Efficient route for oxazolidinone synthesis using heterogeneous biopolymer catalysts from unactivated alkyl aziridine and CO2 under mild conditions

Applied Catalysis A: General 447–448 (2012) 107–114

Contents lists available at SciVerse ScienceDirect

Applied Catalysis A: General journal homepage: www.elsevier.com/locate/apcata

Efficient route for oxazolidinone synthesis using heterogeneous biopolymer catalysts from unactivated alkyl aziridine and CO2 under mild conditions Amal Cherian Kathalikkattil, Jose Tharun, Roshith Roshan, Han-Geul Soek, Dae-Won Park ∗ School of Chemical and Biomolecular Engineering, Pusan National University, Busan 609-735, Republic of Korea

a r t i c l e

i n f o

Article history: Received 30 July 2012 Received in revised form 6 September 2012 Accepted 10 September 2012 Available online 8 October 2012 Keywords: Chitosan Pyridinium ionic liquids Carbondioxide 2-Methylaziridine, 4-Methyloxazolidin-2-one

a b s t r a c t Biopolymers made of polysaccharide chains are emerging as promising materials for designing efficient, cheap, environmental friendly and recyclable heterogeneous catalysts. In this study, we synthesized a series of covalently functionalized chitosan-alkyl pyridinium halides (CS-RPX, R = ethyl, propyl, butyl, hexyl and X = Cl, Br) and evaluated their potential application as catalysts for the chemical transformation of CO2 to 4-methyl-2-oxazolidinone using 2-methylaziridine under mild reaction conditions. The catalysts were characterized using different physicochemical methods, including X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), X-ray photoelectron spectroscopy (XPS), thermo gravimetric analysis (TGA), elemental analysis (EA) and field emission scanning electron microscopy (FE-SEM). 1 H NMR, GC–MS, EA and FT-IR were used to confirm successful oxazolidinone formation. Cycloaddition was found to proceed through the synergistic effect of the hydroxyl and amine groups of chitosan together with the anion. The catalyst was reused five times after the cycloaddition reaction, with a loss of 2–6% in conversion and 1–3% in selectivity per cycle. The effect of different reaction parameters, such as catalyst amount, time, temperature and CO2 pressure were studied to determine the reaction conditions that resulted in the highest conversion and selectivity. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Increasing focus on research for the development of feasible alternative pathways to obtain valuable chemicals using CO2 as a C1 building block is motivated by advantages like its abundance, noninflammability, inexpensiveness and non-toxicity [1,2]. Despite the challenge raised by the kinetic and thermodynamic stability of CO2 , the synthesis of useful products like cyclic and poly carbonates from CO2 using epoxides in presence of catalyst have been well progressed and are continuously being updated [3–5]. Recently, aziridines, the highly ring-strained heterocyclic N-analogues of epoxides, are being emerged as an attractive substrate for CO2 fixation via [2 + 3] coupling to yield oxazolidinones (Scheme 1) [6–12]. Oxazolidinones are a class of 5-membered heterocyclic compounds that find applications as intermediates and chiral auxiliaries in organic synthesis and polymer synthesis [13–17], and biologically active pharmaceutical agents [18,19]. Even though the development of this process had been stagnant for the past three decades [20], researchers have more recently revisited using this

∗ Corresponding author. Tel.: +82 51 510 2399; fax: +82 51 512 8563. E-mail address: [email protected] (D.-W. Park). 0926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apcata.2012.09.031

approach due to increasing concerns over the development of ecofriendly processes. Moreover, among the alternative methods for oxazolidinone synthesis, viz., carbonylation of amino alcohols using phosgene or CO [21–24] and the reaction of propargylamine or propargylic alcohol with CO2 [25–27], the former one (using phosgene or CO) is considered as less attractive C1 feedstock from a green prospective in comparison to the cheap, nontoxic and abundant CO2 . A few homogeneous catalysts like inorganic salts, complexes [12,20,28–32], organic compounds, organic salts, ionic liquids [7–10,33–35], etc. have been used for the cycloaddition of aziridines with CO2 . Despite being the most active homogeneous catalyst, ionic liquids are fraught with drawbacks, including the tedious process required to separate the reaction mixture at the end of synthesis. More recently, polymer supported ionic liquids [12,36] were developed in an attempt to overcome these limitations. Even though these synthetic polymers offer themselves as good platforms for rendering heterogeneity to ionic liquids, they still poses the limitation of acting only as a scaffold and do not promote or enhance the reaction. Recently, the role of vicinal hydroxyl groups in enhancing the cycloaddition of aziridines with CO2 was demonstrated by the high efficiency of polymer supported diol functionalized imidazolium ionic liquids [12].

108

A.C. Kathalikkattil et al. / Applied Catalysis A: General 447–448 (2012) 107–114

Scheme 1. Cycloaddition of aziridines with CO2 .

