An ionic liquid catalyzed probase method for one-pot synthesis of α,β-unsaturated esters from esters and aldehydes under mild conditions

Gang Wang ab, Yiming Xu c, Suojiang Zhang b, Zengxi Li *ab and Chunshan Li *ab
aCollege of Chemistry and Chemical Engineering, University of Chinese Academy of Sciences, Beijing, 100049, People's Republic of China. E-mail: csli@ipe.ac.cn; zxli@ipe.ac.cn
bBeijing Key Laboratory of Green Process and Engineering, Key Laboratory of Ionic Liquids Clean Process and State Key Laboratory of Multiphase Complex Systems, The National Key Laboratory of Clean and Efficient Coking Technology, Institute of Process Engineering, Chinese Academy of Sciences, Beijing, 100190, People's Republic of China
cSchool of Chemical Engineering, East China University of Science and Technology, Shanghai, 200237, People's Republic of China

Received 26th July 2017 , Accepted 24th August 2017

First published on 4th September 2017


A one-pot synthesis of α,β-unsaturated esters from unactivated esters and aldehydes using strong bases, such as sodium alkoxide and potassium tert-butoxide, was reported. However, the ionic liquid (IL) catalyzed probase method for producing α,β-unsaturated esters was not reported until now. In this work, a series of ILs with fluoride anions were firstly prepared and used as catalysts in combination with the probase N,O-bis(trimethylsilyl) acetamide (BSA) for the α,β-unsaturated esters synthesis. This process could also be promoted through the introduction of another IL with Lewis acid sites. The yield and selectivity of the product could reach up to 84.2% and 95.0%, respectively, when [Bmim]F was used in combination with [Bmim]Cl/AlCl3 (the molar fraction of AlCl3 is 0.67). The mechanism investigation through GC-MS indicates that BSA would convert into onium amide, which acted as a strong base for α-H abstraction, with the catalysis of [Bmim]F. Meanwhile, [Bmim]Cl/AlCl3 played an important role in the condensation step between enolates and aldehydes. On the basis of mechanism insights, kinetic and thermodynamic studies were also carried out for a better understanding of this new route.


Introduction

α,β-Unsaturated esters as important chemical intermediates and products are widely used in the field of polymer materials production,1 paints and coatings preparation,2 adhesives and textiles synthesis,3 and pharmaceutical and essences fabrication.4 Approaches to such compounds include direct esterification,5 Reformasky reaction,6 Doebner method,7 dehydrogenation,8 Knoevenagel condensation,9 oxidative esterification,10 Witting reaction,11 Horner–Wadsworth–Emmons reaction12 and aldol condensation.4,13–15 The direct esterification is the simplest among these routes, but α,β-unsaturated acid should be prepared firstly. It could also be obtained from α-haloesters using zinc as the catalyst through the Reformasky reaction and from α-cyanoesters in the presence of a base via the Knoevenagel reaction. But zinc halides cannot be recycled and the preparation of α-cyanoesters often needs a poisonous cyanide reagent. Although the dehydrogenation and oxidative esterification methods seem to be environmentally-friendly, green and sustainable, these technologies usually require novel metal catalysts and the catalysts are hard to recycle. Witting and Horner–Wadsworth–Emmons reactions need phosphine reagents, which are often toxic and environmentally-unfriendly. Aldol condensation is considered as a key and efficient reaction for the construction of the carbon–carbon bond to form α-hydroxyl and α,β-unsaturated carbonyl compounds.16–27

Unactivated esters with high pKa values of α-H, unlike aldehydes or ketones, should be enolized during the aldol condensation. Typically, alkoxide salts are able to deprotonate the α-H of esters, generating active enolate intermediates that are required for the subsequent carbon–carbon coupling reactions.28,29 However, it is difficult to avoid side reactions due to the solubility of these salts in common solvents as well as their nucleophilicity. Otherwise, alkali metal amides, such as lithium hexamethyldisilazide (LiHMDS), lithium diisopropylamide (LDA) and lithium tetramethylpiperidide (LiTMP), are also commonly used in organic chemistry as strong non-nucleophilic bases. But the reaction temperature should be decreased to below 0 °C and the operation should be carried out carefully due to the high activity and air (or water) sensitivity of such strong bases. Followed by Kondo's research on an in situ method to produce a base through the fluoride anion catalysis of aminosilanes,30,31 Teng found the pentanidium- and bisguanidinium-salts catalyzed probase method for the alkylation of lactones using silylamide.32 Compared with strong bases, a probase is relatively stable and it is more convenient for operation. In addition, ILs as green and environmentally-friendly reagents are universally applied for the replacement of traditional solvents and acidic (or basic) catalysts, due to their low vapor pressure, easy recycling and high catalytic performance in most organic reactions.33–39

Inspired from the work of Teng and Kondo, the IL catalyzed probase strategy shown in Scheme 1 was envisioned for producing α,β-unsaturated esters from unactivated esters and aldehydes under mild conditions, which has not been reported until now. In the present work, N,O-bis(trimethylsilyl) acetamide (BSA) was selected as the probase and ILs with fluoride anions (ILs-F) were designed and selected as catalysts to catalytically decompose BSA into a strong base of onium amide. The α-H deprotonation and ester enolization would occur in the presence of onium amide. Other ILs with Lewis acid sites (ILs-L) were also used in combination to catalyze the aldol condensation step between enolates and aldehydes. The effects of cation (for ILs-F), Lewis acidity concentration (for ILs-L) and solvent on this catalytic process were also studied. The reaction mechanism was also analyzed and investigated in detail by using GC-MS. On the basis of mechanism insights, kinetic and thermodynamic studies were also carried out for a better understanding of this new process.


image file: c7gc02265d-s1.tif
Scheme 1 IL catalyzed probase method for the production of α,β-unsaturated esters.

