Lawrence Berkeley National Laboratory

We propose a general solution to the problem of Li dendrite penetration through solid electrolytes such as LLZO. The solution is based on an analogy of Li dendrite penetration to stress corrosion cracking, a process in which a very soft material--water--penetrates through a very hard material--steel--generally at grain boundaries. The problem of stress corrosion cracking was solved many years ago by putting the surface into a state of residual compressive stress. A variety of materials have been strengthened this way, including metals, glasses (e.g., Gorilla glass), and ceramics. However, for this approach to be useful, it is critical, at a minimum, that the presence of a significant residual compressive stress not block ion transport. In this work, we used DFT to show that isotropic compressive residual stresses as high as -10 GPa—the maximum observed in strengthened ceramics—have only a minimal impact on the diffusion rate of Li+ ions through a near-surface region of LLZO. Figure 1