Syntheses of theaspirone and vitispirane via palladium(II)-catalyzed oxaspirocyclization

Article abstract of DOI:10.1021/jo9505031

Total syntheses of theaspirone (A and B) and vitispirane (A and B) are described. The key step in the syntheses is the palladium(II)-catalyzed intramolecular oxaspirocyclization of diene alcohol 4 to either vitispirane or the allylic alcohol 9. The outcome of the oxaspirocyclization is very much dependent on the solvent employed. In water-acetic acid (4:1) a 1:1 mixture of the diastereomeric alcohols 9A and 9B was exclusively formed. In water with 8 equiv of a strong non-nucleophilic acid, vitispiranes A and B (1:1) were obtained. An alternative procedure to obtain vitispirane with the use of LiCl and K₂CO₃ is described. In the latter reaction vitispirane B is formed preferentially. This result is explained by an equilibrium between the two possible π-allyl complexes 5A and 5B, the kinetically favored 5B being transformed into vitispirane 3B before isomerization to 5A occurs.

Full text of DOI:10.1021/jo9505031

J . Org. Chem. 1996, 61, 1825-1829
1825
Syn th eses of Th ea sp ir on e a n d Vitisp ir a n e via
P a lla d iu m (II)-Ca ta lyzed Oxa sp ir ocycliza tion
†
†
†
,†
Ylva I. M. Nilsson, Attila Aranyos, Pher G. Andersson, J an-E. B a¨ ckvall,*
‡
‡
,‡
J ean-Luc Parrain, Christelle Ploteau, and J ean-Paul Quintard*
Department of Organic Chemistry, University of Uppsala, Box 531, S-751 21 Uppsala, Sweden, and
Laboratoire de Synth e` se Organique associ e´ au CNRS, Facult e´ des Sciences et des Techniques de Nantes,
2
rue de la Houssini e` re, 44072 Nantes, Cedex 03, France
Received March 15, 1995X
Total syntheses of theaspirone (A and B) and vitispirane (A and B) are described. The key step in
the syntheses is the palladium(II)-catalyzed intramolecular oxaspirocyclization of diene alcohol 4
to either vitispirane or the allylic alcohol 9. The outcome of the oxaspirocyclization is very much
dependent on the solvent employed. In water-acetic acid (4:1) a 1:1 mixture of the diastereomeric
alcohols 9A and 9B was exclusively formed. In water with 8 equiv of a strong non-nucleophilic
acid, vitispiranes A and B (1:1) were obtained. An alternative procedure to obtain vitispirane with
the use of LiCl and K CO is described. In the latter reaction vitispirane B is formed preferentially.
2 3
This result is explained by an equilibrium between the two possible π-allyl complexes 5A and 5B,
the kinetically favored 5B being transformed into vitispirane 3B before isomerization to 5A occurs.
In tr od u ction
Theaspirone (1) was the first spiro ether isolated from
black tea.1 Oak woods used in wine and spirit matura-
tion as well as yellow passion fruit are other sources for
theaspirone.2 In nature, it occurs as a mixture of
isomers, A and B. Theaspirone A has a sweet, tea-like
3
odor while theaspirone B has a more earthy smell.
Theaspirane (2), found for the first time as a naturally
occurring substance in raspberry,4 is also present in
yellow passion fruit. Vitispirane (3) is a structurally
related compound identified among the volatiles of grape
5
juice or distilled white wines. It also occurs in quince
fruit and vanilla aroma.5 Vitispirane together with
,6
theaspirone and theaspirane belong to a class of nor-
isoprenoid spiro ethers existing in nature in both the A
F igu r e 1.
and B forms (Figure 1). Ever since their isolation these
compounds have been popular synthetic targets.4,6,7
palladium(II), for the formation of [5,6], [5,7], [6,6], and
6,7] spiro ethers (III).9 The first step in the catalytic
[
The palladium-catalyzed intramolecular 1,4-addition
of nucleophiles to conjugated dienes is a useful method
for constructing heterocyclic systems.8 We have recently
developed an oxaspirocyclization of dienes I catalyzed by
cycle involves a regioselective oxypalladation of the
trisubstituted double bond in I to give spirocyclic (π-allyl)-
palladium complex II as an intermediate. This complex
undergoes a p-benzoquinone-induced attack by a second
-
-
nucleophile (AcO , Cl ) resulting in an overall 1,4-
oxidation of the diene (Scheme 1).
†
University of Uppsala.
Facult e´ des Sciences et des Techniques de Nantes.
Abstract published in Advance ACS Abstracts, February 1, 1996.
