Ethylbenzene is an important basic organic monomer raw material in the chemical industry ^[1]. Ethylbenzene is mainly used for catalytic dehydrogenation to produce styrene, which accounts for more than 90% of the total styrene production capacity. A small amount of ethylbenzene is used as a solvent or for dilution and organic synthesis, such as in the production of diethylbenzene. Styrene is an important monomer of high-molecular-weight polymers ^[2], and the market demand for styrene has shown an increasing trend annually, which directly drives the production of ethylbenzene. At present, ethylbenzene is mainly prepared by ethylene and benzene alkylation, including pure ethylene and dilute ethylene (dry gas) as the raw material. The latter is a mature process mainly used in current industry, among which the SGEB (dry gas phase ethylbenzene production process) independently developed by Sinopec is the most typical ^[3]. Ethylene has been widely used as an alkylating agent, but the uneven distribution of ethylene resources makes it difficult to effectively promote or implement in areas lacking ethylene resources ^[4−5]. As a compensation resource, ethanol is a highly active alkylating agent with a wide range of sources. Traditional ethanol mainly comes from grain fermentation, and the successful production of domestic coal-based ethanol products and the important progress in bioethanol technology provided additional possibilities for the further utilization of ethanol. The method of producing ethylbenzene by the reaction of ethanol with benzene has been widely studied ^[6−8]. One-step gas-phase synthesis of ethylbenzene from ethanol and benzene reduces the investment and operation costs of ethanol dehydration equipment. In addition, ethanol, as an alkylating agent, is relatively simple to use in terms of raw material transportation and convenient storage, which provides a new route for ethylbenzene synthesis in areas with excess ethanol capacity and a lack of ethylene resources.

The alkylation of benzene and ethanol follows the mechanism of a positive carbon ion reaction ^[9], and zeolite catalysts (such as ZSM-5, β, Y, and MCM-22 zeolites) are generally used in related research ^[10−14]. Among them, the ZSM-5 zeolite has been studied the most and has attracted the most attention. The alkylation performance of benzene and ethanol is affected by many factors, such as reaction temperature, the benzene-alcohol ratio of the raw material, space velocity, and pressure, among which the acidity of the ZSM-5 catalyst plays a decisive role. Studies have shown that the side reactions of alkylation reactions, such as the cleavage, disproportionation, and isomerization reactions of ethylbenzene and polyalkyl- benzene, occur mainly at the strong acid sites of catalysts, while lower numbers of strong acid sites and higher B-acid density are conducive to improving the selectivity of ethylbenzene ^[15]. Zhang et al.^[15] reported that selective dealuminization or the introduction of P effectively reduced the strong acidity of ZSM-5 zeolite and improved the selectivity of ethylbenzene. Moreover, P also improved the hydrothermal stability of the molecular sieve and increased the life of the catalyst. The ZSM-5 zeolite has two pore structures, straight and sinusoidal. Through molecular dynamics simulations of the alkylation reaction between benzene and methanol ^[16], small molecules tend to diffuse along straight and Z-shaped pore channels, while benzene and alkyl benzene diffuse only along straight pore channels. In addition, the anisotropic length of ZSM-5 zeolite affects the conversion, selectivity, and stability of the catalytic reaction, and differences in synthesis conditions affect the characteristics of the molecular sieve itself, such as zeolite structure and acidity. Huang et al.^[17] investigated the catalytic performance of ZSM-5 zeolites with different morphologies and reported that nanosheet or nanospheroidal ZSM-5 zeolites with larger specific surface areas had greater gasoline selectivity and significantly improved one-way life. Zheng et al. ^[18] described the preparation of multistage porous ZSM-5 molecular sieve and its application in the alkylation reaction of benzene and ethylene in oil refining in detail. Li et al.^[19] introduced the research progress of catalysts for benzene and methanol alkylation from Si/Al, crystal size and modification of ZSM-5. Similarly, Zhu et al.^[20] demonstrated that ZSM-5 with a short b-axis (thickness of 90 nm) exhibited better catalytic performance than traditional ZSM-5 in the alkylation of benzene with ethanol, which was attributed to its short diffusion path and optimized acidity.

