Chiral Mitsunobu reactions with (1S)-(+)-ketopinic acid: kinetic resolutions of secondary alcohols
Introduction
The resolution of a racemate1 followed by the inversion of chirality of one of the enantiomers is a general strategy of considerable interest as it may convert all of the racemate into one enantiomeric form. An attractive and intriguing variant of the above strategy is one that would combine the resolution and inversion steps into a single operation, as would obtain, for instance, when a kinetic resolution2 can be performed with concomitant inversion of chirality. We report herein the design of a chiral version of the Mitsunobu reaction, thereby demonstrating the above ideas within the class of secondary alcohol possessing a single stereogenic centre.
The Mitsunobu reaction is perhaps the most favoured method for the inversion of secondary alcohols.3., 4., 5. It involves the reaction of the alcohol 2 (Scheme 1) with a mixture of triphenylphosphine (TPP), diethyl azodicarboxylate (DEAD) and a carboxylic acid 1, (RCO2H). The reaction is believed to occur via the initial nucleophilic addition of TPP to DEAD to afford the betaine I (step 1); this is protonated by the carboxylic acid 1 to afford the azaphosphonium carboxylate II (step 2), the overall formation of II being irreversible. Nucleophilic attack at the tetracoordinate phosphorus atom of II by the alcohol 2 (step 3), results in a species formally considered to be III (an activated derivative of 2), which suffers nucleophilic attack by the carboxylate anion (of 1). The resulting inverted ester 3 is accompanied by triphenylphosphine oxide (TPPO) (step 4), the formation of which is the thermodynamic driving force for the overall reaction.
The design of a chiral version of the above process is an interesting exercise: among the many possibilities those involving chiral phosphines, chiral azodicarboxylate and chiral carboxylic acid as auxiliaries are obvious. Also, possible enantioselection in Scheme 1 may in principle be expected to occur at either of the steps 3 or 4 or both (which involve both the chiral auxiliary and the alcohol), with the conditions and caveats discussed below. (Note also that the slow step in the Mitsunobu reaction is believed to be step 4; however, the enantioselection may in principle occur in any step, as long as there is no effective racemisation in a further step.)
Enantioselection only at step 4 would require step 3 to be reversible (shown by a dotted arrow), which would allow the ‘wrong’ alcohol enantiomer to dissociate and be deactivated. (Otherwise both enantiomers would be inverted, although at different rates: an essentially null result.) There is apparently no evidence for the reversibility of step 3, but it cannot be ruled out: the equilibrium constant would be largely determined by the relative stability of a cationic phosphorus centre that is bound to an alkoxy oxygen atom or a hydrazinyl-carbamate nitrogen atom, and is not easily predicted. And as any equilibrium is expected to be relatively rapidly attained in view of the ease of nucleophilic substitution at phosphorus, it would allow the required selective dissociation of one of the alcohol enantiomers (mentioned above).
Also, enantioselection at both steps 3 and 4 may or may not be mutually reinforcing. In principle, enantioselection at step 3 is possible with chiral RCO2H 1, chiral azodicarboxylate or chiral phosphine, whereas that at step 4 is possible only with chiral RCO2H 1 or chiral phosphine (azodicarboxylate not being involved). Enantioselection with chiral RCO2H 1 at step 3 may arise from the (known) deprotonation of the alcohol 2 by the carboxylate anion, as also by its proximity (as counterion) to the cationic phosphorus atom (the site of selectivity).
The first chiral Mitsunobu reaction was reported by Kellogg and co-workers,6 who employed two chiral dioxaphosphepanes with a range of alcohols and acids. Although the overall enantiospecificity was apparently moderate (≤39%), the unreacted alcohols could be obtained in high e.e.s at less than complete conversion. (This seems to be due to an equimolar amount of the chiral phosphepane having been employed—half a molar equivalent being ideal for an efficient kinetic resolution—although a competitive racemisation of the chiral ester formed was also suspected.) We are not aware of any further reports of chiral Mitsunobu reactions.
