The redispersion behaviour of Pt on the surface of Fe₂O₃

Abstract

The redispersion behaviour of Pt on ferric oxides under cyclic oxidizing/reducing conditions was observed and studied using several analysis techniques. Pt in a Pt/Fe₂O₃ catalyst sintered after treatment under a single oxidative or reductive atmosphere at 600 °C and the average size of the Pt nanoparticles increased from the original 3.2 nm to 6.8 and 14.6 nm, respectively. In contrast, the average size decreased to 1.1 nm when the catalyst was treated under an alternating oxidizing and reducing atmosphere. It is proposed that the Pt–O–Fe bond formed under an oxidative atmosphere and the related defects or oxygen vacancies formed under a reductive atmosphere play an important role during the redispersion of Pt. In addition, the effect of temperature on the redispersion of Pt is also noted and discussed.

---

Introduction

Among commercial catalysts, supported noble metal catalysts (e.g., platinum, palladium and rhodium) are widely utilized in various industrial processes,^1–3 such as the removal of volatile organic compounds (VOCs),^4,5 the selective hydrogenation or oxidation of hydrocarbons^6–8 and the production of hydrogen.^9,10 In particular, supported Pt catalysts have attracted continuous attention because of their marvelous catalytic performance in emission control/treatment¹¹ and the refining of fuel stock.^12–14 Based on the fact that Pt is fairly expensive and difficult to replace with other components, one of the biggest issues relating to supported Pt catalysts is how to prolong the life-time of the catalysts or regenerate the deactivated ones effectively.

Generally, highly dispersed noble metal nanoparticles are more active than bulk ones for the majority of catalytic reactions. However, they often tend to agglomerate into larger particles, especially at higher reaction temperatures, leading to the degradation of catalytic activity and even a loss of the unique catalytic properties observed in the original nanoparticles. The redispersion of noble metals is an effective method to reverse sintering during reactions or to regenerate spent catalysts for reuse.^15–18 In the cases of Rh and Pd, they can be efficiently redispersed simply using flowing oxygen at certain temperatures.¹⁹ Nevertheless, the redispersion of Pt might be not that easy. For instance, the common method for the redispersion of noble metals mentioned above is not suitable for Pt because even when treated at the optimal operating conditions (around 500 °C), the redispersion of Pt is still at a low level.^20,21 As for other methods, the conditions are comparatively rigorous, for example the support is restricted to substrates containing ceria^22,23 or corrosive chlorine must be introduced during treatment.²⁴

Herein, we find that Pt can be redispersed effectively on the surface of Fe₂O₃ through alternating treatments under oxidative and reductive atmospheres. When treated at 600 °C, most Pt nanoparticles with an average size of 3.2 nm were redispersed to 1.1 nm. In addition, CO oxidation was used as a probe reaction to explore the catalytic performance of the Pt/Fe₂O₃ samples before and after redispersion. As we know, it is an expected trend that supported noble metal catalysts with smaller particle sizes (within a certain limit) are beneficial to the enhancement of catalytic activity, especially for simple catalytic reactions.^15–18 The results of the activity test showed that the catalytic activity of Pt/Fe₂O₃ was obviously enhanced with the decrease in the size of the Pt nanoparticles.

Experimental

Chemicals

All the reagents were used as received without further purification. H₂PtCl₆·6H₂O was obtained from Shanghai Shi Wu Chemical Reagent Science and Technology Co. Ltd. Polyvinylpyrrolidone (PVP) and FeCl₃·6H₂O were obtained from Aladdin Chemicals Co. Ltd.

Catalyst preparation

Synthesis of Pt nanoparticles (NPs). 10 mL of 0.03 M H₂PtCl₆ aqueous solution was added to a solution mixture containing 0.33 g of PVP (K-30) and 90 mL of ethylene glycol with vigorous stirring. The formed solution was heated to 200 °C and kept at this temperature for 1.5 h. After cooling to room temperature, the crude solution of Pt colloid was precipitated with a triple volume of acetone and further isolated using centrifugation at a speed of 8000 rpm. The obtained Pt NPs were dispersed in 50 mL of ethanol for further use and characterization.

