Electrocatalysis of the Oxygen Evolution Reaction: A Comparative Study of Anodically Formed and Nanostructured Iridium Oxides
MetadataVis full innførsel
The thesis deals with electrocatalysts for the oxygen evolution reaction for proton-exchange membrane (PEM) water electrolysis. Water electrolysis is considered the only viable option for large-scale hydrogen production. Hydrogen is a relevant storage medium for renewable energy technologies. The oxygen-evolution reaction (OER) is a major source of loss in PEM water electrolysis, and currently costly catalysts such as iridium oxide, an oxide with a rutile structure, are employed. Optimum use of the catalyst as well as development of alternative ones are considered to be preconditioned by establishing a rational fundamental understanding of these catalysts. The thesis investigates two different forms of iridium oxide, one nanostructured form synthesised through colloidal techniques and passive films formed on iridium metal (AIROF), and compares their electronic structure and electrocatalytic properties. The investigations take departure in the reported ability of AIROF to switch from a metallic behaviour at positive potentials (oxidised state, above approximately 0:8 V versus a reversible hydrogen electrode) to a semiconducting form at low potentials (reduced state). For AIROF in a solution of Fe(CN)6 -3/-4 and sulphuric acid the reduction currents associated with the redox couple are thus much smaller than the oxidation currents, clearly related to a poor alignment of the oxidised part of the redox couple with the valence band of the reduced oxide. The electrochemical characteristics of oxidation and reduction of hexacyanoferrate redox couples at nanostructured iridium oxide produced by hydrolysis (hIrO2) are, however, similar to a metal electrode. In view of also significant differences between the two oxides’ optical and intercalation properties, this strongly indicates that hIrO2 has a different bulk electronic structure. Yet, the charge-normalised electrocatalytic activity for the oxygen-evolution reaction at AIROF and hIrO2 are the same. This indicates that the electrocatalytic properties of iridium oxide are a function of the local structure at the oxide surface and not uniquely related to the bulk electronic structure. Some of the measurements of the nanostructured oxide were performed at Nafion® -containing composite polymer-oxide electrodes, and for interpretation of these it is important to assess the possible influences of the Nafion® electrolyte on the results. Thus, in a solution of the same redox couple and TFMSA the AIROF electrode shows no cathodic ”blocking” of the faradaic current and behaves similarly to a metal electrode. In ferrate-free solutions Mott-Schottky plots revealed no significant changes in the flatband potential in TFMSA as compared to H2SO4. The results still indicate stronger sulphate adsorption since higher rates for the oxygen evolution reaction and formation of insoluble ferrate-decomposition products in solutions containing TFMSA than in H2SO4. The absence of cathodic ”blocking” in TFMSA is suggested to be due to exposure of free iridium substrate, of which the area is estimated to 1% of the electrode area, an inferences about the electronic structure of such oxides should pay due attention to the effects of the supporting electrolyte. However, in terms of electronic structure, the conclusion that the electronic structures of the two oxides are different remains. In view of these results, we also attempted to investigate possible conductivity changes in the nanostructured oxide through impedance analysis. To interpret the results, a mathematical model for the impedance of porous intercalation electrodes of mixed conductivity was employed, part of which was derived within the current work. The model is based on an agglomerate model and dilute-solution theory. The model was used to interpret impedance data collected for the nanostructured iridium oxide (NIRO) at potentials below those at which the oxygen evolution reaction commences. The measurements included thin oxide films covered by a protective NafionR layer and thicker composite Nafion® -oxide electrodes. The time constants for the low frequency diffusion process were approximately the same for both types of electrodes, indicating diffusion in individual particles in the porous electrode rather than across the film. The impedance data indicated that there were no significant variations in conductivity of the oxides with potential, as opposed to what appears to be the case for AIROF. This is interpreted to reflect differences in electronic structure between the nanostructured oxide and AIROF. In conclusion we find strong evidence that the (bulk) electronic structure of the nanostructured iridium oxide is quite different from that of AIROF. In view of the normalised electrocatalytic activity of the two oxides being approximately the same, this restricts the use of correlations based on bulk electronic structure and electrocatalytic activity.