In this context, chitosan like biopolymers, which inherently possesses plenty of hydroxyl groups, could be regarded as efficient and natural support materials. Chitosan (CS), the second most abundant polysaccharide found on earth next to cellulose, is a deacetylated derivative of chitin which occurs in the exoskeleton of shellfish shell such as crab, lobster and shrimp and in food items such as grain, yeast, bananas, and mushrooms. Dry pawn waste contains 23% chitin and dry squid contains 15% chitin [38], which could be extracted using 5% NaOH followed by 30% HCl. Chitosan stands an attractive biopolymer for catalytic applications due to its low cost, biocompatibility, biodegradability, and non-toxicity [37]. Thus, chitosan is a much more environmental friendly support than synthetic polymers. Furthermore, in comparison to cellulose, chitosan is blessed with primary amine group which acts as an adaptor for easy covalent incorporation of catalytically active sites [39,40]. Sun et al. [41] recently succeeded in covalently tethering imidazolium ionic liquids to chitosan, whereby utilizing the potentials of hydroxyl and amine groups to act synergistically with the halide ion in the cycloaddition of epoxides. However, for the cycloaddition of aziridines with CO2 , pyridinium salts have been reported to show better activity than the imidazolium counterparts [9]. Despite the high ring strain energy of aziridines, the readily available unactivated aziridines are less susceptible to nucleophilic ring opening when compared to the aziridines that have been activated by quaternization, N-substitution with electron withdrawing substituents and Lewis base adduct formation [42]. In this study, we designed an efficient catalyst system that could facilitate the cycloaddition of even unactivated aziridine like 2-methylaziridine (MeAz) under mild conditions. This was accomplished by synthesizing chitosan supported alkyl pyridinium halides (CS-RPX, R = Et, Pr, Bu, Hex and X = Cl, Br) and engaging them as catalysts for the synthesis of 4-methyl-2-oxazolidinone from 2-methylaziridine and CO2 . The process represents a recyclable, eco-friendly and energy-saving pathway for the cycloaddition of unactivated aziridines with CO2 under mild conditions. 2. Experimental 2.1. Materials Chitosan, 1-methyl-2-pyrrolidinone (99%), 1,2-dichloroethane (99.8%), 1,2-dibromoethane (98%), 1,3-dibromopropane (99%), 1,4dibromobutane (99%), 1,6-dibromohexane (96%), pyridine (≥99%), acetone (≥99.9%), anhydrous tetrahydrofuran (THF) and anhydrous ethanol were purchased from Aldrich and used as received. 2Methylaziridine (90%) and biphenyl (99.5%) were also obtained from Aldrich. Carbon dioxide of 99.999% purity was used without further purification. 2.2. Catalyst synthesis The chitosan supported alkyl pyridinium halide catalysts, CSRPX, (RPX represents the ionic liquid, where X = Cl− or Br− and R = ethyl (E), propyl (Pr), butyl (B) and hexyl (H)) were synthesized [41] in two steps (Scheme 2) as follows (CS-EPBr): First, N(2-bromoethyl)pyridinium bromide, Br-EPBr, was prepared from

Scheme 2. Schematic representation of the synthesis of chitosan supported alkyl pyridinium halides (CS-RPX).

dibromoethane and pyridine. During this preparation step, dibromoethane (35 g, 186.3 mmol) was dissolved in 20 mL acetone under nitrogen atmosphere in a 100 mL two-neck round bottomed flask at room temperature. Pyridine (5 mL, 62.1 mmol) was then slowly added and refluxed for 24 h. The precipitate formed was filtered, washed in acetone for 6 h using a soxhlet extractor and dried under vacuum at 60 ◦ C for 12 h. The second step in this process was the covalent immobilization of Br-EPBr to chitosan (CS). In this step, 1.78 g chitosan (10.98 mmol) was added to 30 mL 1-methyl2-pyrrolidinone (NMP) and stirred at 50 ◦ C for 12 h in a two-neck flask. Br-EPBr (6.62 g, 24.8 mmol in 10 mL NMP) was added and then stirred at 80 ◦ C for 24 h. Upon completion of the reaction, the product was precipitated by its addition to 100 mL of ethanol, washed 4–5 times with dry ethanol and then rinsed with dry acetone. The sample was dried under vacuum at 80 ◦ C for 24 h to obtain CS-EPBr. The other CS-RPX catalysts, viz., CS-EPCl, CS-EPBr, CS-PrPBr, CS-BPBr, and CS-HPBr were synthesized using a similar process. CS-EMImBr [41] was also synthesized and its activity in the cycloaddition of CO2 with aziridines was compared with the CS-RPX catalysts. 2.3. Characterization The catalysts were characterized by X-ray diffraction (XRD), Xray photoelectron spectroscopy (XPS), fourier transform infra-red spectroscopy (FT-IR), elemental analysis (EA), thermogravimetric analysis (TGA) and field emission scanning electron microscopy (FE-SEM). The characterizations of the various X-RPX salts (EA, 1 H NMR) are provided in the supplementary data. The cyclic product was identified and confirmed using the GC–MS, EA, FT-IR and 1 H NMR techniques. X-ray diffraction (XRD) patterns were obtained on a Philips PANalytical X’pert PRO Model power diffractometer that was operated at 40 kV and 30 mA using Ni filtered Cu-K␣ radiation ˚ The diffractograms were recorded in the 2 range ( = 1.5404 A). 5–55◦ . XPS analyses were performed using an X-ray photoelectron spectrometer (Theta Probe AR-XPS System, Thermo Fisher Scientific, U.K.) with monochromatic Al-K␣ radiation (h = 1486.6 eV). The FT-IR spectra were obtained on Vertex 80 V Microscopic FTIR/Raman Spectrophotometer at a resolution of 4 cm−1 . Elemental analyses of the catalysts and oxazolidinone were carried out on a Vario Micro Cube analyzer. Thermogravimetric analyses (TGA) were performed with 3.78–6.05 mg of CS-RPX using an SDT-Q600 analyzer (TA Instruments Inc.). The surface features of the materials were observed using an S-4200 field emission scanning electron microscope (FE-SEM, Hitachi). The 1 H NMR spectra of X-RPX and oxazolidinone were obtained on a Varian 500 MHz using D2 O and CD3 OD respectively as solvents. Gas chromatography/mass spectrometry (GC–MS) analysis of the cycloaddition products was performed on an Agillent 5975 C GC/MSD analyzer.

