Sol–gel synthesis of nanocrystal-constructed hierarchically porous TiO₂ based composites for lithium ion batteries

Abstract

Hierarchically porous TiO₂ based composites (pure TiO₂ and TiO₂/carbon (TiO₂/C) composite) were synthesized by a facile sol–gel process followed by post-calcination. Poly(vinylpyrrolidone) (PVP) acts as a phase separation inducer as well as a carbon source. The as-prepared TiO₂ based composites possess an interesting hierarchically porous structure constructed of cocontinuous macropores and mesoporous skeletons consisting of interconnected nanocrystals and in situ distributed carbon. The hierarchically porous TiO₂/C composite shows excellent electrochemical performance with fast lithium ion diffusion and electronic transport, resulting from the hierarchically porous structure and conductive carbon material. The TiO₂/C composite calcined at 500 °C exhibits the highest BET surface area of 170 m² g⁻¹, superior cycling stability (delivers a remarkable discharge capacity of 132 mA h g⁻¹ at 1 C after 100 cycles) and excellent rate capability (over 96 mA h g⁻¹ at 30 C rate). The results indicate that these hierarchically porous TiO₂ based composites could be promising anode materials for high performance lithium ion batteries.

---

Introduction

Titanium oxide (TiO₂) has received considerable attention because of its photoelectric properties, biocompatibility and chemical stability, and has been extensively applied in various kinds of fields such as photocatalysis,^1,2 sensing,³ dye-sensitized solar cells,^4–6 and lithium-ion batteries (LIBs),^7–11 etc. Especially, in the case of LIBs applications, TiO₂ is regarded as one of the most promising anode materials due to its long cycle life, low cost, environmental compatibility, and low volume expansion.^7,12,13 Furthermore, compared with the traditional graphite anode materials, the TiO₂ anode shows a relatively higher lithium insertion/extraction voltage (higher than 1.5 V vs. Li⁺/Li) to ensure adequate safety by efficiently avoiding the formation of solid electrolyte interphase (SEI) films and lithium dendrites.^14,15 However, the practical application of the TiO₂ anode is still hampered by limited rate capability arising from its low intrinsic electrical conductivity (10⁻¹² S cm⁻¹).^16,17

Recently, great efforts have been devoted to circumvent the drawback of poor rate capability of TiO₂ electrodes.^18–23 Fabrication of porous structure is an effective approach to improve the electrochemical performance by supporting sufficiently rapid transport of both lithium ions and electrons.^20,24–29 Different types of pores have various effects on the electrochemical of electrodes.^24,30,31 The micropores provide more active sites for Li-insertion by increasing specific surface area, while the mesopores reduce the length of lithium ions and electrons by providing interconnected nanopaths. Meanwhile, the macropores could effectively accommodate electrolytes for the electrochemical reaction. Thus, hierarchically porous structure could offer a good choice to improve electrode properties with multiple benefits arising from different pore size regimes. Additionally, another effective strategy to improve electron transport by decorating TiO₂ materials with high conductivity materials, such as metals,³² metal oxides,^33,34 and carbonaceous materials.^{16,24,26,35–42} Among these conductivity materials, carbonaceous materials have been demonstrated to effectively increase electron transport to further enhance the electrochemical performance of TiO₂ electrode, especially the rate capability. As a result, it is highly desirable to design and fabricate new TiO₂ electrode materials combining hierarchically porous structure with highly conductive carbonaceous materials.

In this paper, we report a facile and convenient method for preparing hierarchically porous pure TiO₂ and TiO₂/C composite by a combination of sol–gel with the aid of PVP, followed by post-calcination in air and argon atmosphere, respectively. The specific procedure is schematically illustrated in Fig. 1. The as-prepared TiO₂ based composites reveal some distinguishing features: (i) both of the pure TiO₂ and TiO₂/C composite show a hierarchically porous structure (Fig. 1b) with a high specific surface area, enlarging the contact surface between electrode and electrolyte. (ii) The cocontinuous skeletons are constructed by many interconnected nanocrystals with sizes of 10–30 nm (Fig. 1c), which shorten the diffusion distance of Li-ion for the fast charge–discharge cycles. (iii) During the procedure, PVP is used as a phase separation inducer as well as a carbon source resulting in the in situ formed TiO₂/C composites, which ensures an intimate contact of carbon and TiO₂, providing excellent electrical conductivity for high rate performance. The structure and electrochemical performance of pure TiO₂ and TiO₂/C composite calcined at different temperatures are investigated. The resultant TiO₂/C composite exhibits excellent rate and cycling performance due to the synergistic effect of the hierarchically porous structure resulting in large specific surface area and enhanced electrical conductivity ascribe to the in situ formed carbon derived from PVP.

Fig. 1 (a) Schematic illustration of the experimental procedure of hierarchically porous TiO₂-based monoliths. (b) Macrochannels are derived from phase separation. (c) Mesoporous skeletons are composed of interconnected nanocrystals and in situ distributed carbon.

Experimental

Materials synthesis

TiOSO₄·xH₂O (Sigma-Aldrich) was utilized as the titanium source. Mixtures of distilled water and ethylene glycol (EG, Sinopharm Chemical Reagent Co., Ltd (China)) were used as the solvents. Formamide (FA, Shanghai Lingfeng Chemical Reagent Co., Ltd) was utilized to initiate the gelation. Poly(vinylpyrrolidone) (PVP, Sigma-Aldrich, M_w = 10 000) was used as a polymer to induce the phase separation and a carbon source after calcination in an argon atmosphere. All chemicals were of analytical grade and used as received without further purification.

