The effect of copper species in copper-ceria catalysts: structure evolution and enhanced performance in CO oxidation

Abstract

In this paper, copper doped ceria porous nanospheres were synthesized using carbon nanospheres as a hard template via a homogenous precipitation method at low temperature. The results showed that the copper-doped ceria has undergone a morphology transformation from the initial double-shell spheres to hollow spheres as the copper doping concentration increased from 0 to 7.5%_mol. Notably, the Cu/Ce + Cu atomic ratio in the final products was approximately five times the initial design ratio, which confirmed an efficient utilization of copper in this system. Furthermore, the copper-ceria catalysts exhibited enhanced catalytic performance towards CO conversion when compared with pure ceria catalysts (e.g., for the optimal catalysts, the complete conversion temperature was 160 °C, for the pure catalysts, complete conversion temperature was 300 °C). Through the analysis of the catalysts structure, we proved that the superior catalytic performance was derived from a combination of CuO_x clusters and copper ions in the Cu–[O_x]–Ce bulk phase.

---

Introduction

Ceria and ceria-based materials, on the basis of their excellent oxygen capacity, unique structural properties and Ce³⁺/Ce⁴⁺ redox cycle, have been widely used in many fields such as three-way catalysts (TWCs),¹ water-gas-shift (WGS) reactions² and light-catalyzed reactions.³ Previous work has demonstrated that more active lattice oxygen can be generated in metal-doped ceria nanomaterials through changing the surface element composition, which resulted in an enhanced catalytic performance.⁴ For example, Ce_0.80M_0.12Sn_0.08O_2−δ (M = Hf, Zr, Pr and La) have received extensive attention because of their superior properties for CO oxidation, which would be promising in future applications in the reduction of automobile emissions.⁵ It's worth noting that copper-ceria catalysts have been widely used in the removal of CO air pollution due to their activity, selectivity and low cost.⁶

In the past decade, many research groups have focused highly on controlling the morphology and size of inorganic nano-materials. To date, there have been several methods on the synthesis of nanostructures utilizing Kirkendall effect⁷ or Ostwald ripening.⁸ However, these methods have repeatability and uniformity issues. From this prospective, template synthesis method is characterized by easy accessibility and narrow size distribution. Different types of templates were used, including silica,⁹ silver,¹⁰ titanium oxide,¹¹ reverse micelles,¹² and gas bubbles.¹³ Template synthesis could provide a way to prepare nanomaterials with desirable morphology and size.

Although the template method has been described in previous literatures¹⁴ to obtain products with uniform morphology, few studies have reported the effect of doping content on structure evolution in the presence of template. Herein, we report the preparation of Cu-doped CeO₂ catalysts via a homogenous precipitation method at low temperature using carbon nanospheres as hard template and investigate their catalytic performance in CO oxidation. The final products transformed from the initial double-shell spheres to hollow spheres clearly, as the copper content increased. Remarkably, on the basis of the combination of CuO_x clusters and copper ions in the Cu–[O_x]–Ce bulk phase, Cu-doped CeO₂ has shown a dramatic catalytic performance towards CO oxidation.

Experimental

Materials

Cerium nitrate hexahydrate (99.9%, Ce(NO₃)₃·6H₂O), urea, glucose and cupric nitrate trihydrate (99%, Cu(NO₃)₂·3H₂O) were purchased from Tianjin Kermel Co. Ltd. All the reactants were of analytical grade and were used without further purification or modification. Deionized water and absolute ethanol were used throughout.

Preparation of carbon nanospheres

The synthesis of carbon nanospheres was described previously in detail.¹⁵ Briefly, glucose (2.5 g) was dissolved in 10 mL of distilled water to form an aqueous solution. Then the solution was transferred to a Teflon-lined stainless steel autoclave (20 mL in total volume) and heated for 6 h at 180 °C. The dark brown precipitates were washed with ethanol and distilled water several times and dried at 80 °C for 4 h.

