Fractal features of pH-sensitive bimodal mesoporous silica-supported bipyridine-proline organocatalysts with core-shell structure and their application in asymmetric aldol reaction

2.1 Chemicals

Tetraethyl orthosilicate (TEOS, 98%) and 3-aminopropyltriethoxysilane (APTES, 98%) were purchased from Bailingwei Technology Beijing Co., Ltd. Cetyltrimethylammonium bromide (CTAB, 99% of purity) was purchased from Sinopharm Chemical Reagent Beijing Co., Ltd. Anhydrous ethanol (99%), toluene (99%), sodium sulfate (Na₂SO₄, 99.5%), and ammonium hydroxide (25%, NH₄OH) were obtained from Beijing chemical works. Polyacrylic acid (PAA, average molecular weights of 2000) was purchased from Shanghai Macklin Biochemical Co., Ltd. 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDCI, 99%) was purchased from Shanghai Saen Chemical Technology Co., Ltd. N-Hydroxysuccinimide (NHS, 99%) was purchased from Shanghai Yuanye Bio-Technology Co., Ltd. Ethanol (GC, 98%), n-Hexane (GC, 98%), ammonium nitrate (NH₄NO₃, 98%), and p-nitrobenzaldehyde (99.8%) were supplied by Shanghai Aladdin Biochemical Technology Co., Ltd. N,N-dimethylformamide (DMF, 99 %), petroleum ether (60–90 °C), ethyl acetate (EtOAc, 99.9%), dichloromethane (CH₂Cl₂, 99.9%), trifluoroacetic acid (TFA, 99.9%), sodium hydroxide, and cyclohexanone (99.9%) were provided by Tianjin Fuchen chemical reagents factory. 0.25 mm SDS silica gel coated glass plates (60F254) and silica gel (200–300 mesh) were purchased from Qingdao Haiyang Chemical Co., Ltd. All chemicals were obtained as reagent grade and used without further purification. All of the solvents and reagents were of A.R. grade, unless otherwise noted. Deionized water was used in all experiments.

2.2 Synthesis of pH-sensitive catalyst

The synthesis routes of pH-sensitive hybrid catalyst (P-Z@BMMs) with core–shell nanostructures are demonstrated, as shown in Scheme 1.

Close

Scheme 1

The synthesis routes of the pH-sensitive catalyst (P-Z@BMMs).

Export to PPT

Due to the weak acylating properties of carboxylic groups, EDCI/NHS coupling systems were used. Herein, EDCI was used to activate the terminal carboxyl groups on the PAA chains for conjugation with primary amines of template-contained BMMs (as shown in Scheme 1a), and NHS was used to improve the efficiency of EDCI-mediated coupling reactions by stabilization of the amine-reactive intermediate (as shown in Scheme 1b and c) (Wang et al., 2011). Particularly, the CTAB templates were removed by NH₄NO₃/EtOH solutions, the advantages of worm-like mesoporous structures of BMMs was highlighted (as shown in Scheme 1d). Then, the Z species were able to be loaded into PAA-coated BMMs (as shown in Scheme 1e). The synthesis procedure of BMMs with template (CTAB) was partially based on the method reported by Sun et al. (Sun et al., 2003). In details: 2.62 g CTAB was dissolved in 105 mL distilled water and stirred magnetically until it was completely dissolved, 8.0 mL of TEOS was then added drop-wise to the abovementioned solution. Subsequently, 3.5 mL of NH₄OH (25%) was quickly injected into the solution. After stirring for about 10 min at room temperature, the solution turned into white gels. The precipitate was filtrated and thoroughly washed with distilled water, and finally dried overnight at 80 °C under vacuum, the resulting product was named as t-BMMs. The sample (0.5012 g) of t-BMMs was weighted and transported into a 100 mL round-bottom flask, and then dehydrated in vacuum oven at 120 °C for 3 h before using. Subsequently, the dried sample was dispersed into anhydrous toluene solution (20 mL) and 0.5 mL of APTES was added drop-wise under magnetic stirring, then placed in an oil bath at 110 °C for 12 h. After centrifuged, the resultant precipitates were collected and washed with absolute ethanol, then dried at 60 °C for 6 h. The resulting product was named as t-BMMs-APS.

As shown in Scheme 1a and b, to activate the carboxyl groups of PAA, 20 mg of PAA was dissolved into 20 mL DMF solution and heated at 80 °C for 0.5 h under magnetic stirring. Then, the corresponding amount of EDCI and NHS were added sequentially according to the molar ratio of EDCI/NHS = 4:1, the mixture was subsequently stirred for 1 h. As shown in Scheme 1c, the dried sample of t-BMMs-APS (150 mg) was added to the above solution and continually stirred for 6 h. After that, the solid was separated from the liquid phase by centrifugation, washed alternately by distilled water and ethanol, and dried at 80 °C for 6 h. The resultant composite nanoparticles were named as P-t-BMMs.

As shown in Scheme 1d, the P-t-BMMs (100 mg) was dispersed in 10 mL of the prepared 10 mg/mL NH₄NO₃/EtOH solutions, then, the reaction mixture was heated at 80 °C under reflux for 12 h, and repeated for three times, the CTAB was removed through washing as described above. After that, the solid was centrifuged, washed alternately by distilled water and ethanol, then dried at 80 °C for 6 h. The obtained products were named as P-BMMs.

The active species (Z) was synthesized according to our previously reported literature procedure (Xu et al., 2020), the Z loading was carried out as follows. As shown in Scheme 1e, the P-BMMs (100 mg) was dispersed in 10 mL of CH₂Cl₂ solution containing 50 mg of Z, and heated at 42 °C for 8 h under magnetic stirring. After that, the mixture solution was centrifuged, the obtained precipitate was washed alternately by CH₂Cl₂ and ethanol, dried at 80 °C for 8 h. The resultant pH-sensitive catalysts were named as P-Z@BMMs.

2.3 pH-dependent catalytic behaviours

The asymmetric aldol reaction was performed under various pHs (pH = 1.53, 7.21, 8.02, 9.24, and 10.18, respectively). A typical procedure was as follows. Cyclohexanone (1.0 mmol, 104 μL) was stirred in various pHs solutions (0.5 mL) at room temperature for 5 min. The desired amounts of pH-sensitive catalyst (P-Z@BMMs, 5 mol% of p-nitrobenzaldehyde) was added to the above solutions, and stirred for 5 min, then the p-nitrobenzaldehyde (0.1 mmol) was added. The mixture solution was constantly stirred for 120 h. The reaction progress was monitored by thin-layer chromatography (TLC system: EtOAc/petroleum ether = 1: 2, v/v), using precoated TLC plates of silica gel 60F254. After that, the solid/liquid was separated by centrifugation, the resultant aqueous solution was extracted with ethyl acetate (3 × 10 mL). The combined organic extracts were dried over anhydrous Na₂SO₄, filtered and concentrated in vacuo. Finally, the crude product was purified by column chromatography on silica gel (eluent: EtOAc/petroleum ether) to afford a mixture of syn- and anti-aldol products, and the dr and ee values were determined by chiral HPLC analysis.

The turnover frequency (TOF) and turnover number (TON) can be described in detail as Eqs. (1) and (2).

Where n₁ denote the moles of the product molecules, n₂ denote the moles of the involved in catalysis Z, m₁ denote the weight of the used catalyst, and t denotes the whole catalytic reaction time, respectively.

A simplified scheme of the experimental and research methodology was given in Scheme S1 in the ESI section. Using PAA as a shell and BMMs as a core, PAA was grafted into the surfaces of the BMMs by a facile grafting strategy, the resultant P-BMMs could be used as bipyridine-proline (Z) support, and thereby a novel pH-responsive P-Z@BMMs catalyst was prepared. Hopefully, the catalytic performances for asymmetric aldol reaction presented a high efficiency by adjusting pH value of the reactive media.

2.4 Characterizations

The powder XRD patterns were recorded on a D6 Advance X-ray diffractometer (Beijing Pu Analysis General Instrument Co., Ltd, Beijing, China) using Cu-Kα radiation (λ = 0.154056 nm, 36 kV, 20 mA) at a scanning speed of 1° and 4° per min in the 1–10° and 5–50° of 2θ range for the experiments, respectively. The textural properties of related samples were determined from N₂ sorption isotherms measured at 77 K using JWGB JW-BK300 analyzer. Prior to each adsorption measurement the samples were outgassed under vacuum at 353 K for at least 5 h. The specific surface area was calculated using the Brunauer-Emmett-Teller (BET) method and the pore size distribution was calculated from the isotherm using the Barrett-Joyner-Halenda (BJH) model. Besides that, the total pore volume was determined from the N2 adsorbed amount at a relative vapor pressure (P/P₀) of approximately 0.99. The SEM images were performed on a Hitachi field-emission scanning electron microscope (S-4800) with an acceleration voltage of 15.0 kV. The morphology of the materials was acquired with a TEM microphotograph (JEOL, JEM-2100F, Japan). The DLS measurements was performed using Zetasizer (Malvern, Herrenberg, Germany). The TGA was carried out on a PerkinElmer Simultaneous Thermal Analyzer (STA-8000) instrument from 30 to 900 °C at a heating rate of 10 °C/min under the N₂ atmosphere with a flow rate of 20 mL/min. The content of sulfur was measured by an elemental analyzer (Vario MICRO, Germany). The FT-IR spectra were measured recorded on a Spectrum 100 spectrometer (PerkinElmer, Waltham, MA, USA) via the KBr tablet method, in which the spectral resolution was 4 cm⁻¹, and 32 scans were recorded for each spectrum. The dr and ee values were determined by HPLC analyses on an Agilent Technologies 1200 system equipped with a photodiode array detector, using Chiralpak AD-H column (25 cm × 0.46 cm) by Daicel Chemical Ind., Ltd (Detection: UV 254 nm, flow rate: 1.0 mL/min, injection: 5.5 μL, mobile phase: n-Hexane/Ethanol). Besides that, the chemical identity of the bipyridine-proline catalyst (Z) and aldol products has been confirmed by nuclear magnetic resonance (¹H- or ¹³C NMR) in our previous report (Xu et al., 2020).

In particular, the SAXS experiments were carried out using synchrotron radiation at the 1W2A X-ray beamline at Beijing Synchrotron Radiation Facility (BSRF). The wavelength of the incident X-ray was 0.154 nm. The sample-to-detector distance for SAXS was 1.59 m, calibrated with the diffraction ring of a standard sample. The scattering vector magnitude q ranged from 0.08 to 3.05 nm⁻¹ for the experiment reported in this paper. The intensity of scattering, I(q) was measured as a function of q, where q was considered as 4πsinθ/λ. The sample was loaded into a sample cell and sealed with scotch tape on a groove. The thickness of the sample cell was approximately 1 mm. The scattering image was collected with an exposure time of 5 min by the single-frame mode with a ‘multi-read’ of 2 times. The two-dimensional SAXS images were transferred to one-dimensional data by using the Fit2D software (https://www.esrf.eu/computing/scientific/FIT2D) (Hammersley, 2016), and further processed with the S program package (Li et al., 2001). In this study, a slit-collimated beam was used, for which the mass fractals (D_m) or surface fractals (D_s) were calculated according to the power law.

3.1 XRD patterns

The small (1–10°) and wide (5–50°) angle XRD patterns of all synthesized samples are depicted in Fig. 1. As can be seen in Fig. 1A, all samples presented the relatively broad and intense diffraction (1 0 0) peaks in the range 2θ = 1.5–2.5°, indicating a uniform and abundant mesoporous structure, being consistent with previously published literature (Sun et al., 2003).

Close

Fig. 1

Small (A) and wide (B) angle XRD patterns of (a) t-BMMs, (b) t-BMMs-APS, (c) P-t-BMMs, (d) P-BMMs, (e) P-Z@BMMs, (f) PAA, and (g) Z.

Export to PPT

As shown in Figure S1 in the Electronic Supplementary Information (ESI) section, the t-BMMs exhibited significant decrease in the peak intensity and left-shift in the peak position as compared to the same samples calcinated at 550 °C, suggesting an enhancement in the degree of ordering of the mesoporous structures during the template removal process. However, the calcinated process for removal of CTAB was accompanied by the shrinkages of the silica frameworks, resulting in a decrease of the crystal plane spacing (d) values (Sun et al., 2018; Tang et al., 2018; Xu et al., 2020; 2021; 2022a; Zhang et al., 2019).

As shown in Fig. 1A-a, -b, after surface-modification of the t-BMMs with the silane coupling agents, the intensity of the diffraction peak_(1 0 0) at around 2° was notably enhanced, indicating that the introduction of the amino groups onto the surface of Si-OH was helpful to increase the ordered mesopore regularity. Similarly, comparing the XRD patterns of t-BMMs-APS (as shown in Fig. 1A-b) and PAA-coated t-BMMs (Fig. 1A-c), the intensity of the characteristic peak increased slightly, probable reason was that t-BMMs-APS and P-t-BMMs were further prepared at high temperature, resulting in the improvements of the mesoporous order degree of t-BMMs. However, after further CTAB-removal via heated reflux extraction method, the intensity of diffractive peak_(1 0 0) of P-BMMs (Fig. 1A-d) was lower than that of P-t-BMMs (Fig. 1A-c), one of the main reasons was due to the decreases of the mesoporous order degree of BMMs. Besides that, the peak_(1 0 0) position (2θ) of P-BMMs (Fig. 1A-d) was shifted to a higher angle region, compared with that of P-t-BMMs (Fig. 1A-c). These observations demonstrate that CTAB was essentially removed by repeated washing, which provided the convenience for active species (Z) immobilization. As shown in Fig. 1A-e, the enhancement in peak_(1 0 0) intensity was observed after Z-immobilization, the peak_(1 0 0) position (2θ) of P-Z@BMMs (Fig. 1A-e) was shifted to a much lower angle region as compared to that of P-BMMs (Fig. 1A-d). These results indicate that the enlarged d space originates from the successful introduction of Z into the mesoporous channels of PAA-coated BMMs. Meanwhile, it could be evidenced by their pore size distributions and fractal features variations deriving from their N₂ sorption isotherms (as shown in Fig. 2) and SAXS patterns (Fig. 3) in the following section. Additionally, the XRD profiles of all samples (Fig. 1A) showed a typical characteristic diffraction pattern of BMMs, indicating the structural integrity of BMMs-based samples was not destroyed during PAA-grafting and Z-loading.

Close

Fig. 2

N₂ adsorption–desorption isotherms and corresponding pore size distributions (inset) of (a) t-BMMs, (b) t-BMMs-APS, (c) P-t-BMMs, (d) P-BMMs, and (e) P-Z@BMMs. The offset values were remarked in the Y-axis (left).

Export to PPT

Close

Fig. 3

(A) Shifted scattering curves (offset values in left Y-axis), and the PDDF profiles (B) of (a) t-BMMs, (b) t-BMMs-APS, (c) P-t-BMMs, (d) P-BMMs, and (e) P-Z@BMMs.

Export to PPT

Noticeably, the wide (5–50°) angle XRD patterns proved the amorphous structure of the prepared BMMs-based materials (as shown in Fig. 1B, from -a to -e), in which, no differences appeared in their characteristic peak around 20° before and after modification, particularly, no peaks assigned to PAA (Fig. 1B-f) and Z (Fig. 1B-g) were observed. In summary, these observations suggest the uniform coating of PAA over the amino-modified surfaces of BMMs and homogeneous Z-loading on the mesoporous channels of BMMs, thus allowing them to retain their catalytic activity. This forecast can be ensured by the catalytic behaviors discussed in the latter section.

3.2 N₂ Sorption isotherms

To investigate the pore structures after PAA-grafting or Z-loading, N₂ adsorption–desorption isotherms and the corresponding pore size distributions of all related samples were conducted in Fig. 2. Their collection of the textural parameters and porosity derived from N₂ sorption isotherms are summarized in Table S1. As shown in Fig. 2, the isotherms of all samples were followed the type-IV behaviors with type-H1 hysteresis loops over the whole range of relative pressures (except Fig. 2a), indicating the presences of the mesoporous structures. Notably, two inflections could be observed in all isotherms (as shown in Fig. 2 from b to e), except for t-BMMs (Fig. 2a) in the low-pressure region. The first hysteresis loops corresponded to a step where the N₂ adsorption within the small pores, besides, the second inflection belonged to the capillary condensation of N₂ within the larger mesopores, being consistent with previously published literature (Sun et al., 2018; Tang et al., 2018; Xu et al., 2020; 2021; 2022a; Zhang et al., 2019).

As shown in Fig. 2 (inset), the pore size distributions were calculated using the Barrett-Joyner-Halenda (BJH) method from the desorption branch of the isotherms. As can be seen in Fig. 2a (inset), the small mesopore of t-BMMs was not be observed, indicating that the structures of t-BMMs was compact with a small specific surface area (S_BET) and pore volume (P_vol), the corresponding values were around 33 m²/g, and 0.21 cm³/g, respectively. Meanwhile, the large mesopore of t-BMMs was around 36.15 nm (Table S1, entry 1), originating from the intra-particle aggregations. As a result of treating with APTES, the amino groups were modified on the surfaces of t-BMMs, the S_BET and P_vol values of t-BMMs-APS (as shown in Fig. 2b (inset) and Table S1, entry 2) were markedly increased to 642 m²/g, and 0.60 cm³/g, respectively, especially, the small mesopores appeared with a sharp peak located at 2.35 nm (Table S1, entry 2), indicating that the partial CTAB located in the mesopores was dissolved. As discussed later, this is compatible with the SAXS demonstrations (as shown in Fig. 3). Besides, after PAA-grafting on the surfaces of t-BMMs-APS, and in turn removal of the CTAB by repetitive organic solvent washing, their S_BET and P_vol, and small mesopore sizes also presented an increasing tendency (Table S1, entries 2–4), in which, the S_BET and P_vol, and small mesopore sizes of P-BMMs were around 913 m²/g, 0.83 cm³/g, and 2.60 nm (Table S1, entry 4), respectively. However, the S_BET and P_vol, and small mesopore sizes of P-Z@BMMs showed a decreasing trend, showing around 877 m²/g, 0.77 cm³/g, and 2.59 nm (Table S1, entry 5), respectively. Obviously, one of the main reasons is due to the Z-occupancies in the mesopore channels. Therefore, we speculate the successful linking between –NH₂ and –COOH groups, these findings can be further evidenced with FT-IR spectra in the following discussion.

3.3 Saxs analysis

The SAXS patterns of BMMs-based samples are displayed in Fig. 3A, their PDDF curves are presented in Fig. 3B, and their related parameters are summarized in Table 1. As shown in Fig. 3A, the slope of each curve (yellow straight lines) was obtained with a linear fit using the least square method in two distinctive linear regions (low-, or high-q region), and thus the following mass (or surface) fractal dimensions (D_m or D_s) deriving from calculations on the basis of above linear curves are generally determined by power-law decay (Hammersley et al., 2016; Xu et al., 2020; 2021; 2022a). In particular, their variations of the fractal features demonstrate the success of the performed covalent modifications.

As seen in Table 1, the slope values of all related samples were between −2 and −3 in the low-q region (0.09 < q < 0.16), their behaviors possessed the D_m features, corresponding their D_m values ranged between 2 and 3, indicating that the fractal structures of BMMs-based samples tended to be compact or loose. According to the reported literature (Hammersley et al., 2016; Xu et al., 2020; 2021; 2022a), the increase of D_m value implies that the fractal structure tends to be closer, otherwise, means that the structure is more open (non-compact), while, the fractal structure of D_m = 3 refers to dense objects.

As listed in Table 1 (entries 1 and 2), the D_m values slightly decreased after amino-modification, from 2.50 for pure t-BMMs (entry 1) to 2.29 for t-BMMs-APS (entry 2). This is due to inevitable partial dissolution of the unstable CTAB within t-BMMs-APS using toluene as the solvent, leading to the variations of the structural compactness. These results are also evident from the pore size distributions (as shown in Fig. 2a, and b) and the following demonstration of their TGA profiles (Fig. 6a, and b). As shown in Table 1 (entries 2 and 3), the fractal structure of P-t-BMMs (entry 3) was more compact than that of t-BMMs-APS (entry 2), due to larger D_m values of polymer-coated particles (2.39 for P-t-BMMs) than that of uncoated (2.29 for t-BMMs-APS). Therefore, these results indicate that the PAA macromolecules were potential to be successfully grafted on the surface of t-BMMs via covalent bonding. After further removal of CTAB from P-t-BMMs by extraction with NH₄NO₃/EtOH solvents, the D_m value of the obtained P-BMMs (entries 4) was about 2.32, smaller than that of P-t-BMMs (D_m = 2.39, entries 3). Obviously, the aforementioned behaviors almost conform to the change rule of fractal features. Moreover, as shown in Table 1 (entries 4 and 5), the D_m value slightly increased afterwards (from 2.32 (P-BMMs, entry 4) to 2.36 (P-Z@BMMs, entry 5)) following loading of Z, suggesting that their greater densifications when the Z were successfully loaded into the mesopores channels of the P-BMMs.

As a whole (as shown in Table 1, from entry 1 to entry 5), the D_m values of above samples gradually decreased from 2.50 (entry 1) to 2.36 (entry 5), which could be ascribed to successful grafting of PAA onto the surfaces of t-BMMs-APS, and the removal of CTAB within the mesoporous structures of p-t-BMMs, and the Z-loading inside the mesopore channels of P-BMMs. Besides, their behaviors possessed the fractal features (D_m or D_s) in the high-q region (0.38 < q < 0.65), the situation with the fractal features appears to be more complicated, however, no obvious correlation of the above features is observed here, further consideration are therefore not considered.

Moreover, their detailed morphological information regarding the particle size, shape, and interfacial structures before and after modification could be derived from the PDDF profiles (Hammersley et al., 2016). As shown in Fig. 3B, judging from the shapes of all P(r) - r curves, as expected, all curves exhibited almost a symmetrical shape-form, the results suggested that the morphology of BMMs-based nanoparticles were approximately spherical, and changed unmarkedly after modification and loading, meanwhile, the following SEM images (as shown in Fig. 4) were also performed to support the PDDF analysis. On the other hand, the intersections of these above curves and X-axis (as show in Fig. 3B, from -a to -e) denotes the maximum diameter (D_max) of the nanoparticles (Hammersley et al., 2016), and the collections of the possible maximum particle sizes (D_max) of the BMMs-based samples are listed in Table 1. As can be seen, the D_max of t-BMMs (Table 1, entry 1) was nearly the same as that of t-BMMs-APS (Table 1, entry 2), corresponding values were around 53 and 52 nm, respectively. These results indicate that the introduction of an amino group appears to have little effect on their particle size. However, the D_max became greater by PAA-grafting onto the surface of the t-BMMs, corresponding values increased from 52 nm for t-BMMs-APS (Table 1, entry 2) to 76 nm for P-t-BMMs (Table 1, entry 3), which were caused by the formation of amide bonds between the amino group of t-BMMs and the carboxylic group of PAA. These results implied that the PAA was successfully coated on the surfaces of t-BMMs, in well agreement with the demonstrations deriving from the fractal features. Furthermore, their D_max values decreased from 76 nm for P-t-BMMs (Table 1, entry 3) to 61 nm for P-BMMs (Table 1, entry 4) after removal of the template (CTAB), which should be attributed to the shrinkages of the silica frameworks. Comparably, further increases in particle size was observed after Z-loading, corresponding D_max values increased from 61 nm for P-BMMs (Table 1, entry 4) to 66 nm for P-Z@BMMs (Table 2, entry 5). Obviously, the enlarged particle size of P-Z@BMMs further indicated the Z-entering mesoporous channels of P-BMMs, being in good accordance with the demonstrations of the XRD patterns (as shown in Fig. 1). Besides, the sizes of P-Z@BMMs nanoparticles were also consistent with the estimated sizes by the following demonstration of the SEM images (as shown in Fig. 4).

Close

Fig. 4

(a) SEM and (b) TEM images of P-Z@BMMs.

Export to PPT

Table 2 Aldol reaction results between p-nitrobenzaldehyde with cyclohexanone in the presence of P-Z@BMMs under different pH conditions.^[a]

Close

Export to PPT

Entry	pH value	Yield[b] (%)	dr[c] (%, anti/syn)	ee[c] (%, anti)	TON	TOF[ (g⋅h)-1]
1	1.53	–	–	–	–	–
2	7.21	81.38	53:47	3.02	16.28	57.35
3	8.02	80.37	51:49	0.32	16.07	56.64
4	9.24	80.49	54:46	0.36	16.10	56.72
5	10.18	85.68	53:47	0.12	17.14	60.38

[a] Reaction conditions: p-nitrobenzaldehyde (0.1 mmol), cyclohexanone (1.0 mmol, 104.0 μL), catalysts (5 mol%), H₂O (0.5 mL), r.t., 120 h.

[b] Isolated yield after separation by silica gel.

[c] The diastereomeric and enantiomeric ratios were determined by chiral HPLC analysis.

In overall, the positive correlation between the fractal characteristics (D_m) of the BMMs-based samples and their PDDF patterns were prominent, especially after PAA-grafting and subsequent Z-loading. As seen in Figure S2 of the ESI section, the scattering properties (including negative or positive deviation, as well as the average thickness of their interfacial layer) of the BMMs-based materials can be roughly estimated on the basis of Porod’s law, their related structural parameters are collected in Table 1. As shown in Figure S2-A, the scattering curve (ln[q⁴I(q)] ∼ q²) of t-BMMs conformed well to Porod’s law with a negative deviation, corresponding the thickness of an interfacial layer between the mesopores and silica matrix was around 1.52 nm (Table 1, entry 1). Herein, these results were a big difference from that of pure BMMs in the absence of CTAB according to the previous reported (Li et al., 2001; Xu et al., 2020; 2021; 2022a). As shown in Figure S2-B, the scattering curve of t-BMMs-APS also showed a negative deviation, corresponding values presented a decreased tendency from 1.52 nm (Table 1, entry 1) to 1.48 nm (Table 1, entry 2), which may be attributed to the decrease of an interface layer thickness due to the part removal of CTAB, being consistent with the fractal features and related PDDF analysis (as shown in Fig. 3). However, the Porod scattering of P-t-BMMs transformed from negative to positive deviation (as shown in Figure S2, from -B to -C), when achievement of the formation of the amide bonds on the surface of t-BMMs. In particular, we could speculate that the existences of the PAA layers caused significantly the additional scatterings, indicating the successful PAA-coating on the surfaces of t-BMMs. Subsequently, with the complete removal of CTAB from P-t-BMMs, the positive deviation would again be back to the negative deviation (as shown in Figure S2, from -C to -D), corresponding the thickness of an interfacial layer was around 0.97 nm (Table 1, entry 4). Comparably, its interface layer thickness was considerably smaller than that of t-BMMs-APS (1.48 nm, Table 1, entry 2), in other words, the average thickness of the grafted-PAA shell was estimated to be around 0.97 nm. Nevertheless, the scattering curve of P-Z@BMMs (Figure S2-E) conformed well to the results of the positive deviation from Porod’s law after Z-loading with low amount, as described in Table 1 (entry 5). Thus, we speculate that the predominant reasons should be attributed to the additional scatterings deriving from Z-loading.

In conclusion, the comprehensive demonstrations on the basis of the XRD patterns, the N₂ sorption isotherms, and the SAXS analysis (including fractal evolution, PDDF profiles, and Porod plots) indicate the successfully uniformly PAA-coating on the outer-surface of the BMMs nanoparticles via covalent bonding, the Z species were also successfully loaded into the inner-mesoporous surface of the BMMs.

In order to fully and more accurately verify the above results, the following demonstrations of the SEM/TEM images, the dynamic light scattering (DLS) measurements, the elemental analysis, the FT-IR spectra, and the thermogravimetric (TG) profiles were robustly performed.

3.4 Sem/TEM images

The SEM and TEM images of P-Z@BMMs nanoparticles are shown in Fig. 4, the resultant nanoparticles were observed to have a nearly spherical morphology in size of around 70 nm with disordered and uniform distribution, similar to what we have previously reported results for parent BMMs or synthesized Z@BMMs (Sun et al., 2018; Tang et al., 2018; Xu et al., 2020; 2021; 2022a; Zhang et al., 2019). Besides, the morphological features of the pH-sensitive catalysts are also well demonstrated on the basis of the N₂ sorption isotherms (as shown in Fig. 2e and Table 1 (entry 5) and PDDF profiles (Fig. 3B and Table 1 (entry 5)). However, the shell thickness of P-BMMs could not be visualized by the TEM image (Fig. 4b), because of too thin (probable<1 nm) according to the comparison results of entry 3 and 4 in Table 1. In present work, the thickness of interfacial layer (PAA shell, around 0.97 nm) was estimated by Porod plot.

3.5 Dls measurements

The DLS measurements of related samples were also performed to support the SEM/TEM images and PDDF analysis, corresponding their apparent hydrodynamic diameter and polydispersity index (PDI) are summarized in Table S2. The mean particle sizes of samples obtained from the Gaussian fit of the histograms as shown in Fig. 5.

Close

Fig. 5

The hydrodynamic diameter distribution of (a) t-BMMs, (b) t-BMMs-APS, (c) P-t-BMMs, (d) P-BMMs, and (e) P-Z@BMMs.

Export to PPT

As can be seen in Table S2 (entries 1 and 2), the mean particle size of t-BMMs and t-BMMs-APS were around 531 and 397 nm, respectively. However, the size of P-t-BMMs increased greatly after PAA-grafting, up to a maximum of 1378 nm (Table S2, entry 3), indicating that the change rule of the hydrodynamic diameter was almost in line with that of PDDF analysis. Likewise, the apparent size measured by DLS was significantly larger than that measured by SEM/TEM (as shown in Fig. 4) due to hydration-induced expand of particles after soaking, originating from the existence of the PAA layers (Zayed et al., 2019) particularly for the swollen PAA networks at pH values higher than pKa (4.25). A detailed discussion regarding this phenomenon was explored in the following sensitivity analysis (as shown in Table 2). Besides, the mean particle sizes of P-BMMs (Table S2 (entry 4)) and P-Z@BMMs (Table S2 (entry 5)) were maintained at a relatively high level, further proving our reliable results.

As shown in Table S2, the relative homogeneity of the BMMs-based nanoparticles can be roughly estimated by the PDI values, where the smaller the PDI values, suggesting a homogeneous and narrow particle size distribution. As a whole (from entries 1 to 5 in Table S2), their PDI values exhibited a trend of obvious decreasing during the surface modification, in which, PDI value of P-Z@BMMs was small of around 0.48, implying the homogeneous distributions of the particle and less tendency to agglomerate. Therefore, it can be concluded that the PAA-coated samples are considered as tend to mono-disperse, and the reliability of the conclusions inferred from the SEM/TEM images (as shown in Fig. 4) and PDDF profiles (as shown in Fig. 3B and Table 1) are also verified.

3.6 Tga profiles

The TG curves of all related samples are presented in Fig. 6. As can be seen, the small weight loss at a temperature range of 30–200 °C was connected to the evaporation of physically or chemically adsorbed water and/or organic solvent on the surface of mesopores, whereas the gradual weight loss profiles of the BMMs-based samples from 200 to 900 °C were associated with the decompositions of the organic functional groups (amino, carboxyl, or amide), which is well agreement with the previous reports (Sun et al., 2018; Tang et al., 2018; Xu et al., 2020; 2021; 2022a; Zhang et al., 2019).

Additionally, the TG profile of t-BMMs exhibited significantly weight loss of around 46.03 wt% (as shown in Fig. 6a), because of the CTAB decompositions. Herein, it was found that the weight loss of CTAB continues to decrease prior to the extracted with mixed solvents (Fig. 6, from a to c), corresponding the weight loss values were around 32.12 wt% for t-BMMs-APS (Fig. 6b), and 22.73 wt% for P-t-BMMs (Fig. 6c), respectively, which may be attributed to the part removal of CTAB. These observations were consistent with the results of the XRD patterns (as shown in Fig. 1), N₂ sorption isotherms (Fig. 2), and SAXS analysis (Fig. 3). However, the weight loss values increased from 20.06 wt% for P-BMMs (Fig. 6d) to 22.33 wt% for P-Z@BMMs (Fig. 6e). Indeed, these substantial alterations give rise to the formations of the interfacial thickness, the fractal evolution, as well as the variability in the pore sizes and porosity of nanoparticles. Consequently, these results indicate that the CTAB are almost completely removed via the solvents treatment, besides the successful Z-loading on the surfaces of mesopores.

3.7 Elemental analysis and FT-IR spectra

Furthermore, the elemental analysis results of N, C, and H in P-Z@BMMs are summarized in Table S3. On the basis of the characteristic element (S) content, the Z-loaded amounts could be roughly estimated to around 2.97 wt% for P-Z@BMMs, in almost agreement with the results of TGA analysis (around 2.27 wt%).

As shown in Fig. 7, the FT-IR spectra were carried out to confirm the complete removal of CTAB and the successful PAA-coating on the surfaces of t-BMMs. As can be seen in Fig. 7a, the t-BMMs presented two intense peaks between 3000 and 2800 cm⁻¹, which could be attributed to the asymmetric (2920 cm⁻¹) and symmetric (2850 cm⁻¹) stretching vibrations of the methylene chains, respectively (Quan et al., 2015). In particular, these intense peaks were also observed in prepared t-BMMs-APS (Fig. 7b) and P-t-BMMs (Fig. 7c), whereas almost absent in the extracted P-BMMs (Fig. 7d) or P-Z@BMMs (Fig. 7e), especially for the peak at 2920 cm⁻¹, as supported by the almost complete removal of CTAB from the fabricated composite through extraction.

Close

Fig. 6

TG curves of (a) t-BMMs, (b) t-BMMs-APS, (c) P-t-BMMs, (d) P-BMMs, and (e) P-Z@BMMs.

Export to PPT

Close

Fig. 7

FT-IR spectra of (a) t-BMMs, (b) t-BMMs-APS, (c) P-t-BMMs, (d) P-BMMs, (e) P-Z@BMMs, (f) PAA, and (g) Z.

