This title appears in the Scientific Report : 2016 

Analytical tools for deciphering PEM fuel cell impedance spectra
Kulikovsky, Andrei (Corresponding author)
Elektrochemische Verfahrenstechnik; IEK-3
2016
67th Annual Meeting International Society of Electrochemistry, The Hague (Netherlands), 2016-08-21 - 2016-08-26
Conference Presentation
Fuel Cells
Analytical Tools for Deciphering PEM Fuel Cell Impedance Spectra Andrei KulikovskyResearch Centre Juelich, IEK-3Juelich, 52425, GermanyA.Kulikovsky@fz-juelich.deEIS is a unique method for in—operando measurement of PEM fuel cell transport and kinetic parameters. Understanding impedance spectra requires modeling. So far, PEMFC developers routinely used equivalent transmission lines (TLs) for fitting experimental spectra. However, relation of TL elements to the cell physical parameters is beyond the scope of this technique. We report analytical solutions for PEMFC impedance. The solutions are obtained from the general system of conservation equations for the cathode catalyst layer, the gas—diffusion layer, and for the flow in the air channel [1—3]. A model with the air flow in the channel predicts the effect of spatial (along the channel) oscillations of the local impedance [2]. Comparison of the models with and without the channel flow shows that the channel impedance rapidly vanishes with the frequency growth and hence the high—frequency part of the spectra can be fitted using much simpler 1D (through—plane) impedance models. All the analytical solutions are “fast” enough for using in accurate least—squares algorithms for experimental spectra fitting [1—4]. Using this technique, we have determined the basic parameters of a PEMFC [4] and high—temperature PEMFC [3]. Typical CPU time for fitting one of the models to a measured spectrum is less than one minute on a standard PC. However, validity of the analytical impedance equations is limited by the cell current density on the order of 100 mA/cm2. This limitation can be relaxed by using a fully numerical impedance model [5]. The model includes numerical solution of the system of linear equations for the perturbation amplitudes in the omega—space, plus solution of nonlinear ODEs for the static shapes of the overpotential and oxygen concentration. The presence of a boundary—value problem solver makes the fitting code much slower; it takes about an hour to fit a single spectrum on a standard PC. Nonetheless, the model is suitable for research purposes. Together with Tatyana Reshetenko (University of Hawaii) we used this model to obtain the dependence of PEMFC parameters on the cell current density in the range from 50 to 400 mA/cm2 [5]. This work reveals an unexpected stepwise growth of the CCL proton conductivity at the cell current of 200 mA/cm2. This growth indicates structural changes of Nafion cluster in the CCL."Missing figures - could not be integrated in the abstract text." 1.A.A.Kulikovsky. Electrochim. Acta, 147 (2014) 773, DOI: 10.1016/j.electacta.2014.09.145; J.Electrochem. Soc. 162 (2015) F217, DOI: 10.1149/2.0151503jes; ibid, 163 (2016) F319, DOI: 10.1149/2.0111605jes; Electrochim. Acta, 2016 (in press).2. A. Kulikovsky, O. Shamardina. J.Electrochem. Soc. 162 (2015) F1068, DOI: 10.1149/2.0911509jes3. O. Shamardina, M. S. Kondratenko, A. V. Chertovich, A. A. Kulikovsky. Int. J. Hydrogen Energy, 39 (2014) 2224. DOI: 10.1016/j.ijhydene.2013.11.058 4. T.Reshetenko and A.Kulikovsky. J.Electrochem. Soc. 162 (2015) F627, DOI: 10.1149/2.1141506jes; ibid, 163 (2016) F238, DOI: 10.1149/2.0871603jes5. T.Reshetenko and A.Kulikovsky. J.Electrochem. Soc. (2016, under review).