Experimental Optimization of Ni-GDC Microstructure with Various Phase Fractions for Solid Oxide Fuel Cell Anode

Abstract

The design of electrode microstructure is important to enhance performance and stability of solid oxide fuel cells (SOFCs). The most common material used for the fuel electrode is nickel-yttria stabilized zirconia (Ni-YSZ) composite, in which electron, ion and gas conduct in the metallic Ni phase, the ceramic YSZ phase and the pores, respectively. The electrochemical reaction therefore takes place at the triple-phase-boundary (TPB), hence the increase in TPB density improves electrode performance. Recently, the ceramic materials, which exhibits mixed ionic-electronic conductivity (MIEC) properties, are considered to be beneficial solution for the SOFC electrodes. They can greatly enhance the available reaction sites by introducing surface reactions on the double phase boundaries (DPB) between MIEC phase and pore. One of the promising materials is considered to be gadolinium doped ceria (GDC). GDC exhibits higher ionic conductivity compared to YSZ and has MIEC property in the reducing atmosphere. In the case of Ni-GDC composites, the electrochemical performance is supported by both DPB and TPB reactions. In general, the active DPB density increases with the share of GDC phase in the composite. On the other hand, the active TPB density achieves maximum value when the share of both solid phases is balanced. The optimal electrode design has to be determined by the ratio between DPB and TPB reactions and their kinetics. In the present study, electrolyte supported SOFC with Ni-Gd0.1Ce0.9O2-δ anodes with various phase fractions were fabricated and their polarization characteristics were experimentally investigated. The ratio of Ni and GDC was controlled by the initial powder composition, and the porosity was varied by isostatically pressing the anode after screen printing. Initially, NiO (AGC Seimi) and GDC (ShinEtsu) powders were mixed to achieve the target volume fractions of Ni:GDC = 70:30, 60:40, 50:50, 40:60, 30:70 vol%. The anodes were then screen-printed on the YSZ discs and calcinated at 1350oC. In order to modify the porosity of the anodes, some of the cells undergoes isostatic pressing treatment at 200 MPa for 30 minutes prior to the calcination. The LSCF (lanthanum strontium cobalt ferrite by Fuel Cell Materials) cathode with GDC (ShinEtsu) barrier layer were screen-printed and sintered on the other side of the electrolyte. Current – voltage curves and electrochemical impedance spectra were measured at 500 – 800 oC. Additionally, the anodes with optimal phase composition were discharged for 100 hrs. 3D microstructural analysis was conducted with the focused ion beam scanning electron microscopy (FIB-SEM). The microstructural parameters, such as phase volume fractions, particle intercept length, TPB and DPB densities and tortuosity factors were quantified. By evaluating the connectivity of each phase, the total and active TPB and DBP densities were calculated. Quantified microstructural parameters were correlated with the polarization characteristics. Neither the DPB reactions nor the TPB reactions could individually explain the measured polarization resistances of the composite anodes. In the case of screen-printed anode samples without isostatic pressing, the high porosity (ca. 52% for all of the samples) lead to low connectivity of the solid phases. The optimal composition of the cermet was found to be Ni:GDC = 0.4:0.6 vol%. Introducing isostatic pressing, improvements in both ohmic and polarization resistances are observed. The decrease in ohmic resistance is mainly attributed to the improved connectivity between the electrode and the electrolyte. The polarization resistance was influenced by the changes in active DPB and TPB densities. As the porosity is reduced and connectivities of the solid phases are enhanced by the isostatic pressing, not only the nominal values of active DPB and TPB densities increased, but also the ratio between them has changed. This effect is most significant for Ni:GDC = 0.3:0.7 vol% sample, which exhibits the smallest polarization resistance among all the investigated electrodes. The discharge testing showed the improved stability of the cells with the decreased porosity. It is therefore concluded that not only the ratio between Ni and GDC but also porosity is a very important parameter for the performance and stability of the anode. Figure 1