Probing the critical point of MgSiO3 using deep potential simulation

FY Xu and ZG Li and XR Chen and HY Geng and L Liu and JB Hu, JOURNAL OF APPLIED PHYSICS, 135, 125901 (2024).

DOI: 10.1063/5.0189696

The giant impact between proto-Earth and a Mars-sized planet called Theia resulted in the formation of the Earth-Moon system, and the silicate mantles of the initial bodies may have partly been vaporized. Here, we develop a machine learning potential for MgSiO3 based on the data from first-principles calculations to estimate its critical point. The variations in pressure along different isotherms yield the position of the critical point of MgSiO3 at 0.54 g cm(-3) and 6750 +/- 250 K, which agrees with the previous theoretical estimation. We also simulate the MgSiO3 melt under a spectrum of critical conditions to understand the changes in coordination environment with density and temperature. The fourfold Si-O coordination hardly changes with increasing density at 3000 K. However, with increasing temperature, the dominance of four- coordinated Si-O diminishes rapidly as density decreases. Regarding Mg-O coordination, the overall trend, which varies with temperature and density, remains largely consistent with Si-O but with a greater diversity in the types of coordination due to more bond breaking events. Our work opens a new avenue by employing machine learning methods to estimate the critical point of silicates. (c) 2024 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution-NonCommercial 4.0International (CC BY-NC) license (https://creativecommons.org/licenses/by-nc/4.0/).

Return to Publications page