Synthesis of 3-Oxa-guaianolides from santonin

Article abstract of DOI:10.1016/S0040-4020(00)00571-8

This article reports on the transformation of santonin into two C₁₀-epimeric 3-oxa-guaianolides which are 8-deoxyderivatives of several natural 3-oxaguaianolides isolated from Achillea species. The synthesis involved the photochemical rearrange

Full text of DOI:10.1016/S0040-4020(00)00571-8

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 6331±6338
Synthesis of 3-Oxa-guaianolides from Santonin
*
  Â
Ä
Gonzalo Blay, Luz Cardona, Begona Garcõa, Luisa Lahoz, Belen Monje and Jose R. Pedro
 Á  Á
Departament de Quõmica Organica, Facultat de Quõmica, Universitat de Valencia, E-46100 Burjassot (Valencia), Spain
Received 6 April 2000; revised 31 May 2000; accepted 15 June 2000
AbstractÐThis article reports on the transformation of santonin into two C10-epimeric 3-oxa-guaianolides which are 8-deoxyderivatives of
several natural 3-oxaguaianolides isolated from Achillea species. The synthesis involved the photochemical rearrangement of the eudesmane
skeleton into a guaiane skeleton and the transformation of the cyclopentane ring into a furan moiety with the concomitant loss of C3.
Comparison of the NMR data of the synthetic products with those of the natural products con®rms the b orientation of the hydroxyl group at
C10 in the products isolated from Achillea. q 2000 Elsevier Science Ltd. All rights reserved.
Introduction
in Mongolia and named 8-acetyl- (6) and 8-angeloylegelo-
lide (7), which differed only in the con®guration assignment
of C10.6 The structure of the compounds 2±4 isolated from
the European species was reinvestigated later7 with the use
of modern 2D NMR techniques and the stereochemistry of
C10 corrected to the con®guration with the hydroxyl group at
C10 b-orientatedÐthe same con®guration as in artabsin
(10)Ðthus establishing that the products isolated from
A. roseo-alba and A. collina were identical to those isolated
from A. millefolium and had structures such as 6±8 (Fig. 1).
The aromatic furan ring, one of the most representative ®ve-
membered heterocycles, can be found in many important
pharmaceuticals as well as in naturally occurring
compounds of vegetable and marine origin, especially terpe-
noids.1,2 For this reason, the synthesis of furans with differ-
ent substitution patterns has attracted the interest of many
synthetic chemists and many strategies for the construction
of this moiety have been devised.3
In this article we report on the ®rst synthetic approach to 10-
hydroxy-3-oxaguaian-6,12-olides starting from santonin
(1). We have prepared the corresponding 8-deoxyderiva-
tives of the above mentioned natural products in both
con®gurations at C10 5 and 9 in a stereoselective and unam-
biguous way. These synthesis have allowed the comparison
of the NMR data of the synthetic materials and the natural
In recent years several unusual 3-nor-guaianolides bearing a
furan moiety, namely 3-oxa-guaianolides, were isolated
from the ¯ower heads of European Achillea roseo-alba
and A. collina, and their structures described as 8a-acetoxy-
(2) (achillicin), 8a-angeloxy- (3) and 8a-tigloxy-3-oxa-10-
epi-artabsin (4).4,5 Almost simultaneously, two 3-oxa-
guaianolides were isolated also from A. millefolium growing
Figure 1.
Keywords: terpenes and terpenoids; lactones; Achillea; furans.
* Corresponding author. Tel.: 134-9638-64329; fax: 134-9638-64328; e-mail: jose.r.pedro@uv.es
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00571-8

6332
G. Blay et al. / Tetrahedron 56 (2000) 6331±6338
Scheme 1. (a) hn/AcOH; (b) 1. TsNHNH2, 2. Catecholborane; (c) 1. O3, 2. NaBH4; (d) Mesyl chloride; (e) o-NO2C6H4SeCN-Bu3P; (f) NaPhSe; (g) H2O2; (h)
1. O3, 2. Ph3P.
products providing additional support to con®rm the b-OH
orientation of the hydroxyl group at C10.
with sodium phenylselenolate to give phenylselenide 16 in
89% yield.11,12 The oxidative elimination of this selenide
yielded 90% of alkene 17 with identical features to the
compound obtained from 15. With compound 17 in our
hands we undertook the cleavage of the C2±C3 double
bond by ozonolysis. The outcome of this reaction was
very dependent on the work up of the ozonolysis mixture.
Ozonization in dichloromethane±methanol followed by
reduction with Zn13 or dimethyl sul®de14 gave the desired
ketoaldehyde 18 in moderate yield (52 and 64%, respec-
tively) together with by-products resulting from methanol
addition. Ozonization in dichloromethane followed by
cleavage of the ozonide with triethylamine15 resulted in a
similar yield (52%) of the expected product.15 Finally,
compound 18 was obtained in very high yield (97%)
when the ozonide prepared in dichloromethane was reduced
with triphenylphosphine.16
Results and Discussion
Santonin (1) was chosen as starting material in our synthetic
scheme. The synthesis of 3-nor-guaianolides from this
material required the transformation of the santonin carbon
framework from eudesmane into guaiane, and the one carbon
(C3) degradation of the cyclopentane ring in the resulting
guaiane with cleavage of the C2±C3 and C3±C4 bonds.
With this aim (Scheme 1), santonin (1) was transformed into
guaianolide 12 as has been described previously, in two
steps involving photochemical rearrangement of santonin
to O-acetylisophotosantonin (11), and reduction of the
tosylhydrazone of the resulting enone with catecholborane.8
The occurrence of the C3±C4 double bond in compound 12
should allow the cleavage of the C3±C4 bond by ozonolysis.
On the other hand it was considered convenient that, after
ozonolysis, both carbon atoms C3 and C4 resulted in a
different functionalization in order to make easier further
selective elaboration of the C2±C3 fragment leading to the
cleavage of this bond. Therefore compound 12 was
subjected to ozonolysis in dichloromethane±methanol
followed by careful reduction with sodium borohydride. In
this way reduction was achieved up to the alcohol at the less
substituted carbon C3 while C4 remained as a ketone giving
hydroxyketone 13 in 68% yield. The following steps were
addressed to the elimination of the hydroxyl group at C3,
which would permit removal of C3 from the molecule, by
ozonolysis of the resulting C2±C3 double bond. Elimination
was carried out in two steps via arylselenide 15. Treatment
of compound 13 with o-nitrophenylselenocyanate and tri-n-
butylphosphine in THF±pyridine gave compound 15 in 96%
yield, which upon treatment with H2O2 followed by sponta-
neous elimination of the resulting selenoxide gave alkene 17
in 91% yield.9,10 Alternatively, compound 17 was also
prepared via a different selenide 16 in a one step longer
sequence which uses diphenyldiselenide instead of the
more toxic and expensive o-nitrophenylselenocyanate.
