The asymmetric aldol reaction is one of the most important C-C bond-forming reactions and creates the optically active β-hydroxy carbonyl structural unit found in many natural products and drugs [1, 2, 3, 4]. Most asymmetric aldol reactions involve the directed aldol reaction in the presence of acidic promoters, including organic and inorganic acids, which are used to activate aldehydes electrophilically. Unfortunately, organic acids such as TfOH and F3CCO2H have high solubility in the organic phase, which results in impurities in the catalytic products owing to difficult separation. Furthermore, liquid inorganic acids are highly corrosive to industrial equipment. Therefore, the choice of the solid acid in the aldol reaction is a crucial decision [5, 6, 7, 8, 9, 10].
Magnetic nanoparticles (MNPs) with large specific surface areas and good textural properties are currently an area of extensive research, particularly for applications in catalysis and adsorption processes owing to their unique superparamagnetism, low toxicity, simple preparation, low cost, and ease of recovery [11, 12, 13]. To combine the advantages of MNPs and solid acids, magnetic solid acids have been developed and widely applied in catalytic nitration [14], the Hantzsch reaction [15], esterification [16, 17], hydrolysis reactions [18, 19], the acetal reaction of benzaldehyde with ethylene glycol [20], and coupling reactions of aryl bromides with heterocycles [21]. However, the application of magnetic solid acids in asymmetric catalytic reactions is seldom reported.
In this paper, the nanomagnetic solid acids SO42−/Zr(OH)4- Fe3O4 [22, 23, 24] were applied to the asymmetric aldol reaction of cyclohexanone with various substituted benzaldehydes for the first time. Good to excellent catalytic performance for the benzaldehydes with strong electron-withdrawing groups (83%- 100% yield, 86.0%-95.6% ee anti, and anti/syn = 88-96/12-4) was satisfactorily achieved. These magnetic solid acids can be easily recovered in quantitative yield using an external magnet and possessed good tolerance in catalytic cycles with no significant loss of catalytic performance.
The suspension of nano-Fe3O4 (2.32 g, 10.0 mmol, Aladdin, 99.5%) in distilled water (20 mL) was well stirred at 65 °C, and aqueous ZrOCl2·8H2O solution (20 mL, Aladdin, 99.9%) with molar ratios of ZrOCl2·8H2O:Fe3O4 = 0.2, 0.5, 1.0, 2.0, or 5.0 was added dropwise. The pH of the reaction medium was adjusted to 13 by ammonia spirit and stirred for another 4 h. The supernatant was decanted by magnetic separation using an external magnet, and the obtained magnetic solids Zr(OH)4/Fe3O4 were washed with distilled water until the solution was neutral and no Cl− ions were present and dried at 110 °C in vacuo. After three nitrogen replacement, grinding, immersing in 0.5 mol/L sulfuric acid for 3 h, and separating using an external magnet, the samples were dried at 110 °C to obtain nanomagnetic solid acids, which were designated as SZF1-5 for ZrOCl2·8H2O:Fe3O4 ratios of 0.2, 0.5, 1.0, 2.0, and 5.0, respectively [25, 26]. SZF3 was calcined in a muffle furnace at 300 or 650 °C for 3 h to give the nanomagnetic solid acids SZF3-300 and SZF3-650, respectively.
Magnetic hysteresis loops were measured on a vibrating sample magnetometer (VSM, HH-15, China) using Ni as the standard substance. The powder was strongly pressed and fixed in a small cylindrical plastic box for the magnetization measurements. Thermogravimetry-differential thermal analysis (TG-DSC) spectra of the as-synthesized samples were measured on an SBTQ600 thermal analyzer (TA, USA) with a heating rate of 20 °C/min from 40 to 800 °C under N2 (100 mL/min). The total acidity of the samples was determined by temperature-programmed desorption (TPD) of NH3. The measurements were carried out by an automatic TPD apparatus (PCA-1200, China). The sample was kept at 50 °C for 60 min in a flow of NH3. The amount of desorbed NH3 was determined after heating the sample up to 700 °C with a heating rate of 10 °C/min. N2 adsorption-desorption analysis was performed at −196 °C on an Autosorb-1 apparatus (Quantachrome, USA). All samples were degassed at 130 °C for 12 h before measurements, and the surface area and pore size distribution were calculated by the BET and BJH models, respectively. After being well-dispersed in water (10 mg sample in 10 mL of H2O) for 10 min under ultrasonic irradiation, sputtered over copper wire, and evaporated under infrared radiation for 10 min, the as-synthesized samples were observed by transmission electron microscopy (TEM) using a JEM 2100 transmission electron microscope under an accelerating rate voltage of 200 kV to show their surface morphologies (JEOL, Japan).