The experimental procedure of TiO₂-based composites is shown in Fig. 1a. The typical synthesis of the titanium sol was as follows in detail. Firstly, 1 g TiOSO₄·xH₂O was dissolved in distilled water and continuously stirred for 2 h to obtain a clear homogeneous solution. Then 0.6 mL of ethylene glycol (EG) and 0.2 g of poly(vinylpyrrolidone) (PVP) were added into the above solution, and after stirring for 30 min to obtain pale yellow solution, 0.5 mL of formamide (FA) was added under vigorous stirring. The mix solution was kept in a sealed container at 60 °C for gelation. The whole gelation process can be finished within about 2 h, during the process the colour of solution was changed from pale yellow to clear and finally to white, and the milk white monolith can be obtained after the gelation. The wet gel was subsequently aged at 60 °C for 12 h, then subjected to solvent exchange to remove the residues by ethanol for 3 times every 12 h and 2-propanol for 12 h, respectively. The dry gel was obtained after evaporation-dried at 40 °C for a week, which presents a about 3–5% volume shrinkage induced by the solvent exchange and subsequent evaporation drying. Some of the dried gels were heat-treated at various temperatures (500, 600 and 800 °C) for 4 h in air to obtain pure TiO₂, respectively, and TiO₂/C composites were prepared by calcination of the dried gel in an argon atmosphere.

Structure characterizations

The crystallinity of the samples was characterized using powder X-ray diffractometry (XRD; X'Pert Pro diffractometer with a Cu Kα radiation, λ = 0.15418 nm). The morphology of the samples was observed using scanning electron microscopy (SEM; Hitachi S-4800) and transmission electron microscopy (TEM; FEI, Tecnai G2 F20), equipped with an energy dispersive spectroscopy (EDS) detector. Fourier transform infrared spectra (FTIR) of powder samples diluted with KBr were recorded with a Bruker VERTEX 70. The composition of the products was performed using X-ray photoelectron spectroscopy (XPS, Escalab 250Xi, UK), equipped with a hemispherical electron energy analyzer using a monochromatic Al Kα X-ray radiation (15 KV, 1486.6 eV) under a vacuum pressure of 5 × 10⁻¹⁰ Torr. A wide-scan survey spectrum was performed on each sample to determine the elements state at the surface. High-resolution spectra of the Ti 2p and O 1s core lines were collected using a pass energy of 20 eV, a step size of 0.05 eV, and a sweep time of 180 s. The thermal decomposition behavior of the dried gel and the carbon content of TiO₂/C composite were examined by thermogravimetric analysis (TGA, TA Q500) at a heating rate of 10 °C min⁻¹ from room temperature (25 °C) to 800 °C in air and N₂ atmosphere, respectively. Measurements of Raman spectra were performed on a Renishaw InVia Raman spectrometer under a backscattering geometry (λ = 532 nm). Nitrogen adsorption–desorption apparatus (AUTOSORB-1-C) was used to characterize the meso- and micropores of the samples. The pore size distribution was calculated using the Barrett–Joyner–Halenda (BJH) method based on the adsorption branch.

Electrochemical measurements

Electrochemical experiments were carried out using standard CR 2025 type coin half cells. The as-prepared samples (pure TiO₂ and TiO₂/C composite) were mixed with carbon black (Super P) and polyvinylidene fluoride (PVDF) binder at a weight ratio of 80 : 10 : 10 in N-methyl-2-pyrrolidene (NMP) solution. The slurry was pasted on a Cu foil and dried in a vacuum oven at 120 °C for 12 h to serve as the working electrode. The weight of the active material in the electrode sheet was about 2.5 mg cm². The pure lithium foil was used as both counter electrode and reference electrode. A solution of 1 M LiPF₆ in ethylene carbonate (EC)/dimethyl carbonate (DMC) (1 : 1 by volume) was used as the electrolyte with a polypropylene microporous film (Cellgard 2300) as the separator. Galvanostatical charge–discharge experiments were tested at different current densities in a voltage range of 1.0–3.0 V on a battery test system (Shenzhen Neware Battery, China). Cyclic voltammetry (CV) measurements and Electrochemical impedance spectroscopy (EIS) tests were carried out on a CHI660b electrochemical workstation (Shanghai Chenhua, China). The CVs were obtained over the potential range of 1.0–3.0 V at a scanning rate of 0.1 mV s⁻¹ and the impedance spectra measurements were tested with the frequency ranging from 0.1 Hz to 100 KHz by applying a sine wave with amplitude of 5.0 mV.

Results and discussion

Fig. 2 shows X-ray diffraction (XRD) patterns of the as-prepared pure TiO₂ and TiO₂/C composite. It can be seen that after calcined at 500 and 600 °C, all the diffraction peaks of both samples can be indexed to anatase TiO₂ with a tetragonal structure (JCPDS 65-5714) without the formation of a new phase, indicating that the amorphous titanium was converted to crystalline TiO₂ completely.²⁴ It can be seen that the peaks appear in TiO₂/C composite samples (Fig. 2b) are broader than those of pure TiO₂ samples (Fig. 2a). It indicates that TiO₂/C composite possesses relatively smaller crystal size, which is possibly owing to the interference of the carbon composition in TiO₂/C composite. The crystal of rutile (JCPDS 21-1276) marked by black diamond appears when the annealing temperature is raised to 800 °C.

Fig. 2 XRD patterns of (a) pure TiO₂ and (b) TiO₂/C composite calcined at different temperatures.

