Small metal nanoparticles are commonly used as materials to accelerate chemical reactions of importance, such as the production and use of fuels and chemicals. However, it is difficult to control the structures of these nanoparticles and how molecules interact with their surfaces. This limits the ability to control and tune their performance. This project will develop computational approaches to examine metal clusters within tiny, ordered pores of aluminosilicate materials called zeolites. The work will combine computer simulations and machine learning to understand how these clusters are formed, and how they change during reaction processes. By changing the relative size of the zeolite pore and metal cluster size, the project will understand how nanometer-scale confinement impacts metal clusters and influences their catalytic reactivity. This will allow the researchers to design materials to perform chemical reactions with lower energy inputs, particularly for reactions that use liquid molecules to store and transport hydrogen. Educational and outreach efforts will train undergraduate and graduate students in new machine learning techniques and engage high school students with direct scientific programs. The public will be engaged through catalysis simulation tools and applications developed to introduce students to model-driven scientific discovery. This project focuses on designing metal nanoclusters encapsulated in nanoporous aluminosilicates where precise control