Exchange-correlation functionals for band gaps of solids: benchmark, reparametrization and machine learning

• Published in: npj computational materials Vol. 6; no. 1
• Main Authors: Borlido, Pedro, Schmidt, Jonathan, Huran, Ahmad W., Tran, Fabien, Marques, Miguel A. L., Botti, Silvana
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 10.07.2020; Nature Publishing Group
• Subjects: 639/301/1034/1038; 639/766/119/995; Approximation; Benchmarks; Characterization and Evaluation of Materials; Chemistry and Materials Science; Computational Intelligence; Computer applications; Density functional theory; Energy gap; Exchanging; Learning algorithms; Machine learning; Materials information; Materials Science; Mathematical and Computational Engineering; Mathematical and Computational Physics; Mathematical Modeling and Industrial Mathematics; Parameter identification; Theoretical
• ISSN: 2057-3960; 2057-3960
• DOI: 10.1038/s41524-020-00360-0

We conducted a large-scale density-functional theory study on the influence of the exchange-correlation functional in the calculation of electronic band gaps of solids. First, we use the large materials data set that we have recently proposed to benchmark 21 different functionals, with a particular focus on approximations of the meta-generalized-gradient family. Combining these data with the results for 12 functionals in our previous work, we can analyze in detail the characteristics of each approximation and identify its strong and/or weak points. Beside confirming that mBJ, HLE16 and HSE06 are the most accurate functionals for band gap calculations, we reveal several other interesting functionals, chief among which are the local Slater potential approximation, the GGA AK13LDA, and the meta-GGAs HLE17 and TASK. We also compare the computational efficiency of these different approximations. Relying on these data, we investigate the potential for improvement of a promising subset of functionals by varying their internal parameters. The identified optimal parameters yield a family of functionals fitted for the calculation of band gaps. Finally, we demonstrate how to train machine learning models for accurate band gap prediction, using as input structural and composition data, as well as approximate band gaps obtained from density-functional theory.