Catalysis Letters

, Volume 137, Issue 1, pp 1–7

Cycloaddition of CO2 to Epoxides Using a Highly Active Co(III) Complex of Tetraamidomacrocyclic Ligand

Authors

    • Department of ChemistryUniversity of Arkansas at Little Rock
  • Punnamchandar Ramidi
    • Department of ChemistryUniversity of Arkansas at Little Rock
  • Sharon Pulla
    • Department of ChemistryUniversity of Arkansas at Little Rock
  • Shane Z. Sullivan
    • Department of ChemistryUniversity of Arkansas at Little Rock
  • Samuel L. Collom
    • Department of ChemistryUniversity of Arkansas at Little Rock
  • Yashraj Gartia
    • Department of ChemistryUniversity of Arkansas at Little Rock
    • Reliance Industries LimitedResearch Center
  • Alexandru S. Biris
    • Department of Applied Science and Nanotechnology CenterUniversity of Arkansas at Little Rock
  • Bruce C. Noll
    • Bruker AXS Inc.
  • Brian C. Berry
    • Department of ChemistryUniversity of Arkansas at Little Rock
Article

DOI: 10.1007/s10562-010-0325-0

Cite this article as:
Ghosh, A., Ramidi, P., Pulla, S. et al. Catal Lett (2010) 137: 1. doi:10.1007/s10562-010-0325-0

Abstract

Synthesis of various cyclic carbonates with yield up to 100% and turn over frequency (TOF) of 351 h−1 using CO2 and epoxides and a cobalt (III) complex of tetraamidomacrocyclic ligand is described. The catalyst was characterized by single crystal X-ray crystallography. A study of reaction conditions indicates that 2 MPa pressure of CO2 without any co-solvent is sufficient to achieve the desired product.

Graphic Abstract

 
https://static-content.springer.com/image/art%3A10.1007%2Fs10562-010-0325-0/MediaObjects/10562_2010_325_Figa_HTML.gif

Keywords

Cyclic carbonatesCO2-epoxidesCo(III)-Tetraamidomacrocyclic ligand catalystCo-catalysts

1 Introduction

Carbon dioxide (CO2), a common green house gas, is produced in large quantities by human activities in addition to other natural phenomena. In recent years, efforts have been underway to reduce CO2 emissions. Then again, carbon dioxide is an abundant, environmentally benign, and renewable source of carbon that can be used for the synthesis of organic molecules. The use of CO2 may potentially end the utilization of toxic C1 synthons, such as phosgene, carbon monoxide, and isocyanate [17]. However, the extreme inertness of CO2 presents a great difficulty for its use in organic synthesis. Developing a better knowledge of the coordination chemistry of CO2 to transition metal centers might provide us with an understanding of the possibilities of incorporating CO2 into discrete organic monomers. In this context, studying the reaction of aliphatic or aromatic epoxides with CO2 to produce cyclic carbonate in an environmentally friendly method is an important area of research (Fig. 1). Cyclic carbonates, are colorless, odorless, and biodegradable chemicals which have found several uses as polar solvents, electrolytic materials, and precursors for the synthesis of many pharmaceutically important and fine chemicals as well as for engineered polymeric materials [812].
https://static-content.springer.com/image/art%3A10.1007%2Fs10562-010-0325-0/MediaObjects/10562_2010_325_Fig1_HTML.gif
Fig. 1

Cyclic carbonate synthesis and the structure of the Cobalt catalyst (1) used for the current study

Both homogeneous and heterogeneous reaction conditions were employed to develop cyclic carbonates. For their industrial production, reactions are generally catalyzed by metal or organic halide salts, such as potassium iodide or tetraethylammonium bromide [13, 14]. Over the years, numerous homogeneous catalysts have been reported in the literature for their ability to synthesize cyclic carbonates with varying yields and TOFs. These include quaternary ammonium salts [15], alkali and alkaline earth metal salts [16, 17], tin and antimony compounds [18], various transition metal salts [1921], and ionic liquids [22] etc.

