Download PDF
Synlett 2024; 35(04): 445-450
DOI: 10.1055/a-2145-8697
cluster
11th Singapore International Chemistry Conference (SICC-11)

Aminoacylation of Alkenes by Cooperative NHC and Photoredox Catalysis

Lena Lezius‡
,
Jannik Reimler‡
,
Nadine Döben
,
Michael Hamm
,
Constantin G. Daniliuc
,
Armido Studer  ∗

This work was funded by the Deutsche Forschungsgemeinschaft (DFG) (German Research Foundation) (GRK 2678 – 437785492) and the Verband der Chemischen Industrie e.V. (VCI) (Ph.D. fellowship to L.L.).
› Further Information
 
 PDF Download Permissions and Reprints All articles of this category


Abstract

Cooperative NHC and photoredox catalysis has gained significant attention as an emerging research field in recent years. Herein, we report a cyclizing aminoacylation of alkenes, which is enabled through the combination of these two catalytic modes. The key step is a radical/radical cross-coupling between a persistent ketyl radical and a transient benzylic or aliphatic C-radical, which is generated through radical cyclization of an oxidatively formed amidyl radical. Several carbamates, amides and sulfonamides containing an alkene moiety and different acyl fluorides can be used as substrates. The resulting products are obtained in moderate to good yields.


#

Key words

NHC catalysis - photoredox catalysis - radical cyclization - radical acylation - nitrogen radicals

N-Heterocyclic carbene (NHC) catalysis has become highly valuable in organic chemistry.[1] In such transformations, mostly aldehydes are used as substrates that react with the NHC catalyst to form the corresponding Breslow intermediates, which then express nucleophilic reactivity as umpoled aldehydes. This approach has been successfully used in various reactions including the important benzoin condensation[2] and the Stetter reaction.[3]

In comparison to classic ionic NHC catalysis, radical NHC catalysis is much less explored.[1d] [4] Although early studies by Fukuzumi et al. in 1997 showed that the Breslow intermediate can readily undergo a single-electron oxidation,[5] the first example of practical radical NHC catalysis was published by our group far later in 2008.[6] In this report, we presented the oxidation of aldehydes to the corresponding TEMPO esters, comprising two subsequent single-electron transfer (SET) oxidations of the Breslow intermediate. 2,2,6,6-Tetramethylpiperidine N-oxyl (TEMPO) was used as a stoichiometric single-electron oxidant in these transformations. Since then, different reactions in the area of oxidative NHC catalysis have been developed,[7] including carbon–carbon bond formations through radical/radical cross-coupling.[8]

Zoom Image
Scheme 1 Methods for two-component couplings via cooperative NHC/photoredox catalysis, carbo- and hydroamination, and the herein reported aminoacylation

More recently, a complementary reductive pathway towards radical NHC catalysis was introduced, which was achieved by the combination of NHC catalysis with photoredox catalysis. In this approach, an acyl azolium ion, generated from a carboxylic acid derivative and the NHC catalyst, is SET-reduced by a photoredox catalyst. Thereby, a persistent ketyl radical is formed that can undergo a radical/radical cross-coupling with a concomitant oxidatively generated C-radical.

The first examples of this approach were independently published by Scheidt et al. and our group in 2020. The Scheidt group reported a two-component coupling using acyl imidazoles and Hantzsch esters for the synthesis of ketones from carboxylic acids (Scheme [1a]).[9] Our contribution involved combining NHC catalysis with photoredox catalysis to realize an alkene acyltrifluoromethylation through a three-component coupling of aroyl fluorides, styrenes and the Langlois reagent.[10] Following these seminal reports, other two- and three-component coupling reactions involving NHC-bound ketyl radicals and various radical precursors were developed.[11] Moreover, the scope of cooperative NHC and photoredox catalysis could be further broadened by combination with a third catalytic mode, thereby realizing triple catalysis.[12]

In addition, photoactivation of acyl azolium ions, either directly or through energy transfer by a photocatalyst, has been achieved. The photoexcited acyl azolium ion can undergo single-electron transfer to generate the corresponding longer-lived ketyl radical, which can then engage in a radical/radical cross-coupling.[13] Recently, photoinduced single-electron reduction of an NHC catalyst by exploiting the π-accepting properties and its subsequent use as a reductant for aryl radical generation was introduced as an additional mode of action in radical NHC catalysis.[14] Furthermore, elegant work has been reported that accomplished the generation of an acyl radical by synergistic Fe, P and photoredox catalysis.[15]