Experimental

Materials

Dichloromethane (purity ≥ 99.5%) was supplied by Xilong Chemical Co., Ltd (China). Esters (purity ≥ 99.0%), aldehydes (purity ≥ 99.0%), AgF (purity of 99.0%) and N,O-bis(trimethylsilyl) acetamide (purity ≥97.0%) were purchased from J&K Scientific Ltd (China). Octane (purity ≥ 99.5%) was provided by Aladdin Industrial Co. (China). ILs (purity ≥ 99.0%) were supplied by Linzhou Keneng Materials Technology Co., Ltd (China). The solvents used for the synthesis reaction should be dehydrated with calcium hydride under reflux for about 24 h and the dry solvents could be obtained after distillation. The reactants should also be dehydrated with 4 Å molecular sieves.

Preparation of ILs-F

The ILs-F were prepared from AgF and ILs with bromine (or chloride) anions (ILs-Br or ILs-Cl) as shown in Scheme 2. The solution of ILs-Br (or ILs-Cl, 1 eq.) was firstly fabricated in ethanol, followed by dropwise addition of the AgF (1.03 eq.) aqueous solution at room temperature at a suitable stirring speed. Another 5 min stirring was performed after the AgF solution was added. The formed light yellow (or white) precipitates were filtered off and the solvent was removed under high vacuum at 30 °C. The obtained crude ILs-F were dissolved in ethanol and filtered off to remove the remaining AgF. Then the ethanol was removed under high vacuum at 30 °C and the relatively high pure ILs-F were obtained. The purity and NMR spectra of each prepared IL-F are provided in the ESI.
image file: c7gc02265d-s2.tif
Scheme 2 Preparation mechanism of ILs-F.

Preparation of ILs-L

In a glove box, metal chlorides (ZnCl2, CuCl, FeCl3, AlCl3) were weighed in a suitable amount and charged into a flask. Then the flask was sealed and kept in an alcohol bath under −20 °C. After that, the dried [Bmim]Cl (under high vacuum at 90 °C for about 24 h) in the corresponding molar ratio was injected into the flask dropwise and mixed with metal chloride powder under an Ar (or N2) atmosphere. The mixture was kept for stirring under −20 °C for about 4 h. The ILs with Lewis acid sites (ILs-L) [Bmim]Cl/MClx (M = Zn2+, Cu+, Fe3+, Al3+) could be obtained (the FT-IR spectra are provided in the ESI). The prepared ILs-L should be stored in a glove box or under an inert gas atmosphere to prevent hydrolysis. The preparation apparatus was shown in Fig. 1.
image file: c7gc02265d-f1.tif
Fig. 1 The apparatus for preparing ILs-L. 1 – flask, 2 – stirrer, 3 – rubber plug, 4 – injector, 5 – glass valve.

General synthesis

All synthesis reactions were carried out in a round-bottomed flask under 1 atm and the air in the flask was replaced with N2 or Ar before the dehydrated solvent and reagents were added into it. The reaction mixture was kept at 283–298 K with a temperature-controlled water bath equipped with a magnetic stirrer. A typical reaction solution consisted of aldehyde (0.1 M in solvent), ester, BSA, ILs and octane which was used as an internal standard. To keep the molar concentrations consistent, solutions for reactions with other reagents were prepared to have the same concentration as that of aldehyde (0.1 M). The aldehyde in a suitable molar ratio to ester, octane and ILs (for the cation effect investigation, only ILs-F were used, and for other synthesis reactions, both ILs-F and ILs-L were used) in an appreciable percentage was firstly mixed in the solvent before being introduced into the flask. Then BSA was injected dropwise into the former solution with a syringe at a suitable stirring speed. Product samples were periodically collected and analyzed with GC-MS.

Analysis methods

1H-NMR, 13C-NMR and 19F-NMR analyses were carried out on a Bruker AVANCE instrument (600 MHz), and the spectra were recorded at ambient temperature. Chemical shifts were reported in ppm relative to the residual solvent signal (D2O: 4.79 ppm).

Product samples were analyzed and quantified in the meantime with GC-MS apparatus (QP2020, Shimadzu) equipped with an Rtx-5MS column (30 m, 0.32 mm, 0.25 μm). Products were identified by comparison with standards and MS information, carrier gas, He; temperature, 40 °C (2 min) to 300 °C at 15 °C min−1 and hold for 5 min; detector temperature, 250 °C; and injector temperature, 300 °C. The concentration of each reactant and product was obtained by using octane as the internal standard. The mass correction factor of each substance relative to octane was determined from the mixtures of the standard sample that were diluted with CH2Cl2. For others without the standard sample, it was estimated by using the effective carbon number method.40,41

All the FT-IR samples of ILs-L were spread into liquid films on a KBr window. The spectra were recorded on a Nicolet Fourier transform infrared spectrophotometer at room temperature. The pyridine probing IR for Lewis acidity concentration measurement was similar to the former operation; the samples were prepared by mixing pyridine, which was dried with solid KOH and distilled over 4 Å molecular sieves, with ILs-L in the volume ratio of 1[thin space (1/6-em)]:[thin space (1/6-em)]5 in a glove box.

