Topotactic conversion-derived Li₄Ti₅O₁₂–rutile TiO₂ hybrid nanowire array for high-performance lithium ion full cells

Abstract

A carbon-free and binder-free Li₄Ti₅O₁₂–rutile TiO₂ (LTO–RTO) hybrid nanowire array electrode has been designed via the partial topotactic conversion of a RTO nanowire array. In such a nanowire electrode: (i) the inner 1D RTO nanowire acts as the electron pathway (conductivity is higher than that of LTO) ultimately providing the electrode with an enhanced conductivity, and RTO also has a higher theoretical capacity (336 mA h g⁻¹) than LTO, thus can essentially improve the capacity of the LTO-based electrode; (ii) the outer LTO is found to have a preferred orientation of (400) planes for facile lithium insertion and its “zero strain” feature is helpful to restrain the slight volumetric expansion of the inner RTO. Accordingly, with the elegant synergistic contribution from LTO and RTO, the optimized 3 : 1 LTO–RTO electrode exhibits high capacity (among the highest: ∼181 mA h g⁻¹ at ∼1 C rate), excellent cyclability and good rate capability. A lithium ion full cell assembled using a LTO–RTO hybrid array as the anode and commercial LCO film as the cathode further shows outstanding performance with an ultralong lifetime (thousands cycles even at ∼10 C) and is capable of powering practical electronic devices, such as Nixie tubes and even an 8 × 8 LED lattice board.

---

1. Introduction

In recent years, many efforts have been made in developing power sources for clean/renewable energy systems, multifunctional portable electronic devices, and nano-/microelectromechanical systems (NEMS/MEMS).^1–5 Since rechargeable lithium ion batteries (LIBs) are considered as attractive energy storage devices for these applications, considerable attention has been focused on how to increase their energy/power density and rate capability, etc.^6–12 Although LIBs have gained current commercial success, they still suffer from some restrictions. Commercial LIBs using graphite as the anode give rise to low volumetric energy density and safety issues, mainly caused by dendritic lithium growth on the anode surface.^13–16 As one promising alternative anode material to graphite, spinel lithium titanate (Li₄Ti₅O₁₂, LTO) has drawn great interest due to several unique advantages,^17–27 which include low cost, a flat and high potential plateau (∼1.5 V vs. Li/Li⁺) for lithium storage (avoids Li plating and electrolyte decomposition), almost no volume change (<1%; “zero strain”) during cycling and high thermal stability upon full charge–discharge. In spite of these advantages, the practical implementation of LTO has been limited by two major drawbacks: a low electrical conductivity (∼10⁻¹³ S cm⁻¹) and small lithium diffusion coefficient (10⁻¹³ to 10⁻⁹ cm² s⁻¹).^28,29 Because of these two drawbacks, bulk LTO typically exhibits a low capacity and very poor rate performance.

To enhance the lithium storage kinetics, the nanostructuring of LTO was extensively employed to shorten the ion diffusion path. Much more attention has been paid to improving the electrical conductivity of LTO by coating with carbon^27,28,30 or anatase²⁹/rutile³¹ TiO₂. Other routes including metal doping,^32,33 nitridation³⁴ and hybridization with CNTs,³⁵ graphene,³⁶ and anatase TiO₂,^37–39 etc. have also been developed. Despite that fast kinetics can be guaranteed in such modified LTO electrodes, a notable amount of carbon black and polymer binder (∼15–20 wt%) are generally included as the additives to form an electrically conductive matrix and a continuous film structure respectively based on the common design rule of conventional electrodes. Since carbon is inactive to lithium at the potential >1 V vs. Li/Li⁺ and the binder is insulating, their presence would drastically lower the energy density of the electrode. For LTO whose theoretical capacity is originally not high (175 mA h g⁻¹), this is a much more serious issue attracting very little attention. Moreover, considering that some LTO nanostructures are already integrated with nanocarbon prior to electrode film processing, the safety issue could not be completely solved.⁴⁰ Therefore, it is highly interesting if a carbon- and binder-free LTO electrode with a high capacity at a high rate and excellent cycling stability could be developed.

An exciting report demonstrated recently that LTO can indeed discharge rapidly and cycle effectively without any conductive carbon.^41,42 The underlying reason for this surprising behavior is the early formation of mixed Ti⁴⁺/Ti³⁺ surface oxidation states that are highly electrically conductive, as well as the rapid propagation of the conducting phase.⁴² Insulating polymer binder, however, was still employed as the particulate LTO needs to be interconnected. To achieve a truly carbon-/binder-free LTO electrode, one promising way is to fabricate a single-crystalline LTO nanostructure directly on the current collector, with appropriate modulation of the LTO surface chemistry, such as hydrogenation,⁴⁰ to enhance the electrical conductivity.

