Li4Ti5O12 nanocomposites with improved electrochemical performance as anode material for lithium-ion batteries

Li4Ti5O12 nanocomposites with improved electrochemical performance as anode material for lithium-ion batteries

Applied Surface Science 403 (2017) 635–644 Contents lists available at ScienceDirect Applied Surface Science journal homepage: www.elsevier.com/loca...

4MB Sizes 11 Downloads 85 Views

Applied Surface Science 403 (2017) 635–644

Contents lists available at ScienceDirect

Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc

Full Length Article

Graphene supported Li2 SiO3 /Li4 Ti5 O12 nanocomposites with improved electrochemical performance as anode material for lithium-ion batteries Qiufen Wang ∗ , Shuai Yang, Juan Miao ∗ , Mengwei Lu, Tao Wen, Jiufang Sun Henan Polytechnic University, Jiaozuo 454000, People’s Republic of China

a r t i c l e

i n f o

Article history: Received 8 November 2016 Received in revised form 24 December 2016 Accepted 21 January 2017 Available online 23 January 2017 Keywords: Graphene Li2 SiO3 Li4 Ti5 O12 Hydrothermal route Electrochemical properties

a b s t r a c t Graphene supported Li2 SiO3 @Li4 Ti5 O12 ([email protected]/LTO) nanocomposites have been synthesized via a hydrothermal route and following calcination. LSO/LTO nanospheres are adhered to the graphene nanosheets with the size of 50–100 nm, in which both LSO and LTO particles are attached together. When tested as the anode for lithium ion batteries, the initial discharge and charge capacities of [email protected]/LTO are 720.6 mAh g−1 and 463.4 mAh g−1 at the current density of 150 mA g−1 . After 200 cycles, the discharge and charge capacities can be remained of 399.2 mAh g−1 and 398.9 mAh g−1 , respectively. Moreover, the charge rate capacities of [email protected]/LTO composites retain 89.1% at the range of current density from 150 mA g−1 to 750 mA g−1 . And its recovery rates are 91.0% when the current density back to 150 mA g−1 . In addition, the reversible capacity and cycle stability of [email protected]/LTO are better than that of LTO and LSO/LTO. The reasons can be attributed to the synergistic effect between GE and LSO/LTO as well as the features of GE supports. © 2017 Elsevier B.V. All rights reserved.

1. Introduction With the development of electric vehicles, communication devices and portable electronics, exploitation of advanced energy storage devices has attracted researchers’ more attention. Lithium ion batteries (LIBs) are considered to be a promising candidate for clean energy storage due to their high electromotive force, long cycling ability, no memory effect, little self-discharge and design flexibility [1–7]. Development of the electrode material is a key factor for LIBs [8,9]. Spinel Li4 Ti5 O12 (LTO), one of the alternative candidates, has been used as an anode material which has facile Li+ storage properties in spite of low theoretical capacity (175 mAh g−1 ). The crystalline structure of LTO could be the structure with Fd3 m space group, where O2− is located at 32e sites, 3/4 of Li+ is located at tetrahedral 8a sites and the rest of Li+ with Ti4+ is located at octahedral 16d sites [10]. When Li+ intercalating, the inserted lithium and Li+ in 8a sites are migrated to the neighboring octahedral 16c sites, and form the structure of a rock salt ([Li2 ]16c [Li1/3 Ti5/3 ]16d O4 ). When

∗ Corresponding authors. E-mail addresses: [email protected] (Q. Wang), [email protected] (J. Miao). http://dx.doi.org/10.1016/j.apsusc.2017.01.221 0169-4332/© 2017 Elsevier B.V. All rights reserved.

Li+ de-intercalating, the lithium ions move back 8a sites again. The reaction can be described as follows [10,11]. Li(8a) [Li1/3 Ti5/3 ](16d) O4(32e) + Li+ + e ↔ Li2(16c) [Li1/3 Ti5/3 ](16d) O4(32e)

In the reaction, 1 mol of Li4 Ti5 O12 may be intercalated in 3 mol of lithium at most. The unit cell parameter of Li4 Ti5 O12 is 0.836 nm while that of the generative Li7 Ti5 O12 is 0.837 nm [12]. Thus, LTO may be regarded as a “zero strain effect” material, indicating its long cycle life. In addition, a charge and discharge potential at ∼1.55 V vs. Li+ /Li of LTO makes it safer [13]. However, it has a low electronic conductivity (∼10−13 S cm−1 ) and a poor lithium ion diffusion coefficient (∼10−15 cm2 s−1 ) which results in a poor rate capability [14–16]. Some approaches have been attempted to solve the problems. Tailoring the particle size of LTO can reduce the ionic and electronic transportation distance and enhances the electronic conductivity [17–19]. Additionally, the composites formed by coating or doping with additives in LTO can also improve the electrochemical properties of the electrodes, such as SiO2 -incorporated Li4 Ti5 O12 [20], Li4 Ti5−x Snx O12 [21], carbon-encapsulated F-doped LTO composites [22], Mg-substituted Li4 Ti5 O12 [23], Li3.8 Cu0.3 Ti4.9 O12 /CNTs [24], carbon-Li4 Ti5 O12 [25,26], and so on. These dopants or coating materials can prevent the agglomeration of LTO particles and provide better electrical conductivity, thus further improve the electrochemical performance of LTO.

636

Q. Wang et al. / Applied Surface Science 403 (2017) 635–644

Furthermore, Li2 SiO3 (LSO) is an ideal dopant due to its unique structure. The SiO4 tetrahedra form zigzag chains, indicating that LSO has a three-dimensional path, which is favorable for the lithium ion diffusion. Moreover, LSO is a kind of chemically inert materials and has good structural stability in organic electrolyte [27,28], which will improve the rate capability and cycling stability and decrease the polarization of LTO. Graphite has been widely used as a kind of commercialized anode materials for LIBs. But its low theoretical capacity (372 mAhg−1 ) limits its further application in the large-scale energy storage fields [29]. At present, various kinds of novel carbon materials such as carbon nanotube [30], mesoporous carbon [31], carbon coating material (typically carbons) [32] and carbon black [33] have been used in LIBs. In all of these, as a two-dimensional crystal lattice sheet of carbon atoms with a honeycomb structure, graphene has been demonstrated to be desirable for good Li ion storage space and charge-transfer rate due to its superior electrical conductivity, large surface-to-volume ratio, high specific surface area of over 2600 m2 g−1 , mechanical strength and structure flexibility [34–36]. Thus, based on the previous research [37], we have developed graphene supported LSO/LTO ([email protected]/LTO) nanocomposite by the hydrothermal method and following sintering as anode materials for lithium-ion batteries. The structure, morphology and the electrochemical properties of LSO/LTO and [email protected]/LTO nanocomposites are characterized and further compared with LTO.

