H₂–D₂ Exchange Activity and Electronic Structure of Ag_xPd_1–x Alloy Catalysts Spanning Composition Space

Abstract

Many computational studies of catalytic surface reaction kinetics have demonstrated the existence of linear scaling relationships between physical descriptors of catalysts and reaction barriers on their surfaces. In this work, the relationship between catalyst activity, electronic structure, and alloy composition was investigated experimentally using a Ag x Pd 1−x Composition Spread Alloy Film (CSAF) and a multichannel reactor array that allows measurement of steady-state reaction kinetics at 100 alloy compositions simultaneously. Steady-state H 2 −D 2 exchange kinetics were measured at atmospheric pressure on Ag x Pd 1−x catalysts over a temperature range of 333−593 K and a range of inlet H 2 and D 2 partial pressures. X-ray photoelectron spectroscopy (XPS) was used to characterize the CSAF by determining the local surface compositions and the valence band electronic structure at each composition. The valence band photoemission spectra showed that the average energy of the valence band, ε ̅ v , shifts linearly with composition from −6.2 eV for pure Ag to −3.4 eV for pure Pd. At all reaction conditions, the H 2 −D 2 exchange activity was found to be highest on pure Pd and gradually decreased as the alloy was diluted with Ag until no activity was observed for compositions with x Pd < 0.58. Measured H 2 −D 2 exchange rates across the CSAF were fit using the Dual Subsurface Hydrogen (2H′) mechanism to extract estimates for the activation energy barriers to dissociative adsorption, ΔE ads ‡ , associative desorption, ΔE des ‡ , and the surface-to-subsurface diffusion energy, ΔE ss , as a function of alloy composition, x Pd . The 2H′ mechanism predicts ΔE ads ‡ = 0−10 kJ/mol, ΔE des ‡ = 30−65 kJ/mol, and ΔE ss = 20−30 kJ/mol for all alloy compositions with x Pd ≥ 0.64, including for the pure Pd catalyst (i.e., x Pd = 1). For these Pd-rich catalysts, ΔE des ‡ and ΔE ss appeared to increase by ∼5 kJ/mol with decreasing x Pd . However, due to the coupling of kinetic parameters in the 2H′ mechanism, we are unable to exclude the possibility that the kinetic parameters predicted when x Pd ≥ 0.64 are identical to those predicted for pure Pd. This suggests that H 2 −D 2 exchange occurs only on bulk-like Pd domains, presumably due to the strong interactions between H 2 and Pd. In this case, the decrease in catalytic activity with decreasing x Pd can be explained by a reduction in the availability of surface Pd at high Ag compositions.