Chenxiao
Qian‡
ab,
Jing
Huang‡
b,
Tingting
Huang‡
a,
Lijuan
Song
c,
Jianwei
Sun
*b and
Pengfei
Li
*a
aDepartment of Chemistry, Guangdong Provincial Key Laboratory of Catalysis, College of Science, Southern University of Science and Technology Guangming Advanced Research Institute, Southern University of Science and Technology (SUSTech), Shenzhen 518055, China. E-mail: lipf@sustech.edu.cn; flyli1980@gmail.com
bDepartment of Chemistry and the Hong Kong Branch of Chinese National Engineering Research Centre for Tissue Restoration & Reconstruction, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong SAR, China. E-mail: sunjw@ust.hk
cSchool of Science, Harbin Institute of Technology (Shenzhen), Shenzhen 518055, China
First published on 9th February 2024
Compared with well-developed construction of Csp2–Csp2 atropisomers, the synthesis of Csp2–N atropisomers remains in its infancy, which is recognized as both appealing and challenging. Herein, we achieved the first organocatalyzed asymmetric synthesis of Csp2–N atropisomers by formal Csp2–O amination. With the aid of a suitable acid, 3-alkynyl-3-hydroxyisoindolinones reacted smoothly with 1-methylnaphthalen-2-ols to afford a wide range of atropisomers by selective formation of the Csp2–N axis. Particularly, both the kinetic (Z)-products and the thermodynamic (E)-products could be selectively formed. Furthermore, the rarely used combination of two chiral Brønsted acid catalysts achieved excellent enantiocontrol, which is intriguing and unusual in organocatalysis. Based on control experiments and DFT calculations, a cascade dehydration/addition/rearrangement process was proposed. More importantly, this work provided a new plat-form for direct atroposelective construction of the chiral Csp2–N axis.
Currently, direct atroposelective construction of chiral Csp2–N axis mainly relies on two strategies: (i) organocatalytic asymmetric amination reactions, including electrophilic amination of electron-rich naphthalenes with azodicarboxylates/quinone diimines11 and nucleophilic amination of azonaphthalenes with electron-rich amines4c (Scheme 1A); (ii) metal-catalyzed asymmetric coupling reactions of secondary amines with aryl halides12 or diazo compounds13 (Scheme 1B). In both scenarios, limited substrate scope and harsh reaction conditions constitute the major issues. Particularly, Lin et al. gracefully disclosed a chiral phosphoric acid (CPA)-catalyzed atroposelective cascade reaction among 2,3-diketoesters, aromatic amines and 1,3-cyclohexanediones for the construction of axially chiral N-arylindoles (Scheme 1C).14 The reaction, however, still necessitated the utilization of dichloromethane at a temperature of 65 °C. Consequently, new efficient methods under mild conditions for the direct atroposelective construction of the Csp2–N bond remain in high demand.
It is known to be challenging to break a Csp2–O bond attached to electron-rich aromatic rings, as a result of the high C–O bond dissociation energy. Generally, transition metal catalysts, including Ni- and Pd-catalysts, are required for such C–O bond activation.15 It is also known that aryl C–O bonds can be converted to C–N bonds with transition-metal-catalyzed catalysis, but no success has been achieved for direct construction of a chiral Csp2–N axis by Csp2–O bond activation.16 Moreover, transition-metal-free amination often necessitates harsh conditions, such as high temperature and strong alkali.17 Based on our previous research on indolinone derivatives18 and in continuation of our efforts on the organocatalytic reactions of functionalized alcohols,19 herein we discovered an intriguing dehydration/addition/rearrangement reaction between 1-substituted naphthols and indolinone derivatives leading to the formation of unusual Csp2–N atropisomeric scaffolds (Scheme 1D). More importantly, the catalytic asymmetric variant of this process has also been achieved by the combination of two chiral Brønsted acid catalysts.20