2. Experimental 2.1. Synthesis procedures 2.1.1. Synthesis of Si-TiO2 sphere The TiO2 spheres were synthesized by the hydrolysis of tetrabutylorthotitanate (Ti(OC4 H9 )4 , TBOT, AR) [37,38]. Firstly, 400 mg of hydroxypropyl cellulose (HPC, AR) was dissolved in 2 mL of distilled water and 300 mL of ethanol under vigorous magnetic stirring. After that, 3 mL of TBOT and 40 mL of ethanol were mixed and then added into the above solution. After reacted at 85 ◦ C for 90 min, the collected product was washed with distilled water and ethanol and dried at 60 ◦ C to obtain amorphous TiO2 spheres. Finally, the products were kept at 500 ◦ C, 600 ◦ C and 700 ◦ C for 3 h in a furnace, which were labeled as T500, T600 and T700, respectively. In addition, Si was provided by Harbin Te Bo Technology Co. Ltd., China Si particles present the nano spheres with a size from 30 to 50 nm and the purity is 99.9%. Si/TiO2 sphere was prepared with the above-mentioned method with 300 mg of silicon as support. The final product was kept in a furnace at 600 ◦ C for 3 h to obtain the precursor of LSO/LTO.

Fig. 1. The schematic diagram of the formation of GE supported LSO/LTO composite.

2.1.3. Synthesis of [email protected] LSO/LTO Graphene oxide was prepared by modified Hummers and Offeman’s method according to the reported method [34]. [email protected] LSO/LTO was prepared through above-mentioned method for LSO/LTO sphere, during which 300 mg of graphene oxide was added as support and the molar ratio of LiOH and TiO2 was set at 6:1. Fig. 1 illustrates schematically the steps for the formation of GE supported LSO/LTO composite. The reaction could be summarized as the following. Ti(OC4 H9 )4 + 4H2 O = Ti(OH)4 + 4C4 H9 OH

(1)

Ti(OH)4 + Ti(OC4 H9 )4 = 2TiO2 + 4C4 H9 OH

(2)

Ti(OH)4 ⇔ TiO2 + 2H2 O

(3)

Si + O2 = SiO2

(4)

5TiO2 + SiO2 + 6LiOH = Li2 SiO3 + Li4 Ti5 O12 + 3H2 O

(5)

In the first step the Si/TiO2 was formed through the hydrolysis of Ti (OC4 H9 )4 as shown in Eqs. ((1)–(3)). SiO2 /TiO2 is formed by the calcination of Si/TiO2 (Eq. (4)). LSO/LTO is formed by hydrothermal method (Eq. (5)). Lastly, [email protected] LSO/LTO is obtained by the calcination in a tube furnace. In addition, LSO/LTO spheres could be grown on the graphene sheets because the effect of an electrostatic adherence could take place between GE and Ti4+ [39]. 2.2. Characterization

2.1.2. Synthesis of LSO/LTO sphere The LSO/LTO sphere was synthesized by the hydrothermal method [18]. 900 mg of Si/TiO2 sphere was dispersed in 20 mL of distilled water and 60 mL of ethanol under vigorous magnetic stirring. After adding a certain molar of LiOH·H2 O, the mixture solution was transferred into a 100 mL of Teflon-lined stainless steel autoclave and heated at 180 ◦ C for 18 h. After the reaction was finished, it was cooled down to room temperature naturally. The solid products were centrifuged, washed with distilled water and ethanol, and then dried at 60 ◦ C under vacuum to obtain the precursors. Finally, the products were kept in a tube furnace at 600 ◦ C for 3 h. The samples prepared at the molar ratio of LiOH:TiO2 = 1.75:1, 4:1, 6:1 and 8:1 were named as LT1.75, LT4, LT6 and LT8, respectively. For comparison, LTO was also prepared via the above method but no silicon was added, and the molar ratio of LiOH and TiO2 is set as1.75:1.

The structure and morphology of the samples were characterized by X-ray diffraction analysis (XRD, X’Pert MPD PRO, PANalytical Company, Almelo, Holland), scanning electron microscope (SEM, NoVaTM Nano 250, FEI Company) and field emission transmission electron microscope (FETEM, Tecnai G2 F20, FEI Company). XPS analysis was characterized by X-Ray photoelectron spectroscopy (K-Alpha, ESCALAB 250, Thermofisher Co., USA). The fourier transform infrared spectroscopy (FTIR) spectra of the samples were obtained by using Model NIcolETiS10 fourier transform spectrometer (Thermo SCIENTIFIC Co., USA) with a 2 cm−1 resolution in the range of 400–4000 cm−1 . The nitrogen adsorption/desorption isotherms of the samples were measured at 77 K by a Quantachrome Instruments Autosorb IQC (USA), and the surface area was determined using the Brunauer-Emmett-Teller (BET) formalism. The pore size distribution was computed from the by the original density functional theory (DFT) model.

Q. Wang et al. / Applied Surface Science 403 (2017) 635–644

637

Fig. 2. XRD patterns of as-prepared samples: (a) TiO2 at different stirring temperature, (b) LSO/LTO prepared at the different molar ratio of LiOH and TiO2 , (c) LTO, LSO/LTO and [email protected]/LTO, (d) FTIR images of LTO, LSO/LTO and [email protected]/LTO.

Electrochemical measurements were evaluated by two electrode cell with lithium metal as a counter electrode. The anode was prepared by blending the active materials (80 wt.%), acetylene black (10 wt.%) and poly-(vinylidene fluoride) (PVDF) binder (10 wt.%). N-methylpyrrolidone (NMP) was used as a solvent to form homogeneous slurry, and microporous polypropylene membrane (Celgard 2400) was used as a separator. The electrolyte was prepared with 1 mol l−1 LiPF6 in a mixture of ethylene carbonate (EC)/diethyl carbonate (DMC)/ethylene methyl carbonate (EMC) (volume ratio of 1:1:1). The cells were assembled in a glove box. The charge-discharge measurements were performed on an eight channel battery test system (Land CT2001A) between 0.01 and 3.0 V. Cyclic voltammetry (CV) and electrochemical impedance spectra (EIS) were obtained on a CHI 760D electrochemical workstation (CH Instruments, USA). CV was carried out at a scan rate of 0.2 mV s−1 between 0.01 and 2.5 V versus Li+ /Li. EIS was carried out by applying an ac cell potential of 0.5 V from 0.01 to 100 kHz. All measurements were carried out at room temperature.

