1. Introduction

Imines are important synthons in the asymmetric synthesis of chiral nitrogen heterocycles, which constitute a privileged structural motif prevalent in biologically active compounds. The rising percentage of nitrogen heterocycles in approved pharmaceuticals^[1] underscores the urgent need to enhance structural diversity in these molecules and advance imine-based synthetic methodologies. In recent years, O-silylated hemiaminals have emerged as highly reactive imine precursors, which can in situ generate various methanimines and thus circumvent several issues inherent to conventional preformed imines (Figure 1A)^[2-22]. However, to the best of our knowledge, the chemical space of these imine precursors remains constrained in asymmetric synthesis^[23-28], particularly for constructing chiral nitrogen heterocycles^[29-31].

Display Full Size

Figure 1. Asymmetric dearomative of indole with methanimines for THQs. THQ: Tetrahydroquinoline; CPA: chiral phosphoric acid.

The [4+2] cycloaddition reaction of N-aryl methanimines with dienophiles has proven effective for assembling the bioactive chiral tetrahydroquinolines (THQs) skeleton^[32-43]. Nevertheless, indoles remain notably underexplored as dienophiles, whether in asymmetric versions or not, although this dearomative strategy^[44-46] provides direct access to indoline-fused THQs^[47-50]. Moreover, despite their unique potential for enhancing the molecular diversity of tetrahydroquinoline scaffolds, methaniminium cations have never been reported as diene partners for this asymmetric transformation, whether derived from O-silylated hemiaminals^[51] or alternative precursors (Figure 1B)^[52-55]. The primary challenges arise from the limited steric differentiation of such methanimines and the inherent complexity of the dearomative transformation, which hinders precise control over regio- and enantioselectivity and limits the improvement of catalytic efficiency.

Considering the excellent catalytic activity and selectivity of chiral phosphoric acids (CPAs) in asymmetric transformations involving imines and indoles^[56-60], together with their significance in activating O-silylated hemiaminals, we reasoned that CPAs could serve as a dual-function catalyst to facilitate the dearomative [4+2] reaction of indoles with O-silylated hemiaminals (Figure 1C): (1) enabling the in-situ formation of the methaniminium cation from O-Silylated hemiaminals as a Brønsted acid; (2) inducing the stereochemical control of the upcoming dearomative cycloaddition reaction as a chiral anion through non-covalent interactions with indoles and the methaniminium cation. Herein, we describe the first example of CPA catalyzed asymmetric dearomative [4+2] reaction of indoles with O-silylated hemiaminals for the target-oriented synthesis of 2-unsubstituted indoline-fused THQs in excellent regio- and enantioselectivities.

2. Methods

2.1 Materials

Chemicals were commercially purchased from Adamas-beta, Tansoole, Bide Pharmatech Ltd., Aladdin, and Daicel Chiral Technologies (China) Co., and were directly used without further purification unless otherwise stated. Reactions were carried out under an atmosphere of nitrogen using a glovebox unless otherwise noted. Toluene and CH₂Cl₂ were freshly distilled from CaH₂ and distilled three times prior to use.

2.2 Typical procedure for asymmetric formal [4+2] cycloaddition reaction of O-silylated hemiaminals and indoles

To a 10 mL tube which was charged with indole (0.1 mmol, 1.0 equiv.), CPA5 (5 mol%) and toluene (1.0 mL), was added the solution of O-Silylated hemiaminals in toluene (0.12M, 1.0 mL), and then the reaction mixture was stirred for 12 h at room temperature. After completion (detected by TLC), the reaction mixture was used directly for column chromatography separation (CH₂Cl₂/ethyl acetate = 20:1 to 8:1 as the eluent) to afford the corresponding pure products.

The corresponding racemic products were synthesized according to the above procedure by replacing the CPA5 with Copper(II) trifluoromethanesulphonate (3.3 mg, 10 mol%) as the catalyst.

