A universal augmentation framework for long-range electrostatics in machine learning interatomic potentials

• URL: http://arxiv.org/abs/2507.14302v1
• Date: Fri, 18 Jul 2025 18:21:45 GMT
• Title: A universal augmentation framework for long-range electrostatics in machine learning interatomic potentials
• Authors: Dongjin Kim, Xiaoyu Wang, Peichen Zhong, Daniel S. King, Theo Jaffrelot Inizan, Bingqing Cheng,
• Abstract summary: Latent Ewald Summation (LES) method infers electrostatic interactions, polarization, and Born effective charges (BECs)<n>We present LES as a standalone library, compatible with any short-range MLIP, and demonstrate its integration with methods such as MACE, NequIP, CACE, and CHGNet.<n>We scale LES to large and chemically diverse data by training MACELES-OFF on the SPICE set containing molecules and clusters, making a universal MLIP with electrostatics for organic systems including biomolecules.
• Score: 8.414405667524889
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Most current machine learning interatomic potentials (MLIPs) rely on short-range approximations, without explicit treatment of long-range electrostatics. To address this, we recently developed the Latent Ewald Summation (LES) method, which infers electrostatic interactions, polarization, and Born effective charges (BECs), just by learning from energy and force training data. Here, we present LES as a standalone library, compatible with any short-range MLIP, and demonstrate its integration with methods such as MACE, NequIP, CACE, and CHGNet. We benchmark LES-enhanced models on distinct systems, including bulk water, polar dipeptides, and gold dimer adsorption on defective substrates, and show that LES not only captures correct electrostatics but also improves accuracy. Additionally, we scale LES to large and chemically diverse data by training MACELES-OFF on the SPICE set containing molecules and clusters, making a universal MLIP with electrostatics for organic systems including biomolecules. MACELES-OFF is more accurate than its short-range counterpart (MACE-OFF) trained on the same dataset, predicts dipoles and BECs reliably, and has better descriptions of bulk liquids. By enabling efficient long-range electrostatics without directly training on electrical properties, LES paves the way for electrostatic foundation MLIPs.

