Electrostatic interactions in atomistic and machine-learned potentials for polar materials

TL;DR

This paper presents a new first-principles-based model to accurately incorporate long-range electrostatic interactions into atomistic potentials for polar materials, improving simulation fidelity without retraining.

Contribution

The study introduces a physically grounded model that seamlessly integrates long-range electrostatics into existing potentials using observable quantities, enhancing accuracy in polar material simulations.

Findings

Reproduces LO-TO splitting in BaTiO3

Accurately models long-wavelength phonons

Integrates with existing force fields without retraining

Abstract

Long-range electrostatic interactions critically affect polar materials. However, state-of-the-art atomistic potentials, such as neural networks or Gaussian approximation potentials employed in large-scale simulations, often neglect the role of these long-range electrostatic interactions. This study introduces a novel model derived from first principles to evaluate the contribution of long-range electrostatic interactions to total energies, forces, and stresses. The model is designed to integrate seamlessly with existing short-range force fields without further first-principles calculations or retraining. The approach relies solely on physical observables, like the dielectric tensor and Born effective charges, that can be consistently calculated from first principles. We demonstrate that the model reproduces critical features, such as the LO-TO splitting and the long-wavelength phonon dispersions of polar materials, with benchmark results on the cubic phase of barium titanate (BaTiO$_3$).