Thus, the C3-hydroxyl in compound 13 was transformed
into its mesylate 14 in 90% yield which was then treated
Although 1,4-dicarbonyl compounds, such as 18, have been
traditionally considered as the direct precursors for furans,17
it was fairly foreseeable that the tertiary acetoxy group at
C10 may cause dif®culties during the cyclization to the furan
moiety. Indeed, it was observed that upon treatment with
p-toluenesulfonic acid in benzene compound 18 led to
²
alkenefuran 19 in very low yield besides polymeric
material (Scheme 2). On the other hand, when the reaction
was attempted using silica gel, basic or acidic aluminum
oxide as dehydrating agents, compound 20, resulting from
acetic acid elimination, was obtained in fair yields (ca.
65%). We assumed that the conjugation of the resulting
double bond with the aldehyde carbonyl was the cause of
the easy elimination of the acetoxy group. The solution to
this problem consisted of the reduction of the aldehyde
group which made necessary an oxidative step later on in
the synthetic sequence (Scheme 3). The selective reduction
of the aldehyde group occurred with concomitant cycli-
zation to hemiacetal 21. The best yield (95%) was obtained
²
Compound 19. 1H NMR (CDCl3) d 7.24 (1H, s, H-1), 5.41 (1H, brd,
Jˆ6.5 Hz, H-9), 5.02 (1H, d, Jˆ8.8 Hz, H-6), 2.61 (1H, dd, Jˆ6.5, 16.8 Hz,
H-8), 2.4±2.1 (3H, m, H-7, H-8, H-11), 2.38 (3H, s, H-15), 1.95 (3H, s, H-
14), 1.24 (3H, d, Jˆ7.0 Hz, H-13); 13C NMR d 178.5 (s, C-12), 148.3 (s, C-
4), 138.8 (d, C-2), 126.6, 123.5 (s, C-1, C-10), 122.8 (d, C-9), 115.3 (s, C-5),
79.3 (d, C-6), 48.3 (d, C-7), 41.3 (d, C-11), 31.7 (t, C-8), 23.7, 23.6 (q, C-14,
C-15), 13.3 (q, C-13).

G. Blay et al. / Tetrahedron 56 (2000) 6331±6338
6333
Scheme 2. (i) TsOH; (j) Al2O3.
with LiAlH4 at low temperature, which proved better than
NaBH4 in CH2Cl2±MeOH (40%) and sodium triacetoxyboro-
hydride in benzene (48%).18 The stereochemistry of C4 was
determined by NOE experiments. NOE between H6 and H5
indicated the b disposition of H5; irradiation of H15 gave
NOEs with H5 and the C4±OH group, and also NOEs were
observed with H1, H2a, and H15 upon irradiation of the C4±
OH signal at d 2.63. These experiments indicated the trans
junction of the dihydrofuran and the cycloheptane ring in
the molecule and the b orientation of the methyl group at C4.
sulfur20 have been used, but they require strong conditions
and do not give good results. Softer conditions have been
described with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone
(DDQ).21 However treatment of compound 23 with DDQ at
408C gave rise to untreatable mixtures. As an alternative, we
attempted the allylic oxidation of the D4,5 double bond with
SeO2 which was also unsuccessful. Finally it was possible to
obtain the furan ring by methylene blue sensitized photo-
oxygenation.10,12 Photooxygenation of alkenes in these
conditions leads to allylic hydroperoxides with migration
of the double bond. In our case, probably this reaction
gave a D1,5-4-hydroperoxyderivative which eliminated
hydrogen peroxide during column chromatography to give
the desired furan 24 with a moderate yield (37%) together
with 35% of recovered starting material 23.
The elimination of the hydroxyl group at C4 was attempted
via its conversion into a mesylate. Unexpectedly, treatment
of compound 21 with mesyl chloride in pyridine gave 96%
of a compound that was identi®ed as the C2-mesylate 22.
Compound 22 is probably formed through the equilibrium
between the hemiacetal and hydroxyketone forms of 21.
Anyway, this fact did not cause any setback since treatment
of mesylate 22 with 1,8-diazabicyclo[5.4.0]undec-7-ene
(DBU) in toluene gave the desired dihydrofuran 23 in
88% yield, via an intramolecular O-alkylation of the ketone
enolate.
The ®nal transformation of compound 24 into 5 required the
hydrolysis of the acetate group. This step was complicated
by the tendency of C10 in compound 5 to epimerize in
aqueous acids. This fact was observed in preliminary assays
of acetate hydrolysis in aqueous media and was con®rmed
in an independent experiment, as we will see below. Since
one of the objectives of this work was to prepare both
epimers at C10 in an unambiguous way, the use of aqueous
acid during reaction or work up had to be avoided. With this
purpose we decided to remove the acetate group not by
hydrolytic but by reductive procedures. Compound 24 was
treated with an excess of Red-Alw in order to ensure
For the construction of the furan moiety the dehydrogena-
tion of the dihydrofuran ring in reaction conditions com-
patible with the tertiary acetoxy group at C10 was
required. Quite surprisingly, not many methods to achieve
this transformation can be found in the literature. NiO219 or
Scheme 3. (d) Mesyl chloride; (k) LiAlH4; (l) DBU; (m) 1. O2, hn, methylene blue, 2. SiO2; (n) 1. Red-Alw, 2. NMO, TPAP; (o) 1. OsO4, 2. 2 M HCl

6334
G. Blay et al. / Tetrahedron 56 (2000) 6331±6338
Table 1. Comparative 1H NMR d values of compounds 5±9
On the other hand, the resonance frequency of H14 will be
also affected because of the magnetic anisotropy due to the
furan ring, depending on its orientation (a or b).24 In
compound 9, H14 with an a orientation is held in the
deshielding zone and it is expected that it resonates at higher
d value than in compound 5 where this methyl group is out
of this zone. The observed values are 1.45 and 1.63 ppm in
compounds 5 and 9, respectively.
H
5a
9a
6a,b
7a,b
8a,b
2
6
14
15
7.27
4.98
1.45
2.36
7.17
5.30
1.63
2.38
7.18
5.41
1.65
2.39
7.19
5.45
1.66
2.39
7.19
5.45
1.64
2.39
a d registered using TMS as internal standard. We have registered the 1H
NMR of compounds 5 and 9 using TMS as internal standard for a more
accurate comparison with the natural products 1H NMR spectra.
b Taken from Refs. 4 and 6.
1
Following this reasoning, a comparative study of the H
NMR data of the oxaguaianolides isolated from Achillea
sp. and of our synthetic materials can be applied to deter-
mine the stereochemistry of C10 in these natural products.
