An updated analysis of U.S. FDA-approved small-molecule drugs in the past 11 years (2013–2023) reveals that >82% of drugs contain at least a nitrogen heterocycle or an amine moiety [
1]. In this context, the continuous development of new methods for the precise synthesis of amine derivatives has emerged as a crucial research field in organic synthesis and medicinal chemistry. Among them, selective C(sp
3)-H functionalization of readily available aliphatic amines could streamline the synthesis of nitrogen-containing heterocycles and complex amine architectures. For example, different strategies for the
α-C(sp
3)-H functionalization of aliphatic amines have been well developed
via oxidative process [
2–
7] or hydrogen atom transfer (HAT) [
8–
13]. In addition,
δ-C(sp
3)-H functionalization based on amidyl radical-mediated 1,5-HAT has been proven to be a robust strategy to access remote functionalized amines and the related nitrogen-containing heterocycles [
14–
17]. In contrast,
γ-amino C(sp
3)-H functionalization of aliphatic amines remains challenging. Current successful
γ-amino C–H functionalization methods mainly rely on directing group-assisted transition metal catalysis [
18–
27]. Alternatively, radical C(sp
3)-H functionalization of amines
via γ-selective HAT process has enabled a new approach to the synthesis of amine derivatives [
28–
33]. However,
γ-C(sp
3)-H acylation of aliphatic amines remains unaccessible, while the products will simplify the synthesis of
γ-aminoketones,
δ-amino alcohols, and pyrrolidines of interest in medicinal chemistry (
Scheme 1a) [
34–
37].
In the past decades,
N-heterocyclic carbene (NHC) catalysis has emerged as a powerful strategy for the synthesis of structurally diverse molecules [
38–
49]. In 2012, Rovis and co-workers developed cooperative NHC and photocatalysis for asymmetric acylation of
α-C(sp
3)-H bond in
N-aryl tetrahydroisoquinoline
via nucleophilic addition of Breslow intermediate to the oxidatively generated iminium (
Scheme 1b) [
50]. Driven by the development of radical NHC catalysis [
51–
58], the direct acylation of C(sp
3)-H bonds has aroused great interest [
59]. Limited successful examples mainly focus on
α- or
δ-C(sp
3)-H acylation of amines. For example, the NHC-catalyzed
α-C(sp
3)-H acylation of amines has been established through a cascade generation of
α-amino radical
via HAT process and the following coupling with ketyl radical [
60–
63]. On the other hand, the remote
δ-C(sp
3)-H acylation has been achieved
via amidyl radical-mediated 1,5-HAT [
64]. However,
γ-amino C(sp
3)-H acylation is highly desirable but still unexplored to date.
Herein, we report the
γ-selective C(sp
3)-H acylation of aliphatic amines with aldehydes as acyl source under cooperative NHC and photoinduced palladium catalysis (
Scheme 1c). 2-Bromophenyl sulfone is used as
γ-selective HAT auxiliary and amino-protecting group [
28,
65]. The photoinduced Pd-promoted generation of aryl radical and the subsequent 1,7-HAT is the key to the success of this methodology. This approach enables direct conversion of a wide array of readily available aliphatic amines to
γ-aminoketones in moderate to good yields with excellent and predictable regioselectivity. The easy conversion of the products to
δ-aminoalcohols and pyrrolidines of interest to medicinal chemists further demonstrates the potential application of the developed method in drug discovery campaigns.
