Grain Boundary Engineering in Solid-State Ionic Conductors

Abstract

In recent years, lithium-based solid-state ionic conductors have begun to find renewed interest for energy storage applications. In part, this interest is driven by the enabling characteristics of the solid-state electrolyte separator; specifically the potential for safer and higher capacity battery cells. The ideal solid-state electrolyte should exhibit conductivity in excess of 10-4 S/cm at room temperature, negligible electronic conductivity, chemical stability against lithium metal, a wide electrochemical stability window > 5V, and high relative density upon sintering. To date, a number of material systems are being pursued in this regard, including but not limited to, various NASICON compositions [1,2], the garnet phase [3], and the Li3PS4based compositions [4]. For all of these cases, the as-sintered resistivity of the grain boundaries is routinely found to comprise a significant part of the total resistivity of the material as measured by electrochemical impedance spectroscopy (EIS). A better understanding of grain boundary structure and composition is required to truly engineer the performance of these materials in terms of consistent and improved grain boundary resistivity. Such efforts may also lead to improved sinterability. Current research work in our laboratory is focused on these topics, and this presentation will detail our efforts to gain a fundamental understanding of the grain boundary in the Lithium Aluminum Germanium Phosphate (LAGP) NASICON system and facilitate its modification in a controlled manner – a transformative approach to materials design and processing. In general, all reported solid-state ion conductors have impurities that typically reside at grain boundaries. However, their presence and levels within the parent solid are rarely discussed. Characterization by X-ray diffraction is almost exclusively used, and can only detect impurity levels above ~ 1% by volume at best. Furthermore, impurities may be amorphous and therefore undetectable by common diffraction methods. For many solid-state systems, dopants and secondary phases are added to facilitate sintering and enhance ionic conductivity [5, 6]. Our ongoing work demonstrates an ability to decouple the effects of doping from the effects of intended or unintentional secondary phases that may act as sintering aids. For example, a typical ionic conductor such as doped LAGP (x= 0.5) is chemically treated prior to sintering to remove all unintended impurities that result from material synthesis and crystallization. The extracted impurities are found to consist of insulating oxide and phosphate phases. Accordingly, the treated material analysis demonstrates a dramatically reduced level of these impurity phases, as shown in Figure 1. We find that sinterability of the LAGP material, thus purified of secondary phases, is just as poor as that of undoped Lithium Germanium Phosphate (LGP). We also find that the ionic conductivity of the material is markedly reduced compared to the untreated solid. The former runs contrary to commonly reported findings that aluminum doping dramatically enhances sinterability in this system. Our current work suggests that the enhanced sinterability of aluminum doped LGP may not be inherently a result of doping, but rather a result of unintended secondary phases formed at grain boundaries acting as sintering aids. Next, we demonstrate the effects of deliberate additions of secondary phases to thusly purified LAGP to characterize their individual contribution to both sinterability and grain boundary conductivity by EIS coupled with a comprehensive structure/composition analysis. Removal of undesired secondary phases substantially reduces the variation in grain boundary resistivity and lowers the average. This work opens the door for targeted modification of grain boundaries leading to improved overall ion conductor performance. We expect these results to translate into other material systems in energy conversion and storage. [1] B.E. Francisco, C.R. Stoldt and J.-C. M’Peko, J. Phys. Chem. C, 119 (2015) 16432–16442. [2] M. Perez-Estebanez, et al., Solid State Ionics266 (2014) 1-8. [3] E. Rangasamy, J. Wolfenstine, J. Sakamoto, Solid State Ionics206 (2012) 28-32. [4] Liu, et al., J. Am. Chem. Soc.135 (2013) 975-978. [5] H. Aono, et al., Solid State Ionics 47 (1991) 257-264. [6] C.J. Leo, G.V. Subba Rao, and B.V.R. Chowdari, Solid State Ionics 159 (2003) 357-367. Figure 1