Synthesis of Cu-doped CeO₂ nanospheres

In a typical synthesis, 4 mmol of metal salt (the total of Ce(NO₃)₃·6H₂O and Cu(NO₃)₂·3H₂O) was added into a three-necked round bottom flask and dissolved in a solvent of 72 mL ethanol and 9 mL water to form a clear solution; then 45 mmol of urea was dissolved in the metal solution after vigorous stirring. The as-prepared carbon nanospheres (800 mg) were added and well dispersed into the above solution with the assistance of sonication for 20 min. Finally, the mixture was kept at 60 °C for 48 h with vigorous stirring before the products were collected by centrifugation. The products washed with deionized water and absolute alcohol three times sequentially. Finally, the products were dried at 80 °C for 4 h and then calcined at 400 °C for 3 h. The above products with different molar ratio of copper in total metal ions (0, 0.5%, 1.5%, 2.5%, 5.0%, 7.5%) were labeled as P₁–P₆ samples, which was shown in Table 1.

Table 1 The compositional data of P₁–P₆ samples and N₂ adsorption–desorption characterization

Samples	Design ratio	SBET (m^2 g^−1)	VBJH (cm^3 g^−1)	DBJH (nm)	EDS
	Cu/Ce + Cu, (%mol)				Cu/Ce + Cu, (%mol)
P1	0	112	0.236	3.8	0
P2	0.5	109	0.232	3.8	2.5
P3	1.5	114	0.220	3.8	7.7
P4	2.5	114	0.236	3.8	11.0
P5	5.0	93	0.220	3.8	23.4
P6	7.5	91	0.175	3.8	36.7

---

Characterization

The phase purity of the samples was examined by a Bruker D8 Advance X-ray diffractometer with a graphite mono chromator and Cu Kα radiation (λ = 0.15418 nm) in the 2θ range from 10° to 80°. The nanostructure and morphology of the products were characterized using a transmission electron microscope (TEM, JEM 100-CXII, 80 kV), a scanning transmission electron microscope (STEM, JEOL-TEM) a field-emission scanning electron microscope (FE-SEM, Hitachi, S4800) equipped with a energy-dispersive X-ray spectrometer (EDS), a scanning electron microscope (SEM, Hitachi, SU8010) and a high-resolution transmission electron microscope (HRTEM, JEM-2100, 200 kV). Raman data were obtained using a Lab RAM HR4800 spectrometer with a 514 nm laser line as an excitation source. The N₂ adsorption–desorption isotherms were measured on a QuadraSorb SI at 77.3 K. Before the measurement, the samples were outgassed at 200 °C under vacuum for 6 h. The surface areas were calculated by the Brunauer–Emmett–Teller (BET) method and the poresize distribution was calculated from the desorption branch using the Barrett–Joyner–Halenda (BJH) theory. Temperature programmed reduction under a H₂ environment (H₂-TPR) was carried out on a PCA-1200 instrument. Typically, 50 mg CeO₂ catalyst was pretreated under 5% O₂–Ar stream at 300 °C for 0.5 h (heating rate = 10 °C min⁻¹). After cooling down to room temperature, a flow of 5% H₂–Ar was introduced into the CeO₂ sample with a flow rate of 30 mL min⁻¹, and then the temperature was increased to 500 °C at a rate of 10 °C min⁻¹.

Catalytic activity measurements

The catalytic activity of the as-obtained samples was evaluated in a continuous flow fixed-bed microreactor operating under atmospheric pressure. In a typical experiment, catalyst particles (25 mg) were placed in the reactor. The reactant gases (1% CO, 10% O₂, and 89% N₂) passed through the reactor at a rate of 60 mL min⁻¹. The composition of the gas exiting the reactor was analyzed with an online infrared gas analyzer (Gasboard, China Wuhan Cubic Co.) which simultaneously detects CO and CO₂ with a resolution of 10 ppm.

Results and discussions

Structural analysis

X-ray diffraction data were collected for pure CeO₂ and the series of Cu-doped CeO₂ samples, as shown in Fig. 1. Four main diffraction peaks of all the products at 28.6, 33.2, 47.9 and 56.4° illustrate that they can be well indexed to a pure phase of face-centered cubic ceria structures (JCPDS no. 34-0394).¹⁶ It is noticed that no other substances, such as Cu₂O, CuO or CeOHCO₃, could be found. Furthermore, the broadening peaks indicate the formation of small nanocrystallites ca. 5 nm (Table S1†) in all products.¹⁷

Fig. 1 The XRD patterns of pure CeO₂ and Cu-doped CeO₂ samples.

