Introduction

In the past few decades, the development of different catalysts which combine the advantages of homogeneous and heterogeneous catalysts has been investigated [1], such as immobilizing the catalyst to organic or inorganic polymeric platforms [2,3,4,5,6,7]. The immobilization of catalysts enables facile separation and recyclability, although usually the catalytic reactivity and selectivity decrease. This probably occurs due to: (1) chemical changes formed in the catalyst structure, in order to enable covalent binding of the catalyst to the immobilizing platform, and (2) less exposed catalytic active sites. Therefore, development of other methods was required and microencapsulation was suggested. By encapsulating the catalyst within the capsule core, the catalyst does not undergo chemical changes and catalytic active sites are more exposed. The main disadvantages of encapsulating a catalyst are firstly that the catalyst is synthesized before encapsulation; therefore, it has to be stable enough to undergo encapsulation conditions. Secondly, changing the catalyst influences the polymeric shell and the formation of the capsules. Since it was first suggested [8], a number of catalytic microcapsules have been employed [9,10,11,12,13,14,15,16,17,18,19,20]. One of the first polyurea microreactors was demonstrated by Ramarao et al. in 2002, where he encapsulated palladium acetate within polyurea capsules and utilized them in the Suzuki reaction [21]. By simple filtration, the catalyst was separated and reused four times without a significant loss in its activity. In the same year, Ley et al. also encapsulated palladium acetate within polyurea microcapsules and demonstrated their applicability and recyclability in Heck, hydrogenation, Stille and Suzuki–Miyaura reactions [22]. Ley’s catalyst was commercialized as PdEnCat and used in other catalytic reaction and enabled reusability with low leaching of the catalyst [23,24,25,26]. However, when the catalytic performance of the commercialized catalyst was examined, it was found that the reaction mainly occured in the solution, indicating that the capsules acted as catalyst reservoirs and not as heterogeneous catalysts [27, 28]. In addition, it was found that the PdEnCat morphology changed after reaction. The capsules crack, indicating that the shell is fragile. One of the solutions to the cracking of the shell was demonstrated in 2008 by Sasson and co-workers who developed PdAlqCat, a heterogeneous catalyst based on Ley’s procedure, where they used N-methyl-N,N,N-trioctylammonium chloride (Aliquat® 366) an ionic liquid (IL), as the emulsifier [29]. It was found that the Aliquat® 366 acts not only as the emulsifier, but also incorporates within the polyurea shell. The incorporation of the aliquat had an important role in the stabilization of the capsules and the catalyst. The PdAlqCat was applied in five consecutive cycles in the hydrogenation reaction with negligible decline in its activity. Most importantly no morphology changes were observed in the microcapsules after the fifth cycle, indicating that the presence of ionic liquid has a great effect on the stabilization of polyurea microreactors. Thus, we have developed a method based on the emulsification and interfacial polyaddition technique to encapsulate the ionic liquid, 1-methyl 3-butylimidazolium hexafoluorophosphate (BMIm[PF6], in polyurea shells (BMIm[PF6]@PU microcapsules) [30]. The technique was recently published and the utilization of these capsules is presented in this paper.

Results and discussion

The formation of BMIm[PF6]@PU microcapsules was based on previous work conducted in our laboratory [30]. Briefly, the BMIm[PF6] phase consisting of 4.5 g BMIm[PF6] and 1.1 g of polymethylene polyphenyl isocyanate (PAPI 27) was emulsified with the water phase composed of 17.0 g of water and 1.0 g of the surfactant butylated polyvinylpyrrolidone (Bu-PVP). Followed by slow addition of 0.7 g of 1,6-hexamethylene diamine (HMAD) and stirring at room temperature for 2 h. The resulted microcapsules were separated by cetrifugation, washed three times with water, dried and analyzed by scanning electron microscopy (SEM, Fig. 1a). The ability of BMIm[PF6]@PU, formed by BMIm[PF6] in water emulsion, to act as microreactors was examined by dissolving an organocatalyst, cinchonine (Scheme 1) or metal catalyst within the BMIm[PF6] pre-encapsulation.