‡
X
In a previous study9b we made preliminary efforts to
apply the palladium(II)-catalyzed oxaspirocyclization
toward synthesis of theaspirone. Unfortunately, we were
not able to obtain a catalytic reaction for the 1,4-oxidation
of 4 to give a spirocycle. However, we did some studies
on the stoichiometric reaction using the isolated (π-allyl)-
palladium complexes 5A and 5B, which were produced
in good diastereoselectivity. In a separate reaction they
were transformed to 6 (A:B, 93:7) (Scheme 2).
(
(
1) Ina, K.; Sakato, Y.; Fukami, H. Tetrahedron Lett. 1968, 2777.
2) Sefton, M. A.; Francis, I. L.; Williams, P. J . J . Agric. Food Chem.
1
990, 38, 2045.
3) (a) Nakatani, Y.; Yamanishi, T. Tetrahedron Lett. 1969, 1995.
b) Weyerstahl, P.; Meisel, T. Liebigs Ann. Chem. 1994, 415.
(
(
(
(
4) Winter, M.; Enggist, P. Helv. Chim. Acta 1971, 54, 1891.
5) (a) Simpson, R. F.; Strauss, C. R.; Williams, P. J . Chem. Ind.
1
977, 663. Wines: (b) Strauss, C. R.; Wilson, B.; Anderson, R.;
Williams, P. J . Am. J . Enol. Vitic. 1987, 38, 23. (c) Winterhalter, P.;
Sefton, M. A., Williams, P. J . Am. J . Enol. Vitic. 1990, 41, 277. Quince
fruit: (d) Tsuneya, T.; Ishihara, M.; Shiota, H.; Shiga, M. Agric. Biol.
Chem. 1983, 47, 2495.
In the present study we have developed a catalytic
reaction for the formation of the theaspirane skeleton and
(
6) Vanilla: Schulte-Elte, K. H.; Gautschi, F.; Renold, W.; Hauser,
A.; Frankhauser, P.; Limacher, J .; Ohloff, G. Helv. Chim. Acta 1978,
1, 1125.
7) (a) Masuda, H.; Mihara, S. Agric. Biol. Chem. 1985, 49, 861. (b)
Bellas, T. E.; Brownlee, R. G.; Silverstein, R. M. Tetrahedron 1974,
6
(
(8) (a) B a¨ ckvall, J . E.; Anderson, P. G.; Vågberg, J . O. Tetrahedron
Lett. 1989, 30, 137. (b) B a¨ ckvall, J . E.; Granberg, K. L.; Andersson, P.
G.; Gatti, R.; Gogoll, A. J . Org. Chem. 1993, 58, 5445. (c) B a¨ ckvall J .
E.; Andersson, P. G. J . Am. Chem. Soc. 1990, 112, 3683; (d) 1992, 114,
6374. (e) B a¨ ckvall, J . E.; Andersson, P. G.; Stone, G. B.; Gogoll, A. J .
Org. Chem. 1991, 56, 2988.
(9) (a) B a¨ ckvall, J . E.; Andersson, P. G. J . Org. Chem. 1991, 56, 2274.
(b) Andersson, P. G.; Nilsson, Y. I. M.; B a¨ ckvall, J . E. Tetrahedron
1994, 50, 559.
3
2
0, 2267. (c) Heckman, R. A.; Roberts, D. L. Tetrahedron Lett. 1969,
701. (d) Sato A.; Mishima, H. Tetrahedron Lett. 1969, 1803. (e) Weiss,
G.; Koreeda, M.; Nakanishi, K. J . Chem. Soc., Chem. Commun. 1973,
65. (f) Etoh, H.; Ina, K.; Iguchi, M. Agric. Biol. Chem. 1980, 44, 2871.
g) K o¨ nst, W. M. B.; Appeldorn, W.; Boelens, H. Synth. Commun. 1980,
0, 899. (h) Marx, J . N. Tetrahedron 1975, 31, 1251. (i) Kato, T.; Kondo,
H. Bull. Chem. Soc. J pn. 1981, 54, 1573.
5
(
1
0
022-3263/96/1961-1825$12.00/0 © 1996 American Chemical Society

1
826 J . Org. Chem., Vol. 61, No. 5, 1996
Nilsson et al.
Sch em e 1. P a lla d iu m (II)-Ca ta lyzed
Oxa sp ir ocycliza tion .
Sch em e 3a
a
(
a) NBS, hν, CCl4. (b) Na2CO3, DMF (77% from â-ionone). (c)
Sch em e 2. Stoich iom etr ic Oxa sp ir ocycliza tion via
Bu3SnH, Pd(PPh3)4 (3-5 mol %) THF/H2O/NH4Cl, 2 h, rt (90%).
