Friedel–Crafts alkylation of toluene with tert-butyl alcohol over Fe₂O₃-modified Hβ

Abstract

Fe₂O₃ (x)/Hβ catalysts with different Fe₂O₃ loadings (x) were successfully prepared and characterized by XRD, SEM, TEM, ICP, BET, NH₃-TPD and Py-IR. With selective alkylation of toluene with tert-butyl alcohol to produce 4-tert-butyltoluene as a probe reaction, the effects of the different Fe₂O₃ loadings on channel structures, acidity and catalytic performance over Hβ were studied. The results showed that modification of Hβ with Fe₂O₃ could adjust pore structures and decrease the total acidity, especially the strong acidity of Hβ. The parent Hβ exhibited the highest toluene conversion of 58.4% with the lowest PTBT selectivity of 67.3%. The low para-selectivity over parent Hβ could be attributed to the low shape-selective action of 12-ring portals of Hβ and the isomerization of the formed PTBT to MTBT. Upon loading of Fe₂O₃, the toluene conversion of Fe₂O₃ (x)/Hβ catalysts decreased slightly, but the PTBT selectivity increased significantly. A toluene conversion of 54.7% with PTBT selectivity of 81.5% was observed over Fe₂O₃ (20%)/Hβ at 190 °C after 4 h. The narrowed pores after loading the Fe₂O₃ were beneficial to increasing the selectivity to PTBT, since PTBT with a lower kinetic diameter (0.58 nm) can diffuse through the narrowed pores of Fe₂O₃ (x)/Hβ more easily than MTBT (0.65 nm). In addition, the decrease in number of strong acid sites and deactivation of acid sites on the surface are beneficial to increasing the selectivity to PTBT by suppressing further isomerization of the formed PTBT on acid sites.

---

Introduction

Acid catalyzed reactions such as Friedel–Crafts alkylation are important processes in organic synthesis and fine chemical production as well as in petrochemical processes.^1–3 The butylated products of toluene, particularly 4-tert-butyltoluene (PTBT), have great commercial significance as their derivatives 4-tert-butylbenzaldehyde and 4-tert-butylbenzoic acid are widely used in industry.^4–6 Conventionally, toluene butylation reactions are carried out with alkylating agents with AlCl₃, sulfuric acid and phosphoric acid as catalysts. However, these acid catalysts are not environmentally benign, not re-usable and also lead to corrosion of equipment. Therefore, considerable attention has been devoted to the development of solid acid catalysts which are environmentally friendly.

It is well known that toluene is not active enough in the Friedel–Crafts alkylation in the absence of a catalyst. So, selection of a suitable catalyst is crucial to the highly efficient alkylation of toluene with tert-butyl alcohol (TBA). Zeolites are regarded as a good choice due to their pore structures and unique acid properties. Among different zeolites, zeolite β with a three-dimensional (3D) interconnecting pore system and high acidity shows a high catalytic activity. For example, Pai et al.⁷ found that Hβ (25) and HY (30) had better catalytic activity than HMCM-22 (52) in vapour-phase butylation of toluene. Mravec et al.⁸ found the most active zeolite catalysts for the liquid-phase butylation of toluene were Hβ (12.5) and HM (17.5). In our previous report,⁹ we also found that Hβ (25) showed high catalytic activity in alkylation of toluene with TBA. However, a para-selectivity (4-tert-butyltoluene, PTBT) of 69.5% and a meta-selectivity (3-tert-butyltoluene, MTBT) of 26.9% over Hβ were observed at 180 °C after 4 h. This showed that the selectivity to the desired product PTBT over Hβ zeolite was low.

A petrochemical process is evaluated not only on the basis of the conversion of a given reactant, but also in terms of selectivity to the desired product. The dimensions of 12-ring portals of parent Hβ are 0.67 nm × 0.66 nm,¹⁰ while the kinetic diameters are 0.58 nm for PTBT and 0.65 nm for MTBT.⁸ The shape-selective action has less effect on the product selectivity to PTBT over Hβ zeolite. However, the shape-selective action could be greatly improved if the 12-ring portals of the parent Hβ were narrowed slightly. Zeolite Hβ contains numerous lattice defects due to its subtle structural disorder, which is believed to create additional acid sites on internal surfaces as well as extra cation exchange positions.¹¹ Modification by metallic oxide is a highly efficient method for improving the para-selectivity by reducing the external acid sites and adjusting the pore entrance.

