Copper(I) mediated highly diastereoselective conjugate addition of Grignard reagents to functionalised cycloalkenols: A general and efficient route for the stereoselective synthesis of 5- and 6-membered ring trisubstituted cycloalkanols

Article abstract of DOI:10.1016/S0040-4020(01)00044-8

The conjugate addition of magnesium cuprates to various 2-silyloxycyclopentene and 2-silyloxycyclohexene carboxylates leads diastereoselectively to related syn-anti cyclopentanols and cyclohexanols in fair overall yields. The β-elimination occurring with free hydroxylic derivatives is also partially or totally avoided by concomitant in situ trapping of the generated enolates. Attempts to rationalise our results are discussed.

Full text of DOI:10.1016/S0040-4020(01)00044-8

TETRAHEDRON
Pergamon
Tetrahedron 57 (2001) 2155±2170
Copper(I) mediated highly diastereoselective conjugate addition
of Grignard reagents to functionalised cycloalkenols: a general
and ef®cient route for the stereoselective synthesis of 5-
and 6-membered ring trisubstituted cycloalkanols
a
Valery Dambrin, Monique Villieras, Pascal Janvier, Loõc Toupet, Hassen Amri,
a
a
b
c
Â
Â
Jacques Lebreton and Jean Villieras
È
a
a,p
Â
a
Á Â Á
Laboratoire de Synthese Organique, UMR CNRS 6513, Faculte des Sciences et Techniques, 2, rue de la Houssiniere,
BP 92208-F44322 Nantes Cedex 03, France
b
Á Â Â Â
Groupe Matiere Condensee et Materiaux, URA CNRS 804, Universite de Rennes 1, Campus de Beaulieu, F-35042 Rennes, France
c
Â
Laboratoire de Chimie Organique, Faculte des Sciences, Campus Universitaire, 1060 Tunis, Tunisie
Received 20 October 2000; revised 18 December 2000; accepted 9 January 2001
AbstractÐThe conjugate addition of magnesium cuprates to various 2-silyloxycyclopentene and 2-silyloxycyclohexene carboxylates leads
diastereoselectively to related syn±anti cyclopentanols and cyclohexanols in fair overall yields. The b-elimination occurring with free
hydroxylic derivatives is also partially or totally avoided by concomitant in situ trapping of the generated enolates. Attempts to rationalise our
results are discussed. q 2001 Elsevier Science Ltd. All rights reserved.
1. Introduction
tively (Scheme 2), probably via an addition/b-elimination
process.
Conjugate addition of organocopper reagents and related
Copper(I) mediated addition of Grignard reagents has
proved to be highly effective for C±C bond formation.1
The ability of such species to carry out the b-alkylation of
a,b-unsaturated carbonyl derivatives in a stereoselective
way has also been extensively used as a key-step in
numerous syntheses of natural products and has been widely
reviewed.2
A few years ago, this SN20 process was applied to the syn-
thesis of various cyclopentenes and cyclohexenes starting
from O-acetyl derivatives.4 In the present paper, we report
our results concerning the method we have developed to
prevent elimination. It leads to stereoselective preparations
of trisubstituted 5- and 6-membered ring cycloalkanols.
We have focused our attention on the conjugate addition
reaction of magnesium organocuprates to hydroxymethyl-
acrylates and to related cyclic analogues (Scheme 1). These
building blocks seem particularly attractive in the ®eld of
bioactive and/or natural compound synthesis (e.g. prosta-
noids, iridoids),3 as previously reported.4
2. Chemoselectivity in conjugate addition reactions to
cyclopentene and cyclohexene carboxylates
First, we have studied the reaction with 5-membered ring
cycloalkenols bearing an ester moiety as electron with-
drawing group. Several attempts were carried out starting
from O-protected derivatives 7 in order to prevent the
Although easily performed on an acyclic substrate5 1 with-
out any protection of the hydroxylic moiety, the conjugate
addition of excess (3.5 equiv.) bromomagnesium di-n-
butylcuprate exclusively leads to the formation of the SN20
like by-products 5 and 6 when performed on 3 and 4, respec-
È
Keywords: conjugate addition; Michael addition; Grignard reagents;
organocuprates; b-elimination; SN20; diastereoselectivity.
Corresponding author. Fax: 133-2-51-12-54-02;
e-mail: jean.villieras@chimie.univ-nantes.fr
p
Scheme 1.
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(01)00044-8

2156
V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
Scheme 2.
b-elimination of the intermediate OM group resulting from
the initial deprotonation step (Table 1). The reactions were
performed with halomagnesium di-dialkylcuprate, with or
without chlorotrimethylsilane6 (TMSCl) (Scheme 3).
The reaction was ®rst investigated with n-butyl organo-
metallic reagents. As shown in Table 1, the protection of
the alcohol moiety with various groups does not, itself,
avoid the b-elimination, whereas, in the presence of excess
TMSCl (2.5 equiv.), the O-silylated 1,4-adducts 8 are
obtained as major products. Indeed, increasing the bulkiness
of the starting O-silyl derivatives 7 greatly favours the
formation of 8. Furthermore, the use of a catalytic amount
of copper (I) salt seems to be quite compatible with our
method since no loss of chemoselectivity (addition vs
b-elimination) is observed.
The use of HMPA (as cosolvent) or more electrophilic halo-
silanes (TMSI, TMSBr, TMSOTf) in order to favour silyl-
ation had no in¯uence.
The catalytic conditions were applied with various Grignard
reagents on cyclopentene 7d. The SiMe2t-Bu (TBDMS)
protecting group has been extensively used in our method
and preferred to the more hindered and stable SiPh2t-Bu
(TBDPS), since epimerisation problems appeared during
the last desilylation step12 from OTBDPS derivatives.
The scope and limitations of this reaction are illustrated in
Table 2.
For primary alkyl chains including functionalized ones, the
addition/b-elimination ratio increases up to 90:10 when the
reaction was run on 7d, as shown by 1H NMR.
Table 1. Organocuprates and copper (I) mediated conjugate addition to O-protected cyclopentenes 7a±e
Cyclopentenes 7
PG
Reaction conditions
Addition/eliminationa
7a
7b
7c
7d
7d
7e
7e
7c
7c
Bn7
(n-Bu)2CuMgBr, 2 equiv.
(n-Bu)2CuMgBr, 2 equiv.
(n-Bu)2CuMgBr, 2 equiv.
(n-Bu)2CuMgBr, 2 equiv.
(n-Bu)2CuLi, 2 equiv.
(n-Bu)2CuMgBr, 2 equiv.
(n-Bu)2CuLi, 2 equiv.
(n-Bu)2CuMgBr, 2 equiv.
n-BuMgBr, 2 equiv.
No TMSCl
No TMSC
No TMSC
No TMSC
No TMSC
No TMSC
No TMSC
TMSCl 2.5 equiv.
No reactionb
7b/5aˆ45:55
0:100
0:100
0:100
0:100
0:100
8c/5aˆ75:25
t-Bu7
8
SiMe3
SiMe2t-Bu8
SiMe2t-Bu8
SiPh2t-Bu8
SiPh2t-Bu8
8
SiMe3
SiMe3
8
LiCuBr2, 2 equiv.c
TMSCl 2.5 equiv.
TMSCl 2.5 equiv.
TMSCl 2.5 equiv.
8c/5aˆ75:25
8d/5aˆ90:10
8e/5aˆ95:5
7d
7e
SiMe2t-Bu8
SiPh2t-Bu8
n-BuMgBr, 2 equiv.
LiCuBr2, 2 equiv.c
n-BuMgBr, 2 equiv.
LiCuBr2, 2 equiv.c
a
Estimated by 1H NMR.
Cyclopentene 7a was totally recovered.
Used as a previously prepared 1N solution in THF.
b
c

