A systematic study on the synthesis of α-Fe₂O₃ multi-shelled hollow spheres

Abstract

Hematite (α-Fe₂O₃) multi-shelled hollow spheres (MSHSs) have been prepared by a facile spray drying method using iron(III) citrate and sucrose as the precursors. The sucrose/iron(III) citrate ratio plays an important role in the morphology of the products. Well-defined MSHSs can be obtained in a wide sucrose/iron(III) citrate ratio of 0.25–1.5. The application of α-Fe₂O₃ MSHSs in lithium storage has been demonstrated. When used as the anode material for lithium-ion batteries, the α-Fe₂O₃ MSHSs exhibit a high reversible capacity of 979 mA h g⁻¹, maintaining 861 mA h g⁻¹ after 50 cycles.

---

1. Introduction

Hematite, also known as α-Fe₂O₃, is the most stable form of iron oxide with n-type semiconducting properties. Due to its low cost, abundance, non-toxicity, and high resistance to corrosion, α-Fe₂O₃ is promising for applications ranging from pigments to adsorption, photoelectrodes, gas sensors, and lithium-ion batteries (LIBs).^1–4 In the particular field of LIBs, α-Fe₂O₃ is one of the most promising alternatives to commercialised graphite anode. Each formula of Fe₂O₃ is able to react with 6 Li⁺, leading to a theoretical capacity (1007 mA h g⁻¹) three times to that of graphite (372 mA h g⁻¹). However, similar to other transition metal oxide based anode materials, α-Fe₂O₃ suffers from poor cycling stability which is associated with the large volume change (∼96%) during lithiation–delithiation.⁵ Various nanostructures, such as nanoparticles,⁶ nanoflakes,⁷ nanodiscs,⁸ nanorods,^5,9 nanotubes,^2,10–12 hierarchical structures^13,14 and hollow structures,^15–20 have been designed to improve the cycling stability of α-Fe₂O₃, among which, hollow structures have attracted the most attention.

Due to their effectiveness in alleviating the volume expansion issue, a number of methodologies have been developed to synthesise α-Fe₂O₃ hollow structures for lithium storage. For example, Lou et al. reported a quasiemulsion-templated method for the synthesis of α-Fe₂O₃ hollow spheres with sheet-like subunits.¹⁵ The same group also developed a thermal decomposition approach towards α-Fe₂O₃ hollow microboxes by using Prussian blue microcubes as the precursor.¹⁶ Wang and co-workers reported the synthesis of α-Fe₂O₃ MSHSs with controlled number of shells by employing carbonaceous microspheres as the sacrificial templates.^3,17 This method is quite versatile and can be generally applied to other transition metal oxide hollow structures.^21–25 Zhou et al. developed a low-cost and scalable spray drying method for the production of α-Fe₂O₃ MSHSs using iron nitrate and sucrose as the precursors.¹⁸ Similar to the spray drying method, spray pyrolysis was also employed to synthesise α-Fe₂O₃ yolk–shell spheres.¹⁹

Although α-Fe₂O₃ MSHSs can be prepared in large scale by spray drying iron nitrate and sucrose followed by calcination,¹⁸ some challenges still exist: (1) iron nitrate is a strong oxidant, while sucrose is a reductant; the spray drying of iron nitrate and sucrose in large scale may pose unacceptable safety hazards such as explosion. (2) Iron nitrate has a strong tendency to absorb moisture (the so-called hygroscopy), so does the iron nitrate–sucrose composite, making the storage of iron nitrate and iron nitrate–sucrose composite a problem. (3) The thermal decomposition of iron nitrate generates highly toxic gases such as NO₂. (4) Well defined α-Fe₂O₃ MSHSs can only be obtained in a narrow sucrose/iron nitrate range (from 0.75 to 1.0) due to the hygroscopy of iron nitrate. These limitations make the mass production of α-Fe₂O₃ MSHSs by spray drying iron nitrate and sucrose industrially unfeasible. Herein, we find that iron(III) citrate, a mild oxidant with zero nitrogen-content, can be used as the iron precursor to replace iron nitrate in spray drying. The above-mentioned four limitations associated with the utilisation of iron nitrate can be overcome, making the mass production viable.

