Introduction

To address the growing energy problem, various energy storage devices have been developed such as supercapacitors, solar cells, and sodium-ion batteries [1,2,3]. Among them, lithium-ion batteries dominate the market due to their high energy density, long cycle life, and eco-friendliness [4, 5]. However, the available commercial graphite anode has an inferior theoretical capacity and may generate the lithium dendrites on the surface of electrodes [6, 7]. In this regard, SiO2 has been considered as the promising anode candidate owing to the low operating voltage, abundant natural resources, and its fivefold increase of theoretical capacity (1965 m Ah g−1) [8, 9]. Nevertheless, the practical application of SiO2 in LIBs is severely hampered by poor electrical conductivity and drastic volume change during lithiation/delithiation processes [10].

Great efforts have been made to enhance the stability and electrical conductivity of SiO2, and nanostructured SiO2 materials have been shown to possess better lithium insertion/extraction kinetics compared to bulk SiO2 [11,12,13]. Another promising approach to achieve remarkable improvement in cycling stability is to obtain a core–shell structure by surface coating with conductive materials, which not only improves the conductivity of the electrode but also avoids direct contact between electrodes and electrolyte, thus reducing the uncontrollable growth of the solid electrolyte interphase (SEI) film [14, 15]. For instance, Jiang et al. prepared a SiO2@C composite by molecular self-assembly method, which showed good rate capability (316 m Ah g−1 at 0.4 A g−1) [16]. Moreover, Dong and co-workers have proposed three-dimensional SiO2/nitrogen-doped graphene aerogels, which achieve high cycling performance (1000 m Ah g−1 at 0.1 A g−1 after 100 cycles) [17]. In the case of most carbonaceous coating layers, however, it is usually frustrating that the low initial coulombic efficiency (ICE) and safe issues [18, 19]. As we know that TiO2 undergoes only slight volume expansion upon lithiation (< 4%), and the resultant lithiated TiO2 during discharge process can promote the electrical conductivity of the electrode [20]. Recently, outer layers of a-TiO2 with structure elasticity is believed to be capable of maintaining structural integrity of SiO2 anodes, and its intrinsically isotropic and open active diffusion channels promote the mobility of Li+ and the diffusion of ions and e−, leading to high ion accessibility and enhanced power and energy density of the electrode materials [21, 22].

Inspired by the above, in this work, we synthesized core–shell structured SiO2@a-TiO2 composite by a simple sol–gel method. This intriguing structure possesses some advantages as follows: (a) the core of SiO2 nanosphere contributes to high capacity; (b) the a-TiO2 shell shows elastic behavior with strain relaxation, thereby maintaining the integrity of the anode and alleviating the volume change during discharging/charging processes. In addition, it provides large number of Li+ and e− diffusion channels to improve Li+ availability and electrochemical reactivity. As expected, the SiO2@a-TiO2 electrode delivers a high ICE of 53% and adorable discharge capacity of 582 m Ah g−1 over 300 cycles at 0.1 A g−1.

Experimental

Materials synthesis

The schematic diagram of the preparation process of SiO2@a-TiO2 composite is shown in Fig. 1. Precisely, 5.8 ml of ammonia solution (28 wt%) 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 with deionized water and ethanol, respectively. Then, the collected SiO2 precipitate was redispersed in 35 ml of ethanol, followed by the addition of 0.1 g hydroxypropyl cellulose (HPC) and 0.5 ml deionized water. Next, 1.2 ml of tetrabutyl titanate (TBOT) dissolved in 15 ml of ethanol was injected into the above solution and reacted at 80 °C for 2 h. The resulting precipitate (SiO2@a-TiO2) was collected by centrifugation and washed with deionized water and ethanol, respectively. And in this composite, the weight percentage of a-TiO2 can be calculated to be about 20.6%.

Fig. 1
figure 1

The schematic of preparing the core–shell SiO2 @a-TiO2 composite

Full size image

As a comparison, SiO2@crystalline TiO2 (SiO2@c-TiO2) composites were synthesized under a similar procedure with the additional process of calcination at 600 °C for 3 h.

Materials characterization

The morphology and microstructure 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 measurement (XRD, Rigaku, D/max-Rbusing Cu Ka radiation). The chemical state was investigated using an X-ray photoelectron spectroscope (XPS) (Thermo Scientific K-Alpha instrument with Al Kα source).