As far as we know, the use of Fe₂O₃ modification of Hβ zeolites as catalysts in tert-butylation of toluene has not been reported. The aim of the present study was to evaluate the catalytic behaviour of Fe₂O₃ supported on Hβ for the alkylation of toluene by TBA under liquid reaction conditions. For this purpose, Fe₂O₃ (x)/Hβ samples containing 8–25% Fe₂O₃ were prepared and were fully characterized by XRD, SEM, TEM, FT-IR, BET, TPD and Py-IR. Our objective was to correlate the channel structures and acidity of the Hβ-supported Fe₂O₃ catalysts with their catalytic properties in the alkylation of toluene.

Experimental

2.1 Materials

Naβ zeolite powder (Si/Al = 25) was purchased from Zeolyst Int., China. Ammonium nitrate, ferric nitrate, toluene, tert-butyl alcohol and cyclohexane (all analytical grade) were bought from ACROS Organics and used without further purification.

2.2 Catalyst preparation

The Hβ zeolite with an Si/Al radio of 25 used in this work was prepared according to the reported procedure.¹² Naβ zeolite was used as a starting material. The Hβ zeolite was prepared by ion exchange of Naβ with NH₄NO₃ aqueous solution. After ion exchange, the zeolite was filtered, and dried at 110 °C for 12 h. Then, the samples were calcined at 550 °C for 4 h in an air atmosphere. Fe₂O₃-modified Hβ zeolite supports with different mass fractions of Fe₂O₃ (4%, 8%, 16%, 20% and 25%) were synthesized by ion exchange of Hβ zeolites with Fe(NO₃)₃ aqueous solutions for 24 h at room temperature.¹³ The resulting materials were calcined at 350 °C for 4 h in an air atmosphere. The obtained Fe₂O₃-modified Hβ zeolites were denoted by Fe₂O₃ (x)/Hβ, where x stands for the mass fraction of Fe₂O₃ in %.

2.3 Catalyst characterization

XRD analysis of material powders was carried out on a D/max-2200PC-X-ray diffractometer using Cu Kα radiation under the setting conditions of 40 kV, 30 mA, and scan range from 10 to 80 at a rate of 4° per min.

The surface morphological details of the catalysts were studied by scanning electron microscopy (SEM, Zeiss-ΣIGMA).

The iron contents of all catalysts were determined by inductively coupled plasma-optical emission spectrometry (ICP-OES) using a PerkinElmer Optima 2000DV instrument.

Nitrogen adsorption–desorption curves were recorded at −196 °C using the Micromeritics adsorption equipment NOVA2000e. Prior to N₂ adsorption, the powder samples were degassed under a secondary vacuum for 1 h at 90 °C and 10 h at 180 °C. The specific surface area of each solid (S_BET in m² g⁻¹) was calculated from the adsorption isotherms using the BET method. The total pore volume (V_tot) was calculated from the adsorbed volume of nitrogen at P/P₀ equal to 0.99.

Sample acidities were measured by NH₃ temperature-programmed desorption (NH₃-TPD) using a Quantachrome Chembet-3000 Characterization System. A 200 mg sample was pre-treated at 550 °C for 1 h in dry helium (50 ml min⁻¹) and cooled to 120 °C. It was then exposed to 10 v% NH₃/He for 1 h. After purging the catalyst with He for 1 h, a TPD plot was obtained at a heating rate of 10 °C min⁻¹ from 120 to 600 °C.

IR spectra with pyridine adsorption of samples were recorded using a Bruker FT-IR spectrometer (Tensor 27) together with a high temperature vacuum cell. The sample powder was pressed into a self-supporting wafer, and the spectra were recorded in a wavenumber range of 4000–400 cm⁻¹ with a 4 cm⁻¹ resolution. Before each experiment, material samples were pressed into thin pellets (10–30 mg) with a diameter of 16 mm and activated in situ during one night under vacuum (10⁻⁵ Pa) at 170 °C. Pyridine was introduced in excess at 150 °C after the activation period. The concentrations of the B sites and L sites were determined from the integrated area bands of the PyH⁺ (located at around 1540 cm⁻¹) and PyL (around 1450 cm⁻¹) species using the values of the molar extinction coefficients of both bands.

