Analytical ab initio hessian from a deep learning potential for transition state optimization
Identifying transition states-saddle points on the potential energy surface connecting reactant and product minima-is central to predicting kinetic barriers and understanding chemical reaction mechanisms. In this work, we train a fully differentiable equivariant neural network potential, NewtonNet, on thousands of organic reactions and derive the analytical Hessians. By reducing the computational cost by several orders of magnitude relative to the density functional theory (DFT) ab initio source, we can afford to use the learned Hessians at every step for the saddle point optimizations. We show that the full machine learned (ML) Hessian robustly finds the transition states of 240 unseen organic reactions, even when the quality of the initial guess structures are degraded, while reducing the number of optimization steps to convergence by 2-3× compared to the quasi-Newton DFT and ML methods. All data generation, NewtonNet model, and ML transition state finding methods are available in an automated workflow.
Authors
- Judit Zádor (ORCID: https://orcid.org/0000-0002-9123-8238)
- Xingyi Guan (ORCID: https://orcid.org/0000-0001-7623-4012)
- Samuel M. Blau (ORCID: https://orcid.org/0000-0003-3132-3032)
- Eric Hermes (ORCID: https://orcid.org/0000-0001-6699-7948)
- Teresa L. Head-Gordon (ORCID: https://orcid.org/0000-0003-0025-8987)
- Anup Kumar (ORCID: https://orcid.org/0000-0002-2068-4695)
- Andrew Rosen (ORCID: https://orcid.org/0000-0002-0141-7006)
- Eric Chung‐Yueh Yuan (ORCID: https://orcid.org/0000-0002-5544-8107)
Institutions
- Lawrence Berkeley National Laboratory (US)
- Sandia National Laboratories California (US)
- University of California, Berkeley (US)
Publication Details
- Journal
- Nature Communications
- Published
- 2024-10-14
- DOI
- https://doi.org/10.1038/s41467-024-52481-5
- Citations
- 22
- Primary Topic
- Machine Learning in Materials Science
- Type
- article
- Field-Weighted Citation Impact
- 2.58
Funders
- U.S. Department of Energy
- Adolph C. and Mary Sprague Miller Institute for Basic Research in Science, University of California Berkeley
- National Energy Research Scientific Computing Center
- Office of Science
- Basic Energy Sciences
- Chemical Sciences, Geosciences, and Biosciences Division
- Lawrence Berkeley National Laboratory