Ceramic Processing Effects on the Multiscale Structure and Transport of Li7La3Zr2O12

Abstract

Solid-state electrolytes are a promising choice for next-generation batteries due to their increased safety, high conductivities approaching that of liquid electrolytes (10-2 S cm-1), and wide electrochemical stability windows (~6 V). While no single solid electrolyte has achieved all these feats, the garnet structured lithium lanthanum zirconium oxide (LLZO) has achieved many. One of the remaining challenges associated with LLZO is its order of magnitude lower conductivity as compared to state-of-the-art liquid electrolytes. The work performed in this thesis aims to address this challenge. To do so, a multipronged approach was taken to improve the understanding of the impacts of the microstructure, structure, and processing on the conductivity of Al-doped LLZO.

First, the effect of the microstructure on the conductivity was investigated as a function of the aluminum doping concentrations. It was shown that beyond the Al solubility limit (x=~0.40), resistive secondary phases formed at the grain boundaries, resulting in a grain boundary resistance increase from 17.6 to 41.2% for 0.25 and 0.55 mol Al. This work demonstrated that aluminum concentration has a significant effect on the microstructure and electrochemical properties of LLZO, improving the understanding of the linkages between processing, microstructure, and electrochemical properties.

Next, the mechanisms for Li ion conductivity were investigated. The metal-oxygen framework structure was observed to remain constant with variations in Al doping, resulting in little effect on the bulk conductivity. Instead, Li concentration, Al trapping of mobile defects, and Li-Li nearest neighbor interactions largely controlled the Al-doped LLZO bulk conductivity, resulting in decreases from 0.73 to 0.22 mS/cm as the Al concentration increased from 0.17 to 0.34 mol. These results differ from those of Ta-doped LLZO, where the framework structure and Li-Li site distances play large roles in controlling the conductivity. The increased understanding of the controlling factors of conductivity allows for greater ability to tailor the design and doping of the LLZO structure for improved conductivity.

Finally, a wide variety of processing variations were simultaneously investigated to determine their aggregate effects on LLZO conductivity. During calcination and sintering, Li losses were limited to ~0.6 mol while unexpected Al, Zr, and La losses occurred, likely to the MgO sintering boat. Cubic LLZO can be stabilized with a wide range of Li concentrations (6.08 mol to 7.61 mol Li) and Al concentrations (>0.06 mol), revealing the wide compositional flexibility of the cubic LLZO structure. This flexibility is extended to conductivity, where Li composition, Al composition, and phase purity do not have clear roles in its definition. The flexibility observed when a confluence of several parameters, such as composition, Li precursors, Li excess, and densification method, are tested and can lead to multiple successful processing routes for achieving ≥0.7 mS/cm bulk conductivity. This work expands the understanding of the relationship of synthesis and composition to the phase purity and conductivity of LLZO.

Efforts from this thesis have resulted in an improved understanding of the mechanisms of conductivity on a hierarchical scale, from structure to microstructure, as a function of compositional and processing variation. Advanced techniques have been used to determine the phase purity as a function of microstructure, the bulk conductivity mechanisms, and the compositional phase space of synthesized garnet electrolytes. The new knowledge gained through this thesis has opened the way for a diverse set of design strategies for improved conductivity performance in solid electrolytes.