Synthesis of Fe₂O₃ nanorods (NRs) and preparation of Pt/Fe₂O₃. 2.7029 g of FeCl₃·6H₂O was dissolved in 100 mL of deionized water in a beaker. The beaker containing the solution was then sealed and kept at 80 °C for 12 h. After filtering and washing several times with deionized water, the obtained sample was dried at 110 °C for 8 h and then calcined at 400 °C for 3 h in air.

Pt/Fe₂O₃ was prepared through the wet impregnation of Fe₂O₃ with a Pt sol. A slurry of Pt NPs and Fe₂O₃ was stirred for 1 h at room temperature and sonicated for 30 min. Then, the mixture was evaporated at 60 °C with stirring and calcined at 400 °C for 3 h.

The Pt/Fe₂O₃ catalyst was also prepared through a conventional impregnation method: an appropriate amount of H₂PtCl₆ solution (0.03 M) was added to the Fe₂O₃ powder and the mixture was evaporated to dryness at 60 °C under stirring to obtain 3.1 wt% Pt/Fe₂O₃. Then the catalyst was dried at 110 °C for 10 h, calcined at 450 °C for 3 h and reduced in a H₂ flow at 200 °C for 2 h. The synthesized catalyst was denoted as Pt/Fe₂O₃(IM).

Characterization

The CO oxidation activities of the catalysts were measured using a fixed-bed flow reactor. 25 mg of sample was loaded into a quartz tube and supported using quartz wool. The reactant gas (1% CO/air vol%) was passed through the catalytic bed at a flow rate of 25 mL min⁻¹ and measured using a gas chromatograph equipped with a thermal conductivity detector.

The samples were treated in a system with a switch valve and temperature controlling system (Fig. 1) which allowed exposure to the various atmospheres under investigation. 25 mg of sample was heated to 500–700 °C at a heating rate of 10 °C min⁻¹ under a certain atmosphere (20% O₂/N₂, 4% H₂/Ar or an alternating atmosphere of these two kinds of gases) at a flow rate of 100 mL min⁻¹ and maintained at that temperature for 30 min, then cooled down to room temperature under the same atmosphere. In the redispersion experiment, samples were treated under an alternating atmosphere of 20% O₂/N₂ for 5 seconds and 4% H₂/Ar for 10 seconds. 10% H₂O vapor was produced by bubbling Ar through water at 46 °C at 0.1 MPa saturated vapor pressure. The samples treated in 20% O₂/N₂, 4% H₂/Ar and alternating gases of 20% O₂/N₂ and 4% H₂/Ar at different temperatures were denoted as Pt/Fe₂O₃(600 O), Pt/Fe₂O₃(600 H) and Pt/Fe₂O₃(500 H–O, 600 H–O or 700 H–O), respectively.

Fig. 1 Sketch map of the system used for treating samples under various atmospheres and CO oxidation.

The powder X-ray diffraction (XRD) patterns were obtained with a Bruker D8 Advance X-ray diffractometer using a Cu Kα radiation source (λ = 1.5406 Å). The powder samples were placed on a glass slide and scanned from 10° to 90° at a rate of 3° min⁻¹. The elemental content was measured using ICP-OES (iCAP 6300 Thermo Scientific USA). The hydrogen temperature-programmed reduction (H₂-TPR) experiments were performed on a Micromeritics AutoChem 2920 instrument. Before H₂-TPR, the samples were heated at 400 °C in nitrogen for 30 min and then cooled to room temperature. The samples were reduced in a 10% H₂/Ar flow with a heating rate of 10 °C min⁻¹. Surface morphological images were obtained using scanning electron microscopy (SEM) (Hitachi S-4800) after gold plating at an accelerating voltage of 10 kV. Transmission electron microscopy (TEM) was performed using an FEI Tecnai G2 at an accelerating voltage of 200 kV. X-Ray Photoelectron Spectroscopy (XPS) was performed using a VG Thermo ESCALAB 250 spectrometer operated at 120 W.

Results and discussion

TEM analyses

The PVP-protected Pt NPs are dispersed uniformly with an average particle size of 3.2 nm (Fig. 2a). In the case of the FeOOH NRs (the precursor of the Fe₂O₃ NRs), each nanorod exhibits an average width of ca. 50 nm and a length of ca. 300 nm (Fig. 2b). As indicated in Fig. 2c, the FeOOH NRs convert into hollow Fe₂O₃ NRs after calcination. The Pt NPs are well dispersed and remain unchanged after loading on the Fe₂O₃ NRs (Fig. 2d). The loading content of Pt is 3.1 wt%, as determined using ICP analysis.

