qNEP: A Highly Efficient Neuroevolution Potential with Dynamic Charges for Large-Scale Atomistic Simulations
Journal article, 2026
Although electrostatics can be incorporated into machine-learned interatomic potentials, existing approaches are computationally very demanding, limiting large-scale, long-time simulations of electrostatics-driven phenomena such as dielectric response, infrared activity, and field-matter coupling. Here, we extend the neuroevolution potential (NEP), a highly efficient machine-learned interatomic potential, to a charge-aware framework (qNEP) by introducing explicit, environment-dependent partial charges. Each ionic partial charge is represented by a neural network as a function of the local descriptor vector, analogous to the NEP site-energy model. This formulation enables the direct prediction of the Born effective charge tensor for each ion and, consequently, the polarization. As a result, dielectric properties, infrared spectra, and coupling to external electric fields can be evaluated within a unified framework. We derive consistent expressions for the forces and virials that explicitly account for the position dependence of the partial charges. The qNEP method has been implemented in the free-and-open-source GPUMD package with support for both Ewald summation and particle-particle particle-mesh treatments of electrostatics. We demonstrate the accuracy and efficiency of the qNEP approach through representative applications to water, Li7La3Zr2O12, BaTiO3, and a magnesium-water interface. These results show that qNEP enables accurate atomistic simulations with explicit long-range electrostatics, scalable to million-atom systems on nanosecond time scales using consumer-grade GPUs.
Author
Zheyong Fan
Bohai University
Suzhou Lab
Benrui Tang
Bohai University
Esmée Berger
Chalmers, Physics, Condensed Matter and Materials Theory
Ethan Berger
Chalmers, Physics, Condensed Matter and Materials Theory
Erik Fransson
Chalmers, Physics, Condensed Matter and Materials Theory
Ke Xu
Bohai University
Zihan Yan
Westlake University
Zhoulin Liu
Harbin Institute of Technology
Zichen Song
Southern University of Science and Technology
City University of Hong Kong
Haikuan Dong
Bohai University
Shunda Chen
George Washington University
Lei Li
Southern University of Science and Technology
Ziliang Wang
Shandong University
Yizhou Zhu
Westlake University
Julia Wiktor
Chalmers, Physics, Condensed Matter and Materials Theory
Paul Erhart
Chalmers, Physics, Condensed Matter and Materials Theory
Journal of Chemical Theory and Computation
1549-9618 (ISSN) 1549-9626 (eISSN)
Vol. In PressProton- och hydridjon-ledning i perovskiter
Swedish Energy Agency (45410-1), 2018-01-01 -- 2021-12-31.
Harnessing Localized Charges for Advancing Polar Materials Engineering (POLARISE)
European Commission (EC) (EC/HE/101162195), 2025-01-01 -- 2029-12-31.
Ab Initio Description of Complete Semiconductor Devices
Swedish Foundation for Strategic Research (SSF) (FFL21-0129), 2022-08-01 -- 2027-12-31.
Phase behavior and electronic properties of mixed halide perovskites from atomic scale simulations
Swedish Research Council (VR) (2020-04935), 2020-12-01 -- 2024-11-30.
Subject Categories (SSIF 2025)
Theoretical Chemistry
Condensed Matter Physics
Physical Chemistry
Infrastructure
Chalmers e-Commons (incl. C3SE, 2020-)
DOI
10.1021/acs.jctc.6c00146
PubMed
42007685
Related datasets
Data and models supporting "qNEP: A highly efficient neuroevolution potential with dynamic charges for large-scale atomistic simulations" [dataset]
DOI: https://zenodo.org/records/18335947