Density-Based Long-Range Electrostatic Descriptors for Machine Learning Force Fields
Abstract
This study presents a long-range descriptor for machine learning force fields (MLFFs) that maintains translational and rotational symmetry, similar to short-range descriptors while being able to incorporate long-range electrostatic interactions. The proposed descriptor is based on an atomic density representation and is structurally similar to classical short-range atom-centered descriptors, making it straightforward to integrate into machine learning schemes. The effectiveness of our model is demonstrated through comparative analysis with the long-distance equivariant (LODE) descriptor. In a toy model with purely electrostatic interactions, our model achieves errors below 0.1%, worse than LODE but still very good. For real materials, we perform tests for liquid NaCl, rock salt NaCl, and solid zirconia. For NaCl, the present descriptors improve on short-range density descriptors, reducing errors by a factor of two to three and coming close to message-passing networks. However, for solid zirconia, no improvements are observed with the present approach, while message-passing networks reduce the error by almost a factor of two to three. Possible shortcomings of the present model are briefly discussed.
AI-Generated Overview
- Research Focus: The study presents a new long-range descriptor for machine learning force fields (MLFFs) that retains rotational and translational symmetry while effectively incorporating long-range electrostatic interactions.
- Methodology: The authors develop a long-range descriptor based on atomic density representations, comparing its performance with existing models such as the long-distance equivariant (LODE) descriptor through tests on simplified and real materials.
- Results: The new descriptor achieved less than 0.1% error in a simple electrostatic system and showed significant error reduction (two to three times) for liquid NaCl compared to short-range density descriptors, although no improvement was observed for solid zirconia.
- Key Contribution(s): The work introduces a framework for including long-range electrostatic interactions in MLFFs while presenting a flexible descriptor that aligns structurally with conventional short-range models.
- Significance: This approach addresses a significant gap in current MLFF methodologies by allowing the accurate representation of long-range interactions, improving predictive accuracy for materials where such effects are crucial.
- Broader Applications: The developed descriptors and methodologies can be applied to various materials sciences problems, especially those involving ionic liquids and other systems where long-range electrostatic interactions play a pivotal role.