hard carbon composite as anode material for Li-ion batteries

hard carbon composite as anode material for Li-ion batteries

Journal of Electroanalytical Chemistry 688 (2013) 86–92 Contents lists available at SciVerse ScienceDirect Journal of Electroanalytical Chemistry jo...

1MB Sizes 0 Downloads 71 Views

Journal of Electroanalytical Chemistry 688 (2013) 86–92

Contents lists available at SciVerse ScienceDirect

Journal of Electroanalytical Chemistry journal homepage: www.elsevier.com/locate/jelechem

Electrochemical profile of lithium titanate/hard carbon composite as anode material for Li-ion batteries Guan-nan Zhu, Yuan-jin Du, Yong-gang Wang, Ai-shui Yu, Yong-yao Xia ⇑ Department of Chemistry, Shanghai Key Laboratory of Molecular Catalysis and Innovative Materials, Institute of New Energy, Fudan University, Shanghai 200438, China

a r t i c l e

i n f o

Article history: Available online 8 August 2012 Keywords: Li4Ti5O12 Hard carbon Carbon coating State of charge Li-ion batteries

a b s t r a c t In this study, we report a carbon-coated Li4Ti5O12/hard carbon composite as an anode for Li-ion batteries. The composite anode provides a clear edge in state-of-charge (SOC) estimation of Li-ion batteries and thus enables efficient energy management and health assessment. The shape of the charge/discharge curve for Li4Ti5O12 (LTO) is significantly modified by the addition of the hard carbon in an optimised amount to present a smoothly sloped voltage profile in the voltage window between 0.01 and 1.5 V. This voltage range is highly effective in avoiding overcharging of a type of Li-ion batteries, where cell voltages remain constant over a large SOC region. Compared with the pristine LTO anode charged/discharged between 3 and 0.01 V, the composite anode, in addition to showing an improved specific capacity and cycling stability, demonstrates a strong enhancement in the rate capability; at a rate as high as 20 C, a capacity of 160 mAh/g is still retained, which has been proven to arise from a dual improvement in the electronic conductivity by the hard carbon through mixing and the soft carbon via carbon coating. The advantages in SOC prediction and the excellent rate capability are expected to greatly promote its practical application. Ó 2012 Elsevier B.V. All rights reserved.

1. Introduction Li4Ti5O12 (LTO) has been demonstrated to be one of the most promising anode materials for Li-ion batteries because of its flat charge/discharge profile at 1.55 V (vs. Li/Li+) and excellent Li-insertion/extraction reversibility with zero structural change [1,2]. Despite these advantages, the rate capability of LTO is relatively low because of a large polarisation at high charge–discharge rates. This polarisation results from both the poor electronic conductivity (<1013 S/cm) and sluggish lithium-ion diffusion [3–7] and precludes its wide application in situations that require high power density, such as hybrid and pure-electric vehicles. Over the past few years, numerous researchers have attempted to enhance its electronic conductivity either by doping of metal or non-metal ions or by surface modification via mixing or coating with conductive species [8–11]. They have also attempted to reduce the distances of the Li-ion diffusion pathways through the synthesis of nanostructured LTO materials with various morphologies [12–16]. Another key factor in exploring the full benefit of Li-ion batteries for practical implementations is efficient energy management through real-time prediction of the SOC (state of charge), the SOH (state of health) and the available power [17–19]. Nevertheless, the voltage-based SOC estimation for a cell comprising electrodes ⇑ Corresponding author. Tel./fax: +86 21 51630318. E-mail address: [email protected] (Y.-y. Xia). 1572-6657/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jelechem.2012.07.035

