No abstract

All-solid-state batteries including a garnet ceramic as electrolyte are potential candidates to replace the currently used Li-ion technology, as they offer safer operation and higher energy storage performances. However, the development of ceramic electrolyte batteries faces several challenges at the electrode/electrolyte interfaces, which need to withstand high current densities to enable competing C-rates. In this work, we investigate the limits of the anode/electrolyte interface in a full cell that includes a Li-metal anode, LiFePO cathode, and garnet ceramic electrolyte. The addition of a liquid interfacial layer between the cathode and the ceramic electrolyte is found to be a prerequisite to achieve low interfacial resistance and to enable full use of the active material contained in the porous electrode. Reproducible and constant discharge capacities are extracted from the cathode active material during the first 20 cycles, revealing high efficiency of the garnet as electrolyte and the interfaces, but prolonged cycling leads to abrupt cell failure. By using a combination of structural and chemical characterization techniques, such as SEM and solid-state NMR, as well as electrochemical and impedance spectroscopy, it is demonstrated that a sudden impedance drop occurs in the cell due to the formation of metallic Li and its propagation within the ceramic electrolyte. This degradation process is originated at the interface between the Li-metal anode and the ceramic electrolyte layer and leads to electromechanical failure and cell short-circuit. Improvement of the performances is observed when cycling the full cell at 55 °C, as the Li-metal softening favors the interfacial contact. Various degradation mechanisms are proposed to explain this behavior.

Ceramic−polymer solid electrolytes, combined with Li metal anodes, hold the promise for safer and more energetically dense battery technologies, as long as key interfacial challenges are fully understood and solved. Here, we investigate a garnet−PEO(LiTFSI) composite electrolyte system, the garnet filler being Li 6.55 Ga 0.15 La 3 Zr 2 O 12 (LLZO) microparticles. A "soft" mechanical milling process ensures good miscibility between the garnet and polymer phases over a wide range of volume fraction (up to 70 vol % garnet). Excellent degree of structural and chemical homogeneity is achieved without degradation nor segregation, even at the local level, as confirmed by solid-state NMR spectroscopy, electron microscopy and gel permeation chromatography. The total Li-ion conductivity of the composites is governed by the polymer matrix, as a consequence of the high interfacial resistance (∼10 4 Ω cm 2 ) between the garnet particles and the PEO(LiTFSI) matrix. However, by using 7 Li NMR 2D exchange spectroscopy (ESXY) in the solid state, it is shown that Li ions can locally exchange between the garnet surfaces to the surrounding polymer chains. This dynamic transfer phenomenon, occurring within the composite, seems to play a key role in kinetically stabilizing the interface with Li metal electrode, as observed from galvanostatic cycling and EIS experiments. Comparison of a garnet-free PEO electrolyte with a PEO−garnet (10 vol %) composite shows key performance improvements in the latter: although the Li-ion conductivity at 70 °C slightly decreases from 7.0 × 10 −4 S cm −1 , for PEO-LiTFSI, to 4.5 × 10 −4 S cm −1 for 10 vol % LLZO, the composite shows up to 1 order of magnitude lower interfacial resistance with Li metal electrode (33 vs 300 Ω cm 2 ), stable Li electrodeposition, and no dendrite formation. In contrast to previously believed, it is demonstrated that these improvements are not related to a change of the mechanical behavior but rather to a structural reorganization in the composite followed by local ion dynamics effects at the vicinity of the Li metal interface.

Solid state electrolytes could address the current safety concerns of lithium-ion batteries as well as provide higher electrochemical stability and energy density.

No abstract