Electrochemical measurements

Electrochemical tests were performed using CR2032-type coin cells. The working electrodes were prepared by mixing the active materials, acetylene black and poly vinylidene fluoride (PVDF) in Nmethyl-2-pyrrolidone (NMP) with a weight ratio of 7:2:1 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 levels below 0.1 ppm. Cyclic voltammetry (CV) data were recorded using an CHI760E electrochemical workstation. Electrochemical performance was collected on a Neware battery test system. Electrochemical impedance spectra (EIS) were also carried out on an CHI760E electrochemical workstation.

Results and discussion

The morphology and microstructure of the prepared samples were characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). As shown in Fig. 2a–c, these samples all exhibit a spherical structure with a uniform diameter of about 200 nm. Note that the surface of SiO2 is relatively smooth compared to SiO2@a-TiO2, which indicates that a-TiO2 is coated on the surface of the SiO2 nanospheres. TEM image (Fig. 2d) also shows that the SiO2 nanospheres display a size of about 200 nm, which is in good agreement with SEM image (Fig. 2a). Compared with the SiO2 nanospheres, a uniform coating layer was observed on the surface of SiO2@a-TiO2 (Fig. 2e), and the thickness of a-TiO2 is about 15 nm, which could be observed from Fig. S1 [23]. Figure 1f further shows the HRTEM image of SiO2@a-TiO2; there is no obvious lattice spacing corresponding to crystalline TiO2, indicating that the TiO2 coating layer is amorphous. Elemental mapping results are displayed in Fig. 2g and reveal the uniform distribution of Si, O, and Ti in the SiO2@a-TiO2 composite.

Fig. 2
figure 2

SEM images of a SiO2 and b–c SiO2@a-TiO2; TEM images of d SiO2 and e SiO2@a-TiO2. f HRTEM image of SiO2@a-TiO2. g TEM image and corresponding elemental mappings (Si, O, and Ti elements) of SiO2@a-TiO2

Full size image

The XRD patterns are shown in Fig. 2c. Note that only a broad peak at 22.3° can be observed in the XRD pattern of SiO2, indicating its amorphous nature [24, 25]. In addition, no obvious diffraction peak can be detected from the pattern of SiO2@a-TiO2, demonstrating that the surface-coated TiO2 shell is also amorphous. Moreover, several new diffraction peaks can be observed in the SiO2@c-TiO2 composite pattern at 24.3°, 37.8°, 47.6°, 54.5°, and 55.3°, corresponding to the (101), (004), (200), (105), and (211) lattice planes of anatase TiO2 (JCPDS no.21–1272), respectively, which proves that the a-TiO2 is successfully converted into crystalline TiO2 after the calcination process [26, 27]. XPS measurements were used to evaluate the chemical states of the elements in the SiO2 and SiO2@a-TiO2 composite. In the XPS survey spectrums, several distinct peaks at ∼ 103, 154, 284, and 534 eV can be found, corresponding to Si 2p, Si 2 s, C 1 s, and O 1 s of SiO2 (Fig. 3b) [28, 29]. The C 1 s peak may be assigned to tape substrate [30]. In contrast to the survey spectrum of SiO2, a unique peak of Ti 2p can be clearly noticed in the XPS survey spectrum of SiO2@a-TiO2 composite, indicating that the SiO2 cores are well encapsulated by the a-TiO2 shells. The Si 2p3/2 peak is evidently prominent at 103.6 eV (Fig. 3c), corresponding to Si4+. In addition, the high-resolution XPS spectrum shows Ti 2p3/2 and Ti 2p1/2 peaks of TiO2 located at 458.2 and 464.6 eV (Fig. 3d) [31, 32].

Fig. 3
figure 3

a XRD patterns of SiO2, SiO2@c-TiO2, and SiO2@a-TiO2; b full XPS spectrum of the SiO2 and SiO2@a-TiO2; Ti 2p c and O1s d XPS high-resolution spectra of the SiO2@a-TiO2

Full size image

The SiO2@a-TiO2 composite was evaluated as anode materials for lithium storage properties in LIBs. The electrochemical redox reactions of this composite were studied by cyclic voltammetry from 0.01 and 3 V at a scan rate of 0.1 mV s−1. Figure 4a shows the CV curves for the first three cycles, and it is obvious that a wide and weak cathodic peak appears about 1.46 V during the initial cycle, which could be caused by the irreversible reaction between Li+ and SiO2 as in Eqs. (1) and (2) [33]. In the subsequent lithiation process, broad cathodic peak at about 0.85 V and 0.56 V can be observed, which could be related to the side reaction of irreversible Li+ insertion together with the formation of SEI film [34]. Another cathodic peak around 0 V corresponds to the alloy reaction between Li+ and Si as in Eq. (3). In addition, two anodic peaks at about 0.25 and 1.05 V corresponding to de-alloy reaction can be observed in the subsequent delithiation process, as in Eq. (4) [35, 36].

