A sandwich structure of mesoporous anatase TiO₂ sheets and reduced graphene oxide and its application as lithium-ion battery electrodes

Abstract

Mesoporous anatase TiO₂ sheets/rGO sandwich-like nanocomposites were facile synthesized by acid-assisted tetrabutyl titanate hydrolysis and subsequent thermal reduction process. Structural, morphological, and compositional properties were characterized by various techniques, such as X-ray diffraction, scanning electron microscopy (SEM), transmission electron microscopy (TEM), Raman spectroscopy, thermal gravimetric (TG) analysis, and BET surface area analysis. When used as the anode materials of Li-ion batteries, the as-prepared sample delivered reasonable capacity, good cycling stability and rate capability. The optimal sample delivered a high reversible lithium-storage capacity of ∼161.4 mA h g⁻¹ after 50 cycles at a current rate of 0.5 C (1 C = 335 mA g⁻¹), with good cycling stability and rate capability. It is believed that the good electrochemical performance can be attributed to the mesoporous feature, the addition of rGO nanosheets, and the special sandwich-like electrode structure. Therefore, rational design of mesoporous structures and compositing with rGO nanosheets are of importance for improving the lithium-storage performance.

---

1. Introduction

Rechargeable lithium-ion batteries (LIBs) are one of the most important energy storage devices for portable electronics and electric vehicles due to their advantages of high energy density, long cycling life, and environmental benignity.^1–3 It is widely accepted that the overall performance of LIBs is highly dependent on the inherent properties of the electrode materials.⁴ Therefore, considerable attention has been paid to developing novel materials for both the cathodes and anodes of LIBs which are inexpensive, safe and environmentally benign. For nanostructured electrode materials (such as nanopowders, nanowires, nanorods and nanotubes), it generally leads to improved energy density, better capacity retention, and superior rate capability, which are due to the large surface area, numerous active sites, short mass and charge diffusion distance, and efficient accommodation of volume changes during charging and discharging processes.^4–7

Anatase titania (TiO₂) is of great importance for both fundamental studies and technological applications in the field of energy storage and conversion.^8–12 Although anatase TiO₂ delivers a lower capacity^13–16 (∼335 mA h g⁻¹, according to the electrochemical reaction TiO₂ + xLi + xe⁻ ↔ Li_xTiO₂ with the maximum insertion coefficient x determined to be ∼0.96) at a higher potential of about 1.7 V vs. Li⁺/Li, the ease of structural tailoring, the low volume expansion upon cycling (<4%), good stability and free of lithium plating, endowing it has great potential to be charged/discharged at high current rates for extended cycling, which can hardly be achieved by other types of anodes.^17,18 The size, shape and assembly of various anatase TiO₂ nanostructures are intensively investigated to optimize the Li-ions storage properties.^19–26 Nevertheless, the main weakness of TiO₂ as anode materials for LIBs lies in the intrinsically slow kinetics of Li⁺ diffusion and low electronic conductivity (∼10⁻¹³ S cm⁻¹), resulting in the deterioration of reversible capacity and rate capability.^27–29 Therefore, considerable attempts (e.g., doping and surface coating) have been made to tackle the above problems.^29–31

Graphene, a single layer of carbon arranged in a honeycomb structure, is recently expected to be promising additives for advanced anode material in LIBs due to many fascinating properties such as giant electron mobility, extremely high thermal conductivity, extraordinary elasticity, stiffness, and ultra large specific surface area.^32–36 In addition, it has attracted much attention because of wide potential applications in nanoelectronics, energy storage and conversion, and biotechnology.^37–39 The graphene sheets provide efficient pathways for electron transfer. However, the graphene sheets can naturally stack into multilayers and thus lose the advantages of high surface area and intrinsic chemical and physical properties. Nanoparticles are usually anchored on the graphene sheets to prevent the stacking problem, and the synthesized nanocomposites are favorable for retaining the stable structures of the electrode materials with enhanced conductivity in the charging and discharging cycling processes.^31,40–44 Yang et al. reported graphene-based TiO₂ nanosheets synthesized via a sol–gel process.⁴² The TiO₂/graphene nanosheets present reversible capacities of 162 and 123 mA h g⁻¹ at 170 and 1700 mA g⁻¹, respectively. Zhao et al. developed a sol–gel design strategy toward ultradispersed TiO₂ nanoparticles on graphene sheet with an unprecedented degree of control based on the precise separation and manipulation of nanoparticles nucleated, grown, anchored, and crystallized and the reduction of graphene oxide (GO), which exhibit a high specific capacity of ∼94 mA h g⁻¹ even at a high current density of ∼10 A g⁻¹.⁴³ Recently, Qiu and co-workers reported a hydrothermal method toward in situ growth of ultradispersed mesoporous TiO₂ nanocrystals with (001) facets on graphene aerogels.⁴⁴ The resultant TiO₂/graphene aerogels composites exhibit a high specific capacity in LIBs. It should be mentioned here that the above obtained monodisperse TiO₂ nanocrystallines on graphene usually possess low packing density, which will lead to poor volumetric storage capacity when considering them for LIBs applications.^45,46 In this respect, mesoporous electrode materials constructed by micro-/nano-structured building blocks are designed to achieve high volumetric energy densities for LIBs. Mesoporous materials are usually micrometres in dimension with nanopores in the range 2–10 nm, thus when employed as electrode materials in LIBs enable easy access for the Li ions from the electrolytes facilitating Li⁺ transport within the bulk grains which are typically in the range 10–20 nm.⁴⁶ Furthermore, the mesoporous materials with micro-/nano-structures can prevent the aggregation of the nanocrystals when compared to the monodispersed case.