(d) NaBH , 0 °C (95%).
th e Isola ted (π-Allyl)P a lla d iu m Com p lex
4
(
BQ ) p-Ben zoqu in on e).
in acetone-acetic acid, 4:1, room temperature) with slow
addition of the diene12 was unsuccessful. Since it was
demonstrated that oxypalladation and nucleophilic attack
on the π-allyl complex 5 (Scheme 2) did take place in
separate stoichiometric reactions,9b although in different
solvents, further studies toward developing a catalytic
reaction were called for. A change in solvent from
acetone-acetic acid to THF-acetic acid did not improve
the reaction. Pure acetic acid (using 10% of the catalyst)
yielded after 24 h 25-30% of the desired cyclized acetate
13
6
together with several side products and the Diels-
Alder adduct between benzoquinone and diene 4.
Some other oxidation systems such as Cu(OAc) /O
CuCl /O were tried in an effort to develop a catalytic
reaction. In most of these reactions, mixtures of 3 and 6
were obtained from 4 in low yield (eq 1). With CuCl /O
2
2
and
2
2
2
2
as the oxidant a prolonged reaction time increased the
selectivity for 3. In all these reactions both the A and B
isomers of the products were obtained and no selectivity
applied it to the total syntheses of theaspirone and
vitispirane. With small variations of the reaction condi-
tions in this new palladium-catalyzed reaction, either
theaspirone or vitispirane was obtained. We have also
discussed the kinetic versus thermodynamic control in
the formation of the intermediate π-allyl complex, which
will determine whether isomer A or B is formed in the
catalytic reaction.
-
in the addition of AcO was achieved (eq 1).
Resu lts a n d Discu ssion
Discouraged by these results we looked into the ben-
zoquinone reaction in more detail. In this reaction we
had observed that alcohol 9 was formed when LiOAc‚
P r ep a r a tion of Sta r tin g Ma ter ia ls. The starting
material for the oxaspirocyclization was prepared ac-
cording to ref 9b with some modifications. An improved
method for the selective reduction of 7 was achieved by
2
2
H O was employed, showing that also water is capable
of attacking the π-allyl intermediate. We therefore
increased the amount of water, and reaction in a 4:1
mixture of water-acetic acid14 gave mainly the alcohol
1
0
using Bu
which gave a higher yield than the radical reduction with
Ph reduction of
SnH previously employed.9b,11 NaBH
3 3 4
SnH and catalytic amounts of Pd(PPh ) ,
9
together with acetate 6, vitispirane 3, and Diels-Alder
product in a ratio of 85:11:2:2. After workup the alcohol
was isolated in 72% yield. In this reaction four
3
4
the keto function in 8 gave the requisite diene alcohol 4
in 95% yield (Scheme 3).
Oxa sp ir ocycliza t ion s t ow a r d Th ea sp ir on e a n d
Vitisp ir a n e. We previously found that the reaction
under standard oxaspirocyclization conditions (5% Pd-
OAc) , 2 equiv of p-benzoquinone, 2 equiv of LiOAc‚2H
9
diastereomeric alcohols of 9 were observed (eq 2). The
ratio A to B of the alcohols was ∼1:1, which shows that
the catalytic oxaspirocyclization reaction is not diaste-
9
b
reoselective. The alcohols were further oxidized by MnO
to theaspirones A and B (1:1) in 88% combined yield.
2
(
2
2
O
To investigate the reason for the low diastereoselec-
tivity at C-8, a control experiment was carried out in
(
(
10) Keinan, E.; Gleize, P. A. Tetrahedron Lett. 1982, 23, 477.
11) (a) Wolf, H. R.; Zink, M. P. Helv. Chim. Acta 1973, 56, 1062.
An additional alternative way for the selective reduction of 7, however
in a lower yield (53%), is the reduction with Cu hydride, prepared by
mixing a 1.0 M solution of LiAlH in THF with CuI. See: (a) Negishi,
4
E.-i.; King, A. O.; Klima, W. L.; Patterson, W.; Silveira, A., J r. J . Org.
Chem. 1980, 45, 2526. (b) Ashby, E. C.; Lin, J . J .; Kovar, R. J . Org.
Chem. 1976, 41, 1939.
(12) The slow addition of the diene was necessary to avoid Diels-
Alder reaction between the diene and p-benzoquinone.
(13) Mainly Diels-Alder product and starting material were ob-
tained.
(14) The alcohol was dissolved in acetone (1 M) for the addition.