Export to PPT

Besides, for P-t-BMMs, P-BMMs, and P-Z@BMMs (Fig. 7, from c to e), the additional band appeared at 1650 cm⁻¹ was associated with the stretching vibrations of amide (OC-NH) bonds, in addition, for P-t-BMMs (Fig. 7c), the intensity of the C–H bonds in carboxyl groups (–COOH) bending vibration at 1450 cm⁻¹ obvious decreased compared to PAA (Fig. 7f). Therefore, the FT-IR spectra strongly support our conclusion that an efficient the covalent attachment of PAA on the surfaces of BMMs via formation of OC-NH bonds between –COOH groups of PAA and –NH₂ groups on the t-BMMs (as shown in Scheme 1). Moreover, the other bands at 1068, 950, and 800 cm⁻¹ are characteristics of bending vibrations of Si-O bonds, and stretching vibrations of Si-OH and Si-O-Si bonds, respectively (Sun et al., 2018; Tang et al., 2018; Xu et al., 2020; 2021; 2022a; Zhang et al., 2019;). Especially for the active species (Z), Fig. 7g indicated that the main bands located at 1677, and 1500 cm⁻¹ could be attributed to the C = O and C = N stretching vibrations in the amide group, respectively, which predominantly originate from the pyridine rings (Xu et al., 2020; 2021; 2022a). The characteristic peak of C = O arising from PAA at around 1652 cm⁻¹ was unobvious due to the strong nearby vibrational absorption peak in the region. Notably, the peak at around 1384 cm⁻¹ in FT-IR spectra represented the nitrate anions stretching vibration of NH₄NO₃ (as shown in Fig. 7d, and e). However, the above bands could not visibly judge the differences between the Z and P-Z@BMMs. Therefore, the successful Z-loading was further elaborated in the discussions below.

3.8 Catalytic performances for asymmetric aldol reaction

In order to evaluate the pH-sensitivity of the grafted-PAA shells and the loaded-Z catalytic properties, the prepared P-Z@BMMs were used as the catalysts, their catalytic performances for asymmetric aldol reaction are explored under various pH (1.53, 7.21, 8.02, 9.24, and 10.18) media.

Meanwhile, reaction progress was monitored by TLC analysis (as shown in Scheme S2 in the ESI section). As shown in Scheme S2-b, the bands of aldol products at acid condition (pH = 1.53) were almost not reflected on TLC plates, similar to that of control group (only reactant: p-nitrobenzaldehyde, as shown in Scheme S2-a). However, as shown in Scheme S2-c, and -d, two bands of aldol products at neutral condition (pH = 7.21) and at alkaline condition (pH = 8.02) were observed on TLC plates, while the band of the reactant practically disappeared. Obviously, these observations further manifested that the PAA shell prevented the release of active species (Z) from P@BMMs matrix under acidic conditions, but promoted its release under neutral and alkaline conditions.

As shown in Table 2 (entry 1), a significant percentage of the unreacted substrates was observed despite in the presence of sufficient amounts (5 mol%) of P-Z@BMMs under the acid condition (pH = 1.53), and almost no reaction was occurred even after 120 h at room temperature. However, the reaction yield reached 80 % (Table 2 (entry 2–5)) under neutral or alkaline conditions (pH = 7.21, 8.02, 9.24 and 10.18), indicating that P-Z@BMMs had a high catalytic activity due to the Z-releasing from P-Z@BMMs. In particular, as mentioned above, the pKa of PAA is about 4.25. When P-Z@BMMs is in a low pH solution, the appearances of hydrogen bond leads to the shrinkage of the PAA shells with the dense networks, and the Z loaded in P-Z@BMMs is difficult to be released, showing a low catalytic activity. As increasing of pH values, the carboxyl groups in the PAA networks are ionized, causing the swelling of the PAA shell due to electrostatic repulsion. As a result, the PAA network channels were opened, and Z in P-Z@BMMs was easily released, showing a high catalytic activity. However, according to our previously published report, when Z or Z@BMMs were used as homogeneous or heterogeneous catalysts under the same acidic conditions, the desired product was obtained in 91% yield (Xu et al., 2020; 2022b). Therefore, as expected, the active species (Z) cannot be released from P-Z@BMMs for exerting catalytic effects due to their shrinkage of the PAA shell.

The TON and TOF values under the catalysis of P-Z@BMMs were provided in Table 2. In particular, the maximum TON and TOF values of the catalytic aldol reaction using p-nitrobenzaldehyde as substrate were calculated to be 17.14 and 60.38 [(g⋅h)^-1] under the alkaline conditions (pH = 10.18).

Consequently, the resulting data strongly support our conclusions that the active species (Z) in P-Z@BMMs was hardly to be released under the acid solutions. Interestingly, this process further results in the controllable release of the active species (Z) by altering their acid-base systems. Nevertheless, it was found that the obtained yields, dr, and ee values were almost unchanged in a limited pH region (from 7 to 10, as shown in Table 2 (entries 2–5)), these results indicate that the PAA shells of the P-Z@BMMs catalyst were insensitive to the above pH region. Overall, the controlled release of the active species (Z) from its protective P-BMMs coating can be achieved by simply adjusting pH values.

The possible catalytic mechanism between p-nitrobenzaldehyde (as an aldol acceptor) and cyclohexanone (as an aldol donor) was demonstrated in our previous reports (Xu et al., 2020; 2021).