Jichao
Xiao
a,
Tingting
Jia
a,
Shuang
Chen
a,
Mengxiao
Pan
a and
Xingwei
Li
*ab
aSchool of Chemistry and Chemical Engineering, Shaanxi Normal University, Xi'an, Shanxi 710062, China. E-mail: lixw@snnu.edu.cn
bInstitute of Molecular Science and Engineering, Institute of Frontier and Interdisciplinary Sciences, Shandong University, Qingdao, Shandong 266237, China
First published on 27th August 2024
Chiral alpha-amino ketones have found extensive applications as functional molecules. A nickel-catalyzed, enantioselective, and fully intermolecular three-component 1,2-alkylacylation of N-acyl enamides has been realized with tertiary alkyl bromides and carboxylic acid-derived electrophiles as the coupling reagents. This reductive coupling strategy is operationally simple, exhibiting broad substrate scope and excellent functional group tolerance using readily available starting materials and allowing rapid access to structurally complex α-amino ketone derivatives in high enantioselectivity. A suitable chiral biimidazoline ligand together with additional chelation of the amide carbonyl group in a Ni alkyl intermediate facilitates the enantioselective control by suppressing the background reaction, accounting for the excellent enantioselectivity. Mechanistic studies indicated intermediacy of radical species.
On the other hand, nickel-catalyzed asymmetric functionalization of olefins stands out as a prominent strategy toward facile construction of chiral complex molecules owing to the unique roles of Ni catalysts that readily participate in single electron transfer (SET) and reductive elimination.24–26 Consequently, catalytic three-component radical difunctionalizations of alkenes represent a powerful strategy for rapid enhancement of molecular complexity by simultaneously forging two functionalities in one single operation, thus enriching the synthetic toolbox to access synthetic building blocks, medicines, and bioactive molecules.27–41 However, enantioselective control of the target stereogenic center remains a formidable challenge. Impressive progress has been made in nickel-catalyzed asymmetric radical three-component difunctionalizations of alkenes. These coupling systems include diarylation,42 alkyl–arylation,43–49 alkyl–alkenylation,50,51 aryl–alkylation,52 and sulfonyl–carbolation53,54 under redox-neutral or reductive conditions, where organohalides serve as the terminal (the 3rd) component.55,56 However, as an important class of dicarbofunctionalization, asymmetric carboacylation reactions have received less attention and are restricted to two categories of intramolecular couplings. A carbamoyl halide-tethered olefin may react with a carbon terminating reagent via cyclization and C–C coupling under redox-neutral or reductive conditions (Scheme 1C).57–66 Alternatively, an aryl iodide-tethered olefin may react with an acylating reagent to fulfil the similar type of reaction.67,68 More recently, the groups of Chu,69 Stanley,70 Wang,71 Yuan72 and other73–77 have developed various methods for fully three-component racemic carboacylation of olefins. In most cases, a nickel acyl species captured a secondary radical species to deliver the final product through reductive elimination. However, the exceptionally high instability and reactivity of open-shell radical intermediates poses significant challenges in enantioselective control in these three-component reactions in general.78–82 In addition, the carbon radical intermediate may also undergo uncatalyzed C–C coupling with a reactive acylating regent, leading to a background reaction. To the best of our knowledge, enantioselective fully three-component 1,2-alkylacylation of alkenes remains unknown. Therefore, the development of 1,2-alkylacylation reactions toward asymmetric synthesis of α-amino ketones will be an intriguing but challenging task.
Herein, we now describe modular synthesis of enantio-enriched α-amino ketones via an efficient nickel-catalyzed enantioselective 1,2-alkylacylation of enamides with an alkyl bromide and an acylating regent, which demonstrates an example of the integration of fully three-component carboacylation of alkenes with a high level of enantioselectivity for the first time (Scheme 1D).