3. Results and discussion 3.1. Characterization of materials The phase structure of the prepared samples was investigated by XRD. Fig. 2a shows the XRD patterns of TiO2 at different sin-

tering temperature. For the curves of T500 and T600, the major diffraction peaks of TiO2 (101), (004), (200), (105), (211) and (204) at 2␪ = 25.28◦ , 37.80◦ , 48.05◦ , 53.89◦ , 55.06◦ and 62.69◦ can be observed, which agree with the standard patterns of the anatase TiO2 (JCPDS 21-1272). Moreover, the diffraction peaks of T600 is narrower than that of T500, indicating the better crystallinity of T600. The major diffraction peaks for the sample of T700 agree with the standard dates of rutile TiO2 (JCPDS 21-1276) expect the standard patterns of anatase TiO2 , indicating that the product is a mixture of the rutile and anatase TiO2 . Thus, the sintering temperature of TiO2 is set as 600 ◦ C in the following process. To prepare the suitable LSO/LTO nanosphere, the effect of different molar ratio of LiOH and TiO2 on the structure of LSO/LTO composites has been investigated, as shown in Fig. 2b. The diffraction peaks intensities of LTO are strengthen with the increasing molar ratio of LiOH and TiO2 . Moreover, the peaks at 2␪ = 18.88◦ , 26.98◦ , 33.15◦ , 38.59◦ , 43.36◦ , 51.69◦ , 55.45◦ and 59.16◦ also agree with the standard patterns of lithium silicate Li2 SiO3 (JCPDS 29-0829), respectively. And other peaks at 2␪ = 25.28◦ , 48.05◦ and 53.89◦ agree with the standard patterns of the anatase TiO2 , respectively. In addition, the diffraction peaks of LTO and LSO are the strongest when the molar ratio of LiOH and TiO2 is 6:1, indicating the best crystallinity of LSO/LTO. Thus, the molar ratio of LiOH and TiO2 is fixed as 6:1 in the following process. Fig. 2c shows the XRD patterns of as-prepared LTO, LSO/LTO and [email protected]/LTO composites under the optimal conditions. After

638

Q. Wang et al. / Applied Surface Science 403 (2017) 635–644

Fig. 3. SEM images of (a) LTO, (b) LSO/LTO and (c) [email protected]/LTO, the elemental mappings of (b–1 ∼ 3) LSO/LTO and (c–1 ∼ 4) GE/LSO/LTO, (d) TEM images of [email protected] LSO/LTO.

the addition of GE, there are not the peaks of TiO2 in the curve of [email protected]/LTO, which indicates that the reaction of LiOH and Si/TiO2 may be facilitated by GE, and further improve the purity of LSO/LTO. Fig. 2d shows FTIR spectra of LTO, LSO/LTO and [email protected]/LTO composites. In LTO, the vibrations around 418 and 526 cm−1 can be attributed to the stretching vibration of the Ti O groups [40,41], and the vibrations around 658 cm−1 corresponds to the stretching vibration of the Li O groups [42]. The weak peaks around 3400 cm−1 are characteristic of bending and stretching vibrations of the −OH group of surface adsorbed water molecules [43]. In the spectrogram of LSO/LTO, except the stretching vibration of the Ti O, Li O and OH groups, the peak at 738 cm−1 corresponds to the stretching symmetric vibration of the Si-Ob -Si groups [44,45]. And the peaks at 869, 985 and 1075 cm−1 correspond to the stretching asymmetric vibrations of the Si Ob Si groups in tetrahedron of Si and O [46,47]. In addition, in the curve of [email protected]/GE, the peaks of [email protected]/GE are similar with that of LSO/LTO. The peaks around 1500 and 1645 cm−1 correspond to the stretching vibration of C C and C H, respectively. The peak at 1640 cm−1 is attributed to the stretching vibration of COOH [34,48].

The morphology and structure of LTO, LSO/LTO and [email protected]/LTO composites were characterized by SEM and TEM. As shown in Fig. 3a, several dozen small nanospheres are connected together to form a larger nanosphere with a diameter of 200–500 nm. When silicon is added, the diameter of LSO/LTO nanospheres is reduced into 50–100 nm (Fig. 3b), and further shorten the transport path of lithium ions and electrons. As shown in Fig. 3c, the LSO/LTO nanospheres are attached on the graphene nanosheets with the consistent sizes. From the TEM image of [email protected]/LTO (Fig. 3d), LSO/LTO nanoparticles are distributed on graphene nanosheets, and the LSO nanoparticle is coated by the LTO nanoparticle, which not only could prevent the re-stacking of LTO nanoparticles, but also buffer the expansion volume of Li-Si during the cycling process. The inset in Fig. 3d displays a selected area electron diffraction (SAED) pattern of [email protected]/LTO. The SAED image shows a few rings composed of many diffraction spots. Those spots can be described as the reflections of LTO and LSO, indicating that the [email protected]/LTO composite has good crystalline characteristic of LTO and LSO. Fig. 3b-1–b-3 show the elemental mapping images of a designated area in Fig. 3b. It clearly displays the distributions of O, Ti

Q. Wang et al. / Applied Surface Science 403 (2017) 635–644

639

[email protected]/LTO, respectively. As shown in Fig. 4b, the peaks from Si 2p3/2 at 103.7–103.8 eV are consistent with Si-O [49–51]. The signals in Fig. 4c present a Ti 2p3/2 peak at around 458.4–459.2 eV and a Ti 2p1/2 peak at 464.1–464.9 eV, indicating that the oxidation state of the Ti cations is mainly Ti (IV) [52–54]. In addition, the binding energies of [email protected]/LTO are roughly higher than that of LSO/LTO due to the existence of GE in every curve. The results above further verify the structures of LSO/LTO and [email protected]/LTO.

