F-assisted synthesis of a hierarchical ZSM-5 zeolite for methanol to propylene reaction: a b-oriented thinner dimensional morphology

Abstract

Under neutral fluoride medium at 373 K, a facile strategy has been developed to synthesise a hierarchical ZSM-5 zeolite (M-ZSM-5) using a solid silica source in a dense system. The resulting material shows a hexagonal lamellar shape with b-oriented thinner dimension and bi-modal porosity containing MFI micropores and intracrystal mesopores. The effect of synthesis factors, such as NH₄F/SiO₂, H₂O/SiO₂, crystallization temperature and time, on the zeolite morphology and size is studied and its primary crystallization process is proposed. Through varying the synthesis parameters, the crystal size could be tuned with the aspect ratio at a range of 6.2–12. Typically, M-ZSM-5 with a thinner thickness (100 nm) and a high aspect ratio (AR = 12) shows a highly effective catalytic performance in methanol to propylene (MTP) reaction. Compared to the bulk C-ZSM-5 sample obtained from a hydroxyl system, M-ZSM-5 shows higher selectivity for propylene (45.1% vs. 38.1%) and butylene (27.2% vs. 21.3%), especially prolonged catalytic lifetime (224 h vs. 98 h). This enhanced performance could be contributed to the optimized acidity and superior diffusivity of M-ZSM-5 with lamellar morphology and bimodal porosity, which are crucial for suppressing secondary reactions and inhibiting coke deposition.

---

1. Introduction

The methanol to propylene (MTP) process, originally developed by Lurgi’s company, provides an alternative route for the production of propylene, which attracts more scientific interest.^1–3 Zeolite ZSM-5 with intricate channels, adjustable acidity and high thermal/hydrothermal stability has been proven to be a remarkably effective material in the MTP reaction.^4,5 Recently, theoretical and experimental results have given many new insights into the extremely complex reaction network^6–8 and the underlying interplay between catalyst microstructure and its MTO/MTP performance.^9–11

The most accepted “dual-cycle” mechanism for MTO/MTP reactions suggests that ethylene is formed from the lower methylbenzenes in the aromatic-based cycle whereas propylene is generated through the olefin-based cycle by alkene methylations, and the light olefins can be further converted to alkanes and aromatics by alkylation, cracking, hydrogen transfer and cyclization reactions.^12,13 The narrow channels strongly impose a restriction on the transport of product molecules, leading to an increase of secondary reactions. Therefore, this implies that the improvement of the diffusivity would be beneficial to increase the selectivity for propylene and reduce the coke deposits.^14,15

Some strategies have been adopted to reduce the diffusion resistance of the pure microporosity in the ZSM-5 zeolite. Decreasing the crystal size to nanometer scale and introducing auxiliary mesopores into the zeolite crystals are beneficial to shorten the diffusion path length and improve accessibility to the internal surface.^16,17 However, separation from the mother liquor is a big difficulty for the nano-sized zeolite in large scale processes, and the preparation of hierarchical zeolite requires multiple synthesis steps and/or expensive organic templates. The ZSM-5 zeolite possesses an anisotropic framework with two intersecting 10-membered ring channels, among which are straight channels parallel to the b-axis and zig-zag channels parallel to the a-axis.¹⁸ Guest molecules can transfer through the two channels in the zeolite crystal for the processes of catalytic reaction or separation. A faster diffusion rate can be observed in the direction of the b-axis due to the diffusion anisotropy.^19–21 Ryoo’s group synthesized ZSM-5 nanosheets consisting of 2 nm-thick layers along the b-axis dimension,^22,23 which performed with high activity in large molecule reactions because the specific crystal with b-axis thin layers facilitated the accessibility of acid sites at the exterior or pore mouths. According to our previous work,²⁴ the high propylene selectivity and low deactivation rate over ZSM-5 nanosheets could mainly be ascribed to its unique morphology with the shortened diffusion path length in MTP reactions. Further insightful studies also embodied the concept of 2D zeolite morphology and its importance in catalysis.²⁵ The ultra-thin nanosheets would lead to lower hydrothermal/physical stability, but this aroused inspiration for controlling the crystal growth rates along the three crystallographic axes, forming ZSM-5 zeolites with thinner b-dimension.

Many works have evidenced that the synthesis conditions (such as the molar composition, Si or Al raw materials, templates, additive, pH value, crystallization method and time) have significant effects on the kinetics and thermodynamics of crystal growth, which accounts for the morphology, preferential growth and orientation of the MFI zeolite crystals.^26,27 By using dimers and trimers of TPAOH as structure-directing agents (SDAs), Bonilla et al. found that the growth rates of the ZSM-5 crystals along each axes of the abc directions depend on the amount of SDAs used in the hydroxyl system.²⁸ Combining the utilization of a co-solvent of diols and microwave heating, Chen et al. fabricated silicate-1 crystals with tuneable sizes, shapes, and aspect ratios.²⁹ Xiao’s group synthesized TS-1 crystals and found that the growth rate of the MFI crystals along the b-axis can be suppressed by the addition of urea.^30,31

Recently, significant progress in zeolite synthesis has been made by using fluoride ions to replace hydroxyl anions as a mineralizer, largely extending the synthetic pathway of the zeolite.³² It is believed that F atoms are occluded inside small cages in the MFI structure progressively which modifies the neighboring silicon³³ to affect zeolite morphology, pore structure and the acidic properties, resulting in an enhancement of the catalytic activity and selectivity. Arichi et al. prepared ZSM-5 zeolite with star-like/hedgehog-like morphology by using polymeric and monomeric silica sources with a certain ratio of F/Si.³⁴ Also, under fluoride medium, the crystal size of ZSM-5 can easily be controlled³⁵ and crystals with a b-axis orientation are synthesized by the addition of co-template.³⁶ For instance, Dose et al. synthesized silicate-1 zeolites with a tuned crystal size ranging from 0.4 to 30 µm along the b direction.³⁷ Yao et al. prepared Silicalite-1 with plate-like morphology and varying axes thicknesses, and they found that anisotropic growth of the crystal faces was mainly governed by the concentration of NH₄F in the gels.³⁸ ZSM-5 zeolites obtained in fluoride or hydroxide medium were comparatively evaluated in MTO/MTP reactions.^39,40 The large ZSM-5 crystals with a size of 15–20 µm obtained in fluoride medium exhibited a lower Brønsted acid site density, thus giving rise to a high propylene to ethylene (P/E) ratio in the MTO reaction.⁴⁰ Through seed addition in F medium at 443 K, the resulting ZSM-5 exhibited cubic small particles, fewer structure defects and thus optimized acidity, which is favourable for a high initial activity and resistance to deactivation in MTH reactions.⁴¹

Herein, under a neutral fluoride medium at 373 K, a hierarchical ZSM-5 zeolite is obtained by using a single TPABr template and solid silica source, and thus forming a dense system. The resulting material showed a hexagonal lamellar shape with a thinner b-axis dimension, containing bi-modal porosity, which provided an excellent catalytic performance in the MTP reaction. Influencing factors in the crystallization process are systematically discussed for fine tuning the zeolite crystal size, aspect ratio and microstructure. In addition, the primary synthesis mechanism of the ZSM-5 crystal is proposed.

