|
|
|
|||||||||||||||||
| JOURNAL HOME | HELP | CONTACT PUBLISHER | SUBSCRIBE | ARCHIVE | SEARCH | TABLE OF CONTENTS |

1 Williamson Research Centre for Molecular Environmental Science, and Earth Sciences Department, University of Manchester, Manchester M13 9PL, UK
2 Department of Chemical Engineering, Imperial College London, London SW7 2AZ, UK
* E-mail: david.vaughan{at}man.ac.uk
| ABSTRACT |
|---|
|
|
|---|
KEYWORDS: acid leaching, sulphides, pyrite, chalcopyrite, galena, sphalerite
| Introduction |
|---|
|
|
|---|
The present study was concerned with the overall mechanisms of dissolution of (air-ground) single-phase pyrite, chalcopyrite, galena and sphalerite particles in chloride media of pH 2.5 and with the complexities that may arise in systems involving two phases: pyrite + chalcopyrite, pyrite + galena, and pyrite + sphalerite.
| Experimental methods |
|---|
|
|
|---|
Acid-leaching experiments
Experiments utilized a Mettler Toledo DL-67 auto-titration instrument, fitted with a Mettler Toledo DG 111-SC pH Electrode. The continuously stirred solution within the reaction vessel was maintained at 25±2°C. Prior to the addition of the mineral powder(s), a pre-titration was performed in which 0.1 M HCl (BDH, Convol) was added to 50 cm3 of air-equilibrated, high-purity water, resulting in a final solution pH of 2.5.
At the start of the experiments, 0.3 g of the sulphide powder was added to the solution. Leaching experiments were conducted for 3 h. During this time the pH of the resultant mineral pulp and the volume of 0.1 M HCl required to maintain a constant pH were recorded continuously. The solution redox potential (mV, SHE) was monitored using a Mettler Toledo InLab 501 combination Pt/Ag-AgCl redox electrode after 0, 30, 45, 60, 90, 120 and 180 min. 1 cm3 solution samples of the mineral pulp were removed at selected time intervals, filtered through 0.22 µm cellulose acetate membrane filters and acidified prior to analysis by inductively coupled plasma atomic emission spectroscopy (ICP-AES). A sample of the unacidified leachate was also analysed for dissolved sulphur species by anion chromatography. At the end of the experiment, the mineral powder was separated by filtration, dried at room temperature and stored under nitrogen prior to analysis by XPS.
Solution analysis
The concentrations of Fe, Cu, Pb, Zn and S were determined by ICP-AES using a VG Elemental Horizon instrument and conventional solution nebulization. Prior to ICP analysis, the solution samples were further acidified to 2% HNO3. The instrument was calibrated using commercial multi-element standards.
Samples of the final leachates from each experiment were also analysed by high performance liquid chromatography (HPLC) which was conducted using a Dionex 4000i gradient ion chromatograph. Sulphur anion separation was accomplished using a combination of Dionex AG11 guard column (75 x 4 mm) and Dionex AS11 analytical column (250 x 4 mm) with sodium hydroxide eluant.
X-ray photoelectron spectroscopy (XPS) analysis
A Kratos XSAM 800 surface analysis instrument, employing a Mg-K
X-ray source, was used to obtain XPS spectra of mineral fracture and powder surfaces. This instrument was calibrated using the Au 4f7/2 83.98 eV, Cu 2p3/2 932.67 eV and Ag 3d5/2 368.27 eV peaks. All peaks were corrected for static charging effects using the 1s electron binding energy of carbon, which was taken as 284.6 eV. Due to instrumental limitations, no measures were taken to reduce potential evaporation of elemental sulphur or other volatile sulphur species within the vacuum chamber.
Fitting of the XPS data was accomplished using XPSPEAK41 software. Unless otherwise stated, all peaks have been assigned 80% Gaussian, 20% Lorentzian character. During fitting, all S(2p) components were assigned as doublets with an intensity ratio of 2:1 and a spin orbit splitting of 1.2 eV. Elemental atomic ratios based on the XPS data have been calculated using Kratos sensitivity factors.
| Results and discussion |
|---|
|
|
|---|
|
|
To these data on the redox potential changes and generation or consumption of protons can be added information on changing solution chemistry with time and surface analyses of the final mineral products for each of the systems studied.
