Zeolites are important heterogeneous catalysts in oil refining and petrochemical industries because of their unique pore structures,large surface area,strong acidity,and excellent hydrothermal stability ^{[1, 2]}. Mordenite (MOR),one of the most important members in the zeolite family,has exhibited excellent performance in catalysis,such as toluene disproportionation,hydroisomerization,alkylation,dimethylamine synthesis,and dimethyl ether carbonylation ^{[3, 4, 5]}. The channel system of MOR zeolite consists of a straight 12-membered ring (MR) channel (0.65 nm × 0.70 nm),a compressed 8-MR channel (0.26 nm × 0.57 nm) parallel to the c-axis,and an additional 8-MR channel (0.34 nm × 0.48 nm) along the b-axis ^[6]. However,MOR is generally referred to as a one-dimensional zeolite as the 8-MR channels are too small for most molecules to ingress. This unique pore structure may impose severe diffusion limitations in catalytic processes,leading to underutilization of the active sites and resulting in rapid deactivation,especially for the reactions involving bulky molecules ^{[7, 8, 9]}. In recent years,the quest for enhancing the catalytic efficiency and prolonging catalyst life has stimulated much research on improving the mass transport properties.

Integrating mesopores into the zeolite structures or decreasing the crystal size has proven to be an effective way to overcome the inherent diffusion limitations and improve the catalysis efficiency of zeolites ^{[10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23]}. Mesoporous MOR zeolites are hitherto mainly synthesized via a top-down approach through dealumination or sequential dealumination- desilication treatment ^{[9, 24, 25]}. A one-step,bottom-up synthesis of mesoporous MOR is highly desirable; however,this remains a challenging task,despite the significant developments in the synthesis of hierarchical zeolites,such as ZSM-5 and beta. Recently,Xiao and co-workers ^[26] successfully synthesized micro-sized MOR nanofiber bundles when assisted by a self-designed cationic copolymer containing quaternary ammonium groups. Jo and Ryoo et al. ^[27] reported a zeolite-seeding route to synthesize mesoporous MOR zeolite with nanorod-like morphology by employing organosilane surfactants as a mesopore-generating agent. Very recently,Li and co-workers ^[28] synthesized H-MOR nanosheet assemblies through a conventional hydrothermal reaction in the absence of templating agents. However,the above-mentioned MOR nanofibers and nanorods show a narrowed a-b cross section,whereas the growth along the c-axis direction is less inhibited,suggesting that the 12-MR channels in these hierarchical MOR zeolites have the longest dimension and implying that the mass transport in this direction remains restricted.

Synthesis of zeolite nanocrystals with reduced size in three dimensions decreases the diffusion pathway irrespective of the direction of travel. There are,however,only a few examples reported where nanocrystalline MOR have been effectively synthesized in such a manner. Kubota’s group ^[29] has reported the preparation of MOR nanocrystals of ~80 nm in diameter using a hydrophobic structure-directing agent,N,N,N’,N’- tetraethylbicyclo[2.2.2]oct-7-ene-2,3:5,6-dipyrrolidinium. In addition,the divalent cation N,N,N,N’,N’,N’- hexaethylpentanediammonium was also reported to direct the synthesis of nanoscale MOR ^{[30, 31]}. Recently,Ryoo et al. ^[32] developed a multivalent surfactant-capping route to synthesize nanocrystalline zeolites including MOR,FAU,CHA,and MFI. By using a self-designed prolinol derivative surfactant C₁₈H₃₇N⁺(Me)₂C₆H₁₂N⁺(CH₃)₂ CH₂C₆H₄CH₂N⁺(CH₃)(C₄H₇CH₂OH) as a capping agent,MOR nanocrystals could be readily prepared. They attributed the nanocrystal formation to the multiple valencies on the surfactant and thereby the introduction of a strong binding effect of the surfactant molecules on the substrate surfaces compared with monovalent surfactants.