2. Experimental

2.1 Synthesis of the hierarchical ZSM-5 zeolite in fluoride medium

The hierarchical ZSM-5 zeolite and a counterpart ZSM-5 sample were synthesized using the following chemicals as raw materials. Solid silica gel (SiO₂, 99 wt%) was purchased from the Qingdao Haiyang Chemical Co. (P. R. China). Sodium aluminate (NaAlO₂, 41 wt% Al₂O₃), aluminum sulfate (Al₂(SO₄)₃·18H₂O, >98 wt%) and tetrapropylammonium bromide (TPABr, >98 wt%) were purchased from the Sinopharm Chemical Reagent Co., Ltd (P. R. China). Ammonium fluoride (NH₄F, >99 wt%) and NaOH were obtained from the Beijing Yili Co. (P. R. China).

A typical synthesis of the hierarchical ZSM-5 zeolite in fluoride medium is performed as follows. NH₄F was completely dissolved in deionized water and Al₂(SO₄)₃·18H₂O, and TPABr were added into the fluoride solution. After that, SiO₂ was added into the mixture slowly under stirring to give the final mole ratio: SiO₂ : 0.02 Al₂O₃ : 0.048 TPABr : 0.8 NH₄F : 6 H₂O. The pH was adjusted to 6.5–7.5 by the dropwise addition of hydrochloric acid solution (HCl, 1 M) under vigorous stirring, while accurate pH paper was used for the pH adjustments. After aging at 323 K for 12 h, the mixture was transferred into a Teflon-lined stainless-steel autoclave and crystallized at 373 K or 453 K for a specific period, respectively. After crystallization, the obtained product was filtered, thoroughly washed with deionized water, and dried overnight at 393 K. By varying the preparation parameters, such as the NH₄F/SiO₂ ratio, H₂O/SiO₂ ratio, crystallization temperature and time, different samples were obtained using the same procedure. The parameters used are listed in Table 1 and the samples are named A–F alphabetically.

Table 1 Synthesis of the ZSM-5 samples in fluoride medium

No.	NH4F/SiO2	H2O/SiO2	Temp. [K]	Time [h]	RC^a [%]	Crystal size^b [L × W × T] µm	AR^c
a Denotes relative crystallinity. b The average crystal sizes of the ZSM-5 crystals, with length (L), width (W) and thickness (T), are estimated using high resolution SEM images. c Denotes the aspect ratio,^37 AR = (L + W)/2T.
A	0.6	6	373	72	62	— × — × 0.10	—
B	0.8	6	373	72	98	1.6 × 0.8 × 0.10	12.0
C	1.2	6	373	72	90	2.8 × 1.5 × 0.25	8.6
D	1.6	6	373	72	85	3.6 × 2.0 × 0.45	6.2
E	0.8	20	373	72	78	3.8 × 1.8 × 0.35	8.0
F	0.8	6	453	12	90	2.4 × 1.0 × 0.20	8.5

---

For comparison, a conventional ZSM-5 zeolite was prepared in an alkaline system. NaAlO₂, SiO₂, TPABr and NaOH were added in order into deionized water with the mole ratio as follows: SiO₂ : 0.0033 Al₂O₃ : 0.24 TPABr : 0.4 Na₂O : 28 H₂O. Finally, the mixture was crystallized at 453 K for 48 h and the product was collected by filtration, washed thoroughly and dried overnight at 393 K.

To obtain H-form zeolites, the synthesized samples in this work were calcined at 823 K for 6 h to remove the organic template, and then ion-exchanged twice in NH₄Cl solution (1 M) at 363 K for 2 h, followed by calcination at 823 K for 6 h. After calcination and ion-exchange, fluoride anions compensating for TPA⁺ in the as-synthesized ZSM-5 are released.

2.2 Catalyst characterization

X-ray diffraction (XRD) patterns were determined using a Bruker D8 Advance X-ray diffractometer, using Cu Kα radiation, operated at 40 kV and 30 mA with a range of 2θ = 5°–50° and at a scanning rate of 2° min⁻¹. The relative crystallinity of the samples was calculated by comparing the sum of the area of the characteristic peaks between 2θ = 22.5°–25° with that of the commercial ZSM-5 zeolite (Nankai Catalyst Plant, China, the crystallinity was considered to be 100%). The elemental compositions of the samples were measured using an AxiosmAX X-ray fluorescence analyser (XRF).

Scanning electron microscopy (SEM) images were obtained on a Quanta 200F for investigating the morphology and crystal size of the samples. Transmission electron microscopy (TEM) images were obtained using a Tecnai G² F20 instrument operating at 200 kV.

Nitrogen adsorption–desorption isotherms were measured on a Micromeritics ASAP 2020 instrument at 77 K. Prior to measurement, the samples were degassed at 573 K under vacuum for 4 h. The specific surface area is calculated using the BET equation. The total pore volume is based on the nitrogen adsorbed volume at P/P₀ = 0.99. The external surface area is calculated via the t-plot method. The mesopore size distribution is analyzed using the Barrett–Joyner–Halenda (BJH) method for the desorption branch of the isotherms.

Temperature programmed desorption of ammonia (NH₃-TPD) measurements were performed on a fixed-bed reactor with a thermal conductivity detector (TCD) for measuring the acidity. The samples (0.1 g, particle size 20–40 mesh) were first pre-treated at 873 K for 1 h in a flow of He, then cooled down to 373 K, and saturated with NH₃ for 30 min. Finally, the sample measurements were conducted in He-flow by raising the temperature from 373 K to 873 K at a rate of 10 K min⁻¹.

