[155268-88-5]  · C21H20BNO  · (313.21)

(catalyst used for the borane-mediated stereoselective reduction of ketones)

Solubility: soluble in most organic solvents, e.g. THF, diethyl ether, CHCl3, toluene.

Analysis of Reagent Purity: 1H NMR (C6D6).

Preparative Methods: (R)-1,1,2-Triphenyl-2-aminoethanol, the precursor of the oxazaborolidine, is prepared in 60-73% yield by portionwise addition of solid methyl (R)-phenylglycinate hydrochloride to an excess of phenylmagnesium bromide (3 M in diethyl ether) at 0 °C. The amino alcohol (>99% ee, after recrystallization from ethanol) is treated with trimethylboroxine in refluxing toluene in a flask provided with a Dean-Stark trap and under argon. Removal of all the volatiles under vacuum gives (R)-B-methyl-4,5,5-triphenyl-1,3,2-oxazaborolidine [(R)-1]as a colorless oil, which is then diluted with toluene up to a known concentration.1For reductions, a sample of that solution is transferred via cannula to another flask and the solvent is removed under vacuum and replaced by THF under argon. (S)-B-Methyl-4,5,5-triphenyl-1,3,2-oxazaborolidine [(S)-1] can be obtained from methyl (S)-phenylglycinate hydrochloride in a similar way. Both enantiomers of methyl phenylglycinate hydrochloride are commercially available at moderate prices.

Purification: Occasionally in the 1H NMR spectrum of the oxazaborolidine (C6D6), besides the expected signal at d 5.38 ppm, a singlet at d 4.93 ppm can be observed which is due to hydrolyzed, ring-cleaved material. The product may be purified by further treatment with a small amount of trimethylboroxine in refluxing toluene.

Handling, Storage, and Precautions: This oxazaborolidine is sensitive to water and to oxygen. However, its toluene solution can be stored at room temperature under argon for months with negligible loss of its catalytic activity. Care should be exercised to avoid contact of this compound with eyes and skin, and it should be manipulated in a well-ventilated fume hood.

Enantioselective Ketone Reduction

After the pioneering work of Itsuno et al.,2 Corey's group isolated the 1,3,2-oxazaborolidine derived from chiral a,a-diphenyl-2-pyrrolidinemethanol (2) and applied it (and also other related B-alkyl compounds) to the stereoselective reduction of ketones with borane-tetrahydrofuran, borane-dimethyl sulfide (BMS) or catecholborane.It was named the CBS method (after Corey, Bakshi, and Shibata).3 Since then, the CBS method has become a standard and has been extensively used, specially for aromatic and a,b-unsaturated ketones, not only in academic laboratories but also in industrial processes.4

Among the diverse 1,2-amino alcohols described as precursors of oxazaborolidines,5 the use of both enantiomers of highly enantioenriched 1,1,2-triphenyl-2-aminoethanol, which lead to oxazaborolidines (R)-1 and (S)-1, is especially attractive as they arise from inexpensive (R)- or (S)-phenylglycine, respectively. Oxazaborolidines (R)-1 and (S)-1are efficient catalysts in the borane-mediated stereoselective reduction of some types of prochiral ketones.

Typically, reductions are performed by slow addition (~15-30 min) of the ketone (1.0 mmol) to a solution of BMS (1.0 mmol) and 0.1-1.0 mmol of 1 (~1 M in THF) under argon at 0 °C (1). Yields are excellent in general after stirring for a further few minutes. The slow addition of ketone appears to enhance the stereoselectivity and in many cases causes the ee noted with 0.1 mmol of 1 to be similar or only slightly lower than that in the stoichiometric case.

Alcohols 3-8, obtained by the reduction of the corresponding ketones with equimolar amounts of BMS and (R)-1, are obtained with high ees (ee values given are obtained using 0.1 equiv of (R)-1). Enantioselectivity is excellent (often similar or only slightly lower than those reported in the CBS reduction) for aromatic and hindered methyl ketones,1a,6 (e.g. 3-5) and is also good for linear and a-monobranched enones7 (e.g. 7 and 8), but lower for linear methyl ketones like 2-octanone (6). In should be noted that in the reduction of unsaturated ketones, the time of addition is critical (the optimum being around 15-20 min) in order to avoid concomitant olefin hydroboration. In sharp contrast to the CBS process, the use of catecholborane (instead of BMS) or alternative solvents proved to be detrimental.

As far as the stereochemical course of the reaction is concerned, the configuration of the emergent stereocenter may be explained in terms of the mechanism proposed by Corey et al. for similar oxazaborolidine mediated reactions.4 Thus, the transition state operates such that the bigger group (RL) is located remotely from the methyl group on the boron atom (2).

Accordingly, the experience gained with oxazaborolidine 1 suggests an order of ‘empirical’ size of R groups.8 Obviously, better enantioselectivities in the reduction of the ketone carbonyl group are achieved when the substituents RL and RS are dissimilar.

The reduction of a-phenylthio enones constitutes a recent application of these findings to the preparation of chiral a-hydroxy thioesters (3).9

Reduction of Acetylenic Ketones

The enantioselective reduction of a,b-acetylenic ketones (R-CO-C&tbond;C-RŽ, RŽ = H or TMS) with BMS and 1affords the corresponding propargylic alcoholsin good to excellent yields and >90% ee (see eq 4 as and example).1b In some cases, the reductions of more sterically crowded hexacarbonyldicobalt complexes (e.g.9) derived from the acetylenic ketones also lead, after decomplexation with Cerium (IV) Ammonium Nitrate (CAN), to the same alcohols. However, the use of an oxazaborolidine with an a-face more available for complexation, such as those derived from commercially available (1S,2R)-2-amino-1,2-diphenylethanol, is required (5).10 Remarkably, the temporary transformation of the acetylenic moiety into its Co2(CO)6 complex, not only reverses the stereoselectivity in the reduction step, but also enhances it.

In addition, the highly enantioenriched propargylic alcohols obtained in such a way are versatile building blocks. Theyhave been applied to the syntheses of the alkyl side chains of zaragozic acids A and C,11several metabolites isolated from marine sponges,12 and the octalactin A ring.13

On the other hand, the BMS/(R)- or (S)-1 mixture is capable of displaying high asymmetric induction in the reduction of acetylenic ketones, thereby overriding the normally small diastereofacial selectivity of a chiral a-substituted 1-trialkylsilyl-1-alkyn-3-one (e.g. 10) in a predictable and controlled manner (reagent control) (6). Remarkably, the stereoselectivity noted in such reductions has shown strong dependence upon the steric requirement of the C(1) substituent. Thus, an increasing stereoselectivity has been noted in the reduction of ketones 10as R changes from Me to Et to i-Pr. An explanation for such an unexpected remote effect has been suggested based on abinitio calculations.14

Further work according to this double asymmetric strategy has led to the establishment of a stereodivergent route to b-hydroxy g-substituted carboxylic acids and a-hydroxy b-substituted carboxylic acids (including N-Boc-statine and N-Boc-norstatine).15

Reduction of 1,4-Diketones

Synthetic access to C2-symmetric 1,4-diols, useful building blocks for the preparation of chiral 2,5-disubstituted pyrrolidines and phospholanes, involves reduction of the parent 2-alkane-1,4-diones, or, even better, reduction of the related (E)-alk-2-ene-1,4-diones (11) (7) or 2-alkyne-1,4-diones (12) (8), followed by catalytic hydrogenation.16