Based on the above, in order to improve both the electric conductivity and packing density, we attempt to incorporate the graphene nanosheets between mesoporous anatase TiO₂ sheets to form sandwich-like nanocomposites. We synthesize the sandwich structure of mesoporous anatase TiO₂ sheet and reduced GO (rGO) by acid-assisted tetrabutyl titanate (Ti(OC₄H₉)₄, TBOT) hydrolysis and subsequent thermal reduction process. When used as electrode materials for LIBs, excellent performance, including high reversible capacity, good rate capability and enhanced cycling performance, has been demonstrated, making it promising for applications in LIBs.

2. Experimental

2.1 Materials synthesis

Concentrated sulfuric acid (98% H₂SO₄, Beijing chemical works), TBOT (Tianjin Yongda chemical reagent development center) were of analytical grade and were used as purchased without further purification, and deionized water was used throughout. Graphene oxide solution (0.5 mg ml⁻¹) was prepared by ultrasonic dispersion GO powder (XF Nano, INC, Nanjing) in deionized water. In a typical synthesis, 17.6 ml of TBOT was mixed with 200 ml of GO solution in a 250 ml sterile bottle with vigorous magnetic stirring. After that a white precipitation was observed (TiO₂/GO composite with poor crystallization, Fig. S1†). Then 10 ml of concentrated H₂SO₄ was added dropwise within 2 min. After 30 min, the solution was transferred to an oil bath at 60 °C and keep the temperature for 12 h with continuous magnetic stirring. The bottle was then taken out and left to cool down to room temperature. The precipitate obtained was washed thoroughly with distilled water and ethanol for three times and then dried at 120 °C for 4 h in an oven. Subsequently, the gray powder was annealed in argon atmosphere at 500 °C for 2 h to give rise to the grain growth of TiO₂ nanoparticles and the thermal reduction of GO (named as TG1). Two other control experiments were performed to tune the proportion of rGO in the final nanocomposites. The procedure was similar to the preparation of TG1 except that the volumes of TBOT and concentrated H₂SO₄ were decreased to 5 and 3.1 ml, 2.5 and 1.6 ml for TG2 and TG3, respectively.

2.2 Materials characterizations

The product morphology, composition and crystal structure was examined using field-emission scanning electron microscopy equipped with an energy dispersive X-ray (EDX) system (FESEM; Hitachi, S5500, 5 kV), transmission electron microscopy (TEM; JEOL, JEM-2011, 200 kV; FEI, Tecnai G20, 200 kV), and thermal gravimetric (TG) analysis (Netzsch-STA 449C, measured from room temperature to 800 °C at a heating rate of 10 °C min⁻¹ under an air flow). Crystallographic information for the samples was collected using a Bruker Model D8 Advance X-ray powder diffractometer (XRD) Cu K_α irradiation (λ = 1.5418 Å). The Barret–Joyner–Halender (BET) surface area of the powders was analyzed by nitrogen adsorption–desorption isotherm at 77 K in a Micromeritics ASAP 2010 system. The sample was degassed at 180 °C before nitrogen adsorption measurements. The BET surface area was determined by a multipoint BET method. A desorption isotherm was used to determine the pore size distribution via the Barret–Joyner–Halender (BJH) method, assuming a cylindrical pore model. The nitrogen adsorption volume at the relative pressure (P/P₀) of 0.994 was used to determine the pore volume and average pore size. Raman spectra were collected by using Raman microscopes (Renishaw, UK) under a 633 nm excitation.

