|
Daniil
N. Chistikov, Vladimir E. Bochenkov, Denis A. Firsov, Vadim V. Korolev, Anastasia V. Bochenkova
Application
of neural networks for investigating potential energy surfaces of
electronically excited states and internal conversion mechanisms of organic
molecules
Abstract
Abstract. Based on
non-empirical data obtained using a high-level multi-configuration quantum
chemical method, neural networks with the architectures of a multilayer
perceptron and an E(3)-equivariant graph network were constructed and trained
to predict the energies of the ground and the first two electronically excited
states of the methylenimmonium cation, CH2NH2+.
It is shown that the E(3)-equivariant graph neural network architecture
demonstrates higher accuracy. Using the trained network, a segment of the
cation’s potential energy surfaces near the region of the conical intersection
between the first excited and the ground states was investigated; this region
plays an important role in the mechanism of internal conversion and
photoisomerization reactions. It is demonstrated that the neural network
accurately reproduces the topography of the potential energy surfaces of the
two electronic states in the region of their conical intersection.
Key words:machine
learning, neural networks, internal conversion, conical intersections,
electronically excited states, quantum chemistry, photochemistry,
photoisomerization, methylenimmonium cation.
Copyright (C) Chemistry Dept., Moscow State University, 2002
|
|