3.2. Electrochemical performances

Fig. 4. XPS of LSO/LTO and [email protected] LSO/LTO: (a) Survey spectra, (b) Si and (c) Ti.

and Si, which demonstrates that LSO and LTO particles are attached together. Fig. 3c-1–c-4 show the elemental mapping images of a designated area in Fig. 3c, which also presents the distributions of O, Ti, C and Si. These further confirm the results of XRD of LSO/LTO and [email protected]/LTO. To further investigate the surface structures and the binding natures of the elements, the XPS spectra of LSO/LTO and [email protected]/LTO have been provided in Fig. 4. The curves of Fig. 4a reveal the presence of Si, Ti and O in the LSO/LTO composite and the existence of Si, Ti, C and O in the [email protected]/LTO composite, which are accorded with the photoelectron spectroscopy of LSO/LTO and

The electrochemical performances of the as-prepared composites had been studied in a two-electrode coin cell. Fig. 5a shows the initial charge and discharge capacities of LTO, LSO/LTO and [email protected]/LTO, respectively. The cell potential window is set between 0.01 V and 3.0 V vs. Li+ /Li at the current density of 150 mA g−1 . The discharge curves of all the electrodes present a potential plateau at 1.5–1.6 V while the charge one emerges at 1.5–1.65 V. The typical characteristic plateau can be assigned to lithium ion intercalation and de-intercalation process of LTO, which indicates that the addition of LSO and GE does not affect the electrochemical reaction process of LTO [9,55]. Moreover, the initial discharge capacities of LTO, LSO/LTO and [email protected]/LTO are 160.8 mAh g−1 , 711.2 mAh g−1 and 720.6 mAh g−1 while their charge capacities are 124.5 mAh g−1 , 391.1 mAh g−1 and 463.4 mAh g−1 , respectively. It can be noticed that the initial discharge and charge capacities of LSO/LTO are higher than that of LTO because of a high theoretical Li-storage capacity of Si-based materials and the interaction between LSO and LTO [37,56]. Firstly, the doping LSO could decrease the size of the LTO particles and further shorten the transport path of lithium ions and electrons. In addition, according to the formation of LSO/LTO, LTO and LSO are closed together because the formations of LSO and LTO are in a same process, and could induce the co-doping between LSO and LTO. So, LSO/LTO composites have better interface bonding, in which could provide interconnected Li+ diffusion channels between LTO and LSO, decrease charge transfer resistance, improve lithium ion diffusion during lithium intercalation and de-intercalation. In addition, the initial discharge and charge capacities of [email protected]/LTO are the highest among the composites because the synergistic effect between GE and LSO/LTO. Firstly, the existence of GE nano-sheets possesses more defects and facilitating electron conductivity for improved lithium ion storage. Secondly, the addition of GE hinders the particles growth, and further shortens the diffusion distance of lithium ions. Fig. 5b shows the cycling performances of as-prepared samples at the current density of 150 mA g−1 between 0.01 V and 3.0 V. After 200 cycles, the discharge/charge capacities of LTO, LSO/LTO and [email protected]/LTO are 108.3 mAh g−1 /108 mAh g−1 , 221.1 mAh g−1 /220.9 mAh g−1 and 399.2 mAh g−1 /398.9 mAh g−1 , respectively. Their capacity retentions are 86.7%, 56.5% and 86.1%, indicating the improved cycling stability of LSO/LTO with GE doping. To further discuss the cycle performances of these composites, the charge capacities of LTO, LSO/LTO and [email protected]/LTO are analyzed at the first, 10th, 50th, 100th, 150th and 200th cycle at the current density of 150 mA g−1 . Fig. 5c shows the relation of the cycle times and the charge capacities of as-prepared samples. It can be shown that their charge capacities decrease with the increase of the cycle times. The curves of LTO, LSO/LTO and [email protected]/LTO may be fitted to different equations, which can be described as follows. y = 125.9 − 0.10x (LTO)

y = 421.3e−x/0.67 + 79.9e−x/49.3 + 218.3(LSO/LTO)

640

Q. Wang et al. / Applied Surface Science 403 (2017) 635–644

Fig. 5. (a) The initial charge and discharge curves, (b) the cycling performance, (c) the relation of the cycle times and the charge capacities at the current density of 150 mA g−1 and (d) the rate capabilities of LTO, LSO/LTO and [email protected]/LTO. The potential window is set between 0.01 and 3.0 V vs. Li+ /Li.

y = 467.3 − 0.34x ([email protected]/LTO) It can be seen that the curves of LTO and [email protected]/LTO can be fitted to the linear equations. The Adj. R-Square of [email protected]/LTO is 98% while that of LTO is 88%. In addition, the curve of LSO/LTO can only be fitted to the nonlinear equation. It indicates the charge capacities of LTO and [email protected]/LTO decrease slowly, which further confirms the cycling performances of LTO, LSO/LTO and [email protected]/LTO. Fig. 6d shows the rate capabilities of LTO, LSO/LTO and [email protected]/LTO after 200 cycles between 0.01 V and 3.0 V at the current density of 150 mA g−1 , 300 mA g−1 and 750 mA g−1 , respectively. The charge rate capacities of LTO, LSO/LTO and [email protected]/LTO composites retain 63.9%, 61.7% and 89.1% at the current density from 150 mA g−1 to 750 mA g−1 while their recovery rates of the charge capacities are 87.9%, 84.8% and 91.0% when the current density back to 150 mA g−1 , respectively. The results demonstrate that the rate capability of LSO/LTO composite is lower than that of the LTO and [email protected]/LTO, which its capacity fading can be ascribed to the large volume expansion occurring and the collapse of electrode during the cycling process [57]. The [email protected]/LTO composites possess higher rate retention ability than that of LTO and LSO/LTO. It could be explained that the addition of GE hinders the particles growth, and further shortens the diffusion distance of lithium ions among the [email protected]/LTO particles [58]. In addition, the graphene sheets act as a buffer matrix to provide enough paths for the entrance of electrolyte during the lithiation and de-lithiation process [34,59,60]. To confirm the rate capabilities and cycling stabilities of LTO, LSO/LTO and [email protected]/LTO, SEM images were used to study the structure of the composites after cycling tests. Fig. 6 shows SEM