Syntheses of Theaspirone and Vitispirane
J . Org. Chem., Vol. 61, No. 5, 1996 1827
Ta ble 1. P a lla d iu m (II)-Ca ta lyzed Oxa sp ir ocycliza tion of
4
w ith th e Ad d ition of LiCl in Aceton e in th e P r esen ce of
Differ en t Acid s. All Rea ction s Wer e Hea ted to 50 °C
conditionsa
(cosolvent)
product ratio
entry
additives
3A:3B
1
2
3
4
acetic acid
dichloroacetic acid
acetic acid
1:1
1:2
1:3
1:3.3
3 equiv of K2CO3
3 equiv of K2CO3
3 equiv of K2CO3
isobutyric acid
a
In all reactions 5% Pd(OAc)2 was used as the catalyst. The
reactions were run for 48 h at 50 °C with the addition of 3 equiv
of LiCl. All reactions were run in acetone together with an acid
as a cosolvent in a 4:1 ratio.
which 9B-syn was treated with aqueous acetic acid (eq
3
). After 24 h at room temperature, a mixture of isomers
9
B-syn and 9B-anti in a ratio of 1.4:1 was isolated. This
shows that the alcohols undergo acid-catalyzed epimer-
ization which would account for the mixture of 1,4-cis
and 1,4-trans addition products observed. However, this
does not affect the synthetic strategy since the C-8
stereogenic center is eliminated in the oxidation to
theaspirone.
Ta ble 2. Differ en t Rea ction Con d ition s for th e
F or m a tion of π-Allyl Com p lex 5 fr om Dien e Alcoh ol 4.
Th e Rea ction s Wer e P er for m ed in THF w ith
P d (P h CN)2Cl2 a s th e Sou r ce for P d
entry
additivesa
reactn time (h) products 5A:5B
1
2
3
4
3
∼14
3
10:1
7:1
2:1
K2CO3b
c
b,d
K2CO3
b
e
f
K2CO3 and water
3
1:1.3
a
The additives were added before the palladium complex. b
2
equiv. c The mixture was stirred 2 h at room temperature and kept
overnight in the refrigerator. The reaction was diluted with THF
d
e
to [substrate] ∼0.006 M (normally 0.06 M). Water (2%) was added
to the solvent. f Not full conversion (60%).
The use of a stronger acid led to diene formation.
Thus, in water with 8 equiv of methanesulfonic acid,
diene alcohol 4 was transformed into vitispiranes 3A and
ratio 3A:3B is essentially 1:1, whereas in buffered acetic
acid (added K
(Table 1, entries 1 and 3). Also the strength of the acid
2 3
CO ) 3B is favored over 3A by a factor of
3
B in moderate yield after 22 h (3A:3B, 1:1). When
3
trifluoroacetic acid (TFA) (8 equiv) was employed, the
reaction of 4 was complete after 15 h and vitispiranes A
and B were isolated in 86% yield (3A:3B, 1:1) (eq 4).
(
entries 2-4) influences the ratio between 3A and 3B.
The isomer preference in entries 2-4 is reversed to that
of the stoichiometric reaction. The use of stronger acids
(
CF
3 3
COOH, CCl COOH) did not lead to cyclization
products.
A likely explanation for the formation of vitispirane
(3) is that water attacks the intermediate (π-allyl)-
palladium complex to give alcohol 9, followed by elimina-
tion of water. In two control experiments, a mixture of
the acetate 6 and the alcohol 9 was treated with (i) acetic
acid at 60 °C or (ii) with 4 equiv of CF
3
COOH in acetic
acid at room temperature. Both reactions led exclusively
Mech a n istic Asp ects. In the stoichiometric forma-
tion of π-allyl complex 5 from diene alcohol 4, a proton
is liberated, which may induce isomerization by cata-
lyzing the retro-oxypalladation. This was investigated
by varying the acidity of the medium in the reaction
to the formation of vitispirane (3A:3B, 1:1) (eq 5).
2 2
between PdCl (PhCN) and 4, and the results are pre-
sented in Table 2. Without base, isomerization is possible
and π-allyl complex 5A predominates (entry 1). Upon
2 3
addition of K CO the 5A:5B ratio decreases (entry 2),
and when the same conditions were used in a diluted
reaction to slow down the isomerization, 5A and 5B were
isolated in a 2:1 mixture (entry 3). Finally, in entry 4,
water was added to the reaction to dissolve the base
which trapped the acid more efficiently, resulting in a
In an alternative procedure for obtaining vitispirane
, LiCl was added to the reaction. It was thought that
3
the chloride anion would be a better nucleophile than the
carboxylate anion and that the allylic chloride intermedi-
ate would directly eliminate hydrogen chloride to form
vitispiranes. Different reaction conditions were tried,
and the results are listed in Table 1. Reaction at room
temperature gave only starting material. However, at
1
:1.3 mixture in favor of isomer 5B.