Herein,we report a facile and direct approach to synthesize nanocrystal-assembled hierarchical MOR zeolites by subjecting the crystallization to relatively low temperatures (130 °C) in the presence of tetraethylammonium hydroxide (TEAOH) and commercial surfactant. The role of TEAOH and surfactants in the generation of the hierarchical assembled structure is investigated in detail and the physicochemical properties of the products are well characterized. The catalytic performance of the hierarchical MOR zeolites is evaluated via the benzylation of benzene with benzyl alcohol (BA) and compared with the control sample and hierarchical MOR having nanosheet- assembled morphology with a longer c-direction dimension. In addition,the hierarchical zeolite is applied as a catalyst for the dimethyl ether (DME) carbonylation to methyl acetate (MA) reaction,which is a novel route to produce ethanol from syngas.

The chemical reagents used in the experiments included NaAlO₂ (54.5% Al₂O₃,40.9% Na₂O,Tianjin Guangfu Chemical Co.),colloidal silica (30.1% SiO₂,Qingdao Chengyu Chemical Co.),TEAOH (25.0% in water,Xiamen Pioneer Technology Co.,Ltd.),NaOH (Tianjin Kemiou Chemical Reagent Co.),C₁₂H₂₅- N⁺(CH₃)₂-C₂H₄-N⁺(CH₃)₂-C₁₂H₂₅Br₂ (C_12-2-12,Henan Titaning Chemical Technology Co.),hexadecyltrimethylammonium bromide (CTAB,Energy Chemical),sodium dodecylbenzenesulphonate (SDBS,Energy Chemical),and Pluronic F-127 (M_r = 12600 g/mol,Sigma). All the above reagents were used without further purification.

A typical synthesis procedure was as follows. NaOH and surfactant (C_12-2-12) were dissolved in H₂O to prepare a clear solution. Thereafter,TEAOH was added to the solution dropwise followed by the addition of NaAlO₂. Then,colloidal silica was added to the mixture under vigorous stirring. The final mixture with a composition of SiO₂:0.033Al₂O₃:0.25Na₂O: 0.10TEAOH:0.05C_12-2-12:40H₂O was stirred for additional 3 h to obtain a homogenous gel prior to being transferred into a stainless steel autoclave. The sol-gel was subjected to hydrothermal crystallization at 130 °C for 5-7 d and rotated at 60 rpm. The products were recovered by centrifugation,washed with deionized water,dried overnight at 120 °C and finally calcined at 550 °C for 5 h to remove the organic species. The protonated samples were obtained by ion exchange with NH₄NO₃ aqueous solution (1 mol/L,1 g calcined sample in 10 mL solution) at 80 °C under stirring. The ion exchange was repeated three times and every ion exchange solution was for 2 h. The samples were recovered by centrifugation,washed with deionized water,and dried at 120 °C before calcination at 550 °C for 4 h.

Powder XRD patterns were recorded on a PANalytical X’Pert PRO X-ray diffractometer with Cu K_α radiation (λ = 0.154059 nm) and a scan step width of 0.026°. The chemical composition of the samples was measured using a Philips fluorescence (XRF,Magix-601 X-ray) spectrometer. Morphology was observed using scanning electron microscopy (SEM,SU8020). N₂ adsorption-desorption isotherms were obtained on a Micromeritics ASAP 2020 system at -196 °C. The total surface area was calculated based on the BET model. Micropore volume and micropore surface area were calculated using the t-plot method. Pore size distribution of the mesopores was determined by the BJH method based on the desorption branch of the isotherms.

Solid state NMR experiments were performed on a Varian Infinity Plus 400 WB spectrometer with a BBO MAS probe operating at a magnetic field strength of 9.4 T. The resonance frequencies were 104.2 and 79.4 MHz for ²⁷Al and ²⁹Si,respectively. Chemical shifts were referenced to Al(NO₃)₃ (1.0 mol/L) for ²⁷Al and 2,2-dimethyl-2-silapentane-5-sulfonate sodium salt (DSS) for ²⁹Si. The spinning rates of the samples at the magic angle were 4 and 6 kHz for ²⁷Al and ²⁹Si,respectively.

Temperature-programmed desorption of ammonia (NH₃- TPD) was measured on a Micromeritics 2920 chemical adsorption instrument. After pre-treating the samples at 650 °C for 1 h under flowing He,the samples (40-60 mesh,0.20 g) were cooled down to 150 °C and saturated with NH₃ gas. Thereafter,NH₃-TPD was performed in a constant flow of He (20 mL/min) from 150 to 650 °C at a heating rate of 10 °C/min.

