Notes
J . Org. Chem., Vol. 64, No. 26, 1999 9743
of 58% (Scheme 1). It is possible to conclude from the
mnemonic of Sharpless,3 depending on how RL, RM, and
RS are assigned, that the major diol 4b is the 3R isomer,5
but 5b is left unassigned.
isomer may be deduced from the published data,8 which
were based on Mosher ester analyses.
The sequence of reactions (outlined in Scheme 1) that
commences with (Z)-â-ocimene (1b) has also been con-
ducted with predominantly (E)-â-ocimene (E:Z 70:30), so
that the corresponding compounds including the E isomer
13 have been acquired. With the availability of 11b and
Flash chromatography led to separation of the required
4b from 5b, and the former was converted to its acetonide
6b, but this was not sufficiently separated into enanti-
omers on the â-cyclodextrin column to assay the induction
level in the AD-reaction. Ozonolysis-reduction (NaBH4)
of 6b provided protected triol 7 with [R]23D +19.6 (c 0.82,
EtOH), which may be compared with [R]22D -21.2 (EtOH)6
and -21.0 (EtOH)7 reported for the enantiomer. This
indicates an ee of at least 96% in the AD reaction, which
is supported by enantioselective gas chromatography of
the derived acetate 8 and its racemate (Scheme 1).
its E isomer 13 ([R]25 -5.0 (c 1.1, CH2Cl2)), of known
D
absolute configuration, GC comparisons, including co-
injection studies, established that the major volatile
component from A. nitida males is 14, i.e., the title
compound with 3R,5E stereochemistry. Behavioral stud-
ies are planned.
Exp er im en ta l Section
Mono-mesylation of the diol (4b) at the secondary
center (C-3) followed by mild base treatment was envis-
aged to form the epoxide with inversion of configuration,
thus providing the (3S)-epoxide. Treatment of the diol
with 1 equiv of freshly distilled mesyl chloride, in the
presence of Et3N, led to monomesylate formation, with
the secondary/tertiary ratio (i.e., 9b:10b) dependent on
the temperature of mesylation. At -15 °C, secondary
mesylation was hardly detectable; at 0 °C the mixture
was ca. 2:1 in favor of the secondary mesylate, but at 23
°C, was ca. 7:1 (Scheme 1). This is based on the observa-
tion that mesylation of the secondary alcohol to form 9b
causes a substantial downfield shift of the methine proton
resonance from δ 3.40 to δ 4.45, whereas mesylation of
the adjacent tert-alcohol to form 10b hardly affects this
chemical shift. This temperature variation in mesylation
selectivity is presumably associated with conformer
populations and hydroxyl group accessibility within
these. There was no NMR evidence that chloride ion
displacement of secondary mesylate intervened prior to
base treatment and epoxide formation. This was further
confirmed by the very high ee’s of the final epoxides, as
chloride intervention prior to cyclization would provide
an overall racemizing effect.
Gen er a l Meth od s. 1H NMR spectra were recorded at 400
and 200 MHz and 13C NMR spectra at 50 and 100 MHz, with
either TMS (δ ) 0) or the signal for residual CHCl3 in the CDCl3
1
solvent (δ7.24) as internal standards for H NMR and the central
peak of the CDCl3 triplet (δ ) 77.00 ppm) for 13C NMR spectra.
J values are reported in Hz. Flash chromatography was per-
formed with Kiesel S (0.032-0.063 mm). Enantioselective gas
chromatography was conducted using a permethylated â-cyclo-
dextrin column (SGE, 50 m; 0.25 µm).
(()-(5E,Z)-2,6-Dim eth yl-2-3-epoxyocta-5,7-dien e (2a, 2b).
A mixture of (E)- and (Z)-â-ocimene (1a /1b ) 70:30) (205 mg,
1.5 mmol) was dissolved in CH2Cl2 (10 mL), and m-chloroper-
benzoic acid (325 mg, 1.5 mmol, technical grade, 80%) was added
to this stirred and cooled (0 °C) solution. After being stirred at
room temperature (1 h), the solution was washed with aqueous
sodium hydrogen carbonate. The separated organic phase was
dried (Na2SO4) and carefully reduced in volume. The regioiso-
meric epoxides 2 and 3 were separated by flash chromatography
on silica gel with 5% ethyl acetate in hexane as eluant, with 3