Solid state ¹⁹F magic angle spinning nuclear magnetic resonance spectroscopy (MAS-NMR) was conducted on a Bruker 500 MHz spectrometer at a resonance frequency of 470.95 MHz using a 2.5 mm rotor. In the experiment, a ¹⁹F MAS NMR spectrum was recorded using a pulse width of 1 µs and spinning rate of 23 kHz. CFCl₃ was used as a reference for the chemical shifts of ¹⁹F.

2.3 Catalyst evaluation

The MTP reaction was conducted in a fixed-bed micro-reactor with an inner diameter of 8 mm at 743 K under atmospheric pressure. For each test, the catalyst (H form zeolite samples, 1 g, particle size 20–40 mesh) was loaded into the reactor. A mixture of methanol and water with the molar ratio of 1 : 1 was fed into the reactor and the weight hourly space velocity (WHSV) was set at 3 h⁻¹. The products were analyzed using an online gas chromatograph (Agilent GC 7890) equipped with a flame ionization detector (FID) and an HP-PLOT Q capillary column. Methanol and DME were considered as reactants in the calculation of conversion and selectivity.

After methanol conversion, the amount of generated coke in the ZSM-5 catalysts was determined using thermogravimetric analysis (TGA), which was performed on a TG-DTAXSTAR6000 instrument. In a typical experiment, the TGA curve was recorded as the catalyst (10 mg) was heated from 298 K to 1073 K with a heating rate of 10 K min⁻¹.

3. Results and discussion

3.1 Synthesis of the ZSM-5 zeolites in the fluoride medium

In the gel mixtures, the solubility of the solid SiO₂ and thus the formed silica/aluminum species depends on the acidic/basic properties of the medium. To a certain extent, the crystallization process is governed by the pH value, which further affects the zeolite morphology and porosity.⁴² Due to using fluoride as an effective mineralizer and a structure-directing agent, the ZSM-5 synthesis could be accomplished under neutral/weakly acidic conditions by using the fluoride route, differing from the traditional synthesis conducted in alkaline solution.⁴³ Herein, experiments were performed in the presence of NH₄F to synthesize the ZSM-5 zeolite at optimal pH (6.5–7.50), NH₄F/SiO₂ ratio, H₂O/SiO₂ ratio and crystallization temperature and time. The relative crystallinity, the crystal size of the prepared ZSM-5 samples and the corresponding synthesis parameters are summarized in Table 1.

On varying the NH₄F/SiO₂ ratio from 0.6, 0.8 and 1.2 to 1.6, the XRD patterns (Fig. 1(i)) showed that all the samples consist of a MFI phase, and the crystallinity of the samples with the various NH₄F/SiO₂ ratios is 62%, 98%, 90% and 85%, respectively. This indicates that a small amount of fluoride is disadvantageous to the formation of the ZSM-5 zeolite. Meanwhile, excess fluoride also interrupts the electronic interactions between the anionic silicates and cationic TPA⁺ in the synthesis gel, leading to the production of some miscellaneous crystals.⁴⁴ Hence, the appropriate amount of fluoride is conducive to promoting the formation of Si–O–Si bonds, influencing the hydrolysis and condensation of the silicates during the crystal growth process, which would play an important role in the ZSM-5 morphology as well.

Fig. 1 i) XRD patterns and (ii) SEM images of the ZSM-5 samples with NH₄F/SiO₂ ratios of (a) 0.6, (b) 0.8, (c) 1.2 and (d) 1.6, respectively. (iii) Crystal dimensions and the corresponding crystal faces.

Fig. 1(ii) presents the morphologies of the synthesized ZSM-5 samples with various NH₄F/SiO₂ ratios. Fig. 1(iii) shows a schematic identifying the crystal size and corresponding crystal faces. When the NH₄F/SiO₂ ratio is 0.6, the crystals display an overall inter-grown and connected lamellar shape with the thickness less than 100 nm. On increasing the ratio of NH₄F/SiO₂ to 0.8, the crystals gradually grow to be individual and form a regularly hexagonal lamellar morphology featuring a high aspect ratio (AR = 12.0). The average size of the length and width is about 1.6 and 0.8 µm and the thickness along the b axis is only 0.1 µm. Some inter-grown hexagonal plates appear when the NH₄F/SiO₂ ratio is increased to 1.2, and the crystal size is increased to 2.8 × 1.5 × 0.25 µm with an AR of 8.6. With further increase of the F⁻ concentration, highly dispersed particles with an inter-grown ellipse shape are obtained. The crystal size becomes larger (3.6 × 2.0 × 0.45 µm), while the AR is lower at 6.2.

The ZSM-5 crystal sizes of length, width and thickness are increased but the aspect ratio decreases with increasing the F⁻ concentration, which indicates that the addition of fluoride anions affects the crystal growth rate along each crystal face. According to the Bravais–Friedel–Donnay–Harker (BFDH) law,^45,46 the growth rate along a crystallographic direction is inversely proportional to its interplanar distance d_hkl, and the fastest growth direction of a MFI zeolite is always along the c-axis of the crystals. For the synthesis of a MFI zeolite with TPA⁺ as the structure-directing agent, the typical morphology of the ZSM-5 crystals is coffin shaped with an order of the crystal dimensions of L_c > L_a > L_b, where L_i indicates the crystal size along the i axis.²⁸ In the initial stage, zeolite nucleation occurs on the surface of gel spheres and the crystals are more likely to be oriented with their ac-plane parallel to the gel sphere surface.^47,48 It is believed that the dissolution and depolymerization of the Si source are closely associated with the concentration of fluoride anions, and the supersaturation degree of the silicate species in the synthesis gel influences the growth rate of the crystal along the b-axis.⁴⁹ When employing solid silica as the silica source under low F⁻ concentration, the dissolution rate of Si is relatively slow and the formation of hydroxylated silica species could be controlled, thus the adsorption of silica species on the nuclei interface is inhibited. Therefore, the crystal growth rate in the direction along the b axis is slow and results in the formation of ZSM-5 crystals with thinner dimensional morphology. With an increased fluoride concentration in the gels, the dissolution rate of silicon is increased and more silica species can be supplied from solution. The crystal growth rate is accelerated in all directions to form inter-grown ellipse crystals with enhancement of the surface energy from the electrostatic forces between F⁻, TPA⁺ and Si species. Herein, it is implied that the F⁻ concentration has an intense effect on the dissolution of the Si source and indirectly controls the growth rate of the crystal face to form ZSM-5 crystals with a b-oriented morphology.