Section snippets
Mitsunobu reactions with (1S)-ketopinic acid
The present work was motivated by the fact that chiral RCO2H is perhaps the simplest of the possible choices as auxiliary, and the availability of (1S)-(+)-ketopinic acid in hand. The secondary alcohols (4, Scheme 2) listed in Table 1 were treated with TPP, DEAD and (1S)-(+)-ketopinic acid 5 (0.5 equiv. each) in dichloromethane solution at −23°C. Upon the usual work-up, the reaction mixture was found to be a 1:1 mixture of the starting alcohol 4 and its ketopinate ester 6 (by IR and NMR), thus
Dynamic kinetic resolution
In another set of experiments, the racemic alcohols 4 were treated with 1 molar equivalent of the Mitsunobu reagent (derived from Ph3P+DEAD+5) in CH2Cl2 at 25°C. Interestingly, it was observed that two of the ketopinates, 6d and 6f, were obtained in yields higher than expected (63–75%, expected ≤50%) and with complete diastereoselectivity as shown by NMR (see in Table 1 under yields of 6 in parentheses; these yields were determined in the reaction mixture by NMR against a p-dinitrobenzene
Mechanism of enantioselection
It would appear that the generally high e.e. values obtained indicate a rigid transition state, whether at step 3 or at step 4 (Scheme 1). In the event that enantioselectivity originates at step 3, a carboxylate anion relatively tightly bound to the cationic phosphorus atom—perhaps making it pentavalent—is indicated. (There is indeed P31 NMR evidence for such intermediates.3) It is, in fact, likely that rigid intermediates such as 7 are involved, in which a pentavalent phosphorus atom is also
Conclusions
A new process of kinetic resolution with concomitant chirality inversion—one of the cherished goals of asymmetric synthesis—has been designed and demonstrated. The process defines a chiral version of the well-known Mitsunobu reaction, employing the readily accessible (1S)-(+)-ketopinic acid as a chiral auxiliary, and affords various secondary alcohols in excellent yields and enantiomeric excesses. Two of the benzylic alcohols also undergo dynamic kinetic resolution during the above Mitsunobu
Acknowledgments
CSIR (New Delhi) is thanked for generous financial support.
References (18)
Tetrahedron: Asymmetry
(1995)- Eliel, E. L.; Wilen, S. H.; Mander, L. N. Stereochemistry of Organic Compounds; John Wiley: New York, 1994; Chapter 7,...
- et al.
Top. Stereochem.
(1988) Org. React. (N.Y.)
(1992)Org. Prep. Proced. Int.
(1996)- et al.
J. Chem. Soc., Perkin Trans. 1
(1995) - et al.
Bull. Chem. Soc. Jpn.
(1995) - Smith, D. J. H. In Comprehensive Organic Chemistry; Barton, D. H. R.; Ollis, W. D.; Sutherland, I. J., Eds.; Pergamon...
Cited by (16)
Critical view on the recent enantioselective synthesis of alcohols, amines and related molecules having tertiary benzylic stereocenter
2018, TetrahedronCitation Excerpt :However, these types of compounds are difficult to access in a stereo selective fashion as routes to access these compounds in enantioselective fashion are rare. To address this issue, Pedro and co-workers developed the first catalytic enantioselective alkylation of pyrroles with 2, 2, 2-trifluoroacetophenones leading to the formation of a tertiary alcohol appended with gem diaryl groups and a trifluoromethyl group [106]. Zirconium complex ligated with 3, 3′-dibromo binol (88) was found to be the most effective Lewis acid catalyst giving the products in high yield (up to 98%) and high enantioselectivity (up to 93% ee) (Scheme 59).
Chemoenzymatic synthesis of (R)- and (S)-1-heteroarylethanols
2008, Tetrahedron AsymmetryEpimers of bicyclo[2.2.2]octan-2-ol derivatives with antiprotozoal activity
2008, European Journal of Medicinal ChemistryCitation Excerpt :For the chemical inversion of secondary alcohols several methods have been reported. The most popular chemical reaction is the Mitsunobu inversion [18]. Some of its drawbacks are the labour-intensive preparation and work-up [19].
Total Synthesis of Clerobungin A via a Cascade Cyclization Reaction
2016, Journal of Organic Chemistry