Automation techniques for exploring potential energy surfaces of chemical reaction with quantum chemistry
Due to the large chemical space in which realistic multi-step chemical transformations take place, the exhaustive exploration of the region of this space that is relevant for the full description of this transformation, including locating all relevant intermediates and transition states, is a very hard task to perform. Not only the computational time needed is large, but also the amount of user time needed to set up and analyze calculations for each of the species.Therefore, it becomes essential to look for new methodologies that allow the mapping of stationary points on the potential energy surface in an automatic way. The computational efficiency development along the last years allows the grow of data management, with the consequent arise of new and faster algorithms that let us tackling the described problem. In this research we will focus on the automation of reaction mechanisms using tools such as graph theory, machine learning, etc. The goal of our research will be to apply these new techniques to realistic chemical reactions of relevance in the chemical industry.