Conference Dates

September 4-8, 2016


Thermoelectric materials often require extensive tailoring of microstructure and composition to optimize their properties. Isoelectronic alloying is one of the most commonly applied techniques used to induce point defect scattering and reduce the lattice thermal conductivity. However, current approaches rely heavily on experiment and are not conducive to the high-throughput methods, which are becoming increasingly commonplace within the thermoelectric community. Herein, we present three computationally inexpensive approaches to the evaluation of point-defect scattering in thermoelectrics. These approaches also weave elements of structural chemistry associated with point defect scattering to provide a clear conceptual connection with classical phonon scattering. Computational results are further validated using bulk synthesis of SnSe and its alloys. Experimental transport measurements serve to assess model efficacy. Herein we develop and integrate three computational metrics with experimental data to find predictive metrics that can predict the relative strength of point defect scattering and associated reductions in the thermal conductivity. Ultimately, we find that all of the computational metrics are successful in predicting the relative strength of the various alloys, enabling computation to play a larger role in the screening and optimization of composition within thermoelectrics.

Please click Additional Files below to see the full abstract.