Jon Ihlefeld1 2 William Meier2 Harlan Brown-Shaklee2 Emily Gurniak2 Andrew Kitahara2 Daniel Drury2 Mia Blea-Kirby2 Leo Small2 Mark Rodriguez2 Bonnie McKenzie2 Farid El Gabaly2 Erik Spoerke2 Anthony McDaniel2

1, University of Virginia, Charlottesville, Virginia, United States
2, Sandia National Laboratories, Albuquerque, New Mexico, United States

Ionic conductivities of solid state fast ion conductors are typically reported as three values for a given temperature: the bulk conductivity, grain boundary conductivity, and the total ionic conductivity, as deduced from electrochemical impedance spectroscopy measurements and fits. In this presentation, we will discuss two factors that greatly affect total ionic conductivity, that can be difficult to discern from conventional impedance spectroscopy measurements: 1) Nano-scale grain size scaling and 2) Porosity. The systems with which these results will be discussed include thin films of Na1+xZr2SixP3-xO12 (sodium super ionic conductor, NaSICon) and LiZr2P3O12 (lithium super ionic conductor, LiSiCon) and Li5La3Ta2O12 (LLTO) ceramics. We will show that when film thicknesses are sub-micron, discerning individual EIS features associated with grain and grain boundary effects is difficult, if not impossible. Further, as grain sizes scale into the nanoscale, unexpected trends in the composition dependence of ionic conductivity in NaSICon will be revealed. Typically, as the silicon content increases toward x=0.2, the ionic conductivity increases. In chemically-derived thin films, however, the competition of grain size scaling and concomitant increase in grain boundary volume and the increased ionic conductivity with silicon content results in a compositionally-dependent peak in room temperature ionic conductivity at a composition of x=0.25. This study demonstrates the significant role that grain boundaries can have on overall ionic conduction properties. Finally, we will show how phase control is vital in the processing of LLTO ceramics and that through proper batching and processing controls, ceramics with densities greater than 98% are possible without the aid of sintering pressure. The resulting phase-pure and dense ceramics possess total room temperature ionic conductivities of 2x10-5 S/cm. Further, we will show that the high activation energy of LLTO makes it competitive with the zirconium garnet counterpart at elevated temperatures.