Cathode durability in PEM fuel cells
Redmond, Erin Leigh
MetadataShow full item record
Proton exchange membrane (PEM) fuel cells are competitive with other emerging technologies that are being considered for automotive transportation. Commercialization of PEM fuel cells would decrease emissions of criteria pollutants and greenhouse gases and reduce US dependence on foreign oil. However, many challenges exist that prevent this technology from being realized, including power requirements, durability, on-board fuel storage, fuel distribution, and cost. This dissertation focuses on fuel-cell durability, or more specifically catalyst stability. New techniques to comprehensively observe and pin-point degradation mechanisms are needed to identify stable catalysts. In this text, an in operando method to measure changes in catalyst particle size at the cathode of a PEM fuel cell is demonstrated. The pair distribution function analysis of X-ray diffraction patterns, generated from an operating fuel cell exposed to accelerated degradation conditions, was used to observe the growth of catalyst particles. The stability of Pt/C and PtCo/C electrodes, with different initial particle sizes, was monitored over 3000 potential cycles. The increase in particle size was fit to a linear trend as a function of cycle number for symmetric linear sweeps of potential. The most stable electrocatalyst was found to be alloyed PtCo with a larger initial particle size. A better understanding of oxide growth kinetics and its role in platinum dissolution is needed to develop a comprehensive fuel-cell performance model. There is an ongoing debate present in the current literature regarding which oxide species are involved in the oxide growth mechanism. This dissertation discusses the results of in operando X-ray absorption spectroscopy studies, where it was found that PtO2 is present at longer hold times. A new method to quantify EXAFS data is presented, and the extent of oxidation is directly compared to electrochemical data. This comparison indicated that PtO2 was formed at the expense of an initial oxide species, and these steps were included in a proposed mechanism for platinum oxidation. Simulations of platinum oxidation in literature have yet to fully replicate an experimental cyclic voltammogram. A modified Butler-Volmer rate equation is presented in this thesis. The effect of including an extra parameter, χ, in the rate equations was explored. It was found that while the χ-parameter allowed the cathodic peak width to be decoupled from the Tafel slope for the platinum-oxide reduction, its inclusion could not address all observed experimental characteristics. Exploration of this concept concluded that current is not a function of only potential and coverage. To that end, a heterogeneous oxide layer was introduced. In this model, place-exchanged PtO2 structures of varying energy states are formed through a single transition state. This treatment allowed, for the first time, the simulation of the correct current-potential behavior under varying scan rates and upper potential limits. Particle size plays a critical role in catalysts stability. The properties of nanoparticles can differ significantly from bulk values, yet few tools exist to measure these values at the nanoscale. Surface stress and surface energy are diagnostic criterion that can be used to differentiate nano from bulk properties. The pair distribution function technique was used to measure lattice strain and particle size of platinum nanoparticles supported on carbon. The effect of adsorbates on surface stress was examined and compared to previous literature studies. Furthermore, a methodology for measuring the surface energy of supported platinum nanoparticles has been developed. While the results of this work are significant, many more challenges need to be addressed before fuel-cell vehicles are marketed. Recommendations for future work in the field of catalyst durability are addressed.