TiO₂ Nanobelt@Co₉S₈ Composites as Promising Anode Materials for Lithium and Sodium Ion Batteries.

TiO₂ anodes have attracted great attention due to their good cycling stability for lithium ion batteries and sodium ion batteries (LIBs and SIBs). Unfortunately, the low specific capacity and poor conductivity limit their practical application. The mixed phase TiO₂ nanobelt (anatase and TiO₂-B) based Co₉S₈ composites have been synthesized via the solvothermal reaction and subsequent calcination. During the formation process of hierarchical composites, glucose between TiO₂ nanobelts and Co₉S₈ serves as a linker to increase the nucleation and growth of sulfides on the surface of TiO₂ nanobelts. As anode materials for LIBs and SIBs, the composites combine the advantages of TiO₂ nanobelts with those of Co₉S₈ nanomaterials. The reversible specific capacity of TiO₂ nanobelt@Co₉S₈ composites is up to 889 and 387 mAh·g-1 at 0.1 A·g-1 after 100 cycles, respectively. The cooperation of excellent cycling stability of TiO₂ nanobelts and high capacities of Co₉S₈ nanoparticles leads to the good electrochemical performances of TiO₂ nanobelt@Co₉S₈ composites.

1. Introduction

Lithium-ion batteries (LIBs), as one of the most important energy storage devices, have attracted extensive attention due to their advantages of high energy density and long cycle life [1,2]. However, the practical application of LIBs is still restricted especially in electrical devices and hybrid electric vehicles (HEV). Electrodes materials are key factors to affect the electrochemical performance for energy storages devices [3]. The commercial graphite anode for LIBs cannot meet the ever-increasing requirement owing to its low specific capacity (372 mAh·g−1) and the safety problems [4]. Therefore, more attention has been focused on designing new anode materials to replace graphite [5,6,7,8,9,10]. Among various transition metal oxides, Titanium dioxide (TiO2) has been considered as new anode materials for both LIBs and SIBs because of its low cost, environmental friendliness, high voltage platform, and long cycling stability [11,12,13,14,15].

One-dimensional (1D) TiO2 nanomaterials such as nanowires, nanotubes, and nanorods have been investigated as anodes, which effectively improved the ionic and electronic transport properties compared to TiO2 nanoparticles [16,17,18,19]. TiO2 nanobelt is one of the potential candidates among these materials in energy storage fields [20,21]. However, the intrinsic low theoretical capacity of TiO2 still limits its wide application. Thus, many effective methods have been adopted to increase its inherent low specific capacity. An efficient way is to hybridize TiO2 with another active material with high capacity. The composites synthesized by TiO2 and metal oxides or sulfides delivered excellent lithium storage performances [22,23,24]. For instance, the 3D electrode by assembling Fe2O3 hollow nanorods onto highly oriented TiO2 nanotube arrays delivered a high capacity of over 600 mAh·cm−2 at a current density of 100 mA·cm−2 after 50 cycles [23]. Furthermore, the TiO2@MoS2 hybrid exhibited a reversible capacity of 710 mAh·g−1 after 100 cycles at 100 mA·g−1 [24]. Another common method is to change the crystal phases of TiO2 to enhance its electrochemical performance. Very recently, TiO2-B anode materials have attracted great interest for both LIBs and SIBs [25,26,27,28]. They presented better electrochemical performances than the other phases of TiO2 due to their open tunnel structure and pseudocapacitive lithium storage properties [29,30]. Besides, it has been reported that the anatase/TiO2-B coherent interfaces could contribute to additional lithium storage, leading to better electrochemical performances than single phase TiO2 [31]. TiO2-B related composites have also become the research focus for LIBs. For example, the hierarchical TiO2-B nanowire@α-Fe2O3 composites exhibited better cycling stability than pure TiO2-B [32]. In addition, the TiO2-B nanoribbons anchored with NiO nanosheets as anode materials also displayed good cycling stability at a large rate of 5 C [33]. Cobalt sulfides as high capacity anodes have so many distinct advantages, such as the good conductivity, low electrode polarization and good thermal stability compared with transition metal oxides. However, to our knowledge, the composites of anatase/TiO2-B nanobelt and cobalt sulfides have not been reported yet for both LIBs and SIBs.

