Steady-state active-site dynamics and mechanism of Au/CeO2 catalysts in room-temperature CO oxidation beyond the activation period

Operando infrared spectroscopy plus DFT identified a single cationic gold site, forming an $O_{\mathrm{lattice}}$–$Au^{+}$–CO carbonyl band (CO bound to a single positively charged gold atom anchored on a ceria lattice oxygen) at $2125$–$2130\ \mathrm{cm}^{-1}$, as the active site of a ceria-supported gold catalyst during room-temperature CO oxidation, but only during the initial activation period: near steady state the direct correlation between this band and CO conversion breaks down, and catalysts with different pretreatment histories converge to a common steady-state behavior through an unknown structural route. Later work confirmed that dynamically formed pseudo-single $Au^{+}$ sites (gold atoms transiently abstracted from clusters by CO) are catalytically active, narrowing the question of which species carries the band, yet the steady-state regime itself remains unexplained: the operative active-site ensemble and mechanism after activation, the structural dynamics of the Au–ceria interface under sustained reaction, the relative importance of true isolated versus pseudo-single $Au^{+}$ sites at steady state, and the mechanism by which different pretreatments converge to identical catalytic behavior are all still open. Resolving this problem would establish how an oxide-supported single-atom/nanoparticle catalyst actually operates at sustained steady state and how its pretreatment history is erased.

Background

Gold dispersed as nanoparticles or single atoms on reducible oxides catalyzes CO oxidation at and below room temperature, a discovery by Haruta and co-workers that made oxide-supported Au a prototypical system of heterogeneous catalysis. Ceria (CeO2) is a reducible support that exchanges lattice oxygen with adsorbates (oxygen vacancies form and are healed at low energy cost), so CO oxidation on Au/CeO2 is commonly described by a Mars–van Krevelen redox picture, in which CO consumes a lattice oxygen and gas-phase O2 later reoxidizes the support. A long-standing controversy concerns the chemical identity of the active gold: metallic nanoparticles, cationic gold at the metal–oxide perimeter, or isolated single Au atoms. Operando DRIFTS (diffuse reflectance infrared Fourier transform spectroscopy performed on the working catalyst, with simultaneous gas-phase conversion measurement) probes this through adsorbed CO as a reporter, since the CO stretch frequency tracks the adsorption site. Schilling, Ziemba, Hess and Ganduglia-Pirovano (J. Catal. 383, 2020, 264–272) combined minute-resolved operando DRIFTS on 0.5 wt% Au/CeO2 at 21 °C with DFT calculations at the PBE+U level (the PBE generalized-gradient density functional amended by a Hubbard-$U$ on-site correction for the localized Ce $4f$ electrons) and the HSE06 hybrid density-functional level, on models of a single Au atom ($Au_1$) and a four-atom Au cluster ($Au_4$) on CeO2(111). They assigned the band at $2125$–$2130\ \mathrm{cm}^{-1}$ to CO on an $O_{\mathrm{lattice}}$–$Au^{+}$–CO species — a single cationic gold atom in contact with ceria lattice oxygen — and showed its intensity rises together with CO conversion during catalyst activation, directly associating the species with activity. A key subtlety emerges from their DFT: CO can abstract a gold atom from an $Au_4$ cluster to form a pseudo-single site, and the CO stretches of both candidate sites are consistent with the $2125$–$2130\ \mathrm{cm}^{-1}$ operando window, so that band alone cannot settle whether the active site is a true isolated $Au^{+}$ or a pseudo-single site abstracted from a cluster — both must be considered active-site candidates (later same-group modulation-excitation DRIFTS work resolved distinct CO–Au contributions near $2122$ and $2133\ \mathrm{cm}^{-1}$ and attributed activity to dynamically formed pseudo-single sites). They also found strong pretreatment dependence: the as-prepared (water-containing) catalyst activates rapidly to about 3% steady conversion, whereas a catalyst outgassed at 200 °C activates over roughly 2 h without reaching steady state within 3 h, with inhibiting adsorbates (CO on ceria at $2160$–$2170\ \mathrm{cm}^{-1}$ and hydrogen carbonates at $1218\ \mathrm{cm}^{-1}$) accumulating; despite different histories, both converge toward a common operation window. Crucially, the paper reports that at later times, as steady state is approached, the conversion–band-intensity correlation is lost, and states that then 'other aspects of the reaction start to gain importance, which will be the focus of future studies'. Later technique development is relevant to this problem: modulation-excitation DRIFTS with phase-sensitive detection (ME-DRIFTS/PSD) periodically switches the feed and separates spectroscopically species that respond to the stimulus (candidates for active participants) from spectators, giving their temporal order in the catalytic cycle.

Problem Statement

Determine how an oxide-supported gold catalyst (Au/CeO2, ~0.5 wt% Au, room-temperature CO oxidation) operates at sustained steady state, beyond the activation period during which the $2125$–$2130\ \mathrm{cm}^{-1}$ $O_{\mathrm{lattice}}$–$Au^{+}$–CO band and CO conversion are correlated (Schilling et al., J. Catal. 383, 2020, 264–272). The objective is a validated structure- and mechanism-resolved picture of the working Au/CeO2 catalyst at steady state: what governs activity once that band–conversion correlation is lost and, per the source, 'other aspects of the reaction start to gain importance', after catalysts with different pretreatment histories (as-prepared/water-containing versus dried or outgassed) have converged to a common steady-state behavior. Facets of this single open goal include the operative active-site ensemble and reaction mechanism in the near-steady-state regime; the structural dynamics of the Au–ceria interface under sustained reaction, including how the true isolated-$Au^{+}$ and pseudo-single-site populations evolve and which sites actually mediate steady-state turnover; and the route by which different pretreatment histories converge to the common steady-state behavior, with possible roles of water/hydroxyls, CO coverage, and inhibiting adsorbates (CO on ceria, hydrogen carbonates). These facets are aspects of one research question rather than separately mandated deliverables, and no particular technique or combination of techniques is prescribed for answering it.

