Learning Long-Range Interactions in Equivariant Machine Learning Interatomic Potentials via Electronic Degrees of Freedom

Machine learning interatomic potentials (MLIPs) provide a computationally efficient alternative to quantum mechanical simulations for predicting material properties. Message-passing graph neural networks, commonly used in these MLIPs, rely on local descriptor-based symmetry functions to model atomic interactions. However, such local descriptor-based approaches struggle with systems exhibiting long-range interactions, charge transfer, and compositional heterogeneity. In this work, we develop a new equivariant MLIP incorporating long-range Coulomb interactions through the explicit treatment of electronic degrees of freedom, specifically global charge distribution within the system. This is achieved using a charge equilibration scheme based on the predicted atomic electronegativities. We systematically evaluate our model across a range of benchmark periodic and nonperiodic data sets, demonstrating that it outperforms both short-range equivariant and long-range invariant MLIPs in energy and force predictions. Due to the explicit treatment of long-range interactions using partial charges, our model achieves higher accuracy using a 4 Å cutoff radius than a short-range model with a 6 Å cutoff. Our approach enables more accurate and efficient simulations of systems with long-range interactions and charge heterogeneity, expanding the applicability of MLIPs in computational materials science.