This comparison is feasible since the 8-acyloxy groups
present in the natural products are far away enough from
H2, H6, and H14; and therefore will not have a noticeable
effect on their NMR chemical shifts. The d values for these
protons and H15 in the 1H NMR spectra of compounds 5, 9,
and of the three natural oxa-guaianolides are shown in Table
1. A good agreement between the d corresponding to H2, H6
and H14 is observed for compounds 6±8 and 9, while they
differ signi®cantly from the observed in compound 5. This
agreement is in accordance with the expected results
according to the reasoning above and therefore constitutes
additional evidence for the b-orientation of the C10±OH
group in the natural products 6±8.
complete reduction of the acetate group. This treatment also
brought about the reduction of the lactone ring to give a
mixture of lactols which were subjected to reoxidation with
4-methylmorpholine N-oxide (NMO) and tetra-n-propylam-
monium perruthenate (TPAP) without separation.22 In this
way we obtained a compound in 75% yield for the two steps.
The NMR data obtained for this compound were in good
agreement with the expected for a 10-hydroxy-3-oxaguaian-
6,12-olide and it was assigned structure 5. The a orientation
of the hydroxyl group was determined by the stereochemistry
of C10 in O-acetylisophotosantonin (11) which is perfectly
known and does not change along the synthetic sequence.
The tendency of compound 5 to epimerize in aqueous acid
was con®rmed in the following way. Compound 5 was
stirred in 2 M HCl for 2 min and then extracted with ethyl
acetate. The extract was analyzed by 1H NMR and
contained ca. 54% of 5, 18% of 9 and 9% of 19.
Experimental
General
For the synthesis of the 3-oxaguaianolide with the b-OH
group 9, the epimerization of C10 was required. Although
this compound can be obtained directly from 5 by epimeri-
zation in acidic media we envisaged a shorter procedure for
its synthesis (Scheme 3). The starting point was compound
20 obtained as mentioned above. Selective reduction of the
aldehyde group with LiAlH4 gave hemiacetal 25 in 81%
yield. The orientation of the C4±OH group in this compound
was determined from the NOEs observed between H5 and
H6, and H5 and H15 signals. Compound 25 was subsequently
subjected to hydroxylation of the double bond with osmium
tetroxide23 followed by work up with 2 M HCl which
brought about dehydration and aromatization of the triols
resulting from hydroxylation. In this way we obtained two
hydroxy furans in 60% and 15% yield respectively after
column chromatography. The minor product showed
spectral data identical to furan 5 obtained as described
above and it was assigned the structure with the C10 a-OH
All melting points are uncorrected. Column chroma-
tography was performed on silica gel (Merck, silica gel
60, 230±400 mesh). Commercial reagents and solvents
were analytical grade or were puri®ed by standard pro-
cedures, prior to use.25 All reactions involving air or
moisture sensitive materials were carried out under argon
atmosphere. Speci®c rotations were measured in CHCl3. IR
were recorded as liquid ®lm in NaCl for oils and as KBr
discs for solids. NMR were run in CDCl3 at 399.94 MHz for
1H and at 50.3, 75.43, or 100.58 MHz for 13C NMR, and
referenced to the solvent as internal standard
(dHˆ7.24 ppm) except for compounds 5 and 9 (TMS as
internal standard). The carbon type was determined by
DEPT experiments. Mass spectra were run by electron
impact at 70 eV or chemical ionization using methane as
ionizing gas.
1
group. On the other hand, the H and 13C NMR spectra of
O-Acetylisophotosantonin (11). A solution of santonin (1)
(9.5 g, 38.6 mmol) in 140 mL of glacial acetic acid was
irradiated under argon with a 150 W medium-pressure
mercury lamp housed in an refrigerated quartz immersion
tube until complete reaction of santonin as determined by 1H
NMR analysis of an aliquot. The acetic acid was evaporated
under reduced pressure with the help of hexane and the
resulting oil dissolved in 16 mL of hot methanol and then
left overnight at 2208C. Filtration afforded 4.5 g (38%) of
compound 11: mp 176±1778C (MeOH) [lit. 175±1788C
(MeOH)];26 [a]D22ˆ149 (c 1.0, CHCl3) [lit. [a]2D4ˆ148.8
the major product were consistent with a 10-hydroxy-3-
oxaguaian-6,12-olide and it was assigned the epimeric
structure 9 with the C10 b-OH group. The most signi®cant
1
differences in the H NMR spectra of both epimers corre-
spond to the chemical shifts of H2 and H6 which are
differently in¯uenced in each epimer by the spatial proxi-
mity of the hydroxyl group at C10 which causes a deshield-
ing effect. This effect is experienced by H2 in compound 5
where this hydroxyl group is a-orientated and by H6 in
compound 9 which has the C10 hydroxyl group with b
con®guration. Thus, H2 appears at d 7.27 in compound 5
and at d 7.17 in compound 9, while H6 appears at d 4.98 in
compound 5 and at d 5.30 in 9.
(CHCl3)];26 IR (KBr) nmax 1775, 1736, 1699, 1235 cm21
;
1H NMR d 4.77 (1H, d, Jˆ10.5 Hz, H-6), 4.13 (1H, dd,
Jˆ3.3, 8.0 Hz, H-1), 2.60 (1H, td, Jˆ4.4, 13.8 Hz, H-9),

G. Blay et al. / Tetrahedron 56 (2000) 6331±6338
6335
2.48 (1H, dd, Jˆ8.0, 19.5 Hz, H-2), 2.37 (1H, dd, Jˆ3.3,
19.5 Hz, H-20), 2.30 (1H, dq, Jˆ6.6, 11.8 Hz, H-11), 2.20±
2.15 (2H, m, H-7, H-90), 2.03 (1H, dt, Jˆ4.4, 16.4 Hz, H-8),
1.97 (3H, s, OAc), 1.87 (3H, t, Jˆ1.5 Hz, H-15), 1.42 (1H,
dddd, Jˆ2.8, 13.0, 13.8, 15.4 Hz, H-80), 1.26 (3H, d,
Jˆ6.6 Hz, H-13), 1.06 (3H, s, H-14); 13C NMR d 206.5
(s, C-3), 176.8 (s, C-12), 169.9 (s, CH3CO), 160.8 (s, C-
5), 142.6 (s, C-4), 85.2 (s, C-10), 80.8 (d, C-6), 47.8 (d, C-7),
46.9 (d, C-1), 40.9 (d, C-11), 37.6 (t, C-9), 36.5 (t, C-2), 24.9
(t, C-8), 21.9 (q, CH3CO), 19.6 (q, C-14), 12.1 (q, C-13), 9.1
(q, C-15).
H-7), 2.76 (1H, m, H-1), 2.31 (1H, m, H-9), 2.22 (3H, s,
H-15), 2.16 (1H, dq, Jˆ1 0.5, 6.8 Hz, H-11), 1.89 (3H, s,
OAc overlapped with H-90), 1.81 (1H, m, H-8), 1.55 (3H, s,
H-14), 1.32 (1H, m, H-80), 1.22 (3H, d, Jˆ6.8 Hz, H-13);
13C NMR d 207.1 (s, C-4), 178.7 (s, C-12), 169.8 (s,
CH3CO), 86.5 (s, C-10), 79.1 (d, C-6), 60.8 (t, C-3), 52.3
(d, C-7), 44.5, 43.6 (d, C-1, C-5), 41.4 (C-11), 31.8, 31.2 (t,
C-2, C-9), 30.1 (q, C-15), 25.5 (q, CH3CO), 25.3 (t, C-8),
21.8 (q, C-14), 13.2 (q, C-13).
