1. Introduction

Lithium ion batteries (LIBs) have gained great commercial success in the field of portable electronic devices due to their outstanding properties such as high energy density and environmental friendliness [1-4] Spinel Li₄Ti₅O₁₂ (LTO) material has been used as anode materials for lithium-ion batteries due to its "zerostrain effect", a high charge/discharge potential of around 1.55 V (vs. Li⁺/Li) and a long cycle life [4, 5]. However, the poor electrical conductivity and the sluggish lithium ion diffusion coefficient of Li₄Ti₅O₁₂ greatly limit its rate capability [6, 7]

Two approaches are usually taken to improve its electrochemical performance. One approach is to shorten the electronic and ionic transport length using porous or hollow structures [5, 8, 9]. Porous or hollow structured LTO materials have been reported to exhibit superior rate performance for their high surface area, good accessibility of the pores and large amount of electro-active sites. Another approach is to dope some materials into LTO materials [10-12] or combine with high conductive materials [7, 13], the composites formed can greatly improve the electrochemical properties of the electrodes. Among all the available conductive additives, graphene which has high conductivity, large surface area and excellent structural stability has attracted particular attention [14, 15]. LTO with graphene has attracted broad interest in the battery community. Kong et al.[13] prepared a Graphene-wrapped dandelion-like LTO. It exhibits high capacity; however the electrode is not stable. Oh et al.[16] prepared a Graphene-wrapped LTO generated by a solid-state reaction. The material exhibited a discharge capacity of 147 mA h g^-1 at the 10 C rate, but, the LTO had a non-uniform size distribution. In addition, the high temperature reaction time is too long.

Herein, wedescribe a novelstrategy forthe synthesis offew-layer reduced graphene oxide-wrapped mesoporous Li₄Ti₅O₁₂ spheres (denoted as m-LTO@FL-RGO) and sintering as anode materials for lithium ion batteries. Such composites with a hierarchical core/shell structure are advantageous for enhancing the electronic conductivity amongthe mesoporous LTO spheres and offer a shortpathway for Li⁺ diffusion within the mesoporous LTO spheres.

2. Experimental 2.1. Material synthesis

Synthesis of m-LTO microspheres: First, amorphous TiO₂ was prepared via a sol-gel process [17]. Second, lithium hydroxide was dissolved in deionized water at room temperature to obtain a 5 mol L^-1 LiOH solution. As-prepared TiO₂ mixed with the LiOH solution and the mixture was transferred into a Teflon-lined autoclave to heat at 100 ℃ for 20 h, without any shaking or stirring. Then the material was collected by centrifugation, washed with deionized water several times and dried on a vacuum oven at 60 ℃ for 12 h. Then, the material was sintered at 600 ℃ for 2 h in a muffle furnace.

Synthesis of m-LTO@FL-RGO composite: The graphite oxide was synthesized from natural graphite powders based on a modified Hummers method [18]. Then, the resulting graphite oxide is exfoliated to give a 2 mg mL^-1 GO solution by ultrasonication of the dispersion for 120 min. Typically, some fresh LTO microsphere powders (80 mg) were added into 1 g L^-1 poly (allylamine hydrochloride) (PAH) solution and then dispersed by ultrasonication for 0.5 h, followed by the addition of the GO solution (0.5 mg mL^-1, 8 mL), The mixture was stirred and treated under ultrasonic conditions to form a homogeneous suspension. After freeze-drying, the obtained gray powder was calcined at 400 ℃ for 2 h in an argon atmosphere. The actual amount of RGO in m-LTO@FL-RGO composite was estimated from carbon contents (wt%) by elemental analysis, which was 4.2 wt%.

2.2. Characterization

The structure of the samples was characterized by XRD (Bruker AXS D8). The morphologies of the materials were observed using FESEM (JEOL JSM-6700F) and a TEM (FEI Tecnai G2). Nitrogen adsorption and desorption isotherms were measured at 77 K on a Micromeritics ASAP2010 instrument. Specific surface area calculations were made using the BET method.

Electrochemical measurements were evaluated in the system of two electrodes. A lithium metal was used as a counter electrode. The working electrode was prepared using the as-prepared materials of 80 wt%, acetylene black of 10 wt% and PVDF of 10 wt%. The charge and discharge were measured on a battery test system (Land CT2001A). Electrochemical impedance spectra (EIS) were performed on a potentiostat (CHI 604C, CH Instrumental Inc.).

3. Results and discussion

Fig. 1a shows the typical XRD patterns of pure m-LTO and mLTO@FL-RGO composites. The main diffraction peaks of m-LTO agreed with the standard patterns of spinel Li₄Ti₅O₁₂ (JCPDS card no. 49-0207), which are sharp and defined, indicating that the samples are highly crystalline [19]. After the graphene wrapping, it can be seen that none of the diffraction peak positions of the m-LTO@FL-RGO changed, and a new and broad peak around 25.6 can be observed, which is attributed to the (0 0 2) plane of reduced graphene oxide [20].

The chemical composition of m-LTO@FL-RGO composite was also confirmed by an analysis of Raman spectrum. Fig. S1 in Supporting information compares the Raman spectrum of GO and m-LTO@FLRGO composites. The Raman bands at 1344 and 1586 cm^-1 for mLTO@FL-RGOcomposite, denoted as the D line and the G line, are also observed. The D/G intensity ratio is 1.12 for RGO in m-LTO@FL-RGO composite, which is higher than the GO (0.86), implying the successful reduction of GO after the thermal treatment.

