A geochemical evaluation of potential magma ocean dynamics using a parameterized model for perovskite crystallization

•A geochemical model for perovskite crystallization from a magma ocean is presented.•Trace elements substitute into MgPv following mechanisms observed for pyroxene.•MgPv floatation (basal magma ocean) imparts unobserved geochemical signatures.•Combined MgPv + CaPv floatation is geochemically favored over MgPv-only flotation.

Magnesium perovskite (MgPv) is likely the first phase to crystallize from a deep magma ocean. Consequently, MgPv crystallization has a strong control on the dynamics and chemical evolution associated with the earliest stages of silicate Earth differentiation. In order to better understand the chemical evolution associated with MgPv crystallization during a magma ocean, a parameterized model for major and trace element partitioning by MgPv has been developed. The parameterization is based on a compilation of published experimental data and is applied to batch and near-fractional crystallization scenarios of ultramafic liquids, allowing for a more complete analysis of the geochemical implications for magma ocean crystallization. The chemical signatures associated with modeled MgPv fractionation are evaluated in the context of possible dynamical outcomes to a magma ocean (e.g. basal magma ocean (BMO) or crystal settling). It is shown that fractionating MgPv from ultramafic liquids imparts diagnostic signatures (e.g. Ca/Al, HFSE anomalies, ε176Hf-ε143Nd) in both the liquid and solid phases. These signatures are not currently observed in the accessible Earth, suggesting that either early-fractionating MgPv was subsequently homogenized or crystal suspension was dominant during the earliest stages of magma ocean crystallization. A BMO that fractionates CaPv and MgPv is also considered and shown to mute many of unobserved geochemical effects associated with a MgPv-only fractionation, offering an alternative possibility for the evolution of a BMO depleted in heat producing elements.