Metal-catalyzed formation of two carbon–carbon (C-C) bonds across vicinal C(sp
2) sites in olefins is a powerful method to construct complex molecules rapidly from readily available starting materials [
1-
8]. Despite the great advances, the development of such a method is generally hindered by
β-H elimination from
β-H−C(sp
3)-[M] intermediates [
9-
15] and regioselectivity placing two discrete carbon sources across the vicinal C(sp
2) sites [
1,
16-
20]. To address these issues, a coordinating group is normally required either by preinstallation or a native functional group to induce a proximity bias at the vicinal sites for regiocontrol and subsequently generate transient metallacycles for addressing
β-H elimination [
21-
25]. An alternative approach that proceeds by reductive coupling
via radical intermediates [
26] is also emerging as a powerful complementary protocol [
27]. In these strategies, pre-activated arenes are often used as aryl sources, including aryl metal reagents, [
13,
28-
32] aryldiazonium salts [
33,
34] and aryl halides (
Scheme 1a) [
35-
37]. Despite significant progress, the direct utilization of Csp
2−H bond as carbon sources instead of pre-activated arenes remains elusive.
With the development of C-H activation catalyzed by transition metals, several types of arenes with special structures have been removed from dependence on leaving groups, and can directly undergo the alkylarylation of alkenes, with atomic economy and step economy, and with good site selectivity (
Scheme 1b). Besides the electron rich
N-heteroarenes (
e.g., indole) [
38-
43], 8-aminoquinoline [
44] and
N,N-disubstituted aniline derivatives [
45] successfully afforded 1,2-alkylarylation products from olefins
via combination of radical pathways with electrophilic substitution at arenes. Additionally,
σ-activation through inner-sphere electron transfer, and cyclometallation enables the difunctionalization of alkene using 2-phenylpyridine/pyrimidine [
46-
52]. Despite these progresses, Csp
2−H coupling sources are limited and alkylarylation of alkenes using this protocol remains underdeveloped.
Aniline is one of the important amines and has a very wide range of uses in industry and manufacturing as well as in the pharmaceutical field [
53-
57]. Modification of anilines by site-selective C-H activation has attracted great attention to tune their function and activity [
58-
61]. Due to the good nucleophilicity of amino group and low steric hindrance of aniline, the alkylarylation of alkenes involving unprotected aniline normally led to alkylamination products [
62-
66], and alkylarylation of alkenes
via para-C-H activation of unprotected aniline remains unsolved (
Scheme 1c). Recently, Ru-catalyzed
para-C-H alkylation of unprotected aniline
via σ-activation was disclosed [
67-
74]. We herein successfully achieved the radical alkylarylation of alkenes
via para-C-H activation of unprotected aniline. Cone angle and minimum percent buried volume (%V
bur) were used to evaluate the ligand effect. The protocol allows absolute C
vs. N site selectivity and good functional group tolerance, which enables the late-stage functionalization of nature products. The mechanism studies reveal a dissociation of
p-cymene during the catalytic processes, which is distinct from most previous
σ-activation pathways.
We commenced our study by investigating the alkylarylation of alkene
2a with aniline
1a and
α–bromo alkane
3a. After rigorous screening of the reaction conditions (For detail screening, see Supporting information), the optimal catalytic system with [RuCl
2(
p-cymene)]
2 (5 mol%), PPh
3 (40 mol%), and CsHCO
3 (2.5 equiv.) in DMF successfully afforded
4a in 70% isolated yield (
Table 1, entry 1). The control experiments revealed that the reaction did not proceed in the absence of PPh
3 (entry 2), which is consistent with previous phosphine ligand involving catalytic systems [
75]. To elucidate the influence of different phosphine ligands, we correlated the experimental selectivities and yields with key descriptors of phosphine ligands, the cone angle (min) and the minimum percent buried volume (min %V
bur). In
Scheme 2, we generated plots illustrating the yields of
4a (blue) and
5 (green) as a function of these parameters. The analysis revealed that phosphines with cone angles within a relatively narrow range (145°-156°) afforded slightly better selectivity for the three-component product
4a over
5. Increasing the cone angles led to the
para-alkylated
5 as predominant product. Additionally, when the min %V
bur of the phosphine was approximately 28.2%, the reaction exhibited better combined yield of
4a and
5. The selectivity of the reaction with different phosphine ligands at the same min %V
bur differ greatly, which might be attributed to their different cone angles. Thus, we proposed in this catalytic system, cone angle might regulate the regioselectivity while min %V
bur would control the combined yields. The highest overall yield of and selectivity for
4a is observed with PPh
3, ligand that have cone angle and min %V
bur in the optimal range (145° and 28.2%, respectively) along with three component products. The reaction failed to proceed in the absence of [RuCl
2(
p-cymene)]
2 underscoring the critical role of Ru catalyst in the reaction (entry 3). Inert gas protection was also required, suggesting that a radical pathway was plausible (entry 4). Substituting [RuCl
2(
p-cymene)]
2 with RuCl
3·3H
2O slightly decreased the reaction yield with better selectivity (entry 5). It suggests the reaction may proceed
via Ru catalyst without
p-cymene. In contrast, Ru(0) catalyst was not effective for the reaction (entry 6). Further optimization revealed that bases play an important role in this transformation (entries 7 and 8). KHCO
3 displayed catalytic activity, but Cs
2CO
3 obtained product with trace yield.