Results and discussion

Cation effect of ILs-F

To investigate this IL catalyzed probase method, the condensation of benzaldehyde with methyl acetate to methyl cinnamate was selected as the model reaction. According to the research results of Kondo and Teng,32,42,43 the cation of the in situ formed strong base onium amide would affect the reaction results. Thus the cation effect of ILs-F was firstly explored. The ILs-F with different cations were prepared as described in the Experimental section and the structure was confirmed through 1H-NMR, 13C-NMR and 19F-NMR, as illustrated in the ESI. The synthetic results obtained using ILs-F with different cations are shown in Fig. 2.
image file: c7gc02265d-f2.tif
Fig. 2 Catalytic performance of ILs-F with different cations. Reaction conditions: Benzaldehyde (0.2 mmol, 0.1 M in CH2Cl2), the molar ratio of methyl acetate to BSA and benzaldehyde was 2[thin space (1/6-em)]:[thin space (1/6-em)]2[thin space (1/6-em)]:[thin space (1/6-em)]1, [Bmim]F: 20 mol%, reaction temperature: 20 °C, reaction time: 4 h.

As mentioned above, the cation has a good influence on the yield and selectivity of methyl cinnamate. The results showed that [Bmim]F has the best performance. In addition, it could be observed that the chain length in the cation could directly affect the results. The data of pyridine- and imidazole-type ILs-F obviously indicate that with the increase of chain length, the catalytic performance rises to the peak and then reduces down. It was considered that the Si–O bond in BSA could be broken with the catalysis of the fluoride anion in ILs-F, followed by the generation of the acetamide anion, after which the onium amide base would be formed through the interaction between the acetamide anion and the cation from ILs-F according to the reported results.32,42,43 Although the catalytic function focuses on the fluoride anion, the molecular voidage of ILs-F would increase with the extension of chain length, which would be beneficial for the capacity of BSA molecules in ILs-F clusters. Therefore, the catalytic decomposition of BSA into the strong base acetamide anion could be promoted to some degree, making it conducive to the enolization of methyl acetate.44,45 The onium amide base, in this reaction, plays the role of a strong base to abstract the α-H of esters, producing the enolate for the next aldol condensation step. So the increasing chain length would also enhance the steric hindrance of the α-H abstraction step. Moreover, for imidazole-type ILs-F, the 2-C of the imidazole ring would transform to carbene due to the active 2-H. The raw material benzaldehyde could react with the carbene, leading to a reduction of the product selectivity. But the increased chain length could also impede this process. Therefore, a suitable chain length would promote the formation of the onium amide base and the enolization of esters.

Effect of ILs-L

On the basis of the cation effect analysis, [Bmim] was confirmed as the optimum one. According to Scheme 1, benzaldehyde would condense with 1-methoxy-1-trimethylsilyloxyethene (enolate of methyl acetate) to produce methyl cinnamate, which is also called the Mukaiyama reaction. And as it is known to all, Lewis acid catalysts could promote the Mukaiyama reaction. When [Bmim]Cl was mixed with a Lewis acidic metal chloride in a suitable molar ratio, the ILs-L ([Bmim]Cl/MClx) exhibited Lewis acidity. Based on the research results of Kou,46 when the molar fraction of the metal chloride (CuCl, FeCl3, ZnCl2 and AlCl3) in [Bmim]Cl/MClx is 0.67, the ILs-L would present a strong Lewis acidity. So a series of ILs-L ([Bmim]Cl/MClx), with a metal chloride molar fraction of 0.67, were prepared and applied in combination with [Bmim]F for the catalytic condensation of methyl acetate with benzaldehyde. The experimental results are shown as Fig. 3. It indicates that with the addition of ILs-L, the yield and selectivity of methyl cinnamate could be enhanced obviously. And the increasing tendency could be ranked as [Bmim]Cl/CuCl < [Bmim]Cl/FeCl3 < [Bmim]Cl/ZnCl2 < [Bmim]Cl/AlCl3. Then pyridine probing IR was conducted for acidic concentration measurement, which is described in Fig. 4. It shows that these four types of ILs-L all have Lewis acid sites; [Bmim]Cl/CuCl and [Bmim]Cl/AlCl3 also have Brønsted acid sites. By comparing the IR spectra, the Lewis acid concentration could also be ranked as [Bmim]Cl/CuCl < [Bmim]Cl/FeCl3 < [Bmim]Cl/ZnCl2 < [Bmim]Cl/AlCl3, which is in accordance with the activity order. So it can be considered that the catalytic performance of ILs-L is related to the amount of Lewis acid sites.
image file: c7gc02265d-f3.tif
Fig. 3 Catalytic performance of different ILs-L. Reaction conditions: Benzaldehyde (0.2 mmol, 0.1 M in CH2Cl2), the molar ratio of methyl acetate to BSA and benzaldehyde was 2[thin space (1/6-em)]:[thin space (1/6-em)]2[thin space (1/6-em)]:[thin space (1/6-em)]1, [Bmim]F: 20 mol%, ILs-L: 20 mol% (the molar fraction of MClx in ILs-L was 0.67), reaction temperature: 20 °C, reaction time: 4 h.

image file: c7gc02265d-f4.tif
Fig. 4 Pyridine probing IR for acidity measurement.

