Introduction

Selective epoxidation of alkenes into high value chemicals (epoxides) which are widely used in the fine chemical industry is of great attention in chemical and pharmaceutical industries [1]. Some transition metal oxides such as NiO, CoO, MoO3, CuO, TiO2–SiO2, Au/SiO2, CuOx/SiO2 and tungstate(VI) or molybdate(VI) have been reported for the epoxidation of alkenes with organic hydroperoxides [2–8]. However, in most cases, either alkene conversion or alkene oxide selectivity/yield was poor and many of these transition metal catalysts are expensive and highly dangerous for the environment as well. From an economic point of view, the separation of ultrascaled and nanosized catalysts from the reaction system via routine methods such as free sedimentation, centrifuging and filtration is difficult, time-consuming and costly. Magnetic nanoparticles (NPs) which are economic friendly can be easily separated by applying a simple external magnet without any significant loss of activity put forward a solution to this problem. In addition, good selectivity and great stability are some especial advantageous of magnetic NPs [9–11]. Magnetic NPs have been reported to catalyze a wide range of reactions including, C–C, C–S, C–N bond formation, asymmetric synthesis and oxidations reactions [12–15]. Usually, silica was utilized to coat the Fe3O4 particles forming Fe3O4/SiO2 core–shell structure. This layer not only protects Fe3O4 from oxidizing or dissolving in the acidic reaction media, but also stabilizes NPs by preventing aggregation of the Fe3O4 particles [16]. The silica shell can also provide numerous surface Si–OH groups for further modification [17, 18].

This article presents the epoxidation of cyclic and acyclic alkenes with m-chloroperbenzoic acid (m-CPBA) in the presence of catalytic amount of Fe3O4/SiO2 NPs in dichloromethane as solvent. By using this catalyst, the corresponding products were obtained in excellent yields after 4 h and magnetic NPs, which are recovered from the reaction mixture by using external magnetic field simplified the isolation of products (Fig. 1).

Fig. 1
figure 1

Epoxidation of alkenes with m-CPBA catalyzed by Fe3O4/SiO2 in dichloromethane solvent. Reaction conditions: alkene (1 mmol), m-CPBA (2 mmol), catalyst (0.06 g), DCM (3 mL) 4 h at room temperature

Full size image

Experimental

General

Tetraethoxysilane (TEOS), FeCl3·6H2O, FeCl2·4H2O, styrene, 4-chlorostyrene, cyclohexene, α-methyl styrene, indene, cis- and trans-stilbene, cyclohexene, cyclooctene, 1-octene, m-CPBA, pyridine N-oxide (PNO), pyridine (Py), N-methylmorpholine-N-oxide (NMNO), 1-methylimidazole (MI), imidazole, dichloromethane (DCM), ethyl acetate (EtOAC), ethanol (EtOH), CH3CN, CHCl3, THF, NaIO4, NH4OAC, t-BOOH, PhI(OAC)2, H2O2 (30 %) and Oxone were purchased from Merck and used without purification. Graphite powder was obtained from Aldrich. The resulting Fe3O4/SiO2 was characterized by IR, TEM, TGA, VSM and X-ray diffraction (XRD) patterns. The IR experiments were carried out on a Perkin-Elmer 783 Infrared spectrophotometer in a KBr pellet, scanning from 4,000 to 600 cm−1 at room temperature. The XRD measurements were carried out by using a Bruker D8-advance X-ray diffractometer with Cu Kα radiation (k = 1.5406 Å). The TEM images measurements were obtained using Philips CM10 instrument. Magnetization measurements were carried out at 300 K on a vibrating sample magnetometer (VSM Leak shore 7200). The TGA analysis was performed by heating the samples in an argon flow at a rate of 100 mL min−1 using a Perkin–Elmer Diamond TG/DTA thermal analyzer with a heating rate of 10 °C min−1. The conversion of products was determined by GC-17A Shimadzu with capillary column (Shimadzu, CBP5, 30 m × 25 mm × 0.25 µm).

