After the cycling tests the cells have been disassembled to

Fig. 8a compares the charge/discharge cycling life span of MicroSi@C and NanoSi@G at a specific current of 450 mA g−1. For the MicroSi@C electrode, the initial specific charge capacity reached as much as 1992 mA h g−1, by far larger than that CD 2314 of the NanoSi@G nanocomposite. It was notable to point out that all the reported capacities were based on the total weight of silicon and carbon in anode materials. However, MicroSi@C underwent a rapid capacity fading in the following cycles and only 10.5% of capacity was left after 150 cycles. The NanoSi@G electrode achieved a reversible initial capacity of 1,138 mA h g−1 at the first cycle. Although the initial value was lower than that of MicroSi@C, NanoSi@G presented a much longer cycle life with 73% capacity retention after 150 cycles. An obvious large voltage hysteresis and distinct flat plateau at around 0.04 V was observed in both material's first-cycle profile curve (Fig. 8b), which corresponded to the characteristic discharge voltage of silicon in half cells. In the meanwhile time, the NanoSi@G electrode maintained higher coulombic efficiencies than the MicroSi@C electrode, in particular, over the first 100 cycles. The first cycle coulombic efficiency for NanoSi@G and MicroSi@C was 93.1% and 93.3%, respectively. After the fourth cycle, the coulombic efficiencies of NanoSi@G maintained over 98.5%.