Accurately weighed nanomagnetic solid acid (100.0 mg) was soaked in 6 mol/L HCl (50 mL) under vigorous stirring for 3 h and filtered. Saturated BaCl2 was added to the filtrate and BaSO4 precipitated out. After BaSO4 was filtered, dried under reduced pressure, and accurately weighed, the SO42− content was calculated.
9-Amino-9-deoxy-epi-cinchonidine (epi-CDNH2) was prepared according to a literature procedure [27]. In a vial (25 mL), the mixture of nanomagnetic solid acid SZF3 (35.0 mg), epi-CDNH2 (11.0 mg, 0.0375 mmol), cyclohexanone (0.98 g, 10.0 mmol, Adamas-beta, 99.5%), and deionized water (2.0 mL) was stirred at room temperature for 15 min. Then, p-nitrobenzaldehyde (76 mg, 0.50 mmol, Adamas-beta, 99%) was added and allowed to react at 25 °C for 48 h. After the reaction was complete, the SZF3 catalyst was magnetically separated using an external magnet and directly reused in the next experiment. The resulting reaction mixture was extracted with ethyl acetate (3 × 20 mL). The combined organic phases were dried over anhydrous Na2SO4 and evaporated under reduced pressure to give the crude product, which was purified by flash column chromatography eluted with petroleum ether/ethyl acetate (v/v = 10/1 to 2/1) to give pure 2-(hydroxy(4- methoxyphenyl)methyl)cyclohexanone. The anti/syn (Dr) ratio was determined by 1H NMR in CDCl3, in which the chemical shifts of syn- and anti-CHOH protons were at δ = 5.32 ppm (d) with 3J = 1.3 Hz and δ = 4.74 ppm (d) with 3J = 8.8 Hz, respectively. The enantiomeric excess (ee) was determined by high-performance liquid chromatography (HPLC) (Agilent Technologies 1200 Series) with a 254 nm UV-vis detector using a Daicel chiralpak Chiral OD column (4.6 mm × 250 mm), eluting with n-hexane/ iso-propanol (95/5) with a flow rate of 0.5 mL/min at 20 °C.
The chemical compositions of the magnetic solid acids SZF1-5 are listed in Table 1. Although the molar ratios of ZrOCl2:Fe3O4 in the preparation of Zr(OH)4-Fe3O4 varied greatly in the range 0.2-5.0, the formed Zr(OH)4 could be loaded onto nano-Fe3O4 in 72%-75% yield. The content of Zr(OH)4 in SZF1-5 determined by the inductively coupled plasma (ICP) method increased from 13.9% to 71.2% with the increasing molar ratio of ZrOCl2:Fe3O4. It is noteworthy that the SO42− contents in the magnetic solid acids SZF1-5 showed a close relationship with the loaded Zr(OH)4. The immersed SO42− contents in SZF1-5 first increased with increasing loaded amount of Zr(OH)4. However, when the loaded Zr(OH)4 exceeded 33.9%, the SO42− content gradually decreased owing to the accumulation of Zr(OH)4 on the outer surface of nano-Fe3O4.
The magnetic properties of the representative solid acids SZF1, SZF3, and SZF3ʹ (SZF3 reused five times) were examined through vibrating sample magnetometry. Their hysteresis loops at room temperature are shown in Fig. 1. VSM analysis showed that the magnetic solid acids SZF1, SZF3, and SZF3ʹ possessed saturated magnetization values of 55.58, 39.21, and 46.44 emu/g, respectively. The saturated magnetization values of SZF1, SZF3, and SZF3ʹ were all sufficiently high to meet the needs of magnetic separation. The zero coercivity and resonance of each magnetization loop confirmed the superparamagnetism behavior at 25 °C for all of the samples, which is very beneficial for the catalyst’s rapid dispersion and magnetic separation. Unexpectedly, the magnetic susceptibility of SZF3ʹ increased from 39.21 to 46.44 emu/g. The most likely reason for this phenomenon is the decrease in SO42− content from 20.7 wt% (SZF3) to 12.8 wt% (SZF3ʹ), which was verified by the decrease in the catalytic performance.
Using solid acid SZF3 as an example, the thermal decomposition was characterized by TG-DSC. From the TG-DSC results shown in Fig. 2, there are three types of peaks for SZF3. An endothermic peak with a gradual weight loss of 6.08% below 200 °C, which was verified by the DSC curve, was mainly attributed to desorption of surface-bound or intercalated water in the pores. Another gradual weight loss peak of 2.41% at 200-600 °C resulted from the dehydration of Zr(OH)4 to form amorphous ZrO2 [28] and the transformation of zirconia from the tetragonal to the monoclinic phase [29]. Accompanied by an endothermic peak centered at 721 °C in the DSC curve, the peak with a sharp weight loss of 8.06% at 600-780 °C corresponded to the removal of the SO42− group [14].