2.3 Removal of the directing or protecting groups

2.3.1 Removal of the directing groups for 3f

To a 25 mL flask, 3f (35.5 mg, 0.1 mmol), KOH (aq.) (1 M, 1.0 mL), and ethanol (2 mL) were refluxed in a 100 ^oC oil bath for 12 h. After completion (detected by TLC), the reaction was cooled down to room temperature, and the resulting solution was neutralized with aqueous HCI and extracted with ethyl acetate (3 × 10 mL). The organic layers were dried over anhydrous Na₂SO₄ and evaporated in vacuo. The desired product 8 was obtained after purification by flash chromatography on silica gel (petroleum ether/ethyl acetate = 8:1) as a yellow oil (19.6 mg, 83% yield). The ee of compound 8 was determined by HPLC using an IC_Daicel column (n-hexane/i-PrOH = 80/20, flow rate = 1.0 mL/min, λ = 254 nm, t_minor = 4.9 min, t_major = 5.5 min).

2.3.2 Removal of the protecting groups for 7a

To a 10 mL flask were charged 7a (43.1mg, 0.1 mmol), Pd/C (10 mg, 10 mol%, 10 wt%), and MeOH (2.0 mL). The flask was evacuated and backfilled with hydrogen (3 times) and then stirred under 1 atm (balloon pressure) of hydrogen at room temperature for 24 h. When the reaction was completed, the Pd/C catalyst was filtered before evaporation of solvent. The solid residue was purified by column chromatography to afford the product 9 as a white solid (31.4 mg, 92% yield) (eluent: petroleum ether/ethyl acetate = 3:1). The ee of compound 9 was determined by HPLC using an IA_Daicel column (n-hexane/i-PrOH = 70/30, flow rate = 1.0 mL/min, λ = 254 nm, t_minor = 11.3 min, t_major = 15.1 min).

To a 10 mL flask charged with the crude product 9, KOH (1 M, 1.0 mL) and ethanol (2 mL) were refluxed in a 100 ^oC oil bath for 12 h. After completion (detected by TLC), the reaction mixture was cooled down to room temperature, and the resulting solution was neutralized with aqueous HCI and extracted with ethyl acetate (3 x 10 mL). The organic layers were dried over anhydrous Na₂SO₄ and evaporated in vacuo. The solid residue was purified by column chromatography to afford the product 10 as a white solid (13.9 mg, 63% yield for two steps) (eluent: petroleum ether/ethyl acetate = 3:1). The ee of compound 10 was determined by HPLC using an IA_Daicel column (n-hexane/i-PrOH = 70/30, flow rate = 1.0 mL/min, λ = 254 nm, t_minor = 10.1 min, t_major = 16.4 min).

3. Results and Discussion

3.1 Optimize the reaction conditions

At the first step, we investigated the feasibility of utilizing hemiaminal silyl ether as a methanimine precursor with N-substituted indole in the presence of CPA. As we reported, the CPA1 catalyzed cycloaddition reaction of hemiaminal silyl ether 1a and simple indole 2a was not observed (Figure 2). The same results were obtained when common indole derivatives, including N-methyl indole 2b, N-acetyl indole 2c, N-Boc indole 2d, and N-Ts indole 2e, were employed in the reaction. Considering that an H atom on the N-atom of the indole moiety may provide a secondary binding site to interact with the Lewis basic site (P=O) on the phosphate, a series of indole-1-carboxamides were prepared and screened in the reaction^[61]. It was delightful to observe that indole-1-carboxamides 2f-2i proceeded smoothly to produce the corresponding indoline-fused THQs 3f-3i in excellent yields (92-96% yield), albeit with low to moderate enantiomeric excess (ee) (33-49% ee). Consistent with our hypothesis, the H atom on the indole-1-carboxamide plays a vital role in the reaction. No reaction was observed when N-methyl-N-phenyl-1H-indole-1-carboxamide 2j was employed as the substrate under the same conditions.

Display Full Size

Figure 2. Optimization of the N-substituted indoles. 0.12 mmol of 1a was added to the solution of 2 (0.1 mmol), CPA1 (5 mol%) in toluene (1.0 mL). Isolated yields. ee was determined by chiral HPLC. CPA: chiral phosphoric acid; HPLC: high-performance liquid chromatography.

Encouraged by the above initial results, the conditions were further optimized using methanimine precursor O-silylated hemiaminal 1a and N-phenyl-indole-1-carboxamide 2f. As depicted in Table 1, toluene was a better solvent than DCM, DCE, THF, and 1,4-dioxane (entries 1-5). Subsequently, various H₈-BINOL-derived CPA catalysts with different steric properties were screened in the reaction (entries 6-9). It was found that the enantioselectivity was strongly responsive to the extent of the substitutions on the 3,3’-positions of the H₈-BINOL-framework, with the 1-pyrenyl catalyst CPA5 being optimal (97% yield, 97% ee). The steric demand of the catalyst, together with the potential aryl–aryl interactions with the substrate, was presumably thought to match the less sterically demanding N-phenylmethaniminium substrate and create a tailored microenvironment for asymmetric annulation. In addition, lowering the temperature to 0 °C did not improve the reaction (entry 10).