In a control experiment a 2:1 mixture of 5A and 5B
was treated in THF with HCl (2 equiv) for 30 min at room
temperature, which led to isomerization of 5B to 5A,
giving a ratio 5A:5B of 7:1 (eq 7). The isomerization
takes place via an acid-catalyzed retro-oxypalladation to
5
(
0 °C the anticipated formation of vitispiranes took place
eq 6, Table 1)). It is interesting to note that the
15,16
diene alcohol 4 and subsequent recyclization.
These
diastereomeric ratio between 3A and 3B varied with
(15) For acid-catalyzed retro-oxypalladation, see: Robinson, S. D.;
slight changes of the pH. In unbuffered acetic acid the
Shaw, B. L. J . Chem. Soc. 1963, 4806.

1
828 J . Org. Chem., Vol. 61, No. 5, 1996
Nilsson et al.
Sch em e 4
acetic anhydride. p-Benzoquinone was recrystallized from
ethanol. Lithium acetate dihydrate, triphenyltin hydride, and
1
.0 M LiAlH
Pd(OAc) was bought from Engelhard. PdCl
from J ohnson Matthey. Tetrakis(triphenylphosphine)pallad-
4
solution in THF were purchased from Aldrich.
2
2
was obtained
ium18 and MnO
19
were prepared according to standard pro-
2
cedures. Tributyltin hydride was prepared from Hydrofugeant
2
0
H68 (Rh oˆ ne-Poulenc) and bis(tributyltin) oxide (Aldrich).
Merck silica gel 60 (240-400 mesh) was used for flash
chromatography.
2
1
3
,4-Deh yd r o-â-ion on e (7), 7,8-d ih yd r o-3,4-d eh yd r o-â-
9b 9b
ion ol (4), a n d π-a llyl com p lex (5) were prepared accord-
ing to literature procedures.
7
,8-Dih yd r o-3,4-d eh yd r o-â-ion on e (8). Bu
.8 mmol) was added dropwise to a stirred solution of 7 (1.9
g, 10 mmol), Pd(PPh (0.34 g, 0.3 mmol) NH Cl (1.07 g, 21
mmol), and H O (0.5 mL, 27 mmol) in 25 mL of THF under
3
SnH (5.26 g,
1
3
)
4
,
4
results show that complex 5A is the favored complex
2
under thermodynamic control.
an atmosphere of N . After 2 h ether (100 mL) was added
2
and the organic phase was separated and washed with brine.
Evaporation of the solvents gave a yellowish oil which was
dissolved in EtOAc (100 mL) and stirred with a saturated
solution of NaF (20 mL) for 3-4 h. The precipitate formed
was filtered off, the solvent was evaporated, and the residue
was subjected to flash chromatography (using pentane:ether
8
5:15 as the eluent), giving 1.7 g (90%) of 8 as a yellow oil.
The spectral data were in agreement with previously reported
values.9b An alternative way for the selective reduction of 7
to 8 was accomplished with triphenyltin hydride as described
in ref 11a.
Sp ir o[4,5]-8-a ce t oxy-2,6,10,10-t e t r a m e t h yl-1-oxa -6-
d ecen e (6) was prepared from (π-allyl)palladium chloro dimer
The acid-catalyzed equilibrium between complexes 5
is shown in Scheme 4. In the catalytic reaction the two
π-allyl complexes 5A′ and 5B′ generated are trapped by
an irreversible nucleophilic attack. Despite the fact that
the reaction is run under acidic conditions, the competing
nucleophilic attack (k2A and k2B) will trap the π-allyl
complexes 5 and hinder equilibration. The results pre-
sented in Table 1 show that the ratio 3A:3B decreases
with decreasing acidity strength in the medium, in line
with a retarded equilibrium between 5A′ and 5B′ at a
higher pH. At a slightly increased pH the reaction
becomes more kinetically controlled and formation of 3B
is favored.17
5
as described in ref 9b. Spectral data are reported for the
+
three major isomers. MS: m/z (mixture) 252 (M , <1%), 196
(38), 177 (1.4), 154 (100).
(2R*,5R*,8S*)-6A-syn :9b 1H NMR (400 MHz) δ 5.33 (m, 1
H), 5.15 (bm, 1 H), 4.15 (m, 1 H), 2.07-1.93 (m, 2 H), 2.02 (s,
3
1
H), 1.84 (m, 1 H), 1.78 (s, 3 H), 1.77-1.64 (m, 2 H), 1.41 (m,
13
H), 1.26 (d, J ) 6 Hz, 3 H), 1.02 (s, 3 H), 0.89 (s, 3 H);
C
NMR (100.5 MHz) δ 171.0, 146.0, 119.5, 88.1, 77.0, 67.4, 39.8,
6.1, 34.8, 32.2, 24.5, 23.8, 21.5, 20.9, 18.7.