2. Experimental

2.1 Preparation of α-Fe₂O₃ MSHSs

In a typical synthesis (Sample 1, see Table 1), 10 mmol of iron(III) citrate (C₆H₅FeO₇) and 5 mmol of sucrose were dissolved in 100 mL of water. The resulting clear solution was spray dried using a Buchi mini spray drier B-290 at an inlet temperature of 220 °C. The as-synthesised iron(III) citrate–sucrose composite was collected and calcined in air at 400 °C for 5 h with a temperature ramp rate of 2 °C min⁻¹. To study the effect of sucrose/iron(III) citrate ratio (defined as x), the sucrose amount was varied from 0 to 20 mmol, while the iron(III) citrate amount was kept constant at 10 mmol. More detailed synthesis parameters can be found in Table 1.

Table 1 Synthesis conditions and textural properties of α-Fe₂O₃ nanostructures

Sample name	Sucrose amount (mmol)	Iron(III) citrate amount (mmol)	x^a	SBET^b m^2 g^−1	Vp^c cc g^−1	Solid content^d %
a x = sucrose/iron(III) citrate ratio.b BET surface area.c Total pore volume.d Solid content of the iron(III) citrate–sucrose composites determined from TGA under air flow.e Sample 8 is prepared by calcination of the grinded iron(III) citrate–sucrose mixture at 400 °C for 5 hours.
Sample 1	5	10	0.5	38.1	0.17	14.23
Sample 2	0	10	0	60.5	0.16	27.91
Sample 3	2.5	10	0.25	49.0	0.15	14.76
Sample 4	7.5	10	0.75	28.5	0.12	13.36
Sample 5	10	10	1.0	39.3	0.15	13.08
Sample 6	15	10	1.5	12.9	0.09	9.01
Sample 7	20	10	2.0	15.5	0.06	7.66
Sample 8^e	5	10	0.5	14.8	0.15	—

---

2.2 Characterization

XRD patterns were obtained on a Rigaku Miniflex X-ray diffractometer with Fe-filtered Co radiation. TEM experiments were conducted on Philips Tecnai F20 and JEOL 1010. The samples for TEM were dispersed in ethanol, gently shaken, and then supported onto the copper grid. SEM images were obtained on JEOL JSM 7800 microscope operated at 15 kV. N₂ adsorption isotherms were measured at 77 K on a nitrogen adsorption apparatus (ASAP Tristar II 3020 surface area and porosity analyser, Micromeritics) after degassing the samples at 180 °C for 6 h. TGA was carried out on a TGA/DSC1 STAR^e System under air flow (25–800 °C, 5 °C min⁻¹).

2.3 Electrochemical measurements

The electrochemical measurements were carried out in homemade two-electrode Swagelok type cells. The working electrodes was prepared by casting a slurry containing α-Fe₂O₃ MSHSs (70 wt%), carbon black (20 wt%), and polyvinylidene fluoride (10 wt%) in n-methyl-2-pyrrolidone onto copper foil substrates and dried in a vacuum oven at 120 °C for 12 h. Lithium chips was employed as the counter electrode. The electrolyte was a mixture of 1 M LiPF₆ in ethylene carbonate (EC), dimethyl carbonate (DMC), and diethyl carbonate (DEC). Half-cells were assembled in an argon-filled glovebox. The cell performance was evaluated galvanostatically at room temperature in a voltage range of 0.05–3.0 V. Cyclic voltammetry (0.01–3.0 V, 0.1 mV s⁻¹) and electrochemical impedance measurements were performed on a CHI660E electrochemical workstation.

