The solubility of platinum and gold in NaCl brines at 1.5 kbar, 600 to 800°C: A laser ablation ICP-MS pilot study of synthetic fluid inclusions

Abstract

The concentration and distribution of Pt and Au in a fluid-melt system has been investigated by reacting the metals with S-free, single-phase aqueous brines (20, 50, 70 wt% eq. NaCl) ± peraluminous melt at a confining pressure of 1.5 kbar and temperatures of 600 to 800 °C, trapping the fluid in synthetic fluid inclusions (quartz-hosted) and vesicles (silicate melt-hosted), and quantifying the metal content of the trapped fluid and glass by laser ablation ICP-MS. HCl concentration was buffered using the assemblage albite-andalusite-quartz and f O 2 was buffered using the assemblage Ni-NiO. Over the range of experimental conditions, measured concentrations of Pt and Au in the brines ( C P t fluid , C A u f l u i d ) are on on the order of 1–10 3 ppm. Concentrations of Pt and Au in the melt ( C P t m e l t , C A u m e l t ) are ∼35–100 ppb and ∼400–1200 ppb, respectively. Nernst partition coefficients ( D P t f l u i d / m e l t , D A u f l u i d / m e l t ) are on the order of 10 2–10 3 and vary as a function of C m e t a l f l u i d (non-Henry’s Law behavior). Trapped fluids show a significant range of metal concentrations within populations of inclusions from single experiments (∼ 1 log unit variability for Au; ∼2–3 log unit variability for Pt). Variability in metal concentration within single inclusion groups is attributed to premature brine entrapment (prior to metal-fluid-melt equilibrium being reached); this allows us to make only minimum estimates of metal solubility using metal concentrations from primary inclusions. The data show two trends: (i) maximum and average values of C A u f l u i d and C P t f l u i d in inclusions decrease ∼2 orders of magnitude as fluid salinity ( m ∑ C l f l u i d ) increases from ∼4 to 40 molal (20 to 70 wt % eq. NaCl) at a constant temperature; (ii) maximum and average values of C A u f l u i d increase approximately 1 order of magnitude for every 100°C increase temperature at a fixed m ∑ C l f l u i d . The observed behavior may be described by the general expression: log ⁡ ( m m e t a l f l u i d , T , 1.5 k b a r ) = x ⋅ log ⁡ ( m ∑ C l f l u i d ) − y where x (slope of the solubility curve) is similar for both Au and Pt (−2), and y increases with increasing T. The strong negative correlation between salinity and metal concentration may result from (i) the “salting-out” of neutral hydrogen chloride complexes [e.g., PtCl 3H] as NaCl is added to the solution and free H 2O is consumed during the production of HCl 0 by the buffer assemblage, or (ii) the consumption of free H 2O by the hydration of non-chloride-based metal complexes such as hydroxides [e.g., Pt(OH) 2 · (H 2O) n ]. The proposed solubility mechanisms would be enhanced if activity coefficients for HCl and/or metal complexes decrease significantly with increasing fluid salinity. Identification of the exact dissolution mechanism at such extreme conditions and confirmation of the equilibrium metal concentrations awaits future study. The results of this pilot study demonstrate that single-phase hypersaline fluids which exsolve from or interact with residual magmatic liquids, partially crystallized rocks, or small volumes of PGE (platinum-group element)-Au-bearing sulfide can potentially dissolve and transport economically-significant amounts of Pt and Au across the magmatic-hydrothermal transition at moderately oxidizing conditions.