sulfur was detected in the MLA analysis that was not associated with pyrite or, in small amounts, with chalcopyrite. The inference from this observation is that fine pyrite particles act as catalysts for the formation and/or nucleation of elemental sulfur. lndependent studies of the oxidation of pyrite under similar conditions of low potential have shown negligible dissolution confirming that the sulfur observed on pyrite particles did not originate with the pyrite. A possible explanation will be given below. The important conclusion from the mineralogy is that the primary product of

the dissolution reaction at the chalcopyrite surface at low potentials is a soluble sulfur species.

Fig. 3. Mineral maps of chalcopyrite (green) and sulfur (blue-grey) in the residue after dissolution of Concentrate 1 in 0.2 M HCl with 0.5 g/L Cu(II) at 35 C and 580 mV. Inset: As above but pH 2.0 with chalcopyrite (black) and sulfur (grey).

Fig. 4. Mineral maps of pyrite (black) and sulfur (blue-grey) in the residue after dissolution of Concentrate 1 in 0.2 M HCl with 0.5 g/L Cu(II) at 35 C and 580 mV in the presence of fine pyrite.

3.3. Catalysis of the dissolution of chalcopyrite It is well known that catalysis of the leaching reaction is possible by the addition of fine solids such as activated carbon, pyrite and even very fine silica and also soluble species such as silver ions. The mechanisms of these observations are not clear and have involved processes such as galvanic interactions and conversion of chalcopyrite to other more stable sulphides. The effect of pyrite on the leaching of chalcopyrite has been investigated in some detail (Abraitis et al., 2004; Al-Harahsheh et al., 2006; Dutrizac, 1982; Harmer et al., 2006; Parker et al., 2003; Dixon et al., 2008). Most of these investigations have focused principally on the role of galvanic interactions in the leaching of sulfide minerals under oxidising conditions at elevated temperatures. It has been postulated that when chalcopyrite and pyrite are in intimate electrical contact with each other they form a galvanic couple as a result of which chalcopyrite behaves as an anode and dissolves more rapidly, while pyrite behaves as a cathode. However, none of these investigations has directly verified the formation of such galvanic couples by electrochemical experiments. It is also difficult to visualize electronic contact through a sulfur layer which coats the chalcopyrite particles under these conditions. This may be possible under conditions in which the sulfur does not form on the surface of the chalcopyrite. The effect of addition of different amounts and particle size of pyrite to constant amounts of Concentrate 1 ( + 25-38 urn) on the extraction of copper was studied. It can be seen from the results in Fig. 5 that a very rapid rate of dissolution of copper was achieved in the presence of 25 um pyrite in a chalcopyrite/pyrite mass ratio of 1:5. However, when coarser ( + 25-38 um) pyrite was used, the rate of copper dissolution obtained was similar to that in

the test carried out in the absence of pyrite. When a smaller particle size (milled 25 um pyrite) was used the rate was lower than in the presence of 25 um of pyrite. It appears therefore that there is an optimum pyrite particle size. The mineralogical results presented in the previous section showed that sulfur was associated with fine pyrite particles and the apparent catalytic role of pyrite in this system is believed to be due to its effect on the kinetics of the oxidation of hydrogen sulfide which is formed during dissolution at these low potentials and this effect is described in more detail in the next section. Several leach experiments were conducted with the addition of small amounts of silver ions but the results showed no significant catalytic effects. However, the low solubility of silver chloride in dilute chloride solutions (approximately 1 ppm in 0.2 M HCl) may have contributed to this observation.

3.4. Kinetics of oxidation ofH2S The catalytic effect of copper ions on the oxidation of H 2Sin initially aerated solutions containing an excess of dissolved oxygen is shown by the data in Fig. 6A which gives typical traces of the response of the oxygen probe with time at 35 C in a solution containing 0.2 M HCl + 0.8 M NaCl with and without copper(II) ions after the addition of 0.1 mM Na 2S solution. Thus, in the absence of copper ions, the rate is very slow under these conditions, while the half life is about 100 s in the presence of 0.5 g/L cupric ions. As indicated previously, no precipitation of CuS was observed in these experiments. The solution

M. Nicol et al. I Hydrometallurgy 103 (2010) 86-95

kinetics of the oxidation of copper(I) by dissolved oxygen in chloride solutions has suggested a modified rate equation of the form d[Cu(l)] / dt = M^ffll [ 2 + k 3 2[ C u ( l ) J] [ H 1 WJ/ 1 v ; ( 1 + K4[Cu(ll)]) "
2 3 2 2

