Alkylation of benzene with methanol over hierarchical porous ZSM-5: synergy effects of hydrogen atmosphere and zinc modification

Abstract

The competitive reaction of methanol to olefins is difficult to be suppressed in benzene alkylation with methanol over hierarchical porous ZSM-5. The influence of ZnO content and different atmospheres on the catalytic performance of hierarchical porous ZSM-5 catalyst was investigated. The results indicated that the introduction of ZnO could form the Lewis acid sites of zinc species (ZnOH⁺) at the expense of the Brönsted acid sites, and the reduction of strong Brönsted acid would help to suppress the side reaction of methanol to olefins. However, the presence of ZnOH⁺ could catalyze the dehydrogenation reaction of light hydrocarbons to olefins which would result in the formation of coke under the nitrogen atmosphere, while the hydrogen atmosphere could inhibit the dehydrogenation ability of ZnOH⁺.

---

1. Introduction

Xylene, an important chemical intermediate, is widely used for the production of fine chemicals.^1–4 Currently, the alkylation of benzene with methanol to produce xylene has attracted a lot of attention due to the decline in crude oil resources.^1,5 While using ZSM-5 zeolite as the catalyst, low conversion of benzene and low selectivity to xylene were observed due to the diffusion limitation in the micropores.^6–8 In order to enhance reagent and product diffusion to and from active sites, mesopores were introduced into ZSM-5 zeolites and excellent catalytic performance was exhibited in benzene alkylation reactions.⁸ Nevertheless, the side reaction of methanol to olefins was difficult to be avoided even over hierarchical porous ZSM-5. Moreover, the formed olefins could lead to the formation of ethylbenzene which was difficult to be separated from xylene.^7–10 Although Pt modified hierarchical porous ZSM-5 could successfully suppress the formation of ethylbenzene by the hydrogenation of ethylene to ethane, the utilization efficiency of methanol was decreased due to that ethylene which was converted from methanol could not further convert to alkylaromatics.^6,7 Obviously, suppressing the side reaction of methanol to olefins was the fundamental method to solve the problem of ethylbenzene formation without losing methanol utilization efficiency.

According to the literatures, the catalytic selectivity and activity of ZSM-5 for the aromatics alkylation mainly depend on the acidity of catalyst, and the zeolites with less Brönsted acidity could help to minimize the side reaction of methanol.^11,12 ZSM-5 catalyst modified with oxides such as MgO, ZnO, P₂O₅, etc. could effectively regulate the acidity, which in turn influenced the reaction activity of catalyst.^13–16 For example, Tan et al. reported that the multiple modification of ZSM-5 with SiO₂, P₂O₅ and MgO in a suitable sequence could efficiently eliminate external surface acid sites and then enhance the selectivity of para-xylene.¹³ Ding et al. observed that Mg (1 wt%) modified mesopores catalyst exhibited the highest conversion of benzene and selectivity to ethylbenzene in the alkylation of benzene with ethanol, which was confirmed to be due to the proper L/B acid proportion.¹⁴ Specially, Zhang et al. reported that as compared to MgO, the introduction of ZnO could effectively improve the activity of conventional ZSM-5, indicating that the acid strength and acid sites adjusted by zinc salt was favorable for benzene alkylation with methanol.¹⁵ This prompts us to investigate the effect of ZnO modification on the catalytic performance of hierarchically porous ZSM-5 in benzene alkylation with methanol.

In present work, the catalytic performance of ZnO modified hierarchical porous ZSM-5 in benzene alkylation with methanol was investigated. The textural properties and acidity of the catalysts were systematically characterized by various techniques (including XRD, BET, XPS, NH₃-TPD, Py-IR and TG) and the conversion of benzene and the selectivity to products versus ZnO content were also investigated. We observed that the introduction of appropriate ZnO content could suppress the side reaction of methanol to olefins by reducing the Brönsted acidity of ZSM-5. However, the implementation of this effect strongly depended on the reaction atmosphere.