Pyrite
In the case of pyrite, no acid addition was required during the experiment to maintain the solution pH at 2.5±0.1, because oxidation of the disulphide by dissolved oxygen at the experimental redox potential (36±35 mV) involves consumption of protons by oxygen reduction. That reaction was assumed to occur by a four-electron reduction (Biegler et al., 1970):
![]() | (1) |
rather than forming H2O2 as an intermediate by a two-electron reduction. Reaction 1 was coupled to pyrite oxidation, assumed to produce S(VI), the thermodynamically most stable sulphide oxidation product under the redox potential conditions applied:
![]() | (2) |
Hence, the overall process would involve the net generation of protons by the reaction:
![]() | (3) |
However, if sulphoxy species with oxidation state +V (S2O26) form, then no excess protons would be formed, whereas formation of S4O26 ions would consume one mole of proton per mole of FeS2. If elemental sulphur were the predominant sulphide oxidation product, then the reaction would consume two moles of protons per mole of FeS2.
Figure 3
shows the time dependences of Fe and S concentrations dissolved oxidatively from pyrite. Linear regression analysis of these data indicate that the total dissolved sulphur concentration increases at twice the rate of Fe(II), consistent with stoichiometric dissolution of pyrite, according e.g. to reaction 3. However, the ratio of dissolved S to Fe concentrations increases from 1.38 to 1.5. The finite intercept given by extrapolating the regression line back to zero time, reflects the aerial oxidation of the pyrite, especially during grinding in air:
|
![]() | (4) |
Hence, dissolution during the initial 15 min was predominantly of the aerial oxidation products rather than the bulk pyrite, which exhibited slower, congruous dissolution thereafter. It is possible that a S-rich surface develops during the initial stages of dissolution, or that the oxidized powder surface is S-rich (relative to the S:Fe ratio in bulk pyrite).
The XPS data for pyrite are shown in Fig. 4
. In each case, the spectrum for a freshly cleaved sample is shown at the bottom, the sample ground (in air) in the middle, and the final reaction product sample spectrum at the top. The dominant peak at 707.3±0.1 eV in the Fe 2p3/2 spectrum of the fracture sample is consistent with binding energies in pyrite. However, an additional peak is required at 709.6±0.1 eV to adequately fit the data. Nesbitt and Muir (1994) and Karthe et al. (1993) have attributed such a feature to surface defects. However, some hydroxyl oxygen was also identified at the pyrite fracture surface, so this feature could be attributable to ferrous hydroxide species, although the presence of surface defects cannot be ruled out (Bonnissel-Gissinger et al., 1998). The spectrum of the oxidized powder sample requires addition of further components at 711.6 eV and 714.3 eV attributable to Fe(III) in oxy(hydr)oxide and sulphate environments. The small component at higher binding energy is attributable to an Fe(III) multiplet structure. The Fe 2p3/2 spectrum of the acid-leached sample was fitted in an identical manner. Peak areas show that Fe(III) comprises 30 and 36% of the total iron at the surface of the air-oxidized and acid-leached pyrite powder surfaces, respectively.
|
Bonnissel-Gissinger et al. (1998) identified Fe and S oxidation products at the surfaces of pyrite samples ground in a nitrogen atmosphere. Therefore, minor oxidation of the pyrite fracture sample within the spectrometer chamber, or during fracturing under nitrogen, cannot be ruled out. Likely oxygen-bearing species in these systems include hydroxyl (OH), oxide (O2), adsorbed molecular water, sulphate species and additional sulphoxy anions (such as thiosulphate and sulphite). The O 1s data for the pyrite fracture sample have been closely simulated with a single peak having a binding energy of 531.1 eV. This is close to the reported binding energy of hydroxyl oxygen (as in Fe hydroxides). As noted above, the presence of ferrous hydroxyl species could explain the feature at 709 eV in the Fe 2p3/2 spectrum of the pyrite fracture surface.
