Layered Tin Chalcogenides SnS and SnSe: Lattice Thermal Conductivity Benchmarks and Thermoelectric Figure of Merit

Abstract

Tin sulfide (SnS) and tin selenide (SnSe) are attractive materials for thermoelectric conversion applications. Favorable small band gap, high carrier mobility, large Seebeck coefficient, and remarkably low lattice thermal conductivity are a consequence of their anisotropic crystal structure of symmetry Pnma, made of corrugated, black phosphorus-like layers. Their internal lattice dynamics combined with chemical bond softening in going from SnS to SnSe make for subtle effects on lattice thermal conductivity. Reliable prediction of phonon transport for these materials must therefore include many-body effects. Using first principles methods and a transferable tight-binding potential for frozen phonon calculations, here, we investigate the evolution of thermal lattice conductivity and thermoelectric figure of merit in Pnma-SnS and-SnSe, also including the high-temperature Cmcm-SnS phase. We show how thermal conductivity lowering in SnS at higher temperatures is largely due to dynamic phonon softening ahead of the Pnma-Cmcm structural phase transition. SnS becomes more similar to SnSe in its lifetime and mean free path profiles as it approaches its high-temperature Cmcm phase. The latter nonetheless intrinsically constraints phonon group velocity modules, preventing SnS to overtake SnSe. Our analysis provides important insights and computational benchmarks for optimization of thermoelectric materials via a more efficient computational strategy compared to previous ab initio attempts, one that can be easily transferred to larger systems for further thermoelectric materials nanoengineering. The good description of anharmonicity at higher temperatures inherent to the tight-binding potential yields calculated lattice conductivity values that are in very good agreement with experiments.