A flexible and scalable scheme for mixing computed formation energies from different levels of theory

Abstract Computational materials discovery efforts are enabled by large databases of properties derived from high-throughput density functional theory (DFT), which now contain millions calculations at the generalized gradient approximation (GGA) level theory. It is feasible to carry out using more accurate methods, such as meta-GGA DFT; however recomputing an entire database with a higher-fidelity method would not effectively leverage enormous investment computational resources embodied in existing calculations. Instead, we propose here general procedure higher-fidelity, low-coverage (e.g., for selected chemical systems) can be combined lower-fidelity, high-coverage GGA calculations) robust and scalable manner. We then use legacy PBE(+ U ) new r 2 SCAN Materials Project demonstrate that our scheme improves solid aqueous phase stability predictions, discuss practical considerations its implementation.