During the alkylation of benzene and ethanol, the distribution of acid sites affects the catalytic performance. Our previous study revealed ^[21] that the existence of annular ZSM-5 zeolite hollow structures changed the distribution of B and L acids in molecular sieves, thus affecting the selectivity of ethylbenzene. In addition to the above influencing factors, the diffusion behavior of raw materials and products also significantly affects the activity, selectivity, and catalyst stability of the reaction.

Based on the above analysis, four kinds of ZSM-5 zeolites with the same Si-Al ratio, similar acidity, and different morphologies and sizes were synthesized for the alkylation of benzene and ethanol by the conventional hydrothermal synthesis method, including ring ZSM-5 zeolites, nano-sheet ZSM-5, nanoaggregate-spheroids, and small-grain ZSM-5 zeolites. The effects of the morphology and size of ZSM-5 zeolites on the alkylation of benzene-ethanol were evaluated from the perspectives of product diffusion and the pore structure of the molecular sieve.

1 Experimental 1.1 Experimental materials

Tetraethyl orthosilicate (TEOS, 98%) was provided by Tianjin Kemiou Chemical Reagent Co., Ltd. Sodium meta-aluminate (NaAlO₂) was supplied by Shanghai Titan Tech-nology Co., Ltd. Tetrapropyl ammonium hydroxide (TPAOH, 25%) was obtained from Zhejiang Kent Chemical Co., Ltd. Sodium hydroxide (NaOH), ammonium chloride (NH₄Cl), benzene (C₆H₆), ethanol (C₂H₅OH), and sodium chloride (NaCl) were purchased from Shanghai Titan Technology Co., Ltd. Deionized water was made in the laboratory.

1.2 Preparation of zeolites

Preparation of cyclic ZSM-5 zeolites: Annular hollow ZSM-5 zeolite powder with n(SiO₂)∶n(Al₂O₃)=60 was synthesized according to the methods described in the literature [21]. NaOH and NaAlO₂ of a certain mass were weighed at room temperature, an appropriate amount of deionized water and TPAOH were added, and the mixture was fully stirred until completely dissolved. During the process of drip and stir, TEOS was added to the precursor solution in batches at a low temperature (0 ℃). After dropping and stirring at a low temperature for 4 h, a white gel molecular sieve crystallization initial gel was obtained. The molar ratio of the raw material in the gel solution was 10 TPAOH∶30 SiO₂∶0.5 Al₂O₃∶1 000H₂O∶2 NaOH. The initial crystallization liquid was transferred to a hydrothermal synthesis reactor and fixed in a homogeneous reactor, and the crystallization product was obtained by dynamic rotation at 30 r∙min⁻¹ at 170 ℃ for 24 h. The crystallization products were subsequently washed in deionized water to neutral and dried in a drying oven at 110 ℃ for 8 h. A Cyclic ZSM-5 zeolite was obtained by calcining at 550 ℃ for 6 h, denoted as C-Z5.

Preparation of lamelliform ZSM-5 zeolites: According to the method ^[20], a thin-sheet ZSM-5 zeolite with n(SiO₂)∶n(Al₂O₃) = 60 was synthesized, and the molar composition of the initial gel was 5.4 TPAOH∶30 SiO₂∶0.5 Al₂O₃∶800 C₂H₅OH∶1340 H₂O∶1 NaOH. The initial gel preparation process was as follows: a molar ratio of NaAlO₂, NaOH, TPAOH, and C₂H₅OH was weighed and mixed with deionized water and stirred at room temperature for 1 h until the cloudy solution became clear. Then, TEOS was slowly added to the solution, and the precursor gel was obtained after aging at room temperature for 4 h. Finally, the precursor gel was transferred to a hydrothermal synthesis vessel and crystallized at 170 ℃ for 24 h, after which the crystallization product was obtained. The product was centrifuged, washed with deionized water to a neutral pH, transferred to a drying oven, dried at 80 ℃ for 8 h, and then roasted at 550 ℃ for 6 h, the lamelliform ZSM-5 zeolites (L-Z5) were obtained.