The effect of the molar fraction of AlCl3 was also investigated and the results are presented in Fig. 5. As it can been seen, with the increasing molar fraction of AlCl3, the yield and selectivity of methyl cinnamate also rise apparently. According to the pyridine probing measurement results published by Kou and listed in Table 1,46 with the molar fraction of AlCl3 changing from 0.5 to 0.67, the predominant anionic species in [Bmim]Cl/AlCl3 varies from [AlCl4] to [Al2Cl7]. The experimental results indicate that the [Al2Cl7] species plays an important role in catalytic condensation between 1-methoxy-1-trimethylsilyloxyethene and benzaldehyde, while [AlCl4] shows no activity. In fact, when the molar fraction is 0.5, [Bmim]Cl/AlCl3 is neutral, but when the molar fraction reaches 0.67, [Bmim]Cl/AlCl3 possesses a strong Lewis acidity, which would be beneficial for the condensation step. Therefore, [Bmim]Cl/AlCl3, in which the molar fraction of AlCl3 is 0.67, could be used in combination with [Bmim]F for the one-pot synthesis of methyl cinnamate from methyl acetate and benzaldehyde under mild conditions.


image file: c7gc02265d-f5.tif
Fig. 5 Catalytic performance of [Bmim]Cl/AlCl3 with different molar fractions of AlCl3. Reaction conditions: Benzaldehyde (0.2 mmol, 0.1 M in CH2Cl2), the molar ratio of methyl acetate to BSA and benzaldehyde was 2[thin space (1/6-em)]:[thin space (1/6-em)]2[thin space (1/6-em)]:[thin space (1/6-em)]1, [Bmim]F: 20 mol%, [Bmim]Cl/AlCl3 (the molar fraction of AlCl3 changes from 0.50 to 0.67): 20 mol%, reaction temperature: 20 °C, reaction time: 4 h.
Table 1 The effect of the molar fraction of AlCl3 on the anionic species
Molar fraction of AlCl3 in [Bmim]Cl/AlCl3 Predominant anionic species
0.5 [AlCl4]
0.55 [AlCl4] and [Al2Cl7]
0.6 [AlCl4] and [Al2Cl7]
0.67 [Al2Cl7]


Effect of solvent

In addition, the effect of solvent on the methyl cinnamate synthesis was also researched. It has been well known to all that a polar solvent would be beneficial for the polarity of the α-(C–H) bond and the α-H abstraction of carbonyl compounds. Thus, a series of polar solvents were selected for our reaction system. Interestingly, the influence of most polar solvents THF and acetonitrile as shown in Table 2 is not as high as that of CH2Cl2 and CH2Cl–CH2Cl. Our experimental results show that [Bmim]F and [Bmim]Cl/AlCl3 have better solubility in CH2Cl2 and CH2Cl–CH2Cl than in other solvents. So it could be considered that in CH2Cl2 or CH2Cl–CH2Cl, the reaction could be treated as homogeneous catalysis, while it is liquid–liquid heterogeneous catalysis in other solvents, otherwise the onium enolate complexes should transfer from the IL phase into the organic phase, which is also inefficient for the whole catalytic procedure. In addition, the polarity of CH2Cl2 is also suitable for α-proton abstraction, which could be supported by the research work of Atsushi and Downey.47,48 So it can be concluded that the solvent selected for the IL catalyzed probase method should be polar which could as far as possible dissolve [Bmim]F and [Bmim]Cl/AlCl3 as well.
Table 2 Effect of solvent on the synthetic results
Entry Solvent Yield/% Selectivity/%
Reaction conditions: Benzaldehyde (0.2 mmol, 0.1 M in solvent), the molar ratio of methyl acetate to BSA and benzaldehyde was 2[thin space (1/6-em)]:[thin space (1/6-em)]2[thin space (1/6-em)]:[thin space (1/6-em)]1, [Bmim]F: 20 mol%, [Bmim]Cl/AlCl3 (the molar fraction of AlCl3 was 0.67): 20 mol%, reaction temperature: 20 °C, reaction time: 4 h.
1 CH2Cl2 82.8 94.6
2 THF 70.4 92.1
3 CH2Cl–CH2Cl 75.5 93.2
4 Toluene 52.7 75.4
5 Acetonitrile 61.2 80.7


Mechanism analysis and kinetic studies

On the basis of the impact analysis of the cation, ILs-L, and solvent, the mechanism of the one-pot synthesis of methyl cinnamate from methyl acetate and benzaldehyde using [Bmim]F in combination with [Bmim]Cl/AlCl3 was investigated. In the light of the intermediates that were detected with GC-MS (for MS information, ESI), the main reaction mechanism was illustrated as given in Scheme 3. BSA firstly converted into the strong base onium amide with the catalysis of [Bmim]F; then the α-H of methyl acetate was abstracted by the onium amide, generating N-trimethylsilyl acetamide and onium enolate. The onium amide could be recycled in situ when onium enolate transformed into 1-methoxy-1-trimethylsilyloxyethene, followed by condensation with benzaldehyde to β-O-(trimethylsilyl) methyl phenylpropionate with the catalysis of [Bmim]Cl/AlCl3. The intermediate β-O-(trimethylsilyl) methyl phenylpropionate could be further converted into methyl cinnamate and hydroxytrimethylsilane with the catalysis of [Bmim]F. The GC-MS testing results also revealed that hydroxytrimethylsilane and methyl cinnamate would convert into hexamethyl disiloxane and trimethylsilyl cinnamate, respectively, with a prolonged reaction time.
image file: c7gc02265d-s3.tif
Scheme 3 Mechanism analysis for the synthesis of methyl cinnamate catalyzed with [Bmim]F and [Bmim]Cl/AlCl3 (the molar fraction of AlCl3 is 0.67) using BSA as the probase.