3. Results and discussion

The α-Fe₂O₃ MSHSs were synthesised by spray drying iron(III) citrate and sucrose followed by calcination in air. As schematically illustrated in Fig. S1,† the non-equilibrium heat treatment induced heterogeneous contraction during the calcination in air is responsible for the formation of the multi-shelled hollow structures.^18,26,27 More specifically, the iron(III) citrate and sucrose are assembled into composite microspheres by spray drying (Fig. S2 and S3†). During the calcination in air, the iron(III) citrate–sucrose composite microspheres are heated from the surface to the center. Due to the existence of a temperature gradient (ΔT) along the radial direction, the outer part of the composite spheres decomposes first, forming a rigid α-Fe₂O₃ shell at the surface. The interface between the inner core and outer shell experiences two forces from opposite directions: the adhesion force (F_a) from the shell and the contraction force (F_c) from the core. While the F_a induces outward shrinkage due to the rigid nature of the α-Fe₂O₃ shell, the F_c promotes continuous shrinkage of the core due to the burn-off of organic species. With a large ΔT at the early stage, F_c exceeds F_a. As a result, the core shrinks and detaches from the shell. The above-mentioned shell formation–core detachment process can be repeated for several times until ΔT become negligible. With a small ΔT, and thus a small F_c, the trend for mass diffusion is reversed. That is the inner core shrinks outwardly, leaving a hollow cavity at the centre.

A typical sample prepared with a sucrose/iron(III) citrate ratio of 0.5 (Sample 1) is shown in Fig. 1. Scanning electron microscopy (SEM) image (Fig. 1a) shows that well-defined Fe₂O₃ spheres with sizes ranging from 400 to 3000 nm can be obtained with high purity. For some spheres, especially those small in size, the electron beam is able to penetrate 1–2 shells, displaying the inner core particles. This phenomenon indicates that the Fe₂O₃ spheres may actually have a yolk–shell or multi-shelled hollow structure. Transmission electron microscopy (TEM) images (Fig. 1b–d) undisputedly demonstrate the multi-shelled hollow structure of the synthesised sample. Most of the Fe₂O₃ spheres have three to four porous shells constructed by numerous nanoparticulate building blocks. The selected area electron diffraction (SAED) pattern (Fig. 1e) of an individual sphere (Fig. 1c) shows a series of discontinuous concentric rings, indicating the polycrystalline characteristic. The diffraction rings from inside to outside can be indexed to the (012), (104), (110), (113), (024) and (116) diffractions of α-Fe₂O₃, respectively. A typical high-resolution TEM (HRTEM) image is shown in Fig. 1f, where the {012} (0.37 nm) and {110} (0.25 nm) lattice fringes of α-Fe₂O₃ can be clearly observed. The crystallographic structure and phase purity of the synthesised Fe₂O₃ is identified by X-ray diffraction (XRD) to be pure rhombohedral phase α-Fe₂O₃ (Fig. 1g, JCPDS card no. 33-0664, space group: R3̄c, a = 0.5035 nm, c = 1.3748 nm). In consistent with the TEM results, the XRD pattern shows relatively broad diffraction peaks, indicating the nanocrystalline characteristic of the sample. The width (full width at half maximum) of the diffraction peaks reflects the crystallite size of the primary nanoparticles. By applying Scherrer's equation to the most intensive (104) diffraction peak, the collective mean crystallite size is estimated to be 18.82 nm. The textural properties of the α-Fe₂O₃ MSHSs are studied by N₂ sorption (Fig. 1h). The α-Fe₂O₃ MSHSs exhibit a typical type IV isotherm with an H₁-type hysteresis loop at high relative pressure, indicating the mesoporous structure. The irregular packing of nanocrystallites in the shells gives rise to a high BET surface area of 38.1 m² g⁻¹, a pore volume of 0.17 cm³ g⁻¹, and a relatively broad pore size distribution centred at ∼20 nm. Compared to the α-Fe₂O₃ MSHSs obtained from iron nitrate,¹⁸ the sample prepared from iron(III) citrate shows a significant increased BET surface area (38.1 vs. 17.3 m² g⁻¹).

Fig. 1 SEM image (a), TEM images (b–d), SAED (e), HRTEM image (f), XRD pattern (g), N₂ adsorption–desorption isotherm (h), and pore size distribution (inset of h) of α-Fe₂O₃ MSHSs (Sample 1).

