Background

High-performance rechargeable batteries are currently being developed to meet the urgent demands of high-specific capacity and superior energy density application devices. Li/S batteries have widely been considered as promising energy storage for power grids and electric devices because of their outstanding theoretical capacity (1672 mAh g−1) and prominent energy density (2600 Wh kg−1) [1, 24, 25].

Fig. 4
figure 4

Corresponding charge/discharge voltage-capacity profiles of cells assembled with CeO2/RGO composite coated separator (a) and pristine separator (b)

Full size image

The cycling performances of cells assembled with and without CeO2/RGO composite modified separator at 0.1 C and 1 C are gathered Fig. 5. At the current rate of 0.1 C, the modified battery achieved a high capacity of 1136 mAh g−1 after the 1st cycle and retained a capacity of 886 mAh g−1 after 100 cycles with high coulombic efficiency throughout the processes. These values were superior to that of the cell assembled with normal separator (992 mAh g−1 and 501 mAh g−1, respectively), suggesting the key role played by the functional separator. In addition, when the current rate increased to 1 C, the modified cells can also delivered an outstanding initial capacity of 917 mAh g−1 and maintained 72.9% of its initial capacity as well as high coulombic efficiency throughout the processes. The well-designed structure would not only allow better transport of electrons by contribute to superior electrical conductivity of RGO. Also, the shuttling of polysulfides could efficiently be impeded by the strong chemical bond between CeO2 and sulfur-related species.

Fig. 5
figure 5

Cycling performance and coulombic efficiency of cells assembled with and without CeO2/RGO composite coated separator

Full size image

The Nyquist plots of the cells assembled with and without CeO2/RGO composite-modified separator were first obtained then fitted with an equivalent circuit model. As shown in Fig. 6, both cells exhibited depressed semicircle in high-frequency region and inclined line at low frequencies. These would correspond to charge-transfer resistance (RCT) for sulfur cathode and Li-ion diffusion or so-called Warburg impedance, respectively [26, 27]. The smaller semicircle represented moderate RCT value of the modified cell, which mainly attributed to the efficiently suppressed shuttling of polysulfides by CeO2 nanoparticles and superior electron transport of RGO. Moreover, the CeO2/RGO composite would improve the electrochemical contact and maximize the utilization of active materials. The larger slope of Warburg impedance in modified cells suggested shortened diffusion of Li ions.

Fig. 6
figure 6

Nyquist plots for cells assembled with and without CeO2/RGO composite coated separator

Full size image

To gain a better understanding about contributions of CeO2/RGO composite-coated separator in impeding the shuttle of sulfur-related species, H-type glass cells were introduced and tested. As displayed in Fig. 7, the dark brown solution in the left side was composed of DOL/DME with 0.05 M Li2S6 as an additive. The right side solution contained pure DOL/DME. Li2S6 would spontaneously diffuse through the membrane from high to low concentration, which can be reflected by changes in color [28, 29]. In cells with normal separator (Fig. 7a), the color of the right cell changed evidently over time to become dark brown after 16 h, confirming that traditional commercial separator was unable to hinder the diffusion of polysulfide. By comparison, in cells with CeO2/RGO composite coated separator (Fig. 7b), no distinct color change took place over time, suggesting the shuttling of polysulfide was inhibited by CeO2/RGO composite modified separator.

Fig. 7
figure 7

Photographs of H-type glass cells assembled with pristine separator (a) and CeO2/RGO composite coated separator (b)

Full size image

XPS was used to confirm the existence of interactions between CeO2 and sulfur-related species. The elemental composition and valence states of CeO2/RGO composite after cycling are displayed in Fig. 8a. Four elements (C, O, Ce, and S) were detected. The peak in S 2p spectrum of CeO2/RGO composite after cycling can be fitted by three parts (Fig. 8b). The peak observed at 166.8 eV was assigned to S–O, and the peaks at 169.0 and 170.2 eV might be caused by metal-SO42− species. The Ce 3d spectrum of CeO2/RGO composite after cycling revealed peaks at binding energies of 882.8, 885.3, 889.1, and 898.6 eV (Fig. 8c), corresponding to CeO2 3d 5/2. The peak at 885.3 eV can be attributed to CeO2 3d 5/2. The peaks located at 901.2, 907.7, and 917.1 eV were associated with CeO2 3d 3/2. The peaks of CeO2/RGO composite-coated separator after cycling appeared slightly shifted to negative values (Fig. 8d). This indicated absorption of sulfur-related species by Ce–S bonding [30], corresponding to S 2p spectrum of CeO2/RGO composite after cycling.

Fig. 8
figure 8

XPS spectra of CeO2/RGO composite after cycling: survey spectrum (a) and S 2p (b). XPS spectra of CeO2/RGO composite: Ce 3d before (c) and after cycling (d)

Full size image

Conclusions

Polymer pyrolysis and hydrothermal method were employed as facile and efficient ways to prepare CeO2/RGO composite with superior structure, where ultrafine CeO2 nanoparticles were anchored on RGO sheets. The chemical suppression of the shuttling effect of polysulfides for CeO2 was confirmed by XPS after electrochemical processes. The performance of Li/S battery was significantly enhanced due to the cooperation of RGO and CeO2. A high initial capacity of 1136 mAh g−1 was obtained at 0.1 C with about 75.7% capacity retention after 100 cycles. The coulombic efficiency of the cell with CeO2/RGO composite-coated separator was also higher than values obtained by traditional commercial separators.