Modeling Oxygen Concentration Oscillation in the Gas Channel of Polymer Electrolyte Fuel Cells: A Comparison between Numerical and Analytical Approaches

Electrochemical Impedance spectroscopy (EIS) is widely used for analyzing the operation of Polymer Electrolyte Fuel Cells. However, the interpretation of the two arcs that appear generally in Nyquist plots (in the case of PEFCs fed with air) can be difficult in the absence of clearly validated model. Indeed, the most widely encountered approaches are mono-dimensional and they rely on the Randles equivalent circuit that associate the high frequency loop to capacitive, kinetic, as well as charge transport effects (1) and the low frequency loop to oxygen transport (2). However, Schneider et al. shown clearly that the local spectra of air-fed PEFC (i.e., the impedance spectra measured locally along the air channel) are coupled to the transport of oxygen along the air channel ((3), (4)). In parallel with the works of Schneider et al., Kramer et al. (5) presented briefly a full numerical two-dimensional model that confirmed, at least qualitatively, the experimental results. Although much useful, this model consider only a half cell (the cathode considered as a surface electrode-, the gas diffusion layer, and the air channel) and being fully numerical, it does not allow a quick estimation of the significance of the oxygen concentration oscillations. On the other hand, we developed recently a semi-analytical model ((6), (7)). This model is based on simplifying assumptions that require to be validated. For this, we compare the results of a full numerical two-dimensional model to those of a pseudo-bidimensionnal approach.