2.3 Electrochemical measurements

To measure the electrochemical performance, composite electrodes were constructed by mixing the active materials, conductive carbon black, and polyvinylidene fluoride (PVDF), in a weight ratio of 70 : 20 : 10. The mixture was prepared as slurry in N-methyl pyrrolidinone and manually spread onto copper foil as a current collector (10 mm in diameter) by using the doctor-blade technique (the typical loading of the active material is in the range of 3.6–4.8 mg cm⁻²). The electrode was dried under vacuum at 120 °C for 5 h to remove the solvent before pressing. The pressing process was conducted at room temperature by using a rolling machine. The rolling pressure and time were 1–3 MPa and 2–5 seconds, respectively. Then the electrodes were cut into disks (12 mm in diameter) and dried at 100 °C for 24 h in vacuum. The typical thickness of the dried electrodes is in the range of 70–100 um. The cells were assembled inside an Ar-filled glove box by using a lithium-metal foil as the counter electrode and the reference electrode and microporous polypropylene as the separator. The electrolyte used was 1 M LiPF₆ in a 1 : 1 weight ratio ethylene carbonate (EC)–dimethyl carbonate (DMC) solvent. Assembled cells were allowed to soak overnight, and then electrochemical tests on a LAND battery testing unit were performed. Galvanostatic charge and discharge of the assembled cells were performed at a current density of 0.5 C between voltage limits of 1 V and 3 V (vs. Li⁺/Li) at room temperature. For the high rate testing, the discharge current gradually increased from 0.5 C to 1, 2.5, and 5 C, then decreased to 0.5 C, step by step. All the charge/discharge testings were performed symmetrically at room temperature. Cyclic voltammograms (CVs) was performed using a CHI 660D electrochemical workstation (Chenhua Instrument, Shanghai). CVs were recorded between 3.0 V and 1.0 V (vs. Li⁺/Li) at a scan rate of 0.5 mV s⁻¹, using the composite as the working electrode and a lithium sheet as both counter electrode and reference electrode.

3. Results and discussion

The crystal structures of the yielding products were characterized by XRD as shown in Fig. 1. Generally, the diffraction peaks at 25.3° and 27.4° are identified as the characteristic peaks of the anatase and rutile crystal phase of TiO₂, respectively.⁴⁷ It can be seen in Fig. 1a that the identified peaks of the TG1 sample can be perfectly indexed to anatase TiO₂ (JCPDS no. 21-1272) without peaks of impurities, indicating the phase-pure nature of TG1 sample. For TG2 and TG3 samples, a small amount of rutile phase (JCPDS no. 21-1276) appeared as indicated by asterisks in Fig. 1b and c. Moreover, the relatively high peak intensities imply that all of the products are highly crystalline. The crystallite sizes (d) of anatase TiO₂ phase can be calculated from the physical breadths of the corresponding diffraction peak by the Scherrer formula: d = 0.89λ/β cos θ, where λ is the X-ray wavelength, β the full width at half maximum (FWHM) of the diffraction peak, and θ is the diffraction angle.^48,49 Here, the crystallite size was calculated using the anatase (101) peak. The obtained crystallite sizes were about 12.1, 10.7, and 9.6 nm for TG1, TG2, and TG3, respectively. This is further confirmed by FESEM observations as discussed later. In addition, no noticeable diffraction peak corresponding to the restacking of graphene to form graphite can be observed, which may be ascribed to the effective prevention of agglomeration of graphene sheets during the thermal reduction process.

Fig. 1 XRD patterns of as-prepared samples and standard pattern of anatase TiO₂ phase. (a) TG1, (b) TG2, and (c) TG3. (Asterisk (*) indicates the rutile phase).

The morphologies of the nanocomposite structures are checked by FESEM as shown in Fig. 2. In general, the final products which present sheet-like characteristic are uniformly distributed on the substrate, and the size of in-plane sheets is in the micrometer/sub-micrometer scale. Moreover, the existence of inserted flexible graphene sheets (indicated by white arrows) demonstrates the sandwich-stacked structural nature of the nanocomposites. Meanwhile, higher magnification FESEM images show that the surface of the nanosheets becomes coarse and porous, and the TiO₂ nanosheets are composed of nanoparticles. The acid-assisted TBOT hydrolysis and subsequent thermal reduction process may facilitate the formation of small nanocrystallites. The gaseous species produced during the annealing process also likely assist in constructing the highly porous texture. As a result, abundant pores between the nanoparticles are generated, which will be confirmed by the following nitrogen adsorption–desorption analysis. Statistics from the higher magnification FESEM images (Fig. S2†) shows that the average particle size is ∼11.8, 11.2, and 10.5 nm for TG1, TG2, and TG3, respectively, in good agreement with the above calculations from XRD patterns. EDX results (the insets of Fig. 2) indicate that except Cu and Al elements from sample stage all of the three samples are composed of Ti, O, and C, and the atomic ratio is ∼1 : 2 for element Ti to element O, further confirming the formation of TiO₂.

Fig. 2 FESEM images of the mesoporous anatase TiO₂ sheets/rGO sandwich-like nanocomposites with different magnifications. (a1, a2) TG1, (b1, b2) TG2, (c1, c2) TG3. The insets are EDX patterns.

