Semiclassical atom theory applied to solid-state physics

TL;DR

This paper extends semiclassical atom theory to develop a new GGA exchange-correlation functional that performs well for solids, large atoms, and chemistry, advancing density functional theory.

Contribution

It introduces a fourth-order gradient expansion for exchange energy and constructs a non-empirical GGA functional applicable to diverse systems.

Findings

The new functional is competitive with state-of-the-art GGAs for solids.

It maintains reasonable accuracy for large atoms.

It is non-empirical and broadly applicable.

Abstract

Using the semiclassical neutral atom theory, we extend to fourth order the modified gradient expansion of the exchange energy of density functional theory. This expansion can be applied both to large atoms and solid-state problems. Moreover, we show that it can be employed to construct a simple and non-empirical generalized gradient approximation (GGA) exchange-correlation functional competitive with state-of-the-art GGAs for solids, but also reasonably accurate for large atoms and ordinary chemistry.