images of as-prepared LTO, LSO/LTO and [email protected]/LTO films on Cu substrates at the current density of 150 mA g−1 after 200 cycles. It is obvious that the spheres become distorted and agglomerated (Fig. 6a–c), indicating the pulverization of the particles during cycling process. The most serious pulverization of LSO/LTO (Fig. 6b) leads to the poorest cycling stability. In Fig. 6c, LSO/LTO particles are distributed on graphene sheets at random, indicating the [email protected]/LTO composite has low polarization. So, [email protected]/LTO composite shows the best reversible capacity and cycle stability. To evaluate the diffusion of lithium ion and further explain the rate capabilities and cycling stabilities of LTO, LSO/LTO and [email protected]/LTO, Fig. 7a shows EIS analysis of as-prepared composites at 0.5 V from 0.01 Hz to 100 kHz after 200 cycles. The equivalent circuit model is shown in the inset of Fig. 7a. In the equivalent circuit model, Rs is ascribed to the ohmic resistance of electrolyte. Rf is ascribed to the surface polarization resistance. Zw is ascribed to the Warburg impedance. Cd1 is ascribed to double-layer capacitance of the electrode and electrolyte interface. Cf is ascribed to the surface capacitance. Rct is ascribed to the charge-transfer resistance [34,38]. Each curve is consisted of a depressed semicircle at the high-middle frequency region and a straight line at the low frequency region. The high frequency semicircle is related to the charge transfer resistance on the electrode and electrolyte interface and the sloping line is assigned to the lithium-ion diffusion process. Obviously, the EIS spectra fitted from the models roughly agree with that of the experiments, suggesting that the equivalent circuit diagram is reasonable. The charge and transfer resistances of LTO, LSO/LTO and [email protected]/LTO composites are 176.3 , 94.2  and 67.1 , respectively. In addition, the semicircle of [email protected]/LTO is the smallest among these composites, which indicates the smallest

Q. Wang et al. / Applied Surface Science 403 (2017) 635–644

641

Fig. 6. SEM images of the films of (a) LTO, (b) LSO/LTO and (c) [email protected]/LTO on Cu substrates after 200 cycles at the current density of 150 mA g−1 .

Fig. 7. (a) The EIS of LTO, LSO/LTO and [email protected]/LTO after the 200 cycles. The inset is the equivalent circuit model. (b) The relationship between Zre and ␻−0.5 at low frequencies for as-prepared samples.

charge transfer resistance or more facile charge transfer process at the electrode/electrolyte interface. The diffusion coefficient of lithium ion can be obtained from the plots in the low-frequency region according to the following equation [5]. D = R 2 T2 /(2A2 n4 F4 C2 бw 2 )

(6)

Where R is the gas constant, T is the temperature, n is the number of electron per molecule oxidized, A is the area of the electrode surface, F is Faraday’s constant, C is the molar concentration of Li+ , D is the diffusion coefficient, and бw is the Warburg coefficient which has the relationship with Zre as follows. Zre = R ct + R s + бw ␻−0.5

(7)

Where ␻ is the angular frequency in the low frequency region. The relationship between Zre and ␻−0.5 for LTO, LSO/LTO and [email protected]/LTO composites in the low frequency region is shown in Fig. 7b. The slope of the fitted line is the Warburg coefficient бw . It is observed that the Warburg coefficients бw of LTO, LSO/LTO and [email protected]/LTO composites are 269.7, 76.12 and 36.64  cm2 s−1/2 , respectively. The relevant diffusion coefficients of lithium ion are roughly calculated to be 4.055 × 10−13 , 5.551 × 10−12 and 2.468 × 10−11 cm2 s−1 , respectively. As a consequence, [email protected]/LTO composite shows the highest lithium ion diffusion coefficient compared with the others. The enhancement of the diffusion coefficient may account for better electrochemical performance of [email protected]/LTO. Moreover, the porous structure and surface area of LTO, LSO/LTO and [email protected]/LTO were measured by the nitrogen adsorption/desorption method. Fig. 8a shows the nitrogen adsorption-

desorption isotherm, and Fig. 8b is the pore diameter distribution of LTO, LSO/LTO and [email protected]/LTO. Results show that the BET specific surfaces of LTO, LSO/LTO and [email protected]/LTO are 3.76 m2 g−1 , 5.05 m2 g−1 and 32.89 m2 g−1 , which indicates that the specific surfaces of LTO, LSO/LTO are much lower than that of [email protected]/LTO. In addition, the pore diameter distribution of [email protected]/LTO is wider than that of LTO and LSO/LTO, indicating that its mesopore portion (>2 nm) became more. An increasing contact surface between electrolyte and electrode materials is preferred for improving the electrochemical performance of the obtained products. Thus, the doped GE can improve the lithium ions diffusion of LSO/LTO. Therefore, [email protected]/LTO shows the best rate capability and cycling stability. The grid structure of GE can not only shorten lithium ion diffusion distances, possess faster lithium ion diffusion and further speed up the migration of lithium ions [57], but also provide electrical donor to improve intrinsic conductivity of LSO/LTO [58–60]. In addition, according to the formation of [email protected]/LTO, GE with a few oxygen-containing functional groups carries few negative charges. LSO/LTO can be grown on the graphene sheets because the effect of an electrostatic adherence taking place between GE and Ti4+ [38]. So, it can depress the electrochemical polarization, decrease charge transfer resistance, improve lithium ion diffusion during lithium intercalation and deintercalation process. Fig. 9a displays the CV curves of [email protected]/LTO electrode for 5 cycles. The scanning rate is 0.2 mV s−1 and the cell potential window is set between 0.01 and 2.5 V vs. Li+ /Li in two electrode systems. The cathodic peaks can be observed at ∼1.5 V and 0.01–0.2 V while the anodic peaks are located at 1.5–1.70 V, ∼0.25 V in the first potential sweep. The cathodic at ∼1.5 V and anodic peaks 1.5–1.70 V

642

Q. Wang et al. / Applied Surface Science 403 (2017) 635–644

Fig. 8. (a) The nitrogen adsorption-desorption isotherm and (b) the pore diameter distribution of LTO, LSO/LTO and [email protected]/LTO.

