Alkenyl aromatics, synthesized by acid-catalyzed alkenylation of aromatic compounds with alkynes, especially solid acid catalysts in terms of its features of high efficiency, low-cost, clean, renewability, easy-separation, and applicability toward large-scale industrial continuous production [1-3], are widely applied to diverse fields such as flavors, dyes, natural products, pharmaceuticals, and agrochemicals [4-6]. Due to oligomerization of alkynes as the result of the poor stability of vinyl cation species [7], alkenylation remains a serious challenge to overcome. Compared with homogeneous catalysts, which leads to heavy pollution and equipment corrosion issues, heterogeneous catalysts, especially solid acid catalysts, show greater application potential for the alkenylation, whose acidity and pore structure have decisive effect on the catalytic activation, selectivity, and coke resistance. Therefore, it is urgent to develop highly efficient solid acid catalysts for alkenylation [8].

Some pioneering work has been done by Sartori who used HSZ-360 to catalyze the alkenylation of aromatics, which gave an unsatisfactory result owing to the existing irreconcilable contradiction between catalytic activity and selectivity. Besides, the use of HY zeolite to alkenylation contributed to a low catalytic efficiency, ascribed that the reaction only took place on the external surface of the catalyst as a result of the narrow pore channels within the HY zeolite [9]. Consequently, the mesoporous solid acid catalyst may be a wise choice for alkenylation. In our previous reports, a series of supported phosphotungstic acid mesoporous solid acid catalysts have been developed for alkenylation, but inevitable use of volatile organic solvent for the recovery of spent catalysts results in economic and environmental issues [10-12]. Besides, a series of sulphated La-mediated ZrO₂-based solid acid catalysts have been developed for the alkenylation [13-15], considering the superacidic properties and good thermal stability of sulphated zirconia [16-18], but SO₄²⁻ tends to leach in the reaction, resulting in the deactivation of catalysts. Moreover, hierarchical Hβ zeolite and ceria-modified hierarchical Hβ zeolite have been developed as highly efficient solid acid catalysts for alkenylation due to their high thermal stability, while coke is easy to form due to the existence of micropores [19, 20]. Therefore, the further work on developing robust alkenylation solid acid catalysts is required.

Tungstated zirconia catalysts have attracted considerable attention in the fields of academia and industry owing to their enhanced regeneration ability and good thermal stability [21], and have been applied to diverse reactions [22-28], but rare reports on alkenylation. Here, we developed highly dispersed silicotungstic acid (STA)-derived WO₃ composited with ZrO₂ supported on SBA-15 (called as WZ/SBA-15) as mesoporous solid acid catalyst prepared by incipient wetness impregnation (IWI) method, that the active components, WO₃ and ZrO₂, were impregnated into the channels of SBA-15 simultaneously to enhance the interaction among them, with a subsequent calcination process for the alkenylation. Besides, alkenylation of p-xylene with phenylacetylene was adopted as a model reaction to test the catalytic properties of as-prepared solid acid catalysts. In addition, the effects of the molar ratio of ZrO₂ to WO₃ (n(ZrO₂/WO₃)) and calcination temperature (T) on the catalysts performance were investigated as well. High resolution transmission electron microscopy (HRTEM), high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), N₂ adsorption-desorption, FT-IR, X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), NH₃ temperature-programmed desorption (NH₃-TPD), and FT-IR of pyridine adsorption (Py-IR) characterization techniques were employed to reveal the relationship between the catalysts' nature and catalytic properties.

The SBA-15 material was prepared according to the procedure described by Zhao [29]. Solid acid catalysts W₁₂Z_X/SBA-15-T (X is defined as the molar content of ZrO₂ in contrast to WO₃, X = 12, 15, and 24; T stands for the calcination temperature) were prepared by IWI method by using the mixed aqueous solution containing zirconium nitrate pentahydrate and STA as an impregnant with a subsequent calcination process at a certain temperature for 3 h, and details of preparation process are described in Supporting information. The WO₃-ZrO₂ composite oxide structure of the obtained product is proved by XRD and XPS. Without specific emphasis, W₁₂Z_X/SBA-15 stands for the samples calcined at 550 ℃. The preparation methods of ZrO₂/SBA-15, WO₃/SBA-15, and W₁₂/Z₁₅/SBA-15 are similar to W₁₂Z₁₅/SBA-15 with some changes. Related characterization parameters and details of catalytic performance tests are showed in Supporting information.