Herein, we provide an alternative protocol to construct an LTO-based nanostructure anode without using any carbon or binder. Our LTO-based anode is designed via the partial topotactic conversion of rutile TiO₂ nanowire arrays growing directly on the current collector. The conversion mechanism is clearly elucidated. The optimized partial conversion leads to a novel ordered nanowire array architecture with dual phases of LTO (outer) and rutile TiO₂ (RTO, inner). In such an LTO–RTO nanowire array electrode, (i) the outer LTO is found to have a preferred orientation of (400) planes and its “zero strain” feature is beneficial to restrain the slight volumetric expansion of the inner RTO. The inner 1D RTO nanowire acts as an electron pathway (its conductivity is higher than that of LTO⁴³) ultimately providing the electrode with an enhanced conductivity, which when combined with the reduction of the lithium diffusion length within the nanowire diameter and the facile lithium insertion of the (400) planes in LTO, is expected to provide the LTO–RTO electrode with fast kinetics; (ii) RTO has a higher theoretical capacity (336 mA h g⁻¹) than LTO,⁴⁴ thus the integration of RTO can essentially improve the capacity of the LTO-based electrode. (iii) the general merits of the nanostructured array electrode in energy storage^6,45–51 have also been fully utilized, such as the sufficient contact of active materials with electrolyte and the facile electron collection, etc. As a result, the high capacity (among the highest: ∼181 mA h g⁻¹ at 175 mA g⁻¹; and superior or comparable capacities to most state-of-the-art LTO electrodes at higher rates) and excellent cycleability can be realized in the absence of any doping, conductive material coating and hydrogenation/nitridation of LTO. Using our LTO–RTO as the anode and commercially available LiCoO₂ (LCO) as the cathode, we further assembled a full-cell LIB that shows a fairly flat plateau, good rate capability and ultralong cycle life (thousands of cycles even at ∼10 C rate) and is capable of powering practical devices, such as Nixie tubes and even an 8 × 8 LED lattice board.

2. Experimental section

2.1 Materials preparation

The tetragonal RTO nanowire array was prepared by a hydrothermal process modified from that reported previously.⁵² Firstly, the carbon cloth was immersed into a solution (1.1 mL titanium tetrachloride in 50 mL ethanol) for 12 h after cleaning by sonication in a 1 : 1 : 1(v/v/v) mixture of acetone, ethanol, and deionized water. Then, the carbon cloth was annealed at 400 °C for 30 min to obtain TiO₂ seeds. The treated carbon cloth was loaded into a sealed Teflon-lined stainless steel reactor (80 mL volume) containing a mixture of acetone (20 mL), 37% hydrochloric acid (20 mL) and tetrabutyl titanate (1.3–1.8 mL). The reactor was then heated at 200 °C for 2 h. After the reaction, the carbon cloth was washed in ethanol and dried in air. The LTO–RTO hybrid nanowire array was synthesized by a solid-state reaction with LiOH. Typically, LiOH·H₂O in a stoichiometric amount was dropped onto the carbon cloth covered with an RTO nanowire array. The cloth was then dried at 80 °C overnight in air and finally calcined at 700 °C for 10 h in argon gas.

2.2 Materials characterizations

The as-prepared products were characterized with X-ray diffraction (XRD, Bruker D-8 Avance), scanning electron microscopy (SEM, JEOL, JSM-6700F), and transmission electron microscopy (TEM, JEM-2100, 200 kV). The mass of electrode materials was measured on a BS 124 S Balance (max: 120 g; d = 0.1 mg).

2.3 Battery assembly and measurements

Batteries were assembled in an argon-filled glove box. For Swagelok-type half cells, pure RTO or LTO–RTO nanowire arrays grown directly on carbon cloth were the working electrode and Li-metal circular foil (0.59 mm thick) was used as the counter and reference electrodes. The two electrodes were separated with polypropylene (PP) film as the separator. A 1 M solution of LiPF₆ in ethylene carbonate (EC) and diethyl carbonate (DEC) (1 : 1 by volume) was used as the electrolyte. The cell was aged for 8 h before measurement. For the laminated-type full cell, the LTO–RTO hybrid array was used as the anode (∼5 mg cm⁻² mass loading) and the commercial LiCoO₂ film on Al foil as the cathode. After soaking with LiPF₆ electrolyte and separation by PP film, the electrodes were finally wrapped with an Al plastic pouch. The discharge–charge testing was performed at room temperature by using a multichannel battery tester (Shenzhen Neware Technology Co., Ltd, China). EIS measurements were performed by applying an AC voltage with 5 mV amplitude in a frequency range from 0.01 Hz to100 kHz using a CS310 electrochemical workstation.