2R*,5R*,8R*)-6A-a n ti:9b 1H NMR (400 MHz) δ 5.35-5.29
m, 1 H), 5.27 (m, 1 H), 4.17-4.08 (m, 1 H), 2.1 (dd, J ) 14.0,
3
(
(
In the stoichiometric reaction without added base, the
π-allyls were allowed to equilibrate, leading to thermo-
dynamical control, and a ratio 5A:5B of 10:1 was ob-
tained.
9
.5 Hz, 1 H), 2.05 (s, 3 H), 1.91-1.98 (m, 1 H), 1.88-1.81 (m,
2 H), 1.78 (s, 3 H, CdCCH ), 1.58 (dd, J ) 8.0, 6.0 Hz, 1 H),
3
1.49-1.38 (m, 1H), 1.28 (d, J ) 6.0 Hz, 3 H), 1.01 (s, 3 H),
0
7
.93 (s, 3 H); 13C NMR (100.5 MHz) δ 170.8, 145.7, 120.4, 88.3,
7.2, 69.5, 40.8, 39.2, 34.8, 34.3, 24.8, 22.2, 21.4, 20.7, 18.1.
(
2S*,5R*,8R*)-6B-a n ti:9 H NMR (400 MHz) δ 5.40 (m, 1
b 1
Con clu sion s
H), 5.24 (m, 1 H), 4.00 (m, 1 H), 2.10-1.98 (m, 2 H), 2.02 (s, 3
H), 1.86 (dd, J ) 9.2, 12.8 Hz, 1 H), 1.78 (m 1 H), 1.75 (m, 3
H), 1.68 (ddd, J ) 1.2, 6.2, 12.5 Hz, 1 H), 1.56 (m, 1 H), 1.26
Palladium(II)-catalyzed oxaspirocyclization can be used
for efficient syntheses of theaspirone and vitispirane. The
reaction proceeds via oxaspirocyclic (π-allyl)palladium
intermediates. An interesting balance between formation
of kinetic and thermodynamic (π-allyl)palladium com-
plexes controls the stereochemistry of the tetrahydrofu-
ran ring (A or B isomer). By controlling the reaction the
isomeric ratio A:B can be changed from 10:1 to 1:3.3.
(
d, J ) 6 Hz, 3 H), 1.01 (s, 3 H), 0.91 (s, 3 H); 13C NMR (100.5
MHz) δ 171.0, 140.8, 123.1, 86.7, 79.1, 69.0, 39.0, 38.7, 36.2,
3
0.6, 24.6, 23.0, 21.4, 21.1, 19.1.
Sp ir o[4,5]-8 h yd r oxy-2,6,10,10-t et r a m et h yl-1-oxa -6-
d ecen e (9). To a stirred mixture of p-benzoquinone (0.104 g,
0.96 mmol) and palladium acetate (0.011 g, 0.05 mmol) in
water (4 mL) and acetic acid (1 mL) was added diene alcohol
4
(0.093 g, 0.48 mmol) in acetone (0.5 mL) over a period of 12
h. The mixture was stirred for another 3 h until the reaction
was complete according to TLC. The reaction was worked up
as described below for the preparation of 3. The crude product
contained a mixture of four diastereoisomeric alcohols (85%)
together with 11% of the acetate 6 according to NMR. The
crude product was heated (60 °C) for 30 min in a mixture of 2
M NaOH (2.4 mL) and MeOH (9.5 mL). The solvents were
removed in vacuo. Ether (100 mL) was added, and the solution
Exp er im en ta l Section
1
NMR spectra were recorded for CDCl
3
solutions ( H at 400
1
3
MHz and C at 100.5 MHz) using tetramethylsilane (0.0 ppm,
1
1
13
H) or chloroform-d (7.26 ppm, H, 77.0 ppm, C) as internal
standards. Mass spectra were obtained in GC/MS mode (EI,
0 eV). Slow addition was performed by the use of a syringe
7
pump. Commercial acetone (99.5%) was dried over molecular
sieves (4 Å). Acetic acid was dried by the addition of 0.5%
was washed with brine (2 × 15 mL) and dried (MgSO
4
).
(
16) A small amount (<10%) of the diene alcohol 4 was also isolated
in this reaction.
17) In the extreme case if the equilibrium is completely stopped
(18) Heck, R. F. Palladium Reagents in Organic Synthesis; Aca-
demic: London, 1985; p 2.
(19) Vogel’s Textbook of Practical Organic Chemistry, 5th ed.;
Wiley: New York, 1989; p 445.
(
the reaction will be under kinetic control and the product ratio will
depend on the ratio k1A/k1B. With a full equilibrium between 5A′ and
(20) Hayashi, K.; Iyoda, I.; Shiihara, I. J . Organomet. Chem. 1967,
10, 81.