The samples were further studied by Raman spectroscopy with an exciting laser wavelength of 514 nm, which is illustrated in Fig. 2, to investigate the effect of copper doping in nanomaterials.¹⁸ The observed main band of P₁ sample (pure ceria) is 463 cm⁻¹, which is ascribed to the triply degenerate F_2g mode of ceria.¹⁹ In contrast with pure CeO₂, slight shifts towards lower frequency of the F_2g Raman peaks of doped samples evidence local structure distortion on embedding copper ions into the lattice sites of CeO₂.²⁰ It's worth noting that the most significant red shift (449 cm⁻¹) appears in P₆ which has the highest doping concentration. Additional weak peaks at around 600 cm⁻¹ in P₂–P₆ and 350 cm⁻¹ in P₆ are also observed. We attribute this mode to the presence of additional oxygen vacancies introduced into the ceria lattice by substitution of Ce ions with copper ions in order to keep neutrality.²¹ Moreover, absence of the Cu–O characteristic peak (in the range 294 cm⁻¹) is indicative of the formation of solid solutions as indicated by the XRD results.²²

Fig. 2 Raman spectra of the synthesized P₁–P₆.

N₂ adsorption–desorption characterization and the compositional data of the as prepared P₁–P₆ samples are shown in Table 1. Calcination has an obvious influence on surface area (Table. S2†). All products exhibit single peaks at ca. 3.8 nm, which proves that the products have narrow pore-size distributions (Fig. S1†). Surprisingly, shown by EDS results, the Cu/Ce + Cu actual atomic ratio in the calcinated samples are approximately five times of the initial design ratio (Fig. S2 and Table S3†), and notably the multiple is higher than reported.²³ This synthesis approach can realize the efficient utilization of Cu species.

The morphologies and structures of the as-synthesized P₁ and P₄ samples were explored by transmission electron microscopy (TEM) and high-resolution transmission electron microscope (HRTEM) (Fig. 3). The absence of free nanoparticles in TEM and HRTEM images indicates that the precipitation of metal cations occurred preferentially in the carbon template. The low magnification TEM and HRTEM images of the precursors of P₁ and P₄ show that the products consist of well-dispersed uniform nanospheres with a size distribution ranging from 220 to 280 nm (Fig. 3A1 and B1), which are similar to carbon nanospheres (Fig. S1†). The TEM images (Fig. 3A2 and B2) show that our products are about 150 nm and make it clear that the surface of the nanosphere is very rough. Notably, an interesting phenomenon appeared when appropriate amount of Cu²⁺ was introduced into the ceria host. Fig. 3 and S2† show that the morphology of samples transformed from double-shell to hollow spheres as the designed doping content increased from 0 to 7.5%_mol. Furthermore, the high magnification technology was applied to analyze the detailed structure of the rough nanospheres (inset in Fig. 3A3 and B3). The polycrystalline properties proved by the selected area electron diffraction (SAED) patterns from the HRTEM results of P1 and P4 samples demonstrate that the products are formed of many small nanoparticles. The lattice fringes measured from HRTEM images is about 0.273 nm (Fig. 3A4) and 0.275 nm (Fig. 3B4), which correspond to the spacing of the (200) planes of ceria. No isolated copper related nanostructures can be detected in HRTEM images, which is also approved by XRD results.

Fig. 3 The TEM and HRTEM images with different magnification of the as-prepared pure CeO₂ and Cu²⁺-doped CeO₂. (A1 and B1) The TEM images of the precursors of P₁ and P₄; (A2 and B2) the TEM images of P₁ and P₄; (A3 and B3) the HRTEM images of P₁ and P₄; (A4 and B4) SAED pattern of P₁ and P₄ taken from a single double-shell and hollow sphere, which are shown as an inset in A3 and B3.

