Solvothermal synthesis and self-assembling mechanism of micro-nano spherical LiFePO₄ with high tap density

Abstract

Although LiFePO₄ has been widely studied and also used as a promising cathode material for Li ion batteries, its inferior tap density is still a big challenge for practical application. Well-defined three-dimensional porous LiFePO₄ microspheres composed of nanosheets were successfully synthesized by a simple one-step solvothermal method. The porous spherical morphology could not only retain the excellent electrochemical performance characteristics of the LiFePO₄ nanosheet but also meet the requirements of high tap density of the powder particles, leading to highly improved volumetric energy density. These micro-nano structured LiFePO₄ microspheres have a high tap density of about 1.4 g cm⁻³. A growth mechanism is also proposed based on time-dependent experiments. This work provides an efficient route for designing a desirable micro–nano structure, which could also be extended to synthesize other hierarchical structures used in different fields.

---

1. Introduction

Lithium iron phosphate (LiFePO₄) has been considered as a main commercial cathode material for Li ion batteries owing to its prominent advantages in production cost,^1,2 service life and safety performance. However, its wide application has always been limited by the bottlenecks of the poor charge-transfer ability (intrinsic conductivity of ca. 10⁻⁹ S m⁻¹, Li⁺ ion diffusion coefficient of ca. 10⁻¹⁴ to 10⁻¹⁶ cm² s⁻¹) and low tap density (generally, ≤1.2 g cm⁻³), resulting in poor rate performance and low volume specific capacity, respectively.^3–5 Several strategies have been developed to effectively overcome these obstacles, e.g., carbon coating or ion doping to improve the electronic conductivity, reducing the LiFePO₄ particle size to enhance the Li⁺ ion diffusivity in LiFePO₄ lattice.^6,7 Nanosized LiFePO₄ with various morphologies, such as nanoparticles, nanowires, and nanosheets have been prepared by various synthesis methods, including hydrothermal/solvothermal, sol–gel, and hard-templating approaches.^8–10 Among these morphologies, nanosheets are regarded as a promising approach to improve the rate performance of LiFePO₄.^11,12 Tang et al. reported LiFePO₄/C nanosheets with a high reversible capacity (164 mA h g⁻¹ at 0.1C) and good cycle stability (capacity retention of near 100% after 100 cycles).¹³ Unfortunately, it exhibited a very low tap density (∼0.76 g cm⁻³) due to the disorderly bridging effect as shown in Fig. S1a.† According to the relevant standards, the tap density of the commercial LiFePO₄ cathode material should be in the range of 1.0–1.2 g cm⁻³ to ensure a suitable volume specific capacity. Therefore, it is essential to assemble the LiFePO₄/C nanosheets from the disorderly and incompact bridging state into an orderly and closely packed state.^12,14,15 A hierarchical micro–nano spherical structure could ensure a high rate capacity and guarantee a high tap density as well.^15–17

Recently, self-assembled LiFePO₄ nanostructures with micro–nano spherical or round-likely morphologies prepared by the solvothermal method, such as the flower-like LiFePO₄/C microspheres (Fig. S1c†) and flower-like LiMnPO₄/C hierarchical microstructures assembled from single-crystalline nanosheets have attracted much attention (Fig. S1d†).^18,19 Cage-like LiFePO₄/C microspheres exhibited a highly reversible capacity and good cycling stability (160 mA h g⁻¹ at 0.1C over 300 cycles) as well as good rate capability (120 mA h g⁻¹ at 10C).^20,21 Sun and his co-workers reported a high tap density LiFePO₄ microsphere (1.2 g cm⁻³) with excellent rate capability and cycle stability, which consisted of nanoplates or nanoparticles with an open three-dimensional (3D) porous microstructure.^22–24 Although these self-assembled hierarchical nanostructures can change the disorderly and incompact bridging state of the primary LiFePO₄/C nanosheets, the regularity and monodispersity of the secondary microspheres still need to be further improved.^25,26 More importantly, exploring the self-assembled mechanism is of great importance since the morphology of the self-assembled LiFePO₄ nanostructures is significantly influenced by the solvothermal reactive conditions, such as reaction time, temperature, solvent, additives, etc.^27–29

