Asymmetric Synthesis of Tetrahydroisoquinoline Derivatives through 1,3-Dipolar Cycloaddition of C,N-Cyclic Azomethine Imines with Allyl Alkyl Ketones

A [3 + 2] 1,3-Dipolar cycloaddition of C,N-cyclic azomethine imines with allyl alkyl ketones has been achieved. The reaction proceeds under mild conditions and tolerates a wide range of functional groups. An array of tetrahydroisoquinoline derivatives is generally constructed with good diastereoselectivities and enantioselectivities (up to >25:1 dr, >95% ee). Moreover, the absolute configuration of the product was previously determined by using the quantum electronic circular dichroism calculation and ECD spectrum method.


Introduction
A variety of isoquinoline alkaloids [1][2][3] exist in many natural products and drugs, and have a broad range of clinical applications, exhibiting a broad range of biological activities such as antitumor, anti-HIV, antibiotic, antifungal, antivirus, anti-inflammatory, anticoagulation, and bronchodilation, and can also act on the central nervous system [4][5][6][7][8].

Scheme 1.
Previous reports and our protocol.

Results
In our initial attempt, we first examined the reaction of N,N-cyclic azomethine imine 1 with deconjugated 3-enone 5a in the presence of DPEN catalyst C1 (Table 1) in CHCl3, but no desired product was observed even at a higher temperature. We then turned our attention to C,N-cyclic azomethine imines 2, 3 and 4a. Unfortunately, no matter what the temperature of the reaction raised from rt to reflux, no desired cycloaddition adduct was observed when we employed the 2 and 3 dipoles. To our delight, the reaction between C,N-cyclic azomethine imine 4a and deconjugated 3-enone 5a proceeded smoothly to give the desired product in high yield (85% yield, 62% ee and dr 3:2, entry 1, Table 1) and the reaction could not afford the desired product if no acid or no catalyst were added to the Scheme 1. Previous reports and our protocol.

Results
In our initial attempt, we first examined the reaction of N,N-cyclic azomethine imine 1 with deconjugated 3-enone 5a in the presence of DPEN catalyst C1 (Table 1) in CHCl 3 , but no desired product was observed even at a higher temperature. We then turned our attention to C,N-cyclic azomethine imines 2, 3 and 4a. Unfortunately, no matter what the temperature of the reaction raised from rt to reflux, no desired cycloaddition adduct was observed when we employed the 2 and 3 dipoles. To our delight, the reaction between C,N-cyclic azomethine imine 4a and deconjugated 3-enone 5a proceeded smoothly to give the desired product in high yield (85% yield, 62% ee and dr 3:2, entry 1, Table 1) and the reaction could not afford the desired product if no acid or no catalyst were added to the reaction [18] (entries 2 and 3). The results indicated that catalysts with acids are critical for the cycloaddition reaction. In addition, the β,γ-C=C bond could act as an inducing group for the formation of more stable extended dienamine species from deconjugated 3-enone substrates 5a and activating the γ-site and furnishing the following vinylogous 1,3-DCs process. reaction [18] (entries 2 and 3). The results indicated that catalysts with acids are critical for the cycloaddition reaction. In addition, the β,γ-C=C bond could act as an inducing group for the formation of more stable extended dienamine species from deconjugated 3-enone substrates 5a and activating the γ-site and furnishing the following vinylogous 1,3-DCs process. We then turned our attention towards the use of chiral primary amine catalyst to improve the reaction enantioselectivity and diastereoselectivity. The reaction was tested in the presence of a number of primary amine catalysts (entries 4-10). The corresponding product 6aa was then isolated in high yield (89%), good enantioselectivity (80% ee), and with good diastereoselectivity (dr 12.5:1) through C2 catalyst in CHCl 3 at room temperature (entry 4). In contrast, the reaction afforded poor to moderate enantioselectivity (11-76%) when used with other catalysts (entries 5-11). The chiral prolinol silyl ethers as the catalyst could not produce the product (entry 12). A screening of the solvents showed that performing the reaction to improve the ee and the dr (entry [13][14][15][16][17][18][19] with C2 as the catalyst. The reaction proceeded with higher yield (91%) and good enantioselectivity (80%) and excellent diastereoselectivity (dr 25:1) in DCE at room temperature (entry 14). However, only poor or moderate diastereoselectivities were observed in other solvents (entry 13,[15][16][17][18][19]. Furthermore, the effect of acid additives was also studied. The presence of 20 mol% o-fluorobenzoic acid (OFBA) enhanced the ee value to 84% and the yield also increased to 92% (entry 20). A screening of the temperature showed that performing the reaction at 0 • C reduced the ee value and the yield (entry 20). The yield and stereoselectivity could not be further improved when the reaction used the other acid additives (entries 21-23). On the basis of the above-mentioned results, the reaction conditions were established to 1.0 equivalent of 4, 2.0 equivalents of 5, 10 mol% of C2 and 20 mol% of OFBA in DCE at room temperature for 12 h.

