Incorporating Long-Range Interactions via the Multipole Expansion into Ground and Excited-State Molecular Simulations

TL;DR

FieldMACE is a novel machine learning framework that efficiently models long-range interactions in molecular simulations by integrating multipole expansion, improving accuracy and scalability for ground and excited states.

Contribution

It introduces FieldMACE, combining multipole expansion with message-passing architectures to better capture long-range effects in molecular simulations.

Findings

Outperforms previous models in accuracy and efficiency

Successfully simulates nonadiabatic excited-state dynamics

Enhanced data efficiency through transfer learning

Abstract

Simulating long-range interactions remains a significant challenge for molecular machine learning potentials due to the need to accurately capture interactions over large spatial regions. In this work, we introduce FieldMACE, an extension of the message-passing atomic cluster expansion (MACE) architecture that integrates the multipole expansion to model long-range interactions more efficiently. By incorporating the multipole expansion, FieldMACE effectively captures environmental and long-range effects in both ground and excited states. Benchmark evaluations demonstrate its superior performance in predictions and computational efficiency compared to previous architectures, as well as its ability to accurately simulate nonadiabatic excited-state dynamics. Furthermore, transfer learning from foundational models enhances data efficiency, making FieldMACE a scalable, robust, and transferable framework for large-scale molecular simulations.