Entrya | A | Solvent | Yield [%]b | |
---|---|---|---|---|
a Unless noted, under Ar atmosphere, a mixture of 1a (0.05 mmol), 2a (0.10 mmol) and A (10 mol%) in the solvent (0.5 mL) was stirred at room temperature (rt) for 24 h. b Determined by 1H NMR with CH2Br2 as internal standard. | ||||
1 | A1 | CH2Cl2 | (Z)-3aa, 63 | (E)-3aa, 15 |
2 | A2 | CH2Cl2 | (Z)-3aa, 10 | (E)-3aa, 60 |
3 | A3 | CH2Cl2 | (Z)-3aa, 14 | (E)-3aa, 83 |
4 | A4 | CH2Cl2 | (Z)-3aa, <5 | (E)-3aa, <5 |
5 | A5 | CH2Cl2 | (Z)-3aa, 10 | (E)-3aa, 76 |
6 | A1 | MeCN | (Z)-3aa, <5 | (E)-3aa, <5 |
7 | A1 | EtOAc | (Z)-3aa, 46 | (E)-3aa, <5 |
8 | A1 | THF | (Z)-3aa, 22 | (E)-3aa, <5 |
9 | A1 | CH2Cl2 (1.0 mL) | (Z)-3aa, 82 | (E)-3aa, <5 |
10 | A1 | CH2Cl2 (1.2 mL) | (Z)-3aa, 90 | (E)-3aa, <5 |
11 | A1 | CH2Cl2 (1.5 mL) | (Z)-3aa, 84 | (E)-3aa, <5 |
Under the optimized conditions, the scope of the cascade dehydration/addition/rearrangement reaction between 1-methylnaphthalen-2-ols 1 and 3-alkynyl-3-hydroxyisoindolinones 2 was examined at a relatively larger scale. Because the target products (Z)-3 could be converted to the thermodynamic products (E)-3, the time of each reaction was monitored by thin-layer chromatography (TLC). In some of these reactions, the amount of A1 was increased to 20 mol% to shorten the reaction time to reduce isomerization. As shown in Table 2A, a range of naphthols 1a–j reacted smoothly with 3-alkynyl-3-hydroxyisoindolinone 2a to provide the corresponding (Z)-2-(naphthalen-2-yl)-isoindolin-1-ones (Z)-3aa–ja in moderate yields. These encouraging results indicated that the protocol could be applied to naphthols bearing different substituents, including trimethylsilyl (SiMe3, Table 2, (Z)-3ia) and terminal olefin (Table 2, (Z)-3ja). In addition, the A1-catalyzed reaction also showed excellent compatibility with 3-alkynyl-3-hydroxyisoindolinones bearing different substituents in the benzene ring, affording the desired products (Z)-3ab–ah in moderate to high yields (Table 2B). However, the methyl group in the 1-position of naphthol seemed to be essential. Replacing it by other substituents (H, Et) would lead to a complex mixture instead of desired product. The reaction of 1,3-dimethylnaphthalen-2-ol also resulted in the complex mixture. The configuration of (Z)-3aa was unambiguously confirmed by X-ray crystallography (CCDC 2294518). The configurations of other products were assigned by analogy.
a A mixture of 1 (0.20 mmol), 2 (0.40 mmol) and A1 (10 mol%) in CH2Cl2 (4.8 mL) was stirred at room temperature (rt) under Ar atmosphere for noted time. b A mixture of 1 (0.20 mmol), 2 (0.40 mmol) and A1 (20 mol%) in CH2Cl2 (6.0 mL) was stirred at 30 °C under Ar atmosphere for the noted time. c A mixture of 1 (0.20 mmol), 2 (0.40 mmol) and A3 (10 mol%) in CH2Cl2 (2.4 mL) was stirred at rt under Ar atmosphere for 24 h. |
---|
The generality for the formation of the thermodynamic products (E)-3 was also examined (Table 2C). Notably, the presented products were obtained in high yields after further optimizations. Exceptionally, some thermodynamic products were either formed in low efficiency or hard to isolated from the reaction mixture. The configuration of (E)-3aa was confirmed by X-ray crystallography (CCDC 2294530). Taken together, a divergent cascade reaction between 1-methylnaphthalen-2-ols and 3-alkynyl-3-hydroxyisoindolinones was achieved for the direct construction of Csp2–N bond from Csp2–O bond to furnish the kinetic products (Z)-3 and the thermodynamic products (E)-3, respectively.