General procedure for the synthesis of Fe3O4/SiO2 core–shell

Black magnetic Fe3O4 NPs were prepared as described in the literature [19]. The core–shell Fe3O4/SiO2 NPs were prepared by a modified Stober method [20]. Briefly, Fe3O4 (0.50 g, 2.1 mmol) was dispersed in the mixture of ethanol/deionized water (50:5 mL). Then, TEOS (0.20 mL) was slowly added to the mixture followed by addition of 5 mL of NaOH (10 wt%). This solution was stirred mechanically for 20 h at room temperature. Then the product, Fe3O4/SiO2, was separated by an external magnet, washed with deionized water and ethanol three times and dried at 80 °C for 10 h.

General procedure for the epoxidation of olefins

In a typical procedure, the catalyst (0.06 g) was dispersed in 3 mL of dichloromethane for 10 min. Then the substrate (1 mmol) and m-CPBA (2 mmol) were added to the mixture at room temperature. The mixture was stirred at room temperature for appropriate times. The progress of the reaction was monitored by TLC. After the completion of the reaction, the catalyst was separated by external magnet. The solution was washed with 1 M NaOH (8 mL) and brine (8 mL) and dried over MgSO4. Then, the solution was concentrated by rotary evaporator to 1 mL. Finally, the conversion of the products was determined by GC. The catalyst was washed twice with ethanol and reused.

Results and discussion

Characterizations of catalyst

The Fe3O4 NPs (A) were simply synthesized by using of FeCl3·6H2O and FeCl2·4H2O in deionized water under nitrogen atmosphere. Then, Fe3O4/SiO2 NPs were obtained by treating the Fe3O4 NPs with TEOS (Fig. 2). The magnetic properties of the Fe3O4 and Fe3O4/SiO2 NPs (B) were studied by a VSM at 300 k. As shown in Fig. 3, both NPs have super paramagnetism at room temperature. Also, no hysteresis phenomenon was observed in Fig. 3 and the saturation magnetization values for Fe3O4 and Fe3O4/SiO2 NPs were 70.495 and 38.30 emu/g, respectively. However, despite the considerable decrease of the magnetization of Fe3O4/SiO2 NPs compared to Fe3O4, Fe3O4/SiO2 NPs can still be separated from the solution by using an external magnetic field on the sidewall of the reactor. Fig. 4a shows the IR spectrum of Fe3O4 with important vibration bands in 560–590 and 3,400 cm−1, which are due to Fe–O and OH, respectively. According to the IR spectrum of Fe3O4/SiO2 in Fig. 4b, there are several important vibration bands in 560–590, 954, and 3,400 cm−1, which are due to Fe–O, Si–OH, and OH. The vibration bands in 755 and 1,100 cm−1 are due to Si–O–Si. These vibration bands (755, 954 and 1,100 cm−1) confirmed coating of the silica shell on the surface of the Fe3O4 NPs.

Fig. 2
figure 2

Schematic illustration for synthesis of Fe3O4 (a) and Fe3O4/SiO2 (b). Reaction conditions: FeCl3·6H2O (4.8 mmol g), FeCl2·4H2O (4.5 mmol), water (30 mL), NH4OH (pH 10), PEG (1 g), N2, 80 °C, 1 h (a) and Fe3O4 (2.1 mmol), ethanol/deionized water (50:5 mL), TEOS (0.20 mL), of NaOH (10 wt%, 5 mL) 20 h at room temperature (b)

Full size image
Fig. 3
figure 3

Magnetization curves of a Fe3O4 NPs. b Fe3O4/SiO2 NPs at 300 K. The Fe3O4/SiO2 NPs can be separated easily with external magnet

Full size image
Fig. 4
figure 4

FT-IR spectra of a fresh Fe3O4 b the fresh Fe3O4/SiO2 catalyst and c the Fe3O4/SiO2 catalyst of cycle 5

