Asymmetric Synthesis of 1,3-Oxazolidine Derivatives with Multi-Component Reaction and Research of Kinetic Resolution

An efficient multi-component reaction to synthesize multi-substituted 1,3-oxazolidine compounds of high optical purity was described. All the products were well-characterized and the absolute configuration of one chiral center was determined. The plausible mechanism was proposed and a kinetic resolution of epoxides process was confirmed.

The reaction of imine with aldehyde is a well-established one, for example, Ishii group described a method to synthesize 1,3-oxazolidines from a multi-step synthesis of imines and epoxides [13]. However, the reaction yields were relatively low and the stereochemistry was not explained. Herein, we describe an efficient multi-component reaction to access 1,3-oxazolidine compounds of high optical purity. The plausible mechanism of the reaction was suggested and a kinetic resolution of the epoxides process was confirmed. All of products were well-characterized and the absolute configuration of the chiral center from the epoxides part was determined. To the best of our knowledge [54,55], this is the first report of synthesis of 1,3-oxazolidine with a multi-component reaction via the kinetic resolution from epoxides, anilines, and ethyl glyoxalate.
The reaction conditions optimization of three-component reactions was carried out, a variety of different chiral Lewis acid (4a-4d) catalysts were screened ( Figure 1). First we examined the reaction of aniline 1a, ethylglyoxalate, and epoxide 2a in dichloromethane at 18 °C in the presence of catalyst 4a (10 mol %), while the reaction gave the product in only 5% yield with poor diastereoselectivity and enantioselectivity after two days ( Table 1, entry 1). Little higher yield and ee value were achieved when catalyst 4b was applied ( Table 1, entry 2). With this encouragement, the product with a higher ee was obtained when the reaction was carried out at −10 °C (Table 1, entry 3). Catalyst 4c/Ti(IV), which was prepared by mixing 4c and Ti(O-i-Pr)4 in 1:1 molar ratio in dichloromethane for 2 h, afforded a moderate ee of 41% (Table 1, entry 4). On the contrary, when the reaction was carried out with catalyst 4c/Ti(IV), which was prepared in a 1:1 molar ratio in toluene, good diastereoselectivity and enantioselectivity trifluoroacetic acid (TFA) (0.5 mol %) was added, the product with a good yield was obtained (Table 1 were provided (Table 1, entry 5). To our delight, after a small amount of, entry 6). Therefore, it is indicated that TFA was very important for the cyclization. TFA might be beneficial for the ring-opening of epoxide. However, when TFA was used as the only catalyst, the reaction rate was not accelerated, and completion of the reaction also needed a long time at room temperature. Catalyst 4c/Ti(IV) was prepared in a 2:1 molar ratio in toluene, affording better ee value (Table 1, entry 7). Slightly lower yield was observed for the reaction carried out with the catalyst 4d/Ti(IV), which was prepared in a 2:1 molar ratio in toluene ( Table 1, entry 8). The influence of temperature to the reaction was also investigated ( Table 1, entries 9 and 10), the decrease of temperature were negative to the yields. It could be concluded that the molar ratio of 4c and Ti(IV) influenced the enantioselectivity and TFA strongly benefited the yields.
Theoretically, the products had C-5 isomer and C-4 isomer, but the C-4 isomer was only separated and actually detected when different reaction scales from 1 mmol to 10 mmol were conducted ( Figure 2). The C-5 isomer was not found. Furthermore, NMR spectra, such as 1 H-1 H NOESY, and HMBC could prove that the structure of the desired product belongs to the C-4 isomer. The H11 had the chemical shift value of 4.56 ppm because it was adjacent to an oxygen atom. It had heteronuclear coupling with C10 and C12. Accordingly, C6 had heteronuclear coupling with H10 and H11 (see Supporting Information).  and catalyst (0.1 mmol) were added, and the system was stirred for 4 days; b Yields of isolated products; c,d were determined by HPLC on a chiral column; e Catalysts were prepared by Ti(IV) and 4c ligand in a 1:1 molar ratio and TFA was added into the reactions as a catalyst; f Catalysts were prepared by Ti(IV) and 4c, 4d ligands in a 1:2 molar ratio and TFA was added into the reactions as a catalyst.
The three-component reactions of anilines, ethyl glyoxalate, and epoxides were expanded under the optimized conditions (Table 2). A series of chiral 1,3-oxazolidine derivatives in moderate yields with good diastereoselectivities and enantioselectivities was provided, such as 4-chloroaniline, affording good ee and diastereoselectivity (Table 2, entries 2, 8, and 9). Product 3h was obtained with a good enantioselectivity ( Table 2, entry 8). Product 3i was obtained with the highest ee (up to 90%) ( Table 2, entry 9). From entries 1, 2, 3, and 11 in Table 2, we found that the ee changed with group R 1 . The electron-withdrawing group was negative for the reaction, such as entry 11 in Table 2. Therefore, substrates bearing electron-donating substituents were advantageous to the reaction. We speculated the reaction is a kinetic resolution as the yields are around 50%. Thus, the reaction of entry 10 ( Table 2) also had been carried out as controlled experiment. Racemic (±)-epoxystyrene was used as substrate to react with ethyl glyoxalate and anisidine under the condition listed in Table 2. When the reaction was finished, unreacted epoxystyrene was subjected into a chiral column on the HPLC. To our delight, the epoxystyrene recollected from the reaction is no longer racemic, according to the HPLC analysis. Instead of the substrate racemic (±)-epoxystyrene, unreacted epoxystyrene with 68% ee was obtained. This means one enantiomer of the racemic epoxystyrene was consumed and the other was kept and the reaction is indeed a kinetic resolution (see Supporting Information).
To understand more about the reaction, effort also has been made to determine the absolute configuration of the products. Firstly, we figured out the retention time of (R)-epoxystyrene and (S)-epoxystyrene by detecting the racemic (±)-epoxystyrene and a standard (R)-epoxystyrene. Then, we reclaimed unreacted epoxystyrene from the reaction of entry 10 in Table 2, by checking the retention time. We found (S)-epoxystyrene was consumed and (R)-epoxystyrene not. This means that the absolute configuration of the chiral center from epoxides in the product is S-configuration [10]. (see Supporting Information).
On the basis of previous studies and the results of our experiments [13,[42][43][44][45][46][47][48][49][50][51][52][53], a possible reaction mechanism of the present reaction is proposed. Firstly, (S)-BINOL reacted with Ti(O-i-Pr)4 for two hours to give a complex A. When racemic epoxide B was added, a transition state C was formed by the complexation of A and (S)-epoxide with the influence of TFA. Immediately, C reacted with the imine from aniline and ethyl glyoxalate to give the target product D, releasing out complexes A to participate the next cycle at the same time ( Figure 3).  Figure 3. Proposed reaction mechanism.

