CHBE 565-International Paper Co Seminar-Prof. Hongliang Xin, Virginia Tech (Host: Prof. Alex Mironenko) "The New Alchemy of Scientific Discovery: From Electrons to AI Agents"

Abstract: Modern science is entering a new form of alchemy, not mystical transmutation but the transformation of knowledge into discovery. In this talk, I will use electrocatalysis for sustainable fuels and chemicals as a lens to examine how domain theory, high-performance computing, and experimentation can be integrated with machine learning and AI for scientific discovery. The discussion begins at electrified interfaces, where site structures, microenvironment, and solvent dynamics determine catalytic outcomes. I will describe how atomistic simulations, electrokinetic modeling, and large-scale machine-learned interatomic potentials can help us understand interfacial charge-transfer reactions, and how theory-infused deep learning can translate them into interpretable design principles beyond conventional methods. Examples will include sustainable carbon, nitrogen, oxygen, and hydrogen chemistries, such as nitrate reduction to ammonia. I will then discuss agentic AI workflows in which AI agents review literature, generate hypothesis, invoke tools to validate or invalidate, and refine the cycle semi-autonomously with human in the loop. By connecting agentic reasoning with fundamental principles, this emerging approach offers a path to discover catalytic systems with speed, scientific depth, and sustainability.

Bio Hongliang Xin is Professor of Chemical Engineering at Virginia Tech. Xin received his Ph.D. in Chemical Engineering from the University of Michigan and completed postdoctoral research at Stanford/SLAC. He is a recipient of the 2019 NSF CAREER Award. In 2026, he received the ACS Catalysis Lectureship for the advancement of catalytic science in heterogeneous catalysis. He serves on the Editorial Board of Chem Catalysis. He is Communications Director of the North American Catalysis Society (NACS). He served as the inaugural co-chair for the 2026 Gordon Research Conference on AI for Materials, Energy, and Chemical Sciences (AIMECS GRC). He chaired the AI for Multidisciplinary Exploration and Discovery (AIMEDx) Workshop in 2024 and has been leading the community efforts to shape transparent reporting for agentic catalysis enabled by artificial intelligence (TRACE-AI).