2. Experimental

2.1. Catalyst preparation

Hierarchical porous ZSM-5 was prepared via solvent evaporation assisted dry-gel route which was reported by the authors.^7,17 The molar ratio of the reaction mixture was as follows: nSiO₂ : nAl₂O₃ : nTPAOH : nHTS : nEtOH = 1 : 0.0028 : 0.2 : 0.05 : 15. Firstly, aluminum isopropoxide (AIP), tetra-n-propylammonium hydroxide (TPAOH) and EtOH were mixed together and stirred until AIP was completely dissolved. Then tetraethylorthosilicate (TEOS), hexadecyltrimethoxysilane (HTS) and EtOH were added to the solution and stirred briefly to obtain a clear gel. The gel was solidified in a Petri dish before being transferred into a PTFE cup and then the cup was moved into a Teflon-lined autoclave. A small amount of water was added outside the cup to create steam for the hydrothermal synthesis conditions. The synthesis temperature and time were 180 °C and 72 h, respectively. The resulting powder was first dried at 110 °C overnight, without ion-exchange steps, and then calcined at 550 °C for 7 h (heating rate of 10 °C min⁻¹) to obtain H-form hierarchical porous ZSM-5.

ZnO modified hierarchical porous ZSM-5 catalysts with a nominal ZnO loading of 1 to 4 wt% were prepared by impregnating method. The hierarchical porous ZSM-5 was firstly tableted, crushed and screened to the particles with diameters of 0.45–0.90 mm. The catalyst was then impregnated with quantitative amount of aqueous solution of Zn(NO₃)₂·6H₂O at ambient temperature for 24 h before dried at 110 °C overnight. The dried sample was later calcined in air at 550 °C for 7 h (heating rate of 10 °C min⁻¹) to obtain the ZnO modified catalyst.

2.2. Catalyst characterization

X-ray diffraction (XRD) (SCINTAG X” TRA) was operated using Cu K-radiation (1542 A) at 30 mA and 40 kV high voltage source with scanning angle (2θ) from 5 to 80°. XPS analysis was performed by Kratos AXIS Ultra DLD spectrometer with monochromatized aluminum X-ray source (1486.6 eV) and the pass energy of 40 eV. The pressure in the sample analysis chamber was lower than 6 × 10⁻⁹ Torr during data acquisition. The specific surface area of the samples was measured using N₂ adsorption–desorption at 77 K with a Micromeritics ASAP 2020 instrument and was calculated employing the BET method. The thermogravimetric analysis (TG) was evaluated by Netzsch STA 449 C apparatus using a temperature ramp of 30–800 °C with a heating rate of 10 °C min⁻¹ in oxygen atmosphere. NH₃-TPD was carried out using QIC characterization system equipped with mass spectra detector (Hiden, England). The sample was evacuated at 500 °C for 1 h in the flow of argon and then cooled down to 50 °C. The adsorption of NH₃ was performed with the flow of 10 vol% NH₃/Ar at 50 °C. The desorption was conducted between 50 and 650 °C with a heating rate of 10 °C min⁻¹. Fourier transform infrared (FT-IR) spectra were measured on a Nexus FT-IR spectrometer. The sample was firstly pressed into a wafer, and then heat-treated in He (H₂) flow (20 mL min⁻¹) at 400 °C for 1 h. The IR spectra were recorded at room temperature. For pyridine adsorption (Py-IR), pyridine vapour was introduced into the cell at room temperature for 1 h; the spectra were then recorded after evacuation at 400 °C for 1 h. The concentrations of Brönsted acid and Lewis acid were calculated by the procedures reported by Emeis et al.¹⁸

2.3. Catalytic activity test

All evaluation experiments were carried out in a continuous-flow fixed-bed reactor with a stainless steel tube (8 mm i.d.) at atmospheric pressure. In each test, 0.5 g catalyst diluted with 5.0 g inert quartz sand were loaded in the middle of the tube reactor. In order to investigate the effect of carrier gas on the catalytic performance of catalyst, the catalyst modified with the same content of ZnO was divided into two parts and tested in hydrogen atmosphere and nitrogen atmosphere, respectively. The catalyst was heat-treated in situ from ambient temperature to 400 °C at 10 °C min⁻¹ and maintained at 400 °C for 1 h in H₂ (N₂) flow with a space velocity of 2400 h⁻¹. Then benzene and methanol mixture (B/M molar ratio = 1 : 1) were fed into the reactor (WHSV = 2.0 h⁻¹) at 400 °C with a co-feed H₂ (N₂) flow of 40 mL min⁻¹. The effluent from the reactor was analyzed online by the gas chromatography (Fuli GC9790) with a DB-1 capillary column (30 m × 0.25 mm × 1.00 μm) and a flame ionization detector. In order to ensure all of the products were in gas phase, the temperature of the effluent line was maintained at 200 °C by heating belt.

