Abstract:Machine learning is employed at an increasing rate in the research field of quantum chemistry. While the majority of approaches target the investigation of chemical systems in their electronic ground state, the inclusion of light into the processes leads to electronically excited states and gives rise to several new challenges. Here, we survey recent advances for excited-state dynamics based on machine learning. In doing so, we highlight successes, pitfalls, challenges and future avenues for machine learning approaches for light-induced molecular processes.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to use machine - learning methods to accelerate the simulation of excited - state dynamics in light - induced molecular processes. Specifically, traditional quantum - chemical calculations face challenges such as high computational costs, limited simulation time, and limited ability to handle complex molecular systems when dealing with excited - state dynamics. These problems limit our in - depth understanding and predictive ability of the mechanism of photochemical reactions. The paper mentions that although experimental techniques can provide some macroscopic information on light - induced reactions, they are often "blind" to the details of the electronic mechanism. Theoretical simulations can supplement experimental findings and provide a more comprehensive understanding, but these simulations require high - performance computing facilities and are particularly dependent on expensive quantum - chemical calculations. Therefore, the focus of the paper is to explore how to improve the efficiency and accuracy of these simulations through machine - learning methods, especially for the learning and prediction of key properties such as excited - state potential energy surfaces (PESs), corresponding forces and non - adiabatic couplings (NACs), and spin - orbit couplings (SOCS) in non - adiabatic molecular dynamics (NAMD) simulations. The authors discuss methods for generating training sets, including how to handle the special properties of excited - state data, such as the arbitrary - phase problem of wave functions, and how to select appropriate reference methods for quantum - chemical calculations. In addition, an adaptive sampling technique is introduced to dynamically expand the training set to cover all important conformational regions that the molecule may visit, thereby improving the predictive ability and reliability of the machine - learning model. In summary, the paper aims to promote the application of machine - learning in the field of photochemistry, especially in improving the efficiency and accuracy of excited - state dynamics simulations.

Machine learning and excited-state molecular dynamics

Machine Learning for Electronically Excited States of Molecules

Molecular excited states through a machine learning lens

Machine learning enables long time scale molecular photodynamics simulations

Machine learning for molecular simulation

Quantum Machine Learning for Chemistry and Physics

Machine Learning in Molecular Dynamics Simulations of Biomolecular Systems

Molecular Dynamics with On-the-Fly Machine Learning of Quantum-Mechanical Forces

Electronic excited states from physically-constrained machine learning

Machine Learning in QM/MM Molecular Dynamics Simulations of Condensed-Phase Systems

Machine learning for the solution of the Schrödinger equation

Machine learning accelerated photodynamics simulations

Electrostatic Embedding Machine Learning for Ground and Excited State Molecular Dynamics of Solvated Molecules

Machine Learning for Quantum Dynamics: Deep Learning of Excitation Energy Transfer Properties

Machine Learning for Quantum Mechanical Properties of Atoms in Molecules

Machine learning for molecular and materials science

Unifying machine learning and quantum chemistry -- a deep neural network for molecular wavefunctions

Molecular Dynamics with Neural-Network Potentials

Unifying machine learning and quantum chemistry with a deep neural network for molecular wavefunctions

Machine Learning for Molecular Dynamics on Long Timescales

Electronic Excited States from Physically Constrained Machine Learning