According to the above mechanism analysis, the reaction for 3 h at 283–298 K was selected for kinetic studies, considering the ignorable side reactions. The elementary reactions from (1-1) to (1-7) are listed in Scheme 4. The reaction steps of (1-1), (1-2) and (1-3) were set as reversible and maintain equilibrium due to the slow rate of (1-5). Reaction (1-6) was also considered an equilibrium reaction due to the high activity of the amide anion and [TMS-O]. According to the reaction steady-equilibrium principle, the equilibrium equations of (1-2), (1-3) and (1-8) could be obtained as formulas (1)–(4).

 
k1CACD = k−1CBCC(1)
 
k2CBCD = k−2CECF(2)
 
k3CACE = k−3CBCG(3)
 
k6CBCK = k−6CFCL(4)


image file: c7gc02265d-s4.tif
Scheme 4 Mechanism-based kinetic model for the one-pot synthesis of methyl cinnamate from methyl acetate and benzaldehyde catalyzed with [Bmim]F and [Bmim]Cl/AlCl3 (the molar fraction of AlCl3 is 0.67), using BSA as the probase.

So the concentration of substrates C, G and L could be expressed as formulas (5) and (6), respectively.

 
image file: c7gc02265d-t1.tif(5)
 
image file: c7gc02265d-t2.tif(6)

Then the kinetic equations could be listed as (7)–(14) based on the model.

 
image file: c7gc02265d-t3.tif(7)
 
image file: c7gc02265d-t4.tif(8)
 
image file: c7gc02265d-t5.tif(9)
 
image file: c7gc02265d-t6.tif(10)
 
image file: c7gc02265d-t7.tif(11)
 
image file: c7gc02265d-t8.tif(12)
 
image file: c7gc02265d-t9.tif(13)
 
image file: c7gc02265d-t10.tif(14)

The computational simulation for the mechanism-based kinetic model was conducted by using the Runge–Kutta method on MATLAB software.14,40 The temperature and concentration distribution under our reaction conditions could not be taken into account. The obtained results were also compared with the experimental concentration for the effect of reaction temperature and time, which are shown in Fig. 6. The experimental concentrations of β-O-(trimethylsilyl) methyl phenylpropionate and hydroxytrimethylsilane were also compared with the calculated data, which are presented in Fig. 7. The pre-exponential factor and activation energy of each elementary reaction are presented in Table 3.


image file: c7gc02265d-f6.tif
Fig. 6 Computational simulation results of the kinetic model.

image file: c7gc02265d-f7.tif
Fig. 7 Comparison of the experimental concentration with calculated data: (A) β-O-(trimethylsilyl) methyl phenylpropionate, (B) hydroxytrimethylsilane.
Table 3 Pre-exponential factor and activation energy of each elementary reaction
Reaction step Reaction name A E a/(kJ mol−1)
(A) → (B) + (C) BSA decomposition 2.8 × 103 23± 1.2
(B) + (C) → (A) 1.2 × 104 28 ± 1.4
(D) + (B) → (E) + (F) Deprotonation 6.2 × 104 33 ± 1.7
(E) + (F) → (D) + (B) 1.4 × 105 36 ± 1.8
(E) + (A) → (G) + (B) Trimethylsilylation 8.1 × 102 18 ± 0.9
(G) + (B) → (E) + (A) 5.3 × 103 25 ± 1.3
(G) + (H) → (I) Condensation 1.4 × 106 43 ± 2.2
(I) → (J) + (K) Product formation 2.0 × 104 30 ± 1.5
(K) + (B) → (L) + (F) 1.4 × 103 20 ± 1.0
(L) + (F) → (K) + (B) 1.3 × 105 35 ± 1.8
(L) + (C) → (M) 2.1 × 102 5 ± 0.3


From Fig. 6, it can be seen that the concentrations of the probase BSA and the raw materials methyl acetate and benzaldehyde drop rapidly in 100 min; after that, the tendency becomes low. While the concentrations of methyl cinnamate and hexamethyl disiloxane increase with the reaction temperature and time. The concentration of N-trimethylsilyl acetamide increases rapidly in 100 min, and then it tends to stabilize. Fig. 7 shows the error of the kinetic model; for β-O-(trimethylsilyl) methyl phenylpropionate and hydroxytrimethylsilane, the deviations between the calculated data and experimental concentration are all around acceptable 5%. So the kinetic model should be believable at temperatures from 283 K to 298 K and a reaction time of 180 min.

The pre-exponential factor and activation energy of each elementary reaction are listed in Table 3. For all equilibrium reactions, the activation energies of all forward reactions are lower than that of the reverse reactions; therefore, it is advantageous for the formation of methyl cinnamate. The activation energy of the decomposition of BSA into the amide anion is relatively low (23 ± 1.2 kJ mol−1), which indicates that [Bmim]F has a good catalytic performance. The decomposition of β-O-(trimethylsilyl) methyl phenylpropionate into methyl cinnamate is more difficult than that of BSA with the catalysis of [Bmim]F. It is well known that the acidity of the α-proton in unsubstituted esters is weaker than that in ketones, β-dicarbonyl compounds or esters that are substituted with electron-accepting groups. So it is difficult to deprotonate the α-proton of methyl acetate, and the activation energy is relatively high (33 ± 1.7 kJ mol−1) in comparison with others. But it still indicates that the in situ generated strong base onium amide is efficient for the α-proton abstraction process. The activation energy of the formation of 1-methoxy-1-trimethylsilyloxyethene, namely the recycling of onium amide, is low (18 ± 0.9 kJ mol−1), which means that this process is easy to conduct and the α-proton abstraction procedure could also be promoted. As for the condensation step between 1-methoxy-1-trimethylsilyloxyethene and benzaldehyde, the activation energy is the highest of all (43 ± 2.2 kJ mol−1). Thus, it is treated as the rate-controlling step. However, the energy barrier of the formation of hexamethyl disiloxane is the lowest among all steps.