The successful construction of the Csp2–N axis inspired us to further develop an enantioselective variant of this process. Thus, the reaction between 1a and 2a was studied for the initial evaluation of some chiral phosphoric acids (Table 3, entries 1–2) and chiral phosphoramides (Table 3, entries 3–7) as potential catalysts. Indeed, chiral phosphoramide C1 was found to be a promising catalyst, affording the desired (R, Z)-3aa in 64% yield with 65% ee (Table 3, entry 3). While a single catalyst could not eventually improve the enantioselectivity to an excellent level, we were curious about the use of combined Brønstred acid systems. The combination of a chiral phosphoric acid and a chiral phosphoramide was then surveyed (Table 3, entries 8–10). Pleasingly, when phosphoramide C1 was used together with chiral phosphoric acid B2, the desired (R, Z)-3aa was formed with 90% ee (Table 3, entry 10). Notably, control experiments indicated that the use of any other combination or any single acid could not reach such a high level of enantiocontrol. Other solvents could not improve the reaction outcome (Table 3, entries 11–13). Notably, the loading of B2 and C1 could be further reduced to 8.0 mol% without compromising the efficiency (Table 3, entry 14). No product was obtained when the reaction was carried out at 0 °C (Table 3, entry 16). It is worth noting that substrate 2a has a low solubility at 0 °C. Therefore, a programmed cooling protocol was used to smoothly afford the target product (R, Z)-3aa with 93% ee (Table 3, entry 17).
Entrya | CPA | Solvent | Temp [oC] | Yield [%]b | ee [%]c |
---|---|---|---|---|---|
a Unless noted, under Ar atmosphere, a mixture of 1a (0.05 mmol), 2a (0.10 mmol) and CPA (10 mol%) in the solvent (2.0 mL) was stirred at room temperature (rt) for 24 h. b Determined by 1H NMR with CH2Br2 as internal standard. c Determined by chiral-HPLC analysis. d C1 (10 mol%), B1 (10 mol%). e C2 (10 mol%), B2 (10 mol%). f C1 (10 mol%), B2 (10 mol%). g C1 (8 mol%), B2 (8 mol%). h C1 (5 mol%), B2 (5 mol%). i The mixture was stirred at rt under Ar atmosphere for 5 min (2a was totally dissolved), then it was cooled to 0 °C for 6 days. | |||||
1 | B1 | CH2Cl2 | rt | Trace | — |
2 | B2 | CH2Cl2 | rt | Trace | — |
3 | C1 | CH2Cl2 | rt | 64 | 65 |
4 | C2 | CH2Cl2 | rt | 20 | 46 |
5 | C3 | CH2Cl2 | rt | 44 | 53 |
6 | C4 | CH2Cl2 | rt | Trace | — |
7 | C5 | CH2Cl2 | rt | 18 | 7 |
8d | C1/B1 | CH2Cl2 | rt | 40 | 61 |
9e | C2/B2 | CH2Cl2 | rt | 34 | 90 |
10f | C1/B2 | CH2Cl2 | rt | 44 | 90 |
11f | C1/B2 | MeCN | rt | Trace | — |
12f | C1/B2 | EtOAc | rt | Trace | — |
13f | C1/B2 | THF | rt | Trace | — |
14g | C1/B2 | CH2Cl2 | rt | 44 | 90 |
15h | C1/B2 | CH2Cl2 | rt | 32 | 88 |
16g | C1/B2 | CH2Cl2 | 0 | Trace | — |
17g,i | C1/B2 | CH2Cl2 | rt to 0 | 56 | 93 |
Under the established conditions, the scope of this combined catalytic system was examined for this intriguing process for the direct formation of atroposelective Csp2–N axis (Table 4). The kinetic products (R, Z)-3 were exclusively obtained under the optimized conditions. More specifically, a range of naphthols 1a–j reacted smoothly with 3-alkynyl-3-hydroxyisoindolinone 2a to provide the corresponding enantioenriched products (R, Z)-3aa–ja in moderate yields. The halogen (Br, Cl) could be introduced into the aromatic ring of naphthol, affording (R, Z)-3ba–fa in 32–68% yields with 79–92% ee. The reaction of naphthol with bulky substituents (3,5-(t-Bu)2C6H3, 2-naphthyl) could also proceed smoothy, furnishing the corresponding products (R, Z)-3ga in 32% yield with 76% ee and (R, Z)-3ha in 64% yield with 94% ee, respectively. Particularly, trimethylsilyl and the terminal olefin unit were also compatible, and (R, Z)-3ia was obtained in 45% yield with 85% ee and (R, Z)-3ja in 40% yield with 94% ee. Moreover, a series of 3-alkynyl-3-hydroxyisoindolinones with different substituents (3-MeC6H4, 3-ClC6H4, 4-FC6H4, 4-ClC6H4, 4-BrC6H4) were also reactive to afford the desired products in generally moderate yields with high enantioselectivities. The reactions could be run at a relatively large scale (0.5 mmol), which afforded the desired products with comparable results as the small-scale reactions. The configuration of (R, Z)-3aa was confirmed by X-ray crystallography (CCDC 2294540).