Under the optimized conditions, we systematically evaluated the functional group compatibility of this transformation using diverse anilines (
Scheme 3). Both electron-withdrawing and electron-donating substituents including halogens (
4a-
4c, 4g), trifluoromethyl (
4d), methyl (
4e, 4h), and methoxy groups (
4f) delivered 1,1-diarylation products
4a-
4h in 45%−70% yields. Notably,
ortho-position substitution with bulkier groups (such as CH
3) rather than halogen led to lower yields. This was likely because the bulky substituent interfered with the coordination between the aniline and the Ru catalyst, thereby affecting the reaction efficiency. The protocol also accommodated disubstituted substrates (
4i-
4j).
NnullHeteroarenes were also accessible substrates, delivering
4l in 22% yield.
Substrate scope for
α–bromo alkane derivatives and alkenes was subsequently investigated (
Scheme 4). Sterically hindered
α–bromo esters (
4m),
α–bromo ketones (
4n), and
α–bromo amides (
4o-4p) proved to be compatible substrates. The reaction scope was further extended to styrene derivatives and
α,β-unsaturated alkenes.
para-Substituted styrenes with both electron-donating (
4s) and electron-withdrawing groups (
4r) exhibited good reactivity.
meta-Substituents also demonstrated similar reactivity in three-component reaction (
4t). The styrenes bearing pyridine moiety were also compatible with the reaction conditions to afford
4u Strained four-membered ring was maintained during the reaction of 4-vinylbenzocyclobutene (
4v), further proving the mildness of the protocol. The reaction of the 1,1-diphenylethylene proceeded, yielding a three-component adduct (
4w) rather than a simple radical adduct.
α,β-Unsaturated esters (
4x) were also viable substrates for 1,2-alkylarylation. Notably, in contrast to other substates, 4-vinylpyridine and 1,1-diphenylethylene afforded
para-substituted anilines as main products. Besides monobromo substrates, the substrates bearing dibromo substituents were synthesized and subjected to the reaction conditions.
3d exhibited reactivity at
α-position of ester group instead of primary bromo site (
4y). When multiple
α-positions of ester group were present in the polybrominated substrate, only monosubstituted products were obtained under our reaction conditions (
4z).
1,1-Diarylalkyl moiety is important building blocks in natural products and drugs [
76]. Switching different 1,1-diarylalkyl moieties might improve the physicochemical properties
via regulating the steric configuration, hydrophobicity, hydrogen bonding,
etc. [
77-
79]. The robustness of this protocol was further demonstrated by the successful late-stage functionalization of
α–bromo esters derived from complex pharmaceuticals and natural products by introducing the 1,1-diarylalkyl moieties (
Scheme 5). The scope of the reaction was striking:
α–bromo esters containing piperonyl alcohol (
4aa), terpenoid (
4ab, 4ad and
4af), sesquiterpenoid (
4ae), steroid (
4ac and
4ag), vitamin E (
4ai), or glucose derivatives (
4ah and
4aj) were shown to be compatible with our system. More specifically, sensitive functional groups, such as ethers (
4aa), ketone (
4ac), alkenes (
4ae and
4ag), ketal (
4aa, 4ah, 4aj) and aryl group (
4ai) were retained. Additionally, the structures of products
4 were explicitly established by single-crystal X-ray diffraction analyses of representative
4ad (CCDC number: 2441323). Notably, the protocol also leaves the intact amino handle for further conversion of these nature product derivatives to antibody drug conjugate similar to previous reports [
80].