Fig. 1 shows the HRTEM and HAADF-STEM images of W₁₂Z₁₅/SBA-15 and W₁₂/Z₁₅/SBA-15. Obviously, the ordered mesoporous structure of SBA-15, proved by XRD and N₂ adsorption-desorption, can be seen in both samples after the loading of active ingredients, ZrO₂ and WO₃. Besides, it is clear that no obvious accumulation of large particles can be observed in W₁₂Z₁₅/SBA-15, while the phenomenon is serious in W₁₂/Z₁₅/SBA-15. The results obtained above implies the enhanced interaction between ZrO₂ and WO₃ in W₁₂Z₁₅/SBA-15, which is in favor of the dispersion of WO₃ in ZrO₂ to produce more acid sites, supported by other characterization results.

Fig. 2a shows the wide XRD patterns of WO₃/SBA-15, ZrO₂/SBA-15, W₁₂/Z₁₅/SBA-15, and W₁₂Z₁₅/SBA-15. Obviously, a broad peak at around 2θ = 20^o-35^o, indicating the amorphous phase of SBA-15 [30], can be observed in all samples. For WO₃/SBA-15, typical diffraction peaks of WO₃ can be observed, meaning that the STA has decomposed into WO₃ due to the high calcination temperature [31-34]. Moreover, no typical diffraction peaks of ZrO₂ in ZrO₂/SBA-15 can be observed, suggesting that ZrO₂ uniformly disperses inside the channels of SBA-15 according to previous reports [35, 36]. Noticeably, no obvious peaks attributed to WO₃ and ZrO₂ can be observed in W₁₂Z₁₅/SBA-15, evidencing that the impregnation method, that active ingredients, ZrO₂ and WO₃, are impregnated into the channels of SBA-15 simultaneously, achieves the uniform dispersion of WO₃ in ZrO₂, which is derived from the strong interaction in the formed composite oxide WO₃-ZrO₂. Nevertheless, some peaks assigned to WO₃ can be observed in W₁₂/Z₁₅/SBA-15 clearly, and the large grains of WO₃ suggest the uneven dispersion of WO₃ due to the relatively weak interaction between ZrO₂ and WO₃ in contrast to W₁₂Z₁₅/SBA-15. The analysis results of XRD are consistent with HRTEM and HAADF-STEM. From the small angle XRD results shown in Fig. S1 (Supporting information), the typical diffraction peaks at 2θ = 0.5^o-2^o are attributed to the diffraction of the (100), (110) and (200) planes of SBA-15 [37], associated with P6mm hexagonal symmetry of SBA-15, which can be observed in W₁₂Z₁₅/SBA-15, indicating the preservation of ordered mesoporous structure, being consistent with the analysis results above. N₂ adsorption-desorption isotherms and pore size distributions of the as-prepared catalysts are presented in Fig. 2b, and SBA-15 is introduced as comparison. The specific surface area (S_BET) and pore volume (V_total) are listed in Table S1 (Supporting information). From Fig. 2b, the characteristic isotherms suggest the mesoporous structure of all the samples. However, the loading of active ingredients results in the remarkable decreases of S_BET, V_total and pore size in all samples, especially W₁₂Z₁₅/SBA-15 and W₁₂/Z₁₅/SBA-15, as shown in Table S1 and Fig 2b. Noticeably, the sharply reduced pore size of WO₃/SBA-15 relative to others indicates the STA tends to aggregate into large particles and evolves into WO₃ after the calcination process at high temperature, resulting in the blockage of channels, which is harmful for mass transfer. Consequently, the increased pore sizes of W₁₂Z₁₅/SBA-15 and W₁₂/Z₁₅/SBA-15, especially the former one, in contrast to WO₃/SBA-15, evidences the effect of ZrO₂ on the dispersion of WO₃. Moreover, compared with W₁₂/Z₁₅/SBA-15, W₁₂Z₁₅/SBA-15 has larger S_BET, V_total and pore size, which is favorable for the dispersion of acidic sites and mass transfer. The acidic properties are the crucial factor for the catalytic performance. Thus, NH₃-TPD method was applied to uncover the acid sites nature of the prepared catalysts. NH₃-TPD profiles are presented in Fig. 2c, and the amount of acid sites (N_a) of the as-prepared catalysts is listed in Table S1. The amount of acidic sites in single component solid acid catalysts ZrO₂/SBA-15 is 71.22 µmol/g and WO₃/SBA-15 possesses a few more acid sites, reaching 136.53 µmol/g. For W₁₂/Z₁₅/SBA-15, the amount of