Experimental Solubility of Gold in Sulphidic Brine at Amphibolite- and Greenschist-Facies Conditions

Robert R. Loucks Research School of Earth Sciences, Australian National Univ., Canberra, ACT 0200, Australia

John A. Mavrogenes Research School of Earth Sciences,Australian National Univ., Canberra, ACT 0200, Australia

Stephen M. Eggins Research School of Earth Sciences, Australian National Univ., Canberra, ACT 0200, Australia

Leslie Kinskey Research School of Earth Sciences, Australian National Univ., Canberra, ACT 0200, Australia

J. Michael Shelley Research School of Earth Sciences, Australian National Univ., Canberra, ACT 0200, Australia

Experimental data on gold speciation and solubility in sulphidic chloride brines have been limited to T £ 350 °C and P £ 165 bars (Hayashi and Ohmoto, 1991). We extend the experimental database to 625 °C and 3700 bars and derive thermodynamic properties of the relevant aqueous species. The reactions studied are

Au + 3/2 H2O + 1/4 FeS + 7/8 FeS2

= 3/8 Fe3O4 + HAu(HS)°2 (aq) (1) and

Au + H2S + 3/8FeS2 + 1/8 Fe3O4

= Au(HS)°(aq) + 1/2 H2O + 3/4 FeS. (2)

The ŸS2, ŸH2, and pH buffer assemblages pyrrhotite + pyrite + magnetite + water (Py-Po-Mt-W) and muscovite + orthoclase + quartz (Mu-Or-Qz) were loaded in a gold foil capsule, together with ~0.06 ml of 1-molal KCl solution spiked with 7.5 ppm Th and 44 ppm U. U and Th are insoluble in the buffer minerals and quantitatively retained in the fluid. In three experiments, the sealed capsules were run in a cold-seal vessel 10 days at 625 ± 3 °C and 3.7 ± 0.2 kbar, at 625 °C and 1.1 kbar, and at 550°C and 3.1 kbar. During the runs, a quartz rind containing primary fluid inclusions up to 50 µm in diameter was precipitated as an overgrowth upon the starting quartz pieces.

A new analytical instrumentation system - the Ultraviolet Laser Trace Element Micro-Analyzer (ULTEMA) - has been developed in the Research School of Earth Sciences.
The ULTEMA uses a 193-nm wavelength excimer laser micro-beam for photo-chemical ablation of solids and fluid inclusions for quantitative analysis of their elemental composition by inductively-coupled-plasma mass spectrometry (ICPMS). Fluid inclusions synthesized in the goldsolubility experiment were opened and analyzed individually by the ULTEMA. The 44 ppm U and 7.4 ppm Th served as internal standards for determination of the Au content of the fluid inclusions. A synthetic glass standard was used to calibrate the relative ionization efficiencies of Au, U, and Th in the ICP. Measured Au concentrations in the fluid inclusions are 540 ± 200 ppm at 625 °C and 3700 bars, 61 ± 25 ppm at 550°C and 3100 bars, and 10 ± 5 ppm at 625 ° and 1100 bars. Using -10 cm3 as the partial molar volume of the thio-gold aqueous complex (derived from Renders and Seward, 1989) and using mineral and water molar volumes (SUPCRT 92) , we evaluate

( log mAu/P)T = - DrV°T / 2.303 RT = 0.609 /kbar

at 250 °C and = 0.354 /kbar at 625 °C. This result represents the change of gold solubility with pressure.

Hayashi and Ohmoto (1991) inferred that at the ŸS2 and ŸH2 of the Py-Po-Mt-W buffer, and acidic pH (Mu-Or-Qz buffer), the HAu(HS)o2) complex accounts for gold solubility at 250-350 °C in vapour-saturated brines of 0-3 molal alkali chloride. Their data at 250 °C (.08 ppb), 300 (1.2 ppb), & 350 (12 ppb) and that of Gibert et al. (1993) - 500 ppb at 450 °C and 500 bars in 0.5 molal chloride brine buffered by Py-Po-Mt-W and Mu-Or-Qz are in excellent accord with our measurements of gold solubility, taking account of the effect of pressure as described above. (Rytuba and Dickson (1977) found Au solubility (1.5 ppm) in sulphidic brine is independent of salinity at 500 °C, 0.4-1m NaCl). These data indicate that log K varies linearly with temperature over the 250-625°C interval, so the dissolution reaction's DrCp°~-0, and DrS°T and DrH°T are constants assessable by linear regression. Provisionally assuming that HAu(HS)°2 is the principal aqueous gold specie, and using

g (HAu(HS)°2 ) ~-1.5
aFeS = 0.52
regression gives
log KT = 4.56 - 7.178 (103/T), DrS °T = 20.8 cal/mol-K

DrH°T = +32,845 cal.
Using thermodynamic properties of minerals and H2O from SUPCRT 92, we derive the following standard partial molar properties of HAu(HS)°2 (aq): -S°298K = 58.8 cal/mol-K, -H°298K = - 11,283 cal/mol, -Cp°298K = 35.7 cal/mol-K. Together with -V°298K = -10 cm3/mole, these may be used to evaluate gold solubility as the HAu(HS)°2 complex over a wide range of T and P relevant to the genesis of porphyry Au-Cu and slate-belt and greenstone-belt gold lodes.


Gibert, F., Pascal, M.-L., & Pichavant, M., Terra Abs. 5, 457 (1993).

Hayashi, K. & Ohmoto, H., Geochim. Cosmochim. Acta 55, 2111-2126 (1991).

Renders, P.J. & Seward, T.M., Geochim. Cosmochim. Acta 53, 245-253 (1989).

Rytuba, J.J. & Dickson, F.W., Proc. 4th IAGOD Symposium, 320-326. (1977).