In this paper, TiO2 nanobelt@Co9S8 composites have been successfully obtained via a solvothermal reaction and high-temperature calcination process. The composites as anode materials for both LIBs and SIBs present good electrochemical performances, which is better than single TiO2 nanobelts and Co9S8 nanoparticles. The high reversible capacity, good cycling stability and rate capability of TiO2 nanobelt@Co9S8 composites are likely attributed to the synergistic effect of Co9S8 nanoparticles and anatase/TiO2-B nanobelts.

2. Materials and Methods

2.1. Synthesis of mixed Phase TiO2 Nanobelt@Co9S8 Composites

Mixed phase TiO2 nanobelts were prepared by a previously reported method [34]. TiO2 nanobelt@Co9S8 composites were synthesized via the following preparation procedure: First, 30 mg TiO2 nanobelts glucose aqueous solution (25 mL, 0.05 M) was dispersed by ultrasonic treatment for 2 min. Then 1.5 mmol of cobalt acetate were added into this solution and stirred for another 20 min to form a homogeneous dispersion. After that, 4.5 mmol of thiourea in ethylene glycol (25 mL) was added into the above mixture. The obtained dispersion was transferred to a Teflon-lined stainless steel autoclave and then heated at 180 °C for 10 h. After it was cooled to room temperature, the precipitate was collected, washed with deionized water and absolute alcohol thoroughly, and dried at 60 °C for 12 h. Finally, the TiO2 nanobelt@Co9S8 composites were obtained through annealing the precipitate at 650 °C for 10 h under Ar/H2 (5%) atmosphere. Co9S8 nanoparticles were also prepared according to the methods reported in previous paper [35].

2.2. Sample Characterization

X-ray diffraction (XRD) patterns were carried out by a Bruker D8 advanced X-ray diffractometer using monochromatic Cu Kα radiation (λ = 1.5418 Å). Transmission electron microscope (TEM) images and high-resolution transmission electron microscopy (HRTEM) images were achieved on a high-resolution transmission electron microscope (JEOL-2100, Akishima, Tokyo, Japan). Scanning electron microscope (SEM) images, mapping images, and energy dispersive spectrometer (EDS) spectrum were taken from a field-emission scanning electron microscope (FEI Nova 450, Hillsboro, OR, USA). Raman spectra were obtained on a MicroRaman spectrometer using a laser of 532 nm as an excitation (LabRAM HR Evolution, Kyoto, Japan). Nitrogen sorption isotherm was examined on a Micromeritics ASAP2020HD88 gas sorptometer at 77.3 K (Micromeritics, Norcross, GA, USA).

2.3. Electrochemical Measurements

Electrochemical performances of various electrodes were evaluated by CR2032 coin cells. The working electrode was fabricated by coating a mixture of 70 wt % of active material, 20 wt % of acetylene black, and 10 wt % of binder CMC (carboxyl methyl cellulose) in deionized water on a clean copper foil. Then the obtained foil was dried in vacuum at 60 °C for 10 h. The resulting foil was roll-pressed and punched into discs with a diameter of 12 mm. The mass loading of active material is estimated to 1.0–1.5 mg·cm−2. The coin cells were assembled in an Ar-filled glovebox. For LIBs, the lithium foil was used as the counter electrode, Celgard 2400 microporous polypropylene membrane as the separator, and a solution of 1 M LiPF6 in ethylene carbonate (EC) and dimethyl carbonate (DMC) (1:1 v/v) as the organic electrolyte. For SIBs, the sodium foil was used as the counter electrode, glass fiber was used as the separator, and the electrolyte is a solution of 1 M NaClO4 in EC and diethyl carbonate (DEC) (1:1 v/v) containing of 2 wt % fluoroethylene carbonate (FEC). Galvanostatic charge-discharge curves were acquired in a range of 0.01 to 3 V on the battery cyclers (Land CT2001A, Wuhan, China). Electrochemical impedance spectra (EIS) were carried out on an electrochemical workstation (AUTOLAB PGSTAT302N, Herisau, Switzerland) over a frequency range of 100 kHz to 0.01 Hz. Cyclic voltammetry (CV) curves were measured on an electrochemical workstation (CHI660E, Shanghai, China) over 0.01 to 3 V at a scanning rate of 0.1 mV·s−1. All the electrochemical tests were performed at 25 °C.