The water content affects the dissociation and migration of raw materials and indirectly influences the concentration of guest species (F⁻ and TPA⁺), which could control the processes of nucleation and crystal growth and further influence the structure and morphology of the zeolite.⁵⁰ Fyfe reported the synthesis of a MFI zeolite with a morphology of inter-grown obloids under NH₄F and minimal water content gels.⁵¹ In this study, on increasing the H₂O/SiO₂ ratio to 20, the sample presents a dispersed plate shape (Fig. S1†) with a larger size (3.8 × 1.8 × 0.35 µm) and lower aspect ratio (AR = 8.0), and some amorphous materials appear leading to low relative crystallinity (78%). This suggests that a high H₂O/SiO₂ ratio reduces the F⁻ and TPA⁺ concentration, leading to increase of the crystal size with length. Optimizing the synthesis composition, by combining a suitable water content with F⁻/TPA⁺ concentration, can modify the dissolution and migration of the silicon and aluminum species, thus affecting the morphology of the synthesized sample.

Using the same synthesis recipe as for sample B, crystal growth curves at 373 K and 453 K were compared and are shown in Fig. 2. When crystallization occurred at 373 K, the nucleation induction period was long and crystal growth started after 36 h, the zeolite crystals with a high aspect ratio obtained after 72 h. However, the crystallization rate largely increased while the crystallization time decreased at 453 K, and the highly crystalline product was obtained after 12 h. An increase in crystallization temperature leads to larger zeolite crystals.⁵² As seen in the SEM image (Fig. S2†), the irregular and dispersed crystals still maintain a lamellar morphology with a larger size of 2.4 × 1.0 × 0.2 µm, showing that the crystal growth increased along the crystal faces corresponding to the abc axes. For the resulting material, its crystal size is strongly related to the number and distribution of crystal nuclei in the synthesis gel. When the rapid crystallite growth overuses the precursor species at the higher temperature (453 K), the number of nucleation sites is relatively reduced, leading to relatively larger crystals. The number of metastable phases increases at 453 K, thus the obtained crystals become of irregular morphology. In contrast, more nuclei evolve at 373 K in the long induction period, and then the crystallite growth starts by employing these viable nuclei to gradually form a highly crystalline zeolite. So the formation of a zeolite with ultimate crystal morphology is mainly governed by the reaction composition and the conditions of the synthesis system, including the concentration of fluoride ions, water content, crystal temperature and time. Herein, by using solid silica as Si source with a relatively dense system, the synthesis process using a 373 K crystallization temperature has significant potential for industrial application.

Fig. 2 Crystal growth curves of sample B synthesized at different temperatures.

3.2 Physicochemical properties of the hierarchical ZSM-5 zeolite

To further understand the textural properties of the fluoride-prepared ZSM-5 zeolites, the well-crystallized sample B (denoted as M-ZSM-5) and the conventional ZSM-5 zeolite (in alkali medium, C-ZSM-5) were comparatively characterized (XRD, Fig. S3†). The C-ZSM-5 presents a cubic shape with a size of about 1 µm (SEM, Fig. S4†).

Fig. 3(a) shows the TEM image of M-ZSM-5 with hexagonal lamellar morphology. The circular streaking in the SAED pattern (inset of Fig. 3(a)) shows that the projection view down the thinner dimension of the M-ZSM-5 crystal corresponds to the [010] direction. This gives confirmation that the b-axis (the straight channels) is the shortest dimension of the crystal. Furthermore, as seen clearly in Fig. 3(b), some randomly distributed mesopores can be observed in M-ZSM-5 zeolite and the pore diameters are in the range of 10–40 nm.

Fig. 3 TEM images of the M-ZSM-5 sample.

The N₂ adsorption/desorption isotherms and the corresponding pore size distributions of M-ZSM-5 and C-ZSM-5 are shown in Fig. 4. Each of the curves has a steep increase at extremely low relative pressure, indicating that the nitrogen is adsorbed in the uniform micropores. Obviously, C-ZSM-5 has characteristics of traditional microporous materials without mesopores. The isotherm of M-ZSM-5 is significantly different from that of C-ZSM-5, and presents a representative modified type IV isotherm, containing an obvious hysteresis loop corresponding to capillary condensation at P/P₀ = 0.45–0.99, which suggests irregular mesoporosity arising from the voids between the particles.⁵³ More interestingly, M-ZSM-5 presents a specific hysteresis loop in the range of P/P₀ = 0.10–0.30, which is usually considered as a fluid-to-crystalline phase transition of the adsorbed phase in the micropores.^54,55 This phase transition phenomenon becomes more pronounced for a high silica MFI zeolite (e.g. silicate-1).⁵⁶ For the similar SiO₂/Al₂O₃ content of M-ZSM-5 and C-ZSM-5, the M-ZSM-5 crystals with lamella-like morphology may be favourable for the phase transition, which is identical with Xiao’s report.³¹ The pore size distribution curve of M-ZSM-5 indicates that the material has bi-modal porosity including MFI micropores and mesopores within the range of 10–40 nm (inset Fig. 4), which is well in agreement with the HRTEM image.

Fig. 4 N₂ adsorption/desorption isotherms and BJH pore size distributions of M-ZSM-5 and C-ZSM-5.

The textural properties of the two synthesized ZSM-5 samples are compared in Table 2. The BET surface areas of M-ZSM-5 and C-ZSM-5 are very close, 358 m² g⁻¹ and 362 m² g⁻¹, respectively. But the external surface area of M-ZSM-5 is higher (105 m² g⁻¹) than that of C-ZSM-5 (87 m² g⁻¹), with an increase of 21%. Especially, the total pore and mesopore volume for M-ZSM-5 are 0.33 cm³ g⁻¹ and 0.20 cm³ g⁻¹, higher than those of C-ZSM-5 (0.22 cm³ g⁻¹ and 0.08 cm³ g⁻¹). Combining this with the N₂ adsorption/desorption isotherms, the larger external surface area and increased pore volume affirmatively result from the creation of mesopores/voids.