Nitrogen-containing heterocycles are important structural motifs in natural products and pharmaceuticals. Nitrogen-centered radicals are highly valuable reactive intermediates, which have been successfully used for C–N bond formation reactions and the construction of N-heterocycles.[16] Accordingly, various methods have been developed for the generation of N-centered radicals. Such heteroatom-centered radicals can be generated reductively, oxidatively, through homolytic cleavage or by oxidative proton-coupled electron transfer.[16a] [b] Considering the latter approach, two elegant processes were developed by Knowles et al. in 2015. In their work, carbo- and hydroamination for the synthesis of N-containing heterocycles by applying photoredox catalysis was described (Scheme [1b]).[17] During the preparation of this manuscript an aminoacylation and an iminoacylation via N-radical generation by N–O bond cleavage were reported by Zhao et al.[18] and Ye et al.[19]

Inspired by the recent work on cooperative NHC/photoredox catalysis and the amidyl radical chemistry developed by Knowles et al.,[17] we decided to combine these two strategies to realize an aminoacylation of alkenes (Scheme [1c]).

Table 1 Optimization of the Reaction Conditionsa

Entry

Deviation from standard conditions

Yieldb

 1

467 nm LED

59%

 2

467 nm LED, NHC precursor B

15%

 3

467 nm LED, NHC precursor C

32%

 4

–

65%

 5

Ir(dFCF3ppy)2(dtbbpy)PF6

46%

 6

Ir(ppy)2(dtbbpy)PF6

38%

 7

MeCN

53%

 8

toluene

28%

 9

DCM (0.2 M)

44%

10

48 h

86% (82%)c

11

no photocatalyst

–

12

no NHC precursor

traces

13

carried out in the dark

–

a Unless otherwise noted, all reactions were performed with carbamate 1a (0.10 mmol), benzoyl fluoride (2a) (0.40 mmol), 4CzIPN (2.0 μmol), NHC precursor A (20 μmol) and Cs2CO3 (0.20 mmol) in DCM (1.0 mL) under irradiation with a 456 nm Kessil lamp without heating for 24 h.

b Yields were determined by 1H NMR using 1,3,5-trimethoxybenzene as an internal standard.

c Isolated yield is given in parentheses.

The investigations commenced using carbamate 1a (1.0 equiv.), benzoyl fluoride (2a) (4.0 equiv.), 4CzIPN (2.0 mol%), NHC precursor A (20 mol%) and Cs2CO3 (2.0 equiv.) as the base in dichloromethane (DCM) (0.1 M) under irradiation with a blue LED (467 nm) for 24 hours. Under these conditions the targeted aminoacylated product 3aa was obtained in 59% yield as a 1:1 mixture of the two diastereomers (Table [1], entry 1). Reaction optimization was started by first varying the NHC precursor. The initial NHC precursor A turned out to be the most efficient in this series (entries 2 and 3). The yield was increased to 65% by using a different blue LED that emitted at 456 nm (entry 4). Variation of the photocatalyst and the solvent did not lead to an improved result (entries 5–8). At a higher concentration, the yield dropped to 44% (entry 9). Pleasingly, upon increasing the reaction time to 48 hours the highest yield of 86% was obtained (82% isolated yield, dr = 1.4:1) (entry 10). Notably, the diastereoselectivity remained low under all the conditions tested. Finally, control reactions were performed. In the absence of a photocatalyst, without the use of an NHC precursor and when carrying out the reaction in the dark, no product formation or only a trace amount of product was observed (entries 11–13).

Zoom Image
Scheme 2 Reaction scope of the aminoacylation of alkenes. Variation of the aroyl fluoride (top) and the alkene (bottom). Reaction conditions: All reactions were performed with alkene 1 (0.10 mmol), aroyl fluoride 2 (0.40 mmol), 4CzIPN (2.0 μmol), NHC precursor A (20 μmol) and Cs2CO3 (0.20 mmol) in DCM (1.0 mL) under irradiation with a 456 nm Kessil lamp without heating for 48 h. Isolated yields are given. For more details see the Supporting Information. a The relative configuration was assigned by analogy.