A.C. Kathalikkattil et al. / Applied Catalysis A: General 447–448 (2012) 107–114

109

1H, NH (d)), 4.51 (t, 1H, CH2 (c )), 3.92–4.02 (m, 2H, CH2 (c) and CH (b)), 1.25(d, 3H, CH3 (a)) C, H, N Data: Calculated (%): C = 47.52; H = 6.98; N = 13.85. Observed (%): C = 47.87; H = 7.18; N = 13.85. IR spectral data: (max /cm−1 ) 3318(b), 2980(s), 2936(w), 1730(s), 1484(m), 1404(m), 1241(s), 1030(m). GC–MS: m/z calcd for C4 H7 NO2 : 101.05. Found: m/z 101.1.

Table 1 Effect of various CS-RPX catalysts on the synthesis of MeAz.a

3. Results and discussion

Entry

Catalyst

Conv (%)b , c

1 2 3 4 5 6 7 8 9 10

None CS EPBre CS-EPBr CS-EPCl CS-PrPBr CS-BPBr CS-HPBr CS-EMImBr CS-EPBrf

0 Trace 77.6 98.3 11.0 98.1 97.9 93.9 84.2 99.8

Yield (%)b

TONd

a

b

(c + d)

0 0.1 66.5 88.4 1.2 84.1 85.2 83.5 67.5 93.4

0 0.0 6.3 5.8 0.0 2.5 12.2 4.5 5.3 5.7

0 9.8 4.8 4.1 9.8 11.5 0.5 5.9 13.5 0.8

0 154.6 195.7 27.8 194.6 194.1 194.1 188.0 198.7

a Reaction conditions: Meaz = 28.3 mmol (2 mL, 25 ◦ C), CS and CS-RPX = 0.2 mmol, PCO2 = 2 MPa, temperature = 100 ◦ C, time = 4 h, 600 rpm. THF = 2 mL. b Determined by GC. c Internal standard = 0.05 g biphenyl. d Turnover number (TON): moles of aziridine converted per mole of ionic liquid. e 0.141 mmol of EPBr, equivalent to the loaded EPBr in 0.2 mmol of the synthesized CS-EPBr. f Semi-batch reaction conditions: Meaz = 28.3 mmol (2 mL, 25 ◦ C), CSEPBr = 0.2 mmol, PCO2 = 0.8 MPa (continuous supply), temperature = 100 ◦ C, time = 4 h, 600 rpm, THF = 2 mL.

2.4. Cycloaddition of 2-methylaziridine with CO2 The synthesis of methyloxazolidinones from 2-methylaziridine (MeAz) and CO2 using various CS-RPX catalysts was carried out in a 60 mL stainless steel reactor equipped with a magnetic stirrer. In a typical batch operation, an appropriate amount of the CS-RPX catalyst was charged in the reactor containing 2 mL (28.27 mmol) of MeAz and 2 mL THF. The reactor was pressurized with CO2 to a preset pressure at room temperature. The reactor was then heated to the desired temperature and stirred at 600 rpm. After the reaction was complete, the reactor was cooled in ice and the remaining CO2 was vented off carefully. 8 mL of THF was added, and the product was filtered and subjected to GC analysis. The catalyst was then recycled. 2.5. Product separation and analysis Product analysis was carried out using EA, FT-IR, 1 H NMR, and GC–MS analyzers. The conversion of MeAz was obtained from gas chromatography (GC, HP 6890, Agilent Technologies) and the yield was calculated using biphenyl (0.05 g) as an internal standard. The residue was purified over a silica gel by column chromatography using hexane/ethyl acetate = 9:1 to 3:7 as eluant to afford the products. The presence of the expected products 4methyl-1,3-oxazolidin-2-one (a), 5-methyl-1,3-oxazolidin-2-one (b), 2,5-dimethyl-piperazine (c) and 2,3-dimethyl-piperazine (d), were detected in the crude mixture by GC-MS analysis (Table 1). The isolated major product 4-methyl-1,3-oxazolidin-2-one (a), a colourless liquid, was further confirmed by 1 H NMR, EA, and FT-IR analysis. 1 H NMR Data (500 MHz, CD3 OD) ı ppm: 4.80 (s,