Table 2 Textural properties of M-ZSM-5 and C-ZSM-5

Sample	SiO2/Al2O3^a	S BET ^b [m^2 g^−1]	S ext ^c [m^2 g^−1]	V total ^d [cm^3 g^−1]	V meso [cm^3 g^−1]
a SiO2/Al2O3 molar ratio of the synthesized hierarchical and the conventional ZSM-5 zeolite as determined using XRF. b S BET (BET surface area) obtained from the adsorption isotherm. c S ext (external surface area) calculated using the t-plot method. d V total (total pore volume) obtained at P/P0 = 0.99.
M-ZSM-5	256	358	105	0.33	0.20
C-ZSM-5	232	362	87	0.22	0.08

---

The NH₃-TPD profiles of M-ZSM-5 and C-ZSM-5 are shown in Fig. 5. Two NH₃ desorption peaks could be observed in the temperature regions 440–470 K and 620–660 K, which were attributed to weak and strong acid sites, respectively. As seen, the total area for M-ZSM-5 is considerably close to that for the corresponding catalyst C-ZSM-5, which demonstrates that there is a similar density of the acid sites on the two samples. Obviously, there is a significant difference between the distribution of the weak and strong acid sites, and M-ZSM-5 has more weak acid and less strong acid sites than C-ZSM-5.

Fig. 5 NH₃-TPD profiles of M-ZSM-5 and C-ZSM-5.

3.3 Formation mechanism of the hierarchical ZSM-5 zeolites

The ¹⁹F MAS-NMR spectrum was obtained for understanding of the incorporation of fluorine species into the framework during the synthesis. For the M-ZSM-5 sample, the spectrum in Fig. 6 displays two significant signals at −64 and −140 ppm, along with some weaker peaks at −79, −105 and −129 ppm. The peak at −64 ppm is assigned to fluoride ions in a [4¹5²6²] cage due to the fluoride being presented as a counterion of TPA⁺ and located in framework interstices, and the signal at −140 ppm can be tentatively attributed to an AlF₆³⁻ species which is present inside or outside the crystal.^57,58 While the weak peak at −79 ppm may be associated with defects in the structure, the signal at about −129 ppm is presumably assigned to silicofluoro complexes (SiF₆²⁻), and the broader and weaker signal at −105 ppm could be interpreted as being due to the presence of non-specifically adsorbed F⁻ (O–H⋯F).^58,59 The results demonstrate that fluoride species in the synthesis system not only perform a charge balance role with the template agent to contribute to the stabilization of the structure, but ensure the formation of T–F (T: Si, Al) species to encourage the crystallization.

Fig. 6 ¹⁹F MAS-NMR spectrum of the hierarchical ZSM-5 sample synthesized in F⁻ medium.

Compared to TEOS as the silica source in a dilute mixture gel system,^34,41 herein, the use of solid silica source, relatively dense gel and neutral fluoride medium could make the crystallization of M-ZSM-5 quite distinguishable. As in the above discussion, the silica/aluminum species are related to the F⁻ concentration and water content in the synthesis gel, which could modify the structure-directing ability of the F⁻/TPA⁺ template and control the rate of nucleation and crystallization, thus have a crucial influence on the morphology of the MFI crystals.

The SEM images in Fig. 7 primarily describe the formation process of M-ZSM-5 during different periods of crystallization at 373 K. In the initial stage of crystallization (12 h), large particles are obtained from the gel mixture (Fig. 7(a)) and no obvious crystalline phase is observed in the XRD patterns (Fig. S5†). According to its crystallization curve, it is suggested that silica sparingly dissolves and some solid remains in the induction stage. After 36 h, a large number of small particles or platelets are observed (Fig. 7(b)), showing weak XRD diffraction peaks at 2θ = 7–9° and 22.5–25°, which suggests that more solid silica dissolves and starts to form the elementary units of ZSM-5. As the time is prolonged to 56 h, some thinner plates around 1 µm in size with a relative crystallinity of ∼54% are clearly observed in Fig. 7(c), indicating the formation of ZSM-5 crystals. The amorphous silica/alumina species are completely disappeared and a highly crystalline ZSM-5 zeolite with lamellar morphology (Fig. 7(d)) is fully accomplished after 72 h. This growth process is consistent with the crystallization curves in Fig. 2 and the N₂ isotherms for the resulting materials in Fig. S6.† At the crystallization times of 12 h and 36 h, the material presents a hysteresis loop in the N₂ isotherms at P/P₀ above 0.8, featuring large mesopores/macropores structures mainly arising from the raw materials. With the crystallization time increased to 56 h, a hysteresis loop occurs at about 0.45 relative pressure which is associated with the presence of mesopores. Finally, two hysteresis loops at P/P₀ = 0.1–0.25 and 0.45–0.99 were found which are closely related to the obtained hierarchical M-ZSM-5 zeolite.

Fig. 7 SEM images of M-ZSM-5 during the crystallization stages at: (a) 12 h; (b) 36 h; (c) 56 h; (d) 72 h.

The primary synthesis mechanism of ZSM-5 is described in Scheme 1. At the beginning, the solid silica material partially solubilizes under the near-neutral pH value and low crystallization temperature, simultaneously, fluoride and the template TPA⁺ spontaneously disperse into the solid silica support due to the increase of entropy.^60,61 On the one hand, fluorine acts as a template with the ion-pair form (F⁻–TPA⁺) inside the [4¹5²6²] cage in the MFI channels and creates a covalent bond with Si to form an energetically stable [SiO_4/2F]⁻ unit due to the high electronegativity,⁶² improving the dissolution of silica species and the formation of T–O–T (T: Si or Al) bonds.⁶³ On the other hand, the neighboring silicon and aluminum raw materials around the fluoride ions gradually form Si–F and Al–F complexes, and thus these complex species self-organize to form local reservoirs of nuclei.³⁴ With the dissolution of the solid silica source and the formation of a large number of nuclei, these aluminosilicate species are locally rearranged to form MFI nanocrystals. Due to the electronic interactions among the fluoride ions, TPA⁺ and Si species (which significantly influence the morphology), zeolite crystals grow along a defined spatial direction.⁶⁴ Numerous nanocrystals are condensed and finally constantly consume neighboring solid silica/aluminum species to form lamellar MFI crystals with a preferential b-orientation. Under the dense system with low water content, it is difficult for the silicon and aluminum species to migrate in the synthesis gel. The assembly of aluminosilicate species around TPA⁺ is non-uniform, resulting in the presence of some mesopores in the surface or interior of the MFI crystals. In addition, a large amount of AlF₆³⁻ and SiF₆²⁻ species are present inside or outside the synthesized crystals and are almost completely removed by calcination (XPS spectra in Fig. S7†). Therefore, it is speculated that the mesoporous structure of M-ZSM-5 might be created from platelet dislocation during the crystallization process and partly from the post-treatment during the removal of Si–F and Al–F complexes by calcination or ion-exchange.