Preparation of spheroidal nano-aggregates ZSM-5 zeolites: According to the methods mentioned in the literature [22], multistage porous ZSM-5 zeolites with n(SiO₂)∶n(Al₂O₃)=60 were synthesized. The molar composition of the initial gel was 5 TPAOH∶60 SiO₂∶1 Al₂O₃∶10 Na₂O∶7856 H₂O. The preparation process of the initial gel was as follows: A certain molar ratio of TPAOH, NaAlO₂, and deionized water was mixed until the clear liquid was clear, TEOS was added dropwise, and the precursor gel was obtained by stirring for 0.5 h. The precursor gel solution was aged at 100 ℃ for 3 h and then transferred to a high-pressure synthesis reactor for crystallization at 180 ℃ for 3 h. The crystallization products were obtained. The crystallization products were washed with deionized water to a neutral pH, dried in a drying oven at 110 ℃ for 8 h, and calcined at 550 ℃ for 6 h to obtain spheroid-nanoaggregate ZSM-5 zeolites (N-Z5).

Preparation of conventional small-grain ZSM-5 zeolites: The preparation method of conventional small-grain zeolites is the same as that in the literature [21]. However, TEOS was added at room temperature to obtain small-particle ZSM-5 zeolites, which were recorded as S-Z5.

1.3 Characterization

The crystal structure of the samples was analyzed by a Rigaku D/Max-3c X-ray diffraction analyzer. The parameters were set as Cu and Kα 1 rays, wavelength λ of 0.154 06 nm, room temperature, tube voltage of 40 kV, tube current of 40 mA, 2θ scanning range of 3°~50°, scanning step of 0.02°, and tracing rate of 0.2 (°)·s⁻¹.

Cold field emission scanning electron microscopy (SEM), a JEM-2100F field emission transmission electron microscope (TEM), and an Oxford Instruments Limited Oxford/ULTIM MAX instrument were employed to observe the morphologies of the samples. An energy dispersive spectrometer (EDS) was used to analyze the composition of the microregion on the surface of the sample. Before the TEM test, the sample was first ground, dissolved in anhydrous ethanol for ultrasonic dis-persion, dropped on a copper net, and then tested after natural air drying at an accelerating voltage of 300 kV.

According to the SEM images, the crystal lengths of the axes a, b, and c of the ZSM-5 zeolites were calculated as L_a, L_b, and L_c (nm), respectively, and the surface area of each crystal face was calculated accordingly. The crystal surface areas S_[100], S_[010], and S_[101] (nm²) of L-Z5, S-Z5, and C-Z5 are calculated as follows ^[23]:

$ S_{[100]}=2 \times L_{{\mathrm{b}}} \times\left(L_{\mathrm{c}}-2 \times \frac{L_{a}}{2 \times \tan \left(118^{\circ} / 2\right)}\right) $

(1)

$ S_{[010]}=2 \times L_{{\mathrm{a}}} \times\left(L_{\mathrm{c}}-\frac{L_{a}}{2 \times \tan \left(118^{\circ} / 2\right)}\right) $

(2)

$ S_{[010]}=4 \times L_{{\mathrm{b}}} \times \frac{L_{a}}{2 \times \sin \left(118^{\circ} / 2\right)} $

(3)

The crystal surface areas S'_[100], S'_[010], and S'_[101] (nm²) of N-ZSM-5 were calculated as follows^[24] :

$ S'_{[100]} = 2 \times L_{{\mathrm{b}}} \times L_{{\mathrm{c}}} $

(4)

$ S'_{[010]} = 2 \times L_{{\mathrm{a}}} \times L_{{\mathrm{c}}} $

(5)

$ S'_{[101]} = 2 \times L_{{\mathrm{a}}} \times L_{{\mathrm{b}}} $

(6)

