As part of a Battery 500 seedling project, MBRDNA partnered with LBNL and Montana State University (MSU) to develop solid-state batteries based on an Al-substituted LLZO (Li7La3Zr2O12) ceramic ion-conductor. All solid-state lithium batteries are attractive contenders for use as power sources for electric vehicles because of their potential for better safety and higher energy density than state-of-the-art lithium-ion batteries, although they are still at early stages of development. Al-substituted LLZO is one of the most promising solid electrolytes for battery applications based on high ionic conductivity, a wide operating voltage window, and apparent stability vs. reduction by lithium. However, difficulties with processing thin (< 20 um and dense LLZO membranes have hampered development of devices based on this material. It is non-compressible and must be sintered at high temperatures to densify components. It is also difficult to maintain contact between the active cathode material and LLZO in the composite cathodes, which is critical for successful operation. This research project aimed to use novel processing methods to overcome these issues. The end result was the first ever-reported truly all-solid-state battery based on LLZO operating at room temperature without application of exogenous pressure. During the course of the project, we also identified a number of issues that need to be addressed to improve the technology readiness level of this battery. The procedure is to fabricate a porous LLZO ceramic scaffold using freeze-tape casting methods and co-sinter it with a tape cast thin dense LLZO layer to make a bilayer. A cathode material (either LiNi0.6Mn0.2Co0.2O2 or LiNi0.8Mn0.1Co0.1O2, hereafter abbreviated NMC-622 or NMC-811), carbon, polyvinylidene fluoride (PVdF) and a secondary solid electrolyte based on succinonitrile complexed with lithium salts is then infiltrated into the porous layer, and lithium metal is attached to the opposite side to form the cell. A number of trouble-shooting configurations that were variants of this procedure were also assembled and tested at LBNL. MSU used their expertise in freeze tape casting to optimize the scaffold-making process, and LBNL developed cell making procedures and used advanced diagnostics such as synchrotron micro-tomography to understand the materials properties. MBRDNA assembled cells from components provided to them by either LBNL or MSU and tested them electrochemically. In some cases, already-assembled coin cells were provided to MBRDNA by LBNL. Tests carried out at MBRDNA included impedance spectroscopy, initial charge and discharge, rate capability tests, and extended cycling, all at room temperature. Some cells were de-crimped and subjected to post-mortem analysis after cycling. Scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDS) and optical imaging were carried out on the de-crimped cells. MBRDNA provided this data to LBNL. Quarterly reports were submitted to DOE summarizing these efforts and results.

Complex chemomechanical interplay exists over a wide range of length scales within the hierarchically structured lithium-ion battery. At the mesoscale, the interdependent structural complexity and chemical heterogeneity collectively govern the local chemistry and, as a result, critically influence the cell level performance. Here we investigate the morphology and state of charge (SOC) inhomogeneity within secondary NCA particles that were cycled in solid polymer batteries. We observe substantial inhomogeneity in the nickel oxidation state (a proxy for SOC) and loss of structural integrity within secondary particles after only 20 cycles due to significant intergranular cracking. The formation of mesoscale cracks causes loss of ionic and electrical contact within cathode particles, triggering increases in local impedance and rearrangement of transport pathways for charge carriers. This can eventually lead to deactivation of sub-particle level domains in solid-state lithium-ion batteries. Our findings highlight the importance of proper mesoscale strain and defect management in polymer lithium-ion batteries.

The appeal of solid-state batteries is clear; there are no flammable liquid components to leak or catch fire, and the possibility of using lithium anodes suggests that higher energy density than currently obtainable with Li-ion configurations should be possible. The emergence of garnet-structured electrolytes based on Li7La3Zr2O12 (LLZO) are particularly attractive for such devices because of their high conductivities and apparent stability against lithium. Still, there are many challenges associated with LLZO, including difficulties in densifying the materials and high interfacial resistances at the lithium electrodes. Work in our group has been directed towards overcoming these obstacles. Classical ceramics processing techniques have been used to produce samples up to 94% dense, as thin as 200 microns, with excellent control over grain size and grain boundary chemistry.  Synchrotron techniques such as soft x-ray absorption spectroscopy (XAS) and x-ray photoemission spectroscopy (XPS) reveal that the main culprit leading to high interfacial resistance at lithium electrodes is lithium carbonate, either present as a by-product of synthesis or produced upon exposure to air.  Polishing in inert atmosphere removes this impurity and lowers the interfacial impedance, but is not practical commercially. Through grain boundary engineering, it is possible to substantially reduce the sensitivity of samples to air exposure, greatly reducing the tendency towards formation of lithium carbonate and resulting in much lower interfacial impedances at lithium electrodes.