Scheme 1 The proposed approach for the formation of the ZSM-5 zeolite under F medium.

3.4 Catalytic performance of the hierarchical ZSM-5 zeolites

The hierarchical ZSM-5 (M-ZSM-5) and the conventional ZSM-5 (C-ZSM-5) were comparatively evaluated for the MTP reaction. The catalytic activity of M-ZSM-5 and C-ZSM-5 as a function of time is displayed in Fig. 8. At the beginning of the reaction, the methanol conversion over both samples is approximately 100%, but the reaction stability of M-ZSM-5 is more prominent. When catalyst deactivation is set as the methanol conversion of 100% falling to around 90%, the catalytic lifetime of M-ZSM-5 (224 h) is nearly twice as long as that of C-ZSM-5 (98 h).

Fig. 8 Methanol conversion with time on stream for the M-ZSM-5 and C-ZSM-5 catalysts. Reaction conditions: T = 743 K, P = 0.1 MPa, n(CH₃OH) : n(H₂O) = 1 : 1, WHSV = 3.0 h⁻¹.

Fig. 9 exhibits the product distribution over M-ZSM-5 and C-ZSM-5 as a function of TOS. At the initial stage (TOS = 0.5 h), the selectivity for propylene and butylene is 42.7% and 28.8% for M-ZSM-5, remarkably higher than those of C-ZSM-5 (33.3% and 21.2%), but the selectivity for ethylene is lower. After about 20 h TOS, the reactions reached a steady state stage with a stable product selectivity over the two ZSM-5 catalysts. Both samples exhibit a regular catalytic behaviour with TOS for the target products, involving the steady formation of propylene, a gradual decrease of ethylene and butylene and a gradual increase of C₅₊ compounds. Obviously, M-ZSM-5 maintains a stable performance for a longer period of more than 200 h, and the steady state stage for C-ZSM-5 is only 80 h. The average values of product selectivity at the steady state stage are shown in Fig. S8.† Compared to C-ZSM-5, the M-ZSM-5 zeolite shows higher selectivity towards propylene (45.1% vs. 38.1%) and butylene (27.2% vs. 21.3%), and lower selectivity towards ethylene (6.5% vs. 9.2%), leading to a higher propylene/ethylene ratio (P/E) (6.91 vs. 4.15). The total amount of light olefins (C₂⁼–C₄⁼) over M-ZSM-5 is nearly 80%, which is greatly higher than that of C-ZSM-5. Furthermore, M-ZSM-5 results in a lower amount of C₁–C₄ alkanes and C₅₊ hydrocarbons. With a prolonged reaction time, the catalysts deactivate and the methanol conversion drops rapidly. Close to the deactivation stage, the propylene selectivity decreased slowly and the C₅₊ increased greatly. It is accepted that the deactivation of a catalyst in the MTP process generally arises from coke deposition.⁶⁵ As shown in Fig. S9,† M-ZSM-5 contains 8.48 wt% coke deposits after 224 h on stream, while for C-ZSM-5 the coke deposit amount is 11.34 wt% after 98 h. Owing to low coke deposition, M-ZSM-5 presents a longer catalytic lifetime than C-ZSM-5.

Fig. 9 Product selectivity of M-ZSM-5 (filled symbols) and C-ZSM-5 (open symbols) for the MTP reaction as a function of time on stream. Reaction condition: T = 743 K, P = 0.1 MPa, n(CH₃OH) : n(H₂O) = 1 : 1, WHSV = 3.0 h⁻¹.

According to the “dual-cycle” MTO/MTP reaction mechanism, the formation of propylene and butylene is governed by alkene methylation/cracking pathways whereas ethylene is mainly formed through the aromatic/ethylene cycle, and the aromatic/ethylene cycle cannot run without the C₃₊ alkene cycle.¹² For the solid-acid catalytic reaction, the acidic properties of the ZSM-5 zeolite affect the activity and selectivity during the MTP process. It is accepted that the strong acid sites are regarded as the main active sites for the conversion of methanol, but they also can accelerate the formation of coke species through secondary reactions.⁶⁶ The weak acid sites promote methylation and alkylation reactions, which is favourable for the formation of alkene-intermediates.^67,68 Moreover, the weak acid sites can restrain hydrogen-transfer reactions to produce saturated hydrocarbons and aromatics, which are the precursors for coke formation.⁶⁹ Therefore, an appropriate weak acid density is beneficial to facilitate the formation of olefins and reduce the further reactions for hindering coke deposition.⁷⁰ As a result, M-ZSM-5 obtained from the fluoride system holding less strong and more weak acid sites has good activity for the C₃₊ alkene cycle in the MTP reaction, thus leading to higher propylene selectivity. Also, the lifetime of M-ZSM-5 is remarkably enhanced due to less coke deposition.

Besides the appropriate catalyst acidity, the high propylene selectivity and long catalytic lifetime strongly depend on the enhanced diffusivity in the zeolite.¹⁴ The hierarchical M-ZSM-5 zeolite can facilitate the removal of light olefins (in particular propylene and butylene) from the channels. With the improvement of diffusivity, the residence time of the primary olefin products is reduced and the secondary reactions (alkylation, cracking, hydrogen transfer and cyclization) are restrained, which is beneficial to suppress the aromatic/ethylene cycle and reduce the formation of coke species. Combining the thinner thickness (∼100 nm) along the b axis and the introduced mesopores in M-ZSM-5, which shorten the diffusion path length and increase the mouth openings, high propylene and butylene selectivity and a long catalytic life time are observed in the MTP reaction. Also, secondary reactions for coke deposition would be suppressed on M-ZSM-5, leading to low selectivities for C₁⁰–C₄⁰ and C₅₊ hydrocarbons.

4. Conclusions

In summary, a hierarchical ZSM-5 zeolite (M-ZSM-5) with a thinner b-oriented morphology has been obtained in fluoride medium by using solid silica and a dense gel at 373 K. The influencing parameters, such as fluoride concentration, water content, crystallization temperature and time, are critical factors for the control of the crystallographic morphology and size. The obtained material with a thickness of 100 nm possesses abundant mesopores, resulting in a higher external surface area and extremely larger volume. Moreover, the synthesis mechanism of M-ZSM-5 is discussed. Fluoride anions with the strong electronegativity promote the dissolution of the solid silica and act as a structure directing agent in the form of an ion-pair (F⁻–TPA⁺), leading to the formation of ZSM-5 nuclei/microcrystals. These microcrystals gradually generate the platelets and fabricate the hierarchical ZSM-5 zeolite with a thinner b-axis oriented dimension.

