Abstract: Metal–Organic Frameworks (MOFs) are highly versatile materials for gas separation, storage, sensing, and catalysis. Their modular design allows almost unlimited structural variation, but finding the best material for a specific use remains a significant challenge. The design options are immense, objectives can conflict, and even ideal structures in theory can be hard to produce. Artificial intelligence is now changing how we handle this complexity. By combining large experimental and computational databases, AI helps us explore chemical questions and design spaces far beyond traditional methods. In this talk, I will explain how AI can uncover the hidden structure–property relationships that determine high-performance materials, especially for carbon capture. Instead of replacing human intuition, AI enhances it—helping us discover more targeted, efficient, and innovative materials.

Bio: Berend Smit is Professor of Chemical Engineering at EPFL, where he leads the Laboratory of Molecular Simulation. His research focuses on developing and applying molecular simulation techniques to address major challenges in energy and sustainability, including carbon capture, gas separation, and catalysis. After earning degrees in Chemical Engineering and Physics from Delft University of Technology and a PhD in Chemistry from Utrecht University, he worked at Shell Research. He later became Professor of Computational Chemistry at the University of Amsterdam. He subsequently held joint appointments at UC Berkeley and Lawrence Berkeley National Laboratory before joining EPFL in 2014.

Smit is co-author of Understanding Molecular Simulations, a seminal text that has shaped the field, and Introduction to Carbon Capture and Sequestration. His group combines molecular modeling, machine learning, and process design to create digital twins for materials discovery, thereby accelerating the transition to sustainable technologies.