A new iron-mediated strategy for the synthesis of α-lipoic acid and analogues

Article abstract of DOI:10.1002/(SICI)1099-0690(199809)1998:9<1949::AID-EJOC1949>3.0.CO;2-8

Dithioester 10 was synthesized in 6 steps from tricarbonyl(diene)iron complex 3. Compound 10 is not only a key intermediate for the synthesis of methyl lipoate (13) but could also be of interest for the preparation of various labelled compounds and struct

Full text of DOI:10.1002/(SICI)1099-0690(199809)1998:9<1949::AID-EJOC1949>3.0.CO;2-8

FULL PAPER
A New Iron-Mediated Strategy for the Synthesis of ␣-Lipoic Acid and
Analogues
a
a
a
b
a
´
¨
´
´
Christophe Crevisy , Benjamin Herbage , Marie-Laure Marrel , Loıc Toupet , and Rene Gree*
Laboratoire de Chimie Organique Biologique, CNRS ESA 6052, E. N. S. Chimie de Rennesa,
´ ´
Avenue du General Leclerc, F-35700 Rennes, France
b
`
´
´
´
Groupe Matiere Condensee et Materiaux, Universite de Rennes ,
´ ´
Avenue du General Leclerc, F-35042 Rennes cedex, France
Received April 14, 1998
Keywords: α-Lipoic acid / Structural analogues / Tricarbonyl(η5-pentadienyl)iron cations / Carbonyliron complexes /
Diimide reduction
Dithioester 10 was synthesized in 6 steps from tricarbonyl- also be of interest for the preparation of various labelled
(diene)iron complex 3. Compound 10 is not only a key compounds and structural analogues.
intermediate for the synthesis of methyl lipoate (13) but could
Introduction
syntheses reported, the chirality is introduced either directly
from the “chiral pool”[6b][6d][6e], or by using chiral auxilia-
ries[6a][6c], and by asymmetric epoxidation[6f]; enzymatic
(R)-(ϩ)-α-lipoic acid[1] (1) (Scheme 1) is widely distrib-
uted in plant and animal tissues.[2] As lipoamide, it func-
tions as a cofactor in the oxidative decarboxylation of α-
keto acids.[3] It is also a growth factor for many bacteria
and protozoa.[4]
methods have been also used: reduction[6g][6j]
dation[6h][6k] or resolution.[6i]
, oxi-
Sulfur atoms are usually introduced in two steps from the
key intermediate 2 (Scheme 2).
Scheme 1
Scheme 2
Lipoic acid exhibits antioxidant functions[5]: it scavenges
hydroxyl radicals, hypochlorous acid, and singlet oxygen,
and also chelates transition metals. Lipoic acid and its re-
duced form (dihydrolipoic acid) appeared to be able to re-
generate other antioxidants such as ascorbate and vitamin
E. Thanks to these antioxidant properties, lipoic acid is use-
We planned a new strategy that would permit the syn-
ful or potentially useful for the treatment of various dis- thesis of labelled derivatives and structural analogues which
eases[5]: α-lipoic acid has beneficial effects in prevention would be very useful in biochemical experiments. We be-
and treatment of both type-I and type-II diabetes; α-lipoic lieved that a strategy using tricarbonyl(diene)iron com-
acid and dihydrolipoic acid have been reported to be effec- plexes could offer an original and powerful access to lipoic
tive in ameliorating or preventing damage in myocardial acid and analogues (Scheme 3). Indeed, intermediate 10
and cerebral ischemia-reperfusion injury in rat; α-lipoic acid should allow access to many analogues of lipoic acid by 1,4-
also protects against damages produced by radiation expo- and 1,6-additions to the diene or to labelled compounds by
sure.
It has also be shown to be potentially useful for the treat-
deuteration or tritiation (at a late stage in the synthesis).[7]
Tricarbonyl(diene)iron complexes have already found
ment of heavy-metal poisoning, degenerative diseases of the many applications in the asymmetric synthesis of organic
central nervous system, AIDS, and diabetic cataractogen- molecules.[8] Indeed, their preparation is easy and these
esis.
High doses of α-lipoic acid are approved in Germany for conditions, the Fe(CO)3 moiety acting as an efficient pro-
the treatment of diabetic polyneuropathy.[5]
tective group for the 1,3-diene. They are chiral and complex
complexes are stable under a large variety of experimental
Lipoic acid is then a biologically important molecule. Its 3, for example, can be obtained in optically active form
extraction from natural sources yields only small quantities after resolution.[9] So, starting from an optically pure 3, it
of material, and so its synthesis has been largely investi- would be possible to obtain both enantiomers of lipoic acid.
gated[6] since its isolation in 1951.[1] In the asymmetric Finally, due to the properties of tricarbonyl(η5-pentadie-
Eur. J. Org. Chem. 1998, 1949Ϫ1954
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1434Ϫ193X/98/0909Ϫ1949 $ 17.50ϩ.50/0
1949

´
´
C. Crevisy, B. Herbage, M.-L. Marrel, L. Toupet, R. Gree
FULL PAPER
Scheme 3
Scheme 4
nyl)iron cations[10] the stereocontrolled formation of the
CϪS bond may be envisaged.
