Iridium Oxide Coordinatively Unsaturated Active Sites Govern the Electrocatalytic Oxidation of Water

Abstract

A special membrane electrode assembly to measure operando X‐ray absorption spectra and resonant photoemission spectra of mesoporous templated iridium oxide films is used. These films are calcined to different temperatures to mediate the catalyst activity. By combining operando resonant photoemission measurements of different films with ab initio simulations these are able to unambiguously distinguish µ2‐O (bridging oxygen) and µ1‐O (terminal oxygen) in the near‐surface regions of the catalysts. The intrinsic activity of iridium oxide scales with the formation of µ1‐O (terminal oxygen) is found. Importantly, it is shown that the peroxo species do not accumulate under reaction conditions. Rather, the formation of µ1‐O species, which are active in O−O bond formation during the OER, is the most oxidized oxygen species observed, which is consistent with an O−O rate‐limiting step. Thus, the oxygen species taking part in the electrochemical oxidation of water on iridium electrodes are more involved and complex than previously stated. This result highlights the importance of employing theory together with true and complementary operando measurements capable of probing different aspects of catalysts surfaces during operation.

Toward Realistic Models of the Electrocatalytic Oxygen Evolution Reaction

“…While most CHE studies have restricted allowed surface phases to *O and *OH, ,, a similar claim was made for rutile-type RuO 2 , , though operando XAS suggests *OO(H) does not accumulate on RuO 2 . And while there is then debate about the experimental evidence for *OO(H) on RuO 2 under OER conditions, adsorbed superoxo (or peroxo) are not seen on IrO 2 either indirectly at the Ir L 3 -edge or directly through resonant photoemission at the O K-edge . This latter observation places an upper limit of ≪1/10 ML for OO(H) species on Ir at potentials as high as 2 V vs RHE.…”

“…401 And while there is then debate about the experimental evidence for *OO(H) on RuO 2 under OER conditions, adsorbed superoxo (or peroxo) are not seen on IrO 2 either indirectly at the Ir L 3 -edge 224 or directly through resonant photoemission at the O K-edge. 402 This latter observation places an upper limit of ≪1/10 ML for OO(H) species on Ir at potentials as high as 2 V vs RHE. The authors of ref 127 point out that O-O overbinding in DFT may artificially stabilize the superoxo species, resulting in an incorrect prediction, as noted previously.…”