CNT@TiO2 nanohybrids for high-performance anode of lithium-ion batteries

This work describes a potential anode material for lithium-ion batteries (LIBs), namely, anatase TiO2 nanoparticle-decorated carbon nanotubes (CNTs@TiO2). The electrochemical properties of CNTs@TiO2 were thoroughly investigated using various electrochemical techniques, including cyclic voltammetry, electrochemical impedance spectroscopy, galvanostatic cycling, and rate experiments. It was revealed that compared with pure TiO2 nanoparticles and CNTs alone, the CNT@TiO2 nanohybrids offered superior rate capability and achieved better cycling performance when used as anodes of LIBs. The CNT@TiO2 nanohybrids exhibited a cycling stability with high reversible capacity of about 190 mAh g-1 after 120 cycles at a current density of 100 mA g-1 and an excellent rate capability (up to 100 mAh g-1 at a current density of 1,000 mA g-1).


Background
The use of limited fossil fuel resources and their negative impact on the environment are significant challenges facing world economies today, creating an urgent demand for new technologies that enable high efficiencies in energy harvesting, conversion, and storage devices [1,2]. Various technologies, including fuel cells, batteries, solar cells, and capacitors, show great promise to significantly reduce carbon footprints, decrease reliance on fossil fuels, and develop new driving forces for economic growth [3,4]. Lithium-ion batteries (LIBs) have been regarded as one of the most promising energy storage technologies for various portable electronics devices [5], and one of the key goals in developing LIBs systems is to design and fabricate functional electrode materials that can lower costs, increase capacity, and improve rate capability and cycle performance [6][7][8][9].
It has been extensively reported that TiO 2 is a promising candidate to compete with commercial graphite anode for LIBs due to its multiple advantages of high abundance, low cost, high Li-insertion potential (1.5 to 1.8 V vs. Li + /Li), structural stability, and excellent safety during cycling [10]. Practical applications of TiO 2 in LIBs, however, face significant challenges of poor electrical conductivity and low chemical diffusivity of Li, which are two key factors for the lithium insertion-deinsertion reaction. Therefore, it is highly desirable to develop reliable strategies to advance electrical conductivity and Li + diffusivity in TiO 2 [11,12]. In fact, continued breakthroughs have been made in the preparation and modification of TiO 2based nanomaterials for high performance energy conversion and storage devices [13,14].
It is generally acknowledged that there are three routes available to tune the properties of TiO 2 for its corresponding applications: (1) preparation of TiO 2 -based nanostructures with specific morphology [11]. For example, TiO 2 -based nanorods were reported to show enhanced rate capability and improved stability as electrodes in LIBs due to their one-dimensional (1D) structure and high surface area [15,16]. (2) Synthesis of TiO 2 nanocrystals with specific crystal surface orientations [17]. It was reported that TiO 2 -based nanocubes dominated by (001) planes had much higher catalytic activity for photo-degradation of organic dyes than the conventional TiO 2 with mixed crystallographic facets [18,19]. (3) Fabricating TiO 2 -based nanohybrids with other functional materials. Carbon nanostructures, such as carbon nanotubes (CNTs) and graphene, are the most appealing functional materials for improving the performance of TiO 2 nanostructures due to their unique structure, excellent electrical conductivity, high stability, and great mechanical properties [20,21].
We recently developed a convenient procedure to synthesize TiO 2 nanoparticle-decorated CNT hybrid structures (CNTs@TiO 2 ) through annealing treatment of carbonaceous polymer-modified CNTs with adsorbed Ti 4+ . The as-prepared CNT@TiO 2 nanocomposites exhibit multiple favorable features, such as excellent electrical conductivity and considerable high surface area, which make them to be potentially used for promising electrode material of electrochemical energy storage and conversion devices. We systematically investigated the electrochemical properties of CNT@TiO 2 nanohybrids as anodes of LIBs, and demonstrated that the unique properties of both CNTs and TiO 2 can merge well in the CNT@TiO 2 nanohybrids with synergetic effects. In this way, the CNTs@TiO 2 can potentially address the intrinsic issues associated with TiO 2 anodes in LIBs, namely poor electrical conductivity and low chemical diffusivity of Li ions, and thus significantly improve performance in term of capacity, cycle performance, and rate capability.

