DOI: 10.1039/d6dd00167j ISSN: 2635-098X
Stability and thermophysical properties of solid and molten BeF 2 via density functional theory and machine learning potentials
Thiago J. Pinheiro dos Santos, Luke D. Gibson, Rajni Chahal-Crockett, Shubhojit Banerjee, Mathieu Taillefumier, Ada Sedova, Juliano Schorne-Pinto, Richard Messerly, Vyacheslav S. BryantsevKernel- and graph-based machine learning potentials enable DFT-accurate property predictions of solid and liquid BeF 2 . Using PBE0-D3 improves agreement with experiments compared to PBE-D3 for crystal lattices and molten thermophysical properties.