Pyrrolidine acidity

The initial asymmetric synthesis (see Scheme 5.2) of pyrrolidine acid 3 suffered from a chiral HPLC bottleneck. As a result, chiral salt resolution was investigated. The rapid discovery of a crystalline di-p-toluoyl-D-tartaric acid salt provided the necessary means to resolve and purify the desired diastereomer. Using 0.65 equivalent of the acid in methyl (cr(-butyl ether (MTBE), the crystallized salt was shown to be a 92 8 ratio of (3S,4R) (3R,4S) diastereomers. The resolved tartaric acid salts were then recrystalhzed from n-butanol to ratios of >99 1 in a 42% overall yield on a 2-kg scale. Further improvements were made in the preparation of the azomethine ylide precursor 38. In step 1, using dimethyl sulfoxide (DMSO) as the solvent, the reaction temperature of trimethylsilylmethylation of tert-butylamine was lowered from 200°C used in the original synthesis to 80°C. In step 2, substituting n-butanol in place of methanol reduced the amount of oligomerization observed and increased the yield to 69%. Overall, these improvements allowed for the preparation of pyrrolidine acid 3 in 22% overall yield in 99% ee from cinnamate 39 (Scheme... 

SCHEME 5.17 First-generation process pyrrolidine acid synthesis. 

Our initial improvement in the synthesis of pyrrolidine acid 3 relied on a racemic 1,3 dipolar cycloaddition followed by resolution. Attempts to devise asymmetric protocols of this reaction using chiral auxiliaries were not productive. The results from our laboratories were consistent with literature findings, with a moderate diastereoselectivity of 3 to 4 1 at best obtained even when double chiral auxiliaries were used. Several other approaches, such as Aza-Cope/Mannich reaction, intramolecular C-H insertion, and asymmetric aryl 1,4 addition, did not bear fruit. 

Initially, the 80 20 transxis mixture of pyrrolidine nitrile 50 was converted to a mixture of -butyl esters under snlfnric acid catalysis, followed by epimerization with n-BuONa, and then hydrolysis to the acid (Scheme 5.27). The best trans. cis ratio of n-bntyl esters achieved was 95 5. Hydrolysis of esters with HCl afforded the HCl salt of 3 in 89% overall yield, which led to a minor upgrade in the transxis ratio (97 3). On the other hand, hydrolysis of the n-bntyl ester by NaOH and snbseqnent pH adjnstment to 6.5 afforded pyrrolidine acid 3 in 86% yield and >99.9 chemical and optical parities. 

The basic hydrolysis of nitrile 50 with aqneons NaOH in ethanol was examined, which proceeded throngh intermediate amides 57, and reached 98% conversion to acid 3 within 4 hr with <1% of amides remaining. Both cis- and trani-amides were observed by liqnid chromatogra-phy/mass spectrometry (LC/MS) during reaction, and the strnctnre of trans-amide Sit was confirmed by LC/MS and independent synthesis (by treatment of trani-pyrrolidine acid with CDI and ammoninm hydroxide). It is reasonable to postnlate that the hydrolysis of both di-nitrile 50 and di-amide Sic to the corresponding di-acids was slow relative to epimerization, which provided a mechanism for complete conversion of di-nitrile 50 to trani-acid 3. 

This procednre prodnced crude pyrrolidine acid 3 that was >99.6% trans with an optical purity of 99.5% ee, indicating that there was no chirality leakage in the cyclization-hydrolysis-epimer-ization process. This 3 was isolated by crystallization from IPA/MTBE in 95% yield with 99.97 LCAP purity and >99.9% ee on multi-kg scale. Thns, an efficient direct one-step basic epimeriza-tion-hydrolysis of the pyrrolidine nitrile mixtnre to the trani-pyrrolidine acid 3 was achieved (Scheme 5.28). 

SCHEME 5.28 One-step conversion of trans cis mixture pyrrolidine nitrile 50 to rra -pyrrolidine acid 3. Source-. Chung, John Y.L. et al. The Journal of Organic Chemistry, Vol. 70(9), 3592-3601, ACS, 2005. 

Ji, C., Peng, Y, Huang, C., Wang, N. and Jiang, Y, The influence of acidity on direct aldol reactions catalyzed by pyrrolidine/acid bifunctional organocatalyst, Synlett, 2005,986-990. 

---