3. Results and Discussion

Figure 1 shows the preparation process of TiO2 nanobelt@Co9S8 composites. Firstly, the mixed phase TiO2 nanobelts were prepared via a three step chemical reaction and annealing [34]. Secondly, the intermediate products labeled as TiO2 nanobelt@CoS were obtained by a solvothermal reaction, using mixed phase TiO2 nanobelts as growth templates and glucose as a linker. Finally, TiO2 nanobelt@Co9S8 composites were formed with the initial amorphous CoS changing into crystalline Co9S8 after an annealing treatment of 650 °C for 10 h under Ar/H2 atmosphere.

Figure 1

Preparation process of TiO2 nanobelt@Co9S8 composites.

3.1. Characterization of Samples

Figure 2 shows the XRD pattern of obtained TiO2 nanobelt@Co9S8 composites. Three diffraction peaks marked with blue rhombus can be assigned to (311), (222) and (440) planes of cubic-phase Co9S8 (JCPDS Card, No. 65-1765) [8]. The diffraction peaks located at 2θ~25.3°, 48.0°, 53.9°, 55.1° and 62.7° can be identified as anatase TiO2 (JCPDS Card, No. 21-1272) [13]. The weak diffraction peak at 44.5° is attributed to TiO2-B (JCPDS Card, No. 46-1237) [16]. Meanwhile, mixed phase TiO2 nanobelt has been clearly demonstrated by the XRD patterns (Figure S1a and b). The Raman spectra of as-obtained TiO2 nanobelt and annealed treated TiO2 nanobelt also verify above results. As shown in Figure S1c and Figure S1d, two strong peaks located at 145 and 640 cm−1 are attributed to both anatase and TiO2-B, and one weak peak at 517 cm−1 is ascribed to anatase TiO2. All the remaining peaks are originated from TiO2-B [36]. EDS spectrum was measured to further verify the element composition and contents of the composites (Figure S2). The TiO2 contents (wt %) in the composites is estimated as ~64%, in which C and Al is from the conductive substrates, and the signal of Pt arises from the conductive coating of Pt on the sample surface by sputtering [37]. The EDS elemental mapping of composites also provides an even element distribution of Ti, Co and S (Figure S3).

Figure 2

XRD pattern of the TiO2 nanobelt@Co9S8 composites.

