Introduction

Various energy storage devices have been developed to solve the growing energy problem, such as supercapacitors, solar cells, and sodium-ion batteries. Among them, lithium-ion batteries are dominating the market due to their high-energy density, high cycle life, and eco-friendliness [1,2,3,4,5]. Titanium-based materials are regarded as promising anode materials in LIBs, among which titanium dioxide has drawn intensive interest because of its low cost, non-toxicity, and small volume change (< 4%) [6,7,8]. More importantly, titanium dioxide is highly safe as anode for LIBs owning to is electrochemically stable during Li+ insertion/extraction processes and can avoid the occurrence of lithium electroplating [9, 10]. Nevertheless, the practical application of TiO2 in LIBs has been severely hindered by poor electronic conductivity and low theoretical capacity [11, 12].

Fortunately, it has been demonstrated by many previous studies that nanostructured materials possess better lithium insertion/extraction kinetics and higher lithium storage capacity, improving the electrochemical performance by reducing the particle size of the electrode materials, has become a research hotspot [13,14,15]. In addition, constructing the hollow structure TiO2 materials has also been proposed to enhance the lithium storage performance. As is known, hollow structures exhibit large specific area and abundant pores, which can efficiently enhance the electrochemical properties of electrode including specific capacity, rate capability, and cycling stability [16, 17]. For example, Tian et al. [18] designed the TiO2 hollow nanowires with the diameter of 70 nm via chemical method followed by the calcination in a muffle furnace. The material shows the discharge capacity of 180 mA h g−1 at the current density of 0.2 C after 50 cycles. Gao et al. [19] prepared TiO2 microboxes by template-free method, and the obtained material exhibits rate performance with the discharge capacity of 150 mA h g−1 at the current density of 2 C.

In this work, TiO2 hollow nanospheres were successfully fabricated by a hard-template method. Compared with the solid TiO2 nanoparticles, it is suggested that the as-prepared H-TiO2 has unique advantages. (1) The large specific area of the H-TiO2 can not only provide more active sites for lithium storage but also keep an increased contact area between the electrodes and the electrolyte. (2) The hollow structure with abundant mesoporous of the H-TiO2 can efficiently promote transport rate of Li+ and e− in the electrodes. As expected, the HNS TiO2 used as anode materials for LIBs exhibit superior rate ability with a capacity of 101 mA h g−1 at a current density of 2 A g−1 and an admirable discharge capacity of 196 mA h g−1 at a current density of 0.1 A g−1 after 300 cycles.

Experimental sections

Synthesis of H-TiO2

A total of 5.8 ml of 28% ammonia solution and 2 ml of deionized water were added into 60 ml of ethanol under magnetic stirring, and then 4 ml of tetraethyl orthosilicate (TEOS) dispersed in 20 ml of ethanol was mixed into this solution. After stirring for 5 h, the white precipitate (SiO2) was obtained by centrifugation and washed three times with deionized water and ethanol, respectively. Then, the collected precipitate was redispersed in 35 ml of ethanol, followed by the addition 0.1 g of hydroxypropyl cellulose (HPC) and 0.5 ml deionized water. Next, 1.2 ml of titanium butoxide (TBOT) dissolved in 15 ml of ethanol was injected into above solution and reacted at 80 °C for 2.5 h. The resulting precipitate (TiO2@SiO2) was collected by centrifugation and washed three times with deionized water and ethanol, respectively. After that, the precipitate was calcined under argon gas atmosphere, and then the calcined powder was added into 15 ml of 0.1 M NaOH solution stirring for 3 h. Finally, the H-TiO2 were obtained by centrifugation and washed three times with deionized water and ethanol, respectively.

Synthesis of TiO2 nanoparticles

For comparison, TiO2 nanoparticles (N-TiO2) were also prepared; 2 ml TBOT was mixed with 60 ml acetone stirring for 0.5 h at room temperature and then transferred to a PTFE-lined reaction kettle and reacted at 200 °C for 2 h. Next, the white precipitate (TiO2) was obtained by centrifugation and washed three times with deionized water and acetone, respectively, followed by dried at 60 °C for 12 h and calcined at 600 °C for 3 h.

Materials characterization

The morphology and microstructural were analyzed with the scanning electron microscopy (SEM, Hitachi S4800) and transmission electron microscope (TEM, Tecnai-G2-F30 FEI with image corrector). The composition and crystal structure were characterized by X-ray diffraction (XRD, Rigaku, D/max-Rbusing Cu Ka radiation) measurement. The N2 adsorption/desorption isotherms were measured with Micromeritics ASAP 2010 instrument.

