¹Emily L. Fischer, ¹Stephen W. Parman
Icarus (in Press) Link to Article [https://doi.org/10.1016/j.icarus.2025.116664]
¹Department of Earth, Environmental, and Planetary Sciences, Brown University, 324 Brook St, Providence, RI 02912, United States of America
Copyright Elsevier

Enstatite chondrites are often used as models for the bulk composition of Mercury because they have similarly low oxygen fugacities. However, e-chondrites are too Si-rich to explain the observed composition of Mercury’s lavas. Here we explore a model in which an initially enstatite chondrite-like Mercurian silicate magma ocean loses Si to the large Fe core during early differentiation. We define a Mercury Fractionation Line (MFL) based on average basaltic geochemical terrane compositions and assume Mercury’s bulk silicate composition must fall along this line. We estimate that 26.5–36.7 ± 7.5 % (1σ) Si must be lost from an initial mantle to bring the e-chondrite compositions up to the MFL. Assuming that the Si is partitioned into the core, this implies a core Si content of 2.8–3.9 ± 0.8 wt% and an oxygen fugacity of IW–4.5 ± 1.0. We also show that a model where Mercury was initially ~2 times larger is consistent with more reducing oxygen fugacities (IW–5.0 ± 1.0) and a higher core Si content (~15 wt%). This estimated initial Mercury size is also consistent with predictions from dynamical simulations. We consider how Si partitioning into the core affects the δ30Si composition of the mantle. Though uncertainties are large, we show that as the initial radius of Mercury increases, δ30Si decreases, trending towards the δ30Si composition of enstatite chondrites. Our calculations do not constrain the mechanism by which Mercury’s mantle may have been lost. However, if they are correct, they imply that the mantle loss must have happened after core formation.