Arsenopyrite dissolution rates in O2-bearing solutions

Arsenopyrite dissolution was studied by means of long-term, stirred and non-stirred flow-through experiments in the pH range of 1 to 9 at 25, 50 and 70 degrees C and at different input dissolved-O-2 concentrations (from 0.2 to 8.7 mg L-1), At pH lower than 4, aqueous iron, which is mainly in the ferrous form, and arsenic are stoichiometrically released. Sulphur concentrations released were lower than stoichiometrically expected (S/As<1). X-ray Photoelectron Spectroscopy (XPS) and MicroRaman Spectroscopy surface analyses on reacted and unreacted samples showed an enrichment of the reacted arsenopyrite surface in sulphur and arsenic under acidic conditions. In the light of these results, the steady-state dissolution rates were estimated by the release of arsenic at pH<4 and were used to derive an empirical dissolution rate law expressed as: Rarsenopyrite(mol m(-2) s(-1))(25 degrees C) = 10(-7.41 +/- 0.47).a(O2)(0.76 +/- 0.11).a(H+)(-0.12 +/- 0.07) where a(O2) and a(H+) are the activities of hydrogen ions and dissolved oxygen, respectively and their exponents were estimated from multiple linear regression of the dissolution rates. Temperature increase from 25 to 70 C yields an apparent activation energy for the arsenopyrite oxidation by dissolved oxygen of 18.5 +/- 1.6 kJ mol(-1). At pH>6, aqueous iron is mainly in the ferric form and is depleted as it precipitates as Fe-oxyhydroxide onto arsenopyrite surfaces, yielding Fe/As and Fe/S less than one: between pHs 7 and 9, iron depletion is complete, and sulphur released is more abundant than arsenic released, which is precipitated as As-O phases, as confirmed by MicroRaman spectroscopy. At pHs 6-9, iron-oxyhydroxide phases and arsenic oxide phases upon the arsenopyrite surface provide an effective layer that reduces diffusion of dissolved oxygen and arsenopyrite dissolution. As coating on the arsenopyrite surface becomes the rate-limiting step, the Shrinking Core Model (SCM) allows quantification of the surface dissolution rate, especially from data obtained where the effect of coating was still negligible. The SCM also allowed us to calculate the effective coefficient for oxygen diffusion through the coating, which can vary from 10(-17) to 1.5 . 10(-16)m(2)s(-1). The formation of such a coating produced a decrease in arsenic and sulphur release over time and a final surface passivation. (C) 2010 Elsevier B.V. All rights reserved.