Acid catalysts have been widely used in both academia and industry, and can be categorized into two major types: homogeneous liquid acids and heterogeneous solid acids ^{[1, 2, 3]}. Although homogeneous mineral acids, such as H₂SO₄ and HF, which are corrosive and toxic, have been widely adopted by chemical industry, with increasing pressure coming from environmental protection there is a need to develop green synthetic systems based on environmentally benign acid catalysts ^{[4, 5, 6, 7, 8]}. The use of heterogeneous solid acid catalysts has the general advantages of easier operation, milder conditions, reduced equipment corrosion, and minimized contamination of the waste streams as well as reusability of the catalyst ^{[9, 10, 11]}. For this reason, heterogeneous solid acids have been extensively investigated ^{[12, 13, 14, 15]}, and in some chemical processes, their use has already contributed to significant improvement of both the synthetic efficiency and the greenness of the system ^{[16, 17, 18, 19, 20, 21, 22, 23]}. Despite promising and sometime spectacular results, the practical application of solid acids in industrial processes is still hindered by many factors. Among all the methods used to prepare solid acids, the binding of catalysts to organic polymers or inorganic solids is widely used to produce a stable covalent bond between the acid center and the solid support, and is preferable to reduce the chance of leaching ^{[24, 25, 26]}. The synthesis of some solid acids with this method, however, involves several steps which are rather time-consuming. Furthermore, functionalization of a solid support using covalent bonds often requires harsh reaction conditions and as a consequence, calls for both a support and an acid precursor with sufficient chemical and thermal stability ^[27]. In many cases, a sophisticated operational procedure is also needed ^[28]. Therefore, some current solid acid catalysts are rather expensive and are therefore not widely used in synthesis. There thus exists a need to develop new, innovative app roaches toward the design of recoverable and reusable solid acid catalysts.

Recently, the concept of solid-supported liquid phase catalysis has experienced a rekindled interest in the catalytic and synthetic community for the preparation of efficient heterogeneous catalytic materials ^{[29, 30, 31, 32, 33, 34]}. With the aid of a thin layer of less volatile fluid coated onto the surface of solid support, some homogeneous catalysts have been easily immobilized without employing sophisticated operational procedures. In a selected reaction, solid form of the catalyst enables an easy recycling, but the reaction proceeded inmicrocosmically homogeneous conditions ^{[35, 36]}. Inspired by this elegant strategy, we envision that it may be possible to transfer this concept to the development of a solid acid catalyst. To realize this idea, an acidic species that is amenable to the immobilization step and can stay stable on the surface of solid support is needed.

Policresulen is prepared by the condensation of metacresolsulfonic acid and formaldehyde and is commercially available as a clear red-brown solution. It is available at a low price in a number of countries worldwide ^[37]. Its systematic name is 2-hydroxy-3,5-bis(4-hydroxy-2-methyl-5-sulfobenzyl)- 4-methylbenzenesulfonic acid. Policresulen is a topical hemostatic and antisepticwidely used as medicine for treating common anal disorders, such as hemorrhoids, and for gynecological infections ^[38]. Because the backbone of this reagent contains many hydroxyl and sulfonic acid groups, it is barely soluble in many organic solvents, such as toluene, ethyl acetate, and dichloroethane. The thermal stability and chemical stability of policresulen are quite good, making the extension of its applications to other fields possible. Considering that policresulen is a strong organic acid, and also taking into account its hydrophilic properties, we propose that it may be possible to construct a new heterogeneous acid catalyst using this reagent as a homogeneous counterpart in conjunction with silica as a supporting material. Thus, we herein report a new solid acid composed of silica and policresulen that displays remarkable catalytic activity in many organic reactions. Compared with conventional solid acids, such as silica-supported sulfonic acid catalysts and Amberlyst-15, the materials obtained by our method have several advantages including simple and time-efficient preparation, low cost, and controllable sulfonic acid loading as well as tunable catalytic activity.

In a typical procedure used to prepare silica/policresulen composite catalyst 1b, silica gel (10 g) was added to a 250-mL round-bottomed flask containing a mixture of aqueous policresulen solution (52 wt%, 2.09 g) and ethanol (150 mL). The suspension was stirred at room temperature for 30 min prior to the removal of the ethanol and water under reduced pressure, affording free-flowing solid 1b. Catalysts with different amounts of loaded policresulen (1a, 1c, 1d) were prepared in an analogous way.