Figure 1
figure 1

SEM images of a BMIm[PF6]@PU microcapsules, b cinchonine, BMIm[PF6]@PU microreactors and c Pt,BMIm[PF6]@PU microreactors

Scheme 1
scheme 1

Cinchonine structure

When the encapsulation of cinchonine, an organocatalyst, was examined by dissolving 100 mg of cinchonine in the BMIm[PF6] phase pre-emulsification, spherical capsules were formed (cinchonine, BMIm[PF6]@PU microreactors) (Fig. 1b). These microreactors were dried and utilized in the Michael addition reaction between dimethylmalonate and trans-β-nitrostyrene (Scheme 2), at room temperature for 3 days.

Scheme 2
scheme 2

Michael addition between dimethylmalonate and trans-β-nitrostyrene

The product, dimethyl 2-(2-nitro-1-phenylethyl)malonate, was achieved with moderate conversion (50%), as validated by NMR analysis. In addition, ionic liquid characteristic peaks were observed, indicating that BMIm[PF6] leaches from the capsules. SEM images showed that the capsules were destroyed (Fig. 2a), thus explaining the presence of ionic liquid in the 1H-NMR spectrum.

Figure 2
figure 2

SEM images of a cinchonine, BMIm[PF6]@PU microreactors and b Pt,BMIm[PF6]@PU after the Michael addition and hydrosilylation reactions, respectively

In addition, 30 mg of platinum acetylacetonate were dissolved in 300 mg of CHCl3, which was added to the BMIm[PF6] phase pre-emulsification. The capsules formed (Pt, BMIm[PF6]@PU) were analyzed by SEM. Spherical microreactors were formed (Fig. 1c), with no detection of the ionic liquid outside the capsules. These microreactors were dried and utilized in the hydrosilylation reaction between phenylacetylene and triethylsilane at room temperature (Scheme 3). A conversion of 70% was detected by NMR. Unfortunately, BMIm[PF6]’s characteristic peaks were observed as well. SEM images indicated that the capsules collapse and break, and the presence of ionic liquid was detected outside the capsules (Fig. 2b).

Scheme 3
scheme 3

Hydrosilylation reaction between phenylacetylene and triethylsilane

These results led us to conclude that the BMIm[PF6]@PU microreactors break due to mechanical stress that develops in the polyurea shell, as was noted in the literature. Therefore, we examined the feasibility of encapsulating BMIm[PF6] within polyurethane, which is known to be more flexible than polyurea. This was performed by changing the surfactant to polyvinyl alcohol (Mw = 20–30 K, PVA), which also acts as a substituting monomer in the polyaddition with isocyanate, to form the polyurethane shell. The BMIm[PF6] phase, consisting of 3.75 g of BMIm[PF6] and 1.0 g of PAPI 27, was emulsified with the water phase, consisting of 1.5 g of PVA and 18.0 g of water, which was then stirred at room temperature for 2 h. The resulting material was washed three times with water by centrifugation and analyzed by SEM (Fig. 3).

Figure 3
figure 3

BMIm[PF6]@polyurethane microcapsules before optimization

Spherical microcapsules, polyurethane chunks and non-encapsulated ionic liquid were observed. Thereafter, the system was optimized by examining a number of parameters, the isocyanate type, the BMIm[PF6] percentage and the PVA type as well as the amount. First, four isocyanate types were tested: 2,4-toluene diisocyanate (TDI), PAPI 27, 1,6-hexamethylene diisocyanate (HDI), and 4,4′-methylenebis(cyclohexyl isocyanate) (MBDI) in the presence of two different PVA surfactants, 20–30 K and 30–70 K (Figs. 4 and 5, respectively). In the presence of MBDI phase separation occurred (Figs. 4c and 5c), and in the presence of HDI, polymeric chunks were formed (Figs. 4b and 5b). When PAPI 27 was used, capsule features were observed, although in the presence of PVA 20–30 K, non-encapsulated ionic liquid was observed as well, and in the presence of PVA 30–70 K, mainly polymeric chunks were obtained (Figs. 4a and 5a). In the presence of TDI, capsule formation was indicated in both PVA types (Figs. 4d and 5d), although in the presence of PVA 30–70 K, non-encapsulated ionic liquid was observed as well (Fig. 4d). The morphology obtained in the presence of TDI and 20–30 K PVA resembles the desired spherical microcapsules. Thus, further optimization was performed in the presence of TDI and PVA 20–30 K.