Discussion
With the optimized conditions in hand, the generality of the reaction was evaluated. A range of substrates was shown to be compatible with the developed protocol for the [3 + 2] cycloaddition of C,N-cyclic azomethine imines to allyl alkyl ketone by using C2 as catalyst. As summarized in Figure 2 (Figures S2 and S3 Supplementary Material, respectively), not only aromatic alkyl allyl ketones but also long aliphatic chain allyl ketones could all be employed successfully to afford the products 6aa-ah in high yields (88-96%), moderate to high enantioselectivities (50-96%), and high diastereoselectivities (dr 10:1 to >25:1) From these results, we found that the conditions were applicable to a wide variety of allyl alkyl ketones. With this promising result in hand, we then investigated the generality of the C,N-cyclic azomethine imines. The influence of the substituent of the benzoyl group on the nitrogen was examined. The benzoyl group bearing electron-withdrawing or -donating groups at the para position and meta position afforded the corresponding products in high yields and steroselectivities (6ba to 6ga). We found that some electron-withdrawing groups (−NO 2 , −Cl, −Br) on the aromatic moiety were more effective than some donating groups (−CH 3 , −OCH 3 , 3,5-Me 2 ) to afford the products in yields and steroselectivities, respectively (6aa to 6ga). From these results, it was determined that the electron density of the N'-acyl moiety played an important role in trapping the β-position of the carbon−carbon double bond intermediate for the 1,3-DCs. This study prompted us to examine the influence of structurally different N'-acyl moiety azomethine imines. The furoyl and naphthoyl groups on the nitrogen were also tolerated, giving the desired products in high yields, moderate enantioselectivities, and high diastereoselectivities, respectively (6ha 95% yield, 60% ee, dr > 25:1 and 6ia 93% yield, 71% ee, dr > 25:1). An increase in the steric bulk of the N'-acyl moiety was also tolerated to afford the product 6ja in high yields (90%) with 72% ee value and dr 10:1. As we failed to obtain single crystals suitable for X-ray crystallographic analysis to determine the absolute configuration of the products, the electronic circular dichroism (ECD) spectrum of chiral product 6ba was recorded in methanol and compared with the theoretically calculated results [37][38][39]. As depicted in Figure 3, the experimental ECD spectrum matched quite well to the calculated one of (R,S)-6ba. Therefore, the stereogenic center of product 6ba is probably in the (R,S) configuration. As we failed to obtain single crystals suitable for X-ray crystallographic analysis to determine the absolute configuration of the products, the electronic circular dichroism (ECD) spectrum of chiral product 6ba was recorded in methanol and compared with the theoretically calculated results [37][38][39]. As depicted in Figure 3, the experimental ECD spectrum matched quite well to the calculated one of (R,S)-6ba. Therefore, the stereogenic center of product 6ba is probably in the (R,S) configuration. As shown in Scheme 2, we proposed a plausible catalytic cycle to explain the reaction mechanism. The condensation of a chiral primary amine catalyst with allyl phenylethyl ketone 5a would lead to the formation of the iminium-ion, which could form the dienamine intermediate A [40][41][42].