In order to increase the selectivity, yields and reaction rates for catalytic cyclic carbonate synthesis, several metal complexes of well-defined ligands that act as Lewis acids along with Lewis base co-catalysts were introduced by various groups. For example, metal porphyrins and phthalocyanines are attractive catalysts since they are cheap, easily available, and highly reactive catalysts [2327]. Metal complexes of salen ligands are other types of well defined metal catalysts which have been studied extensively for the activation of CO2 for synthesizing polycarbonates [7, 28] and cyclic carbonates. Among many salen complexes, those derived from aluminum, nickel, copper, zinc, magnesium, cobalt, and chromium are noteworthy [2934]. In terms of reactivity, aluminum and chromium salen complexes are very active in producing the cyclic carbonates. M. North and co-workers developed one of the most highly reactive catalysts that can operate at relatively mild reaction conditions. This catalyst is a μ-oxo-dimeric Al-salen complex which produces cyclic carbonates in high yields at a low temperature and 1 atm of CO2 pressure [30, 31].

The synthesis of cyclic carbonates from CO2 and similar epoxides has been applied on an industrial scale and a number of the works cited above report on the catalyst development and its reaction mechanisms [35]. However, the activity, stability, and recovery of the catalysts still need to be improved. In many cases, metal-based homogenous catalysts, besides being active at high pressure and temperature, are moisture sensitive (salen complexes) [36] and oxygen (low-valent metal complexes) [37]. These catalysts (metal porphyrin and phthalocyanine complexes) agglomerate with each other and need to be used in large quantities and require co-solvents to achieve good activity [37]. Heterogeneous catalysts have the inherent problem of lower activity and they also need the use of co-solvents [37]. Considering these facts, there is still room to develop novel and efficient homogenous catalysts for synthesizing cyclic carbonates under relatively mild reaction conditions.

In this present work, we report the use and detailed study of a cobalt-tetraamidomacrocyclic complex (1, Fig. 1) as a new catalyst for the synthesis of aromatic and aliphatic cyclic carbonates in excellent yields, high selectivity, and fast reaction rates. Structurally the catalyst resembles porphyrin or salen complexes but has four deprotonated N atoms which act as strong σ donor ligands to coordinate to Co(III) ion. Although similar complexes have been known for some time [3840], but no catalytic activity has been tested using such complexes in cyclic carbonate synthesis. We inferred that the catalyst would provide the following advantageous characteristics: (a) much more robustness than any salen complexes, especially hydrolytic stability; (b) possesses planar four coordinate geometry (as revealed by the crystal structure of the cobalt complex), and thus leaves two axial positions still available for binding, a crucial feature for allowing the docking of epoxide on the Lewis acidic metal center and facilitate the reaction with carbon dioxide; (c) can easily be synthesized; (d) is uncreative towards oxygen and moisture, and thus can be handled without any specialized equipment, and (e) shows no agglomeration (unlike metal phthalocyanine or porphyrin complexes) and has flexible solubility including in many epoxides. Therefore, the use of 1 may be beneficial to investigate in detail for the synthesis of cyclic carbonates.

2 Experimental

2.1 General

Chemicals were purchased from Aldrich chemical company, USA and Acros chemical company and used without further purifications unless otherwise mentioned. Propylene epoxide, cyclohexene epoxide were purified using CaH2. CO2 was obtained from Airgas (99.9%). All high pressure reactions were carried out in 100 mL Parr reactor connected to a 4842 controller. 1H and 13C-NMR spectra were obtained using 200 and 600 MHz Bruker instruments equipped with a 5 mm triple resonance inverse probe. The spectra were collected at 25 °C and chemical shifts are in ppm relative to TMS as external standard unless otherwise stated. Infrared spectra were obtained using a Shimadzu FT-IR Affinity-1 spectrophotometer. Electrospray ionization mass spectra (ESI-MS) were obtained using an Agilent 100 series MSD VL spectrometer. Gas Chromatography Mass spectra (GC/MS) were obtained using an Agilent technologies 6890N network GC system and equipped with Agilent Technologies 5975 inert XL mass selective detector. Elemental analysis was performed at Midwest Microlab LLC., Indianapolis.

2.2 Synthesis of Catalyst

The details of catalyst synthesis and characterizations are presented in the supplementary information. Both lithium and tetraphenylphosphonium salts of the catalyst were used in the reaction and characterization purposes as required.

