Next-Generation Anodes for High Performing, Redox- and Sulfur-Tolerant Solid Oxide Fuel Cells and Stacks

The worlds increasing energy demand calls for transitioning to renewable energy sources to reduce greenhouse gas emissions and limit global warming. This transition is only possible with technologies that can convert and store the energy. A key enabling technology is Solid Oxide Cells, which are high-temperature electrochemical cells that convert chemical energy to electric energy with high electrical efficiencies in fuel cell mode, or reverse in electrolysis mode. The state-of-the-art solid oxide fuel cell (SOFC) anodes are based on Ni-ceramic composites. These electrodes experience various degradation phenomena during operation, including Ni-coarsening, intolerance to fuel impurities, and low stability towards redox cycling, which all limit the lifetime of the cells. Hence, alternative anode materials that can mitigate these challenges are highly desirable. Within this field, anodes based on all-ceramic electronic conducting doped strontium titanates are especially interesting due to their structural stability and tolerance to fuel impurities. Their low performance compared to SoA can be improved by catalyst exsolution from the ceramic itself, alone or coupled with infiltration of a second, catalytically active phase.

In this study, anodes based on Ni:CGO infiltrated La_0.4Sr_0.4Ti_0.94Fe_0.03Ni_0.03 were evaluated on symmetric cell, full cell, and stack level as candidate materials for next-generation redox- and impurity tolerant SOFC anodes. Electrode optimization studies on symmetric cell level revealed no effect on the electrochemical performance when Fe was infiltrated together with the Ni:CGO electrocatalyst, as long as high operating temperature and dry fuel was used. However, Fe induced a performance decrease at lower temperatures and in more humid atmospheres. The size of the infiltrated Ni could be decreased by increasing the infiltration solution pH. This resulted in a better performance in terms of lower activation energies compared to using neutral Ni solutions. Superior performance, especially at lower operating temperatures of 650 °C, was achieved by high temperature pre-reduction of the cells in pure hydrogen at 900 °C. This effect relates not to the infiltrated catalysts, but to the LSFNT backbone due to increased Fe/Ni exsolution and possibly higher electronic conductivities.

Based on the symmetric cell studies, strategies for full cell integration and testing could be advised. Metal supported cells (MSCs) with Ni:CGO infiltrated LSFNT anodes were tested after in-situ anode reduction (and cathode sintering) and compared to SoA cells with a Ni/YSZ anode. A known challenge for these MSCs are their low activity for internal reforming of hydrocarbons. Since many SOFC systems use a pre-reformer that converts hydrocarbon fuels into a mixture of CO, H₂, and H₂O, the MSC and SoA cell were tested in such a mixture. However, the MSC performance remained low compared to SoA, which means that further optimization studies are necessary. The difference in cell performance between the two cell types was significantly lower in steam/hydrogen fuel.

One of the aims of this work was to develop a redox-tolerant anode. The redox tolerance was measured on full cell level through simulated failure of fuel supply (hydrogen) during galvanostatic durability testing. SoA cells with Ni/YSZ anodes showed the expected redox intolerance related to microstructural changes in the anode. In contrast, the redox tolerance of the MSCs depended on the fuel utilization, and superior redox tolerance was observed at lower fuel utilizations. Impedance data and post-test microstructural analysis revealed that the degradation on the MSC relates to corrosion of the metal support at high steam partial pressures and fracturing of the electrolyte rather than degradation of the LSFNTbased anode, which is itself redox tolerant.

The relevance of the infiltrated LSFNT anodes to industry was evaluated directly through integration into novel, large area (100 cm²) electrolyte supported cells, which were tested on stack level at the Swiss SOFC system manufacturer Hexis. Both Ni:CGO and FeNi:CGO infiltrated LSFNT anodes were evaluated. The cells were tested for their performance and tolerance to operation without a desulfurization unit. Electrochemical measurements showed similar initial performance and stability to a SoA cell with Ni/CGO composite anode. The results showed that cells with LSFNT-based anodes were tolerant to short-term breakdown of the desulfurization unit, whereas degradation was observed for the SoA cell. The long-term sulfur tolerance was high for both the SoA cell (after initial degradation) and the Ni:CGO infiltrated cells, but not the FeNi:CGO infiltrated cells. This is in contrast to the known synergistic effect of co-exsoluted Fe and Ni on the sulfur tolerance observed in literature, and was attributed a different behaviour of infiltrated compared to exsoluted particles. The cells with Ni:CGO infiltrated LSFNT anodes showed remarkable sulfur tolerance. Together with their high electrochemical performance without sulfur, similar to SoA, it makes them promising candidates for next generation sulfur tolerant anodes for cheaper and simpler SOFC systems.