Xixi Zhang,
Lei Liu and
Chunbao Li*
Department of Chemistry, School of Science, Tianjin University, Tianjin 300072, China. E-mail: lichunbao@tju.edu.cn; Fax: +86-022-27403475; Tel: +86-022-27892351
First published on 3rd March 2016
The use of large amounts of volatile organic solvents in industrial chemical processes contributes to widespread environmental pollution. To help solve this problem, water and a phase transfer catalyst were used to replace organic solvents in the transformations of bromoacetophenones into chloroacetophenones and aroyl epoxides into aroyl chlorohydrins. The reactions were promoted by sulfonyl chlorides and gave quantitative or close to quantitative yields. Notably, chromatographic purification, which is laborious and consumes large amounts of organic solvents, was not needed. These two processes have opened a green and cost-effective channel to prepare the chemical intermediates chloroacetophenones and aroyl chlorohydrins. The reaction mechanisms are discussed based on control experiments.
Chloroacetophenones and aroylchlorohydrins are two important intermediates in organic synthesis27–30 and the latter are bioactive compounds.31–33 Chloroacetophenones are accessible via the oxidation of acetophenones using chloride sources such as N-chlorosuccinimide,13,34 HCl,35,36 Cl2 (ref. 37) or NaCl38,39 and oxidants such as O2,35 H2O2 (ref. 40) or K2S2O8,39 or via halogen-exchange reactions starting with bromoacetophenones using LiCl,41 NaCl,38,42 HCl,43 TiCl3 (ref. 44) or S4N5·SbCl5 (ref. 45) as chlorination agents. Aroylchlorohydrins are accessible by reacting aroyl epoxides with reagents such as HCl,46 TiCl4,47,48 F3CCO2H/LiCl,49 AcOH/LiCl50 or CeCl3.48
Most of the above reactions are mediated by volatile organic solvents and the yields are less than quantitative. Further, for some aroylchlorohydrins, the stereoselectivities are not satisfactory.47,48 Green solvents such as aqueous CH3CN and ionic liquids have both been applied to these reactions, but generally the yields are poor.13,38
Herein, the preparations of chloroacetophenones using bromoacetophenones and of aroylchlorohydrins using aroyl epoxides are reported. Both reactions were promoted by sulfonyl chlorides mediated by water and most of the reaction yields were quantitative or close to quantitative. Chromatography was not needed to purify the products in these reactions.
Entry | 1a (equiv.) | PhSO2Cl (equiv.) | PTCd (equiv.) | H2O (mL) | Li2CO3 (equiv.) | t (h) | 2a yieldb (%) |
---|---|---|---|---|---|---|---|
a Reaction conditions: bromoacetophenone (300 mg), benzenesulfonyl chloride, base, phase transfer catalyst (PTC), H2O at rt.b All yields were isolated yields.c NR: no reaction.d TBAB = tetrabutyl ammonium bromide; Aliquat 336 = tricapryl methyl ammonium chloride; TBAF = tetrabutyl ammonium fluoride; SDS = sodium dodecyl sulfonate; DTAC = dodecyl trimethyl ammonium chloride. | |||||||
1 | 1.0 | 6.0 | CTMAB (0.1) | 5.0 | 1.2 | 1.8 | 83 |
2 | 1.0 | 6.0 | CTMAB (0.1) | — | 1.2 | 2.3 | 91 |
3 | 1.0 | 6.0 | — | 5.0 | 1.2 | 24 | NRc |
4 | 1.0 | 6.0 | CTMAB (0.1) | 5.0 | — | 4.0 | >99 |
5 | 1.0 | 6.0 | TBAB (0.1) | 5.0 | — | 3.5 | >99 |
6 | 1.0 | 6.0 | Aliquat 336 (0.1) | 5.0 | — | 3.5 | >99 |