3. Results and discussion

3.1. Catalyst characterization and tests

Fig. 1 presents XRD patterns of unmodified and ZnO modified hierarchical porous ZSM-5 catalysts. Both unmodified and modified catalysts displayed two distinct diffraction peaks in 8–10° and 20–25° (2θ) ranges which correspond to the typical characteristic pattern of MFI structure, indicating that the crystals of hierarchical porous ZSM-5 were preserved during the modification. In addition, no diffraction peaks corresponding to ZnO (2θ = 31.6, 34.2, 36.1 and 56.6°)¹⁹ were detected for modified catalysts, suggesting that ZnO was well dispersed on the surface of ZSM-5 zeolites.

Fig. 1 XRD patterns of unmodified and ZnO modified hierarchical porous ZSM-5 catalysts.

The catalytic performance of ZnO modified ZSM-5 was investigated in benzene alkylation with methanol under N₂ and H₂ atmospheres, respectively (Fig. 2). It can be seen that with the increase of ZnO content from 0 to 1 wt%, the conversion of benzene increased from 51.6 to 53.6% in H₂ and to 51.8% in N₂. This result indicated that the modification of hierarchical porous ZSM-5 with ZnO could indeed improve the catalytic activity, and the atmospheres had different influence. Further increasing the amount of ZnO to 4 wt% led to the decrease of benzene conversion under both H₂ and N₂ atmospheres, which suggested that excess ZnO was adverse for the alkylation of benzene with methanol. Therefore, 1 wt% of ZnO was considered as a good loading for the modification of hierarchical porous ZSM-5 and was used in the following experiments. In order to fully understand the influence of atmosphere, the conversion of benzene and selectivity to products over the catalysts under different atmospheres were carefully examined (Fig. 3). It was clear that the change of carrier gas (H₂ and N₂) had no obvious effect on the performance of unmodified sample. While for modified catalyst under hydrogen atmosphere, the conversion of benzene and the selectivity to xylene were both increased as compared to those of unmodified catalyst, but the similar phenomenon was not observed under nitrogen atmosphere. It should be noted that modified catalyst could suppress the formation of ethylbenzene under both atmospheres. In general, the increase of benzene conversion meant the alkylation of aromatic with methanol was promoted and the side reaction of methanol to olefins was suppressed. It has been recognized that ethylene, the main constituent of olefins, is the key reactant for the formation of ethylbenzene.^6,20 Therefore, the suppression of ethylbenzene formation on ZnO modified ZSM-5 under hydrogen atmosphere might be attributed to the inhibition of the reaction of methanol to olefins. However, the benzene conversion was not increased under nitrogen atmosphere and it was hard to explain how the formation of ethylbenzene was suppressed. Thus, it was necessary to compare the total content of light hydrocarbons in the products on corresponding catalysts.

Fig. 2 Catalytic performance of ZnO modified catalysts in benzene alkylation with methanol under different atmospheres.

Fig. 3 The influence of atmosphere on the catalytic performance of unmodified and ZnO modified catalysts in benzene alkylation with methanol ((a) 0 wt% ZnO, H₂; (b) 0 wt% ZnO, N₂; (c) 1 wt% ZnO, H₂; (d) 1 wt% ZnO, N₂).