Vinyl ketone 17. (a) Via o-nitrophenylselenide 15: To a
solution of compound 13 (611 mg, 1.87 mmol) and o-nitro-
phenylselenocyanate (975 mg, 4.2 mmol) in 19 mL of 1:1
THF±pyridine was added via syringe n-Bu3P (1.4 mL,
5.62 mmol) under argon. After 1 h, the reaction mixture
was diluted with ethyl acetate (100 mL), washed with 2 M
HCl (25 mL) and brine (25 mL), and dried with MgSO4.
After ®ltration and evaporation of the solvent under reduced
pressure, the residue was chromatographed on silica gel (6:4
to 3:7 hexane±EtOAc) to give 910 mg (96%) of a yellow
solid identi®ed as compound 15: mp 96±988C (ether);
[a]1D9ˆ17 (c 0.8, CHCl3); IR (KBr) nmax 1764, 1729,
10a-Acetoxy-1,7aH,5,6,11bH-guaia-3-en-6,12-olide (12).
A solution of compound 11 (3.02, 9.85 mmol) and
p-toluenesulfonylhydrazine (2.42 g, 12.9 mmol) in 10 mL
of absolute ethanol was heated at 80±858C under argon
for 1 h. After this time the solvent was evaporated under
vacuum and the resulting yellow solid dissolved in 30 mL of
chloroform, cooled at 08C and treated with 20 mL of an 1 M
solution of catecholborane in THF (20 mmol) under argon.
The resulting mixture was stirred for 1 h at 08C and at room
temperature for two additional hours. After this time,
sodium acetate trihydrate (10 g, 73 mmol) was added and
the mixture re¯uxed under argon for 1 h. Then it was
®ltered, concentrated and the residue chromatographed on
silica gel with 8:2 hexane±EtOAc to give 2.55 g (89%) of
compound 12: mp 114±1168C (CHCl3) [lit. 116±1178C];8
[a]2D2ˆ233 (c 0.8, CHCl3) [lit. [a]D24ˆ235 (CHCl3)];8 IR
(KBr) nmax 1768, 1739 cm21; HRMS (EI) m/z 232.1456
(M12CH3COOH, 100, C15H20O2 required 232.1463), 217
1
1700 cm21; H NMR d 8.28 (1H, dd, Jˆ1.2, 8.5 Hz, Ar),
7.53 (1H, td, Jˆ8.5, 1.2 Hz, Ar), 7.44 (1H, dd, Jˆ1.2,
8.5 Hz, Ar), 7.34 (1H, td, Jˆ8.5, 1.2 Hz, Ar), 4.24 (1H,
dd, Jˆ5.5, 10.0 Hz, H-6), 3.20 (1H, t, Jˆ5.5 Hz, H-5),
3.00 (2H, m, 2H-3), 2.76 (1H, m, H-1), 2.63 (1H, m, H-7),
2.24 (3H, s, H-15), 2.09 (4H, m, H-2, H-8, H-9, H-11), 1.91
(3H, s, OAc overlapped with H-90), 1.61 (1H, m, H-20), 1.51
(3H, s, H-14), 1.30 (1H, m, H-80), 1.21 (3H, d, Jˆ7.2 Hz,
H-13); 13C NMR d 206.6 (s, C-4), 177.6 (s, C-12), 169.6 (s,
CH3CO), 146.0 (s, Ar), 133.8 (s, d, Ar), 132.2 (s, Ar), 128.8,
126.5, 125.8 (d, Ar), 86.5 (s, C-10), 79.7 (d, C-6), 52.6 (d,
C-7), 48.6, 42.7, 42.5 (d, C-1, C-5, C-11), 33.7 (t, C-3), 31.8
(q, C-15), 29.4 (t, C-9), 25.7, 24.9 (t, C-2, C-8), 23.4 (q,
CH3CO), 22.1 (q, C-14), 13.0 (q, C-13).
1
(37), 159 (31), 105 (22); H NMR d 5.34 (1H, br m, H-3),
4.24 (1H, dd, Jˆ9.0, 10.0 Hz, H-6), 3.27 (1H, dt, Jˆ9.0,
10.0 Hz, H-1), 2.97 (1H, t, Jˆ9.0, H-5), 2.44 (1H, td,
Jˆ10.0, 3.2 Hz, H-9), 2.3±2.1 (4H, m, 2H-2, H-7, H-11),
1.96 (3H, s, OAc overlapped with H-90), 1.86 (3H, s, H-15
overlapped with H-8), 1.35 (3H, s, H-14 overlapped with
H-80), 1.21 (3H, d, Jˆ6.4 Hz, H-13); 13C NMR d 178.5 (s,
C-12), 170.3 (s, CH3CO), 139.1 (s, C-4), 125.4 (d, C-3),
86.7 (s, C-10), 82.9 (d, C-6), 49.0, 48.6, 47.7 (d, C-1, C-5,
C-7), 42.3 (d, C-11), 37.7 (t, C-2), 32.4 (t, C-9), 22.4 (d,
C-8), 22.4 (q, CH3CO), 21.2 (q, C-15), 16.9 (q, C-14), 12.1
(q, C-13).
To a solution containing compound 15 (2.25 g, 4.4 mmol) in
34 mL of THF cooled at 2788C was added 5 mL of 30%
H2O2. The mixture was stirred at room temperature for 7 h
and then diluted with 200 mL of EtOAc, washed with
50 mL of 8% aqueous Na2S2O3 and 50 mL of brine.
Removal of the solvent under reduced pressure and
chromatography on silica gel (6:4 hexane±EtOAc) yielded
1.23 g (91%) of compound 17: mp 144±1468C (hexane±
EtOAc); [a]2D1ˆ264 (c 0.8, CHCl3); IR (KBr) nmax 1776,
1727 cm21; HRMS (CI) m/z 309.1701 (M111, 1.4,
C17H25O5 required 309.1702), 249 (100), 207 (48), 189
Hydroxy ketone 13. Ozone enriched oxygen was bubbled
through a solution of compound 12 (1.52 g, 5.19 mmol) in
415 mL of 1:1 MeOH±CH2Cl2 at 2788C until the appear-
ance of a blue color in the solution. The excess ozone was
removed with an argon purge. A ®vefold excess of NaBH4
(0.21 g, 5.5 mmol, total 27.5 mmol) was added at intervals
of 15 min at 2788C. After the last addition, the reaction
¯ask was transferred to an ice bath and stirred for 45 min.
