Artificial Intelligence‐assisted Heterogeneous Catalyst Design, Discovery, and Synthesis Utilizing Experimental Data

Experimental data are more suitable for developing heterogeneous catalysis in practical applications since the current theoretical and computational approaches still lack the inherent complexity of real-world catalysts. This chapter focuses on the up-to-date artificial intelligence approaches utilized in the literature for heterogeneous catalysis using experimental data. These include catalyst design by machine learning model interpretation and data analysis, catalyst discovery to reach performance goals, and catalyst synthesis to achieve target morphologies via active learning. These approaches are discussed using key studies in the literature, along with advantages, drawbacks, and aspects that need to be considered for their successful implementation.