As shown in Fig. 4, for unmodified catalyst, the atmosphere did not exhibit significant impact on the content of light hydrocarbons. For ZnO modified ZSM-5, the total content of light hydrocarbons was decreased under either nitrogen or hydrogen atmosphere. However, we could not directly conclude that the side reaction of methanol to olefins was suppressed under both atmospheres due to that olefins could further convert to coke via polymerization.²⁰ TG-DTA was performed to evaluate the coke content (removed within 200 to 600 °C) on the catalysts. As seen in Fig. 5, it was clear that the atmosphere did not have significant influence on the content of coke over unmodified catalyst. For ZnO modified ZSM-5, the coke content (2.1%) under hydrogen atmosphere was similar to that of unmodified catalyst, indicating that the total content of light hydrocarbons was actually decreased and the side reaction of methanol to olefins was suppressed. However, the coke content (6.26%) under nitrogen atmosphere was much higher than that of the catalyst with same ZnO content under hydrogen atmosphere, suggesting that a large number of olefins were converted to coke. Considering the suppression of ethylbenzene formation and the large amounts of coke on modified catalyst under nitrogen atmosphere, it was reasonable to conclude that the promotion of olefins to coke also could help to suppress the formation of ethylbenzene. It was worth noting that the coke would lead to the deactivation of catalyst. As expected, the catalytic activity of ZnO modified catalyst under nitrogen atmosphere decreased quickly (Fig. 6).

Fig. 4 The content of light hydrocarbons on unmodified and ZnO modified catalysts under different atmospheres ((a) 0 wt% ZnO, H₂; (b) 0 wt% ZnO, N₂; (c) 1 wt% ZnO, H₂; (d) 1 wt% ZnO, N₂).

Fig. 5 TG-DTA profiles of unmodified and ZnO modified catalysts after successive reaction time (10 h).

Fig. 6 The stability of unmodified and ZnO modified catalysts under different atmospheres.

In conclusion, the hierarchical porous ZSM-5 catalyst modified with appropriate content of ZnO could improve the catalytic activity for the alkylation of benzene with methanol. However, the atmosphere had a significant effect on the performance of modified catalyst. For ZnO modified ZSM-5, the suppression of side reaction of methanol to olefins was observed under hydrogen atmosphere and the promotion of coke formation was exhibited under nitrogen atmosphere. The mechanism of how atmosphere influences the performance of ZnO modified catalyst will be discussed in details in the next section.

3.2. Mechanism of the effect of atmosphere on the performance of ZnO modified catalyst

According to the literatures, the introduction of zinc to ZSM-5 could exhibit significant influence on the distribution of acid sites across their acidic strength.^19,21 Specifically, Niu et al. reported that the introduction of zinc into ZSM-5 could form Lewis acid sites of zinc species (ZnOH⁺) at the expense of the silanol hydroxyl and Brönsted acidity.²¹ In order to confirm the change of acid sites and acidic strength, NH₃-TPD (Fig. 7) and Py-IR (Fig. 8) were performed. In general, the NH₃-TPD spectrum of ZSM-5 zeolites has two main types of desorption peaks, named low-temperature peak (located below 200 °C) and high-temperature peak (situated above 300 °C), which are ascribed to the ammonia that had been adsorbed onto the weak and strong acid sites, respectively.²² The amounts of ammonia desorbed from the catalyst surface can be shown via TPD peak areas. The quantities of strong, medium and weak acid sites were measured by the amounts of ammonia desorbed at 300–550, 200–300 and 120–200 °C, respectively. As compared with the unmodified ZSM-5 (Fig. 7), the amount of medium acid sites on ZnO modified ZSM-5 increased at the expense of the strong acid sites which was consistent with the result reported by Niu et al.²¹ Moreover, Py-IR (Fig. 8) showed the change of Brönsted acid sites (1545 cm⁻¹) and Lewis acid sites (1454 cm⁻¹) (expressed by B/L). It can be observed that with the increase of ZnO content, the ratio of B/L decreased significantly, indicating that the introduction of zinc into ZSM-5 formed Lewis acid sites of zinc species (ZnOH⁺) at the expense of Brönsted acid sites. Considering the change in the quantity of strong acid sites and in B/L ratio, we could conclude that the modification of ZSM-5 with ZnO could effectively reduce the quantity of strong Brönsted acid sites. In our work, the reaction of methanol to olefins and the alkylation of aromatics were both catalyzed by Brönsted acid. Moreover, Adebajo et al. reported that reducing Brönsted acid sites would help to minimize the side reaction of methanol and enhance the benzene methylation with methanol.²³ Therefore, the suppression of the side reaction of methanol to olefins over modified ZSM-5 could be attributed to the reduced quantity of strong Brönsted acid sites.