The detailed structural investigations of the mesoporous sandwich-like nanocomposites are studied by TEM and HRTEM. Fig. 3a1 and a2 are typical low and high magnifications TEM images for TG1, which disclose the porous and sheet-like natures of the sample. Moreover, the porous nanosheets consist of interconnected nanoparticles. The corresponding selected area electron diffraction (SAED) pattern further confirms that the as obtained mesoporous sandwich-like nanocomposites are polycrystal in nature as shown in the inset of Fig. 3a2. Fig. 3a3 shows a typical HRTEM image of an individual nanosheet. The lattice spacings of d ∼ 3.58 Å is determined (see Fig. 3a3), which corresponds to the (101) plane of anatase TiO₂. TEM and HRTEM results for TG2 (Fig. 3b1–b4) and TG3 (Fig. 3c1–c4) possess similar structural features with TG1, i.e., mesoporous anatase TiO₂ sheets composed of nanoparticles as building blocks. Furthermore, flexible rGO sheets are occasionally observed on the edge of the samples, also confirming the sandwich-like characteristic of the nanocomposites.

Fig. 3 TEM and HRTEM images of the mesoporous anatase TiO₂ sheets/rGO sandwich-like nanocomposites: (a1–a3) TG1, (b1–b4) TG2, (c1–c4) TG3. The insets in (a2, b3 and c3) are corresponding SAED patterns.

The presence of graphene in the final products was further confirmed by their Raman spectrum as shown in Fig. 4a. Apart from the TiO₂ Raman feature (Fig. S3†), there were another two Raman peaks centered at ∼1345 and ∼1590 cm⁻¹ for all the three sandwich-like nanocomposites. The peak at 1590 cm⁻¹ (G band) is attributed to the vibration of sp² hybridized C–C bond of a in-plane hexagonal lattice. The peak at 1345 cm⁻¹ (D band) is associated with the vibrations of carbon atoms with dangling bonds in plane terminations of the disordered graphite from the defects and disorders of structures in carbon materials.⁵⁰ It was generally accepted that the relative intensity ratio between the D and G band (I_D/I_G) was found to correlate to the nature of carbon.^51,52 The I_D/I_G values for TG1, TG2, and TG3 samples are 0.44, 0.87, and 0.54, respectively, indicated the formation of a reasonable degree of graphitization. In order to determine the amount of graphene in the sandwich-like nanocomposites, TGA experiments were performed (Fig. S4†). As the temperature rises from room temperature to ∼400 °C the weight loss could be ascribed to surface water adsorption. The weight loss after ∼400 °C could be ascribed to the oxidation of graphene in the nanocomposites, which yielding the weight fraction of graphene in the nanocomposites of about 1.2%, 2.3%, and 5.4% for TG1, TG2, and TG3 samples, respectively. The presence of graphene in the sandwich-like nanocomposites could dramatically enhance their electronic conductivity and is responsible for improving lithium storage ability. On the other hand, the contents of TiO₂ in the nanocomposites is larger than most of the reported nanocomposite of TiO₂ and various carbon nanostructures, which is important for the LIB applications.^40–44

Fig. 4 (a) Raman spectra, (b) nitrogen adsorption–desorption isotherms and (c) the corresponding pore size distribution curves of the mesoporous anatase TiO₂ sheets/rGO sandwich-like nanocomposites. TG1 (black line), TG2 (red line), and TG3 (blue line).

The pore structures, including specific surface areas and the porous feature, of the mesoporous anatase TiO₂ sheets/rGO sandwich-like nanocomposites were studied by measuring nitrogen adsorption–desorption isotherms at 77 K (Fig. 4b). The BET specific surface area is 39.8, 38.3, and 42.9 m² g⁻¹ for TG1, TG2, and TG3, respectively. In addition, the mesopore size distributions based on the BJH method of all the three samples are further confirmed by the corresponding pore size distribution curves (Fig. 4c). These results show that the sandwich-like nanocomposites have large surface areas which mainly attribute to the mesoporous nature of the TiO₂ nanosheets. In addition, all the three samples show similar porous structures. The present mesoporous anatase TiO₂ sheets/rGO sandwich-like nanocomposites are of importance in lithium-storage processes, due to their capability of providing extra active sites for the storage of lithium ions and facilitating mass diffusion and ion transport, which are induced by the synergistic reactions of mesoporous structures and the addition of rGO nanosheets. Therefore, the synthesized samples are anticipate to show good lithium-storage properties.

The electrochemical behavior of the assembled cells is first investigated by CV experiments between 1.0 and 3 V (vs. Li⁺/Li) at a scan rate of 0.5 mV s⁻¹, using the composite as the working electrode and a lithium sheet as both counter electrode and reference electrode. Fig. 5a shows the first three cyclic voltammograms (CVs) for the TG3 sample. In the first cycle, there are two well-defined peaks at voltage positions of 1.60 (cathodic sweep) and 2.09 V (anodic sweep), which correspond to the Li insertion and Li extraction, respectively. The sharp oxidation/reduction peaks reveal the two-phase reaction mechanism during electrochemical lithium insertion/extraction.^24–26 Compared to the initial cycle, in the subsequent cycles, small deviations in the peak positions are noted, possibly due to structural rearrangement of TiO₂ crystal lattice,^53,54 and almost no change in the peak shape is observed, suggesting high reversibility of the electrode.