V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
2157
Scheme 3.
Table 2. Copper(I) mediated addition of Grignard reagents to cyclopentene 7d
RM
T 8C
235
235
235
210
210
235
Addition (%)
Elimination (%)a
MeMgBr
EtMgBr
n-BuMgBr
i-PrMgCl
t-BuMgCl
EtO(H3C)CH±O± (CH2)5±MgCl9
9a (90)
9b (90)
8d (90)
9c (68)
9d (70)
9e (90)
10a (10)
10b (10)
5a (10)
10c (32)
10d (30)
10e (10)
235
235
9f (90)
9g (90)
10f (10)
10g (10)
Aryl and vinyl Grignard reagent
EtO2C± (CH2)3±Zn±I11
No reaction
No reaction
rt
a
1
Estimated by H NMR.
With secondary (i-Pr) and tertiary (t-Bu) alkyl chains, the
reaction gives a ca. 1:1 mixture of addition/elimination
compounds when performed at 2408C or at a lower
temperature. Raising the temperature to 2108C leads to
ca. 70:30 ratio of 9c/10e or 9d/10d, respectively, probably
by favouring the O-silylation kinetics. From these results, it
is clear that the chemoselectivity of such a reaction must be
discussed in terms of concurrent kinetics of b-elimination
and O-silylation processes.
case, the reaction must be performed at 2108C, using a
larger excess (2.8 equiv.) of Grignard reagent. It leads to
the exclusive formation of the conjugate addition products,
except with the less bulky 4a (20% elimination, Scheme 4),
as silyl ketene acetals 11, which are quite stable under
hydrolysis conditions of the Grignard reaction (NH4Cl/
H2O) (Table 3). Under these conditions, no cyclohexene
carboxylate 4 is recovered. In the absence of TMSCl, only
the elimination product is obtained.
This reaction has been applied with success to various
protected functional (alcohol, ketone, aldehyde) primary
Grignard reagents, leading to the formation of polyfunc-
tional compounds of high synthetic potential. However, it
always failed with aryl and vinyl organomagnesium
reagents.
As noted for cyclopentene analogues, the protection of the
hydroxyl group as a bulky silyl ether and the presence of an
excess TMSCl to trap the intermediate enolate are required
to improve the formation of 11, although no b-elimination
occurs with OTBDMS or OTBDPS derivatives in this case.
The particular stability of compounds 11 can obviously be
related to these results and it can be concluded that exo-
cyclic silyl ketene acetals are much more stable in the cyclo-
hexane than in the cyclopentane series.
With O-silylcyclohexene carboxylates 4a±c, no reaction
occurred when carried out under the same catalytic con-
ditions (n-BuMgBr (2 equiv.) in the presence of LiCuBr2
(0.1 equiv.) and TMSCl (2.5 equiv.) at 2408C). In this
Besides, with regard to cyclopentene 3, the lowered reactivity
Scheme 4.

2158
V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
Table 3. LiCuBr2 mediated conjugate addition of Grignard reagents to cyclohexenes 4a±c
Cyclohexenes 4
PG
SiMe3
R
Addition/elimination
Silyl ketene acetal
4a
4b
4b
4c
4b
4b
4b
4b
4b
n-Bu
n-Bu
n-Bu
n-Bu
Me
80:20a
11a
±
11b
11c
±
11d
11e
±
SiMe2t-Bu
SiMe2t-Bu
SiPh2t-Bu
SiMe2t-Bu
SiMe2t-Bu
SiMe2t-Bu
SiMe2t-Bu
SiMe2t-Bu
0:100b
100:0
100:0
No reaction
100:0
100:0
No reaction
100:0
Et
i-Pr
t-Bu
EtO(CH3)CH±O± (CH2)5±
11f
4b
4b
SiMe2t-Bu
SiMe2t-Bu
100:0
100:0
11g
11h
a
Estimated by 1H NMR.
Attempt performed without TMSCl.
b
been established by performing nOe experiments in 1H
NMR at 400 MHz. Several measurements have shown a
positive and mutual effect (16%) between H±C2 and the
protons of SiMe3, whereas no spatial interaction have been
measured between the latter and H±C6 (Scheme 5). These
observations lead us to assign the exclusive (Z) con®gu-
ration implying the conjugate addition reaction to occur
only on the s-trans form of the substrate.
Scheme 5.
3. Steric course and diastereoselectivity of the reactions
of cyclohexene 4 (harder conditions of reaction and no reac-
tion with methyl and t-butyl reagents) could be considered
as the result of an initial lower cycle strain and a larger steric
hindrance in the initial cyclohexene ring.
In addition to the observed chemoselectivity with O-silyl
derivatives 7c±d and 4b±c, we have found that the corre-
sponding 1,4-adducts 9 and 11 were obtained with high
diastereoselectivities, since no splitting of the signals were
1
2.1. Con®guration determination of 11b
observed, according to H (400 MHz) and 13C (100 MHz)
NMR. Consequently, it can be deduced that conjugate
addition reactions (and hydrolysis in the case of 9) lead,
in all cases, to one single diastereoisomer in the conditions
In a continuation of the pioneering work of C. Ainsworth et
al.13 our own results seem to be the ®rst example of stereo-
selective synthesis of disubstituted cyclohexylidene silyl
ketene acetal (one single diastereoisomer obtained, accord-
1
described. Indeed, in H NMR, b-silyloxyesters 9 have
exhibited an explicit dd (3J<5 and 10 Hz) around 2.2 ppm
for H±C1 whereas, after silica gel chromatography using
hexane/CH2Cl2 (95:5), 11c collapsed to ester 12c (one
single diastereoisomer) in 78% yield (Scheme 6), which
1
ing to H and 13C NMR). Apart from the relative con®gu-
ration of H±C2 and H±C6 that will be discussed in the
following part of this paper, the conformation of 11b has
Scheme 6.
Scheme 7.

V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
2159
Scheme 8.
shows similar spectral data, e.g. dd (3J<3 and 10 Hz)
around 2.15 ppm for H±C1.
X-Ray analyses of compounds 12b and 13b have ®nally
allowed the exclusive relative con®guration of H±C1±C2±
H (syn) and H±C1±C5±H (anti) (or H±C1±C6±H (anti)) to
be established unambiguously (Scheme 8).
While 12c could easily be crystallised as white needles,
efforts to obtain a 5-membered crystallised analogue were
unsuccessfull. Nevertheless, this problem has been
bypassed by the preparation of a 3,5-dinitrobenzoyl ester
13b, as white needles, starting from 9b (RˆEt), after
desilylation14 and DMAP-mediated esteri®cation with 3,5-
dinitrobenzoyl chloride (Scheme 7).
According to the absence of signi®cant modi®cation or
splitting of the signals (particularly concerning H±C1) in
1H NMR along the different subsequent steps, that con®gu-
ration could be related to corresponding cyclopentanes 9
and to cyclohexanes 11 (vide infra). Desilylation with
Table 4. Diastereoselective synthesis of cyclopentanols carboxylates
silylether
Alcohol 14
R
Yield (%)a
d.e. (%)b
9a
9b
8d
9c
9d
9e
9f
a
b
c
d
e
f
Me2c
Et2c
50
55
77
53
62
69
$95
$95
$95
$95
$95
$95
$95
$95
n-Bu2c
i-Pr2c
t-Bu2c
HO±(CH2)5±2d15
g
h
H3C±C(O)±(CH2)3±2d16
H±C(O)±(CH2)2±16
75
e
9g
a
Overall isolated yield calculated from 7d.
b
c
d
e
1
Estimated by H and 13C NMR.
Desilylation with TBAF.
Desilylation in acidic medium.
Degradation occurred during the deprotection step.
Scheme 9.

