Enhanced Power Stability for Proton Conducting Solid Oxides Fuel Cells

Abstract

In order to provide the basis for a rational approach to improving the performance of Y-doped BaZrO{sub 3} electrolytes for proton conducting ceramic fuel cells, we carried out a series of coupled computational and experimental studies to arrive at a consensus view of the characteristics affecting the proton conductivity of these systems. The computational part of the project developed a practical first principles approach to predicting the proton mobility as a function of temperature and doping for polycrystalline systems. This is a significant breakthrough representing the first time that first principles methods have been used to study diffusion across grain boundaries in such systems. The basis for this breakthrough was the development of the ReaxFF reactive force field that accurately describes the structure and energetics of Y-doped BaZrO{sub 3} as the proton hops from site to site. The ReaxFF parameters are all derived from an extensive set of quantum mechanics calculations on various clusters, two dimensionally infinite slabs, and three dimensionally infinite periodic systems for combinations of metals, metal alloys, metal oxides, pure and Y-doped BaZrO{sub 3}, including chemical reaction pathways and proton transport pathways, structures. The ReaxFF force field enables molecular dynamics simulations to be carried out quickly for systems with {approx} 10,000 atoms rather than the {approx}100 or so practical for QM. The first 2.5 years were spent on developing and validating the ReaxFF and we have only had an opportunity to apply these methods to only a few test cases. However these simulations lead to transport properties (diffusion coefficients and activation energy) for multi-granular systems in good agreement with current experimental results. Now that we have validated the ReaxFF for diffusion across grain boundaries, we are in the position of being able to use computation to explore strategies to improve the diffusion of protons across grain boundaries, which both theory and experiment agree is the cause of the low conductivity of multi-granular systems. Our plan for a future project is to use the theory to optimize the additives and processing conditions and following this with experiment on the most promising systems. The experimental part of this project focused on improving the synthetic techniques for controlling the grain size and making measurements on the properties of these systems as a function of doping of impurities and of process conditions. A significant attention was paid to screening potential cathode materials (transition metal perovskites) and anode electrocatalysts (metals) for reactivity with Y-doped BaZrO{sub 3}, fabrication compatibility, and chemical stability in fuel cell environment. A robust method for fabricating crack-free thin membranes, as well as methods for sealing anode and cathode chambers, have been successfully developed. Our Pt|BYZ|Pt fuel cell, with a 100 {micro}m thick Y-doped BaZrO{sub 3} electrolyte layer, demonstrates the peak power density and short circuit current density of 28 mW/cm{sup 2} and 130mA/cm{sup 2}, respectively. These are the highest values of this type of fuel cell. All of these provide the basis for a future project in which theory and computation are combined to develop modified ceramic electrolytes capable of both high proton conductivity and excellent mechanical and chemical stability.