acid sites is basically equal to the sum of ZrO₂/SBA-15 and WO₃/SBA-15. Surprisingly, the amount of acid sites increased sharply in W₁₂Z₁₅SBA-15, reaching to 232.46 µmol/g due to the formation of more composite oxide WO₃-ZrO₂ originated from the special impregnation way of active species described above, which is conductive to enhance solid acidity proved by previous reports [21, 38]. The large number of acid sites in W₁₂Z₁₅SBA-15 evidences the strong interaction of WO₃ and ZrO₂. Moreover, the acidic properties of W₁₂Z₁₅/SBA-15 and W₁₂/Z₁₅/SBA-15 solid acid catalysts were also tested by Py-IR at 150 ℃ desorption temperature, which can be assigned to total acid sites. As shown in Fig. S2 (Supporting information), the vibration bands at 1540 cm⁻¹ and 1450 cm⁻¹ are ascribed to the B and L acid sites, respectively. Obviously, both B and L acidic centers can be observed on the two samples, and the L/B acid site on W₁₂Z₁₅/SBA-15 is much larger than that on W₁₂/Z₁₅/SBA-15. Combined with NH₃-TPD results, the increased acid sites in W₁₂Z₁₅/SBA-15 relative to W₁₂/Z₁₅/SBA-15 are attributed to L acid sites. Both B and L acid sites can promote this reaction according to previous reports [1-3, 11, 13]. Fig. 2d shows the framework FT-IR spectra of as-prepared solid acid catalysts. All samples display three absorption peaks at 1088, 816 and 461 cm⁻¹, which are attributed to Si-O-Si stretching, Si-O stretching, and Si-O-Si bending modes of vibration in SBA-15, respectively [39-41]. These bands, together with the band at about 3458 cm⁻¹ and the band at 1635 cm⁻¹ assigned to vibration of the -OH stretching vibration mode and adsorbed water respectively [40, 42], can be observed in all samples. In the spectra of ZrO₂/SBA-15, no obvious vibration mode of ZrO₂ can be observed due to overlapping with SBA-15 [43, 44]. And for WO₃/SBA-15, the band at 949 cm⁻¹ is assigned to the Si-O-W linkage [45-47]. Obviously, the peak intensity of the band reduces sharply and shifts to 975 cm⁻¹ in W₁₂Z₁₅/SBA-15 and W₁₂/Z₁₅/SBA-15 after the introduction of ZrO₂, evidencing the strong interaction between WO₃ and ZrO₂. Noticeably, the strong absorption peaks of −OH stretching vibration mode and adsorbed water can be observed in WO₃/SBA-15 and W₁₂/Z₁₅/SBA-15, meaning the existence of a considerable amount of water, which is harmful for alkenylation, because the acid sites are easily covered by the water molecules from the competitive adsorption of aromatic compound, resulting in a situation that the aromatic compound are hindered to approach the active centers [48]. The electronic environment of active sites can affect the adsorption properties of reactant [49-52], so the prepared catalysts were investigated by XPS. Fig. 2e shows the XPS spectra of the W 4f region. The peaks at binding energies of 35.93 eV and 37.94 eV in WO₃/SBA-15 can be assigned to W⁶⁺ species [53-56]. A small peak at ~40 eV is due to W 5P_3/2 [53]. Clearly, these peaks shift to lower region in W₁₂/Z₁₅/SBA-15 and W₁₂Z₁₅/SBA-15, implying the strong interaction between WO₃ and ZrO₂, which is further confirmed by the analysis of Zr 3d spectra. The much lower binding energies of W⁶⁺ species in W₁₂Z₁₅/SBA-15 suggest the enhanced interaction between WO₃ and ZrO₂, resulting in the formation of more composite oxide WO₃-ZrO₂. Fig. 2f shows the XPS spectra of Zr 3d region. The binding energies of 182.82 eV and 185.12 eV are assigned to Zr⁴⁺ species in W₁₂/Z₁₅/SBA-15 [53], which shift to higher region in W₁₂Z₁₅/SBA-15 due to the high electron attractor effect of the neighboring W atoms [57], which demonstrates the enhanced interaction between WO₃ and ZrO₂ in contrast to W₁₂/Z₁₅/SBA-15 again. According to the analysis results above, W₁₂Z₁₅/SBA-15 prepared by the special impregnation method mentioned above possesses more acid sites, relatively high S_BET, large V_total and moderate pore size due to the uniform dispersion of WO₃ in ZrO₂ originated from the strong interaction among them, which is in favor of alkenylation.

