Mercury's formation within the Early Instability Scenario

Abstract

The inner solar system's modern orbital architecture provides inferences into the epoch of terrestrial planet formation; a ∼100 Myr time period of planet growth via collisions with planetesimals and other proto-planets. While classic numerical simulations of this scenario adequately reproduced the correct number of terrestrial worlds, their semi-major axes and approximate formation timescales, they struggled to replicate the Earth-Mars and Venus-Mercury mass ratios (∼9 and 15, respectively). In a series of past independent investigations, we demonstrated that Mars' mass is possibly the result of Jupiter and Saturn's early orbital evolution, while Mercury's diminutive size might be the consequence of a primordial mass deficit in the region (potentially the result of the growing Earth's early outward migration). Here, we combine these ideas in a single modeled scenario designed to simultaneously reproduce the formation of all four terrestrial planets and the modern orbits of the giant planets in broad strokes. By evaluating our Mercury analogs' core mass fractions, masses, and orbital offsets from Venus, we favor a scenario where Mercury forms through a series of violent erosive collisions between a number of ∼Mercury-mass embryos in the inner part of the terrestrial disk. We also compare cases where the gas giants begin the simulation locked in a compact 3:2 resonant configuration to a more relaxed 2:1 orientation and find the former to be more successful. In 2:1 cases, the entire Mercury-forming region is often depleted due to strong sweeping secular resonances that also tend to overly excite the orbits of Earth and Venus as they grow. While our model is quite successful at replicating Mercury's massive core and dynamically isolated orbit, the planets' low mass remains extremely challenging to match. Indeed, the majority of our Mercury analogs have masses that are 2-4 times that of the real planet. Finally, we discuss the merits and drawbacks of alternative evolutionary scenarios and initial disk conditions (specifically a narrow annulus of material between 0.7-1.0 au). We argue that the results of our N-body accretion models are not sufficient to break degeneracies between these different models, and implore future studies to apply further cosmochemical and dynamical constraints on terrestrial planet formation models.