Reproducible density functional theory predictions of bandgaps for materials
Even though reproducible computational procedures for density-functional-theory (DFT) calculations of molecular properties are well established, the additional complexities for calculations of materials properties present significant current issues. Considering a randomly selected set of 340 3D materials, we demonstrate that standard computational protocols lead to c.a. 20% occurrences of significant failures during bandgap calculations. The bandgap is a quintessential materials property that underpins the prediction of most other properties. Examined herein are the effects of the choice of the pseudopotential to describe core electrons, the plane-wave basis-set cutoff energy, and the Brillouin-zone integration. For the pseudopotential and the cutoff energy, optimization of internal computational parameters is performed. For the Brillouin-zone integration, a new computational protocol is developed that chooses grids by minimization of interpolation errors using the second-derivative matrix of the orbital energies. This is shown to provide significant enhancement over established procedures that seek merely to maximize integration-grid densities.
Authors
- Roger D. Amos (ORCID: https://orcid.org/0000-0003-2868-7994)
- Michael J. Ford (ORCID: https://orcid.org/0000-0002-8595-5900)
- Rika Kobayashi (ORCID: https://orcid.org/0000-0002-0672-833X)
- Jeffrey R. Reimers (ORCID: https://orcid.org/0000-0001-5157-7422)
- Musen Li
- Chenxi Lu (ORCID: https://orcid.org/0009-0001-5233-7823)
Institutions
- Shanghai University (CN)
- University of Technology Sydney (AU)
- University of Canberra (AU)
- IP Australia (AU)
Publication Details
- Journal
- Computational Condensed Matter
- Published
- 2025-08-29
- DOI
- https://doi.org/10.1016/j.cocom.2025.e01122
- Citations
- 1
- Primary Topic
- Machine Learning in Materials Science
- Type
- article
- Field-Weighted Citation Impact
- 0.52
Funders
- National Computational Infrastructure
- University of Technology Sydney
- National Natural Science Foundation of China
- China Postdoctoral Science Foundation
- Australian Research Council
- Young Scientists Fund