a Under Ar atmosphere, a mixture of 1 (0.20 mmol), 2 (0.40 mmol), C1 (8 mol%) and B2 (8 mol%) in CH2Cl2 (8.0 mL) was stirred at room temperature (rt) for 5 min, then it was cooled to 0 °C for 6 h. Products (R, Z)-3 were obtained in isolated yields. The ee was determined by chiral-HPLC analysis. |
---|
To gain more information about the reaction mechanism, we have performed a series of control experiments. First of all, product (Z)-3aa could be converted to (E)-3aa in the presence of catalyst A3, which confirmed that the former is the kinetic product and the latter is the thermodynamic product (Scheme 2A).
Furthermore, no reaction took place between naphthalen-2-ol 1a and enone 4 from 3-alkynyl-3-hydroxyisoindolinone 2a under the standard conditions, suggesting that enone 4 was not a viable intermediate (Scheme 2B, eqn (1)). When H218O was added to the A1-mediated reaction, the 18O atom was incorporated into the product (Z)-3aa-18O (Scheme 2B, eqn (2)). Importantly, with 4 Å molecular sieves as additive, this process provided an allene as product, which was a mixture of diastereomers IM-up (weak polarity) and IM-down (strong polarity). The structure of this allene product was confirmed un-ambiguously by X-ray crystallography (CCDC 2294537, Scheme 2B, eqn (3)). Interestingly, the diastereomeric ratio (IM-up and IM-down) changed over time, likely via reverse reaction. Furthermore, with B2 as catalyst, the allene diastereomers could be isolated as enantioenriched form (59% ee for IM-up, 39% ee for IM-down), both of which could lead to enantioenriched product (R, Z)-3aa in the presence of the racemic catalyst C1 (Scheme 2B, eqn (4)), suggesting that the conversion from allene to product could be stereospecific. Moreover, the racemic allene diastereomers were also generated in the presence of catalyst A1 combined with 4 Å molecular sieves. Notably, treated with the combined catalysts, chiral C1 and B2, enantioenriched (R, Z)-3aa could be still obtained (Scheme 2B, eqn (5)). We reasoned that, even though racemic, the allene intermediate could reverse to their precursors in the presence of the chiral acid to generate enantioenriched allene, which resulted in the enantioenrichment in the final product. In this case, the rate-determining step is likely at a later stage (after allene formation), which would allow reversible steps before the addition of water to allene.
Interestingly, enantiopurity of product (R, Z)-3aa has a linear correlation to both the enantiopurity of C1 (Scheme 3A) and B2 (Scheme 3B), suggesting that both of them are contributing to the key enantiodetermining transition state. It is worth noting that such a scenario with two different chiral Brønsted acids contributing to enantiocontrol is extremely uncommon, to the best of our knowledge.
To understand the mechanism, density functional theory (DFT) calculations were performed on the racemic reaction. The results indicated that this intriguing process comprises of dehydration, allene formation, water addition, cyclization, and ring-opening (Fig. 2). It starts substrate coordination to naphthalene-2-sulfonic acid (A1) by hydrogen bonding followed by elimination of water to form intermediate INT3 with an energy barrier of 16.7 kcal mol−1. Next, a nucleophilic attack from the naphthol to the resulting imine INT4 occurs together with synergistic proton transfer, leading to the allene intermediate INT5, with an energy barrier 12.4 kcal mol−1. In the presence of water, the electron-rich allene motif can be further activated by acid and trigger a water addition to form INT9. Next, an intramolecular cyclization takes place by the nucleophilic attack of the N-atom to the carbonyl group to form a C–N bond. Further dehydration forms iminium INT13. Finally, a facile ring-opening process driven by the aromatization furnishes the final product. It is worth noting that the water elimination from INT11 is the rate-determining step, which is consistent with the control experiments showing that the allene formation might be likely reversible.
Fig. 2 DFT study. Energetic profiles of the reaction at the M06-2X-D3/6-311G(d,p)-SMD(DCM)//M06-2X-D3/6-31G(d) level of theory and the energy values are in given kcal mol−1. |
Footnotes |
† Electronic supplementary information (ESI) available. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d3sc06707f |
‡ These authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2024 |