Titanium oxide (5 g) and toluene (100 mL) were added to a 250-mL round-bottomed flask equipped with magnetic stirring and a Dean-Stark apparatus. After 10 h of reflux, 4-hydroxybenzenesulfonic acid (3.87 g) dissolved in acetonitrile (7 mL) was added to the system and refluxed for another 16 h. At the end of the reaction, the solid was filtered, washed with ethanol (50 mL × 5) and dried under vacuum conditions, yielding catalyst 1f. The ion exchange capacity of this catalyst was measured by titration. The obtained result was in good agreement with elemental analysis of the sulfur and carbon content.

SiO₂/policresulen composite catalyzed β-alkylation of styrene with benzhydrol. All reactions were conducted in a 10-mL V-type vessel equipped with a triangular magnetic stirrer. In a typical reaction, benzhydrol (92.1 mg, 0.5 mmol), styrene (62.4 mg, 0.6 mmol), and 1b (20 mg, 2 mol%) were mixed in 1,2-dibromoethane (1.0 mL) and stirred for 5 h at 80 °C. After the reaction, the mixture was cooled to room temperature and petroleum ether (3.0 mL) was added prior to centrifugation. The liquid phase was subjected to isolation with PTLC (PE as eluent). The desired product, 4a, was obtained in a yield of 92%. As for the solid phase, after washing with PE (3.0 mL × 3) and drying under vacuum, 1b was used in the next run with a similar procedure. Tests of the substrate scope were conducted in the same manner.

SiO₂/policresulen composite catalyzed oxa-Pictet-Spengler reaction. All reactions were conducted in a 10-mL V-type vessel equipped with a triangular magnetic stirrer. In a typical reaction, 2-(3,4-dimethoxyphenyl)ethanol (91.1 mg, 0.5 mmol), benzaldehyde (64.8 mg, 0.6 mmol), and 1b (10 mg, 1 mol%) were mixed in 1,2-dichloroethane (1.0 mL) and stirred for 5 h at 80 °C. After the reaction, the mixture was cooled to room temperature and then centrifuged, and the liquid phase was subjected to isolation with PTLC (eluting solvent: ethyl acetate/petroleum ether = 1/8). The desired product, 7a, was obtained in a yield of 96%. Tests of the substrate scope were conducted in the same manner. The recycling of 1b was realized after doubling the dose. After washing the solid phase with PE (3.0 mL × 3) and drying under vacuum, the recovered 1b was used in the next run with a similar procedure.

SiO₂/policresulen composite catalyzed C3-alkenylation of 2-methylindoles with methyl acetoacetate. The reaction was conducted in a 10-mL V-type vessel equipped with a triangular magnetic stirrer. In a typical reaction, 2-methylindole (65.6, 0.5 mmol), methyl acetoacetate (116.1 mg, 1.0 mmol), and 1c (50 mg, 10 mol%) were mixed in toluene (1.0 mL) and stirred for 16 h at 100 °C. After the reaction, the mixture was cooled to room temperature and then centrifuged, and the liquid phase was subjected to isolation with PTLC. The desired product, 10a, was obtained in a yield of 90%. After washing the solid phase with PE (3.0 mL × 3) and drying under vacuum, the recovered 1c was used in the next run with a similar procedure.

SiO₂/policresulen composite catalyzed C3-alkenylation of 2-methylindole with phenylacetaldehyde. The reaction was conducted in a 10-mL of V-type vessel equipped with triangle magnetic stirring. 2-methylindole (65.6, 0.5 mmol), phenylacetaldehyde (90.1 mg, 0.75 mmol), ethanol anhydrous (50.6 mg, 1.1 mmol) and 1d (33.3 mg, 10 mol%) were mixed in dichloromethane (1.0 mL) and stirred for 11 h at room temperature. After reaction, petroleum ether (3.0 ml) was added prior to centrifuge, the liquid phase was submitted to isolation with PTLC. The desired product, 10b, was obtained in a yield of 94%. After washing the solid phase with PE (3.0 mL × 3) and drying under vacuum, the recovered 1d was used in the next run with a similar procedure.