To study the effects of sucrose/iron(III) citrate ratio on the formation of α-Fe₂O₃ MSHSs, the sucrose amount is tuned from 0 to 20 mmol in the synthesis, while other synthesis parameters are kept constant. When no sucrose is used in the synthesis (x = 0), crumpled microspheres with a deflated balloon-like morphology are obtained (Fig. 2a). Crumpled microspheres have been frequently found in spray drying and its formation is believed to be fast solvent evaporation induced deformation.^28,29 Well-defined MSHSs can be synthesised in a wide sucrose/iron(III) citrate range of 0.25–1.50, corresponding to a sucrose amount of 2.5–15 mmol (Fig. 2b–e). This is in sharp contrast to the results obtained from iron nitrate precursor, where MSHSs can only be obtained in a narrow iron nitrate/sucrose ratio of 0.75–1.0 due to the hygroscopy of iron nitrate.¹⁸ Further increase of the sucrose amount to 20 mmol (x = 2.0) leads to aggregation and collapse of the MSHSs (Fig. 2f).

Fig. 2 SEM images of Fe₂O₃ nanostructures prepared with a sucrose/iron(III) citrate ratio of 0 (a, Sample 2), 0.25 (b, Sample 3), 0.75 (c, Sample 4), 1.0 (d, Sample 5), 1.5 (e, Sample 6) and 2.0 (f, Sample 7).

TEM characterisation is further utilised to obtain more information on the internal structure of the samples. As shown in Fig. 3a, deflated balloon-like microspheres comprised of numerous nanoparticles are obtained with an x value of 0. MSHSs with 2–3 shells and 3–4 shells can be obtained with a x value of 0.25 (Fig. 3b) and 0.50–1.5 (Fig. 1c and 3c–e), respectively. When the x value reaches 1.5, the shells of the MSHSs become less continuous. When x further increases to 2.0, severe aggregation and collapse of the MSHSs are observed, which is in agreement with the SEM results.

Fig. 3 TEM images of Fe₂O₃ nanostructures prepared with a sucrose/iron(III) citrate ratio of 0 (a, Sample 2), 0.25 (b, Sample 3), 0.75 (c, Sample 4), 1.0 (d, Sample 5), 1.5 (e, Sample 6) and 2.0 (f, Sample 7).

Thermal gravimetric analysis (TGA) analysis is used to obtain information on the decomposition behaviour of the as-synthesised iron(III) citrate–sucrose composites in air. TGA curves for selected samples (as-synthesised Sample 1, 2, 5, 6, and 7) are shown in Fig. S4.† The as-synthesised Sample 2 has a solid content of 27.91%. With the increase of sucrose/iron(III) citrate ratio (x), the solid content decreases continuously (Fig. S5† and Table 1). The collapse of the MSHSs in Sample 7 might be related to its low solid content (7.66%).

N₂ sorption is employed to gain information on the textural properties of the synthesised samples. N₂ adsorption–desorption isotherms and pore size distributions for selected samples are shown in Fig. S6.† The BET surface area and total pore volume of the samples are summarised in Table 1. Samples 1–5 with a x of 0–1.0 show relatively high surface area (25–60 m² g⁻¹) and pore volume (0.12–0.17 cm³ g⁻¹). Although the MSHS structure can be maintained in Sample 6 (x = 1.5), it shows significantly reduced BET surface area and total pore volume. Due to the aggregation and collapse of the MSHSs, Sample 7 with an x value of 2.0 also shows low BET surface area and total pore volume as well.

From the above results, it is safe to draw the conclusion that the sucrose/iron(III) citrate ratio (x) plays an important role in the morphology of the products. Well-defined α-Fe₂O₃ MSHSs can be obtained with an x value of 0.25–1.5. Above this value, the solid content of the iron(III) citrate–sucrose composites is too low to support the multi-shelled hollow structure. As a result, the α-Fe₂O₃ MSHSs tend to collapse (Fig. 3f). Below this value, the porosity generated by burn-off of the organic species is not enough to induce the formation of MSHSs. Instead, crumpled microspheres with a deflated balloon-like morphology are obtained (Fig. 3a).