1,3,5-triisopropylbenzene (TIPB) cracking was performed in a fixed-bed reactor at atmospheric pressure. The sample (40-60 mesh,200 mg) was loaded into the reactor and dehydrated in N₂ (50 mL/min) at 500 °C for 1 h prior to the reaction. The temperature of the catalyst bed was adjusted to 210 °C under N₂ and fed with vaporized TIPB (85 °C) in N₂ (50 mL/min). The products were analyzed online by gas chromatography (Agilent 6890) attached to an FID detector using an HP-5 column.

DME carbonylation experiments were performed in a continuous flow fixed-bed stainless steel reactor. The catalyst (40-60 mesh,1.0 g) was packed in the reactor and pre-treated under N₂ at 500 °C for 2 h. The catalyst sample was cooled to 200 °C prior to a reactant gas mixture (5% DME,35% CO,and 60% H₂) being introduced into the reactor at a gas hourly space velocity (GHSV) of 1500 mL/(g·h). Reaction pressure was 2.0 MPa. The outlet gas was analyzed online using a gas chromatograph (Agilent 7890A) equipped with TCD and FID detectors.

The liquid phase benzylation of benzene with BA was performed in a three-necked round-bottom flask equipped with a reflux condenser and heated in an oil bath at 80 °C under atmospheric pressure. In a typical run,34 mL of benzene was added with 0.20 g catalyst,which had been pre-calcined at 500 °C for 4 h. The reaction mixture was stirred at 80 °C for 30 min,followed by the addition of 2.0 mL of BA. Liquid samples were withdrawn at intervals. The liquid products were analyzed by gas chromatography using an Agilent 7890 connected to an FID detector using a 50-m HP-5 column. The products were identified by GC-MS analysis.

Nanocrystal-assembled hierarchical MOR zeolite (designated as MOR-M) was hydrothermally synthesized in the presence of TEAOH and surfactant C_12-2-12 at 130 °C. The reproducible synthetic conditions are summarized in Table 1. For comparison,the material synthesized under similar conditions in the absence of TEAOH and surfactant is used as a control sample (designated as MOR-C,sample 12 in Table 1). Fig. 1 shows the XRD patterns of MOR-M (sample 1) and MOR-C,which exhibit characteristic diffraction peaks of mordenite structure and confirm the phase purity of the samples. SEM micrographs provide detailed information on the morphology of the samples. As shown in Fig. 2(a),MOR-M consists of near-spherical particles with rough surfaces ranging from 10 to 20 μm. Increased magnification (Fig. 2(b)) reveals that these aggregates are composed of interconnected nanocrystallites,forming an abundance of interstitial voids,generating a hierarchical structure. Fig. 2(c) displays the TEM micrograph of MOR-M,from which nanocrystals possessing 20 to 50-nm dimensions could be clearly observed. This nanocrystal-assembled morphology is believed to be favorable for catalytic reactions because of the shortened diffusion path. Furthermore,the formation of the nanocrystal assemblies benefits solid-liquid separation after the synthesis. Indeed,the spherical architectures of MOR-M are very stable,preserving their morphology almost intact after ultrasonic treatment for 30 min. Conversely,the MOR-C sample possesses plate-like morphology consisting of closely packed crystallites (Fig. 2(d)). The distinct morphology between the two samples suggests that the nucleation and growth of MOR crystals are significantly modified by the introduction of TEAOH and C_12-2-12.

Table 1 Summary of synthetic conditions and textual properties of the MOR products. Sample Starting gel composition a Si/Al b Surface area c (m2/g) Pore volume c (cm3/g) x y SBET Smicro Sext Vmicro Vtotal 1 (MOR-M) 0.1 0.05 (C12-2-12) 9.1 471 369 102 0.17 0.62 2 0.1 0 8.1 428 352 76 0.16 0.43 3 0 0.05 (C12-2-12) 8.8 524 302 222 0.14 0.76 4 0.1 0.10 (C12-2-12) — 477 371 106 0.17 0.69 5 0.23 0.05 (C12-2-12) — 425 330 95 0.16 0.71 6 0.1 0.02 (CTAB) — 493 316 177 0.15 0.48 7 0.1 0.05 (SDBS) — 478 343 135 0.16 0.44 8 0.1 0.40 d (F-127) — 460 340 120 0.16 0.40 9 e 0 0 — 390 367 23 0.17 0.20 10 e 0.1 0 — 399 364 35 0.17 0.21 11 e 0.1 0.05 (C12-2-12) — 408 370 38 0.17 0.23 12 (MOR-C) 0 0 7.0 384 344 40 0.16 0.26 a The starting gel composition is SiO2 : 0.033 Al2O3 : 0.25 Na2O : x TEAOH : y surfactant : 40 H2O (130 °C,7 d). b Determined by XRF. c SBET corresponds to the BET surface area. Smicro and Vmicro are micropore surface area and micropore volume determined by the t-plot method,respectively. Sext = SBET - Smicro. Vtotal is calculated at p/p0 = 0.99. d Mass ratio of F-127/SiO2. e Control experiments conducted at 180 °C for 20 h.