3. Results and discussion

The LTO–RTO nanowire array was grown by two steps and the structural details are illustrated in Fig. 1a. The RTO nanowire array was firstly synthesized on a carbon cloth current collector by a facile hydrothermal method. The reason we chose carbon cloth is because of its high conductivity, lightweight, high chemical stability, and robust flexibility, which makes it still applicable where flexible electrodes are required. After the reaction, the carbon fibers were uniformly covered by RTO nanowire array, which has a relatively smooth nanowire surface and has the nanowire diameter of ca. 40–70 nm (see ESI, Fig. S1†). The final LTO–RTO nanowire array was further attained by the controlled partial conversion of surface RTO into LTO, achieved via a solid state reaction between RTO and LiOH at ∼700 °C. Fig. 1b shows an optical photo of the optimized LTO–RTO hybrid array electrode, indicating that the entire surface of the carbon cloth was covered homogeneously with a white film. Fig. 1c–e display the typical SEM images of the film at different resolutions. After the phase transformation reaction, the nanowire array architecture is kept unchanged. However, the LTO–RTO nanowires have an apparently rougher surface, as compared to the pristine RTO nanowires. This feature is helpful as it may provide more active sites for lithium storage. The XRD pattern in Fig. 1f confirms the existence of both LTO and RTO. For LTO, the intensity ratio of the (400) to (111) planes is much larger than that in the standard pattern (JCPDS Card no. 49-0207), suggesting that the LTO has a preferred crystal orientation of (400). The (400) surface was reported to be the most energetically favorable for Li insertion.⁵³ A TEM image of the LTO–RTO hybrid nanowires is further shown in Fig. S2.† High-resolution TEM pictures are also displayed in Fig. 1g and h. There are generally two types of LTO (also see the structural illustration of a single hybrid nanowire in Fig. 1a): one is particulate-shaped with (111) plane observed (Fig. 1g), resulting from the complete transformation from RTO; the other type has a layered coating on the remaining RTO surface, forming a clear interface (Fig. 1h). In Fig. 1h, two interplanar spacings of 0.21 and 0.32 nm can be detected, corresponding to the (400) plane of LTO and (110) plane of RTO, respectively. Both the HRTEM image and the inserted fast Fourier transform (FFT) pattern confirm the high crystallinity of LTO and RTO in the hybrid nanowires.

Fig. 1 (a) Schematic diagram illustrating the structural details of the LTO–RTO nanowire array on carbon cloth. (b) Optical picture of the optimized array electrode. (c–e) SEM images of the LTO–RTO nanowire array at different resolutions. (f) XRD pattern of the nanowire array. High-resolution TEM images of particulate-shaped LTO (g) and layer-shaped LTO coating on RTO (h), and the corresponding FFT pattern.

The exact mechanism of the transformation from tetragonal RTO to cubic spinel LTO was not completely understood previously. Nevertheless, it is chemically different from that of the conversion from layered titanate to LTO.^13,40 The model transformation from rutile β-MnO₂ to spinel LiMn₂O₄ (ref. 54–56) is considered as the key in our case. Based on the crystal structures of RTO and LTO, Scheme 1 shows a schematic illustration of the proposed lithium intercalation process from RTO to LTO. Basically, primitive tetragonal packing opens up the interstitial 1D channels to give them a square cross-section in the basal plane (Scheme 1a) and thus optimizes the possibility of 1D Li⁺ mobility within a channel parallel to the z-axis. During the transformation process, the Li⁺ enters the channels, accompanied by the formation of LiO₄ tetrahedra, LiO₆ octahedra and new TiO₆ octahedra. This leads to the spinel structured LTO, which can be precisely described as [Li]_{8a 16c}[Ti]_16d[Li_1/3Ti_2/3]_16d⁻O₄, similar to the case of LiMn₂O₄ that can be written as [Li]_{8a 16c}[Mn⁴⁺]_16d[Mn³⁺]_16d⁻O₄. LTO structures viewed along the z and x axis are shown in Scheme 1b and c, respectively. In addition to the 8a site the Li⁺ also occupied the 16d⁻ site. With lithium intercalation into the 16d⁻ site, the old TiO₆ octahedron in RTO evolved into a new one in LTO. Nevertheless, the distance between the nearest Ti–Ti along the c-axis in crystal cell (0.2959 nm) only changes slightly to 0.2946 nm, complying with the theory of the minimal reorganization of structure.⁵⁴ This topotactic conversion process is believed to be the main reason for LTO–RTO reserving the initial nanowire morphology of RTO.

