Carbon dioxide fixation is a very active research area because of the greenhouse effect [1]. An important product of carbon dioxide fixation is a cyclic carbonate, which has broad applications in polycarbonate synthesis, pharmaceutical intermediates, and aprotic solvents [2, 3, 4]. Many catalysts for the synthesis of cyclic carbonates have been developed, including salen metal compounds [5, 6, 7, 8, 9, 10], metalloporphyrins [11, 12, 13], and others [14, 15, 16, 17]. However, these catalysts suffer from low catalyst activity or the need for a co-solvent or a metal ion.
In recent years, ionic liquids as catalysts have gained much interest [18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28]. Some literature reports demonstrated that hydrogen bonding can activate an epoxide and allow its coupling reaction [29, 30, 31, 32, 33, 34, 35, 36, 37]. These investigations implied that hydrogen bonding between the catalyst and epoxide may be the key factor in the coupling of epoxides and CO2. However, Brönsted acid ionic liquids (BAILs) with imidazolium as the cation are expensive, which hindered their industrial application. Furthermore, the commonly used dialkylimidazolium- based ionic liquids showed negligible biodegradability. Thus, it is necessary to synthesize a less expensive ionic liquid. Meanwhile, the factors such as the structure and acidity of the catalyst have not been studied well. It was reported in our previous reports that the structure of the ionic liquid has an important influence on the catalytic activity [38, 39, 40]. Here, several novel long chain BAILs functionalized with a carboxyl group (Scheme 1) were synthesized. A detailed study of the effects of alkyl chain length and nucleophilicity of the anion on the coupling reaction and recycle performance was also conducted. To understand the relationship between acidity and activity, the Hammett method was used to evaluate the acidity of the BAILs.
1-Methylimidazole (99%), triethylamine, N,N- dimethylbutylamine (97%), N,N-dimethyloctylamine (95%), N,N- dimethyldodecylamine(97%),N,N-dimethylhexadecylamine (97%), p-toluenesulfonic acid monohydrate, 1,3-propanesultone, methyl chloroacetate, and methyl bromoacetate were obtained from Aladdin Reagent Co. (Shanghai, China). Propylene oxide, epichlorohydrin, styrene oxide, butylene oxide, phenyl glycidyl ether, and epoxy cyclohexane were obtained from Energy Chemicals (Shanghai, China). The other reagents were purchased from Tianjin Kemio Fine Chemical Institute (China) and used without purification. NMR spectra were recorded on a Bruker 500 MHz spectrometer. ESI-MS was performed on an Esquire 6000 mass spectrometer. The Hammett acidity of BAILs was measured using a PerkinElmer Lambda 35 UV/Vis spectrometer with a basic indicator as reported in our previous paper [39]. The melting point was determined on an XT-4 melting point apparatus without calibration.
The BAILs were prepared according to reported procedures [41]. A typical synthesis route of [(CH2COOH)DMDA]Br (5a) is as follows. Under an inert atmosphere of N2, a mixture of N,N-dimethyldodecylamine (0.01 mol, 1.27 mL) and methyl bromoacetate (0.01 mol, 0.6 mL) was stirred at room temperature for 12 h, during which time the reaction mixture turned into a solid. A mixture of the solid and HCl (37% H2O solution, 0.012 mol) was refluxed for 1 h. The solvent was removed under vacuum, and the remaining solid was washed three times with cold diethyl ether to give the product as a white powder.
[(CH2COOH)TEA]Br (2). mp 188-190 °C, yield 86%, a white solid; 1H NMR (500 M, D2O): δ = 1.12 (dd, J = 7.5 Hz, 9H,CH3CH2-), 3.36-3.40 (m, 6H, CH2N), 3.89 (s, 2H, NCH2COOH); 1C NMR (125 M, D2O): δ = 7.02, 54.20, 55.37, 167.34; ESI-MS: calcd for C8H18NO2Br m/z [M]+ = 160.2, found 160.3.
[(CH2COOH)DMBA]Br (3). mp 156-158 °C, yield 90%, a white solid; 1H NMR (500 M, D2O): δ = 0.76 (dd, J = 7.5 Hz, 3H, CH3CH2-), 1.16-1.23 (m, 2H, CH3CH2CH2CH2N), 1.53-1.69 (m, 2H, CH3CH2CH2CH2N), 3.07 (s, 6H, NCH3), 3.86 (dd, J = 8.5 Hz, 2H, CH2CH2N), 3.97 (s, 2H, NCH2COOH); 1CNMR (125 M, D2O): δ = 12.91, 19.11, 24.13, 51.57, 65.41, 167.50; ESI-MS: calcd for C8H18NO2Br m/z [M]+ = 160.2, found 160.3.
