The Nature of Interfaces in All Solid-State Batteries

All solid-state battery (ASSB) systems using ceramic-based electrolytes have the potential to revolutionize the battery consumer market - from electric vehicles, consumer appliances and power tools, to miniaturize rechargeable cells on electronic chips - because of potential benefits in energy density, operable temperature range, and safety in comparison to traditional liquid electrolyte based systems. One of the most promising solid electrolytes to realize ASSBs are Li7La3Zr2O12 garnets and variants (LLZO). LLZO is related to the class of garnets known as gem stones (e.g. almandine) and is one of the most promising solid electrolytes with high Li-ion conductivities and superior stability. Preliminary tests to implement LLZO into ASSBs, however, suggest plenty of room for improvement. High interfacial resistances currently limit the applicability of LLZO. In order to improve LLZO a profound understanding of the key factors related to its properties is highly needed. In previous studies polycrystalline LLZO samples were used. Polycrystalline samples, however, suffer very often from significant compositional inhomogeneities and a strong variation between different samples, even if made in the same batch (or even within one sample). This variation in composition limits the significance of experimental results and therefore the understanding of the underlying processes. These strong compositional variations could be overcome by using large, homogeneous single crystals. In this study, we have the great opportunity to use such single crystals with a size of several inches as a model system to systematically study the chemical and physical processes at the electrode- electrolyte interface and to test approaches to improve the interface with regard to a working device. Furthermore, the impact of chemical and physical inhomogeneities, which are one of the major reasons of failure in liquid electrolyte based lithium ion batteries, will be evaluated by highly sophisticated locally resolved analysis techniques. This will lead to a very accurate description of the LLZO-electrode interface and will provide a deep understanding on the underlying processes, thus creating a kind of roadmap to improvements of the interface for a future working ASSB.

The main goal of the project was the investigation of the chemical and physical processes taking place at interfaces between Li-ion conducting ceramic solid electrolytes and typical electrodes or other phases. Garnet-type Li7La3Zr2O12 (LLZO) and perovskite-type Li0.29La0.57TiO3 (LLTO) were used as electrolytes. LLZO turned out to be highly prone to proton/Li+ exchange when being exposed to water or steam. Chemical analysis of this ion interdiffusion process revealed the corresponding bulk diffusion coefficient and also showed existence of fast diffusion along grain boundaries. Moreover, severe interaction of LLZO with electrode materials may take place: Annealing an LLZO/LCO-interface to several hundred degrees lead to substantial Co diffusion into LLZO, which is very detrimental for the performance. The corresponding electrodes (e.g. LCO = LiCoO2) were sputter-deposited and separately investigated by impedance spectroscopy. Such impedance measurements revealed the electrochemical properties of our sputter-deposited thin film electrodes (LiCoO2 and LiMn2O4) in contact with liquid electrolytes and allowed an in-depth defect chemical interpretation of all relevant electrochemical parameters of the electrode materials (charge transfer, ionic conductivity, chemical capacitance, Li chemical diffusion coefficient) in dependence of their charging state. A second type of stability studies was performed on LLZO and LLTO by applying a voltage between two ion-blocking electrodes. This yielded the electrochemical stability window of the corresponding electrolytes. In both cases, a novel degradation mechanism was found when exceeding the stability limit at the oxidizing side of the electrolyte: Before full decomposition of the respective phase, oxide ion transport and oxygen evolution, and thus a Li2O depletion in the electrolyte, leads to a very localized but severe lowering of the Li content. At the reduction side of LLTO, on the other hand, a coloration front indicates formation of Ti3+. This decreases the electrolytic domain by introducing electronic conductivity. However, this also allowed a very detailed analysis of the electrode-like Li-insertion regime of LLTO, with quantitative data on the potential-dependent electronic conductivity and non-stoichiometry.