Ab initio calculation of polarons: algorithms and benchmarks
ORAL
Abstract
Calculations of polarons using density-functional theory and supercell approaches face two key challenges: (i) large computational cells are required to describe intermediate and large polarons, (ii) the formation energy and localization of the polaron wavefunction are sensitive to the exchange-correlation functional. We developed a new approach where the polaron is expressed as a superposition of Bloch states, and the calculation of wavefunctions and energies is cast into the solution of a nonlinear system involving Kohn-Sham energies, phonon frequencies, and el-ph matrix elements from density-functional perturbation theory [PRL 122, 246403 (2019)]. Here we report on further optimization, specifically improvements of the iterative eigensolver, parallelism, memory management, and increased modularity using EPW and Quantum ESPRESSO. We analyze the performance of the method in terms of Brillouin-zone sampling and wavefunction initialization and report benchmark on prototypical polaronic systems, from ionic to covalent insulators.
–
Presenters
-
Chao Lian
University of Texas at Austin
Authors
-
Chao Lian
University of Texas at Austin
-
Weng Hong Sio
University of Texas at Austin
-
Feliciano Giustino
Physics, University of Texas at Austin, University of Texas at Austin, Oden Institute for Computational Engineering and Sciences, University of Texas at Austin, Department of Physic, The University of Texas at Austin, Austin, Texas 78712, USA, Oden Institute for Computational Engineering and Sciences, Oden Institute, University of Texas at Austin, Department of Materials, University of Oxford, Department of Physics, University of Texas at Austin, ODEN Institute for Computational Engineering and Sciences, University of Texas at Austin