We have first investigated the synthesis of racemic methyl
lipoate to validate our strategy and we report here the cor-
responding results.
Results and Discussion
Synthesis of the Key Intermediate 10 (Scheme 4)
The addition of vinylmagnesium bromine to readily avail-
able[9] complex 3 gave a mixture of easily separable alcohols
Ψ-endo-4a (22%) and Ψ-exo-4b (56%). The stereochemistry
of the major, more polar, isomer was attributed according
to previous studies[8][11] in that field. After protecting the
OH group of 4b as a tert-butyldimethylsilyl ether 5, the lat-
ter compound was subjected to an hydroboration/oxidation
sequence to afford alcohol 6 in 47% overall yield. The first
sulfur atom was then introduced by a Mitsunobu reaction:
6 treated with PPh3 and DIAD and then with thiobenzoic
acid afforded thiobenzoate 7 in high yield (86%).
Scheme 5
We planned to introduce stereoselectively the sulfur atom
in the secondary position by allowing an appropriate sulfur
nucleophile to react with the tricarbonyl(η5-pentadienyl)-
iron cation 8t (Scheme 5). It is well-known[8][10] that tricar-
bonyl(η5-pentadienyl)iron cations are usually prepared by
the acid-assisted dehydration of alcohol or ether precursors
such as 7. This cation may exist in two forms: transoid 8t
and cisoid 8c (Scheme 5).
Addition of nucleophiles to such cations is usually highly
stereoselective; therefore, addition of thiobenzoic acid in
our case could a priori give two products; attack of the
transoid form leading to (E,E) complex 9, and attack of the
cisoid form leading to (E,Z) compound 9b. However, it is
known that under the “in situ” reaction conditions, the
(E,E) complexes are usually obtained. Thus, compound 7
was treated with Amberlyst 15 and a large excess (ca. 50
equiv.) of thiobenzoic acid to afford only the (E,E)-diene
complex 9 in high yield[12] (81%).
Finally, oxidative decomplexation with ceric ammonium
nitrate gave diene 10 in 87% yield. This key intermediate
should open the route to labelled derivatives and analogues
of lipoic acid.
Synthesis of Methyl Lipoate (13) (Scheme 6)
The first step was the reduction of a diene in a molecule
containing two sulfur atoms. The difficulty of reducing a
X-ray crystallographic analysis[13] confirmed the ex- double or triple bond in a molecule containing a sulfur
pected stereochemistry (Figure 1): departure of the leaving atom is well-known, so this step proved to be challenging.
group and the attack of the nucleophile occur on the face
We first tried to perform the catalytic hydrogenation of
opposite to the carbonyliron unit[8][10] so that the reaction the diene over palladium on charcoal or platinum oxide.
proceeds with overall retention of configuration.
1950
The reaction led only to recovery of the starting material.
Eur. J. Org. Chem. 1998, 1949Ϫ1954

A New Iron-Mediated Strategy for the Synthesis of α-Lipoic Acid and Analogues
FULL PAPER
Figure 1. Crystal structure of 9
Thus, treatment of diene 10 with a large excess of TPSH in
MeOH at 38°C during 12 days gave saturated compound
12 in 61% yield.
Finally, 12 was converted into racemic methyl lipoate (13)
by treatment with potassium carbonate in MeOH (82%):
the intermediate bis-thiolate was probably further oxidized
by oxygen dissolved in the solvent to afford the cyclic disulf-
ide. 1H- and 13C-NMR spectral values were in excellent
agreement with literature data.[6k] Conversion of methyl
lipoate to lipoic acid has already been reported in the litera-
ture.[6d]
Synthesis of Analogue 14 (Scheme 7)
The preparation of a carbonylmetal derivative of lipoic
acid could be biologically interesting since a cold FT-IR
assay technique has been developed which relies on the
strong carbonyl IR absorbance of organometallic spe-
cies.[18] Thus, the same reaction that yielded 13 was per-
formed with complex 9 to give the tricarbonyl(diene)iron
analogue 14 of methyl lipoate in 62% yield.
Photodecomplexation using an improved[14] variant of
the Franck-Neumann reaction[15] was attempted with the
dithioacetate complex. The reaction yielded a complex mix-
ture of compounds where the expected β,γ-unsaturated es-
ter complex 11 (Figure 2) was clearly identified. Unfortu-
nately, the separation of the products was very difficult.
Moreover, the yield of expected compound 11 was low (<
50%) and this approach was abandoned.
