Undesired stereoisomer

Schemem 4.29 Lipase-catalyzed purification by removal of undesired stereoisomers from the product. Chirazyme L2 (CALB), vinyl acetate, n-heptane, mol. sieves , 25°C.

Enzymes can also be used to eliminate an undesired stereoisomer. The enzyme will process only one isomer in a racemic mixture and leave the other stereoisomer untouched. 

The carbometallation of alkynes is also unique in its stereochemical pattern, since it is executed as an exclusive cis addition. For example, the stereoisomeric purity of pheromone 104, prepared as shown, is higher than 99.9% This feature is of special importance in the synthesis of natural pheromones, as the biological activity of these compounds is dramatically affected by the presence of an even negligible amount (less than 0.5% ) of the undesirable stereoisomer. Not surprisingly, this reaction has found numerous areas of application, especially in the stereospecific synthesis of the tri- and tetrasubstituted alkenes, a goal difficult to achieve by other methods. To illustrate the effectiveness of this approach, the synthesis of faranal 105, the trail-marking pheromone of the ant species Monomoriumpharaonis, is shown in Scheme 2.33. 

From an industrial perspective the most elegant chemistry is the most economical chemistry. Synthesis through resolutions are less glamorous than those in which the stereochemistry is set through asymmetric synthesis nonetheless, resolutions have been shown to be more economically effective in some cases [8,45, 60]. With asymmetric synthesis the desired stereochemistry must be efficiently introduced. If it is not, the undesired stereoisomers may complicate isolations and reduce the yield of the desired product. Many of the features of a route that need to be considered were mentioned in Chapter 2. Comparing product costs through spreadsheets may be extremely valuable [61]. 

Many pharmaceutical products are specific stereoisomers, and in many cases the inactive enantiomer must be removed to minimize side effects. In recent years, advances in selective stereochemistry have reduced the need for isomer separation. However, many synthetic processes continue to have racemic output (desired and undesired stereoisomers). 

Subsequently, the enantioselective route was replaced by a resolution process (33—>rac-34—>rac-36—>rac-31—>(S,S,i )-31—>32) [26]. This synthesis proved superior due to the reduced number of steps, although the undesired stereoisomer R,R,S)-3l could not be recycled. Ecological considerations led to the investigation of another enantioselective approach which is based on the hydrogenation of a-pyrone 35 to dihydropyrone (i )-34 [27]. Ee-values up to 96% were achieved with a cationic Ru catalyst derived from the electron-rich and sterically bulky f-Bu-MeOBIPHEP Hgand. The relatively low TON of 1,000 calls for further catalyst improvement in this hitherto unprecedented type of reduction. 

Dihydropyridinium salt 31 was initially selected to check the viability of the proposal. This salt was prepared from tetrahydropyridine 30, which was accessible in six steps from methyl nicotinate, by treatment of the corresponding N-oxide derivative (mixture of two diastereoisomers) under Polonovski-Potier reaction conditions (Scheme 11). Reaction of the crude salt 31 with the sodium salt of diethyl 1,3-acetonedicarboxylate gave a mixture of two diastereoisomeric adducts 32 (two isolable enol forms) and 33 in 87% overall yield, the undesired stereoisomers 32 (axial amino-methyl chain) predominating (69 31 ratio). 

Note that if a reagent selects for the undesired stereoisomer, as in the conversion of 72 73, this can sometimes be corrected by inverting config-... 

The next stage of the synthesis required reduction of the Cj-Cs double bond with control over stereochemistry at Cs- The tactics ultimately used to accomplish this transformation involved conjugate addition of thiophenoxide to the enone to provide 58 with Cj stereochemistry that was never established. The critical stereochemistry (Cs), however, was clean and presmnably controlled by kinetic protonation of the intermediate enolate. Reduction of the C9 ketone was followed by esterification to provide acetate 59 as a single stereoisomer (C7 stereochemistry still not defined). Reduction of the C7 thiol was followed by excision of the extra carbon in the usual manner to provide aldehyde 60. The final carbons of the seco- dA were introduced via crossed condensation of the enolate derived from a thioester of propionic acid, with aldehyde 60. This reaction provided the proper stereochemistry at C3, but the undesired stereoisomer at C2. The C2 stereochemistry was corrected by kinetic protonation of the enolate derived from 61 with acetic acid. The structure of the resulting seco-zcid derivative (62) was established by X-ray crystallography. 