The primary role of ionic liquid based catalysts in the cycloaddition of aziridine with CO2 is to promote the nucleophilic ring opening of aziridine to yield an intermediate that can undergo cycloaddition with CO2 [9,10]. Pinhas et al. [11] previously reported that, generally the cycloaddition of CO2 with 2-aryl aziridines easily led to the regioselective product 5-aryl-2-oxazolidinones, while 2-alkyl aziridines yielded 4-alkyl and 5-alkyl isomers in a ratio of 2:1 ratio. Six membered ring dimers of aziridine, viz., 2,3- and 2,5-dimethyl piperazines were also reported to be generated along with the major products. Generally, chitosan functionalization is synthetically achieved by dispersing chitosan in iso-propanol. However, chitosan has low dispersion in iso-propanol and remains clumped until the final step. To address this issue, we found that chitosan could be better dispersed by using1-methyl-2pyrrolidinone (NMP) as the solvent for synthesis. 3.1. Characterization of chitosan supported pyridinium catalysts Various physicochemical methods such as XRD, XPS, FT-IR, TGA, EA and FE-SEM were used to characterize the synthesized catalysts. However, due to the structural similarity between the various CS-RPX catalysts, only the catalytic features of CS-EPBr will be described in detail. The characteristics of all other CS-RPX catalysts and their precursors are provided in the supporting data. The successful covalent immobilization of the pyridinium IL (EPBr) on the surface of chitosan was confirmed using XPS analysis by comparing the Br 3d spectrum of EPBr, Br-EPBr and CS-EPBr (Fig. 1(a)). The parent IL (EPBr) which contained bromine as Br− anion, only appeared to have a Br 3d peak at 65 eV whereas Br-EPBr, the precursor for the functionalization of chitosan which contained both Br− and CH2 Br in the 1:1 ratio exhibited two Br 3d peaks with the same intensities at 65 eV and 67.5 eV respectively. The catalyst material CS-EPBr only exhibited a peak at 65 eV, conclusively demonstrating the successful tethering of the precursor to the chitosan surface [43]. The Br 3d spectra for the other CS-RPBr catalysts and the Cl 2p spectra of CS-EPCl produced similar peaks and are shown in Fig. S1(a) and (b), respectively, in the supporting data. Further evidence of successful covalent immobilization was obtained by comparing the N 1s spectra of CS, Br-EPBr and CS-EPBr (Fig. 1(b)). The N 1s peak of chitosan was at 396.5 eV, which corresponded to the primary amine group, while the N 1s peak of Br-EPBr for pyridinium N+ was at 399.5 eV [45]. As expected, CS-EPBr contained two major peaks at 397.4 eV and 399.5 eV, which was due to the secondary amine group and pyridinium N+ , respectively. The N 1s band, corresponding to the unreacted amine groups of chitosan, was observed as a weak shoulder peak in the primary amine region at 396.4 eV. The N 1s spectra of various CS-RPX materials and their X-RPX counterparts are shown in Fig. S2 of the supporting data. Further characterization of the bulk CS-RPX catalysts were carried out using the X-ray diffraction technique. Two distinct peaks around 2 values of 10◦ and 19◦ were observed in the XRD pattern of chitosan. These peaks were attributed to the diffraction at the plane of the crystal region in the chitosan structure [44]. These peaks, which resulted from the crystallization capacity of chitosan due to the extensive intramolecular H-bonding between NH2 and OH groups in the repeating hexosaminide residues [45], became less

110

A.C. Kathalikkattil et al. / Applied Catalysis A: General 447–448 (2012) 107–114

Fig. 2. XRD patterns of CS and CS-EPBr catalyst.

Fig. 1. Comparison of XPS data of EPBr, Br-EPBr and CS-EPBr using (a) Br3d and (b) N1s spectra illustrating the successful covalent immobilization of EPBr on chitosan surface.

intense upon covalent immobilization of pyridinium ILs to chitosan. Representative patterns comparing the XRD of CS and CS-EPBr are shown in Fig. 2, while that of the remaining catalysts are depicted in Fig. S3 of the supporting data. CS-RPX compounds gave weak and broad XRD peaks at 2 = 20◦ [45]. The FT-IR spectra of CS and CS-EPBr are shown in Fig. 3. The peaks corresponding to the saccharide structure were found around 1155–1040 cm−1 (C O C in glycosidic linkage) and 890 cm−1 [37,46]. The broad peak around 3500 cm−1 was assigned to both O H as well as N H stretching vibrations [45,46]. The peak around 1660 cm−1 corresponded to the C O stretching, which was observed in both CS and CS-RPX materials. However, the NH bending vibration of the primary NH2 group of chitosan, which appeared at 1600 cm−1 , migrated to 1540 cm−1 due to the formation of the secondary >N H in CS-EPBr [47]. These findings confirm the successful covalent immobilization of EPBr on CS-EPBr. The other CS-RPX catalysts exhibited similar peaks and the results are shown in Fig. S4 of the supporting data. The results of the thermal analysis (TGA) of CS and CS-EPBr are shown in Fig. 4. For CS, the first weight loss around 100 ◦ C resulted from the removal of absorbed water molecules. The further weight loss occurring around 300 ◦ C was likely due to the

Fig. 3. FT-IR spectra of CS and CS-EPBr.

decomposition of the polysaccharide chain in chitosan [40]. CSRPX, including CS-EPBr, was synthesized under dry conditions and showed comparatively less water loss around 100 ◦ C (Fig. S5, supporting data). The decomposition of the polysaccharide chain for all

Fig. 4. Thermogravimetric plots (TGA–DTG) of CS and CS-EPBr.

