The use of atomically dispersed metals on oxide supports as catalysts has become increasingly appealing as the world moves toward limiting the use of precious metal resources, such as Rh or Pt, in a wide variety of applications. However, the dynamic chemical environment a catalyst is exposed to, for example during pre-treatment or in a catalytic converter, may alter the intentional structure synthesized at an atomic scale. Although the idea that atomically dispersed metals may change the local coordination to a support surface under different conditions may seem chemically intuitive, details of the link between the local coordination of the metal atom (or ion) and its ability to perform as a catalyst are not understood.

Our group, Phillip Christopher’s group at UCSB, has been working on synthesizing and characterizing atomically dispersed Pt and Rh catalysts on oxide supports. We often apply CO probe molecule infrared (IR) spectroscopy, where the C-O stretching vibration serves as a tool to tell us about the electronic and physical structure of the metal to which it is bound. Interestingly, we have found that the CO IR signatures of atomically dispersed Rh catalysts are sensitive to the environment the catalyst has previously been exposed to. Fortunately, Phillipe Sautet’s group at UCLA, approached us to understand how environmental conditions influence the bonding of Rh to an oxide support by comparing our results to their models using density functional theory (DFT) of single atom Rh on TiO₂ under different chemical potentials. His visiting student, Yan Tang, found that the most stable location for Rh on (or in) TiO₂ was quite different when the catalyst was exposed to O₂ or H₂ (typical of catalyst pretreatment), CO (typical of characterization), and the combination of H₂ and CO₂ (typical of the reverse water-gas shift reaction conditions). This finding makes the important point that understanding the reactivity of atomically dispersed metal catalysts requires an understanding of the atomic scale structure under reaction conditions.

This revelation is likely applicable to many catalytic systems, not only this specific catalyst and reaction. Our collaboration has been able to take experimental characterization and reactivity results to the next level by comparing detailed DFT calculations of the catalyst in different environments to provide an explanation of what the catalyst looks like while it is actually operating. While these studies are challenging, only through detailed understanding of what the catalyst structure looks like as it operates can we hope to understand its catalytic behavior. As I advance in my research to study the role of atomically dispersed Rh active sites in catalytic converter chemistry, I must be cognizant of these effects and that the structure I look at during characterization may very well be different than the one present during the reaction. I will trace how the structure changes due to the exposed conditions and connect the structure of interest in reactive conditions to my conclusions. I hope that the article, "Rh single atoms on TiO₂ dynamically respond to reaction conditions by adapting their site" also helps other scientists to approach characterization techniques with thought-provoking enquiries before assigning results.

https://www.nature.com/articles/s41467-019-12461-6