] ( 12)

'

which is based on the following mechanism, in which copper(I) is oxidised in two parallel reactions, i.e. Eqs. (13) and (14) by oxygen and Eq. (15) by peroxide.
Cu(I) + O2 iCuO2 ( K ) ( 1 3 )

Cu-O2 + Cu(I)2H+ 2Cu(II) + H2O2 (

2Cu(I) + 2Cu(II) + H2O + H+ (

Cu(I) + Cu(II)iCu2 (

The inhibiting effect of copper(II) is ascribed to the formation of a kinetically inert adduct between copper(I) and copper(II) ions in chloride solutions, for which there is evidence in the literature. Reaction (15) is comparable in rate to reactions (13) and (14) and therefore peroxide accumulates in the solution giving rise to enhanced rates at long times. The effect of complexation of copper

remained colorless for several hundred seconds after which a faint white milkiness was observed. Fig. 6B shows the result of a similar experiment in which the addition of sulfide was replaced by that of 0.18 mM copper(I) ions. The curve describing the rate of consumption of dissolved oxygen is almost identical to that in Fig. 6A confirming that the oxidation of H 2S by copper(II) ions (Eq. (9)) is rapid under these conditions and is followed by slower oxidation of the copper(I) formed by dissolved oxygen (Eq. (10)) i.e. H 2S is not directly oxidised by dissolved oxygen.
H2S + 2Cu(II)= 2Cu(I) + S + 2H+ ( r a s l p o i w d ) ) ( ( 1 9 0 ) )

2Cu(I) + 1/2O2 + 2H+ = 2Cu(II)+ H2O (

and the overall reaction is

H2S + 1/2O2 = S + H2O . (11)

At low concentrations ofcopper (less than about 0.2 g/L), the rate decreases significantly and is controlled at least partially by the rate of reaction (9) under these conditions. The kinetic analysis is therefore different at high and low copper concentrations and each will be dealt with separately. 3.4.1. High (> 0.2 g/L) concentrations of copper Under these conditions the rate is controlled by the rate of reaction (8) and the kinetic data can be analyzed by assuming that the applicable rate law is that for this reaction i.e. the H 2S concentration is effectively zero soon after injection into the solution and [Cu(I)]o = 2 [H2S]o. A recent investigation (Miki and Nicol, 2008) of previously published information (Nicol, 1984) on the