To elucidate the radical generation of this reaction, we carried out a series free radical trapping experiments. In the presence of 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO), the reaction was completely inhibited and a radical adduct
6 of
3a and TEMPO was obtained (
Table 2, entry 1), indicating that the reaction involved a radical process. The absence of
1a led to less production of
6 (entry 2). According to previous literatures [
46,
48-
51,
67,
69,
73,
81-
86], styrene is not readily coordinate with Ru in this
σ-activation reaction. Indeed, the yield of
6 is not decreased without styrene
2a (entry 3). Notably,
6 was not generated in the absence of PPh
3 or Ru (entries 4 and 5), highlighting PPh
3 and Ru played a crucial role in the production of free radicals. Thus, we considered that Ru and PPh
3 would form a complex which is responsible for reduction of alkyl bromide
3a to generate corresponding radical. Furthermore, complex
A, RuCl
2(PPh
3)
3, which is readily generated from RuCl
3 and PPh
3 proved more effective in radical generation (entry 6), which is consistent with our optimal results in
Table 1 (entry 5). It also suggests
p-cymene might not be involved in the catalytic Ru species as ligands.
According to literatures [
87-
89], C-H functionalization using [RuCl
2(
p-cymene)]
2 normally undergoes two pathways: (1)
p-cymene dissociates from [RuCl
2(
p-cymene)]
2 and remaining Ru complex coordinate with substrates [
90–
95]; (2) [RuCl
2(
p-cymene)]
2 directly coordinates with substrates [
67,
69,
82-
84,
86]. To further identify the catalytic Ru species generated from [RuCl
2(
p-cymene)]
2, a series of ruthenium complexes including complexes
B-
E with
p-cymene were synthesized to compare with complex
A, RuCl
2(PPh
3)
3 (
Table 3, and Supporting information for detail of their synthesis). When subjecting to the reaction without additional PPh
3, complexes
B-
E exhibited poor catalytic efficiency (entries 2–5). Despite the increase of the yield of
4a by
B and
C after the addition of PPh
3 (entries 7 and 8), the complex
A, RuCl
2(PPh
3)
3, without additional PPh
3 almost afforded the same reaction yield to the optimal conditions (entry 6). Therefore, complex
A without
p-cymene was more likely to be a catalytic species in the reaction process, which is transformed from [RuCl
2(
p-cymene)]
2. Notably, similar dissociation of
p-cymene from [RuCl
2(
p-cymene)]
2 was observed in many visible-light mediated processes [
90-
95].
Based on these results and previous literatures [
67], a plausible mechanism for this three-component reaction was proposed, as illustrated in
Scheme 6. The reaction initiates with the dissociation of
p-cymene to form complex
A. Subsequently,
A reacts with alkyl halide
3a to generate intermediate
I, which is further converted to alkyl radical
Ⅲ and a ruthenium(Ⅲ) intermediate
Ⅱ via ligand-to-metal charge transfer (LMCT) processes. The alkyl radical
Ⅲ promptly adds to alkene
2a, resulting in the formation of benzyl radical
Ⅳ. Meanwhile, the ruthenium(Ⅲ) intermediate
Ⅱ coordinates with aniline to form complex
Ⅴ. The benzyl radical
Ⅳ attacks the aromatic moiety
para to the ruthenium center, leading to the formation of intermediate
Ⅵ. This intermediate undergoes a LMCT process to yield complex
Ⅶ. Finally, rearomatization of
Ⅶ occurs, releasing the difunctionalization product
4k and regenerating the catalytically active ruthenium(Ⅱ) species
A.
In summary, we report a mild ruthenium-catalyzed three-component reaction enabling alkylarylation of alkenes via para-selective C-H activation of unprotected anilines. Substrate scope encompasses electron-rich/depleted anilines, hindered α-bromo derivatives, and styrenes, with late-stage functionalization for complex natural products. Mechanistic studies reveal a dissociation of p-cymene during the catalytic processes, which is distinct from most previous σ-activation pathways. The method circumvents traditional limitations (e.g., protecting groups, harsh conditions) and offers applicability in pharmaceutical and natural product modification.
PDF
Full
54
30
关闭全屏
No. 86 Xueyuan South Road, Haidian District, Beijing
010-62662699(Operation and Technical Support)
qkjq@cast.org.cn