(E)-1-chloro-4-(1-(4-methoxyphenyl)-3-phenylallyl)benzene (4j). Pale yellow liquid; ¹H NMR (400 MHz, CDCl₃, 25 °C): δ = 7.34 (d, J = 7.4 Hz, 2H), 7.27-7.23 (m, 3H), 7.19 (t, J = 2.7 Hz, 2H), 7.13 (d, J = 8.5 Hz, 2H), 7.10 (d, J = 8.6 Hz, 2H), 6.84 (d, J = 8.6 Hz, 2H), 6.58 (dd, J_a = 7.3 Hz, J_b = 15.8, 1H), 6.29 (d, J = 15.9 Hz, 1H), 4.79 (d, J = 7.3 Hz, 1H), 3.75 ppm (s, 3H); ¹³C NMR (100 MHz, CDCl₃): δ = 158.4, 142.4, 137.2, 135.2, 132.4, 132.2, 131.6, 130.1, 129.6, 128.7, 127.5, 126.4, 114.1., 55.3, 52.8 ppm; HRMS m/z (ESI) calcd. for C₂₂H₁₉ClNaO [M + Na]⁺: 357.1022, found 357.1020.

(E)-1-fluoro-2-(1-(4-methoxyphenyl)-3-phenylallyl)benzene (4k). Pale yellow liquid; ¹H NMR (400 MHz, CDCl₃, 25 °C): δ = 7.36 (d, J = 7.4 Hz, 2H), 7.28 (t, J = 7.6 Hz, 2H), 7.22-7.12 (m, 5H), 7.10-7.00 (m, 2H), 6.84 (d, J = 8.7 Hz, 2H), 6.64 (dd, J_a = 7.2 Hz, J_b = 15.8 Hz, 1H), 6.31 (d, J = 15.8 Hz, 1H), 5.15 (d, J = 7.1 Hz, 1H), 3.76 ppm (s, 3H); ¹³C NMR (100 MHz, CDCl₃): δ = 161.9 159.4, 158.3, 137.2, 134.4, 131.6, 131.4, 130.9, 130.8, 130.0, 129.9, 129.5, 128.6, 128.2, 128.1, 127.4, 126.4, 124.2, 124.1, 115.7, 115.5, 113.9, 55.3, 46.2, 46.2 ppm; HRMS m/z (ESI) calcd. for C₂₂H₁₉FNaO [M + Na]⁺: 341.1318, found 341.1319.

(E)-1-chloro-2-(1-(4-methoxyphenyl)-3-phenylallyl)benzene (4l). Pale yellow liquid; ¹H NMR (400 MHz, CDCl₃, 25 °C): δ = 7.36-7.30 (m, 3H), 7.28-7.22 (m, 3H), 7.19-7.09 (m, 5H), 6.82 (d, J = 8.7 Hz, 1H), 6.59 (dd, J_a = 1.2 Hz, J_b = 15.9 Hz, 1H), 6.24 (d, J = 15.8 Hz, 1H), 5.33 (d, J = 6.6 Hz, 1H), 3.73 ppm (s, 3H); ¹³C NMR (100 MHz, CDCl₃): δ = 158.4, 141.3, 137.3, 134.3, 134.2, 131.9, 131.8, 130.2, 129.9, 128.7, 128.6, 127.9, 127.5, 127.0, 126.5, 113.9, 55.3, 49.5 ppm; HRMS m/z (ESI) calcd. for C₂₂H₁₉ClNaO [M + Na]⁺: 357.1022, found 357.1025.

(E)-1-chloro-4-(3-phenyl-1-(p-tolyl)allyl)benzene (4m). Pale yellow liquid; ¹H NMR (400 MHz, CDCl₃, 25 °C): δ = 7.33 (d, J = 7.2 Hz, 2H), 7.28-7.21 (m, 4H), 7.21-7.16 (m, 1H), 7.15-7.06 (m, 6H), 6.59 (dd, J_a = 7.4 Hz, J_b = 15.8 Hz, 1H), 6.30 (d, J = 15.8 Hz, 1H), 4.79 (d, J = 7.4 Hz, 1H), 2.30 ppm (s, 3H); ¹³C NMR (100 MHz, CDCl₃): δ = 142.4, 140.1, 137.2, 136.3, 132.3, 132.3, 131.7, 130.1, 129.4, 128.7, 128.6, 128.6, 127.5, 126.4, 53.3, 21.1 ppm; HRMS m/z (ESI) calcd. for C₂₂H₁₉ClNa [M + Na]⁺: 341.1073, found 341.1071.