Fig. 9. (a) The CV of [email protected]/LTO electrode at a rate of 0.2 mV s−1 and (b) XRD pattern of [email protected]/LTO at the current density of 150 mA g−1 after 200 cycles.

are attributed to the insertion and de-insertion process between Li and Li4 Ti5 O12 (Li4 Ti5 O12 ↔ Li7 Ti5 O12 ) [9,10–17].The weak cathodic peak at ∼0.70 V could be ascribed to the solid electrolyte interface (SEI) film formation, the reduction of Li2 SiO3 to Si and the synchronous formation of Li2 O, which could be responsible for the first irreversible capacity loss [54,55]. The cathodic peak at 0.01–0.2 V may be ascribed to the insertion process between Li and Si, respectively. The anodic peak at ∼0.25 V is attributed to the de-insertion process between Li and Si [54,61]. In addition, the redox peaks at 0.01–0.2 V and ∼0.25 V may be also ascribed to the insertion and de-insertion process between Li and C [62]. The existence of few C is attributed to GE and acetylene black in the process prepared electrodes. After the first sweep, the cathodic peaks shift to rough positive migration, indicating the electrochemical polarization declines. To confirm the results of the CV, XRD pattern was used to study the structure of the [email protected]/LTO composite after 200 cycles. As shown in Fig. 9b, the major diffraction peaks of the LTO and Si (JCPDS 17-0901) can be observed in the pattern except the peaks of Cu substrate, which further confirms the results of CV.

4. Conclusions [email protected]/LTO nanocomposites have been synthesized via a hydrothermal route and further calcination. LSO/LTO nanospheres with the size of 50–100 nm are distributed on the graphene nanosheets, and LSO and LTO particles are attached together.

When tested as the anode for lithium ion batteries, the initial discharge of LTO, LSO/LTO and [email protected]/LTO are 160.8 mAh g−1 , 711.2 mAh g−1 and 720.6 mAh g−1 while their charge capacities are 124.5 mAh g−1 , 391.1 mAh g−1 and 463.4 mAh g−1 at the current density of 150 mA g−1 . After 200 cycles, their charge capacities retain 86.7%, 56.5% and 86.1%, respectively. Moreover, the charge rate capacities of retain 63.9%, 61.7% and 89.1% at the current density from 150 mA g−1 to 750 mA g−1 . And their recovery rates of the charge capacities are 87.9%, 84.8% and 91.0% when returned the current density of 150 mA g−1 , respectively. This strategy, with GE constituting a good supporting structure, is an effective way to improve the cycling performance of anode materials for lithium ion batteries.

Acknowledgements This work was supported by the Key Science and Technology Research Program of Henan province under Grant No. 152102210106 and the Key Science Research Program of High School of Henan province under Grant No. 16A480007.

References [1] J.M. Tarascon, M. Armand, Issues and challenges facing rechargeable lithium batteries, Nature 414 (2001) 359–367.

Q. Wang et al. / Applied Surface Science 403 (2017) 635–644 [2] H.G. Wang, D.L. Ma, X.L. Huang, Y. Huang, X.B. Zhang, General and controllable synthesis strategy of metal Oxide/TiO2 hierarchical heterostructures with improved lithium-ion battery performance, Sci. Rep. 2 (2012) 136. [3] X.L. Huang, X. Zhao, Z.L. Wang, L.M. Wang, X.B. Zhang, Facile and controllable one-pot synthesis of an ordered nanostructure of Co(OH)2 nanosheets and their modification byoxidation for high-performance lithium-ion batteries, J. Mater. Chem. 22 (2012) 3764–3769. [4] A.S. Arico, P.G. Bruce, B. Scrosati, J.M. Tarascon, W.V. Schalkwijk, Nanostructured materials for advanced energy conversion and storage devices, Nat. Mater. 4 (2005) 366–377. [5] Q.F. Wang, Y. Huang, J. Miao, Y. Zhao, Y. Wang, Synthesis and electrochemical characterizations of Ce doped SnS2 anode materials for rechargeable lithium ion batteries, Electrochim. Acta 93 (2013) 120–130. [6] Y. Zhao, X.F. Li, B. Yan, D.B. Xiong, D.J. Li, S. Lawes, X.L. Sun, Recent developments and understanding of novel mixed transition-metal oxides as anodes in lithium ion batteries, Adv. Energy. Mater. 