Table 1 Summary of synthetic conditions and textual properties of the MOR products.

Fig. 1. XRD patterns of the as-synthesized samples MOR-M and MOR-C.

Fig. 2. (a) Low-magnification SEM micrograph of MOR-M; (b) High-magnification SEM micrograph of MOR-M; (c) TEM micrograph of MOR-M; (d) SEM micrograph of MOR-C.

Fig. 3 presents the N₂ sorption isotherms and BJH pore size distribution curve of the samples. MOR-M exhibits the type IV isotherm with a hysteresis loop at a relative pressure of 0.80-0.95,which demonstrates the existence of intercrystal meso/macroporosity ^[33]. The corresponding BJH pore size distribution curve derived from the desorption branch indicates that it is rather broad with a mean pore size of ~30 nm. Comparatively,MOR-C presents a type I adsorption curve typical of microporous materials with a small uptake of N₂ at high relative pressure. Table 1 shows the textual properties of the samples. In comparison with the control sample MOR-C,MOR-M exhibits a higher BET surface area (471 m²/g) and larger pore volume (0.62 cm³/g) owing to the decreased crystal size and the formation of meso- and macropores. It is envisaged that the abundant porosity associated with the nano-sized crystals in MOR-M offers improved ingress of guest molecules into the pores,enhancing diffusion and mass transfer,thus improving the catalytic efficiency.

Fig. 3. (a) N2 adsorption-desorption isotherms of MOR-M and MOR-C; (b) The BJH pore size distribution curve of MOR-M calculated from the desorption branch.

Solid state ²⁷Al and ²⁹Si MAS NMR spectra were employed to investigate the local atomic coordination environments of MOR-M and MOR-C. As shown in Fig. 4(a),there only exists one sharp peak centered at δ= 54 in the ²⁷Al MAS NMR,corresponding to tetrahedral coordinated aluminum species. No peak around δ = 0 is observed,which indicates that all the Al atoms were incorporated in the framework. Two major peaks centered at -112 and -105 and a less intense peak at -99 are evident in the ²⁹Si MAS NMR spectra,which are ascribed to the Si(0Al),Si(1Al),and Si(2Al) species,respectively. Notably,the Si(0Al) signal in the spectrum of the MOR-M sample is more intense than that of MOR-C,implying a higher Si/Al ratio of the former material ^[34]. The results are consistent with the XRF data—the Si/Al ratio being 9.1 and 7.0 for MOR-M and MOR-C,respectively.

Fig. 4. 27Al and 29Si MAS NMR spectra of MOR-M and MOR-C.

