Vapor Deposition Process for Engineering of Dispersed PEMFC ORR Pt/NbO<sub>x</sub>/C Catalysts

Abstract

EXECUTIVE SUMMARY With support from the Department of Energy, and in partnership with Oak Ridge National Laboratory, Exothermics, Northeastern University, the University of Michigan, and IRD Fuel Cells, Ford has led an effort to use physical vapor deposition (PVD) to produce oxygen reduction reaction (ORR) catalysts for proton exchange membrane fuel cells (PEMFCs). Conventional ORR catalysts are made by wet chemical processes, and experience loss of electrochemical surface area and activity over voltage cycling due to Pt dissolution and Pt particle coalescence. Not only was PVD processing researched as a means to reduce solvent waste and enhance reproducibility in catalyst manufacturing, but it was also perceived as a means to deposit preferential phases of niobium oxide onto the catalyst, which would then serve to prevent Pt particle movement and extend PEMFC lifetime. The project was able to generate samples of Pt/NbOx/C from PVD that far exceeded the mass activities of commercial Pt/C, as measured in rotating disk electrode (RDE) experiments in conventional glass cells. These samples also avoided the presence of the electrically insulating phase of niobium oxide, crystalline Nb2O5. However, magnetron sputtering was unable to provide for a narrow particle size distribution, and the average Pt particle size was smaller than for commercial catalysts (~ 2 nm versus 4-5 nm in commercial catalysts). Furthermore, X-ray absorption spectroscopy techniques showed that Pt-Nb interactions did not exist for samples for which Nb and Pt had been sequentially sputtered, and that no strain effect existed to enhance activity. PVD samples were shown to be worse than commercial Pt/C for both mass activity and power density. Samples that were made through co-sputtering did show improved Pt-Nb interactions at small batch scale. However, the co-sputtered samples showed poor RDE mass activity. The use of niobium oxide did show signs of improved durability, despite its low performance. A PVD Pt/NbOx/C sample containing 35 wt% Pt and 7.2 wt% NbOx endured 30,000 cycles between 0.6 and 0.95 V while losing only 1.7% of its mass activity and only 20 mV performance at 0.8 A/cm2. In contrast, the commercial Pt/C lost 46% mass activity and 31 mV performance under the same test metrics. While PVD Pt/NbOx/C catalysts are clearly unfit for usage as ORR catalysts, they could be a candidate for further study as PEMFC anode catalysts.