Benchmarking Universal Machine Learning Interatomic Potentials for Alkali-Ion Battery Kinetics
ACS Materials Letters · 06.07.2026 · 8 (7) · 1891-1899
DOI: 10.1021/acsmaterialslett.6c00134 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACS Materials Letters, online erschienen am 28.05.2026. Verfasst von Xingyu Guo und 2 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 34 Literaturangaben. Gefördert von: City University of Hong Kong, National Natural Science Foundation of China.
Abstract (Originalsprache)
Accelerating alkali-ion battery discovery requires accurate modeling of atomic-scale kinetics, yet the reliability of universal machine learning interatomic potentials (uMLIPs) in capturing these high-energy landscapes remains uncertain. Here, we systematically benchmark state-of-the-art uMLIPs, including M3GNet, CHGNet, MatterSim, MACE, SevenNet, GRACE, and Orb, against DFT baselines for cathodes and solid electrolytes. We find that the Orb-v3 family excels in static migration barrier predictions (MAE ≈ 75–111 meV), driven primarily by architectural refinements. Conversely, for dynamic transport, the GRACE model trained on the OMat24 dataset demonstrates superior fidelity in reproducing ion diffusivities and structural correlations. Our results reveal that while architectural sophistication (e.g., equivariance) is beneficial, the inclusion of high-temperature, non-equilibrium training data is the dominant driver of kinetic accuracy. These findings establish that modern uMLIPs are sufficiently robust to serve as zero-shot surrogates for high-throughput kinetic screening of next-generation energy storage materials.
Förderung
- City University of Hong Kong
- National Natural Science Foundation of China
Zitieren
APA
Guo, X., Gui, C., & Wang, Z. (2026). Benchmarking Universal Machine Learning Interatomic Potentials for Alkali-Ion Battery Kinetics. ACS Materials Letters, 8(7), 1891-1899. https://doi.org/10.1021/acsmaterialslett.6c00134
Vancouver
Guo X, Gui C, Wang Z. Benchmarking Universal Machine Learning Interatomic Potentials for Alkali-Ion Battery Kinetics. ACS Materials Letters. 2026;8(7):1891-1899. doi:10.1021/acsmaterialslett.6c00134
BibTeX
@article{Guo2026,
author = {Guo, Xingyu and Gui, Cheng and Wang, Zhenbin},
title = {Benchmarking Universal Machine Learning Interatomic Potentials for Alkali-Ion Battery Kinetics},
journal = {ACS Materials Letters},
year = {2026},
volume = {8},
number = {7},
pages = {1891--1899},
doi = {10.1021/acsmaterialslett.6c00134}
}
EUVIMED