[(CH2COOH)DMOA]Br (4). mp 134-135 °C, yield 92%, a white solid; 1H NMR (500 M, D2O): δ = 0.68 (dd, J = 7.0 Hz, 3H, CH3CH2-), 1.10-1.18 (m, 10H, CH3(CH2)5CH2CH2N), 1.59 (dd, J = 8.5 Hz, 2H, -CH2CH2N), 3.07 (s, 6H, NCH3), 3.34-3.37 (m, 2H, CH2CH2N), 3.99 (s, 2H, NCH2COOH); 1C NMR (125 M, D2O): δ = 13.64, 22.22, 25.58, 28.36, 31.20, 51.71, 65.37, 167.21; ESI-MS: calcd for C12H26NO2Br m/z [M]+ = 216.3, found 216.3.
[(CH2COOH)DMDA]Br (5a). mp 140-141 °C, yield 95%, a white solid; 1H NMR (500 M, D2O): δ = 0.70 (d, 3H, CH3CH2-), 1.16 (d, 20H, CH3(CH2)10CH2CH2N), 1.59 (s, 2H, -CH2CH2N), 3.11 (s, 6H, NCH3), 3.47 (m, 2H, CH2CH2N), 3.92 (s, 2H, NCH2COOH); 1C NMR (125 M, D2O): δ = 14.01, 22.72, 22.82, 26.41, 29.30, 29.69, 29.81, 29.90, 30.00, 30.07, 32.16, 52.12, 63.96, 167.50; ESI-MS: calcd for C16H34NO2Br m/z [M]+ = 272.4, found 272.4.
[(CH2COOH)DMDA]Cl (5b). mp 150-152 °C, yield 95%, a white solid; 1H NMR (500 M, D2O): δ = 0.66 (t, 3H, CH3CH2-), 1.12 (t, 20H, CH3(CH2)10CH2CH2N), 1.53 (d, 2H, -CH2CH2N), 3.06 (s, 6H, NCH3), 3.41 (t, 2H, CH2CH2N), 3.91 (s, 2H, NCH2COOH); 1C NMR (125 M, D2O): δ = 14.01, 22.72, 22.82, 26.41, 29.30, 29.69, 29.81, 29.90, 30.00, 30.07, 32.16, 52.12, 63.96, 167.50; ESI-MS: calcd for C16H34NO2Cl m/z [M]+ = 272.4, found 272.4.
[DDPA]Br (6). mp 164-165 °C, yield 96%, a white solid; 1H NMR (500 M, D2O): δ = 0.65 (t, 3H, CH3CH2-), 1.06 (s, 16H, CH3(CH2)8CH2CH2N), 1.15 (s, 2H, -CH2CH2CH2N ), 1.54 (s, 2H, -CH2CH2N), 1.94-2.00 (m, 2H, NCH2CH2CH2SO3H), 2.71 (t, 2H, NCH2CH2CH2SO3H), 2.91 (s, 6H, NCH3), 3.13 (t, 2H, -CH2CH2N), 3.25-3.28 (m, 2H, NCH2CH2CH2SO3H); 1C NMR (125 M, D2O): δ = 13.99, 18.44, 22.41, 22.79, 26.25, 29.26, 29.67, 29.76, 29.90, 29.99, 30.04, 32.12, 47.51, 50.77, 62.22, 64.32; ESI-MS: calcd for C17H38NSO3Br m/z [M]+ = 336.6, found 336.5.
[(CH2COOH)DMHA]Br (7). mp 146-147 °C, yield 93%, a white solid; 1H NMR (500 M, DMSO): δ = 0.84 (t, 3H, CH3CH2-), 1.22 (s, 28, CH3(CH2)14CH2CH2N), 1.63 (s, 2H, -CH2CH2N), 2.44-2.49 (m, 6H, NCH3), 3.21 (t, 2H, CH2CH2N), 3.34-3.74 (m, 2H, NCH2COOH); ESI-MS: calcd for C20H42NO2Br m/z [M]+ = 328.6, found 328.6.
The Hammett acidity (H0) function was calculated by the equation: H0 = pK(Iaq) + lg([I]s/[IH+]s). Here, "I" represents the base indicator, and [IH+]s and [I]s are, respectively, the molar concentration of the protonated and non-protonated forms of the indicator. The value of [I]s/[IH+]s was determined and calculated from the UV-Vis spectrum. In the experiment, dimethyl yellow (pKa = 3.3) was chosen as the base indicator and ethanol as the solvent. The maximum absorbance of the non- protonated form of dimethyl yellow was observed at 432.5 nm in ethanol. With the increase in acidity of the BAILs, the absorbance of the non-protonated form of the base indicator decreased, while the protonated form of the indicator could not be observed because of its small molar absorptivity. So [I]s/[IH+]s was determined from the difference in measured absorbance after the addition of catalyst, and then H0 can be calculated.
A solution of BAIL (0.1 mmol) in epoxide (0.1 mol) was introduced into a 100 mL stainless steel autoclave. It was purged three times with CO2 and then filled with CO2 to 0.9 MPa at room temperature. Then, the mixture was stirred and heated to 120 °C. After the reaction, the reactor was cooled and vented. The remaining mixture was distilled under reduced pressure or recrystallized with ethanol to obtain the pure cyclic carbonate.
