Different Approaches for the Determination of Degradation Rates in PEMFC Durability Tests

Abstract

In our presentation we show results obtained in the scope of the automotive FCH-JU project Impact. Thereby we focus on a detailed description of voltage loss rates of PEM fuel cells and particularly on the discrimination between reversible and irreversible voltage losses. A major motivation of our work is the lack of common description procedures and determination approaches of voltage losses in durability tests of fuel cell; this issue leads to severe difficulties in the comparison of results obtained by different institutions or within different projects especially if only one value for a degradation rate is reported. In order so address voltage losses we have performed single cell durability measurements of several hundreds of hours in 25 cm2 lab-scale cells using specific test protocols containing regular refresh procedures allowing distinguishing between reversible and irreversible losses. To test possible refresh procedures and analyze their effect on cell performance recovery a systematic analysis of e.g. duration of OCV, potential changes, gas interruption, changes of humidification was studied. Between these refresh steps the cells were typically operated using NEDC-based dynamic load cycles. As samples conventional 5-layer MEAs were used with PFSA membranes, Pt-based catalysts and hydrophobized carbon fiber substrates with micro porous layers as GDLs. For in-situ diagnostics of the operated cells I-V characteristics, EIS spectra, and CVs were recorded in order to determine the impact of the refresh procedures on the cells. Additionally we provide a comparison of various mathematical approaches to determine voltage decay rates. For instance, voltage values before or after voltage recovery steps can be used to calculate an irreversible loss rate. The advantages and drawbacks of both approaches are discussed. One clear conclusion is that short time tests in the range of 100 hour are not conclusive since this time is too short to make a reliable discrimination between reversible and irreversible losses; also, the decay rate of reversible loss observed after each refresh step may increases substantially upon long time operation. In summary, in our presentation operation strategies for determination of fuel cell voltages loss rates are compared, evaluated and assessed according to their suitability to distinguish between reversible and irreversible degradation rates; a description of voltage loss rates is proposed.