Scheme 1 Schematic illustration of the Li⁺ intercalation process from a tetragonal to cubic structure.

The mass ratio of LTO to RTO (m_LTO : m_RTO) in the hybrid nanowire array can be readily manipulated by adjusting the amount of LiOH. To find out the optimal component ratio for lithium storage, we performed half-cell tests with anodes of the pure RTO sample and samples with different m_LTO : m_RTO ratios. Herein, three typical ratios (1 : 1, 3 : 1 and 5 : 1) were chosen for discussion. Firstly, the electrochemical impedance spectra (EIS) for the four samples were investigated and the Nyquist plots are comparably depicted in Fig. 2a. All the plots consist of one semicircle at the high frequency region and a spike at the low frequency region. The equivalent circuit and key element values are provided in the ESI (Table S1†). The diameter of the semicircle is indicative of the charge-transfer resistance R_ct. Apparently, the R_ct increases with increasing the m_LTO : m_RTO. This phenomenon confirms the presence of RTO is beneficial to the electrical conductivity of our nanowire array electrode. However, the optimal amount of RTO should be determined by further considering the capacity and cycling stability. The charge–discharge voltage profiles and cycling performance of the samples at the current density of 350 mA g⁻¹ are displayed in Fig. 2b and c (see Table S2† for detailed data). Note that the carbon cloth has no contribution to the capacity within the potential range of 1.0–2.5 V vs. Li/Li⁺. The charge–discharge curve from the pure RTO electrode shows a slope profile of the potential–capacity relationship, similar to that of RTO reported previously.^56,57 Differently, the 1 : 1, 3 : 1 and 5 : 1 LTO–RTO array electrodes include a relatively flat potential plateau around 1.55 V, mainly characteristic of the two phase equilibrium between Li₄Ti₅O₁₂ and Li₇Ti₅O₁₂.^18,53 The length of the plateau is indicative of the LTO's contribution to the total capacity. It is apparent that the 5 : 1 LTO–RTO electrode exhibits a too large potential polarization (∼0.25 V) and a too small capacity (<100 mA h g⁻¹) although its plateau is obvious. For the pure RTO and the 1 : 1 LTO–RTO electrodes, although their capacities are relatively high, the Coulombic efficiency (Table S3†) and especially the cycling performance are inferior. By contrast, the 3 : 1 LTO–RTO electrode shows a high capacity, high cycling stability, high Coulombic efficiency, and moderate polarization and resistance, which are attributed to the optimized synergy between LTO (reversible Li⁺ insertion/desertion; “zero strain”, which renders LTO a protective function to restrict the volumetric change of inner RTO) and RTO (high capacity and conductivity). Thus, we used the 3 : 1 LTO–RTO electrode to conduct the following study.

Fig. 2 (a) Nyquist plots of different LTO–RTO electrode samples. (b) Charge–discharge voltage profiles and (c) cycle performances of the four samples at 350 mA g⁻¹ between 1.0 and 2.5 V (vs. Li/Li⁺).