Scheme 7
Figure 2. β,γ-Unsaturated ester 11
Though we were not able to find any example reported
of the reduction of unsaturated carbonϪcarbon bonds by
diimide in the presence of a sulfur atom, this reagent
seemed a priori promising. The two main ways to generate
diimide were tried: 2,4,6-triisopropylbenzenesulfonyl hydra-
zide (TPSH) in MeOH[16] was found to be more effective
than potassium azodicarboxylate/acetic acid in CH2Cl2.[17]
Conclusion
We have shown that tricarbonyl(diene)iron chemistry fur-
nishes a new stereoselective route to 10, diene analogue of
protected dihydrolipoic acid. We have already synthesized
racemic methyl lipoate in only 8 steps from readily available
complex 3, together with its interesting carbonyl(diene)iron
analogue 14. Starting from optically active 3 it should be
possible to obtain optically active 10 which could be con-
verted into various analogues of (R)-(ϩ)-α-lipoic acid.
Scheme 6
We thank Dr. P. Guenot (CRMPO, Rennes) for performing the
mass-spectral experiments.
Experimental Section
General: THF was distilled from sodium benzophenone, CH2Cl2
from P2O5 or CaH2. MeOH was distilled from Na and stored over
˚
molecular sieves while DMF was dried over 4-A molecular sieves.
Ϫ Melting points are uncorrected. Ϫ IR spectra were recorded with
a Nicolet 205 spectrometer. Ϫ 1H NMR (400 MHz) and 13C NMR
(100 MHz) spectra were recorded using a Bruker ARX 400 instru-
ment (internal standard TMS, δH and δC ϭ 0 or CDCl3, δH ϭ 7.26
and δC ϭ 77.0). The coupling constants (J) are in Hertz (Hz). High-
resolution mass measurements were performed at the CRMPO
(Rennes) with a Varian Mat 311 spectrometer. Ϫ Microanalyses
were performed at the “Service de Microanalyse” I.C.S.N.-
Eur. J. Org. Chem. 1998, 1949Ϫ1954
1951

´
´
C. Crevisy, B. Herbage, M.-L. Marrel, L. Toupet, R. Gree
FULL PAPER
C.N.R.S., Gif-sur-Yvette, France. Ϫ Crystallographic data were ob- crude product was purified by flash chromatography on silica gel
tained using a CAD4 ENRAF-NONIUS diffractometer. Crystal
data are given in Table 1[13]
(petroleum ether/ether, 97:3) to give 5 as a yellow solid (1.89 g,
94%). Ϫ IR (nujol): ν˜ ϭ 2064, 2006, 1988, 1722 cmϪ1. Ϫ 1H NMR
(CDCl3): δ ϭ 0.05 (s, 3 H, SiCH3), 0.09 (s, 3 H, SiCH3), 0.89 (s, 9
H, tBu), 1.08 (dd, 1 H, J ϭ 1.1, 8.1, ECH), 1.33 (dt, 1 H, J ϭ 1.1,
8.4, CHϭCHCHOH), 3.66 (s, 3 H, E), 3.90 (br. tdd, 1 H, J ഠ 1,
7, 8, CHOSitBu), 5.12 (ddd, 1 H, J ϭ 0.8, 1.4, 10.2 CHϭCH2),
5.19 (ddd, 1 H, J ϭ 1.1, 1.4, 17.1, CHϭCH2), 5.36 (ddd, 1 H, J ϭ
0.8, 5.0, 8.5, ECHϭCHCHϭCH), 5.80 (ddd, 1 H, J ϭ 1.1, 5.0, 8.1,
ECHϭCHCHϭCH), 5.89 (m, 1 H, J ϭ 7.2, 10.2, 17.1, CHϭCH2).
.
Table 1. Crystallograpic data for 9
formula
mol. wt.
C26H23FeS2O7 radiation
Mo-Kα
567.44
µ (Mo-Kα)
7.73
[cmϪ1
]
cryst. system
space group
crystal size
monoclinic
P21/n
variance of stan- 0.2%
dards
Ϫ
13C NMR (CDCl3): δ ϭ Ϫ4.4, Ϫ4.1, 18.1, 25.8, 46.2, 51.7, 67.0,
tmax (for one ne- 60
asure) [s]
0.22ϫ0.25ϫ0.33 range of h; k; l Ϫ47/47; 0/7;
76.2, 83.9, 85.6, 114.8, 140.1, 172.5. Ϫ C18H26FeO6Si (422.3): calcd.
C 51.19, H 6.20; found C 51.29, H 6.23.