7 | 1.0 | 6.0 | TBAF (0.1) | 5.0 | — | 4.0 | >99 |
8 | 1.0 | 6.0 | SDS (0.1) | 5.0 | — | 6.7 | >99 |
9 | 1.0 | 6.0 | DTAC (0.1) | 5.0 | — | 3.6 | >99 |
10 | 1.0 | 6.0 | TEBAC (0.1) | 5.0 | — | 3.0 | >99 |
11 | 1.0 | 0.1 | TEBAC (0.1) | 5.0 | — | 24 | 9.5 |
12 | 1.0 | 0.5 | TEBAC (0.1) | 5.0 | — | 24 | 19 |
13 | 1.0 | 1.0 | TEBAC (0.1) | 5.0 | — | 24 | 52 |
14 | 1.0 | 4.0 | TEBAC (0.1) | 5.0 | — | 7.0 | >99 |
15 | 1.0 | 8.0 | TEBAC (0.1) | 5.0 | — | 2.8 | >99 |
16 | 1.0 | 10.0 | TEBAC (0.1) | 5.0 | — | 2.8 | >99 |
17 | 1.0 | 20.0 | TEBAC (0.1) | 5.0 | — | 2.8 | >99 |
18 | 1.0 | 8.0 | TEBAC (0.01) | 5.0 | — | 24 | 21 |
19 | 1.0 | 8.0 | TEBAC (0.03) | 5.0 | — | 24 | 65 |
20 | 1.0 | 8.0 | TEBAC (0.05) | 5.0 | — | 6.0 | 91 |
21 | 1.0 | 8.0 | TEBAC (0.08) | 5.0 | — | 3.7 | >99 |
22 | 1.0 | 8.0 | TEBAC (0.3) | 5.0 | — | 2.1 | >99 |
23 | 1.0 | 8.0 | TEBAC (0.5) | 5.0 | — | 1.5 | >99 |
24 | 1.0 | 8.0 | TEBAC (2.0) | 5.0 | — | 1.5 | >99 |
25 | 1.0 | 8.0 | TEBAC (0.5) | 2.0 | — | 1.5 | >99 |
26 | 1.0 | 8.0 | TEBAC (0.5) | 8.0 | — | 2.0 | >99 |
27 | 1.0 | 8.0 | TEBAC (0.5) | 10.0 | — | 2.0 | >99 |
Under the optimized conditions, fourteen bromoacetophenones were successfully transformed into chloroacetophenones (Scheme 1, 2a–2n). All the yields were quantitative and the reaction rates fast (0.5–3.5 h). At the end of the reactions, the excess PhSO2Cl was hydrolyzed using Na2CO3 and then the products were extracted. Then simple concentrations resulted in pure products. The reactions could be scaled up to 10 g with the same results. All the reaction mixtures became milky liquids under agitation in both 0.3 g and 10 g scales, which probably indicates that they are micelle-mediated reactions (ESI, S2, Pictures 1 and 2†).51,52
The yields for the synthesis of these products reported in the literature are in the range of 36–100% and all the procedures utilize volatile organic solvents and require aqueous workups and chromatography to purify the products.13,34–40,42–45,53 In this study, external heating, organic solvents as reaction media and as elutes for chromatography were saved.
As shown in Scheme 1, the reaction rates for the substrates bearing electron-donating groups such as alkyl or alkoxyl groups were faster than those for the substrates bearing electron-withdrawing groups such as halo and nitro groups (2f–2j vs. 2b–2e). The rates for meta-substituted groups were similar to those for para-substituted groups (2k vs. 2b, 2l vs. 2e). However, ortho-substituted groups had substantially slower reaction rates (2b vs. 2m, 2i vs. 2n), which is probably due to steric effects.
For comparison, nine common organic solvents were used as the reaction media instead of water and PTC (Table 2, entries 1–9). For eight of the solvents, no reactions occurred (Table 2, entries 2–9). Only the reaction mediated by acetonitrile gave the desired product quantitatively (Table 2, entry 1). The amount of PhSO2Cl was then varied between 0.1 and 10.0 equiv. (Table 2, entries 1 and 10–19) and the acetonitrile volume was adjusted from 2.0 to 30.0 mL (Table 2, entries 1 and 20–25). The optimized conditions were PhSO2Cl (6.0 equiv.) and CH3CN (4.0 mL), which produced the product in quantitative yields in 2.5 h (Table 2, entry 21).