Moreover, the morphologies and structures of precursors of P₁, P₁, precursors of P₄ and P₄ are explored by scanning electron microscope (SEM) (Fig. S3†) and scanning transmission electron microscope (STEM). Elemental mapping analysis of STEM provides a direct elemental distribution of the samples, which confirms the homogeneous distribution of Ce, Cu, O and C in precursors of P₄ (Fig. 4a) and Ce, Cu and O in P₄, respectively (Fig. 4b). The results clearly indicate that the Cu and Ce species are highly dispersed, and no obvious Cu-rich or Ce-rich region is observed, which is consistent with the XRD results. In addition, elemental mapping analysis of P₁ sample is similar to P₄ (Fig. S4†).

Fig. 4 The STEM images of precursors of P₄ samples (a) and P₄ (b); EDS-mapping image of an individual nanosphere, which is marked in (a) and (b).

Morphology mechanism

In order to understand the formation mechanism of the samples better, we studied their temporal morphological evolution by taking TEM images of carbon spheres (Fig. S5†), samples obtained at different doping concentration (Fig. S6†) and calcination process of P₁ and P₅ (Fig. S7†). Results show that the nanospheres of P₁ samples are completely double-shelled, but as the Cu concentration increases, the proportion of double-shell nanospheres decreases, meanwhile hollow sphere pattern appears.

The proposed formation route is illustrated in Fig. 5. The whole process is explored as follows: (1) metal cations were adsorbed in the carbon nanosperes because of electrostatic force; (2) the metal salts were precipitated as the urea decomposed slowly; (3) the carbon templates were removed during the calcination step, leading to the formation of metal oxide hollow structure. Notably that the values of the Cu/Ce + Cu atomic ratio are approximately five times of the initial design ratio. It can be described as follows: in the process of ion adsorption in the carbon nanospheres²⁴ and low temperature urea homogenous precipitation,²⁵ The adsorption rate for Cu²⁺ is faster than Ce³⁺ due to their different valence, ionic radius and ion mass. Then the metal ions are precipitated in the form of basic precursors²⁶ in the alkaline condition provided by urea. In the process of urea homogenous precipitation, metal ions are adsorbed and precipitated in the carbon nanospheres.²⁷ During the calcination stage, the lessen of double-shell nanospheres may be greatly dependent on the quantity of Cu deposited in the carbon spheres during the reaction in the solution and it is postulated that the presence of Cu²⁺ hindered the formation of double-shell structure during the calcination process.²⁸ According to results from calcination, we find that the removal process of carbon template for P₁ and P₅ are different (Fig. S7 and S8†). Template of the P₅ sample was removed completely before 280 °C, while P₁ sample was still solid sphere. It indicate that the presence of copper promoted the oxidation of carbon template, during which process copper induces a similar effect as in enhancement for catalytical oxidation of CO. The slower the remove rate of carbon template, the easier some nanoislands of catalysts in the shell migrate to or be ‘stuck’ on the surface of the shrinking carbon cores during the calcination process.²⁹

Fig. 5 Schematic illustration of the formation of hollow spheres and double-shell spheres at the calcination process.

H₂-TPR reduction behaviors

The H₂-TPR profiles of the samples with different copper content are shown in Fig. 6, in the temperature range 50–400 °C. The wide band at low temperature around 170 °C (called α peak) can be attributed to the reduction of highly dispersed surface CuO_x clusters,³⁰ while the sharper one (called β peak) located around 200 °C is due to the strong interaction of Cu–[O_x]–Ce structure.^16c For the samples of P₃–P₆, the area of the third peak (called γ peak) at higher temperature about 220 °C is originated from the reduction of CuO particles.³¹ The 150–250 °C reduction region is commonly associated with Cu(II) centers and normalized with respect to the total copper content of each sample, and the area of reduction peaks increases with the increase of copper content, demonstrating that copper is dispersed in the crystal lattice of ceria to enhance its reducibility, which is in accordance with the XRD results. A previous report revealed that the pure ceria has two reduction peaks: one at 500 °C and the other above 700 °C, which can be assigned to the reduction of surface oxygen and bulk oxygen respectively.³² Notably the position and relative intensity of these three peaks change with the copper content of the catalysts. The α peak positions of all samples are of no obvious difference, indicating that the reduction properties of surface CuO_x clusters are similar. The presence of copper, even as less as the copper content in P₂, affects surface redox properties. The redox properties of Cu²⁺ and Ce⁴⁺ are mutually affected by the presence of each other, and their sequential reduction cannot be regarded as independent processes.³³

Fig. 6 The H₂-TPR profiles corresponding to the synthesized P₁–P₆ samples.