2160
V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
TBAF14 for others. Although this latter method led to a
partial epimerization of the C1, such a drawback has been
easily solved by careful chromatography, affording pure
single diastereoisomers in good overall yields (Scheme
11, Table 5). The cyclohexanols carboxylates 15d±g were
obtained without isolation of the corresponding inter-
mediate 12.
Scheme 10.
tetra-n-butylammonium ¯uoride14 (TBAF) of 8d and 9a±d
compounds or in acidic medium15,16 for 9e±g yielded syn±
anti alcohols in moderate-to-good overall yields and with
excellent diastereoselectivities (d.e.$95%) (Table 4,
Scheme 9).
4. Conjugate addition reactions on a b,b0-disubstituted
acyclic analogue
We have also explored the behaviour of a b,b0-disubstituted
acyclic analogue of cyclopentenes 3 and cyclohexenes 4 in
conjugate addition reactions. The compound 17b was selec-
tively prepared by a known procedure17 (Scheme 12) as 3:97
E/Z mixture.
Generating the cyclohexanols 15 from the corresponding
silylketene acetals 11 has proved to be more tricky, as all
attempts to chromatography compounds 11b,d±g over
silica gel have led to the formation of the corresponding
b-elimination compounds. Neither the presence of triethyl-
amine, nor the use of basic alumina led to a successful
puri®cation without such problems. We next decided to
perform the silylketene acetal hydrolysis by other ways.
Among the methods tried for this purpose (various organic
and inorganic acids, TBAF, Lewis acids,¼), ®nally, glacial
acetic acid (1 equiv.) in `wet' THF smoothly afforded syn±
anti b-silyloxyesters 12 without the formation of b-elimi-
nation side-product (Scheme 10).
Conjugate addition of excess (3.5 equiv.) bromomagnesium
di-n-butyl cuprate was performed on 17b, affording exclu-
sively (E) ethyl 4-methyl-oct-2-ene-3-carboxylate 18 in a
66% yield, while using our improved procedure with 19
®nally yielded the desired b-hydroxy ester 20 as a ca.
5:3:2 mixture of isomers (Scheme 13).
5. Discussion
As described for 5-membered ring analogues, the last depro-
tection step on 12 was accomplished in acidic medium15,16
for b-silyloxyester bearing a functionalised chain or with
5.1. Chemoselectivity and reactivity
As described in the last example (vide supra, Scheme 13),
Scheme 11.
Table 5. Diastereoselective synthesis of cyclohexanols carboxylates
R
Major epimer
Minor epimer
15/16
Yield (%)a
d.e. (%)
n-Bu
Et
i-Pr
15b
15d
15e
16b
16d
16e
95:5
88:12
92:8
±
±
±
70
84
64
67
65
96e
90
76
84
b
HO±(CH2)5±
H3C±C(O)±(CH2)3±
H±C(O)±(CH2)2±
15f15
15g15
15h16,d
±
±
$95c
$95c
$95c
b
±
a
Overall isolated yield calculated from 4b.
b
c
d
No epimerization occurs, since deprotection was performed in acidic medium.
Estimated by 1H and 13C NMR.
The hydrolysis of the ketene acetal and the dioxolane moieties were performed simultaneously using the (HCl/dioxan/KIO4) procedure and gave crude b-
silyloxyester 15h which is unstable.
Crude yield.
e
Scheme 12.

V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
2161
Scheme 13.
Scheme 14.
the SN20 by-product formation seems to be a quite general
case when performing conjugate addition of magnesium
organocuprates on b-substituted 2-hydroxyalk-1-ene-1-
carboxylates.
Interestingly, a comparison between 1 and 17b leads us
to the conclusion that the b-substitution of the alkene is
of great in¯uence on the global course of the conjugate
addition reaction.
Ê
For the cuprate (or Grignard reagent) addition, initial depro-
tonation of unprotected substrates seems likely to happen,
resulting in the formation of a potentially leaving group, e.g.
XMgO2. Further, organometallic reagent addition leads to
the conjugate addition reaction itself, affording enolate 21
that stands in balance with the related carbeniate 22, respon-
sible of the b-elimination (Scheme 14).
In fact, a methyl group (Van der Waals radiusˆ2 A) seems
to involve strong enough gauche interaction so as to dis-
favour the formation of the enolate, whereas an hydrogen
atom (Van der Waals radiusˆ1.2 A) does not induce this
Ê
kind of allylic strain (Scheme 15).
The exclusive (Z) con®guration of the silylketene acetal 11
clearly demonstrates that s-trans is the exclusive reactive
conformation of the substrate. In fact, calculations
performed with AM-1 program on the two conformations
of 4b have shown virtually no DHf difference. On the other
hand, in both cases, the a,b-unsaturated carbonyl systems is
found not to be planar in the initial state, allowing 66 and
1408 dihedral angles, respectively, for s-cis 4b and s-trans
4b. In addition to those considerations, we found that
the lowered LUMO energy level of 7d, compared with
the one of 4b (0.10152 eV vs 0.19865 eV), is compatible
with our experimental observations that 5-membered ring
In the case of 7c±e or 4a±c, the bulky O-silyl moiety tends
to shift this equilibrium in the direction of the preferential
formation of 21, due to a rate-lowering effect implied by
strong steric interactions. It may also be considered
that bulky O-silyl derivatives prevent 22 from reaching a
conformation, favourable to a b-elimination favourable
conformation (syn or anti).
Consequently, the 21±22 tautomeric balance rate is greatly
slowed down and subsequent rate limiting trapping with
TMSCl can more easily occur (Scheme 14).
From this point of view, the absence of b-elimination in
6-membered ring series (compared with 5-membered
ones) can be regarded as resulting from global, a lowering
of steric strains leading to a more stable enolate form and
enabling an easier trapping.
Scheme 15.