As a proof-of-concept application, the synthesised α-Fe₂O₃ MSHSs are assembled into half cells for lithium storage study. Fig. 4 shows the first five cyclic voltammograms (CVs) of the α-Fe₂O₃ MSHSs. Two reduction peaks can be observed in the first cathodic process: a minor peak centred at 1.50 V corresponding to the reduction of Fe³⁺ to Fe²⁺, and a pronounced peak located at 0.47 V corresponding to the reduction of Fe²⁺ to Fe and the irreversible reduction of the electrolyte.¹⁸ Only one broad peak located between 1.75 and 2.00 V can be observed in the first anodic process, corresponding to the oxidation of Fe to Fe³⁺. The CV curves for the subsequent cycles differ substantially from that of the first cycle. From the second cycle onward, only a pair of broad reduction/oxidation peaks can be observed and the CV curves generally overlap, suggesting the good reversibility of the redox reaction.

Fig. 4 First five CVs of the α-Fe₂O₃ MSHSs at a scan rate of 0.1 mV s⁻¹ in the voltage window of 0.01–3.0 V vs. Li/Li⁺.

Representative galvonastatic discharge/charge curves are shown in Fig. 5a. The α-Fe₂O₃ MSHSs manifest an initial discharge and charge capacity of 1360 and 979 mA h g⁻¹, respectively. The irreversible capacity loss is mainly due to the formation of solid electrolyte interface (SEI) layer.⁵ The discharge capacity decreases to 908 mA h g⁻¹ in the first 5 cycles and then even off (Fig. 5b). Even after 50 discharge/charge cycles at 400 mA g⁻¹, a high capacity of 861 mA h g⁻¹ can still be retained. The coulombic efficiency as a function of cycle number is shown in Fig. S7.† The coulombic efficiency for the first cycle is 72%, it increases to 99.6% in 5 cycles and then stabilises. For the purpose of comparison, randomly aggregated α-Fe₂O₃ nanoparticles are prepared by calcination of the grinded iron(III) citrate–sucrose mixture at 400 °C (Sample 8, Fig. S8†). In contrast to α-Fe₂O₃ MSHSs, the discharge capacity of α-Fe₂O₃ nanoparticles is only 331 mA h g⁻¹ after 40 cycles (Fig. S9†).

Fig. 5 (a) Representative discharge/charge profiles at 400 mA g⁻¹; (b) cycling performance at 400 mA g⁻¹; (c) discharge/charge profiles at current densities of 400–4000 mA g⁻¹; (d) rate performance of α-Fe₂O₃ MSHSs (Sample 1).

To evaluate the rate capability of the α-Fe₂O₃ MSHSs, the current density is increased stepwise from 400 to 4000 mA g⁻¹ (Fig. 5c and d). The α-Fe₂O₃ MSHSs deliver a discharge capacity of 993, 821, 557, and 294 mA h g⁻¹ at a current density of 400, 1000, 2000, and 4000 mA g⁻¹, respectively. After the high rate discharge/charge cycling, a high specific capacity of 1052 mA h g⁻¹ can be restored when the current density is reduced to 400 mA g⁻¹.

Electrochemical impedance spectroscopy (EIS) is used to provide insight into the kinetics of lithium ion transfer. The Nyquist plots of the α-Fe₂O₃ MSHSs before and after 5 CV cycles are shown in Fig. 6. The two spectra share the common features of a depressed semicircle in the medium-to-high frequency region and an inclined linear tail in the low-frequency region. The former is related to the charge transfer resistance, while the later represents the Warburg impedance associated with the diffusion of lithium ions in the electrode materials.^30,31 The diameter of the semicircle decreases significantly after 5 CV cycles, indicating the reduction in charge transfer resistance.

Fig. 6 Nyquist plots of the α-Fe₂O₃ MSHSs electrode before and after 5 CV cycles.