The reaction was quenched with 200 mL of std aqueous
NH4Cl. The organic layer was separated and the aqueous
layer concentrated to half volume and extracted twice with
dichloromethane (2£100 mL). The combined organic layers
were dried, concentrated and chromatographed on silica gel
(1:1 hexane±EtOAc) to give 1.15 g (68%) of compound 13:
mp 141±1438C (hexane±EtOAc); [a]1D9ˆ253 (c 1.0,
1
(15); H NMR d 5.69 (1H, ddd, Jˆ10.0, 10.2, 16.5 Hz, H-
2), 5.04 (1H, dd, Jˆ1.5, 10.2 Hz, H-3), 4.96 (1H, dd, Jˆ1.5,
16.5 Hz, H-30), 4.28 (1H, dd, Jˆ7.8, 10.6 Hz, H-6), 3.63
(1H, t, Jˆ10.0 Hz, H-1), 3.28 (1H, dd, Jˆ7.8, 10.0 Hz,
H-5), 2.74 (1H, td, Jˆ11.5, 4.1 Hz, H-9), 2.47 (1H, qd,
Jˆ10.6, 1.9 Hz, H-7), 2.15 (3H, s, H-15), 2.10 (1H, dq,
Jˆ7.2, 10.5 Hz, H-11), 1.95 (2H, m, H-8, H-90), 1.85
(3H, s, OAc), 1.38, (3H, s, H-14), 1.26 (1H, m, H-80),
1.21 (3H, d, Jˆ7.2 Hz, H-13); 13C NMR d 207.0 (s, C-4),
177.3 (s, C-12), 169.7 (s, CH3CO), 135.0 (d, C-2), 118.4 (t,
C-3), 85.7 (s, C-10), 81.8 (d, C-6), 52.1, 50.9 (d, C-1, C-7),
44.4 (d, C-5), 41.6 (d, C-11), 37.2 (t, C-9), 34.3 (q, C-15),
26.9 (t, C-8), 22.4 (q, CH3CO), 21.7 (q, C-14), 12.8 (q,
C-13).
CHCl3); IR (KBr) nmax 3496, 1765, 1739, 1695 cm21
;
HRMS (CI) m/z 327.1802 (M111, 0.6, C17H27O6 required
327.1808), 309 (29), 267 (100), 249 (34), 207 (36); 1H NMR
d 4.25 (1H, dd, Jˆ4.2, 10.5 Hz, H-6), 3.79 (2H, m, 2H-3),
3.19 (1H, t, Jˆ4.2 Hz, H-5), 2.87 (1H, dq, Jˆ5.5, 10.5 Hz,

6336
G. Blay et al. / Tetrahedron 56 (2000) 6331±6338
(b) Via phenylselenide 16: A solution of compound 13
(79 mg, 0.24 mmol) and triethylamine (48 mL, 0.35 mmol)
in 1.2 mL of CH2Cl2 at 08C was treated with mesyl
chloride (22 mL, 0.28 mmol). After 85 min, the reaction
mixture was diluted with 60 mL of EtOAc, washed with
20 mL of 2 M HCl and 20 mL of brine. Chromato-
graphy on silica gel with 8:2 hexane±EtOAc yielded
88 mg (90%) of mesylate 14: 1H NMR d 4.35 (1H,
dd, Jˆ5.9, 10.0 Hz, H-6), 4.20 (2H, m, 2H-3), 3.17
(1H, t, Jˆ5.9 Hz, H-5), 3.00 (3H, s, OMs), 2.6±2.4
(2H, m, H-1, H-7), 2.2±2.0 (4H, m, H-2, H-8, H-9,
H-11) 2.25 (3H, s, H-15), 1.91 (3H, s, OAc overlapped
with H-90), 1.51 (3H, s, H-14 overlapped with H-20),
1.25 (1H, m, H-80), 1.17 (3H, d, Jˆ6.9 Hz, H-13).
activated, weakly acidic, in 2 mL of benzene was stirred
at room temperature under argon for 6 h. The solvent was
removed and the resulting powder was chromatographed on
silica gel to give 3.3 mg (16%) of starting material and
10.6 mg (66%) of a product identi®ed as 20: mp 79±818C
(EtOAc); [a]2D5ˆ2200 (c 1.0, CHCl3); IR (KBr) nmax 1784,
1707, 1661 cm21; HRMS (EI) m/z 250.1201 (M1, 3,
C14H18O4 required 250.1205), 208 (68), 190 (73), 162
1
(57), 134 (100); H NMR d 10.07 (1H, s, H-2), 4.96 (1H,
d, Jˆ4.0 Hz, H-5), 3.85 (1H, dd, Jˆ4.0, 10.8 Hz, H-6), 2.84
(1H, td, Jˆ13.5, 2.0 Hz, H-9), 2.44 (1H, qd, Jˆ10.8, 3.2 Hz,
H-7), 2.33 (3H, s, H-15), 2.20 (3H, s, H-14 overlapped with
H-90 and H-11), 2.00 (1H, m, H-8), 1.31 (1H, m, H-80), 1.21
(3H, d, Jˆ6.9 Hz, H-13), 13C NMR d 205.6 (s, C-4), 189.2
(d, C-2), 177.1 (s, C-12), 165.4 (s, C-1), 133.0 (s, C-10),
81.7 (d, C-6), 48.8, 48.3 (d, C-5, C-7), 42.2 (d, C-11), 37.1
(t, C-9), 32.2 (q, C-15), 26.9 (t, C-8), 22.1 (q, C-14), 12.2 (q,
C-13).
To a mixture of NaBH4 (9.4 mg, 0.25 mmol) and diphenyl-
diselenide (45 mg, 0.14 mmol) under argon was added
0.5 mL of DMF via syringe. A rapid evolution of hydrogen
took place and the mixture was decolorized. After evolution
of hydrogen ceased, a solution of mesylate 14 (23 mg,
0.058 mmol) in 1 mL of DMF was injected and the mixture
stirred for 2 h at room temperature. The reaction was
quenched with water (10 mL) and extracted with EtOAc
(3£20 mL). Chromatography of the extract allowed to
obtain 24 mg (89%) of phenylselenide 16. 1H NMR
(300 MHz) d 7.50 (2H, m, Ar), 7.26 (3H, m, Ar), 4.17
(1H, dd, Jˆ5.3, 10.4 Hz, H-6), 3.09 (1H, t, Jˆ5.3 Hz,
H-5), 3.1±2.9 (3H, m, H-1, 2H-3), 2.8±2.6 (2H, m, H-7,
H-9), 2.15 (3H, s, H-15 overlapped with H-2, H-8, H-11),
1.86 (3H, s, OAc overlapped with H-90), 1.62 (1H, m, H-20),
1.46 (3H, s, H-14), 1.21 (4H, d, Jˆ6.9 Hz, H-13, overlapped
with H-80).
By the same procedure using aluminum oxide, activated,
basic, Brockmann I, compound 20 was obtained in 66%
yield.
By the same procedure using silica gel at benzene re¯ux
temperature, compound 20 was obtained in 46±67%
yield.
Hemiacetal 21. To a solution of compound 18 (150 mg,
0.48 mmol) in 14 mL of dry THF cooled at 08C under
argon was added LiAlH4 (23 mg, 0.60 mmol) portionwise.