2162
V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
Scheme 16.
Scheme 17.
cycloalkenols appear to be more reactive than 6-membered
ring analogues.
6. Conclusion
In the present paper, we have described a new route to
functionalised 5- and 6-membered ring cycloalkanols
substituted on three contiguous atoms leading to total
stereocontrol in the formation of two asymmetric centres
in high overall yields. This ef®cient and rapid methodology
also induces high chemoselectivity by enabling the mini-
misation or the absence of the competitive SN20 side-
reaction.
5.2. Diastereoselectivity
From the global course of the reaction, both cuprate addition
and silylketene acetal hydrolysis have proved to be highly
stereoselective steps.
Quite unexpected results from the hydrolysis reaction
remain unclear, while selectivity in cuprate addition can
readily be explained. In the case of almost planar18 cyclo-
pentene carboxylates 7c±e, initial attack occurs by the less
hindered face of the substrates. Hence, the anti adducts
(with regard to TBDMSO) are exclusively obtained
(Scheme 16).
The use of a copper (I) catalyst and the ability to add new
functions with Grignard reagents of choice seems to be
particularly attractive in the ®eld of multi-step synthesis.
Furthermore, preliminary experiments have shown a total
compatibility while starting from enantioenriched substrate
(S)-3a20 since initial e.e. and absolute con®guration21 was
recovered on (S)-14c (Scheme 17).
Diastereoselectivity in cuprate addition to 6-membered ring
analogues seems to be less predictable, since increasing the
size of the ring unarguably lead to a greater conformational
freedom. In our special case, the bulky t-butyldimethylsilyl-
oxy group tends to be in a pseudoaxial position in order to
avoid strong allylic strain19 with the ester moiety (Scheme
16). This conformation allows the cycle to behave like the
cyclopentene analogue in conjugate addition reaction,
involving 1,3-stereocontrol leading to kinetic axial anti
alkylation. However, the below-situated C4 methylene
group can undergo steric interactions, which could explain
the harder conditions required for the conjugate addition
(compared to the cyclopentene analogue) and the impossi-
bility to add the bulky t-butyl group (see Tables 2 and 3).
7. Experimental
7.1. General
All reactions involving anhydrous conditions were con-
ducted in dry glassware under a positive argon (oxygen-
free) atmosphere and monitored by TLC (Kieselgel 60F254
MERCK Art. 5735 aluminium sheets) and/or GC unless
otherwise indicated. Solvents were distilled under nitrogen
immediately prior to use. THF and ethyl ether were dried by
distillation from sodium benzophenone ketyl, triethylamine
and methylene chloride from calcium hydride, ethyl acetate
from phosphorous pentoxide, and methanol from mag-
nesium methoxide. n-Hexane was stirred with conc.
sulphuric acid for a period of two days, treated with
Concerning the anti hydrolysis (with regard to TBDMSO)
of the related silylketene acetals, 1,2-stereocontrol from the
O-silyl group may also reasonably be assumed.

V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
2163
potassium carbonate, washed with brine and dried over
magnesium sulphate before distillation. Grignard and
organolithium reagents except MeMgBr, purchased from
Aldrichw, were prepared by known procedures and stored
under argon atmosphere. They were titrated prior to use,
e.g. with a 1 M solution of benzyl alcohol in toluene and
2,20-biquinoline (Grignard reagents) or o-phenantrolin
(organolithium reagents) as indicators.
added. The mixture was stirred at rt for 20 min. before
hydolysis with 20 mL of aqueous NaHCO3 and extraction
with diethyl ether. After removing the solvents, the 2,2,2-
trichloroacetamide was precipitated with cold hexane and
®ltrated, the resulting solution was concentrated and the
residue was puri®ed by a Kugelrohr distillation. Colourless
liquid (1.52 g, 62%), bp 50±758C/0.1 mmHg. 1H NMR
(200 MHz, CDCl3/TMS) d 7.35 (5H, m, aromatic ring),
6.95 (1H, td, Jˆ2.6, 0.6 Hz), 5.2 (2H, s), 5.10 (1H, m),
4.2 (2H, q, Jˆ7.2 Hz), 2.6 (1H, m), 2.4±2.2 (2H, m), 1.9
(1H, m), 1.3 (3H, t, Jˆ7.2 Hz). 13C NMR (50 MHz, CDCl3)
d 164.2 (CO2), 145.8, 136.1, 129.3 (Ar), 129.1 (2£Ar),
128.9 (Ar), 128.7 (Ar), 128.6 (Ar), 77.8, 61.5, 46.9, 32.0,
30.9. MS m/z (CI) 250 (M2NH41), 233 (M2H1). IR nmax
(thin ®lm) 1734 (s, CvO), 1653 (s, CvC). Calculated for
C15H18O3: C, 73.15; H, 7.37. Found: C, 73.28; H, 7.25.
NMR spectra were recorded on a BRUKER AC200 (1H:
200 MHz and 13C (fully decoupled): 50 MHz) or a
BRUKER ARX400 (1H: 400 MHz and 13C (fully
decoupled): 100 MHz) apparatus. The chemical shifts are
given in ppm using tetramethylsilane (TMS) as internal
standard. The coupling constants (J) are expressed in Hz,