Fig. 3a illustrates the first charge–discharge curve of the 3 : 1 LTO–RTO hybrid array electrode at 175 mA g⁻¹. It displays relatively flat plateaus which can be further evidenced by the corresponding differential capacity versus voltage plot in Fig. 3b. The two intense peaks in Fig. 3b correspond to the redox reactions between Li₄Ti₅O₁₂ and Li₇Ti₅O₁₂ associated with Li⁺, demonstrating good electrochemical reversibility. The first charge–discharge curves of the electrode at various current densities from 175 to 1750 mA g⁻¹ (Fig. 3c) all exhibit perfect plateau and high Coulombic efficiencies (95–100%). The cycling response at continuously variable currents was further evaluated and displayed in Fig. 3d. At the current density of 175 mA g⁻¹, our hybrid nanowire array electrode delivers an impressive capacity of 181 mA h g⁻¹, among the highest reported to date for LTO-based electrodes. The capacity slightly decreases to 165, 137 and 122 mA h g⁻¹ respectively at current densities of 350, 875 and 1750 mA g⁻¹. Nevertheless, it should be pointed out that, at the same current densities, our hybrid array electrode generally exhibits superior or comparable capacities to the reported state-of-the-art LTO electrodes, such as carbon-LTO porous microspheres,^27,58 LTO embedded in mesoporous carbon,^18,59 LTO nanowires,⁴⁰ and hollow LTO spheres,⁶⁰ etc. (see Table S4† for details). Fig. 3e presents the cycling performance of the LTO–RTO hybrid electrode up to 400 times at a high current density of 1750 mA g⁻¹ (∼10 C; charge–discharge within several minutes). It is observed that ∼83% of its initial capacity can be retained. The coulombic efficiency is maintained at almost 100% after the 10th cycle. A comparison of the charge–discharge profiles of the first and the 400th cycle in Fig. 3f clearly reveals that long-term cycling slightly increases the potential polarization, but does not deteriorate the flat plateau. An SEM image of the 400 times-cycled electrode (Fig. S3†) demonstrates that the nanowire array architecture and the surface morphology and diameter of LTO–RTO nanowire are almost preserved, indicating that 400 cycles is still far from its lifetime.

Fig. 3 The 3 : 1 LTO–RTO electrode: (a) the first charge–discharge curve at 175 mA g⁻¹. (b) The corresponding differential capacity versus voltage plot. (c) The first charge–discharge curves at different currents. (d) Cycling performance at continuously variable current densities. (e) Cycling performance at a constant current of 1750 mA g⁻¹. (f) The first and the 400th discharge–charge curves at 1750 mA g⁻¹.

In an effort to establish a more complete and practical energy storage platform, a 2.5 × 2.5 cm² full-cell LIB was constructed with the LTO–RTO hybrid array as an anode and the commercial LiCoO₂ film on Al foil as a cathode. Since the capacity of the used LiCoO₂ cathode (∼2 mA h cm⁻²) is higher than that of our anode (0.8–1 mA h cm⁻²) the cell is anode-limited and all the following reported data is based on the mass of the LTO–RTO array. The charge–discharge curves of the LCO/LTO–RTO full cell at different current densities are displayed in Fig. 4a. The potential plateau is around 2.3 V; flat plateaus can be observed over almost the entire capacity at every current density, which is very important for practical application.^61,62 Fig. 4b shows the good rate performance of the full cell. The specific capacity is as high as 180 mA h g⁻¹ at 45 mA g⁻¹. With the current density increased ∼40 times to 1750 mA g⁻¹, the capacity of 113 mA h g⁻¹ can still be delivered. Even suffering from a rapid change of the current density, the full cell exhibits a stable capacity at each current. When the current was turned back to 45 mA g⁻¹, ∼100% of the capacity is recovered. The excellent cycle performance and coulombic efficiency stability of the full cell at a current density of 450 mA g⁻¹ are subsequently illustrated in Fig. 4c. The capacity can still be retained at about 96%, after even 3000 cycles and the coulombic efficiency is maintained at almost 100%, and no gas evolution problem was observed. At a high current density of 1750 mA g⁻¹(∼10 C), a cycling stability of up to 2000 times was also demonstrated (Fig. S4†). The LCO/LTO–RTO full-cell performance is quite remarkable and demonstrates again the advantages of the carbon- and binder-free binary nanowire array architecture design. In Fig. S5, S6 and Table S5,† we further provide the cell capacities, cell energy densities and energy efficiencies for reference.

Fig. 4 The LCO/LTO–RTO full cell: (a) the first charge–discharge curves at different currents. (b) Rate performance at progressively increased current densities. (c) Cycling performance and coulombic efficiency variation at 450 mA g⁻¹. (d and e) The optical images showing the “CCNU” pattern and an 8 × 8 LED lattice board could be lit.