In our group, we have comprehensively investigated the influence of the solvothermal reactive conditions on the self-assembly process, and thus obtained various LiFePO₄ nanostructures. Among these nanostructures, micro–nano spherical LiFePO₄ with high regularity, uniformity and monodispersity is more interesting, whose tap density was measured to be as high as 1.4 g cm⁻³ (Fig. S1b†). According to the relevance between the solvothermal reactive conditions and the morphology of the LiFePO₄ nanostructures, we also proposed a self-assembly mechanism involving the synergistic effect of Ostwald ripening, crystal inhibiting and orientated assembly process for the formation of porous LiFePO₄ microspheres composed of nanosheets.

2. Experimental section

2.1 Synthesis of LiFePO₄ microspheres

A monodisperse LiFePO₄ microsphere was prepared by a glycol-based solvothermal process combined with carbothermal reduction. All the reagents were of analytical grade and used without further purification. Under typical synthesis conditions, LiH₂PO₄ (10 mmol), FeSO₄·7H₂O (10 mmol) and citric acid (0.5 mmol) were dissolved in 40 mL of glycol in a beaker under inert gas protection by continuous magnetic stirring at room temperature. When the mixture was completely dissolved, 8 mmol urea was added and a celandine green solution was obtained. The autoclave was sealed and then kept at 200 °C for 24 h. After the autoclave was cooled down to room temperature, the precipitate was filtered and washed with distilled water and directly dried in a vacuum at 80 °C for 10 h to obtain a blackgreen precursor. The obtained product was calcined at 600 °C under a flowing Ar/H₂ atmosphere (100 mL min⁻¹, 95 vol% Ar and 5 vol% H₂) for 10 h.

To investigate the formation process of the microsphere LiFePO₄, time-dependent experiments were conducted in detail. The experimental parameters, such as solvent and different kinds of organic additives (urea, citric acid) were also varied during the synthesis to study their effects on the morphology of the final product.

2.2 Materials characterization

The crystal structure of the samples was investigated by X-ray diffraction (XRD) on a RigaKu D/max2550VB⁺ 18 kW using graphite-monochromatized CuKα radiation (40 kV, 250 mA) with a scan range from 10° to 80°. The morphology and microstructure of the products were investigated using field-emission scanning electron microscopy (FE-SEM, Sirion 200) and high-resolution transmission electron microscopy (HRTEM, Tecnai G2 20 s) with an energy dispersive X-ray analysis (EDX) IXRF EDS-2000 System. The Raman spectrum was obtained from a Labram-010 laser Raman spectrometer.

2.3 Electrochemical measurements

The electrochemical performance of the as-prepared crystalline LiFePO₄/C composite was evaluated using CR2016 coin cells during 2.5–4.2 V with a Land 2001 battery test system. The coin cells were fabricated with a LiFePO₄ cathode, metallic lithium foil anode, 1 M LiPF₆ in ethylene carbonate (EC)–dimethyl carbonate (DMC)–diethyl carbonate (DEC) (volume ratio of 1 : 1 : 1) electrolyte, and a Celgard polypropylene separator. The cathode was prepared by mixing LiFePO₄, acetylene black and polytetrafluorethylene (PTFE) binder with weight ratio of 8 : 1 : 1, the mixture was rolled into thin sheets and then cut into circular electrodes of 0.64 cm². The typical electrode contained active materials of 3–5 mg, and was dried in a vacuum at 120 °C for 12 h before being assembled into coin cells in an argon-filled glovebox. The assembled cells were galvanostatically charged and discharged at various rates from 0.1C to 10C (1C = 170 mA g⁻¹). Cyclic voltammetry (CV) curves were recorded at a scan rate of 0.5 mV s⁻¹. Electrochemical impedance spectroscopy (EIS) was plotted for frequencies from 0.01 Hz to 100 kHz using Zahner elektrik (IM6ex).