Figure 4
figure 4

SEM images demonstrating the effect of different isocyanates on the formation of BMIm[PF6]@polyurethane microcapsules, in the presence of PVA 20–30 K as the surfactant. a PAPI 27, b HDI, c MBDI and d TDI

Figure 5
figure 5

SEM images demonstrating the effect of different isocyanates on the formation of BMIm[PF6]@polyurethane microcapsules, in the presence of PVA 30–70 K as the surfactant. a PAPI 27, b HDI, c MBDI and d TDI

Then, the effect that the BMIm[PF6] percentage has on the formation of the BMIm[PF6]@polyurethane microcapsules was examined. Five percentages were tested: 2, 6, 10, 14 and 18%. The SEM images showed that the spherical morphology is impaired with increasing amounts of the ionic liquid (Fig. 6). Thus, 2% of BMIm[PF6] was chosen as the optimal amount.

Figure 6
figure 6

SEM images showing the effect that the BMIm[PF6] percentage has on the formation of BMIm[PF6]@polyurethane microcapsules. a 2%, b 4%, c 10%, d 14% and e 18%

Next, the amount of PVA was examined; six values were tested: 0.5, 1.0, 1.5, 2.0, 2.5 and 3.0 g (Fig. 7). SEM images show that a minimum of 1.0 g PVA is required to form spherical capsules. In addition, in amounts higher than 1.5 g, the spherical morphology is impaired. PVA serves as one of the monomers required for the formation of the polyurethane shell, as well as the surfactant. Higher PVA quantities promote the formation of a more cross-linked shell; thus, we preferred using 1.5 g rather than 1.0 g of PVA.

Figure 7
figure 7

SEM images showing the effect of the amount of PVA on the formation of BMIm[PF6]@polyurethane microcapsules. a 0.5 g, b 1.0 g, c 1.5 g, d 2.0 g, e 2.5 g and f 3.0 g

The optimized conditions were found to be 2% BMIm[PF6] and 1.5 g PVA, which were used in the formation of the cinchonine, BMIm[PF6]@polyurethane microcapsules. Briefly, the ionic liquid phase, consisting of 30 mg of cinchonine and 0.5 g of BMIm[PF6], was emulsified with the water phase, consisting of 1.5 g of PVA at 10000 rpm for 1 min. This was followed by stirring for 2 h at room temperature to form the desired polyurethane microreactors. The formed microreactors were washed three times with water, dried and characterized using SEM, IR, TGA and solid-state NMR. SEM demonstrated that spherical capsules were formed (Fig. 8). To ensure the absence of ionic liquid out of the capsules, the dried capsules were dissolved in water, stirred and separated by centrifugation. CDCl3 was added to the water phase, extracted and was examined using 1H-NMR. Ionic liquid was not detected in the NMR indicating particulated ionic liquid within polyurethane is formed.

Figure 8
figure 8

SEM images of cinchonine, BMIm[PF6]@polyurethane microcapsules

IR analysis (Fig. 9) indicated that polyurethane was formed, by the presence of the polyurethane characteristic peaks: 3271 cm−1 representing the N–H stretching band; the C=O stretching band is represented by the peak at 1637 cm−1. The peak at 1218 cm−1 represents the C–O stretching band. An additional peak is observed at 2274 cm−1 belonging to the C=O stretch of an isocyanate group, indicating that not all of the isocyanate motifs react during the polyaddition process. To prove this assumption, we washed the capsules with ether, a solvent that can extract unreacted TDI monomers and examined them in IR again (Fig. 10). Both spectra overlap one another, indicating the peak at 2274 cm−1 represents unreacted isocyanate groups incorporated within the polyurethane shell.

