Electrochemical Impedance Spectroscopy (EIS) is a well-established technique for studying Polymer Exchange Membrane Fuel Cells (PEMFC) but data interpretation remains delicate, mostly because impedance models are either based on oversimplified equations or conversely, include too many correlated parameters. It is thus crucial to carefully choose the models to interpret impedance data, according to FC materials and operation conditions. Most of PEMFC impedance spectra are composed of two loops in Nyquist plot that can be perfectly represented by classical Randles Electrical Equivalent Circuit (EEC). However, several spectra show a straight line at high frequencies associated with proton conduction in the cathode catalyst layer. Assuming an interface electrode, the Randles EEC is poorly adapted to such spectra and one will rather use Transmission Line Models (TLM). However, since TLM do not usually consider mass transport, it is necessary to adapt the EEC, especially at the cathode. Such EEC can then be used as general FC models independently of the occurrence of the straight line at high frequencies, i.e. independently of the ratio between proton conduction and reaction kinetics limitations. These TLM EECs are then used to analyze the layer(s) at the origin of oxygen transport limitations: catalyst and/or the gas diffusion layer.

No abstract

No abstract

One common way to interpret the data of Electrochemical Impedance Spectroscopy (EIS) with Polymer Exchange Membrane Fuel Cells (PEMFC) consists in using an Equivalent Electrical Circuit (EEC). There are however various issues in EEC modeling, among which the location and expression of the oxygen transport impedance. In this work, we compare the results obtained using a Randles circuit with those of an EEC where the oxygen diffusion impedance is connected in series with the circuit of the Cathode Catalyst Layer (CCL). In the Randles circuit, the oxygen transport impedance is in series with the charge transfer resistance of the Oxygen Reduction Reaction (ORR), implying that the CCL (pores and/or ionomer) is governing oxygen diffusion. In the other case, the oxygen diffusion impedance is outside of the CCL circuit, which implicates that the Gas Diffusion Layer (GDL) governs oxygen diffusion. In addition, two expressions of the GDL oxygen diffusion impedance were tested: the usual finite Warburg impedance and an alternative expression derived by Kulikovsky that considers the impact of the double-layer capacity on oxygen concentration at the CCL/GDL interface. The parameters obtained with these EEC are used to estimate the main characteristic diffusion length, for various cells and operating conditions. The same trend was observed in all cases: the values of the characteristic diffusion length are found to be of the order of the GDL thickness.