The NH3-TPD profiles of the magnetic solid acids SZF3, SZF3-300, and SZF3-650 are shown in Fig. 3. Magnetic solid acid SZF3 with 0.952 mmol/g acidity showed a broad NH3 desorption peak, including four peaks centered at 146, 257, 389, and 539 °C, which indicated that the surface acid strengths were not homogeneously distributed. In general, the broad NH3 desorption peak was divided into two types of desorption peaks based on Gaussian fitting: the first broad medium temperature (MT) desorption signal in the range 100-300 °C, corresponding to NH3 adsorbed on the acid sites at medium strengths; and the second broad high-temperature (HT) peak in the range of 300-650 °C, suggesting the presence of very strong acid sites [29]. Unfortunately, the MT and HT desorption signals of SZF3-300 and SZF3-650 sharply decreased with increasing calcination temperature from 110 to 650 °C for 3 h. Especially for magnetic solid acid SZF3-650, little NH3 adsorption demonstrated weak acidity (0.178 mmol/g), which was also evidenced by its poor catalytic performance in the following asymmetric aldol reaction.
The TEM images of nano-Fe3O4 and magnetic solid acid SZF3 are shown in Fig. 4. The micrograph of nano-Fe3O4 clearly shows uniform square crystallites of about 60-80 nm in diameter. After Fe3O4 was coated with Zr(OH)4 and immersed in 0.5 mol/L sulfuric acid, the obtained magnetic solid acid SZF3 with 80-100 nm in diameter shows about 10-20 nm growth of particle size. The N2 adsorption-desorption isotherm of magnetic solid acid SZF3 (Fig. 5) is a type II isotherm in classic definitions, agreed well with the normal form obtained with a non-porous or macroporous adsorbent. This indicates that magnetic solid acid SZF3 was a typical microporous material, which was verified by its BJH pore size distribution plot (inset of Fig. 5). Based on the multipoint BET method and the BJH method, the BET specific surface area and pore volume of magnetic solid acid SZF3 were calculated to be 15.8 m2/g and 0.029 mL/g, respectively.
Figure 6 shows the high-angle powder XRD patterns of nano-Fe3O4, Zr(OH)4-Fe3O4, and SO42−/Zr(OH)4-Fe3O4. The XRD pattern of nano-Fe3O4 shows the typical peaks at 18.12°, 30.08°, 35.45°, 43.06°, 53.49°, 57.00°, 62.62°, and 74.05°, which correspond to the (111) (d = 5.68 Å), (220) (d = 3.44 Å), (311) (d = 2.94 Å), (400) (d = 2.44 Å), (422) (d = 1.99 Å), (511) (d = 1.87 Å), (440) (d = 1.72 Å), and (533) (d = 1.48 Å) reflections, respectively [30]. After Zr(OH)4 and SO42− were consecutively loaded on nano-Fe3O4, there was no significant change in the various characteristic diffraction peaks, which indicated that Zr(OH)4 and SO42− were well-dispersed on/in the internal and external surfaces of nano-Fe3O4.
The asymmetric aldol reaction between 4-nitrobenzaldehyde and cyclohexanone was used as a model reaction to investigate the catalytic performance of the as-synthesized magnetic solid acids [31, 32]. From Fig. 7, the catalytic reaction temperature had a great influence on the yield and enantioselectivity of the aldol adduct. With increasing reaction temperature, the yield of the aldol adduct increased, whereas the enantioselectivity markedly decreased. Considering these two contradictory factors, the optimal temperature of 25 °C was chosen for the following screening tests.
Furthermore, the catalytic performance of various magnetic solid acids (SZF1-5) in the aldol reaction of 4-nitrobenzaldehyde with cyclohexanone was evaluated, and the results are listed in Table 2. It was found that the SO42− content had a significant influence on the yield of the aldol adduct. Magnetic solid acid SZF3 with 20.7% SO42− content produced the highest yield (96%, entry 3). However, the stereoselectivity was insensitive to the SO42− content. The enantioselectivity (82.9%-85.1% ee anti) and diastereoselectivity (anti/syn = 79-82/21-18) changed little when the SO42− content varied in the range 6.8%-20.7%. Unfortunately, in the absence of magnetic solid acid, the aldol reaction afforded unsatisfactory catalytic results (54% yield, 20% ee anti, and anti/syn = 70/30, entry 6). In particular, compared with an organic acid (>99% yield, 96.8% ee anti, and anti/syn = 93/7), magnetic solid acid SZF3 gave similar diastereoselectivity (anti/syn = 91/9) and lower enantioselectivity (90.8% ee anti) in excellent yield (>99%) (entries 7 and 8). Considering the possible effect of the bonding interaction between Zr(OH)4 and Fe3O4 on the catalytic performance, Zr(OH)4/SO42− was prepared according to the same procedure as SZF3 in the absence of Fe3O4. A similar yield and stereoselectivity of aldol products were obtained (entry 9), which indicated that the bonding interaction between Zr(OH)4 and Fe3O4 had little effect on the catalytic performance.