The exposure ratio d_[010] of the [010] crystal plane was calculated by the following formula^[23] :

$ {d}_{\left[010\right]}=\frac{{S}_{\left[010\right]}}{{S}_{\left[100\right]}+{S}_{\left[010\right]}+{S}_{\left[100\right]}}\times 100\% $

(7)

The formula for calculating the proportion of straight channel p_str and sinusoidal channel p_sin of the zeolite is as follows:

$ {p}_{\mathrm{s}\mathrm{t}\mathrm{r}} = \frac{{S}_{\left[010\right]} \times {n}_{\left[010\right]}}{{S}_{\left[100\right]} \times {n}_{\left[100\right]}+{S}_{\left[010\right]} \times {n}_{\left[010\right]}+{S}_{\left[101\right]} \times {n}_{\left[101\right]}} \times 100\% $

(8)

$ p_{{\mathrm{sin}}} = (1-{p_{{\mathrm{str}}}} ) \times 100\% $

(9)

where, n is the number of holes per unit area of the corresponding crystal face, where n_[100] = 0.748, n_[010] = 0.743, and n_[101] = 0.416.

N₂ adsorption-desorption isotherm and specific surface area: Using N₂ as the adsorption gas, the adsorption desorption curve of N₂ on the sample surface was determined by the static volume method using an ASAP2020 Version 4.03 physical adsorption instrument of (Micromeritics) at −196 ℃. The BET equation method was used to calculate the specific surface area (S_BET), the BJH method was used to calculate the aperture distribution, and the t-plot method was utilized to calculate the specific surface area (S_Mic) and micropore volume (V_Mic).

The characterization of NH₃-TPD was carried out on a ChemiSorb 2720 multifunctional automatic temperature pro-grammed chemisorbent produced by Micromeritics Company in the United States. A sample of approximately 0.1 g was loaded into a U-shaped quartz tube, the carrier gas was He (25 mL∙min⁻¹), and the sample was pretreated at 550 ℃ for 1 h. The temperature was decreased below 50 ℃ and maintained for 30 min. In addition, 10% of the NH₃-He was absorbed to saturation, and the temperature was increased to 150 ℃ under a He atmosphere and maintained for 30 min to prevent physical adsorption of NH₃. Finally, the mixture was heated to 700 ℃ at a rate of 10 ℃∙min⁻¹, and the NH₃-TPD curves were recorded synchronously.

The elemental composition of the product was analyzed by Bruker S8 TIGER X-ray fluorescence spectrometer(XRF). The sample was dried, pressed and scanned. The content of Si-Al oxide in the sample was determined to calculate the Si-Al ratio of the sample.

The ²⁷Al and ²⁹Si MAS NMR of zeolite samples were measured by Agilent 600M nuclear magnetic resonance analyzer with sodium aluminate (NaAlO₂) and tetramethylsilane (TMS) as reference materials, respectively. The Si-Al ratio of the molecular sieve skeleton was obtained by the following formula^[23] :

$ n_{\mathrm{SiO}_2}/n_{\mathrm{Al}_2\mathrm{O}_3}=\frac{\sum\limits_{n=0}^4I_{\mathrm{Si}(n\mathrm{Al})}}{\sum\limits_{n=0}^40.25n[I_{\mathrm{Si}(n\mathrm{Al})}]} $

(10)

where, I_Si(nAl) is the area of the formant corresponding to Si(nAl), and n is the number of Al atoms in different structures.