Fig. 2 TEM images of Pt NPs (a), FeOOH NRs (b), Fe₂O₃ NRs (c) and Pt/Fe₂O₃ (d). The inset in (a) shows the statistical Pt NP distribution and N indicates the total statistical number of Pt NPs.

As shown in Fig. 3a and b, growth of the Pt NPs is observed when the Pt/Fe₂O₃ samples are treated in 20% O₂/N₂ or 4% H₂/Ar. Concretely, the average size of the Pt NPs increases to 6.8 nm under an oxidative atmosphere and 14.6 nm under a reductive atmosphere. Meanwhile, the support Fe₂O₃ is destroyed to some extent, especially under a reductive atmosphere. In sharp contrast, a large amount of Pt NPs with small sizes appear after H–O treatment (Fig. 3c–h). The average sizes of the Pt NPs are 2.5, 1.1 and 2.8 nm after H–O treatment at 500, 600 and 700 °C, respectively. Particularly, few large Pt NPs could be observed in the Pt/Fe₂O₃(600 H–O) sample and the proportion of Pt NPs with sizes smaller than 1.5 nm reaches 80%. As for the fresh Pt/Fe₂O₃(IM) sample, the sizes of the Pt NPs are not uniform and range from 1 to 9 nm (Fig. S1†). After H–O treatment at 600 °C, most Pt NPs are around 2 nm in diameter, indicating that the redispersion behaviour of Pt on Fe₂O₃ applies to the common catalyst prepared through the impregnation method as well.

Fig. 3 TEM images of Pt/Fe₂O₃ samples after treatment under various conditions. (a) In 20% O₂/N₂ at 600 °C; (b) in 4% H₂/Ar at 600 °C; (c, d) in alternating H–O at 500 °C; (e, f), in alternating H–O at 600 °C; (g, h), in alternating H–O at 700 °C.

Catalytic performance for CO oxidation

Fig. 4 exhibits the catalytic activities of fresh Pt/Fe₂O₃ and Pt/Fe₂O₃ after different treatments. Compared to the fresh Pt/Fe₂O₃ catalyst, which could completely oxidize CO at 210 °C, the catalytic activities of catalysts after H–O treatment at different temperatures are improved to different degrees, and the temperatures required for complete conversion of CO (TCC) are 190 (500 H–O), 150 (600 H–O) and 180 °C (700 H–O). However, a reverse trend occurs for Pt/Fe₂O₃(600 O) and Pt/Fe₂O₃(600 H) and the TCC are 260 and 450 °C, respectively. The redispersion of Pt on the surface of Fe₂O₃ results in the production of a large amount of Pt NPs with smaller sizes, which means that there are more active centres for CO oxidation and the size effect of Pt is more significant.

Fig. 4 CO conversion as a function of reaction temperature for Pt/Fe₂O₃ catalysts before and after various treatments.

Redox properties

H₂-TPR is an effective method to test the reducibility of samples. For a fresh Fe₂O₃ sample (as shown in Fig. 5), the reduction peaks located around 415 and 600 °C are attributed to the reduction of Fe₂O₃ → Fe₃O₄ and Fe₃O₄ → FeO → Fe.²⁵ After the loading of Pt, the reduction temperatures of these two peaks shift to lower temperature ranges, indicating that the presence of Pt promotes the reduction of Fe₂O₃ significantly. A very weak reduction peak at about 140 °C is observed, which is primarily attributed to the reduction of Pt²⁺ to Pt⁰, along with some contribution from the partial reduction of iron oxide linked with the Pt species directly (i.e., Fe–O–Pt).²⁶ Additionally, the peak at 220 °C is assigned to the reduction of oxidized Pt species (e.g., Pt⁴⁺ to Pt⁰),²⁶ which partially overlaps with the reduction of Fe₂O₃ to Fe₃O₄ (at 270 °C). In the case of the Pt/Fe₂O₃ samples after H–O treatment, the low-temperature reduction peak centred at 140 °C disappears. Furthermore, the two reduction peaks (220 and 270 °C) integrate into a single one, the hydrogen consumption of which is remarkably less than the sum of the former two peaks, suggesting that the oxidation states of the Pt species decrease after H–O treatment. It should be noted that the single reduction peak shifts to a lower temperature region for Pt/Fe₂O₃(600 H–O), which is due to the fact that the small Pt NPs formed during H–O treatment at 600 °C could improve the spillover of H atoms from the Pt species to Fe₂O₃ more effectively.