3. Results and discussion

Poor electronic conductivity (∼10⁻⁹ S cm⁻¹) and low Li-ion diffusion coefficient (∼1.8 × 10⁻¹⁴ cm² s⁻¹) are the two big issues for LiFePO₄ which can be addressed by the following two methods.^30,31 One is coating the LiFePO₄ particles with an electron-conducting layer or doping with supervalent ions.³² The other is to shorten the Li⁺ diffusion distance by minimizing the particle size.³³ However, a down-sizing strategy adversely affects the tap density and volumetric energy density.^34,35 In our previous work, LiFePO₄ with various interesting shapes have been successfully synthesized. Herein, we report a novel strategy to synthesize LiFePO₄ microspheres with high regularity, uniformity and monodispersity via a solvothermal approach.

The chemical composition and phase purity of the products were analyzed by X-ray diffraction. The XRD pattern of LiFePO₄ microspheres is presented in Fig. 1. All the identified diffraction peaks can be unambiguously assigned to the phase-pure orthorhombic structure of LiFePO₄ (JCPDS 81-1173).^27,28 These sharp diffraction peaks indicate good crystallization of the as-synthesized samples.²⁹

Fig. 1 XRD pattern of the as-prepared LiFePO₄.

Furthermore, the compositions of LiFePO₄ microspheres were confirmed by EDS analysis. As shown in Fig. S2,† Fe, P and O elements can be clearly observed, and there are no other impurity elements except C which should be derived from the decomposition of organic compounds.

Fig. 2 shows the microstructure of the samples. As seen in Fig. 2a, uniform and monodisperse microsphere morphology assembled by nanosheets appears, whose average diameter is about 15 μm. Note that the LiFePO₄ microspheres with unique morphology here promise an ultrahigh powder tap density of 1.4 g cm⁻³, higher than those reported values in references.^22–25 Note that the high tap density of LiFePO₄ is a key parameter for its practical application. Good particle size distribution is also proved in Fig. S3.† A closer look at these particles is demonstrated in Fig. 2b. Interestingly, LiFePO₄ microspheres are composed of nanosheets with a thickness of several tens of nanometres. These nanosheets interweave together to form a microsphere with a three-dimensional open porous structure. Fig. 2c shows the surface structure of the individual microspheres. Such ordered porous structure is expected to facilitate electrolyte penetration into the electrode particles, thus providing more interface area between the electrode material and the electrolyte. Meanwhile, this unique microsphere structure is very stable such that it cannot be easily destroyed or be dispersed into nanoflakes under the ultrasound conditions or during the Li ion insertion/extraction process.

Fig. 2 Microstructure of the as-prepared LiFePO₄: (a, b) FE-SEM images; (c) the surface structure of an individual microsphere; (d) HRTEM images. The inset shows the SAED spot pattern of a single nanosheet.

The morphology of LiFePO₄ microsphere was further characterized by high-resolution TEM. Fig. 2d shows a very thin carbon layer with a thickness of several nanometres on the surface of the microsphere LiFePO₄ and this could also be proved by the following Raman spectroscopy result. The thin carbon layer should be derived from carbonization of citric acid attached to the surface of LiFePO₄ during the calcining process. The clear lattice fringes indicate the high crystallinity of the LiFePO₄. Furthermore, the clear and regular diffraction spot array in the corresponding SAED pattern obtained from the individual nanosheet (inset of Fig. 2d) demonstrates the single-crystal nature of the original nanosheets.^11,12

Raman spectroscopy was performed to analyze the phase composition of LiFePO₄ (Fig. 3). Clearly, the sample displays two broad peaks at about 1310 cm⁻¹ and 1583 cm⁻¹, which are assigned to the disordered carbon (D-band) and the graphite carbon (G-band), respectively.^28,29 The Raman result further confirms the existence of a thin carbon layer on the LiFePO₄.

Fig. 3 Raman spectrum of the as-prepared LiFePO₄ sample.

The nitrogen adsorption–desorption isotherms and the corresponding Barrett–Joyner–Halenda (BJH) pore size distribution curve of LiFePO₄ microspheres are shown in Fig. 4. The adsorption isotherm of LiFePO₄ microspheres shows a type-IV adsorption–desorption isotherm with H3-type hysteresis, a feature of mesoporous material. The measured Brunauer–Emmett–Teller (BET) area is 11.8 m² g⁻¹. And the average pore diameter is 3.3 nm calculated from the desorption branch of the nitrogen isotherm using the BJH method. The corresponding BJH desorption cumulative pore volume is 0.045 cm³ g⁻¹ which is relatively low for the electrolyte penetration into the electrode materials via the porous structure.