M. Nicol etat. /Hydrornetallwgy 103 (2010) 86-95

10

(II) and, particularly, copper(I) by chloride ions has not been included in the above mechanism. Although this overall reaction is important in that re-oxidation of copper(I) is a necessary step in the overall leaching process, it is not directly involved in the oxidation of H2S which is required as part of the non-oxidative dissolution mechanism. Full details of the kinetics and mechanism of the oxidation of copper(I) ions by dissolved oxygen and of the oxidation of H 2S under typical leach conditions will be published elsewhere. 3.4.2. Low (< 0.05 g/L) concentrations of copper At low concentrations of copper(II), the rate of oxygen consumption decreased significantly with decreasing concentration of copper (II) ions which is contrary to the trend at high concentrations. It is therefore suggested that under these conditions, the rate becomes controlled by the rate of the reduction of copper(II) by hydrogen sulfide i.e. reaction (9). Analysis of the kinetic data for the rate of consumption of oxygen under conditions of low copper concentration shows that the use of a first-order rate equation gives a good fit to the data while a second-order equation does not as is obvious from the plots in Fig. 7. Comparison of the data in Fig. 7 with that in Fig. 6 shows that the rate at the low concentration of copper(II) ions is about 25 times slower than at higher concentrations. Thus, one could expect that at low copper(II) concentrations, copper(I) ions generated by oxidation of H2S are rapidly re-oxidised to copper(II) ions by dissolved oxygen and that one can therefore assume that the concentration of copper(II) is maintained constant during an experimental run. The first-order dependence observed therefore suggests that the rate of reaction (9) is first-order in hydrogen sulfide. In the presence of copper ions, the first-order rate constants for the consumption of oxygen obtained as shown in Fig. 7, increase with increasing copper(II) concentration but decrease with increasing acidity as shown by the data in Fig. 8. Inspection of the data reveals an approximately inverse dependence of the rate constant on the acid concentration while the reaction order with respect to the copper(II) concentration is greater than 1 and is 2 in the case of two sets of data. It is interesting to note that an analysis of published data (Tekin et al., 1999) on the kinetics of the similar reaction involving the oxidation of H 2S by iron(II) showed that the rate is 1st-order in iron (III) and inversely dependent on the acidity. The strong inverse dependence of the rate on the acidity can be used to explain the observations in Fig. 3 (inset) in which concentric elemental sulfur rims, around but not on chalcopyrite particles, were only observed in experiments conducted at relatively high pH (around 2). Under these conditions of high oxidation rates (relative to mass transfer from the surface) of H2S, sulfur is formed close to the mineral particles; while at low pH values, oxidation will only be expected to occur further from the mineral surface. The effect of fine pyrite on the consumption of oxygen in the absence and presence of H2S (1 mM) in 4 mL of a solution of 0.2 M HCl is shown in Fig. 9. In the absence of sulfide ions, the rate of consumption of oxygen i.e. the rate of oxidation of pyrite is very low. As shown in Fig. 6, the rate of oxidation of sulfide ion is similarly very low. However, in the presence of pyrite, the oxidation of sulfide is rapid as shown by the complete reduction of dissolved oxygen in less than 300 s. It is obvious that the oxidation of H2S by dissolved oxygen is catalysed by the presence of fine pyrite particles even in the absence of copper ions. This observation is consistent with the results of leaching experiments described above that showed enhanced dissolution rates of chalcopyrite in the presence of fine ( 25 urn) pyrite and the mineralogical observation of sulfur coatings around fine pyrite particles in the residues from the leaching experiments. The observation that only fine pyrite particles are active could be related to the enhanced mass transport of H2S and possibly copper ions and oxygen to the surface of fine particles relative to larger particles. It would be interesting to establish whether pyrite also acts as a catalyst for the oxidation of H 2S by copper(II) ions in the chloride system. It is possible that the mechanism of the action of fine pyrite and other solids on the rate ofdissolution ofchalcopyrite

and has been applied in the Galvanox process (Dixon et al., 2008) also involves this catalytic oxidation of H2S and is not, as suggested in these papers, only a galvanic coupling of chalcopyrite with pyrite in a typical oxidative model.

TO Ci

3.5. The role of oxygen The observation that enhanced rates of dissolution are only observed in the potential "window" in the presence of dissolved oxygen requires that any mechanism should identify the essential role of oxygen. A similar observation in regard to the necessity for dissolved oxygen was made (Hiroyoshi et al., 2000) in studies of the dissolution of chalcopyrite in sulfate solutions and the rather inadequate conclusion made that chalcopyrite was simply oxidised by dissolved oxygen. In all subsequent publications involving their mechanism, the role of oxygen has been largely ignored. It should be recognized that the solutions used in these and the present studies are meta-stable in that the measured EH does not reflect the true equilibrium solution potential in the presence of dissolved oxygen. Thus, it is possible to have a solution which may be almost saturated with respect to dissolved oxygen (Eo = 1.23 V) but still have a measured EH as low as 0.6 V. The EH in a meta-stable system consisting of two or more redox couples is determined mainly by that of the couple with the greatest exchange (17) current density. In the present case, both the copper(II)/copper(I) and iron(III)/iron(II) couples have exchange current densities on a platinum surface which are orders of magnitude greater than that of the oxygen/water couple. In terms of the present proposed mechanism involving a non-oxidative dissolution reaction, it has been demonstrated that the crucial step involving the oxidation of H2S by dissolved oxygen is rapid in the presence of copper(II) ions, but that at the concentrations of copper ions normally present in leach solutions, the rate of the oxidation of H2S by copper(II) ions in the presence of dissolved oxygen is rapid compared to the rate of oxidation of the copper(I) ions formed in this reaction by dissolved oxygen. Thus, H 2S is removed rapidly by a reaction which does not appear to involve oxygen. One possible mechanism involves the initial formation of a copper (II)-HS complex (Rickard and Luther, 2006) which can undergo oxidation by a second copper(II) ion or by oxygen. Thus, writing copper(II) as Cu 2+ and copper(I) as Cu+.
Cu2+ + H2 S = CuHS+ + H+ CuHS+ + Cu2+ = 2Cu+ + H+ + S CuHS+ + 0.5O2 + H+ = Cu2+ + H2O + S Rapid equilibrium Slow Fast