1.4 Catalyst performance evaluation and product analysis

Catalyst evaluation: A continuous-flow fixed bed reaction device was used to evaluate catalyst performance, and the evaluation process is shown in Fig. 1. The inner diameter of the reactor was 10 mm, the length was 300 mm, and the three-stage temperature control ensured that the temperature difference of the bed was controlled at ±1 ℃. Before the reaction, the molecular sieve samples were exchanged with NH₄Cl for ammonium (zeolite∶solution= 1∶20, g∙mL⁻¹, in which c(NH₄Cl)=1.0 mol∙L⁻¹) and calcined to obtain an acidic H-type zeolite catalyst. Then, the mixture was pressed, broken, and screened to 20~40 mesh particles of a certain particle size and loaded into the reaction tube. The catalyst was loaded at 1.0 g each time, and the catalyst bed was filled with an equal amount of quartz sand. The reaction mixture was driven into the reactor by an EPP 010S high-pressure infusion pump (Dalian Yilite Analytical Instruments Co. Ltd.). The reaction products were collected by cold trapping, and liquid samples were taken once every 4 h for analysis.

Product analysis: The composition of the product was analyzed by a Thermo Trace 1310 Gas Chromatograph. The chromatographic column used was a TR-Wax capillary column (60 m × 0.32 mm × 1.0 μm) with an FID detector. N₂ was selected as the carrier gas, and the pressure was 0.5 MPa. The injection volume was 1 μL, the shunt ratio was 50∶1, and the temperature of the injector and detector was set at 180 ℃. The temperature of the column temperature box was programmed to increase, the initial temperature was 80 ℃ for 2 min, and then the temperature was increased at 10 °C∙min⁻¹ to 145 ℃ for 10 min.

The quality correction factor was used to quantitatively calculate the product. The reaction performance of the catalysts was evaluated by the conversion rate of the raw materials, the selectivity of the main and byproducts, and the relative content of xylene, as calculated according to formulas (11)−(14).

$ X_{\mathrm{B}}=\left(n_{\mathrm{B},{\mathrm{in}}}-n_{\mathrm{B},\mathrm{out}}\right) / n_{\mathrm{B}, \mathrm{in} } \times 100 \% $

(11)

$ X_{\mathrm{E}}=\left(n_{\mathrm{E},{\mathrm{in}}}-n_{\mathrm{E},{\mathrm{out}}}\right) / n_{\mathrm{E},{\mathrm{in}}} \times 100 \% $

(12)

$ w_{\mathrm{X}}=m_{\mathrm{X}} / m_{\mathrm{EB}} \times 100 \% $

(13)

$ \begin{array}{l} S_{\mathrm{i}}=n_{\mathrm{i}} / \sum n_{\mathrm{i}} \times 100 \% \\ \end{array} $

(14)

where X_B and X_E are the conversion rates of the raw material benzene and ethanol, respectively; S_i is the selectivity of product i (toluene, ethylbenzene, propylbenzene, or diethyl- benzene). n_B,in and n_E,in are the amounts of benzene and ethanol in the raw materials, mol, n_B,out and n_E,out are the amounts of benzene and ethanol in the products, mol, n_i and ∑n_i are the amounts of compound i and all products, mol, m_X and m_EB are the masses of xylene and ethylbenzene in the product (g); w_X is the relative content of xylene(%).

2 Results and discussion 2.1 Characterization of ZSM-5

The ZSM-5 zeolite has a typical two-dimensional cross-pore structure. The sinusoidal (or "Z" font) ten-membered ring channel is oriented along the a-axis direction (intersecting the [100] crystal plane), the straight channel is parallel to the b-axis direction (intersecting the [010] crystal plane), and the c-axis direction is the intersection point of the two kinds of channels. The structural characteristics of zeolites determine their acidity, which affects their catalytic activity and selectivity. Differences in the morphology and size of ZSM-5 zeolites affect the length of the two kinds of pore channels, which affects the diffusion behavior of the reactants and catalytic products in the pore channels, thus affecting the catalytic performance of the ZSM-5 catalyst.

Fig. 2 shows the XRD patterns of the zeolite samples. Four types of zeolites exhibited characteristic diffraction peaks with distinct MFI topology at 7.86°, 8.78°, 23.18°, 23.90°, and 24.40°, double peaks at 45.20°, and no diffraction peaks corresponding to other impurities appear, which suggested a relatively high crystallinity. In addition, the XRD peaks of L-Z5 were broadened obviously, indicating that its b-axis thickness was small^[25].