Entry | Deviation from standard conditions | Yieldb (%) | eec (%) |
---|---|---|---|
a Optimal reaction conditions: enamide (1a, 0.1 mmol), alkyl bromide (2a, 0.2 mmol), acid anhydride (3a, 0.15 mmol), NiBr2·diglyme (10 mol%), L1 (15 mol%), picoline, MgCl2, nanopowder Zn (4.0 equiv.), and DMAc/THF was stirred at 25 °C for 48 h. b Isolated yield. c The ee was determined using HPLC with a chiral stationary phase. | |||
1 | None | 82 | 92 |
2 | L2 instead of L1 | 81 | 91 |
3 | L3 instead of L1 | 75 | 87 |
4 | L4 instead of L1 | 77 | 87 |
5 | L5 instead of L1 | 65 | 77 |
6 | L6 instead of L1 | 70 | 79 |
7 | L7 instead of L1 | 23 | 21 |
8 | L8 instead of L1 | 43 | 12 |
9 | L9 instead of L1 | 67 | 31 |
10 | L10 instead of L1 | 32 | 15 |
11 | L11 instead of L1 | 13 | 45 |
12 | Zinc instead of nanopowder zinc | 53 | 91 |
13 | No L1 | 26 | 0 |
14 | No NiBr2·diglyme | 21 | 0 |
15 | No nanopowder zinc | 0 | — |
16 | No MgCl2 | 18 | 93 |
17 | No 2-picoline | 44 | 92 |
18 | THF only | 38 | 72 |
19 | DMAC only | 85 | 89 |
![]() |
With the optimized conditions in hand, we next investigated the generality of this three-component carboacylation. As illustrated in Table 2, vinyl amides bearing a wide variety of substituents delivered the desired products in good to excellent yields with good to excellent enantioselectivities. Enamides with an electron-donating (5–9 and 21) or -withdrawing (10–20 and 30) substituent at different positions of the benzene ring coupled efficiently, affording the α-amino ketones in 55–90% yields and 87–94% ee. Such mild reaction conditions allowed the use of a diverse spectrum of functional groups, including ether (7 and 12), thiolether (8), amine (9), ketone (13), ester (14), nitrile (15), sulfone (16), and aminosulfonyl (30). Notably, some potentially reactive functionality such as aryl bromide (17) and chlorides (18, 20 and 25) were left intact, offering opportunities for further useful transformations. A series of heterocycles including furan (23), thiophene (24), pyridine (25), benzofuran (26), and pyrimidine (27), which are frequently found in pharmaceutically active molecules, were also compatible. To our delight, cycloalkyl-substituted amides were also viable (28 and 29), albeit with attenuated enantioselectivity (82–90% ee). In contrast, 1,1- and 1,2-disubstituted internal olefins turned out to inapplicable under the standard conditions (see the ESI† for unsuccessful olefins).
a Reaction conditions: enamide (0.1 mmol), alkyl bromide (0.2 mmol), acid anhydride (0.15 mmol), NiBr2·diglyme (10 mol%), ligand L1 (15 mol%), picoline, MgCl2, and nanopowder Zn (4.0 equiv.) in DMAc/THF was stirred at 25 °C for 48 h. Isolated yield. The ee was determined using HPLC with a chiral stationary phase. |
---|
![]() |
After defining the scope of vinyl amides, attention was then turned to the scope of the alkyl bromide component. Both cyclic tertiary bromide (31) and open-chained tertiary bromides (32–46) proved to be amenable to the coupling conditions (Table 3 (A)). We were pleased to find excellent functional group compatibility for the tertiary alkyl bromides component. Functional groups such as ether (32), primary alkyl bromide/chloride (33 and 34), aryl chloride (37), sulfone (38), tosylate (39), phthalimide (40), strained ring (44 and 45), and an internal alkene (46) were compatible, affording the corresponding products in moderate to good yields (58–82%) and excellent enantioselectivities (up to 94% ee). The absolute configuration for products 33 and 47 has been confirmed by X-ray crystallographic analyses. Interestingly, when the substrate contains both tertiary and primary alkyl bromide, the reaction is completely selective for the more sterically congested C–Br bond to give product 33 in 82% yield and 93% ee, leaving primary alkyl bromide unchanged, indicative of a stable radical species. Notably, heterocyclic substrates such as furan (41), thiophene (42), and benzofuran (43) were equally suitable for this chemistry. Unfortunately, primary and secondary alkyl bromides were found to be unreactive in our protocol, giving rise to only two-component coupling reaction with the benzoic acid anhydride.
a Reaction conditions: enamide (0.1 mmol), alkyl bromide (0.2 mmol), acid anhydride or 2-pyridyl ester (0.15 mmol), NiBr2·diglyme (10 mol%), ligand L1 (15 mol%), picoline (or K3PO4), MgCl2, and nanopowder Zn (4.0 equiv.) in DMAc/THF was stirred at 25 °C for 48 h. Isolated yield. The ee was determined using HPLC with a chiral stationary phase. |
---|
![]() |
We next demonstrated the substrate spectrum with respect to the acyl donors. Both electron-rich and -deficient groups on the different positions of the aromatic ring of the benzoic anhydrides were tolerated, furnishing the desired α-amino ketones 47–53 in moderate to good efficiency (Table 3 (B)). Additionally, we wondered whether other acyl precursors are suitable for this asymmetric three-component reductive alkylacylation. Gratifyingly, 2-pyridyl esters66 were identified to be compatible with our protocol after subtle changes of the reaction conditions (K3PO4 instead of 2-picoline as a base, Table 3 (C)). Accordingly, a variety of 2-pyridyl esters were examined. A series of functional groups, including ether (54 and 55), thioether (56), ester (58), 2-naphthyl (59), benzofuran (60), and benzothiophene (61), were nicely accommodated. Besides aryl 2-pyridyl esters, an aliphatic acid-derived 2-pyridyl ester (62) also exhibited moderate reactivity.