The specific surface area and pore size distribution of the samples were further investigated by nitrogen adsorption-desorption isotherms. As shown in Fig. 1b, both samples show Ⅳ-type isotherm curve with a distinct hysteresis loop in the range of 0.4-1.0 P/P₀, which is indicative of mesoporous materials [21]. All samples exhibit uniform mesoporous with an average pore diameter of 7.1 nm, indicating that RGO has no impact on the mesoporous property of the LTO microspheres. We find that the specific surface area of m-LTO is 136.1 m² g^-1; however the mLTO@FL-RGO composite possesses a higher specific surface area of 148.4 m² g^-1, which may be attributed to the contribution of RGO [22]. The high surface area in association with its mesoporous features facilitates faster lithium-ion diffusion due to efficient contact between active materials and electrolytes in the electrochemical reaction mixtures, thereby improving the electrochemical performance in LIBs.

The morphology and microstructure of LTO@ RGO composite were examined by SEM, TEM and high-resolution TEM (HRTEM) observations. As shown in Fig. 2a and c, we found that the surface of the pure LTO microspheres is rough and the whole microspheres with a diameter of about 1.4 μm. Different from that of the pure mLTO microspheres, the surface of the m-LTO@FL-RGO core/shell structure presents a wrinkle-like morphology (Fig. 2b), which confirms the hypothesis that the reduced graphene oxide (RGO) sheets have combined with the m-LTO microspheres successfully. Fig. 2d shows typical TEM images of the m-LTO@FL-RGO core/shell structure microspheres, revealing that the mesoporous LTO microspheres are composed of tiny LTO particles. The contrast of RGO is low; so it can only be viewed under a higher magnification. A further HRTEM image (Fig. 2e) confirms that few-layered graphene nanosheets (4-6 nm) indeed wrap around the LTO microspheres.

Fig. 3a and b shows the discharge and charge profiles of m-LTO and m-LTO@FL-RGO composites at a rate of 1 C (1 C=175 mA g^-1). The initial discharge and charge capacities are 195.3 and 156.1 mA h g^-1 for m-LTO microspheres, and 198.3 and 175 mA h g^-1 for m-LTO@FL-RGO composite, respectively, which lead to an irreversible capacity loss of~20.7% and~11.8%, respectively. The large irreversible capacity for the two electrodes can be due to the Li⁺ storage in the irreversible sites and the decomposition of the electrolyte caused by the adsorbed moisture in the mesoporous samples [22, 23]. It can be seen that the reversible capacity of m-LTO decreases from 156.1 mA h g^-1 to 131.5 mA h g^-1 after 100 cycles with a capacity retention rate of only 84%. On the contrast, the reversible capacity of m-LTO@FLRGO composite slightly decreases with cycling and reaches 165.4 mA h g^-1 after 100 cycles, showing high capacity retention of 94.5%, which exhibits much better cycling performance than mLTO microspheres. We also studied the cycling performance of other m-LTO@FL-RGO samples with different RGO contents (Fig. 3c). We find that the best sample is m-LTO@FL-RGO-4.2, the account of RGO of which is 4.2%. Too little RGO cannot improve the electrical conductivity. At the same time too much RGO hinder the capacity of composites [24].

Fig. 4a compares the rate performance of m-LTO and mLTO@FL-RGO composites at different rate. It can be clearly observed that the reversible capacity of m-LTO@FL-RGO composite was kept at 165.3 mA h g^-1 after the 10th cycle at a 1 C current. Upon increasing the discharge-charge rate to 2, 5 10 and 20 C, the reversible capacity was maintained at about 162, 157.9, 153.6 and 128.4 mA h g^-1, respectively. It was obvious that the rate capacity of m-LTO@ RGO composite was higher than that of m-LTO microspheres. Even at a high rate of 30 C, the specific capacity remained at about 115.1 mA h g^-1, whereas that of m-LTO dropped to only 36.7 mA h g^-1. When the current rate was again reduced back to 1 C, the specific capacity of m-LTO@RGO composite returned to about 160.5 mA h g^-1, which did not ultimately change in the subsequent cycles. Obviously, the result clearly demonstrates that the m-LTO@FL-RGO composite could tolerate varied discharge current densities and has a good application prospect in high power lithium-ion batteries.

Fig. 4b shows the EIS analysis of the samples. Apparently, the radius of the semicircle of m-LTO@FL-RGO composite electrode is significantly smaller than that of m-LTO microspheres electrode, suggesting that the charge transfer resistance of m-LTO@FL-RGO composite is lower than that of m-LTO microspheres [25]. This suggests that the wrapping of conductive graphene in the composite can facilitate the electron transfer from wrapped mesoporous LTO microspheres within the whole electrode and contribute to the higher rate capability of m-LTO@FL-RGO composite electrode compared with the m-LTO microsphere electrode.

4. Conclusion

In conclusion, we have fabricated m-LTO@FL-RGO by the integration of m-LTO microspheres with reduced graphene oxide nanosheets via a simple solution route. The m-LTO@FL-RGO composite exhibits improved electrical conductivity, rate capability and cyclic stability compared with that of pure m-LTO, which exhibits high reversible capacity of 165.4 mA h g^-1 after 100 cycles. Even at a rate of 30 C, a high discharge capacity of 115.1 mA h g^-1 is still obtained, which is three times higher than that of pristine m-LTO microspheres. The good cycling performance and high discharge capacity, as well as the simple synthetic route and low cost of m-LTO@FL-RGO composites are expected to lead to better commercial applications.

Acknowledgment

This work was financially supported by Program for Changjiang Scholars and Innovative Research Team in University (PCSIRT, No. IRT1161), Program of Science and Technology Innovation Team in Bingtuan (No. 2011CC001), and the National Natural Science Foundation of China (Nos. 21263021, U1303291).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cclet.2016.02.028.