0/15
5072
ρcalcd. [g cmϪ3
]
1.448
reflections me-
asured
Alcohol 6: To a solution of 5 (296 mg, 0.701 mmol) in THF (5
ml) was added at 0°C a solution of 9-BBN in THF (0.5 , 2.9 ml,
1.45 mmol). After the solution was stirred at room temperature for
8 h, an aqueous solution of NaOH (1 , 2 ml, 2 mmol), then a
35% solution of H2O2 (240 ml, 2.74 mmol) was added dropwise at
0°C. The reaction mixture was further stirred at room temperature
for 15 min, then poured into a saturated NH4Cl solution and ex-
tracted with ether. The organic layer was washed with a saturated
NH4Cl solution and water, dried (MgSO4), and concentrated under
reduced pressure. The crude product was purified by flash chroma-
tography on silica gel (petroleum ether/ether, 75:25) to give 6 as a
yellow solid (154 mg, 50%). Ϫ IR (nujol): ν˜ ϭ 3532, 2063, 1989,
1980, 1689 cmϪ1. Ϫ 1H NMR (CDCl3): δ ϭ 0.14 (2s, 6 H, 2
SiCH3), 0.89 (s, 9 H, SitBu), 1.16 (dd, 1 H, J ϭ 0.9, 8.1, ECH),
1.42 (dt, 1 H, J ϭ 0.8, 8.9, CHϭCHCHOSi), 1.80Ϫ1.89 (m, 1 H,
CH2), 1.93Ϫ2.02 (m, 1 H, CH2), 2.10 (t, 1 H, J ϭ 5.2, OH), 3.67
(s, 3 H, E), 3.70 (ddd, 1 H, J ϭ 3.3, 7.4, 9.2, CHOSi), 3.75Ϫ3.91
˚
a [A]
34.516(4)
5.956(1)
refl. observed [I 2856 (4σ)
> σ(I)]
˚
b [A]
Rint (merging
equiv. refl.)
R (isotropic)
0.011
˚
c [A]
12.707(2)
94.93
2603(1)
0.082
β [°]
R (anisotropic) 0.062
V [A3]
Fourier diffe-
rence
0.46Ϫ0.15
˚
Z
4
N(obs)/N(var)
final R
2856/391
0.029
F(000)
T [K]
2Θmax [°]
scan
1172
294
Rw[a]
0.028
50
Sw
1.116
ω/2θ ϭ 1
max. resid. e
0.28, 0.12
Ϫ3, ∆/σ
˚
A
w ϭ 1/σ(Fo)2 ϭ [σ2(I) ϩ (0.04 Fo2)2]Ϫ1/2
.
[a]
Alcohols 4a and 4b: To a solution of aldehyde 3 (2.72 g, 9.7
mmol) in THF (80 ml) was added at Ϫ65°C, a solution of vinyl-
magnesium bromide (10 , 15 ml, 15 mmol). The reaction mixture (m, 2 H, CH2OH), 5.32 (ddd, 1 H, J ϭ 1.0, 5.1, 8.6, CHϭ
was warmed to Ϫ45°C over a period of 50 min and quenched with CHCHOSi), 5.82 (ddd, 1 H, J ϭ 1.0, 5.1, 8.1, ECHϭCH). Ϫ 13C
a saturated aqueous NH4Cl solution then diluted with ether. The
NMR (CDCl3): δ ϭ Ϫ4.4, Ϫ3.7, 18.0, 25.7, 40.9, 46.7, 51.7, 59.5,
organic layer was washed with water, then dried (MgSO4). The sol- 66.2, 73.8, 84.0, 86.2, 172.3. Ϫ C18H28FeO7Si (440.3): calcd. C
vents were removed in vacuo. The crude product was purified by 49.10, H 6.41; found C 48.82, 6.39.
flash chromatography on silica gel (petroleum ether/ether, 3:2) to
Thioester 7: To a solution of PPh3 (2.8 g, 10.6 mmol) in THF
furnish alcohol 4a as a yellow oil (0.65 g, 22%) then alcohol 4b as
(40 ml) was added dropwise at 0°C DIAD (2.1 ml, 10.7 mmol).
The reaction mixture was stirred 20 min at 0°C. Then a solution
of alcohol 6 (2.35 g, 5.3 mmol) in THF (40 ml) and thiobenzoic
a yellow solid (1.68 g, 56%).
Alcohol 4a: 1H NMR (CDCl3): δ ϭ 0.98 (d, 1 H, J ϭ 7.9, CHE),
1.32 (t, 1 H, J ϭ 7.5, CHϭCHCHOH), 1.92 (br. s, 1 H, OH), 3.66
(s, 3 H, E), 4.16 (br. t, 1 H, CHOH), 5.16 (d, 1 H, J ϭ 10.3, CHϭ
CH2), 5.26 (d, 1 H, J ϭ 17.0, CHϭCH2), 5.43 (dd, 1 H, J ϭ 5.1,
8.3, ECHϭCH), 5.82 (dd, 1 H, J ϭ 5.3, 7.5, ECHϭCHCH), 5.91
(m, 1 H, J ϭ 6.2, 10.2, 16.8, CHϭCH2). Ϫ 13C NMR (CDCl3):
δ ϭ 45.6, 51.7, 67.6, 74.0, 83.2, 83.8, 114.8, 140.6, 172.7. Ϫ HRMS:
calcd. for C12H12FeO6 307.9983; found 307.9987.