Related papers

• Accurate Ab-initio Neural-network Solutions to Large-Scale Electronic Structure Problems [52.19558333652367]
  We present finite-range embeddings (FiRE) for accurate large-scale ab-initio electronic structure calculations.<n>FiRE reduces the complexity of neural-network variational Monte Carlo (NN-VMC) by $sim ntextel$, the number of electrons.<n>We validate our method's accuracy on various challenging systems, including biochemical compounds and organometallic compounds.
  arXiv  Detail & Related papers  (2025-04-08T14:28:54Z)
• Machine learning interatomic potential can infer electrical response [3.502816712907136]
  Machine learning interatomic potentials (MLIPs) offer an efficient and scalable alternative to quantum mechanical methods.<n>We show that polarization and Born effective charge (BEC) tensors can be directly extracted from long-range MLIPs.<n>This work extends the capability MLIPs to predict electrical response--without training on charges or polarization or BECs--and enables accurate modeling of electric-field-driven processes in diverse systems at scale.
  arXiv  Detail & Related papers  (2025-04-07T15:14:07Z)
• Equivariant Machine Learning Interatomic Potentials with Global Charge Redistribution [1.6112718683989882]
  Machine learning interatomic potentials (MLIPs) provide a computationally efficient alternative to quantum mechanical simulations for predicting material properties.<n>We develop a new equivariant MLIP incorporating long-range Coulomb interactions through explicit treatment of electronic degrees of freedom.<n>We systematically evaluate our model across a range of benchmark periodic and non-periodic datasets, demonstrating that it outperforms both short-range equivariant and long-range invariant MLIPs in energy and force predictions.
  arXiv  Detail & Related papers  (2025-03-23T05:26:55Z)
• Learning charges and long-range interactions from energies and forces [3.502816712907136]
  We introduce the Latent Ewald Summation method, which captures long-range electrostatics without explicitly learning atomic charges or charge equilibration.<n>We benchmark LES on diverse and challenging systems, including charged molecules, ionic liquid, electrolyte solution, polar dipeptides, surface adsorber, electrolyte/solid interfaces, and solid-solid interfaces.<n>Our results show that LES can effectively infer physical partial charges, dipole and quadrupole moments, as well as achieve better accuracy compared to methods that explicitly learn charges.
  arXiv  Detail & Related papers  (2024-12-19T23:24:44Z)
• Electron-Electron Interactions in Device Simulation via Non-equilibrium Green's Functions and the GW Approximation [71.63026504030766]
  electron-electron (e-e) interactions must be explicitly incorporated in quantum transport simulation.<n>This study is the first one reporting large-scale atomistic quantum transport simulations of nano-devices under non-equilibrium conditions.
  arXiv  Detail & Related papers  (2024-12-17T15:05:33Z)
• Predicting ionic conductivity in solids from the machine-learned potential energy landscape [68.25662704255433]
  We propose an approach for the quick and reliable screening of ionic conductors through the analysis of a universal interatomic potential.<n>Eight out of the ten highest-ranked materials are confirmed to be superionic at room temperature in first-principles calculations.<n>Our method achieves a speed-up factor of approximately 50 compared to molecular dynamics driven by a machine-learning potential, and is at least 3,000 times faster compared to first-principles molecular dynamics.
  arXiv  Detail & Related papers  (2024-11-11T09:01:36Z)
• Learning Equivariant Non-Local Electron Density Functionals [51.721844709174206]
  We introduce Equivariant Graph Exchange Correlation (EG-XC), a novel non-local XC functional based on equivariant graph neural networks (GNNs)<n>By applying an equivariant GNN on this point cloud, we capture molecular-range interactions in a scalable and accurate manner.<n> EG-XC excels in data efficiency and molecular size extrapolation on QM9, matching force fields trained on 5 times more and larger molecules.
  arXiv  Detail & Related papers  (2024-10-10T14:31:45Z)
• Latent Ewald summation for machine learning of long-range interactions [0.0]
  We introduce a straightforward and efficient method to account for long-range interactions by learning a latent variable from local atomic descriptors.<n>We demonstrate that in systems including charged and polar molecular dimers, bulk water, and water-vapor interface, standard short-ranged MLIPs can lead to unphysical predictions.<n>The long-range models effectively eliminate these artifacts, with only about twice the computational cost of short-range MLIPs.
  arXiv  Detail & Related papers  (2024-08-27T16:03:18Z)
• Neural Pfaffians: Solving Many Many-Electron Schrödinger Equations [58.130170155147205]
  Neural wave functions accomplished unprecedented accuracies in approximating the ground state of many-electron systems, though at a high computational cost. Recent works proposed amortizing the cost by learning generalized wave functions across different structures and compounds instead of solving each problem independently. This work tackles the problem by defining overparametrized, fully learnable neural wave functions suitable for generalization across molecules.
  arXiv  Detail & Related papers  (2024-05-23T16:30:51Z)
• A predictive machine learning force field framework for liquid electrolyte development [11.463808946378743]
  We introduce BAMBOO, a predictive framework for molecular dynamics (MD) simulations, with a demonstration of its capability in the context of liquid electrolyte for lithium batteries.<n>We design a physics-inspired graph equivariant transformer architecture as the backbone of BAMBOO to learn from quantum mechanical simulations.<n>We also introduce an ensemble knowledge distillation approach and apply it to MLFFs to reduce the fluctuation of observations from MD simulations.
  arXiv  Detail & Related papers  (2024-04-10T17:31:49Z)
• A Hybrid Quantum-Classical Method for Electron-Phonon Systems [40.80274768055247]
  We develop a hybrid quantum-classical algorithm suitable for this type of correlated systems. This hybrid method tackles with arbitrarily strong electron-phonon coupling without increasing the number of required qubits and quantum gates. We benchmark the new method by applying it to the paradigmatic Hubbard-Holstein model at half filling, and show that it correctly captures the competition between charge density wave and antiferromagnetic phases.
  arXiv  Detail & Related papers  (2023-02-20T08:08:51Z)
• Topological Josephson Junctions in the Integer Quantum Hall Regime [42.408991654684876]
  tunable Josephson junctions (TJJs) are desirable platforms for investigating the anomalous Josephson effect and topological quantum insulator applications. We propose a robust and electrostatically tunable TJJ by combining the physics of the integer quantum Hall (IQH) regime and of superconductors. They are of particular relevance towards scalable and robust Andreev-qubit platforms, and also for efficient phase batteries.
  arXiv  Detail & Related papers  (2022-11-04T16:45:07Z)
• BIGDML: Towards Exact Machine Learning Force Fields for Materials [55.944221055171276]
  Machine-learning force fields (MLFF) should be accurate, computationally and data efficient, and applicable to molecules, materials, and interfaces thereof. Here, we introduce the Bravais-Inspired Gradient-Domain Machine Learning approach and demonstrate its ability to construct reliable force fields using a training set with just 10-200 atoms.
  arXiv  Detail & Related papers  (2021-06-08T10:14:57Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.