Answers are grounded in this paper and cite their sources.
Bi2Te3-based alloys, particularly (Bi1–xSbx)2Te3, remain the most effective thermoelectric materials near room temperature. In this work, we employ neuroevolution interatomic potentials trained on first-principles data to perform large-scale atomistic simulations of (Bi1–xSbx)2Te3 across a wide range of compositions. The developed potential enables accurate and efficient evaluation of structural stability, phonon transport, and thermoelectric-relevant properties in disordered systems. The potential model reproduces ab initio elastic constants, phonon dispersions, and equilibrium lattice parameters. With the help of molecular dynamics-based dynamic tensile simulations, we show the existence of a factor-of-two anisotropy for the Young’s modulus between in-plane and cross-plane loading directions across the entire composition range. The lattice thermal conductivity across different compositions of (Bi1–xSbx)2Te3 follows the Klemens-type alloy-scattering model, whose values nearly match the experimental results. Our results reveal that alloying-induced atomic disorder plays a dominant role in suppressing lattice thermal conductivity through enhanced phonon scattering, with a non-linear dependence on Sb concentration arising from the competition between mass-disorder scattering and changes in phonon dispersion. We show pronounced anisotropy for in-plane and cross-plane thermal conductivity values in Bi1.08Sb0.92Te3, which has a minimum bulk-average thermal conductivity at 300 K across all compositions. We also develop an MD-assisted model to calculate bipolar thermal conductivity in Bi0.4Sb1.6Te3, which can be extended to any alloy composition. Finally, the transferability of the NEP model is explored with the help of interfacial thermal transport between Bi2Te3 and Sb2Te3 lattices at 300 K. These findings provide a unified atomistic picture of how alloy composition governs thermal transport in Bi–Sb–Te systems and highlight the critical role of disorder-driven phonon scattering in optimizing lattice thermal conductivity for high thermoelectric performance. The present study establishes neuroevolution potentials as a potent tool for studying thermomechanical properties of high-performance thermoelectric materials.
No takes yet. Share an insight, caveat, or question.
Kushwaha et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: