Highly diastereoselective dihydroxylation of cis-substituted sulfonyl vinyl oxiranes

Article abstract of DOI:10.1016/S0040-4039(00)00275-6

An expedient route to rare carbohydrate-like fragments that relies on the highly stereoselective osmium-catalyzed dihydroxylation of 2-sulfonyl-2- hydroxyalkyl-3-vinyloxiranes is outlined. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00275-6

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 2871–2874
Highly diastereoselective dihydroxylation of cis-substituted
sulfonyl vinyl oxiranes1
Roberto Fernández de la Pradilla,a,∗ Paloma Méndez a and Alma Viso b
aInstituto de Química Orgánica, CSIC, Juan de la Cierva, 3, 28006 Madrid, Spain
bDepartamento de Química Orgánica I, Facultad de Químicas, Universidad Complutense, 28040 Madrid, Spain
Received 3 January 2000; accepted 11 February 2000
Abstract
An expedient route to rare carbohydrate-like fragments that relies on the highly stereoselective osmium-
catalyzed dihydroxylation of 2-sulfonyl-2-hydroxyalkyl-3-vinyloxiranes is outlined. © 2000 Elsevier Science Ltd.
All rights reserved.
Keywords: hydroxylation; oxiranes; carbohydrates; sulfones.
The osmium-catalyzed dihydroxylation of alkenes bearing an allylic oxygen takes place with high
anti stereoselectivity in most cases.2 Interestingly, to our knowledge, there are just three reports on
the dihydroxylation of alkenyl trans-oxiranes that occurred with low stereoselectivity.3 While the use
of reagent-controlled catalytic asymmetric dihydroxylation procedures4 provided good to excellent
diastereocontrol in those cases,3,5 we felt that cis-substituted oxiranes could display a useful level of
diastereoselectivity even in the absence of chiral ligands. Enantiopure sulfonyl vinyl oxiranes B, readily
available by metal-catalyzed oxidation at sulfur followed by hydroxyl-directed epoxidation at the most
electron deficient double bond,6 appeared as good test substrates for this study since the dihydroxylation
to give adducts C would also produce rare carbohydrate-like fragments in an expedient fashion (four steps
from commercially available menthyl sulfinate) (Scheme 1). The recent disclosure of the work by Cossy
et al. on the dihydroxylation of cis-substituted isopropenyl cyclopropanes, oxiranes and aziridines,7
prompted us to report our preliminary results on this subject.8
Scheme 1.
∗
Corresponding author. Fax: 34 91 5644853; e-mail: iqofp19@fresno.csic.es (R. Fernández de la Pradilla)
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00275-6
tetl 16555

