Study for material analogs of FeSb$_{2}$: material design for thermoelectric materials

TL;DR

This study uses computational methods to design and analyze a new thermoelectric material FeSbAs, showing it can be stable under high pressure and has improved thermoelectric properties compared to FeSb2, with insights into related compounds.

Contribution

The paper introduces FeSbAs as a novel thermoelectric material analog of FeSb2, demonstrating its potential stability and superior thermoelectric performance through computational analysis.

Findings

FeSbAs can be thermodynamically stable above ~30 GPa.

FeSbAs exhibits higher Seebeck coefficients than FeSb2 above 50 K.

Doping enhances the figure of merit for FeSbAs compared to FeSb2.

Abstract

Using the \emph{ab initio} evolutionary algorithm (implemented in USPEX) and electronic structure calculations we investigate the properties of a new thermoelectric material FeSbAs, which is a material analog of the enigmatic thermoelectric FeSb$_{2}$. We utilize the density functional theory and the Gutzwiller method to check the energetics. We find that FeSbAs can be made thermodynamically stable above $\sim$30 GPa. We investigate the electronic structure and thermoelectric properties of FeSbAs based on the density functional theory and compare with those of FeSb$_{2}$. Above 50 K, FeSbAs has higher Seebeck coefficients than FeSb$_{2}$. Upon doping, the figure of merit becomes larger for FeSbAs than for FeSb$_{2}$. Another material analog FeSbP, was also investigated, and found thermodynamically unstable even at very high pressure. Regarding FeSb$_{2}$ as a member of a family of compounds (FeSb$_{2}$, FeSbAs, and FeSbP) we elucidate what are the chemical handles that control the gaps in this series. We also investigate solubility (As or P for Sb in FeSb$_{2}$) we found As to be more soluble. Finally, we study a two-band model for thermoelectric properties and find that the temperature dependent chemical potential and the presence of the ionized impurities are important to explain the extremum in the Seebeck coefficient exhibited in experiments for FeSb$_{2}$.