Peng-Qiao
Chen
a,
You-Cai
Xiao
a,
Cai-Zhen
Yue
a and
Ying-Chun
Chen
*ab
aKey Laboratory of Drug-Targeting and Drug Delivery System of the Education Ministry, West China School of Pharmacy, Sichuan University, Chengdu 610041, China. E-mail: ycchen@scu.edu.cn; Fax: +86 28 85502609
bCollege of Pharmacy, Third Military Medical University, Chongqing 400038, China
First published on 1st April 2014
An array of deconjugated linear 3,5-dienones with substantial substitutions have been successfully used in β,ε-regioselective Diels–Alder cycloadditions with 3-olefinic oxindole-based dienophiles via trienamine catalysis of cinchona-based primary amines, efficiently producing spirocyclic oxindole architectures with dense and diverse substitutions in high stereoselectivity (up to 99% ee, >19:1 d.r.).
In fact, almost no reaction occurred for the combination of linear 2,4-dienone 1 with an α′-enolizable methyl group and diverse dienophiles in the presence of a primary amine catalyst, probably because the cross-conjugated trienamine intermediate I would be generated, as outlined in Scheme 1. Very recently, we developed a new strategy, in which the previously positioned δ,ε-CC bond could perform as an inducing group for the formation of linear trienamines from interrupted cyclic 2,5-dienones bearing an α′-CH group.4 Inspired by such observation, we envisaged that similar inducing effect would be applicable to the deconjugated linear 3,5-dienone 2a, rendering the formation of the requisite extended trienamine intermediate and facilitating the later β,ε-regioselective cycloaddition process.
Based on the above-mentioned considerations, the initial investigation was set up for the model reaction of 6-methylhepta-4,6-dien-2-one 2a, which was effectively obtained by the deconjugation of 6-methylhepta-3,5-dien-2-one 1,5 with 3-olefinic oxindole 3a,6 in the presence of 9-amino-9-deoxyepiquinine (C1, 20 mol%) and salicylic acid (SA, 40 mol%).7 To our gratification, the desired Diels–Alder reaction occurred smoothly in toluene at room temperature, and the spirocyclic oxindole 4a was produced in remarkable stereoselectivity and with a good yield (Table 1, entry 1).8 It should be pointed out that the cycloaddition reaction did not occur in the absence of a primary amine catalyst, indicating that the HOMO-activation of the β,γ,δ,ε-diene moiety by the formation of a conjugated enamine intermediate is necessary. Consequently, other catalytic parameters were briefly screened. Amine C2 with a free OH group provided a decreased ee value and lower yield (entry 2). In addition, both 9-amino-9-deoxyepiquinidine (C3) and 9-amino-9-deoxyepicinchonine (C4) delivered the product with an opposite configuration, also in high stereoselectivity (entries 3 and 4). The cycloaddition could take place in an array of solvents, though diminished enantioselectivity was generally observed (entries 5–10). Lower enantiocontrol was also produced when benzoic acid (BA) or o-fluorobenzoic acid (OFBA) was used (entries 11 and 12). In addition, the reaction proceeded smoothly with 10 mol% of the catalyst, while the enantiocontrol was slightly reduced (entry 13). Similar inferior data were observed on a larger scale under the optimized conditions (entry 14).
