The extended commercialization of electrochemical energy conversion devices (i.e., fuel cells and (co-)electrolyzers) relies on the development of improved catalysts for the reactions taking place in their electrodes (i.e., O₂-reduction/evolution, H₂-oxidation/evolution or CO₂-reduction). This in turn passes by deepening our understanding of these reactions’ mechanisms and of the parameters that determine the performance and stability of the electrocatalysts speeding up their kinetics. As a result, great efforts are being devoted to investigate these reactions and materials under operation-relevant conditions. Such in situ / operando studies are often carried out using X-ray absorption spectrocospy (XAS), which is a bulk-sensitive technique realizable at atmospheric pressure and that can provide insight on an analyte’s oxidation state, electronic properties and coordination environment. Moreover, recent developments in fast X-ray energy scanning and spectral acquisition have resulted in a surge of interest in time-resolved XAS studies that can yield additional information on the changes undergone by the catalysts in the course of the reaction [1]. However, the large majority of these works are performed using electrochemical cells and/or experimental conditions (e.g., convective properties, catalyst loadings) optimized to assure the successful completion of the spectroscopic measurements, but that may significantly differ from those encountered by these materials in the real devices. Most importantly, little is known about how these differences can affect the results derived from such spectroelectrochemical studies, as well as on the extent to which their conclusions are amenable to application-relevant operative conditions.

With this motivation, this contribution will start by discussing how XAS-results may be affected by the thickness of the studied catalyst layers (CLs). This was achieved by studying the formation of palladium hydride on electrodes with similar Pd-loadings but ≈ 5-fold different CL-thicknesses, which were prepared using an unsupported Pd-aerogel vs. a carbon-supported Pd-nanoparticle catalyst [2]. In a subsequent step, we will show how this CL-thickness and the intensity of the incident X-ray beam can caused the accumulation of evolved O₂-bubbles within the pores of an Ir-oxide-based CL, leading to a localized loss of potential control that can ultimately result in an erroneous assignment of the catalyst’s oxidation state [3]. Finally, we will present our recent efforts at developing a novel electrochemical cell allowing operando spectroscopic studies with ultra-low catalyst loadings, enhanced mass transport properties and the online analysis of the reaction products.

In summary, this contribution will provide novel insight on the impact of various experimental parameters on operando XAS measurements, while providing guidelines for the successful completion of such experiments and the subsequent interpretation of their results.