Full size image

The XRD pattern of Fe3O4 and Fe3O4/SiO2 NPs were determined by powder XRD. As shown in Fig. 5a, the pattern of Fe3O4 indicates a crystallized structure at 2θ: 30.2°, 35.4°, 43.3°, 53.6°, 57.5° and 63.1° which are assigned to the (220), (311), (400), (422), (511) and (440) crystallographic faces of magnetite, which is in good agreement with the literature value (JCPDS Card No. 19-0629). The XRD pattern of Fe3O4/SiO2 presents almost the same feature as shown in Fig. 5b. The broad peak at 2θ = 15–27°, which depends on amorphous silica. The average diameter of Fe3O4 was about 11 nm, while the diameter of Fe3O4/SiO2 was about 14 nm, which is due to the agglomeration of Fe3O4 inside nanospheres and surface growth of silica on the shell. The SEM image of Fe3O4 shows the morphology and average product size of Fe3O4 NPs (Fig. 6a). The TEM images of Fe3O4 and Fe3O4/SiO2 NPs are shown in Fig. 6b, c. The results showed the average product size of Fe3O4 and Fe3O4/SiO2 NPs 10 and 13 nm similar to the results of XRD patterns. The TEM images of Fe3O4/SiO2 NPs indicate the successful coating of magnetic Fe3O4 NPs (Fig. 6c). The TGA curve of Fe3O4/SiO2 shows a weight loss over the range of 90–160 °C of about 3 %. These losses can be attributed to the loss of adsorbed water and dehydroxylation of internal OH groups. The second weight loss step is over the range 250–590 °C, which can be ascribed to even further decomposition of the materials. The total weight losses are approximately 10 % (Fig. 7).

Fig. 5
figure 5

XRD pattern of a fresh Fe3O4 b the fresh Fe3O4/SiO2 catalyst which are in good agreement with the literature value (JCPDS Card No. 19-0629)

Full size image
Fig. 6
figure 6

SEM images of a Fe3O4, b TEM micrographs of Fe3O4 and c TEM images of Fe3O4/SiO2

Full size image
Fig. 7
figure 7

Thermogravimetric weight loss pattern of Fe3O4/SiO2NPs with temperature raised of 10 °C/min and initial sample weigh of 4.9 mg

Full size image

Catalytic activity

In this study, Fe3O4/SiO2 NPs were tested to catalyze the selective oxidation of alkenes using m-CPBA as an oxidant. Styrene was selected as model substrate and treated with Fe3O4/SiO2 NPs in the presence of m-CPBA as the oxygen donor at room temperature under different conditions. This reaction was best carried out using 2 equiv. of m-CPBA for 4 h at room temperature in dichloromethane. The catalytic activity of the Fe3O4 NPs was also investigated in the epoxidation of styrene, and low yield (45 %) was observed. Blank experiments showed that Fe3O4 and Fe3O4/SiO2 NPs alone are inactive towards the styrene epoxidation. To optimize the amount of catalyst, the reaction was carried out in the presence of different amount of Fe3O4/SiO2 NPs (0.01–0.08 g) at room temperature. It was found that 0.06 g of Fe3O4/SiO2 was sufficient enough to afford styrene oxide with 97 % isolated yield (Table 1, entry 6).

Table 1 Optimization of the catalyst amount of Fe3O4/SiO2 NPs
Full size table

To explore the effect of solvent on the reaction, the same reactions were performed in different solvents. Despite the moderate yields in some solvent such as EtOAC and CH3CN (Table 2, entries 1, 2), the best conversion was observed when the reaction was performed in DCM (Table 1, entry 7, 97 %).