It is well known that the calcination temperature of the solid acid strongly affects the form of the SO42− group [14]. To investigate the influence of the calcination temperature and the amount of SZF3, the aldol reaction was performed under the optimized conditions. From Fig. 8, after calcining at 110 °C, magnetic solid acid SZF3 gave the highest catalytic performance (>99% yield, 92.1% ee anti, and anti/syn = 89/11). With increasing the calcination temperature from 110 to 650 °C, the enantioselectivity and yield decreased for different amounts of catalyst SZF3 (5.0-55.0 mg). In particular, the enantioselectivity and yield of the aldol adduct sharply decreased when the calcination temperature reached 650 °C. The main reason might be the removal of the SO42− group with increasing temperature, which was verified by the sharp mass loss at 600-780 °C (Fig. 2) and the decreased acidity of catalyst SZF3-650 (Fig. 3).
Encouraged by the remarkable results under the above reaction conditions, the substrate scope was extended to a wide variety of benzaldehydes with ortho, meta, and para substituents. The catalytic results are summarized in Table 3. For the benzaldehydes with strong electron-withdrawing substituents (R = NO2 and CN), the aldol reactions exhibited good to excellent enantioselectivity (86.0%-95.6% ee anti) and diastereoselectivity (anti/syn = 88-96/12-4) with 83%-100% yield (entries 1-6). However, the benzaldehydes with weak electron-withdrawing substituents (R = Cl and Br) gave moderate to good enantioselectivity (70.9%-85.9% ee anti) and disappointing yield (19%-76%) although good to excellent diastereoselectivity (anti/syn = 88-96/12-4) was achieved. Unfortunately, the benzaldehydes with electron-donating substituents (R = CH3 and OCH3) did not proceed smoothly.
At the end of the catalytic aldol reaction, magnetic solid acid SZF3 could be easily and quantitatively separated from the reaction mixture using an external magnet and directly reused in cycle tests. Figure 9 shows the catalytic results of reused magnetic solid acid SZF3 in the aqueous asymmetric aldol reaction of cyclohexanone with 4-nitrobenzaldehyde. To our surprise, the yield and enantioselectivity unexpectedly increased during the first three cycles. In the third cycle, magnetic solid acid SZF3 produced the highest enantioselectivity (94.7% ee anti) owing to the increased dispersion in the aqueous medium. It is noteworthy that 91.3% yield, 90.2% ee anti, and anti/syn = 78/22 were obtained in the fifth run. The slight decrease in catalytic performance was because of the loss of SO42− (20.7 to 12.8 wt%) in the catalytic process.
A large-scale asymmetric aldol reaction between cyclohexanone (9.8 g, 0.10 mol) and 4-nitrobenzaldehyde (3.8 g, 25 mmol) was performed at 25 °C for 48‒72 h in the presence of epi-CDNH2 (110.0 mg, 0.375 mmol) and SZF3 (350.0 mg). Good catalytic performance (96% yield, anti/syn = 90/10, 93.2% ee anti) was observed. After the completion of the aldol reaction, solid acid SZF3 was filtered for reuse. Furthermore, epi-CDNH2 was also recovered by extraction with dilute hydrochloric acid (10%), alkalization with an excess of concentrated ammonia, and extraction with CH2Cl2. The catalytic system could be reused and gave satisfactory results (87% yield, anti/syn = 89/11, 90.4% ee anti) in the third cycle.
With the purpose of realizing environmentally benign and friendly processes in asymmetric reactions, a series of sulfated zirconium hydroxide compounds loaded on nano-Fe3O4 were prepared, characterized, and applied in the aqueous asymmetric aldol reaction of various benzaldehydes with cyclohexanone for the first time. These magnetic solid acids, which could be easily and quantitatively separated from the reaction mixture using an external magnet, showed good to excellent catalytic performance for benzaldehydes with strong electron- withdrawing substituents and possessed good tolerance in five consecutive runs without a significant loss of catalytic performance.
2015, Vol. 36 Issue (3): 425-431
PDF (695 KB)
2015, Vol. 36 Issue (3): 425-431
PDF (695 KB)