To investigate the potential of our full cell for practical applications, popular electronic devices such as Nixie tubes and LED lattice boards were tested. As demonstrated in Fig. 4d and e, a single cell could efficiently light four Nixie tubes to display the “CCNU” pattern and could also easily power a large 8 × 8 LED lattice board. The ∼2.3 V output voltage of the cell is suitable for many currently available portable electronics. We believe that the cell voltage could be further increased by coupling our hybrid array anode with alternative high-potential cathode materials.⁶³

4. Conclusions

In summary, a carbon-/binder-free LTO–RTO hybrid nanowire array electrode has been designed via partial topotactic conversion of an RTO nanowire array. The optimized 3 : 1 LTO–RTO electrode exhibits both high capacity and excellent cyclability. A lithium ion full cell assembled using an LTO–RTO hybrid array as the anode and commercial LCO film as the cathode further manifests outstanding performance in terms of high capacity, high rate and ultralong lifespan (capacity retention over 96% after 3000 cycles). Our work will provide new opportunities in the search of alternative high-performance LTO anodes. Furthermore, the concept of developing a synergistic dual-phase nanowire array architecture can be extended to other LIB electrode designs, especially in the case that a binder-free electrode is required.

Acknowledgements

This work was supported by grants from the National Natural Science Foundation of China (no. 51102105, 11104088), the Key Project of Natural Science Foundation of Hubei Province (no. 2013CFA023) and Self-determined Research Funds of CCNU from the Colleges' Basic Research and Operation of MOE (CCNU12A01009).