Electrochemical measurements

Electrochemical tests were performed using CR2032-type coin cells. The working electrodes were prepared by mixing the active materials, acetylene black, and polyvinylidene fluoride (PVDF) with a weight ratio of 7:2:1 in N-methyl-2-pyrrolidone (NMP) to form a slurry. The slurry was uniformly spread on a copper foil. Pure lithium foil was used as the counter electrode. Celgard2400 was used as separator. A 1 M solution of LiPF6 dissolved in ethylene carbonate and dimethyl carbonate (1:1 in volume ratio) was used as the electrolyte. The lithium half-cells were assembled in an argon-filled glovebox with both water and oxygen contents below 0.1 ppm. Cyclic voltammetry (CV) data were recorded using a PGSTAT302N electrochemical workstation. Galvanostatic discharge-charge curves were collected on a Neware battery test system within a voltage range of 1–3 V (vs Li+/Li). Electrochemical impedance spectra (EIS) were also carried out on a PGSTAT302N electrochemical workstation in the frequency range of 0.1 Hz–100 kHz.

Results and discussion

A brief schematic diagram of the preparation process of H-TiO2 is shown in Fig. 1a. The as-prepared uniformly sized SiO2 was used as template to synthesize the TiO2@SiO2 precursor, which was then etched with NaOH solution to remove the SiO2, resulting in H-TiO2, and the detailed growth mechanism of H-TiO2 as shown in supplementary information (SI). Fig. 2b shows the SEM image of the H-TiO2. It can be clearly seen that these samples exhibit spherical structure with a uniform diameter of ∼200 nm. Interestingly, several broken spheres can be observed, which reveals the hollow structure of the obtained TiO2 materials. Their hollow interiors are further elucidated by TEM. Fig. 3c reveals a clear inner cavity by obvious comparison of the hollow inner cavity and the hollow outer cavity, which indicating that TiO2@SiO2 precursor were completely converted into TiO2 hollow spherical structure, and the thickness of the H-TiO2 shell is about 15 nm. The HRTEM image of the H-TiO2 is also provided in Fig. 3d, a clear lattice with an interlayer spacing of 0.35 nm can be observed, which coinciding well with the (101) crystal planes of anatase TiO2. The phase purity and crystalline structure of the H-TiO2 were tested by X-ray diffraction (XRD) measurement, and the corresponding XRD pattern as shown in Fig. 2c. As can be seen, all the intensive diffraction peaks were well assigned to anatase TiO2 (JCPDS no.21-1272) [20, 21]. And no peaks were observed for the other phases, indicating their high purity. Nitrogen adsorption-desorption measurements were used to investigate the specific surface area and pore size distribution of the H-TiO2. The N2 adsorption/desorption isotherms in Fig. 2d depict typical Type IV curves, corresponding to the characteristic isotherms of mesoporous materials [11, 22]. The Brunner-Emmett-Teller (BET) specific surface area of the H-TiO2 yields to be ∼225 m2 g−1. The pore size distribution curve of the H-TiO2 (inset of Fig. 2d) confirms the existence of mesopores with size distribution centering at ∼7.8 nm. It is worth noting that mesopores can further facilitate Li+ diffusion in the electrodes and shorten the Li+ and e− transport length [23].

Fig. 1
figure 1

a Schematic illustrations of synthesized process of H-TiO2; SEM images of b SiO2, c TiO2@SiO2, and d H-TiO2

Full size image
Fig. 2
figure 2

SEM images of a N-TiO2, b H-TiO2; c XRD patterns of H-TiO2 and N-TiO2; d nitrogen adsorption–desorption isotherms of H-TiO2, the inset shows the pore size distribution

Full size image
Fig. 3
figure 3

TEM images of a SiO2, b TiO2@SiO2, and c H-TiO2; d HRTEM image of H-TiO2

Full size image

The H-TiO2 were evaluated as anode materials for lithium storage properties in LIBs. The electrochemical properties of the H-TiO2 were investigated by cyclic voltammetry (CV) in the voltage range of 1–3 V vs Li+/Li. Fig. 4a shows the CV curves of the H-TiO2 for the first three cycles at scan rate of 0.1 mV s−1. In the first cycle, a couple of current peaks located at 1.68 V and 2.05 V can be observed, corresponding to the insertion and extraction of lithium ions, respectively [24, 25]. In the second cycle, the reduction peak shifted to a higher potential of 1.7 V and the peak current increased slightly, indicating an activation process. Besides, both the reduction and oxidation peaks of the third cycle almost overlap with the second cycle, which implies that the H-TiO2 exhibits good reversibility of electrochemical reactions.