Materials and Methods
NMR data were obtained for 1 H at 400 MHz or 600 MHz, and for 13 C at 100 MHz or 150 MHz. The data are presented as follows: chemical shift was reported in ppm from tetramethylsilane with the solvent resonance as the internal standard in CDCl3 solution, integration, multiplicity (br = broad, s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet) and coupling constant in Hertz (Hz). ESI HRMS was recorded on a Waters SYNAPT G2. In each case, enantiomeric excesses (ee) were determined by chiral high- As shown in Scheme 2, we proposed a plausible catalytic cycle to explain the reaction mechanism. The condensation of a chiral primary amine catalyst with allyl phenylethyl ketone 5a would lead to the formation of the iminium-ion, which could form the dienamine intermediate A [40][41][42]. As shown in Scheme 2, we proposed a plausible catalytic cycle to explain the reaction mechanism. The condensation of a chiral primary amine catalyst with allyl phenylethyl ketone 5a would lead to the formation of the iminium-ion, which could form the dienamine intermediate A [40][41][42]

Materials and Methods
NMR data were obtained for 1 H at 400 MHz or 600 MHz, and for 13 C at 100 MHz or 150 MHz. The data are presented as follows: chemical shift was reported in ppm from tetramethylsilane with the solvent resonance as the internal standard in CDCl3 solution, integration, multiplicity (br = broad, s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet) and coupling constant in Hertz (Hz). ESI HRMS was recorded on a Waters SYNAPT G2. In each case, enantiomeric excesses (ee) were determined by chiral high-Scheme 2. Plausible mechanism.

Materials and Methods
NMR data were obtained for 1 H at 400 MHz or 600 MHz, and for 13 C at 100 MHz or 150 MHz. The data are presented as follows: chemical shift was reported in ppm from tetramethylsilane with the solvent resonance as the internal standard in CDCl3 solution, integration, multiplicity (br = broad, s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet) and coupling constant in Hertz (Hz). ESI HRMS was recorded on a Waters SYNAPT G2. In each case, enantiomeric excesses (ee) were determined by chiral highperformance liquid chromatography (chiral HPLC) that were Daicel Chiralpak AD-H Column (250 × 4.6 mm), Chiralpak IA Column (250 × 4.6 mm), or Chiralpak IC Column (250 × 4.6 mm). UV detection was monitored at 220 nm or 254 nm. Optical rotation was measured in CHCl3 solution at 20 • C. Column chromatography was performed on silica gel (200-300 mesh) eluting with ethyl acetate and petroleum ether. TLC was performed on glass-backed silica plates. UV light, I 2 , and solution of potassium permanganate were used to visualize products. All chemicals were used without purification as commercially available unless otherwise noted. Petroleum ether (PE) and ethyl acetate (EtOAc) were distilled. THF was freshly distilled from sodium/benzophenone before use. Experiments involving moisture and/or air sensitive components were performed under a positive pressure of argon in oven-dried glassware equipped with a rubber septum inlet. Dried solvents and liquid reagents were transferred by oven-dried syringes.

Conclusions
In conclusion, we have developed the first chiral primary amine-catalyzed enantioselective [3 + 2] 1,3-dipolar cycloaddition of allyl alkyl ketones with C,N-cyclic azomethine imines via induced dienamine catalysts to give a novel class of dinitrogen-fused heterocycles combining the biologically important tetrahydroisoquinoline core and pyrazolidine core. The reaction affords a tetrahydroisoquinoline derivative in high yield with moderate to high enantioselectivities (up to 96%) and high diastereoselectivities. In addition, this reaction provides an efficient method for constructing diverse and complex chiral tetrahydroisoquinolines compounds. Research into further applications of this enantioselective 1,3-dipolar cycloaddition with C,N-cyclic azomethine imines is in progress.

Supplementary Materials:
The following are available online. Figure S1: General procedure for the preparation of the C,N-cyclic azomethines imines and its analogs, Figure S2: General procedure for catalytic asymmetric 1,3-DCs, Figure S3: Copies of NMR spectra and HPLC spectra.
Author Contributions: G.F., G.M., W.C. and S.X. participated in the synthesis, purification and characterization of the new compound. K.W. and S.W. participated in the interpretation of spectroscopy of the new compounds and the review of the manuscript. K.W. and S.W. participated in the interpretation of the results, writing, revision and correspondence to the Journal of Molecules until the manuscript was accepted. All authors have read and agreed to the published version of the manuscript.