Miao
Ding
,
Feng
Zhou
,
Yun-Lin
Liu
,
Cui-Hong
Wang
,
Xiao-Li
Zhao
and
Jian
Zhou
*
Shanghai Key Laboratory of Green Chemistry and Chemical Processes, Department of Chemistry, East China Normal University, 3663N, Zhongshan Road, Shanghai 200062, China. E-mail: jzhou@chem.ecnu.edu.cn; Fax: (+86) 21-6223-4560
First published on 3rd August 2011
A newly developed cinchonidine-derived phosphoramide 6b, simple and easily available, was identified as a powerful catalyst for the highly enantioselective Michael addition of both unprotected 3-aryl and 3-alkyloxindoles to β-substituted nitroalkenes to furnish the C3 quaternary stereogenic carbon center with an adjacent tertiary stereocenter in up to 21:
1 diastereoselectivity and up to 99% enantioselectivity.
Recently, the Michael addition of 3-substituted oxindoles 1 to nitroolefins 3 emerged as a versatile method for the synthesis of oxindole and indoline derivatives bearing a quaternary center at the C3 position.5,6 The highly enantioselective conjugate addition of N-Boc protected 3-alkyloxindoles to nitroolefins was pioneered by Barbas III5a and Shibasaki,5b independently. Later, Maruoka found that the high reactivity of N-Boc protected 3-aryloxindoles allowed a novel enantioselective base-free phase-transfer reaction with nitroolefins in a water rich solvent.5c Luo & Cheng reported the addition of N-phenyl 3-methyloxindole and nitroolefins.5d Most recently, Yuan reported the highly enantioselective conjugate addition of N-Boc protected 3-alkyloxindoles to protected 2-amino-1-nitroethenes.5e Despite these achievements, there is still room for further improvement judging by the guidelines of the “ideal synthesis”.4b First, it would be more atom-efficient to use unprotected 3-prochiral oxindoles 2 as the nucleophile, because the synthesis of N-Boc oxindoles 1 from isatins involves three steps, with the sacrifice of one equiv. of (Boc)2O (eqn (1)), and the direct protection of 2 with (Boc)2O was known to be problematic due to the N- and O-selectivity.7
![]() | (1) |
Since Shibata and Toru pioneered the catalytic asymmetric aldol-type reaction of unprotected 3-alkyloxindoles with trifluoropyruvate catalyzed by cinchona alkaloids,8a unprotected 3-prochiral oxindoles have been utilized in the catalytic asymmetric synthesis of 3,3-disubstituted oxindolesvia aldol,8a–c Mannich,8damination8e–g and Michael addition reactions.6b–c However, to the best of our knowledge, most of these protocols were limited to either unprotected 3-aryl or 3-alkyl oxindoles. While a variety of Michael addition reactions of 3-prochiral oxindoles to electron-deficient olefins have been developed,5,6 very limited examples were based on unprotected oxindole 2,6b–c and no method enabled both 3-alkyl and 3-aryloxindoles to work well with β-substituted Michael acceptors in a highly enantioselective manner. Here we wish to report a newly developed phosphoramide 6b as a powerful catalyst for the highly enantioselective Michael addition of unprotected 3-aryl and 3-alkyloxindoles to nitroolefins.
Generally, unprotected 3-substituted oxindoles were less reactive nucleophiles, and the pKa value of the C–H bond at the C3 position could be expected to be substantially high, as that of oxindole was 18.2.10 Accordingly, the use of a bifunctional catalyst to activate both the oxindole and nitroalkene might be important for reaction development. Therefore, we designed cinchona alkaloid-derived phosphoramides 6 for the following reasons: 1) amide substituents R1 could serve as shielding groups for two directions, helpful for the stereocontrolled creation of adjacent quaternary and tertiary stereocenters.11 More, it was easy to vary the R1group to tune the steric and electronic properties of the catalysts; 2) the pKa value of the N–H bond of the amide could be easily tuned by varying the R1group from an aryl to an alkoxy group, very useful for the electrophile activation;12 3) these kinds of catalysts were largely unexplored,13 although they could be easily accessed in one step from a cinchona alkaloid-derived primary amine. Only a quinidine-derived phosphinamide catalyst was reported by our group for the enantioselective Strecker reaction of isatin-derived ketoimines using TMSCN.9c
Phosphinamide
6a and phosphoramides 6b–c, obtained in one step from a cinchonidine-derived primary amine and the corresponding phosphinic chloride or chlorophosphates, were first examined in the reaction of unprotected 3-phenyloxindole 2a and nitroalkene 3a in dichloromethane at −10 °C (Table 1). To our delight, all the three catalysts afforded the desired product 4a in excellent ee (entries 1–3). Although phosphinamide 6a could deliver 99% ee for the major diastereomer, the diastereoselectivity was poor (entry 1). Interestingly, phosphoramide 6b gave product 4a in excellent yield, 8:
1 dr and 95% ee for the major diastereomer (entry 2). The diphenyl substituted catalyst 6c provided product 4a in slightly lower dr and ee than catalyst 6b (entry 3). These initial results encouraged us to prepare quinine- and cinchonine-derived catalysts 6d and 6e to further improve the selectivity, but both of them were inferior to 6b (entries 4–5). We also synthesized cinchonidine-derived sulfamide catalysts 7a–d14 to compare with catalyst 6b, but they were less diastereo- and enantioselective than catalyst 6b (entries 6–9). On the basis of the above results, the simple and easily available catalyst 6b was chosen for further optimization. Diethyl ether turned out to be the best solvent, and up to 14.0
:
1.0 dr was achieved for product 4a with 95% ee (entry 10). Lowering the temperature to −20 °C could further improve the dr, but it took four days for the reaction to complete (entry 11). The addition of powdered MS 5 Å could promote the reaction to finish within two days at −20 °C, affording the desired product in 96% yield, with up to 21.0
:
1.0 dr and 98% ee for the major diastereomer (entry 12).