References

1. Schilling, C.; Ziemba, M.; Hess, C.; Ganduglia-Pirovano, M. V. (2020). Identification of single-atom active sites in CO oxidation over oxide-supported Au catalysts. Journal of Catalysis 383, 264–272. DOI: 10.1016/j.jcat.2020.01.022. https://doi.org/10.1016/j.jcat.2020.01.022 (primary source; full text closed access)
2. Weyel, J.; Ziemba, M.; Hess, C. (2022). Elucidating Active CO–Au Species on Au/CeO2(111): A Combined Modulation Excitation DRIFTS and Density Functional Theory Study. Topics in Catalysis 65(7–8), 779–787. DOI: 10.1007/s11244-022-01599-1. https://doi.org/10.1007/s11244-022-01599-1
3. Ziemba, M.; Schilling, C.; Ganduglia-Pirovano, M. V.; Hess, C. (2021). Toward an Atomic-Level Understanding of Ceria-Based Catalysts: When Experiment and Theory Go Hand in Hand. Accounts of Chemical Research 54(13), 2884–2893. DOI: 10.1021/acs.accounts.1c00226. https://doi.org/10.1021/acs.accounts.1c00226
4. Su, Y.-Q.; Qin, Y.-Y.; Wu, T.; Wu, D.-Y. (2022). Structure sensitivity of ceria-supported Au catalysts for CO oxidation. Journal of Catalysis 407, 353–363. DOI: 10.1016/j.jcat.2022.02.008. https://doi.org/10.1016/j.jcat.2022.02.008
5. Ziemba, M.; Hess, C. (2020). Influence of gold on the reactivity behaviour of ceria nanorods in CO oxidation: combining operando spectroscopies and DFT calculations. Catalysis Science & Technology 10(11), 3720–3730. DOI: 10.1039/d0cy00392a. https://doi.org/10.1039/d0cy00392a
6. Fiuza, T. E. R.; Gonçalves, D. S.; Gomes, I. F.; Zanchet, D. (2021). CeO2-supported Au and AuCu catalysts for CO oxidation: Impact of activation protocol and residual chlorine on the active sites. Catalysis Today 381, 171–180. DOI: 10.1016/j.cattod.2020.07.034. https://doi.org/10.1016/j.cattod.2020.07.034
7. Abdel-Mageed, A. M.; Chen, S.; Fauth, C.; Häring, T.; Bansmann, J. (2021). Fundamental Aspects of Ceria Supported Au Catalysts Probed by In Situ/Operando Spectroscopy and TAP Reactor Studies. ChemPhysChem 22(13), 1302–1315. DOI: 10.1002/cphc.202100027. https://doi.org/10.1002/cphc.202100027
8. Li, Y.; Li, S.; Bäumer, M.; Moskaleva, L. V. (2021). Transient Au–CO Complexes Promote the Activity of an Inverse Ceria/Gold Catalyst: An Insight from Ab Initio Molecular Dynamics. The Journal of Physical Chemistry C 125(48), 26406–26417. DOI: 10.1021/acs.jpcc.1c07040. https://doi.org/10.1021/acs.jpcc.1c07040
9. Liu, R.; Qu, W.; Hu, X.; Chen, J.; Dong, Y.; Xu, D.; Liu, J.; Ma, Z.; Tang, X. (2022). Valence states of single Au atoms dictate the catalytic activity of Au1/CeO2(100). Chemical Communications 58(82), 11587–11590. DOI: 10.1039/d2cc04219c. https://doi.org/10.1039/d2cc04219c
10. Huang, L.; Xu, Q.; Tu, Y.; Cheng, X.; Zhang, D.; Hu, J.; Ding, H.; Zhu, J. (2025). O2 Activation on Au/CeOX(111) Model Catalysts and Its Role in CO Oxidation. ChemCatChem 17(20), e01024. DOI: 10.1002/cctc.202501024. https://doi.org/10.1002/cctc.202501024
11. Weyel, J.; Schumacher, L.; Ziemba, M.; Pfeiffer, M.; Hess, C. (2024). Modulation Excitation Spectroscopy: A Powerful Tool to Elucidate Active Species and Sites in Catalytic Reactions. Accounts of Chemical Research 57(18), 2643–2652. DOI: 10.1021/acs.accounts.4c00351. https://doi.org/10.1021/acs.accounts.4c00351
12. Tang, Y.; Ma, X.; Du, X.; Liu, X.; Chen, R.; Shan, B. (2023). Breaking the activity–stability trade-off of Au catalysts by depth-controlled TiO2 nanotraps. Journal of Catalysis 423, 145–153. DOI: 10.1016/j.jcat.2023.05.002. https://doi.org/10.1016/j.jcat.2023.05.002
13. Zhou, T.; Zhang, N.; Xiao, Z.; Zhang, W.; Li, Z.; Zhang, Y.; Lei, H.; Wu, W.; Li, H.; Yan, H.; Zeng, J. (2026). Advanced Single-Atom Catalysts for Thermal-Catalytic C1 Chemistry. Chemical Reviews 126(14), 8106–8201. DOI: 10.1021/acs.chemrev.6c00203. https://doi.org/10.1021/acs.chemrev.6c00203