Catalytic properties

The catalytic activity of the copper doped ceria nanospheres towards CO oxidation was evaluated via a continuous flow fixed-bed microreactor. Fig. 7 displays the catalytic activities of the Cu-doped CeO₂, along with the pure monodispersed ceria nanospheres and commercial CeO₂ powder as comparisons. It can be obviously seen that the catalytic performance of our products has been sharply improved by comparison with commercial ceria. The conversion rate of commercial ceria failed to reach 40% at 300 °C, while all of our samples have achieved full conversion. The surface areas of all our products are larger than that of the commercial ceria, which means they can support more active sites, thus leading to enhanced catalytic property. The P₄ product is much more active than the others, which become active at 55 °C whereas the pure CeO₂ nanospheres become active at 160 °C. The T₁₀₀ temperatures (the temperature at 100% conversion) of the prepared samples are as follows: 160 °C (P₄) < 180 °C (P₆ and P₅) < 215 °C (P₃) < 250 °C (P₂) < 300 °C (P₁). According to the previous research, surface area and pore size distribution are of little importance.³⁴ The results clearly indicate that the T₁₀₀ temperature witnesses a sharp drop as the doping ratio increases, and reaches its lowest point at sample P₄. However, a further increase in the doping ratio will lead to its increase again, while still much lower than the sample without doping. To make this explanation clear, we employ the date from H₂-TPR results and BET measurements. For P₁–P₄, the surface areas are similar, the increase of doping concentration resulted in more active sites (CuO_x clusters and Cu–[O_x]–Ce structures),^6a,35 so the catalytical performance from P₁ to P₄ improved obviously; for P₅ and P₆, the morphology of products transforms to hollow spheres, and the surface areas are smaller than P₁–P₄, so the quantities of active sites supported by catalysts with the same quality are lesser, which result in the weakness of catalytical performance slightly. In brief, the key to catalytical process is mainly two active copper species which provide probably the same contribution: CuO_x clusters and Cu–[O_x]–Ce structures.

Fig. 7 Conversion of CO over commercial ceria and P₁–P₆ samples.

In order to investigate the thermal stability of the samples, the recycling tests were performed in six cycles. Fig. 8 demonstrates the catalytic profiles of the P₄ sample after each cycle. It is obvious that the sample retains high catalytic activity throughout all those runs. The initial conversion temperatures are about 55 °C. And all the final conversion temperatures are about 160 °C. In order to confirm our conclusion, we tested the catalytical properties of P₅ and P₆, which indicate that P₅ and P₆ are as stable as P₄ (Fig. S9†).

Fig. 8 Catalytical performance of P₄ in different runs.

Conclusion

In summary, copper doped ceria porous nanospheres were synthesized using carbon nanospheres as hard template via a homogenous precipitation method. The morphology of the final products changed from the initial double-shell spheres to hollow spheres obviously, as the copper doping concentration increased. Cu-doped ceria with high utilization of copper species showed excellent catalytic activity and cycling properties towards CO oxidation. With detailed assessment of the structure, it can be concluded that the superior catalytic performance was attributed to the combination of CuO_x clusters and copper ions in the Cu–[O_x]–Ce bulk phase. We hope this study could provide a rational preparation way for novel doped CeO₂ with the desirable morphology and size, as well as other functional materials.

Acknowledgements

This work was supported by the Natural Science Foundation of China (grant no. 21276142 and 21476129), the Science & Technology Development Projects of Shandong Province (grant no. 2014GSF117024) and Special Fund for “Taishan Scholar” construction engineering “agricultural nonpoint source pollution prevention and control” position.