Fig. 5 H₂-TPR profiles of Fe₂O₃, Pt/Fe₂O₃ and Pt/Fe₂O₃ after H–O treatment at different temperatures.

XRD characterization

Wide angle XRD patterns for FeOOH, Fe₂O₃, Pt/Fe₂O₃ and Pt/Fe₂O₃ after H–O treatment are shown in Fig. 6. For the Fe₂O₃ and Pt/Fe₂O₃ samples, only peaks ascribed to α-Fe₂O₃ (PDF#33-0664) are observed and no signals corresponding to Pt species are detected. This is presumably attributed to the fact that the Pt NPs in the measured sample are not large enough to generate a detectable signal or the diffraction peaks from Fe₂O₃ have covered the weak ones of Pt. Moreover, in the case of the Pt/Fe₂O₃ samples after H–O treatment, some new peaks located at approximately 30.4°, 43.3°, 53.6° and 57.2° appear, which are ascribed to γ-Fe₂O₃ (further confirmed by XPS analyses later), and the partial magnification patterns are presented in Fig. S2.† It is well known that the crystal phase transformation between α-Fe₂O₃ and γ-Fe₂O₃ could occur easily under reductive and oxidative atmospheres in the range from 400–600 °C.^27,28 The results suggest that on the surface of Fe₂O₃, continuous crystal phase transformation between α-Fe₂O₃ and γ-Fe₂O₃ happens during the H–O treatment, accompanied by the generation of a large amount of relative defects. Compared to Pt/Fe₂O₃(500 H–O) and Pt/Fe₂O₃(700 H–O), the diffraction intensity of γ-Fe₂O₃ for the Pt/Fe₂O₃(600 H–O) sample is much higher, indicating that more defects are generated when treated at 600 °C. Furthermore, as we know, the transformation between γ-Fe₂O₃ and Fe₃O₄ is quite easy under the experimental conditions due to their similar structure (i.e., Fe₃O₄ can transform to γ-Fe₂O₃ slowly in air, even at room temperature).^29–33 Both γ-Fe₂O₃ and Fe₃O₄ have the same inverse spinel structure, excluding that some oxygen ions in γ-Fe₂O₃ are substituted with oxygen vacancies in Fe₃O₄. Namely, the transformation between γ-Fe₂O₃ and Fe₃O₄ just refers to the storage and release of O atoms without any change in the basic crystal structure. Therefore, it is reasonable to speculate that the Fe₃O₄ phase is generated during H–O treatment and transforms to γ-Fe₂O₃ gradually in the later cooling process. As a result, the Fe₃O₄ phase disappears completely in the test samples.

Fig. 6 XRD patterns of FeOOH, Fe₂O₃, Pt/Fe₂O₃ and Pt/Fe₂O₃ after H–O treatment at different temperatures. ● and ♦ represent the diffraction peaks ascribed to α-Fe₂O₃ and γ-Fe₂O₃, respectively.