acid (1.3 ml, ca. 10 mmol) were added dropwise. The reaction mix-
ture was stirred 1 h at 0°C then concentrated under reduced pres-
sure. The resulting residue was filtered through silica gel (petroleum
ether/ether, 95:5), then purified by flash chromatography on silica
gel (petroleum ether/ether, 95:5) to give thioester 7 (2.55 g, 85%)
as a yellow syrup. Ϫ IR (film): ν˜ ϭ 2050, 1995, 1988, 1715, 1666
1
cmϪ1. Ϫ H NMR (CDCl3): δ ϭ 0.14 (s, 3 H, SiCH3), 0.15 (s, 3
H, SiCH3) 0.91 (s, 9 H, SitBu), 1.12 (dd, 1 H, J ϭ 0.8, 8.1, ECH),
1.36 (dt, 1 H, J ϭ 0.8, 8.6, CHϭCHCHOSi), 1.93Ϫ2.09 (m, 2 H,
CH2), 3.04 (td, 1 H, J ϭ 7.9, 13.3, CH2S), 3.29 (m, 1 H, J ϭ 4.8,
7.6, 13.3, CH2S), 3.61 (dt, 1 H, J ϭ 3.5, 8.4, CHOSi), 3.66 (s, 3 H,
E), 5.33 (ddd, 1 H, J ϭ 0.8, 5.1, 8.6, CHϭCHCHOSi), 5.80 (ddd,
1 H, J ϭ 1.0, 5.1, 8.1, ECHϭCH), 7.41Ϫ7.47 (m, 2 H, Ph),
7.53Ϫ7.59 (m, 1 H, Ph), 7.91Ϫ7.96 (m, 2 H, Ph). Ϫ 13C NMR
(CDCl3): δ ϭ Ϫ4.2, Ϫ3.5, 18.1, 25.4, 25.8, 38.8, 46.6, 51.7, 66.5,
73.6, 83.8, 86.1, 127.1, 128.5, 133.2, 137.1, 172.3, 191.7. Ϫ
C25H32FeO7SSi (560.5): calcd. C 53.57, H 5.75; found C 53.84, H
5.99.
Alcohol 4b: M.p. 65°C (hexane/ethyl acetate). Ϫ IR (nujol): ν˜ ϭ
1
2064, 1999, 1966, 1712 cmϪ1. Ϫ H NMR (CDCl3): δ ϭ 1.08 (dd,
1 H, J ϭ 1.1, 8.1, CHE), 1.27 (dt, 1 H, J ϭ 1.1, 8.2, CHϭ
CHCHOH), 1.87 (d, 1 H, J ϭ 3.7, OH), 3.67 (s, 3 H, E), 3.99 (br.
dt, 1 H, J ഠ 3.5, 7.5, CHOH), 5.19 (td, 1 H, J ϭ 0.9, 10.2, CHϭ
CH2), 5.29 (td, 1 H, J ϭ 1.1, 17.1, CHϭCH2), 5.51 (ddd, 1 H, J ϭ
0.9, 5.0, 8.5, CHϭCHCHOH), 5.85 (ddd, 1 H, J ϭ 1.1, 5.0, 8.1,
ECHϭCH), 5.96 (m, 1 H, J ϭ 7.2, 10.2, 17.1, CHϭCH2). Ϫ 13C
NMR (CDCl3): δ ϭ 46.1, 51.7, 64.7, 75.2, 84.4, 85.2, 115.8, 139.5,
172.5. Ϫ HRMS: calcd. for C12H12FeO6 307.9983; found 307.9987.
Silyl Ether 5: To a solution of alcohol 4b (1.48 g, 4.80 mmol) in
anhydrous DMF (4 ml) was added imidazole (1.14 g, 16.7 mmol)
and tert-butyldimethylsilyl chloride (1.59 g, 10.5 mmol). The reac-
Dithioester 9: To a solution of thioester 7 (300 mg, 0.53 mmol)
in CH2Cl2 (10 ml) was added at 0°C, thiobenzoic acid (3 ml, ca.