X-ray photoelectron spectroscopy measurements were performed to confirm the oxidation states of Ti in pure TiO₂ and TiO₂/C composite. All the binding energies (BE) were measured in reference to the C1s peak at 284.80 eV. Pure TiO₂ and TiO₂/C composite display the similar survey spectrum, as shown in Fig. 3a. The signals originating from Ti 2p, O 1s and C 1s are easily detected, indicating the existence of Ti, O, and C. The specific information of binding energies of Ti and O are tabulated in Table S1.† Fig. 3b displays the XPS spectrum of Ti 2p doublet peaks; the binding energies of Ti 2p_1/2 and Ti 2p_3/2 were observed at approximately 464 eV and 459 eV, respectively, which are characteristic peaks of Ti⁴⁺.⁴³ Additionally, it can be seen that the peak of Ti 2p_1/2 and Ti 2p_3/2 shifts toward lower binding energy with increasing calcination temperature, which may be contributed to trace amount of Ti³⁺.⁴⁴ However, none of the characteristic peaks of Ti³⁺ (457.2 eV, BE) and/or Ti²⁺ (455.3 eV, BE) can be observed from the Ti 2p spectrum.⁴⁵ It indicates that the main composition phase of the as-prepared products is TiO₂. The result can be further confirmed by the O1s spectrum (Fig. 3c) with a main strong binding energy peak at about 530 eV (as shown in Table S1†), characteristic of the Ti–O–Ti bonds in TiO₂. The weak peak at 532.18 eV (TiO₂ sample, 500 °C) can be attributed to surface –OH or adsorbed water.^16,36,46

Fig. 3 XPS spectrum of pure TiO₂ and TiO₂/C composite calcined at different temperatures: (a) full survey scan spectrum, (b) Ti 2p (left) and an enlarged figure (right) of Ti 2p_3/2 state in valence band marked by the dashed rectangle in left and (c) O1s peak. (d) FT-IR spectra of the dried gel and pure TiO₂ samples calcined at different temperatures.

The FTIR spectra of the dried gel and pure TiO₂ samples collected in the range of 400–4000 cm⁻¹ is shown in Fig. 3d. The broad absorption band at 3134 cm⁻¹ is assigned to the stretching vibration mode of N–H resulting from FA.^47,48 The peak at 1625 cm⁻¹ corresponds to the plane bending vibration of O–H bending vibrations of water which adsorbed physically on the surface of the TiO₂ samples due to the strong hydrophilic quality of the TiO₂.⁴⁹ The peak at 1120 cm⁻¹ is due to the C–O stretching vibrations, and the peak at 1399 cm⁻¹ is assigned to the C–H stretching vibration derived from PVP (Fig. S1a†), indicating that PVP is preferentially distributed to the gel solid phase^50,51 and these peaks disappeared after calcination above 500 °C.^28,36 The peak locates at 621 cm⁻¹ is characteristic bending vibrations of Ti–OH bonds, which shifts to a low peak at 472 cm⁻¹ after calcination, corresponding to the bending vibration of Ti–O bond.⁵² These FT-IR spectra results indicate that the amorphous titanium gel completely transforms to TiO₂ after calcination above 500 °C. Moreover, similar FT-IR spectra curves (Fig. S1b†) can be observed from TiO₂/C composites calcined at different temperatures. During the reaction process, ethylene glycol works as a chelating agent which coordinates with titanium ion through hydrogen bonding to attain the better control of the gelation. FA is a highly polar solvent with hydrogen bonding and is hydrolyzed to produce ammonia and formic acid in the presence SO₄²⁻ (hydrolysis of TiOSO₄·xH₂O), increasing the pH of solution gradually, which accelerates the polycondensation reaction to adjust the gelation time. Surfactant of PVP was applied as an phase separation inducer, which leads to phase separation into the solvent-rich and gel-rich phases due to spinodal decomposition, resulting in well-defined macroporous gels (as discussed below) after gelation and drying.^53–55

Fig. 4a reveals the Raman spectra of pure TiO₂ and TiO₂/C composite. Two strong vibrational bands at 512 and 634 cm⁻¹ are observed in the spectra of both samples, which belong to A_1g and E_g vibration mode of anatase TiO₂, respectively.⁵⁶ Moreover, two broad peaks located at 1362 and 1585 cm⁻¹ are detected in the spectra of TiO₂/C composite, corresponding to the A_1g vibration mode of the disordered carbon (D bond) and the E_2g vibration mode of the ordered graphitic carbon (G bond), respectively.⁵⁷ The Raman spectra results confirm that PVP was converted to carbon after calcination in an argon atmosphere. Additionally, the I_D/I_G value (ratio of the intensity of D Raman peak and G Raman peak) of the TiO₂/C composite is 0.84, which is favorable to provide good electrical conductivity and maintain the structural integrity of TiO₂/C composite during repeated charge/discharge cycling.

Fig. 4 (a) Raman spectra of pure TiO₂ and TiO₂/C composite. (b) TGA curves of dried gel and TiO₂/C composite in air, all the samples obtained calcined at 500 °C.

Thermal analysis was performed to evaluate the thermal decomposition behavior of dried gel and the actual carbon content of TiO₂/C composite. Fig. 4b exhibits the thermal decomposition characteristic curve of dried gel and TiO₂/C composite. For dried gel, there is a distinct weight loss of about 39% occurred from 120 to 500 °C, and the weight keeps stable later. The big weight loss can be attributed to the elimination of absorbed/trapped water, chemical reaction of dried gel converts to TiO₂ and the decomposition of PVP. The TGA curve of dried gel in N₂ atmosphere (Fig. S2†) shows the similar thermal behavior with a weight loss of about 28% during the calcination process. The difference value of weight after calcined in different atmosphere (air and N₂) is calculated to about 11%, which is ascribed to the carbon resulting from the pyrolysis of PVP. Furthermore, under the same heat treatment process, a 10% weight loss is observed for the TiO₂/C composite, which is mainly ascribed to the combustion of carbon to CO₂. That is to say, the carbon content of TiO₂/C composite is calculated to 10%, which is consistent with the results based on TGA curves obtained in air and N₂ atmosphere.