Table 1. Optimization of the reaction^a.

Display Full Size

entry	catalyst	solvent	T (oC)	yield(%)	ee (%)
1	CPA1	toluene	rt	95	49
2	CPA1	CH2Cl2	rt	89	32
3	CPA1	DCE	rt	87	29
4	CPA1	THF	rt	85	39
5	CPA1	dioxane	rt	92	47
6	CPA2	toluene	rt	94	60
7	CPA3	toluene	rt	96	86
8	CPA4	toluene	rt	95	75
9	CPA5	toluene	rt	97	97
10	CPA5	toluene	0	95	96

^a: 0.12 mmol 1a was added to the solution of 2f (0.1 mmol), CPA (5 mol%) in toluene (1.0 mL). Isolated yields. ee was determined by chiral HPLC; CPA: chiral phosphoric acid; HPLC: high-performance liquid chromatography.

3.2 Expand the reaction substrates

With the optimal conditions established, we focused on the investigation of indole ring variants (Figure 3). Notably, the reaction demonstrated excellent tolerance toward various functional groups in the indole moiety, irrespective of their electronic properties. A wide range of groups, including methyl, methoxy, fluoro, chloro, and bromo, were well compatible with this asymmetric dearomative cycloaddition reaction, regardless of their position at the 4-, 5-, or 6-position of the indole. As a result, the corresponding indoline-fused tetrahydroquinoline products 4a-4p were obtained in high to excellent yields and enantioselectivities (82 to 97% yield, 88 to 99% ee). It is noted that the product 4q, bearing an all-carbon quaternary stereocenter, was also prepared from 3-methyl indole, although both the yield and ee were reduced presumably due to the steric effects.

Display Full Size

Figure 3. Asymmetric dearomative cycloaddition with various indoles. Isolated yields. ee was determined by chiral HPLC. CPA: chiral phosphoric acid; HPLC: high-performance liquid chromatography.

Subsequently, the scope of O-silylated hemiaminals was investigated (Figure 4). Hemiaminals with electron-donating substituents, such as methyl, iso-propyl, tert-butyl, methoxy and ethoxy, underwent the reaction smoothly to afford the indoline-fused THQs with excellent enantioselectivities in high yields (5a-5g, 95 to 98% ee, 88 to 97% yield). Halide substituents including F, Cl, Br and I were well compatible with the reaction to give the corresponding halide-substituted indoline-fused tetrahydroquinolines in high yields and enantioselectivities (5h-5k, 91 to 94% ee, 81 to 87% yield). As expected, 3,5-disubstituted or 3,4,5-trisubstituted N-methyl hemiaminals proceeded smoothly under the conditions to produce the desired indoline-fused tetrahydroquinolines 5l-5n with high enantioselectivities in high yields. Specifically, hemiaminals containing a meta-substituent (methyl and chloro) gave rise to single regioselective products 5o (97% ee, 82% yield) and 5p (96% ee, 93% yield), respectively. Moreover, a 2-naphthyl hemiaminal also produced exclusively tetrahydro-7H-benzo[f]indolo[3,2-c]quinoline 5q with 98% ee and 96% yield. The absolute configuration of 5i was determined by X-ray crystallographic analysis (CCDC 2189145), and other products were assigned by analogy.

Display Full Size

Figure 4. Asymmetric dearomative cycloaddition with various hemiaminals. Isolated yields. ee was determined by chiral HPLC. CPA: chiral phosphoric acid; HPLC: high-performance liquid chromatography.

Furthermore, substituent effects of O-silylated hemiaminals on the N atom were also investigated (Figure 5). Remarkably, 6a bearing the removable benzyl (Bn) group showed similar reactivity and enantioselectivity to 1a, giving 7a with 95% ee and 97% yield. Diphenyl substitution was compatible with the reaction to give the corresponding 7b with 92% ee and 85% yield. In addition, the cyclic O-silylated hemiaminals, including 6c bearing tetrahydroquinoline, 6d bearing benzomorpholine, 6e bearing phenoxazine and 6f bearing phenothiazine, were also examined in the reaction. It was found that these substrates were well-compatible with the reaction to afford the polycyclic products 7c-7f with 90-98% ee and 63-96% yields.