Entrya | Cat. | Solvent | T /°C | Yieldb (%) | Drc | Ee d (%) (syn/anti) |
---|---|---|---|---|---|---|
a Run on a 0.10 mmol scale. b Isolated yield. c Determined by 1H NMR analysis of the crude reaction mixture. d Determined by chiral HPLC analysis. e 1.0 g/1.0 mmol powdered MS 5 Å added. | ||||||
1 | 6a | DCM | −10 | 87 | 2.0/1.0 | 99/85 |
2 | 6b | DCM | −10 | 96 | 8.0/1.0 | 95/70 |
3 | 6c | DCM | −10 | 97 | 6.4/1.0 | 92/72 |
4 | 6d | DCM | −10 | 96 | 10.0/1.0 | 87/49 |
5 | 6e | DCM | −10 | 76 | 6.0/1.0 | 92/84 |
6 | 7a | DCM | −10 | 95 | 8.0/1.0 | 83/70 |
7 | 7b | DCM | −10 | 77 | 7.3/1.0 | 81/71 |
8 | 7c | DCM | −10 | 95 | 7.0/1.0 | 81/62 |
9 | 7d | DCM | −10 | 95 | 7.1/1.0 | 77/64 |
10 | 6b | Et2O | −10 | 90 | 14.0/1.0 | 95/66 |
11 | 6b | Et2O | −20 | 89 | 18.0/1.0 | 96/71 |
12e | 6b | Et2O | −20 | 96 | 21.0/1.0 | 98/73 |
Based on the above results, the substrate scope with respect to both the electrophile and nucleophile was investigated under the optimal conditions: to run the reaction in air, using Et2O as the solvent in the presence of 10 mol% catalyst 6b and MS 5 Å (Table 2). The diastereoselectivity of product 4 was determined by 1H NMR analysis of the crude reaction mixtures, or by HPLC analysis when NMR analysis failed. The reaction of unprotected 3-aryloxindoles with nitroalkenes 3 was first examined. Various nitrostyrene derivatives with electron-rich or -poor substituents on the aromatic ring afforded the desired products 4a–d in excellent ee for the major diastereomer, with high to excellent dr (entries 1–4). Due to the problems in separating the diastereomers and enantiomers by HPLC analysis, (E)-2-(2-nitrovinyl)furan had to work with 5-fluoro-3-phenyloxindole, and product 4e was obtained in excellent ee with moderate dr (entry 5). Aliphatic nitroolefins also worked well at 0 °C to afford the desired products 4f–g in good to high yield and dr, with up to 98% ee for the major diastereomer (entries 6–7). Different 3-aryloxindoles were also examined, and the corresponding products 4h–n were all obtained in high yield and dr, with excellent ee for the major diastereomer (entries 8–14). It should be noted that both the substrate scope and selectivity obtained by phosphoramide 6b in the Michael addition of unprotected 3-aryloxindoles to nitroolefins were obviously better than the literature report which was based on N-Boc protected 3-aryloxindoles.5c
Entrya | 2 | R2 | 4 | Yieldb(%) | Drc | Ee d (%) |
---|---|---|---|---|---|---|
a Run on a 0.25 mmol scale. b Isolated yield. c Determined by 1H NMR analysis of the crude reaction mixture or by HPLC analysis. d Determined by HPLC analysis, for major diastereomer. e −20 °C. f −40 °C. g 0 °C. h Room temperature. i 20 mol% catalyst. | ||||||
1e | R = H, R1 = Ph | 2,4-Cl2C6H3 | 4a | 96 | 21![]() ![]() |
98 |
2e | R = H, R1 = Ph | p-BrC6H4 | 4b | 89 | 13![]() ![]() |
99 |
3e | R = H, R1 = Ph | p-CF3C6H4 | 4c | 92 | 13![]() ![]() |
98 |
4e | R = H, R1 = Ph | p-MeOC6H4 | 4d | 85 | 7![]() ![]() |
97 |
5f | R = F, R1 = Ph | 2-Furanyl | 4e | 84 | 3![]() ![]() |
96 |
6g | R = H, R1 = Ph | n-Pr | 4f | 76 | 6![]() ![]() |
98 |
7g | R = H, R1 = Ph | i-Bu | 4g | 84 | 6![]() ![]() |
98 |
8f | R = F, R1 = Ph | Ph | 4h | 84 | 10![]() ![]() |
96 |
9f | R = Cl, R1 = Ph | Ph | 4i | 92 | 8![]() ![]() |
97 |
10f | R = Br, R1 = Ph | Ph | 4j | 78 | 7![]() ![]() |
97 |
11e | R = OMe, R1 = Ph | Ph | 4k | 83 | 11![]() ![]() |
98 |
12f | R = H,R1 = p-ClC6H4 | Ph | 4l | 84 | 18![]() ![]() |
97 |