The morphology of TiO2 nanobelt@Co9S8 composites was characterized by SEM and TEM techniques (Figure 3). The SEM image in Figure 3a clearly shows that some nanoparticles are evenly anchored on TiO2 nanobelts, implying the formation of TiO2 nanobelt@Co9S8 composites. Figure 3b exhibits its low magnified TEM images. It can be clearly observed that the TiO2 nanobelt is uniformly coated with some small nanoparticles, which agrees well with the SEM data. The high magnified TEM image in Figure 3c shows the clear morphology of single TiO2 nanobelt@Co9S8 composite. As can be seen, TiO2 nanobelt presents an irregular porous nanostructure after acidizing and annealing. The related HRTEM image is demonstrated in Figure 3d, the d-spacings at 0.299 and 0.34 nm agree well with those from (311) planes of Co9S8 and (101) planes of TiO2. The overlap of lattice fringes shows that small Co9S8 nanoparticles are grown on the surface of TiO2 nanobelt likely induced by some chemical force effect. The SEM and TEM images of TiO2 nanobelts and Co9S8 nanoparticles are also given as control samples (Figure S4). The nitrogen adsorption-desorption measurement was carried out to determine the surface area of TiO2 nanobelt@Co9S8 composites. As shown in Figure S5a, a type-IV isotherm with a distinct hysteretic loop for P/P0 ranges from 0.5 to 1.0 could be observed, suggesting the mesoporous structure in the product. The surface area of TiO2 nanobelt@Co9S8 composites is estimated as ~32.2 m2·g−1. The pore size ranges from 23 to 38 nm and the main peak is located at 38 nm (Figure S5b), which is in agreement with what observed in TEM images. As seen in the previous paper, the low surface area can suppress the unnecessary side reaction, such as inevitable electrolyte decomposition and formation of solid electrolyte interface (SEI) [38,39]. Thus, it could be supposed that the hierarchical mesoporous TiO2 nanobelt@Co9S8 composites as anode materials should present good lithium storage performance.

Figure 3

SEM image (a), TEM images (b,c) and HRTEM image (d) of the TiO2 nanobelt@Co9S8 composites.

3.2. Electrochemical Performance of TiO2 Nanobelt@Co9S8 Composites for LIBs

The electrochemical performances of TiO2 nanobelt@Co9S8 composites for LIBs were tested (Figure 4). Figure 4a shows the cyclic voltammetry (CV) curves in a voltage window of 0.01–3.0 V at a scan rate of 0.1 mV·s−1. In the first cathodic scan, the strong cathodic peak located at 1.13 V is likely to come from the reduction of Co9S8 to metallic Co, which is ascribed to the conversion reaction: Co9S8 + 16Li+ → 9Co + 8Li2S [40]. A small wide peak located at~1.70 V is attributed to the formation of anatase LiTiO2, corresponding to reaction equations: xLi+ + xe− + TiO2 → LiTiO2 [31]. The wide peak at 0.76 V is related to the electrolyte decomposition and formation of SEI layer [38,39]. Inversely, a sharp anodic peak around 2.02 V can be assigned to the reversible oxidation of metallic Co, which overlaps with the charge process of Li+ deintercalation from the anatase framework (LiTiO2) [31]. Another two pairs of peaks located at 1.48/1.57 V and 1.54/1.65 V are ascribed to the surfaced-confined charge-transfer process (faradic pseudocapacitive lithium storage behavior) of TiO2-B [31]. The CV curves of TiO2 nanobelt further confirm the coexistence of anatase TiO2 and TiO2-B (Figure S6a). Figure 4b shows the galvanostatic discharge/charge voltage profiles of TiO2 nanobelt@Co9S8 composites for the first, second and fifth cycles over 0.01–3.0 V at 0.1 A·g−1. The first discharge curve shows multiple voltage plateaus mainly located at 1.6 and 1.2 V, which is in agreement with the redox peaks observed from CV results. The initial reversible capacity and coulombic efficiency of TiO2 nanobelt@Co9S8 composites could reach 714 mAh·g−1 and 74%, respectively. The irreversible capacity for the first cycle could result from the electrolyte decomposition and formation of SEI layer, which is very similar to that of transition metal oxides-based anodes [41,42]. Besides, it also arises from the solvated lithium intercalation and subsequent reduction of the solvent [43]. The irreversible capacities are largely dependent on the external surface area of the electrode and also plausibly related to the irreversible conversion reaction of Co9S8 for the first cycle and volume change of the electrode during the conversion process [44]. The cycling performance of TiO2 nanobelt@Co9S8 composites shows that a high reversible capacity of 889 mAh·g−1 can be achieved after 100 cycles (Figure 4c), which is far higher than that of single TiO2 nanobelt (Figure S6b). The capacity increase upon cycling is mainly from the pseudocapacitive lithium storage of TiO2 nanobelt and Co9S8 nanoparticles [8]. Such good electrochemical performance could be attributed to the designed hierarchical composites. The Co9S8 nanoparticles attached to TiO2 nanobelt can provide high capacity and improve conductivity for overall composites. Their small particle size and uniform dispersion of Co9S8 can effectively inhibit volume changes during cycling. Moreover, 1D TiO2 nanobelt can enhance the electron transfer efficiency, and the TiO2-B in the TiO2 nanobelt will lead to a higher reversible capacity compared to pure anatase TiO2 [45]. The interfaces between anatase and TiO2-B nanodomains can also contribute to additional lithium storage capacity [31]. More importantly, the anatase/TiO2-B nanobelt as backbone for Co9S8 can suppress the separation of Co9S8, its inherent cycling stability of TiO2 can also hinder the capacity loss of Co9S8 upon cycling. All the above factors facilitate the good electrochemical performances of TiO2 nanobelt@Co9S8 composites. The rate capability is another important kinetic factor to evaluate the electrochemical performance of TiO2 nanobelt@Co9S8 composites. As presented in Figure 4d, the average reversible capacity of TiO2 nanobelt@Co9S8 composites at a current density of 0.1, 0.2, 0.5, 1, 2 and 5 A·g−1 is 707, 676, 605, 502, 473 and 260 mAh·g−1, respectively, which is superior to those of the TiO2 nanobelt. The electrochemical performances of TiO2 nanobelt@Co9S8 composites at large current densities such as 1, 2 and 5 A·g−1 are superior to those of Co9S8 nanoparticles. Surprisingly, when the current density returns back to 0.1 A·g−1, the reversible capacity is still as high as 776 mAh·g−1, which is higher than initial capacity value. This phenomenon is likely attributed to the enhanced capacitance contribution resulting from the so-called electrochemical milling effect. All the results mentioned show that the TiO2 nanobelt@Co9S8 composites could be considered as potential anode materials to be applied for LIBs.