The reaction mixture was stirred for 3 h and quenched with
std aqueous NH4Cl (5 mL) at this temperature. The tempera-
ture was allowed to rise to room temperature, 2 M HCl
(5 mL) added and the mixture extracted with EtOAc
(3£25 mL) and washed with brine (25 mL). Removal of
the solvent under reduced pressure followed by chroma-
tography (4:6 hexane±EtOAc) afforded 144 mg (95%) of
compound 21: mp 134±1368C (hexane±EtOAc); [a]2D5ˆ
287 (c 1.0, CHCl3); IR (KBr) nmax 3471, 1772,
1729 cm21; HRMS (EI) m/z 297.1347 (M12CH3, 7,
C15H21O6 required 297.1338), 209 (34), 192 (44), 164
Compound 16 (55 mg, 0.12) in THF was treated with 30%
H2O2 for 16 h as described above for compound 15 to give
33 mg (90%) of 17.
Dicarbonyl compound 18. A solution of compound 17
(1.10 g, 3.58 mmol) in 95 mL of dry dichloromethane was
treated with ozone as described in the synthesis of 13. The
resulting solution was treated with Ph3P (1.22 g,
5.66 mmol), the temperature was allowed to rise to room
temperature and stirred for 5 h. After this time the solvent
was removed and the residue chromatographed (6:4 to 4:6
hexane±EtOAc) to give 1.08 g (97%) of compound 18 as a
white solid: mp 136±1388C (CH2Cl2±hexane); [a]2D1ˆ
2197 (c 1.2, CHCl3); HRMS (CI) m/z 311.1491 (M111,
95, C16H23O6 required 311.1495), 251 (100), 233 (86), 209
1
(100), 119 (77); H NMR d 4.18 (1H, t, Jˆ10.4 Hz, H-6),
4.00 (1H, t, Jˆ7.5 Hz, H-2), 3.67 (1H, dd, Jˆ7.5, 10.5 Hz,
H-20), 3.55 (1H, ddd, Jˆ7.5, 10.5, 12.5 Hz, H-1), 2.63 (1H,
br s, OH); 2.38 (1H, dt, Jˆ13.5, 4.0 Hz, H-9), 2.27 (2H, dd,
Jˆ10.0, 12.5 Hz, H-5 overlapped with H-7), 2.19 (1H, td,
Jˆ13.5, 4.0 Hz, H-9), 2.11 (1H, dq, Jˆ11.5, 7.0 Hz, H-11),
1.93 (3H, s, OAc, overlapped with H-8), 1.65 (3H, s, H-15),
1.38 (3H, s, H-14), 1.24 (1H, m, H-80), 1.21 (3H, d,
Jˆ7.0 Hz, H-13); 13C NMR d 177.9 (s, C-12), 170.2 (s,
CH3CO), 105.0 (s, C-4), 84.8 (s, C-10), 80.9 (d, C-6),
67.6 (t, C-2), 48.1, 47.2, 47.1 (d, C-1, C-5, C-7), 41.7 (d,
C-11), 38.2 (t, C-9), 27.5 (q, C-14), 26.0 (t, C-8), 22.3 (q,
CH3CO), 18.6 (q, C-15), 12.4 (q, C-13).
(62), 191 (76); IR (KBr) nmax 1777, 1760, 1720, 1717 cm21
;
1H NMR d 9.80 (1H, s, H-2), 4.29 (1H, m, H-6), 3.89 (1H, s,
H-1), 3.88 (1H, d, Jˆ2.0 Hz, H-5), 2.50 (1H, dt, Jˆ4.5,
14.0 Hz, H-9), 2.32 (3H, s, H-15 overlapped with H-90),
2.19 (1H, qd, Jˆ10.5, 4.0 Hz, H-7), 2.12 (1H, dq, Jˆ6.5,
10.5 Hz, H-11), 2.01 (4H, s, OAc overlapped with H-8),
1.50 (3H, s, H-14), 1.27 (1H, m, H-80), 1.18 (3H, d,
Jˆ6.5 Hz, H-13); 13C NMR d 207.3 (s, C-4), 201.6 (s,
C-2), 176.8 (s, C-12), 169.8 (s, CH3CO), 86.4 (s, C-10),
80.7 (d, C-6), 58.6 (d, C-1), 46.4 (d, C-5), 44.6 (d, C-7),
41.4 (d, C-11), 38.2 (t, C-9), 33.5 (q, C-15), 26.4 (t, C-8),
22.3 (q, CH3CO), 21.5 (q, C-14), 12.7 (q, C-13).
Dihydrofuran 23. A solution of compound 18 (148 mg,
0.47 mmol) and MsCl (357 mL, 4.60 mmol) in 10 mL of
pyridine was stirred at 08C for 1 h. After this time the
mixture was diluted with 75 mL of EtOAc, washed with
25 mL of 2 M HCl and 25 mL of brine, and dried. Solvent
removal and chromatography with 1:1 hexane±EtOAc gave
176 mg (96%) of mesylate 22: [a]2D8ˆ290 (c 0.8, CHCl3);
Dicarbonyl compound 20. A suspension of compound 18
(20 mg, 0.065 mmol) and 188 mg of aluminum oxide,
1
IR (KBr) nmax 1765, 1722 cm21, H NMR (250 MHz) d

G. Blay et al. / Tetrahedron 56 (2000) 6331±6338
6337
4.5±4.1 (3H, m, H-6, 2H-2), 3.62 (1H, t, Jˆ8.0 Hz, H-5),
3.17 (1H, m, H-1), 2.98 (3H, s, OMs), 2.41 (1H, dt, Jˆ13.5,
4.0 Hz, H-9), 2.34 (3H, s, H-15), 2.3±1.9 (4H, m, H-7, H-11,
H-8, H-90), 1.99 (3H, s, OAc), 1.49 (3H, s, H-14), 1.25 (1H,
m, H-80), 1.17 (3H, d, Jˆ6.6 Hz, H-13); 13C NMR d 208.0
(s, C-4), 176.9 (s, C-12), 169.8 (s, CH3CO), 85.5 (s, C-10),
81.1 (d, C-6), 69.6 (t, C-2), 49.8 (d, C-7), 46.3, 44.2 (d, C-1,
C-5), 41.3 (d, C-11), 38.2 (t, C-9), 37.3 (q, CH3SO2), 33.9
(q, C-15), 26.4 (t, C-8), 22.4 (q, CH3CO), 21.6 (q, C-14),
12.7 (q, C-13).