while the multiplicity of the signals are abreviated as s for
singlet, d for doublet, dt for doublet of triplets, m for multi-
plet. IR spectra were recorded neat in KBr cells on a
BRUKER IFS 45 WHR Fourier Transform Spectrometer.
The wave numbers (n) are given in cm21. GC-MS spectra
were obtained on a HP 5890 chromatograph ®tted with HP 1
(0.33 mm£12 m) capillary column and to a HP 5889A
quadripolar spectrometer in electronic impact (70 ev) or in
chemical ionization (500 ev) with NH3 gas. The fragmen-
tation peaks are given in relative intensity (%). MERCK
Geduran SI60 Art. 10832 silica gel was used for preparative
scale chromatography and MERCK Geduran SI60 Art.
11567 for ¯ash-chromatography. Gas chromatography
were performed on a CARLO ERBA 4100 chromatograph
®tted with a RLS 300 ALLTECH (polymethylphenyl-
siloxane) capillary column (25 mm£0.25 mm). The maxi-
mum temperature allowed was 2808C, the carrier gas was
N2 and the detection was made by FID. Melting points were
determined on a RCH (C. Reichert) microscope equipped
with a KOFLER heating system, and are uncorrected.
Optical rotation values were measured in a 100 mm cell
on a Perkin±Elmer 341 polarimeter under Na lamp radia-
tion. X-Ray analyses were performed by L. Toupet from the
7.2.2. 5-tert-Butoxy-cyclopent-1-enecarboxylic acid ethyl
ester 7b7. The title compound was prepared by the same
procedure used for the synthesis of 7a using tert-butyl-
2,2,2-trichloroacetimidate, chromatographied with hexane/
ethyl acetate (95:5). Colourless liquid (4.3 g, 92%). 1H
NMR (200 MHz, CDCl3/TMS) d 6.9 (1H, td, Jˆ2.48 Hz),
4.9 (1H, m), 4.18 (2H, qd, Jˆ7.12, 5.3 Hz), 2.5 (1H, m),
2.2±1.8 (3H, m) 1.4 (9H, s), 1.15 (3H, t, Jˆ7.12 Hz). 13C
NMR (50 MHz, CDCl3) d 164.6, 142.0, 137.8, 75.5, 60.5,
33.9, 31.9, 30.8, 28.4, 14.2. MS m/z (CI) 230 (M2NH41),
213 (M2H1). IR nmax (thin ®lm) 1732 (s, CvO), 1655 (s,
CvC). Calculated for C12H20O3: C, 67.89; H, 9.50. Found:
C, 67.55; H, 9.32.
7.2.3. 5-(Trimethyl-silanyloxy)-cyclopent-1-enecarboxylic
acid ethyl ester 7c. Typical O-silylation procedure: To a
solution of 3b (3.9 g, 25 mmol, 1 equiv.) and imidazole
(2.12 g, 31.25 mmol, 1.25 equiv.) in dry N,N-dimethyl-
formamide (20 mL) at 08C chlorotrimethylsilane (TMSCl)
(3.8 mL, 30 mmol, 1.2 equiv.) was slowly added. The reac-
tion mixture was allowed to warm up to rt and stirred for an
additional 24 h period. Then, supplementary TMSCl
(30 mmol) was added to the mixture and stirred until
completion (24 h). After hydrolysis with water (100 mL),
extraction with ethyl ether, the combined organic layers
were washed with brine and dried over MgSO4. Removal
of the solvents in vacuo yielded 7c, which was distillated
under reduced pressure. Colourless liquid (4.02 g, 74%), bp
Á
University of Rennes I (Groupe Matiere Condensee et
Materiaux, URA CNRS 804). The samples were studied
Â
Â
on an automatic diffractometer CAD4 ENRAF-NONIUS
with graphite monochromatised Mo Ka radiation. Atomic
scattering factors are from International Tables for X-ray
Crystallography (1974). The calculations were performed
on a Hewlett Packard 9000-710 for structure determination
(Sheldrick, 1985) and on a Digital Micro VAX 3100 com-
puter with the MOLEN package (Enraf-Nonius, 1990) for
re®nement and ORTEP calculations. Elemental analyses
were performed at the `Service Central des Analyses du
CNRS', B.P. 22, 69390 VERNAISON (France). The
samples were sent in sealed tubes.
1
53±548C/0.1 mmHg. H NMR (200 MHz, CDCl3/TMS) d
6.9 (1H, t, Jˆ2.6 Hz), 5.0 (1H, m), 4.2 (2H, q, Jˆ7.1 Hz),
2.8±2.6 (1H, m), 2.4±2.2 (2H, m), 1.9±1.7 (1H, m), 1.3 (3H,
t, Jˆ7.1 Hz), 0.14 (9H, s, SiMe3). 13C NMR (50 MHz,
CDCl3) d 164.3, 146.7, 138.8, 75.1, 60.3, 34.0, 30.8, 14.1,
20.1 (SiMe3). MS m/z (rel. int., %) 213 (82), 185 (57), 75
(100). IR nmax (thin ®lm) 1719 (s, CvO), 1637 (CvC),
1065 (OSi). Calculated for C11H20O3Si: C, 57.85; H. 8.83.
Found: C, 57.82; H, 9.03.
7.2. Starting material syntheses
Cycloalkenols 3 and 4 as well as acrylate 1 were prepared
according to the previously described procedure22 improved
in our laboratory.5
7.2.4. 5-(tert-Butyl-dimethyl-silanyloxy)-cyclopent-1-ene-
carboxylic acid methyl ester 7d. The same procedure
described for the preparation of 7c was used with only
1.1 equiv. of dimethyl-tert-butylchlorosilane. Colourless
7.2.1. 5-Benzyloxy-cyclopent-1-enecarboxylic acid ethyl
ester 7a7. To a solution of 3b (1.42 g, 10 mmol, 1 equiv.) in
hexane (20 mL), prepared benzyl-2,2,2-trichloroaceti-
midate (5.05 g, 20 mmol, 2 equiv.) and boron tri¯uoride
diethyl etherate (200 mL, 1.6 mmol, 0.15 equiv.) were
1
liquid (10.35 g, 93%), bp 66±688C/0.09 mmHg. H NMR
(200 MHz, CDCl3/TMS) d 6.90 (1H, t, Jˆ2.4 Hz), 5.05
(1H, m), 3.73 (3H, s), 2.8±2.5 (1H, m), 2.35 (1H, m),
2.3±2.1 (1H, m), 1.8 (1H, m), 0.88 (9H, s), 0.2 (3H, s),