1502175 (2016) 1–19. [7] L.L. Wang, K.B. Tang, M. Zhang, X.Z. Zhang, J.L. Xu, Facile synthesis of CuO nanoparticles as anode for lithium ion batteries with enhanced performance, Func. Mater. Lett. 1440008 (2014) 1–3. [8] Y. Zhao, Y. Huang, X. Sun, H.J. Huang, K. Wang, M. Zong, Q.F. Wang, Hollow Zn2 SnO4 boxes wrapped with flexible graphene as anode materials for lithium batteries, Electrochim. Acta. 120 (2014) 128–132. [9] Q.F. Wang, M.W. Lu, J. Miao, Y.Y. Hao, Synthesis and electrochemical performance of nanodiamond supported Li4 Ti5 O12 -Fe2 O3 composites as anode materials for lithium ion batteries, Mater. Lett. 164 (2016) 1–4. [10] K. Ariyoshi, R. Yamato, T. Ohzuku, Zero-strain insertion mechanism of Li[Li1/3 Ti5/3 ]O4 for advanced lithium-ion (shuttlecock) batteries, Electrochim. Acta 51 (2005) 1125–1129. [11] A. Mahmoud, J.M. Amarilla, I. Saadoune, Effect of thermal treatment used in the sol-gel synthesis of Li4 Ti5 O12 spinel on its electrochemical properties as anode for lithium ion batteries, Electrochim. Acta 163 (2015) 213–222. [12] E. Pohjalainen, T. Rauhala, M. Valkeapää, J.K. Kallioinen, T. Kallio, Effect of Li4 Ti5 O12 particle size on the performance of lithium ion battery electrodes at high C-rates and low temperatures, J. Phys. Chem. C 119 (2015) 2277–2283. [13] Y.B. Shen, M. Søndergaard, M. Christensen, S. Birgisson, B.B. Iversen, Solid state formation mechanism of Li4 Ti5 O12 from an anatase TiO2 source, Chem. Mater. 26 (2014) 3679–3686. [14] X. Lu, L. Zhao, X. He, R. Xiao, L. Gu, Y. Hu, H. Li, Z. Wang, X. Duan, L. Chen, et al., Lithium storage in Li4 Ti5 O12 spinel: the full static picture from electron microscopy, Adv. Mater. 24 (2012) 3233–3238. [15] J.L. Allen, T.R. Jow, J. Wolfenstine, Low temperature performance of nanophase Li4 Ti5 O12 , J. Power Sources 159 (2006) 1340–1345. [16] S. Ganapathy, M. Wagemaker, Nanosize storage properties in spinel Li4 Ti5 O12 explained by anisotropic surface lithium insertion, ACS Nano 6 (2012) 8702–8712. [17] K. Song, D.H. Seo, M.R. Jo, Y.I. Kim, K. Kang, Y.M. Kang, Tailored oxygen framework of Li4 Ti5 O12 nanorods for high-power Li ion battery, J. Phys. Chem. Lett. 5 (2014) 1368–1373. [18] Y.F. Tang, L. Yang, S.H. Fang, Z. Qiu, Li4 Ti5 O12 hollow microspheres assembled by nanosheets as an anode material for high-rate lithium ion batteries, Electrochim. Acta 54 (2009) 6244–6249. [19] T.F. Yi, Z.K. Fang, Y. Xie, Y.R. Zhu, S.Y. Yang, Rapid charge-discharge property of Li4 Ti5 O12 -TiO2 nanosheet and nanotube composites as anode material for power lithium-ion batteries, ACS Appl. Mater. Interfaces 6 (2014) 20205–20213. [20] S.M. Jiang, B.T. Zhao, Y.B. Chen, R. Cai, Z.P. Shao, Li4 Ti5 O12 electrodes operated under hurdle conditions and SiO2 incorporation effect, J. Power Sources 238 (2013) 356–365. [21] S. Sharmila, B. Senthilkumar, V.D. Nithya, K. Vediappan, C.W. Lee, R.K. Selvan, Electrical and electrochemical properties of molten salt-synthesized Li4 Ti5-x Snx O12 (x = 0.0, 0.05 and 0.1) as anodes for Li-ion batteries, J. Phys and Chem. Solids 74 (2013) 1515–1521. [22] Y. Ma, B. Ding, G. Ji, J.Y. Lee, Carbon-encapsulated F-doped Li4 Ti5 O12 as a high rate anode material for Li+ batteries, ACS Nano 7 (2013) 10870–10878. [23] C.H. Chen, J.T. Vaughey, A.N. Jansen, D.W. Dees, A.J. Kahaian, T. Goacher, M.M. Thackeray, Studies of Mg-substituted Li4-x MgxTi5 O12 spinel electrodes (0 ≤ x ≤ 1) for lithium batteries, J. Electrochem. Soc. 148 (2001) A102–1040. [24] C.F. Lin, B. Ding, Y.L. Xin, F.Q. Cheng, M.O. Lai, L. Lu, H.H. Zhou, Advanced electrochemical performance of Li4 Ti5 O12 -based materials for lithium-ion battery: Synergistic effect of doping and compositing, J. Power Sources 248 (2014) 1034–1104. [25] A. Nugroho, K.Y. Chung, J. Kim, A facile supercritical alcohol route for synthesizing carbon coated hierarchically mesoporous Li4 Ti5 O12 microspheres, J. Phys. Chem. C 118 (2014) 183–193. [26] J. Liu, K.P. Song, P.A.V. Aken, J. Maier, Y. Yu, Self-supported Li4 Ti5 O12 -C nanotube arrays as high-rate and long-life anode materials for flexible Li-ion batteries, Nano. Lett. 14 (2014) 2597–2603. [27] E. Zhao, X. Liu, H. Zhao, X. Xiao, Z. Hu, Ion conducting Li2 SiO3 -coated lithium-rich layered oxide exhibiting high rate capability and low polarization, Chem. Commun. 51 (2015) 9093–9096. [28] I.D. Raistrick, C. Ho, R.A. Huggins, Ionic Conductivity of some lithium silicates and aluminosilicates, J. Electrochem. Soc. 123 (1976) 1469–1476. [29] S. Goriparti, E. Miele, F.D. Angelis, E.D. Fabrizio, R.P. Zaccaria, C. Capiglia, Review on recent progress of nanostructured anode materials for Li-ion batteries, J. Power Sources 257 (2014) 421–443.

