| dc.description.abstract | Proton ceramic fuel cells based on an yttrium-doped barium zirconate electrolyte might pose as a viable option for future hydrogen applications. High resistances in the electrolyte however hampers the performance of the material. Therefore, a novel heterostructure engineering strategy is applied in order to investigate alternative solutions to reducing the resistances. This thesis aims to deepen our understanding of how heterostructures and their interfaces might affect the electrolyte resistances. The goal of the following approach is to circumvent proton trapping at acceptor dopant sites. A job-sharing model has been investigated where one phase “supplies” the protons while the other phase conducts them. The proton donor phase is acceptor doped and more acidic than the other phase, resulting in a net transfer of protons over to the conductive phase. The latter phase, free of trap sites and enriched in charge carriers, is then hypothesized to exhibit a conductivity increase. Modern computational methods allow for compositions to be investigated theoretically without having to perform tedious experiments. While experiments measure the actual effects of the system of interest, the results might be ambiguous as they often are a combination of several contributions. Theoretical computations on the other hand might give insight into trends, often on the atomistic scale, else impossible to separate from other effects experimentally. Combining the two methods, in depth knowledge and understanding of the system is acquirable. The model system, an alternating multi layered BaZr1-xYxO3/SrTi1-xScxO3 film was fabricated by pulsed laser deposition onto a (100) MgO substrate. The thin (60 nm) films were by X-ray diffraction confirmed to be grown epitaxially, made possible by a good lattice match between the substrate and the films. Impedance spectroscopy measurements of the BaZrO3/SrTi0.9Sc0.1O3 film showed a conductivity of 0.27 mS/cm, comparable to the conductivity of the reference BaZr0.9Y0.1O3 film (0.28 mS/cm). The activation energy of the heterostructure was measured to 0.45 eV, lower than for the reference BaZrO3 and BaZr0.9Y0.1O3 films and in the range of the proton migration activation energy. The SrTi0.9Sc0.1O3 had a larger activation energy of 0.64 eV, expected for oxide ion conduction mechanism. VII When going to dry from humid atmosphere for the BaZr1-xYxO3 containing films was a decrease in conductivity of 55 % to 65 %, attributed to a decrease in charge carrier (proton) concentrations which was further confirmed by a hydrogen isotope exchange. A slope of 0.143 was observed in the Arrhenius plot of the SrTi0.9Sc0.1O3 film, indicating that ionic defects dominate concentration-wise whilst minority holes contribute significantly to the conductivity in the measured pO2 range. First principles calculations of a BaZr0.984Y0.016O3H0.016 4 by 4 supercell showed the energy difference a at set of Y-H + distances. The trapping energy of the protons was calculated as a function of in-plane strain and was found to increase with more negative (compressive) strain. Additionally, strain in general decreases the long-range mobility of protons in the yttrium dopants because of an energy barrier, larger than or equal to the trapping energy. Removing -0.5 % of strain was found to result in an activation energy decrease, increasing the conductivity by a factor of two. The calculated trapping energy change for different levels of strain agree well with experimental activation energies from literature [1]. Comparisons of measurements between reference films and the BaZrO3/SrTi0.9Sc0.1O3 heterostructure supports the job-sharing model. By assuming that the space charge region of the BaZrO3 and SrTi0.9Sc0.1O3 does not affect the conductivity of the latter, a conductivity increase of a factor of 14 was calculated for the BaZrO3 layer of the heterostructure compared to the reference BaZrO3 film. | eng |