Abstract
As society has grown to demand more portable power in segments including, but not limited to, portable computing, communications, power tools, transportation, and grid-scale energy storage, lithium ion batteries have become functionally omnipresent due to their high energy density and long cycle life. The vast majority of these batteries are constructed with a polymer separator that is impregnated with a liquid electrolyte to prevent contact of the electrodes and facilitate the motion of lithium across the battery. While this construction offers an enticing blend of manufacturability, performance, and cost, the liquid electrolytes are also a safety hazard and act as one of the limiting factors for increases in energy density. Generally, the organic liquid electrolytes that are used are highly reactive and do not cope well with high temperatures, which can lead to fires that are difficult and dangerous to extinguish. Additionally, the stability of the electrolyte against electrode materials with high electrochemical potentials is limited, which prevents the integration of new electrode materials that could drive increased energy density. Solid state lithium ion batteries make use of solid electrolytes that can be less reactive, more heat tolerant, and allow for the engineering of interface phases to enable stability against high energy electrodes. One key challenge to the commercial implementation of solid state lithium ion batteries is the low ion conductivity relative to liquid electrolytes, which prevents current solid state batteries from being used in applications with high power demands. A number of ceramic materials have been identified as solid state electrolyte candidates due to their high lithium ion conductivity in the bulk but exhibit low total ionic conductivity in polycrystalline forms due to high lithium ion impedance at the grain boundaries.
It is known that grain boundary energy varies widely with misorientation due to different structural, chemical, and electrical configurations. As such, it is expected that the boundaries separating differently oriented grains might also have different ion conducting behaviors. This has been supported by a number of theoretical studies and has been observed indirectly in experimental work but has not been quantitatively experimentally confirmed. Traditional ion conductivity measurements observe an ensemble response of all grain boundaries in the conduction pathway of a polycrystalline sample, removing any possibility of understanding the individual grain boundary behaviors. This dissertation seeks to develop methods for the quantitative measurement of ionic conduction across single grain boundaries via the epitaxial growth of ion conducting films on substrates with well characterized grain boundaries. To this end, this dissertation includes studies on the following subjects to enable the deposition and characterization of lithium ion conducting grain boundaries: 1) generation of Li3xLa1/3-xTaO3 bulk ceramics for characterization and for use as pulsed laser deposition targets; 2) the effects of deposition parameters on the structure, stoichiometry, and ionic conductivity of epitaxial Li3xLa1/3-xTaO3 films on single crystalline SrTiO3 substrates; 3) the ionic conductivity of single grain boundaries in Li3xLa1/3-xTaO3 films grown on bicrystal SrTiO3 substrates; and 4) the synthesis of polycrystalline SrTiO3 with an average grain size on the order of several hundred microns to enable combinatorial substrate epitaxy of Li3xLa1/3-xTaO3.
Trials on the synthesis of Li3xLa1/3-xTaO3 ceramics found that the ion conductivity was modified as a function of the lithium ion concentration in the sample due to the formation of a secondary LaTaO4 phase, driven by the non-stoichiometry of lithium in the sample. Sintering conditions that reliably resulted in dense, polycrystalline Li3xLa1/3-xTaO3 were determined and pulsed laser deposition targets were synthesized. During pulsed laser deposition, the background gas atmosphere was determined to change the transfer of lithium from the target to the substrate. Additionally, it was determined that the composition of the target changed significantly during deposition due to preferential ablation of lithium from the target. Commercially procured SrTiO3 bicrystals were used with epitaxial deposition conditions to generate thin films with grain boundaries that could be quantitatively measured using impedance spectroscopy. It was found that, of the available misorientations, 24°- and 6° tilt were not highly disruptive to ion conductivity but there was a significant increase in the impedance of the 45°-tilt grain boundary. A process for the generation of polycrystalline SrTiO3 with large grains was determined based on the requirements that the SrTiO3 remain insulating. It was found that air sintering followed by reduction annealing and a controlled reoxidation procedure resulted in grains with an average diameter greater than 100 μm without inducing conductive behavior. Finally, results from a related study on LiZr2P3O12 thin films are included, where it was determined that the ion conductivity was influenced by a number of competing effects depending on the annealing and synthesis procedures.