Fig. 5 shows SEM images of pure TiO₂ and TiO₂/C composite calcined at different temperatures. It can be seen that all of them demonstrate a interconnected macroporous structure, which are formed by phase separation induced by PVP at the same instant as the sol–gel transition.^54,58–60 As shown in Fig. 5, the typical size of cocontinuous macropores is in the range of 1–3 μm in diameter. The high-magnification SEM images (insets) reveal that these skeleton surface consists of many fine nanocrystals. With the increase of temperature, the porosity decreases due to the increase of average grain size, resulting in more compact skeletons. Larger nanoparticles and some aggregations occur at the high temperature of 800 °C (inset of Fig. 5c), which could be attributed to the effect of sintering. Compared to pure TiO₂, the surface of TiO₂/C composite becomes rough and some nanoparticles emerge with the increasing of calcination temperature. Some isolated particles coating appear on the surface of skeleton after calcination at 800 °C (inset of Fig. 5f) due to the formation of carbon originated from the pyrolysis of PVP.

Fig. 5 SEM images of pure TiO₂ (a–c) and TiO₂/C composite (d–f) calcined at different temperatures. Insets are the corresponding high magnification images.

The detail microstructure of pure TiO₂ and TiO₂/C composite was further investigated by TEM and HRTEM analysis. As revealed in Fig. 6, both of samples display the morphology characteristic of porous structure assembled with fine nanocrystals with the size of 10–30 nm (Fig. 6a and c). The result is consistent well with SEM images. The polycrystalline structure of both samples can be confirmed by the selected area electronic diffraction pattern (inset in Fig. 6a and c), moreover, the distinct lattice stripes in the HRTEM images (Fig. 6b and d) reveal the pure TiO₂ and TiO₂/C composite are well crystallized, and the lattice interplanar spacings of approximately 0.36 nm correspond to the (101) plane of tetragonal TiO₂, in good agreement with the results of XRD (Fig. 2). Furthermore, compared to pure TiO₂, some grey regions marked by white arrow in Fig. 6d could be detected for TiO₂/C composite, which could be attributed to the carbon derived from PVP.

Fig. 6 (a and b) TEM and HRTEM images of pure TiO₂. (c and d) TEM and HRTEM images of TiO₂/C composite. The insets are corresponding SAED images and both the samples calcined at 500 °C.

Scanning transmission electron microscope (STEM) and area-scan elemental mapping tests were carried out to verify the composition of pure TiO₂ and TiO₂/C composite. As seen in Fig. 7a, pure TiO₂ exhibits rough surface, which is similar to the SEM results. The energy-dispersive spectroscopy (EDS) mappings (Fig. 7a) corresponding to the area outlined by the red square demonstrate titanium and oxygen elements are dispersed homogeneously, implying the uniform chemical composition of pure TiO₂ without any other elements. Meanwhile, the detailed microstructure and elements mappings of TiO₂/C composite are investigated. As shown in Fig. 7b, the uniform element distribution of C, Ti and O can be detected, indicating the in situ formed carbon is tightly contacted with TiO₂ particles, which authentically confirms the above mentioned results of Raman spectra and TEM images.

Fig. 7 STEM image and EDX mapping images of pure TiO₂ (a) and TiO₂/C composite (b), both the samples calcined at 500 °C.

The mesoporous characteristic was investigated by nitrogen isothermal absorption–desorption measurements. Fig. 8a shows the isotherm curve of TiO₂/C composite calcined at different temperature, the obtained specific parameters of TiO₂/C composite samples are shown in Table 1. Samples calcined at 500 and 600 °C present a typical type IV with H1 hysteresis loops at relative pressure of 0.65–0.85 according to IUPAC classification, representing the mesoporous structure.^28,61,62 The BET specific surface area of samples calcined at 500, 600 and 800 °C are identified to be 170, 132 and 7 m² g⁻¹, respectively. The surface area and pore volume decrease as temperature increases, which could be ascribed to the effect of sintering and the growth of grains. Additionally, the pore size distribution curves (Fig. 8b) indicate most pores are around 10 nm for the samples calcined at 500 and 600 °C. Compared with pure TiO₂ (Fig. S3 and Table S2†), TiO₂/C composite presents a larger specific surface area with a smaller average pore size due to the smaller grain size. The results are consistent with XRD patterns shown in Fig. 2.

Fig. 8 Nitrogen adsorption–desorption isotherms (a) and pore size distributions curves (b) of TiO₂/C composite measured at 77 K.

Table 1 Pore characteristics of TiO₂/C composite calcined at different temperature

Sample	Calcination temperature (°C)	Time (h)	SBET (m^2 g^−1)	Vp (cm^3 g^−1)
TiO2/C composite	500	4	170	0.338
TiO2/C composite	600	4	132	0.258
TiO2/C composite	800	4	7	0.063

---

Cyclic voltammetry is a complementary technique to galvanostatic cycling and helps in understanding the nature and potentials at which the discharge–charge reactions take place in the electrode material. Fig. 9 shows the CV curves of pure TiO₂ and TiO₂/C composite electrodes at a scanning rate of 0.1 mV s⁻¹ from 1 to 3 V. It can be seen that all the samples display typical electrochemical characteristics of anatase TiO₂ with a sharp pair of redox peaks at about 1.68 and 2.1 V, corresponding to the two-phase processes of lithium insertion/extraction in the anatase phase (xLi⁺ + TiO₂ + xe⁻ ↔ Li_xTiO₂).^63,64 The intensity of peaks reflect the electrochemical performance of electrode materials, as shown in Fig. 9a and b, the intensity of peaks decreases with the increase of temperature, agrees with the result of galvanostatic charge–discharge tests (shown below). Moreover, there are two small redox peaks between 1.5 and 1.6 V are observed for pure TiO₂ calcined at 500 °C, which could be attributed to the trace amount of TiO₂-B, a low temperature phase of TiO₂, which could be irreversible transformed to anatase with the increase of temperature.^9,19,64 Compared to pure TiO₂ samples, the background of the redox peaks for TiO₂/C composites (Fig. 9b) are much more wider than that of pure ones (Fig. 9a), which can be assigned to the pseudocapacitive, faradaic-limited process derived from the carbon.⁶⁵

