A first-principles study of ternary metal chalcogenides [formula omitted][formula omitted] for efficient thermoelectric applications empowered by machine-learning interatomic potential

Abstract

Metal chalcogenides have already been recognized as promising candidates for thermoelectric applications in the past decade. This study systematically analyzes the ternary metal chalcogenide Ba2MnX3 (X = Te,Se,S) materials to check their suitability for thermoelectric applications. Ba2MnX3 materials belong to the orthorhombic and centrosymmetric Pnma space group, with Ba atoms distributed inside the three-dimensional framework of MnX4 tetrahedron units, forming a quasi-one-dimensional crystal structure. The electronic band structures of the materials investigated using density functional theory (GGA+U) reveal that the materials have direct band gaps ranging between 1.3 eV to 1.9 eV. The structural stability and phonon transport properties are studied using a combination of density functional theory and machine-learning approach. The study shows that all Ba2MnX3 materials possess low lattice thermal conductivity (κL). For Ba2MnTe3, at 700 K κL∼0.45Wm−1K−1 is the lowest achieved κL in this work. The thermoelectric transport parameters of the materials are estimated by working out the semiclassical Boltzmann transport equation. Finally, the Figure of merit (ZT) and thermoelectric conversion efficiency values of the materials are calculated. The maximum ZT≈ 0.74 at 700 K at a carrier concentration of 2.43×1020cm−3, and an efficiency ∼8% in the temperature interval 300 – 700 K is achieved for Ba2MnSe3.