Characterizing Electrode-Level Oxygen Transport in Polymer Electrolyte Fuel Cells

Polymer electrolyte fuel cells (PEFCs) are a promising technology for environmentally friendly automobiles, among other applications. However, performance losses due to oxygen transport hindrances in the PEFC’s cathode continue to be an issue in widespread commercialization. This dissertation focuses on the transport of oxygen through the thickness of the PEFC cathode, and the effect of the cathode’s microstructure on that transport. In order to react in the cathode, oxygen travels from gas flow channels down through a diffusion medium, through the pores of a catalyst layer, and finally, into and through the ionomer covering the catalyst particles. Transport resistances throughout this path lead to oxygen starvation at some of the catalyst particles. Due to these transport resistances, much of the platinum is underutilized when the fuel cell is operating at appreciable currents. This dissertation aims to characterize the transport resistance in each of these phases. We study (1) oxygen transport throughout the diffusion medium using a commercial electrochemical microsensor at multiple points, (2) oxygen transport through the entire diffusion medium using a thin film oxygen microsensor at one point, (3) transport through the catalyst layer pores using a device that allows oxygen microsensors to contact the side of the catalyst layer at multiple points, (4) the effect of the catalyst layer’s microstructure on oxygen transport using x-ray computed tomography, and (5) transport into and through ionomer-covered catalyst agglomerates using ex-situ experiments. We go on to discuss the application of similar methods to solid oxide fuel cells. Using the methods developed in this work, we determine that the two dominant oxygen transport resistances are the diffusion medium, and a “local” resistance at the interface of the platinum catalyst and the ionomer binder that has previously generated some controversy in the field. The oxygen transport resistance of the diffusion medium in this work (defined as the ratio of the drop in concentration across a component to the flux through it) is 65 s/m, with about 2/3 of that coming from its microporous layer. This value can rise to double or more in the case of liquid water condensation in the diffusion medium’s pores. We find that the oxygen transport resistance in the catalyst layer’s pores is an order of magnitude less than that of the diffusion medium. That value, too, can change depending on liquid water flooding. In previous works, the “local” oxygen transport resistance at the level of the platinum catalyst and ionomer binder was of unclear origin. We have determined that it arises at the Pt|ionomer interface – it does not originate from the ionomer|gas interface, nor is it due to nanoscale confinement effects. In our investigation of the catalyst layer’s morphology, we find that a popular approach to modelling PEFC performance – the agglomerate model – changes significantly when one incorporates a realistic distribution of agglomerate sizes instead of assuming a uniform agglomerate size. This effect, however, is small compared to the oxygen transport resistance of the diffusion medium oxygen resistance and the Pt|ionomer interfacial oxygen resistance.