5
B′ (thermodynamically controlled intermediates) the product ratio
will depend on (k1A
k
-1B/k-1A
k
1B)(k2A/k2B).
(21) Findley, J . A.; Mackay, M. D. Can. J . Chem. 1971, 49, 2369.

Syntheses of Theaspirone and Vitispirane
J . Org. Chem., Vol. 61, No. 5, 1996 1829
Purification by flash chromatography (pentane/ether, 80:20-
reaction was ready according to TLC. Water (4 mL) was
added, and the resulting mixture was extracted with ether (3
× 3 mL). The combined organic layers were washed with 2
M NaOH until the aqueous extract was colorless. The water
layers were back-extracted with ether (2 × 10 mL). The
combined organic layers were washed with brine (7 mL) and
5
0:50) gave two fractions, the first containing three isomers
and the second one single diastereoisomer (9B-anti). The
combined alcohols were obtained in 72% yield as colorless oils.
The isomers were isolated by HPLC.
2R*,5R*,8S*)-9A-syn : 1H NMR (400 MHz) δ 5.43 (m, 1
9
b
(
H), 4.17 (m, 1 H), 4.09 (bm, 1 H), 2.00 (m, 2 H), 1.81 (m, 1 H),
.78 (m, 3 H), 1.74 (m, 1 H), 1.66 (m, 1 H), 1.44 (m, 1 H), 1.26
dried (MgSO
4
). The crude product was purified by flash
chromatography (pentane:ether, 95:5) to give 0.079 g (86%) of
1
1
3
(
d, J ) 6 Hz, 3 H), 1.05 (s, 3 H), 0.88 (s, 3 H); C NMR (100.5
MHz) δ 142.4, 124.6, 88.2, 76.8, 64.9, 43.3, 36.4, 34.8, 25.8,
4.6, 22.1, 21.0, 19.1.
vitispirane 3 (3A:3B ) 1:1) as a colorless oil. Spectral data
1
( H NMR, MS, IR) were in accord with those reported in ref 6.
2
(2R*,5R*)-3A:
6
1
H NMR (400 MHz) δ 6.05 (mdd, J ) 2.7,
(
2R*,5R*,8R*)-9A-a n ti:9b 1H NMR (400 MHz) δ 5.33 (m, 1
9.5 Hz, 1 H), 5.59 (m, 1 H), 5.06 (s, 1 H), 4.86 (s, 1 H), 4.31
(app hex, J ) 6 Hz, 1 H), 2.16 (md, J ) 18.5 Hz, 1 H), 2.12-
1.89 (m, 3 H), 1.50-1.39 (m, 1 H), 1.24 (m, 1 H), 1.23 (d, J )
H), 4.23 (bm, 1 H), 4.12 (m, 1 H), 2.08 (dd, J ) 8.5, 13.2 Hz, 1
H), 1.93 (m, 1 H), 1.82 (m, 2 H), 1.75 (m, 3 H), 1.44 (m, 2 H),
1
.26 (d, J ) 6 Hz, 3 H), 0.95 (s, 3 H), 0.91 (s, 3 H); 13C NMR
13
6 Hz, 3 H), 0.95 (s, 3 H), 0.89 (s, 3 H); C NMR (100.5 MHz)
(
100.5 MHz) δ 143.3, 124.8, 88.5, 77.0, 65.9, 45.1, 39.2, 34.4,
δ 149.5, 128.6, 127.0, 109.0, 88.8, 77.2, 41.1, 36.1, 32.8, 31.6,
3
1.2, 24.8, 22.1, 20.6, 18.0.
(
23.5, 23.3, 21.4.
2S*,5R*,8S*)-9B-syn : 1H NMR (400 MHz) δ 5.41-5.38
m, 1 H), 4.25-4.19 (m, 1 H), 4.17 (dp, J ) 8.5, 6.0 Hz, 1H),
.18 (dt J ) 13.0, 8.5 Hz, 1 H), 2.10-2.02 (m, 1 H), 1.87-1.80
m, 2 H), 1.74 (s, 3 H), 1.57 (ddt, J ) 12.0, 9.5, 8.2 Hz, 1 H),
.44 (dd, J ) 13.0, 8.0 Hz, 1 H), 1.25 (d, J ) 6.0 Hz, 3 H), 0.99
9b
2S*,5R*)-3B:6 1H NMR (400 MHz) δ 6.07 (mdd, J ) 2.9,
(
(
2
(
1
9
.6 Hz, 1 H), 5.59 (m, 1 H), 5.18 (s, 1 H), 4.89 (s, 1 H), 4.04 (m,
J ) 6 Hz, 1 H), 2.17 (md, J ) 18.2 Hz, 1 H), 2.12-1.82 (m, 3
H), 1.62-1.51 (m, 1 H), 1.34 (d, J ) 6 Hz, 3 H), 0.98 (s, 3 H),
13
0
8
.90 (s, 3 H); C NMR (100.5 MHz) δ 150.2, 128.9, 126.8, 110.3,
1
3
(
s, 3 H), 0.96 (s, 3 H); C NMR (100.5 MHz) δ 142.4, 124.7,
8.5, 75.6, 40.3, 37.3, 32.8, 32.5, 23.33, 23.28, 20.6.