2872
To enhance the synthetic usefulness of our proposed approach to carbohydrate-like fragments we
sought an efficient procedure for preparing diastereomeric oxiranes, such as 3 (Scheme 2) from
1. After many fruitless attempts under Mitsunobu conditions we turned our attention to an oxida-
tion–reduction protocol. After considerable experimentation, we found that the reduction of ketone 2
9
with NaBH4·CeCl3 gave an excellent yield of the desired product 3, readily separated from a small
amount of 1.10 It should be pointed out that the stereoselectivity of this reduction is opposite to the
literature data,9 probably due to the presence of the sulfonyl moiety in our substrate 2.
Scheme 2.
Table 1 gathers our preliminary results on the osmium-catalyzed dihydroxylation of our sulfonyl vinyl
oxiranes.11,12 To evaluate the stereodirecting capabilities of the benzylic and/or sulfinyl chiral centers,
in the absence of the oxirane, the dihydroxylation of diene 4 was studied and a good yield of a 60:40
mixture of sulfinyl triacetates 5 was obtained along with some sulfone 6, also as a 60:40 mixture; standard
oxidation of 5 gave 6 in good yield. Under similar conditions, hydroxy vinyl oxirane 3 led to a fair yield
of a 65:35 mixture of triacetate 7 and ketone 8, derived from overoxidation at the benzylic position, both
obtained as single isomers. In contrast, benzyl ether 9 gave the corresponding protected triol 10 with
diminished diastereoselectivity (85:15). Entries 4 and 5 show that diastereomeric hydroxy vinyl oxirane
1 is also a good substrate for this protocol giving triacetate 12 as a single isomer along with a small
amount of ketone 8, provided that osmylation is not allowed to proceed for a long time.
While the observed overoxidation is likely related with the reaction time and the benzylic nature of
these substrates, we sought an alternative protecting group that would prevent overoxidation without
compromising the facial selectivity of the process. In this regard the osmylation of acetoxy vinyl oxirane
13 was explored to give, after removal of the acetate and peracetylation, a single isomer of triacetate 12
(entry 6). Finally, we addressed the ‘simultaneous’ dihydroxylation of 14 at both vinyl residues that gave
just two of the four possible isomers of the rare protected heptitols 15 and 16 in a highly stereoselective
manner (entry 7).
The isolation of ketone 8 from both diastereomers 3 and 1 indicates that the stereochemical outcome
of the process is independent of the configuration of the hydroxyalkyl substituent cis to the vinyl moiety.
The very high selectivity found for these substrates is in sharp contrast to the recent findings of Cossy et
al. for a cis-2-alkyl-3-vinyloxirane that, unlike the related isopropenyl oxiranes, gave a 50:50 mixture of
diastereomers upon catalytic dihydroxylation.7 These results may be accounted for largely as described
by these authors in terms of electrophilic addition to the less hindered face of the gauche conformation
D (Scheme 3).13 In our case, the presence of the bulky sulfonyl moiety is likely to exert a determining
influence on the conformational distribution of the key secondary alcohol center of R, and that in turn
controls the conformation of the vinyl residue.
Scheme 3.

2873
Table 1
Osmium-catalyzed dihydroxylation of sulfonyl vinyl oxiranes
In conclusion, the osmium-catalyzed dihydroxylation of readily available enantiopure 2-sulfonyl-2-
hydroxyalkyl-3-vinyloxiranes takes place with excellent selectivity to produce unusual carbohydrate
fragments. We are currently studying the scope and limitations of this process as well as additional
applications of the methodology.