The superior lithium storage performance of α-Fe₂O₃ MSHSs can be attributed to the unique structural features: (1) the shells constructed by nanoparticles significantly shorten the distances for Li⁺ diffusion, contributing to the high rate capability; (2) the hollow cavity effectively accommodates the large volume change during lithiation–delithiation, which enhances the cycling stability.

When compared to the previous report on spray drying synthesis of α-Fe₂O₃ MSHSs, the simple replacement of iron nitrate with iron(III) citrate allows a number of benefits. (1) Iron(III) citrate is a mild oxidant; the spray drying of iron(III) citrate in large scale is much safer than the spray drying of iron nitrate, which has a stronger oxidation capability. (2) Iron(III) citrate has a weaker tendency to absorb moisture from the atmosphere than iron nitrate, making the storage much easier. (3) Iron(III) citrate is an environment-friendly precursor with zero-nitrogen content, the thermal decomposition of which generates no toxic gases. (4) Well-defined α-Fe₂O₃ MSHSs can be obtained in a wide range of sucrose/iron(III) citrate ratio of 0.25–1.5. This is in sharp contrast to the iron nitrate system, where α-Fe₂O₃ MSHSs can only be obtained in a narrow iron nitrate/sucrose ratio of 0.75–1.0.¹⁸

4. Conclusions

In conclusion, well-defined α-Fe₂O₃ MSHSs has been synthesised by a facile spray drying method using iron(III) citrate and sucrose as the precursors. When used as the anode material for LIBs, the resulting α-Fe₂O₃ MSHSs show a high reversible capacity and good cycling performance. Considering the cheap, widely available and environment friendly precursors, and easy, scalable, and highly reproducible synthesis procedure, we believe that the α-Fe₂O₃ MSHSs are promising for a broad spectrum of applications such as LIBs, gas sensing, and water remediation.

Acknowledgements

The authors acknowledge the Australian Research Council for financial support. L. Z. gratefully acknowledges the award of UQ Postdoctoral Research Fellowship and UQ Foundation Research Excellence Awards. The Australian Microscopy & Microanalysis Research Facility and Australian National Fabrication Facility are acknowledged for providing structural characterization facilities.

