Abstract:
The impact of various mixtures of CO2, H2O, H2, and CH4 upon Fe, Ni, Co, W. Ge, P, and Mo partitioning between silicate and metal alloy liquids was evaluated experimentally at 1400°C and 10 kb in graphite. The effects observed are consistent with the control expected from redox-driven transfer between metal and silicate liquids. Oxidation moves these metals to the silicate liquid. Oxidizing C-O-H or C-O vapors generated by decomposition of either brucite or magnesite or both have a stronger, opposite effect on element distribution than the reducing C-H vapors generated by anthracene decomposition. Special effects associated with particular volatile and siderophile elements were largely unobserved. Individual C-O-H species involved in the redox equilibria are not so important as their redox effect, except for phosphorous which responds differently to CO2- and H2O-inspired vapors. An analysis of the redox effects of several M/MOy equilibria confirm the presence of large departures from ideality in the solutions involved.Oxidants might be quite potent in remediating some of the excesses in the siderophile element chemistry of the mantle. Because little special effect for any volatile species is observed, siderophile elements should respond in a coherent manner regardless of which volatile species is the immediate cause of the redox transfer. In consequence, oxidizing the system with any of the plausible volatile agents we have investigated to make Ni or Ge or Mo less siderophile than they are in the absence of volatile elements will also make Fe, Co, W, and P too lithophile to be consistent with their observed abundances in the mantle. The observed Fe abundances in the mantle put limits on redox ability to explain the excesses of other siderophile elements. Furthermore it is unlikely that redox effects can be fine tuned to give chondritic relative abundance values in the mantle as needed. Nevertheless, the possible effects of redox equilibria have long been known to be commensurate in importance with pressure and temperature effects upon siderophile element distribution and should continue to be considered in any analysis of core formation.