698
In Situ Electron Microscope Observations of Electrochemical Lithium Deposition/Dissolution with a Glassy Solid Electrolyte
The recent discoveries of inorganic solid electrolytes with large Li+ ionic conductivities have greatly propelled the research on all-solid-state lithium batteries. Solid electrolyte blocks Li dendrite growth toward the cathode if Li metal is used as the negative electrode. This is certainly attractive because the theoretical energy density of Li metal (2062 Ah L-1) is much greater than those of presently commercialized negative electrodes (e.g. graphite). It is hence important to understand how Li metal grows and dissolves through solid electrolyte interfaces in the charge-discharge processes [1]. This study reports in-situscanning electron microscope (SEM) observations of the initial stage of Li growth at electrode/amorphous solid electolyte interfaces.
Experimental
The mojority of the electrolyte (1.25 cm × 1.25 cm) is a mirror-polished Li1.3Al0.3Ti1.7(PO4)3 (LATP) sheet (OHARA Co.) with a thickness of 150 μm. Its top and bottom surfaces were sputter-coated with lithium phosphorus oxynitride (LiPON) layers with thicknesses of 2.5 μm. The working electrode is a 30-nm-thick-Cu film onto a circular area with a diameter of 5.0 mm in the center of the top LiPON surface. The counter electrode is a 3-4-μm-thick-Li film on the bottom LiPON surface. Electrochemical Li deposition and dissolution were conducted in a Keyence VE-9800 SEM under galvanostatic conditions. Li deposition/dissolution are carried out at 50 μA cm-2 to 3 mA cm-2. The total charges are 180 mC cm-2 for all curent densities. This charge amount corresponds to a plating of a 242-nm-thick Li film if the plated film formed a contiguous solid thin film.
Results and discussion
Figure 1 shows SEM images of the growth trajectories of different Li nuclei at 100 μA cm-2. Li nuclei under a Cu film eventually break the Cu film to grow Li rods. Li deposits break a Cu film at approximately 240 s. Once a Cu film is cracked, new Li nucleation no longer occurs. Thus, strain energies in Li and Cu are considered to influence the Li nucleation and growth under a Cu film. At 1.0 mA cm-2, the number of Li nucleation sites increases. Additionally, rapid agglomeration of Li nuclei under a Cu film is observed (Figure 2). In contrast to the case at 100 μA cm-2, Li rods do not grow through cracks of a Cu film.
Li dissolution reactions were also observed by in-situ SEM, and Li self-diffusion in Li rods seems to dominate the stripped length within a rod. Morphological changes of a Cu film at the first cycle influence the preferential sites for Li growth at subsequent cycles. Li deposition/dissolution reactions will be further discussed in more detail with in-situ SEM movies.
Reference
[1] F. Sagane, K. Ikeda, K. Okita, H. Sano, H. Sakaebe, and Y. Iriyama, J. Power Sources, 233 (2013) 34.