The influence of spin-orbit resonances on the evolution of Mercury's mantle and crust

Abstract

Mercury's record of large impact basins and spin evolution models suggest that its present-day 3:2 spin-orbit resonance may not be primordial. It could have been established up to hundreds of millions of years after planet formation, possibly triggered by the impact that created the Caloris basin about 3.7 billion years ago. Before this, Mercury may have been in a synchronous rotation or a 2:1 resonance, which would have induced strong hemispheric surface temperature variations, influencing the thermal structure of the lithosphere and mantle. Using 3D thermochemical mantle convection models, we simulate Mercury's mantle evolution and volcanic crust formation over one billion years, incorporating surface temperature distributions from different spin-orbit resonances. We assess whether these variations can generate large-scale lateral differences in crustal thickness, as inferred from gravity, topography and surface composition data, and compare predicted radius changes due to mantle and core cooling with existing estimates from compressional tectonic features. Crustal thickness, interior cooling rate, and radius change are primarily controlled by internal heat production, with models using intermediate to high heat production rates (characteristic of CI and EH chondrites) best matching observations. The mantle reference viscosity, low thermal conductivity attained at Mercury's mantle conditions, and cooling due to melt extraction exert first-order controls on the timing of crust emplacement and its final extent. Regardless of surface temperature patterns, mantle convection is dominated by small, stable cells. While surface temperature variations influence the location of hot and cold regions at large scales, they do not alter the spatial scale of convection. Assuming vertical melt extraction, crustal thickness locally follows the convection pattern. The present-day 3:2 resonance does not induce significant large-scale variations in crustal thickness, but a past synchronous rotation could have produced hemispheric differences, with crust up to 10-15 km thicker on the dayside. Similarly, radial contraction is hemispherical, with the hot hemisphere contracting less and at a slower rate than the cold one as long as the resonance persists. The surface record does not clearly support these hemispheric patterns. This suggests that past spin-orbit resonances may have been short lived, or that they may have affected Mercury's interior more subtly than our models predict, or that subsequent geological processes erased or modified early large-scale asymmetries. Future high-resolution imaging and surface composition data from BepiColombo, particularly of Mercury's poorly-mapped southern hemisphere, will be critical in testing this hypothesis and in refining the available constraints on the planet's crustal evolution and tectonic history.