Isotope Effects at Classical Cost through Mass-Differentiable Machine Learning
Abstract
Isotope effects govern fractionation and modulate reactivity, with applications from hydrogen energy to environmental science and catalysis, yet predicting them requires resolving small isotope-dependent free-energy differences that remain very challenging for conventional path-integral simulations in complex systems. Here we introduce iso-EPIGS, a path-integral coarse-graining framework built on a mass-differentiable neural network that reconstructs the mass- and temperature-dependent path integral centroid free-energy surface. Classical molecular dynamics on the learned surface yields rigorous isotope-resolved thermodynamics without explicit path-integral sampling. Benchmarks spanning gas, liquid, and crystalline phases, including liquid water and oxalic acid crystal, reproduce reference path-integral isotope free-energy differences, enthalpies, and lattice parameters at near classical computational cost. Crucially, iso-EPIGS trained solely on all-H and all-D isotopologues retains high accuracy for unseen partially deuterated isotopologues, demonstrating robust transferability across nuclear masses. Iso-EPIGS makes accurate isotope-effect simulations feasible for complex systems and lays the groundwork for foundation models of isotope effects across chemical space.
Related
- Machine learning for sample-based quantum diagonalization: a review of generative configuration recovery and the classical-simulability frontier
- Hybrid quantum-classical machine learning potential with variational quantum circuits
- Machine learning-driven inverse design of high-hardness multi-principal element alloys through symbolic regression
- Growth driven phase transitions in Zinc Oxide nanoparticles through machine-learning assisted simulations
- Thermodynamically Constrained Deep Learning for Hydrogen Sorption Prediction: Overcoming Classical Isotherm Generalization Failure through Physics-Informed Cross-Lithology Transferability
BibTeX
@misc{du2026isotope,
title = {{Isotope Effects at Classical Cost through Mass-Differentiable Machine Learning}},
author = {Ming-Zheng Du and Shi-Yu He and Jing Shen and Ziyan Ye and Jia-Xi Zeng and Dong H. Zhang and Venkat Kapil and Wei Fang},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.11325},
eprint = {2610.11325},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.11325},
}
RIS
TY - UNPB
TI - Isotope Effects at Classical Cost through Mass-Differentiable Machine Learning
AU - Ming-Zheng Du
AU - Shi-Yu He
AU - Jing Shen
AU - Ziyan Ye
AU - Jia-Xi Zeng
AU - Dong H. Zhang
AU - Venkat Kapil
AU - Wei Fang
PY - 2026
DA - 2026/10/08
PB - arXiv
DO - 10.48550/arxiv.2610.11325
UR - https://arxiv.org/abs/2610.11325
AB - Isotope effects govern fractionation and modulate reactivity, with applications from hydrogen energy to environmental science and catalysis, yet predicting them requires resolving small isotope-dependent free-energy differences that remain very challenging for conventional path-integral simulations in complex systems. Here we introduce iso-EPIGS, a path-integral coarse-graining framework built on a mass-differentiable neural network that reconstructs the mass- and temperature-dependent path integral centroid free-energy surface. Classical molecular dynamics on the learned surface yields rigorous isotope-resolved thermodynamics without explicit path-integral sampling. Benchmarks spanning gas, liquid, and crystalline phases, including liquid water and oxalic acid crystal, reproduce reference path-integral isotope free-energy differences, enthalpies, and lattice parameters at near classical computational cost. Crucially, iso-EPIGS trained solely on all-H and all-D isotopologues retains high accuracy for unseen partially deuterated isotopologues, demonstrating robust transferability across nuclear masses. Iso-EPIGS makes accurate isotope-effect simulations feasible for complex systems and lays the groundwork for foundation models of isotope effects across chemical space.
ER -