Correlation between the Coherence Length and Ionic Conductivity in LiNbOCl<sub>4</sub> via the Anion Stoichiometry
Jon A. Newnham, Jędrzej Kondek, Johannes Härtel, Carolin Rosenbach, Cheng Li, Vasiliki Faka, Lara M. Gronych, Dana Glikman, Fabian Schreiner, Domenik D. Wind, Björn Braunschweig, Michael Ryan Hansen, Wolfgang G. Zeier
Abstract
LiNbOCl 4 is a recently reported material with high Li + conductivities of ∼10 mS·cm –1 at room temperature. Here, we explore how changing the anion ratio and the Li + content in the Li 1– x NbO 1– x Cl 4+ x series (−0.4 ≤ x ≤ 0.2) affects the ionic conductivity of the material. In doing so, we find that the maximum coherence length and ionic conductivity of LiNbOCl 4 are highly dependent on the O 2– /Cl – anion ratio in the material. Specifically, we show that, while an amorphous phase fraction of LiNbOCl 4 remains constant throughout the substitution series, any excess of O 2– results in a rapid decrease in the maximum coherence length of the crystaline fraction in each sample. Through a combination of diffraction and spectroscopic techniques, we show that this occurs because the O 2– anions cannot exist on the terminal sites of the [NbOCl 4 ] ∞ – chains in the material, even when it is made with an excess of O 2– resulting in a shortening of those chains. In contrast, it was observed that Cl – can occupy the bridging sites resulting in a dependence of the coherence length to the anion ratio. As such, the ionic conductivity of LiNbOCl 4 can be maximized by controlling the maximum coherence length in the material through the anion ratio. Notably, we achieved high ionic conductivities for LiNbOCl 4 consistent with literature reports only when the material was slightly Li + and O 2– deficient, suggesting that the literature samples may also have been off-stoichiometry. In addition, we highlight the features missing from the current structural models of LiNbOCl 4 including the presence of mixed Cl – /O 2– sites, even in the stoichiometric material, which were previously thought to not exist. Finally, we show that slightly reducing the Li + and O 2– contents in LiNbOCl 4 also translates to higher capacities when it is used as a catholyte in solid-state batteries. These findings show the importance of careful control of the stoichiometry in LiNbOCl 4 to optimize its properties and highlights the potential of LiNbOCl 4 for use as a catholyte in solid-state batteries.