Effect of ensemble generalization on the highest-occupied Kohn-Sham eigenvalue
Eli Kraisler, Tobias Schmidt, Stephan K\"ummel, Leeor Kronik

TL;DR
This study demonstrates that ensemble generalization improves the accuracy of ionization potential predictions in density-functional theory across various exchange-correlation functionals without sacrificing total energy accuracy, and reduces parameter dependence in hybrid functionals.
Contribution
The paper applies ensemble generalization to various exchange-correlation functionals, showing systematic improvements in ionization potential predictions and reduced parameter sensitivity in hybrid functionals.
Findings
Ensemble generalization improves ionization potential predictions.
It maintains total energy accuracy across tested systems.
Reduces parameter dependence in hybrid functionals.
Abstract
There are several approximations to the exchange-correlation functional in density-functional theory that accurately predict total energy-related properties of many-electron systems, such as binding energies, bond lengths, and crystal structures. Other approximations are designed to describe potential-related processes, such as charge transfer and photoemission. However, the development of a functional which can serve the two purposes simultaneously is a long-standing challenge. Trying to address it, we employ in the current work the ensemble generalization procedure proposed in Phys. Rev. Lett. 110, 126403 (2013). Focusing on the prediction of the ionization potential via the highest occupied Kohn-Sham eigenvalue, we examine a variety of exchange-correlation approximations: the local spin-density approximation, semi-local generalized gradient approximations, and global and local hybrid…
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.
