CATHODES FOR LOW TEMPERATURE SOFC: ISSUES CONCERNING INTERFERENCE FROM INERT GAS ADSORPTION AND CHARGE TRANSFER

Abstract

This report summarizes the work done on the project over the duration of the project, from October 1, 2002 through December 31, 2003, which includes a three month no-cost extension. Effort was directed in the following areas: (1) Fabrication of Sr-doped LaCoO3 (LSC) dense and porous samples. (2) Design and construction of a conductivity relaxation apparatus for the estimation of surface exchange coefficient, k{sub chem}, which depends on adsorption, and oxygen chemical diffusion coefficient, {tilde D}{sub 0}, the parameters which are thought to describe the cathodic activation polarization (overall charge transfer) in mixed ionic electronic conducting (MIEC) cathodes. (3) The measurement of and K{sub chem} and {tilde D}{sub 0} on LSC by conductivity relaxation, as a function of temperature and oxygen partial pressure, p{sub O{sub 2}}. (4) Fabrication of YSZ electrolyte discs with patterned LSM and LSC electrodes with three-phase boundary (TPB) length, l{sub TPB}, varying between 50 and 1200 cm{sup -1}. (5) The measurement of charge transfer resistance, R{sub ct}, and estimation of the charge transfer resistivity, {rho}{sub ct}, as a function of temperature and p{sub O{sub 2}}, and the incorporation of the adsorption step in the analysis. (6) Preliminary cell tests with oxidants having different inert gas diluents; N{sub 2}, Ar, and CO{sub 2}. Dense samples of LSC of thickness as small as 150 microns were fabricated by sintering followed by grinding. Porous samples of LSC were also fabricated wherein the porosity was {approx}30%. Both samples were used in conductivity relaxation experiments. Analysis of data from the dense samples gives both and k{sub chem} and {tilde D}{sub 0}, while that of porous samples gives k{sub chem}. It was observed that at a given temperature, k{sub chem} increases with increasing p{sub O{sub 2}}, while the {tilde D}{sub 0} is essentially a constant. The dependence of k{sub chem} on p{sub O{sub 2}} is attributed to the adsorption step. It was also observed that the porous samples gave a more accurate measurement of k{sub chem}, as the data were not influenced by {tilde D}{sub 0}. By contrast, the results on dense samples were influenced by {tilde D}{sub 0}, especially at lower temperatures. It is thus concluded that the use of porous samples is preferred for the measurement of k{sub chem}. In the case of composite electrodes, such as LSM + YSZ, the relevant parameters are the {rho}{sub ct} (or R{sub ct}) and the ionic resistivity of YSZ {rho}{sub i}, where 1/{rho}{sub ct} is analogous to k{sub chem} and 1/{rho}{sub i} is analogous to {tilde D}{sub 0}. LSM patterned electrodes were deposited on YSZ discs using photomicrolithography. The R{sub ct} was measured as a function of temperature and p{sub O{sub 2}}using complex impedance techniques, on samples with l{sub TPB} varying between 50 and 1200 cm{sup -1}. The plot of 1/R{sub ct} vs. l{sub TPB} was linear, consistent with the occurrence of charge transfer at TPB. Also, the data plotted on the assumption of dissociative adsorption was consistent with the model. The significance of the role of adsorption is discussed. Similar results were observed with LSC, indicating a similar role of adsorption. In the case of LSC, however, a significant transport of oxygen also occurs through the dense part of the electrode. Preliminary work was conducted on the testing of button cells with mixtures of O{sub 2} + N{sub 2}, O{sub 2} + Ar, and O{sub 2} + CO{sub 2} as oxidants, wherein the p{sub O{sub 2}} was varied between {approx}0.05 and {approx}1.0 atm. As expected, the results showed that the higher the p{sub O{sub 2}}, the better was the performance. In pure oxygen, the maximum power density at 800 C was {approx}2.9 W/cm{sup 2}. However, in 5% O{sub 2}, it was {approx}0.6 W/cm{sup 2}. This difference is attributed to adsorption, indicating that both charge transfer and adsorption needs to be addressed in order to improve cathode performance at lower temperatures and under high oxidant utilization (in low p{sub O{sub 2}} atmospheres). Data at low current densities was analyzed to deduce the polarization resistance, R{sub p}, and its dependence on p{sub O{sub 2}}. Although preliminary, the results show differences in R{sub p}, which are attributed to possible differences in adsorption characteristics. Specifically, it was observed that the polarization resistance was lower in O{sub 2} + CO{sub 2} atmospheres compared to O{sub 2} + N{sub 2} and O{sub 2} + Ar atmospheres. This suggests that CO{sub 2} has a lower tendency for adsorption compared to the other two diluents.