Compared to the C-ZSM-5 sample, the M-ZSM-5 zeolite performs with a longer catalytic lifetime in the MTP reaction, and the initial selectivity for propylene and butylene of the M-ZSM-5 zeolite was increased by 9.4% and 7.6%, respectively. The selectivity for propylene and butylene is more than 45% and 27%, and the total amount of olefin (ethylene, propylene and butylene) selectivity is nearly 80% at the steady state stage. The excellent performance should be contributed to by its optimal acidity and the superior diffusivity of the specific structure with a thinner b-axis path and intracrystal secondary mesopores.

Acknowledgements

The authors acknowledge the State Key Development Program for Basic Research of China (2012CB215002) and the National Natural Science Foundation of China (21176255, 21276278) for financial support of this work.

Notes and references

1. C. D. Chang, Catal. Rev., 1983, 25, 1–118.
2. F. J. Keil, Microporous Mesoporous Mater., 1999, 29, 49–66.
3. U. Olsbye, S. Svelle, M. Bjørgen, P. Beato, T. V. W. Janssens, F. Joensen, S. Bordiga and K. P. Lillerud, Angew. Chem., Int. Ed., 2012, 51, 5810–5831.
4. C. D. Chang, C. T. W. Chu and R. F. Socha, J. Catal., 1984, 86, 289–296.
5. M. Stöcker, Microporous Mesoporous Mater., 1999, 29, 3–48.
6. J. F. Haw, W. Song, D. M. Marcus and J. B. Nicholas, Acc. Chem. Res., 2003, 36, 317–326.
7. C. Wang, Y. Chu, A. Zheng, J. Xu, Q. Wang, P. Gao, G. Qi, Y. Gong and F. Deng, Chem.–Eur. J., 2014, 20, 12432–12443.
8. S. Müller, Y. Liu, M. Vishnuvarthan, X. Sun, A. C. van Veen, G. L. Haller, M. Sanchez-Sanchez and J. A. Lercher, J. Catal., 2015, 325, 48–59.
9. R. Khare, D. Millar and A. Bhan, J. Catal., 2015, 321, 23–31.
10. Q. Wang, S. Xu, J. Chen, Y. Wei, J. Li, D. Fan, Z. Yu, Y. Qi, Y. He, S. Xu, C. Yuan, Y. Zhou, J. Wang, M. Zhang, B. Su and Z. Liu, RSC Adv., 2014, 4, 21479–21491.
11. Q. Zhang, S. Hu, L. Zhang, Z. Wu, Y. Gong and T. Dou, Green Chem., 2014, 16, 77–81.
12. S. Svelle, F. Joensen, J. Nerlov, U. Olsbye, K.-P. Lillerud, S. Kolboe and M. Bjørgen, J. Am. Chem. Soc., 2006, 128, 14770–14771.
13. M. Bjørgen, S. Svelle, F. Joensen, J. Nerlov, S. Kolboe, F. Bonino, L. Palumbo, S. Bordiga and U. Olsbye, J. Catal., 2007, 249, 195–207.
14. C. Mei, P. Wen, Z. Liu, H. Liu, Y. Wang, W. Yang, Z. Xie, W. Hua and Z. Gao, J. Catal., 2008, 258, 243–249.
15. J. Kim, M. Choi and R. Ryoo, J. Catal., 2010, 269, 219–228.
16. A. Corma, Chem. Rev., 1997, 97, 2373–2420.
17. L. H. Chen, X. Y. Li, J. C. Rooke, Y. H. Zhang, X. Y. Yang, Y. Tang, F. S. Xiao and B. L. Su, J. Mater. Chem., 2012, 22, 17381.
18. D. H. Olson, G. T. Kokotailo, S. L. Lawton and W. M. Meier, J. Phys. Chem., 1981, 85, 2238–2243.
19. J. Caro, M. Noack, J. Richter Mendau, F. Marlow, D. Petersohn, M. Griepentrog and J. Kornatowski, J. Phys. Chem., 1993, 97, 13685–13690.
20. G. Muller, T. Narbeshuber, G. Mirth and J. A. Lercher, J. Phys. Chem., 1994, 98, 7436–7439.
21. L. Gueudré, T. Binder, C. Chmelik, F. Hibbe, D. M. Ruthven and J. Kärger, Materials, 2012, 5, 721–740.
22. M. Choi, K. Na, J. Kim, Y. Sakamoto, O. Terasaki and R. Ryoo, Nat. Mater., 2009, 461, 246–249.
23. K. Na, M. Choi, W. Park, Y. Sakamoto, O. Terasaki and R. Ryoo, J. Am. Chem. Soc., 2010, 132, 4169–4177.
24. S. Hu, J. Shan, Q. Zhang, Y. Wang, Y. Liu, Y. Gong, Z. Wu and T. Dou, Appl. Catal., A, 2012, 445–446, 215–220.
25. M. Tsapatsis, AIChE J., 2014, 60, 2374–2381.
26. R. Mostowicz and J. M. Berak, in Studies in Surface Science and Catalysis, Elsevier, 1985, vol. 24, pp. 65–72.
27. L. F. Petrik, C. T. O’Connor and S. Schwarz, in Studies in Surface Science and Catalysis, Elsevier, 1995, vol. 94, pp. 517–524.
28. G. Bonilla, I. Díaz, M. Tsapatsis, H. K. Jeong, Y. Lee and D. G. Vlachos, Chem. Mater., 2004, 16, 5697–5705.
29. X. Chen, W. Yan, X. Cao, J. Yu and R. Xu, Microporous Mesoporous Mater., 2009, 119, 217–222.
30. Z. C. Shan, H. Wang, X. J. Meng, S. Y. Liu, L. A. Wang, C. Y. Wang, F. Li, J. P. Lewis and F. S. Xiao, Chem. Commun., 2011, 47, 1048–1050.
31. Y. Liu, X. Zhou, X. Pang, Y. Jin, X. Meng, X. Zheng, X. Gao and F. S. Xiao, ChemCatChem, 2013, 5, 1517–1523.
32. P. Caullet, J. L. Paillaud, A. Simon-Masseron, M. Soulard and J. Patarin, C. R. Chim., 2005, 8, 245–266.