13e | R = H, R1 = 2-Naphthyl | Ph | 4m | 77 | 11![]() ![]() |
99 |
14f | R = F, R1 = Ph | p-CF3C6H4 | 4n | 85 | 9![]() ![]() |
98 |
15h,i |
![]() |
p-BrC6H4 | 4o | 78 | 8![]() ![]() |
94 |
16h,i |
![]() |
p-BrC6H4 | 4p | 89 | 8![]() ![]() |
95 |
17h,i |
![]() |
p-BrC6H4 | 4q | 95 | 7![]() ![]() |
96 |
18h,i |
![]() |
p-BrC6H4 | 4r | 95 | 8![]() ![]() |
91 |
19h,i | R = Br, R1 = Me | p-BrC6H4 | 4s | 77 | 9![]() ![]() |
95 |
20h,i | R = Br, R1 = Me | 2-Thienyl | 4t | 86 | 12![]() ![]() |
95 |
Most importantly, unprotected 3-alkyloxindole turned out to be viable substrates under this reaction condition. Unprotected 3-alkyloxindoles were much less reactive than unprotected 3-aryloxindoles, because their pKa value might be substantially higher than that of oxindole (18.2). As a result, the reaction of unprotected 3-alkyloxindole and nitroolefins was run at room temperature with 20 mol% of catalyst 6b. Because of the problem in separating isomers by HPLC analysis, bromo-substituted nitrostyrene was used as the acceptor, and we were pleased to find that different 3-alkyloxindoles worked well to give the major diastereomer in high diastereoselectivity and excellent enantioselectivity (entries 15–19). Heteroaromatic nitroalkene also worked well with unprotected 3-alkyloxindole to give the major diastereomer of product 4t in high dr and excellent ee (entry 20).
The relative configuration of the major diastereomer of product 4a was confirmed by the X-ray analysis (Fig. 1). The configuration of the stereogenic center of the major diastereomer of product 4m at the C3 position was also assigned to be R and that at the remaining stereocenter was R (see ESI†). Those of other products 4 were tentatively assigned by analogy.
![]() | ||
Fig. 1 X-ray structure of 4a. |
The high diastereo- and enantioselectivity realized by catalyst 6b in this reaction was very impressive. Especially, our method was workable for both unprotected 3-aryl and 3-alkyloxindoles, which was more atom-efficient than literature methods that are based on N-Boc protected 3-prochiral oxindoles. Furthermore, catalyst 6b was very simple and could be easily obtained in only two steps from cheap cinchonidine. These facts made our method very attractive for the synthesis of enantioenriched oxindole or indoline derivatives bearing a quaternary center at the C3 position. These natural product analogues might possess important bioactivity, helpful for drug development.
For example, the reaction of oxindole 2a and nitrostyrene 3h afforded product 4u in 95% yield and 11:
1 dr, which could be readily oxidized to acid 8 by a literature method (see Scheme 1).15 The major diastereomer of the oxindole based acid 8 could be isolated in 75% yield with 95% ee, which was an interesting structural motif for medicinal research.2a–g The Michael adduct 4u could also be converted to compound 9 after the reduction and the following cyclization, affording the major diastereomer in 65% yield with 96% ee. The relative and absolute configuration of compound 9 was in accordance with literature report.5c It also should be noted that the isomers of product 4u could not be separated by chiral HPLC analysis, but the ee of compound 8 and 9 was readily determined.
![]() | ||
Scheme 1 Product elaboration. |
Footnote |
† Electronic supplementary information (ESI) available. CCDC reference numbers 827254. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c1sc00390a |
This journal is © The Royal Society of Chemistry 2011 |