Journal of Power Sources 128 (2004) 45–53 Engineered cathodes for high performance SOFCs R.E. Williford ∗ , P. Singh Pacific Northwest National Laboratory, PO Box 999, Mail Stop K2-44, Richland, WA 99352, USA Received 21 August 2003; accepted 15 September 2003 Abstract Computational design analysis of a high performance cathode is a cost-effective means of exploring new microstructure and material options for solid oxide fuel cells. A two-layered porous cathode design has been developed that includes a thinner layer with smaller grain diameters at the cathode/electrolyte interface followed by a relatively thicker outer layer with larger grains at the electrode/oxidant interface. Results are presented for the determination of spatially dependent current generation distributions, assessment of the importance of concentration polarization, and sensitivity to measurable microstructural variables. Estimates of the electrode performance in air at 700 ◦ C indicate that performance approaching 3.1 A/cm 2 at 0.078 V is theoretically possible. The limitations of the model are described, along with efforts needed to verify and refine the predictions. The feasibility of fabricating the electrode configuration is also discussed. © 2003 Published by Elsevier B.V. Keywords: Solid oxide fuel cells; Cathodes; Microstructure 1. Introduction A solid oxide fuel cell (SOFC) is composed of a dense electrolyte sandwiched between porous electrodes. The key electrochemical reactions occur mostly on the surfaces at the electrolyte/electrode interfaces, thus enabling the harvesting of electrons in an electrical circuit to produce useful power. The dense electrolyte conducts oxygen ions from the cath- ode to the anode, but prevents mixing of the fuel and oxidant in the gas phase, where electrons cannot be harvested. The porous electrodes permit passage of the gases to the elec- trolyte/electrode interfaces where the reactions occur. A rel- atively thick porous anode is often used to provide structural support for the assembly. The cathode is often relatively thin to minimize its contribution to the overpotential by polar- ization losses. It is generally recognized that a significant portion of these polarization losses originate in the cathode. Consequently, much attention has recently been focused on improving cathode performance through two means: im- proved materials and improved microstructural designs. Im- proved materials include mixed ionic electronic conductors (MIECs), which essentially increase the surface area avail- able for conversion of gaseous oxygen molecules to oxygen ions. Such a material has been developed in our laboratory [1], and was employed in this work. The present paper fo- ∗ Corresponding author. Tel.: +1-509-375-2956; fax: +1-509-375-2186. E-mail address: rick.williford@pnl.gov (R.E. Williford). cuses on methods to improve the microstructural design of the cathode. An objective of this work is to design a solid oxide fuel cell (SOFC) cathode exhibiting a low area specific resistance (ASR = 0.1  cm 2 ) and capable of high current output. Such an objective can be attained best by a coupled experimental-modeling approach: the modeling helps to guide the experiments and the experiments provide data for fitting the parameters of the model. Since modeling is often cheaper than a long series of Edisonian experiments, costs are often minimized with this approach. This paper de- scribes modeling efforts and includes selected experimental data, which are reported in more detail elsewhere [1]. An initial step was to review the existing models in the literature, and select an approach that was both pragmatic and thorough, in addition to exhibiting direct linkages with the experimental data. We found four categories of models, which are described below in terms of complexity and in- put data requirements. The following paragraphs are not an exhaustive review, and contain only representative examples of each model type. In the first category are detailed models by Svensson et al. [2]. This model treats the classic three phase bound- ary (TPB) problem explicitly by addressing the individual mechanisms involved (surface adsorption, dissociation, elec- tronic exchange, surface diffusion). Non-linear second order differential equations are derived and solved numerically, with the proper boundary conditions (six are required). Al- though the model certainly contains enough technical depth, 0378-7753/$ – see front matter © 2003 Published by Elsevier B.V. doi:10.1016/j.powsour.2003.09.055