2874
Acknowledgements
This research was supported by DGICYT (PB96-0822 and AGF98-0805-CO2-02) and CAM
(08.5/0046/1998). We also thank Janssen-Cilag for additional support.
References
1. Taken in part from the MS Thesis of P.M.
2. (a) Cha, J. K.; Christ, W. J.; Kishi, Y. Tetrahedron 1984, 40, 2247–2255. (b) For a recent review, see: Cha, J. K.; Kim, N.-S.
Chem. Rev. 1995, 95, 1761–1795. (c) For a computational treatment, see: Haller, J.; Strassner, T.; Houk, K. N. J. Am. Chem.
Soc. 1997, 119, 8031–8034. (d) For an exception to this trend using stoichiometric conditions, see: Donohoe, T. J.; Moore,
P. R.; Waring, M. J.; Newcombe, N. J. Tetrahedron Lett. 1997, 38, 5027–5030. (e) For enhanced selectivities in some cases,
see: Donohoe, T. J.; Waring, M. J.; Newcombe, N. J. Synlett 2000, 149–151.
3. (a) Allen, J. V.; Cappi, M. W.; Kary, P. D.; Roberts, S. M.; Williamson, N. M.; Wu, L. E. J. Chem. Soc., Perkin Trans. 1 1997,
3297–3298. (b) Allen, J. V.; Bergeron, S.; Griffiths, M. J.; Mukherjee, S.; Roberts, S. M.; Williamson, N. M.; Wu, L. E. J.
Chem. Soc., Perkin Trans. 1 1998, 3171–3179. (c) Kim, N.-S.; Choi, J.-R.; Cha, J. K. J. Org. Chem. 1993, 58, 7096–7099.
4. For a review, see: Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. Rev. 1994, 94, 2483–2547.
5. Xiang, A. X.; Watson, D. A.; Ling, T.; Theodorakis, E. A. J. Org. Chem. 1998, 63, 6774–6775.
6. Fernández de la Pradilla, R.; Méndez, P.; Priego, J.; Viso, A. J. Chem. Soc., Perkin Trans. 1 1999, 1247–1249.
7. Cossy, J.; Blanchard, N.; Meyer, C. Tetrahedron Lett. 1999, 40, 8361–8364.
8. See footnote 11 of Ref. 6.
9. Li, K.; Hamann, L. G.; Koreeda, M. Tetrahedron Lett. 1992, 33, 6569–6570.
10. A variety of reducing agents were tested with mediocre selectivities. Complete details will be published in a full account of
this work.
11. VanRheenen, V.; Kelly, R. C.; Cha, D. Y. Tetrahedron Lett. 1976, 1973–1976.
12. To facilitate evaluation of the data, the crude osmylation mixtures (5% OsO4, 3 equiv. Me3NO, 9:1 acetone:water, rt)
were filtered through silica gel (gradient elution, 5–80% MeOH:EtOAc) and acetylated (Ac2O, Et3N, DMAP, CH2Cl2, rt).
1
The diastereomeric ratios were determined by integration of well resolved signals in the H NMR of the crude mixtures.
All products described here have been fully characterized. These stereochemical assignments are tentative and rely on
examination of molecular models and on careful correlations of the 1H and 13C NMR data of our products, particularly 12,
with related oxiranes obtained by allylation of an epoxy aldehyde (Ref. 6) and with literature precedence (Nikitenko, A. N.;
Arshava, B. M.; Taran, I. G.; Mikerin, I. E.; Shvets, V. I.; Raifeld, Y. E.; Lang Jr., S. A.; Lee, V. J. Tetrahedron 1998, 54,
11 731–11 740. We are currently seeking confirmation to these assignments in the context of developing applications of the
1
methodology. Data of 12: Rf=0.16 (30% EtOAc–hexane). [α]D20=+19.7 (c 0.38, CHCl3). H NMR (CDCl3, 300 MHz) δ
2.05 (s, 3H, CH3 Ac), 2.07 (s, 3H, CH3 Ac), 2.08 (s, 3H, CH3 Ac), 2.41 (s, 3H, CH3 p-Tol), 3.82 (dd, 1H, J=12.4, 5.2 Hz,
H-β0), 4.09 (d, 1H, J=8.9 Hz, H-3), 4.12 (dd, 1H, J=12.4, 3.2 Hz, H-β0), 5.02 (ddd, 1H, J=8.7, 5.2, 3.2 Hz, H-α0), 6.40 (s,
1H, H-α), 7.20–7.27 (m, 5H, ArH), 7.35–7.38 (m, 2H, ArH), 7.57 (d, 2H, J=8.3 Hz, ArH). 13C NMR (CDCl3, 50 MHz) δ
20.6, 20.6, 20.8, 21.7, 59.5, 62.6, 67.7, 69.7, 76.7, 127.1 (2C), 128.4 (2C), 128.8, 129.5 (2C), 129.8, 133.8, 134.0, 145.4,
169.0, 169.5, 170.2. IR (film): 3040, 3000, 2920, 1730, 1580, 1480, 1430, 1350, 1310, 1200, 1130, 1070, 1030, 800, 740,
710, 680, 650 cm−1. EM (electrospray): 513 [M+Na]+ (100%).
13. Conformation D follows Vedejs’ model for dihydroxylation of allylic ethers, which is the commonly accepted model for
most acyclic cases. See: Vedejs, E.; McClure, C. K. J. Am. Chem. Soc. 1986, 108, 1094–1096.