Initially, the model reaction of 2-bromophenyl sulfonamide
1a and commercially available picolinaldehyde
2a was investigated under photoredox cooperative NHC/Pd catalysis (
Table 1). We were pleased to find that the reaction gave the desired product
3a in 62% yield when carried out in the presence of 20 mol% thioazolium preNHC
N1, 10 mol% of Pd(OAc)
2 with
tBu-Xantphos
L1 (15 mol%) as ligand and Cs
2CO
3 (2.0 equiv.) as base in PhCl under blue LED irradiation at room temperature (entry 1). The screening of preNHC catalysts revealed that
N-benzyl and
N-aryl substituted thioazolium preNHC
N2-
N4 resulted in some loss of yields (entries 2 and 3), while triazolium preNHC
N5 did not work for this reaction (entry 4). Other palladium catalysts such as PdCl
2 and Pd(PPh
3)
4 were less effective (entries 5 and 6). The use of ligand Xantphos resulted in decreased yield (entry 7), and PCy
3 was ineffective (entry 8). Solvent screening revealed that the reaction in PhCF
3, THF or DCM did not give better yields (entries 9 and 10). Evaluation of bases showed that other inorganic bases, such as K
2CO
3 and CsF, and organic base led to decreased yields (entries 11 and 12). The improvement of the yield was achieved by reducing reaction concentration to 0.05 mol/L (entry 13). The reaction with other photocatalysts, such as 4CzIPN,
fac-Ir(ppy)
3, Ru(phen)
3(PF
6)
2, gave the desired product in <7% yield (entries 14–16).
With the optimized reaction conditions in hand, the scope of aldehydes was evaluated (
Scheme 2). It was found that picolinaldehydes bearing electron-donating or electron-withdrawing groups such as OMe, Me, F, and Cl at various positions of pyridine ring all worked well to afford desired ketones
3b−
3k in moderate to good yields. Picolinaldehyde bearing 9-anthracenyl at 5-position was compatible for this reaction giving the desired product
3l in 55% yield. The reactions of other heteroaromatic aldehydes such as quinoline-2-carbaldehyde, isoquinoline-3-carbaldehyde, quinoxaline-2-carbaldehyde and thiazole-2-carbaldehyde all went smoothly to furnish the desired ketones
3m−
3p. However, the reaction of benzaldehyde gave no desired product which promoted us reoptimized the conditions of the reaction (see Supporting information for details). It was found that the use of preNHC
N4 afforded the desired ketone
3q in 65% yield. The scope of benzaldehydes was then examined. Benzaldehydes bearing electron-donating (Me) or electron-withdrawing groups (F, Cl, and Br) on the
para-position all worked well, leading to ketones
3r−
3u in moderate yields. It is worth noting that functional groups, such as internal alkyne, acetyl, carbomethoxy, cyano, pinacol boronic ester (Bpin), trimethylsilyl (TMS) and methylsulfonyl, were all tolerable to afford the corresponding ketones
3v−
3ab.
meta-Substituted benzaldehyde bearing Me or CF
3 groups gave the products
3ac−
3ad in good yields. Besides, the reaction of furfural and 2-thenaldehyde also proceeded smoothly to give ketones
3ae−
3af in good yields. The structure of the product
3ae was confirmed by single crystal X-ray diffraction analysis. However, cinnamaldehyde and aliphatic aldehydes, such as pivaldehyde and benzenepropanal, did not work for this reaction under current reaction conditions.
Then, various amines were examined. 3-Phenylpropylamines bearing OMe, iPr, Me, Cl, Br, and CF3 on the para-position afforded corresponding ketones 3ag-3al in moderate to good yields. The reaction of 3-phenylpropylamines bearing OMe and CF3 at meta- or ortho-position all worked well, leading to products 3am-3ap in moderate yields. 3-Phenylpropylamines bearing Me and ester at amino α-position were also tolerable (3aq-3ar). Besides acylation of benzyl C(sp3)-H bond, acylation of unactivated aliphatic C(sp3)-H bond with various (hetero)aromatic aldehydes could also be achieved on this catalytic platform (3as−3aw). The protecting group Boc could be replaced by methyl without affecting the reaction efficiency (3ax). However, unsubstituted substrate 1 did not work for this reaction under current reaction conditions.
The excellent
γ-selectivity observed in these reactions led us to further investigate the limits of the regioselectivity in the key HAT step (
Scheme 3).