Table 2 The effect of solvent on the conversion of styrene to the corresponding styrene oxide
Full size table

The effect of different oxidants and additives in the epoxidation of styrene

The effect of various oxidants such as NaIO4, m-CPBA, UHP, H2O2, Oxone, PhIO, PhI(OAC)2 and tert-BuOOH was investigated in the epoxidation of styrene. The results showed that m-CPBA is the best oxygen source because this oxidant can give better yield (Table 3, entry 1) while other oxidants such as UHP, H2O2, PhIO, PhI(OAC)2 gave low yields (Table 3, entries 5–8). In addition, different equivalents of the oxidant were tested, and the best catalytic activity was obtained with 2 equivalents of the oxidant which provide moderate source of oxygen for the catalytic reaction. Also, dichloromethane was chosen as the best reaction medium. Consequently, the optimum molar ratio of olefin to oxidant is 1:2.

Table 3 The effect of different oxidants on the conversion of styrene to the corresponding Styrene oxide
Full size table

The effect of various additives such as NMNO, PNO, NH4OAC, MI, Py and imidazole was investigated in the epoxidation of styrene in the DCM/m-CPBA system. Generally, additives such as NMNO and imidazole in the Mn(III) salen reaction mixture facilitate faster reaction rates and higher epoxide yields. However, in this test, the catalytic activity did not increase in the presence of PNO and other additives.

The reactions were also carried out in 0, 40 and 60 °C. The yield was lower at 0 °C than at room temperature, but both of them showed good selectivity. When the reactions were carried out in 40 and 60 °C, the reaction rates increased but the selectivity of epoxides decreased. So, considering the economic point of view and selectivity, room temperature was chosen as the best condition for this reaction. Therefore, we employed the optimized conditions (0.06 g nanocatalyst, 2 mmol m-CPBA and DCM at room temperature) for the conversion of several alkenes into the corresponding products. Table 4 lists a group of alkenes that were investigated by magnetic NPs catalysts. The catalyst showed excellent activity toward alkenes oxidation with an average isolation yield of 95 %.

Table 4 Epoxidation of different alkenes catalyzed by Fe3O4/SiO2 nanoparticles
Full size table

The efficiency of Fe3O4/SiO2 is compared in Table 5 with the earlier reported ones for their styrene epoxidation activity, expressed in terms of selectivity and yields for the styrene oxide formation. Usually, transition metals are not very highly efficient catalysts for alkene epoxidation (Table 5, entries 1–12). The comparisons of the catalysts reveal that Fe3O4/SiO2 catalyst with faster reaction rate, higher yield and selectivity shows better performance as compared to the earlier ones (Table 5, entry 13).

Table 5 Performance of different transition metal oxides for the epoxidation of styrene
Full size table

We also investigated the possibility of reusing of recovered catalysts for new reaction. Therefore, the catalyst was separated by external magnet, washed with EtOH, dried and reused directly for a subsequent round of reaction without further purification with no significant loss of activity, which validates its recyclability (Fig. 8). Moreover, the IR spectrum of Fe3O4/SiO2 NPs after five reuses show that IR bands of the original skeletal vibration of Fe3O4/SiO2 have no obvious change compared to fresh catalyst (Fig. 4c).

Fig. 8
figure 8

A comparison of the catalytic activity of fresh and reused catalyst for alkenes epoxidation. Reaction conditions: substrate (1 mmol), m-CPBA (2 mmol), DCM (3 mL), fresh or reused catalyst (0.06 g) 4 h at room temperature

Full size image

Conclusions

In summary, we developed a facile, highly efficient, and eco-friendly procedure for the epoxidation of olefins in the presence of Fe3O4/SiO2 as a heterogeneous catalyst at mild reaction conditions. The prepared catalyst is found to be efficient catalyst for the selective epoxidation of olefins to their corresponding products. This heterogeneous catalyst is highly reactive in the epoxidation of a wide range of alkenes such as linear and cyclic ones. Moreover, easy preparation, handling and recovery, reusability and long-term stability of the catalyst, as well as excellent yields in shorter reaction time under mild reaction conditions, which are some advantages of this heterogeneous catalyst, make it a useful catalyst for further applications in the area of catalysis.