(Invited) Performance Loss in PEM-WE – Assessing the Applicability of the Tafel Equation

Polarization curves in PEM water electrolyzers are typically considered to consist of contributions from electrochemical equilibrium, reaction kinetics, and transport phenomena for both charge and mass. However, breaking down these contributions can be difficult due to limited experimental data, making it challenging to accurately resolve each component. To mitigate this underdetermination, models are employed to provide insight into the specific contributions. A common approach assumes that, at low current densities, reaction kinetics adhere to the Tafel equation. By fitting the iR-corrected polarization curves, Tafel slopes can be derived, with deviations at mid- and high-current densities often attributed to mass transport effects. Yet, given the complexities of porous electrodes and necessary simplifications, it remains uncertain whether this traditional method is entirely applicable. In this study, we introduce a statistical method for voltage breakdown analysis that does not assume adherence to the Tafel equation for reaction kinetics. We hypothesize that the precise gas concentration within the catalyst layers is unknown but must exceed that in the flow fields, in accordance with mass transport principles. Mass transport should, therefore, result in a concentration increase in the catalyst layer with higher production rates. Additionally, we assume that increasing the cathode gas pressure in a PEM water electrolyzer minimally impacts the kinetic overpotentials of both the oxygen and hydrogen evolution reactions. Consequently, when cathode flow field gas concentration varies across otherwise identical experiments, we anticipate systematic shifts between polarization curves. These shifts should align with both transport laws and the Nernst equation. By correcting the iR-free cell voltage based on mass transport effects, reaction kinetics—potentially displaying a Tafel slope—can be unveiled without presupposing it. Surprisingly, our experimental data reveals that the overpotential attributed to reaction kinetics aligns closely with a Tafel slope across the full current density range (R² > 0.999). This approach reduces the unexplained overpotential by 90% and reveals a slightly steeper Tafel slope (+5%) compared to conventional methods. Additionally, the calculated increase in gas concentration within the catalyst layer agrees with transport laws, predicting a gas pressure increase of approximately 380 mbar·cm²·A⁻¹, indicating a transport resistance intermediate between those of membranes and porous transport layers and pointing to the catalyst layer as a source. While these findings are unexpectedly precise and potentially groundbreaking, they contradict initial assumptions, necessitating a careful re-evaluation and further discussion.