Fig. 5 (a) CV curves of the mesoporous anatase TiO₂ sheets/rGO sandwich-like nanocomposite electrode (sample TG3) between 3.0 V and 1.0 V (vs. Li⁺/Li) at a scan rate of 5 mV s⁻¹. (b) Galvanostatic charge–discharge profiles of the first cycle of the sandwich-like nanocomposite electrodes (TG1, TG2, and TG3) at a charge–discharge rate of 0.5 C in the voltage range of 1.0–3.0 V (vs. Li⁺/Li). (c) The cycling performance (reversible capacity) of the sandwich-like nanocomposite electrodes (TG1, TG2, and TG3) at a charge–discharge rate of 0.5 C in the voltage range of 1.0–3.0 V (vs. Li⁺/Li) up to 50 cycles. (d) Cycling performance (reversible capacity) of the sandwich-like nanocomposite electrodes (TG1, TG2, and TG3) at various current rates. TG1 (black), TG2 (red), and TG3 (blue).

Fig. 5b shows the galvanostatic cycling profiles of the three samples for the first charge/discharge cycle at a current rate of 0.5 C in the voltage range of 1.0–3.0 V (vs. Li⁺/Li) at room temperature, in which all three electrodes display similar electrochemical behaviors. Two dominant voltage plateaus appear at ∼1.62 and ∼2.10 V during the discharge and charge processes, which correspond to the lithium insertion and deinsertion, respectively. The potential difference is around 0.5 V between charge and discharge reaction and the results are generally consistent with the above CV analysis as well as previous reports.^24–26 For example, in the studies by Sun and Yang et al., two well-defined current peaks located at about 1.75 V and 2.10 V in the CV curves of anatase TiO₂ nanosheets with 80% (001) surface.²⁵ Chen and Low et al. also observed two current peaks at 1.55 V (cathodic sweep) and 2.2 V (anodic sweep) when studying the electrochemical lithium insertion/extraction behavior anatase TiO₂ nanosheets hierarchical spheres.²⁴ The potential differences between charge and discharge reaction are 0.35 and 0.65 V for the above two cases, respectively. However, Ortiz and Djenizian et al. found that the potential difference was only 0.28 V for the annealed self-organized TiO₂ nanotubes cells, where the TiO₂ nanotube layers were directly used as binder-free electrodes, and no additives such as PVDF and carbon black were added.⁵⁵ Therefore, the polarization is much lower than that of the present work and other studies.^24–26

The theoretical capacity of the TiO₂–rGO composite, C_theor, can be calculated through the equation:

C_theor = C_TiO2 × mass percentage of TiO₂ + C_graphene × mass percentage of graphene

where C_TiO2 (335 mA h g⁻¹) and C_graphene are the theoretical capacity of TiO₂ and graphene, respectively. To the best of our knowledge, the value of C_graphene in the potential range of 1–3 V can not be obtained. Taking into account approximation, the value of C_graphene in the potential range of 1–3 V (744 mA h g⁻¹)⁴⁰ is employed to estimate the theoretical capacity of the composites, which yields the theoretical capacity of ∼339.9, 344.4, and 357.1 mA h g⁻¹ for TG1, TG2, and TG3, respectively. From the profiles in Fig. 5b, the initial discharge capacities are 180.1, 196, and 231 mA h g⁻¹ for TG1, TG2, and TG3, which are lower than that of the corresponding theoretical capacities. It also found that the initial charge capacities are 142.2, 167, and 206 mA h g⁻¹, yielding the Coulombic efficiency values (the ratio of charge capacity to discharge capacity) of 75%, 85%, and 89%, respectively.

Fig. 5c displays the comparison of cycling performance of the three samples at a current rate of 0.5 C in the voltage range of 1.0–3.0 V (vs. Li⁺/Li) up to 50 cycles. It can be seen that after the first several cycles all the three sandwich-like nanocomposites show good cyclic capacity retention during cycling. It is obvious that the TG3 cell is superior, with the highest lithium-storage capacity. The results are also supported by the performance measurements of identical cells (Fig. S5†). The reversible capacity reaches ∼161.4 mA h g⁻¹ after 50 cycles with about 70% retention as compared with the first reversible capacity. In comparison, the reversible capacity of TG2 cell decreases to ∼133.5 mA h g⁻¹ after 50 cycles of operation, and the reversible capacity of TG1 cell drops down to ∼127.9 mA h g⁻¹, with retention rates of about 67% and 71% with respect to the first reversible capacity for TG2 and TG1 cells. Accordingly, the irreversible capacity losses after 50 cycles are 29%, 33%, and 30% for TG1, TG2, and TG3, respectively, which is comparable with the other reports.^17,24 The cycling performance of anatase TiO₂ nanoparticles (TiO₂ NPs, ∼25 nm in diameter) reported by Sun et al. is also shown in Fig. 5c for comparison.⁵⁶ We mentioned here that the experimental conditions for TiO₂ NPs are similar to our case except that they were cycled with a low current rate of 0.1 C. However, the reversible capacity is only 57 mA h g⁻¹ after 50 cycles. The compared results further confirm the superior lithium-storage performance of the sandwich-like nanocomposites.