Materials and synthesis
All chemicals were purchased from Sigma-Aldrich (St. Louis, MO, USA) and used without further purification, except CNTs (200 nm in diameter) which were purchased from Carbon Nanotechnologies, Inc. (Sunnyvale, CA, USA). CNTs@TiO 2 were prepared through a modified route reported previously [22]. Typically, 0.15-g CNTs were completely mixed with a 60-ml glucose solution (0.5 mg/ml) under sonication. The mixed turbid liquid was then placed in a 100-ml Teflon-lined stainless steel autoclave and heated at 180°C for 5 h. Next, 0.2 g of the product after centrifuging and drying, namely carbonaceous polymer-modified CNTs (CNTs@C polymer ), was then dispersed in 15 ml ethanol with the addition of 1 ml of titanium isopropoxide (TIP, 97%) under vigorous agitation. After centrifuging and drying, the solid products were then calcined at 400°C and exposed in an air atmosphere to evolve into CNTs@TiO 2 . Powder Xray diffraction (XRD) was conducted on a Scintag XDS 2000 X-ray powder diffractometer (Scintag Inc., Santa Clara, CA, USA) using monochromatized CuKα as radiation (λ = 1.5418 Å); the data were collected by scanning angles (2θ) from 20°to 60°. N 2 adsorptiondesorption experiments were tested at 77 K by a Quantachrome autosorb gas-sorption system (Boynton Beach, FL, USA). The morphologies of the as-prepared samples were observed using a Hitachi (H 9000 NAR, Tokyo, Japan) transmission electron microscope (TEM) and a Hitachi S-4800 scan electron microscope (SEM).

Characterization
The working electrode of LIB was prepared by compressing a mixture of active materials (80%), acetylene black (10%), and polyvinylidene fluoride (10%) as a binder dissolved in 1-methyl-2-pyrrolidinone solution onto a copper foil. The pellet was dried in vacuum at 120°C for 10 h and then assembled into a coin cell in an Ar-protected glove box. The electrolyte solution was 1 M LiPF 6 dissolved in a mixture of ethylene carbonate (EC) and dimethyl carbonate (DMC), with a volume ratio of EC/DMC = 4:6. Galvanostatic cycling experiments were conducted to measure the electrode activities using a Maccor battery tester system (Tulsa, OK, USA) at room temperature. Cyclic voltammograms (CVs) were carried out with three-electrode cells and recorded from 3.0 to 1.0 V at a scan rate of 0.1 mV s −1 using a CHI 600 electrochemical station (CHI Inc., Austin, TX, USA). Discharge-charge curves were recorded at fixed voltage limits between 3.0 and 1.0 V at various current densities. The specific capacity was calculated based on the total mass of the active materials. Electrochemical impedance spectroscopy (EIS) measurements were carried out at the open-circuit voltage state of fresh cells using a CHI600 (Austin, TX, USA) electrochemical workstation. The impedance spectra were recorded potentiostatically by applying an AC voltage of 5-mV amplitude over a frequency range from 100 kHz to 5 mHz.