tion mixture was stirred at room temperature for 1 h then diluted 25 mmol) and Amberlyst 15 (600 mg). The reaction mixture was
(Et2O). The organic layer was washed with a saturated solution of stirred at 0°C for 4.5 h then filtered, poured into a saturated aque-
NH4Cl and with water, dried (MgSO4), and concentrated. The
1952
ous NaHCO3 solution and extracted with ether. The organic layer
Eur. J. Org. Chem. 1998, 1949Ϫ1954

A New Iron-Mediated Strategy for the Synthesis of α-Lipoic Acid and Analogues
FULL PAPER
was washed successively three times with saturated aqueous reaction mixture was then stirred for 3.5 h under air at room tem-
NaHCO3, solution then with water, dried (MgSO4), and concen- perature. A saturated NH4Cl solution was then added and the mix-
trated. The crude product was purified by flash chromatography
(petroleum ether/ether, 4:1) to give compound 9 (245 mg, 81%) as
ture was extracted three times with CH2Cl2. The organic layer was
dried (MgSO4) and the solvents were evaporated. The crude residue
a yellow foam which was crystallized from hexane, m.p. 94Ϫ96°C was purified by flash chromatography on silica gel (petroleum
(hexane). Ϫ IR (film): ν˜ ϭ 2058, 1996, 1990, 1716, 1666, 1662 ether/ether 10:1) to give methyl lipoate (13) (14.4 mg, 82%) as a
cmϪ1. Ϫ 1H NMR (CDCl3): δ ϭ 1.08 (dd, 1 H, J ϭ 1.0, 8.2, CHE),
clear yellow oil. Ϫ HRMS: calcd. for C9H16S2O2 220.0592; found
1.47 (ddd, 1 H, J ϭ 0.9, 8.3, 10.4, CHϭCHCHS), 2.15Ϫ2.36 (m, 202.0586. Ϫ 1H- and 13C-NMR spectral values were comparable
2 H, CH2), 3.06 (td, 1 H, J ϭ 8.1, 13.7, CH2S), 3.38 (ddd, 1 H,
J ϭ 4.3, 7.9, 13.5; CH2S), 3.66 (s, 3 H, E), 3.84 (dt, 1 H, J ϭ 3.5,
10.7, CHS), 5.68 (ddd, 1 H, J ϭ 0.9, 5.2, 8.5, CHϭCH), 5.79 (ddd,
1 H, J ϭ 1.1, 5.2, 8.2, CHϭCH), 7.41Ϫ7.51 (m, 4 H, Ph),
7.54Ϫ7.63 (m, 2 H, Ph), 7.92Ϫ8.01 (m, 4 H, Ph). Ϫ 13C NMR
(CDCl3): δ ϭ 26.5, 35.7, 46.3, 47.0, 51.8, 66.8, 84.2, 85.6, 127.2,
127.5, 128.6, 128.7, 133.4, 133.8, 136.6, 136.9, 172.2, 190.8, 191.5.
Ϫ C26H22FeO7S2 (566.4): calcd. C 55.13, H 3.91; found C 55.35,
H 3.98.
with those reported[6k]
.
Disulfur Compound 14: To a solution of dithioester 9 (70 mg,
0.124 mmol) in MeOH (5 ml) was added K2CO3 (100 mg, 0.723
mmol). The reaction mixture was stirred at room temperature for
1.5 h, poured into saturated aqueous NH4Cl solution and extracted
with ether. The organic layer was washed with an aqueous
NaHCO3 solution and with H2O before being dried (MgSO4) and
concentrated under reduced pressure. The crude product was puri-
fied by flash chromatography on silica gel (petroleum ether/ether,
9:1) to give compound 14 (27.5 mg, 63%) as a yellow syrup. Ϫ IR
Diene 10: To a solution of complex 9 (1.0 g, 1.76 mmol) in
MeOH (130 ml) was added at 0°C ceric ammonium nitrate (3.55
g, 6.48 mmol). The reaction mixture was stirred at 0°C for 1.5 h,
poured into H2O and extracted with ether. The organic layer was
dried (MgSO4) and concentrated under reduced pressure. The
crude product was purified by flash chromatography (petroleum
ether/ether 4:1) to give diene 10 (685 mg, 91%) as a colorless oil.
Ϫ IR (film): ν˜ ϭ 1719, 1664 cmϪ1. Ϫ 1H NMR (CDCl3): δ ϭ
2.12Ϫ2.27 (m, 2 H, CH2CH2S), 3.17 (t, 2 H, J ϭ 7.4, CH2S), 3.74
(s, 3 H, E), 4.52 (q, 1 H, J ϭ 7.8, CHS), 5.93 (d, 1 H, J ϭ 15.3,
ECH), 6.14 (dd, 1 H, J ϭ 8.6, 15.1, CHϭCHCHS), 6.52 (dd, 1 H,
J ϭ 11.1, 15.1, CHϭCHCHϭCH), 7.27 (ddd, 1 H, J ϭ 0.5, 11.1,
15.2, CHϭCHCHϭCH), 7.43Ϫ7.49 (m, 4 H, Ph), 7.55Ϫ7.61 (m,
2 H, Ph), 7.92Ϫ7.99 (m, 4 H, Ph). Ϫ 13C NMR (CDCl3): δ ϭ
26.5, 33.8, 44.7, 51.5, 121.6, 127.2, 127.3, 128.6, 128.7, 130.0, 133.4,
133.6, 136.6, 136.8, 140.5, 143.7, 167.2, 190.1, 191.4. Ϫ C23H22O4S2
(426.6): calcd. C 64.76, H 5.20; found C 64.89, H 5.31.