To elucidate the role of TEAOH and C_{12-2 -12} surfactant in the formation of nanocrystal-assembled hierarchical structures,control experiments were conducted with a single TEAOH template,a single C_12-2-12 surfactant and different ratios of the two organics while keeping other conditions unchanged. The experimental conditions and textual properties of the products are listed in Table 1. Fig. 5 illustrates the SEM micrographs of the samples. Sample 2 synthesized with a TEAOH/SiO₂ ratio of 0.1 shows an aggregated morphology similar to MOR-M,implying that nanocrystal-assembled mordenite could be obtained with low concentrations of TEAOH at 130 °C. Thermal analysis confirms the existence of TEA⁺ in the product with a mass loss of 8.2% in the temperature range of 300-700 °C. The surface area (428 m²/g) and pore volume (0.43 cm³/g) of sample 2 are smaller than that of MOR-M,possibly owing to larger nanocrystal sizes (50-70 nm). Sample 3 presents variation in its morphology as nanosheet assemblies are formed upon the addition of C_12-2-12 surfactant only in the initial gel. From Fig. 5(b),it can be seen that each individual nanosheet has a thickness of ~30 nm and a length of ~500 nm. The corresponding FFT diffractograms reveal that the longest dimension of the nanosheet runs along the c-axis direction. This is expected considering that multivalent surfactants could serve as a capping agent to direct the formation of MOR nanorods and nanosheets (having a preferential growth along the c-direction) as reported by Ryoo and co-workers ^[32]. The nanosheet assembly exhibits much larger external surface area and higher mesopore volume than MOR-M and MOR-C. Further increasing the concentration of C_12-2-12 or TEAOH in the gel has little effect on enhancing the mesoporosity of the products (samples 4 and 5). It is thus concluded that the smaller molecule TEAOH alone can decrease the MOR crystal size and induce hierarchical structure formation. The addition of the surfactant to the sol-gel is not essential for the formation of nanocrystal assemblies; however,the combined presence of TEAOH further inhibits crystal growth improving the external surface area. TEAOH is generally used in the synthesis of high-silica MOR zeolites ^{[35, 36]},while the use of TEAOH to form hierarchical MOR remains less explored. Herein,it is suggested that TEAOH functions as a micropore-directing agent,which induces significantly more nuclei,leading to smaller crystal sizes. Additionally,TEA⁺ has been reported to adsorb on the crystal surface ^{[37, 38]},preventing excessive interconnection among nanocrystallites,prompting the formation of the final morphology.

Fig. 5. SEM micrographs of sample 2 (a),3 (b),6 (c),7 (d),8 (e),9 (f),10 (g),and 11 (h) in Table 1. The inset in (b) shows the HRTEM micrograph and the corresponding selected FFT diffractogram viewed along the (100) axis.

Interestingly,the growth inhibition effect of the surfactant is not limited to C_12-2-12. By applying the present synthetic strategy,nanocrystal-assembled MOR could be hydrothermally synthesized with various conventional surfactants,such as CTAB,SDBS,and F-127 (samples 6-8 in Table 1). All three samples present irregular micrometer-sized morphology composed of very small nanoparticles (Fig. 5). The type IV isotherms associated with the BJH pore size distribution of the samples are given in Fig. 6,which clearly indicate the presence of mesopores in the structures. Comparatively,the single use of each surfactant in the absence of TEAOH could not produce the nanocrystal-assembled hierarchical structure. This confirms again that the presence of TEAOH is vital to the generation of the nanocrystal assembly,whereas the surfactant serves more as a growth-inhibiting agent possibly through interaction with the negatively charged crystal surface ^[39] or the TEA⁺ adsorbed on the surface ^{[37, 38]}.

Fig. 6. (a) N2 adsorption-desorption isotherms and (b) the pore size distribution curves of the samples. The pore size distributions are derived from the desorption branches of the isotherms.

Further samples were synthesized at higher temperatures (180 °C) to investigate the effect of crystal growth behavior as a function of crystallization temperature (samples 9-11 in Table 1). All products exhibit distinct morphology from the corresponding samples synthesized at 130 °C possessing the same sol-gel compositions. Sample 9 synthesized in the absence of TEAOH and C_12-2-12 displays elliptic micrometer-sized morphology with smooth surfaces. After the introduction of TEAOH,the product (sample 10) possesses rough surfaces composed of closely packed nanoscale particles. A low external surface area of 35 m²/g and a dramatically reduced pore volume of 0.21 cm³/g are observed,indicating that the adsorption of TEA⁺ on the crystal surface and its inhibition on crystal intergrowth may be significantly weakened at higher temperatures. In addition,sample 11 synthesized with the addition of both TEAOH and C_12-2-12 displays well-defined morphology with overlapped slabs on the particle surface. Here the external surface area decreases to 38 m²/g. These results demonstrate that the morphology and porosity of the products are sensitive to the crystallization temperature with lower temperatures favoring the formation of the present hierarchical structure ^{[40, 41]}.

NH₃-TPD experiments were performed to investigate the acid properties of the samples. Fig. 7 presents the NH₃-TPD profiles of MOR-M and MOR-C. Both data sets exhibit two peaks assigned to weak and strong acid sites,respectively. The stronger acid strength associated with the MOR-M sample may relate to variation in framework Al distribution resulting from the use of the TEAOH template ^[42]. In addition,MOR-M has a lower acid concentration compared with the control sample,owing to its slightly higher Si/Al ratio.