with a flat charge/discharge plateau is particularly challenging because the cell potential remains substantially flat over a broad SOC range and an abrupt voltage drop/rise occurs near the end of the charge/discharge plateau, which easily leads to overcharging/ overdischarging [20–23]. A typical example is the LTO/LFP (LiFePO4) battery system, which has been considered as a promising candidate because of its smooth voltage output and long cycle life. Until now, however, few studies have focused on this direction. One suggestion is to control the cut-off potential of the lithiation step for LTO from the traditional 1–0 V (vs. Li/Li+). In the case of 1 V as the cut-off potential, small errors in open-circuit voltage (OCV) estimation in the 1.55–1 V range can result in significantly large errors in the prediction of the SOC; while in the case of 0 V (vs. Li/Li+), the previously described sharp voltage drop is replaced by an extended sloped curve between 1.55 and 0 V [24], which serves as a continuous voltage marker that allows for a significantly larger deviation in the estimation of the OCV. However, the inflection point has been observed at approximately 0.6 V (vs. Li/Li+) during charge/discharge, which is also unfavourable for practical OCV detection. In this study, for the first time, a solution has been developed to overcome the barriers that hinder the practical application of LTO anode through the preparation of a carbon-coated LTO/Hard C (hard carbon) composite. The advantages of the composite anode over the pristine LTO anode in electrochemical performances within the 3–0.01 V potential range, including the specific capacity, the

G.-n. Zhu et al. / Journal of Electroanalytical Chemistry 688 (2013) 86–92


cycling stability and the rate capability, are discussed in detail. Specifically, the concept of the composite as an effective voltagebased SOC estimation tool is investigated and verified in the full cell using LFP as the cathode. 2. Experimental 2.1. Synthesis of the carbon-coated LTO/Hard C composite Li4Ti5O12 (LTO) was prepared by a conventional solid-state reaction as follows: TiO2 ((AR, Zhoushan Mingri New Material) and Li2CO3 (AR, Shanghai Chemical Agents) were mixed in a molar ratio of 2.5/1 and then heated at 800 °C for 24 h to obtain a well-crystallised pure phase. The hard carbon (Hard C) was prepared according to the previously reported process [25]. For the preparation of the carbon-coated LTO/Hard C composite, the LTO and Hard C were first ball-milled in a weight ratio of 7/3 (Fritsch Pulverisette 6), and the resulting mixed powder was subsequently carbon coated at 700 °C for 2 h via the CVD (chemical vapour deposition) method described in our previous work [11], The composite was then heattreated at different temperatures of 800, 900, 950 or 1000 °C for 2 h to obtain a series of carbon-coated samples. In addition, to clarify the influence of the carbon coating on each component of the composite, the LTO and Hard C were carbon-coated under a similar two-step route of 700 °C for 2 h and 800 °C for 2 h, respectively. 2.2. Characterisations The phase purity of the different carbon-coated LTO/Hard C composites was analyzed by powder X-ray diffraction (XRD, Bruker D2 Phaser Table-top Diffractometer) using Cu Ka radiation. The morphologies of the different samples were characterized using transmission electron microscopy (TEM, Joel JEM2010). The content of the coated carbon in the carbon-coated LTO/Hard C composite was determined by a thermogravimetric measurement (TG, Perkin-Elmer TGA 7) from 20 to 1000 °C under an air flow of 60 mL/min with a heating rate of 5 °C/min. 2.3. Electrochemical tests All of the working electrodes, including the pristine and carboncoated LTOs, pristine and carbon-coated Hard Cs, and carboncoated LTO/Hard C composite, were prepared by mixing 85 wt% active material, 5 wt% carbon black and 10 wt% polyvinyl difluoride (PVDF) dispersed in 1-methyl-2-pyrrolidinone (NMP). The slurry was subsequently cast uniformly onto a piece of Cu foil to prepare the electrode film, which was vacuum-dried at 80 °C for 12 h before being assembled in a glove box filled with argon gas. The typical mass load of active material was controlled to be approximately 5 mg/cm2. A coin-type cell (CR2016) was used in a half-cell test and was assembled with a configuration of working electrode/ separator/lithium-metal counter electrode, as well as in a full-cell test with a configuration of working anode/separator/LFP cathode. The electrolyte solution in both cells was 1 M LiPF6 in a mixture of ethylene carbonate (EC), dimethyl carbonate (DMC) and ethyl methyl carbonate (EMC) (1:1:1 by volume). The electrochemical tests were evaluated under an automatic battery testing system (LAND model CT2001A). 3. Results and discussion The XRD pattern of the carbon-coated LTO/Hard C composite treated at 700 °C is shown in Fig. 1a. Major diffraction peaks at 18.4°, 35.6°, 43.4°, 47.4°, 57.2°, 62.8°, 66.1°, 74.3°, 75.4°, and 79.4° (2h) are found to be indexed as a spinel-type Li4Ti5O12

Fig. 1. XRD patterns of the LTO/Hard C composite carbon-coated at 700 °C for 2 h (a) and the patterns of the samples heat-treated for 2 h at 800 °C (b), 900 °C (c), 950 °C (d) and 1000 °C (e).