6,7-Dimethoxy-1-phenethyl-isochroman (7k). Yellow liquid; ¹H NMR (400 MHz, CDCl₃, 25 °C): δ = 7.28 (t, J = 7.2 Hz, 2H), 7.22 (d, J = 6.0 Hz, 2H), 7.16 (tt, J_a = 1.6 Hz, J_b = 8.4 Hz, 1H), 6.58 (s, 1H), 6.52 (s, 1H), 4.70 (d, J = 6.4 Hz, 1H), 4.20-4.11 (m, 1H), 3.83 (s, 3H), 3.80 (s, 3H), 3.78-3.71 (m, 1H), 2.98-2.86 (m, 1H), 2.83-2.71 (m, 2H), 2.65-2.56 (m, 1H), 2.23-2.01 ppm (m, 2H); ¹³C NMR (100 MHz, CDCl₃): δ = 147.5, 147.5, 142.4, 129.9, 128.5, 128.3, 126.1, 125.7, 111.5, 107.8, 74.8, 63.2, 56.0, 55.8, 37.8, 31.4, 28.6 ppm; HRMS m/z (ESI) calcd. for C₁₉H₂₂NaO₃ [M + Na]⁺: 321.1467, found 321.1465.

1-(4-Trifluoromethyl-phenyl)-1,3,4,9-tetrahydro-pyrano[3,4-b]indole (7m). Pale yellow liquid; ¹H NMR (400 MHz, CDCl₃, 25 °C): δ = 7.61 (d, J = 8.1 Hz, 2H), 7. 55 (d, J = 6.8 Hz, 2H), 7.47 (d, J = 8.0 Hz, 2H), 7.24-7.19 (m, 1H), 7.18-7.09 (m, 2H), 5.82 (s, 1H), 4.36-4.22 (m, 1H), 4.03-3.88 (m, 1H), 3.15-3.01 (m, 1H), 2.89-2.77 ppm (m, 1H); ¹³C NMR (100 MHz, CDCl₃): δ = 143.6, 136.3, 132.7, 131.3, 131.0, 130.0, 128.8, 128.4, 128.0, 127.0, 126.1, 125.9, 125.8, 125.5, 125.4, 122.7, 122.4, 120.0, 118.6, 111.2, 109.2, 75.5, 64.9, 22.3 ppm; HRMS m/z (ESI) calcd. for C₁₈H₁₄F₃NNaO [M + Na]⁺: 340.0925, found 340.0922.

Initially, four SiO₂-supported policresulen catalysts were prepared through a simple procedure, which involved (1) mixing an aqueous solution of policresulen (52 wt%) and silica together in ethanol solvent; and (2) removing the volatile solvent under reduced pressure after 30 min of stirring. The SO₃H loadings of the catalysts were 0.20, 0.50, 1.00 and 1.50 mmol/g. The catalysts produced were all fine powders (see the photo in Fig. 1). It should be noted that, with this method, the obtained catalysts also contained a small amount of water. Fortunately, this was easily quantified using the following procedure: (1) unmodified silica was subjected to TG analysis to determine the initial amount of absorbed water; (2) the increase in the weight of the silica material after the immobilization of policresulen was measured, and the weight contributed by pure policresulen was removed. The water contents of all the catalysts were found to be nearly the same, ranging from 15wt% to 17 wt% of the hybrid composite. Two reference catalysts, SiO₂-supported sulfonic acid and TiO₂-supported benzenesulfonic acid were used for comparison in this study. The former was prepared according to our previous method ^{[39, 40]}, whereas the latter was prepared based on a modified literature method (see Electronic Supplementary Information (ESI)) ^{[41, 42]}. The schematic structures of all the prepared catalysts are shown in Fig. 1. Some physico-chemical properties of the materials are given in the ESI. Compared with those of the reference material, unmodified SiO₂, a new set of absorption peaks appeared in the IR spectra of policresulen/SiO₂. The bands and their assignments are as follows: The hydroxyl group stretching vibration v(O−H) at 3451 cm⁻¹ increased in intensity, indicating the presence of hydroxyl groups in the structure of policresulen. The absorption peaks at 1579, 1541, 1491, 1456, and 1408 cm⁻¹ can be ascribed to stretching vibrations of the aromatic ring v(C−C)/v(C=C). Peaks appearing at 691, 577, 550 cm⁻¹also indicated the existence of aromatic structure in the policresulen/SiO₂ catalyst. Meanwhile, the BET surface areas decreased with increase of policresulen loading.

Fig. 1 . Schematic representation of solid-supported catalysts used in this study and photo of prepared catalyst 1c.