With optimized conditions in hand, we next investigated the scope of the cyclizing aminoacylation.[20] First, different acyl fluorides were tested keeping carbamate 1a as the reaction partner (Scheme [2]). The attachment of a methyl group at the para or meta position of the aroyl fluoride did not have a significant influence on the yield or the diastereoselectivity of the corresponding products 3ab and 3ac. 2-Naphthoyl fluoride and 2-thienoyl fluoride were tolerated in the reaction leading to the desired products in moderate yields but with low or no diastereoselectivity (3ad, 47%, dr = 1.3:1; 3ae, 48%, dr = 1:1). On using para-methoxy- or the para-cyanobenzoyl fluoride, besides the desired aminoacylation products 3af and 3ag, side products 3af1 and 3ag1 were identified in significant amounts. These side products derive from deprotonation of the desired aminoacylation products with subsequent O-acylation of the intermediate enolates with the acyl fluoride. With the less acidic product 3af bearing the electron-donating methoxy group at the para position, the desired product was obtained as the major compound in good yield (67%, dr = 1.4:1), and only a small amount of the side product 3af1 was formed (11%, E-product only). However, for the more acidic cyano derivative, the desired product 3ag was obtained in only 31% yield (dr = 4.5:1) with O-acylated 3ag1 being formed as the major product in 63% yield (E/Z = 1.2:1). We assume that enolate formation also explains the low diastereoselectivity of these transformations, as epimerization of the formed products leads to decreased diastereoselectivity.

Next, the scope of the alkene component was investigated by using benzoyl fluoride (2a) as the coupling partner (Scheme [2]). Carbamates with electron-donating and electron-withdrawing groups at the para position of the aryl ring attached to the alkene or the aryl moiety attached to the nitrogen atom were tolerated, and the desired products 3ba–ea were obtained in moderate to good yields. When varying the substituent on the aryl ring of the cinnamyl alcohol moiety, again the formation of the O-acylated side product was observed. As expected, larger amounts of the O-acylated side product were formed for systems bearing an acidifying electron-withdrawing substituent (compare 3ba1 , 15%, only Z with 3ca1 , 31%, Z/E = 3.2:1). Formation of quaternary C-centers was also possible, as documented by the successful preparation of 3fa (44%) and 3ha (34%). Considering 3fa, a high diastereoselectivity was obtained (>19:1), as epimerization is not possible through enolate formation. For the less activated carbamate 1g derived from E-crotyl alcohol, efficient aminoacylation was achieved and 3ga was obtained in 77% yield with low diastereoselectivity (dr = 1.7:1).

Next, δ,ε-unsaturated amide 1i and δ,ε-unsaturated sulfonamides 1j and 1k were tested in the aminoacylation reaction with 2a (Scheme [2]). The cyclic amide 3ia was obtained in a good yield (62%) and excellent diastereoselectivity (dr > 19:1). A similar outcome was noted for the reactions with the sulfonamides, leading to 3ja (61%, dr > 19:1) and 3ka (41%, dr > 19:1). The relative configuration of 3ja was assigned by X-ray structure analysis[21] and the relative configurations of 3ia and 3ka were assigned by analogy. In contrast to the carbamates discussed above, high diastereoselectivities were obtained for all these substrates. Apparently, products that arise from amides and sulfonamides are less acidic than their carbamate congeners and epimerization does not occur.

The proposed mechanism is depicted in Scheme [3a]. The aroyl fluoride first forms a bisacyl carbonate intermediate through reaction with Cs2CO3.[11p] The NHC, which is formed in situ by deprotonation, can react with this intermediate to afford the acyl azolium ion I. This azolium ion (E 1/2 = –0.81 V vs SCE)[22a] can then be reduced by the excited photocatalyst 4CzIPN* (E 1/2(4CzIPN*/4CzIPN•+) = –1.04 V vs SCE)[22b] to generate the persistent ketyl radical II. After NH deprotonation of substrate 1, the anion (for carbamate with R1 = Me, R2 = H, R3 = Ph: E 1/2 = 0.83 V vs SCE)[22c] is oxidized by the oxidized form of the photocatalyst (E 1/2(4CzIPN•+/4CzIPN) = 1.52 V vs SCE)[22b] so that the amidyl radical III is formed, closing the photoredox catalysis cycle. Next, 5-exo cyclization of the amidyl radical III leads to the corresponding cyclized C-radical IV. The transient C-radical IV and the persistent ketyl radical II can then undergo a radical/radical cross-coupling to give the intermediate V. Finally, NHC fragmentation leads to formation of the aminoacylated product 3, thereby closing the NHC catalysis cycle.

