9
078
T. Yamamoto et al. / Tetrahedron Letters 43 (2002) 9077–9080
Scheme 2.
to complete the reaction. After acetonitrile (1.6 L) and
Aliquat™ (3.12 g) were added to the reaction mixture
at room temperature, KOH powder (224 g) was added
over 30 minutes under a nitrogen atmosphere. The
mixture was then heated at reflux for 16 h. After
removal of acetonitrile in vacuo, the residue was com-
bined with hexane (600 mL), and the solution was
neutralized with 5% hydrochloric acid. The hexane
layer was separated out, washed with water and brine,
dried over Na SO , and evaporated in vacuo to give the
The typical procedure is as follows; the (3S)-allenyl
alcohol 5 (5.0 g, 32.5 mmol) was stirred with catalytic
amounts of palladium(II) acetate (73.0 mg, 0.33 mmol)
and
(S)-BINAP{2,2%-bis(diphenylphosphino)-1,1%-
binaphthyl} (0.51 mg, 0.66 mmol) at room temperature
for 16 h, and the reaction mixture was distilled (63–
6
5°C/6 mmHg) to give 2.1 g of the l-oxide 1 (cis/
trans=37/63).
2
4
Recent remarkable progress in the methods and tech-
nologies for both synthesis and analysis has accelerated
crude (3S)-5 (158 g, GC; 80%) which was subjected to
cyclization without any purification. That is, toluene
13
olfactory studies of optically active compounds.
(
(
223 g) and KHSO (3 g) was added to the crude (3S)-5
4
Regarding the odor of optically active citronellyl
derivatives, we had reported that the cyclic compounds
show fairly large differences in odor properties between
the diastereomeric pairs compared with the linear com-
158 g), and the mixture was heated at refluxed for 8 h.
The reaction mixture was washed with water and brine,
dried over Na SO , and evaporated in vacuo followed
2
4
by distillation to yield the l-oxide 1 (74 g, bp 75–77°C/
1
4
4
pounds. In the case of rose oxide, Ohloff had
reported that the d-cis-oxide 1 has a sweet scent,
whereas the l-cis-oxide 1 has a powerful fruity odor.
8
0–85 mmHg, GC; 99.8%, l-cis-1/l-trans-1=90/10) in
4% from the (S)-2. The optical purity of the l-oxide 1
4
was confirmed to be same as that of the starting
material (3S)-2 by gas chromatographic analysis with a
chiral stationary phase [Chiraldex G-TA: 0.25 mm
I.D.×30 m; conditions: 30°C (20 min) to 70°C (40 min)
at 2°C/min]. In the same manner, d-oxide 1 can be
synthesized starting from the (3R)-2 (98% ee) as a raw
material.
15
Further detailed odor evaluation
of the four
diastereomers synthesized in the present work has been
conducted. (Table 2)
4
As suggested in previous reports, there have been great
differences in both odor quality and thresholds among
the diastereomers. Namely, the l-cis-oxide 1, a major
diastereomer in Bulgarian rose, has shown the lowest
odor threshold value of which the ratios relative to the
other three diastereomers were 1/100 to 1/320 and has
been evaluated to be superior in olfactory qualities to
The synthesis of 1 from conjugated diene alcohol 6 by
acid-catalyzed cyclization has been well studied, but the
products always consist of a mixture of cis-1 as a major
product and trans-1 as a minor product in different
1
2
ratios depending on the reaction conditions. Because
the thermodynamically unstable trans-isomer was pro-
duced in less than 40% under ordinary conditions, it
was troublesome to obtain pure trans-1 for odor
evaluation.
Thus we tried to produce optically active trans-rich 1
by diastereoselective Pd–BINAP-catalyzed cyclization
of the (R)- and (S)-5. As a result, the l-cis-rich oxide 1
was obtained from the (3S)-allenyl alcohol 5 using the
(
3R)-BINAP–Pd complex. On the other hand, the l-
trans-rich mixture was produced from the (3S)-allenyl
alcohol 5 using the (S)-BINAP–Pd complex (Table 1).
It is interesting to note that diastereomeric combina-
tions between the substrates and the ligands are very
important to produce the cis- and trans-configurations
selectively as shown in Scheme 4.
Scheme 3. A practical synthesis of l-rose oxide.