In recent years, new-generation energy storage systems such as Li-S and Li-O2 batteries have attracted increasing attention because of the high energy densities over 500 Wh kg-1[1]. The high energy storage systems do not basically have Li+ ion in the positive electrodes and need Li metal negative electrode (NE), whose theoretical capacity is ca. 3860 mAh g-1. However, Li metal NE has a risk of short circuit by Li dendrite growth and the poor rate capability by the small surface area due to the flat morphology, i.e. foil or thin film. On the contrary, Si is one of the good candidates for the NEs because of the high theoretical capacity of ca. 4200 mAh g(Si)-1, and the suppression of Li dendrite growth by the Li alloying/de-alloying reactions. In addition, morphology of Si active materials, e.g. nanoparticles, nano-wires, nano-flakes, not only suppresses the capacity fading due to the crack and pulverization but also enhances the rate capability of Si NEs by increasing the surface area. However, for applying the Si NEs to next-generation batteries, it is necessary to Li pre-dope into the Si NEs. Namely, the easier and more convenient Li pre-doping technique must be developed for the practical use. In this study, we prepared two types of Li pre-doped Si NEs by direct Li pre-doping (DP) [2] and electrochemical Li charging (EC) methods without and with fluoroethylene carbonate (FEC) as an additive, and investigated the charge/discharge properties by a half-cell operation at limiting capacity of 2000 mAh g (Si)-1. The morphology and crystalline phase changes were investigated by a SEM, SPM and XRD analyses. The electrochemical properties also evaluated by constant current charge/discharge test and cyclic voltammetry by using Li pre-doped Si NE | 1.0 M LiPF6/ EC+DMC(1:1 by vol.) | Li metal half-cells at 30oC. To investigate the morphology change of Si nanoparaticles, SPM images of the Li pre-doped Si NEs were compared (Fig. 1). In the case of EC method, volume expansion of Si nanoparticles by Li pre-doping was clearly confirmed. Especially by the FEC addition, the volume expansion was more promoted, indicating deeper Li pre-doping. In contrast, Si nanoparticles treated by the DP method rather became smaller than the pristine one. This implies that the Si nanoparticles were cracked and pulverizated by the mechanical stress due to rapid Li alloying. Namely, the cracking is considered to shorten the Li+ diffusion distance to accelerate the Li pre-doping. In fact, XRD analysis exhibited much larger amount of Li15Si4, i.e. more Li-rich alloying state than Li x Si y amorphous phase, for the DP treated Si NEs, and the degree of Li alloying was enhanced by the FEC addition. Fig. 2 shows the cyclic voltammograms (CVs) for the Li pre-doped Si NEs between 1.5 and 0.02 V vs. Li/Li+. For EC treated Si NEs, two current peaks were observed for both charging at 0.20 and 0.10 V and discharging at 0.35 and 0.50 V, indicating Li alloying/de-alloying during the amorphous Si phase. Also, the charge/discharge capacities increased after adding FEC. In contrast, for DP treated Si NEs exhibited only one sharp current peak at 0.53 V during 1st cycle, meaning Li de-alloying from Li15Si4 crystalline phase. After that, the oxidation current due to electrolyte decomposition was continuously observed and relatively suppressed by the FEC addition. This indicates that the SEI film derived from FEC additive were more stable and tough against the large volume change by the phase transition between the Li15Si4 crystalline and Li x Si y amorphous phases. The charge/discharge capacities estimated from CVs at 3rd cycle was summarized in Table 1. The specific capacities were enhanced by the DP method and FEC addition. Therefore, the combination was found to be quite effective to improve the bulk activation and interphase stabilization of Si NE. The effect mechanism in more detail will be discussed in the meeting. This study was partially supported by NEDO Project “RISING2”, Japan. [1] P. G. Bruce et al., Nature materials, 11, 19 (2012). [2] M. Saito et al., Electrochemistry, 85(10), 656 (2017). Figure 1
Read full abstract