Zoom Image
Scheme 3 (a) Proposed mechanism of the aminoacylation of alkenes. (b) The A1,3-strain model to explain the stereochemical outcome of the reaction with sulfonamide 3ja.

The selectivity of the 5-exo cyclization can be understood considering the A1,3-strain model[23] for benzylic radicals (Scheme [3b]). The conjugated phenyl ring of the benzylic radical affects the substituent orientation at the adjacent stereocenter. Thus, the hydrogen atom at the chirality center is oriented in the plane of the planar C-radical toward the phenyl ring in order to minimize the allylic strain. As the methylene moiety is smaller than the NSO2Me substituent, the bulky ketyl radical approaches the benzylic radical syn to the methylene entity leading to isomer 3ja.

The assumption that oxidative quenching of 4CzIPN takes place is supported by Stern–Volmer quenching studies. The acyl azolium ion shows stronger luminescence quenching than a mixture of the carbamate 1a and tetrabutylammonium carbonate (see the Supporting Information). Due to the lower solubility of Cs2CO3 in comparison to tetrabutylammonium carbonate, it can be inferred that the excited photocatalyst is oxidatively quenched by the acyl azolium ion. In addition, the reaction was inhibited when TEMPO was present in the reaction mixture, supporting a radical mechanism (see the Supporting Information).

In this work, we have developed a new strategy for the aminoacylation of alkenes, which is enabled by cooperative NHC and photoredox catalysis. The key step is a radical/radical cross-coupling of a transient C-radical with a persistent ketyl radical. The C-radical is generated through radical cyclization of an amidyl radical, while the persistent ketyl radical is formed by SET reduction of an acyl azolium ion. Carbamates readily derived from cinnamyl and crotyl alcohols can be used as substrates and aroyl fluorides serve as acylation reagents. Cyclic carbamates are obtained with low diastereoselectivity due to product epimerization. However, with δ,ε-unsaturated amides and sulfonamides as the N-radical precursors, aminoacylation occurs with excellent diastereoselectivity. The resulting N-heterocycles are not only important structural motifs in organic chemistry, but also in natural products and pharmaceuticals.


#

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

We thank Dr. Qing-Yuan Meng for providing some of the starting materials.

Supporting Information

    Supporting information for this article is available online at https://doi-org.accesdistant.sorbonne-universite.fr/10.1055/a-2145-8697.
  • Supporting Information

Corresponding Author

Armido Studer
University of Münster, Organisch-Chemisches Institut
Corrensstraße 40, 48149 Münster
Germany   

Publication History

Received: 22 June 2023

Accepted after revision: 01 August 2023

Accepted Manuscript online:
01 August 2023

Article published online:
12 September 2023

© 2023. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


   Permissions and Reprints All articles of this category
Zoom Image
Scheme 1 Methods for two-component couplings via cooperative NHC/photoredox catalysis, carbo- and hydroamination, and the herein reported aminoacylation
Zoom Image
Scheme 2 Reaction scope of the aminoacylation of alkenes. Variation of the aroyl fluoride (top) and the alkene (bottom). Reaction conditions: All reactions were performed with alkene 1 (0.10 mmol), aroyl fluoride 2 (0.40 mmol), 4CzIPN (2.0 μmol), NHC precursor A (20 μmol) and Cs2CO3 (0.20 mmol) in DCM (1.0 mL) under irradiation with a 456 nm Kessil lamp without heating for 48 h. Isolated yields are given. For more details see the Supporting Information. a The relative configuration was assigned by analogy.
Zoom Image
Scheme 3 (a) Proposed mechanism of the aminoacylation of alkenes. (b) The A1,3-strain model to explain the stereochemical outcome of the reaction with sulfonamide 3ja.