4d/5d Transition Metals in Perovskite Cathode Materials
Abstract
Solid oxide fuel cells (SOFCs) have the potential to convert chemical energy to electrical energy in an efficient and environmentally friendly manner. The high operational temperatures of the first generation of SOFCs caused problems with thermal and chemical stability, while the most recent cell designs are based on metal supported SOFCs operated at intermediated temperatures, where the oxygenion conducting electrolyte may also be replaced by proton conducting oxides. Regardless of the type of electrolyte material used, new electrode materials are needed to achieve sufficiently low polarisation resistance at intermediate temperatures, and the development of alternative cathode materials with improved performance is therefore an important current research topic in this field.
The aim of this thesis was initially to develop cathode materials compatible with LaNbO4, which was reported as a promising proton conducting electrolyte. The first part of the work was devoted to find cathode materials based on the LaNbO4 – LaCoO3 system, which previously had been identified to be chemically compatible with LaNbO4. LaCoO3 has been extensively studied as a potential cathode material due to its excellent electrical transport properties, but excessive thermal expansion and poor chemical compatibility with electrolytes have proven to be problematic. The crystal structure, thermal expansion and electrical properties of materials in the solid solutions LaCo1-x NbxO3 and La1-x SrxCo0.8 Nb0.203 were characterised by X-ray diffraction (XRD), dilatometry and 4-point DC-conductivity. Nb-substituted LaCoO3 was prepared both by solid state synthesis and by an aqueous spray pyrolysis route. Nb-substitution was shown to reduce the crystal symmetry and lower both the thermal expansion and the electronic conductivity compared to pure LaCoO3. It was also confirmed that the average oxidation state of Co was depressed to charge compensate for the higher valence of Nb. This was especially interesting with regards to thermal expansion, since a high fraction of Co3+ is known to cause a high, non-linear thermal expansion.
The findings from studying the LaCo1-xNbxO3 system motivated further attempts to control the oxidation state of the host B-cation in perovskite systems by doping with 4d or 5d transition metals. The approach was further pursued in the ternary LaNiO3 – LaCoO3 – LaNbO4 system. Ni was introduced to lower the Co-content in order to suppress the high thermal expansion caused by Co3+. Two series of materials was designed in the LaCoxNiyNbzO3 system, where one combined isovalent oxidation states of Co and Ni, and the other series combined a mixed valence state of Ni and Co. The crystal structure, thermal expansion, electronic conductivity and thermopower were characterised as a function of composition in both series of materials. Temperature programmed reduction (TPR) confirmed a stabilisation of the materials in reducing conditions due to Nb-substitution. The materials with mixed oxidation state on the B-site possessed superior electrical transport properties compared to solid solutions with Ni/Co in mainly isovalent oxidation states. Promising electronic conductivity could be obtained in materials, accompanied by reduced thermal expansion and increased chemical stability. Finally, a related effect of co-doping with Mo6+ and Ni2+ was investigated in the LaCo2/3Ni1/3O3 – LaNi3/4Mo1/4O3 system. Further optimisation of the LaCoxNiyNbzO3 and LaCoxNiyMozO3 systems was attempted by Sr-substitution on the A-site in order to introduce oxygen vacancies and possibly induce oxygen ionic conductivity. Initial attempts of Sr-substitution were promising for materials with low Ni content, while formation of NiO and La2-xSrxNiO 4+δ were identified in materials with higher Ni content due to the low thermal stability of LaNiO3 at elevated temperatures.
La1-xSrxCo0.2M0.6Zr0.2O 3-δ (M = Mn, Fe) were the final systems investigated. Mn and Fe are known to be more stable in higher oxidation states than Co and Ni, and Zr 4+ was chosen to replace Nb5+ or Mo6+ in order to compensate for the higher oxidation state of Fe/Mn. Single phase perovskite materials were obtained for all the compositions in the La1-xSrxCo0.2Fe0.6Zr0.2O3-δ series. It was demonstrated that both the electronic conductivity and the oxygen vacancy concentration could be optimised through adjustment of the Sr-content. The electronic conductivity reached a maximum for x = 0.5, while at higher Sr-substitution levels the electric conductivity was reduced and onset of chemical expansion was observed due to an increasing oxygen vacancy concentration. Finally, single phase materials could not be prepared in the series La1-xSrxCo0.2Mn0.6 Zr0.2O3-δ as La2Zr2O7, and/or SrZrO3 were formed, which could be rationalised by the high stability of Mn4+ in the main perovskite phase.
The most promising materials from each series described above were tested in symmetrical fuel cells with Ce1-xGdxO2-δ as an electrolyte using impedance spectroscopy. The impedance data showed that the materials, which were inferred to be essentially stoichiometric with respect to oxygen, demonstrated a relatively high area specific resistance despite an acceptable electronic conductivity. Excessive chemical impedance due to the lack of ionic conductivity was inferred for these materials. The symmetrical cell with a La0.5Sr0.5Co0.2Fe0.6Zr0.2O3-δ cathode demonstrated a promising cathode performance, and low polarisation resistance was obtained at intermediate temperatures.
This thesis has demonstrated that 4d and 5d substitutions in perovskite materials have the potential to stabilise the materials chemically, tailor the thermal expansion and tune the electronic properties; particularly in materials with a mixed valence state of the B-cations. Further optimisation of the composition may allow the development of cathode materials with the desired combination of both high electronic and ionic conductivity at intermediate temperatures.