Magnetic Moment Tensor Potentials for collinear spin-polarized materials reproduce different magnetic states of bcc Fe

TL;DR

This paper introduces magnetic Moment Tensor Potentials (mMTPs), a machine-learning approach that accurately models vibrational and magnetic properties of collinear spin-polarized materials like bcc iron, enabling advanced simulations.

Contribution

The paper presents a novel two-step minimization scheme for mMTPs that effectively captures both atomic and spin degrees of freedom in magnetic materials.

Findings

Successfully reproduces vibrational and magnetic properties of bcc Fe

Enables phonon calculations for different magnetic states

Supports molecular dynamics with fluctuating magnetic moments

Abstract

We present the magnetic Moment Tensor Potentials (mMTPs), a class of machine-learning interatomic potentials, accurately reproducing both vibrational and magnetic degrees of freedom as provided, e.g., from first-principles calculations. The accuracy is achieved by a two-step minimization scheme that coarse-grains the atomic and the spin space. The performance of the mMTPs is demonstrated for the prototype magnetic system bcc iron, with applications to phonon calculations for different magnetic states, and molecular dynamics simulations with fluctuating magnetic moments.