2.4 Toluene alkylation with tert-butyl alcohol

The side-chain alkylation of toluene with TBA was carried out in a laboratory autoclave reactor (300 ml) at 190 °C. In a typical experiment, 10 ml of toluene (94 mmol, Aladdin, 98%) and 28 ml of tert-butyl alcohol (283 mmol Aladdin, 98%) were mixed with 60 ml of cyclohexane (Aladdin, 99%) solvent. Then 1.0 g of zeolite catalyst (1.35%) was added. The reaction system was sealed and purged with continuous N₂ purging for 30 min. Then the mixture was heated to a certain temperature to start the alkylation reaction. After reaction, the mixture was cooled down to room temperature, and the zeolite catalyst was removed by filtration. The liquid phase was analyzed on a GC-14C gas chromatograph equipped with an SE-30 capillary column (φ 0.25 mm × 50 m). The measurement was performed at an initial temperature of 60 °C for 2 min, a ramping rate of 10 °C min⁻¹, a final temperature of 220 °C for 10 min, with nitrogen as carrier gas. During the course of the investigation, a number of runs were repeated to check for reproducibility in the experimental results, which was found to be excellent. Typical errors were in the range of ±2%.

Results and discussion

3.1 Physicochemical characterization

Fig. 1 shows typical XRD patterns of parent Hβ and Fe₂O₃ (x)/Hβ. The peak positions of Hβ did not change even with high Fe₂O₃ content. These results showed that the structure of Hβ was retained after Fe₂O₃ modification. The peaks of Fe₂O₃ were not observed with Fe₂O₃ contents less than 20%. This result indicated that the Fe₂O₃ was highly dispersed. Nevertheless, the peaks corresponding to Fe₂O₃ appeared at 2θ of 33.1° and 35.6° (ref. 14–16) with Fe₂O₃ contents up to 25%. In addition, the crystallite size of Fe₂O₃ (20%)/Hβ, which was calculated using the Debye–Sherrer equation applied to the reflection at 2θ of 22.5°, was estimated to be 249 nm.¹⁷

Fig. 1 XRD of parent Hβ and Fe₂O₃ (x)/Hβ (a) parent Hβ, (b) Fe₂O₃ (8%)/Hβ, (c) Fe₂O₃ (16%)/Hβ, (d)Fe₂O₃ (20%)/Hβ and (e) Fe₂O₃ (25%)/Hβ.

Fig. 2(a) shows the scanning electron micrograph of a fresh Fe₂O₃ (20%)/Hβ sample. The crystal of fresh Fe₂O₃ (20%)/Hβ showed typical agglomerates of small cubic particles with sizes around 200–300 nm, which was in good agreement with the calculated crystallite size of Fe₂O₃ (20%)/Hβ from XRD analysis. Fig. 2(b) shows the scanning electron micrograph of a spent Fe₂O₃ (20%)/Hβ sample. There were some carbonaceous deposits on the surface of the spent Fe₂O₃ (20%)/Hβ zeolite as compared with the fresh one.

Fig. 2 SEM images obtained Fe₂O₃ (20%)/Hβ (a) fresh Fe₂O₃ (20%)/Hβ; (b) spent Fe₂O₃ (20%)/Hβ.

The transmission electron microscope image of the Fe₂O₃ (20%)/Hβ catalyst is shown in Fig. 3. Fe₂O₃ (20%)/Hβ showed a honeycomb structure. The Fe₂O₃ was highly dispersed on Hβ, and could be distinguished as dark dots in the TEM images.

Fig. 3 TEM images obtained Fe₂O₃ (20%)/Hβ.

As can be seen from Table 1, the actual loading amounts of Fe₂O₃ in the fresh Fe₂O₃ (x)/Hβ materials obtained from ICP measurement were close to their corresponding theoretical amounts. However, the amount of Fe₂O₃ in spent Fe₂O₃ (20%)/Hβ was 19.5%, which was slightly lower than that found in the fresh sample (19.8%). The leaching of Fe₂O₃ into the reaction mixture was responsible for the low amount of Fe₂O₃ in the spent Fe₂O₃ (20%)/Hβ.

