Melting phase relations in the system Fe-FeO to the inner core boundary pressure

Abstract:

Oxygen has been a plausible candidate for a light element dissolved in the Earth’s core mostly due to its large abundance in the mantle and low eutectic temperature at low pressure. In order to place constraints on the existence/absence of oxygen in the core, which are the key to understand the origin, evolution, and dynamics of the Earth, many experimental efforts have been made on oxygen-bearing iron systems, including measurements of phase relations and of the density and sound velocity of phases. I made thermodynamic analysis on the experimentally constrained melting phase relations in the system Fe-FeO, which helps us to refine the physical properties of the system. From the analysis I found two important issues in the melting relations: (i) the melting temperatures of the end-member systems at 15-16 GPa and (ii) mixing properties for liquids. (i) In order for the thermodynamic calculations to reproduce phase relations determined in multianvil experiments at 15-16 GPa which include an assemblage of FeO + metallic liquid just above the solidus, the melting temperature of iron needs to be lower than that of FeO. If the iron melting temperature is greater, another assemblage of Fe + ionic liquid would be stabilized instead of FeO + metallic liquid. (ii) I tested existing mixing models for liquids by comparing calculated eutectic relations with experimental data to 300 GPa. I found that the Fe-FeO liquids show nonideal mixing behavior at least to 50 GPa and become nearly ideal mixtures under the core pressures. The constructed thermodynamic database to the inner core boundary pressure (330 GPa) yields the eutectic compositions of Fe–7.2–9.1wt% O and eutectic temperatures of 2990–4330 K under the outer core pressures (136-330 GPa). I will also discuss the nature of an oxygen-bearing outer core.