Two components were required to adequately simulate the O 1s data obtained for the air-oxidized pyrite sample. A peak centred at 531.1 eV is attributable to hydroxyl species and a second peak centred at 532 eV is attributed to sulphate. The O 1s spectrum of the acid leached pyrite powder was similarly fitted using two peaks at binding energies 531.1 eV and 532 eV. The greater relative area of the higher binding energy component in the O 1s spectrum of the acid leached sample is a reflection of the greater amount of sulphate present at the surface of this sample, consistent with the S 2p data.
The pyrite used in this study had a bulk Fe:S ratio very close to the ideal stoichiometry Fe:S = 0.5, whereas, based on the XPS data obtained for the pyrite fracture surface, Fe:S = 0.6. The net reaction stoichiometry for pyrite dissolution may be estimated from the XPS and solution chemistry data, the main reaction products being dissolved sulphate, ferrous ions and protons, leaving Fe-depleted sulphide and sulphate on the pyrite surface:
![]() | (5) |
and
![]() | (6) |
so overall:
![]() | (7) |
driven by:
![]() | (8) |
Hence, reactions 7 and 8 predict a net production of 2(yx+xy) mol H+(mol FeS2)1, (y+xxy) mol Fe2+ (mol FeS2)1 and 2y mol SO24(mol FeS2) )1 However, in the experiment, 8.68 x 105 mol H+ were produced and 4.70 x 105 mol Fe2+ dissolved, giving a H+:Fe2+ molar ratio of 1.85, rather than the value of 2.0 predicted. Similarly, the SO4 2:Fe2+ molar ratio is predicted as 2y:(y+xxy), whereas, according to the data in Fig. 3
, after the initial dissolution of air-formed products, the experimental value was 2, implying that x was very small or zero in the steady state.
| Sphalerite and galena |
|---|
|
|
|---|
|
![]() | (9) |
![]() | (10) |
When coupled in electron balance with reaction 1, the overall chemical reaction:
![]() | (11) |
results in a net proton consumption of 2(1 y) mol H+ (mol ZnS)1. At and beyond 1.5 hours of leaching, the dissolved Zn(II) : S molar ratio (y1) was essentially constant with a value of ~10 (i.e. y = ~0.1), the time-dependence arising primarily from dissolution of products of aerial oxidation formed during grinding prior to the leaching experiments.
As shown in Fig. 6
, dissolution of oxidation products formed during grinding galena produced dissolved Pb(II) and sulphur concentrations significantly higher than in the case of sphalerite leaching. However, concentrations of both species decreased over the first hour of the experiment, probably due to restricted solubility of PbCl2 and PbSO4, the latter phase having a particularly low solubility product Ksp(PbSO4) = 107.86
|
![]() | (12) |
![]() | (13) |
![]() | (14) |
Though the dissolved metal to sulphur ratios were not as high as in the case of sphalerite leaching, they increased from 1.3 (y = 0.7) after 19 min to 2.5 (y = 0.4) after 182 min. Though this may imply that the rate of reaction was greater than that of reactions such as 13, the insolubility of PbSO4 would have restricted concentrations of both Pb(II) and dissolved sulphur, as implied in Fig. 6
.
Based on the XPS data, no significant changes in S speciation occur at the surfaces of these minerals as a result of atmospheric oxidation and acid leaching. Elemental S has been reported at the surfaces of both air-oxidized and acid-leached sphalerite and galena, but no evidence for the presence of elemental S was obtained here, possibly due to sulphur desorption in the UHV chamber of the instrument, which had no low-temperature stage.