2.3 Synthesis of Cyclic Carbonates

A representative method for synthesis of cyclic carbonate (Propylene carbonate (PC)) using propylene oxide (PO) and co-catalyst 4-dimetheylaminopyridine (DMAP) is described below. Propylene Oxide (2.08 g, 0.036 mol), 1 (15 mg, 0.035 mmol) and co-catalyst DMAP (8.6 mg, 0.07 mmol) were added into a 100 mL stainless steel Parr high pressure reactor. The reactor was pressurized with CO2 (Purity 99.9%, Airgas) and the temperature of the reactor was quickly increased and maintained at 120 °C. The final pressure of the reaction vessel was read as 2 MPa. The reaction mixture was allowed to stir in these conditions for 3 h. The reactor was quickly brought to a low temperature using an ice water bath and pressure was released slowly. The reaction mass was collected using dichloromethane for a complete transfer and the catalyst was separated by passing through a small plug of silica gel (2 g, 60–100 mesh) and eluting with dichloromethane. The solvent and excess epoxide were removed at reduced pressure. The yield was calculated with respect to the authentic sample by the external standard method from the GC peak. The isolated yield was calculated by taking the weight of the isolated product. The product was characterized by IR, 1H-NMR and GC-MS. The reaction yield was also calculated at different time intervals.

In accordance to the above-mentioned conditions, the amount of DMAP was varied from 1 to 2.5 mol equivalents with respect to 1 while the effect of the co-catalyst was studied. The pressure of CO2 was optimized at 2 MPa based upon the experiments carried in the range of 0.068–4.8 MPa. Similarly, the study of temperature was executed in the range of 70–120 °C.

After filtration through a short column of silica and washing with dichloromethane the silica adsorbed catalyst was further used to perform the synthesis of PC. Reusability tests were performed for 3 times consecutively.

3 Results and Discussion

3.1 Catalyst Synthesis and Structure Determination

The tetraamidomacrocyclic ligand and its cobalt complex (1, Fig. 1) were synthesized following the literature procedure of similar cobalt complexes with readily available starting materials (see supporting information for details) [41]. The tetraamidomacrocyclic ligand was characterized by 1H-NMR, electrospray ionization mass spectrometer (ESI-MS, m/z 373.3 (M-H+, 100)), infrared spectroscopy (IR) and elemental analysis. Using n-butyllithium as a strong base, under an inert atmosphere, the ligand was deprotonated and reacted with cobalt(II) chloride in tetrahydrofuran. Finally, the desired cobalt(III) complex was obtained under air exposure. The cobalt complex is negatively charged and both the lithium and the tetraphenylphosphonium versions of the cobalt complex were obtained. The tetraphenylphosphonium salt was obtained by the metathesis of 1 with tetraphenylphosphonium chloride in water. ESI-MS, elemental, UV/Vis studies were performed to characterize the cobalt complex. The cobalt complex was found to be highly air and moisture stable. In fact, a high hydrolytic stability of the complex was observed by dissolving the complex in dilute sulphuric acid solution and finally analyzing the solution using ESI-MS and UV/Vis studies. All these properties are suitable for achieving a robust homogeneous catalyst for cyclic carbonate synthesis.