Fig. 3 shows SEM images of four kinds ZSM-5 zeolites with different morphologies and sizes. The spheroidal nano-aggregates of the ZSM-5 zeolite samples demonstrated different crystal structures and sizes due to crystal agglomeration, while the crystals of the other three zeolites had regular morphologies and relatively uniform particle sizes. The four zeolites exhibited obvious differences in morphology and size. C-Z5 (Fig. 3(a)) and S-Z5 (Fig. 3(d)) have typical hexagonal prismatic morpho-logies, but there are certain differences in crystal size. C-Z5 is approximately 1.4 times larger than S-Z5 in the a-axis and c-axis directions, and approximately 1.2 times larger in the b-axis. More importantly, C-Z5 crystals have an obvious cavity structure in the middle (cavity radius between 120 and 160 nm). The crystal sizes S-Z5 and C-Z5 were approximately 380 and 600 nm, respectively. The L-Z5 zeolite with a sheet structure was a sheet with a grain length of approximately 1500 nm, a width of approximately 400 nm, and a thickness of appro-ximately 70 nm (Fig. 3(b)), indicating that a very short b-axis, which was also consistent with the XRD characterization results. As shown in Fig. 3(c), the spheroid N-Z5 zeolite is an approximately spherical aggregate assembled of small particles with a particle size of 50~100 nm, a particle size of appro-ximately 1~1.5 μm, and a rough surface.

Table 1 shows the calculation results of the crystal structure parameters and pore proportions of the four ZSM-5 zeolites with different morphologies and sizes. Among the four zeolites, the L-Z5 zeolites possessed the largest [010] crystal surface area, reaching more than 82%, accompanied by 83.75% of straight pores. The pores of C-Z5 and S-Z5 were relatively small with more than 64% of straight pores, which indicated that molecular diffusion mainly spreads along straight pores in L-Z5, C-Z5, and S-Z5. N-Z5 is quasispherical with three axes located very close to each other, resulting in a significant reduction in the exposure ratio of the [010] crystal surface area, and the proportion of straight channels is only 42.86%.

Fig. 4 shows the N₂ adsorption-desorption isotherms of zeolites with different morphologies. As shown in Fig. 4(a), the C-Z5, S-Z5, and L-Z5 zeolites exhibited mixed isotherms ofⅠand Ⅳ, with adsorption hysteresis loops between the relative pressures P/P₀=0.4 and 1.0, indicating a certain degree of mesopores. As shown in Fig. 4(b), the mesopore size of the C-Z5, S-Z5, and L-Z5 zeolites was approximately 3.8 nm (Fig. 4(b)). In contrast to C-Z5, S-Z5, and L-Z5, the N-Z5 prepared by stepwise crystallization has typical type IV adsorption and desorption isotherms. There are sharp jumps and obvious hysteresis rings in the adsorption-to-desorption isotherms of N-Z5 at higher relative pressures, suggesting a multistage porous zeolite with different mesoporous structures ^[26]. This observation is also consistent with the electron microscopy results of N-Z5 in Fig. 3(c). Combined with the pore size distribution diagram (Fig. 4(b)) the mesoporous structure of N-Z5 is different in size and disorderly in shape, with relatively wide mesoporous pore size distribution.

Table 2 lists the pore structure parameters of zeolites with different morphologies. C-Z5 and S-Z5 had similar pore structure characteristics, as well as similar S_Ext, V_Mic, and V_Tatal to each other. The micropore volume of the sample is 0.15 cm³∙g⁻¹, the external specific surface area is between 46~48 m²∙g⁻¹, and the total pore volume is between 0.22~0.25 cm³∙g⁻¹, indicating that the shell layer of C-Z5 is also highly crystallized. Due to the hollow structure of C-Z5, the specific surface area and micropore area of the C-Z5 zeolite were slightly greater than those of S-Z5. The specific surface area and pore volume of the N-Z5 samples containing multistage pore structures were significantly greater than those of C-Z5 and S-Z5. Among the four zeolites, the specific surface area and pore volume of thin sheet L-Z5 prepared by adding alcohol were the largest, with S_BET and S_Ext values up to 595 and 211 m²∙g⁻¹, respectively. The highly exposed outer surface of the L-Z5 zeolite had abundant mesoporous structures, which indicated more active sites for alkylation reactions^[27].

