Classical-to-DFT multi-fidelity approach for developing machine learning potentials of aqueous systems

Machine learning interatomic potentials (MLIPs) offer near-ab initio accuracy at a fraction of the cost of conventional electronic structure calculations, but their reliability depends critically on the quality and diversity of training data. Here, we present a classical-to-density functional theory (DFT) multi-fidelity workflow for efficient dataset generation, in which long classical molecular dynamics trajectories are supplemented with selectively DFT-optimized configurations to train neural network potentials. We demonstrate this approach on four systems: bulk water, water at the air-water interface, CaSO4 in bulk water, and CaSO4 at the interface. In all cases, the trained models faithfully reproduce reference DFT properties, validating the workflow across both simple and chemically complex environments. We further probe the extrapolation capabilities of the models, showing that bulk-trained networks can generalize to different temperatures, pressures, and interfacial geometries when datasets include targeted augmented configurations. This multi-fidelity strategy thus enables accurate and transferable MLIPs for complex solvated and interfacial systems, bridging the gap between computational efficiency and first-principles accuracy.

Authors

Institutions

Publication Details

Journal
The Journal of Chemical Physics
Published
2026-11-16
DOI
https://doi.org/10.1063/5.0341814
Primary Topic
Machine Learning in Materials Science
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Classical-to-DFT multi-fidelity approach for developing machine learning potentials of aqueous systems

Kendra Coronado, Mirza Galib, Md. Abu Bakar Siddique, Mohammad Mohib Ullah Badhon
5 links
The Journal of Chemical Physics
Machine Learning in Materials Science
article

Classical-to-DFT multi-fidelity approach for developing machine learning potentials of aqueous systems

Kendra Coronado, Mirza Galib, Md. Abu Bakar Siddique, Mohammad Mohib Ullah Badhon
article en

Abstract

Machine learning interatomic potentials (MLIPs) offer near-ab initio accuracy at a fraction of the cost of conventional electronic structure calculations, but their reliability depends critically on the quality and diversity of training data. Here, we present a classical-to-density functional theory (DFT) multi-fidelity workflow for efficient dataset generation, in which long classical molecular dynamics trajectories are supplemented with selectively DFT-optimized configurations to train neural network potentials. We demonstrate this approach on four systems: bulk water, water at the air-water interface, CaSO4 in bulk water, and CaSO4 at the interface. In all cases, the trained models faithfully reproduce reference DFT properties, validating the workflow across both simple and chemically complex environments. We further probe the extrapolation capabilities of the models, showing that bulk-trained networks can generalize to different temperatures, pressures, and interfacial geometries when datasets include targeted augmented configurations. This multi-fidelity strategy thus enables accurate and transferable MLIPs for complex solvated and interfacial systems, bridging the gap between computational efficiency and first-principles accuracy.

The Journal of Chemical PhysicsVol. 165(19)
Howard University (US)
National Energy Research Scientific Computing Center, Office of Science, Office of Advanced Cyberinfrastructure
Machine Learning in Materials Science
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon