Fewest-Switches Surface Hopping with Combined Deep Learning Potential and Long Short-Term Memory Network Propagator for Simulating Realistic Photochemical Processes

TL;DR

This paper introduces an extended LSTM-FSSH method that combines deep learning and neural networks to efficiently simulate realistic photochemical reactions, accurately predicting excited-state lifetimes and product yields with minimal training data.

Contribution

The work develops an advanced LSTM-FSSH framework with redesigned input features and equivariant neural networks for potential energy surfaces, enabling efficient and accurate photochemical simulations.

Findings

Accurately reproduces excited-state lifetimes and product yields.

Requires only 10 reference trajectories for training.

Efficiently simulates complex photochemical reactions.

Abstract

Fewest-switches surface hopping (FSSH) is the most popular method for simulating photochemical processes of molecular systems. Recently, we have constructed long short-term memory (LSTM) networks as a propagator for electronic subsystems in FSSH dynamics simulations. The collective results on Tully's three models have been reproduced satisfactorily. In the present work, we develop an extended LSTM-FSSH framework to simulate realistic photochemical reactions. The input features of LSTM as well as the training procedure are redesigned to represent high-dimensional nuclear degrees of freedom in an effective way. Equivariant neural networks are integrated with LSTM to build adiabatic potential energy surfaces in ground and excited states. Photoisomerizations of $\mathrm{CH_2NH}$ and azobenzene are simulated, showing that our new proposed LSTM-FSSH method can produce excited-state lifetimes and product yields accurately in comparison with conventional FSSH simulations as reference. Only 10 reference trajectories are required for training LSTM networks, and then a trajectory ensemble can be generated with very efficient LSTM-FSSH dynamics simulations to obtain collective results.