(LTO) (PDF card no. 49-0207). However, no carbon is detected in the XRD pattern because of its low content or amorphous characteristics. To evaluate the influence of temperature on the phase purity, the composite was further treated under a series of temperatures that ranged from 800 to 1000 °C, and the corresponding XRD results are illustrated in Fig. 1b–e. According to Fig. 1b, the XRD pattern of the sample treated at 800 °C exhibits no detectable differences from that of the 700 °C-treated composite. When the temperature is increased to 900 °C, the original XRD pattern is retained, except for the presence of a new small peak at 20.6° (2h), which continuously intensifies as the temperature is increased to 950 and 1000 °C. In addition to the tiny peak, other new diffractions at 20.4°, 26.0°, 33.6° and 40.3° (2h), accompanied by a significant decrease in the intensity of some previous peaks at 18.4°, 35.6°, and 43.4° (2h), are also observed for the 950 °C- and 1000 °C-treated samples. These changes are likely due to the partial reduction of Ti4+ or oxygen deficiencies that result from the high-temperature treatment during the carbon coating process. The significant changes in the XRD pattern are also reflected in the colour of the sample, which turns from brown to dark blue. The TEM results in Fig. 2 give clear details about the distribution and structure of the coated carbon layer. Based on Fig. 2a, the deposited carbon forms a uniform coating over the LTO/Hard C active material, which suggests that the two components, LTO and Hard C, are firmly attached by the carbon bridge rather than being separated from each other. Furthermore, the HRTEM micrographs in Fig. 2b–d reflects the variations in the carbon structure with the increased annealing temperature. From 700 to 900 °C, the coated carbon presents a more clearly layered and compact structure that is typical of graphitized carbon, confirming that a higher applied temperature increases the attained degree of graphitization [26,27]. Fig. 3a shows the first and second charge/discharge curves of the pristine LTO anode tested in a voltage window of 3.0–0.01 V (vs. Li/Li+). On the first discharge, the curve shows a flat voltage plateau at 1.55 V (vs. Li/Li+) until 160 mAh/g; during the further discharge, the curve exhibits a sloped voltage profile until the voltage reaches 0.01 V (vs. Li/Li+), contributing to a capacity of 102 mAh/g in three steps distinguished by two inflection points at 0.85 and 0.55 V (vs. Li/Li+). On the first charge, the operating voltage increases from 0.01 to 0.55 V (vs. Li/Li+) with a charge capacity of 40 mAh/g, and the charge curve shows a linear slope; the voltage then increases and remains at 1.58 V (vs. Li/Li+) to give a plateau charge capacity of 158 mAh/g, and at the end of charge, the operating voltage rises sharply to 3.0 V (vs. Li/Li+), which con-


G.-n. Zhu et al. / Journal of Electroanalytical Chemistry 688 (2013) 86–92

Fig. 2. TEM image of the carbon-coated LTO/Hard C composite heat-treated at 800 °C (a). HRTEM image of the carbon-coated LTO/Hard C composites heat-treated at 800 °C (b), 900 °C (c) and 1000 °C and (d).

Fig. 3. The first and second charge/discharge curves (a) and the cycling performance for 100 cycles (b) of the pristine LTO at a rate of 0.5 C.

tributes to a reversible capacity of 218 mAh/g. For the subsequent cycles, the inflection point at 0.85 V (vs. Li/Li+) observed during the first discharge has disappeared, and the coulombic efficiency shown in Fig. 3b exhibits a significant increase from 85.5% for the 1st cycle to 96.8%, 98.2%, 99.5% and 99.7% for the 2nd, the 3rd, the 6th and the 100th cycles, respectively. Compared with the electrochemical performance of a cell operated between 1 and 3 V in our previous studies [11,26], the current cell with an operating voltage range between 0.01 and 3 V shows a larger discharge capacity of 220 mAh/g but a higher irreversibility of 35 mAh/g and an inferior cycling performance, with a capacity retention of 90.9% after 100 cycles. For the discharge test between 3 and 0.01 V, the Li-ion intercalation of LTO is considered to proceed in a topotactic manner by two steps [24]. The first step could be represented by a one-electron transfer reaction [1] in which one

