Scientists engaged in heterogeneous catalysis often cite enzymes as their model catalysts. Enzymes can efficiently catalyze multistep processes that give various types of biomolecules.1 Remarkably, many enzymes combine two antagonistic catalytic functions, such as acid and base functions.2 Attracted by this challenge, several groups synthesized homogeneous catalysts that can successfully combine chemically hostile functions.3 However, the difficulty lies in controlling the separation between these groups and simultaneously working at practical catalyst concentrations. In principle, both of these problems can be solved using heterogeneous catalysis. But synthesizing solid catalysts that combine hostile functions is no easy task. Not surprisingly, examples of solids with both acidic and basic functions are limited. The reported examples combine acids such as sulfonic acids, silanols, ureas, and thiols with amines.4 However, problems such as low catalytic activity because of weak acidity, complicated preparation, and lack of a continuous range of acidic and basic catalytic sites often significantly limit their application in organic reactions.
Here, we present a simple and straightforward route to such bifunctional solids. We combine amine base functions and heteropolyacid functions on periodic mesoporous silica, obtaining an efficient and robust bifunctional acid–base solid catalyst. This new material enables one-pot cooperative catalysis. Moreover, the synthesis permits easy tailoring of the acid/base properties by controlling the number and surface concentration of the acid and base sites, respectively. We demonstrate the catalytic efficiency and robustness of this new system in two cascade reactions: a tandem deprotection–Henry reaction, and a tandem deprotection–aldol reaction. Excellent yields and selectivities are obtained in both cases.
Figure 1 shows the synthesis of the catalyst. We started by making a mesoporous silica support S (Figure 1, top; pore diameter of 4.9 nm and surface area of 870 m2 g−1, see the Experimental Section). The base functions were then added by grafting 3-aminopropyl groups, giving the base catalyst SB (Figure 1, middle). This material was then immersed in a methanolic solution of phosphotungstic acid (H3PW12O40). The 3-ammoniumpropyl groups immobilized the acid polyanions, creating the acid/base catalyst SAB (Figure 1, bottom). By controlling the ratio of immobilized heteropolyacids and aminopropyl tethers, we succeeded in reacting only part of the amino groups, creating bifunctional catalytic sites inside the silica mesopores.
The catalysts were characterized before and after each functionalization step. Figure 2 shows the N2 adsorption isotherms of S and SAB. Both materials are mesoporous, with average pore diameters of 4.9 and 3.4 nm, respectively. The isotherms show that grafting the base and immobilizing the acid do not change the structure. SAB adsorbs less nitrogen than S, because a fraction of the pores is already occupied by the aminopropyl and phosphotungstic groups. However, the pores are still wide enough to allow easy diffusion of reactant and product molecules. The structure retention after immobilization was also confirmed by X-ray diffraction (see the Supporting Information).
13C cross-polarization (CP) magic angle spinning (MAS) NMR spectra showed peaks at 11, 22, and 43 ppm (Figure 3). These peaks are typical for the SiCH2CH2CH2NH2 chain, indicating the successful grafting of the aminopropyl groups on silica support S. The presence of T3 and T2 functionalities in the 29Si CP-MAS NMR spectra (Figure 4) confirmed a strong covalent linkage between the organic groups and the silica surface (see also Figure S3 in the Supporting Information). A comparison of the W-L3 edge region in the X-ray absorption near-edge structure (XANES) spectrum of SAB with that of bulk phosphotungstic acid showed no major shifts in position and amplitude. This confirms the absence of geometrical or electronic changes in the immobilized groups (Figure 5). Moreover, 31P NMR spectroscopy also showed that the Keggin structure of the phosphotungstic acid was maintained after immobilization (singlet at −15.5 ppm, see the Supporting Information).
We then investigated the bifunctionality of catalyst SAB in the one-pot tandem conversion of dimethoxymethylbenzene (benzaldehydedimethylacetal) 1 to trans-1-nitro-2-phenylethylene 3 (Scheme 1; the functionality that catalyzes each step is shown in blue). The first step of this reaction is the acid-catalyzed deacetalization of 1 to benzaldehyde 2. In the second step, which is base-catalyzed, the benzaldehyde reacts with nitromethane, giving the nitro product 3. Table 1 shows the results. High conversions and yields were observed in the presence of the SAB catalyst (entry 1 in Table 1). When SB was used instead of SAB, no reaction occurred with 1 (entry 2 in Table 1). However, control reactions starting from 2 confirmed that SB is an excellent base catalyst for the second step, converting benzaldehyde with nitromethane to 3 in near-quantitative yield (entry 4 in Table 1). These results show that both the amino groups and the acid polyanions in SAB retain their reactivity. The immobilized polyanions, which are still acidic, catalyze the deacetalization of 1 to 2 (Figure 6). This is followed by the Henry reaction of 2 with nitromethane at the free amino groups. The key requirement for the tandem deacetalization–nitroaldol reaction is the ratio between amino groups and polyanions. If all the amino groups are used for polyanion immobilization, the reaction stops at the first step and benzaldehyde is the sole product (entry 3, Table 1). In our bifunctional catalyst, only part of the amine groups are used for immobilizing polyacids; the remaining half are free. We also confirmed that immobilization does not affect the intrinsic activity of polyacid groups by comparing the catalytic activities of SAB and bulk phosphotungstic acid in ethyl acetate hydrolysis reaction. The catalytic activities, calculated based on the amount of acidic protons (measured by pyridine adsorption), were similar (102 and 95 mmol mol−1 min−1).
|Entry||Catalyst||Substrate||Conversion [%]||Yield GC/isolated [%]|
Control reactions confirmed that S alone did not catalyze the reaction of 1 under the same conditions (entry 5, Table 1). Moreover, the SAB catalyst was easily recovered by filtration and reused several times without the loss of the catalytic activity and selectivity (entry 6, Table 1; the reactions stopped as soon as the catalyst was filtered off the hot reacting mixture). This rules out the possibility of leaching of active species into the reaction solution.
Similarly, we evaluated the SAB catalyst in the tandem reaction of 1 through 2 to benzylidene malononitrile 4 (Scheme 2). Following the first deacetalization step, 2 undergoes a base-catalyzed aldol condensation with malononitrile. SAB efficiently catalyzed this tandem deacetalization–aldol reaction as well, which yielded 91 % of 4 and shows the versatility of our catalyst (Table 2). Control reactions in the presence of homogeneous p-toluenesulfonic acid gave no product because of the neutralization of the amino groups of the catalyst.
|Entry||Catalyst||Substrate||Conversion [%]||Yield GC/isolated [%]|
In conclusion, we showed that phosphotungstic acid immobilized on amine-grafted mesoporous silica is an efficient bifunctional catalyst. The catalyst combines two antagonistic active functions on one solid catalyst, allowing acid–base tandem conversions in a single pot. We can prepare this catalyst as well as tune it easily. The catalyst can be made predominantly basic, or predominantly acidic, or equally acidic and basic by changing the ratio of polyacid and amine groups. Similar types of catalysts can be synthesized using other heteropolyacids, which further widens the scope of these materials.