Proceedings of MATSUS Fall 2023 Conference (MATSUSFall23)
DOI: https://doi.org/10.29363/nanoge.matsus.2023.094
Publication date: 18th July 2023
In the future energy scenario, hydrogen is seen as a promising alternative to currently used fossil fuels. Proton exchange membrane (PEM) electrolysis is recognized as the best technology for the sustainable production of hydrogen, however, its widespread utilization is jeopardized by the fact that the catalyst, used for the anodic oxygen evolution reaction, is based on scarce iridium. One of the strategies to reduce its loading in the electrolyzer is the dispersion of Ir-based nanoparticles on ceramic supports, which can withstand harsh conditions in the electrolyzer.
The newly designed materials for electrocatalytic applications, pursuing this goal, are generally first tested on the laboratory scale. The most critical parameters that reveal their viability are their activity, stability, and selectivity. The best metric, indicating the activity of the catalyst is the so-called turnover frequency (TOF), which is in reality difficult to obtain. Instead, normalization of the current by the electrochemically active surface area (ECSA) is used. Unfortunately, there are currently no methodologies for its determination in the case of supported Ir-based catalysts. A promising methodology was recently proposed by Watzele et al. [1], which suggested the use of electrochemical impedance spectroscopy (EIS) to measure the so-called adsorption capacitance of reaction intermediates, which is directly related to the active surface area. In their work, the authors showcased this methodology on thin films.
Here, we applied this methodology to a commercial catalyst Ir on Vulcan (Premetek) to test its suitability for the measurement of ECSA of supported nanostructured catalysts. We prepared electrodes with different loadings of the investigated material and performed a detailed analysis of the impedance spectra. Results showed significant discrepancies between expected and measured values of obtained parameters, which cannot be attributed solely to the experimental errors. Based on the results, we discuss possible reasons for the anomalies and confirm the main hypotheses by impedance simulations using the full physical model ‐ instead of the simplified equivalent circuit.