The protection of the hemiacetal hydroxyl in Step B-2 was followed by a purification of the dominant stereoisomer. In Step C-l, the addition of the C(6) methyl group gave predominantly the undesired a-stereoisomer. The enolate was trapped as the trimethylsilyl ether and oxidized to the enone by Pd(OAc)2. The enone from sequence C was then subjected to a Wittig reaction. As in several of the other syntheses, the hydrogenation in Step D-2 was used to establish the configuration at C(4) and C(6). 

Even this brief list may suffice to show that it would be a formidable task to develop a system of factorization free of avoidable redundancies, and that such a system would not be satisfactory even if it avoids arbitrary choices. It would require a rule disqualifying certain centers or lines of stereoisomerism on the basis of their relationships to other potential elements in the same molecule. Such definitions would not be self-contained. Moreover, the products of factorization that would take the place of those dropped cannot be limited to points or lines that are merely differently defined. There would have to be a virtually open-ended proliferation of new elements. This highly undesirable feature would not be offset by a major benefit of the revised system such as a correlation between the numbers of elements and of stereoisomers, because a complete elimination of all redundancies does not seem possible. We conclude that the system of choice is the one based on the principle that the elements of stereoisomerism allow no further factoring. Accordingly we think it best to retain the definitions given in Sects. IV and VI and their strictures that all centers and lines be occupied by atoms or bonds. 

Isomerizations may take place over some catalysts more than over others. Regio- as well as stereoisomers may be formed [13, 49. Such side reactions are undesirable in the reduction of pro-chiral alkenes, and especially in the hydrogenation of some cycloalkenes. 

However, the isopulegol mixture can also be hydrogenated to produce a mixture of menthols the individual stereoisomers are then separated by distillation. To obtain optically pure (-)-menthol, a resolution step involving a suitable crystalline derivative (such as the benzoate) is required. The undesired stereoisomeric menthols mainly (+)-neomenthol and (+)-isomenthol, are epimerized to an equilibrium mixture (e.g., by heating in the presence of sodium menthylate). (-)-Menthol is then again separated from the mixture. 

Korte and coworkers15,5 and H. Newman86 did not separate their mixtures of stereoisomeric alkyl 2,3-anhydro-4,6-dideoxy-DL-hexopyrano-sides (248) in which the undesired a-lyxo stereoisomer was certainly the preponderant component. In consequence, racemic alkyl desosa-minide was obtained only as a minor product in admixture with alkyl 3,4,6-trideoxy-3-(dimethylamino)-DL-flr / ino-hexopyranoside (250),... 

Hydrolase-catalyzed acylation can be used to purify a diastereo- and enantiomerically enriched product. For example dimethylzinc addition to the racemic aldehyde 77 furnishes the racemic phenylsulfanylbutanol 78 (Scheme 4.29) in a 95/5 (2R, 3R )/(2R, 3S )-mtio. When this is treated with Chirazyme L2 (CALB) and vinyl acetate in heptane it is resolved with a high E-value (>400) [91]. However the diastereomeric ratio in the remaining substrate and produced ester is virtually unchanged. To circumvent the problematic contamination with the undesired diastereomers, enantiomerically enriched aldehyde 77 was reacted with dimethyl-zinc to furnish one major stereoisomer of 78 contaminated with a small amount of a mixture of the other three (Scheme 4.29). Because the two major contaminants had the opposite configuration at position 2 relative to the major product, these contaminants were efficiently removed from the major product and the trace byproduct by treatment with the 2R-selective Chirazyme L2 (CALB) and vinyl acetate in heptane to furnish virtually diastereo- and enantiomerically pure acetate (2R,3R)-79 or the alcohol (2S,3S)-78 (Scheme 4.29) [91]. 

Thus, the need has arisen for larger amounts of unnatural amino acids—those that contain unusual side chains or are in the D-series. Because a-amino acids contain an epimerizable center, D-amino acids are usually accessible through epimerization of the natural isomer, followed by a resolution. Of course, the racemic mixture can also be accessed by synthesis. Because resolution can be either wasteful—if the undesired isomer is discarded—or clumsy—when the other isomer is recycled through an epimerization protocol—many large-scale methods now rely on a dynamic resolution, where all of the starting material is converted to the desired isomer (vide infra, Chapters 6 and 7). With the advent of asymmetric reactions that can be performed at large scale, a substrate can now be converted to the required stereoisomer without the need for any extra steps associated with a resolution approach. 

---