Results and discussion
The crystalline structure, morphology, and nanostructure of the products were firstly investigated using XRD, SEM, and TEM, as shown in Figure 1. Figure 1a shows the XRD pattern of the CNTs@TiO 2 , which shows typical peaks that can be well assigned to anatase TiO 2 with characteristic peaks of CNTs, indicating the successful decoration of anatase TiO 2 nanoparticles on CNTs. Figure 1b exhibits the typical SEM image of the asprepared CNTs@TiO 2 , demonstrating that the samples have a 1D structure with an average diameter of around 200 nm. Figure 1c presents the SEM image of one single CNT@TiO 2 ; one can observe a large number of nanoparticles uniformly decorated on the surface of the nanofiber, which stands in sharp contrast to the carbonaceous modified CNT with a relative smooth surface (Additional file 1: Figure S1). The TiO 2 -decorated CNTs were additionally confirmed by a typical TEM image (Figure 1d). Figure 1e shows the TEM image recording the nanostructure of the tip of the CNTs@TiO 2 ; it was revealed that the size of TiO 2 nanoparticles on the CNTs is around 5~10 nm. For comparison, we prepared TiO 2 nanoparticles with an average diameter of 50 nm through a sol-gel method (Figure 1f ).
The present CNTs@TiO 2 feature a favorable porous structure and improved electrical conductivity, which are attractive for addressing the existing issues for TiO 2 as anodes of LIBs; therefore, we systematically investigated the electrochemical performance of the CNTs@TiO 2 as anode of LIBs. We first applied the techniques of galvanostatic charge/discharge and CV to compare and study the electrochemical properties of lithium insertion/deinsertion in half-cells based on CNT, TiO 2 , and CNT@TiO 2 materials. Figure 2a,b,c and Figure 2d,e,f display the initial two charge-discharge profiles and CV curves for the CNT, TiO 2 , and CNT@TiO 2 electrodes, respectively. The initial two charge-discharge profiles are generally consistent with the corresponding CV results. For CNTs, there is no pronounced peak in the range of 1.0 to 3.0 V with a remarkable discharge capacity loss from 55 mAh g −1 in the first cycle to 20 mAh g −1 in the second cycle. In contrast, both TiO 2 and CNT@TiO 2 electrodes show a discharge plateau at around 1.70 V and a charge plateau at about 1.90 V in the first cycle, which is basically consistent with those reported previously [20,21]. In particular, the TiO 2 electrode exhibits a pronounced capacity loss of 20.0% in the second discharge process, while the CNT@TiO 2 electrode only shows a capacity loss of less than 10.0% in the initial two cycles. As expected, there is a pair of peaks in the CV curves of the TiO 2 and CNT@TiO 2 electrodes, namely, the cathodic peak at 1.69 V and the anodic peak at 2.08 V, corresponding with the reversible biphasic transition between the tetragonal anatase and orthorhombic Li x TiO 2 , respectively (Equation 1).
There is an observable decrease of cathodic current in the second CV compared with the first CV for the TiO 2 electrode, which agrees with the previous report on TiO 2 anode materials and can be attributed to the irreversible lithium insertion-deinsertion reaction, indicating a large capacity loss during the first two cycles. The CNTs@TiO 2 , however, only display a small change during the initial two CVs, suggesting a small capacity loss in the initial two cycles. Figure 3a compares the cycling performance and the rate capability between the TiO 2 and CNTs@TiO 2 . The CNTs@TiO 2 show significantly improved performance in terms of the capacity (except the first discharge capacity), rate capability, and stability. First, the CNTs@TiO 2 showed a remarkable improvement in cycling performance compared with TiO 2 . The CNTs@TiO 2 delivered a specific capacity of 251.9 mAh/g in the first cycle at a current density of 100 mA g −1 . This value is slightly lower than the corresponding capacity of the TiO 2 (263.0 mAh/g); however, the CNTs@TiO 2 discharged a higher capacity than TiO 2 in the following cycle. One can observe that the discharge capacity gradually decreased in the initial several cycles for both CNTs@TiO 2 and TiO 2 . The CNT@TiO 2 electrode achieved a stable capacity of around 195.5 mAh/g in the tenth cycle, while the TiO 