Table 1 The structural parameters of parent Hβ zeolite and Fe₂O₃ (x)/Hβ zeolites

Zeolites	Content of Fe2O3^a (%)	BET surface area^b (m^2 g^−1)	Pore volume^c (cm^3 g^−1)	Pore size (nm)
a Deduced from ICP analysis.b Specific surface area calculated by the BET method.c Total pore volume determined at P/P0 = 0.99.
Hβ	—	492	0.48	3.90
Fe2O3 (8%)/Hβ	8.2	446	0.46	3.81
Fe2O3 (16%)/Hβ	16.0	425	0.43	3.75
Fe2O3 (20%)/Hβ	19.8	396	0.41	3.62
Fe2O3 (25%)/Hβ	25.3	352	0.31	3.36
Spent Fe2O3 (20%)/Hβ	19.5	177	0.20	4.58

---

The N₂ adsorption/desorption curves in Fig. 4 obtained from parent Hβ and Fe₂O₃ (x)/Hβ samples showed a typical structure of mesoporosity, namely type-IV isotherms with well-defined hysteresis loops. A typical H4 type hysteresis loop was observed in the P/P₀ range from 0.4 to 0.9, indicating the existence of large mesopores.^18,19

Fig. 4 N₂ adsorption/desorption curves of Hβ and Fe₂O₃ (x)/Hβ (a) parent Hβ; (b) Fe₂O₃ (8%)/Hβ; (c) Fe₂O₃ (16%)/Hβ; (d) Fe₂O₃ (20%)/Hβ and (e) Fe₂O₃ (25%)/Hβ.

The textural properties including BET surface area, pore volume, and pore size of parent Hβ and Fe₂O₃ (x)/Hβ sample distributions are summarized in Table 1. The parent Hβ zeolite possessed the highest specific surface area (492 m² g⁻¹), pore volume (0.48 cm³ g⁻¹) and pore size (3.90 nm). The increase in the Fe₂O₃ loading from 8% to 25% led to a decrease in the specific surface area from 446 m² g⁻¹ to 352 m² g⁻¹, in the total pore volume from 0.46 cm³ g⁻¹ to 0.31 cm³ g⁻¹ and in the pore size from 3.81 nm to 3.36 nm. The decrease in pore volume and pore size after loading Fe₂O₃ is expected to improve the para-selectivity, since narrowed pores are beneficial to shape-selective action. The severe decrease in specific surface area and total porous volume observed for the Fe₂O₃ (25%)/Hβ solid sample could be explained by the blocking of pores due to the high loading of Fe₂O₃. The pore size decreased in the order: parent Hβ > Fe₂O₃ (8%)/Hβ > Fe₂O₃ (16%)/Hβ > Fe₂O₃ (20%)/Hβ > Fe₂O₃ (25%)/Hβ.

The specific surface area and pore volume of the spent Fe₂O₃ (20%)/Hβ decreased remarkably to 177 m² g⁻¹ and 0.20 cm³ g⁻¹, which could be caused by carbonaceous deposits on the surface and inside the pores of the spent Fe₂O₃ (20%)/Hβ zeolite. The increase in pore size (4.58 nm) observed for the spent Fe₂O₃ (20%)/Hβ further confirmed the blockage of micropores during the reaction by carbonaceous deposition.

3.2 Acidity characterization

The acid properties of zeolites are important in determining the catalytic activity and product selectivity for any catalyzed reaction. Acidic measurement was achieved by NH₃-TPD and Py-IR characterization. NH₃-TPD curves of parent Hβ and Fe₂O₃ (x)/Hβ are shown in Fig. 5.

Fig. 5 NH₃-TPD patterns of Hβ and Fe₂O₃ (x)/Hβ (a) parent Hβ; (b) Fe₂O₃ (8%)/Hβ; (c) Fe₂O₃ (16%)/Hβ; (d) Fe₂O₃ (20%)/Hβ and (e) Fe₂O₃ (25%)/Hβ.

