资源环境科技发展态势分析平台

Transfer of oxygen between Earth's core and lowermost mantle is important for determining the chemistry and nature of stratification on both sides of the core-mantle boundary (CMB). Previous studies have found that oxygen enters the metal when Fe-O liquid equilibrates with representative lower mantle materials. However, experiments have not yet been conducted at CMB pressure-temperature conditions. Here we use density functional theory to obtain the first estimates of oxygen partitioning between liquid Fe-O-Si metals and ferropericlase at CMB conditions. Our method successfully reproduces experimentally derived partitioning data at 134GPa and 3200K, while our calculations show a strong increase of oxygen partitioning into metal with temperature and a weaker increase with pressure, consistent with previous work. At CMB conditions of 135GPa and 4000-4700K oxygen partitioning into metal is higher than previous estimates and increases strongly with metal oxygen concentration. Analysis of the lower mantle chemical boundary layer shows that oxygen transport through the solid is severely limited even with the enhanced partitioning and is unlikely to explain the thickness of a stably stratified layer below the CMB inferred from seismology. However, if the lower mantle was molten in early times, as suggested by core evolution models with high thermal conductivity, then the mass flux and stable layer thickness are significantly increased.

Plain Language Summary Earth's core is composed primarily of iron, silicon, and oxygen; it is directly below the solid mantle, which is mainly composed of two different minerals called bridgmanite and ferropericlase. Here we present the first calculations of iron oxide partitioning between ferropericlase and liquid iron-silicon-oxygen mixtures at core-mantle boundary (CMB) pressure-temperature-concentration conditions. Partitioning of iron oxide between the core and mantle is important for constraining the chemistry on either side of the CMB, determining the composition of the core, and elucidating the origin of the seismically detected stable layer at the top of the core (which has previously been ascribed to FeO transfer from the mantle). We find that FeO partitioning into the core is stronger than found by previous studies at lower pressures and temperatures and is particularly sensitive to oxygen content in the metal. We analyze transfer of O through the lower mantle chemical boundary layer by diffusion and dynamic instability and find that in both cases, oxygen flux is smaller than previous estimates even with the greater partitioning. The resulting thickness of the chemically stable layer that arises below the CMB is too small to explain the seismic observations.