Fig. 4
figure 4

a CV curves of the H-TiO2 at a scan rate of 0.1 mV s−1; b charge and discharge profiles of the H-TiO2 for the first three cycles at 0.1 A g−1; c cycling performance of H-TiO2 and N-TiO2 at 0.1 A g−1; d rate performance of H-TiO2 and N-TiO2 at different current rates from 0.1 to 2 A g−1

Full size image

The charge and discharge curves of the H-TiO2 at a current density of 0.1 A g−1 are shown in Fig. 3b. The first discharge and charge capacities are 289 and 225 mA h g−1, respectively, and the initial coulombic efficiency (CE) was 77.9%; the loss of capacity is caused by the formation of the solid electrolyte interface (SEI) [26, 27]. And the subsequent charge and discharge curves coincide very well, suggesting the excellent electrochemical reversibility of the H-TiO2. In addition, the curves exhibit two obvious voltage plateaus, 1.7 V for lithium insertion and 2.1 V for lithium extraction, which is in good agreement with the CV curves. Fig. 3c shows the cycling performance of H-TiO2 and N-TiO2 at a current rate of 0.1 A g−1. It is obviously observed that the H-TiO2 exhibits higher discharge capacity of 202 mA h g−1, and there is no rapid capacity decay during the first 15 cycles, which suggesting that H-TiO2 has a superior cycling performance than N-TiO2. The cycling performance of the H-TiO2 electrode is also superior to that of many similar TiO2-based electrodes, as shown in Table 1. H-TiO2 and N-TiO2 were also investigated for rate capability (Fig. 4d). As expected, the H-TiO2 shows higher discharge capacities of 198, 180, 158, and 135 mA h g−1 at current rates of 0.1, 0.2, 0.5, and 1 A g−1, respectively. Even at a very high current rate of 2 A g−1, a capacity of 98 mA h g−1 can be still achieved. Compared with N-TiO2, a discharge capacity of 197 mA h g−1 can be recovered when the current rate reduces back to 0.1 A g−1. This demonstrates the superior rate performance and structure stability of H-TiO2, which could be ascribed to that the hollow structure can shorten the diffusion path for Li+ and ensure increased contact area between electrodes and electrolyte.

Table 1 Performance comparison of related TiO2-based materials
Full size table

The electrochemical impedance spectroscopy was performed to study the resistance property of H-TiO2 and N-TiO2. The Nyquist plots display a semicircle in high to medium frequency and a slope line in the low frequency, attributing to charge transfer resistances (Rct) and Li-ion diffusion resistances, as shown in Fig. 5 [30, 31]. The corresponding Rct values were obtained by measuring the diameter of semicircle that H-TiO2 and N-TiO2 before and after cycling 100th show the values of 78 Ω/97 Ω and 83 Ω/112 Ω, respectively. The H-TiO2 presents lower Rct value than N-TiO2 before and after cycling, indicating better Li-ion transfer ability of H-TiO2. And the Rct values of H-TiO2 only slightly increase, demonstrating a stable charge/discharge reaction [32, 33]. The basis of lithium storage from the Li-ion diffusion of the two electrodes was investigated by CV measurements at various scan rates ranging from 0.1 to 5 mV s−1 (Fig. 6a and b). The linear relationship between peak current density (Ip) and the square root of scan rates is correlated to the corresponding Li-ion diffusion. As can be observed from Fig. 6c, the H-TiO2 electrode exhibits larger slope than N-TiO2 electrode, indicating better Li-ion diffusion in the H-TiO2 electrode. In addition, based on the classical Randles-Sevcik equation, the corresponding Li-ion diffusion coefficient can be calculated [34]:

$$ {\mathrm{I}}_{\mathrm{p}}=2.6\times {10}^5{\mathrm{n}}^{1.5}{{\mathrm{AD}}_{\mathrm{Li}}}^{0.5}{\mathrm{v}}^{0.5}\mathrm{C} $$

where Ip is the peak current density (A g−1), n is the number of reaction electrons in LIBs, A is the electrode area (cm−2), v is the scan rates (V s−1), DLi is the Li-ion diffusion coefficient (cm2 s−1), and C is the Li-ion concentration (mol ml−1). The corresponding Li-ion diffusion coefficients of H-TiO2 electrode are larger than N-TiO2 electrode, which further suggesting the superior Li-ion diffusion property in the H-TiO2 electrode. This may be attributed to the hollow structure with abundant mesoporous, which can not only provide more channels for Li-ion diffusion but also shorten the transport pathways for Li-ion.

Fig. 5
figure 5

Nyquist plot of H-TiO2 and N-TiO2 a before and b after 100th cycles (inset is the equivalent circuit model)

Full size image
Fig. 6
figure 6

CV curves of a H-TiO2 and b N-TiO2 at different scan rates; c the cathodic reaction versus the square root of scan rates

Full size image

Conclusion

In summary, the TiO2 hollow nanospheres have been efficiently prepared via a hard-template method. Owning to large specific area and rich mesoporous of the hollow spherical structure, the as-obtained material used as anode for LIBs exhibits high reversible capacity, superior rate capability, and excellent long-term cycling stability. The excellent electrochemical performance makes the H-TiO2 an ideal candidate for high-energy anode materials in LIBs.