Polarizable atomic multipoles for learning long-range electrostatics

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kim, Dongjin, King, Daniel S., Park, Yoonjae, Savoj, Roya, Hamel, Sebastien, Wang, Xiaoyu, Cheng, Bingqing
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911655055589376
author Kim, Dongjin
King, Daniel S.
Park, Yoonjae
Savoj, Roya
Hamel, Sebastien
Wang, Xiaoyu
Cheng, Bingqing
author_facet Kim, Dongjin
King, Daniel S.
Park, Yoonjae
Savoj, Roya
Hamel, Sebastien
Wang, Xiaoyu
Cheng, Bingqing
contents Long-range electrostatics and polarization remain central obstacles to extending machine learning interatomic potentials (MLIPs) to ionic, polar, and interfacial systems. Here, we introduce a semi-local framework for learning electrostatics from energies and forces using polarizable atomic multipoles. Local equivariant descriptors predict environment-dependent latent monopoles, dipoles, and quadrupoles, while residual non-local charge transfer and polarization are captured by non-self-consistent linear response in induced charges and dipoles. Across four diverse benchmarks and four short-range MLIP architectures, the multipole hierarchy and response terms systematically improve potential energy surface accuracy, with the largest gains in systems where long-range effects are essential. More importantly, the learned latent variables recover physically meaningful electrical responses: accurate Born effective charge tensors, emergent polarizabilities, infrared spectra in close agreement with experiments, and semi-quantitative Raman spectra for bulk water and hybrid MAPbI$_3$ perovskite. This systematically improvable, physically transparent framework enables MLIPs trained on standard energy and force labels to predict polarization-sensitive observables.
format Preprint
id arxiv_https___arxiv_org_abs_2605_05746
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Polarizable atomic multipoles for learning long-range electrostatics
Kim, Dongjin
King, Daniel S.
Park, Yoonjae
Savoj, Roya
Hamel, Sebastien
Wang, Xiaoyu
Cheng, Bingqing
Materials Science
Machine Learning
Chemical Physics
Computational Physics
Long-range electrostatics and polarization remain central obstacles to extending machine learning interatomic potentials (MLIPs) to ionic, polar, and interfacial systems. Here, we introduce a semi-local framework for learning electrostatics from energies and forces using polarizable atomic multipoles. Local equivariant descriptors predict environment-dependent latent monopoles, dipoles, and quadrupoles, while residual non-local charge transfer and polarization are captured by non-self-consistent linear response in induced charges and dipoles. Across four diverse benchmarks and four short-range MLIP architectures, the multipole hierarchy and response terms systematically improve potential energy surface accuracy, with the largest gains in systems where long-range effects are essential. More importantly, the learned latent variables recover physically meaningful electrical responses: accurate Born effective charge tensors, emergent polarizabilities, infrared spectra in close agreement with experiments, and semi-quantitative Raman spectra for bulk water and hybrid MAPbI$_3$ perovskite. This systematically improvable, physically transparent framework enables MLIPs trained on standard energy and force labels to predict polarization-sensitive observables.
title Polarizable atomic multipoles for learning long-range electrostatics
topic Materials Science
Machine Learning
Chemical Physics
Computational Physics
url https://arxiv.org/abs/2605.05746