eluting earlier. The E,Z isomers 2a ,b and 3a ,b were separated
by normal-phase HPLC using 1% diethyl ether in hexane. In
this way, 115 mg (0.76 mmol) (51%) of 2a ,b and 45 mg (0.3
mmol) (20%) of 3a ,b were obtained prior to HPLC separation of
these E,Z pairs. 2a : 1H NMR (CDCl3, 200 MHz) δ 6.38 (dd, J
17.3, 10.0, H7), 5.51 (t, J 7.7, H5), 5.12 (d, J 17.3, H8a), 4.97 (d,
J 10.8, H8b), 2.76 (t, J 6.4, H3), 2.46 (ddd, J 15.4, 6.7, 6.7, H4a),
2.29 (ddd, J 15.4, 6.7, 6.7, H4b), 1.75 (s, CH3), 1.30 (s, CH3),
1.29 (s, CH3); 13C NMR (CDCl3, 50 MHz) δ 141.1 (C7), 136.1 (C6),
127.0 (C5), 111.6 (C8), 63.4 (C3), 58.4 (C2), 28.3 (C4), 24.8 (CH3),
18.7 (CH3), 11.9 (CH3). 2b: 1H NMR (CDCl3, 200 MHz) δ 6.72
(ddd, J 17.3, 10.8, 8.0, H7), 5.41 (t, J 7.6, H5), 5.23 (d, J 17.2,
H8a), 5.12 (d, J 10.8, H8b), 2.74 (t, J 6.4, H3), 2.47 (ddd, J 15.4,
7.0, 7.0, H4a), 2.32 (ddd, J 15.4, 7.0, 7.0, H4b), 1.83 (dd, J 2.4,
1.2, CH3), 1.30 (s, CH3), 1.29 (s, CH3); 13C NMR (CDCl3, 50 MHz)
δ 134.6 (C7), 133.3 (C6), 124.9 (C5), 114.4 (C8), 63.6 (C3), 58.4
(C2), 27.4 (C4), 24.8 (CH3), 18.7 (CH3), 19.8 (CH3); HREIMS calcd
for C10H16O 152.1201, found 152.1199. 3a : 1H NMR (CDCl3, 400
MHz) δ 5.65 (dd, J 17.4, 10.7, H7), 5.26 (dd, J 17.4, 1.1, H8a),
5.15 (dd, J 10.7, 1.1, H8b), 2.78 (t, J 6.4, H5), 2.37 (m, H4), 2.19
(m, H4b), 1.71 (d, J 1.1, CH3), 1.52 (s, CH3), 1.40 (s, CH3); 13C
NMR (CDCl3, 50 HMz) δ 140.9, 136.1, 118.6, 115.7, 64.8, 59.5,
28.0, 25.7, 17.9, 15.0. 3b: 1H NMR (CDCl3, 400 MHz) δ 5.82
(dd, J 17.4, 10.9, H7), 2.87 (t, J 6.4, H5), 2.26 (m, H4a), 2.08 (m,
H4b), 1.69 (d, J 1.1, CH3), 1.40 (s, CH3); 13C NMR (CDCl3, 50
MHz) δ 134.5, 118.8, 117.7, 65.4, 60.4, 29.7, 25.7, 21.6, 17.9 (one
signal not located). The data for the minor isomers 3a and 3b
are incomplete because of some signal overlapping and low
quantities. The 1H NMR and mass spectral data match those
reported for 2a and 2b acquired by H2O2 epoxidation in the
presence of sodium tungstate.10
Stirring a predominantly secondary mesylate mixture
(9b, 10b) with K2CO3 in MeOH caused rapid conversion
(∼30 min, rt) to the epoxide 11b (Scheme 1), whereas
the minor tert-mesylate 10b reacted much more slowly
and did not form an epoxide. The pure epoxide 11b (flash
chromatography), exhibited [R]25 -3.5 (c 2.38, CHCl3)
D
and was nicely separated into its enantiomers on a
â-cyclodextrin phase and showed >95% ee. Confirmation
that 11b is S-configured was provided in the following
way. The epoxide on ozonolysis, reduction, and acetyla-
tion provided 12 of >97% ee based on enantioselective
gas chromatographic comparisons with the separately
synthesized racemic acetate and with [R]23D -19.5 (c 0.4,
CHCl3). This latter value may be compared with the
literature value of [R]25 +3.70, for the predominantly
D
3R isomer (27% ee,)8 which was stereochemically cor-
related with authentic (S)-(-)-4-methyl-1,3-pentanediol
in turn previously related to L-(-)-glyceraldehyde.9
A
calculated value of [R]25D +13.7 for the optically pure 3R
(5) This is consistent with the observation that treatment of
5-bromo-2-methyl-2-pentene with AD-Mix-R etc. forms (3S)-5-bromo-
2-methylpentane-2,3-diol. Vidari, G.; Landranchi, G.; Sartore, P.; Serra,
S. Tetrahedron: Asymmetry 1995, 6, 2977.
(6) Meier, V. H.; Ue¨belhart, P.; Eugster, C. H. Helv. Chim. Acta
1986, 69, 106.
(9) Bu¨chi, G.; Crombie, L.; Godin, P. J .; Kaltenbronn, J . S.; Siddal-
ingaiah, K. S.; Whiting, D. A. J . Chem. Soc. 1961, 2843. (R)-(+)-4-
methyl-1,3-pentanediol derived from (R)-(+)-4-methyl-3,4-epoxypen-
tanol of 27% ee had [R]22 +6.67 (CHCl3),8 whereas (S)-(-)4-methyl-
D
1,3-pentanediol exhibited [R]27 -6.9 (CHCl3), indicating incomplete
D
optical resolution of (()-3-hydroxy-4-methylpentanoic acid.
(10) Anisimov, A. V.; Chau, F. L.; Tarakanova, A. V.; Lebedev, M.
Yu.; Berentoreig, V. V. Zh. Org. Kh. 1992, 28, 1403.
(7) Nakagawa, N.; Mori, K. Agric. Biol. Chem. 1984, 48, 2799.
(8) Rossiter, B. E.; Sharpless, K. B. J . Org. Chem. 1984, 49, 3707.