The sensitivity analysis of significant reaction steps, including BSA decomposition, ester deprotonation, enolate trimethylsilylation, condensation and product formation, was carried out and the results are depicted in Fig. 8. It could be easily observed that the condensation step was the most sensitive to the reaction temperature, while the enolate trimethylsilylation step was the most adverse. The sequence could be ranked as condensation > ester deprotonation > product formation > BSA decomposition > enolate trimethylsilylation, and this is in accordance with the energy barrier of these reaction steps.


image file: c7gc02265d-f8.tif
Fig. 8 Temperature sensitivity analysis for significant reaction steps.

On the bases of the above kinetic studies, the equilibrium constant of each reversible reaction was calculated with the reaction rate constant at different temperatures, which is presented in Table 4. It could be believed that the rate constants of the forward reactions are all higher than that of the reverse reactions, which are all beneficial for the production of methyl cinnamate. By comparing these constants at different temperatures, the order can be ranked as K6 > K3 > K1 > K2, which means that the protonation of onium acetamide proceeds more thoroughly than the other steps.

Table 4 Equilibrium constant of each reversible reaction at different temperatures
ln[thin space (1/6-em)]K ln[thin space (1/6-em)]K1 ln[thin space (1/6-em)]K2 ln[thin space (1/6-em)]K3 ln[thin space (1/6-em)]K6
T/K
283 0.693 0.471 1.089 1.851
288 0.656 0.449 1.038 1.740
293 0.620 0.427 0.988 1.633
298 0.586 0.407 0.939 1.530


The enthalpy change (ΔH) is one of the most important thermodynamic parameters of reversible reactions. Thus, the enthalpy change of each reversible reaction, shown in Table 5, was also calculated from the equilibrium constant at different temperatures through the Van't Hoff formula that is expressed as formula (15). The Van't Hoff plot over reaction temperature is shown in Fig. 9. The results were also creditable owing to the acceptable error listed in Table 5.

 
image file: c7gc02265d-t11.tif(15)


image file: c7gc02265d-f9.tif
Fig. 9 Van't Hoff plot for the enthalpy change of each reversible reaction.
Table 5 Enthalpy change of each equilibrium reaction
Equilibrium reaction ΔH/(kJ mol−1) C R 2
image file: c7gc02265d-t12.tif −5.00 −1.43 0.99
image file: c7gc02265d-t13.tif −3.00 −0.80 0.99
image file: c7gc02265d-t14.tif −7.00 −1.89 0.99
image file: c7gc02265d-t15.tif −15.00 −4.52 0.99


As shown in Table 5, it could be obtained that all the reversible reactions are exothermal. The enthalpy changes of the first three equilibrium reactions are close to each other and the reason could be concluded as follows. The thermal effects of these reactions are small, so the absolute values of enthalpy change are all small, which is in agreement with that of the room-temperature reaction. By comparing the absolute values of enthalpy change with each other, the order could be ranked as ΔH6 > ΔH3 > ΔH1 > ΔH2. The interaction between TMS+ and [TMS-O] is the highest, which indicates that this reaction is the most sensitive to the reaction temperature among the reversible reactions. However, as noted above, the absolute values of enthalpy change are all relatively low. This reveals that the reaction temperature has little influence on these equilibrium constants and the previous assumption of steady-equilibrium is reasonable. These equilibrium reactions are very important for the formation of methyl cinnamate. Therefore, the thermodynamic studies on equilibrium reactions involving the equilibrium constant and enthalpy change could help in understanding this new catalytic process.