A greater reactivity of the bisulfide complex towards oxygen than copper(II) ions could thereby result in greater rates of mineral dissolution in the presence of dissolved oxygen. A similar mechanism has been invoked (Byerley et al., 1975) for the copper-catalysed oxidation of thiosulfate ions in ammoniacal solutions in which thiosulfate is more rapidly oxidised in the presence of dissolved oxygen. 3.6. Simulation of the kinetics of dissolution As described in a previous publication (Nicol and Lazaro, 2003), it is possible to estimate the rate ofdissolution ofchalcopyrite assuming that reaction (5) is at equilibrium at the mineral surface and that the H2S is transported from the surface during which time it is oxidised by copper(II) ions as in reaction (9). Thus, in this case, the

flux of H2S (or twice the rate of dissolution of copper) from the mineral surface due to combined mass transfer and chemical oxidation is given by Eq. (7). In this case, the pseudo-1st-order rate constant k can be obtained from the data in Fig. 8. The results of such calculations, assuming reaction orders of 2 and 1 for copper(II) and acid respectively are shown in Fig. 10. The relatively small variation in the predicted rate at copper concentrations below about 4 g/L, which is the range normally encounCuFeS2 + 2H+ = CuS + Fe2+ + H2S .

The formation of covellite after an extended period of dissolution of chalcopyrite at low (550 mV) potentials was observed mineralogically as described in Part 1. Fig. 11 compares the rates of dissolution of copper from both chalcopyrite and covellite particles of the same nominal size under

0.0E+00

Fig. 10. Estimated rates of dissolution of chalcopyrite at 35 C in solutions of various initial copper(II) concentrations at acid concentrations of () 0.2 M HCl, (A)1.0 M HCl.

teredinheapleach operations, is partially consistent with the observation in Part 2 that the rate of leaching is insensitive to the initial concentration of copper ions. On the other hand, the relatively large predicted effect of the acidity on the rate does not agree with the negligible effect of acidity on the measured rate as shown in Part 2. From the slope of the linear dissolution curves for Concentrate 1 in a solution initially containing 0.5 g/ L copper ions in 0.2 M HCl at 35 C given in Part 2, an initial rate of dissolution of approximately 1.3x10 12molCucm2 s1 can be estimated which is about two orders of magnitude greater than that predicted under similar conditions in Fig. 10. The observation in Part 2 that high concentrations of iron(II) also inhibited the rate is consistent with this mechanism. Thus, although most aspects of the non-oxidative dissolution mechanism are in agreement with the experimental observations, there are some experimental aspects which are not consistent with the mechanism. 3.7. An alternative mechanism Although the non-oxidative dissolution Eq. (5) appears to be consistent with some of the data, an alternative reaction to Eq. (5) which is thermodynamically considerably more favourable ( K =2.8x10 3at35 C) is as follows This mechanism is consistent copper(II) and 1 for protons as with the reaction orders of 2 for observed.

similar conditions. On the basis of the very similar rates, it is tempting to suggest that reaction (17) is occurring rapidly on the surface of the mineral and that the CuS-like layer is oxidatively dissolved at a rate similar to that for bulk covellite. Furthermore, mineralogical (MLA) examination of the residue after 63% dissolution of synthetic covellite under the conditions of Fig. 11, showed that a significant amount of the sulfur is formed on the covellite particles. However, it occurs as isolated islands and greater than 95% of the covellite surface is free of sulfur. In terms of reaction (17) at least 50% of the sulfur from the chalcopyrite is initially formed as a soluble intermediate. Furthermore, our yet to be published studies on covellite dissolution under these conditions, show that the rate of dissolution is independent of the initial concentration of copper ions, the acidity and the chloride concentration within the ranges to be expected for a heap leaching operation. Although there are still some uncertainties, it is therefore suggested that the non-oxidative reaction (17) is the preferred first step in the overall reaction of dissolution of chalcopyrite in chloride solutions at potentials below about 600 mV.

chalcopyrite in both chloride and sulfate solutions. While the alternative reductive/oxidative mechanism described in Section 1.2 is similar in some respects, it is not consistent with several experimental observations made during this investigation. The extent to which the proposed mechanism is also appropriate at elevated temperatures remains to be established. The initial results at elevated temperatures presented in Part 2 suggest that enhanced rates of dissolution at low potentials are indicative of a similar mechanism under these conditions. It is also possible that this mechanism applies in the same way to dissolution in sulfate solutions.

