I will summarize our efforts to develop methods for computing electronic and thermal transport properties of materials from first principles, with application to thermoelectric semiconductors. The new computational approaches achieve good predictive accuracy and transferability while reducing complexity and computation cost compared to the existing methods. By applying our electron-phonon averaged (EPA) approximation in the Boltzmann transport formalism for electrons we discovered several new thermoelectric alloy compositions that have recently been confirmed to have record performance and stability in power generation devices. To address transport in narrow gap and disordered materials, we generalize the semiclassical Boltzmann equation to describe quantum transport effects and study transport properties in the Bi2Se3 topological semiconductor. We also study the coupled dynamics of electrons and phonons from first principles in the combined Boltzmann formalism and find strong enhancement of mobility and Seebeck coefficient due to drag effects even at room temperature in SiC, as well as significant deviations from the Wiedemann-Franz law. I will also discuss recent progress in developing highly accurate and fast machine learning energy models for enabling large-scale molecular dynamics simulations of structural evolution and transport in complex materials. We apply these methods to study dynamic surface reconstruction, 2D->3D phase transitions and transport of ions and heat in superionic phases.
Zoom link will be forwarded to Condensed Matter Seminar email list. For a copy of the link or to be added to the email list, please contact Stephen Bullwinkel (email@example.com).
Recordings of Condensed Matter Seminar events can be found on mediaspace: https://mediaspace.illinois.edu/channel/Condensed%2BMatter%2BSeminar%2BTalks%2BFall%2B2020/178724052