Experimental Constraints on the Sulfide- and Chromite-Silicate Melt Partitioning Behavior of Rhenium and Platinum-Group Elements

摘要

We investigate the origin of high platinum-group element (PGE) abundances associated with chromite-rich rocks by determining the relative partitioning of these elements between chromite- and sulfide-silicate liquids. Chromites were crystallized in the presence of immiscible sulfide and silicate melts in experiments at 1 GPa, producing a few, relatively large (20-50 mu m) crystals, which were analyzed by laser ablation ICP-MS. Our results show that the PGE inventory of chromite and silicate melt produced in experiments is dominated by sulfide and/or alloy micronuggets and that the intrinsic PGE content of these phases is low (sub-ppm), despite high concentrations in coexisting sulfide liquid (i.e., up to alloy saturation). Lower bounds on minimum sulfide-silicate melt PGE partition coefficients (D_(PGE)) calculated from this data are 0.4 to 10 X 10~4, which are similar to values determined in previous studies, confirming the extreme fractionation of these elements into the sulfide phase. Rhenium, which was added to experiments in order to constrain Re-Os fractionation, is highly concentrated in sulfide liquid, present at low but uniform levels in silicate melt, and undetectable in chromite. Calculated sulfide-silicate melt D_(Re), are 3.3 to 5.2 X 10~4, and experiments yielded lower bounds for D_(Os)/D_(Re) of 3, indicating that sulfide-silicate melt equilibrium can fractionate Re from Os. Minimum sulfide melt-chromite partition coefficients are 1,000 or more, indicating that coexisting sulfide melt will be the dominant host for the PGE. Using this partitioning data, we have calculated the mass balance for Ir in chromite-sulfide mixtures and show that for rocks with greater than 200 ppm sulfur, less than 24 percent of the Ir will be in chromite, illustrating that chromite is not the significant PGE host even in low sulfur chromitites. In an experiment saturated in Ir-Re alloy, we measured a maximum iridium concentration in run-product chromites of 150 ppb, which, when combined with an estimate of the Ir activity in the coexisting alloy, yields a maximum Ir solubility of approx 210 ppb. We have found examples of chromitites with Ir contents exceeding this value, indicating that these samples have accumulated an additional PGE-bearing phase. Such results support the notion that interstitial sulfide liquid, or accessory minerals included at the magmatic stage (i.e., laurite, alloys), are most likely to be the dominant primary hosts for PGE in chromite-rich rocks.