In order to verify the effect of synergy of WO₃ with ZrO₂ on the boost for the alkenylation of p-xylene with phenylacetylene, the catalytic properties of prepared catalysts were investigated and the reaction results are depicted in Table 1. The alkenylation reaction of p-xylene with phenylacetylene is a quite complex competition process, and the mixture consists of main product α-(2, 5-dimethylphenyl) styrene (Ⅰ), also called as α-arylstyrene, and side-products acetophenone (Ⅱ), α-(2, 5-dimethylphenyl)ethylbenzene (Ⅲ), β-(2, 5-dimethylphenyl)styrene (Ⅳ), and oligomers (Ⅴ). From Table 1, W₁₂Z₁₅/SBA-15 exhibits the best catalytic performance due to more acid sites, relatively high S_BET, large V_total and suitable pore size. The conversion of phenylacetylene can reach 94.7% with a considerable selectivity of α-arylstyrene, reaching 88.4%. Besides, fewer oligomers (Ⅴ), being easy to evolve into coke during the reaction, can be attained over W₁₂Z₁₅/SBA-15 compared with the other three samples, evidencing that the oligomerization is limited and therefore coke reduces, preventing the deactivation of catalysts consequently. However, W₁₂/Z₁₅/SBA-15 presents poorer catalytic properties compared with W₁₂Z₁₅/SBA-15, both the conversion of phenylacetylene and the selectivity of α-arylstyrene, due to the relatively weak interaction between WO₃ and ZrO₂. Moreover, the specific activity results shown in Table S2 (Supporting information) suggest that the excellent performance of W₁₂Z₁₅/SBA-15 is attributed to the increase of number of active sites and the enhanced intrinsic activity of acid sites, evidencing the superiority of W₁₂Z₁₅/SBA-15 again. Not surprisingly, it is anticipated that both ZrO₂/SBA-15 and WO₃/SBA-15 show bad results for the alkenylation, even though ZrO₂/SBA-15 presents considerable α-arylstyrene selectivity, reaching 93.3%. Combined with characterization and reaction results, the enhanced interaction of WO₃ with ZrO₂ in W₁₂Z₁₅/SBA-15, contributing to the formation of more composite oxide WO₃-ZrO₂ resulting from the special impregnation way of active component, has a decisive effect on the boost for alkenylation of p-xylene with phenylacetylene. Clearly, W₁₂Z₁₅/SBA-15 prepared by the method described above is more conducive for alkenylation. Consequently, the composition of active species and the calcination temperature of the solid acid catalysts were optimized to obtain the optimal catalyst.

