High-precision identification of polarization processes of proton exchange membrane fuel cells through relaxation time analysis: Targeted experimental design and verification

Abstract

Distribution of relaxation time (DRT) technology can achieve more precise decoupling of polarization process of proton exchange membrane fuel cell (PEMFC). However, except for the two peaks of DRT at low frequency, which are widely acknowledged to the cathode charge transfer impedance and the oxygen mass transfer impedance, there has been ongoing debate regarding the interpretation of other peaks observed at high frequency. In this work, the single-state variable control method is introduced by design of the membrane electrode assembly (MEA) and corresponding working operating conditions to simulate performance degradation factors of PEMFC, enabling accurate identification and verification of each peak in DRT. Based on the previous understanding that peaks P1 and P2 are oxygen mass transfer impedance and cathode charge transfer impedance, peaks P3-P5, ranging from high to low time scales, are determined to correspond to the other three polarization impedances. Specifically, peaks P3 and P4 are influenced by ionomer content in cathode and anode catalyst layers, respectively, and exhibit a decline with increasing humidity. They correspond to proton transfer processes within their respective catalyst layers. And the peak P5 is affected by anode catalyst loading and displays an increase with rising current density at low loading, signifying its involvement in the charge transfer process of anodic hydrogen oxidation reaction. A novel ECM was developed, consisting of five series-connected Randles circuits with polarization significance. The cathode and anode were separated in this design. The results demonstrate the enhanced sensitivity of DRT analysis to the performance degradation of PEMFC caused by oxygen transfer, charge transfer, and proton migration. This perspective provides a clear refinement for prior work of DRT polarization characteristic peaks.