XPS analyses

To gain better insight into the nature of the crystal structure and oxidation states, the Fe 2p and Pt 4f XPS spectra of Pt/Fe₂O₃ samples before and after H–O treatment are displayed in Fig. 7 (the survey scans are given in Fig. S3†). For fresh Pt/Fe₂O₃, the XPS spectrum of Fe 2p exhibits two peaks at 710.4 and 723.7 eV,³⁴ which can be attributed to Fe 2p_3/2 and Fe 2p_1/2, respectively. Furthermore, a satellite peak located at 718.3 eV could also be observed and the binding energy between the satellite and Fe 2p_3/2 peak is 7.8 eV. The results above suggest that Fe is in a +3 oxidation state, which is consistent with the XRD tests. As reported in the literature,³⁵ the core level spectra of α-Fe₂O₃ and γ-Fe₂O₃ are almost identical. Therefore, the partial transformation from α-Fe₂O₃ to γ-Fe₂O₃ would not shift the binding energies significantly. Compared to α-Fe₂O₃ or γ-Fe₂O₃, the Fe 2p_3/2 and Fe 2p_1/2 peaks of Fe₃O₄ shift to high binding energies and broaden obviously due to the appearance of Fe²⁺. Moreover, for the Fe 2p peaks of Fe₃O₄, there is no satellite at about 719 eV.^33–35 In our experiments, after H–O treatment, although the binding energy in the Fe 2p spectra shifts to high values, the presence of a satellite peak located at 718.7 eV and the unchanged peak width at half height of Fe 2p indicate that the Fe₃O₄ phase does not exist in the measured samples and the newly-appearing diffraction peaks for Pt/Fe₂O₃(500–700 H–O) in the XRD patterns are attributed to γ-Fe₂O₃. The shift of the Fe 2p peaks could be ascribed to the strong interaction between the support and the small Pt NPs formed during the treatment. As for the spectra of Pt 4f, the peaks shift to a lower binding energy slightly after H–O treatment, suggesting that the oxidation states of Pt decrease, which is consistent with the TPR results. Besides, it is noteworthy that the intensity of Pt/Fe₂O₃(600 H–O) is lower than the others. It is inferred that the Pt NPs redisperse adequately under these circumstances and some of the small Pt NPs move into a crack in the support during the redispersion process, and cannot be detected using XPS measurements.³⁶

Fig. 7 XPS spectra of Fe 2p (a) and Pt 4f (b) for Pt/Fe₂O₃ samples before and after H–O treatment.

Discussion

On the basis of the above analysis, a possible process for the Pt NPs redispersion on Fe₂O₃ under alternating atmospheres of oxidation and reduction is proposed as follows: under an oxidative atmosphere, Pt–O–Fe or Pt–O–O–Fe forms on the interface of the Pt NPs and Fe₂O₃. During this course, the Pt-NPs are flattened to some extent and a part of the Pt atoms on the Pt NPs would be pulled onto the interface; under a reductive atmosphere, PtO is reduced to metallic Pt and a hydrogen spillover from Pt to Fe occurs, followed by the breaking of the Pt–O–Fe or Pt–O–O–Fe bonds. Meanwhile, related defects and oxygen vacancies are formed due to the crystal phase transformation of α-Fe₂O₃ to γ-Fe₂O₃ and the partial slight reduction of surface Fe₂O₃ through hydrogen spillover, respectively. Finally, Pt atoms are trapped at defects and oxygen vacancies. Furthermore, during H–O treatment, the transformation between γ-Fe₂O₃ and Fe₃O₄ happens continuously, accompanied by the storage and release of O atoms, just like the “oxygen storage material” ceria. Namely, the crystal transformation between γ-Fe₂O₃ and Fe₃O₄ can improve the oxygen mobility of catalysts, which is beneficial for the oxidation of CO and the redispersion of Pt.²²

In addition, the degrees of redispersion and sintering influenced by temperature are different. Sintering exists throughout the whole H–O treatment and would be more serious with an increase in temperature. Moreover, the H₂O generated during the H–O treatment can accelerate the sintering of Pt NPs (Fig. S4†). However, the energy required for the redispersion of Pt is sufficient when the treatment temperature reaches 600 °C and a subsequent increase in temperature has no significant influence on the further improvement of the redispersion degree. In detail, when treated at 500 °C, both the redispersion and sintering are weak, which is reflected by statistical analysis in that the size of Pt NPs with a maximum number is 1.75 nm (larger than 1.05 and 1.65 nm when treated at 600 and 700 °C, respectively) and the proportion of Pt NPs with sizes larger than 3.2 nm is just 14.7% (that would be as much as 21.5% when treated at 700 °C). During H–O treatment at 600 °C, although sintering is aggravated to some extent, redispersion is improved enormously, resulting in redispersion dominating the treatment process. It is speculated that compared with treatment at 500 °C, the formation and breakup of Pt–O–Fe or Pt–O–O–Fe bonds are much easier at 600 °C under oxidative and reductive atmospheres, respectively. When the treatment temperature reaches 700 °C, sintering aggravates greatly and the proportion of Pt NPs with sizes larger than 3.2 nm reaches 21.5% as mentioned above, which is much larger than 3% according to the statistical results of samples treated at 600 °C. Moreover, the rising temperature could not further promote the degree of redispersion obviously, leading to the fact that the size of Pt NPs with maximum number is 1.65 nm (larger than 1.05 nm when treated at 600 °C). Therefore, it could be seen that at 600 °C, sintering is not so serious and almost “overcome” by redispersion during the treatment process, leading to the final result that the proportions of Pt NPs with sizes smaller than 1.5 and 3.2 nm are 81.1% and 98.2%, respectively.

