How Semilocal Are Semilocal Density Functional Approximations? -Tackling Self-Interaction Error in One-Electron Systems
Akilan Ramasamy, Lin Hou, Jorge Vega Bazantes, Tom J. P. Irons, Andrew M. Wibowo-Teale, Timo Lebeda, Jianwei Sun

TL;DR
This paper introduces a new meta-GGA density functional that significantly reduces self-interaction error in one-electron systems, improving accuracy over existing semilocal functionals while maintaining computational efficiency.
Contribution
A non-empirical meta-GGA functional incorporating the Laplacian of the density that mitigates self-interaction error within semilocal approximations.
Findings
Reduces self-interaction error in $H_2^+$
Matches exact binding energy at equilibrium
Outperforms PBE and SCAN across bond lengths
Abstract
Self-interaction error (SIE), arising from the imperfect cancellation of the spurious classical Coulomb interaction between an electron and itself, is a persistent challenge in modern density functional approximations. This issue is illustrated using the prototypical one-electron system . While significant efforts have been made to eliminate SIE through the development of computationally expensive nonlocal density functionals, it is equally important to explore whether SIE can be mitigated within the framework of more efficient semilocal density functionals. In this study, we present a non-empirical meta-generalized gradient approximation (meta-GGA) that incorporates the Laplacian of the electron density. Our results demonstrate that the meta-GGA significantly reduces SIE, yielding a binding energy curve for that matches the exact solution at equilibrium and improves…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdvanced Chemical Physics Studies · Surface and Thin Film Phenomena · Machine Learning in Materials Science