Entry | 1 | Solvent | Acid | Yieldb (%) | eec (%) |
---|---|---|---|---|---|
a Unless noted otherwise, reactions were performed with 3,5-dienone 2a (0.2 mmol), 3-olefinic oxindole 3a (0.1 mmol), amine C (20 mol%) and acid (40 mol%) in a solvent (1.0 mL) at rt in 10 h. b Yield of the isolated product. c Determined by HPLC analysis on a chiral column; d.r. > 19:1 by 1H NMR analysis. d With 10 mol% of C1 and 20 mol% SA for 18 h. e At 1.0 mmol scale for 12 h. | |||||
1 | C1 | Toluene | SA | 91 | 96 |
2 | C2 | Toluene | SA | 78 | 87 |
3 | C3 | Toluene | SA | 86 | −94 |
4 | C4 | Toluene | SA | 86 | −90 |
5 | C1 | CHCl3 | SA | 84 | 92 |
6 | C1 | DCE | SA | 86 | 87 |
7 | C1 | DCM | SA | 86 | 89 |
8 | C1 | THF | SA | 91 | 90 |
9 | C1 | Dioxane | SA | 84 | 92 |
10 | C1 | MeCN | SA | 78 | 67 |
11 | C1 | Toluene | BA | 84 | 87 |
12 | C1 | Toluene | OFBA | 89 | 89 |
13d | C1 | Toluene | SA | 86 | 87 |
14e | C1 | Toluene | SA | 89 | 87 |
With the optimal reaction conditions in hand, we then investigated a variety of 3,5-dienones and 3-olefinic oxindoles catalyzed by amine C1 (20 mol%) and SA (40 mol%) in toluene at room temperature. The results are summarized in Table 2. At first, an array of deconjugated 3,5-dienones 2 with diverse substituted patterns was investigated in the reaction with dienophile 3a. Both aliphatic and aromatic substituents could be well compatible in the γ- or δ-position of 3,5-dienones 2, delivering the corresponding cycloadducts 4a–4e without apparent influence on the reaction outcome. It was gratifying that γ,ε-disubstituted 3,5-dienones exhibited good reactivity, and products 4f and 4g were obtained in good yields with excellent diastereo- and enantioselectivity. When a single ethyl group was introduced at the ε-position of a 3,5-dienone, the desired cycloadduct 4h was obtained in a moderate yield but still with outstanding enantioselectivity. Notably, other 3,5-dienones carrying a larger α′-alkyl group could also be successfully employed to obtain cycloadducts 4i and 4j in good yields. A 3,5-dienone with an α′-phenyl group without substitution on the backbone still exhibited acceptable reactivity, though modest yield and enantioselectivity were attained for product 4k. Unfortunately, a simple hepta-4,6-dien-2-one showed much lower reactivity, and a very poor yield with bad stereoselectivity was observed. Moreover, the results seemed reasonable when the benzoyl group was replaced by an ethoxycarbonyl group, and the cycloadduct 4l was produced in high yield and with prominent enantioselectivity, but 3-olefinic oxindole with a β-phenyl group (R5 = Ph) showed no reactivity. On the other hand, a spectrum of oxindole-based dienophiles 3 with diverse electron-withdrawing or -donating substituents was tested in the reaction with 6-methylhepta-4,6-dien-2-one 2a. Pleasingly, the corresponding cycloadducts 4m–4u were generally obtained with similar good results. A few substrates were further explored by the catalysis of amine C3, generally affording the cycloadducts with an opposite configuration in excellent enantioselectivity (see more data in parentheses). It should be pointed out that remarkable diastereoselectivity (d.r. > 19:1) was universally achieved in the illustrated reactions.
a Unless noted otherwise, reactions were performed with 3,5-dienones 2 (0.2 mmol), dienophiles 3 (0.1 mmol), amine C1 (20 mol%) and SA (40 mol%) in toluene (1.0 mL) at rt for 10 h. b Isolated yield. c ee was determined by chiral HPLC analysis; d.r. >19:1. d The absolute configuration of 4i was determined by X-ray analysis, see ESI. Other products were assigned by analogy. e Data in parentheses were obtained with amine C3. |
---|
To further expand the utility of this strategy, more electron-deficient dienophiles were explored in the reactions with 3,5-dienone 2a under the similar catalytic conditions. It was found that a good ee value with a high yield could be obtained for cycloadduct 6 using maleimide 5 with an N-aryl group (Scheme 2). However, attempts to improve the stereocontrol at a lower temperature resulted in no success. On the other hand, other potential dienophiles, such as β-nitrostyrene, were further investigated, but generally failed to give the desired cycloadducts, and decomposition or other transformations of 3,5-dienone 2a were observed when harsher conditions were applied.9 Therefore, more research remains to be conducted to expand the application of the current catalytic protocol.
In conclusion, we have successfully accomplished the trienamine catalysis of linear polyunsaturated ketones with substantial substitutions using deconjugated 3,5-dienones as the substrates, which was previously not feasible with conjugated 2,4-dienone analogues. Highly stereoselective and β,ε-regioselective Diels–Alder cycloadditions were developed with 3-olefinic oxindole-based dienophiles under the catalysis of readily available cinchona-based primary amines, efficiently producing a spectrum of spirocyclic oxindole architectures with dense and diverse substitutions. Currently, more investigation is being conducted to develop asymmetric reactions with polyunsaturated carbonyl substances via aminocatalysis in this laboratory.
We are grateful for the financial support from the National Natural Science Foundation of China (21125206, 21372160 and 21321061) and the National Basic Research Program of China (973 program, 2010CB833300).
Footnote |
† Electronic supplementary information (ESI) available: Experimental procedures, structural proofs, the CIF file of the enantiopure product 4i. CCDC 992018. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c4qo00079j |
This journal is © the Partner Organisations 2014 |