Synlett 2024; 35(11): 1279-1284
DOI: 10.1055/a-2214-5512
cluster
Japan/Netherlands Gratama Workshop

Visible-Light-Induced Three-Component Radical Coupling of Selenocarbamates, Enones, and Allylstannanes with Diphenyl (2,4,6-trimethylbenzoyl)phosphine Oxide

Keita Komine
,
Yuta Yamazaki
,
Taiga Iwanaga
,
Hikaru Sakaguchi
,
Hayato Fukuda
,

This work was supported by JSPS KAKENHI grants numbers 22K15254 (K.K.) and 23K04736 (J.I.). This work was the result of using research equipment shared in MEXT project for promoting public utilization of advanced research infrastructure (program for supporting introduction of the new sharing system) Grant Number JPMXS0422500320.
 


Abstract

A blue LED-induced three-component coupling of a carbamoyl radical, cyclic enone, and allylstannane was developed. The use of blue LEDs and diphenyl(2,4,6-trimethylbenzoyl)phosphine oxide (TPO) as a radical initiator permitted the three-component radical coupling to proceed with a high chemoselectivity. An elucidation of the mechanism revealed a pathway for the formation of a tributyltin radical from TPO and allylstannane. This tandem radical reaction is expected to be applicable in natural-product synthesis.


#

Recently, a number of radical reactions that proceed under mild conditions have been reported.[1] [2] Radical reactions are useful methods for C–C bond formation in crowded positions, permitting the construction of complex molecules, such as natural products.[3,4] Previously, we reported a one-step construction of the azabicyclo[3.3.1]nonanone skeleton by a tandem radical reaction involving a carbamoyl radical, to achieve a formal synthesis of haliclonin A (Scheme [1a]).[5] This tandem radical reaction proceeded under neutral and thermal conditions, to afford the target product in excellent yield with high stereoselectivity. On the other hand, carbamoyl radicals,[6] [7] [8] [9] [10] [11] [12] which are known for their high reactivity, are relatively stable radicals compared with acyl[12] [13] or alkoxycarbonyl[14] radicals. This is because decarbonylation and decarboxylation hardly occur from carbamoyl radicals (Scheme [1b]).[12]

Zoom Image
Scheme 1 (a) Previous work: tandem radical reaction and formal synthesis of haliclonin A; (b) acyl radicals; and (c) this work: a three-component radical coupling

By considering the stability of carbamoyl radicals, we hypothesized that a three-component radical coupling in intermolecular reactions might be feasible. However, to our knowledge, there have been no reports of multicomponent radical couplings of carbamoyl radicals. The resulting products would share common structural features found in various natural products and are expected to have potential applications in natural-products synthesis.[15] Here, we describe a new method for the visible-light-induced three-component coupling an enone, an allylstannane, and a carbamoyl radical.

To test our hypothesis, we first compared the reaction of selenocarbonates with that of a selenocarbamate. The three-component radical coupling was performed under thermal conditions using selenocarbonate or selenocarbamate 1, cyclic enone 2, and allylstannane 3 according to the protocol reported previously (Table [1]).[5] When a mixture of selenocarbonate 1a (1 equiv), 2a (4 equiv), and 3a (4 equiv) was heated at 130 °C in the presence of 1,1′-azobis(cyclohexane-1-carbonitrile) (V-40) in chlorobenzene (PhCl), compound 4a was obtained as a 4:1 anti/syn diastereomeric mixture in 43% yield (Table [1], entry 1). The major isomer was anti-4a, which was characterized by NOESY analysis. Although various conditions were investigated for the thermal reaction, none of the changes in the substrate amounts, reaction temperature, or substrate concentration improved the reaction yield.[16] Next, other selenocarbonates were examined. The reaction with O,Se-diphenyl selenocarbonate (1b) yielded a trace amount of 4b (entry 2), whereas the reaction with O-isobutyl Se-phenyl selenocarbonate (1c), gave compound 4c as a 9:1 diastereomeric mixture in 11% yield (entry 3). In contrast, the reaction of the selenocarbamate 1d, which generates a more stable carbamoyl radical, proceeded smoothly to afford 4d as a 4:1 anti/syn mixture in 50% yield (entry 4). These results suggested that increasing the stability of the acyl radical lengthens the free-radical lifetime, facilitating the reaction of 2a.[12] Note that for 1d, other variations in the reaction condition, such as the use of 2,2′-azobis(4-methoxy-2,4-dimethylvaleronitrile) (V-70) at 40 °C or of Et3B and O2 at 25 °C proved fruitless.

