Ceramic processing and electrochemical analysis of proton conductive solid oxide fuel cell

Abstract

Ba(Zr 0.8-xCe xY 0.2)O 3-delta (0 < or = x < or = 0.4) (BZCYs) powders were successfully fabricated by both solid state reaction and glycine-nitrate process. Lithium fluoride (LiF) was selected as a liquid phase sintering additive to lower the sintering temperature of BZCYs. Using LiF as an additive, high density BZCYs ceramics can be obtained at sintering temperatures 200~300 °C lower than the usual 1700 °C with much shorter soaking time. Nuclear reaction investigations showed no lithium and a small amount of fluorine reside in the sample which indicates the non-concomitant evaporation of lithium and fluorine during the sintering process. Scanning electron microscopic investigations showed the bimodal structure of BZCY ceramics and grain growth as Ce content increases. In a water saturated hydrogen containing atmosphere, BZCY ceramics have higher conductivity when LiF is used in the sintering process. LiF-added BZCY electrolyte-supported fuel cells with different cathodes were tested at temperatures from 500 ~ 850 °C. Results show that Pt cathode gives much higher power output than ceramic cathodes, indicating much larger polarization from ceramic cathodes than Pt. Ba(Zr 0.6Ce 0.2Y 0.2)O 3-delta anode supported proton conductive solid oxide fuel cells (H-SOFCs) show low power output due to its low proton conductivity. Ba(Ce 0.8Y 0.2)O 3-delta anode supported H-SOFCs show excellent power output. Different H 2 and O 2 partial pressures were used for fuel and oxidative gas, respectively, to obtain information for V(i) modeling. Different thicknesses of supporting anode were used to obtain saturation current densities of H-SOFC. Using the dusty-gas model which includes Stefan-Maxwell equation and Knudsen terms, the calculation gave tortuosity of our supporting anode 1.95 ± 0.1. The gas concentrations across the anode were also calculated by knowing the tortuosity of the supporting anode. An electrochemical model of H-SOFC was developed. The excellent agreement between model and experimental data implies that our model is close to the true physical picture of H-SOFC. The more accurate prediction of our model, based on a physical picture of electrochemical processes, also provides a replacement for using the Butler-Volmer equation in SOFC modeling. In the parametric analysis, our model shows that ohmic polarization and cathodic polarization limit the performance of H-SOFC. Research for improving H-SOFC performance should be focused on reducing electrolyte thickness, increasing proton conductivity of electrolyte and finding a compatible cathode material.