Fig. 9 The CV curves (2nd cycle) of pure TiO₂ (a) and TiO₂/C composite (b) electrodes at a scan rate of 0.1 mV s⁻¹ in the potential window of 1.0–3.0 V.

The electrochemical properties of pure TiO₂ and TiO₂/C composite calcined at different temperatures for Li-ion battery applications were evaluated by galvanostatic charge–discharge tests, and the specific capacities are calculated based on the weight of the whole active material in this work. Fig. 10a displays the first galvanostatic charge–discharge profiles at 0.5 C. Agree well with the result of CV curves, a discharge voltage plateau at around 1.70 ± 0.5 V and a charge voltage plateau at around 1.97 ± 0.5 V are detected in all samples, the detail information of voltage plateau is shown in Table S3.†^,40,41 Large irreversible capacities were observed in all samples, which may be ascribed to the formation of Li₂O by H₂O/OH groups on the TiO₂ surfaces for the samples calcined at 500 and 600 °C, because these samples possess relatively high specific surface areas.^28,37 Additionally, the irreversible capacity of the sample calcined at 800 °C, composed of anatase and rutile, is due to the irreversible phase transformation from rutile to LiTiO₂.⁶⁶ TiO₂/C composite presents higher capacities than pure TiO₂, which could be attributed to the higher electronic conductivity contributed from the in situ produced carbon. The first discharge capacities of pure TiO₂ samples calcined at 500, 600, and 800 °C are 263, 243, and 189 mA h g⁻¹, respectively, corresponding to insertion of 0.78, 0.72 and 0.56 mol of Li per mol of TiO₂. Compared with pure TiO₂, TiO₂/C composite delivers a higher initial discharge capacity with 371 (1.1 mol of Li), 286 (0.85 mol of Li), and 217 mA h g⁻¹ (0.64 mol of Li) calcined at 500, 600, and 800 °C, respectively. The high discharge capacity of TiO₂/C composite is contributed to the synthetic effect of insertion of Li for TiO₂ and surface lithium storage resulting from carbon.^49,65

Fig. 10 Electrochemical properties of pure TiO₂ and TiO₂/C composite: (a) initial charge and discharge curves at 0.5 C (1 C = 168 mA h g⁻¹). (b) Charge and discharge profiles (2nd cycle) of TiO₂/C composite calcined at 500 °C at different rates. (c) Cycling performance at 1 C. (d) Rate performance at various rates. All the tests are conducted with a voltage window of 1.0–3.0 V.

Fig. 10b presents the discharge–charge (2nd cycle) curves of the TiO₂/C composite electrode calcined at 500 °C at various rates (0.5–20 C). The capacity and the potential plateau (Table S4†) both decrease with increasing current density due to the polarization of electrode materials. The discharge capacities at various rates of 0.5, 1, 3, 5, 10 and 20 C are 217, 173, 142, 133, 127 and 89 mA h g⁻¹, respectively. A plateau at about 1.7 V, corresponding to a biphasic transition from the tetragonal TiO₂ to the orthorhombic Li_0.5TiO₂, and disappears completely at the high rate of 20 C, which could be attributed to the hierarchical porous anatase TiO₂ is capacitive lithium storage dominant at high rates instead of the solid-state diffusion reaction.^26,28,67 The capacitive effect may make cycling performance more stable (as shown in Fig. 10d) due to reversible surface lithium storage, which avoids the lattice expansion.

Fig. 10c compares the cyclic stability between pure TiO₂ and TiO₂/C composite electrodes in the voltage range of 1–3 V at 1 C. It can be seen that the discharge capacities of all samples decrease with increasing temperature. The discharge capacities of pure TiO₂ electrode calcined at 500, 600 and 800 °C decay quickly upon cycles, which could only remain 85, 66 and 41 mA h g⁻¹ after 100 cycles, respectively. Compared to pure TiO₂ electrode, TiO₂/C composite present better cycling performance, the discharge capacities remain relatively stable along with cycles. Discharge capacities of TiO₂/C composite calcined at 500, 600 and 800 °C could keep 132, 110 and 93 mA h g⁻¹ after 100 cycles, respectively. The distinct difference between TiO₂/C composite and pure TiO₂ is ascribed to the higher electronic conductivity contributed from the in situ produced carbon and the larger contact area of the electrode and electrolyte.

