CHEMISTRY AND MATERIAL SCIENCE

Bridging accuracy and efficiency: assessing universal ML potentials for niobiumoxygen clusters

I. S. Popov, A. A. Valeeva, A. N. Enyashin

Institute of Solid State Chemistry, Ural Branch of Russian Academy of Sciences, Yekaterinburg, Russia

Abstract: Machine Learning Interatomic Potentials (MLIPs) promise to combine the accuracy of DFT with the speed of classical force fields. However, their reliability for complex, multi-component systems requires rigorous validation. Here, we perform a targeted evaluation of three leading universal MLIPs using niobium oxide clusters (Nb$_n$O$_m$, $n\le 6$, $m\le 6$) as a challenging test case. The Nb–O system is very well suited for this task due to its complex electronic interactions, manifested in existence of the bulk phase with 25% vacancy-ordered lattice and, at nanoscale, by a diverse range of non-stoichiometric clusters. We employ a dataset of global minima structures identified via DFT-based evolutionary search as a strict reference. A comparative analysis is then performed by executing evolutionary searches with the MLIPs. By directly comparing predicted structures, energies, and relative stability, we provide a comprehensive assessment of the accuracy and limitations of current universal potentials for modeling complex nanoscale oxides.

Keywords: DFT, Machine learning potential, MLIP, Evolutionary algorithm, Niobium oxide, NbO, Nanoparticle, Cluster, USPEX.

Received: 09.10.2025
Accepted: 15.10.2025

Language: English

DOI: 10.17586/2220-8054-2025-16-5-619-627