Fig. 7. NH3-TPD curves of the protonated MOR-M and MOR-C samples.

The acid sites located on the external surface of the samples were further evaluated by the TIPB cracking reaction. As the TIPB molecule possesses a kinetic diameter of 0.95 nm,ingress to the 12-MR channel of the mordenite zeolite is prevented and thus the first-order cracking of TIPB selectively occurs on the external acid sites ^[43]. Fig. 8 illustrates the conversion of TIPB as a function of time on stream. MOR-M shows a high initial conversion of 77.5%,whereas only 29.1% conversion is obtained on MOR-C. After the reaction had progressed for 3 h,a clear decrease of the TIPB conversion occurs over both catalysts. The conversion on MOR-M drops to 34.4%,maintaining 44% of the initial conversion. However,the MOR-C standard after 3 h on stream gives only 6.8% TIPB conversion,corresponding to 23.3% of the fresh conversion value. The superior catalytic activity of MOR-M and the ability to resist deactivation in TIPB cracking when compared with the conventional standard confirms the higher concentration of acid sites located on the external surface of the sample.

Fig. 8. Catalytic cracking conversion of 1,3,5-triisopropylbenzene (TIPB) over protonated MOR-M and MOR-C catalysts. Reaction conditions: vaporized TIPB in nitrogen (50 mL/min,85 °C) as feed,210 °C,0.20 g catalyst.

The liquid phase benzylation of benzene with BA was performed to investigate the catalytic properties of the samples in the acid-catalyzed reactions involving bulky molecules. As presented in Fig. 9,MOR-M exhibits significantly improved catalytic activity than the control sample,MOR-C,during the entire reaction time. The BA conversion is 2.8% at t = 0.5 h and gradually rises to 12.0% with reaction time,prolonged to 4 h,when using MOR-C as the catalyst. Comparatively,the conversion of BA over the hierarchical MOR-M increases rapidly as a function of time and a complete conversion of 100% is readily achieved at a reaction time of 3 h,showing the enhanced catalytic efficiency over the conventional standard. Moreover,the selectivity of the target product diphenylmethane (DPM) over MOR-M reaches close to 90%,15%-30% higher than the corresponding MOR-C sample throughout the reaction. In principle,the acid sites responsible to polarize BA to produce an electrophilic intermediate play a crucial role in the present reaction ^{[9, 44]}. However,the observed catalytic performance does not correlate well to the acid properties of the catalysts. Considering that both MOR-M and MOR-C possess similar acid strength,and the former possesses a lower acid concentration than the latter,the improved activity and selectivity than MOR-C is thought to be attributed to the shortened diffusion path together with a higher concentration of acid sites on the external surface.

Fig. 9. A comparison of the catalytic performance of the protonated samples in the benzylation of benzene with BA. Reaction conditions: 1.0 mL BA,34 mL benzene,0.20 g catalyst,80 °C.

To further demonstrate the excellent catalytic performance of the present nanocrystal-assembled hierarchical MOR-M,samples 2 and 3 were also tested as catalysts and the results are given in Fig. 9. Sample 2,synthesized from a TEAOH/SiO₂ ratio of 0.1,displays good catalytic activity,being only slightly lower than MOR-M. This is somewhat expected as sample 2 has similar morphology,but with a smaller external surface area and pore volume compared with MOR-M. A further reduction in activity is observed over sample 3,which converts less BA than both MOR-M and sample 2 until a conversion of 100% is reached at t = 5 h. Here,the diffusion limitation is believed to be the main reason for the inferior catalytic activity of sample 3 with the highest external surface area and largest pore volume among the samples. The nanosheet-assembled morphology of sample 3,in which the 12-MR channels run parallel to the longest dimension,would result in a relatively slow diffusion and thus slow down the reaction inside the channels. These results clearly demonstrate that decreasing the crystal size of MOR zeolites,especially in the c-direction,effectively enhances mass transfer in the 12-MR channels,producing superior catalytic activity as observed with MOR-M.

