SpaiNN: equivariant message passing for excited-state nonadiabatic molecular dynamics

Mausenberger S, Müller C, Tkatchenko A, Marquetand P, González L, Westermayr J (2024)

Publication Type: Journal article

Publication year: 2024

Journal

Chemical Science Royal Society of Chemistry

Book Volume: 15

Pages Range: 15880-15890

Journal Issue: 38

DOI: 10.1039/d4sc04164j

Abstract

Excited-state molecular dynamics simulations are crucial for understanding processes like photosynthesis, vision, and radiation damage. However, the computational complexity of quantum chemical calculations restricts their scope. Machine learning offers a solution by delivering high-accuracy properties at lower computational costs. We present SpaiNN, an open-source Python software for ML-driven surface hopping nonadiabatic molecular dynamics simulations. SpaiNN combines the invariant and equivariant neural network architectures of SchNetPack with SHARC for surface hopping dynamics. Its modular design allows users to implement and adapt modules easily. We compare rotationally-invariant and equivariant representations in fitting potential energy surfaces of multiple electronic states and properties arising from the interaction of two electronic states. Simulations of the methyleneimmonium cation and various alkenes demonstrate the superior performance of equivariant SpaiNN models, improving accuracy, generalization, and efficiency in both training and inference.

Authors with CRIS profile

Carolin Müller Juniorprofessur für die Theorie elektronisch angeregter Zustände

Involved external institutions

Universität Wien / University of Vienna AT Austria (AT) University of Luxembourg LU Luxembourg (LU) Universität Leipzig DE Germany (DE)

How to cite

APA:

Mausenberger, S., Müller, C., Tkatchenko, A., Marquetand, P., González, L., & Westermayr, J. (2024). SpaiNN: equivariant message passing for excited-state nonadiabatic molecular dynamics. Chemical Science, 15(38), 15880-15890. https://doi.org/10.1039/d4sc04164j

MLA:

Mausenberger, Sascha, et al. "SpaiNN: equivariant message passing for excited-state nonadiabatic molecular dynamics." Chemical Science 15.38 (2024): 15880-15890.

BibTeX: Download