To further demonstrate the synthetic utility of our protocol, a gram-scale (5 mmol scale) reaction was carried out. Product 4 was isolated in a synthetically useful yield (1.11 g) with excellent enantioselectivity (92% ee, Scheme 2A). Chemoselective reduction of the ketone carbonyl group afforded product 63 in good yield and high diastereoselectivity (Scheme 2B). The reduction-Mitsunobu reaction of 4 afforded a chiral oxazoline 64 bearing two chiral centers in high dr, which could be a potential chiral ligand. Reduction of the carbonyls in both the ketone and the amide moieties afforded the amino alcohol 65 in 7.5:
1 dr. Condensation between 4 and hydroxyamine followed by a Beckmann rearrangement afforded amide 66 in good yield. In all cases, only slight, if any, erosion of the enantiopurity was observed.
To shed light on the mechanism of this novel nickel-catalyzed asymmetric alkylacylation system, a series of experiments were conducted. The desired alkylacylation reaction was completely inhibited when a radical scavenger was added (Scheme 3A). Next, a radical-clock experiment using a α-cyclopropyl styrene 67 was conducted (Scheme 3B), and only the ring-opened product 68 was obtained through a sequential radical addition, ring opening, and acylation process, whereas the formation of the direct cross-coupled product 69 was not observed. This result indicated the formation of a tert-butyl radical which, upon addition to the alkene, delivers a new carbon radical that can be intercepted by the acyl-Ni species. The formation of radical species was further confirmed by an electron paramagnetic resonance (EPR) study using a spin-trapping agent phenyl tert-butyl nitrone (PBN, Scheme 3C). Additionally, a nickel(II) acyl complex 72 (ref. 83) was prepared to elucidate the nature of the active nickel species in the catalytic cycle. The stoichiometric reaction of Ni(II)-complex 72, N-vinylbenzamide 1a, and 2-bromo-2-methylpropane 2a in the absence of any zinc reductant gave the desired cross-coupled product 62 in 18% yield, whereas only traces of 62 were detected when a stoichiometric amount of nanopowder zinc was introduced (Scheme 3D). Taken together, these results suggest that the putative acyl-Ni(II) complex could be a productive intermediate in this catalytic cycle. No corresponding product 4 could be observed when employing stoichiometric Ni(cod)2 but lacking the zinc reductant, whereas catalytic amount Ni(cod)2 with stoichiometric zinc resulted in the formation of product 4 in 63% yield (Scheme 3E). These outcome suggests that the reduction of Ni(III) acyl species III to Ni(I) acyl intermediate IV is likely required in the current protocol.
On the basis of our mechanistic investigations and insight from prior studies,71,72,83,84 a plausible mechanism of this asymmetric three-component alkylacylation reaction is proposed (Scheme 3F). Initially, Ni(I) species I may be formed via reduction of the Ni(II) precatalyst. Oxidative addition of the Ni(I) species I to the anhydrides or 2-pyridyl esters II generates a Ni(III) intermediate III, which reduced by nanopowder zinc as facilitated by MgCl2 (ref. 83 and 84) to give Ni(I) intermediate IV. Then, the tertiary alkyl bromide 2a is reduced by a Ni(I) complex IV to give a tert-butyl radical together with a Ni(II) species V (reduction of 2a by nanopowder zinc to give the tertiary alkyl radical cannot be rule out now). Subsequent addition of the tert-butyl radical to the vinyl amides 1 affords a secondary alkyl radical VI. At this juncture, oxidative addition of the radical VI to Ni(II) V delivers Ni(III) species VII, which then undergoes reductive elimination to release the chiral α-amino ketone product VIII and regenerates the Ni(I) species I to complete the catalytic cycle. We anticipate that the chelation of the carbonyl group to the BiIM-ligated nickel center would be beneficial to control the stereoselectivity in the radical capture step (V + VI → VII).43–45,50,85
Footnote |
† Electronic supplementary information (ESI) available. CCDC 2356761 and 2356762. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4sc04561k |
This journal is © The Royal Society of Chemistry 2024 |