Li-ion can be intercalated electrochemically into Li[Li1/3Ti5/3]O4 at a flat potential of 1.55 V (vs. Li/Li+) until no vacant octahedral sites remain for Li-ion accommodation; because the remaining tetravalent Ti ions are available to accept electrons, Li-ions further intercalate into the tetrahedral sites through reaction [2] to deliver the additional discharge capacity. According to XRD analyses performed by several research groups [23,24], the lattice dimensions experience a 0.5% change when the LTO electrodes are operated in a voltage window from 3 to 0.01 V, which partially explains the inferior cycling stability given that almost no lattice distortion occurs when the cells are charge/discharged between 3 and 1 V [24]. However, a root cause in terms of cyclability is expected to be discovered through the use of another tool to detect structural variations and the formation of a new phase at potentials less than 1 V that XRD characterisation fails to detect [24,28].

G.-n. Zhu et al. / Journal of Electroanalytical Chemistry 688 (2013) 86–92 þ

Li½Li1=3 Ti5=3 O4 þ Li þ e ! Li2 ½Li1=3 Ti5=3 O4 þ

Li2 ½Li1=3 Ti5=3 O4 þ dLi þ de ! Li2þd ½Li1=3 Ti5=3 O4

ð1Þ ð2Þ

It has been widely recognised that the Hard C material has a large Li-ion storage capacity, a stable cyclic performance, and a high electronic conductivity [25,29,30]. In addition, Hard C gives a sloped charge/discharge curve in the main capacity-contribution potential range of 1–0 V (vs. Li/Li+). These features are expected to compensate for deficiencies observed in pristine LTO. The first two charge/discharge curves of Hard C between 0.01 and 3 V are illustrated in Fig. 4a. The voltage profiles present a sloped feature rather than a distinct plateau or a staging mechanism observed for graphitic carbon, which can be a clear advantage for allowing a voltage-based SOC estimation [31]. In the first cycle, the anode delivers a discharge capacity of 329 mAh/g and a reversible charge capacity of 247 mAh/g, which results in a low initial coulombic efficiency of 75.1%. A large irreversible capacity is common for Hard C materials, which is mainly attributed to the formation of SEI (Solid Electrolyte Interface) film on the carbon surface and the reaction of lithium with surface functional groups resulted from air exposure [29]. To reduce the irreversibility, we modified the surface with a soft-carbon coating via the CVD method, which helps to suppress the diffusion of air and water into Hard C [29,30] and thus leads to a significant enhancement in the initial coulombic efficiency from 75.1% to 81.2%, as demonstrated in Fig. 4b. Furthermore, the carbon-coated Hard C shows a cycling performance superior to that of pristine Hard C by maintaining a discharge capacity of 252 mAh/g after 100 cycles with a capacity retention of 94.7% and a high coulombic efficiency of above 98.8% for each cycle. These results prove the effectiveness of the carbon-coating treatment in improving the coulombic efficiency and cycling stability of the Hard C material. The CVD method was also used for the preparation of the carbon-coated LTO/Hard carbon composite. In addition to improving the initial reversibility and cycling stability of Hard C, the resulting soft carbon is expected to improve the electronic conductivity of the LTO phase and make it closely attach to Hard C [11,26]. The electrochemical performance of the carbon-coated LTO/Hard C composite was extensively evaluated, including its charge/discharge curves, cycling stability and rate capability. As shown in Fig. 5a, the inflection point at 0.55 V (vs. Li/Li+) observed in the charge/discharge curve of the pristine LTO anode (Fig. 3a) has been successfully eliminated by the addition of hard carbon, and as a result, the composite anode exhibits a smooth sloped voltage profile before the end of discharge, which enables the easy estimation of the voltage-based SOC. Notably, the amount of hard carbon in