Fig. 7 NH₃-TPD profiles of unmodified and ZnO modified catalysts.

Fig. 8 FT-IR spectra of pyridine adsorption on unmodified and ZnO modified catalysts: (a) 0 wt% ZnO, (b) 1 wt% ZnO, (c) 2 wt% ZnO, (d) 3 wt% ZnO, and (e) 4 wt% ZnO.

It was worth noting that although the modified catalysts were with the identical content of ZnO, the atmosphere exhibited a profound effect on the performance of the catalyst, indicating that the atmosphere might have different influence on how the introduced zinc species affect the reaction. As reported in literatures,^21,24–28 Zn–ZSM-5 zeolites prepared by impregnation method will contain (i) ZnOH⁺ ions located on Brönsted sites, (ii) (Zn–O–Zn)²⁺ species formed from the condensation of two ZnOH⁺, (iii) (O⁻–Zn²⁺–O⁻) species in which Zn²⁺ cations replaced two protons and interacted with two Al sites bridged by oxygen, (iv) more bulky intrazeolite or extrazeolite clusters of zinc oxide. However, ZSM-5 zeolites used in our reaction contained high Si/Al ratios which was unfavourable to the formation of (Zn–O–Zn)²⁺.²⁹ In addition, if zinc was presented as (O⁻–Zn²⁺–O⁻) species, the overall ammonia uptake on Zn modified ZSM-5 would decrease significantly due to that one zinc atom concerned with two Al sites.²⁵ According to Fig. 7, the overall peak area of unmodified catalyst was close to that of ZnO modified catalyst (1 wt%), indicating that the overall ammonia uptake on the samples was similar and the presence of (O⁻–Zn²⁺–O⁻) species could be eliminated. XPS spectra of Zn (2p_3/2) of 1 wt% ZnO modified ZSM-5 was measured (Fig. 10), and the binding energies around 1023.4 and 1022.8 eV correspond to ZnOH⁺ and ZnO were detected.²¹ Therefore, it was reasonable to conclude that the possible state of zinc in Zn modified ZSM-5 was ZnOH⁺ and ZnO. Studies have pointed out that as for ZnO species, the main effect was to narrow the pore size of ZSM-5 due to that the alkylation of benzene and the side reaction of methanol to olefins were catalyzed by Brönsted acid sites.^12,21 In our research, the influence of ZnO on the pore size of ZSM-5 between the catalysts with the same content of ZnO was similar. Moreover, the reduction of ZnO was difficult in hydrogen atmosphere at 400 °C.³⁰ Therefore, the presence of ZnO specie could not explain the effect of the atmosphere on the performance of the catalyst.

Fig. 9 FT-IR spectra of pyridine adsorption on 1 wt% ZnO modified catalysts: (a) 400 °C, N₂ treated; (b) 400 °C, H₂ treated.

Fig. 10 XPS spectra of Zn (2p_3/2) of 1 wt% ZnO/ZSM-5 zeolites.

As compared to ZnO species, the formation of ZnOH⁺ species could reduce the Brönsted acid sites which in turn would suppress the side reaction of methanol to olefins.^12,21 In addition, Triwahyono et al. had reported the formation of coke on Zn–ZSM-5 in nitrogen stream might be attributed to the strong Lewis acid sites being the active sites for the formation of dehydrogenated carbonaceous species.³¹ We noted that the ZnOH⁺ species were introduced as Lewis acid sites in ZSM-5 and widely used in alkane aromatization to promote the dehydrogenation of hydrocarbons.^21,32 This reminded us that the effect of atmosphere might be related to the presence of ZnOH⁺ species.

