Ceramic Energy: Material Trends in SOFC Systems
The accepted U.S. Department of Energy Solid State Energy Conversion Alliance (SECA) cost target for SOFCs to be commercially viable is $400/kW.1 Although some of the new low-temperature (500-700°C) electrolyte materials contain expensive raw materials, such as scandia or gallia, the largest cost reduction challenge is in the fuel reforming or "balance of plant," rather than the material cost in the SOFC stack. A potential breakthrough area is in sulfur-tolerant anodes that will operate with fuels requiring minimal refining. This is one case where a materials development could be an enabling technology that might dramatically reduce the cost of the SOFC system and push the state-of-the-art one step closer to the SECA target.
However, even in cathode and anode materials, some potential exists to increase SOFC performance and drive down costs. As new materials are developed, researchers will have even more options for creating commercially viable systems.
ElectrolytesThe state-of-the-art electrolyte material is yttria-stabilized zirconia (YSZ). This material has acceptable ionic conductivity in the 800-1000°C range but falls short in the more desirable 600-800°C range.2 YSZ is inexpensive, and its properties are well known. It is a very refractory material, typically requiring higher sintering temperatures than the other SOFC components. For this reason, YSZ electrolytes are typically derived from chemically co-precipitated powders, since solid-state processed ("shake and bake") material is unable to attain the high surface areas required to co-sinter with the cathode and anode material (1200-1500°C)3. This processing method obviously produces a more expensive powder than "shake and bake;" however, most planar-designed SOFCs use a very thin electrolyte layer (microns) and require a minimal amount of material.
Two materials that have better ionic conductivity than YSZ in the 600-800°C range are scandia-stabilized zirconia (ScSZ) and lanthanum strontium magnesium gallate (LSGM).4 Both of these materials contain expensive oxides-the 2004 market price of scandium oxide was around $4000/kg, while gallium oxide was $1000/kg-but they have much greater ionic conductivities than YSZ. The LSGM material has issues with Ga volatility, as well as chemical incompatibility with Ni anodes.2 One approach to this problem involves depositing a barrier layer of doped cerium oxide between the electrolyte and the electrode to protect the electrolyte from chemical reactions. In the past, doped ceria had been considered an electrolyte, but it was susceptible to reduction and electronic conductivity in reducing atmospheres near the anode.5
Cathode MaterialsThe classic SOFC cathode material has been lanthanum strontium manganite (LSM), which has good phase stability, a good thermal conductivity match with YSZ electrolytes and is well characterized in the SOFC system. The chief drawback of LSM is that its conductivity is low (200 S/cm) and primarily electronic.3 This limits the active areas (regions where oxygen catalysis, electronic and ionic conduction occur simultaneously) to those triple regions where LSM, YSZ and air are in contact.
New materials called mixed ionic-electronic conductors (MIEC) are being investigated as potential cathode materials for next-generation SOFCs. These materials can conduct a charge through both ionic and electronic motion. Lanthanum strontium ferrite (LSF) and lanthanum strontium ferrite cobaltite (LSFC) are among the more celebrated examples of these materials. The mixed conduction provides a larger active volume for the reduction of gaseous oxygen to the oxygen ion, oxygen ion conduction and electron conduction because the cathode serves a triple role as a catalyst, an electronic conductor and an oxygen ion conductor. This dramatically reduces the electrical resistance due to polarization overpotential at the cathode.2
However, these materials are not as well studied as the LSM material, and they possess phase-stability and interface-reaction issues that need to be addressed prior to commercialization. One of these issues is chromium poisoning. With the trend to lower operating temperatures, stainless steel is placed either in contact with or in a close proximity to the fuel cell stack from balance of plant components such as heat exchangers and air inlet piping. The chromium from the stainless steel may react with the LSF cathode during SOFC operation, causing a severe degradation in stack performance.6
Among the best oxide conductors considered for cathode operations are the perovskite cobaltites. These are MIECs with typical conductivity values above 1000 S/cm2 . Figure 1 shows a comparison of the conductivity of lanthanum strontium manganite with lanthanum strontium cobaltite. The chief drawback of the perovskite cobaltites is their high thermal expansion values (20 ppm/°C), which are a mismatch with electrolyte materials. Extensive research in Japan is being conducted on Sm5Sr5CoO3-x, which has a better thermal expansion mismatch than most other cobaltites while retaining high conductivity values.7
Finally, although cathode powders have been traditionally produced by chemical co-precipitation, there is a trend toward producing this material by the less expensive solid-state method, provided the surface area targets can be reached. Additionally, some materials may be plasma-sprayed, which requires a chemically pure feedstock while putting less emphasis on phase purity and requiring a lower surface area.