the carbon-coated LTO/Hard C composite has been systematically tuned, and the optimal amount was determined to be 30 wt%. This hard-carbon content eliminates the inflection in the charge/discharge profile while introducing the least amount of adverse irreversibility effects. Based on the TG result in Fig. 6, which gives a total weight loss of 32.9 wt%, the soft carbon that resulted from the CVD method is calculated to account for approximately 3 wt% in both the carbon-coated Hard C and the composite. In addition to a slightly increased stable discharge capacity of 233 mAh/g (Fig. 5a) and improved cycling stability with a capacity retention of 92.5% after 100 cycles (Fig. 5b), the composite anode demonstrates a significantly better rate capability than the pristine LTO anode. As illustrated in Fig. 5c and d, the composite anode delivers a capacity of 232, 222, 202 and 177 mAh/g when the rate is progressively increased from 0.2, 2, 5 to 10 C, and even at a rate as high as 20 C, it still delivers a discharge capacity of 160 mAh/g with a well-preserved flat plateau region. By contrast, the pristine LTO anode shows such a rapid capacity decay in both the flat and sloped voltage profiles that it hardly works at the rate of 10 C because of the large polarization. These results indicate that the soft-carbon coating undoubtedly plays a significant role in improving the electronic conductivity of LTO; however, the question remains about whether the Hard C influences the electrochemical properties of LTO. To acquire deep insight into the possible interactions that exist between the LTO and Hard C components in the composite, a modelling method is applied [20]. The SOC curves of the individual carbon-coated LTO and Hard C anodes under an extremely low rate of 0.05 C, which we treat as an equilibrium condition, are shown in Fig. 7a; correspondingly, the combined SOC curves calculated from Eq. (1) are defined as ‘‘model results’’ and are also presented:




where Q, QLTO and QHC represent the capacity of the composite, the carbon-coated LTO and the hard carbon, respectively. As a consequence, the modelling result (red line in Fig. 7a) agrees well with the experimental data (blue line in Fig. 7a) for the composite anode, which indicates that no voltage interference is detected between the LTO and the Hard C in the composite or that electrochemical reactions for them could occur in an independent manner in avoidance of anomalous battery behaviours. Meanwhile, the respective and combined SOC curves at a high rate of 10 C were also performed, and the results are illustrated in Fig. 7b. However, in the high-rate case, the experimental curve of the composite shows an evident difference with the model prediction by delivering a far smaller polarization and better maintenance of the flat plateau typ-

Fig. 4. The first and second charge/discharge curves (a) and the cycling performance for 100 cycles (b) of pristine Hard C and 800 °C-treated carbon-coated Hard C at a rate of 0.5 C.


G.-n. Zhu et al. / Journal of Electroanalytical Chemistry 688 (2013) 86–92

Fig. 5. The first and second charge/discharge curves (a) and cycling performance for 100 cycles (b) of the 800 °C-treated carbon-coated LTO/Hard C composite at a rate of 0.5 C. A comparison of the rate capability between the pristine LTO (c) and the 800 °C-treated carbon-coated LTO/Hard C composite (d) under different current rates (from right to left): 0.5, 1, 2, 5, 10 and 20 C.

Fig. 6. TG curves of the LTO/Hard C mixture (a) and of the carbon-coated LTO/Hard C composite (b) measured from 20 to 1000 °C under an air flow of 60 mL/min with a heating rate of 5 °C/min.

ical of the LTO phase. Since the improvement in electronic conductivity by carbon coating is similar to both the experimental and modelling results, this significant difference strongly demonstrates the positive influence of Hard C on the electronic conductivity of LTO and further on the rate capability of the composite, which is assumed to be closely related with the firm contact between LTO and Hard C linked by the coated carbon. Therefore, the two components in the composite interact with respect to the electronic conductivity but perform no structural interference, which enables the compos-

Fig. 7. A combined curve of the voltage versus the SOC for the individual 800 °Ctreated carbon-coated LTO and pristine Hard C (red line), and an experimental plot of the voltage versus the SOC for the 800 °C-treated carbon-coated LTO/Hard C composite (blue line) at a rate of 1/20 C (a) and a rate of 10 C (b).(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

G.-n. Zhu et al. / Journal of Electroanalytical Chemistry 688 (2013) 86–92


Fig. 8. The typical charge/discharge curves of the carbon-coated LTO/Hard C composite heat-treated at 900, 950 and 1000 °C for 2 h (a). A comparison of the voltage versus the SOC curves between different carbon-coated composites heat-treated at 800, 900 and 950 °C for 2 h (b).

