About the Necessity to Consider Membrane Electrolyte Degradation Statistically

Abstract

Listen

Translate article

One key factor for the commercialization of fuel cell vehicles is a sufficient membrane durability. In this context, car manufactures work on reliability accelerated stress test procedures (AST) to measure membrane lifetime. There have been many lifetime investigations for polymer electrolyte membrane fuel cells (PEMFCs) especially for the perfluorinated sulfonic acid (PFSA) membrane [1-3]. For comparison of membrane lifetimes, it is necessary to ensure reproducibility of the experimental lifetime data. However, in part of the literature lifetime of membranes was derived from few repetitions – if not none – of single cell experiments only [e.g. 3-5]. Random failures due to membrane production variations, cell assembly inaccuracies and statistical defects cannot be quantified without statistical analysis. Accordingly, some conclusions in literature might have been blurred due to insufficient repetition of measurements. In this work, we stress the necessity to generate statistically relevant membrane degradation data. The development of the open circuit voltage (OCV) and the H2-crossover current (iH2) is shown in Figs. 1-2 for one specific AST and nine cells. It can be seen, that in the range of reproducibility lifetime can be 1000 or 20000 cycles. At the conference more evidence for the wide nature of the lifetime distribution will be presented. Accordingly, the distribution in voltage loss and hydrogen crossover development cannot be neglected if clear conclusions about membrane life shall be derived. Without a high number of repetitions the statistical significance is not assured. Some of the recent conclusions in literature about impact on membrane lifetime will be considered critically from this point of view.