8
8.8, 77.4, 65.9, 44.5, 39.8, 35.1, 33.6, 25.8, 23.1, 21.7, 19.1.
2S*,5R*,8R*)-9B-a n ti: H NMR (400 MHz) δ 5.50 (m, 1
Oxa sp ir ocycliza tion of 4 to Vitisp ir a n e (3) in th e
9
b 1
(
P r esen ce of K
with a reflux condenser under nitrogen were successively
placed Pd(OAc) (20 mg, 0.089 mmol), p-benzoquinone (390 mg,
.61 mmol), potassium carbonate (400 mg, 5.41 mmol), and
lithium chloride (128 mg, 3.01 mmol) in acetone-acetic acid
4:1, 7.5 mL). After subsequent addition of the diene alcohol
2 3
CO a n d LiCl. In a two-necked flask fitted
H), 4.10 (bm, 1 H), 3.98 (m, 1 H), 2.02 (m, 2 H), 1.78 (m, 1 H),
1
.74 (dd, J ) 2, 1.6 Hz, 3 H), 1.70 (md, J ) 8 Hz, 2 H), 1.63
2
(bm, 1 H), 1.57 (m, 1 H), 1.26 (d, J ) 6 Hz, 3 H), 1.00 (s, 3 H),
3
1
3
0
6
.85 (s, 3 H); C NMR (100.5 MHz) δ 138.3, 128.3, 86.8, 76.5,
5.9, 43.7, 39.0, 36.3, 30.2, 24.7, 22.7, 21.0, 19.1.
(
Sp ir o[4,5]-2,6,10,10-t e t r a m e t h yl-1-oxa -8-oxo-6-d e c-
4
4
(350 mg, 1.8 mmol), the reaction mixture was heated during
8 h at 50 °C (the reaction was followed by TLC). The usual
en e (Th ea sp ir on e, 1). Oxidation of the alcohols 9 with MnO
2
performed as described in ref 9b afforded theaspirones A and
B. The NMR data of theaspirone A were in accord with those
reported in ref 9b.
workup followed by flash chromatography (hexane:ethyl ac-
etate, 90:10) afforded 195 mg (56%) of 3 as a colorless oil (3A:
3
B, 26:74). Further experiments to evaluate the influence of
the acidity of the medium were conducted employing different
acids and K
CO under similar reaction conditions (Table 1).
(
2S*,5R*)-Th ea sp ir on e B:3b 1H NMR (400 MHz) δ 5.76 (m,
1
(
H), 4.22 (m, 1 H), 2.40 (m, 1 H), 2.35-2.26 (m, 2 H), 2.15
2
3
m, 1 H), 1.96 (d, J ) 1.3 Hz, 3 H), 1.87 (ddd, J ) 3, 9.3, 13.5
Hz, 1 H), 1.64 (m, 1 H), 1.30 (d, J ) 6 Hz, 3 H), 1.07 (s, 3 H),
1
3
Ack n ow led gm en t. We are grateful to the Swedish
Natural Science Research Council, the Swedish Re-
search Council for Engineering Science, the MESR, the
CNRS, and The R e´ my-Martin Society for financial
support. We thank Agismed for grants (A.A. and C.P.).
We also thank J ohnson Matthey for a loan of PdCl2.
1
7
.01 (s, 3 H); C NMR (100.5 MHz) δ 198.8, 168.2, 125.3, 88.6,
8.0, 49.9, 41.6, 35.0, 32.7, 24.5, 23.7, 21.3, 20.5.
S p ir o [4,5]-2,10,10-t r im e t h y l-6-m e t h y le n e -1-o x a -7-
d ecen e (Vitisp ir a n e, 3). To a stirred mixture of p-benzo-
quinone (0.104 g, 0.96 mmol) and palladium acetate (0.011 g,
0
.05 mmol) in water (4 mL) and trifluoroacetic acid (0.44 mL)
was added diene alcohol 4 (0.093 g, 0.48 mmol) in acetone (0.5
mL) over a period of 12 h. After stirring for another 4 h, the
J O9505031