2020 Volume 68 Issue 6 Pages 552-554
α,β-Unsaturated amides were incorporated as viable dipolarophiles in a catalytic asymmetric 1,3-dipolar cycloaddition of azomethine imines. The use of a 7-azaindoline auxiliary was essential to acquire sufficient reactivity with excellent diastereoselectivity, likely due to the chelating activation of the amide by the In(III)/bishydroxamic acid complex. Although the enantioselectivity remains unsatisfactory, this work is an important step toward the development of an asymmetric catalysis utilizing stable and low-reactive substrates.
Heterocyclic units are common structural subsets in a plethora of biologically active compounds, such as agrochemicals and therapeutics.1–4) 1,3-Dipolar cycloaddition reactions offer expeditious construction of heterocycles by simultaneously forging two bonds in an intermolecular manner.5–9) Azomethine imine 1 derived from 3-pyrazolidinone is a unique dipole that can produce fused bicyclic compounds hinged by two nitrogen atoms.10,11) The first disclosure of catalytic asymmetric 1,3-dipolar cycloaddition of 1 was reported by Fu and colleagues, in which 1 and terminal alkynes were enantioselectively coupled by the action of a chiral Cu(I) complex and an amine.12,13) This initial discovery was followed by 1,3-dipolar cycloadditions using electron-deficient olefins as dipolarophiles, affording a series of enantioenriched products with the expected regioselectivity (N1–Cb)14–32) (Chart 1). By taking advantage of the iminium character of 1, electron-rich olefins, including enolate species, can reverse the regioselectivity, allowing for the formation of a diverse set of bicyclic compounds with the N–N hinge.33–42) For the reaction with electron-poor dipolarophiles, electron deficiency is a primary determinant that controls the reactivity, and thereby the scope of dipolarophiles has been largely limited to highly electrophilic olefins conjugated with aldehyde,14,18) ketone,15,26) imide,16,17) oxindole,21,25) or malonyl functionalities.22,24) In our continuing efforts to engage less electrophilic α,β-unsaturated amides in catalytic asymmetric reactions,43–45) we reasoned that our 7-azaindoline auxiliary allows for 1,3-dipolar cycloaddition of 1 with less reactive amide-conjugated olefins. Herein we report the incorporation of α,β-unsaturated amides as a viable dipolarophile in the 1,3-dipolar cycloaddition of azomethine imine 1.
On the basis of the previous 1,3-dipolar cycloaddition of α,β-unsaturated 7-azaindoline amides 2 promoted by the In(III)/bishydroxamic acid (BHA)46–54) catalytic system,44) we began our investigation by screening the BHA scaffold in the reaction using azomethine imine 1 (Table 1). Intriguingly, both chemical yield and enantioselectivity changed drastically depending on the BHA ligand used, while the diastereoselectivity was largely determined by substrates and remained uniformly high. Unexpectedly, BHA-1,3, optimal BHA ligands in the reaction with 2a and nitrones,44) resulted in poor reaction outcomes. Structural modifications to increase or decrease the steric factor (BHA-2,4) marginally improved both the yield of 3a and enantioselectivity. Ligand-bearing fluorene substituents (BHA-5) afforded higher yields but the enantioselectivity remained the same. The introduction of bulkier trityl groups (BHA-6) improved the enantioselectivity (45% enantiomeric excess (ee)). BHA-7,8, in which the methylene units were truncated to induce more steric bias, delivered almost racemic product 3a in excellent yield. An aliphatic BHA ligand armed with adamantly groups (BHA-9) exhibited poor catalytic performance. Further screening of the conditions using optimal ligand BHA-6 was not fruitful.55) The 7-azaindoline auxiliary proved indispensable for acquiring sufficient reactivity with azomethine imine 1 (Chart 2). Structurally similar indoline amide 2b or 5-azaindoline amide 2c was unreactive under the current catalytic conditions (2b,c), indicating that chelating coordination of the 7-azaindoline amide moiety (2a) had a pivotal role in the in situ activation.44) Intriguingly, an acyclic amide (2d) capable of chelation via a pyridyl nitrogen failed in the reaction, implying the uniqueness of the cyclic 7-azaindoline auxiliary for strong activation. As expected, a simple dimethylamide (2e) was unreactive, which further highlights the utility of 7-azaindoline.
a) 1: 0.12 mmol, 2a: 0.1 mmol. Diastereoselectivity (dr) was determined by 1H-NMR analysis of the crude mixture. Enantioselectivity (ee) was determined by chiral stationary phase HPLC analysis.
With the optimal ligand in hand, we investigated the scope of the present 1,3-dipolar cycloaddition of α,β-unsaturated 7-azaindoline amides 2 (Table 2). Reaction could be run with lower catalyst loading (5 mol%) to afford 3a in high yield. A methyl substitution on the β-phenyl group at m- and p-positions was tolerated to give product 3f,g with comparable enantioselectivity. Amides with electron-withdrawing substituents at the para-position were compatible, albeit with slightly diminished enantioselectivity (3h–j). The high reactivity was maintained with amides having electron-donating functionalities (p-OTBS, p-OMe, m-OMe), delivering product 3k–m with higher enantioselectivity than the products 3h–j with electron-withdrawing substituents.
a) 1: 0.12 mmol, 2: 0.1 mmol. Isolated yield is reported. Diastereoselectivity (dr) was determined by 1H-NMR analysis of the crude mixture. Enantioselectivity (ee) was determined by chiral stationary phase HPLC analysis.
In summary, we developed a catalytic asymmetric 1,3-dipolar cycloaddition of an azomethine imine incorporating α,β-unsaturated amides as viable electrophiles. Although the enantioselectivity remains unsatisfactory from a synthetic point of view, this study is an important step toward developing catalytic asymmetric reactions utilizing generally unreactive α,β-unsaturated amides.
This work was financially supported by KAKENHI (17H03025 and 18H04276 in Precisely Designed Catalysts with Customized Scaffolding) from JSPS and Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan. We are grateful to Dr. Tomoyuki Kimura, Dr. Ryuichi Sawa, Ms. Yumiko Kubota, and Dr. Kiyoko Iijima at the Institute of Microbial Chemistry for technical support in NMR and X-ray crystallographic analysis.
The authors declare no conflict of interest.
The online version of this article contains supplementary materials.