The formation of Earth’s core has left a compositional imprint on the mantle, depleting and fractionating most of its siderophile (iron-loving) elements. Gallium is a moderately siderophile, hence it should be strongly depleted in the mantle. However, gallium concentration in the mantle matches that of lithophile (silicate-loving) elements having the same volatility. That is to say that either gallium behaves as a lithophile element during core formation, or a large influx of gallium was brought to the Earth after the core had formed. Geochemical evidence does not support the latter hypothesis, as it would require all other lithophile elements with similar volatility to be enriched in the mantle, or for late accretion to be composed of anomalously gallium-rich objects. In order to mitigate this issue, experimental studies have tried to understand how gallium behaves during core segregation by gauging the effects of pressure, temperature and oxygen fugacity on the partitioning of gallium between metal and silicate. None of these parameters provided the first-order change required to match the observation.
We investigated the influence of core composition on gallium partitioning. The core in known to contain light-elements (oxygen, silicon sulfur and carbon), and those can change the activity of gallium in the metal, and strongly affect the behavior of gallium during core formation. We performed a series of metal-silicate partitioning experiments (2 GPa, 1673–2073 K) in a piston-cylinder press. We varied the light-element composition of the metal and observed that Si and O have a very strong influence on the activity of gallium, making it more lithophile. We then modeled terrestrial accretion as a continuous process and tested different accretion histories; we can reproduce the mantle concentration of gallium if the core segregates in a deep magma ocean (>40 GPa) and contains large amounts of silicon or oxygen.