Metal-Silicate Equilibration at Super-Liquidus Temperatures During Core Formation

Thursday, 18 December 2014: 5:30 PM
John W Hernlund1, Hiroki Ichikawa2, Stephane Labrosse3 and Masanori Kameyama2, (1)Earth-Life Science Institute, Meguro, Japan, (2)Ehime University, Matsuyama, Japan, (3)Ecole Normale Supérieure Lyon, Lyon, France
Experimental constraints on the partitioning of moderately siderophile elements between metal and silicates during core formation suggest equilibration temperatures significantly greater than the liquidus of the silicate Earth (e.g., Wade and Wood, 2005). However, because equilibration was considered to occur in a ponded metal at the silicate solidus, such high temperature equilibration was rejected as implausible. Instead, lower temperature equilibration with variable oxygen fugacity was proposed as an alternative, although the plausibility of the physical mechanisms invoked in this scenario is also questionable. We have re-visited the model of metal-silicate separation in large molten pockets following energetic accretion events, and find that silicate-metal equlibration is most rapid when the iron rains out of the magma, and the release of gravitational potential energy by this rain heats the mixture by as much as 1000 K above the liquidus. However, the first drops of iron rain to pond at the base of the molten pocket will equilibrate at lower temperatures, and only the final drops will be subject to the highest temperatures. We model rain fall and heating of the magma by viscous dissipation to calculate the effective pressure-temperature conditions for partitioning in this scenario, and find that effective pressure conditions are smaller than the pressure at the base of the molten pocket. The ponded metal itself is gravitationally stratified (both in composition and temperature), and is not expected to convect or mix until it undergoes subsequent downward transport into the Earth's core. We also suggest that such a process operating during the very largest giant impact events (extending into the deep mantle) may have given rise to a buoyant oxygen-enriched metal layer atop the outer core, as suggested by some seismological models of the present-day Earth (e.g., Helffrich and Kaneshima, 2010).


Helffrich, G. and S. Kaneshima (2010), Outer-core compositional stratification from observed core wave speed profiles, 468: 807-810.

Wade, J., and B.J. (2005), Core formation and the oxidation state of the Earth, Earth Planet, Sci. Lett., 236:78–95.