Further characterization of the cobalt complex was performed by single-crystal X-ray crystallographic analysis. X-ray crystallographic quality crystals were obtained using 1 (with tetraphenylphosphonium ion as a counter cation) from a solvent mixture of acetonitrile and water (1:1). Figure 2 shows the ORTEP structure of 1. The tetraphenylphosphonium ion is not shown in the structure. Details of X-ray crystallographic studies, including bond angles and lengths, are given in the supporting information. X-ray crystallographic studies demonstrate that Co(III) in the monoclinic crystal, is in a square-planar environment comprised of 4 N atoms of the deprotonated ligand moiety. The average Co–N bond distances in 1 is 1.828 (3) Å, which is comparable to the reported distances of similar cobalt(III) complexes [36, 37]. The average N–Co–N bond angle is 90.01º which reflects the square-planarity formed by deprotonated N donor atoms of the ligand and Co(III) atoms. However, differences in the N–Co–N angle in the five-membered near-phenylinic moiety (84.9º) compared to the ring formed by malonic moiety (101.4º) can be seen. On the other hand, the Co–N distance of phenylinic nitrogens (1.822 Å) is slightly shorter than the Co–N (1.834 Å) bonds, which are attached to malonic moiety nitrogen. The shorter distance in case of phenylinic nitrogen indicates high bond order due to the back bonding originated from the aromatic ring. It is notable that the crystal structure shows a planar 4-coordinate cobalt complex that provides open axial positions for access to the epoxides for activation during catalysis. The strong donation capacity of deprotonated amide ligands may be responsible for providing the Co center optimum electrophilicity for the reaction without being inhibited by any epoxides.
https://static-content.springer.com/image/art%3A10.1007%2Fs10562-010-0325-0/MediaObjects/10562_2010_325_Fig2_HTML.gif
Fig. 2

X-ray crystallographic structure of the Co-complex (1) for the present study. The counter cation tetraphenylphosphonium part is not shown. Selected bond lengths and angles are given in the supporting information

3.2 Catalytic Activity

The catalyst and co-catalyst were found to be soluble in all epoxides thus no co-solvents were used for this reaction. Cyclic carbonates were found to form from CO2 and epoxides using 0.1 mol% catalyst and 0.2 mol% of co-catalyst with respect to epoxides. The required pressure of CO2 was found to be 2 MPa, which is notably lower than many literature-reported values for this reaction. A blank reaction at similar conditions in absence of the catalyst and co-catalyst showed little product formed. This not only indicates the need for both catalyst and co-catalyst to be present for an optimum activity but also that the reactor has no influence in the reaction. The products were confirmed by respective molecular ion peaks in GC/MS. The formation of the product was also supported by the broad infra-red peak at ~1800 cm−1 (see supplementary information). The absence of any other carbonate peaks in the IR spectra eliminates the possible formation of polycarbonate during the reaction and thus indicating high selectivity. Gas chromatrography of the reaction mixture shows the presence of the product (RT 6.6 min, PC) as the sole component other than epoxide (RT 1.2 min, PO), suggesting a high reaction selectivity. The reaction yield was also calculated at different time intervals.

When PC was synthesized from CO2 and PO, the co-catalyst was observed to have a strong role on the activity as shown in Table 1. Low product yield was observed in the absence of any co-catalyst. We also verified that the co-catalyst itself does not show any reactivity under the reaction conditions. From Table 1, it is evident that DMAP has the highest influence on PC production among the various co-catalysts as indicated by TOF of 351 h−1. Other co-catalysts were also found to achieve significant activity: tetraphenylphosphonium bromide was found to show a TOF of almost 300 h−1 whereas DMAP was identified as the best co-catalyst for the reaction. Further studies were performed using DMAP along with 1. Methyl imidazole and pyridine were observed to have much lower activities when compared to DMAP and tetraphenylphosphonium bromide. This is due to low Lewis basicity of the amine or anion. In other words, a strong basic nature of co-catalyst is required to obtain enhanced activity.
Table 1

Formation of PC from PO and CO2 using 1 and various co-catalysts

Entry

Catalyst, co-catalysta

Efficiency

Yieldb (%)

Isolated yieldc (%)

TOF (h−1)

1

1

8.90

8.60

31

2

4-Dimethylamino pyridine (DMAP)

4.00

3.85

14

3

1/DMAP

100

98.9

351

4

1/Pyridine

50.7

49.1

177

5

1/Triethylamine

64.6

63.2

226

6

1/Tetraphenyl phosphonium bromide

84.6

 

296

7

1/1-methyl imidazole

 

53.2

186

PC propylene carbonate, PO propylene oxide. CO2 2 MPa, Temp 120 °C, Time 3 h

aCatalyst:co catalyst = 1:2 mol ratio, External standard method in gas chromatography, Isolated yield, measured as per physical weight of product recovered