After determining that we had successfully synthesized the desired catalytic materials, we applied them to selected organic reactions. β-Alkylations of styrene with alcohols are important C-C bond forming reactions, and many metal complexes or acids have been investigated as catalysts ^{[43, 44, 45]}. Because the only by-product of this type of reaction is water, it has been considered as a green route for the synthesis of multi-substituted olefins ^{[46, 47]}. However, most of the reported catalysts are homogeneous species and cannot be recycled. Although a few heterogeneous catalysts have also been employed ^[48], they suffer from either a time-consuming preparation procedure or call for task-specific experimental tools that are generally inaccessible in a typical organic synthesis lab. To address these drawbacks, our silica/policresulen catalysts, which can be prepared within 1 h without using special instruments, were employed in the nucleophilic substitution of benzhydrol 2a with styrene 3a. As shown in Table 1, no product was obtained in the absence of catalyst (entry 1). The present silica/policresulen composite catalyst was indeed effective for promoting this reaction. With catalyst 1a, which contained a low policresulen loading, the desired product 4a was obtained in only moderate yield after 5 h of reaction at 80 °C in 1,2-dibromoethane (entry 2). The yield of 4a increased with policresulen loading, and when 1b was used, it reached 92% under identical conditions (entry 3). However, no significant improvement in yield was observed when 1b was replaced with 1c (entry 4). It should be noted that blank SiO₂ and TiO₂ were proven to be inactive for this reaction (entry 5). Therefore, the catalytic activity of 1b can be attributed to its coating of policresulen. The reaction also proceeded smoothly with the SiO₂-supported sulfonic acid catalyst 1e. However, the obtained yield was far inferior to that obtained with 1b (entry 6). To our surprise, although the acidic species of the TiO₂-supported benzenesulfonic acid catalyst 1f has a similar molecular structure, a very low yield was obtained with this catalyst (entry 7). Although we are uncertain as to the reasons behind this discrepancy in performance, the poor catalytic activity of 1f can be partially ascribed to insufficient contact between the catalyst and substrate, which is a result of the poor flexibility of the chemically bonded benzenesulfonic acid. A commonly used resin-supported acid, Amberlyst-15, was also examined, and only moderate yield was obtained under identical conditions (entry 8). These results indicate that silica/policresulen composites are indeed able to act as acid catalysts to promote organic reactions. Because the policresulen component is insoluble in DBE, the silica/policresulen composite catalyst 1b could be recovered easily by filtration at the end of the reaction. After 20 min of treatment at 80 °C under vacuum conditions (10 mmHg), the recovered catalyst was subjected to the next run. With this procedure, 1b could be reused at least four times without significant loss in its activity (entry 9). A congener of 1b produced using toluenesulfonic acid (PTSA) as the acidic component (SiO₂-PTSA) was also used as a catalyst for the model reaction. Although good catalytic activity was obtained in the first run, its recyclability was far from satisfactory (entry 10). This result demonstrated that the policresulen component was key to rendering this immobilization protocol possible. Notably, the reaction could also be effectively scaled up by 10 times with similar efficiency (entry 11).

Table 1 Nucleophilic substitution of benzhydrol with styrene over different catalysts. Entry Catalyst Yield (%) 1 — 0 2 1a 78 3 1b 92 4 1c 93 5 SiO2 or TiO2 0 6 1e 63 7 1f 10 8 Amberlyst-15 46 9 a 1b 91 10 congener of 1b with PTSA 89 (55b) 11c 1b 91 Reaction conditions: 2a 0.5 mmol, 3a 0.6 mmol, DBE 1.0 mL, 80 °C, 5 h. a 1b was reused in the fourth run. b the catalyst was reused in the first run. c 5.0 mmol scale reaction.

Table 1 Nucleophilic substitution of benzhydrol with styrene over different catalysts.

With the optimized conditions in hand, we probed the scope of the reaction with respect to both the styrene and the benzhydrol components. As evidenced by the results shown in Table 2, both electron-rich and moderately electro-poor styrenes reacted smoothly with benzhydrol, producing the substitution products in generally good to excellent yields. Styrene with methyl or phenyl at the α-position of the double bond was afforded with the aid of catalyst 1b in moderate or excellent yield. The scope of the reaction with respect to the benzhydrol component was then investigated and was found to be good also. Several benzhydrol derivatives with methyl, methoxy, fluoro or chloro groups readily participated in the reaction with styrene, affording the corresponding tri-substituted ethylene in good to excellent yields.

