Accelerated Stress Tests to Project PEM Fuel Cell Durability

Abstract

One of the major degradation mechanisms limiting the long-term durability of proton exchange membrane fuel cells (PEMFCs) is the loss of platinum electrochemically active surface area (ECSA) of the carbon-supported platinum (Pt/C) cathode catalyst, caused by Pt dissolution that is followed by both Ostwald ripening of the Pt nanoparticles and loss of Pt into the ionomer phase [1]. The Pt ECSA loss is accelerated when subjecting PEMFCs to extended load-cycling inducing concomitant cycling of the cathode potential. To this end, accelerated stress tests (ASTs) can be conducted either by controlling the cell/stack current (“load-cycling” AST) under H2/air (anode/cathode) or the potential (“voltage-cycling” AST) under H2/N2 (anode/cathode).Most of the experiments studying the effect of load-cycling on catalyst durability have been based on voltage-cycling in a H2/N2 configuration, showing that Pt ECSA loss is aggravated with increasing upper potential limit (UPL), temperature, and relative humidity (RH) [2, 3, 4]. Comparing voltage-cycling induced degradation under H2/N2 versus H2/air, a recent study has found an essentially identical Pt ECSA loss, but a slightly higher H2/air performance decay when cycling under H2/N2 [4]. This study by General Motors Corporation indicates that Pt ECSA loss may not be a unique descriptor for H2/air performance loss, contrary to what was observed in recent work by Toyota Motor Corporation and Kyushu University [5] as well as in our own studies [6].In this talk, we will discuss the correlation between H2/air performance loss and Pt ECSA loss during voltage-cycling ASTs at different conditions (UPL, RH, and gas-feed) conducted in 5 cm2 active area single-cell PEMFCs, complemented by a voltage-loss analysis to deconvolute oxygen reduction reaction (ORR) activity losses and mass transport losses (due to oxygen mass transport and proton conduction in the cathode catalyst layer). We will also discuss whether the H2/air performance loss is governed by the Pt ECSA loss (independent of catalyst loading) or, as we had proposed previously, by the cathode electrode roughness factor (rf) loss (in cm2 Pt/cm2 cathode, i.e., the product of ECSA and Pt loading) [7]. As roughly 100,000 [8] or even more voltage-cycles are expected for heavy-duty applications, requiring very long measurement times, an approach to relate the degradation under harsh AST conditions with those under application-relevant conditions will be discussed. Experiments are conducted with cathode catalysts based on different carbon supports (Vulcan, Ketjenblack, or so-called accessible carbon supports [9, 10]), on catalysts with different initial Pt ECSAs (i.e., different Pt nanoparticle sizes) and with different initial Pt-loadings (i.e. different initial rf). Finally, exploratory experiments to evaluate the effect of start-up/shut-down on the correlation between electrode rf and H2/air performance decay will be discussed.