Two peaks could be detected at 240 °C and 460 °C in the curve of parent Hβ, which were attributed to weak acid sites and strong acid sites, respectively. The number of weak acid sites was obviously greater than that of strong acid sites. Compared to parent Hβ, the desorption peak of weak acid sites in Fe₂O₃ (x)/Hβ shifted slightly toward higher temperatures with the increase in Fe₂O₃ loading. This indicated that the strength of the weak acid increased slightly. The amount of acid also changed after the addition of Fe₂O₃: the weak acid sites in Fe₂O₃ (x)/Hβ decreased gradually with an increase in x, while the strong acid sites of all the Fe₂O₃ (x)/Hβ almost vanish. This result showed that Fe₂O₃ preferentially acted with the strong acid sites of the Hβ zeolite. In addition, the loading of Fe₂O₃ is beneficial to increasing the selectivity to PTBT by suppressing the further isomerization of formed PTBT to MTBT and this will be discussed later in Section 3.3. The total amount of acid decreased in the order of parent Hβ > Fe₂O₃ (8%)/Hβ > Fe₂O₃ (16%)/Hβ > Fe₂O₃ (20%)/Hβ > Fe₂O₃ (25%)/Hβ.

Table 2 shows Py-IR spectra of parent Hβ and Fe₂O₃ (x)/Hβ. It can be clearly seen from Table 2 that the total amount of acid determined by Py-IR correlated well with the NH₃-TPD results and the relative amounts of B acid and L acid of Hβ changed significantly after loading Fe₂O₃. Upon increasing the Fe₂O₃ content, both the B acid and L acid decreased. This is possibly because Fe₂O₃ would cover some acid sites on the Hβ zeolite. However, it is worth noting that the B/L ratio first increased then dropped. This meant that the L acid decreased more drastically than the B acid. This may be attributed to the electronic structure of transition metal elements. Most transition metal elements have an empty orbit, which can be used as the “transfer station” for electrons. This special electronic structure can cause transition metal cations to interact with the framework via a polarizing or inductive effect, withdrawing electrons from O–H bonds, thereby increasing the B acidity.²⁰ As a result, the increased B acidity can to some degree compensate for the original B acidity covered by Fe₂O₃. In addition, both B acid and L acid of Fe₂O₃ (25%)/Hβ were sharply reduced compared to those of Fe₂O₃ (20%)/Hβ, which can be attributed to the blocking of pores of Hβ owing to the high Fe₂O₃ loading in the former. As proposed above, based on the N₂ adsorption analysis, this decreases the total acid amount. The amount of B acid decreased in the order of parent Hβ > Fe₂O₃ (8%)/Hβ > Fe₂O₃ (16%)/Hβ > Fe₂O₃ (20%)/Hβ > Fe₂O₃ (25%)/Hβ.

Table 2 Acidic properties of catalysts determined by Py-IR at 150 °C

Catalysts	B acid^a (μmol g^−1)	L acid^b (μmol g^−1)	Total acid (μmol g^−1)	B/L ratio
a Deduced from the intensity of the band located at around 1540 cm^−1.b Deduced from the intensity of the band located at around 1450 cm^−1.
Parent Hβ	84.2	47.2	131.4	1.78
Fe2O3 (8%)/Hβ	81.4	44.3	125.7	1.84
Fe2O3 (16%)/Hβ	78.9	41.2	120.1	1.92
Fe2O3 (20%)/Hβ	75.4	37.3	112.7	2.02
Fe2O3 (25%)/Hβ	63.2	32.5	95.7	1.94

---

3.3 Catalytic activity evaluation

Table 3 presents the results of toluene conversion and PTBT selectivity in the alkylation of toluene with TBA. Two monoalkylated products were detected. One was PTBT (para-isomer), which is in a large amount, and the other was MTBT (meta-isomer). However, the OTBT (ortho-isomer) was not detected. The PTBT is a desirable product since the para-position is favored by the influence of the steric hindrance of the methyl group and voluminous tert-butyl group on the one hand, and by shape-selective action due to the regular channel structure of zeolite catalysts on the other hand. The formation of OTBT is hindered by the ortho-position of the methyl group and the tert-butyl group. A trace amount of di-alkylated product 3,5-di-tert-butyltoluene (3,5-DTBT) from further alkylation of the monoalkylated products PTBT and MTBT was also detected. This can be explained by the same steric effects, where all the alkyl groups are in the meta-position.