traditional LTO/LFP cell was also assembled (Fig. 9). The end of the charge/discharge profile for this cell exhibits an inevitable and abrupt voltage rise/drop, which creates a significant safety concern of overcharging/overdischarging. The carbon-coated LTO/ Hard C composite/LFP full cell, however, delivers an extended inclined curve during charge (1.9–3.5 V) and discharge (1.8–1.0 V) to a different extent. This behaviour provides an effective indicator to signal the real-time SOC and to predict the end of the charge/discharge in avoidance of overcharging/overdischarging. 4. Conclusions

Fig. 9. The typical charge/discharge curves of the pristine LTO/LFP full cell (a) and 800 °C-treated carbon-coated composite/LFP full cell (b) at a rate of 0.5 C.

ite to exhibit stable performance in practical high-rate use and in voltage-based SOC estimation. Furthermore, the electrochemical profiles of composite samples heat-treated under 900, 950 and 1000 °C were investigated, and the results are illustrated in Fig. 8a. Consistent with the XRD results in Fig. 1, the heat-treated composites exhibit significant changes in their charge/discharge curves, particularly in the flat plateau region and at the end of charge. As the temperature is increased from 900 to 1000 °C, the flat plateau at approximately 1.55 V (vs. Li/Li+) gradually disappears until the voltage profiles exhibit a generally inclined feature with several stages characterized by different slopes, whereas the capacity decreases to 215, 206 and 130 mAh/ g for 900, 950 and 1000 °C, respectively. Despite some capacity loss for the samples treated at 900 and 950 °C, the shape modification at the end of the charge plateau is nonetheless encouraging. The SOC curves for samples treated at 800, 900 and 950 °C are plotted in Fig. 8b to demonstrate the differences in their end-of-charge behaviours. Compared with the 800 °C-treated sample, the other two samples demonstrate a more pronounced sloping tendency with increased annealing temperature, which is favourable for the prevention of overdischarging in full cells. In addition, because of the existence of the sloping area, an even greater discrepancy in the OCV prediction would not significantly affect the accuracy of the SOC estimation, which is expected to facilitate the design of estimation approaches and models. The optimization of the voltage-based SOC estimation for the carbon-coated LTO/Hard C composite anode is further verified in a full cell using LiFePO4 (LFP) as the cathode. For comparison, the

In the present study, we report a carbon-coated Li4Ti5O12/Hard C composite anode and perform extensive electrochemical tests in a voltage window of 3–0.01 V (vs. Li/Li+). Compared with the pristine LTO, the composite anode shows a larger specific capacity of 233 mAh/g and improved cycling stability, with a capacity retention of 92.5% after 100 cycles. Furthermore, the composite anode demonstrates a significantly better rate capability, and a capacity of 160 mAh/g is still retained at a rate as high as 20 C. This improved rate capability is attributed to dual improvements in the electronic conductivity resulted from hard-carbon mixing and soft-carbon coating. Specifically, the LTO-based composite successfully addresses the overcharging/overdischarging issue common to Li-ion batteries where cell voltages remain constant over a large SOC region, demonstrating its potential for improving the voltage-based SOC estimation for the LTO-based composite/LFP full cell by taking advantage of the modified voltage profile near the end of the charge/discharge process. The enhancements in the energy density, power density and the SOC prediction are certain to promote the practical application of the LTO-based composite material. Acknowledgements This work was partially supported by the National Natural Science Foundation of China (20925312), the State Key Basic Research Program of PRC (2011CB935903), Shanghai Science & Technology Committee (10JC1401500, 08DZ2270500), and Fudan University Doctoral Scientific Funding Project (EHH1615203). References [1] A.N. Jansen, A.J. Kahaian, K.D. Kepler, P.A. Nelson, K. Amine, D.W. Dees, D.R. Vissers, M.M. Thackeray, J. Power Sources 81–82 (1999) 902. [2] K. Nakahara, R. Nakajima, T. Matsushima, H. Majima, J. Power Sources 117 (2003) 131.


