Xiaojun Dai,
Renjie Mao,
Baochuan Guan,
Xiaoliang Xu* and
Xiaonian Li*
Institute of Industrial Catalysis, College of Chemical Engineering, Zhejiang University of Technology, Hangzhou 310014, People's Republic of China. E-mail: xuxiaoliang@zjut.edu.cn; xnli@zjut.edu.cn
First published on 18th June 2015
The regioselective addition of α-aminoalkyl radicals to 2,3-allenoates by visible-light-mediated electron transfer using 1 mol% of Ru(bpy)3(BF4)2 as a photocatalyst was successfully established. This photoredox protocol is a simple and effective method for the synthesis of unsaturated γ-aminobutyric ester derivatives.
Recently visible-light photoredox catalysis has been paid much attention due to its particular characteristics.4 Compared with the traditional ionic and radical reactions, visible-light photoredox catalysis, which follows the “ion radical” reaction path, provides a new view for studying the reactions of allenes. According to visible light photoredox reaction path, it may proceed either by oxidative or by reductive quenching, and the in situ formed radicals have to lose or obtain an electron to complete the catalytic cycle.5 Compared with vinyl radical B and C, the allylic radical A is more easily oxidized or reduced. However, from the stability of the radical intermediates, A should be more stable than B and C. Therefore, we hypothesize that radical reaction of allenes may occur in the middle sp hybridized carbon atom under visible light photoredox catalysis (Scheme 1).
Generally, tertiary amines are popular electron donors which can readily undergo single-electron oxidation, and thus have been widely used in visible light catalysis.6–8 There are two paths for the photoredox oxidation of tertiary amine, which give the reactive intermediates electrophilic iminium ion or nucleophilic α-aminoalkyl radical according to the reaction system. The routes of addition of various nucleophiles to iminium ion, for example Strecker reaction, Mannich reaction, aza-Henry reactions, etc. have been successfully realized and extensively studied.7 Comparatively the research of α-aminoalkyl radical is less investigated due to its easily oxidation to form the iminium ions,9 and currently electron-deficient olefins, isocyanates, azodicarboxylates, alkynes (intramolecular reaction) function as appropriate coupling partners.8 Recently we reported the photoredox coupling of tertiary amines with acrylate derivatives including Baylis–Hillman adducts under visible light irradiation (Scheme 2a).10 Based on the similarity of allenoates and electron-deficient olefins, the photoredox coupling reaction with tertiary amine, mediated by visible light photoredox, was examined (Scheme 2b).
At first, the reaction of ethyl buta-2,3-dienoate (1a) with 1.2 equiv. of N,N-dimethylaniline (2a) was carried out (Table 1). When a solution of 1a and 2a in the presence of 1 mol% of Ru(bpy)3(BF4)2 and 1 equiv. of NaOAc in N-methylpyrrolidone (NMP) was illuminated with a 45 W compact fluorescent lamp at room temperature for 24 h, ethyl 3-((methyl(phenyl)amino)methyl)but-3-enoate (3a) and (E)-ethyl 3-methyl-4-(methyl(phenyl)amino)but-2-enoate (4a) were obtained in 52% yield (entry 1, 3a/4a = 6.0/1). Next, the control experiment showed that no formation of 3a or 4a was observed in the absence of a photocatalyst or light (entries 2–3), and switch on-off experiment also revealed that the light was essential. Notably, without base the reaction afforded the products with diminished yield (entry 4). Examination of a range of photocatalysts showed Ru(bpy)3(BF4)2 was the most effective catalyst (entries 5–9). Solvents affected the reaction greatly, and polar aprotic solvents facilitated the reaction (entries 10–12). The reaction was difficult to proceed in MeOH (entry 13). A survey of base demonstrated that K2HPO4·3H2O was the best choice (entries 14–19). Curiously, the use of K2HPO4 gave the worse result than the one containing crystal water. However, the use of the mixture of NMP and H2O as the solvent did not improve the yield (entry 20). Finally, 1 mol% of Ru(bpy)3(BF4)2 as photocatalyst, K2HPO4·3H2O as a base and NMP as a solvent were shown to be optimal to provide the isomer mixtures of 3a and 4a in 67% yield. The product of N,N-dimethylaniline coupling with the sp2 hybridized carbon atom of ethyl buta-2,3-dienoate was not found in all of these reactions, indicating the high regioselectivity of the reaction.