Table 3 Product selectivity and catalyst activity^a

Catalyst	Toluene conversion (%)	Distribution of products (%)			PTBT selectivity (%)	MTBT selectivity (%)
		PTBT	MTBT	Other
a Reaction conditions: n (TBA) : n(toluene)-3 : 1, catalyst-1.0 g, temperature-190 °C, time-4 h, cyclohexane-60 ml.
Parent Hβ	58.4	39.3	16.7	2.4	67.3	28.6
Fe2O3 (8%)/Hβ	56.0	43.4	10.9	1.7	77.5	19.5
Fe2O3 (16%)/Hβ	55.5	44.2	10.2	1.1	79.6	18.4
Fe2O3 (20%)/Hβ	54.7	44.6	9.8	0.3	81.5	17.9
Fe2O3 (25%)/Hβ	46.3	37.8	8.3	0.2	81.6	17.9

---

Of the catalysts investigated, parent Hβ exhibited a toluene conversion of 58.4% at 190 °C after 4 h. The toluene conversions of all the Fe₂O₃ (x)/Hβ catalysts were slightly lower than that of parent Hβ. In addition, with an increase in the Fe₂O₃ loading the toluene conversion by the Fe₂O₃ (x)/Hβ catalyst decreased. This result agrees with the trend of B acidity of the Fe₂O₃ (x)/Hβ catalyst (Table 2), which also decreased with an increase in the Fe₂O₃ loading. As is well known, the alkylation of toluene with alcohols catalyzed by zeolites is commonly considered to proceed via carbonium ion mechanisms²¹ and the B acid sites are active sites.^22,23 Therefore, lower B acidity leads to a lower catalytic activity. A toluene conversion of 54.7% was observed over the Fe₂O₃ (20%)/Hβ catalyst. With further increases in the Fe₂O₃ content to 25%, a sharp decrease in toluene conversion was observed. This is probably due to a sharp decrease in the amount of B acid sites, which is caused by blockage of the pores by excess Fe₂O₃. In addition, the severe decrease in pore size would lead to poor adsorption and diffusion of reactants and products, and this is another reason for the poor catalytic activity of the Fe₂O₃ (25%)/Hβ catalyst.

The selectivity to PTBT over parent Hβ is only 67.3%, which is much lower than the values found for Fe₂O₃ (x)/Hβ (around 80.0%). As shown in Fig. 6, the lower selectivity for PTBT over Hβ can be explained in three ways. (i) The low para-selectivity can be attributed to the pore size of the Hβ zeolite. The dimensions of the 12-ring portals of parent Hβ are 0.67 × 0.66 nm, while the kinetic sizes are 0.58 nm for PTBT and 0.65 nm for MTBT. Therefore, the shape-selective action has less effect on the product selectivity to PTBT over Hβ zeolite. (ii) The large amount of strong acid of the Hβ (Fig. 4) is also responsible for the low para-selectivity. Because the MTBT is the most thermodynamically favored, the kinetically favored PTBT is gradually transformed to MTBT by isomerization on the acid sites. And the isomerization reaction occurs easily on the strong acidic sites.^24,25 (iii) The high of acidity of the external surface also promotes isomerization of the formed PTBT to MTBT. The isomerization of formed PTBT to MTBT runs more easily over the active sites on the external surface, where it is not sterically hindered as in the pores of the zeolite. The above three reasons lead to the low PTBT selectivity over parent Hβ.

Fig. 6 The mechanism for low PTBT selectivity over parent Hβ zeolite.

The PTBT selectivities over all the Fe₂O₃ (x)/Hβ catalysts were around 80.0%, showing that upon loading of Fe₂O₃, the PTBT selectivity increased significantly. A possible mechanism was proposed to explain the effect of Fe₂O₃ on the PTBT selectivity of the catalyst, as sketched in Fig. 7, and explains the increased selectivity in three ways. (i) The first reason is the narrowed pore size upon loading the Fe₂O₃. The pore size of the Fe₂O₃ (x)/Hβ catalyst decreased with an increase in the Fe₂O₃ loading (Table 1). The narrowed pores of the Fe₂O₃ (x)/Hβ catalysts were beneficial to the formation of PTBT. Because of the steric restriction with the narrowed pores of Fe₂O₃ (x)/Hβ, the PTBT with a lower kinetic diameter (0.58 nm) can diffuse through the narrowed pores of Fe₂O₃ (x)/Hβ more easily than MTBT with a higher kinetic diameter (0.65 nm). (ii) The decrease in strong acid sites could improve the para-selectivity. As discussed in Section 3.2 (Fig. 5), the weak acid sites in Fe₂O₃ (x)/Hβ decreased gradually with an increase in x, while the strong acid sites of all the Fe₂O₃ (x)/Hβ almost vanish. Fewer strong acid sites caused by the loading of Fe₂O₃ is beneficial to increasing the selectivity to PTBT by suppressing the further isomerization of PTBT on acid sites. (iii) In addition, the increase in PTBT selectivity could also be attributed to the deactivation of acid sites on the surface of Fe₂O₃ (x)/Hβ caused by the loading of Fe₂O₃. The isomerization of the formed PTBT to MTBT runs more easily over the active sites on the external surface, and fewer acid sites on the surface reduces the isomerization reaction. As a result, loading Fe₂O₃ on Hβ catalysts could enhance and retain constant high para-selectivity.