According to results above, the introduction of ZrO₂ is favorable for the dispersion of WO₃, and the formation of composite oxide WO₃-ZrO₂ can produce more acid sites. Therefore, a series of W₁₂Z_X/SBA-15 solid acid catalysts with various n(ZrO₂/WO₃) were prepared. The effect of n(ZrO₂/WO₃) on the catalyst structure was investigated by XRD experiments, and the characterization results are shown in Fig. S3 (Supporting information). Obviously, two weak peaks assigned to WO₃ can be observed in W₁₂Z₁₂/SBA-15, not seen in W₁₂Z₁₅/SBA-15 and W₁₂Z₂₄/SBA-15. Therefore, it is clear that the high loading of ZrO₂ helps to disperse WO₃ through the strong interaction among them. Figs. S4a and b (Supporting information) present N₂ adsorption-desorption isotherms and pore size distributions of the as-prepared W₁₂Z_X/SBA-15 catalysts. The S_BET, and V_total are listed in Table S3 (Supporting information). With the rise of n(ZrO₂/WO₃) from 12/12 to 24/12, the S_BET decreases monotonically due to the high loading of active ingredients. Moreover, V_total and pore size increase when n(ZrO₂/WO₃) increases from 12/12 to 15/12 due to the more uniform dispersion of WO₃ in ZrO₂ reducing the blockage of channels, resulting from the strong interaction among them. However, smaller V_total and pore size can be obtained in W₁₂Z₂₄/SBA-15 due to excessive addition of ZrO₂ compared with W₁₂Z₁₅/SBA-15, which is harmful for mass transfer. According to the results above, the conclusion can be obtained that the appropriate addition of ZrO₂ is beneficial to the dispersion of WO₃ to form composite oxide WO₃-ZrO₂, which is conductive to produce more accessible acid sites. Nevertheless, excessive addition of ZrO₂ results in the reducing of exposure of composite oxide WO₃-ZrO₂ to surface, fewer accessible acid sites producing consequently confirmed by NH₃-TPD results. Fig. S5 (Supporting information) shows the NH₃-TPD profiles of the as-prepared catalysts, and the number of acid sites is listed in Table S3 (Supporting information). Clearly, W₁₂Z₁₅/SBA-15 has the acidest sites due to the appropriate loading of ZrO₂, being consistent with analysis results above. Fig. 3a and Table S4 (Supporting information) shows the catalytic performance of as-prepared W₁₂Z_X/SBA-15 catalysts with diverse n(ZrO₂/WO₃). Based on more acid sites, larger V_total and pore size compared with the other two samples, W₁₂Z₁₅/SBA-15 shows best catalytic performance for alkenylation. Due to the suitable pore size, the least oligomers (Ⅴ) can be obtained over W₁₂Z₁₅/SBA-15. According to the results above, W₁₂Z₁₅/SBA-15 stands for the optimum composition of ZrO₂ and WO₃.