Table 2 Substrate scope of 1b-catalyzed nucleophilic substitution of benzhydrols with styrenes. Entry Benzhydrol Styrene Time (h) Product Yield (%) 1 2a 3a 5 4a 92 2 2a 3b 5 4b 70 3 2a 3c 5 4c 50 4 2a 3d 5 4d >99 5 2a 3e 7 4e 74 6 2a 3f 1.5 4f >99 7 2a 3g 5 4g 75 8 2a 3h 1 4h >99 9 2b 3a 15 4i 86 10 2c 3a 9 4j 92 11 2d 3a 5 4k 60 12 2e 3a 5 4l 90 13 2f 3a 5 4m 98 Reaction conditions: 2a 0.5 mmol, 3a 0.6 mmol, 1b 2 mol%, DBE 1 mL, 80 °C.

Table 2 Substrate scope of 1b-catalyzed nucleophilic substitution of benzhydrols with styrenes.

Tertiary benzyl alcohols can undergo a dehydration reaction under acidic conditions to form styrene-type products ^[49]. Based on this mechanism, Han et al. ^[50] recently developed a tandem dehydration/nucleophilic substitution reaction involving two different alcohols. We found that this type of transformation can also be performed in our system with the aid of 1b. As shown in Scheme 1, two benzhydrol-type alcohols reacted smoothly under the above-mentioned optimum conditions, providing the expected multi-substituted olefin 4h in 96% yield, which is quite competitive compared with the result obtained for the condensation reaction of 1,1-diphenylethylene with benzhydrol over the same catalyst.

Scheme 1 . Tandem dehydration/nucleophilic substitution reaction of two different alcohols over catalyst 1b.

Encouraged by the promising results obtained in the above-mentioned reactions, we then applied the silica/policresulen composite catalyst to the oxa-Pictet-Spengler reaction of 2-phenylethanol and benzaldehyde. This type of condensation relies heavily on the use of acid catalysts ^{[51, 52, 53, 54, 55]}, and has been extensively used for the synthesis of isochromans. Although heterogeneous catalysts are preferable from the viewpoint of green chemistry, they are still rarely used in oxa-Pictet-Spengler reactions. Until now, there has been only one report on the use of a solid catalyst, a small pore size zeolite Ersorb-4 ^[56]. This is a clinoptilolite-type zeolite material with high Si content (Si:Al ratio 5:1). Unfortunately, this kind of zeolite is inaccessible to most labs working in organic chemistry. We found that, in the presence of a catalytic amount of 1b, 2-(3,4-dimethoxyphenyl)ethanol 5a reacted readily in 1,2-dichloroethane with benzaldehyde 6a at 80 °C, affording the expected isochroman 7a in 96% yield after 5 h (Table 3, entry 1). In this reaction, catalyst 1b could also be recycled at least three times without significant loss in activity. Increasing the reaction scale from 0.5 to 5.0 mmol did not significantly influence the synthetic efficiency. However, when SiO₂-supported sulfonic acid 1e was used as catalyst, the yield of 7a reached only 78% (entry 2). The reaction with TiO₂-supported benzenesulfonic acid catalyst 1f proceeded sluggishly, affording 7a in only 41% under identical conditions (entry 3).

Table 3 Oxa-Pictet-Spengler reaction catalyzed by solid acid catalysts. Entry Alcohol R3 Product Catalyst Time (h) Yield (%) 1 5a C6H5 7a 1b 5 96(95a, 93b) 2 5a C6H5 7a 1e 5 78 3 5a C6H5 7a 1f 5 41 4 5a 4-MeOC6H4 7b 1b 12 88 5 5a 2-MeOC6H4 7c 1b 5 90 6 5a 4-NO2C6H4 7d 1b 3 98 7 5a 2-NO2C6H4 7e 1b 3 98 8 5a 4-ClC6H4 7f 1b 5 97 9 5a 2-ClC6H4 7g 1b 5 96 10 5a 4-BrC6H4 7h 1b 3 98 11 5a 4-CNC6H4 7i 1b 5 >99 12 5a H 7j 1b 5 93 13 5a C6H5(CH2)2 7k 1b 16 98 14 5b C6H5 7l 1b 16 91 15 5c 4-CF3C6H5 7m 1b 36 75 Reaction conditions: alcohol 0.5 mmol, aldehyde 0.6 mmol, 1,2-dichloroethane 1.0 mL, 80 °C, 5 h. a 1b was reused in the third run. b reaction scale 5.0 mmol.

