Composition and Layering in Earth's Outer Core

We use results from ab initio simulations on the properties of Fe-Ni-S-C-O-Si liquids under core pressures and temperatures to put constraints on the composition of the low-velocity layer at the top of Earth's outer core. We find that increasing the concentration of any light-element in liquid iron always increases velocity. Our data rules out the possibility of making a low-velocity and low-density layer (required for dynamical stability) by changing the concentration of a single light element. However, replacing of one light element with another could produce such a layer. We find that replacing silicon with oxygen, sulfur, or carbon produces a layer with lower bulk sound velocity and lower density. Similarly, replacing carbon with sulfur or oxygen produces similar results. On the other hand, sulfur and oxygen cannot be replaced by anything that produces such a layer, thus precluding them from being the only light elements in the core. Swapping elements for one another requires chemical reactions, and the obvious one at the CMB is interaction with the mantle. Si and O can swap for one another depending on the FeO and SiO2 content of the mantle at the CMB, whereas reactions involving S and C are less obvious. Using thermodynamic models for O and Si partitioning in the metal, we find that core-mantle equilibrium with a pyrolitic mantle will always enrich the core in Si and O under the CMB, and hence won't make a low-velocity layer. On the other hand, equilibrium with a SiO2-depleted and FeO-enriched mantle. An iron (or ferropriclase) enriched lowermost mantle is compatible with the ULVZ and with the basal magma ocean hypothesis; the composition of the slow layer atop the outer core brings an independent support for the BMO hypothesis.