References

1. (a) M. Haneda, T. Kaneko, N. Kamiuchi and M. Ozawa, Catal. Sci. Technol., 2015, 5, 1792–1800; (b) K. Tang, W. Liu, J. Li, J. Guo, J. Zhang, S. Wang, S. Niu and Y. Yang, ACS Appl. Mater. Interfaces, 2015, 7, 26839–26849.
2. D.-W. Jeong, H.-S. Na, J.-O. Shim, W.-J. Jang and H.-S. Roh, Catal. Sci. Technol., 2015, 5, 3706–3713.
3. N. S. Arul, D. Mangalaraj, R. Ramachandran, A. N. Grace and J. I. Han, J. Mater. Chem. A, 2015, 3, 15248–15258.
4. (a) M. D. Krcha and M. J. Janik, Catal. Sci. Technol., 2014, 4, 3278–3289; (b) W. Cen, Y. Liu, Z. Wu, H. Wang and X. Weng, Phys. Chem. Chem. Phys., 2012, 14, 5769–5777.
5. D. Devaiah, T. Tsuzuki, T. Boningari, P. G. Smirniotis and B. M. Reddy, RSC Adv., 2015, 5, 30275–30285.
6. (a) G. Chen, Q. Xu, Y. Yang, C. Li, T. Huang, G. Sun, S. Zhang, D. Ma and X. Li, ACS Appl. Mater. Interfaces, 2015, 7, 23538–23544; (b) Z. Liu, Z. Wu, X. Peng, A. Binder, S. Chai and S. Dai, J. Phys. Chem. C, 2014, 118, 27870–27877; (c) S. Sun, D. Mao, J. Yu, Z. Yang, G. Lu and Z. Ma, Catal. Sci. Technol., 2015, 5, 3166–3181.
7. N. Du, H. Zhang, B. Chen, X. Ma and D. Yang, Chem. Commun., 2008, 3028–3030.
8. J. Li and H. C. Zeng, J. Am. Chem. Soc., 2007, 129, 15839–15847.
9. K. H. Rhodes, S. A. Davis, F. Caruso, B. Zhang and S. Mann, Chem. Mater., 2000, 12, 2832–2834.
10. Y. Sun, B. T. Mayers and Y. Xia, Nano Lett., 2002, 2, 481–485.
11. F. Caruso, X. Shi, R. Caruso and A. Susha, Adv. Mater., 2001, 13, 740–744.
12. T. Chen, I. Öçsoy, Q. Yuan, R. Wang, M. You, Z. Zhao, E. Song, X. Zhang and W. Tan, J. Am. Chem. Soc., 2012, 134, 13164–13167.
13. J. Yang and T. Sasaki, Chem. Mater., 2008, 20, 2049–2056.
14. B. Liu, Q. Wang, S. Yu, T. Zhao, J. Han, P. Jing, W. Hu, L. Liu, J. Zhang, L. D. Sun and C. H. Yan, Nanoscale, 2013, 5, 9747–9757.
15. X. Sun and Y. Li, Angew. Chem., Int. Ed., 2004, 43, 597–601.
16. (a) Q. Dai, S. Bai, H. Li, W. Liu, X. Wang and G. Lu, CrystEngComm, 2014, 16, 9817–9827; (b) W. Liu, X. Liu, L. Feng, J. Guo, A. Xie, S. Wang, J. Zhang and Y. Yang, Nanoscale, 2014, 6, 10693–10700; (c) R. Si, J. Raitano, N. Yi, L. Zhang, S.-W. Chan and M. Flytzani-Stephanopoulos, Catal. Today, 2012, 180, 68–80.
17. X. Zhang, J. Wei, H. Yang, X. Liu, W. Liu, C. Zhang and Y. Yang, Eur. J. Inorg. Chem., 2013, 2013, 4443–4449.
18. Y. Lee, G. He, A. J. Akey, R. Si, M. Flytzani-Stephanopoulos and I. P. Herman, J. Am. Chem. Soc., 2011, 133, 12952–12955.
19. P. Fornasiero, G. Balducci, R. Di Monte, J. Kašpar, V. Sergo, G. Gubitosa, A. Ferrero and M. Graziani, J. Catal., 1996, 164, 173–183.
20. (a) H. Yen, Y. Seo, S. Kaliaguine and F. Kleitz, Angew. Chem., Int. Ed., 2012, 51, 12032–12035; (b) B. Choudhury and A. Choudhury, Curr. Appl. Phys., 2013, 13, 217–223.