Substrates extension

In addition to methyl cinnamate, other unsubstituted acetate and propionate esters for this one-pot condensation reaction in our IL catalyzed system were also evaluated, and the experimental results are presented in Table 6. A comparison between benzaldehyde and acetate derivatives shows a decrease in yield with the increasing length of the carbon chain that combined with the oxygen atom in the ester bond. This observed decreasing tendency could be related to the increasing difficulty in the enolization of esters. The substituted methyl group of the α-carbon atom and the length of the carbon chain that combined with the oxygen atom in the ester bond could contribute to the steric hindrance and difficulty in polarizing carbonyl groups, sequentially affecting the acidification and abstraction of the α-proton. The addition of an electron-donating group in benzaldehydes could promote the yield, while the electron-withdrawing group decreases the yield. These results prove the catalytic performance of ILs-L. However, the trend is not very sharp, which shows a cooperative effect between the strong basicity of onium amide and the Lewis acidity of [Bmim]Cl/AlCl3 (the molar fraction of AlCl3 is 0.67). The enolate anion generated from the deprotonation of esters with onium amide could also condense with benzaldehydes directly, during which the substituted electron-withdrawing group is more advantageous.
Table 6 One-pot synthesis of methyl cinnamate from esters and benzaldehyde catalyzed with [Bmim]F and [Bmim]Cl/AlCl3a
Entry Product Yieldb/% Selectivityb/%
a Reaction conditions: Benzaldehyde (0.2 mmol, 0.1 M in CH2Cl2), the molar ratio of methyl acetate to BSA and benzaldehyde was 2[thin space (1/6-em)]:[thin space (1/6-em)]2[thin space (1/6-em)]:[thin space (1/6-em)]1, [Bmim]F: 20 mol%, [Bmim]Cl/AlCl3 (the molar fraction of AlCl3 was 0.67): 20 mol%, reaction temperature: 20 °C, reaction time: 4 h. b Determined by GC-MS, GC part for quantification and MS part for structure confirmation.
1 image file: c7gc02265d-u1.tif 79.5 94.3
2 image file: c7gc02265d-u2.tif 80.7 93.6
3 image file: c7gc02265d-u3.tif 80.2 92.9
4 image file: c7gc02265d-u4.tif 78.6 94.1
5 image file: c7gc02265d-u5.tif 79.4 93.7
6 image file: c7gc02265d-u6.tif 75.2 93.2
7 image file: c7gc02265d-u7.tif 74.3 92.6
8 image file: c7gc02265d-u8.tif 73.9 93.0
9 image file: c7gc02265d-u9.tif 70.1 93.2
10 image file: c7gc02265d-u10.tif 68.8 89.9
11 image file: c7gc02265d-u11.tif 84.2 95.0
12 image file: c7gc02265d-u12.tif 79.5 94.0
13 image file: c7gc02265d-u13.tif 80.3 93.8
14 image file: c7gc02265d-u14.tif 81.2 92.9


Conclusions

A series of ILs-F were firstly prepared and applied in combination with ILs-L for the one-pot catalytic synthesis of α,β-unsaturated esters from esters and aldehydes. In this catalytic system, BSA as a probase could be converted into the strong base onium amide with the catalysis of ILs-F; meanwhile, esters could be deprotonated by such a strong base and transformed into enolates. The onium amide could also be recycled through the trimethylsilyl etherification of enolates with BSA. ILs-L play an important role in the condensation between enol silyl ether and aldehyde. The cations of ILs-F, the Lewis acidity concentration and strength of ILs-L as well as the solvent have an effect on this reaction. The kinetic parameters and equilibrium constants of each reaction step were calculated through mechanism-based kinetic analysis and simulation. The results also revealed that the condensation process was the rate-controlling step; all the equilibrium reactions were exothermal and insensitive to the reaction temperature. This IL catalyzed probase method was also universal for the synthesis of other α,β-unsaturated esters besides methyl cinnamate. The yield and selectivity could reach up 84.2% and 95.0% respectively.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors wish to acknowledge the National Key Projects for Fundamental Research and Development of China (2016YFB0601303). This work was also financed by the National Science Fund for Excellent Young Scholars 4(no. 21422607), the National Natural Science Foundation of China (no. 21676270), the National Natural Science Foundation of China (no. 21576261) and the NSFC-Key Projects of Shanxi Coal Based Low Carbon Joint Foundation (no. U1610222). This research was also supported by the CAS/SAFEA International Partnership Program for Creative Research Teams.