The NMR spectra of four ZSM-5 zeolites were showed in Fig. 5. The ²⁷Al NMR spectrum of the sample (Fig. 5(a)), where the formant at the chemical shift δ=50 corresponds to quad-coordination skeleton aluminum, and the formant at δ=0 corresponds to six-coordination non-skeleton aluminum. The four ZSM-5 zeolites all showed strong signal peaks at the chemical shift δ=50, and weak signal peaks at δ=0, indicating that aluminum atoms of the four zeolites mainly exist in the form of tetrad coordination skeleton aluminum. In addition, the signal peak of C-Z5 sample at δ=0 was the weakest, demonstated that almost all aluminum atoms entered the hollow ZSM-5 zeolite in the treatment during the crystallization process, which was basically consistent with the characterization results obtained by Qi^[28] who prepared the hollow ZSM-5 zeolite with pure silicon silicalite-1 as the crystal seed. The ²⁹Si NMR spectrum of the sample (Fig. 5(b)) was fitted by peak-differentiation-imitating and obtained four formants. The chemical shift at −107 and −115 were attributed to Q⁴[Si(0Al)], and the chemical shift at −103 were corresponded to Q³[Si(1Al)] or the silicon oxygen potential connected to a skeleton aluminum. The fitted peak at −98 corresponded to Q²[Si(2Al)] or the silicon bit connecting two skeleton Aluminum^[29]. The Si/Al ratios corresponding to the four samples were calculated, and the results were listed in Table 3. It can be seen that the Si/Al ratio of the four zeolites was close, and was slightly higher than the ratio of feed, indicating that few aluminum atoms entered the ZSM-5 framework, which was consistent with ²⁷Al NMR spectra.

2.2 NH₃-TPD characterization

Fig. 6 shows the NH₃-TPD curves of zeolites with different morphologies. The four kinds of zeolites possessed two desorption peaks at 208 and 441 ℃, which were attributed to the weak acid and strong acid centers, respectively. The acidity of the zeolites is closely related to the SiO₂∶Al₂O₃ ratio of the zeolite samples. Due to the similar silicon aluminum ratio during the synthesis process, the four kinds of molecular sieves had similar acid strengths. Among them, the S-Z5 zeolite had the strongest high-temperature peak and low-temperature peak. The low-temperature peak (weak acid peak) of the other three zeolites was relatively close, and the strong acid peak was N-Z5>C-Z5>L-Z5. The acidity results of the four zeolites were shown in Table 4.

2.3 Catalytic performance of ZSM-5 catalysts

The catalytic alkylation performance of benzene and ethanol by ZSM-5 with different morphologies is listed in Table 3. The conversion rates of four catalysts for ethanol were all above 99.9%, but there is a certain difference in the conversion rate of benzene. As for the conversion rate of benzene, S-Z5 was only 9.45%, and L-Z5 reached 12.48%, which was ascribed to the direct influence of the strong acid peak intensity of the zeolite on the selectivity of ethylbenzene. There was no significant decrease or deactivation in the reaction after 24 h, but the distribution of alkylation products on four catalysts showed significant differences, as shown in Table 3. For comparison, the data in the table are the average of six experimental results within 24 h of continuous reaction.