Conclusions

In this work, the redispersion behaviour of Pt on the surface of Fe₂O₃ was observed and studied. After treatment under alternating oxidation and reduction atmospheres at 600 °C, the average size of the Pt NPs decreased from the original 3.2 nm to 1.1 nm. Besides, we investigated the influence of treatment temperature on the redispersion of Pt and found that the optimal treatment temperature for redispersion was 600 °C. Based on a variety of characterizations, it is proposed that the Pt–O–Fe bonds formed under an oxidative atmosphere and related defects or oxygen vacancies formed under a reductive atmosphere are key factors for Pt redispersion on the surface of Fe₂O₃. We hope that our work can be helpful for research on anti-sintering, the regeneration of deactivated catalyst and “tuning selectivity” through changing the dispersion of noble metals.^{15,16,18,37–39}

Acknowledgements

The authors thank the National Natural Science Foundation of China. This work was supported by the National Natural Science Foundation of China (21273221).

Notes and references

1. L. F. Liotta, Appl. Catal., B, 2010, 100, 403–412.
2. X. Wang, C. Wang, L. Cheng, S. T. Lee and Z. Liu, J. Am. Chem. Soc., 2012, 134, 7414–7422.
3. C. Chen, C. Nan, D. Wang, Q. Su, H. Duan, X. Liu, L. Zhang, D. Chu, W. Song, Q. Peng and Y. Li, Angew. Chem., Int. Ed., 2011, 50, 3725–3729.
4. V. P. Santos, S. A. C. Carabineiro, P. B. Tavares, M. F. R. Pereira, J. J. M. Órfão and J. L. Figueiredo, Appl. Catal., B, 2010, 99, 198–205.
5. G. Li and L. Li, RSC Adv., 2015, 5, 36428–36433.
6. Z. Miao, T. Wu, J. Li, T. Yi, Y. Zhang and X. Yang, RSC Adv., 2015, 5, 19823–19829.
7. Z. Miao, Y. Zhang, X. Pan, T. Wu, B. Zhang, J. Li, T. Yi, Z. Zhang and X. Yang, Catal. Sci. Technol., 2015, 5, 1314–1322.
8. H. Hong, L. Hu, M. Li, J. Zheng, X. Sun, X. Lu, X. Cao, J. Lu and H. Gu, Chem. - Eur. J., 2011, 17, 8726–8730.
9. A. M. Pasqualeti, P.-Y. Olu, M. Chatenet and F. H. B. Lima, ACS Catal., 2015, 5, 2778–2787.
10. Z. Zhang, Y. Jiang, M. Chi, Z. Yang, C. Wang and X. Lu, RSC Adv., 2015, 5, 94456–94461.
11. F. Diehl, J. Barbier, D. Duprez, I. Guibard and G. Mabilon, Appl. Catal., A, 2015, 504, 37–43.
12. M. O. Kazakov, A. V. Lavrenov, M. S. Mikhailova, N. A. Allert, T. I. Gulyaeva, I. V. Muromtsev, V. A. Drozdov and V. K. Duplyakin, Kinet. Catal., 2010, 51, 438–443.
13. M. O. Kazakov, A. V. Lavrenov, I. G. Danilova, O. B. Belskaya and V. K. Duplyakin, Kinet. Catal., 2011, 52, 573–578.
14. M. O. Kazakov, A. V. Lavrenov, O. B. Belskaya, I. G. Danilova, A. B. Arbuzov, T. I. Gulyaeva, V. A. Drozdov and V. K. Duplyakin, Kinet. Catal., 2012, 53, 101–106.
15. K. Morgan, A. Goguet and C. Hardacre, ACS Catal., 2015, 5, 3430–3445.
