Assessment of a nonempirical semilocal density functional on solids and surfaces

TL;DR

This paper evaluates the Tao-Mo nonempirical semilocal density functional, demonstrating its high accuracy in predicting properties of solids and surfaces, and showing it outperforms previous similar functionals.

Contribution

The study provides a comprehensive assessment of the Tao-Mo functional's performance on solids and surfaces, highlighting its improved accuracy over existing nonempirical functionals.

Findings

Mean absolute error of 0.017 Å for lattice constants

Bulk modulus errors averaging 7.0 GPa

Cohesive energy errors around 0.08 eV

Abstract

Recently, Tao and Mo developed a new nonempirical semilocal exchange-correlation density functional. The exchange part of this functional is derived from a density matrix expansion corrected to reproduce the fourth-order gradient expansion in the slowly varying limit, while the correlation part is based on the TPSS correlation model with a modification for the low-density limit. In the present work, the Tao-Mo functional is assessed by calculations on a variety of solids and jellium surfaces. This includes 22 lattice constants and bulk moduli, 7 cohesive energies, and jellium surface exchange and correlation energies for the density parameter rs in the range from 2 to 3 bohrs. Our calculations show that this meta-generalized gradient approximation can yield consistently remarkable accuracy for the properties considered here, with mean absolute errors of 0.017 {\AA} for lattice constants, 7.0 GPa for bulk moduli, 0.08 eV for cohesive energies, and 35 erg/cm2 for surface exchange-correlation energies, substantially improving upon existing nonempirical semilocal density functionals.