Entry | Cat. | Solvent | Base | Yieldb (%) | 3a![]() ![]() |
---|---|---|---|---|---|
a Reaction conditions: 1a (0.5 mmol), 2a (0.6 mmol), photocatalyst (0.005 mmol), base (0.5 mmol), solvent (2 mL), 24 h, 45 W compact fluorescent lamp irradiation under N2 and room temperature until otherwise noted.b Isolated yield of 3a and 4a based on 1a.c the ratio of 3a![]() ![]() |
|||||
1 | Ru(bpy)3(BF4)2 | NMP | NaOAc | 52 | 6.0![]() ![]() |
2d | Ru(bpy)3(BF4)2 | NMP | NaOAc | 0 | |
3 | — | NMP | NaOAc | 0 | |
4 | Ru(bpy)3(BF4)2 | NMP | — | 35 | 6.7![]() ![]() |
5 | Ru(bpy)3Cl2 | NMP | NaOAc | 42 | 5.6![]() ![]() |
6 | Ru(phen)3(BF4)2 | NMP | NaOAc | 12 | 9.0![]() ![]() |
7 | Ir(ppy)3 | NMP | NaOAc | Trace | |
8 | [Ir(ppy)2bpy]BF4 | NMP | NaOAc | 46 | 5.6![]() ![]() |
9 | Eosin Y | NMP | NaOAc | Trace | |
10 | Ru(bpy)3(BF4)2 | DMSO | NaOAc | 38 | 4.2![]() ![]() |
11 | Ru(bpy)3(BF4)2 | MeCN | NaOAc | 27 | 2.8![]() ![]() |
12 | Ru(bpy)3(BF4)2 | DCM | NaOAc | 31 | 3.0![]() ![]() |
13 | Ru(bpy)3(BF4)2 | MeOH | K2HPO4·3H2O | Trace | |
14 | Ru(bpy)3(BF4)2 | NMP | Cs2CO3 | Trace | |
15 | Ru(bpy)3(BF4)2 | NMP | DABCO | Trace | |
16 | Ru(bpy)3(BF4)2 | NMP | 2,6-Lutidine | 28 | 1.3![]() ![]() |
17 | Ru(bpy)3(BF4)2 | NMP | NaF | 27 | 3.5![]() ![]() |
18 | Ru(bpy)3(BF4)2 | NMP | K2HPO4·3H2O | 67 | 5.9![]() ![]() |
19 | Ru(bpy)3(BF4)2 | NMP | K2HPO4 | 60 | 6.1![]() ![]() |
20e | Ru(bpy)3(BF4)2 | NMP/H2O | K2HPO4·3H2O | 54 | 12.8![]() ![]() |
Other reactions of a variety of 2,3-allenoates 1 and tertiary amines 2 were investigated under the optimized reaction conditions (Table 2). When the fourth position of 2,3-allenoates with substituent groups were reacted with N,N-dimethylaniline, only isomer 3 was obtained and the regional isomer 4 was not found. The reactions of 1b–1d bearing an aryl group at the 4-position proceeded smoothly to give the corresponding products (3b–3d) in moderate yields (entries 2–4). Ethyl penta-2,3-dienoate was also applicable to this reaction system, giving the product 3e in 47% isolated yield (entry 5). Due to the steric effect, 4,4-disubstituted 2,3-dienoates gave the desired products in low yield (entries 6–7). The yield of methyl ester was slightly lower than the yield of ethyl ester (entry 8). On the other hand, a variety of tertiary amines 2 was examined as well. When methyldiphenylamine 2b and N,N-diethylaniline 2c reacted with ethyl buta-2,3-dienoate 1a, it afforded the desired products 3i/4i and 3j/4j in low isolated yield (entries 9–10). N,N-Dimethylaniline with an electron-donating group on the benzene ring, afforded 3k/4k and 3l in reasonable yields (entries 11–12). The substrates with weak electron-withdrawing groups, such as F, Cl and Br, successfully afforded the desired products in moderate yields (entries 13–15). Cyclic amine N-phenylpyrrolidine (2i) was transformed into 3p in 12% yield (entry 16). However, aromatic tertiary amines with a strong electron-withdrawing group such as 4-(dimethylamino)benzonitrile and 4-dimethylaminopyridine deter the reaction. Unfortunately, aliphatic tertiary amines could not give the desired products.