To demonstrate the rate performance, the pure TiO₂ and TiO₂/C composite samples were progressively charged and discharged from 0.5 to 30 C for 10 cycles at each rate, and then again at 1 and 0.5 C for 10 cycles, as shown in Fig. 10d. The result is consistent with Fig. 10c, TiO₂/C composites exhibit superior rate capability than pure TiO₂. It can be seen that the discharge and charge capacities of TiO₂/C composite samples remain stable and decrease regularly when the rate is increased from 0.5 to 30 C. While the capacities of pure TiO₂ fade rapidly even at low rates of 0.5 and 1 C, which could be attributed to slow electron transport resulting from poor electrical conductivity. Meanwhile, compared with the sample calcined at 500 °C, a substantial decrease in capacity was observed when the calcination temperature was raised to 600 and 800 °C respectively, which is ascribed to the decrease of BET surface area resulting from the increment in crystal size and the disappearing of small mesopores. Moreover, when the current density is restored to 1 and 0.5 C again, specific capacities can be recovered to a large extent due to the unique hierarchically porous structure, and keep stable during continuous discharge/charge cycles. The performance of hierarchically porous TiO₂ based composite reported here is compared favourably to those of TiO₂ based electrodes reported in previous works.^{27,38,39,57,68,69}

Electrochemical impedance spectroscopy (EIS) of pure TiO₂ and TiO₂/C composite after three cycles was investigated, the corresponding Nyquist plots and enlarged Nyquist plots (inset) are shown in Fig. 11a. All the Nyquist plots exhibit a semicircle in the high frequency region and a sloping straight line in the low frequency region. The high frequency semicircle is ascribed to the charge-transfer impedance at the electrolyte/electrode interface, and the sloping straight line in low frequency region corresponds to the lithium ion diffusion process in the electrodes.^70,71 Fig. 11b presents the equivalent circuit, R_s is the solution resistance, R_sl (i) and CPE (i) (i = 1, 2) represent the migration of hydrogen ions and the capacity of the layer, respectively.⁷² R_ct and C_dl denote the charge-transfer resistance and a double-layer capacitance, respectively.⁷³ Z_W is the Warburg impedance. It can be seen from the inset of Fig. 11a that all the samples show a similar shape in the low frequency region, while in the case of the high frequency region, the size of semicircle increases with the increase of temperature. Meanwhile, compared with pure TiO₂, TiO₂/C composite exhibits smaller semicircle. Therefore, the TiO₂/C composite calcined at 500 °C shows the smallest semicircle, and pure TiO₂ calcined at 800 °C displays the biggest semicircle. The EIS measurement results indicate that TiO₂/C composite possesses lower charge-transfer resistance and contact resistance than those of pure TiO₂, resulting in better electrochemical performance due to higher reactivity and reaction kinetics. It could be attributed to the improved electronic conductivity of the TiO₂/C composite derived from carbon and large specific surface area resulting from small crystal particles.^74,75

Fig. 11 (a) Electrochemical impedance spectra (EIS) of pure TiO₂ and TiO₂/C composite after three cycles (the inset shows the enlarged Nyquist plots). (b) The equivalent circuit used for fitting the EIS curves.

The outstanding electrochemical performance of TiO₂/C composite can be ascribed to its unique hierarchically porous nanoarchitecture, as shown in Fig. 12. Firstly, the interconnected porosity provides rich open channels, which improves the wettability of the electrolyte and offer more favorable paths for electrolyte penetration. In addition, the porous structure may also alleviate the strain induced by the volume change during the Li⁺ insertion/extraction processes. Secondly, the assembled TiO₂ nanocrystals (10–30 nm) reduce the Li⁺ diffusion distance and provide large BET specific surface area, ensuring better contact between the electrode material and the electrolyte. Additionally, the carbon derived from the pyrolysis of PVP effectively enhances the electrical conductivity of TiO₂ for the rapid electrochemical reactions, therefore leading to superior rate capability. Consequently, the synergistic effects of the hierarchically porous structure and the incorporation of conductive carbon materials result in excellent cycling stability and rate capability.

Fig. 12 Schematic description of the paths for the electron and lithium-ion transport in the hierarchically porous TiO₂ based composites electrode.

Conclusions

In summary, unique porous TiO₂ and TiO₂/C composite were successfully prepared by a facile sol–gel process followed by post-calcination treatment. These TiO₂ based composites display a characteristic of hierarchically porous structure constructed by interconnecting nanocrystals, shortening lithium ion transmission pathway for high rate performance, and then the porous structure provides large space for alleviating volume change of TiO₂ during cycling. Moreover, the in situ formed carbon derived from PVP guarantees a rapid electrons transport during repeated charge/discharge processes. Thus, the resultant TiO₂/C composite exhibits excellent cycling stability and superior rate capability as anode materials for LIBs. We believe that this facile synthesis method can be extended to other composites (such as Li₄Ti₅O₁₂, TiO_x, etc.) for a variety of energy conversion and storage applications.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (51372225), and the Natural Science Foundation of Zhejiang Province, China (LY13B010001).

