The information below is updated, and contains new data beyond the GSA 1996 talk. The research is still ongoing. Before citing these results formally, please email Phil Candela. Informal citations can be made to this web page.
CANDELA, P.A., PICCOLI, P.M., and WILLIAMS, T.J., Laboratory for Mineral Deposits Research, Department of Geology, University of Maryland at College Park, MD 20742, candela(piccoli;williams)@geol.umd.edu; http://www.geol.umd.edu/.
The association of some Au deposits with felsic intrusive rocks suggests that magmatic processes, including the scavenging of Au from magmas by a buoyant Magmatic Volatile Phase, {MVP}, may be active in concentrating Au in ore deposits or protore in the upper levels of the continental crust. The MVP (gas-like or liquid-like) may transport Au toward and/or through the roof of a magma chamber, where the Au may be deposited in ore or protore. Au may also be transported to higher levels in magmatic-geothermal systems where dominantly non-magmatic waters may transport Au to sites of deposition located far from the genetically-related magma.
Experimentation on the solubility of Au at magmatic temperatures, and the partitioning of Au among magmatic phases is necessary (though not sufficient) for the formulation of quantitative hypotheses of ore genesis (which can be tested in the field). In this study, we have taken advantage of the many aqueous phase inclusions trapped in rhyolitic glass run products to capture samples of the experimental MVP that is in equilibrium with both the rhyolitic melt and the noble metal capsule (either X(Au) =1, or X(Au) = 0.1 {balance Pt}). Capsules were loaded with ground pumice or synthetic rhyolite + a single grain of obsidian or more synthetic glass + aqueous solution so that the glass run product possessed a variable proportion of aqueous inclusions. Experiments were run in cold seal vessels at 800oC and 140 MPa; the intrinsic oxygen fugacity of the vessels was determined by H2 fugacity sensors and is = NN0+0.5 (water pressure medium). At the end of an experiment, inclusion-rich and inclusion-poor glass were removed from the capsule, inspected, separated, and washed in aqua regia. All lab materials that contacted the glass run products were also washed with aqua regia before hand. Glasses with and without aqueous inclusions were then analyzed by INAA for Au and Cl. The INAA analysis of inclusion-poor glass yielded the solubility of Au in the melt and the concentration of Cl in the melt. The proportion of aqueous phase trapped in the inclusion-rich glass fragment was calculated from the difference between the Cl concentration of the inclusion-rich and inclusion-free glasses. The concentration of Au in the MVP was calculated from the Au concentration of the two glasses, and the proportion of the aqueous phase trapped in the inclusion-rich glass .
Our experiments at 800oC, 140MPa, f(O2) = NNO+0.5, aqueous NaCl eq. concentration = 3 wt.%, and HCl/(NaCl+KCl) = 1:100, suggest, to date , that 1} the solubility of Au in a water-saturated rhyolite melt is on the order of 2-4 ppm (microg/g) IN BOTH LOW SULFUR, AND PO-SATURATED SYSTEMS; 2} the solubility of Au in the aqueous phase is hundreds of ppm (microg/g) IN THE LOW-SULFUR SYSTEMS; and 3} the MVP/melt partition coefficient for Au is 300(+/- 175) (1 sigma).
These data suggest that oxidized (low H2S) felsic magmatic systems, exsolving chloride- and HCl-bearing volatile phases, have the potential to deliver Au to high temperature hydrothermal systems. Reduced, sulfur-bearing conditions are not required for efficient partitioning of Au between melt and the MVP, or high aqueous solubilities for Au at magmatic temperatures.