Thermal conductivity of Fe-bearing bridgmanite and post-perovskite: Implications for the heat flux from the core

Thermal Conductivity of MgSiO3‐H2O System Determined by Machine Learning Potentials

“…Our results suggest a weak anisotropy of κ of Brg (Figure S1 in Supporting Information S1), whereas pPv shows a strong anisotropic profile following κ 100 > κ 001 > κ 010 (Figure 3), consistent with previous ab initio and MLP predictions (Dekura & Tsuchiya, 2019;D. Wang et al, 2023).…”

“…After averaging the thermal conductivity based on the Hashin-Shtrikman scheme (Hashin & Shtrikman, 1963), we obtain that the Brg to pPv phase transition enhances the thermal conductivity of the lowermost mantle by ∼19%, and this value reduces to ∼10% considering the presence of 0.14 wt% water in both phase. Using the effect of iron on κ of Brg and pPv predicted by D. Wang et al (2023) as well as their assumptions about the iron content in the lowermost mantle, the values for water-free and water-bearing systems changes to 23% and 10%, respectively. Therefore, we conclude that the presence of water significantly reduces the thermal conductivity enhancement effect caused by the phase transition from Brg to pPv under the CMB conditions.…”

“…The reduction of κ of hydrous pPv is more than 8% from 2,000 to 4,000 K, while κ changes are only discernible at 2,000 K and 140 GPa for Brg. Similarly, D. Wang et al (2023) identified that iron impurities also have a greater effect on κ of pPv relative to Brg. Okuda et al (2020) argue that the additional phonon scattering by Fe impurities would be stronger in a more symmetrical mineral and pPv possesses better symmetry than Brg.…”

“…Zhang and Wentzcovitch (2021), and other experimental and theoretical studies that capture full phonon anharmonicity also show wilder temperature dependence than T 1 in 4. Results from previous experimental and theoretical studies are plotted for comparison (Manthilake et al, 2011;Ohta et al, 2012;Stackhouse et al, 2015;Dekura et al, 2013;Zhang & Wentzcovitch, 2021;Yang et al, 2022;D. Wang et al, 2023).…”

“…On the one hand, Dekura et al (2013), Tang et al (2014), and Dekura and Tsuchiya (2019) obtained the κ of MgSiO 3 using ab initio anharmonic lattice dynamics, but this method ignores high-order anharmonicity of lattice vibrations and is therefore only suited for low temperatures with small atomic displacements from equilibrium. On the other hand, while non-equilibrium molecular dynamics (NEMD) simulations (Stackhouse et al, 2015;D. Wang et al, 2023) fully incorporate anharmonicity, they grapple with significant finite-size effects and need extrapolations to infinite system sizes, introducing a further layer of uncertainty (Sellan et al, 2010).…”