Subsequently, a series of solid acid catalysts calcined at various T were prepared. The textural properties of the as-prepared catalysts were investigated by N₂ adsorption-desorption experiment, and relevant results are illustrated in Figs. S6a and b (Supporting information). S_BET and V_total are listed in Table S5 (Supporting Information). It is obvious that the S_BET decreases monotonically with the rise of T. Moreover, V_total presents a completely opposite tendency in contrast to S_BET. The pore size increases when T rises from 450 ℃ to 500 ℃, and it decreases when T further increases from 500 ℃ to 550 ℃. Fig. S7 (Supporting information) shows the NH₃-TPD profiles of the as-prepared solid acid catalysts calcined at diverse temperature, and the amount of acid sits of relevant catalysts is listed in Table S5. Clearly, W₁₂Z₁₅/SBA-15-500 has the acidest sites due to suitable T, reaching 234.75 µmol/g, which is greater than W₁₂Z₁₅/SBA-15-550. Fig. S8 (Supporting information) shows the FT-IR spectra of the as-prepared catalysts, and ZrO₂/SBA-15 and WO₃/SBA-15 are introduced as reference. Obviously, for W₁₂Z₁₅/SBA-15-450, the intensity of bands at around 1635 cm⁻¹ and 3444 cm⁻¹ are obviously stronger compared with the other samples, meaning the existence of quantity of crystalline water, which is harmful to alkenylation. Combined with the results above, appropriate T of solid acid catalyst can produce more acid sites, and larger V_total, which is favorable for alkenylation. As is shown in Fig. 3b and Table S6 (Supporting information), W₁₂Z₁₅/SBA-15-500 shows the best catalytic performance with a 99.4% conversion of phenylacetylene and a 92.3% selectivity of α-arylstyrene due to the acidic sites, large V_total. Moreover, very low level of oligomers (Ⅴ) can be obtained at the same time, which is attributed that suitable pore size reduces the formation of coke and avoids the blockage of channels, being conductive to mass transfer. Besides, W₁₂Z₁₅/SBA-15-450 shows obviously worse catalytic properties compared with the other three samples, resulting from the overlay of active sites by the considerable crystalline water proved by FT-IR. Therefore, the calcination temperature of 500 ℃ is most favorable for obtaining optimized solid acid catalyst for alkenylation.

In addition, the stability and regeneration of the obtained optimized solid acid catalyst W₁₂Z₁₅/SBA-15-500 was investigated, and the conversion and selectivity as a function of time on steam for the alkenylation of p-xylene with phenylacetylene over the fresh and regenerated W₁₂Z₁₅/SBA-15-500 is shown in Figs. 3c and d, with WO₃/SBA-15 introduced as comparison. At 300 min, W₁₂Z₁₅/SBA-15-500 presents the highest conversion of phenylacetylene with the highest selectivity of α-arylstyrene, reaching 99.4% and 92.9% respectively. 77.9% conversion can be maintained for up to 480 min of time on stream over W₁₂Z₁₅/SBA-15-500 solid acid catalyst with excellent 91.3% selectivity of α-arylstyrene. Moreover, the selectivity of α-arylstyrene still reaches 87.8% at 720 min with a poor catalytic activity, implying the deactivation of part of catalysts. As for regenerated one, recovered from spent W₁₂Z₁₅/SBA-15-500 by a simple calcination process, the activity and selectivity of α-arylstyrene are just a little less than the fresh one with the same variation tendency. It is evident that the deactivation of catalyst is derived from carbon deposition, which can be removed by a simple calcination process. However, the catalytic performance of WO₃/SBA-15 is bad due to more oligomers producing during the reaction, which will evolve into coke to overlay the active sites, resulting in the rapid deactivation of catalyst. According to the evidence described above, W₁₂Z₁₅/SBA-15-500 possesses excellent long-time stability and renewability.

In conclusion, the developed ordered mesoporous solid acid catalyst W₁₂Z₁₅/SBA-15-500 prepared by IWI method with a subsequent calcination process presents excellent catalytic performance for the alkenylation of p-xylene with phenylacetylene, with a 99.4% conversion of phenylacetylene and a 92.3% selectivity of main product (I) α-arylstyrene, which is attributed that the special preparation method that active components, WO₃ and ZrO₂, are impregnated into the channels of SBA-15 simultaneously achieves the uniform dispersion of WO₃ in ZrO₂, resulting in the formation of more composite oxide WO₃-ZrO₂, more accessible acid sites producing consequently. Besides, the relatively high specific surface area, large pore volume and ordered-mesoporous structure of SBA-15 are favorable for mass transfer. Moreover, the regenerated catalyst obtained by a simple calcination process shows comparable catalytic properties with the fresh one, meaning the excellent regeneration of the developed solid acid catalyst. Consequently, W₁₂Z₁₅/SBA-15-500 can be regarded as an outstanding candidate for the alkenylation.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

This work is financially supported by the National Natural Science Foundation of China (No. 21276041) and by the Chinese Ministry of Education via the Program for New Century Excellent Talents in University (No. NCET-12-0079).

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2021.07.071