Oxygen electrodes for ceramic fuel cells with proton and oxide ion conducting electrolytes

Abstract

The overall aim of this work is to contribute to a better understanding of the reactions taking place at the oxygen electrode in proton ceramic fuel cells (PCFCs) and, moreover, to develop new materials with improved performance for this electrode. PCFCs and their cathode reactions are the main focus of the study, but these reactions are often running in parallel with reactions associated with other charge carriers that also need to be addressed. Most proton conducting ceramics exhibit also transport of oxide ions, and although small at intermediate temperatures, the relative contribution from partial oxide ion conductivity increases with temperature and eventually dominates at higher temperatures. Hence, characterization of the performance of a PCFC cathode may at higher temperatures in reality be affected by or even directly reflect the cathode reactions of an oxide ion conducting solid oxide fuel cell (SOFC) rather than that of a PCFC. The crossover between SOFCs and PCFCs with respect to the oxygen electrode reactions is emphasized in this work. The first manuscript presents status and challenges of PCFC research undertaken in Norway by the start of 2010. The work comprises manufacturing of single cells and cell stacking, focusing on the performance, the mechanical and thermal properties, as well as, the chemical stability of the different PCFC component materials. State-of-the-art cathode material at that time, La0.8Sr0.2MnO3 (LSM), showed a polarization resistance (Rp) of 30 Ωcm2 800°C on proton conducting Ca doped LaNbO4 electrolyte, revealing the necessity for a significant improvement in the cathode performance. New materials had to be found and their microstructural design optimized, based on the requirements specific for the proton conductor oxygen reaction. Reaction kinetics, with particular emphasis on the features specific for the PCFC oxygen electrode is investigated in manuscripts II, III and V. In manuscript IV, the experimental conditions are such that the SOFC reactions dominate the electrode process. The electrolyte/electrode interfacial exchange of protons instead of oxide ions distinguishes PCFCs from oxide ion conducting SOFCs and entails that water is formed on the oxygen side of the electrolyte. A major challenge with the PCFC cathode candidate materials studied so far is the confinement of the electrochemical process to pass the electrolyte / electrode / gas triple phase boundary (tpb), instead of utilizing the whole electrode area as for the best mixed conducting SOFC electrodes. The challenges related to tpbs as a bottleneck are addressed by microstructural improvements. Moreover, a novel material with simultaneous transport of electrons and protons is introduced that will enable also the PCFC cathode reactions to occur over the electrode surface, thereby extending the tpb reaction zone. The effect of water formation on the cathode reaction is studied in detail on a Pt model electrode. The results show higher reaction rates upon increased water vapor partial pressure, pH2O. Since the Pt electrode is rate limited by surface diffusion both under dry and wet conditions, the pH2O effect is explained by the formation of surface hydroxyls with high surface mobility relative to the adsorbed oxide ions which dominate under drier conditions. The presence of surface hydroxyls is confirmed by X-ray photoelectron spectroscopy. Water is looped in the oxygen reaction series, acting both as reactant and product. In manuscript V and in the results part of the thesis it is shown that ambient water vapor gives the same positive effect for the mixed conducting electrodes BaGd0.8La0.2Co2O6-δ (BGLC) and BaPrCo2O6- (BPC) when operated on a BaZr0.7Ce0.2Y0.1O3 (BZCY72) proton conducting electrolyte. At higher temperatures where BZCY72 is mainly oxide ion conducting, water vapor on the other has an adverse effect on the electrode reaction rate for the same mixed conducting electrodes. With the mixed oxide ion-p-type electron conductor La2NiO4+δ (LNiO) as electrode and La27.16W4.85O55.27, (La/W ≈ 5.6; LWO56) as electrolyte (manuscript IV), the electrode performance was independent of pH2O under conditions where oxide ion conductivity dominates in the electrolyte (above 700°C). Three well-established routes to improve the electrode microstructure were followed in this work; (i) addition of nano-sized catalysts by infiltration, (ii) improvement of the functional layer close to the electrolyte and (iii) manufacturing of composite electrodes by mixing electrode and electrolyte materials. The two first methods showed promising results: Addition of Pt nanoparticles in the LSM electrode lowered significantly the polarization resistance; from 260 to 40 Ωcm2 at 650°C. Characterization of the microstructure of BGLC and BPC electrodes showed that a fine-grained functional layer was successfully manufactured. The composite electrode approach did, however, not prove to enhance the performance of an electrode rate limited by surface reactions. The materials investigated in this work range from well-known pure electron conductors such as Pt and LSM, used first and formerly for the detailed characterization of the electrode reactions, via the promising mixed conducting candidate LNiO, to the novel mixed conducting double perovskites BGLC, BPC and their B-site iron-substituted variants BaGdCo1.8Fe0.2O6-δ (BGCF) and BaPrCo1.4Fe0.6O6-δ (BPCF). For Pt and LSM, high capacitance processes like surface diffusion is limiting the overall electrode reaction rate. For the mixed conducting electrodes LNiO and BGLC, the oxide ion transfer is shown to happen through the electrode interior. The latter also shows indications of partial bulk proton conductivity concluded based on the pH2O dependencies encountered for Rp and supported by hydration of the material at low temperatures with a hydration enthalpy of -50 kJ/mol. Bulk proton transport would facilitate the low temperature PCFC cathode reaction and widen the triple phase reaction zone improving the electrode performance. The behavior of these mixed conducting double perovskites, especially BGLC but possibly also BPC, with polarization resistances measured to 0.05 and 0.09 Ωcm2 at 650°C for BGLC and BPC, consequently gives indications of the first established mixed proton / electron conducting materials with sufficient electrochemical performance on a proton conducting electrolyte. To account correctly for mixed conductivity in the electrolyte is challenging when studying electrode reactions. In manuscript III and V, a model for the separation of the measured polarization resistance into the contributions from more than one charge carrier is developed. The model accounts also for the effect of parallel non-faradaic current during high temperature measurements under oxidizing conditions. The results of the modelling show that the measured polarization resistance for the system investigated here and reported above for 650°C is underestimated by approximately one order of magnitude. The same underestimation would apply to any other oxygen electrode measured on BZCY72 if the effect of electrolyte p-type partial conductivity was not properly addressed. In a running fuel cell or electrolyzer cell, the fuel-side reducing conditions are expected to induce a blocking layer for electronic conductivity in the electrolyte. The "true" polarization resistance will therefore be higher when the partial short circuit is absent. At lower temperatures, this effect of parallel non-faradaic current is less pronounced during half-cell electrode characterization. BGLC exhibits a total polarisation resistance for proton transport of only 10 Ωcm2 at 350°C, with an activation energy of 50 kJmol-1 ascribed mainly to the surface electrode reaction. Based on this, there is reason to believe that further improvements of the cathode performance can be achieved by enhanced microstructural processing, such as infiltration of BGLC in a BZCY backbone.