In order to verify this relationship, the stability of ZnOH⁺ under different atmospheres was investigated. Py-IR (Fig. 9) showed the change of Brönsted acid sites and Lewis acid sites (expressed by B/L) of modified catalysts which were heat-treated under different atmospheres. It can be seen that after treating ZnO modified catalyst in hydrogen atmosphere at 400 °C (same as reaction temperature), the ratio of Brönsted acid to Lewis acid was remained, indicating that ZnOH⁺ was not reduced in H₂ at 400 °C and would be stable under reaction conditions. This result was consistent with the finding that the reduction temperature required for Zn²⁺ on ZSM-5 (Zn content <2 wt%) prepared by impregnation method was above 500 °C.²⁸ Thus, we could conclude that the introduced zinc species (ZnOH⁺) could remain in the hydrogen atmosphere. According to the literatures, ZnOH⁺ species was active for the dehydrogenation of light hydrocarbons to olefins, and coke was formed from unsaturated species: alkenes, cyclic alkenes and cyclic dienes, but not from alkanes.^21,33–35 In the alkylation of aromatics, a lot amount of alkenes and alkanes would be formed by the side reaction of methanol to olefins and the subsequent hydrogen transfer reactions.^11,21 Therefore, the dehydrogenation of light hydrocarbons to olefins on ZnOH⁺ species would promote the formation of coke via the polymerization of olefins and this well explained the higher coke content on the ZnO modified ZSM-5 under nitrogen atmosphere. However, the regeneration of the active sites (ZnOH⁺ species) for dehydrogenation was based on the desorption of H₂, the hydrogen atmosphere might inhibit the desorption of H₂ which in turn led to the deactivation of ZnOH⁺.^36,37 At this point, the reduction of strong Brönsted acid suppressed the side reaction of methanol to olefins and promoted the alkylation of benzene.

4. Conclusions

The alkylation of benzene with methanol was investigated over unmodified and ZnO modified hierarchical porous ZSM-5 in a fixed-bed reactor under nitrogen and hydrogen atmosphere. The synergy effects of hydrogen atmosphere and zinc modification was observed, in which the introduction of ZnO could suppress the side reaction of methanol to olefins by the reduction of strong Brönsted acid and the hydrogen atmosphere could inhibit the formation of coke by suppressing the dehydrogenation of light hydrocarbons to olefins on ZnOH⁺.

Acknowledgements

We acknowledge financial support from the National Natural Science Foundation of China (NSFC-21476207) and National Basic Research Program of China (973 Program) (no. 2011CB710800).

