Alkenyl aromatics compounds are widely applied in many areas such as dyes, pharmaceuticals, natural products and agrochemicals [1, 2]. The alkenylation of arenes with alkynes provides an environmentally friendly strategy for production of alkenyl aromatic with high efficiency and atom economy [3, 4]. Homogeneous metal complex catalysts show prominent activity and selectivity for production of alkenyl aromatics [5-10]; however, these catalysts are expensive, have low recyclability as homogeneous catalysts, and feature difficulties in purification of the desired products. Hence, it would be desirable to develop a more clean and economical method. Heterogeneous solid acid catalysts provide an easily recycled, separable, and environmentally friendly alternative approach [11-14]. Currently, only a few kinds of solid acid catalysts have been used for aromatic alkenylation reaction. Zeolites are a type of solid acid catalysts, which are widely used in various fields, having been first applied to the alkenylation of aromatic reactions by Sartori and coworkers. Unfortunately, the catalytic results with these materials have been far from satisfactory, and the yield of alkenyl aromatics compounds is quite low. Furthermore, the low catalytic efficiency cannot be overcome while HY is used as catalyst for the alkenylation. The reaction has been shown to take place on the external surface of the catalyst because the narrow channels of the zeolite limit diffusion of reactants and products [15-17]. Therefore, much attention has been paid to developing mesoporous solid acid catalysts.

Phosphotungstic acid (PTA) catalysts supported on MCM-41, modified activated carbon and mesoporous silica nanospheres have been developed as attractive mesoporous solid acid catalysts for alkenylation of a diverse range of aromatics [18-20]. However, the regeneration of supported PTA catalysts requires a large amount of organic solvent, which limits their cost effectiveness and contributes to environmental problems. Moreover, sulfated ZrO₂, as a common solid super-acid, has been widely used in various reactions. A series of SO₄^2-/meso-La₂O₃-ZrO_2, as novel and efficient solid acid catalysts for alkenylation, have been developed [21-23]. Although the catalyst could be recovered by a calcination treatment, the simple low-cost preparation of well-ordered mesoporous La₂O₃-ZrO₂ remains challenging. Furthermore, Fe-containing mesoporous alumina-silicate exhibits a high activity for the alkenylation aromatic reaction under mild conditions. However, the selectivity of the reaction requires further improvements [24].

Hierarchical zeolites are solid acid catalysts that show excellent performance, owing to a combination of intrinsic zeolite and mesoporous properties, which improve mass transfer [25, 26]. Thus, this material shows great promise as a hierarchical zeolite for solid acid catalysis for alkenylation. Various strategies have been developed to create hierarchical porous zeolites [25, 27, 28], and post-synthesis treatments by acid or alkali treatment through dealumination or desilication are considered to be simple and scalable approaches [29-31]. The modification of zeolites by metal oxides, such as alkaline earth and rare earth metal oxides has been demonstrated to be an efficient method for improving their catalytic performance in various transformations [32-38]; however, there have been few reports on promoted alkenylation.

In this work, CeO₂ modified hierarchical mesoporous Hβ zeolites were prepared by a desilication-dealumination procedure with subsequent cerium impregnation and then used to catalyze the alkenylation reaction. The effect of preparation parameters such as alkali concentration, acid concentration and CeO₂ loading on the nature and catalytic performance were investigated. The as-synthesized CeO₂ modified hierarchical Hβ demonstrated superior catalytic performance to that of pristine beta zeolite for alkenylation of aromatics, owing to more acidic sites and improved accessibility to the active sites through the interconnected mesopores and micropores of the hierarchical zeolites.

Commercially available Hβ zeolite calcined at 550 ℃ for 6 h, and is denoted as Hβ. The Hβ zeolite was treated with a NaOH aqueous solution with a range of concentrations (0.1, 0.2, 0.3 mol/L) and a liquid-to-solid ratio of 25 mL/g at 65 ℃ for 0.5 h. The resulting solids were filtered, washed with deionized water until the washings became neutral and then dried in the oven. The as-prepared alkaline treated zeolites were labeled as Hβ-B_x, where x denotes the NaOH concentration. The alkaline-treated Hβ-B_x zeolites were further treated by HNO₃ aqueous solution at certain concentrations (0.1, 0.2, and 0.3 mol/L) and a liquid-to-solid ratio of 15 mL/g at 65 ℃ for 5 h. The resulting solids were filtered, washed with deionized water until the washings were neutral and then dried in an oven. These samples are denoted as Hβ-B_xA_y, where y indicates the HNO₃ concentration. Ceria-modified hierarchical Hβ-B_0.2A_0.2 zeolites were prepared by an impregnation method with a series of Ce(NO₃)₃ concentration solutions at 65 ℃ for 5 h, and then dried in an oven and calcined at 550 ℃. The obtained catalysts are denoted as z% Hβ-B_0.2A_0.2, where z indicates the mass ratio of ceria to zeolites.

