Reaction prediction via atomistic simulation: from quantum mechanics to machine learning

It is an ultimate goal in chemistry to predict reaction without recourse to experiment. Reaction prediction is not just the reaction rate determination of known reactions but, more broadly, the reaction exploration to identify new reaction routes. This review briefly overviews the theory on chemical reaction and the current methods for computing/estimating reaction rate and exploring reaction space. We particularly focus on the atomistic simulation methods for reaction exploration, which are benefited significantly by recently emerged machine learning potentials. We elaborate the stochastic surface walking global pathway sampling based on the global neural network (SSW-NN) potential, developed in our group since 2013, which can explore complex reactions systems unbiasedly and automatedly. Two examples, molecular reaction and heterogeneous catalytic reactions, are presented to illustrate the current status for reaction prediction using SSW-NN.

Automated summary

Predicting chemical reactions without experimental verification is an ultimate goal in chemistry, involving not only known reaction rate determination but also exploration of new reaction routes. This review discussed the theory of chemical reactions, methods for computing/estimating reaction rates, and exploring reaction spaces, with a focus on atomistic simulation methods and machine learning potentials. The development of the stochastic surface walking global pathway sampling based on global neural network (SSW-NN) potential is highlighted as an unbiased and automated approach for exploring complex reaction systems.