Table 1 Three-Component Radical Coupling under Thermal Conditions

Entry

X

Time (h)

Product

Yield (%)

anti/syn a

1

1

4a

43

4:1

2

3.5

4b

trace

3

1.5

4c

11

9:1

4

1.5

4d

50

4:1

a Determined by 1H NMR spectroscopy.

Table 2 Three-Component Radical Coupling under Photochemical Conditions

Entry

Initiator

Solvent

Temp (°C)

Yield (%)

anti/syn a

4d

5a

1

AIBN

PhCl (0.1 M)

 25

51

10b

8:1

2

TPO

PhCl (0.1 M)

 25

56

18b

5:1

3

TPO

PhCl (0.1 M)

  3

69

22b

4:1

4

TPOc

PhCl (0.5 M)

–40

25

5b

5:1

5d

TPO

THF (0.1 M)

  3

6

TPO

PhCl (0.6 M)

  3

70

17b

4:1

7e

TPOc

PhCl (0.6 M)

  3

46

 8b

4:1

8f

TPOg

PhCl (0.2 M)

  3

74

20b

4:1

a Determined by 1H NMR analysis.

b Compounds 6a and 7 were also detected.

c 2 × 0.2 equiv.

d Eu(OTf)3 (50 mol%) was added.

e A violet LED was used instead of a blue LED.

f 1d/2a/3a = 1:2:2.

g 3 × 0.1 equiv.

Next, the radical reaction under photochemical conditions was explored. Because visible-light-mediated reactions of carbamoyl radicals have been developed in recent years,[9] [10] [11] we focused on the three-component coupling with blue LED illumination, which should generate free radicals, even at low temperatures (Table [2]). When 2,2′-azobis(isobutyronitrile) (AIBN), which has an absorption maximum at λ = 347 nm,[17] was used as an initiator at 25 °C, compound 4d was obtained in 51% yield as an 8:1 mixture of the anti and syn diastereomers (Table [2], entry 1). Interestingly, the reaction at 25 °C with diphenyl(2,4,6-trimethylbenzoyl)phosphine oxide (TPO), which exhibits a strong absorption at λ = 380 nm[18] [19] resulted in a slight improvement (entry 2). To our delight, the same reaction at 3 °C afforded 4d in a 69% yield (entry 3); however, lowering of the temperature did not improve the yield. (entry 4) We also found that the addition of the Lewis acid Eu(OTf)3 was incompatible with the photochemical radical reaction (entry 5).[20] On the other hand, when the reaction was irradiated with a violet LED (λ = 390 nm) instead of a blue LED, the yield decreased (entry 7). Eventually, we found that the reaction of one equivalent of 1d, two equivalents of 2a, and two equivalents of 3a in the presence of a catalytic amount of TPO in PhCl solution (0.2 M) at 3 °C under blue LED irradiation provided 4d as a 4:1 anti/syn diastereomeric mixture in a 74% yield (entry 8).[16] [21] Note that compounds 6a and 7 were also observed in the crude mixture from the three-component coupling (entries 1–4 and 6–8).