The carbonylation of DME to MA is a key step to produce ethanol from syngas. Mordenite zeolite has proven to be the most promising catalyst with high activity and MA selectivity ^[45]. According to the literature,the target reaction exclusively proceeds in the 8-MR side-pockets of MOR zeolite,whereas the acid sites in the 12-MR main channels prompt the occurrence of side reactions resulting in coke formation and fast deactivation ^{[46, 47]}. It is expected that the shortened diffusion path in MOR-M facilitates mass transfer and enhances catalyst activity and stability.

Pre-adsorption of pyridine on the acid sites located in the straight channels has been verified to be an effective way to retard the catalyst deactivation and significantly prolong the lifetime ^[46]. However,here,the catalytic reactions are conducted without pyridine pre-adsorption,to clearly compare the catalytic performance of the hierarchical MOR-M and MOR-C. Fig. 10 presents the DME conversion and MA selectivity as a function of time on stream over the samples. It is evident that MOR-M exhibits significantly improved activity compared with MOR-C during the entire run time. An induction period of about 4 h is observed on MOR-M,during which the conversion of DME increases from 35.2% at 1.5 h to 40.4% at 4 h. Thereafter,the conversion of DME gradually drops to 13.6% at 23.5 h,maintaining 33.6% conversion. For MOR-C standard material,the conversion of DME is relatively stable during the first 3 h (~30%),which then finally decreases to 4.2% after a reaction time of 23.5 h. This value amounts to only 13.9% of the initial conversion,implying a faster deactivation over MOR-C than on the hierarchical sample. The improved activity and anti-deactivation ability of MOR-M should arise from the reduced diffusion limitations and the abundant porosity,which leads to a more efficient utilization of the acid sites in the side-pocket 8-MR channels ^[31].

Fig. 10. DME conversion and MA selectivity of DME carbonylation over protonated MOR-M and MOR-C catalysts. Reaction conditions: 200 °C,2.0 MPa,DME/CO/H2 = 5/35/60 (vol%),GHSV = 1500 mL/(g·h).

The selectivity of MA over MOR-M and MOR-C also shows differences,though both samples display a similar trend following the decrease of DME conversion. MOR-M is able to sustain a high level of MA selectivity over a longer period of time than on MOR-C. The selectivity drops to 59.8% on the hierarchical zeolite at 23.5 h,whereas only a selectivity of 40.3% on the conventional zeolite is observed. The results indicate that the side reactions,which occur in the 12-MR channels,are likely to be suppressed by the improved mass transport of the product. The coke content on the catalysts after 23.5 h of reaction was further detected by TG analysis. Three peaks are observed in the DTG curves (Fig. 11),corresponding to physisorbed water,soft cokes,and hard cokes,respectively ^[48]. The total coke amount was 6.5 wt% for the deactivated MOR-M,while a much higher coke deposition was found on MOR-C (11.3 wt%). This implies that the coke formation is much slower on MOR-M,further confirming the advantage of the shortened diffusion length.

Fig. 11. TG-DTG curves of the coked catalysts after 23.5 h of DME carbonylation reaction.

We developed a facile synthetic strategy to fabricate hierarchical MOR zeolites that results in nanocrystal-assembled morphology composed of 20-50-nm crystallites dramatically reducing mass transport limitation. Detailed investigation shows that the use of TEAOH itself at lower crystallization temperatures could decrease the crystal size and generate hierarchical structures constructed by the interconnection of nanocrystals. The introduction of the cationic C_12-2-12 surfactant in the TEAOH-containing sol-gel further reduces the crystal size and improves the porosity of the products. However,the surfactant used in the present system is not limited to C_12-2-12. Several commercial surfactants including CTAB,anionic SDBS and neutral F-127 also exhibit the ability to decrease the nanocrystal size when used in conjunction with TEAOH. More importantly,the nanocrystal assemblies show much higher catalytic activity in benzylation of benzene with BA in comparison with the control sample,and the hierarchical MOR possesses a longer c-axis dimension and larger external surface area. The results demonstrate that the decreased crystal size of the MOR zeolite,especially in the c-direction,could effectively prompt mass transport in the 12-MR channels,leading to superior catalytic efficiency. In addition,the material also displays enhanced activity and anti-deactivation behavior in the DME carbonylation reaction. This attractive approach,employing commercially available templates and surfactants,provides an effective and viable way to synthesize MOR nanocrystal assemblies with excellent catalytic performance.