G.-n. Zhu et al. / Journal of Electroanalytical Chemistry 688 (2013) 86–92

[3] C.H. Chen, J.T. Vaughey, A.N. Jansen, D.W. Dees, A.J. Kahaian, T. Goacher, M.M. Thackeray, J. Electrochem. Soc. 148 (2001) A102. [4] L. Cheng, H.J. Liu, J.J. Zhang, H.M. Xiong, Y.Y. Xia, J. Electrochem. Soc. 153 (2006) A1472. [5] C.H. Jiang, M. Ichihara, I. Honma, H.S. Zhou, Electrochim. Acta 52 (2007) 6470. [6] J.R. Li, Z.L. Tang, Z.T. Zhang, Electrochem. Commun. 7 (2005) 894. [7] C.H. Jiang, Y. Zhou, I. Honma, T. Kudo, H.S. Zhou, J. Power Sources 166 (2007) 514. [8] S. Huang, M. Woodson, R. Smalley, J. Liu, Nano Lett. 4 (2004) 1025. [9] H.M. Xie, R.S. Wang, J.R. Ying, L.Y. Zhang, A.F. Jalbout, H.Y. Yu, G.L. Yang, X.M. Pan, Z.M. Su, Adv. Mater. 18 (2006) 2609. [10] S.A. Al-Muhtaseb, J.A. Ritter, Adv. Mater. 15 (2003) 101. [11] G.N. Zhu, C.X. Wang, Y.Y. Xia, J. Electrochem. Soc. 158 (2011) A102. [12] A. Jaiswal, C.R. Horne, O. Chang, W. Zhang, W. Kong, E. Wang, T. Cherm, M.M. Doeff, J. Electrochem. Soc. 156 (2009) A1041. [13] H.W. Lu, W. Zeng, Y.S. Li, Z.W. Fu, J. Power Sources 164 (2007) 874. [14] E.M. Sorensen, S.J. Barry, H.K. Jung, J.R. Rondinelli, J.T. Vaughey, K.R. Poeppelmeier, Chem. Mater. 18 (2006) 482. [15] A.S. Prakash, P. Manikandan, K. Ramesha, M. Sathiya, J.M. Tarascon, A.K. Shukla, Chem. Mater. 22 (2010) 2857. [16] G.N. Zhu, H.J. Liu, J.H. Zhuang, C.X. Wang, Y.G. Wang, Y.Y. Xia, Energy Environ. Sci. 4 (2011) 4016. [17] M. Verbrugge, Modern Aspects of Electrochemistry, first ed., M. Schelsinger, New York, 2009.

[18] B. Kang, G. Ceder, Nature 190 (2009) 458. [19] A. Yamada, H. Koizumi, S.I. Nishimura, N. Sonoyama, R. Kanno, M. Yonemura, T. Nakamura, Y. Kobayashi, Nat. Mater. 5 (2006) 357. [20] J. Wang, M.W. Verbrugge, P. Liu, J. Electrochem. Soc. 157 (2010) A185. [21] N. Meethong, H.Y.S. Huang, W.C. Carter, Y.M. Chiang, Electrochem. Solid-State Lett. 10 (2007) A134. [22] N. Meethong, H.Y.S. Huang, S.A. Speakman, W.C. Carter, Y.M. Chiang, Adv. Funct. Mater. 17 (2007) 1115. [23] V. Srinivasan, J. Newman, J. Electrochem. Soc. 151 (2004) A1517. [24] X.L. Yao, S. Xie, H.Q. Nian, C.H. Chen, J. Alloys Compd. 465 (2008) 375. [25] K. Tokumitsu, H. Fujimoto, A. Mabuchi, T. Kasuh, J. Power Sources 90 (2000) 206. [26] L. Cheng, X.L. Li, H.J. Liu, H.M. Xiong, P.W. Zhang, Y.Y. Xia, J. Electrochem. Soc. 154 (2007) A692. [27] R. Dominko, M. Bele, M. Gaberscek, M. Remskar, D. Hanzel, S. Pejovnik, J. Jamnik, J. Electrochem. Soc. 152 (2005) A607. [28] T. Ohzuku, A. Ueda, N.J. Yamamoto, J. Electrochem. Soc. 142 (1995) 1431. [29] J.H. Lee, H.Y. Lee, S.M. Oh, S.J. Lee, K.Y. Lee, S.M. Lee, J. Power Sources 166 (2007) 250. [30] W. Xing, J.R. Dahn, J. Electrochem. Soc. 144 (1997) 1195. [31] H. Groulf, B. Kaplan, F. Lantelme, S. Komaba, N. Kumagai, H. Yashiro, T. Nakajima, B. Simon, A. Barhoun, Solid-State Ionics 177 (2006) 869.