To investigate the rate capability, the three samples were discharged and charged at various current rates between 0.5 C and 5 C, and the results are shown in Fig. 5d. The charge/discharge rates are programmably modified from 0.5 C to 1 C, 2.5 C, 5 C and then back to 0.5 C for 10 cycles. It can be clearly observed that the nanocomposites, especially TG3 sample, show excellent capacity retention at different rates. The reversible capacity of TG3 cell varies from 161.1 mA h g⁻¹ to 85.6 mA h g⁻¹ at current rates of 0.5 C and 5 C, respectively. However, the reversible capacity of the TG1 and TG2 cells rapidly drops from 126 to 58 mA h g⁻¹ and 131 to 66 mA h g⁻¹ respectively. When the rate return to the initial 0.5 C after 40 cycles, TG3 composite cell recovers its original capacity a little bit lower (155.2 mA h g⁻¹ for the 50th cycle). The superior rate performance of the TG3 cells mainly comes from the highest content of graphene among the three samples, which is favourable for increasing the conductivity and decreasing the inner resistance of LIBs. It is obvious that the nanostructuration and optimum porosity also account for the high rate capability of TG3 cells. The morphology and structure of the sandwich-like nanocomposite electrodes after rate capability testing (50 cycles) are characterized by TEM (Fig. S6†). It can be seen that all of the three samples still maintain the initial morphologies with anatase phase after cycling testing, demonstrating the good structural and morphological stabilities of the nanocomposites during charge/discharge cycling.

The above electrochemical measurements show that the sandwich-like nanocomposite electrodes, especially TG3 sample, possess good lithium-storage properties in terms of specific capacity, rate capability, and cycling stability. There are several possible reasons which may be responsible for the superior lithium-storage performance of the mesoporous anatase TiO₂ sheets/rGO sandwich-like nanocomposites. (1) The mesoporous anatase TiO₂ nanosheet with micro-/nano-structures can prevent the aggregation of the nanocrystals when compared to the monodispersed case, which is favorable for preventing the undesirable aggregation of these nanobuildings and ensure the stability of the electrodes.^26,56 (2) The anatase TiO₂ sheets with mesoporous feature can provide extra active sites for the storage of lithium ions, accommodate the local volume change upon charge/discharge cycling, and reduce the effective diffusion distance for lithium ions and electrons, which are responsible for enhancing the specific capacity, improving the cycling performance and rate capabilities. (3) The addition of rGO nanosheets with good electrical conductivity can serve as the conductive channels, which decrease the inner resistance of LIBs and are essential for stabilizing the electronic and ionic conductivity, therefore leading to a higher reversible capacity.^32,40–44 Therefore, abundant mesopores with large specific surface area and suitable amount of rGO nanosheets are of importance for the lithium-storage properties, which is the main reason that TG3 sample shows superior performance.

4. Conclusions

In summary, mesoporous anatase TiO₂ sheets/rGO sandwich-like nanocomposites were successfully synthesized by acid-assisted TBOT hydrolysis and subsequent thermal reduction process. When used as the anode materials for LIBs, the as-prepared samples delivered reasonable capacity, good cycling stability and rate capability. The specific surface area, pore nature and the content of rGO of the sandwich-like nanocomposites have important influences on their electrochemical performances. The optimal sample possessed a reversible capacity of ∼161.4 mA h g⁻¹ after 50 cycles at a current rate of 0.5 C, and showed good cycling stability and rate capability. It is believed that the good electrochemical performance can be attributed to the mesoporous feature, the addition of rGO nanosheets, and the special designed sandwich-like electrode structure.

Acknowledgements

The authors would like to appreciate the financial supports from Natural Scientific Foundation of China (no. 51401114) and China Postdoctoral Science Foundation (20110490024). This work made use of the resources of the Beijing National Center for Electron Microscopy.