2164
V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
0.1 (3H, s). 13C NMR (50 MHz, CDCl3) d 165.8, 147.2,
139.8, 76.6, 51.9, 35.2, 31.7, 26.7, 19.0, 24.0. MS m/z
(rel. int., %) 199 (69), 171 (18), 89 (100), 75 (41), 73
(18), 65 (17), 59 (20). IR nmax (thin ®lm) 1726 (s, CvO),
1635 (s, CvC), 1069 (OSi). Calculated for C13H24O3Si: C,
60.89; H, 9.43. Found: C, 60.71, H, 9.48.
ABX3 system, Jˆ7.1 Hz). 13C NMR (100 MHz, C6D6) d
166.7, 142.1, 136.6, 136.5, 135.1, 134.7, 133.5, 130.1,
129.7, 128.4, 127.8, 64.1, 60.3, 31.5, 27.4, 26.1, 19.8,
16.4, 14.4. MS m/z (rel. int., %) 351 (M257, 100), 199
(68), 139 (59), 77 (51). IR nmax (thin ®lm) 3071 (Ar),
1717 (CvO), 1647 (CvC), 1259 (C±O). Calculated for
C25H32O3Si: C, 73.49; H, 7.89. Found: C, 73.65; H, 7.67.
7.2.5. 5-(tert-Butyl-diphenyl-silanyloxy)-cyclopent-1-ene-
carboxylic acid methyl ester 7e. The same procedure
described for the preparation of 7d was used with diphenyl-
tert-butylchlorosilane affording crude O-silyl ether 7e
which was puri®ed by chromatography on silica gel using
Hexane/CH2Cl21ˆ1:1 as eluent. Colourless visqueous oil
(23.1 g, 69%). H NMR (200 MHz, CDCl3/TMS) d 7.7±
7.4 (10H, m, Ar), 6.97 (1H, t, Jˆ2.7 Hz), 5.13 (1H, m),
3.54 (3H, s), 2.7±2.5 (1H, m), 2.4±2.2 (1H, m), 2.0±1.8
(2H, m), 1.06 (9H, s). 13C NMR (50 MHz, CDCl3) d
165.0, 147.0, 139.0, 135.9±127.4 (Ar), 76.4, 50.9, 34.0,
30.9, 26.9, 19.3. MS m/z (rel. int., %) 323 (100), 295 (36),
199 (27), 139 (89). IR nmax (thin ®lm) 3071 (Ar), 1721
(CvO), 1635 (CvC), 1258 (C±O). Calculated for
C23H28O3Si: C, 72.59; H, 7.42. Found: C, 7. 55; H, 7.23.
7.2.9. (2S) 5-Hydroxy-cyclopent-1-enecarboxylic acid
methyl ester (S)-3a.20 A solution of (R,S) alcohol 3a
(2.5 g, 17.6 mmol, 1 equiv.) in 25 mL of vinyl acetate was
stirred at rt with 250 mg of lipase Pseudomonas sp. The
reaction was monitored by gas chromatography and stopped
after ca. 50% of conversion (20 h). The lipase was removed
by ®ltration over celite and the mixture was concentrated
under reduced pressure. Puri®cation by a ¯ash chroma-
tography with hexane/ether (4:1) afforded (S) alcohol (and
(R) acetate). Colourless liquid (1.24 g, 49%), e.e.21 50%.
20
[a]D ˆ217.4 (cˆ0.99, CHCl3). 1H NMR (200 MHz,
CDCl3/TMS) d 6.92 (1H, t, Jˆ2.3 Hz), 5.1 (1H, m), 3.8
(3H, s), 2.7±2.55 (1H, m), 2.5±2.25 (2H, m), 2.0±1.75
(1H, m). 13C NMR (50 MHz, CDCl3) d 165.0, 146.38,
137.88, 74.75, 51.11, 32.08, 30.63. MS m/z (rel. int., %)
115 (100), 114 (45), 77 (37), 71 (32), 60 (36), 53 (24), 51
(24). IR nmax (thin ®lm) 3480 (broad, OH), 1720 (CvO),
1640 (CvC).
7.2.6. 6-(Trimethyl-silanyloxy)-cyclohex-1-enecarboxylic
acid ethyl ester 4a. The same procedure described for the
preparation of 7c was used, yielding to O-silyl ether 4a,
which was puri®ed by Kugelrohr distillation under reduced
pressure. Colourless liquid (8.72 g, 72%), bp 80±1008C/4
mmHg. 1H NMR (200 MHz, CDCl3/TMS) d 7.0 (1, m), 4.5
(1H, m), 4.1 (2H, ABX3 system), 2.35±1.30 (6H, m), 1.1
(3H, t, Jˆ7.2 Hz), 0.15 (9H, s). 13C NMR (50 MHz, CDCl3)
d 166.7, 142.1, 132.9, 62.6, 60.2, 31.5, 25.9, 17.5, 14.3, 0.4.
MS m/z (rel. int., %) 242 (M, 1), 228 (17), 227 (100), 199
(49), 181 (41), 75 (66), 73 (35). IR nmax (thin ®lm) 1715 (s,
CvO), 1646 (CvC), 1260 (C±O), 1060 (O±Si). Calculated
for C12H22O3Si: C, 59.46; H, 9.15. Found: C, 59.51; H, 9.37.
7.2.10. (2S) 5-(tert-Butyl-dimethyl-silanyloxy)-cyclopent-
1-enecarboxylic acid methyl ester (S)-7d. The title silyl
ether (S)-7d was prepared from (S)-3a and according to the
same procedure used for the racemic material. Spectral data
are identical to racemic material 7d. Colourless liquid
20
(1.65 g, 92%), e.e. 50%. bp 59±608C/0.1 mmHg. [a]D
16.2 (cˆ1.06, CHCl3).
ˆ
7.3. Conjugate addition and desilylation reactions to
5-membered ring cycloalkenols
7.2.7. 6-(tert-Butyl-dimethyl-silanyloxy)-cyclohex-1-ene-
carboxylic acid ethyl ester 4b. The same procedure
described for the preparation of 7d was used, affording
O-silyl ether 4b which was distilled under reduced pressure.
Colourless liquid (3.52 g, 86%), bp 63±648C/0.09 mmHg.
1H NMR (400 MHz, C6D6/TMS) d 6.9 (1H,m), 4.8 (1H, t,
Jˆ3.1 Hz), 4.2 (2H, ABX3 system, Jˆ7.1 Hz), 2.05 (1H,
m), 1.90 (3H, m), 1.45 (2H, m), 1.2 (3H, t, Jˆ7.1 Hz), 1.0
(9H, s), 0.27 (3H, s), 0.25 (3H, s,). 13C NMR (100 MHz,
C6D6) d 166.8, 142.2, 133.8, 63.2, 60.5, 32.2, 26.5, 26.4,
20.9, 16.6, 14.9, 23.9, 24.3. MS m/z (rel. int., %) 227
(M257, 100), 199 (56), 181 (47), 75 (79). IR nmax (thin
®lm) 1718 (s, CvO), 1648 (CvC), 1269 (C±O), 1062
(O±Si). Calculated for C15H28O3Si: C, 63.33; H, 9.92.
Found: C, 63.41; H 9.97.
All conjugate addition reactions were typically performed in
dry glassware under argon atmosphere using a 50 mL four-
necked round bottom ¯ask ®tted with a mechanical stirrer, a
thermometer, a re¯ux condenser and an addition funnel.
7.3.1. Typical Copper(I) mediated conjugate addition of
Grignard reagent procedure. To a solution of 5 mmol of
cyclopentenol in 40 mL THF at a temperature between 235
and 2108C (depending on the nature of the Grignard
reagent) was added 0.5 mL of a 1N solution of LiCuBr2 in
THF and with TMSCl (1.6 mL, 12.5 mmol, 2.5 equiv.).
After 5 min, a solution of the Grignard reagent (10 mmol,
2 equiv.) was slowly added during ca. 1 h and the mixture
was stirred for a remaining hour until completion. Then, the
mixture was quenched at 2308C with saturated aqueous
NH4Cl. After extraction with ethyl ether, the combined
organic layers were washed with brine and dried
(MgSO4). Removal of the solvents in vacuo yielded the
crude conjugate adduct which was generally used in the
desilylation step with no further puri®cation.
7.2.8. 6-(tert-Butyl-diphenyl-silanyloxy)-cyclohex-1-ene-
carboxylic acid ethyl ester 4c. The same procedure
described for the preparation of 7e was used, affording
O-silyl ether 4c which was puri®ed by chromatography on
silica gel (hexane/CH2Cl2ˆ1:1). Colourless visquous oil
(15.6 g, 77%). 1H NMR (400 MHz, C6D6/TMS) d 8.1
(2H, m, Ar), 7.9 (2H, m, Ar), 7.4 (6H, m, Ar), 7.1 (1H, m,
H6), 4.9 (1H, m), 4.15 (1H, ABX3 system), 3.85 (1H, ABX3
system), 2.1±2.0 (2H, m), 1.9 (1H, m), 1.75 (1H, m), 1.40
(1H, m), 1.30 (9H, 2s), 1.20 (1H, m), 1.0 (3H, X3 part of a
7.3.2. (anti, syn) 2-(tert-Butyl-dimethyl-silanyloxy)-5-ethyl-
cyclopentanecarboxylic acid methyl ester 9b. The title
cyclopentane was obtained from 7d following typical pro-
cedure. Colourless visquous oil (1.62 g, 88%), d.e.$95%.

