SOFC Anodes Based on Infiltration of La0.3Sr0.7TiO3

Composites formed by infiltration of 45 wt % La0.3Sr0.7TiO3 (LST) into 65% porous yttria-stabilized zirconia (YSZ) were examined for application as solid oxide fuel cell (SOFC) anodes. Although LST does not react with YSZ, the structure of the LST deposits was strongly affected by the calcination temperature. At 1373 K, the LST formed loosely packed, 0.1 μm particles that filled the YSZ pores. The conductivity of this composite depended strongly on the pretreatment conditions but was greater than 0.4 S/cm after heating to 1173 K in humidified (3% H2O)H2. Following calcination at 1573 K, the LST had sintered significantly, decreasing the conductivity of the composite by a factor of approximately 5. The addition of a catalyst was critical for achieving reasonable electrochemical performance, with the addition of 0.5 wt % Pd and 5 wt % ceria increasing the power density of otherwise identical cells from less than 20 to 780 mW/cm2 for operation in humidified (3% H2O)H2 at 1073 K. Electrodes prepared from LST deposits calcined at 1373 K were found to exhibit a much better performance than those prepared from LST deposits calcined at 1573 K, demonstrating that the structure of the composite is critical for achieving high performance. Comments © The Electrochemical Society, Inc. 2008. All rights reserved. Except as provided under U.S. copyright law, this work may not be reproduced, resold, distributed, or modified without the express permission of The Electrochemical Society (ECS). The archival version of this work was published in Journal of the Electrochemical Society, Volume 155, Issue 11, 2008, pages B1179-B1183. Publisher URL: http://doi.dx.org/10.1149/1.2976775 This journal article is available at ScholarlyCommons: http://repository.upenn.edu/cbe_papers/124 SOFC Anodes Based on Infiltration of La0.3Sr0.7TiO3 Shiwoo Lee, Guntae Kim, J. M. Vohs,* and R. J. Gorte* Department of Chemical and Biomolecular Engineering, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA Korea Institute of Energy Research, Daejeon 305-343, Korea Composites formed by infiltration of 45 wt % La0.3Sr0.7TiO3 LST into 65% porous yttria-stabilized zirconia YSZ were examined for application as solid oxide fuel cell SOFC anodes. Although LST does not react with YSZ, the structure of the LST deposits was strongly affected by the calcination temperature. At 1373 K, the LST formed loosely packed, 0.1 m particles that filled the YSZ pores. The conductivity of this composite depended strongly on the pretreatment conditions but was greater than 0.4 S/cm after heating to 1173 K in humidified 3% H2O H2. Following calcination at 1573 K, the LST had sintered significantly, decreasing the conductivity of the composite by a factor of approximately 5. The addition of a catalyst was critical for achieving reasonable electrochemical performance, with the addition of 0.5 wt % Pd and 5 wt % ceria increasing the power density of otherwise identical cells from less than 20 to 780 mW/cm2 for operation in humidified 3% H2O H2 at 1073 K. Electrodes prepared from LST deposits calcined at 1373 K were found to exhibit a much better performance than those prepared from LST deposits calcined at 1573 K, demonstrating that the structure of the composite is critical for achieving high performance. © 2008 The Electrochemical Society. DOI: 10.1149/1.2976775 All rights reserved. Manuscript submitted June 12, 2008; revised manuscript received July 22, 2008. Published September 22, 2008. Increased efficiency for the conversion of chemical energy to electrical energy is going to be very important for many applications in the future. Because of this, solid oxide fuel cells SOFCs are attractive for the intrinsically high efficiency they exhibit. This efficiency derives in part from their high operating temperatures between 873 and 1073 K , which decrease electrode overpotentials compared to that found with other types of fuel cells. The electrode overpotential is defined as the difference between the ideal Nernst potential and the actual electrode potential. Equally important, SOFCs are “fuel” flexible, partly because of their high operating temperatures but also because the electrolytes are oxygen-ion conductors rather than proton conductors. In principle, any combustible fuel can react with the oxygen ions to produce electrons. However, to take full advantage of the intrinsic fuel flexibility, it is necessary that the material used for the anode be stable in the combustible fuel. For SOFCs with yttria-stabilized zirconia YSZ electrolytes, the state-of-the-art anode is a mixture of YSZ with Ni, usually referred to as a ceramic–metallic cermet composite. Ni– YSZ cermets perform very well under some conditions but have several important limitations. First, the cermets are not stable to reoxidation. Second, Ni-based electrodes cannot be