Notes and references

1. N. Roy, Y. Sohn and D. Pradhan, ACS Nano, 2013, 7, 2532.
2. Z. Y. Liu, D. D. L. Sun, P. Guo and J. O. Leckie, Nano Lett., 2007, 7, 1081.
3. G. Wang, Q. Wang, W. Lu and J. H. Li, J. Phys. Chem. B, 2006, 110, 22029.
4. J. H. Huang, P. Y. Hung, S. F. Hu and R. S. Liu, J. Mater. Chem., 2010, 20, 6505.
5. K. Pan, Y. Z. Dong, C. G. Tian, W. Zhou, G. H. Tian, B. F. Zhao and H. G. Fu, Electrochim. Acta, 2009, 54, 7350.
6. M. Stefik, F. J. Heiligtag, M. Niederberger and M. Gratzel, ACS Nano, 2013, 7, 8981.
7. D. Deng, M. G. Kim, J. Y. Lee and J. Cho, Energy Environ. Sci., 2009, 2, 818.
8. M. Wagemaker, A. P. M. Kentgens and F. M. Mulder, Nature, 2002, 418, 397.
9. H. Huang, Z. Y. Yu, W. J. Zhu, Y. P. Gan, Y. Xia, X. Y. Tao and W. K. Zhang, J. Phys. Chem. Solids, 2014, 75, 619.
10. H. Huang, J. W. Fang, Y. Xia, X. Y. Tao, Y. P. Gan, J. Du, W. J. Zhu and W. K. Zhang, J. Mater. Chem. A, 2013, 1, 2495.
11. Z. H. Bi, M. P. Paranthaman, P. A. Menchhofer, R. R. Dehoff, C. A. Bridges, M. F. Chi, B. K. Guo, X. G. Sun and S. Dai, J. Power Sources, 2013, 222, 461.
12. X. Chen and S. S. Mao, Chem. Rev., 2007, 107, 2891.
13. J. S. Chen, D. Y. Luan, C. M. Li, F. Y. C. Boey, S. Z. Qiao and X. W. Lou, Chem. Commun., 2010, 46, 8252.
14. J. W. Zhang, X. X. Yan, J. W. Zhang, W. Cai, Z. S. Wu and Z. J. Zhang, J. Power Sources, 2012, 198, 223.
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. 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.
17. I. Moriguchi, R. Hidaka, H. Yamada, T. Kudo, H. Murakami and N. Nakashima, Adv. Mater., 2006, 18, 69.
18. X. Su, Q. L. Wu, X. Zhan, J. Wu, S. Y. Wei and Z. H. Guo, J. Mater. Sci., 2012, 47, 2519.
19. S. H. Liu, H. P. Jia, L. Han, J. L. Wang, P. F. Gao, D. D. Xu, J. Yang and S. N. Che, Adv. Mater., 2012, 24, 3201.
20. Y. G. Guo, Y. S. Hu, W. Sigle and J. Maier, Adv. Mater., 2007, 19, 2087.
21. J. Wang, Y. K. Zhou, Y. Y. Hu, R. O'Hayre and Z. P. Shao, J. Phys. Chem. C, 2011, 115, 2529.
22. J. F. Ye, W. Liu, J. G. Cai, S. A. Chen, X. W. Zhao, H. H. Zhou and L. M. Qi, J. Am. Chem. Soc., 2011, 133, 933.
23. A. I. Hochbaum and P. D. Yang, Chem. Rev., 2010, 110, 527.
24. H. J. Zhou, L. Liu, X. C. Wang, F. X. Liang, S. J. Bao, D. M. Lv, Y. K. Tang and D. Z. Jia, J. Mater. Chem. A, 2013, 1, 8525.
25. V. Gentili, S. Brutti, L. J. Hardwick, A. R. Armstrong, S. Panero and P. G. Bruce, Chem. Mater., 2012, 24, 4468.
26. B. T. Zhao, S. M. Jiang, C. Su, R. Cai, R. Ran, M. O. Tade and Z. P. Shao, J. Mater. Chem. A, 2013, 1, 12310.
27. H. E. Wang, H. Cheng, C. P. Liu, X. Chen, Q. L. Jiang, Z. G. Lu, Y. Y. Li, C. Y. Chung, W. Y. Zhang, J. A. Zapien, L. Martinu and I. Bello, J. Power Sources, 2011, 196, 6394.
28. G. Hasegawa, T. Sato, K. Kanamori, K. Nakanishi and T. Abe, New J. Chem., 2014, 38, 1380.
29. H. B. Wu, J. S. Chen, H. H. Hng and X. W. Lou, Nanoscale, 2012, 4, 2526.
30. X. Xu, B. Tian, J. Kong, S. Zhang, B. Liu and D. Zhao, Adv. Mater., 2003, 15, 1932–1936.
31. I. Moriguchi, R. Hidaka, H. Yamada, T. Kudo, H. Murakami and N. Nakashima, Adv. Mater., 2006, 18, 69–73.
32. W. Wang, M. Tian, A. Abdulagatov, S. M. George, Y. C. Lee and R. G. Yang, Nano Lett., 2012, 12, 655.
33. Y. S. Luo, J. S. Luo, J. Jiang, W. W. Zhou, H. P. Yang, X. Y. Qi, H. Zhang, H. J. Fan, D. Y. W. Yu, C. M. Li and T. Yu, Energy Environ. Sci., 2012, 5, 6559.
34. K. S. Park, J. G. Kang, Y. J. Choi, S. Lee, D. W. Kim and J. G. Park, Energy Environ. Sci., 2011, 4, 1796.
35. Z. X. Yang, G. D. Du, Z. P. Guo, X. B. Yu, Z. X. Chen, T. L. Guo and R. Zeng, Nanoscale, 2011, 3, 4440.
36. Y. Qiao, X. L. Hu, Y. Liu, C. J. Chen, H. H. Xu, D. F. Hou, P. Hu and Y. H. Huang, J. Mater. Chem. A, 2013, 1, 10375.
37. W. S. Wang, Q. N. Sa, J. H. Chen, Y. Wang, H. J. Jung and Y. D. Yin, ACS Appl. Mater. Interfaces, 2013, 5, 6478.
38. Z. X. Yang, Q. Meng, Z. P. Guo, X. B. Yu, T. L. Guo and R. Zeng, J. Mater. Chem. A, 2013, 1, 10395.