Entry | 1a (equiv.) | PhSO2Cl (equiv.) | Solventd (mL) | t (h) | 2a yieldb (%) |
---|---|---|---|---|---|
a Reaction conditions: bromoacetophenone (300 mg), benzenesulfonyl chloride, CH3CN at 85 °C.b All yields were isolated yields.c NR: no reaction.d THF = tetrahydrofuran; EA = ethyl acetate. | |||||
1 | 1.0 | 6.0 | CH3CN (10.0) | 3.0 | >99 |
2 | 1.0 | 6.0 | CH3OH (10.0) | 24 | NRc |
3 | 1.0 | 6.0 | CH2Cl2 (10.0) | 24 | NR |
4 | 1.0 | 6.0 | CHCl3 (10.0) | 24 | NR |
5 | 1.0 | 6.0 | THF (10.0) | 24 | NR |
6 | 1.0 | 6.0 | EA (10.0) | 24 | NR |
7 | 1.0 | 6.0 | Toluene (10.0) | 24 | NR |
8 | 1.0 | 6.0 | Cyclohexane (10.0) | 24 | NR |
9 | 1.0 | 6.0 | 1,4-Dioxane (10.0) | 24 | NR |
10 | 1.0 | 0.1 | CH3CN (10.0) | 24 | Trace |
11 | 1.0 | 0.5 | CH3CN (10.0) | 24 | 10 |
12 | 1.0 | 1.0 | CH3CN (10.0) | 24 | 65 |
13 | 1.0 | 1.2 | CH3CN (10.0) | 17 | 84 |
14 | 1.0 | 1.5 | CH3CN (10.0) | 9.0 | 94 |
15 | 1.0 | 2.0 | CH3CN (10.0) | 5.0 | 96 |
16 | 1.0 | 3.0 | CH3CN (10.0) | 5.0 | 96 |
17 | 1.0 | 4.0 | CH3CN (10.0) | 4.0 | >99 |
18 | 1.0 | 8.0 | CH3CN (10.0) | 3.5 | >99 |
19 | 1.0 | 10.0 | CH3CN (10.0) | 3.5 | >99 |
20 | 1.0 | 6.0 | CH3CN (2.0) | 5.5 | >99 |
21 | 1.0 | 6.0 | CH3CN (4.0) | 2.5 | >99 |
22 | 1.0 | 6.0 | CH3CN (6.0) | 2.5 | >99 |
23 | 1.0 | 6.0 | CH3CN (8.0) | 3.0 | >99 |
24 | 1.0 | 6.0 | CH3CN (20.0) | 7.0 | >99 |
25 | 1.0 | 6.0 | CH3CN (30.0) | 12 | >99 |
These optimized conditions were then applied to the substrates 2a–2n in Scheme 1 again. As in the water-mediated reactions, all the reactions gave quantitative yields of products, and a similar electronic effect on the reaction rates was observed. However, in most cases, the acetonitrile-mediated reactions had substantial slower reaction rates. The fact that halogen exchange reactions do not occur in most of the organic solvents and occur in water indicates that the hydrophobic core effect51,52 for the possible micelle-mediated reactions enables the reactions.
A plausible mechanism for the reaction between bromoacetophenone and benzenesulfonyl chloride is shown in Scheme 2. First, benzenesulfonate I is formed via a reaction between bromoacetophenone and benzenesulfonyl chloride. The opening of the bromonium ring by chloride then leads to the formation of II. This type of compounds (II) is not known in the literature, supposed to be very reactive with water because of the two geminal good leaving groups on the benzylic position. The hydrolysis of II results the final product.