General Procedure for the Synthesis of All 1,3-Oxazolidines
Ti(O-i-Pr)4 (0.05 mmol) and chiral binaphthalene ligand (4c, 0.10 mmol) were dissolved in 2.0 mL toluene, and the mixture was stirred for 2 h at room temperature, then aniline (1.1 mmol) and ethyl glyoxalate (1.0 mmol) were added into the mixture, and the result system was stirred for 30 min. Finally epoxide (1.2 mmol) and TFA (0.5 mol %) were added into the system and were stirred at −40 °C for four days. Then, the solvent was evaporated under vacuum. The residue was purified by silica gel column chromatography using 1:5 ethyl acetate/petroleum ether as eluent, giving a light yellow liquid. Enantiomeric excess (ee) were determined by HPLC analysis on a L-7420 (UV-VIS Detector with an L-7110 pump and a Chiralcel OD-H column). We determined the retention time of the product is based on significant changes in HPLC on a chiral column.   1837 (m, 3H[3Hʹ]). 13

Conclusions
In conclusion, we disclosed an efficient asymmetric three-component reaction of anilines, ethyl glyoxalates, and epoxides, yielding 1,3-oxazolidine derivatives with high diastereo and enantioselectivities (up to 20:1 d.r., 90% ee) by cascade process. This reaction provides a convenient method to synthesize multi-substituted 1,3-oxazolidine compounds of high optical purity. The plausible mechanism was suggested and a kinetic resolution process was confirmed. All of the products in this paper were well-characterized and the absolute configuration of the chiral center from epoxides was determined. We believe this study will enrich the methodologies for the synthesis of five-membered oxacycles and natural products. Further studies are underway and other synthetic applications will be reported in a due time.