PARTITIONING OF ORE METALS AMONG FE-S-O MELT, RHYOLITE MELT, AND PYRRHOTITE
Experimental starting materials consisted of powdered rhyolite glass (glass) (Bishop Tuff) and magnetite (mt) (Essex, NY); pyrrhotite (po) (Sudbury, Ontario); and metallic silver. All charges were sealed in evacuated, fused quartz tubes. Experiments were performed at temperatures from 1000 to 1070°C in a tube furnace. The final assemblage was mt+po+glass+quenched Fe-S-O melt. Vapor pressure was established by the composition of the starting materials. Run times varied from 15 minutes to 3 days, with runs quenched in a water bath. Phase compositions were determined by using a JEOL JXA-8900 superprobe for EDS/WDS and an Element 2 single collector LA-ICP-MS at the University of Maryland.
Experiments at and above 1039°C showed rounding and reduction in grain size of the po and mt, accompanied by the appearance of complex, fine-textured patches of sulfide-oxide intergrowth. The boundary of these patches and the silicate melt are consistent with a liquid-liquid contact, suggesting the intergrowth was a homogenous immiscible Fe-S-O liquid at temperature which subsequently formed two phases (mt + po) on quench. Following established S2 and O2 fugacity relationships in systems containing mt + po, final compositions suggest O2 ~FMQ. Silver concentration in the Fe-S-O melt was found to be at higher levels than in pyrrhotite and was measured at 5700 ppm ±700. Preliminary data show: Fe-S-O/poDAg= 2.5±0.8; Cu = 0.9±0.2; Zn = 2.4 ±0.3; Pb = 70±30. Fe-S-O/glassDAg = 400±200; Cu = 500±200; Zn = 130±50; Pb = 7±3.
The Fe-S-O melt is seemingly more efficient than po at concentrating Ag, Zn, and Pb; and as efficient at concentrating Cu. Fe-S-O melts will control substantially more of the metal budget than silicic melts.