Notes and references

1. J. B. Goodenough and K.-S. Park, J. Am. Chem. Soc., 2013, 135, 1167.
2. N.-S. Choi, Z. H. Chen, S. A. Freunberger, X. L. Ji, Y.-K. Sun, K. Amine, G. Yushin, L. F. Nazar, J. Cho and P. G. Bruce, Angew. Chem., Int. Ed., 2012, 51, 2.
3. Q. F. Zhang, E. Uchaker, S. L. Candelariaza and G. Z. Cao, Chem. Soc. Rev., 2013, 42, 3127.
4. M. K. Song, S. Park, F. M. Alamgir, J. Cho and M. L. Liu, Mater. Sci. Eng., R, 2011, 72, 203.
5. P. Poizot, S. Laruelle, S. Grugeon, L. Dupont and J. M. Tarascon, Nature, 2000, 407, 496.
6. J. N. Tiwari, R. N. Tiwari and K. S. Kim, Prog. Mater. Sci., 2012, 57, 724.
7. D. R. Rolison, J. W. Long, J. C. Lytle, A. E. Fischer, C. P. Rhodes, T. M. McEvoy, M. E. Bourg and A. M. Lubers, Chem. Soc. Rev., 2009, 38, 226.
8. M. Wagemaker, A. P. M. Kentgens and F. M. Mulder, Nature, 2002, 418, 397.
9. E. J. W. Crossland, N. Noel, V. Sivaram, T. Leijtens, J. A. Alexander-Webber and H. J. Snaith, Nature, 2013, 495, 215.
10. I. Chung, B. Lee, J. Q. He, R. P. H. Chang and M. G. Kanatzidis, Nature, 2012, 485, 486.
11. S. W. Liu, J. G. Yu and M. Jaroniec, Chem. Mater., 2011, 23, 4085.
12. V. Subramanian, A. Karki, K. I. Gnanasekar, F. Posey Eddy and B. Rambabu, J. Power Sources, 2006, 159, 186.
13. C. Kim, R. Buonsanti, R. Yaylian, D. J. Milliron and J. Cabana, Adv. Energy Mater., 2013, 3, 1286.
14. G. S. Zakharova, C. Jähne, A. Popa, Ch. Täschner, Th. Gemming, A. Leonhardt, B. Büchner and R. Klingeler, J. Phys. Chem. C, 2012, 116, 8714.
15. S.-T. Myung, N. Takahashi, S. Komaba, C. S. Yoon, Y.-K. Sun, K. Amine and H. Yashiro, Adv. Funct. Mater., 2011, 21, 3231.
16. W. J. H. Borghols, D. Lützenkirchen-Hecht, U. Haake, W. Chan, U. Lafont, E. M. Kelder, E. R. H. van Eck, A. P. M. Kentgens, F. M. Mulder and M. Wagemaker, J. Electrochem. Soc., 2010, 157, A582.
17. X. W. Lou and L. A. Archer, Adv. Mater., 2008, 20, 1853.
18. H. B. Wu, J. S. Chen, H. H. Hng and X. W. Lou, Nanoscale, 2012, 4, 2526.
19. J. Wang, Y. K. Zhou, Y. Y. Hu, R. Hayre and Z. P. Shao, J. Phys. Chem. C, 2011, 115, 2529.
20. J. F. Ye, W. Liu, J. G. Cai, S. Chen, X. W. Zhao, H. H. Zhou and L. M. Qi, J. Am. Chem. Soc., 2011, 133, 933.
21. Y. Wang, X. W. Su and S. Lu, J. Mater. Chem., 2012, 22, 1969.
22. Z. X. Yang, G. D. Du, Q. Meng, Z. P. Guo, X. B. Yu, Z. X. Chen, T. L. Guo and R. Zeng, J. Mater. Chem., 2012, 22, 5848.
23. S. Myung, N. Takahashi, S. Komaba, C. S. Yoon, Y. Sun, K. Amine and H. Yashiro, Adv. Funct. Mater., 2011, 21, 3231.
24. J. S. Chen, Y. L. Tan, C. M. Li, Y. L. Cheah, D. Luan, S. Madhavi, F. Y. C. Boey, L. A. Archer and X. W. Lou, J. Am. Chem. Soc., 2010, 132, 6124.
25. C. H. Sun, X. H. Yang, J. S. Chen, Z. Li, X. W. Lou, C. Z. Li, S. C. Smith, G. Q. Lu and H. G. Yang, Chem. Commun., 2010, 46, 6129.
26. Y. L. Yu, X. L. Wang, H. Y. Sun and M. Ahmad, RSC Adv., 2012, 2, 7901.
27. Y. Wang, B. M. Smarsly and I. Djerdj, Chem. Mater., 2010, 22, 6624.
28. Y. G. Guo, Y. S. Hu, W. Sigle and J. Maier, Adv. Mater., 2007, 19, 2087.
29. H. Han, T. Song, J.-Y. Bae, L. F. Nazar, H. Kim and U. Paik, Energy Environ. Sci., 2011, 4, 4532.
30. W. Jiao, N. Li, L. Z. Wang, L. Wen, F. Li, G. Liu and H.-M. Cheng, Chem. Commun., 2013, 49, 3461.
31. X. Xin, X. F. Zhou, J. H. Wu, X. Y. Yao and Z. P. Liu, ACS Nano, 2012, 6, 11035.