Fig. 4
figure 4

a CV curves for the first three cycles of SiO2@a-TiO2 at 0.1 mV s−1; b discharge and charge profiles for the 1st, 2nd, and 300th of SiO2@a-TiO2 at 0.1 A g−1; c rate capability at different current rates, and d cycling performances and CE of SiO2, SiO2@c-TiO2, and SiO2@a-TiO2 at 0.1 A g−1

Full size image
$$2Si{O}_{2}+4{Li}^{+}+4{e}^{\_}\to {Li}_{4}Si{O}_{4}+Si$$
(1)
$$Si{O}_{2}+4{Li}^{+}+4{e}^{\_}\to {2Li}_{2}O+Si$$
(2)
$$Si+x{Li}^{+}+x{e}^{\_}\to {Li}_{x}Si$$
(3)
$${Li}_{x}Si\to Si+x{Li}^{+}+x{e}^{\_}$$
(4)

Figure 4b shows the 1st, 2nd, and 300th galvanostatic discharge/charge profiles of the SiO2@a-TiO2 composite. As can be observed, the composite delivers an initial discharge and charge capacity of 1125 m Ah g−1 and 597 m Ah g−1, respectively. Meanwhile, the initial discharge/charge capacity of SiO2 and SiO2@c-TiO2 were 1046/481 m Ah g−1 (Fig. S2a) and 891/338 m Ah g−1 (Fig. S2b), respectively. These samples all suffered serious capacity loss, which could be caused by the irreversible reaction and the formation of SEI film. However, what is gratifying is that the SiO2@a-TiO2 composite also exhibits intriguing stability, with capacity retention of 92% from the 2nd to 300th cycle, which is superior to that 77.6% of SiO2 and 63.8% of SiO2@c-TiO2. Figure 4c shows the rate curves of the three electrodes; it can be observed that the SiO2@a-TiO2 composite delivers discharge capacities of 581, 548, 472, and 395 m Ah g−1 at current density of 0.1, 0.2, 0.5, and 1 A g−1. Even at an ultrahigh current density of 2 A g−1, it can still achieve a capacity of 387 m Ah g−1, with a retention of 56% relative to the value at 0.1 A g−1. Moreover, a discharge capacity of 578 m Ah g−1 can be recovered when the current density goes back to 0.1 A g−1, indicating the superior rate performance of the SiO2@a-TiO2 composite. The composite also exhibits excellent cycling performance as shown in Fig. 4d. It delivers an initial discharge capacity of 1125 m Ah g−1, then decays to 580 m Ah g−1 in the first 50 cycles, and remains very stable value of 572 m Ah g−1 up to 300 cycles. In addition, SiO2@a-TiO2 composites not only provide a high ICE of 53%, higher than that 37% of SiO2 and 46% of SiO2@c-TiO2, but also maintain a stable CE value of nearly 100% over 2 to 300 cycles. The SiO2@a-TiO2 shows superior cycling performance than SiO2 and SiO2@c-TiO2, resulting from the improvement of electrical conductivity and slighter volume variation of the electrode.

Electrochemical impedance spectroscopy measurements were also carried out on three samples to study the resistance property. Figure 5a shows the Nyquist plots and equivalent circuit; it is obvious that all the plots contain a semicircle in high- to middle-frequency region and an inclined line in low-frequency region, of which the diameter correlates with the charge transfer resistance at the electrodes and electrolyte interface (Rct), and a low-frequency sloping line, which can associate with Warburg impedance (Zw), is corresponding to the spread of Li+ in a large proportion of the electrode materials [37]. The simplified equivalent circuit for the simulation of EIS, in which Rs, Rct, CPE, and Wo accordingly represent the solution resistance, charge transfer resistance, constant phase element, and Warburg resistance [38]. The fitting values of various electrodes are exhibited in Table 1. The value of Rs for each electrode is about 3 Ω. It is distinct that the SiO2@a-TiO2 has lower Rct (80.5 Ω) comparing with SiO2@c-TiO2 (115.2 Ω) and SiO2 nanospheres (132.6 Ω). In addition, as shown in Fig. 5b, the Rct value of SiO2@a-TiO2 electrode slightly increased after cycling, indicating that a-TiO2 could not only maintain the stable structure of the electrode during cycling, but also ensure the rapid transport of Li+ and e− in the electrode.