The concentration of co-catalysts, namely DMAP, was found to have a profound effect in the TOF for PC production. Figure 3 shows the effect of DMAP to catalysts ratio on PC production. It is evident from the graph that, with increasing DMAP concentration, the PC yield increases and reached its maximum when DMAP:1 is 2:1. However further increase in DMAP concentration reduces the activity of 1 which shows a similar trend to the activity of Co-salen or porphyrin complexes with increasing co-catalyst concentration [25]. The reduction of activity may arise from excess DMAP co-ordinately saturating the catalyst binding positions, thus leaving no room for the binding of epoxide to the catalyst for further activity.
https://static-content.springer.com/image/art%3A10.1007%2Fs10562-010-0325-0/MediaObjects/10562_2010_325_Fig3_HTML.gif
Fig. 3

TOF of propylene carbonate synthesis for the 1/DMAP system as a function of DMAP equivalents. Reaction conditions: Temperature 120 °C, CO2 pressure 2 MPa

Studies of various epoxides, including aliphatic and aromatic epoxides, show that the lower epoxide, in particular propylene oxide, gives a significant yield as shown in the Table 2. Some of the sterically hindered epoxides, such as styrene oxide and cyclohexene oxide also produce their corresponding cyclic carbonates in good yields and turn over frequencies. It can be noted that the catalyst is equally efficient with aromatic as well as aliphatic epoxides. In case of cyclooctene oxide almost no product was detected when the reaction was run at 120 °C for 3 h. However a TOF of 47 h−1 was obtained for the formation of cyclooctene carbonate when the reaction was run at 180 °C for 8 h. Poor activity in case of cyclooctene oxide can be correlated to its sterically hindered and puckered structure which hinders the coordination of the epoxide to 1.
Table 2

Formation of cyclic carbonates using different epoxides

Entry

Epoxides

Yield (%)a

TOF (h−1)

1

Propylene oxide

~100

351

2

Epichlorohydrine

76.2

261

3

Styrene oxide

85.8

294

4

Cyclohexene oxide

72.6

249

5

Cyclooctene oxideb

76.2

47

Catalyst = 1, Co-catalyst = DMAP, 1: DMAP = 1:2, CO2 2 MPa, Temp 120 °C, Time 3 h

aYield = isolated, Temp 180 °C and Time 8 h

In order to understand the effect of pressure on activity, we have also tested the activity of the catalyst with varying CO2 pressure. Figure 4 shows the effect of CO2 pressure on the TOF of propylene carbonate production. With increasing CO2 pressures, the TOF increases and reaches a maximum at a CO2 pressure of 2 MPa. TOF of 312 h−1 at 0.7 MPa indictates that the catalyst retains sizable activity even at low pressure. Further increase in pressure from 2 MPa diminishes the overall activity of the catalyst. The decrease in activity is possibly due to the fact that with increasing CO2 pressure, the polarity of the reaction mixture (Carbon dioxide expanded liquids) decreases, which most likely lowers the solubility of the catalyst in the reaction mixture [9]. A similar pressure effect was also observed previously by researchers [25].
https://static-content.springer.com/image/art%3A10.1007%2Fs10562-010-0325-0/MediaObjects/10562_2010_325_Fig4_HTML.gif
Fig. 4

Turnover frequency (h−1) of propylene carbonate synthesis at different pressures. Reactions were performed at 120 °C for 3 h while maintaining a 1:DMAP ratio of 1:2

The activity of the catalyst with DMAP was checked at different temperatures for the synthesis of PC. Figure 5 shows the % yield of PC production at various temperatures. The percentage yield increases with an increase in temperature and at 120 °C at a reaction time of 3 h, it was found to have completely converted to the product. This finding compelled us to choose 120 °C as the standard temperature to use for the rest of the reactions.
https://static-content.springer.com/image/art%3A10.1007%2Fs10562-010-0325-0/MediaObjects/10562_2010_325_Fig5_HTML.gif
Fig. 5

Percentage yield of propylene carbonate at different temperatures at 2 MPa, 3 h, with 1: DMAP ratio of 1:2

We also measured the % yield PC production at different time intervals to understand how much time is required at a particular temperature. Figure 6 shows the results at 100 °C and at 120 °C. The reaction completes within 3 h at 120 °C whereas at 100 °C it takes slightly longer for the reaction to complete. Therefore we established our optimum reaction time and temperature of 3 h at 120 °C respectively and at 2 MPa of CO2 pressure based on the above-mentioned findings.
https://static-content.springer.com/image/art%3A10.1007%2Fs10562-010-0325-0/MediaObjects/10562_2010_325_Fig6_HTML.gif
Fig. 6