Notes and references

1. L. S. Zhong, J. S. Hu, H. P. Liang, A. M. Cao, W. G. Song and L. J. Wan, Adv. Mater., 2006, 18, 2426.
2. J. Chen, L. N. Xu, W. Y. Li and X. L. Gou, Adv. Mater., 2005, 17, 582.
3. X. Y. Lai, J. Li, B. A. Korgel, Z. H. Dong, Z. M. Li, F. B. Su, J. A. Du and D. Wang, Angew. Chem., Int. Ed., 2011, 50, 2738.
4. L. Zhang, H. B. Wu and X. W. Lou, Adv. Energy Mater., 2014, 4, 1300958.
5. Y. M. Lin, P. R. Abel, A. Heller and C. B. Mullins, J. Phys. Chem. Lett., 2011, 2, 2885.
6. D. Larcher, C. Masquelier, D. Bonnin, Y. Chabre, V. Masson, J. B. Leriche and J. M. Tarascon, J. Electrochem. Soc., 2003, 150, A133.
7. M. V. Reddy, T. Yu, C. H. Sow, Z. X. Shen, C. T. Lim, G. V. S. Rao and B. V. R. Chowdari, Adv. Funct. Mater., 2007, 17, 2792.
8. J. S. Chen, T. Zhu, X. H. Yang, H. G. Yang and X. W. Lou, J. Am. Chem. Soc., 2010, 132, 13162.
9. X. Y. Yao, C. L. Tang, G. X. Yuan, P. Cui, X. X. Xu and Z. P. Liu, Electrochem. Commun., 2011, 13, 1439.
10. N. Kang, J. H. Park, J. Choi, J. Jin, J. Chun, I. G. Jung, J. Jeong, J. G. Park, S. M. Lee, H. J. Kim and S. U. Son, Angew. Chem., Int. Ed., 2012, 51, 6626.
11. Z. Y. Wang, D. Y. Luan, S. Madhavi, C. M. Li and X. W. Lou, Chem. Commun., 2011, 47, 8061.
12. J. P. Liu, Y. Y. Li, H. J. Fan, Z. H. Zhu, J. Jiang, R. M. Ding, Y. Y. Hu and X. T. Huang, Chem. Mater., 2010, 22, 212.
13. J. M. Jeong, B. G. Choi, S. C. Lee, K. G. Lee, S. J. Chang, Y. K. Han, Y. B. Lee, H. U. Lee, S. Kwon, G. Lee, C. S. Lee and Y. S. Huh, Adv. Mater., 2013, 25, 6250.
14. J. X. Zhu, Z. Y. Yin, D. Yang, T. Sun, H. Yu, H. E. Hoster, H. H. Hng, H. Zhang and Q. Y. Yan, Energy Environ. Sci., 2013, 6, 987.
15. B. Wang, J. S. Chen, H. B. Wu, Z. Y. Wang and X. W. Lou, J. Am. Chem. Soc., 2011, 133, 17146.
16. L. Zhang, H. B. Wu, S. Madhavi, H. H. Hng and X. W. Lou, J. Am. Chem. Soc., 2012, 134, 17388.
17. S. M. Xu, C. M. Hessel, H. Ren, R. B. Yu, Q. Jin, M. Yang, H. J. Zhao and D. Wang, Energy Environ. Sci., 2014, 7, 632.
18. L. Zhou, H. Y. Xu, H. W. Zhang, J. Yang, S. B. Hartono, K. Qian, J. Zou and C. Z. Yu, Chem. Commun., 2013, 49, 8695.
19. M. Y. Son, Y. J. Hong, J. K. Lee and Y. C. Kang, Nanoscale, 2013, 5, 11592.
20. S. L. Chou, J. Z. Wang, D. Wexler, K. Konstantinov, C. Zhong, H. K. Liu and S. X. Dou, J. Mater. Chem., 2010, 20, 2092.
21. Z. H. Dong, X. Y. Lai, J. E. Halpert, N. L. Yang, L. X. Yi, J. Zhai, D. Wang, Z. Y. Tang and L. Jiang, Adv. Mater., 2012, 24, 1046.
22. Z. H. Dong, H. Ren, C. M. Hessel, J. Y. Wang, R. B. Yu, Q. Jin, M. Yang, Z. D. Hu, Y. F. Chen, Z. Y. Tang, H. J. Zhao and D. Wang, Adv. Mater., 2014, 26, 905.
23. J. Y. Wang, N. L. Yang, H. J. Tang, Z. H. Dong, Q. Jin, M. Yang, D. Kisailus, H. J. Zhao, Z. Y. Tang and D. Wang, Angew. Chem., Int. Ed., 2013, 52, 6417.
24. X. Y. Lai, J. E. Halpert and D. Wang, Energy Environ. Sci., 2012, 5, 5604.
25. Z. M. Li, X. Y. Lai, H. Wang, D. Mao, C. J. Xing and D. Wang, J. Phys. Chem. C, 2009, 113, 2792.
26. L. Zhou, D. Y. Zhao and X. W. Lou, Adv. Mater., 2012, 24, 745.
27. J. G. Guan, F. Z. Mou, Z. G. Sun and W. D. Shi, Chem. Commun., 2010, 46, 6605.
28. Z. X. Wu, W. D. Wu, W. J. Liu, C. Selomulya, X. D. Chen and D. Y. Zhao, Angew. Chem., Int. Ed., 2013, 52, 13764.
29. M. Eslamian and N. Ashgriz, J. Eng. Mater. Tech., 2007, 129, 130.
30. W. J. Yu, P. X. Hou, F. Li and C. Liu, J. Mater. Chem., 2012, 22, 13756.
31. L. Zhou, L. C. Yang, P. Yuan, J. Zou, Y. P. Wu and C. Z. Yu, J. Phys. Chem. C, 2010, 114, 21868.

---

Footnote

† Electronic supplementary information (ESI) available: TGA curves, N₂ sorption results, and SEM images. See DOI: 10.1039/c4ra13790f

---