32. H. Y. Sun, Y. G. Liu, Y. L. Yu, M. Ahmad, D. Nan and J. Zhu, Electrochim. Acta, 2014, 118, 1.
33. D. Li, D. Q. Shi, Z. X. Chen, H. K. Liu, D. Z. Jia and Z. P. Guo, RSC Adv., 2013, 3, 5003.
34. W. B. Yue, Z. Z. Lin, S. H. Jiang and X. J. Yang, J. Mater. Chem., 2012, 22, 16318.
35. X. L. Yang, K. C. Fan, Y. H. Zhu, J. S. Shen, X. Jiang, P. Zhao and C. Z. Li, J. Mater. Chem., 2010, 22, 17278.
36. S. M. Paek, E. Yoo and I. Honma, Nano Lett., 2009, 9, 72.
37. R. M. Westervelt, Science, 2008, 320, 324.
38. A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183.
39. Y. W. Zhu, S. Murali, M. D. Stoller, K. J. Ganesh, W. W. Cai, P. J. Ferreira, A. Pirkle, R. M. Wallace, K. A. Cychosz, M. Thommes, D. Su, E. A. Stach and R. S. Ruoff, Science, 2011, 332, 1537.
40. H. Q. Cao, B. J. Li, J. X. Zhang, F. Lian, X. H. Kong and M. Z. Qu, J. Mater. Chem., 2012, 22, 9759.
41. Y. F. Sun, J. B. Zhu, L. F. Bai, Q. Y. Li, X. Zhang, W. Tong and Y. Xie, Inorg. Chem. Front., 2014, 1, 58.
42. S. B. Yang, X. L. Feng and K. Müllen, Adv. Mater., 2011, 23, 3575.
43. W. Li, F. Wang, S. S. Feng, J. X. Wang, Z. K. Sun, B. Li, Y. H. Li, J. P. Yang, A. A. Elzatahry, Y. Y. Xia and D. Y. Zhao, J. Am. Chem. Soc., 2013, 135, 18300.
44. B. C. Qiu, M. Y. Xing and J. L. Zhang, J. Am. Chem. Soc., 2014, 136, 5852.
45. B. K. Guo, X. Q. Yu, X.-G. Sun, M. F. Chi, Z.-A. Qiao, J. Liu, Y.-S. Hu, X.-Q. Yang, J. B. Goodenough and S. Dai, Energy Environ. Sci., 2014, 7, 2220.
46. K. Saravanan, K. Ananthanarayanan and P. Balaya, Energy Environ. Sci., 2010, 3, 939.
47. X. X. Xue, W. Ji, Z. Mao, H. J. Mao, Y. Wang, X. Wang, W. D. Ruan, B. Zhao and J. R. Lombardi, J. Phys. Chem. C, 2012, 116, 8792.
48. P. Scherrer, Ber. Verh. Saechs. Akad. Wiss. Leipzig, Math.-Phys. Kl., 1918, 26, 98.
49. M. F. Ashby, P. J. Ferreira and D. L. Schodek, Nanomaterials, Nanotechnologies and Design: An Introduction for Engineers and Architects, Butterworth-Heinemann, 2009, p. 288.
50. Y. J. Zhang, M. Chu, L. Yang, W. F. Deng, Y. M. Tan, M. Ma and Q. J. Xie, Chem. Commun., 2014, 50, 6382.
51. Y. Lin, K. A. Watson, M. J. Fallbach, S. Ghose Jr, J. G. Smith, D. M. Delozier, W. Cao, R. E. Crooks and J. W. Connell, ACS Nano, 2009, 3, 871.
52. H. Y. Sun, M. Ahmad, J. Luo, Y. Y. Shi, W. C. Shen and J. Zhu, Mater. Res. Bull., 2014, 49, 319.
53. W. Wei, G. Oltean, C.-W. Tai, K. Edström, F. Björeforsa and L. Nyholma, J. Mater. Chem. A, 2013, 1, 8160.
54. P. Suresh Kumar, V. Aravindan, J. Sundaramurthy, V. Thavasi, S. G. Mhaisalkar, S. Ramakrishna and S. Madhavi, RSC Adv., 2012, 2, 7983.
55. G. F. Ortiz, I. Hanzu, T. Djenizian, P. Lavela, J. L. Tirado and P. Knauth, Chem. Mater., 2009, 21, 63.
56. Z. Q. Sun, J. H. Kim, Y. Zhao, F. Bijarbooneh, V. Malgras, Y. Lee, Y.-M. Kang and S. X. Dou, J. Am. Chem. Soc., 2011, 133, 19314.

---

Footnotes

† Electronic supplementary information (ESI) available: XRD pattern of the white precipitation, size distribution histograms, Raman spectra, TGA curves of the mesoporous anatase TiO₂ sheets/rGO sandwich-like nanocomposites, TEM images of the mesoporous anatase TiO₂ sheets/rGO sandwich-like nanocomposite electrodes after cycling testing. See DOI: 10.1039/c4ra04979a

‡ These authors contributed equally to this work.

---