Display Full Size

Figure 5. Asymmetric dearomative cycloaddition with different substituents on the nitrogen atom of O-silylated hemiaminals. Isolated yields. ee was determined by chiral HPLC. CPA: chiral phosphoric acid; HPLC: high-performance liquid chromatography.

3.3 Synthetic transformation

For further synthetic transformation, the directing group PhNHCO- at the indole moiety was conveniently removed in 83% yield and 97% ee (Figure 6). Notably, PhNHCO- at the indole moiety and Bn- at the tetrahydroquinoline moiety of the final products could be selectively removed to afford indoline-fused tetrahydroquinolines bearing free NH groups in high yields without loss of enantiomeric purity, which are more valuable from the perspective of molecular complexity in synthetic chemistry and drug discovery. This merit would also pave the way for late-stage derivatization in more complex polycyclic frameworks.

Display Full Size

Figure 6. Selective removal of the directing and protecting groups.

3.4 Mechanistic studies

To gain mechanistic insight into the catalytic cycle, we conducted a series of control experiments. The reaction of hemiaminal silyl ether 1a with 2f did not proceed in the absence of phosphoric acid, but the formal cycloaddition occurred in the presence of p-toluenesulfonic acid (Figure 7A). These experiments suggest that the CPA plays a vital role in the desilanolation of hemiaminal silyl ether 1a to generate the active iminium phosphate for the cycloaddition reaction. In principle, both the stepwise process (a, b and d) and the concerted pathway (c and d) are possible for the formal [4+2] cycloaddition of N-aryl imines and electron-rich olefins (Figure 7B). To gain further insight into the mechanism, kinetic studies were performed. Notably, primary kinetic isotope effect (KIE) was not observed for the hydrogens of the aniline group of D-1a (step d) (Figure 7C, Figure S1).

Display Full Size

Figure 7. Mechanistic investigation. CPA: chiral phosphoric acid.

Besides, the analysis of the effect of para substituents on the aniline group on the reaction rate reveals a negative linear Hammett correlation (ρ = -2.4349 ± 0.2284, Figure S2). Similarly, the analysis of the effect of substituents on indoles on the reaction rate also shows a negative linear Hammett correlation (ρ = -1.9203 ± 0.2117, Figure S3). These results are consistent with a concerted process.

4. Conclusion

In summary, we have developed an efficient and modular asymmetric dearomative formal [4+2] reaction for synthesizing enantioenriched indoline-fused tetrahydroquinolines. This method employs O-silylated hemiaminals as precursors to in situ generate highly reactive methaniminium salts in the presence of CPA, which then catalyze the sequential dearomative cycloaddition of methaniminium salts and diverse substituted indoles under mild conditions. The reaction tolerates a wide range of functional groups with high efficiency (up to 97% yield and 99% ee). The PhNHCO- group at the indole moiety plays a crucial role in both reactivity and enantioselectivity, probably through non-covalent interactions with CPA. In addition, a concerted mechanism was implied by the analysis of KIE experiments and Hammett correlation.

Supplementary materials

The supplementary material for this article is available at: Supplementary materials.

Authors contribution

Guan Z: Conceptualization, supervision, formal analysis, writing-original draft.

Zhao Q, Ren Z: Formal analysis, writing-original draft, writing-review & editing.

Mo N, Zhang Y: Methodology, investigation.

Conflicts of interest

The authors declare no conflicts of interest.

Ethical approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

The data reported in this paper are available in the main text or Supplementary Information, including methods, NMR data, HRMS data, HPLC spectra Crystal data and NMR spectra. Crystallographic data for the structures reported in this article have been deposited at the Cambridge Crystallographic Data Centre, under deposition numbers CCDC 2189145 (5i).

Funding

This work was supported by the financial support from the National Natural Science Foundation of China (Grant Nos. NSFC 22171224, 21971204, 21602172), the Fund of Education Department of Shaanxi Provincial Government (Grant No. 24JP187), and the Medical Science and Technology Foundation of Guangdong Province (Grant No. A2025079).

Copyright

© The Author(s) 2025.