Notes and references

1. M. Armand and J. M. Tarascon, Nature, 2008, 451, 652–657.
2. Y. Wang and G. Z. Cao, Adv. Mater., 2008, 20, 2251–2269.
3. L. S. Li, F. Meng and S. Jin, Nano Lett., 2012, 12, 6030–6037.
4. C. Zhou, Y. W. Zhang, Y. Y. Li and J. P. Liu, Nano Lett., 2013, 13, 2078–2085.
5. H. S. Zhou, Energy Environ. Sci., 2013, 6, 2256.
6. C. K. Chan, H. Peng, G. Liu, K. McIlwrath, X. F. Zhang, R. A. Huggins and Y. Cui, Nat. Nanotechnol., 2008, 3, 31–35.
7. Y. L. Liu, Y. H. Xu, X. G. Han, C. Pellegrinelli, Y. J. Zhu, H. L. Zhu, J. Y. Wan, A. C. Chung, O. Vaaland, C. S. Wang and L. B. Hu, Nano Lett., 2012, 12, 5664–5668.
8. H. L. Pan, Y.-S. Hu and L. Q. Chen, Energy Environ. Sci., 2013, 6, 2338–2360.
9. X. H. Cao, Y. M. Shi, W. H. Shi, X. H. Rui, J. Kong, Q. Y. Yan and H. Zhang, Small, 2013, 9, 1703–1707.
10. H. X. Ji, L. L. Zhang, M. T. Pettes, H. F. Li, S. S. Chen, L. Shi, R. Piner and R. S. Ruoff, Nano Lett., 2012, 12, 2446–2451.
11. H. G. Wang, D. L. Ma, X. L. Huang, Y. Huang and X. B. Zhang, Sci. Rep., 2012, 2, 701.
12. K. T. Nam, D. W. Kim, P. J. Yoo, C. Y. Chiang, N. Meethong, P. T. Hammond, Y. M. Chiang and A. M. Belcher, Science, 2006, 312, 885–888.
13. J. Kim and J. Cho, Electrochem. Solid-State Lett., 2007, 10, A81–A84.
14. J. W. Deng, C. L. Yan, L. C. Yang, S. Baunack, S. Oswald, H. Wendrock, Y. F. Mei and O. G. Schmidt, ACS Nano, 2013, 7, 6948–6954.
15. (a) D. N. Wang, J. L. Yang, X. F. Li, D. S. Geng, R. Y. Li, M. Cai, T. K. Sham and X. L. Sun, Energy Environ. Sci., 2013, 6, 2900–2906; (b) W. Tang, Y. S. Zhu, Y. Y. Hou, L. L. Liu, Y. P. Wu, K. P. Loh, H. P. Zhang and K. Zhu, Energy Environ. Sci., 2013, 6, 2093–2104.
16. (a) H. N. Song, S. W. Yun, H. H. Chun, M. G. Kim, K. Y. Chung, H. S. Kim, B. W. Cho and Y. T. Kim, Energy Environ. Sci., 2012, 5, 9903–9913; (b) B. H. Li, C. P. Han, Y. B. He, C. Yang, H. D. Du, Q. H. Yang and F. Y. Kang, Energy Environ. Sci., 2012, 5, 9595–9602.
17. (a) Y. Sun, L. Zhao, H. L. Pan, X. Lu, L. Gu, Y. S. Hu, H. Li, M. Armand, Y. C. Lkuhara, L. Q. Chen and X. J. Huang, Nat. Commun., 2013, 4, 1870–1880; (b) Y. S. Hu, L. Kienle, Y. G. Guo and J. Maier, Adv. Mater., 2006, 18, 1421–1426; (c) Z. J. Ding, L. Zhao, L. M. Suo, Y. Jiao, S. Meng, Y. S. Hu, Z. X. Wang and L. Q. Chen, Phys. Chem. Chem. Phys., 2011, 13, 15127–15133; (d) H. L. Pan, L. Zhao, Y. S. Hu, H. Li and L. Q. Chen, ChemSusChem, 2012, 5, 526–529; (e) X. Lu, L. Zhao, X. He, R. Xiao, L. Gu, Y. S. Hu, H. Li, Z. Wang, X. Duan, L. Q. Chen, J. Maier and Y. Ikuhara, Adv. Mater., 2012, 24, 3233–3238.
18. E. Kang, Y. S. Jung, G. H. Kim, J. Y. Chun, U. Wiesner, A. C. Dillon, J. K. Kim and J. Lee, Adv. Funct. Mater., 2011, 21, 4349–4357.
19. Z. G. Yang, D. Choi, S. Kerisit, K. M. Rosso, D. H. Wang, J. Zhang, G. Graff and J. Liu, J. Power Sources, 2009, 129, 588–598.
20. N. Li, Z. P. Chen, W. C. Ren, F. Li and H. M. Cheng, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 17360–17365.
21. K. Amine, I. Belharouak, Z. Chen, T. Tran, H. Yumoto, N. Ota, S. T. Myung and Y. K. Sun, Adv. Mater., 2010, 22, 3052–3057.
22. H. G. Jung, M. W. Jang, J. Hassoun, Y. K. Sun and B. Scrosati, Nat. Commun., 2011, 2, 516.
23. L. F. Shen, H. S. Li, E. Uchaker, X. G. Zhang and G. Z. Cao, Nano Lett., 2012, 12, 5673–5678.
24. X. W. Zhang, Y. Li, S. A. Khan and P. S. Fedkiw, J. Electrochem. Soc., 2004, 151, A1257–A1263.
25. X. G. Hao and B. M. Bartlett, Adv. Energy Mater., 2013, 3, 753–761.
26. M. Wagemaker, D. R. Simom, E. M. Kelder, J. Schoonman, C. Ringpfeil, U. Haake, D. Lützenkirchen-HeCht, R. Frahm and F. M. Mulder, Adv. Mater., 2006, 18, 3169–3173.
27. L. Zhao, Y. S. Hu, H. Li, Z. X. Wang and L. Q. Chen, Adv. Mater., 2011, 23, 1385–1388.