Fig. 4 The nitrogen adsorption–desorption isotherms and the BJH pore size distribution curves of the spherical LiFePO₄.

4. Growth mechanism of the spherical LiFePO₄

To investigate the formation mechanism of three-dimensional hierarchical LiFePO₄ microspheres, samples prepared after various reaction times were collected and investigated by XRD and FE-SEM. As seen in Fig. S4,† no crystalline LiFePO₄ phase appears within a short reaction time (e.g., 1 h and 2 h). After 4 h, the diffraction peaks of the samples at 25.6° and 29.7° and 35.6° (2θ) are indexed to the (201), (020) and (311), revealing the formation of the LiFePO₄ phase. Note that the crystallinity of the sample is inferior since the diffraction peak intensity is not high.^36,37 As the reaction proceeded (24 h), the crystallinity of the sample increased significantly. Fig. 5 shows the corresponding SEM images which clearly illustrate the morphology evolution process. In the initial stage, the product consisted of tiny particles. When the reaction time prolonged to 4 h, the amorphous precursor underwent a dissolving-precipitation process, and most were transformed to nanosheets with the thickness of about 20 nm. Also a gap between adjacent nanosheets was observed. As follows, saddle-shaped structures were observed through the piling up of nanosheets driven by the further reducing of the surface tension. However, these saddle-shaped structures were only the intermediates. Round-like spheres by the orientated assembling of the nanosheets were observed after 16 h. Finally, these microstructures continuously grew into monodispersed microspheres with a diameter of about 15 μm through the piling up of nanosheets to reduce their surface energy. Furthermore, Fig. 6 shows the effect of reaction temperature on the morphology of the self-assembled LiFePO₄. As the hydrothermal reaction temperature increased, most of the saddle-shaped particles grew into roundish structures.

Fig. 5 SEM images of samples after different heating time in the solvothermal process: (a) 1 h, (b) 2 h, (c) 4 h, (d) 8 h, (e) 16 h, (f) 24 h.

Fig. 6 SEM images of samples obtained after hydrothermal reaction for: (a) 140 °C, (b) 160 °C, (c) 180 °C.

In this work, one of the challenges was the controllable synthesis of the nano-micro spherical morphology. It was reported that glycol can not only act as a solvent in the process, but also as a stabilizer to limit particle growth and prohibit agglomeration.^38–40 Based on the results in references, we believe that glycol plays an important role in this process. When we changed the length of the carbon chain and the number of the hydroxyl functional groups of alcohol solvents by using propylene glycol, isopropyl alcohol, 1,2-propylene glycol or deionized water to replace the glycol, the products were different and even no LiFePO₄ formed in some cases (Fig. S6 and S7†). The orientated-chaining effect of the hydroxyl oxygen in the ethylene glycol was also investigated by varying the water content added to the solvent in Fig. 7. It is demonstrated that adding small amount of water into solvent resulted in slight particle morphology changes with some fragments appearing during the synthesis, and the crystallinity of the products was also decreased. We speculate that the preemption of the orientated-chaining site of ethylene glycol hydroxyl oxygen and Fe²⁺ may be taken over by the hydroxyl oxygen in deionized water thus the orientated-chaining effect of glycol could be cut off and the cambers could not be orientated assembly. Therefore, no perfect micro-nano spherical LiFePO₄ could be obtained.

Fig. 7 (a) and (b) Schematic illustration of the influence of different water content on the morphology of the LiFePO₄ and the corresponding XRD results (specific SEM images are shown in Fig. S4†).