exposed to hydrocarbon fuels unless sufficient steam is present to prevent Ni from catalyzing the formation of carbon fibers, especially with hydrocarbons larger than methane. Carbon-fiber formation is a serious problem because it can destroy the electrode by loss of Ni due to metal dusting and by producing stresses within the electrode that can fracture the cell. Carbon formation in the presence of dry hydrocarbons can be avoided by replacing Ni with a metal that does not catalyze fiber formation, like Cu, but the thermal stability of alternative electrodes tends to be worse. Electrodes based on conducting ceramics could provide an almost ideal solution to the problems associated with Ni–YSZ cermets if the ceramic electrodes could provide comparably low electrode losses. Oxides would not be expected to catalyze the formation of carbon fibers the way that Ni does because carbon dissolution is a key step in the formation of carbon fibers on Ni, and oxides do not dissolve carbon. Because many oxides have very high melting temperatures, the thermal stability of ceramic anodes is likely to be good. Unfortunately, the electrochemical performance of fuel cells based on ceramic anodes tends to be poor due to the fact that few oxides exhibit good electronic conductivities at the highly reducing conditions present in SOFC anodes. The oxides that do show reasonably good conductivity under these conditions exhibit poor ionic conductivity and catalytic activity. Anode performance comparable to that observed with Ni–YSZ composites has recently been observed in an electrode for which the functional layer was prepared by infiltration of La0.8Sr0.2Cr0.5Mn0.5O3 LSCM and catalytic amounts of Pd 0.5 wt % and ceria 5 wt % into porous YSZ. These composite electrodes have electronic conductivity due to LSCM, ionic conductivity due to the YSZ scaffold, and catalytic activity due to the Pd/ceria. An SOFC with this anode exhibited maximum power densities at 1073 K of 1.1 W/cm2 in H2 and 0.71 W/cm2 in methane. An especially interesting feature of these electrodes is their relatively high electronic conductivity. Although the intrinsic conductivity of LSCM in humidified H2 is only between 1 and 2 S/cm at 1173 K, the 45 wt % LSCM–YSZ composite, for which the LSCM phase was only 30 vol % of the solid, exhibited a conductivity greater than 0.1 S/cm at this temperature. A conductivity of 0.1 S/cm is sufficient for the functional layer of an electrode, so long as a conduction layer is present for lateral conduction. A random composite consisting of only 30 vol % of the conductive phase would normally have a conductivity much less than 20 times that of the conductive phase; however, the nonrandom nature of the composite leads to a much higher conductivity than expected. Under reducing conditions, oxides based on the doping of SrTiO3 exhibit some of the highest conductivities among ceramics. In particular, La0.3Sr0.7TiO3 LST has been reported to have conductivities greater than 20 S/cm at 1173 K under anode conditions. LST does not undergo solid state reactions with YSZ, even after calcination at 1823 K. Modest success has been reported with SOFC anodes that were conventional composites of LST and ceria, but the reported anode overpotentials and fuel cell power densities were clearly not comparable to that achieved with the best Ni–YSZ anodes. In a study from our laboratory, SOFC anodes were prepared by infiltrating a Pd/ceria catalyst into a porous scaffold that was itself a composite of LST and YSZ. Again, the performance of these electrodes was only modest. Furthermore, the conductivities of composites with 35 vol % LST were only 0.1 S/cm at 1173 in humidified H2, a factor of more than 200 lower than the value reported for bulk LST. Because structure seems to play a large role in the performance of the infiltrated-LSCM electrodes, and because LSCM is required primarily for its electronic conductivity, it is interesting to consider whether infiltrated-LST electrodes might perform even better, due to the higher electronic conductivity of LST. Furthermore, because LST is less reactive with YSZ than is LSCM, a wider range of processing conditions can be used for these materials. However, we will show in this paper that the structure of the infiltrated LST is very different from that of the LSCM, possibly because of weaker surface interactions. These structural differences lead to the compos* Electrochemical Society Active Member. z E-mail: [email protected] Journal of The Electrochemical Society, 155 11 B1179-B1183 2008 0013-4651/2008/155 11 /B1179/5/$23.00 © The Electrochemical Society B1179 Downloaded 31 Oct 2008 to 130.91.116.168. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp ites exhibiting a lower conductivity than expected. Also, while the performance of the LST-based electrodes is reasonable, it is not comparable to that achieved with LSCM.