28. G. N. Zhu, H. J. Liu, J. H. Zhuang, C. X. Wang, Y. G. Wang and Y. Y. Xia, Energy Environ. Sci., 2011, 4, 4016–4022.
29. G. Du, N. Sharma, V. K. Peterson, J. A. Kimpton, D. Jia and Z. P. Guo, Adv. Funct. Mater., 2011, 21, 3990–3997.
30. L. F. Shen, C. Z. Yuan, H. J. Luo, X. G. Zhang, L. Chen and H. S. Li, J. Mater. Chem., 2011, 21, 14414–14416.
31. Y. Q. Wang, L. Gu, Y. G. Guo, H. Li, X. Q. He, S. Tsukimoto, Y. Ikuhara and L. J. Wan, J. Am. Chem. Soc., 2012, 134, 7874–7879.
32. B. B. Tian, H. F. Xiang, L. Zhang and H. H. Wang, J. Solid State Electrochem., 2012, 16, 205–211.
33. S. Huang, Z. Wen, X. Zhu and Z. Gu, Electrochem. Commun., 2004, 6, 1093–1097.
34. K. S. Park, A. Benayad, D. J. Kang and S. G. Doo, J. Am. Chem. Soc., 2008, 130, 14930–14931.
35. L. F. Shen, C. Z. Yuan, H. J. Luo, X. G. Zhang, K. Xu and F. Zhang, J. Mater. Chem., 2011, 21, 761–767.
36. L. F. Shen, C. Z. Yuan, H. J. Luo, X. G. Zhang, S. D. Yang and X. J. Lu, Nanoscale, 2011, 3, 572–574.
37. X. Li, C. Lai, C. W. Xiao and X. P. Gao, Electrochim. Acta, 2011, 56, 9152–9158.
38. M. M. Rahman, J. Z. Wang, M. F. Hassan, D. Wexler and H. K. Liu, Adv. Energy Mater., 2011, 1, 212–220.
39. J. Wang, H. L. Zhao, Q. Yang, C. M. Wang, P. P. Lv and Q. Xia, J. Power Sources, 2013, 222, 196–201.
40. L. F. Shen, E. Uchaker, X. G. Zhang and G. Z. Cao, Adv. Mater., 2012, 24, 6502–6506.
41. M. S. Song, A. Benayad, Y. M. Choi and K. S. Park, Chem. Commun., 2012, 48, 516.
42. C. J. Kim, N. S. Norberg, C. T. Alexander, R. Kostecki and J. Cabana, Adv. Funct. Mater., 2013, 23, 1214–1222.
43. O. Byl and J. T. Yates, J. Phys. Chem. B, 2006, 110, 22966–22967.
44. H. J. Kim, M. G. Kim and J. Cho, Adv. Energy Mater., 2012, 2, 1425–1432.
45. J. Jiang, Y. Y. Li, J. P. Liu, X. T. Huang, C. Z. Yuan and X. W. Lou, Adv. Mater., 2012, 24, 5166–5180.
46. J. P. Liu, J. Jiang, C. W. Cheng, H. X. Li, J. X. Zhang, H. Gong and H. J. Fan, Adv. Mater., 2011, 23, 2076–2081.
47. J. Jiang, Y. Y. Li, J. P. Liu and X. T. Huang, Nanoscale, 2011, 3, 45–58.
48. C. Q. Shang, S. M. Dong, S. Wang, D. D. Xiao, P. X. Han, X. G. Wang, L. Gu and G. L. Cui, ACS Nano, 2013, 7, 5430–5436.
49. B. Liu, J. Zhang, X. F. Wang, G. Chen, D. Chen, C. W. Zhou and G. Z. Shen, Nano Lett., 2012, 12, 3005–3011.
50. Y. G. Li, B. Tan and Y. Y. Wu, Nano Lett., 2008, 8, 265–270.
51. S. R. Gowda, A. L. Reddy, X. B. Zhan and P. M. Ajayan, Nano Lett., 2011, 11, 3329–3333.
52. X. J. Feng, K. Zhu, A. J. Frank, C. A. Grimes and T. E. Mallouk, Angew. Chem., Int. Ed., 2012, 51, 2727–2730.
53. S. Ganapathy and M. Wagemaker, ACS Nano, 2012, 6, 8702–8712.
54. B. Gates, B. B. Mayers, Y. Wu, Y. Sun, B. Cattle, P. Yang and Y. Xia, Adv. Funct. Mater., 2002, 12, 679–686.
55. W. I. F. David, M. M. Thackeray, P. G. Bruce and J. B. Goodenough, Mater. Res. Bull., 1984, 19, 99–106.
56. Y. L. Ding, J. Xie, G. S. Cao, T. J. Zhu, H. M. Yu and X. B. Zhao, Adv. Funct. Mater., 2011, 21, 348–355.
57. D. H. Wang, D. W. Choi, Z. G. Yang, V. V. Viswannathan, Z. M. Nie, C. M. Wang, Y. J. Song, J. G. Zhang and J. Liu, Chem. Mater., 2008, 20, 3435–3442.
58. H. G. Jung, S. T. Myung, C. S. Yoon, S. B. Son, K. H. Oh, K. Amine, B. Scrosatiand and Y. K. Sun, Energy Environ. Sci., 2011, 4, 1345–1351.
59. L. F. Shen, X. G. Zhang, E. Uchaker, C. Z. Yuan and G. Z. Cao, Adv. Energy Mater., 2012, 2, 691–698.
60. L. Yu, H. B. Wu and X. W. Lou, Adv. Mater., 2013, 25, 2296–2300.
61. L. El. Ouatani, R. Dedryvère, C. Siret, P. Biensan and D. Gonbeau, J. Electrochem. Soc., 2009, 156, A468–A477.
62. J. H. Kim, S. Y. Bae, J. H. Min, S. W. Song and D. W. Kim, Electrochim. Acta, 2012, 78, 11–16.
63. J. Kim, J. K. Yoo, Y. S. Jung and K. Kang, Adv. Energy Mater., 2013, 3, 1004–1007.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra00111g

---