On the basis of the investigations described above and the experimental results, a possible camber-orientated assembly mechanism was proposed for the micro-nano spherical LiFePO₄ (Fig. 8). Fig. 9 shows the SEM images of samples with different kinds of organic additives. As is well-known, citric acid is a good biological ligand for metal ions and it possesses three carboxylic acid groups and one hydroxyl functional group which can provide chelating ability.^35,36 In the initial stage, crystal nuclei were separated out from the precursor solution. A uniform citric acid membrane formed on the surface of the micrinite and the formation of acidic system increased the solubility of the precursor due to the addition of citric acid, which was a favourable factor for the Ostwald ripening process.^37,38 Thus, the rapid growth of the precursor micrinite was limited. Meanwhile, the aggregation of the micrinites was also restricted by the chelation steric effect of citric acid molecules, resulting in the formation of little flakes with neat morphology due to the reduced surface energy. For comparison, the product with large blocked morphology was observed without the addition of citric acid, which is not in favour of the orientated assembly process.

Fig. 8 Schematic illustration of the formation of the spherical LiFePO₄.

Fig. 9 SEM images of samples with different kinds of organic additives: (a) without any organic additive; (b) only citric acid; (c) the intermediate of the final micro-nano LiFePO₄; (d) citric acid and urea together.

Crystal plane orientation and crystal growth have an important relationship. Here we consider that the N atoms of urea have an important effect on the surface energy of the Fe₃(PO₄)₂ crystal, resulting in the change of the direction of crystal growth.^35,36 The lateral growth of the lamellar structure was better inhibited, and then the nanosheet was formed. The thinner nanosheet was named camber. According to our analysis results, two features of nanosheets' orientated assembly into the spheres are summarized. First, the nanosheet is very thin and the surface energy is easily reduced. Second, the nanosheet has a certain radian, thus it can be easily deformed during the camber-orientated assembly process.

The reduced surface free energy of the system provides a driving force of the camber-orientated assembly process. In this work, the formation of micro-nano spherical LiFePO₄ may be attributed to the strong synergistic effect between urea and glycol. Glycol reacted with Fe²⁺ ions to form chelation, making the camber-orientated assembly take place along the cambered surface (axis is perpendicular to the initial base of nanosheet) thus the surface energy of the system was significantly decreased. If a certain proportion of distilled water was added into the reaction system, a monodisperse and non-orientated assembling camber appeared. It is believed that a water film covers the surface of the camber, leading to the failure of the orientated-chaining effect of glycol. As the orientated assembly continued, a saddle-shaped structure was formed. At the last stage, the Ostwald ripening process is expected to proceed in the active site (the red area) of the saddle-shaped structure due to the existence of citric acid. The smaller camber dissolved to form a larger camber and perfect micro-nano spherical LiFePO₄ was thus obtained. In this system, the camber was distorted oriented along the axis direction due to the citric acid. This is a critical factor for the morphology change from round-like spherical to perfect spherical morphology. This distorting effect further drew the camber over to the axis, and a relatively dark suture on the particle surface was formed (the part of the red oval marked). Fig. S5† shows the TEM images of the intermediate of the saddle-shaped structure. It is clear that the colour on both sides is deeper than the middle part, which proves that the camber was assembled along the direction of the vertical axis and played a basic role in the appearance of the suture in the ultimate structure. According to the proposed growth mechanism, we can tune the morphology and the porosity of the LiFePO₄ materials in our future work by adjusting the particle size and the space between the adjacent cambers to achieve the best electrochemical properties.

5. Electrochemical properties

As cathode material for Li ion batteries, the electrochemical properties of LiFePO₄ microspheres without an additional carbon coating were investigated. Fig. 10a shows the discharge curves at various rates. As seen the pristine LiFePO₄ composite delivers the discharge capacities of 132.5 mA h g⁻¹, 108.6 mA h g⁻¹, 81.0 mA h g⁻¹ and 56.5 mA h g⁻¹, at 0.1C, 0.2C, 0.5C and 1C, respectively. It is well-known that the pristine LiFePO₄ generally exhibits very poor electrochemical properties due to its ultralow electronic conductivity (∼10⁻⁹ S cm⁻¹) and Li-ion diffusion coefficient (∼1.8 × 10⁻¹⁴ cm² s⁻¹).³⁰ In this work, although the rate capability of the as-prepared LiFePO₄ was not very good, the initial reversible capacity at low current density reached was high (132.5 mA h g⁻¹), which is due to the thermal decomposition of residual carbon source on the sample. This was proved by the HRTEM and Raman results. CV curves are demonstrated in Fig. 9b. The anodic oxidation and cathodic reduction for LiFePO₄ appear at 3.65 V and 3.25 V, respectively, with a potential interval of 0.30 V. Fig. 11 shows the discharge curves of the LiFePO₄ composite obtained under different conditions. Without any organic additive, the product delivers poor discharge capacities. With only citric acid added, the obtained LiFePO₄ nanoparticles deliver an initial discharge capacity of 142 mA h g⁻¹ at 0.1C. These phenomena can be mainly attributed to the increased active surface area of LiFePO₄ for the electrochemical reaction resulting from the gradual penetration of electrolyte into the interior of the particles.