XRD data of the samples were collected in the 2θ range of 5°-80° with a step of 0.02° on a Rigaku D/max-2400 apparatus with Cu K_α radiation. Ratios of Si/Al in the samples were determined by X-ray fluorescence (XRF) conducted on a Bruker SRS3400 instrument. Nitrogen adsorption experiments at -196 ℃ were measured on samples previously evacuated at 150 ℃ overnight with the use of a Beishide 3H-2000PS1. Micropore volumes were determined from t-plots. The specific surface areas were determined with the Brunauer-Emmett-Teller (BET) model. Pore size calculations were performed from the sorption isotherms based on the BJH method. NH₃ temperature-programmed desorption (NH₃-TPD) data were acquired on a Builder Chemisorption (PCA-1200) instrument with a thermal conductivity detector (TCD) to measure the desorbed NH₃. A 50-mg sample was loaded into a quartz reactor between two quartz wool plugs, and then pretreated under an Ar atmosphere at 500 ℃ (a ramp rate of 10 ℃/min) for 1 h, followed by cooling to room temperature. The pretreated sample was saturated with ammonia at 100 ℃ via pulse injection of ammonia. The ammonia desorption process was performed from 100 ℃ to 600 ℃ at a heating rate of 10 ℃/min under an Ar flow. TGA analysis was conducted to study the coking behavior of spent Hβ, Hβ-B_0.2A_0.2 and 5 wt% CeO₂ Hβ-B_0.2A_0.2 catalysts for 6 h time on stream with the use of a Perkin-Elmer STA 6000 instrument at a heating rate of 10 ℃/min from 25 to 800 ℃ under an air flow (60 mL/min).

The alkenylation of p-xylene with phenylacetylene was performed in a stainless steel fixed-bed continuous-flow reactor at 140 ℃. A 1.0-g portion of catalyst with 20-60 mesh was loaded into the reactor for all performance tests. The residual space in the reactor was filled with 20-60 mesh quartz granules. The catalyst was pre-activated in situ for 1 h under an N₂ flow at 30 mL/min, and then input into the fixed-bed reactor by a syringe measuring pump. N₂ (99.999% purity) was used to maintain the reaction system pressure. The product was analyzed on FULI 9790 II GC with FID detector using HP-5 column (30 m × 0.32 mm × 0.25 μm). The GC chromatogram of the products was found and the molecular structure of alpha-(2, 5-dimethylphenyl) styrene (Ⅰ) was further identified by ¹H NMR, as described in our previous report [18]. As evaluation criterion of the alkenylation reaction, the conversion of phenylacetylene was calculated as the weight percentage of converted phenylacetylene in the total amount of introduced phenylacetylene in the feed. The main product alpha-arylstyrene were labeled as (Ⅰ), together with a series of side products, including acetophenone (Ⅱ), (2, 5-dimethylphenyl) ethylbenzene (Ⅲ), beta-(2, 5-dimethylphenyl) styrene (Ⅳ), and oligomers (Ⅴ). Moreover, the stability of the developed ceria-modified and unmodified hierarchical beta zeolites as well as that of the pristine beta zeolite was investigated.

Fig. 1 shows the XRD patterns of Hβ and Hβ-B_xA_0.2 processed with a range of NaOH concentrations, which featured the typical diffraction pattern of a beta zeolite. These results indicated that sequential alkali and acid treatments led to a gradual reduction in crystallinity, as seen by the decreasing diffraction intensity relative to that of pristine Hβ. The decrease in crystallinity of the as-prepared Hβ-B_xA_0.2 zeolites can be clearly seen as the concentration of the NaOH solution was increased.