643

[30] B.J. Landi, M.J. Ganter, C.D. Cress, R.A. DiLeo, R.P. Raffaelle, Carbon nanotubes for lithium ion batteries, Energy Environ. Sci. 2 (2009) 638–654. [31] H.S. Zhou, S.M. Zhu, M. Hibino, I. Honma, M. Ichihara, Lithium storage in ordered mesoporous carbon (CMK-3) with high reversible specific energy capacity and good cycling performance, Adv. Mater. 15 (2003) 2107–2111. [32] W.-M. Zhang, X.-L. Wu, J.-S. Hu, Y.-G. Guo, L.-J. Wan, Carbon coated Fe3 O4 nanospindles as a superior anode material for lithium-ion batteries, Adv. Funct. Mater. 18 (2008) 3941–3946. [33] X.L. Jia, Y.H. Cheng, Y.F. Lu, F. Wei, Building robust carbon nanotube interweaved nanocrystal architecture for high performance anode materials, ACS Nano 8 (2014) 9265–9273. [34] Q.F. Wang, Y. Huang, J. Miao, Y. Zhao, W. Zhang, Y. Wang, Graphene-supported Ce-SnS2 nanocomposite as anode material for lithium-ion batteries, J. Am. Ceram. Soc. 96 (2013) 2190–2196. [35] T. Yuan, W.T. Li, W.M. Zhang, Y.S. He, C.M. Zhang, X.Z. Liao, Z.F. Ma, One-pot spray-dried graphene sheets-encapsulated nano-Li4 Ti5 O12 microspheres for a hybrid bat cap system, Ind. Eng. Chem. Res. 53 (2014) 10849–10857. [36] Z.L. Jian, L. Zhao, R. Wang, Y.S. Hu, H. Li, W. Chen, L.Q. Chen, The low-temperature (400 ◦ C) coating of few-layer graphene on porous Li4 Ti5 O12 via C28 H16 Br2 pyrolysis for lithium-ion batteries, RSC. Adv. 2 (2012) 1751–1754. [37] Q.F. Wang, M.W. Lu, J. Miao, Li2 SiO3 @Li4 Ti5 O12 nanocomposites as anode material for lithium-ion batteries, Mater. Technol. 31 (2016) 471–476. [38] S. Fang, L.F. Shen, G.Y. Xu, P. Nie, J. Wang, H. Dou, X.G. Zhan, Rational design of void-involved [email protected] nanospheres as high-performance anode material for lithium-ion batteries, ACS Appl. Mater. Interfaces 6 (2014) 6497–6503. [39] S. Yang, J. Miao, Q.F. Wang, M.W. Lu, J.F. Sun, T. Wen, Synthesis of graphitized carbon, nanodiamond and graphene supported Li4 Ti5 O12 and comparison of their electrochemical performance as anodes for lithium ion batteries, Appl. Surf. Sci. 389 (2016) 428–437. [40] J.S. Shin, C.H. Han, U.H. Jung, S.I. Lee, H.J. Kim, K. Kim, Effect of Li2 CO3 additive on gas generation in lithium-ion batteries, J. Power Sources 109 (2002) 47–52. [41] V. Etacheri, G. Michlits, M.K. Seery, S.J. Hinder, S.C. Pillai, A highly efficient TiO2-x Cx nano-heterojunction photocatalyst for visible light induced antibacterial applications, ACS Appl. Mater. Interfaces 5 (2013) 1663–1672. [42] F.L. Souza, P.R. Bueno, E. Longo, E.R. Leite, Sol-gel nonhydrolytic synthesis of a hybrid organic-inorganic electrolyte for application in lithium-ion devices, Solid State Ion. 166 (2004) 83–88. [43] A. Nugroho, K.Y. Chung, J. Kim, A facile supercritical alcohol route for synthesizing carbon coated hierarchically mesoporous Li4 Ti5 O12 Microspheres, J. Phys. Chem. C 118 (2014) 183–193. [44] K.S. Amanpreet, S. Surinder, P.P. Om, Neutron irradiation effects on optical and structural properties of silicate glasses, Mater. Chem. Phys. 115 (2009) 783–788. [45] C.A. Rfos, C.D. Willianms, Synthesis of zeolitic materials from natural clinker: a new alternative for recycling coal combustion by-products, Fuel 87 (2008) 2482–2492. [46] S.L. Lin, C.S. Hwang, Structures of CeO2 -Al2 O3 -SiO2 glasses, J. Non-Crystalline Solids 202 (1996) 61–67. [47] B.N. Roy, Assessment of the CaO-Al2 O3 system, J. Am. Ceram. Soc 73 (1990) 15–23. [48] V. Etacheri, J.E. Yourey, B.M. Bartlett, Chemically bonded TiO2 -bronze nanosheet/reduced graphene oxide hybrid for high-power lithium ion batteries, ACS NANO 8 (2004) 1491–1499. [49] D. Yan, Y. Bai, C.Y. Yu, X.G. Li, W.F. Zhang, A novel pineapple-structured Si/TiO2 composite as anode material for lithium ion batteries, J. Alloys. Compd. 609 (2014) 86–92. [50] Q. Li, L.W. Yin, J.Y. Ma, Z.Q. Li, Z.W. Zhang, A.L. Chen, C.X. Li, Mesoporous silicon/carbon hybrids with ordered pore channel retention and tunable crbon incorporated content as high performance anode materials for lithium-ion batteries, Energy 85 (2015) 159–166. [51] S. Guo, H.X. Li, H.M. Bai, Z.L. Tao, J. Chen, Ti/Si/Ti sandwich-like thin film as the anode of lithium-ion batteries, J. Power Sources 248 (2014) 1141–1148. [52] J. Mosa, J.F. Vélez, J.J. Reinosa, M. Aparicio, A. Yamaguchi, K. Tadanaga, M. Tatsumisago, Li4 Ti5O12 thin-film electrodes by sol-gel for lithium-ion microbatteries, J. Power Sources 244 (2013) 482–487. [53] W. Li, M.Z. Chen, J.J. Jiang, R. Wu, F. Wang, W.J. Liu, G.C. Peng, M.Z. Qu, Structural and electrochemical characteristics of SiO2 modified Li4 Ti5 O12 as anode for lithium-ion batteries, J.Alloys. Compd. 637 (2015) 476–482. [54] R.G. Lv, J. Yang, J.L. Wang, Y.N. NuLi, Electrodeposited porous-microspheres Li-Si films as negative electrodes in lithium-ion batteries, J. Power Sources 196 (2011) 3868–3873. [55] B. Philippe, R. Dedryvère, J. Allouche, F. Lindgren, M. Gorgoi, H. Rensmo, D. Gonbeau, K. Edström, Nanosilicon electrodes for lithium-ion batteries: interfacial mechanisms studied by hard and soft X-ray photoelectron spectroscopy, Chem. Mater. 24 (2012) 1107–1115. [56] S. Mao, X.K. Huang, J.B. Chang, S.M. Cui, G.H. Zhou, J.H. Chen, One-step, continuous synthesis of a spherical Li4 Ti5 O12 /graphene composite as an ultra-long cycle life lithium-ion battery anode, NPG. Asia. Mater. 7 (2015) 1–8. [57] W.T. Li, T. Yuan, W.M. Zhang, J.J. Ma, C.M. Zhang, Y.S. He, X.Z. Liao, Z.F. Ma, Influence of lithium precursors and calcination atmospheres on graphene sheets-modified nano-Li4 Ti5 O12 anode material, J. Power Sources 285 (2015) 51–62. [58] G.B. Xu, L.W. Yang, X.L. Wei, J.W. Ding, J.X. Zhong, P.K. Chu, Highly-crystalline ultrathin gadolinium doped and carbon-coated Li4 Ti5 O12 nanosheets for enhanced lithium storage, J. Power Sources 295 (2015) 305–313.

644

Q. Wang et al. / Applied Surface Science 403 (2017) 635–644

[59] Q.Y. Zhang, Y. Liu, H.S. Lua, D.P. Tang, C.Y. Ouyang, L.Z. Zhang, Ce3+ -doped Li4 Ti5 O12 with CeO2 surface modification by a sol-gel method for high-performance lithium-ion batteries, Electrochim. Acta 189 (2016) 147–157. [60] Q.F. Wang, M.W. Lu, J. Miao, S. Yang, T. Wen, J.F. Sun, Graphitized carbon and graphene modified Fe2 O3 /Li4 Ti5 O12 as anode material for lithium ion batteries, Surf. Interface Anal. 49 (2017) 63–70.

[61] Z.H. Wen, G.H. Lu, S. Mao, H. Kim, S.M. Cui, K.H. Yu, X.K. Huang, P.T. Hurley, O. Mao, J.H. Chen, Silicon nanotube anode for lithium-ion batteries, Electrochem. Commun. 29 (2013) 67–70. [62] I. Cameán, P. Lavela, J.L. Tirado, A.B. García, On the electrochemical performance of anthracite-based graphite materials as anodes in lithium-ion batteries, Fuel 89 (2010) 986–991.