Effect of temperature on propylene carbonate yield. Reactions were run at 2 MPa of CO2 pressure for 3 h maintaining 1 and DMAP ratio of 1:2

We attempted to reuse the catalyst after adsorbing onto silica gel. Though a diminishing trend in efficiency after each recycle (TONrecycle : 5701st, 4892nd and 3013rd) is observed, the selectivity of the product remained unaltered indicating a retention of identity of the catalyst.

3.3 Mechanism

As revealed by the X-ray crystallographic structure, the catalyst is a square-planar complex with no axial ligands attached to the cobalt complex. Even though the cobalt complex is a 4-coordinate complex, a weak interaction between the Co complex and any neutral ligands like DMAP cannot be ruled out, similar to many Co-salen or Co-porphyrin complexes. Thus, the mechanism of this reaction using a cobalt-based catalyst can be envisioned as being very similar to mechanisms previously proposed by Inoue [24] and more closely by Nguyen and co-workers [25, 32]. Figure 7 shows the 1H and 13C-NMR of cyclohexene carbonate. As revealed by the NMR spectra, there is no split for the 1-C, 3-C (δ = 75.67 ppm) and 1-H, 3-H (4.71–4.77 ppm). These observations indicate that 1-C, 3-C, 1-H, and 3-H are in same chemical environment which confirms the production of cis isomer of the cyclohexene carbonate exclusively [42]. Mechanistically NMR results imply that the epoxide ring opening as well as the ring closure to produce the cyclic carbonate on 1 is taking place under stereo-controlled fashion. Thus based on these information and study of reaction conditions, it can be concluded that after epoxide ring opening and insertion of CO2, the active species undergoes cyclilization to give rise to the cyclic carbonates as shown in the Fig. 8.
https://static-content.springer.com/image/art%3A10.1007%2Fs10562-010-0325-0/MediaObjects/10562_2010_325_Fig7_HTML.gif
Fig. 7

The 1H, 13C (inset) NMR spectra of Cyclohexyl-[1, 3]-dioxolan-2-one (Cyclohexene carbonate). Reaction was run at 2 MPa of CO2 pressure for 3 h maintaining 1 and DMAP ratio of 1:2

https://static-content.springer.com/image/art%3A10.1007%2Fs10562-010-0325-0/MediaObjects/10562_2010_325_Fig8_HTML.gif
Fig. 8

Proposed mechanism of the reaction

4 Conclusion

In conclusion, we have found that the cobalt complex of tetraamidomacrocyclic ligand is an efficient catalyst for the production cyclic carbonate at TOF of 351 h−1. The catalyst is air stable, easily synthesized and environmentally benign metal complex and found to be active to produce cyclic carbonates without any added co-solvent. The catalyst is capable of producing cyclic carbonates using epoxide and CO2 at 120 °C and 2 MPa of pressure. Efficiency to form cyclic carbonates was studied using both aromatic and aliphatic epoxides and the catalyst showed activity even at relatively low CO2 pressure (~0.7 MPa). The effectiveness of the catalyst is believed to be governed by the unique electronic and geometric properties offered by the tetraamidomacrocyclic ligand. Although the outcome of the catalyst recovery method is not satisfactory, the retention of product selectivity leaves us further opportunity to develop alternative techniques. Further studies are needed for designing other catalysts for better efficiency and recovery to synthesize cyclic carbonate and polycarbonates.

Acknowledgments

AG likes to thank UALR faculty start up grant and Department of Energy (Grant number DE-FG36-06GO86072) for financial assistance to complete the work. AG and SLC also like to thank Arkansas State Technology Authority for a summer scholarship (09-EPSCoR-0072) grant for financial support.

Supplementary material

10562_2010_325_MOESM1_ESM.doc (1.6 mb)
Supplementary material 1 (DOC 1638 kb)

Copyright information

© Springer Science+Business Media, LLC 2010