LOW-TEMPERATURE, ANODE-SUPPORTED HIGH POWER DENSITY SOLID OXIDE FUEL CELLS WITH NANOSTRUCTURED ELECTRODES

Abstract

This report summarizes the work done during the entire project period, between October 1, 1999 and March 31, 2003, which includes a six-month no-cost extension. During the project, eight research papers have, either been, published, accepted for publication, or submitted for publication. In addition, several presentations have been made in technical meetings and workshops. The project also has provided support for four graduate students working towards advanced degrees. The principal technical objective of the project was to analyze the role of electrode microstructure on solid oxide fuel cell performance. Prior theoretical work conducted in our laboratory demonstrated that the particle size of composite electrodes has a profound effect on cell performance; the finer the particle size, the lower the activation polarization, the better the performance. The composite cathodes examined consisted of electronically conducting perovskites such as Sr-doped LaMnO{sub 3} (LSM) or Sr-doped LaCoO{sub 3} (LSC), which is also a mixed conductor, as the electrocatalyst, and yttria-stabilized zirconia (YSZ) or rare earth oxide doped CeO{sub 2} as the ionic conductor. The composite anodes examined were mixtures of Ni and YSZ. A procedure was developed for the synthesis of nanosize YSZ by molecular decomposition, in which unwanted species were removed by leaching, leaving behind nanosize YSZ. Anode-supported cells were made using the as-synthesized powders, or using commercially acquired powders. The electrolyte was usually a thin ({approx}10 microns), dense layer of YSZ, supported on a thick ({approx}1 mm), porous Ni + YSZ anode. The cathode was a porous mixture of electrocatalyst and an ionic conductor. Most of the cell testing was done at 800 C with hydrogen as fuel and air as the oxidant. Maximum power densities as high as 1.8 W/cm{sup 2} were demonstrated. Polarization behavior of the cells was theoretically analyzed. A limited amount of cell testing was done using liquid hydrocarbon fuels where reforming was achieved internally. Significant polarization losses also occur at the anode, especially at high fuel utilizations. An analysis of polarization losses requires that various contributions are isolated, and their dependence on pertinent parameters is quantitatively described. An investigation of fuel composition on gas transport through porous anodes was investigated and the role of fuel diluents was explored. This work showed that the molecular weight of the diluent has a significant effect on anode concentration polarization. This further showed that the presence of some molecular hydrogen is necessary to minimize polarization losses. Theoretical analysis has shown that the electrode microstructure has a profound effect on cell performance. In a series of experiments, cathode microstructural parameters were varied, without altering other parameters. Cathode microstructural parameters, especially three phase boundary (TPB) length, were estimated using techniques in quantitative stereology. Cell performance was quantitatively correlated with the relevant microstructural parameters, and charge transfer resistivity was explicitly evaluated. This is the first time that a fundamental parameter, which governs the activation polarization, has been quantitatively determined. An important parameter, which governs the cathodic activation polarization, and thus cell performance, is the ionic conductivity of the composite cathode. The traditional composite cathode is a mixture of LSM and YSZ. It is well known that Sr and Mg-doped LaGaO{sub 3} (LSGM), exhibits higher oxygen ion conductivity compared to YSZ. Cells were fabricated with composite cathodes comprising a mixture of LSM and LSGM. Studies demonstrated that LSGM-based composite cathodes exhibit excellent behavior. Studies have shown that Ni + YSZ is an excellent anode. In fact, in most cells, the principal polarization losses, at least at low fuel utilizations, are associated with the cathode. Theoretical analysis conducted in our group has also shown that anode-supported cells exhibit lower polarization losses compared to cathode-supported or electrolyte-supported cells. It is essential, however, that Ni + YSZ anode be not subjected to a re-oxidation step, such as may occur accidentally, as this can cause anode disintegration. A theoretical model was developed for reduction and re-oxidation kinetics of Ni-based anodes. The model was experimentally verified. In addition, criteria for the development of redox tolerant, yet Ni-based, anodes were established.