Fig. 5
figure 5

Nyquist plots of a SiO2, SiO2@c-TiO2, and SiO2@a-TiO2 before cycling, and b SiO2@a-TiO2 after cycling 1st, 50th, and 300th; c the line relationship of Z′ ~ ω−1/2; d the diagram of battery

Full size image
Table 1 Impedance parameters and lithium-ion diffusion coefficient of various electrode
Full size table

The slope of the line located at the lower frequency corresponds to the diffusion process of Li+ within the electrode [20]. And the DLi+ of the corresponding sample can be calculated by Eqs. (5) and (6) [39].

$${D}_{{(Li}^{+})}=\frac{{R}^{2}{T}^{2}}{2{A}^{2}{n}^{4}{F}^{4 }{C}^{2}{\sigma }^{2}}$$
(5)
$${Z}_{re}= {R}_{ct}+ {R}_{s}+\sigma {\omega }^{-1/2}$$
(6)

where the R, T, A, and F stood for the gas constant, absolute temperature, surface area of the pole piece, and Faraday constant, respectively. C denoted the volume concentration of Li+. n represented the number of transferred electrons. For Eq. (4), σ donated the slope of Zre and ω−1/2. The ω represents the angular frequency. As shown in Table 1, the SiO2@a-TiO2 electrode possesses the highest DLi+ value of 1.51 × 10−13, which is superior to the value of the pure SiO2 electrode (2.85 × 10−14) and SiO2@c-TiO2 electrode (7.38 × 10−14). The results may be attributed to the surface coating of a-TiO2 with more diffusion channels, which is beneficial for improving the electrical conductivity of electrode by providing faster lithium ions transfer.

To further observe the structural changes of the electrodes during long-term cycling, the morphologies of the electrode surfaces of SiO2, SiO2@c-TiO2, and SiO2@a-TiO2 were investigated after 300 cycles, as shown in Fig. 6a–c. It can be clearly seen that more obvious and wider cracks appear on the electrode surfaces of SiO2 and SiO2 @c-TiO2 after long-term cycling. In contrast, the cracks of the SiO2@a-TiO2 electrode are much less and smaller than those of the other two electrodes. In addition, Figs. 6d–f show the microstructure of the high magnification images of the various electrodes, and it is obvious that there are distinct granules on the SiO2 and SiO2@c-TiO2 electrodes, which may represent aggregates of electrode materials and conductive agents. The aggregation of powders leads to a non-uniform distribution on the electrode, which in turn loses contact with the conducting agent and leads to a decrease in electronic conductivity. Therefore, a-TiO2 coating layer can effectively mitigate the volume expansion of SiO2 and form a stable SEI film on the surface of SiO2. And a mechanism diagram of material surface on the performance of these electrodes is also demonstrated in Fig. 7.

Fig. 6
figure 6

SEM images of a and d SiO2, b and e SiO2@c-TiO2, and c and f SiO2@a-TiO2 after 300 cycles

Full size image
Fig. 7
figure 7

Mechanism diagram of material surface of a SiO2, b SiO2@c-TiO2, and c SiO2@a-TiO2

Full size image

Conclusion

In summary, the core–shell nano-structured SiO2@a-TiO2 composite was successfully obtained by sol–gel method. The composite, coated by a-TiO2, which is beneficial for alleviating the volume expansion of SiO2 and providing more diffusion channels for the transfer of Li+ and e−, delivered excellent discharge capacity of 582 m Ah g−1 at 0.1 A g−1 after 300 cycles, and superior rate performance with a discharge capacity of 387 m Ah g−1 at 2 A g−1. These results demonstrated that nano SiO2 coated with a-TiO2 is promising anode materials for LIBs. In addition, this facial sol–gel synthetic route could inspire further capability on constructing core–shell structured materials for energy storage.