In new paper, Mavrikakis shares a pathway to more accurate models of catalysis

In an invited review article for the journal Nature Chemical Engineering posted on February 17, 2025, Manos Mavrikakis, the Ernest Micek Distinguished Chair, James A. Dumesic Professor, and Vilas Distinguished Achievement Professor in the Department of Chemical and Biological Engineering at the University of Wisconsin-Madison, discusses new modeling methods for understanding the dynamic evolution of active sites generated by reactants and/or intermediates on catalyst surfaces under realistic reaction conditions, including temperature and pressure.

Catalysis is the process of adding a substance to facilitate a chemical reaction, often by reducing the temperature, pressure or overall energy needed for chemical transformations to take place. Typically, this means that molecules involved in a reaction adsorb, or temporarily stick, to the surface of the solid catalyst, at atomic-scale ensembles called “active sites” where the reactions take place before the final products desorb, or are released from the catalyst.

This model of catalysis, however, is too simple; in fact, per experimental evidence, even at low pressures and temperatures, the adsorbed species can dramatically affect the structure and reactivity of the solid catalysts themselves, altering the atomic-scale identity of the active sites and the dynamics of the entire process. Understanding exactly how the new active sites are formed under reaction conditions and how this change in catalyst structure impacts reactions, has proved challenging.

That’s why in their paper, Mavrikakis and his co-author, Benjamin W.J. Chen of the Institute of High Performance Computing at the Agency for Science, Technology, and Research in Singapore, evaluate the current state of computationally modeling these complex reactive environments. They also share a vision of how cutting-edge computation, artificial intelligence, informatics and new modeling frameworks can combine with experimentation to unravel the dynamic interplay between adsorbates and catalysts. This type of modeling could then be used in the rational design of new classes of catalysts that could be implemented for accelerating a wide range of thermally catalyzed and electrocatalyzed reactions.