A Machine Learning Based Approach to Accelerate Catalyst Discovery

Computational catalysis, in contrast to experimental catalysis, uses approximations such as density functional theory (DFT) to compute properties of reaction intermediates. But DFT calculations for a large number of surface species on variety of active site models are resource intensive. In this work, we are building a machine learning based predictive framework for adsorption energies of intermediate species, which can reduce the computational overhead significantly. Our work includes the study and development of appropriate machine learning models and effective fingerprints or descriptors to predict energies accurately for different scenarios. Furthermore, Bayesian inverse problem, that integrates experimental catalysis with its computational counterpart, uses Markov chain Monte Carlo (MCMC) methods to refine the uncertainties on the quantities-of-interest such as turnover frequency. However, large number of forward simulations required by MCMC can become a bottleneck, especially in computational catalysis, where the evaluation of likelihood functions involves finding the solution to microkinetic models. A novel and faster MCMC method is proposed to reduce the number of expensive target evaluations and to shorten the burn-in period by emulating the target along with using a better informed proposal distribution