Lécaille, Manon ; Tosi, Nicola ; Rivoldini, Attilio, ; Baumeister, Philipp ; Namur, Olivier ; Charlier, Bernard

Talk presented at 46th COSPAR Scientific Assembly 2026, Florence, Italy on 2026-08-04

Abstract: Mercury’s bulk composition remains uncertain, although several observations are consistent with an enstatite-chondrite-like origin. Formation under highly reducing conditions likely pro- moted significant partitioning of silicon into the core, leaving the mantle variably enriched or depleted in silicon. Depending on this residual silicon abundance, the mantle mineralogy may range from forsterite- to enstatite-dominated [1]. Because these endmembers exhibit strongly contrasting lattice thermal conductivities, mantle mineralogy is expected to exert a first-order control on Mercury’s long-term thermal and magnetic evolution. To quantify these effects, we investigate Mercury’s coupled mantle and core evolution using the 1-D parameterized mantle convection model TEMPURA [2], linked to a core evolution framework that accounts for inner core formation and the development of thermally stable stratification within the outer core [3]. The mantle composition is modeled by fixing the diopside fraction based on CaO in enstatite chondrites while systematically varying the forsterite-to-enstatite ratio. For each composi- tion, depth-dependent elastic and transport properties are calculated from mineral equations of state and recent experimental studies [4,5]. We assume an Fe–Si core and determine the interior structure for each model realization by determining the silicon concentration required to match Mercury’s mass for a given initial thermal state and mantle composition. The resulting core thermophysical properties are then evolved self-consistently through time. We explore the pa- rameter space using Monte Carlo simulations that vary initial temperature conditions, mantle composition and rheological properties, and the partitioning of heat-producing elements upon melting and crust formation. We evaluate model outcomes against key observational bench- marks: (1) present-day crustal thickness [6]; (2) early onset of global contraction [7]; and (3) dynamo activity between 4 and 3.5 Ga and at present day [8]. We deliberately avoid imposing constraints on total radial contraction because its magnitude remains debated. Reproducing early dynamo activity represents one of the most restrictive conditions on accept- able evolutionary scenarios: it necessitates either an early inner core nucleation, or a late start of outer core thermal stratification. Successful solutions preferentially involve mantles with high enstatite content, elevated crustal concentrations of heat-producing elements, and rela- tively high mantle reference viscosities, all consistent with an enstatite-rich mantle mineralogy. Since the onset and persistence of dynamo action are strongly influenced by the formation of thermally stratified layers, we further examine how uncertainties in the thermal conductivity of the core alloy influence the viability of otherwise successful models. These findings indicate that Mercury’s thermal and magnetic histories are highly sensitive to both mantle mineralogy and core heat transport efficiency, offering independent constraints on mantle composition as well as core thermophysical properties. References [1] Xu et al. (2024). Nature Comm., 15, 5061, doi:10.1038/s41467-024-49305-x. [2] Baumeister et al. (2023). AA, 675, A122, doi:10.1051/0004-6361/202245791. [3] Davies et al. (2024). Earth Planet. Sci. Lett., 641, 118812, doi:10.1016/j.epsl.2024.118812. [4] Zhang et al. (2019). Earth Planet. Sci. Lett., 519, 109-119, doi:10.1016/j.epsl.2019.04.048. [5] Guo et al. (2024). Geochem. Geophys. Geosyst.,25(6),e2023GC011419,doi:10.1029/2023GC011419. [6]Padovan et al. (2015). Geophys. Res. Lett., 42(4), 1029-1038, doi:10.1002/2014GL062487. [7] Crane et al. (2017). Geophys. Res. Lett., 44(7), 3082-3089, doi:10.1002/2017GL072711. [8] Johnson et al. (2015). Science, 348(6237), 892-895, doi:10.1126/science.aaa8720

Funding: 3PRODPLANINT/3PRODPLANINT/3PRODPLANINT