Preparation of SOFC Anodes by Electrodeposition

Anodes for solid oxide fuel cells (SOFCs) have been prepared by electrodeposition of either Co or Ni into a layer of porous yttria-stabilized zirconia (YSZ), 60 μm thick. The YSZ, having 65% porosity, was prepared by tape casting with graphite pore formers and was attached to the dense YSZ electrolyte. After adding 10 vol % CeO2 by impregnation of aqueous solutions of CeNO3)3, followed by calcination at 723 K, the porous YSZ was made conductive by exposing it to n-butane at 1123 K to form a coating of carbon. As much as 40 vol % metal could be added to the porous layers, while the carbon could then be removed by exposing the anode to humidified H2 at SOFC operating temperatures. The ohmic losses in cells containing 40 vol % Co or 30 vol % Ni were unaffected by heating to 1173 K. Finally, a cell with 15 vol % Cu and 15 vol % Co was prepared by electrodeposition of Cu onto electrodeposited Co. No carbon formation was observed on the Cu–Co anode following exposure to dry methane at 1073 K. Comments © 2007 The Electrochemical Society. This article may be downloaded for personal use only. Any other use requires prior permission of the author and The Electrochemical Society. Reprinted from Journal of The Electrochemical Society, Volume 154, Issue 12, October 2007, pages B1270-B1275. Publisher URL: http://dx.doi.org/10.1149/1.2790280 This journal article is available at ScholarlyCommons: http://repository.upenn.edu/cbe_papers/98 Preparation of SOFC Anodes by Electrodeposition S.-W. Jung, J. M. Vohs,* and R. J. Gorte* Department of Chemical and Biomolecular Engineering, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA Anodes for solid oxide fuel cells SOFCs have been prepared by electrodeposition of either Co or Ni into a layer of porous yttria-stabilized zirconia YSZ , 60 m thick. The YSZ, having 65% porosity, was prepared by tape casting with graphite pore formers and was attached to the dense YSZ electrolyte. After adding 10 vol % CeO2 by impregnation of aqueous solutions of Ce NO3 3, followed by calcination at 723 K, the porous YSZ was made conductive by exposing it to n-butane at 1123 K to form a coating of carbon. As much as 40 vol % metal could be added to the porous layers, while the carbon could then be removed by exposing the anode to humidified H2 at SOFC operating temperatures. The ohmic losses in cells containing 40 vol % Co or 30 vol % Ni were unaffected by heating to 1173 K. Finally, a cell with 15 vol % Cu and 15 vol % Co was prepared by electrodeposition of Cu onto electrodeposited Co. No carbon formation was observed on the Cu–Co anode following exposure to dry methane at 1073 K. © 2007 The Electrochemical Society. DOI: 10.1149/1.2790280 All rights reserved. Manuscript submitted June 26, 2007; revised manuscript received September 6, 2007. Available electronically October 11, 2007. Solid oxide fuel cells SOFCs have received a great deal of attention in recent years because they offer the promise of high efficiency and fuel flexibility. These properties result primarily from the high operating temperatures that are required for diffusion of oxygen anions in the ceramic electrolyte, usually yttria-stabilized zirconia YSZ . The best SOFC electrodes are composites of YSZ with an electronically conductive material, usually Ni at the anode and Sr-doped LaMnO3 LSM at the cathode. 2 The YSZ in the composite electrodes plays a number of important roles, the most important being that it provides a good ion-conducting interface with the YSZ electrolyte, transporting oxygen ions into the electrodes and increasing the length of the three-phase boundary TPB . While electrodes based on Ni and LSM offer reasonably good performance, their widespread use is also due to the relative ease with which they can be fabricated. Because the reactivity of YSZ with both NiO and LSM is relatively low, Ni–YSZ and LSM–YSZ electrodes can be fabricated by simply mixing YSZ powder with either NiO or LSM, followed by high-temperature sintering. NiO–YSZ composites are easily reduced to form Ni–YSZ electrodes with good conductivity and porosity. Some alternatives to Ni, such as Co, cannot be produced in this way because Co cations from CoOx diffuse into YSZ at the temperatures required for sintering the YSZ component of the electrode. Because Co has similar catalyticreforming properties as that of Ni, a slightly higher melting temperature than Ni, and a lower affinity for sulfur than Ni, there are potential advantages to Co–YSZ electrodes compared to that of Ni–YSZ. Finally, Co and Ni have different alloying properties with other metals, so that Co composites offer new opportunities for mixed metal electrodes. Recent studies have suggested that Cu–Co anodes exhibit good thermal stability and a high tolerance against carbon formation because Cu and Co do not form solutions and Cu tends to segregate to the surface of Cu–Co mixtures due to its lower surface energy. In order to produce composite electrodes with different compositions and microstructures, our laboratory has investigated an alternative fabrication method that does not require cofiring the components. For SOFC with YSZ electrolytes, the method involves synthesizing porous YSZ layers on either side of a dense YSZ electrolyte layer, followed by addition of the conducting and catalytic components by impregnation of soluble salts into the electrode layer. By sintering the porous YSZ together with the dense YSZ electrolyte, prior to the addition of the infiltrated salts, the treatment temperatures for the various components of the cermet can be different, thus avoiding the problems associated with solid-state reactions. The impregnation method has been used to make both anodes and cathodes with a wide range of compositions. In addition to avoiding solid-state reactions, composites formed by impregnation have the additional advantage of forming nonrandom structures in which the impregnated phases coat the walls of the YSZ pores. This structure results in composites that have a thermal expansion match close to that of the YSZ backbone and good conductivities at loadings of the conductive phase that are well below the normal percolation threshold for random media. One drawback of the impregnation method is the difficulty of fabrication. While porous-dense, YSZ structures can be produced by inexpensive tape-casting methods, the impregnation procedure is tedious due to the requirement of using multiple steps. For example, to achieve sufficient conductivity and stability for LSM–YSZ composites by impregnation, it is necessary to fill a minimum of 15% of the pore volume with LSM. Based on studies of Cu anodes, even higher loadings appear to be required for metals. However, the maximum amount of the electronic conductor that can be added in a single impregnation step is limited by the cation content of the solution volume that fills the pores. In the case of LSM prepared using nitrate salts, it was necessary to use multiple impregnation steps, even when using molten salts with essentially no water, due to the large volume of the nitrate anions. A higher concentration of metal could be obtained using solutions with less bulky anions, such as Cl−, but removal of these anions during subsequent calcinations would be difficult. The use of nanoparticle solutions avoids the need for charge-neutralizing anions, but the concentrations of metal ions in nanoparticle solutions tend to be low compared to salt solutions. A potentially simple method for adding large amounts of metal to the porous electrode in a single step involves electrodeposition. In past work from our laboratory, bimetallic composites have been prepared by electrodeposition of Cu into highly porous, 1.2 mm thick Ni–YSZ cermets and both Cr and Co into thin Cu–YSZ composites. Because electrodeposition requires that the substrate onto which the metals are to be deposited be conductive, past work from our laboratory required a pre-existing ceramic–metallic cermet composite. Because it would be desirable to electrodeposit directly into the porous ceramic, we set out to determine whether the substrate could be made conductive through the use of carbon coatings. In this paper, we demonstrate the synthesis of Ni and Co cermets by electrodeposition into carbon-coated, porous YSZ. While we found it difficult to electroplate Cu in this way, we show that Cu can be plated onto an electroplated-Co layer so as to produce a composite that is stable against the formation of carbon when exposed to hydrocarbon fuels at high temperatures.