Tracking Dendrites and Solid Electrolyte Interphase Formation in Composite Electrolyte using Solid State NMR Spectroscopy

Solid-state lithium batteries have garnered significant attention in recent years as they are a promising technology with immense potential for development with diverse applications. Utilization of solid electrolytes is particularly appealing as it enables the use of lithium metal anodes and thus offer superior energy density with enhanced safety compared to traditional liquid electrolytes. Among the different types of solid-state batteries, polymer-based electrolytes are attractive due to their flexibility and non-flammable nature [1]. However, the practical implementation of polymer electrolytes in rechargeable batteries for high-energy applications faces challenges. These include limited room temperature Li-ion conductivity and formation of lithium dendrites at high current densities. To address these issues, one potential approach is to incorporate solid-state ceramic particles into the polymer matrix. However, there is still a limited understanding as to how ceramics incorporation impacts dendrite formation and propagation, as well as the composition and properties of the solid electrolyte interphase (SEI). The SEI plays a crucial role in the battery chemistry and gaining insights into its atomic-level structure and composition has the potential to transform the development of composites, enhancing their ability to suppress dendrite formation [2]. Unfortunately, there is a shortage of analytical tools capable of precisely identifying the SEI's chemical constituents at the atomic scale, and even more so, of understanding how these constituents affect ionic transport across the SEI. Here we introduce an innovative approach which allows to (i) quantify dendrites formation, (ii) determine SEI composition and its properties which allows us to (iii) determine the dendrite’s propagation path within the composite electrolyte.<br/>Using Li NMR spectroscopy, we successfully quantified the formation of dendrites in the cycled composites. Remarkably, we find a consistent rise in dendrite formation by increasing ceramic content up to 40 wt. %, when using Li_1.5Al_0.5Ge_1.5(PO₄)₃ (LAGP) and Li_6.4Al_0.2La₃Zr₂O₁₂ (LLZO) as ceramic fillers in polyethylene oxide. To determine the effect of ceramic content on dendrites formation, we employed dynamic nuclear polarization (DNP) [3], a method in which the high electron spin polarization is used to increase NMR sensitivity, to selectively enhance the SEI signal [4]. We make use of the inherent conduction electrons of the dendrites for DNP, enabling us to identify the chemical components and structure of the SEI, as well as the SEI permeability to Li-ions. We determined the chemical components of the SEI formed on dendrites in a pure polymer electrolyte and with the addition of varying content of LAGP and LLZO. Surprisingly, we found that SEI structures significantly differ not just with varying ceramic types but also with different ceramic contents, suggesting the nature of possible interaction of the dendrites with ceramic and polymer matrix. Furthermore, analysis of the SEI composition allows us to trace the dendrite propagation within the composite. We showed that in composites with 40 wt. % LAGP, dendrites chemically react with LAGP, hindering their growth towards the opposite electrode and impacting battery lifespan. Conversely, with 40 wt. % LLZO, dendrites are physically blocked without any chemical reaction. In summary, our study offers valuable insights into the SEI's composition, structure, and its correlation with dendrites and ceramic components. The approach can be used to identify the optimal ceramic material to reduce dendrite formation and enhance SEI's Li-ion permeability for improved battery performance.<br/><br/><u>References</u><br/>1. Yu et al<i>. Energy Storage Materials</i> <b>2021</b>, <i>34</i>, 282–300.<br/>2. E. Peled and S. Menkin, <i>J. Electrochem. Soc,</i> <b>2017</b>, <i>164</i>, 1703-1719.<br/>3. Rossini et al <i>Acc. Chem. Res</i>. <b>2013</b>, <i>46</i>, 1942.<br/>4. Hope et al. <i>Nat. Commun</i>, <b>2020</b>, 11, 2224.