AnodesThe anode or fuel side electrode typically is composed of a cermet-containing YSZ and Ni metal. The Ni metal acts as a catalyst for the oxidation of the fuel. Among the problems faced with a Ni cermet anode is Ni metal coarsening (sintering)2,3 during use, as well as the buildup of carbon deposits on the surface (coking) during internal reforming of the fuel.8 Sintering reduces the surface area available for catalysis, while coking covers the active area with graphite deposits. Both of these cause the resistance (overpotential) to increase during operation.
Researchers at the University of Pennsylvania have shown that the use of copper (Cu)-based cermets dramatically reduces coking and loss of performance during service due to coking.8 Although one of the greatest advantages of the SOFC over other fuel cell types is the lack of fuel selectivity, the quality of the fuel remains an important parameter with SOFC materials. The presence of impurities such as sulfur in dirty fuels will quickly poison metallic Ni in the anode. Cu-based anodes show improved sulfur tolerance relative to Ni.
Recent activity has focused on the use of ceramic anode materials based on titanates, such as lanthanum strontium titanate.9 Titanates are stable in low oxygen partial pressure at the anode, and are also resistant to coking and attack by sulfur. At present, the conductivities are too low for these materials to stand alone as the anode without a Ni-based current collector. However, a number of new compositions are being investigated. If a material can be found that is stable in the reducing atmosphere at the anode, resistant to coarsening or degradation due to coking or fuel impurities, and has a large electronic conductivity, a massive effort will follow to successfully incorporate this material into SOFC designs.
Continuing AdvancesThe ceramic forming and processing challenges for SOFCs are considerable. Add these issues to the already complex undertaking of creating composite structures of dissimilar materials exposed to temperatures up to 800°C in a range of oxygen partial pressures, and it's easy to see how demanding the task engineering and commercializing SOFC materials can be. Nonetheless, a material breakthrough in ceramic anodes, MIECs or a new low-temperature conductor could vastly change the state-of-the-art in SOFCs, enabling them to approaching the target of $400/kW and moving them closer to commercialization.
For more information about SOFC materials, contact Trans-Tech Inc., 5520 Adamstown Rd., Adamstown, MD 21710; (301)-874-6422; fax (301)-695-7065; e-mail email@example.com; or visit http://trans-techinc.com/.
References1. SECA Core Technology Program website (http://www.seca.doe.gov).
2. Ivers-Tifee, Weber and Herbstritt, J. Eur. Ceram. Soc., Vol. 21, 2001, pp. 1805-1811.
3. Minh, Nguyen, J. Am. Ceram. Soc., Vol. 76, No. 3, 1993, pp. 563-588.
4. Huang, Tichy and Goodenough, JACerS, Vol. 81, No. 10, 1998, pp. 2565-2575.
5. Goodenough, J.B., Ann., Rev. Mater. Res., Vol. 33, 2003, pp. 91-128.
6. Jiang, Zhang and Zheng, J. Eur. Ceram. Soc., Vol. 22, 2002, pp. 361-373.
7. Ishihara, Honda, Shibayama, Minami, Nishiguchi and Amikura, J. Electrochem Soc., Vol. 145, 1998, p. 3177.
8. Kim, Lu, Worrell, Vohs and Gorte, J. Electrochem. Soc., Vol. 149, No. 3, 2002, pp. A247-A250.
9. Marina, O., SECA Core Technology Program Review Meeting, Albany N.Y., Sept. 30, 2003.