Fig. 2(a) shows that the N₂ adsorption-desorption isotherms measured for the pristine Hβ zeolite were type I isotherms of a purely microporous material [39]. From Fig. 2(b) and Table 1, the parent Hβ zeolite exhibited a minor mesoporous surface area and pore volume with pore size in the range of 5-30 nm, owing to the contributions of inter-particle voids originating from the disordered agglomeration of small crystals of the Hβ zeolite [30, 31]. Processed Hβ zeolites with alkali-acid solution resulted in a clear hysteresis loop and a greater uptake of N₂ adsorption-desorption isotherms at high relative pressures, giving a relatively regular mesoporous distribution in the major peak profile compared with that of pristine Hβ. The N₂ adsorption-desorption isotherms of processed Hβ-B_xA_0.2 zeolites contained type I and IV features, indicating a typical hierarchical structure [40]. The values of V_micro and S_micro decreased as the NaOH concentration was increased from 0.1 to 0.3 mo/L. In addition, the values of V_meso and S_meso increased as NaOH concentration was increased from 0.1 to 0.2 mol/L. Conversely, the S_meso value of Hβ-B_0.3A_0.2 decreased as the NaOH concentration was increased up to 0.3 mol/L, attributed to the resulting framework of the zeolites collapsing at the higher alkali concentration [36]. From Table 1, while the HNO₃ concentration was fixed with 0.2 mol/L, the Si/Al ratio increased as the alkali concentration was increased, and reached a maximum of 79 at 0.1 mol/L NaOH. Further increase of the concentration from 0.1 to 0.3 mol/L led to a continuous decrease of the Si/Al ratio. This result could be mainly attributed to the strengthened dealumination as the alkali concentration was increased [32].

The NH₃-TPD technique was used to study the acidic properties of the series of Hβ-B_xA_0.2, and pristine Hβ was examined for comparison. The NH₃-TPD profiles are presented in Fig. 3. On the basis of previous reports [18-23], the NH₃-TPD profiles could be divided into three regions, which can be respectively assigned as weak (150-300 ℃), medium (300-450 ℃), and strong (450-600 ℃) acid sites. The amount of acid sites of the pristine Hβ zeolite was 722 μmol/g. However, a lower amount of 561 μmol/g acid sites was observed on Hβ-B_0.1A_0.2, which was attributed to the loss of Al³⁺. Owing to the decrease of the Si/Al ratio through desilication at higher alkali concentrations, the amount of acid sites increased from 561 to 840 μmol/g as the alkali concentration was increased from 0.1 to 0.2 mol/L. Although the Si/Al ratio decreased, a further increase in the alkali concentration from 0.2 to 0.3 mol/L resulted in a decrease in the amount of acid sites, mainly because of the collapsed framework of the zeolite owing to the excessively high NaOH concentration.

From Table 2, unlike pristine Hβ zeolite, hierarchical zeolites prepared with different alkali concentrations exhibited much higher catalytic activity for alkenylation, attributed to the increase in the amount of acid sites and the strengthened mass transfer of the as-formed mesopores. The Hβ-B_0.2A_0.2 sample showed the highest conversion. Owing to the lower amount of acid sites and collapsed framework resulting from an excessively high alkaline concentration, the Hβ-B_0.3A_0.2 catalyst also showed a lower conversion compared with Hβ-B_0.2A_0.2 prepared at an optimum alkali concentration. Moreover, narrow micropores are beneficial for producing the beta-(2, 5-dimethylphenyl) styrene (Ⅳ) because of the thermodynamic stability compared with the main production. As a result, both Hβ-B_0.2A_0.2 and Hβ-B_0.3A_0.2 demonstrated a lower percentage of the product Ⅳ. Furthermore, the collapsed framework of Hβ-B_0.3A_0.2 limited mass transport, which led to formation of more oligomers [41]. Thus, the hierarchical Hβ-B_0.2A_0.2 zeolite treated by alkali and acid at 0.2 mol/L of alkali demonstrated a much higher catalytic activity and selectivity than both pristine Hβ and the hierarchical zeolites prepared under other alkali concentrations, attributed to more acidic sites and greater accessibility to acidic sites.

On the basis of the optimized alkali concentration for the alkali/acid treatment for synthesizing hierarchical beta zeolite, the effect of the acid concentration on the nature and catalytic performance of the as-synthesized hierarchical beta zeolites was investigated by adjusting the acid concentration from 0 to 0.3 mol/L. The as-synthesized hierarchical beta zeolite catalysts are listed as Hβ-B_0.2, Hβ-B_0.2A_0.2 and Hβ-B_0.2A_0.3. Powder XRD patterns of the Hβ-B_0.2, Hβ-B_0.2A_0.2, and Hβ-B_0.2A_0.3 are shown in Fig. 4. All the samples featured typical XRD patterns corresponding to beta zeolite. Moreover, the Hβ-B_0.2 and Hβ-B_0.2A_0.2 featured a lower crystallinity than that of Hβ-B_0.2A_0.3, which can be attributed to the residual amorphous aluminosilicate species produced by alkali treatment of beta zeolite in the framework. The acid treatment can efficiently remove the as-formed aluminosilicate species from the hierarchical zeolite framework.