Fig. 7 The mechanism for high PTBT selectivity over Fe₂O₃ (x)/Hβ zeolite.

Conclusions

In the present work, Fe₂O₃ (x)/Hβ catalysts with different Fe₂O₃ loadings were successfully prepared, and the effects of Fe₂O₃ loading on the catalytic properties for alkylation of toluene with TBA were studied. The acidity characterization results showed that modification of Hβ with Fe₂O₃ could increase the weak acid strength. However, the weak acid sites in Fe₂O₃ (x)/Hβ decreased gradually with an increase in x, while the strong acid sites of all the Fe₂O₃ (x)/Hβ almost vanish. Both the amount of B acid and the total amount of acid decreased in the order of parent Hβ > Fe₂O₃ (8%)/Hβ > Fe₂O₃ (16%)/Hβ > Fe₂O₃ (20%)/Hβ > Fe₂O₃ (25%)/Hβ. The textural properties characterization results showed that the parent Hβ zeolite possessed the highest specific surface area (492 m² g⁻¹), pore volume (0.48 cm³ g⁻¹) and pore size (3.90 nm). An increase in the Fe₂O₃ loading from 8% to 25% led to a decrease in the pore size from 3.81 nm to 3.36 nm. The pore size decreased in the order of parent Hβ > Fe₂O₃ (8%)/Hβ > Fe₂O₃ (16%)/Hβ > Fe₂O₃ (20%)/Hβ > Fe₂O₃ (25%)/Hβ.

The catalytic activity results showed that the toluene conversion over parent Hβ, which possessed the largest B acidity (84.2 μmol g⁻¹), exhibited the highest activity of 58.4% at 190 °C after 4 h. However, the PTBT selectivity over parent Hβ was the lowest (67.3%). The dimensions of the 12-ring portals of parent Hβ are 0.67 nm × 0.66 nm, while the kinetic sizes are 0.58 nm for PTBT and 0.65 nm for MTBT. Therefore, the shape-selective action has less effect on the product selectivity to PTBT over Hβ zeolite. In addition, further isomerization of formed PTBT to MTBT on acid sites also decreased the para-selectivity. A toluene conversion of 54.7% with PTBT selectivity of 81.5% was observed over a Fe₂O₃ (20%)/Hβ catalyst at 190 °C after 4 h. The selectivity to PTBT was improved significantly from 67.3% to 81.5%. A possible mechanism was proposed to explain the increased selectivity in three ways. (i) The narrowed pores of the Fe₂O₃ (x)/Hβ catalysts were beneficial to the formation of PTBT, since the PTBT with its lower kinetic diameter (0.58 nm) can diffuse through the narrowed pores of Fe₂O₃ (x)/Hβ more easily than MTBT with its higher kinetic diameter (0.65 nm) because of its steric restriction with the narrowed pores of Fe₂O₃ (x)/Hβ. (ii) The deactivation of acid sites caused by the loading of Fe₂O₃ is beneficial to increasing the selectivity to PTBT by suppressing the further isomerization of PTBT on acid sites. As a result, the Fe₂O₃ (x)/Hβ catalysts could enhance and retain constant high para-selectivity. However, a significant deactivation of catalysts containing 25% of Fe₂O₃ was observed and attributed to a pore blocking phenomenon by excess Fe₂O₃. With further increases in the Fe₂O₃ content to 25%, a sharp decrease in toluene conversion was observed, which is mainly due to the sharp decrease in the amount of B acid sites, which was caused by the blockage of pores by excess Fe₂O₃.