Acknowledgements
The authors would like to acknowledge the financial support from BHP Billiton and the Parker Centre. Thanks also to Dr. Puru Shrestha and his mineralogy team at BHP Billiton for their excellent mineralogical support of this project.

4. Conclusions
As demonstrated in Parts 1 and 2, the rate of dissolution of chalcopyrite in chloride solutions under ambient conditions is enhanced within a potential window between 560 and 600 mV (SHE) in the presence of copper ions and dissolved oxygen. This study has aimed at the establishment of a mechanism which is consistent with the data presented in Parts 1 and 2. Mineralogical investigations of chalcopyrite surfaces and the residues from dissolution experiments have shown that the surface of the mineral is converted to a covellite-like phase during leaching at potentials below the potential window. Except at relatively high pH values, the product of elemental sulfur forms largely as isolated spherical globules, but seldom on the surface of chalcopyrite, confirming that initial dissolution involves a soluble sulfur intermediate. Addition of fine pyrite was shown to catalyse the reaction and mineralogy of the residues showed that most of the sulfur formed as layers around smaller pyrite particles. The rate of dissolution of synthetic covellite particles was found to be similar to that of chalcopyrite particles of the same nominal size under the same conditions. These phenomena can be explained by an extended reaction model in which chalcopyrite partially dissolves non-oxidatively to form H2S and a covellite-like surface species within the potential window. The H 2S that is in equilibrium with the mineral is oxidised by oxygen in a reaction catalysed by copper ions and other minerals such as pyrite. Confirmation of this reaction model has been enhanced by a detailed study of the kinetics of the coppercatalysed oxidation of H2S with the following main conclusions The oxidation of hydrogen sulfide in acidic chloride solutions by dissolved oxygen is very slow in the absence of catalysts.

References
Abraitis, P.K., Pattrick, R.A.D., Kelsall, G.H., Vaughan, D.J., 2004. Acid leaching and dissolution of major sulphide ore minerals: processes and galvanic effects in complex systems. Mineralogical Magazine 68 (2), 343-351. Al-Harahsheh, M., Kingman, S., Rutten, F., Briggs, D., 2006. ToF and SEM study on the preferential oxidation of chalcopyrite. International Journal of Mineralogy 80, 205-214. Ammou-Chokroum, M., Sen, P.K., Fouques, F., 1981. Electrooxidation of chalcopyrite in acid chloride medium: kinetics, stoichiometry and reaction mechanism. Developments in Mineral Processing 2, 759-809. Byerley, J.J., Fouda, S.A., Rempel, G.L., 1975. Activation of copper(ll) ammine complexes by molecular oxygen for the oxidation of thiosulfate ions. Journal of the Chemical Society, Dalton Transactions 1329-1338. Dixon, D.G., Mayne, D.D., Baxter, K.G., 2008. Galvanox a novel galvanically assisted atmospheric leaching technology for copper concentrates. Canadian Metallurgical Quarterly 47, 327-336. Dutrizac, J.E., 1978. The kinetics of dissolution of chalcopyrite in ferric ion media. Metallurgical Transactions B 9B, 431-439. Dutrizac, J.E., 1981. The dissolution ofchalcopyrite in ferric sulphate and ferric chloride media. Metallurgical Transactions B 12B, 371 -378. Dutrizac, J.E., 1982. Ferric ion leaching of chalcopyrite from different localities. Metallurgical Transactions B 13B, 303-309. Dutrizac, J.E., 1990. Elemental sulfur formation during the ferric chloride leaching of chalcopyrite. Hydrometallurgy 23,153-176. Harmer, S.L., Thomas, J.E., Fornasiero, D., Gerson, A.R., 2006. The evolution of surfaces layers formed during chalcopyrite leaching. Geochimica et Cosmochimica Acta 70, 4392-4402. Hiroyoshi, N., Miki, H., Hirajima, T., Tsunekawa, M., 2000. A model for ferrous promoted chalcopyrite leaching. Hydrometallurgy 57, 31-38. Hiroyoshi, N., Maeda, H., Mild, H., Hirajima, T., Tsunekawa, M., 2001. Enhancement of chalcopyrite leaching by ferrous ions in acidic ferric sulfate solutions. Hydrometallurgy 60, 185-197. Hiroyoshi, N., Arai, M., Miki, H., Tsunekawa, M., Hirajima, T., 2002. A new reaction model for the catalytic effect of silver ions on chalcopyrite leaching in sulfuric acid solutions. Hydrometallurgy 63, 257-267. Hiroyoshi, N., Kuroiwa, S., Miki, H., Tsunekawa, M., Hirajima, T., 2004. Synergistic effect of cupric and ferrous ions on active-passive behaviour in anodic dissolution of chalcopyrite in sulfuric acid solutions. Hydrometallurgy 74,103-116. Jones, D.L., Peters, E., 1976. The leaching of chalcopyrite with ferric sulphate and ferric chloride. In: Yannopoulos, J.C., Agarwaal, J.C. (Eds.), Extractive Metallurgy of Copper. AIME, New York, pp. 633-653. Kametani, H., Aoki, A., 1985. Effect of suspension potential on the oxidation rate of copper concentrate in a sulfuric acid solution. Metallurgical Transactions B 16B, 695-705. Luther, G.W., 2002. Aqueous copper sulfide clusters as intermediates during copper sulfide formation. Environmental Science and Technology 36, 394-402. Martell, A.E., Smith, R.M., 2004. NlST Standard Reference Database 46, Ver 8. National Institute of Standards and Technology, Gaithersburg, MD, USA. Miki, H., Nicol, M.J., 2008. The kinetics of the copper-catalysed oxidation of iron(ll) in chloride solutions. ln: Young, C., et al. (Ed.), Hydrometallurgy 2008. The Minerals, Metals and Materials Society, Warrendale, Pennsylvania, pp. 971-979.