A.C. Kathalikkattil et al. / Applied Catalysis A: General 447–448 (2012) 107–114 Table 2 Loading of ionic liquids on various CS-RPX catalysts (obtained from elemental analysis data). Entry

Sample

%C

%H

%N

%O

IL loading (mmol g−1 )

1 2 3 4 5 6

CS CS-EPCl CS-EPBr CS-PrPBr CS-BPBr CS-HPBr

40.3 41.86 37.93 37.10 38.73 38.82

6.92 7.31 6.27 6.43 6.82 6.71

7.2 6.18 5.64 5.45 5.24 5.59

42.88 38.10 33.14 34.65 33.37 34.81

– 1.85 2.13 2.05 1.98 1.76

CS-RPX materials occurred at around 260 ◦ C, which was in agreement with the assumptions from the XRD analysis regarding the reduced crystallinity. Based on the elemental analysis of the various CS-RPX materials (Table 2), the loading of various X-RPX ionic liquids, viz., EPCl, EPBr, PrPBr, BPBr and HPBr in the CS-RPX were calculated (based on halides) to be 1.85, 2.13, 2.05, 1.98 and 1.76 mmol per gram of chitosan, respectively. The highest loading was observed for CS-EPBr (Table 2, entry 3). The SEM images of CS-RPX (Fig. S6, supporting data) showed that the material was uniform with a lower crystallinity than pristine CS, which is in agreement with the surface modification analyses of CS-RPX. 3.2. Characterization of the cyclic product, oxazolidinone A very strong absorption band at 1730 cm−1 was observed in the FT-IR spectrum of 4-methyl-1,3-oxazolidin-2-one (Fig. S7, supporting data), which was characteristic of the amide linkage in oxazolidinone and corresponded to the C O stretching vibrations [17,20,48,49]. A secondary amine vibration was also observed at 1484 cm−1 . In addition, the absorption band at 3318 cm−1 corresponded to the N H in the oxazolidinone. The other bands at 2980 and 2936 cm−1 for alkane ( CH2 and CH3 ) stretching vibrations, 1404 cm−1 for alkyl bending vibrations, and 1241 and 1030 cm−1 for C O and C N stretching vibrations further confirmed oxazolidinone formation. 1 H NMR analysis in CD3 OD (Fig. S8, supporting data) also confirmed successful 4-methyl-1,3oxazolidin-2-one formation. ı ppm: 4.80 (s, 1H, NH (d)), 4.51 (t, 1H, CH2 (c )), 3.92–4.02 (m, 2H, CH2 (c) and CH (b)), 1.25 (d, 3H, CH3 (a)). Product formation was further confirmed by elemental analysis and GC–MS (Table S1 and Fig. S9, supporting data). 3.3. Influence of various chitosan-pyridinium (CS-RPX) catalysts on oxazolidinone synthesis In the absence of a catalyst, the reaction of CO2 and aziridine did not afford any product (entry 1). The biopolymer support material chitosan resulted in only trace amounts of MeAz conversion and no 4-methyl-2-oxazolidinone (a) was produced (entry 2). Catalysis of the cycloaddition of 2-methyl aziridine (MeAz) with CO2 was conducted in a batch-wise operation in the presence of 0.71 mol% of the catalyst relative to MeAz at temperatures, CO2 pressures and reaction times varying from 40–100 ◦ C, 0.8 to 2 MPa and 1–5 h at 600 rpm (Table 1). Reactions were performed with the ionic liquid precursor, 1-ethylpyridinium bromide (EPBr, entry 3) and a similar heterogeneous catalyst CS-EMImBr (entry 9) [41] in order to compare the efficiencies of the CS-RPX catalysts. The effect of molecular structure and composition of the CSRPXs on the cycloaddition reactions of CO2 with MeAz are shown in Table 1 (entries 4–8). The nucleophilicity of the anion associated with the CS-RPX was found to play a key role in dictating the catalyst activity. Of the two anions, the more nucleophilic bromide ion of CS-EPBr (entry 4) yielded a conversion of 98% MeAz with a high TON of 196, whereas the conversion and the TON using chloride ion

