Electrical Properties of Solid State Electrolytes and Their Grain Boundaries

Methods of energy production have been relying excessively on burning of fossil fuels for a sustained period of time; hence, development of new sources of energy production is becoming a critical issue to be addressed. As one of the most viable solutions to the issue, solid oxide fuel cells (SOFCs) have been intensively researched for more than a decade. There exists many obstacles to realizing the SOFC technology, and one of them is discovering the right material to serve as a stable fuel cell electrolyte with acceptable performance. In parallel, the research on secondary batteries is also heavily pursued lately as a method for efficient energy storage in portable electronic devices and transportations.;Polycrystalline ceramic ionic conductors are likely most preferably selected as the electrolyte materials of solid oxide fuel cells (SOFC) due to their chemical/mechanical stability (even for high temperature applications required for SOFC electrolytes) with reasonably simple and economical synthesis/processing methods. Currently, many polycrystalline ceramics such as the rare-earth doped ceria, yttria stabilized zirconia, doped barium zirconates, and doped strontium zirconates are popularly researched materials for such applications. However, the grain boundaries of these polycrystalline ceramics are often highly resistive; hence, it is important to understand the origin of this resistive behavior to further enhance the performance of these materials as solid electrolytes.;Results from numerous studies suggest that the formation of space charge layers due to the positively charged grain boundary core is responsible for the existence of highly resistive electrical grain boundaries. Conventionally, these electrical grain boundaries were expected to behave as a set of back-to-back Schottky barriers, and the ions transport through the grain boundaries by thermionic emission (TE). Hence, the TE model formed based on the Schottky barrier model with the thermionic emission theory was applied frequently to quantify the potential barriers.;The TE model allows calculation of the height of potential barrier using the ratio between resistivities of the bulk and the grain boundary. However, the barrier height computed based on the model increases with increasing temperature, which is a trend that cannot be understood logically. Additionally, the ionic current should exponentially depend on the applied voltage according to the thermionic emission theory, but instead, a power function dependence has been typically observed from the actual experimental I-V results. Hence, Kim and Lubomirsky have devised the linear diffusion model (LDM) to calculate the potential barrier height, which is described in detail in this dissertation. A set of Matlab simulations was performed using this newly devised linear diffusion model, and the resulting I-V characteristics have shown a very good match with the experimental I-Vs. From the simulation, a power (power of nth) relation of the grain boundary potential barrier height (psigb) in the I-V at high voltage region (where Ugb > 2Vth, the relation becomes I∝Vn) could be observed. Hence, if I-V plots are available from experimental results, it has become possible to extract the psigb values more reliably based on this model.;From the experimental I-V characteristic plots of various rare-earth (Y, La, Gd, Sm) doped ceria and Y-doped barium zirconate, we were able to extract the psigb using the LDM and the resistivity ratio model for comparison. Surprisingly, two models have produced different results, where psigb values extracted from the LDM were significantly lower than those from the resistivity ratio model. Interestingly the LDM based psi gb values from all rare-earth doped ceria were constant over the measurement temperatures, which are much more logical compared to the TE model based psi gb values that are clearly temperature dependent. On the other hand, an obvious temperature dependence was observed in the psigb determined for Y-doped BaZrO3 even from the LDM. This dependence can be attributed to a change in the charge carrier concentration for the proton conductors with respect to temperature. In the case of the proton conductor, the psi gb from the LDM was significantly lower than that from the resistivity ratio model. However, the overall resulting trends in the psigb determined from the experimental results indicate that the LDM enables a more accurate quantification of the electrical grain boundary properties.;Lastly, the electrical property of the sodium conducting glass-ceramic solid chalcogenide electrolyte was measured using the electrical impedance spectroscopy (EIS). To improve the safety as well as the stability of the batteries, replacing the widely used liquid-based electrolytes with much more stable solid electrolytes would be extremely beneficial. Sodium chalcogenide electrolytes are potential candidates for realizing batteries based on solid electrolytes, but it currently requires further improvement in its ionic conductivity and material stability for practical applications. The glass-ceramic electrolyte was synthesized by mixing Na2Se, Ga2Se3, and GeSe2 in a molar ratio of 25%, 15%, and 60% respectively. This glass-ceramic electrolyte produced a total ionic conductivity of ~10-5 at 25 °C, which is significantly higher (2 orders of magnitude) compared to a glass phase with a similar composition (15%-25%-60%). It is well known that the crystalline phase of this material is hydroscopic, but the chemical stability was significantly improved in this glass-ceramic phase, because the portions in their crystalline phase were protected by surrounding stable glass. This new discovery of a highly conductive glass-ceramic electrolyte shows promise in producing a solid battery electrolyte with high stability and suitable ionic conductivity for commercial usages.