Fig. 5(a) depicts the N₂ adsorption-desorption isotherms of the Hβ-B_0.2, Hβ-B_0.2A_0.2 and Hβ-B_0.2A_0.3 samples, and the mesoporous size distribution is shown in the Fig. 5(b). The textural properties are listed in Table 3. From Fig. 5(a), no sharp rise on the isotherm of Hβ-B_0.2 at lower P/P₀ could be observed, suggesting that no micropores were present in this sample. This could be attributed to the blockage of inherent micropores in the beta zeolite by the as-formed amorphous aluminosilicate species in the process of the alkali treatment. From Table 3, both the surface area and the pore volume of the micropores were close to 0. From Fig. 5(a), the isotherms of the three samples showed type Ⅳ characteristic with a hysteresis loop, implying that mesopores were induced in the as-prepared hierarchical zeolites by the alkali/acid treatment. Fig. 5(b) featured mesopores with sizes in the range of 2-30 nm for all three samples. From Table 3, after the alkali-treated beta zeolites were subjected to the acid treatment, the samples featured a greatly increased specific surface area together with an increase in the pore volumes of the micropore and mesopore regimes in the as-prepared hierarchical zeolites. This result can be attributed to the role of the acid treatment in removing amorphous aluminosilicate and the further dealumination, which leads to formation of additional mesopores. The considerable increase in surface area and pore volume, of mesopores in particular, is of great importance for enhancing the alkenylation by improving accessibility to acidic sites.

From Fig. 6 and Table 3, owing to the as-formed amorphous aluminosilicate having been cleaned up through the acid treatment, both the Hβ-B_0.2A_0.2 and Hβ-B_0.2A_0.3 catalysts showed a greater increase in the number of acidic sites than that of the alkali treated Hβ-B_0.2 without the acid treatment procedure; those catalysts also showed a much higher Si/Al ratio than Hβ-B_0.2. As the concentration was increased, the acid treatment led to a considerable decrease in the Si/Al ratio through dealumination, which markedly affected the acidic properties [42]. Hβ-B_0.2A_0.2 zeolite prepared by alkali and acid treatment in sequence with a 0.2 mol/L of alkali and acid concentration showed a much higher amount of acidic sites than those of the other two samples, which suggests that it might be a highly active alkenylation catalyst.

Table 4 presents the catalytic results of the alkenylation of p-xylene with phenylacetylene over a series of Hβ-B_0.2A_y zeolites treated with acid over a range of concentrations. From Table 4, Hβ-B_0.2 showed lower activity and selectivity for alkenylation than those behaviors of the acid treated hierarchical Hβ-B_0.2 zeolites. This result was attributed to the lower surface area and pore volume, and the blockage of acidic sites resulting by residual amorphous aluminosilicate formed in the alkali treatment process. As the acid concentration was increased from 0 to 0.2 mol/L, an increase in catalytic activity was found, which could be attributed to more acidic sites generated by the clean-up of blockages in the amorphous aluminosilicate. However, further increasing the acid concentration up to 0.3 mol/L led to a decrease in catalytic activity, attributed to a decrease in the amount of acidic sites caused by dealumination in the acid treatment. Hβ-B_0.2A_0.2 featured the highest catalytic activity and selectivity among the four catalysts, which was attributed to this sample having the largest amount of acidic sites and the highest specific surface area and pore volume. The 0.2 mol/L acid concentration treatment was found to be the optimum acid concentration for achieving excellent catalytic performance for the alkenylation reaction.

Fig. 7 demonstrates the catalytic stability of the as-prepared Hβ-B_0.2A_0.2 zeolite solid acid catalyst for alkenylation of p-xylene with phenylacetylene. The performance of pristine Hβ is also included for comparison.