References

1. X. M. Wang, X. Jun, G. D. Qi, B. J. Li, C. Wang and F. Deng, J. Phys. Chem., 2013, 117, 4018–4023.
2. W. Alabi, L. Atanda, R. Jermy and S. Al-Khattaf, Chem. Eng. J., 2012, 195–196, 276–288.
3. T. Y. Yan, Ind. Eng. Chem. Res., 1989, 28, 572–576.
4. M. P. M. Nicolau, P. S. Bárcia, J. M. Gallegos, J. A. C. Silva, A. E. Rodrigues and B. L. Chen, J. Phys. Chem. C, 2009, 113, 13173–13179.
5. F. Schmidt, C. Hoffmann, F. Giordanino, S. Bordiga, P. Simon, W. Carrillo-Cabrera and S. Kaskel, J. Catal., 2013, 307, 238–245.
6. H. L. Hu, Q. F. Zhang, J. Cen and X. N. Li, Catal. Commun., 2014, 57, 129–133.
7. H. L. Hu, Q. F. Zhang, J. Cen and X. N. Li, Catal. Lett., 2015, 145, 715–722.
8. W. Deng, X. He, C. Zhang, Y. Gao, X. D. Zhu, K. K. Zhu, Q. S. Huo and Z. J. Zhou, Chin. J. Chem. Eng., 2014, 22, 921–929.
9. V. Finsy, H. Verelst, L. Alaerts, D. D. Vos, P. A. Jacobs, G. V. Baron and J. F. M. Denayer, J. Am. Chem. Soc., 2008, 130, 7110–7118.
10. T. C. Tsai, I. Wang, G. K. Huang and S. D. Liu, Appl. Catal., A, 2007, 321, 125–134.
11. Z. R. Zhu, Q. L. Chen, Z. K. Xie, W. M. Yang and C. Li, Microporous Mesoporous Mater., 2006, 88, 16–21.
12. M. O. Adebajo and M. A. Long, Catal. Commun., 2003, 4, 71–76.
13. W. Tan, M. Liu, Y. Zhao, K. K. Hou, H. Y. Wu, A. F. Zhang, H. Liu, Y. R. Wang, C. S. Song and X. W. Guo, Microporous Mesoporous Mater., 2014, 196, 18–30.
14. W. Ding, Y. Y. Cui, J. J. Li, Y. Q. Yang and W. P. Fang, RSC Adv., 2014, 4, 50123–50129.
15. Y. Y. Zhang, Y. F. Li, L. Chen, C. T. Au and S. F. Yin, Catal. Commun., 2014, 54, 6–10.
16. S. Phatanasri, P. Praserthdam and T. Punsupsawat, Korean J. Chem. Eng., 2000, 17, 414–419.
17. K. K. Zhu, J. M. Sun, J. Liu, L. Q. Wang, H. Y. Wan, J. Z. Hu, Y. Wang, C. H. F. Peden and Z. M. Nie, ACS Catal., 2011, 1, 682–690.
18. C. A. Emeis, J. Catal., 1993, 141, 347–354.
19. J. H. Gao, L. D. Zhang, J. X. Hu, W. H. Li and J. G. Wang, Catal. Commun., 2009, 10, 1615–1619.
20. Y. Zhao, W. Tan, H. Y. Wu, A. F. Zhang, M. Liu, G. M. Li, X. S. Wang, C. S. Song and X. W. Guo, Catal. Today, 2011, 160, 179–183.
21. X. J. Niu, J. Gao, Q. Miao, M. Dong, G. F. Wang, W. B. Fan, Z. F. Qin and J. G. Wang, Microporous Mesoporous Mater., 2014, 197, 252–261.
22. Y. Ni, A. Sun, X. Wu, G. Hai, J. Hu, T. Li and G. Li, Microporous Mesoporous Mater., 2011, 143, 435–442.
23. M. O. Adebajo, R. F. Howe and M. A. Long, Catal. Today, 2000, 63, 471–478.
24. A. L. Yakovlev, A. A. Shubin, G. M. Zhidomirov and R. A. Santen, Catal. Lett., 2000, 70, 175–181.
25. J. A. Biscardi, G. D. Meitzner and E. Iglesia, J. Catal., 1998, 179, 192–202.
26. J. A. Biscardi and E. Iglesia, Catal. Today, 1996, 31, 207–231.
27. H. Berndt, G. Lietz and J. Völter, Appl. Catal., A, 1996, 146, 365–379.
28. S. M. T. Almutairi, B. Mezari, P. C. M. M. Magusin, E. A. Pidko and E. J. M. Hensen, ACS Catal., 2012, 2, 71–83.
29. V. B. Kazansky, A. I. Serykh, B. G. Anderson and R. A. Santen, Catal. Lett., 2003, 88, 211–217.
30. A. K. Aboul-Gheit, A. E. Awadallah, A. A. Aboul-Enein and A. H. Mahmoud, Fuel, 2011, 10, 3040–3046.
31. S. Triwahyono, A. A. Jalil, R. R. Mukti, M. Musthofa, N. A. M. Razali and M. A. A. Aziz, Appl. Catal., A, 2011, 407, 91–99.
32. Y. G. Kolyagin, V. V. Ordomsky, Y. Z. Khimyak, A. I. Rebrov, F. Fajula and I. I. Ivanov, J. Catal., 2006, 238, 122–133.
33. H. A. Aleksandrov, G. N. Vayssilov and N. Rösch, J. Mol. Catal. A: Chem., 2006, 256, 149–155.
34. G. Q. Zhang, T. Bai, T. F. Chen, W. T. Fan and X. Zhang, Ind. Eng. Chem. Res., 2014, 53, 14932–14940.
35. M. Guisnet and P. Magnoux, Appl. Catal., A, 2001, 212, 83–96.
36. E. A. Pidko and R. A. Santen, J. Phys. Chem. C, 2007, 111, 2643–2655.
37. Y. Y. Sara, J. A. Biscardi and E. Iglesia, J. Phys. Chem. B, 2002, 106, 9642–9648.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra03048j

---