Notes and references

  1. E. S. Dragan, Chem. Eng. J., 2014, 243, 572 CrossRef CAS .
  2. J. Bryjak, K. Bachmann, B. Pawlówand and I. Maliszewska, Chem. Eng. J., 1997, 65, 249 CrossRef CAS .
  3. A. Mamoru, Appl. Catal., A, 2005, 288, 211 CrossRef .
  4. J. K. Augustine, C. Boodappa, S. Venkatachaliah and A. Mariappan, Tetrahedron Lett., 2004, 55, 3503 CrossRef .
  5. K. V. N. S. Srinivas and B. Das, J. Org. Chem., 2003, 68, 1165 CrossRef CAS PubMed .
  6. C. Liu, S. Tang, L. Zheng, D. Liu, H. Zhang and A. Lei, Angew. Chem., Int. Ed., 2012, 51, 5662 CrossRef CAS PubMed .
  7. A. Galat, J. Am. Chem. Soc., 2002, 68, 376 CrossRef .
  8. K. C. Nicolaou, T. Montagnon and P. S. Baran, Angew. Chem., Int. Ed., 2002, 41, 993 CrossRef CAS PubMed .
  9. L. F. Tietzeand and A. Steinmetz, Angew. Chem., Int. Ed., 1996, 35, 651 CrossRef .
  10. L. Ta, A. Axelsson and H. Sundén, Green Chem., 2016, 18, 686 RSC .
  11. D. B. Denneyand and S. T. Ross, J. Org. Chem., 1962, 27, 998 CrossRef .
  12. B. E and M. A. B. Reitz, Chem. Rev., 1989, 89, 863 CrossRef .
  13. D. Yang, C. Sararuk, K. Suzuki, Z. X. Li and C. S. Li, Chem. Eng. J., 2016, 300, 160 CrossRef CAS .
  14. G. Wang, Z. X. Li, C. S. Li and H. Wang, Chem. Eng. J., 2017, 319, 297 CrossRef CAS .
  15. T. Mukaiyama, K. Banno and K. Narasaka, J. Am. Chem. Soc., 1974, 96, 7503 CrossRef CAS .
  16. H. L. Fan, Y. Y. Yang, J. L. Song, G. D. Ding, C. Y. Wu, G. Y. Yang and B. X. Han, Green Chem., 2014, 16, 600 RSC .
  17. A. L. Zhu, T. Jiang, D. Wang, B. X. Han, L. Liu, J. Huang, J. C. Zhang and D. H. Sun, Green Chem., 2005, 7, 514 RSC .
  18. R. Lee, J. R. Vanderveen, P. Champagne and P. G. Jessop, Green Chem., 2016, 18, 5118 RSC .
  19. G. F. Liang, A. Q. Wang, X. C. Zhao, N. Lei and T. Zhang, Green Chem., 2016, 18, 3430 RSC .
  20. H. X. Li, Z. W. Xu, P. F. Yan and Z. C. Zhang, Green Chem., 2017, 19, 1751 RSC .
  21. W. Gati and H. Yamamoto, Acc. Chem. Res., 2016, 49, 1757 CrossRef CAS PubMed .
  22. X. P. Guo, D. Yang, Z. J. Peng, C. S. Li and S. J. Zhang, Ind. Eng. Chem. Res., 2017, 56, 5860 CrossRef CAS .
  23. C. Sararuk, D. Yang, G. L. Zhang, C. S. Li and S. J. Zhang, J. Ind. Eng. Chem., 2017, 46, 342 CrossRef CAS .
  24. G. Wang, H. Wang, C. S. Li, C. C. Zuo, Z. X. Li and S. J. Zhang, J. Ind. Eng. Chem., 2017, 55, 173 CrossRef CAS .
  25. H. Zhao, C. C. Zuo, D. Yang, C. S. Li and S. J. Zhang, Ind. Eng. Chem. Res., 2016, 55, 12693 CrossRef CAS .
  26. G. L. Zhang, H. H. Zhang, D. Yang, C. S. Li, Z. J. Pengand and S. J. Zhang, Catal. Sci. Technol., 2016, 6, 6417 CAS .
  27. D. Yang, D. Li, H. Y. Yao, G. L. Zhang, T. T. Jiao, Z. X. Li, C. S. Li and S. J. Zhang, Ind. Eng. Chem. Res., 2015, 54, 6865 CrossRef CAS .
  28. G. Wittig and H. D. Frommeld, Angew. Chem., Int. Ed. Engl., 1963, 2, 683 CrossRef .
  29. H. O. House, D. S. Crumrine, A. Y. Teranishi and H. D. Olmstead, J. Am. Chem. Soc., 1973, 95, 3310 CrossRef CAS .
  30. S. Kikkawa and Y. Kondo, Chem. Commun., 2012, 48, 9771 RSC .
  31. H. Taneda, K. Inamoto and Y. Kondo, Chem. Commun., 2014, 50, 6523 RSC .
  32. B. Teng, W. C. Chen, S. Dong, C. W. Kee, D. A. Gandamana, L. L. Zong and C. H. Tan, J. Am. Chem. Soc., 2016, 138, 9935 CrossRef CAS PubMed .
  33. P. C. Marr and A. C. Marr, Green Chem., 2016, 18, 105 RSC .
  34. S. Dewilde, W. Dehaen and K. Binnemans, Green Chem., 2016, 18, 1639 RSC .
  35. K. Erfurt, I. Wandzik, K. Walczak, K. Matuszek and A. Chrobok, Green Chem., 2014, 16, 3508 RSC .
  36. G. H. Tao, L. He, W. S. Liu, L. Xu, W. Xiong, T. Wang and Y. Kou, Green Chem., 2006, 8, 639 RSC .
  37. J. J. Wang, Y. C. Pei, Y. Zhao and Z. G. Hu, Green Chem., 2005, 7, 196 RSC .
  38. G. Y. Zhao, T. Jiang, H. X. Gao, B. X. Han, J. Huang and D. H. Sun, Green Chem., 2004, 6, 75 RSC .
  39. J. L. Yu, Y. Yang, W. T. Chen, D. Xu, H. Guo, K. Li and H. Q. Liu, Green Energy Environ., 2016, 1, 166 CrossRef .
  40. G. Wang, H. Yin, S. F. Yuan and Z. R. Chen, J. Anal. Appl. Pyrolysis, 2015, 116, 27 CrossRef CAS .
  41. G. Wang, H. Yin, S. F. Yuan and Z. R. Chen, J. Anal. Appl. Pyrolysis, 2017, 124, 89 CrossRef CAS .
  42. K. Inamoto, H. Okawa, H. Taneda, M. Sato, Y. Hirono, M. Yonemoto, S. Kikkawa and Y. Kondo, Chem. Commun., 2012, 48, 9771 RSC .
  43. K. Inamoto, Y. Araki, S. Kikkawa, M. Yonemoto, Y. Tanaka and Y. Kondo, Org. Biomol. Chem., 2013, 11, 4438 CAS .
  44. A. Skrzypczak and P. Neta, J. Phys. Chem. A, 2003, 107, 7800 CrossRef CAS .
  45. J. D. Holbrey, W. M. Reichert, M. Nieuwenhuyzen, O. Sheppard, C. Hardacre and R. D. Rogers, Chem. Commun., 2003, 476 RSC .
  46. Y. L. Yang and Y. Kou, Chem. Commun., 2004, 226 RSC .
  47. A. Atsushi and J. F. Liu, J. Org. Chem., 1996, 61, 2590 CrossRef .
  48. C. W. Downey, J. A. Ingersoll, H. M. Glist, C. M. Dombrowski and A. T. Barnett, Eur. J. Org. Chem., 2015, 7287 CrossRef CAS .

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c7gc02265d

This journal is © The Royal Society of Chemistry 2017