2 showed a continuous decrease, even in the initial 20 cycles. In fact, when the current density was switched back to 100 mA g −1 in the 81st cycle, the CNTs@TiO 2 reached a reversible capacity of around 191.0 mAh g −1 and maintained this capacity in the subsequent cycles, while the TiO 2 discharged a corresponding capacity of 163.3 mAh g −1 and showed a slow decrease with the continuous cycling. In addition, the CNTs@TiO 2 also exhibited a greatly improved rate performance compared with TiO 2 , with varying current densities from 100 to 1,000 mA g −1 . For instance, the CNTs@TiO 2 maintained a capacity of 110 mAh g −1 at a current density of as high as 1,000 mA g −1 , while the TiO 2 only had a capacity of around 85 mAh g −1 under this current density. It should be noted that the CNTs@TiO 2 , as an anode of LIBs, also show improved electrochemical performance compared with the TiO 2 nanostructures reported previously [23][24][25], signifying that the as-designed CNTs@TiO 2 show great promise to advance electrochemical performance. In addition, the CNTs@TiO 2 can compete with or outperform the TiO 2 /CNT composites reported previously in terms of capacity and cycling performance [26,27]. For instance, the CNTs@TiO 2 still retained a specific capacity of about 190 mAh g −1 at a current density of 100 mA g −1 [28], which shows a remarkable contrast to the blended TiO 2 /CNT that only retained a capacity of about 170 mAh g −1 at the same current density. Figure 3b schematically illustrates the Li + insertion/deinsertion in CNT@TiO 2 nanohybrids and demonstrates advantages of the high electrical conductivity and facile transport of Li + in CNT@TiO 2 nanohybrids. The improved electrochemical performance in CNT@TiO 2 nanohybrids can be attributed to the following factors: above all, the TiO 2 nanoparticles were uniformly decorated on the surface of one-dimensional CNTs, which offered excellent flexibility and enough space for alleviating the effects of electrode degradation and volume change upon cycling. Additionally, the large surface area (109.9 m 2 g −1 ) and suitable pore size (11.5 nm) in CNTs@TiO 2 can facilitate the transport of electrolytes and Li + on the interface of electrodes, leading to good rate capability. Furthermore, the electrical conductivity, thanks to the CNT's core, is expected to be greatly enhanced, which can significantly decrease the capacity loss from Ohmic resistance.
The EIS measurements were carried out to investigate the resistance associated with the TiO 2 and the CNTs@TiO 2 . Figure 4 shows the Nyquist plots recorded for the TiO 2 and the CNTs@TiO 2 , respectively, which typically consists of a high-frequency semicircle corresponding with the charge transfer resistances (R ct ). The Nyquist data were then fitted to a hypothetical equivalent circuit (inset of Figure 4a) to evaluate the R ct and the resistance of the film formed on the electrode surface (R f ). It was revealed that the R ct and R f for the CNTs@TiO 2 were 48.8 and 21.3 Ω, respectively, much lower than the corresponding R ct (117.95 Ω) and R f (72.0 Ω) for the TiO 2 electrode, indicating that the CNTs@TiO 2 have a significantly lower overall impedance, which might be one of the key factors responsible for the improved electrochemical performance of the CNTs@TiO 2 . We further investigated the impedance change after cycling; it was revealed that the TiO 2 / CNT only shows a slight change in impedance spectroscopy, while the TiO 2 exhibits an evident change in impedance spectroscopy after 120 cycles (Figure 4b). These results additionally confirmed that the former can well maintain the high conductivity upon cycling.

Conclusion
In summary, we demonstrated the electrochemical properties of the nanohybrids of TiO 2 nanoparticle-decorated CNTs as an anode of lithium-ion batteries. The CNT@TiO 2 hybrids showed better electrochemical performance than the pure TiO 2 nanoparticles with regard to specific capacity (except the initial cycle), rate capability, and cycling stability. The improved electrochemical performance can be ascribed to the synergetic effects of combined properties, including the one-dimensional structure, high-strength with flexibility, excellent electrical conductivity, and large surface area.