References

1. Y. Sun and R. Prins, Appl. Catal., A, 2008, 336, 11–16.
2. C. Sievers, J. S. Liebert, M. M. Stratmann, R. Olindo and J. A. Lercher, Appl. Catal., A, 2008, 336, 89–100.
3. H. Nur, Z. Ramli, J. Efendi, A. N. A. Rahman, S. Chandren and L. S. Yuan, Catal. Commun., 2011, 12, 822–825.
4. W. Y. Li, Y. H. Xu, J. Q. Wang, Z. B. Zhai, Z. Y. Yan and Y. L. Yang, Catal. Lett., 2007, 119, 327–331.
5. W. Y. Zhou, J. G. Pan, F. A. Sun, K. Huang, M. Y. He and Q. Chen, React. Kinet., Mech. Catal., 2016, 117, 789–799.
6. W. H. Yu, Z. R. Zhang, H. Wang, Z. H. Ge and T. J. Pinnavaia, Microporous Mesoporous Mater., 2007, 104, 151–158.
7. S. Pai, U. Gupta and S. Chilukuri, J. Mol. Catal. A: Chem., 2007, 265, 109–116.
8. D. Mravec, P. Zavadan and A. Kaszonyi, Appl. Catal., A, 2004, 257, 49–55.
9. Y. Y. Wang, H. Song, H. L. Song, X. L. Sun and X. Q. Wang, Prog. React. Kinet. Mech., 2016, 41, 126–134.
10. H. K. Min, S. H. Cha and S. B. Hong, ACS Catal., 2012, 2, 971–981.
11. J. C. Jansen, E. J. Creyghton, S. L. Njo, H. V. Koningsveld and H. V. Bekkum, Catal. Today, 1997, 38, 205–212.
12. K. F. Liu, S. J. Xie, S. L. Liu, G. L. Xu, N. N. Gao and L. Y. Xu, J. Catal., 2011, 283, 68–74.
13. B. Xue, Y. X. Li and L. J. Deng, Catal. Commun., 2009, 10, 1609–1614.
14. T. Nobukawa, M. Yoshida, S. Kameoka, S. Ito, A. Keiichi Tomishige and K. Kunimori, J. Phys. Chem. B, 2004, 108, 4071–4079.
15. M. Mauvezin, G. Delahay, B. Coq, S. Kieger, A. J. C. Jumas and J. OlivierFourcade, J. Phys. Chem. B, 2001, 105, 928–935.
16. C. H. He, Y. H. Wang, Y. S. Cheng, C. K. Lambert and R. T. Yang, Appl. Catal., A, 2009, 368, 121–126.
17. F. Richard, S. Célérier, M. Vilette, J. D. Comparot and V. Montouillout, Appl. Catal. B-Environ., 2014, 152–153, 241–249.
18. S. S. Vieira, Z. M. Magriotis, M. F. Ribeiro, I. Graça, A. Fernandes, J. M. F. M. Lopes, S. M. Coelho, N. A. V. Santos and A. A. Saczk, Microporous Mesoporous Mater., 2015, 201, 160–168.
19. N. Kasian, E. Verheyen, G. Vanbutsele, K. Houthoofd, T. I. Koranyi, J. A. Martens and C. E. A. Kirschhock, Microporous Mesoporous Mater., 2013, 166, 153–160.
20. R. Carvajal, P. J. Chu and J. H. Lunsford, J. Catal., 1990, 125, 123–131.
21. M. Selvaraj, S. H. Jeon, J. Han, P. K. Sinha and T. G. Lee, Appl. Catal., A, 2005, 286, 44–51.
22. M. T. Portilla, F. J. Llopis and C. Martínez, Appl. Catal., A, 2011, 393, 257–268.
23. L. Chen, H. Dong and L. Shi, Ind. Eng. Chem. Res., 2010, 49, 7234–7238.
24. I. Fechete, P. Caullet, E. Dumitriu, V. Hulea and H. Kessler, Appl. Catal., A, 2005, 280, 245–254.
25. E. Dumitriu, D. Meloni, R. Monaci and V. Solinas, C. R. Chim., 2005, 8, 441–456.

---