1. Copper ions are very effective in catalysing the reaction by oxidation of 2. The rate-determining step is re-oxidation of copper(l) by oxygen at copper
concentrations above about 0.1 g/L while at lower concentrations, reduction of copper(ll) by hydrogen sulfide becomes rate-limiting. 3. At high copper concentrations, the rate of the reaction is similar to the published rate law for the kinetics of the oxidation of copper(l) by oxygen in chloride solutions. 4. Rates of dissolution, estimated using a modified previously published model involving mass transport coupled with chemical oxidation of H 2S formed at the surface of the mineral, compare favourably with those measured under similar conditions. 6. Fine pyrite catalyses the oxidation of hydrogen sulfide. This confirms observations of the effect of fine pyrite on the leaching ofchalcopyrite concentrates and the accumulation ofsulfur around fine pyrite particles during dissolution of chalcopyrite. This mechanism provides an alternative which is more consistent with the observed data than the conventional mixed-potential electrochemical model which has previously been applied to the kinetics of the dissolution of sulfide by copper(ll) and re-oxidation of copper(l) by dissolved oxygen.

M. Nicol et al. / Hydrometallurgy 103 (2010) 86-95 Miller, J.D., McDonough, P.J., Portillo, H.Q., 1981. Electrochemical model in silver catalysed ferric sulfate leaching of chalcopyrite. ln: Laskowski, J. (Ed.), 13th lnternational Mineral Processing Congress. Elsevier, Amsterdam, pp. 851 -894. Nicol, M.J., 1984. The kinetics of the oxidation of copper(l) by oxygen in acidic chloride solutions. South African Journal of Chemistry 37, 77-80. Nicol, M.J., Lazaro, l., 2003. The role of non-oxidative processes in the leaching of chalcopyrite. Copper 2003, The Canadian lnstitute of Mining, Metallurgy and Petroleum, Montreal, Quebec, Book 1, pp. 383-394. Parker, A.J., Paul, R.L., Power, G.P., 1981. Electrochemical aspect ofleaching copper from chalcopyrite in ferric and cupric salt solutions. Australian Journal of Chemistry 34, 13-34. Parker, A., Klauber, C., Kougianos, A., Watling, H.R., van Bronswijk, W., 2003. An X-ray photoelectron spectroscopy study of the mechanism of oxidative dissolution of chalcopyrite. Hydrometallurgy 71, 265-276. Rickard, D., Luther, G.W., 2006. Metal sulfide complexes and clusters reviews in mineralogy and geochemistry. Mineralogical Society of America 61,427-504. Tekin, T., Boyabat, N., Bayramoglu, M., 1999. Kinetics and mechanism of aqueous oxidation of H2Sby Fe3+. International Journal of Chemical Kinetics 31, 331-335.

13

M. Nicol et al. / Hydrometallurgy 103 (2010) 86-95