Reliable and Efficient Automated Transition-State Searches with Machine-Learned Interatomic Potentials

TL;DR

This paper demonstrates that machine-learned interatomic potentials can reliably and efficiently perform transition-state searches, significantly reducing computational costs while maintaining high accuracy across diverse chemical reactions.

Contribution

The study systematically benchmarks MLIP-based workflows with reaction-path algorithms, showing their effectiveness and transferability for automated transition-state searches in various chemical systems.

Findings

MACE-OMol25 achieves 96.6% success rate with fewer than four DFT-gradient evaluations.

MLIP workflows reduce computational cost by 94-96% compared to traditional DFT methods.

Low-level refinement on MLIP surfaces enables three-fold cost reduction with minimal reliability loss.

Abstract

Transition-state searches are central to understanding reaction mechanisms, but the high computational cost of density-functional theory (DFT) limits their application in high-throughput catalyst and materials discovery. Machine-learned interatomic potentials (MLIPs) offer near-DFT accuracy at orders-of-magnitude lower cost, yet their reliability for transition-state searches remains underexplored. Here, we systematically benchmark hybrid transition-state-search workflows combining six freely available potentials (MACE-OMol25, UMA-Small, UMA-Medium, eSEN-S, AIMNet2, and GFN2-xTB) with two reaction-path-finding algorithms (the freezing-string method and climbing-image nudged elastic band) across 58 diverse reactions spanning small organics, polymerization chemistry, and transition-metal catalysis. We find that models trained on the Open Molecules 2025 dataset exhibit markedly superior performance, with MACE-OMol25 achieving a 96.6% success rate while requiring fewer than four DFT-gradient evaluations per reaction on organic systems - a 94-96% reduction compared to conventional DFT-based searches. Low-level refinement on the MLIP surface before high-level DFT optimization reduces computational cost three-fold with minimal loss in reliability. For transition-metal systems, UMA-Medium demonstrates promising transferability to in-distribution transition metal complex reactions and out-of-distribution organometallic C-H activation. These results establish MLIP-accelerated workflows as practical tools for automated reaction discovery, enabling near-DFT accuracy at a fraction of traditional expense.