Fig. 10 (a) Discharge curves of the as-prepared LiFePO₄ composite at different current densities in the potential between 2.5 and 4.2 V; (b) cycling performance of LiFePO₄.

Fig. 11 Discharge curves of the LiFePO₄ composite obtained under different conditions.

This work demonstrates a facile strategy to achieve a high-tap density microsphere LiFePO₄ consisting of uniform nanosheets, which exhibited relatively high reversible capacity. Carbon coating has been considered a promising strategy to further improve the electrochemical properties for LiFePO₄. In this work, we paid more attention to the synthesis of micro-nano spherical LiFePO₄ and the self-assembly mechanism. And the micro-nano spherical LiFePO₄ with a suitable carbon coating for further performance improvement is under way in our group.

6. Conclusion

Well-defined 3D hierarchical LiFePO₄ microspheres were successfully synthesized by a simple one-step solvothermal method. The LiFePO₄ microspheres were composed of nanosheets which were self-assembly via the Ostwald ripening-Crystal inhibiting-Orientated assembly process under the comprehensive influence of glycol and the organic additives of citric acid and urea. Owing to the closely stacked and unique self-assembly, microspherical LiFePO₄ has a quite high tap density of 1.4 g cm⁻³, which is a key parameter for commercial application. A camber-orientated assembly mechanism was proposed based on the investigations described above and the experimental results. Moreover, the presented strategy here opens a new avenue to synthesize various other electrode materials with a high tap density.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant No. 21276286 and No. 21271187) and the Open-End Fund for Valuable and Precision Instruments of Central South University (CSUZC201622).