33. C. A. Fyfe, D. H. Brouwer, A. R. Lewis, L. A. Villaescusa and R. E. Morris, J. Am. Chem. Soc., 2002, 124, 7770–7778.
34. J. Arichi and B. Louis, Cryst. Growth Des., 2008, 8, 3999–4005.
35. B. Louis and L. K. Minsker, Microporous Mesoporous Mater., 2004, 74, 171–178.
36. H. Yu, X. Wang and Y. Long, Microporous Mesoporous Mater., 2006, 95, 234–240.
37. M. E. Dose, K. Zhang, J. A. Thompson, J. Leisen, R. R. Chance, W. J. Koros, B. A. McCool and R. P. Lively, Chem. Mater., 2014, 26, 4368–4376.
38. X. Q. Yao, X. Y. Xu, Z. Lu, K. Jiao, J. Q. Song, Z. F. Li, Q. Wang, L. J. Yan and M. Y. He, Acta Phys.-Chim. Sin., 2013, 29, 1809–1813.
39. S. Ivanova, C. Lebrun, E. Vanhaecke, C. Pham-Huu and B. Louis, J. Catal., 2009, 265, 1–7.
40. F. L. Bleken, S. Chavan, U. Olsbye, M. Boltz, F. Ocampo and B. Louis, Appl. Catal., A, 2012, 447–448, 178–185.
41. Z. Qin, L. Lakiss, L. Tosheva, J.-P. Gilson, A. Vicente, C. Fernandez and V. Valtchev, Adv. Funct. Mater., 2014, 24, 257–264.
42. H. Lechert, Microporous Mesoporous Mater., 1998, 22, 519–523.
43. J. L. Guth, H. Kessler, J. M. Higel, J. M. Lamblin, J. Patarin, A. Seive, J. M. Chezeau and R. Wey, in Zeolite Synthesis, American Chemical Society, 1989, vol. 398, pp. 176–195.
44. L. Wang, Y. Shao, J. Zhang and M. Anpo, Microporous Mesoporous Mater., 2007, 100, 241–249.
45. J. D. H. Donnay and D. Harker, Am. Mineral., 1937, 22, 446–467.
46. P. Hartman and W. Perdok, Acta Crystallogr., 1955, 8, 49–52.
47. J. H. Koegler, H. van Bekkum and J. C. Jansen, Zeolites, 1997, 19, 262–269.
48. D. G. Hay, H. Jaeger and K. G. Wilshier, Zeolites, 1990, 10, 571–576.
49. T. Ban, H. Mitaku, C. Suzuki, J. Matsuba, Y. Ohya and Y. Takahashi, J. Cryst. Growth, 2005, 274, 594–602.
50. M. A. Camblor, L. A. Villaescusa and M. J. Díaz-Cabañas, Top. Catal., 1999, 9, 59–76.
51. C. A. Fyfe, R. J. Darton, H. Mowatt and Z. S. Lin, Microporous Mesoporous Mater., 2011, 144, 57–66.
52. C. S. Cundy, B. M. Lowe and D. M. Sinclair, Faraday Discuss., 1993, 95, 235–252.
53. N. Alam and R. Mokaya, Microporous Mesoporous Mater., 2011, 143, 104–114.
54. P. L. Llewellyn, J. P. Coulomb, Y. Grillet, J. Patarin, G. Andre and J. Rouquerol, Langmuir, 1993, 9, 1852–1856.
55. G. Kyriakou and C. R. Theocharis, in Studies in Surface Science and Catalysis, Elsevier, 2002, vol. 144, pp. 709–716.
56. U. Müller and K. K. Unger, in Studies in Surface Science and Catalysis, Elsevier, 1988, vol. 39, pp. 101–108.
57. L. Delmotte, M. Soulard, F. Guth, A. Seive, A. Lopez and J. L. Guth, Zeolites, 1990, 10, 778–783.
58. J. L. Guth, L. Delmonte, M. Soulard, B. Brunard, J. F. Joly and D. Espinat, Zeolites, 1992, 12, 929–935.
59. J. I. Corredor, A. Cota, E. Pavon and M. D. Alba, Am. Mineral., 2013, 98, 1000–1007.
60. F. S. Xiao, S. Zheng, J. Sun, R. Yu, S. Qiu and R. Xu, J. Catal., 1998, 176, 474–487.
61. L. Ren, Q. Wu, C. Yang, L. Zhu, C. Li, P. Zhang, H. Zhang, X. Meng and F. S. Xiao, J. Am. Chem. Soc., 2012, 134, 15173–15176.
62. C. A. Fyfe, D. H. Brouwer, A. R. Lewis and J.-M. Chézeau, J. Am. Chem. Soc., 2001, 123, 6882–6891.
63. B. Louis, G. Laugel, P. Pale and M. M. Pereira, ChemCatChem, 2011, 3, 1263–1272.
64. K. B. Yoon, Acc. Chem. Res., 2007, 40, 29–40.
65. F. Schmidt, C. Hoffmann, F. Giordanino, S. Bordiga, P. Simon, W. Carrillo-Cabrera and S. Kaskel, J. Catal., 2013, 307, 238–245.
66. F. Yaripour, Z. Shariatinia, S. Sahebdelfar and A. Irandoukht, Microporous Mesoporous Mater., 2015, 203, 41–53.
67. Y. Yang, C. Sun, J. Du, Y. Yue, W. Hua, C. Zhang, W. Shen and H. Xu, Catal. Commun., 2012, 24, 44–47.
68. M. Westgård Erichsen, S. Svelle and U. Olsbye, Catal. Today, 2013, 215, 216–223.
69. Q. Zhu, J. N. Kondo, T. Setoyama, M. Yamaguchi, K. Domen and T. Tatsumi, Chem. Commun., 2008, 5164–5166.
70. S. Zhang, Y. Gong, L. Zhang, Y. Liu, T. Dou, J. Xu and F. Deng, Fuel Process. Technol., 2015, 129, 130–138.

---

Footnote

† Electronic supplementary information (ESI) available: SEM images of ZSM-5 synthesized with different factors; XRD patterns of the ZSM-5 zeolite synthesized in F⁻ and OH⁻ media; XRD patterns and N₂ adsorption/desorption isotherms of M-ZSM-5 obtained at different crystallization times; product selectivity for the MTP reaction and thermogravimetric curves of the deactivated zeolites are provided in Fig. S1–S9. See DOI: 10.1039/c5ra09561a

---