The measured impedance spectra are commonly analyzed by a complex nonlinear least square (CNLS) fit, delivering a model function represented by an equivalent circuit model (ECM). In this case the ECM has to be defined a priori, without knowing number, size and time constants of the contributing polarization processes. The assessment of a physico-chemically motivated ECM remains challenging, especially if polarization processes with close time constants occur. This leads to an overlap of processes in the frequency domain and thus makes it difficult to separate them, which may result in wrong model assumptions and instable and ambivalent fitting results.
To overcome this issue we will present an alternate approach, where ECM and starting parameters for the CNLS fit are obtained by a pre-identification of the impedance response by the distribution of relaxation times (DRT) method [1],[2]. This approach was established in the framework of our solid oxide fuel cells research [3], later on transferred to lithium-ion batteries [4] and recently to high temperature polymer electrolyte fuel cells (HT-PEMFC) [5].
A typical result of our approach on a low-temperature PEM fuel cell is displayed in Figure 1: Whereas in the imaginary part of the spectrum only one broad peak is visible, the DRT facilitates a deconvolution of up to five clearly separable peaks (P1 - P5). In order to assign each peak to a physicochemical mechanism, systematic and comprehensive variations of operating parameters as current density, humidity and gas composition at anode and cathode have been conducted. By that we are able to distinguish between (i) charge transfer kinetics at the Pt-catalyst, (ii) the proton transport coupled therewith in the ionomer of the catalyst layer and (iii) the gas transport mechanisms in the porous media of the electrode and gas diffusion layer (GDL). Furthermore, our analysis enables us to assign the polarization processes to the respective electrode and to quantify their share to the overall polarization.
References
[1] H. Schichlein, A. C. Müller, M. Voigts, A. Krügel, and E. Ivers-Tiffée, J. Appl. Electrochem., 32, 875–882 (2002).
[2] E. Ivers-Tiffée and A. Weber, J. Ceram. Soc. Japan, 125, 193–201 (2017).
[3] A. Leonide, V. Sonn, A. Weber, and E. Ivers-Tiffée, J. Electrochem. Soc., 155, B36–B41 (2008).
[4] J. Illig, M. Ender, T. Chrobak, J. P. Schmidt, D. Klotz, and E. IVers-Tiffée, J. Electrochem. Soc., 159, A952–A960 (2012).
[5] A. Weiß, S. Schindler, S. Galbiati, M. A. Danzer, and R. Zeis, Electrochim. Acta, 230, 391–398 (2017).