Nonlinear Impedance Spectra Analysis of CO Poisoning on PEM Fuel Cell Performance

Abstract

In polymer electrolyte membrane (PEM) fuel cells, hydrogen is the preferred fuel with air or oxygen as oxidant. However, many a time, the presence of carbon monoxide (CO), even at low ppm levels, inhibits the hydrogen oxidation reaction [1, 2]. In this work, we describe nonlinear impedance spectra analysis (NLEIS) of CO poisoning in PEM fuel cells. In NLEIS method, large amplitude perturbations are employed and the resulting signal have significant higher harmonic components, which, if acquired, can provide additional information about the system under investigation. Here we illustrate a NLEIS methodology in which governing equations were solved without linearization, using numerical method. The response signal was subjected to fast Fourier transform (FFT) to extract magnitude and phase at fundamental and higher harmonics. In this study, linear adsorption model, in which a CO molecule occupies one adsorption site was employed to generate NLEIS data [3]. Figure 1 shows the concentration effect on NLEIS results, simulated under galvanostatic mode employing large (Iac = 200 mA) amplitude perturbation at Idc =200 mA.cm-2 . The results show that the CO poisoning effects are more easily distinguishable at higher harmonics, when compared to the response at fundamental frequency. Key words: Nonlinear EIS, PEM fuel cell, CO poisoning, linear adsorption model