Table 5 shows that the ethylbenzene selectivity of the H-ZSM-5 catalysts with different morphologies decreases in the order of L-Z5, C-Z5, N-Z5, and S-Z5 under the same reaction conditions. The ethylbenzene selectivity is not directly related to the acid density or surface area of the zeolite samples. The Si/Al ratios of the four zeolites were similar. The NH₃-TPD results showed that the weak acid peaks of the four zeolites were similar, but the selectivity of ethylbenzene and the relative content of xylene were very different. Yang et al.^[30] studied the use of synthesis gas to produce aromatics and reported that the catalytic performance of zeolites with different Si/Al ratios did not show significant differences, which further indicated that the catalytic performance of different H-ZSM-5 catalysts was also somewhat related to the morphology of the zeolites. In the process of alkylation of benzene and ethanol, raw benzene and the product alkylbenzene (ethylbenzene, diethylbenzene, toluene, and xylene) required different micropore lengths for diffusion in zeolites. Differences in zeolite morphology and grain size inevitably lead to changes in the diffusion paths of reactants and products, thus affecting the selectivity of the products ^[19]. Molecular dynamics simulations of the alkylation reaction between benzene and methanol were performed. Yan et al.^[16] reported that small molecules diffused along straight and sinusoidal channels, but benzene and alkyl benzene diffused only along straight channels. Table 1 demonstrated that the proportion of straight pores in the thin ZSM-5 zeolite reached 83.75%, with the shortest b-axis length, which was conducive to the diffusion of ethylbenzene and prevented secondary alkylation of ethylbenzene. Moreover, the outer surface of the ZSM-5 zeolite was highly exposed, which promoted benzene and ethanol molecules to contact the acidic sites on the surface. Therefore, the highest selectivity of ethylbenzene was obtained in the alkylation process, reaching 94.56%, and the relative content of xylene was reduced to 0.15% (approximately 1 500 mg∙kg⁻¹). The cyclic ZSM-5 zeolite possessed a large cavity structure and relatively thin wall thickness for the rapid diffusion of reaction intermediates, as well as a large number of straight pores for improving the selectivity of ethylbenzene. Compared with straight pores, the proportion of sinusoidal pores in spherical nanoagglomerated ZSM-5 zeolite was larger, and the products diffused along a zigzag path in the sinusoidal pores, which led to a low diffusion efficiency theoretically. However, the selectivity of ethyl-benzene was greater than that of small-grain ZSM-5 zeolite, which has more straight pores but a smaller external specific area. The surface area is large, and the surface is enriched with multistage pores. Notably, this did not indicate that the influence of the external specific area was greater than the influence of the direct channel. As shown in Table 2, the outer area of annular hollow ZSM-5 was smaller than that of small-grain ZSM-5 and much smaller than that of nanospherical ZSM-5, but its catalytic performance was greater than that of nano-spherical ZSM-5, which was due to its unique hollow structure. The existence of an annular hollow structure in ZSM-5 resulted in a molecular sieve with nanoscale structural characteristics, multistage pores, and short b-axis ZSM-5, thus greatly improving its catalytic performance.

3 Conclusions

In this paper, four ZSM-5 zeolites with different morphologies and sizes were prepared by the hydrothermal method. The structural differences of the four zeolites were compared, and the effects of their morphologies and sizes on the performance of benzene and ethanolation alkylation reactions were explored. The main conclusions were as follows:

The morphology and size of zeolite particles significantly affected the acidity and alkylation properties of zeolites. The weak acid peaks of the four zeolites were close to each other, but the existence of short B-axes and hollow structures decreased the strength of the molecular sieves, which directly affected the selectivity of ethylbenzene.

The morphology and size of zeolite particles affected the pore structure of zeolites. The shorter the b-axis is, the larger the proportion of straight pores in the molecular sieve. The specific surface area, empty volume, and proportion of straight pores in the HZSM-5 zeolites affected the alkylation performance to a certain extent. The influence of four mor-phologies on the catalytic alkylation performance of benzene and ethanol decreased in the order of sheet ZSM-5, annular ZSM-5, spheroidal nanoagglomerates ZSM-5, and small-grain ZSM-5.

Lower acid strength, larger specific surface area and pore volume, and straighter pore proportion improved the alkylation performance of zeolites, thus enhancing the selectivity of ethylbenzene product.