A Water-Promoted Mars-van Krevelen Reaction Dominates Low-Temperature CO Oxidation over Au-Fe2O3 but Not over Au-TiO2

Abstract

We provide experimental evidence that is inconsistent with often proposed Langmuir-Hinshelwood (LH) mechanistic hypotheses for water-promoted CO oxidation over Au-Fe₂O₃. Passing CO and H₂O, but no O₂, over Au-γ-Fe₂O₃ at 25 °C, we observe significant CO₂ production, inconsistent with LH mechanistic hypotheses. Experiments with H₂¹⁸O further show that previous LH mechanistic proposals cannot account for water-promoted CO oxidation over Au-γ-Fe₂O₃. Guided by density functional theory, we instead postulate a water-promoted Mars-van Krevelen (w-MvK) reaction. Our proposed w-MvK mechanism is consistent both with observed CO₂ production in the absence of O₂ and with CO oxidation in the presence of H₂¹⁸O and ¹⁶O₂. In contrast, for Au-TiO₂, our data is consistent with previous LH mechanistic hypotheses.

Scheme 1. Elementary Reaction Steps for the Postulated (by Chandler et al.) LH Reaction Mechanism of Water-Promoted CO Oxidation over Au-Al₂O₃ and Au-TiO₂, Which Has Also Been Proposed to Be Dominant on Au-Fe₂O₃

Note: * denotes an active site on the Au NP, away from the NP–support interface. ^† denotes a Au site at the NP–support interface, and ^‡ denotes a support site at the NP–support interface. Formal charges that form during the reaction are assumed to be balanced by formal charges distributed over the Au NP.

Scheme 2. Elementary Reaction Steps for a Previously Postulated (by Iglesia et al.) LH Reaction Mechanism of Water-Promoted CO Oxidation over Au-Al₂O₃, Au-TiO₂, and Au-Fe₂O₃

Note: * denotes an active site on the Au NP.

Figure 1

TEM micrographs of (A) Au-TiO₂ and (B) Au-γ-Fe₂O₃. Au NP size histograms for (C) Au-TiO₂ and (D) Au-γ-Fe₂O₃. (E–G) Typical high-resolution TEM images of Au-γ-Fe₂O₃. The Au NPs are polyhedral, with the (111) facet dominating at the Au/γ-Fe₂O₃ interface.

Figure 2

(A) Transient CO oxidation rates over Au-TiO₂. (B) Transient CO oxidation rates over Au-γ-Fe₂O₃. Reaction starts immediately upon introduction of CO and terminates immediately upon removal of CO (at 3 h 45 min). Orange squares: 1 vol % CO, 2.8 vol % H₂O, 20 vol % O₂, balance N₂. Blue circles: 1 vol % CO, 2.8 vol % H₂O, balance N₂. Reaction temperature was 25 °C and pressure 1 atm. The gas hourly space velocity (GHSV) was 21 L h^–1 g^–1_cat, and the CO conversion was below 20% (ensuring data was collected under differential conditions).^, Reported curves are averages of three independent measurements. Error bars are 2 standard deviations wide. For some data points, the error bars are so small, they are obscured by the data labels. Full data sets in Figure S9.

Scheme 3. Elementary Reaction Steps for Our Proposed w-MvK CO Oxidation Mechanism (R15, R17–R21); the (Non-Feasible) Non-Water-Promoted MvK Mechanism Is Represented by R15, R16, R20, and R21

Note: * denotes an active site on the Au NP. O_lat denotes a lattice-oxygen near the Au NP. □_Olat denotes a lattice-oxygen vacancy. OH_ad denotes a hydroxyl on a lattice Fe, and OH_lat denotes a hydroxyl in a lattice-oxygen position.

Figure 3

Schematic representation of our proposed mechanism for w-MvK CO oxidation (R15, R17–R21) over Au-γ-Fe₂O₃. The non-feasible, non-water-promoted, MvK mechanism is represented by R15, R16, R20, and R21.

Figure 4

Energy diagrams of previously proposed LH-mechanisms and of our proposed w-MvK mechanism. The reactions are labeled according to Schemes 1–3. Boxed values (in eV) are activation energies for the respective elementary reaction. Refer to the Supporting Information, Section S2, for computational details.

Figure 5

C¹⁶O oxidation over Au-γ-Fe₂O₃ with H₂¹⁸O and ¹⁶O₂. (A) Transient production rates of different isotopic CO₂ species, and the total rate. (B) Transient relative abundances of different isotopic CO₂ species in the reactor effluent. Relative abundances were calculated (e.g., for C¹⁶O₂) as. The relative abundances predicted from our proposed w-MvK mechanism are indicated by dashed lines. (C) Comparison between predicted (from different mechanistic hypotheses) and experimentally observed relative abundances in the reactor effluent. Reaction conditions: 1 vol % C¹⁶O, 2.8 vol % H₂¹⁸O, 20 vol % ¹⁶O₂, balance N₂. Reaction temperature was 25 °C, pressure 1 atm, and space velocity (GHSV) 10.5 L h^–1 g^–1_cat.