Tailoring microstructure engineering to decouple electron-phonon transport in Se-based thermoelectric alloys

Abstract

Selenide compounds are known for their intrinsically low thermal conductivity, yet their thermoelectric application is often hindered by poor electrical transport. Here we examine how this limitation manifests in three different systems, Sn-Se, Sb-Se, and Bi-Se, by correlating their microstructures with measured transport properties. Sn-Se forms a composite containing crystalline SnSe, elemental Se, and amorphous regions, resulting in ultralow lattice thermal conductivity [0.08 W/(m·K) at 650 K] but limited carrier mobility [~7.39 cm2/(V·s)]. Sb-Se is largely amorphous with scattered Sb2Se3 crystallites, yielding extremely low electrical conductivity (~6.86 × 10-2 S/m at 300 K) despite a high Seebeck coefficient (1,902 μV/K). Bi-Se develops a distinct two-phase structure comprising metallic Bi nanoprecipitates embedded in a semiconducting Bi2Se3 matrix. This configuration enables efficient charge transport through interconnected Bi pathways [μ ≈ 2,170 cm2/(V·s)]. The electrical conductivity reaches 1.56 × 104 S/m at room temperature and exceeds 3.7 × 104 S/m at elevated temperatures, while the Bi2Se3 matrix preserves a reasonable Seebeck coefficient. Additionally, atomic-scale disorder at the Bi/Bi2Se3 interfaces enhances phonon scattering, reducing lattice thermal conductivity without severely impeding electron flow. As a result, the Bi-Se system achieves a peak thermoelectric figure of merit (ZT ) of approximately 0.85 at 400 K, demonstrating excellent low-temperature performance, while the Sn-Se system reaches a ZT of ~1.0 at 650 K, suitable for mid-temperature applications. These findings suggest that incorporating a metallic secondary phase into a semiconducting matrix can help balance the competing requirements for thermoelectric performance, although the optimal operating temperature range depends on the specific microstructure.