References

1. A. K. Padhi, K. Nanjundaswamy and J. Goodenough, J. Electrochem. Soc., 1997, 144, 1188–1194.
2. J.-M. Tarascon and M. Armand, Nature, 2001, 414, 359–367.
3. J. Zhang, Y. Wang, Z. Lin and F. Huang, Cryst. Growth Des., 2010, 10, 4285–4291.
4. C. Sun, S. Rajasekhara, J. B. Goodenough and F. Zhou, J. Am. Chem. Soc., 2011, 133, 2132–2135.
5. M. Armand and J.-M. Tarascon, Nature, 2008, 451, 652–657.
6. M. S. Whittingham, Chem. Rev., 2004, 104, 4271–4302.
7. B. Ellis, W. Makahnouk, Y. Makimura, K. Toghill and L. Nazar, Nat. Mater., 2007, 6, 749–753.
8. D. Morgan, A. Van der Ven and G. Ceder, Electrochem. Solid-State Lett., 2004, 7, A30–A32.
9. M. Thackeray, Nat. Mater., 2002, 1, 81–82.
10. A. Padhi, K. Nanjundaswamy, C. Masquelier, S. Okada and J. Goodenough, J. Electrochem. Soc., 1997, 144, 1609–1613.
11. A. Churikov, A. Ivanishchev, I. Ivanishcheva, V. Sycheva, N. Khasanova and E. Antipov, Electrochim. Acta, 2010, 55, 2939–2950.
12. V. Srinivasan and J. Newman, Electrochem. Solid-State Lett., 2006, 9, A110–A114.
13. L. Li, X. Tang, H. Liu, Y. Qu and Z. Lu, Electrochim. Acta, 2010, 56, 995–999.
14. L. Wang, X. He, W. Sun, J. Wang, Y. Li and S. Fan, Nano Lett., 2012, 12, 5632–5636.
15. S.-Y. Chung, J. T. Bloking and Y.-M. Chiang, Nat. Mater., 2002, 1, 123–128.
16. C. Delacourt, L. Laffont, R. Bouchet, C. Wurm, J.-B. Leriche, M. Morcrette, J.-M. Tarascon and C. Masquelier, J. Electrochem. Soc., 2005, 152, A913–A921.
17. N. Ravet, Y. Chouinard, J. F. Magnan, S. Besner, M. Gauthier and M. Armand, J. Power Sources, 2001, 97, 503–507.
18. B. Kang and G. Ceder, Nature, 2009, 458, 190–193.
19. Y. J. Lee, H. Yi, W. J. Kim, K. Kang, D. S. Yun, M. S. Strano, G. Ceder and A. M. Belcher, Science, 2009, 324, 1051–1055.
20. J. Zhao, J. He, J. Zhou, Y. Guo, T. Wang, S. Wu, X. Ding, R. Huang and H. Xue, J. Phys. Chem. C, 2011, 115, 2888–2894.
21. J. M. Chen, C. H. Hsu, Y. R. Lin, M. H. Hsiao and T. K. Fey, J. Power Sources, 2008, 184, 498–502.
22. P. P. Prosini, D. Zane and M. Pasquali, Electrochim. Acta, 2001, 46, 3517–3523.
23. X. L. Wu, L. Y. Jiang, F. F. Cao, Y. G. Guo and L. J. Wan, Adv. Mater., 2009, 21, 2710–2714.
24. A. A. Salah, P. Jozwiak, J. Garbarczyk, K. Benkhouja, K. Zaghib, F. Gendron and C. M. Julien, J. Power Sources, 2005, 140, 370–375.
25. P. S. Herle, B. Ellis, N. Coombs and L. F. Nazar, Nat. Mater., 2004, 3, 147–152.
26. R. Malik, D. Burch, M. Bazant and G. Ceder, Nano Lett., 2010, 10, 4123–4127.
27. J. Qian, M. Zhou, Y. Cao, X. Ai and H. Yang, J. Phys. Chem. C, 2010, 114, 3477–3482.
28. S. Lim, S. Y. Chong and J. Cho, Chem. Mater., 2008, 39, 4560–4564.
29. Y. Wang, Y. Wang, E. Hosono, K. Wang and H. Zhou, Angew. Chem., 2008, 47, 7461–7465.
30. C. Sun, S. Rajasekhara, J. B. Goodenough and F. Zhou, J. Am. Chem. Soc., 2011, 133, 2132–2135.
31. X. Lou and Y. Zhang, J. Mater. Chem., 2011, 21, 4156–4160.
32. H. S. Kim, B. W. Cho and W. I. Cho, J. Power Sources, 2004, 132, 235–239.
33. K. S. W. Sing, Pure Appl. Chem., 1982, 54, 2201–2218.
34. J. Tang and A. P. Alivisatos, Nano Lett., 2006, 6, 2701–2706.
35. X. F. Shen and X. P. Yan, Angew. Chem., Int. Ed., 2007, 46, 7659–7663.
36. Y. Xiao, S. Liu, F. Li, A. Zhang, J. Zhao, S. Fang and D. Jia, Adv. Funct. Mater., 2012, 22, 4052–4059.
37. X. W. Xu, X. M. Zhang, C. Liu, Y. L. Yang, J. W. Liu, H. P. Cong, C. H. Dong, X. F. Ren and S. H. Yu, Adv. Funct. Mater., 2013, 135, 12928–12931.
38. Y. Gu, W. Liu, L. Wang, G. Li and Y. Yang, CrystEngComm, 2013, 15, 4865–4870.
39. H. Peng, C. K. Chan, S. Meister, X. F. Zhang and Y. Cui, Chem. Mater., 2009, 21(2), 247–252.
40. Y. Han, S. Liu, M. Han, J. Bao and Z. Dai, Cryst. Growth Des., 2009, 9, 3941–3947.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra13907h

---