![]() | ||
Scheme 2 A possible mechanism for the reaction between bromoacetophenone and benzenesulfonyl chloride. |
Next, a series of experiments were performed to verify the mechanism and to better understand the role of sulfonyl chloride (Scheme 3). Treating bromoacetophenone with MsOH/LiCl, HCl, or PhSO3H/LiCl in water (Scheme 3, (A), conditions 1, 2 and 3) resulted in no reaction. This excludes that proton acids catalyzed the reaction since proton acids can be generated from the hydrolysis of sulfonyl chlorides. No reaction took place in the presence of LiCl or NaCl (Scheme 3, (A), conditions 4 and 5) when no sulfonyl chloride was present. This points to the indispensible role of sulfonyl chloride. Moreover, chloroacetophenone could not be transformed into bromoacetophenone in the presence of PhSO2Cl and KBr in micellar media (Scheme 3, (A), condition 6). The possible reason is that the three-membered bromonium (Scheme 2, I) is present in substantial amounts which facilitates the reaction while the three-membered chloronium is only present in trace amounts according to previous reports.53,54 For this reason, acetoxyacetophenone and α-hydroxyacetophenone could not be transformed into chloroacetophenones in the presence of PhSO2Cl in water (Scheme 3, (B) and (C)). Since the electron-donating group helps stabilize the positive charge of I to facilitate the reaction, the reaction rates for the substrates bearing electron-donating groups such as alkyl or alkoxyl groups were faster than those for the substrates bearing electron-withdrawing groups such as halo and nitro groups (Scheme 1, 2f–2j vs. 2b–2e).
Condition 1: bromoacetophenone (100 mg), methylsulfonic acid (8.0 equiv.), LiCl (8.0 equiv.), TEBAC (0.5 equiv.), H2O (0.7 mL), rt, NR. Condition 2: bromoacetophenone (100 mg), HCl (aq.) (8.0 equiv.), TEBAC (0.5 equiv.), H2O (0.7 mL), rt, NR. Condition 3: bromoacetophenone (100 mg), benzenesulfonic acid (8.0 equiv.), LiCl (8.0 equiv.), TEBAC (0.5 equiv.), H2O (0.7 mL), rt, NR. Condition 4: bromoacetophenone (100 mg), LiCl (8.0 equiv.), TEBAC (0.5 equiv.), H2O (0.7 mL), rt, NR. Condition 5: bromoacetophenone (100 mg), NaCl (8.0 equiv.), TEBAC (0.5 equiv.), H2O (0.7 mL), rt, NR. Condition 6: chloroacetophenone (100 mg), benzenesulfonyl chloride (8.0 equiv.), KBr (8.0 equiv.), benzyltriethylammonium bromide (BTEAB) (0.5 equiv.), H2O (0.7 mL), rt, NR.
It is known that bromoacetophenones are more reactive than ordinary alkyl bromides. A number of mechanisms have been proposed to explain this.55–60 Lewis et al. reported that an enolate structure contributes to the transition state (Scheme 4, A).55 If this mechanism were operating in our case, an enol benzenesulfonate (Scheme 4, B) from A and benzenesulfonyl chloride would be the final product, which is known to be a type of stable compounds.61,62 However it was not found in our crude products. Although our mechanism is different, it is in agreement with all the experimental data. The following experiments on the synthesis of chlorohydrins were performed in order to provide further support for our mechanism.
Similar aqueous reaction conditions were used to transform benzoyl epoxide 3a to benzoyl chlorohydrin 4a (Table 3). Neither the yields nor reaction rates were satisfactory with 1.5 or 6.0 equiv. of PhSO2Cl (Table 3, entries 1 and 2). With 1.5 equiv. of MsCl, the reaction yield was 69% and when the amount of MsCl was increased to 3.0, 4.5, 6.0 or 7.5 equiv., the yields were improved up to 97% (Table 3, entries 3–7). Using 6.0 or 7.5 equiv. of MsCl gave the same reaction rate, so the optimized reaction conditions use 6.0 equiv. of MsCl (Table 3, entry 6).