111

containing CS-EPCl (entry 5) was only 11% and 28, respectively. A high order of regioselectivity (around 94:6) was achieved using CSEPBr, which parallels the high regioselectivity reported by Pinhas et al. with alkyl aziridines [11]. The turnover number obtained with equimolar quantities of EPBr (TON = 155, entry 10) clearly supplements the expected role of synergism played by OH and NH2 groups of CS-EPBr in catalyzing the reaction. In contrast, the alkyl chain length of the pyridinium cation was found to play only a negligible role in the conversion and selectivity of cycloaddition (entries 4, 6–8). However, CS-EPBr was the most active catalyst among them, with highest conversion, yield and TON. To obtain a quantitative information regarding the difference in activity with the imidazolium and pyridinium cation, the cycloaddition reactions of aziridine were performed using a recently reported biopolymer catalyst, CS-EMImBr [41] (CSEMImBr is chitosan immobilized imidazolium ionic liquid (entry 9)). In these experiments, all of the CS-RPBr catalysts displayed a better conversion, selectivity and TON than CS-EMImBr. Based on these combined findings, CS-EPBr was selected for further studies. 3.4. Proposed mechanism of cycloaddition reaction As per the previous reports, the general mechanism of ionic liquid based catalysis for the cycloaddition of alkylaziridine with CO2 commences with attack by the anion at the least hindered ␤-carbon, leading to the ring opening of aziridines [9,10,36]. To rationalize our experimental results of the catalytic activities of CS-RPX catalysts in the cycloaddition of CO2 and MeAz (Scheme 3), we are herewith proposing a plausible mechanism which also involves additional intermediates in the pathway. The N-atom of MeAz binds strongly with the hydroxyl group of chitosan (CS) through both intramolecular N—H· · ·O and H—O· · ·N hydrogen bonding interactions. This strong H-bonding owes to the NH group of MeAz that can behave as intramolecular H-bonding donor as well as acceptor unlike N-substituted aziridines which can act only as a H-bonding acceptor at the heterocyclic N-atom. The halide ion of the bound chitosan generates synchronized nucleophilic attack at the least sterically hindered ␤-carbon atom of MeAz (Scheme 3) paving way to the formation of the ring-opened intermediate 2 [45]. Parallel to this, the secondary NH group of the catalyst interacts with CO2 leading to the carbamate salt 3 formation [41]. It is noteworthy that the potential of such a carbamate salt formed between catalyst and CO2 in synergistically boosting the cycloaddition [41] is unexplored in the cycloaddition of CO2 and aziridines. In the next step, the nucleophilic attack of the N− of 2 on carbamate salt 3 at its carbonyl centre produces the alkyl urethane anion 4 along with the regeneration of the catalyst CS-RPX. Further, the ring closure takes place on the alkyl urethane anion 4 when the O− atom attacks the ␤-carbon atom, so that the halide gets eliminated, resulting in the regeneration of the catalyst and the formation of the product 4-methyl-2-oxazolidinone a. The comparatively disfavored isomer 5-methyl-2-oxazolidinone b is formed upon the nucleophilic attack of halide on the more substituted carbon (␣-carbon) atom. The two other side products, piperazines c and d (Table 1) are formed by the nucleophilic attack of intermediate 2 on the ␤- or ␣-carbon atom of intermediate 1. 3.5. Influence of reaction parameters The effects of reaction parameters, viz., catalyst amount, reaction time, temperature, and CO2 pressure on the cycloaddition of MeAz with CO2 were studied using CS-EPBr as the catalyst (Fig. 5). The conversion rate of MeAz which was low at a catalyst amount of 0.18 mol% was found to increase when the catalyst amount was increased up to 0.71 mol%. However, a further increase in the catalyst amount did not result in any significant increase in the

112

A.C. Kathalikkattil et al. / Applied Catalysis A: General 447–448 (2012) 107–114

Scheme 3. Proposed reaction mechanism for the cycloaddition of 2-methylaziridine with CO2 .

yield of the major product 4-methyl-2-oxazolidinone (a) (Table 1), which may be ascribed to the hindrance in mass transfer occurring between active site and reagent resulting from the low dispersity of excess catalyst in the reaction mixture [50]. The influence of catalyst amount on the cycloaddition of CO2 with MeAz is shown in Fig. 5. The 0.71 catalyst mol% is the lowest reported for MeAz–CO2 cycloaddition to date [29,31]. Moreover, this is one of the lowest reported catalyst mole percentages for heterogeneous catalysts used in the cycloaddition of aziridines and CO2 (including activated aziridines). A study on the effect of reaction time on CS-RPX catalyzed cycloaddition is depicted in Fig. 6. The conversion increased with an increase in the reaction time and a 98.3% conversion was obtained at 4 h. However, there was only a slight increase in the yield of 4methyl-2-oxazolidinone (a) (Table 1) at reaction times longer than 4 h. We also investigated the effect of temperature (40–100 ◦ C) on cycloaddition. As illustrated in Fig. 7, temperature had a positive effect on the yield of oxazolidinone. A steep increase in the MeAz conversion and yield of a was observed when the temperature was increased from 40 ◦ C to 100 ◦ C and the highest conversion of 98.3% was observed at 100 ◦ C. However, a further increase in the temperature did not have a significant effect on the reaction and only a slight increase in the selectivity of 4-methyl-2-oxazolidinone (a)

Fig. 6. Effect of reaction time on MeAz conversion and selectivity of (a).

Fig. 7. Effect of reaction temperature on MeAz conversion and selectivity of (a).

Fig. 5. Dependence of CS-EPBr catalyst amount on the conversion of MeAz and selectivity of 4-methyl-2-oxazolidinone (a).

was observed at higher temperatures, which was presumably due to the lower solubility of the CO2 gas phase in the system [50]. Thus, the optimum temperature for this process was determined to be 100 ◦ C.

A.C. Kathalikkattil et al. / Applied Catalysis A: General 447–448 (2012) 107–114

Fig. 8. Effect of CO2 pressure on MeAz conversion and selectivity of (a).

113

Fig. 9. Catalyst recycle test.

4. Conclusions The dependence of the cycloaddition reactions on CO2 pressure was also studied between 0.8 MPa and 2.4 MPa (Fig. 8). Lower CO2 pressures of the order of 0.8 MPa gave low conversions with a selective production of the side products, piperazines c and d (Table 1). The formation of low energy chitosan-CO2 complex 3 (Scheme 3) has a direct dependence on the available amount of CO2 , whereas the H-bonded intermediates 1 and 2 (Scheme 3) are independent of CO2 concentrations. Therefore at low CO2 pressures, the formation of chitosan-CO2 complex 3 was retarded, whereas the increased chance for nucleophilic attack of intermediate 2 on the ␤- or ␣carbon atom of intermediate 1 lead to the dimerized aziridines, viz., 2,5- and 2,3-dimethyl piperazines, c and d respectively. At higher CO2 pressures (1.2–2.4 MPa), where more of intermediate 3 (Scheme 3) might have been formed, the reaction completes with 4-methyl-2-oxazolidinone (a) as the major product. However, beyond 2 MPa CO2 pressures, there was no significant rise in the yield, probably due to the effect of dilution by excessive quantities of CO2 [50]. In order to establish the effect of a continuous supply of CO2 at lower pressures, we performed a reaction at 0.8 MPa using CS-EPBr under semi-batch conditions (entry 10). It was observed that the conversion and yield (a) and TON increased to >99%, 93.4 and 199 respectively under these conditions.