4,5-Tetramethylene-1,3-dioxolan-2-one. 1H NMR (500 M, CDCl3): δ = 1.38-1.43 (m, 2H), 1.56-1.61 (m, 2H), 1.86-1.89 (m, 4H), 4.66-4.68 (m, 2H); 1C NMR (125 M, CDCl3): δ = 19.39, 26.99, 76.02, 155.84.
The H0 function expresses the acidity strength of an acid in an organic solvent. The H0 of the BAILs is listed in Table 1. Unfortunately two ionic liquids showed no protonation, but the other results effectively distinguished the acidity of the various ionic liquids. The Brönsted acidity of the BAILs with different cations was determined (entries 2-5, 7, and 8). The acidity of the BAILs followed the order 6 > 7 > 5a > 4 > 2, which clearly indicated that the length of the carbon chain in the cation influenced the acidity of the BAILs. A longer carbon chain in the catalyst gave a stronger acidity, which was in agreement with our previous reported work (entries 2-5 and 8) [38, 39, 40]. It was postulated that an increase in the bulkiness of the alkyl chain results in a decrease in the electrostatic interaction between the cation and anion, consequently, the acidity increased [36]. Also, the acidity order was correlated with the order of the BAILs catalytic activity for the coupling reaction. The BAIL of 6 incorporated with a sulfonic acidic group instead of a carboxylic one showed the strongest acidity (entry 7 versus entries 2-6 and 8). However, 6 exhibited the lowest activity even at a high temperature (entry 7). The result implied that moderate acidity was beneficial for the reaction. According to the reaction mechanism, hydrogen bonding between the proton of the carbonyl group and oxygen of the epoxide has a positive effect on activating the ring opening of the epoxide. However, hydrogen bonding that is too strong, formed when the acidity of the ionic liquid was strong, impeded the insertion of CO2 into the hydrogen bond between the hydrogen atom and oxygen atom [35]. Moreover, we propose that the long chain catalysts 5a and 7 have the ability to increase the solubility of CO2. Zhang et al. [42] reported that the solubility of CO2 increased with the length of the polyethoxyl chain. Hence, considering the cost, we selected 5a as a model catalyst for further investigation.
The effect of reaction temperature on the coupling reaction of ECH and CO2 is shown in Fig. 1. The catalytic activity was sensitive to temperature. The yield increased sharply when the temperature increased from 60 °C to 120 °C. Further temperature increase had no significant effect on the conversion with a slight decrease of the yield because of side reactions [43].Therefore, the optimum reaction temperature is 120 °C.
The catalyst amount was also of great importance for the coupling reaction. The effect of the molar ratio of the catalyst on the reaction illustrated that there were not enough active sites for the reaction when the catalyst amount was low (Fig. 2). The yield increased with the catalyst amount from 0.2 mol% to 0.8 mol%, and the yield remained constant when the molar ratio was above 1.0 mol%. On further increase in the amount of catalyst, the yield decreased slightly, therefore, 1 mol% was selected as the best catalyst amount.
It is well known that the stability and reusability of a catalyst system are two key factors that identify whether it can be applied in industry. To test these characteristics for the catalyst system, the coupling reaction was carried out under the optimum conditions with the best catalyst 5a. After the reaction, the catalyst was separated from the volatile organic products and starting materials by distillation and reused for the next cycle. As shown in Fig. 3, after five successive recycles, no significant drop in the yield was detected.
To further extend the scope of the reaction, several other epoxides were utilized as the substrate in the coupling reaction in the presence of the best catalyst 5a. The results are summarized in Table 2. The cyclic carbonates were synthesized with yields from good to excellent (entries 1-6). Aliphatic epoxides were preferred substrates for the reaction (entries 1-3 and 5), in particular, the reaction of ECH and CO2 was completed within 30 min (entry 2). Aromatic epoxides, due to hindrance from the phenyl group, need a longer reaction time to reach a good conversion (entry 4). A more hindered epoxide needs a longer reaction time (entry 6). It can be seen that the steric effect is very important in this coupling reaction.
A mechanism for the BAIL-catalyzed coupling of epoxides with CO2 in accordance with previous reports [29, 30, 31, 32, 33, 34, 35, 36, 37] is proposed in Scheme 2. The Brönsted acid center of the catalyst activates the epoxide through a hydrogen bonding interaction. The Lewis basic center (bromide anion) attacks the activated carbon of the epoxide leading to ring opening to generate the alkoxide intermediate. Finally, the alkoxide intermediate undergoes the insertion of CO2 and subsequent ring closure to form the cyclic carbonate along with the regeneration of the catalyst.
Long chain BAILs are efficient catalysts for the coupling of epoxides and CO2. They are easy to synthesize and are thermally stable and environmentally benign. The relationship between the acidity and catalytic activity of the catalyst was investigated. [(CH2COOH)DMDA]Br with moderate acidity and a long carbon chain exhibited the best activity, and it can be reused at least five times without loss of activity.
2015, Vol. 36 Issue (3): 408-413
PDF (553 KB)
2015, Vol. 36 Issue (3): 408-413
PDF (553 KB)