Chalcopyrite
Figure 7
shows that, after the initial dissolution, presumably of oxidation products formed mainly during grinding chalcopyrite in air, concentrations of dissolved S, Cu and Fe species decreased slightly, before increasing after the first hour and reaching time-independent values after 2 h. The redox potential (Fig. 1
) decreased between 18 and 30 min and reached a relatively constant value of 472 mV. E0Cu2+/CuCl2 (SHE)/V = 0.457, so that in 0.1 M HCl, the electrode potential for equal activities/concentrations of Cu2+ and CuCl2 species is 0.339 V (SHE). Hence, the measured redox potentials correspond to conditions in which Cu2+ ions predominate.
|
The S 2p spectra for the chalcopyrite powder sample and the acid-reacted powder clearly show the presence of sulphate at the powder surfaces (Fig. 8
). In the case of the ground chalcopyrite sample, sulphate accounts for ~30% of the total surface S. In the case of the acid-reacted chalcopyrite powder, sulphate comprises ~3% of the total surface S. Spectra for the air-oxidized powder and the acid-reacted powder are also consistent with the presence of a CuS phase, resulting in an additional component centred at ~162.5 eV in the S 2p spectra.
|
While Fe(II), Cu(II) and sulphate are the thermodynamically favoured aqueous reaction products, elemental S, non-stoichiometric Fe(1x)CuS2 phases and covellite (CuS) have been reported as solid reaction products. Hence, a higher sulphur:sulphate molar ratio was expected than for pyrite. Based on mean values, the solution concentration ratios during the experiment were [Fe]:[Cu]:[S] = 1:0.3:0.8. This suggests that a Cu,S-rich surface alteration layer formed at the chalcopyrite surface. During the experiment, acid addition was required to maintain pH 2.5. Therefore, the mineral-solution reactions resulted in net proton consumption (cf. the pyrite experiment).
![]() | (15) |
![]() | (16) |
![]() | (17) |
Overall:
![]() | (18) |
![]() | (19) |
Hence, in coupling reaction 19 to the oxygen reduction reaction (1), the net consumption of protons would be {8y (2 + 6y + 2z)} = 2(y z 31) mol H+ (mol FeCuS2)1 and the predicted molar ratios [Fe(II)]:[Cu(II)]:[S(VI)] = 1:z:y. Values derived from experimental data in Fig. 7
were z = 0.32±0.01 and y = 0.73±0.06, omitting the first point at 18 min. The theoretically predicted value of 1.18 mol H+ (mol FeCuS2)1 can be compared with the measured value of 1.59 mol H+ (mol FeCuS2)1
Rates of leaching in single- and two-phase systems
Net mineral dissolution rates have been calculated on the basis of the temporal solution chemistry data (see Fig. 9
). The surface areas of the 45 to +150 µm powder size fraction samples have been simply estimated by assuming that the powders comprise spherical mineral grains with an average uniform diameter of (45+150)/2 = 97.5 µm. Dissolution rates have then been estimated on the basis of the increase of the Fe (pyrite experiment), Cu (chalcopyrite experiment), Pb (galena experiment), Zn (sphalerite experiment) concentrations between the first and last sampling times in a given experiment.
|
The leaching rates of galena, chalcopyrite and sphalerite increased by factors of 31, 18 and 1.5, respectively, in the presence of pyrite, due to its superior catalytic properties (Biegler et al., 1970) for the oxygen reduction reaction 1. In mixed-sulphide mineral pulps and ore-bearing waste heaps, particles of different sulphides are in electronic contact and galvanic interactions must occur (Subrahmanyam and Forssberg, 1993).
In the galena + pyrite experiment, the concentration of Fe did not increase appreciably between the first and final sampling times, whilst the Pb concentration did increase significantly. Hence, galvanically-promoted dissolution of galena + pyrite decreases the pyrite electrode potential and its dissolution rate. Also in the galena + pyrite experiment, 75% of the total S in solution as measured by ICP-AES was not detectable by HPLC, which detects only anionic species; this could be due to the presence of colloidal elemental S. There was clear evidence of galvanically promoted chalcopyrite leaching in chalcopyrite + pyrite pulps; again in these experiments, 70% of the S in solution (measured by ICP-AES) was not detectable by HPLC, and so could be present largely as colloidal S rather than anionic sulphur species.