V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
2165
1H NMR (400 MHz, C6D6/TMS) d 4.5 (1H, dt, Jcisˆ6.2,
Jˆ5.0 Hz, HC±OSi), 3.5 (3H, s), 3.2 (2H, pseudo q, Jˆ
7.0 Hz), 2.7 (1H, m, CH±Et), 2.3 (1H, dd, Jcisˆ6.2,
Jtransˆ8.8 Hz, H±C±CO2), 2.0±1.3 (4H, m), 1.1 (9H, s),
1.1 (3H, t, Jˆ7.0 Hz, CH3), 0.09 (3H, s), 0.08 (3H, s). 13C
NMR (100 MHz, C6D6) d 173.2 (CO2), 76.8 (SiO±C), 65.8,
56.9 (C±CO2), 50.6, 41.6 (CH±Et), 35.3, 28.7, 27.1, 19.2,
15.5 (CH3), 24.3. IR nmax (thin ®lm) 1728 (CvO), 1067
(Si±O).
(Ar±NO2). Calculated for C16H18N2O8: C, 52.46; H, 4,95;
N, 7.65. Found: C, 52.35; H, 5.07; N, 7.55.
7.3.5. (anti,syn) 2-Hydroxy-5-methyl-cyclopentanecar-
boxylic acid methyl ester 14a. The title alcohol was
prepared from 7d and desilylated by the same procedure
described for the preparation of 14b. It was puri®ed on silica
gel using hexane/ethyl acetateˆ7:3. Colourless liquid
(0.39 g, 50%), d.e.$95%. 1H NMR (400 MHz, C6D6/
TMS) d 4.4 (1H, dt, Jcisˆ4.9, 3.7 Hz, CH±OH), 3.5 (3H,
s), 2.88 (1H, broad s, OH), 2.6 (1H, dtq, Jtransˆ10.3, 8.5,
6.7 Hz, CH±Me), 2.16 (1H, dd, Jtransˆ10.3, Jcisˆ4.9 Hz,
CH±CO2), 2.1 (1H, ddt, Jˆ12.8, 8.5, 6.7 Hz), 1.78 (2H,
m), 1.1 (4H, m). 13C NMR (100 MHz, C6D6) d 174.3,
74.7 (CH±OH), 57.9 (CH±CO2), 51.0, 35.4 (CH±Me),
34.2, 31.8, 19.9. MS m/z (rel. int., %) 101 (M257, 100),
87 (34), 69 (47), 41 (17). IR nmax (thin ®lm) 3467 (broad,
OH), 1740 (CvO). Calculated for C8H14O3: C, 60.74; H,
8.92. Found: C, 60.35; H, 9.23.
7.3.3. (anti, syn) 2-Hydroxy-5-ethyl-cyclopentanecarboxylic
acid methyl ester 14b. Typical desilylation procedure.14 To
a solution of crude O-silyl 5-ethylcyclopentane 9b (ca.
5 mmol) in 10 mL THF at 08C was added over a period of
30 min, a solution of tetra-n-butylammonium ¯uoride tri-
hydrate (4.1 g, 13 mmol, 2.6 equiv.) in 10 mL THF. The
mixture was allowed to warm up to rt and stirred for the
remaining 2 h until completion. Then, the mixture was
quenched with saturated dilute HCl. After extraction with
ethyl acetate, the combined organic layers were washed
with brine and dried (MgSO4). Removal of the solvents in
vacuo yielded the crude alcohol 14b as a single diastereo-
isomer which was puri®ed by chromatography on silica gel
(hexane/ethyl acetateˆ4:1). Colourless liquid (0.47 g,
7.3.6. (anti,syn) 2-Hydroxy-5-n-butyl-cyclopentanecar-
boxylic acid methyl ester 14c. The title alcohol was
prepared by previously described procedures from 8d. It
was puri®ed on silica gel using hexane/ethyl acetateˆ3:2
1
1
55%), d.e.$95%. H NMR (400 MHz, C6D6/TMS) d 4.33
as eluent. Pale yellow liquid (0.77 g, 77%), d.e.$95%. H
(1H, dt, Jcisˆ5.2, 4.0 Hz, CH±OH), 3.40 (3H, s,), 2.44 (1H,
dtt, Jtransˆ9.9, 8.5, 5.8 Hz, CH±Et), 2.2 (1H, dd, Jtransˆ9.9,
Jcisˆ5.2 Hz, CH±CO2), 1.97 (1H, ddt, Jˆ14.3, 8.5, 6.6 Hz),
1.70±1.65 (2H, m), 1.50 (1H, ddq, Jˆ14.8, 7.4, 7.3 Hz), 1.1
(1H, ddq, Jˆ14.8, 7.4, 7.3 Hz), 1.0 (1H, m), 0.82 (3H, 2t,
Jˆ7.3 Hz). 13C NMR (100 MHz, C6D6) d 174.6, 74.7 (CH±
OH), 55.9 (CH±CO2), 51.2, 42.1 (CH±Et), 34.1, 29, 28.3,
12.4. MS m/z (rel. int., %) 115 (M257, 100), 87 (32), 83
(44). IR nmax (thin ®lm) 3478 (broad, OH), 1737 (CvO),
1261 (C±O). Calculated for C9H16O3: C, 62.77; H, 9.36.
Found: C, 62.35; H, 9.73.
NMR (400 MHz, C6D6/TMS) d 4.4 (1H, dt, Jcisˆ5.1,
3.9 Hz, CH±OH), 3.5 (3H, s), 3.2 (1H, broad s, OH), 2.6
(1H, dtt, Jtransˆ9.9, 8.3, 5.1 Hz, CH±n-Bu), 2.3 (1H, dd,
Jtransˆ9.9, Jcisˆ5.1 Hz, CH±CO2), 2.1 (1H, ddt, Jˆ12.7,
8.3, 6.6 Hz), 1.8 (1H, m), 1.7 (1H, m), 1.6 (1H, m), 1.4±
1.1 (6H, m), 0.9 (3H, t, Jˆ6.8 Hz, CH3). 13C NMR
(100 MHz, C6D6) d 174.6, 74.7 (CH±OH), 56.3 (CH±
CO2), 51.2, 40.6 (CH±n-Bu), 35.5, 34.2, 30.5, 29.4, 23.1,
13.9. MS m/z (rel. int., %) 143 (100%, M257), 115 (57),87
(56), 83 (41), 55 (41). IR nmax (thin ®lm) 3488 (broad, OH),
1740 (CvO), 1261 (C±O). Calculated for C11H20O3: C,
65.97; H, 10.07. Found: C, 65.29; H, 10.03.
7.3.4. (anti,syn) 3,5 Dinitrobenzoic acid 3-ethyl-2-meth-
oxycarbonyl-cyclopentyl ester 13b. To a solution of 3,5-
dinitrobenzoyl chloride (0.428 g, 1.86 mmol, 2 equiv.) in
35 mL of methylene chloride, 4-dimethylaminopyridine
(DMAP) (10 mg, 0.09 mmol, 0.01 equiv.) was added and
stirred at rt for 5 min. before the addition of triethylamine
(2.53 mL, 1.86 mmol, 2 equiv.) and alcohol 14b (0.2 g,
0.93 mmol, 1equiv.). After 30 h at rt, the mixture was
quenched with 10 mL of water, extracted with diethyl
ether and washed with brine until pH 7. After drying over
MgSO4 and removing of the solvents, the crude product was
chromatographed on silica gel (hexane/CH2Cl2ˆ1:4) and
recrystalised (ether/pentaneˆ2:1). White crystals (0.310 g,
7.3.7. (anti,syn) 2-Hydroxy-5-iso-propyl-cyclopentane-
carboxylic acid methyl ester 14d. The title alcohol was
prepared by previously described procedures. It was puri®ed
on silica gel using hexane/ethyl acetateˆ1:1 as eluent. Pale
yellow oil solidi®ed at 2188C (0.53 g, 57%), d.e.$95%. 1H
NMR (400 MHz, C6D6/TMS) d 4.3 (1H, dt, Jcisˆ5.3,
4.1 Hz, CH±OH), 3.5 (3H, s), 2.8 (1H, broad s, OH), 2.6
(1H, dm, Jtransˆ9.8 Hz, CH±i-Pr), 2.5 (1H, dd, Jtransˆ9.8,
Jcisˆ5.3 Hz, CH±CO2), 1.95 (1H, dddd, Jˆ11.1, 6.3, 6.3,
5.0 Hz), 1.85 (2H, m), 1.75±1.55 (2H, m), 0.95 (3H, d, Jˆ
6.7 Hz), 0.9 (3H, d, Jˆ6.7 Hz). 13C NMR (100 MHz, C6D6)
d 175.4, 75.2 (CH±OH), 53.5 (CH±CO2), 51.5, 47.8 (CH±i-
Pr), 34.8, 32.0, 26.3, 21.4, 19.5. MS m/z (rel. int., %) 143
(38%), 129 (M257, 43), 115 (67), 111 (100), 83 (57), 55
(45), (CI) 187 (M2H). IR nmax (thin ®lm) 3466 (broad, OH),
1717 (CvO). Calculated for C10H18O3: C, 64.49; H, 9.74.
Found: C, 64.48; H, 10.15.
1
91%), mp 124±1268C (ether/pentaneˆ2:1), d.e.$95%. H
NMR (400 MHz, C6D6/TMS) d 8.7 (2H, d, Ar), 8.4 (1H, t,
Ar), 5.52 (1H, dt, Jcisˆ5.9, 3.3 Hz, CH±OPh), 3.5 (3H, s),
2.5 (1H, m, CH±Et), 2.38 (1H, dd, Jtransˆ10.1, Jcisˆ5.9 Hz,
CH±CO2), 1.9±1.7 (3H, m), 1.4 (1H, dqd, Jˆ15.3, 7.4,
7.1 Hz, CH2(Et)), 1.1 (1H, dqd, Jˆ15.3, 7.4, 7.1 Hz,
CH2(Et)), 0.9 (1H, m), 0.8 (3H, t, Jˆ7.4 Hz). 13C NMR
(100 MHz, C6D6) d 192.1 (Ph±CO2), 171.9 (CO2Me),
161.9 (Ar), 148.4 (Ar),33.5 (Ar), 128.7 (Ar), 128.3 (Ar),
122.1 (Ar), 79.5 (CH±OPh), 54.6 (CH±CO2), 51.6, 42.4
(CH±Et), 31.8, 28.8, 27.9, 12.3. MS m/z (rel. int., %) 195
(M2171, 69), 171 (17), 149 (32), 115 (100), 95 (80). IR
7.3.8. (anti,syn) 2-Hydroxy-5-tert-butyl-cyclopentanecar-
boxylic acid methyl ester 14e. The title alcohol was
prepared by previously described procedures. It was puri®ed
on silica gel using hexane/ethyl acetateˆ1:1 as eluent.
Colourless liquid (0.62 g, 62%), d.e.$95%. 1H NMR
(400 MHz, C6D6/TMS) d 4.29 (1H, dt, Jcisˆ5.7, 4.8 Hz,
CH±OH), 3.5 (3H, s), 2.74 (1H, broad s, OH), 2.70 (1H,
n
max (thin ®lm) 1738 (CvO), 1723 (CvO), 1633 (Ar), 1549