16. K. Morgan, R. Burch, M. Daous, J. J. Delgado, A. Goguet, C. Hardacre, L. A. Petrov and D. W. Rooney, Catal. Sci. Technol., 2014, 4, 729.
17. J. Sa, A. Goguet, S. F. Taylor, R. Tiruvalam, C. J. Kiely, M. Nachtegaal, G. J. Hutchings and C. Hardacre, Angew. Chem., Int. Ed., 2011, 50, 8912–8916.
18. K. Morgan, R. Burch, M. Daous, J. J. Delgado, A. Goguet, C. Hardacre, L. A. Petrov and D. W. Rooney, Catal. Struct. React., 2015, 1, 120–124.
19. H. Lieske and J. Voelter, J. Phys. Chem., 1985, 89, 1841–1842.
20. R. M. J. Fiedorow and S. E. Wanke, J. Catal., 1976, 43, 34–42.
21. R. M. J. Fiedorow, B. S. Chahar and S. E. Wanke, J. Catal., 1978, 51, 193–202.
22. T. Wu, X. Pan, Y. Zhang, Z. Miao, B. Zhang, J. Li and X. Yang, J. Phys. Chem. Lett., 2014, 5, 2479–2483.
23. Y. Nagai, N. Takagi, Y. Ikeda, K. Dohmae, T. Tanabe, G. Guilera, S. Pascarelli, M. Newton, H. Shinjoh and S. i. Matsumoto, AIP Conf. Proc., 2007, 882, 594–596.
24. M. J. D’Aniello Jr, D. R. Monroe, C. J. Carr and M. H. Krueger, J. Catal., 1988, 109, 407–422.
25. G. Wang, H. Lian, W. Zhang, D. Jiang and T. Wu, Kinet. Catal., 2002, 43, 433–442.
26. N. An, Q. Yu, G. Liu, S. Li, M. Jia and W. Zhang, J. Hazard. Mater., 2011, 186, 1392–1397.
27. Y. El Mendili, J.-F. Bardeau, N. Randrianantoandro, F. Grasset and J.-M. Greneche, J. Phys. Chem. C, 2012, 116, 23785–23792.
28. X. Zhang, Y. Niu, X. Meng, Y. Li and J. Zhao, CrystEngComm, 2013, 15, 8166–8172.
29. F. Jiao, J.-C. Jumas, M. Womes, A. V. Chadwick, A. Harrison and P. G. Bruce, J. Am. Chem. Soc., 2006, 128, 12905–12909.
30. L. Zhong, J. Hu, H. Liang, A. Cao, W. Song and L. Wan, Adv. Mater., 2006, 18, 2426–2431.
31. Y. Chueh, M. Lai, J. Liang, L. Chou and Z. Wang, Adv. Funct. Mater., 2006, 16, 2243–2251.
32. I. Mitov, Z. Cherkezova-Zheleva and V. Mitrov, Phys. Status Solidi A, 1997, 161, 475–482.
33. Y. Chen, G. Xiao, T. Wang, Q. Ouyang, L. Qi, Y. Ma, P. Gao, C. Zhu, M. Cao and H. Jin, J. Phys. Chem. C, 2011, 115, 13603–13608.
34. J. Lu, X. Jiao, D. Chen and W. Li, J. Phys. Chem. C, 2009, 113, 4012–4017.
35. G. Sun, B. Dong, M. Cao, B. Wei and C. Hu, Chem. Mater., 2011, 23, 1587–1593.
36. A. S. Ivanova, E. M. Slavinskaya, O. A. Stonkus, R. V. Gulyaev, T. S. Glazneva, A. S. Noskov and A. I. Boronin, Catal. Sci. Technol., 2015 10.1039/c5cy01588j.
37. J. Okal, H. Kubicka, L. K. epiński and L. Krajczyk, Appl. Catal., A, 1997, 162, 161–169.
38. A. Monzón, T. F. Garetto and A. Borgna, Appl. Catal., A, 2003, 248, 279–289.
39. A. K. Aboul-Gheit, A. A. Aboul-Enein, A. E. Awadallah, S. A. Ghoneim and E. A. Emam, Chin. J. Catal., 2010, 31, 1209–1216.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra01803c

---