TPAP (1.8 mg, 0.005 mmol) were added. The reaction
mixture was stirred at room temperature for 30 min and
then chromatographed on silica gel with 1:1 hexane±
EtOAc to give 10.7 mg (75%) of compound 5: mp 125±
1278C (hexane±EtOAc); [a]2D8ˆ224 (c 0.3, CHCl3); IR
(KBr) nmax 3473, 1770 cm21; HRMS (EI) m/z 250.1212
(M1, 100, C14H18O4 required 250.1205), 235 (86), 162
1
(89), 150 (58), 109 (28); H NMR d 7.27 (1H, s, H-2),
4.98 (1H, dq, Jˆ1.2, 10.4 Hz, H-6), 2.36 (3H, d, Jˆ
1.2 Hz, H-15), 2.31 (1H, dq, Jˆ11.0, 6.8 Hz, H-11),
2.11 (1H, m, H-9), 2.00 (1H, m, H-8), 1.86 (1H, qd,
Jˆ10.4, 2.2 Hz, H-7), 1.81 (1H, td, Jˆ12.8, 2.4, H-90),
1.69 (1H, m, H-80), 1.45 (3H, s, H-14), 1.25 (3H, d,
Jˆ6.8 Hz, H-13); 13C NMR d 178.0 (s, C-12), 147.6 (s,
C-4), 136.4 (d, C-2), 132.2 (s, C-1), 114.1 (s, C-5), 79.5
(d, C-6), 72.3 (s, C-10), 52.4 (d, C-7), 42.5 (t, C-9),
41.2 (d, C-11), 28.6 (q, C-14), 26.4 (t, C-8), 13.4 (q,
C-15), 12.5 (q, C-13).
A solution of the above mesylate (170 mg, 0.44 mmol) and
DBU (68 mL, 0.45 mmol) in 45 mL of benzene was heated
at re¯ux for 80 min. The usual work up and chromatography
on silica gel with 1:1 hexane±EtOAc gave 113 mg (88%) of
compound 23: mp 134±1368C (hexane±EtOAc); [a]2D9ˆ
171 (c 1.2, CHCl3); HRMS (EI) m/z294.1482 (M1, 2,
C16H22O5 required 294.1467), 234 (100), 219 (27), 190
1
(83), 161 (24); IR (KBr) nmax 1764, 1722 cm21; H NMR
d 4.61 (1H, d, Jˆ10.8 Hz, H-6), 4.33 (2H, m, 2H-2), 3.98,
(1H, t, Jˆ8.4 Hz, H-1), 2.39 (1H, dt, Jˆ13.5, 4.0 Hz, H-9),
2.20 (1H, dq, Jˆ11.5, 6.8 Hz, H-11), 2.1±1.8 (3H, m, H-7,
H-8, H-9), 1.97 (3H, s, OAc), 1.89 (3H, s, H-15), 1.31 (3H,
s, H-14), 1.26 (1H, m, H-8), 1.22 (3H, d, Jˆ6.8 Hz, H-13
overlapped with- H-8); 13C NMR d 178.4 (s, C-12), 170.3
(s, CH3CO), 158.3 (s, C-4), 101.6 (s, C-5), 86.9 (s, C-10),
79.7 (d, C-6), 70.8 (t, C-2), 52.6, 49.8 (d, C-1, C-7), 41.8 (d,
C-11), 37.9 (t, C-9), 24.1 (t, C-8), 22.5 (q, CH3CO), 18.7 (q,
C-14), 12.7, 12.3 (q, C-13, C-15).
Hemiacetal 25. By the same procedure used in the synthesis
of 21, compound 20 (14 mg, 0.06 mmol) gave 11.4 mg
(81%) of compound 25: mp 127±1288C (EtOAc); [a]2D5ˆ
2143 (c 0.9, CHCl3); IR (KBr) nmax 3494, 1759 cm21
;
HRMS (EI) m/z 253.1448 (M111, 2, C14H21O4 required
1
253.1440), 192 (100), 119 (99), 107 (58); H NMR d 4.69
(1H, dd, Jˆ8.5, 10.5 Hz, H-6), 4.49 (1H, dd, Jˆ1.5,
12.5 Hz, H-2), 4.32 (1H, dd, Jˆ1.5, 12.5 Hz, H-20), 3.25
(1H, d, Jˆ10.5 Hz, H-5), 2.62 (1H, m, H-7), 2.34 (1H, m,
H-9), 2.23 (1H, dq, Jˆ11.0, 7.0 Hz, H-11), 2.10 (1H, qd,
Jˆ12.0, 4.8 Hz, H-8), 1.82 (1H, ddd, Jˆ1.5, 6.0, 14.0 Hz,
H-90), 1.66 (3H, s, H-15), 1.63 (3H, d, Jˆ1.5 Hz, H-14),
1.35 (1H, m, H-80), 1.16 (3H, d, Jˆ7.0 Hz, H-13); 13C
NMR d 179.3 (s, C-12), 132.9 (s, C-10), 124.7 (s, C-1),
106.5 (s, C-4), 77.6 (d, C-6), 68.5 (t, C-2), 51.4 (d, C-7),
42.4, 42.3 (d, C-11, C-5), 30.2, 28.4 (t, C-8, C-9), 27.4 (q, C-
14), 21.1 (q, C-15), 13.4 (q, C-13).
Acetoxyfuran 24. Oxygen was gently bubbled through a
solution of compound 23 (26 mg, 0.09 mmol) and methyl-
ene blue (2.4 mg) in absolute ethanol. The reaction tube was
submerged in an ice bath and irradiated by two lamps
(MAZDA, 400 W each) for 2 h. After this time, the solvent
was removed under vacuum and the residue adsorbed on
silica gel (500 mg). The resulting powder was placed on
the top of a chromatography column for 3 h and the products
eluted with 8:2 to 1:1 hexane±EtOAc mixtures to give
9.5 mg (37%) of compound 24 and 9 mg (35%) of starting
material. Compound 24: an oil, [a]2D6ˆ234 (c 1.0, CHCl3);
HRMS (EI) m/z 292.1316 (M1, 80, C16H20O5 required
292.1311), 250 (28), 232 (M12CH3COOH, 100), 217
Hydroxyfuran 9. To a solution of compound 25 (34 mg,
0.14 mmol) in 1.3 mL of THF and 0.3 mL of pyridine
cooled at 08C was added OsO4 (68 mg, 0.27 mmol) and
the reaction mixture stirred at this temperature for 3 h.
After this time, a mixture containing 12 mg of NaHSO3,
1.9 mL of pyridine and 1.6 mL of water was added and
stirring continued for 1 h at 08C and for two additional
hours at room temperature. Then, 2 M HCl (15 mL) was
added and after 5 min the mixture extracted with CH2Cl2
(3£20 mL). The organic layer was washed with brine
(20 mL), dried, and concentrated under reduced pressure.
