An Efficient High-Throughput Approach for Investigating Anharmonic Lattice Dynamics

This work introduces a high-throughput framework specifically designed for anharmonic phonon calculations. The method is based on the efficient fitting of the force constant (FC) matrix implemented in the hiPhive ¹ framework and subsequent calculation of harmonic and anharmonic properties. In particular, we determined the set of optimal parameters: FC fitting methods, cutoffs, supercell size for finite displacement DFT calculations, the magnitude of atomic displacements, and the number of perturbed supercells. Our workflow can calculate phonon band structure, phonon DOS, finite temperature free energy, heat capacity, coefficient of thermal expansion (��), and lattice thermal conductivity (). A temperature-induced anharmonic phonon renormalization (APRN) ² scheme to tackle dynamical instability is also included in the workflow, which is shown to predict the finite temperature stability of cubic GeTe ³ and ZrO₂ ⁴ phases.<br/><br/>Our workflow exhibits an impressive 100 to 1000X speedup compared to conventional approaches when calculating anharmonic vibrational properties. We conduct a rigorous benchmarking process against experimental measurements, encompassing ��, , and computationally calculated phonon DOS. The �� and obtained from the workflow agree well with the experimentally reported values with a Pearson correlation coefficient of 0.94 and 0.99, respectively. The Pearson correlation coefficient of phonon DOS varies from 0.63 in CdSe to 0.96 in BP.<br/><br/>In this workflow, we integrate lattice dynamic packages like hiPhive with the Materials Project (MP) ⁵ tech stack using Atomate2. We aim to use this workflow to curate a high-quality dataset of materials with finite temperature phase diagrams and anharmonic vibrational properties in a high-throughput way. This dataset will make it possible to identify novel thermoelectric materials, solid-state ionic conductors for solid-state batteries, photovoltaic absorber materials, and solar cells. It will also aid in the predictive synthesis of solid-state reactions.<br/><br/>1. Eriksson, F., Fransson, E. & Erhart, P. The Hiphive Package for the Extraction of High-Order Force Constants by Machine Learning. <i>Adv. Theory Simul.</i> <b>2</b>, 1800184 (2019).<br/>2. Xia, Y. Revisiting lattice thermal transport in PbTe: The crucial role of quartic anharmonicity. <i>Appl. Phys. Lett.</i> <b>113</b>, 073901 (2018).<br/>3. Anharmonic stabilization and lattice heat transport in rocksalt β-GeTe | Applied Physics Letters | AIP Publishing. https://pubs.aip.org/aip/apl/article/113/19/193902/35993.<br/>4. Tolborg, K. & Walsh, A. Exploring the High-Temperature Stabilization of Cubic Zirconia from Anharmonic Lattice Dynamics. <i>Cryst. Growth Des.</i> <b>23</b>, 3314–3319 (2023).<br/>5. Jain, A. <i>et al.</i> Commentary: The Materials Project: A materials genome approach to accelerating materials innovation. <i>APL Mater.</i> <b>1</b>, 011002 (2013).