Figure 9
figure 9

IR spectrum of cinchonine, BMIm[PF6]@polyurethane microcapsules

Figure 10
figure 10

IR spectra of cinchonine, BMIm[PF6]@polyurethane microcapsules before and after extraction with ether black and red curves, respectively

Solid-state NMR (Fig. 11) indicates that polyurethane is formed by the presence of a peak at 156 ppm, representing the C=O of the urethane group. In addition, aromatic peaks are present, at 136, 125 and 123 ppm, representing the imidazolium groups of the BMIm[PF6]. These peaks overlap with the aromatic peaks of TDI, present in the polyurethane shell, and cinchonine entrapped within the polyurethane microcapsules. Furthermore, the aliphatic peaks obtained overlap the PVA present in the polyurethane shell and the aliphatic component of BMIm[PF6]. The broad peaks from 224 to 236 are presumed to be satellites of the peaks presented from 123 to 136 ppm. It should be noted that IR analysis showed that some of the isocyanate groups did not react, and the characteristic isocyanate peak overlaps with the ionic liquid peaks between 136 and 123 ppm.

Figure 11
figure 11

a 13C-NMR spectrum of cinchonine in CDCl3. b 13C-NMR spectrum of PVA 20–30 K in D2O. c 13C-NMR spectrum of BMIm[PF6] in CDCl3 and d 13C CP-MAS NMR of cinchonine, BMIm[PF6]@polyurethane microcapsules

TGA analysis revealed two major decomposition peaks (Fig. 12). The first, from 200 to 550 °C, represents the overlap of the decomposition of cinchonine, BMIm[PF6] and the polyurethane present in the polyurethane microcapsules. The second occurs from 600 to 950 °C, representing the decomposition of species formed during the first step of the microcapsules’ decomposition. A total decomposition of 79% was observed.

Figure 12
figure 12

TGA curve of a BMIm[PF6], b cinchonine, c PVA 20–30 K, d cinchonine, BMIm[PF6]@polyurethane microcapsules

The cinchonine, BMIm[PF6]@polyurethane microcapsules were tested in the Michael addition reaction between dimethylmalonate and trans-β-nitrostyrene at room temperature for 3 days and 20% of the desired product was obtained. The microreactors were separated from the reaction medium by centrifugation, washed and examined by SEM. SEM images indicate that the cinchonine, BMIm[PF6]@polyurethane microcapsules were not destroyed after the Michael addition reaction (Fig. 13b). The used polyurethane microreactors were examined by solid-state NMR. The spectra obtained before and after the reaction indicate that the solid does not change. This result supports our assumption that mechanical stress most likely destroyed the BMIm[PF6]@PU microcapsule during the reaction. In addition, the used capsules were examined in three continuous cycles. In the first two cycles, no decrease in the activity of the catalyst was detected. In the third cycle, the conversion of the reaction was reduced to 5%.

Figure 13
figure 13

SEM images of cinchonine, BMIm[PF6]@polyurethane microcapsules. a Before the reaction and b after the reaction

Conclusion

The ability of particulated ionic liquid in polyurea to act as microreactors, by encapsulating the desired catalyst simultaneously to the BMIm[PF6], was examined. Although Sasson and co-worker showed that the presence of an ionic liquid increases the capsule stability, when the ionic liquid acts as the core, they are fragile. Thus, polyurethane a less fragile polymer was chosen as the polymeric shell. The encapsulation of ionic liquid within a polyurethane shell was examined and the optimized system was used to encapsulate cinchonine, an organocatalyst simultaneously to ionic liquid (cinchonine, BMIm[PF6]@polyurethane microreactors). The system was characterized and their ability to act as microreactors in the Michael addition reaction was tested. The desired product was achieved with a 20% conversion after 3 days at room temperature. In addition, the microcapsules were examined under SEM. Their morphology after the reaction was not changed furthermore. These results show the importance of the polymeric shell on the stability and applicability of the microreactors.

Experimental section

Materials

All materials were purchased from Sigma–Aldrich or Acros unless otherwise noted. The ionic liquid, BMIm[PF6] was purchased from Chemada Fine Chemicals.