We next turned our attention to the blue LED-induced three-component coupling of various substrates 1, 2, and 3 (Scheme [2]). First, selenocarbamates 1df underwent a coupling reaction with 2a and 3a,[22] The reaction of the piperidine derivative 1e gave product 4e as a 6:1 diastereomeric mixture in a moderate 48% yield, whereas amine 1f gave product 4f was in 67% yield as a 5:1 mixture. On the other hand, when selenocarbonate 1a was subjected to the blue-LED-induced three-component coupling, the reaction did not proceed well, and the yield of 4a was low (entry 4).

Zoom Image
Scheme 2 Substrate scope. a Compounds 6b and 7 were also obtained.

To further examine the substrate scope, the reactions of alkenes 2be were then investigated. The reaction of the vinyl sulfone 2b proceeded smoothly to form 4g in 78% yield (entry 5); however, the reaction with cyclohex-2-en-1-one (2c) gave compound 4h in a poor 27% yield, with no selectivity. In the case of cyclohept-2-en-1-one (2d), the reaction afforded compound 4i with excellent diastereoselectivity, but the yield was only 29%. A similar yield and diastereoselectivity to that obtained with 2d were observed in the reaction of the butenolide 2e. We also found that substrates having electron-donating groups at the α- or β-position, such as 2f and 2g, were unsuitable substrates for the three-component coupling. We also examined the radical coupling of stannane 3b, which afforded 4d in 40% yield as a 4:1 mixture. Notably, the use of a triphenyltin radical instead of a tributyltin radical had almost no effect on the diastereoselectivity. However, in the case of 3c, the desired product 4m was not obtained, and 1d was recovered in 88% yield.

The plausible mechanism for the three-component radical coupling is depicted in Scheme [3]. The maximum absorption wavelength of TPO is 380 nm, and TPO is excited by the blue LED irradiation to generate radicals 8 and 9.[19] Nucleophilic radical 8 reacts sequentially with 2a and 3a to afford compound 6 and a tributyltin radical (10). Electrophilic radical 9 is also intercepted by 3a to give compound 7 and a tributyltin radical (10). The two reactions shown in Scheme [3] rationalize the detection of compounds 6 and 7. The resulting tributyltin radical (10) reacts with 1d to give the carbamoyl radical 11 and the selenostannane 12.[23] Nucleophilic radical 11 undergoes coupling with enone 2a to afford the electrophilic α-radical 13, which subsequently reacts with stannane 3a to furnish product 4d via intermediate 14. The coupling between 13 and 3a proceeds on the opposite side of the amide group, giving anti-4d preferentially. In addition, the unfavorable mismatched coupling between 11 and 3a yields 5a as a byproduct.

Zoom Image
Scheme 3 Plausible mechanism for the three-component radical coupling

In conclusion, we have demonstrated the three-component radical coupling of selenocarbamates, cyclic enones, and allystannanes to afford ketoamides. The use of blue LED irradiation and TPO as a radical initiator enabled the radical reactions to take place at 3 °C and under mild conditions. Consequently, this reaction proceeds with high chemoselectivity and is therefore suitable for natural-product synthesis. Natural-product syntheses using this three-component radical coupling are currently underway in our group.

Zoom Image
Figure 1 Telluride 1d-Te

#

Conflict of Interest

The authors declare no conflict of interest.

Supporting Information


Corresponding Author

Jun Ishihara
Graduate School of Biomedical Sciences, Nagasaki University
1-14 Bunkyo-machi, Nagasaki, Nagasaki 852-8521
Japan   

Publication History

Received: 31 October 2023

Accepted after revision: 20 November 2023

Accepted Manuscript online:
20 November 2023

Article published online:
21 December 2023

© 2023. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


Zoom Image
Scheme 1 (a) Previous work: tandem radical reaction and formal synthesis of haliclonin A; (b) acyl radicals; and (c) this work: a three-component radical coupling
Zoom Image
Scheme 2 Substrate scope. a Compounds 6b and 7 were also obtained.
Zoom Image
Scheme 3 Plausible mechanism for the three-component radical coupling
Zoom Image
Figure 1 Telluride 1d-Te