The rate of sphalerite leaching was increased only slightly in the presence of pyrite, though the mineral pulp pH decreased during the experiment. This suggests strongly that pyrite dissolution also occurred in this system, since pyrite dissolution alone decreased the solution pH in the single-mineral power experiments. However, the concentration of dissolved Fe did not increase significantly during the experiment, perhaps due to precipitation of Fe(III) species.
| Conclusions |
|---|
|
|
|---|
| Acknowledgements |
|---|
|
|
|---|
|
|
|---|
Present address: BNFL plc, BTC (B170), Sellafield, Seascale, Cumbria, UK Dedicated to the memory of Dr A. J. Criddle, Natural History Museum, London, who died in May 2002
[Manuscript received 9 July 2003: revised 13 November 2003]
| References |
|---|
|
|
|---|
Biegler, T., Rand, D.A.J. and Woods, R. (1970) Oxygen reduction on sulphide minerals. Part 1. Kinetics and mechanism at rotated pyrite electrodes. Electroanalytical Chemistry and Interfacial Electrochemistry, 60, 151162.
Bonnissel-Gissinger, P., Alnot, M., Erhardt, J. and Behra, P. (1998) Surface oxidation of pyrite as a function of pH. Environmental Science and Technology, 32, 28392845.[CrossRef]
Craig, J.R. and Vaughan, D.J. (1994) Ore Microscopy and Ore Petrography (2nd edition). Wiley Interscience, New York.
Karthe, S., Szargan, R. and Suoninen, E. (1993) Oxidation of pyrite surfaces: a photoelectron spectroscopic study. Applied Surface Science, 72, 157170.[CrossRef]
Keith, C.N. and Vaughan, D.J. (2000) Mechanisms and rates of sulphide oxidation in relation to the problems of acid rock (mine) drainage. Pp. 197219 in: Environmental Mineralogy: Microbial Interactions, Anthropogenic Influences, Contaminated Land and Waste Management (J.D. Cotter-Howells, L.S. Campbell, E. Valsami-Jones and M. Batchelder, editors). Mineralogical Society Series 9, Mineralogical Society, London.
Nesbitt, H.W. and Muir, I.J. (1994) X-ray photoelectron spectroscopic study of a pristine pyrite surface reacted with water vapour and air. Geochimica et Cosmochimica Acta, 59, 46674679.
Smith, E.E. and Schumate, K.S. (1970) Sulfide to Sulfate Reaction Mechanism. Federal Water Quality Administration Water Polution Control Research Series Report # 14010, FPS, 02/70.
Subrahmanyam, T.V. and Forssberg, K.S.E. (1993) Mineralsolution interface chemistry in minerals engineering. Minerals Engineering, 6, 439454[CrossRef]
Vaughan, D.J., England, K.E.R., Kelsall, G.H. and Yin, Q. (1995) Electrochemical oxidation of chalcopyrite (CuFeS2) and the related metal-enriched derivatives Cu4Fe5S8, Cu9Fe9S16 and Cu9Fe8S16. American Mineralogist, 80, 725731.[Abstract][Web of Science][GeoRef]
Williamson, M.A. and Rimstidt, J.D. (1994) The kinetics and electrochemical rate-determining step of aqucous pyrite oxidation. Geochimica et Cosmochimica Acta, 58, 54435454.[CrossRef][Web of Science][GeoRef]
Yin, Q., Kelsall, G.H., Vaughan, D.J. and England, K.E.R. (1995) Atmospheric and electrochemical oxidation of the surface of chalcopyrite (CuFeS2). Geochimica et Cosmochimica Acta, 59, 10911100.[CrossRef][Web of Science][GeoRef]
This article has been cited by other articles:
![]() |
M. P. Asta, J. Cama, J.M. Soler, R .S. Arvidson, and A. Luttge Interferometric study of pyrite surface reactivity in acidic conditions American Mineralogist, April 1, 2008; 93(4): 508 - 519. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. M. Rosso and D. J. Vaughan Reactivity of Sulfide Mineral Surfaces Reviews in Mineralogy and Geochemistry, January 1, 2006; 61(1): 557 - 607. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| JOURNAL HOME | HELP | CONTACT PUBLISHER | SUBSCRIBE | ARCHIVE | SEARCH | TABLE OF CONTENTS |