β-C(sp
3)-H acylation occurs
via 1,6-HAT to furnish
β-aminoketones
3ay in moderate yield, when no
γ-C(sp
3)-H is present in the substrate. The reaction of
n-amylamine exhibits exclusive
γ-regioselectivity, and
β- and
δ-C(sp
3)-H are untouched (
3az). The
γ-C(sp
3)-H acylation of aliphatic amine bearing benzyl
δ-C(sp
3)-H further confirms 1,7-HAT is highly favored over 1,8-HAT (
3ba). The excellent regioselectivity may be attribute to the rigid sulfonamide tether, which makes 1,7-HAT favored over the corresponding 1,5-, 1,6- or 1,8-HAT [
28,
66,
67].
To further demonstrate the synthetic potential of this method, gram-scale synthesis and synthetic transformations were carried out (
Scheme 4). The efficiency of this reaction was maintained in gram-scale synthesis, affording 1.24 g of ketone
3a in 64% yield. The removal of the Boc protective group of ketone
3a gave sulfonamide
4, which was reduced to afford
δ-aminoalcohol
5. Pyrrolidine
6 was obtained
via the following intramolecular Mitsunobu reaction. Similar deprotection of ketone
3q and the following reductive amination afforded pyrrolidine
7 in good yield with 6:1
dr. Similarly, ketone
3ag could be converted into pyrrolidine
8, which is the key intermediate for the synthesis of BIRZ227, a potentially useful therapeutic agent for treating inflammatory disorders [
68].
To go insight into the mechanism, a series of control experiments were carried out (
Scheme 5). It is found that preNHC, palladium and ligand are all crucial for the reaction (
Scheme 5a). The light on/off experiments reveals that light is essential (see Supporting information for details). When the radical scavenger TEMPO was added, the reaction was completely inhibited and the adduct
9 of aryl radical with TEMPO could be isolated in 6% yield (
Scheme 5b). When introducing a deuterium at benzyl position, only the benzene ring of sulfonyl group was partially deuterated, demonstrating the 1,7-HAT between aryl radical and
γ-C(sp
3)-H (
Scheme 5c). The
in situ electron paramagnetic resonance (EPR) analysis of the model reaction mixture suggests the generation of ketyl radical
10 (
Scheme 5d). In addition, Stern-Volmer quenching studies demonstrated that
1a quenched the excited palladium catalyst (see Supporting information for details).
Based on the mechanistic experiments and previous reports [
69–
71], a plausible catalytic cycle is shown in
Fig. 1. An inner-sphere single electron transfer from photo-excited Pd
(0)L species to aryl bromide
1 produces aryl radical
Ⅰ and Pd
(Ⅰ)L complex. Site-selective
γ-C(sp
3)-H activation through aryl radical-mediated 1,7-HAT affords transient alkyl radical
Ⅱ. Meanwhile, single electron oxidation of the Breslow enolate intermediate
Ⅲ by Pd
(Ⅰ)L complex gives persistent ketyl radical
Ⅳ and regenerates Pd
(0)L complex. The radical/radical coupling of ketyl radical
Ⅳ and alkyl radical
Ⅱ furnishes the adduct
Ⅴ, which is fragmented to afford
γ-amino C(sp
3)-H acylation products and release the NHC catalyst.
In conclusion, we have developed a practical acylation of γ-amino C(sp3)-H bonds with aldehydes as acyl source through cooperative photoredox NHC/Pd catalysis. This approach enables direct conversion of a variety of readily available aliphatic amines to γ-aminoketones in moderate to good yields with excellent and predictable regioselectivity. Mechanistic studies of the γ-amino C(sp3)-H acylation are consistent with a radical/radical coupling process that entails (a) photoinduced Pd-promoted formation of aryl radical, (b) generation of transient γ-amino alkyl radical through aryl radical-mediated 1,7-HAT, (c) single-electron oxidation of Breslow enolate intermediate to persistent ketyl radical, and (d) radical/radical coupling of γ-amino alkyl radical with ketyl radical. Additionally, we have shown that the γ-aminoketones are easily manipulated to enable access to pyrrolines and δ-amino alcohols of interest in medicinal chemistry, thus highlighting the synthetic potential of this methodology.
PDF
Full
87
48
关闭全屏
No. 86 Xueyuan South Road, Haidian District, Beijing
010-62662699(Operation and Technical Support)
qkjq@cast.org.cn