Entry | R1, R2, R3 | R4, R5, R6 | Yieldb (%) |
---|---|---|---|
a Reaction conditions: 1 (0.5 mmol), 2 (0.6 mmol), photocatalyst (0.005 mmol), K2HPO4·3H2O (0.5 mmol), NMP (2 mL), 24 h, 45 W compact fluorescent lamp irradiation under N2 and room temperature.b Isolated yield based on 1.c The configuration of E/Z was determined by NOESY spectra, please see ESI.d Only E configuration product was obtained. | |||
1 | H, H, Et (1a) | Ph, Me, H (2a) | 67 (3a/4a = 85/15) |
2 | Ph, H, Et (1b) | 2a | 52 (3b), E/Z = 1/1c |
3 | 4-ClC6H4, H, Et (1c) | 2a | 33 (3c), E/Z = 1/1 |
4 | 4-MeOC6H4, H, Et (1d) | 2a | 40 (3d), E/Z = 1/1 |
5 | Me, H, Et (1e) | 2a | 47 (3e), E/Z = 77/23 |
6 | Ph, Ph, Et (1f) | 2a | 29 (3f) |
7 | Me, Me, Et (1g) | 2a | 32 (3g) |
8 | Me, Me, Me (1h) | 2a | 20 (3h) |
9 | 1a | Ph, Ph, H (2b) | 23 (3i/4i = 91/9) |
10 | 1a | Ph, Et, Me (2c) | 14 (3j/4j = 88/12) |
11 | 1a | 4-MeOC6H4, Me, H (2d) | 38 (3k/4k = 95/5) |
12 | 1b | 4-MeC6H4, Me, H (2e) | 41 (3l), E/Z = 48/52 |
13 | 1b | 4-FC6H4, Me, H (2f) | 38 (3m), E/Z = 51/49 |
14 | 1b | 4-ClC6H4, Me, H (2g) | 34 (3n), E/Z = 40/60 |
15 | 1b | 4-BrC6H4, Me, H (2h) | 38 (3o), E/Z = 46/54 |
16 | 1b | N-Phenylpyrrolidine (2i) | 12 (3p)d |
Based on experimental results and related literatures,9 a plausible mechanism is proposed in Scheme 3. First tertiary amine 2 is oxidized by visible light excited Ru*2+ to give amino radical cation 5 and Ru+. In the presence of base, 5 loses a proton to afford α-aminoalkyl radical 6, which undergoes nucleophilic radical addition to 2,3-allenoate 1 to form a new C–C bond and afford the alkyl radical 7. When R4 is hydrogen atom, a part of 7 can be converted into 8. Then, a single electron transfer from the Ru+ to the alkyl radical 7 and 8 completes the photoredox cycle and generates the corresponding alkyl anion 9 and 10. Subsequent protonation of 9 leads to the product 3, and the same with 10 to 4.
Footnote |
† Electronic supplementary information (ESI) available: Experimental details and spectral data. See DOI: 10.1039/c5ra10491b |
This journal is © The Royal Society of Chemistry 2015 |