1
(film): ν˜ ϭ 2055, 1979, 1712 cmϪ1. Ϫ H NMR (CDCl3): δ ϭ 1.17
(dd, 1 H, J ϭ 1.0, 8.2, ECH), 1.44 (ddd, 1 H, J ϭ 0.9, 8.5, 10.1,
CHϭCHCHS), 2.07Ϫ2.18 (m, 1 H, CH2CH2S), 2.57Ϫ2.67 (m, 1
H, CH2CH2S), 3.15 (ddd, 1 H, J ϭ 6.5, 7.9, 11.4, CH2S), 3.31 (m,
1 H, J ϭ 4.6, 6.8, 11.4, CH2S), 3.51 (td, 1 H, J ϭ 6.9, 10.3, CHS),
3.67 (s, 3 H, E), 5.37 (ddd, 1 H, J ϭ 1.0, 5.1, 8.5, CHϭCHCHS),
5.84 (ddd, 1 H, J ϭ 1.1, 5.1, 8.2, ECHϭCH). Ϫ13C NMR (CDCl3):
δ ϭ 39.8, 42.6, 46.7, 51.8, 59.2, 63.8, 84.8, 85.7, 172.1. Ϫ HRMS:
calcd. for C12H12FeO5S2 355.9475; found 355.9483.
[1]
L. J. Reed, I. C. Gunsalus, B. G. DeBusk, C. S. Hornberger,
Jr., Science 1951, 114, 93Ϫ94.
[2]
U. Schmidt, P. Grafen, H. W. Goedde, Angew. Chem. 1965, 77,
900Ϫ911; Angew. Chem. Int. Ed. Engl. 1965, 4, 846Ϫ856.
[3] [3a]
[3b]
L. J. Reed, Acc. Chem. Res. 1974, 7, 40Ϫ46. Ϫ
H. Sigel,
Angew. Chem. 1982, 94, 421Ϫ432; Angew. Chem. Int. Ed. Engl.
1982, 21, 389Ϫ400.
C. V. Natraj, V. M. Gandhi, K. K. G. Menon, J. Biosci. 1984,
6, 37Ϫ46.
Saturated Dithiobenzoate 12: To a solution of diene 10 (55.5 mg,
0.130 mmol) in MeOH (25 ml) at 38°C, was added every day for 5
d triisopropylbenzenesulfonyl hydrazide (860 mg, 2.9 mmol). The
reaction mixture was heated at 38°C for 6 h, then kept at room
temperature for 48 h after the fifth addition. It was then diluted in
CH2Cl2. The organic layer was treated with an aqueous NaHCO3
solution. The aqueous layer was extracted three times with CH2Cl2.
The organic layer was washed with water, dried (MgSO4) and the
solvents were evaporated. The crude residue was filtered through
silica gel (petroleum ether/ether, 9:1 to 3:1). The colorless crude
syrup obtained was taken up in MeOH (25 ml). To the solution
heated at 38°C was added every day for 6 d TPSH (850 mg, 2.8
mmol). The reaction mixture was heated at 38°C for 48 h after the
sixth addition. The work-up (vide infra) gave a crude product that
was purified by flash chromatography on silica gel (petroleum
ether/ether, 8:1 to 4:1) to give compound 12 as a colorless oil (34
mg, 61%). Ϫ IR (film): ν˜ ϭ 2933, 2859, 1737, 1664 cmϪ1. Ϫ 1H
NMR (CDCl3): δ ϭ 1.40Ϫ1,85 (m, 6 H, CH2), 1.97Ϫ2.14 (m, 2 H,
CH2), 2.32 (t, 2 H, J ϭ 7.5, CH2E), 3.09 (ddd, 1 H, J ϭ 6.6, 9.1,
13.5, CH2S), 3.26 (ddd, 1 H, J ϭ 5.3, 9.1, 13.5, CH2S), 3.65 (s, 3
H, E), 3.89 (tt, 1 H, J ϭ 5.3, 8.4, CHS), 7.41Ϫ7.48 (m, 4 H, Ph),
7.53Ϫ7.60 (m, 2 H, Ph), 7.94Ϫ8.01 (m, 4 H, Ph). Ϫ 13C NMR
(CDCl3): δ ϭ 24.7, 26.4, 26.5, 33.9, 34.7, 35.1, 43.7, 51.5, 127.2,
127.3, 128.6, 133.35, 133.4, 137.0, 137.1, 174.0, 191.6, 191.8. Ϫ
HRMS: calcd. for C16H21O3S2 325.0932; found 325.0918.
[4]
[5]
[6]
L. Packer, E. H. Witt, H. J. Tritschler, Free Rad. Biol. Med.
1995, 19, 227Ϫ250 and references cited therein.
[6a] J. D. Elliott, J. Steele, W. S. Johnson, Tetrahedron Lett. 1985,
26, 2535Ϫ2538. Ϫ [6b] A. V. Rama Rao, M. K. Gurjar, K. Gary-
ali, T. Ravindranathan, Carbohydr. Res. 1986, 148, 51Ϫ55. Ϫ
[6c] R. B. Menon, M. A. Kumar, T. Ravindranathan, Tetrahedron
Lett. 1987, 28, 5313Ϫ5314. Ϫ [6d] M.H. Brookes, B. T. Golding,
[6e]
J. Chem. Soc., Perkin Trans. 1 1988, 9Ϫ12. Ϫ
J. S. Yadav, S.
V. Mysorekar, S. M. Pawar, M. K. Gurjar, J. Carbohydr. Chem.
1990, 9, 307Ϫ316. Ϫ [6f] P.C.B. Page, C.M. Rayner, I. O. Suther-
[6g]
land, J. Chem. Soc., Perkin Trans. 11990, 1615Ϫ1618. Ϫ
A.