We initiated biopolymer based material as a cheap, recyclable and environmentally benign catalyst for the synthesis of oxazolidinone by the cycloaddition of aziridine with CO2 . The pyridinium halide functionalized chitosan based biopolymer catalyst (CS-RPX) in the present study offered a synergistic mechanism for the cycloaddition process by means of hydroxyl, amine and halide anion which in turn eased the successful cycloaddition of an unactivated alkyl aziridine with CO2 under mild reaction conditions and a low catalyst mol percentage of 0.71% with respect to the substrate. Moreover, CS-RPX catalysts are the first biopolymer based catalysts used in the cycloaddition of aziridines with CO2 and exploits the H-bonded assisted low energy pathways without performing a synthetic modification to achieve cooperating functional groups. A high turnover number (TON) was achieved with the catalyst and the optimum reaction conditions were 4 h, 100 ◦ C, 2 MPa CO2 pressure and 0.2 mmol catalyst. A further increase in the yield and TON was obtained even at 0.8 MPa under similar reaction conditions in a semi-batch operation. In addition, the catalyst could be reused up to five times by simple filtration without considerable loss in the activity and yield. Acknowledgements

3.6. Reusability of the catalyst It is reported previously that ionic liquids used in the cycloaddition of aziridines and CO2 have been recycled up to four times without significant loss of activity or selectivity [10]. However, the ionic liquid catalyst was reported to be recovered after the product was separated out by distillation under reduced pressure from the reaction mixture, which is both economically and technologically unviable. In contrast, the heterogeneous catalyst CS-EPBr in our present study was recovered easily by simple filtration from the crude mixture after the reaction. The recovered catalyst was washed with dichloromethane, dried at 80 ◦ C for 24 h and reused for the cycloaddition of MeAz and CO2 under similar reaction conditions. Fig. 9 shows the reusability of the CS-EPBr catalyst up to five cycles. The conversion of MeAz was 98.3% and selectivity of a was 89.8% when using the fresh catalyst. The catalyst was found to remain active up to the fifth cycle, with a slow decrease in conversion and selectivities during each cycle. The final (fifth) recycle of the catalyst resulted in a MeAz conversion of 85.3% and selectivity of 82.7% a.

This study was supported by the Korean Ministry of Education, Science and Technology through the National Research Foundation (2012-001507) and the Brain Korea 21 Project. The authors are grateful to KBSI for performing characterization studies. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.apcata.2012.09.031. References [1] H. Arakawa, M. Aresta, J.N. Armor, M.A. Barteau, E.J. Beckman, A.T. Bell, J.E. Bercaw, C. Creutz, E. Dinjus, D.A. Dixon, K. Domen, D.L. DuBois, J. Eckert, E. Fujita, D.H. Gibson, W.A. Goddard, D.W. Goodman, J. Keller, G.J. Kubas, H.H. Kung, J.E. Lyons, L.E. Manzer, T.J. Marks, K. Morokuma, K.M. Nicholas, R. Periana, L. Que, J. Rostrup-Nielson, W.M.H. Sachtler, L.D. Schmidt, A. Sen, G.A. Somorjai, P.C. Stair, B.R. Stults, W. Tumas, Chem. Rev. 101 (2001) 953–996. [2] T. Sakakura, J.C. Choi, H. Yasuda, Chem. Rev. 107 (2007) 2365–2387. [3] J. Sun, J. Ren, S. Zhang, W. Cheng, Tetrahedron Lett. 50 (2009) 423–426. [4] L. Han, S.W. Park, D.W. Park, Energy Environ. Sci. 2 (2009) 1286–1292.