Chromatography of the residue with 7:3 to 1:1 mixtures of
hexane±EtOAc eluted 18.5 mg (55%) of compound 9 and
4.5 mg (14%) of compound 5. Compound 9: mp 110±1128C
(hexane±EtOAc); [a]D27ˆ234 (c 1.6, CHCl3); HRMS (EI)
m/z 250.1212 (M1, 41, C14H18O4 required 250.1205), 232
(100), 217 (60), 162 (51), 150(38); IR (KBr) nmax 3527,
1
(29), 176 (22); IR (NaCl) nmax 1787, 1766, 1728 cm21; H
NMR d 7.24 (1H, s, H-2), 4.97 (1H, dq, Jˆ10.5, 1.0 Hz, H-
6), 2.36 (3H, d, Jˆ1.0 Hz, H-15), 2.27 (1H, dq, Jˆ11.0,
7.0 Hz, H-11), 2.20 (1H, m, H-9), 2.1±1.9 (3H, m, H-7,
H-8, H-90), 2.04 (3H, s, OAc), 1.75 (3H, s, H-14), 1.57
(1H, m, H-80), 1.23 (3H, d, Jˆ7.0 Hz, H-13); 13C NMR d
178.1 (s, C-12), 169.9 (s, CH3CO), 148.2 (s, C-4), 137.9 (d,
C-2), 128.5 (s, C-1), 114.5 (s, C-5), 81.5 (s, C-10), 78.5 (d,
C-6), 51.0 (d, C-7), 41.6 (d, C-11), 36.1 (t, C-9), 26.2 (q, C-
14), 25.6 (t, C-8), 22.4 (q, CH3CO), 13.5 (q, C-15), 12.6 (q,
C-13).
1
1765 cm21; H NMR d 7.17 (1H, s, H-2), 5.30 (1H, qd,
Hydroxyfuran 5. A solution of compound 24 (16.5 mg,
0.057 mmol) in 1.8 mL of dry THF was cooled at 08C and
treated with 60 mL (0.19 mmol) of a 65% solution of
Red-Alw in toluene under argon. After 4 h, the reaction
mixture was diluted with EtOAc (75 mL), washed with
brine (15 mL) and dried with MgSO4, and the solvent
removed under reduced pressure. The resulting oil was
diluted in CH2Cl2 and NMO (15.5 mg, 0.13 mmol) and
Jˆ1.5, 10.5 Hz, H-6), 2.38 (3H, d, Jˆ1.5 Hz, H-15), 2.34
(1H, dq, Jˆ10.5, 7.2 Hz, H-11), 2.05 (1H, m, H-9), 1.5±1.9
(3H, m, H-7, H-8, H-90), 1.63 (4H, s, H-14 overlapped with
H-80), 1.26 (3H, d, Jˆ7.2 Hz, H-13); 13C NMR d 178.6 (s,
C-12), 147.0 (s, C-4), 136.3 (d, C-2), 130.0 (s, C-1), 116.8
(s, C-5), 79.5 (d, C-6), 69.2 (s, C-10), 52.7 (d, C-7), 41.4 (d,
C-11), 40.0 (t, C-9), 30.1 (q, C-14), 25.4 (t, C-8), 13.3 (q,
C-15), 12.5 (q, C-13).

6338
G. Blay et al. / Tetrahedron 56 (2000) 6331±6338
11. Miyashita, M.; Suzuki, T.; Yoshikoshi, A. J. Am. Chem. Soc.
1989, 111, 3728±3734.
Acknowledgements
Financial support from European Commission (FAIR
Â
12. Blay, G.; Cardona, L.; Garcõa, B.; Pedro, J. R. J. Org. Chem.
1993, 58, 7204±7208.
Â
CT96-1781) and from Direccion General de Investigacion
Cientõ®ca y Tecnica (PB 94-0985) is grateful acknowl-
Â
Â
Â
13. Patel, D. V.; VanMiddlesworth, F.; Donaubauer, J.; Gannett,
P.; Sih, C. J. J. Am. Chem. Soc. 1986, 106, 4603±4614.
14. Brunnelle, W. H.; Rafferty, M. A.; Hodges, S. L. J. Org. Chem.
1987, 52, 1603±1605.
edged. One of us (L. L.) is especially thankful to Conselleria
 Á
de Cultura, Educacio i Ciencia (Generalitat Valenciana) for
a grant.
15. Aurell, M. J.; Ceita, L.; Mestres, R.; Tortajada, A. Tetrahedron
1997, 53, 10883±10898.
References
16. Lorenz, O.; Parks, C. R. J. Org. Chem. 1965, 30, 1976±1981.
17. Joule, J. A.; Mills, K.; Smith, G. F. Heterocyclic Chemistry,
3rd ed.; Chapman & Hall: London, 1995; pp 290±300.
18. Cribble, G. W.; Ferguson, D. C. J. Chem. Soc. Chem.
Commun. 1975, 535±536.
1. Lipshutz, B. H. Chem. Rev. 1986, 86, 795±819.
2. Nakanishi, K. Natural Products Chemistry, Kondasha Ltd:
Tokyo, 1974.
3. Hou, X. L.; Cheung, H. Y.; Hon, T. Y.; Kwan, P. L.; Lo, T. H.;
Tong, S. Y.; Wong, H. N. C. Tetrahedron 1998, 54, 1955±2020.
4. Kastner, U.; Jurenitsch, J.; Baumann, A.; Robien, W.; Kubelka,
W. Pharm. Pharmacol. Lett. 1991, 1, 55±56.
5. Kastner, U.; Glasl, S.; Jurenitsch, J.; Baumann, A.; Robien, W.;
Kubelka, W. Sci. Pharm. 1992, 60, 163.
19. Evans, D. L.; Minster, D. K.; Jordis, U.; Hecht, S. M.; Mazzu,
A. L.; Meyers, A. I. J. Org. Chem. 1979, 44, 497±501.
20. Wynberg, H. J. Am. Chem. Soc. 1958, 80, 364±366.
21. Hae¯iger, W.; Kloppner, E. Helv. Chim. Acta 1982, 65, 1837±
1852.
22. Grif®th, W. P.; Ley, S. V.; Whitcombe, G. P.; White, A. D.
J. Chem. Soc. Chem. Commun. 1987, 1625±1627.
23. Ando, M.; Isogai, K.; Azami, H.; Hirato, N.; Yanagi, Y. J. Nat.
Prod. 1991, 54, 1017±1024.
6. Ochir, G.; Budesinsky, M.; Motl, O. Phytochemistry 1991, 30,
4163±4165.
7. Schroeder, H.; Kastner, U.; Gargula, K.; Budesinsky, M.;
Haslinger, E.; Jurenitsch, J.; Kubelka, W. Phytochemistry 1994,
36, 1449±1451.
24. Abraham, R. J.; Fisher, J.; Loftus, P. Introduction to NMR
Spectroscopy, Wiley: Chichester, 1988; pp 22±23.
25. Perrin, D. D.; Armarego, W. L. F. Puri®cation of Laboratory
Chemicals, 3rd ed.; Pergamon: Oxford, 1988.
8. Greene, A. E. J. Am. Chem. Soc. 1980, 102, 5337±5343.
9. Grieco, P. A.; Gilman, S.; Nishizawa, M. J. Org. Chem. 1976,
41, 1485±1486.
26. Barton, D. H. R.; De Mayo, P.; Sha®q, M. J. Chem. Soc. 1957,
929±935.
Â
10. Blay, G.; Cardona, L.; Garcõa, B.; Pedro, J. R. Can. J. Chem.
1991, 70, 817±822.