39. L. C. Liu, Q. Fan, C. Z. Sun, X. R. Gu, H. Li, F. Gao, Y. F. Chen and L. Dong, J. Power Sources, 2013, 221, 141.
40. S. B. Yang, X. L. Feng and K. Mullen, Adv. Mater., 2011, 23, 3575.
41. X. Xin, X. F. Zhou, J. H. Wu, X. Y. Yao and Z. P. Liu, ACS Nano, 2012, 6, 11035.
42. F. Sordello, G. Zeb, K. W. Hu, P. Calza, C. Minero, T. Szkopek and M. Cerruti, Nanoscale, 2014, 6, 6710.
43. Y. Luo, J. Luo, W. Zhou, X. Qi, H. Zhang, Y. Denis, C. M. Li, H. J. Fan and T. Yu, J. Mater. Chem. A, 2013, 1, 273.
44. C. M. Greenlief, J. M. White, C. S. Ko and R. J. Gorte, J. Phys. Chem., 1985, 89, 5025.
45. D. Song, J. Hrbek and R. Osgood, Nano Lett., 2005, 5, 1327.
46. J. Y. Ruzick, F. Abu Bakar, L. Thomsen, B. C. Cowie, C. McNicoll, T. Kemmitt, H. E. A. Brand, B. Ingham, G. G. Andersson and V. B. Golovko, RSC Adv., 2014, 4, 20649.
47. A. D. C. Permana, A. Nugroho, K. Y. Chung, W. Chang and J. Kim, Chem. Eng. J., 2014, 241, 216.
48. H. Yang, W. Zhu, S. Sun and X. Guo, RSC Adv., 2014, 4, 32934.
49. K. Hemalatha, A. S. Prakash, K. Guruprakash and M. Jayakumar, J. Mater. Chem. A, 2014, 2, 1757.
50. C. Lu, Y. Sun, J. Liu, X. Wang, S. L. Liu and W. Feng, J. Appl. Polym. Sci., 2015, 132, 41583.
51. X. B. Wang, Y. Y. Wang, L. Yang, K. Wang, X. D. Lou and B. B. Cai, J. Power Sources, 2014, 262, 72.
52. R. Nakamura and Y. Nakato, J. Am. Chem. Soc., 2004, 126, 1290.
53. K. Nakanishi and K. Kanamori, J. Mater. Chem., 2005, 15, 3776.
54. K. Nakanishi, J. Porous Mater., 1997, 4, 67.
55. G. Hasegawa, K. Kanamori, K. Nakanishi and T. Hanada, J. Sol-Gel Sci. Technol., 2010, 53, 59.
56. H. C. Choi, Y. M. Jung and S. B. Kim, Vib. Spectrosc., 2005, 37, 33.
57. P. Y. Chang, C. H. Huang and R. A. Doong, Carbon, 2012, 50, 4259.
58. W. Y. Li, X. Z. Guo, Y. Zhu, Y. Hui, K. Kanamori and K. Nakanishi, J. Sol-Gel Sci. Technol., 2013, 67, 639.
59. Y. Tokudome, N. Tarutani, K. Nakanishi and M. Takahashi, J. Mater. Chem. A, 2013, 1, 7702.
60. K. Nakanishi and N. Tanaka, Acc. Chem. Res., 2007, 40, 863.
61. H. Huang, W. Zhu, X. Tao, Y. Xia, Z. Yu, J. Fang, Y. Gan and W. Zhang, ACS Appl. Mater. Interfaces, 2012, 4, 5974.
62. W. Zhu, H. Huang, Y. Gan, X. Tao, Y. Xia and W. Zhang, Electrochim. Acta, 2014, 138, 376.
63. C. T. Cherian, M. V. Reddy, T. Magdaleno, C. H. Sow, K. V. Ramanujachary, G. V. S. Rao and B. V. R. Chowdari, CrystEngComm, 2012, 14, 978.
64. P. C. Chen, M. C. Tsai, H. C. Chen, I. N. Lin, H. S. Sheu, Y. S. Lin, J. G. Duh, H. T. Chiu and C. Y. Lee, J. Mater. Chem., 2012, 22, 5349.
65. Y. Gan, L. Zhu, H. Qin, Y. Xia, H. Xiao, L. Xu, L. Ruan, C. Liang, X. Tao and H. Huang, Solid State Ionics, 2015, 269, 44.
66. E. Baudrin, S. Cassaignon, M. Koelsch, J.-P. Jolivet, L. Dupont and J.-M. Tarascon, Electrochem. Commun., 2007, 9, 337.
67. J. Wang, Y. K. Zhou, Y. Y. Hu, R. O'Hayre and Z. P. Shao, J. Mater. Sci., 2013, 48, 2733.
68. Y. C. Qiu, K. Y. Yan, S. H. Yang, L. M. Jin, H. Deng and W. S. Li, ACS Nano, 2010, 4, 6515.
69. D. H. Wang, D. W. Choi, J. Li, Z. G. Yang, Z. M. Nie, R. Kou, D. H. Hu, C. M. Wang, L. V. Saraf, J. G. Zhang, I. A. Aksay and J. Liu, ACS Nano, 2009, 3, 907.
70. R. Mo, Z. Lei, K. Sun and D. Rooney, Adv. Mater., 2014, 26, 2084.
71. C. L. Zhang, Q. Y. Zhang, S. F. Kang and X. Li, J. Mater. Chem. A, 2014, 2, 2801.
72. G. F. Cai, J. P. Tu, D. Zhou, J. H. Zhang, Q. Q. Xiong, X. Y. Zhao, X. L. Wang and C. D. Guts, J. Phys. Chem. C, 2013, 117, 15967.
73. Y. Xia, Z. Xiao, X. Dou, H. Huang, X. H. Lu, R. J. Yan, Y. P. Gan, W. J. Zhu, J. P. Tu, W. K. Zhang and X. Y. Tao, ACS Nano, 2013, 7, 7083.
74. J. Jin, S. Z. Huang, J. Liu, Y. Li, D. S. Chen, H. E. Wang, Y. Yu, L. H. Chen and B. L. Su, J. Mater. Chem. A, 2014, 2, 9699.
75. L. Tan, L. Pan, C. Y. Cao, B. F. Wang and L. Li, J. Power Sources, 2014, 253, 193.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra03491d

---