114

A.C. Kathalikkattil et al. / Applied Catalysis A: General 447–448 (2012) 107–114

[5] D.J. Darensbourg, S.J. Lewis, J.L. Rodgers, J.C. Yarbrough, Inorg. Chem. 42 (2003) 581–589. [6] Z.Z. Yang, L.N. He, J. Gao, A.H. Liu, B. Yu, Energy Environ. Sci. 5 (2012) 6602–6639. [7] J. Seayad, A.M. Seayad, J.K.P. Ng, C.L.L. Chai, ChemCatChem 4 (2012) 774–777. [8] Z.Z. Yang, Y.N. Li, Y.Y. Wei, L.N. He, Green Chem. 13 (2011) 2351–2353. [9] Y. Wu, G. Liu, Tetrahedron Lett. 52 (2011) 6450–6452. [10] Z.Z. Yang, L.N. He, S.Y. Peng, A.H. Liu, Green Chem. 12 (2010) 1850–1854. [11] C. Phung, A.R. Pinhas, Tetrahedron Lett. 51 (2010) 4552–4554. [12] R.A. Watile, D.B. Bagal, K.M. Deshmukh, K.P. Dhake, B.M. Bhanage, J. Mol. Catal. A: Chem. 351 (2011) 196–203. [13] L. Aurelio, R.T.C. Brownlee, A.B. Hughus, Chem. Rev. 104 (2004) 5823–5846. [14] T.M. Makhtar, G.D. Wright, Chem. Rev. 105 (2005) 529–542. [15] R.E. Gawley, S.A. Campagna, M. Santiago, T. Ren, Tetrahedron Asymmetry 13 (2002) 29–36. [16] M. Prashad, Y.G. Liu, H.Y. Kim, O. Repic, T.J. Blacklock, Tetrahedron Asymmetry 10 (1999) 3479–3482. [17] S. Neffgen, H. Keul, H. Höcker, Macromol. Rapid Commun. 17 (1996) 373–382. [18] L. Aurelio, R.T.C. Brownlee, A.B. Hughus, Chem. Rev. 104 (2004) 5823–5862. [19] M.R. Barbachyn, C.W. Ford, Angew. Chem. Int. Ed. 42 (2003) 2010–2023. [20] K. Soga, S. Hosoda, H. Nakamura, S. Ikeda, J. Chem. Soc. Chem. Commun. (1976) 617. [21] W.J. Close, J. Am. Chem. Soc. 73 (1951) 95–98. [22] D. Ben-Ishai, J. Am. Chem. Soc. 78 (1956) 4962–4965. [23] L.V.J. Ciula, O.W. Gooding, Org. Process Res. Dev. 7 (2003) 514–520. [24] M. Costa, G.P. Chiusoli, M. Rizzardi, Chem. Commun. (1996) 1699–1701. [25] M. Shi, Y.M. Shen, J. Org. Chem. 67 (2002) 16–21. [26] M. Feroci, M. Orsini, G. Sotgiu, L. Rossi, A. Inesi, J. Org. Chem. 70 (2005) 7795–7798. [27] Y. Kayaki, M. Yamamoto, T. Suzuki, T. Ikariya, Green Chem. 8 (2006) 1019–1021. [28] A.W. Miller, S.T. Nguyen, Org. Lett. 6 (2004) 2301–2304. [29] H. Kawanami, H. Matsumoto, Y. Ikushima, Chem. Lett. 34 (2005) 60–61.

[30] M.T. Hancock, A.R. Pinhas, Tetrahedron Lett. 44 (2003) 5457–5460. [31] A. Sudo, Y. Morioka, E. Koizumi, F. Sanda, T. Endo, Tetrahedron Lett. 44 (2003) 7889–7891. [32] Y. Wu, L.N. He, Y. Du, J.Q. Wang, C.X. Miao, W. Li, Tetrahedron 65 (2009) 6204–6210. [33] T. Sudo, Y. Morioka, F. Sanda, T. Endo, Tetrahedron Lett. 45 (2004) 1363–1365. [34] Y.M. Shen, W.L. Duan, M. Shi, Eur. J. Org. Chem. (2004) 3080–3089. [35] H.F. Jiang, J.W. Ye, C.R. Qi, L.B. Huang, Tetrahedron Lett. 51 (2010) 928–932. [36] Y. Du, Y. Wu, A.-H. Liu, L.N. He, J. Org. Chem. 73 (2008) 4709–4712. [37] Z. Guo, R. Xing, S. Liu, Z. Zhong, P. Li, Carbohydr. Polym. 73 (2008) 173–177. [38] P. Madhavan, K.G.R. Nair, Fish. Technol. 11 (1974) 50–53. [39] R. Moucel, K. Perrigaud, J.M. Goupil, P.J. Madec, S. Marinel, E. Guibal, A.C. Goumont, I. Dez, Adv. Synth. Catal. 352 (2010) 433–439. [40] H. Xie, S. Zhang, S. Li, Green Chem. 8 (2006) 630–633. [41] J. Sun, J. Wang, W. Cheng, J. Zhang, X. Li, S. Zhang, Y. She, Green Chem. 14 (2012) 654–660. [42] D. Tanner, Angew. Chem. Int. Ed. Engl. 33 (1994) 599–619. [43] L. Cen, K.G. Neoh, L. Ying, E.T. Kang, Surf. Interface Anal. 36 (2004) 716–719. [44] S. Lu, X. Song, D. Cao, Y. Chen, K. Yao, J. Appl. Polym. Sci. 91 (2004) 3497–3503. [45] J. Tharun, Y. Hwang, R. Roshan, S. Ahn, A.C. Kathalikkattil, D.W. Park, Catal. Sci. Technol. 2 (2012) 1674–1680. [46] S.S. Silva, M.I. Santos, O.P. Coutinho, J.F. Mano, R.L. Reis, J. Mater. Sci. – Mater. Med. 16 (2005) 575–579. [47] J. Coates, R.A. Meyers (Eds.), Encyclopedia of Analytical Chemistry, John Wiley & Sons Ltd, Chichester, UK, 2000, pp. 10815–10837. [48] Q.W. Lu, T.R. Hoye, C.W. Macosko, J. Polym. Sci., Part A: Polym. Chem. 40 (2002) 2310–2328. [49] T. Endo, M. Okawara, Die Makrornolekulare Chemie 146 (1971) 237–245. [50] J. Sun, W. Cheng, W. Fan, Y. Wang, Z. Meng, S. Zhang, Catal. Today 148 (2009) 361–367.