S. Gopalan, H.K. Jacobs, J. Chem. Soc., Perkin Trans 1 1990,
[6h]
1897Ϫ1900. Ϫ
B. Adger, M. T. Bes, G. Grogan, R.
McCague, S. Pedragosa-Moreau, S. M. Roberts, R. Villa, P. W.
H. Wan, A. J. Willets, J. Chem. Soc., Chem. Commun. 1995,
[6i]
1563Ϫ1564. Ϫ
Y.R. Santosh Laxmi, D.S. Iyengar, Synthesis
[6j]
1996, 594Ϫ596. Ϫ
M. Bezbarua, A.K. Saikia, N. C. Barua,
[6k]
D. Kalita, A.C. Ghosh, Synthesis 1996, 1289Ϫ1290. Ϫ
B.
Adger, M. T. Bes, G. Grogan, R. McCague, S. Pedragosa-Mo-
reau, S. M. Roberts, R. Villa, P. W. H. Wan, A. J. Willetts,
Bioorg. Med. Chem. 1997, 5, 253Ϫ261.
[7]
A similar strategy has been reported for the synthesis of (ϩ)-
5Ϫ6-Gingerol: T. Le Gall, J. P. Lellouche, J. P. Beaucourt, Tetra-
hedron Lett. 1989, 30, 6521Ϫ6524.
[8] [8a]
´
R. Gree, J. P. Lellouche, Adv. Met.-Org. Chem. 1995, 4,
[8b]
´
R. Gree, Synthesis 1989, 341Ϫ355.
129Ϫ273 Ϫ
[9]
´
´
A. Monpert, J. Martelli, R. Gree, R. Carrie, Tetrahedron Lett.
1981, 22, 1961Ϫ1965.
[10]
W. A. Donaldson, Aldrichim. Acta 1997, 30, 17Ϫ24.
[11] [11a]
´
´
R. Gree, M. Laabassi, P. Mosset, R. Carrie, Tetrahedron
Methyl Lipoate (13): To a suspension of potassium carbonate (55
mg, 0.40 mmol) in MeOH (4 ml) was added for 2.5 h a solution of
dithiobenzoate 12 (34.3 mg, 0.08 mmol) in MeOH (12 ml). The
[11b]
Lett. 1984, 25, 3693Ϫ3696. Ϫ
Thesis, University of Rennes, 1984.
P. Mosset, Doctor Ingenior
[12]
It is worth noting that the reaction carried out with the
Eur. J. Org. Chem. 1998, 1949Ϫ1954
1953

´
´
C. Crevisy, B. Herbage, M.-L. Marrel, L. Toupet, R. Gree
FULL PAPER
[15]
[16]
[17]
thioacetate analogue of 7 and thioacetic acid led to the transoid
dithioacetate with a lower yield (< 55%), along with a cisoid
isomer when a slight excess of acid was used or with a third
unidentified by-product when a large excess of acid was used.
M. Franck-Neumann, D. Martin, F. Brion, Angew. Chem. 1978,
90, 736Ϫ737; Angew. Chem. Int. Ed. Engl. 1978, 17, 690Ϫ691.
N. J. Cusack, C.B. Reese, A. C. Risius, B. Roozpeikar, Tetra-
hedron 1976, 32, 2157Ϫ2162.
[13]
Atomic coordinates for structure 9 have been deposited with
J. W. Hamersma, E.I. Snyder, J. Org. Chem. 1965, 30,
3985Ϫ3988.
the Cambridge Crystallographic Data Centre. The coordinates
can be obtained on request from Cambridge Crystallographic
Data Centre, 12 Union Road, Cambridge, CB2 1EZ, U.K.
[18] [18a]
G. Jaouen, A. Vessieres, S. Top, J. Am. Chem. Soc. 1985,
[18b]
`
S. Tondue, S. Top, A. Vessieres, G. Ja-
[18c]
107, 4778Ϫ4780. Ϫ
[14]
Franck-Neumann accomplishes the reactions in acetic acid as
ouen, J. Chem. Soc., Chem. Commun. 1985, 326Ϫ328. Ϫ
M. Savignac, G. Jaouen, C. A. Rodger, R. E. Perrier, B. J. Sayer,
M. J. Mc Glinchey, J. Org. Chem. 1986, 51, 2328Ϫ2332.
[98173]
solvent[15]. We have shown that the reaction can be performed
under less drastic conditions by using only 2 or 3 equivalents
of acetic acid in CH2Cl2 or hexane as solvent.
1954
Eur. J. Org. Chem. 1998, 1949Ϫ1954