Rehabilitation of PBE-GGA for Layered Materials

TL;DR

This paper introduces PBE+rVV10L, a modified density functional that significantly improves the prediction of structural and energetic properties of layered materials, aiding theoretical studies.

Contribution

The paper develops and validates a new functional, PBE+rVV10L, optimized for layered materials, demonstrating improved accuracy over existing methods.

Findings

Accurately predicts structural properties of 28 layered materials.

Achieves excellent energetic property predictions.

Outperforms previous functionals in layered material modeling.

Abstract

The structural and energetic properties of layered materials propose a challenge to density functional theory with common semilocal approximations to the exchange-correlation. By combining the most-widely used semilocal generalized gradient approximation (GGA), Perdew--Burke--Ernzerhof (PBE), with the revised Vydrov--Van Voorhis non-local correlation functional (rVV10), both excellent structural and energetic properties of 28 layered materials were recovered with a judicious parameter selection. We term the resulting functional as PBE+rVV10L with "L" denoting for layered materials. Such combination is not new, and involves only refitting a single global parameter, however the resulting excellency suggests such corrected PBE for many aspects of theoretical studies on layered materials. For comparison, we also present the results for PBE+rVV10 where the parameter is determined by the 22 interaction energies between molecules.