21. (a) Z. D. Dohčević-Mitrović, M. J. Šćepanović, M. U. Grujić-Brojčin, Z. V. Popović, S. B. Bošković, B. M. Matović, M. V. Zinkevich and F. Aldinger, Solid State Commun., 2006, 137, 387–390; (b) X. Wang, W. Wen, Y. Su and R. Wang, RSC Adv., 2015, 5, 63135–63141.
22. (a) N. K. Gamboa-Rosales, J. L. Ayastuy, M. P. González-Marcos and M. A. Gutiérrez-Ortiz, Int. J. Hydrogen Energy, 2012, 37, 7005–7016; (b) W. Liu, X. Liu, L. Feng, J. Guo, A. Xie, S. Wang, J. Zhang and Y. Yang, Nanoscale, 2014, 6, 10693–10700.
23. (a) G. Avgouropoulos and T. Ioannides, Appl. Catal., A, 2003, 244, 155–167; (b) M. S. P. Francisco, V. R. Mastelaro, P. A. Nascente and A. O. Florentino, J. Phys. Chem. B, 2001, 105, 10515–10522.
24. (a) X. Sun and Y. Li, Angew. Chem., Int. Ed., 2004, 43, 3827–3831; (b) R. Yang, H. Li, X. Qiu and L. Chen, Chem.–Eur. J., 2006, 12, 4083–4090.
25. J. Yu, X. Yu, B. Huang, X. Zhang and A. Y. Dai, Cryst. Growth Des., 2009, 9, 7.
26. A. M. Kaczmarek, L. Miermans and R. Van Deun, Dalton Trans., 2013, 42, 4639–4649.
27. (a) J. Ming, H. Cheng, Y. Yu, Y. Wu and F. Zhao, J. Mater. Chem., 2011, 21, 6654; (b) B. Hu, K. Wang, L. Wu, S.-H. Yu, M. Antonietti and M.-M. Titirici, Adv. Mater., 2010, 22, 813–828; (c) F. Teng, W. Yao, Y. Zheng, Y. Ma, T. Xu, G. Gao, S. Liang, Y. Teng and Y. Zhu, Talanta, 2008, 76, 1058–1064.
28. (a) X. W. Lou, C. Yuan and L. A. Archer, Small, 2007, 3, 261–265; (b) H. Qian, G. Lin, Y. Zhang, P. Gunawan and R. Xu, Nanotechnol., 2007, 18, 355602.
29. L. Wang, Z. Lou, T. Fei and T. Zhang, J. Mater. Chem., 2011, 21, 19331.
30. W.-W. Wang, P.-P. Du, S.-H. Zou, H.-Y. He, R.-X. Wang, Z. Jin, S. Shi, Y.-Y. Huang, R. Si, Q.-S. Song, C.-J. Jia and C.-H. Yan, ACS Catal., 2015, 5, 2088–2099.
31. (a) D. Gu, C.-J. Jia, H. Bongard, B. Spliethoff, C. Weidenthaler, W. Schmidt and F. Schüth, Appl. Catal., B, 2014, 152–153, 11–18; (b) G. Avgouropoulos and T. Ioannides, Appl. Catal., A, 2003, 244, 155–167.
32. Z. Zhang, D. Han, S. Wei and Y. Zhang, J. Catal., 2010, 276, 16–23.
33. M. Jobbagy, F. Marino, B. Schönbrod, G. Baronetti and M. Laborde, Chem. Mater., 2006, 18, 1945–1950.
34. H.-T. Chen and J.-G. Chang, J. Phys. Chem. C, 2011, 115, 14745–14753.
35. Z. Yang, B. He, Z. Lu and K. Hermansson, J. Phys. Chem. C, 2010, 114, 4486–4494.

---

Footnote

† Electronic supplementary information (ESI) available: Primary particle size of samples, N₂ adsorption–desorption isotherms and corresponding BJH pore size distribution curves, XRD patterns, EDS results, AES-ICP data and additional TEM and SEM images. See DOI: 10.1039/c6ra08204a

---