Instruments

Emulsification was carried out using Kinematica Polytron homogenizer PT-6100 equipped with dispersing aggregate 3030/4EC. High-resolution scanning electron microscopy (HR SEM) Sirion (FEI company) was used to examine the formation of the polymeric microcapsules and their morphology. Images were taken using Schottky-type field emission source and secondary electron detector, at 5 kV. FTIR measurements were performed on a PerkinElmer FTIR spectrometer model Spectrum 65. 1H-NMR and 13C-NMR spectra were recorded with Bruker DRX-400 instrument. Solid-state 13C CP-MAS NMR spectrum was recorded with Bruker DRX-500 instrument. The cinchonine, BMIm[PF6]@polyurethane microcapsules composition was determined by thermogravimetric analysis (TGA), performed under a nitrogen atmosphere at temperatures ranging from 50 to 950 °C at a heating range of 10 °C/min on a TGA/SDTA851e instrument.

Preparation of polymeric microreactors

BMIm[PF6]@polyurea microreactors

The formation was based on previous work conducted in our laboratory [30]. Briefly, the BMIm[PF6] phase consisting of 100 mg of cinchonine or 30 mg of platinum acetylacetonate, 1.1 g PAPI 27 and 4.5 g BMIm[PF6] was emulsified with the water phase consisting of 1.0 g surfactant, Bu-PVP and 17.0 g water, at 10000 rpm. After 30 s, 0.7 g HMDA was added slowly and the resulted mixture was stirred at room temperature for 2 h. Then, the microreactors were separated by centrifugation and washed three times with water and dried.

BMIm[PF6]@polyurethane microcapsules

The ionic liquid phase, consisting of BMIm[PF6] (0.5–4.5 g) and the desired isocyanate monomer (0.65–1.1 g) was emulsified with 17.0 g of water containing the desired PVA (0.5–3.0 g), at 10000 for 1 min. The resulted mixture was stirred at room temperature for 2 h. Then, the BMIm[PF6]@polyurethane microcapsules were separated via centrifugation, washed with water and dried to give a powder.

Cinchonine, BMIm[PF6]@polyurethane microcapsules

The BMIm[PF6] phase containing 30 mg cinchonine, 0.65 g TDI and 0.5 g BMIm[PF6] was emulsified with the water phase consisting of 22.0 mg of water and 1.5 PVA 20–30 K for 1 min at 10000 rpm. Followed by 2 h room temperature stirring. Then, the resulted microreactors were separated and washed by centrifugation and dried, to give a power-like material.

Catalytic reactions

Hydrosilylation reaction

A total of 300 mg of dried Pd, BMIm[PF6]@PU microreactors were redispersed in 4 mL of ether. Then, phenylacetylene (30 mg, 0.5 mmol) and triethylsilane (80 mg, 0.7 mmol) were added and the reaction was conducted at room temperature for 24 h. The microcapsules were then separated from the reaction medium by centrifugation and washed with ether three times. The collected ether was then evaporated and the resulting material was examined by NMR. 1H NMR (400 MHz, CDCl3, δ): trans: 7.5–7.1 (m, 5H, Ar H), 6.9 (d, J = 19 Hz, 1H; CH), 6.4 (d, J = 19 Hz, 1H; CH), (m, 6H, Si–CH2), 0.6 (m, 9H, CH3); cis 7.5–7.1 (m, 5H, Ar H), 5.8 (d, J = 3 Hz, 1H; CH), 5.6 (d, J = 3 Hz, 1H; CH), 0.9 (m, 6H, Si–CH2), 0.6 (m, 9H, CH3).

Michael addition reaction

A total of 300 mg of cinchonine, BMIm[PF6]@PU/polyurethane microreactors were dissolved in water, followed by the addition of dimethylmalonate (45 mg, 0.3 mmol) and trans-β-nitrostyrene (20 mg, 0.15 mmol) and stirred at room temperature for 3 days. The microcapsules were separated from the reaction medium by centrifugation, and the product was extracted with ether. In addition, the microreactors were washed three times with ether to ensure that all of the reagents were extracted from the microcapsule. The desired product, dimethyl 2-(2-nitro-1-phenylethyl)malonate was obtained (50 or 20% depending on the system used) as determined from 1H-NMR. 1H NMR (400 MHz, CDCl3, δ): 7.4–7.2 (m, 5H, Ar H), 4.9–4.8 (m, 2H, CH2NO2), 4.3–4.2 (m, 1H, PhCH), 3.8 (d, J = 3.8 Hz, 1H COCHCO), 3.7 (s, 3H, OCH3) and 3.5 (s, 3H, OCH3).