Uncertainty Quantification for Materials Properties in Density Functional Theory with k-Point Density

TL;DR

This study analyzes how k-point density affects the precision of material properties in density functional theory calculations, recommending it as an efficient convergence parameter for high-throughput materials databases.

Contribution

It establishes the correlation between k-point density and property precision, providing practical guidelines for setting convergence criteria in DFT calculations.

Findings

k-point density correlates well with property precision

recommended k-point density as a convergence parameter

predicted typical precisions for high-throughput DFT data

Abstract

Many computational databases emerged over the last five years that report material properties calculated with density functional theory. The properties in these databases are commonly calculated to a precision that is set by choice of the basis set and the k-point density for the Brillouin zone integration. We determine how the precision of properties obtained from the Birch equation of state for 29 transition metals and aluminum in the three common structures -- fcc, bcc, and hcp -- correlate with the k-point density and the precision of the energy. We show that the precision of the equilibrium volume, bulk modulus, and the pressure derivative of the bulk modulus correlate comparably well with the k-point density and the precision of the energy, following an approximate power law. We recommend the k-point density as the convergence parameter because it is computationally efficient, easy to use as a direct input parameter, and correlates with property precision at least as well as the energy precision. We predict that common k-point density choices in high throughput DFT databases result in precision for the volume of 0.1%, the bulk modulus of 1%, and the pressure derivative of 10%.