Search for: All records

New high-throughput search strategies based on statistical identification of promising chemical subspaces show potential for accelerating half-Heusler thermoelectric materials discovery. 

The solid solution behavior of ZnTe in CuInTe2 complicates defect analysis. 

Bi1−xSbx alloys are classic thermoelectric materials for near-cryogenic applications. Despite more than half a century of study, unraveling the underlying transport physics within this space has been nontrivial due to the complex electronic structure, disorder, and small bandgap within these alloys. Furthermore, as Peltier coolers, Bi1−xSbx alloys operate in a bipolar regime; as such, understanding the impact of minority carriers is critical for further improvements in device performance. This study unites first principles calculations with low-temperature experimental measurements to create a generalized model for transport within semiconducting Bi-Sb alloys. Our exploration reveals the interplay between the complex, degenerate valence band structure with the extremely light conduction bands. By building a hybrid computational/experimental model, an understanding of both the electron and hole relaxation times emerges both as a function of temperature and energy. Special quasi-random supercell calculations reveal that, despite significant atomic disorder, the electronic band structures within the alloy remains largely unaffected and electron–phonon scattering dominates. For charge carriers near the band edges, the relaxation times are thus extremely long, consistent with cyclotronic behavior appearing at low magnetic fields (≪ 1 T). Modeling thermoelectric performance suggests that the valence band edge deformation potential is significantly weaker and highlights the potential for p-type compositions to meet or exceed the current n-type alloys. 

Abstract Data-driven approaches to materials exploration and discovery are building momentum due to emerging advances in machine learning. However, parsimonious representations of crystals for navigating the vast materials search space remain limited. To address this limitation, we introduce a materials discovery framework that utilizes natural language embeddings from language models as representations of compositional and structural features. The contextual knowledge encoded in these language representations conveys information about material properties and structures, enabling both similarity analysis to recall relevant candidates based on a query material and multi-task learning to share information across related properties. Applying this framework to thermoelectrics, we demonstrate diversified recommendations of prototype crystal structures and identify under-studied material spaces. Validation through first-principles calculations and experiments confirms the potential of the recommended materials as high-performance thermoelectrics. Language-based frameworks offer versatile and adaptable embedding structures for effective materials exploration and discovery, applicable across diverse material systems.