Table 3 Oxa-Pictet-Spengler reaction catalyzed by solid acid catalysts.

Subsequently, other aldehydes and alcohols were subjected to the conditions used for entry 1 (Table 3). First, we selected a set of aldehydes spanning a convenient spectrum of lipophicity and reactivity and used 5a as the only source of the required alcohol. When 1b was used as the catalyst, the results for all the aldehydes, which included both aromatic and aliphatic aldehydes, largely paralleled those obtained with benzaldehyde (entries 4-13). The reaction appeared to be insusceptible to the presence of a substituent in the C2 position of benzaldehyde; 2-methoxybenzaldehyde, 2-nitrobenzaldehyde, and 2-chlorobenzaldehyde reacted as well as their C4-substituted congeners (entries 4-9). Under the present conditions, a cyano group could be delivered into the final product without any structural damage (entry 11). 3-Methoxyphenethyl alcohol 5b and tryptophol 5c could also be used as counter substrates of the aldehyde component. Thus, catalyst 1b was proved again to be effective for these oxa-Pictet-Spengler reactions (entries 14 and 15).

The silica/policresulen composite catalystwas also used in the 3-alkenylation of indoles with 1,3-dicarbonyl compounds, which can produce valuable C3-substituted indoles. The use of some Lewis or Brønsted acids, such as FeCl₃ ^[57], I₂ ^[58], acidic ionic liquid ^[59], and an Au (III) salt catalyst ^[60] has been examined for this reaction. However, all these catalysts are homogeneous and cannot be recycled. Using 1c as the catalyst, 2-methylindole 8a reacted smoothly with methyl acetoacetate 9a in toluene, affording the corresponding C3-alkenylated indole 10a in 90% yield (Scheme 2). At the end of the reaction, catalyst 1c was easily recovered and reused at least two times without significant loss in activity. It should be noted that the same reaction over silica-supported sulfonic acid catalyst 1e failed to give a satisfactory result. Because only 55% yield was obtained with catalyst 1b under identical conditions, we deduced that the reaction might be significantly influenced by the catalyst loading. Fortunately, it is very easy to increase the loading of the policresulen component on the present catalyst. In contrast, an increase in acid loading with the conventional silica-supported sulfonic acid catalyst is not easy because it generally needs a sophisticated procedure and strict control of the preparation parameters. 2-Methylindole can also readily undergo an alkenylation with 2-phenylacetaldehyde over 1d (Scheme 2). A similar reaction using FeCl₃ as the catalyst has been reported by Yang et al. ^[61]. An excellent yield, 94%, was obtained after 11 h of stirring at room temperature. In this reaction, the recovered 1d also showed good recyclability. We have thus established the first heterogeneous catalytic system for 3-alkenylation of indoles with 1,3-dicarbonyl compounds and 2-phenylaldehyde.

Scheme 2 . C3-Alkenylation of 2-methylindoleover silica/policresulen composite catalyst.

Conventional methods for heterogenizing homogeneous molecular catalysts must covalently anchor the desired catalyst molecule on a solid support, and are thus rather complex and sometimes need sophisticated operations. The present silica/policresulen system achieves the ultimate goal for moving from a homogeneous catalyst towards heterogeneous conditions: easy separation of the catalyst and products while maintaining the same activity. The distinction in our case is that the catalytically active component, policresulen, could be easily immobilized on silica without forming chemical bonds. The unique structure of policresulen is the key factor that renders this protocol possible. The hydroxyl groups of policresulen likely enable the formation of hydrogen bonds that facilitate its immobilization, while the sulfonic acid ensures a good catalytic activity of the composite. This catalyst exhibited good performance in various organic reactions, including β-alkylation of styrenes with benzhydrol, oxa-Pictet-Spengler reactions, and C3-alkenylation of indoles with 1,3-dicarbonyl compounds. However, we wish to point out that the catalytic reaction might occur in a homogeneous aqueous phase formed at the surface of the silica, and that the use of an organic solvent that cannot dissolve the policresulen component allows the catalyst to remain stable on the solid component. Because this unique composite catalyst can be easily prepared without the use of any special instruments, researchers working in both catalytic reaction and organic synthesis may find it of use in the future.We are actively working on exploration of other reactions with this catalyst with the hope of clarifying its application scope and limitations.