From Fig. 7, unlike the pristine Hβ solid acid catalyst, the as-prepared hierarchical Hβ zeolite featured superior stability for alkenylation of p-xylene with phenylacetylene. However, the catalytic stability of the hierarchical Hβ zeolite was unsatisfactory. A further increase of the stability would be highly desirable. From previous reports, the modification of zeolites by metal oxides, such as alkaline earth and rare earth metal oxides has been demonstrated to be an efficient method for improving their catalytic performance and stability for various transformations [32-38]. However, there have been few reports on the promotion of alkenylation through metal oxide modification. Herein, CeO₂ modified hierarchical Hβ zeolites with a range of loadings were prepared by impregnation, and the catalytic stability of the ceria-modified hierarchical Hβ zeolites for alkenylation was investigated. As shown in Fig. 7, the 5 wt% CeO₂ modified hierarchical Hβ zeolite showed higher stability for alkenylation than the unmodified Hβ-B_0.2A_0.2 zeolite and the pristine Hβ zeolite. Both hierarchical zeolites show higher selectivity than that of the pristine zeolite, owing to enhanced mass transfer from the as-formed hierarchical pores. The ceria-modified Hβ-B_0.2A_0.2 zeolite showed a slightly lower selectivity than that of the unmodified Hβ-B_0.2A_0.2 zeolite, which may be attributed to the partial blockage of hierarchical pores by the introduced ceria.

Fig. 8 shows that the hierarchical Hβ-B_0.2A_0.2 zeolite featured more acidic sites than did the pristine Hβ, which allowed Hβ-B_0.2A_0.2 to exhibit a higher initial catalytic activity. The higher stability of Hβ-B_0.2A_0.2 could be ascribed to its hierarchical pore structure. Furthermore, owing to possible blockage of Bronsted acidic sites by the ceria modification [38], the 5 wt% CeO₂ Hβ-B_0.2A_0.2 zeolite featured fewer weak, medium and strong acidic sites compared with those of the unmodified Hβ-B_0.2A_0.2 zeolite. The loss of Bronsted acidic sites of ceria-modified Hβ-B_0.2A_0.2 zeolite might have reduced polymerization and oligomerization, which in turn reduced the formation of carbonaceous deposits. As a result, the 5 wt% CeO₂ Hβ-B_0.2A_0.2 zeolite demonstrated much higher catalytic stability for alkenylation than that of the unmodified Hβ-B_0.2A_0.2 zeolite. Moreover, some of the acidic sites on Hβ-B_0.2A_0.2 could become covered by carbonaceous species that rapidly formed owing to the higher concentration of strong acidic sites. This effect in turn led to a decrease in catalytic activity. As a result, from Fig. 7, the lower amount of acidic sites on the 5 wt% CeO₂ Hβ-B_0.2A_0.2 zeolite did not result in a decrease of its catalytic activity over the investigated time range on stream (from 240 to 480 min).

TGA was used to reveal the possible mechanism for the remarkably improved catalytic stability of our optimized catalyst compared with those of the unmodified Hβ-B_0.2A_0.2 zeolite and the pristine Hβ zeolite. From Fig. 9, the ceria-modified and unmodified Hβ-B_0.2A_0.2 zeolites featured similar total weight losses, which was less than that of the pristine Hβ zeolite. This result was attributed to the hierarchical pore structure of the former. As a result, both hierarchical zeolites featured superior stability to that of the pristine Hβ zeolite. Furthermore, the spent ceria-modified Hβ-B_0.2A_0.2 zeolite showed lower oligomer content (weight loss at low temperature) and more eupolymer (weight loss at high temperature) content compared with those features of the spent unmodified Hβ-B_0.2A_0.2 zeolite catalyst. The reason for this effect is unclear and will be further investigated in our future work. Moreover, the temperature of carbon combustion of the ceria zeolite was clearly lower than that of the unmodified Hβ-B_0.2A_0.2 zeolite and the pristine Hβ, which might be the main reason for its higher catalytic stability. Hence, the developed 5 wt% CeO₂ Hβ-B_0.2A_0.2 zeolite shows great potential as a solid acid catalyst for producing alkenyl aromatics through alkenylation.

Ceria-modified hierarchical Hβ zeolites were prepared by a desilication-dealumination procedure with subsequent cerium impregnation. Our results show that the ceria-modified hierarchical Hβ zeolite catalyst prepared with optimized parameters demonstrated high catalytic activity, selectivity, and stability for alkenylation of p-xylene with phenylacetylene, attributed to the more acidic sites and the improved accessibility to active sites owing to enlarged pores, a higher mesoporous surface area, and a mesoporous volume resulting from the hierarchical pore architecture and ceria modification. The combination of excellent catalytic performance and the easy recovery of the spent catalyst by calcination, make the 5 wt% CeO₂-Hβ-B_0.2A_0.2 catalyst with a hierarchical pore architecture attractive for producing alkenyl aromatics through acid catalyzed alkenylation of aromatics.