Experimental study of gold-hydrosulphide complexing in aqueous solutions at 350–500°C, 500 and 1000 bars using mineral buffers

The solubility of gold was measured in KCl solutions (0.001–0.1 m) at near-neutral to weakly acidic pH in the presence of the K-feldspar-muscovite-quartz, andalusite-muscovite-quartz, and pyrite-pyrrhotite-magnetite buffers at temperatures 350 to 500°C and pressures 0.5 and 1 kbar. These mineral buffers were used to simultaneously constrain pH, f(S2), and f(H2). The experiments were performed using a CORETEST flexible Ti-cell rocking hydrothermal reactor enabling solution sampling at experimental conditions. Measured log m(Au) (mol/kg H2O) ranges from −7.5 at weakly acid pH to −5.9 in near-neutral solutions, and increases slightly with temperature. Gold solubility in weakly basic and near-neutral solutions decreases with decreasing pH at all temperatures, which implies that Au(HS)2Au(HS)2 is the dominant Au species in solution. In more acidic solutions, solubility is independent of pH. Comparison of the experimentally measured solubilities with literature values for Au hydrolysis constants demonstrates that at 350°C AuHS(aq)0 dominates Au aqueous speciation at the weakly acidic pH and f(S2)/f(H2) conditions imposed by the pyrite-pyrrhotite-magnetite buffer. In contrast, at temperatures >400°C AuHS(aq)0 becomes less important and AuOH(aq)0 predominates in weakly acid solutions. Solubility data collected in this study were used to calculate the following equilibrium reaction constants: Au(cr)+H2S(aq)0=AuHS(aq)0+0.5 H2(g),K(AuHS0) and Au(cr)+H2S(aq)0+HS−=Au(HS)2−+0.5 H2(g),K(Au(HS)2−)The calculated log K(AuHS0)K(AuHS0) values are −5.25 ± 0.35 and −5.92 ± 0.36 at 350°C (500–1000 bars) and 400°C/1000 bars, respectively. The log K(Au(HS)2-) values are −1.68 ± 0.42 at 350°C/500 bars, −1.40 ± 0.41 at 350°C/1000 bars, −2.14 ± 0.42 at 400°C/ 500 bars, −1.75 ± 0.40 at 400°C/1000 bars, −1.87 ± 0.40 at 450°C/1000 bars, and −1.80 ± 0.40 at 500°C/1000 bars. Equilibrium constants for both reactions are retrograde with respect to temperature, but this effect is more pronounced for AuHS0(aq) than for Au(HS)2−. The K   values’ dependence on temperature, as well as the calculated values of K(Au(HS)2-) are consistent with the data of Benning and Seward (1996). The calculated values of K(AuHS0)K(AuHS0) are between those determined by Benning and Seward (1996) and Gibert et al. (1998).