Figure 4

Cyclic voltammograms (a), galvanostatic discharge/charge profiles (b), cycling performances (c) of the TiO2 nanobelt@Co9S8 composites and (d) rate performances of the TiO2 nanobelt@Co9S8 composites, TiO2 nanobelts and Co9S8 nanoparticles.

Figure 5a presents the cycling performances of TiO2 nanobelt@Co9S8 composites, TiO2 nanobelts and Co9S8 nanoparticles at 1 A·g−1. Although the cycling stability of TiO2 nanobelt is the best among three kinds of materials, its reversible capacity is only 216 mAh·g−1. The reversible capacity of TiO2 nanobelt@Co9S8 composites could retain at 369 mAh·g−1 after 100 cycles. The electrochemical impedance spectra (EIS) were tested to further investigate the electrode process kinetics of three kinds of electrodes (Figure 5b). For all the electrodes before cycling, the depressed semicircle at high-to-medium frequencies is attributed to charge-transfer impedance (Rct), and a slope at low frequencies is associated with ion diffusion process inside the electrode (constant phase element, CPE) [46]. After cycling, the depressed semicircles are related to two overlapped interface impedances (SEI and Rct) [35]. The phase angle of slope for TiO2 is close to 45°, suggesting a diffusion-controlled feature of lithium insertion/extraction. The phase angles of slope for both TiO2 nanobelt@Co9S8 composites and Co9S8 are greater than 45°, indicating significant capacitive component in lithium insertion/extraction [8]. Compared to the fresh electrodes, the decrease of Rct for all the cycled electrodes suggests the electrochemically activation of anodes [47]. The relatively small Rs, SEI and Rct verify effective lithium and electron transfer of this composite during cycling.

Figure 5

Cycling performances (a), electrochemical impedance spectra (EIS) of TiO2 nanobelts, Co9S8 nanoparticles and TiO2 nanobelt@Co9S8 composites: before cycling (b), and after cycling (c).

