Dynamic kinetic resolution of ( S )-mandelate-derived α-bromo esters in nucleophilic substitution and asymmetric syntheses of 3-substituted morpholin-2-ones

Dynamic kinetic resolution of (S)-mandelate-derived α-bromo esters in nucleophilic substitution reaction has been investigated. Substitutions with various alkyl amine nucleophiles in the presence of TBAI and DIEA can provide various α-amino esters up to 81% yield and 97:3 dr. Also, the substitution of α-bromo esters with N-substituted 2-aminoethanol nucleophiles and following spontaneous cyclization provides a practical protocol for asymmetric syntheses of 3substituted morpholin-2-ones up to 95:5 er.


Introduction
Dynamic kinetic resolution of chiral alcohol-derived α-halo esters has been recently recognized as an effective synthetic method for asymmetric syntheses of α-heteroatom substituted carboxylic acid derivatives. 1-3While the chiral auxiliaries can achieve a useful level of stereoselectivity, it is still desirable to find novel ways to utilize the methodology for practical asymmetric syntheses.We have previously reported (S)-mandelate-mediated dynamic kinetic resolution of α-bromo esters with various aryl amines for asymmetric syntheses of dihydroquinoxalinones and dihydrobenzoxazinones. 4Herein we report our recent results on the asymmetric nucleophilic substitution of (S)-mandelate-derived α-bromo esters with various alkyl amines for practical asymmetric syntheses of morpholin-2-ones.

Results and Discussion
Initial studies on (S)-mandelate-mediated dynamic kinetic resolution were carried out with αbromo phenylacetate 1 and dibenzylamine (Bn 2 NH).When the diastereomeric mixture (1:1) of (αRS)-1 was treated with tetrabutylammonium iodide (TBAI, 1.0 equiv), diisopropylethylamine (DIEA, 1.0 equiv) and dibenzylamine (1.5 equiv) in CH 2 Cl 2 at room temperature for 12 h, the amino acid derivative (αR)-2 was produced in 81% yield with 97:3 diastereomeric ratio (dr) as shown in Scheme 1. Subsequent reductive cleavage of (αR)-2 using LiAlH 4 furnished the enantioenriched N,N-dibenzyl 2-aminoalcohol (R)-3 in 86% yield with 97:3 enantiomeric ratio (er).The results imply that the α-bromo stereogenic center is configurationally labile with respect to the rate of substitution and (αRS)-1 is dynamically resolved under the reaction condition.The scope of the observed dynamic kinetic resolution has been examined with α-bromo propionate 4 as shown in Table 1.When the diastereomeric mixture (1:1) of (αRS)-4 was treated with dibenzylamine (1.5 equiv), TBAI (1.0 equiv) and DIEA (1.0 equiv) for 24 h, the amino ester (αR)-11 was obtained in 62% yield with 90:10 dr.(entry 1) In the absence of DIEA, the rate of the substitution was substantially decreased to provide 11 in 38% yield after 24 h with the same selectivity (entry 2).The reaction of 4 in the absence of TBAI, however, gave 11 with a lower stereoselectivity.(entry 3) The substitution of 4 was very slow in the absence of both TBAI and DIEA to provide 11 in 19% yield with 86:14 dr.(entry 4).The results in entries 2-4 pointed to the importance of the presence of halide ion and base for sufficient rate acceleration and selectivity.The lack of an extraneous halide ion would inhibit the rapid epimerization of 4 thereby decreasing the stereoselection.As shown in entries 5-9, most of the solvents explored gave similar selectivities to give 11 with 90:10 dr in CH 3 CN, 87:13 dr in THF, 89:11 dr in dioxane, 87:13 dr in acetone and 88:12 dr in ethyl acetate.However, the selectivity was reduced in DMSO or in DMF (entries 10-11) and the reaction was very slow in n-hexane or in diethyl ether.Decreasing the reaction temperature reduced the rate of the reaction significantly and a slightly lower selectivity (88:12dr) of 11 was observed at 0°C.In an effort to improve the stereoselectivity, we examined the substitutions of α-bromo propionates with three different chiral auxiliaries derived from (S)-mandelic acid.No substantial difference has been found in the reaction of 5 derived from isopropyl mandelate.However, much lower stereoselectivities were noted for the substitution reactions of 6 and 7 derived from mandelamides.(entries 12-14) Also, a similar trend in stereoselectivity was observed with αbromo phenylacetates 8, 9 and 10. (entries 15-17) Next, we examined six different alkyl amine nucleophiles to evaluate the scope of the dynamic kinetic resolution as shown in Table 2.The treatment of α-bromo phenylacetate 1 with dibutylamine (1.5 equiv) in CH 2 Cl 2 for 24 h at room temperature gave 20 in 92% yield with 91:9 dr.(entry 1) Under the same reaction conditions, the reaction of α-bromo propionate 4 with the secondary amine gave N,N-dibutyl amino ester 21 with a slightly lower dr.(entry 2) In the reactions with diallylamine nucleophile, similar stereoselectivities were obtained in the reactions of α-bromo acetates 1 and 4. (entries 3-4) Notably, when cyclic secondary amine nucleophiles were used, the reactions gave much lower stereoselectivities (entries [5][6][7][8][9].Also, the reactions with two primary amines provided the products 29-32 with lower selectivities compared to the reactions of the corresponding α-bromo acetates with dibenzylamine (entries 10-13).Limited results in Table 2 indicate that both the size of amine nucleophiles and the nature of the αsubstituent of α-bromo acetates significantly affect the stereoselectivity of the nucleophilic substitution.

Conclusions
We conclude that dynamic kinetic resolution of (S)-mandelate-derived α-bromo esters in nucleophilic substitution with alkyl amines can be successfully applied towards the preparation of various enantioenriched amino acid derivatives.The results showed that stereoselectivity depends critically on the α-substituent of the α-bromo acetate and the structure of the amine nucleophiles.The substitution with N-substituted 2-aminoethanol nucleophiles and subsequent spontaneous cyclization can provide a general procedure for asymmetric syntheses of 3-substituted (R)-morpholin-2-ones.The mild and practical synthetic method for optically active amino acid derivatives suggests that this DKR approach should be further developed.

Experimental Section
General.All reactions were performed in oven-dried glassware under nitrogen atmosphere with freshly distilled solvents.Analytical thin layer chromatography (TLC) was performed on silica gel plates with QF-254 indicator and TLC visualization was carried out with UV-light.Flash column chromatography was performed with 230-400 mesh silica gel.Analytical chiral stationary phase HPLC was performed on pump system coupled to absorbance detector (254nm).Chiral stationary phase columns (25cm×4.6mmi.d.) with isopropanol/hexane mobile phase were used to determine enantiomeric ratios. 1 H and 13 C NMR spectra were acquired on U400 (400MHz 1 H, 100.6MHz 13 C) spectrometer using chloroform-d (CDCl 3 ) and purity of the compounds was established to be >95% based on 1 H and 13 C NMR spectra.
General procedure for the preparation of α-halo esters 1, 4-10, 18, 19 and 40.Mandelic acid derived ester or amide (1.0 equiv), racemic α-bromo acid (1.0 equiv), DCC (1.0 equiv) and DMAP (0.2 equiv) were dissolved in CH 2 Cl 2 and stirred at room temperature for 3-10 h.The precipitate was filtered off and the organic phase was washed with water.The organic phase was dried over MgSO 4 , filtered and concentrated to provide the crude product that was purified by column chromatography on silica gel.

Scheme 2 .
Scheme 2. Asymmetric synthesis of key intermediates of aprepitant.

Table 1 .
Substitutions under various reaction conditions