Generally, reduction of diketones 12 (or in some cases their hexacarbonyldicobalt complexes) yields better stereoselectivities than the related ethylenic diketones 11, especially when R is a sterically demanding group. In addition, the propargylic diols obtained can be easily transformed not only into the saturated 1,4-diols, but also into (Z)- or (E)-alk-2-ene-1,4-diols. The C2-symmetric allylic 1,4-diols have been very recently used as building blocks in a formal synthesis of (-)-methylenolactocin and (-)-phaseolinic acid.17Related to this, it should be noted that reduction of the unstable (Z)-alk-2-ene-1,4-diones (13) is an unsuitable route to (Z)-alk-2-ene-1,4-diols since a considerable amount of 1,4-reduction (9) is also observed.17, 18

1. (a) Berenguer, R.; Garcia, J.; Vilarrasa, J., Tetrahedron: Asymmetry 1994, 5, 165. (b) Bach, J.; Berenguer, R.; Garcia, J.; Loscertales, T.; Vilarrasa, J., J. Org. Chem. 1996, 61, 9021.
2. Itsuno, S.; Sakurai, Y.; Ito, A.; Hirao, S.; Nakahama, S., Bull. Chem. Soc. Jpn 1987, 60, 395 and references therein.
3. Corey, E. J.; Bakshi, R. K.; Shibata, S., J. Am. Chem. Soc. 1987, 109, 5551.
4. For a review, see: Corey E. J.; Helal, C. J., Angew. Chem., Int. Ed. Engl. 1998, 37, 1986.
5. For reviews, see: (a) Deloux, L.; Srebnik M., Chem. Rev. 1993, 93, 763. (b) Wallbaum, J.; Martens, J., Tetrahedron: Asymmetry 1992, 3, 1475.
6. Berenguer, R.; Garcia, J.; Gonzàlez, M.; Vilarrasa, J., Tetrahedron: Asymmetry 1993, 4, 13.
7. (a) Bach, J.; Berenguer, R.; Garcia, J.; Vilarrasa, J., Tetrahedron Lett. 1995, 36, 3425. (b) Bach, J.; Berenguer, R.; Farràs, J.; Garcia, J.; Meseguer, J.; Vilarrasa, J., Tetrahedron: Asymmetry 1995, 6, 2683.
8. Bach, J.; Berenguer, R.; Garcia, J.; López, M.; Vilarrasa, J., unpublished results.
9. Berenguer, R.; Cavero, M.; Garcia, J.; Muñoz, M., Tetrahedron Lett. 1998, 39, 2183.
10. Quallich, G. J.; Woodall, T. M., Tetrahedron Lett. 1993, 34, 4145.
11. Bach, J.; Galobardes, M.; Garcia, J.; Romea, P.; Tey, C.; Urpí, F.; Vilarrasa, J., Tetrahedron Lett. 1998, 39, 6765.
12. (a) Garcia, J.; López, M.; Romeu, J., Tetrahedron: Asymmetry 1999, 10, 2617. (b) Garcia, J.; López, M.; Romeu, J., Synlett 1999, 4, 429.
13. Bach, J.; Garcia, J., Tetrahedron Lett. 1998, 39, 6761.
14. Alemany, C.; Bach, J.; Farràs, J.; Garcia, J., Org. Lett. 1999, 1, 1831.
15. Alemany, C.; Bach, J.; Garcia, J.; López, M.; Rodriguez, A. B., Tetrahedron 2000, 56, 9305.
16. (a) Bach, J.; Berenguer, R.; Garcia, J.; Loscertales, T.; Manzanal, J.; Vilarrasa, J., Tetrahedron Lett. 1997, 38, 1091. (b) Bach, J.; Berenguer, R.; Garcia, J.; López, M.; Manzanal, J.; Vilarrasa, J., Tetrahedron 1998, 54, 14947.
17. Ariza, X.; Garcia, J.; López, M., Synlett 2001, 6, 9305.
18. Berenguer, R.; Garcia, J., unpublished results.

Jordi Garcia

University of Barcelona, Barcelona, Spain

Copyright © 1995-2000 by John Wiley & Sons, Ltd. All rights reserved.