3.3. Electrochemical Performance of TiO2 Nanobelt@Co9S8 Composites for SIBs

The electrochemical performances of TiO2 nanobelt@Co9S8 composites, TiO2 nanobelt and Co9S8 nanoparticles for SIBs were also investigated (Figure 6). The CV curves are shown in Figure 6a, in the first cathodic scan, a strong peak located at ~0.46 V is likely to come from the formation of SEI film, and a small peak at 0.25 V is attributed to the intercalation process of Na+ into the TiO2 lattice [28]. The wide peaks located at ~1.13 V and 0.86 V are commonly attributed to the formation of NaxCo9S8 and further reduction process to Co and Na2S, respectively. The cathodic peak for the second cycle shifts to ~0.89 V, owing to irreversible structural rearrangement of Co9S8. For the anodic process, an intense oxidation peak at ~1.76 V could be assigned to the oxidation reaction of Co metal to form Co9S8 or NaCo9S8 due to the irreversible reaction [48], which is in good accordance with previous reports [49]. The wide peak at ~0.52 V is associated with oxidation process of NaTiO2 to TiO2 and Na1−TiO2 [50]. The overlap of all the cathodic and anodic peaks in subsequent cycles demonstrates the good reversibility of TiO2 nanobelt@Co9S8 composites. Moreover, two sloping voltage plateau located at about 0.9 and 0.5 V appear in the first discharge curves, while for the following cycles, only one obvious discharge voltage plateau can be found, which is located at around 0.9 V. For all of the charge process, just a wide voltage plateau is observed. The wide voltage plateau indicate the overlap of electrochemical reaction of TiO2 and Co9S8 electrodes. All of the voltage plateaus in the discharge/charge curves of TiO2 nanobelt@Co9S8 composites can agree with the corresponding CV curves (Figure 6b).

Figure 6

Cyclic voltammograms (a), galvanostatic discharge/charge profiles (b), cycling performances (c,d) rate performances of the TiO2 nanobelt@Co9S8 composites for SIBs.

Figure 6c shows the cycling performance of TiO2 naobelt@Co9S8 composites at 0.1 A·g−1. The initial discharge and charge capacity can reach 554 and 387 mAh·g−1, respectively. After 100 cycles, the reversible capacity can still maintain at 258 mAh·g−1, displaying good cycling stability of TiO2 naobelt@Co9S8 composites. The rate capability of TiO2 nanobelt@Co9S8 composites are shown in Figure 6d. The reversible capacity is 388, 358, 322, 288 and 237 mAh·g−1 at 0.1, 0.2, 0.5, 1 and 2 A·g−1, respectively. When the current density goes back to 0.1 A·g−1, the reversible capacity can return to 374 mAh·g−1. The cycling performance and rate capacity of TiO2 nanobelt@Co9S8 composites are better than those of single TiO2 and Co9S8 for SIBs (Figure S7).

The EIS of three kinds of electrodes for SIBs were also measured (Figure S8), compared to fresh electrodes, all of the cycled electrodes exhibit smaller Rct, and the Rct of TiO2@Co9S8 is located between TiO2 and Co9S8, which is possibly resulted from the good conductivity of Co9S8. The phase angles of slope for these electrodes are similar to those of LIBs. The TiO2 nanobelt@Co9S8 composites with good electrochemical performances are of great potential to be used for both LIBs and SIBs.

4. Conclusions

In summary, we have successfully synthesized mixed phase TiO2 nanobelt@Co9S8 composites. As new anode materials, they show high specific capacities and good cycling performances for both LIBs and SIBs. After 100 cycles, their reversible capacity for LIBs can retain at 889 and 369 mAh·g−1 at 0.1 and 1 A·g−1, respectively. Besides, an initial reversible capacity of 387 mAh·g−1 can be obtained at 0.1 A·g−1 for SIBs. Such good electrochemical performances of TiO2 nanobelt@Co9S8 composites could be attributed to the hierarchical nanostructure and synergistic effect of Co9S8 nanoparticles and anatase/TiO2-B nanobelts.