Entry | 3a (equiv.) | RSO2Cl (equiv.) | t (h) | 4a yieldb (%) |
---|---|---|---|---|
a Reaction conditions: 3a (500 mg), RSO2Cl, TEBAC (0.5 equiv.), H2O (4 mL) at rt.b All yields were isolated yields. | ||||
1 | 1.0 | PhSO2Cl (1.5) | 24 | 57 |
2 | 1.0 | PhSO2Cl (6.0) | 16 | 93 |
3 | 1.0 | MsCl (1.5) | 24 | 69 |
4 | 1.0 | MsCl (3.0) | 13 | 96 |
5 | 1.0 | MsCl (4.5) | 8.5 | 97 |
6 | 1.0 | MsCl (6.0) | 5.5 | 97 |
7 | 1.0 | MsCl (7.5) | 5.5 | 97 |
Altogether 8 aroyl epoxides were subjected to the optimized conditions (Scheme 5, 4a–4h) and chlorohydrins were obtained in close to quantitative yields. The reactions were stereoselective, leading to anti-stereoisomers as the major products (anti/syn = 15.7:
1 to 3.8
:
1). All the reactions except the one for 4b were conducted at rt and were complete in 4.5–12 h. The reactions were worked up by washing the extracted ethyl acetate layer with water to get rid of excess MsCl. Simple concentrations yielded pure products.
The syntheses of open chain chlorohydrins reported in the literature require organic solvent-mediated reactions and chromatography for purification.46–50 Those reaction yields range from 40–95%47–50 and in some cases, the stereoselectivities of the products were poor.47,48 Therefore, our procedure compares favorably to current procedures in terms of yield, product purification, stereoselectivity and greenness of reaction media. In addition, the reaction scale can easily be increased to 10 g with the same results.
Based on these results, the following mechanism is plausible (Scheme 6). The addition of Cl− to the SO bond of MsCl leads to I, which then adds to the C
O bond of aroyl epoxide to form II. Mesylation of II by MsCl forms III. The intramolecular opening of the epoxy ring by chloride then gives IV. After hydrolysis, the chlorohydrin is produced.
To verify the mechanism, additional investigations were performed with aroyl epoxide 3f (Scheme 7). The ring-opening reaction did not occur with MsOH/LiCl, HCl, LiCl or NaCl in water, indicating that this reaction is not catalyzed by proton acids and that MsCl is indispensable for promoting the reaction. The enolate mechanism put forward by Lewis et al.55 cannot explain the higher reaction rates for substrates bearing electron-deficient phenyl rings compared to those for substrates bearing electron-rich phenyl rings (Scheme 5, 4d–4g vs. 4h). However, this fact is in agreement with our mechanism. The hydrophobic core effect51,52 is assumed to be crucial for the formation of intermediates I and II (Scheme 6).
![]() | ||
Scheme 7 Reactions to verify the mechanism in Scheme 6. Condition 1: aroyl epoxide 3f (100 mg), methylsulfonic acid (8.0 equiv.), LiCl (8.0 equiv.), TEBAC (0.5 equiv.), H2O (0.7 mL), rt, NR. Condition 2: aroyl epoxide 3f (100 mg), HCl (aq.) (8.0 equiv.), TEBAC (0.5 equiv.), H2O (0.7 mL), rt, NR. Condition 3: aroyl epoxide 3f (100 mg), LiCl (8.0 equiv.), TEBAC (0.5 equiv.), H2O (0.7 mL), rt, NR. Condition 4: aroyl epoxide 3f (100 mg), NaCl (8.0 equiv.), TEBAC (0.5 equiv.), H2O (0.7 mL), rt, NR. |
Footnote |
† Electronic supplementary information (ESI) available: Pictures of the synthesis of 2a, general experimental information, general procedure for all products, analytic data of all products, references and NMR spectra for all compounds. See DOI: 10.1039/c6ra00433d |
This journal is © The Royal Society of Chemistry 2016 |