[S-IT16] Deep Earth Sciences

[SIT16-P18] Ab initio lattice thermal conductivity of pyrolytic lower mantle to infer the core-mantle boundary heat flow

*Haruhiko Dekura¹, Taku Tsuchiya¹ (1.Geodynamics Research Center, Ehime University)

Keywords:Lower mantle, Thermal transport property, Theory

Thermal conduction is an essential physical process in controlling the Earth’s energy transport in the Earth’s interior. The heat flux from the core to mantle is dominated by the thermal conductivity (κ) above the core-mantle boundary (CMB), and it is an important indicator representing the thermophysical coupling strength between the core and mantle. The primary component in κ of the solid mantle is the lattice part (κ_lat). Those of lower mantle (LM) minerals are, however, still poorly understood because experimental measurements are technically difficult under LM in-situ pressure (P) and temperature (T) condition. Meanwhile, we established an ab initio technique to compute κ_lat of based on the density-functional theory (DFT) which was successfully applied to the Fe-free systems, MgO, MgSiO₃ bridgmanite (Brg) and MgSiO₃ post-perovskite (PPv) (Dekura and Tsuchiya, 2013, 2017, 2019). Recently, our method was extended to more realistic LM mineral compositions, Fe-bearing LM minerals, based on a new state-of-the-art computational technique combining the DFT, the internally consistent LDA+U method, the anharmonic lattice dynamics theory, and the self-consistent approach to solving the phonon Boltzmann transport equation (Dekura and Tsuchiya, under review). Using the newly developed method, in this study, we calculate κ_lat of (Mg,Fe)O ferropericlase (FP), (Mg,Fe)SiO₃ Brg and PPv under LM in-situ P and T condition, and construct an effective thermal conductivity of LM to estimate the CMB heat flow. Our results show that the κ_lat of those minerals are substantially smaller than Fe-free systems. Such a strong negative effect of Fe incorporation is found due to decreases in phonon group velocity and lifetime. An effective κ_lat of the lowermost mantle is then estimated for pyrolytic aggregate (FP+PPv) with a pyrolytic ratio to be ~2-3 Wm^-1K^-1, which produces a net heat flow from the core to mantle ~2-3 TW. The values of estimated heat flow are comparable to that required to sustain the geodynamo at the current magnitude (~2-3.5 TW). However, this value is found to be substantially smaller than that estimated from the core with high thermal conductivity of iron (~12-15 TW), although the heat flows estimated from the mantle and from the core should in principle match at the CMB. We discuss the possibility of reconciling the discrepancy.