2166
V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
dd, Jtransˆ8.8, Jcisˆ5.7 Hz, CH±CO2), 2.67 (1H, ddd, Jtrans
ˆ
Colourless liquid (1.8 g, 93%), d.e.$95%. 1H NMR
(400 MHz, C6D6/TMS) d 4.4 (1H, m, CH±OSi), 3.87 (4H,
m), 3.6 (3H, s), 2.54 (1H, m), 2.4 (1H, dd, Jtransˆ9.5, Jcisˆ
5.8 Hz, CH±CO2), 2.04 (1H, m), 1.84 (1H, m), 1.70±1.58
(4H, m), 1.45±1.1 (4H, m), 1.3 (3H, s), 0.86 (9H, s), 0.02
(3H, s), 20.01 (3H, s). 13C NMR (100 MHz, C6D6) d 173.0,
110.4, 75.9 (CH±OSi), 64.9, 57.8 (CH±CO2), 51.5, 39.6,
39.2, 36.2, 35.3, 29.4, 25.9, 24.0, 22.9, 18.2, 24.3, 24.9.
MS m/z (rel. int., %) 371 (5), 330 (9), 239 (38), 133 (13),
119 (16), 89 (38), 87 (100), 75 (25), 73 (14), 43 (10), (CI)
387 (M2H). IR nmax (thin ®lm) 1745 (s, CvO), 1070 (OSi).
8.8, 16.0, 7.8 Hz, CH±t-Bu), 1.89 (2H, m), 1. 66 (1H, m),
1.29 (1H, m), 0.88 (9H, s). 13C NMR (100 MHz, C6D6) d
175.5, 75.5 (CH±OH), 51.6 (CH±t-Bu), 51.3, 51.0 (CH±
CO2), 35.1, 32.5, 27.3, 27.4. MS m/z (rel. int., %) 143
(M257, 98), 111 (100), 83 (40), 57 (28), (CI) 218
(M2NH41), 201 (M2H). IR nmax (thin ®lm) 3401 (broad,
OH), 1735 (CvO). Calculated for C11H20O3: C, 65.97; H,
10.07. Found: C, 65.52; H, 10.17.
7.3.9. (anti,syn) 2-(tert-Butyl-dimethyl-silanyloxy)-5-[5-
(1-ethoxy-ethoxy)-pentyl]-cyclopentanecarboxylic acid
methyl ester 9e. A solution of 1-chloro 7-methyl-6,8-dioxa-
decane (2.4 g, 12.5 mmol, 2.5 equiv.), prepared following
the literature, in 10 mL of THF was stirred with magnesium
(0.48 g, 20 mmol, 2 equiv.) at 45±508C for 3 h then added
to 2-silyloxycyclopentene 7d according to typical proce-
dure. The crude product obtained was used in the next
step with no further puri®cation. Colourless liquid (quant.
7.3.12. (anti,syn) 2-Hydroxy-5-(4-oxo-pentyl)-cyclo-
pentanecarboxylic acid methyl ester 14g. The dioxolane
9f (1.8 g, 4.8 mmol, 1 equiv.) was stirred at re¯ux for 5 min.
in a mixture of 65 mL of dioxane and 30 mL of 0.1N
aqueous hydrochloride with potassium periodate (1.38 g,
6 mmol, 1.25 equiv.). After cooling to rt, the mixture was
washed with water and saturated aqueous NaHCO3 until
pH$7 and extracted with ethyl acetate. After removal of
the solvents in vacuo, the crude product was treated at 08C
with 15 mL of 2N aqueous hydrochloride in 10 mL acetone
for 1 h. Then, the mixture was extracted with chloroform,
washed with aqueous NaHCO3, dried over MgSO4 and
evaporated under reduced pressure. The crude product was
¯ash-chromatographied over silica gel with hexane/diethyl
etherˆ1:1 as eluent. Colourless liquid (0.81 g, 75%),
1
yield), d.e.$95%. H NMR (400 MHz, C6D6/TMS) d 4.6
(1H, q, Jˆ5.3 Hz), 4.4 (1H, m, CH±OSi), 3.6 (3H, s), 3.58±
3.45 (4H, m), 2.5 (1H, m), 2.36 (1H, dd, Jtransˆ9.5, Jcisˆ
5.6 Hz, CH±CO2), 2.02±1.28 (12H, m), 1.30 (3H, d, Jˆ
7.1 Hz), 1.18 (3H, t, Jˆ7.0 Hz), 0.82 (9H, s), 0.00 (3H, s),
20.02 (3H, s). 13C NMR (100 MHz, C6D6) d 172.6, 99.4,
75.5 (CH±OSi), 65.1, 60.5, 57.4 (CH±CO2), 51.0, 38.8,
35.6, 34.8, 29.7, 28.9, 27.8, 26.3, 23.6, 19.7, 17.7, 15.1,
24.8, 25.4. MS m/z (rel. int., %) 359 (10), 327 (10), 313
(48), 269 (20), 255 (10), 199 (10), 159 (8), 115 (14), 89 (39),
75 (47), 73 (100), 45 (36), (CI) 434 (M2NH41). IR nmax
(thin ®lm) 1747 (s, CvO), 1065 (s, OSi).
1
d.e.$95%. H NMR (400 MHz, CD3COCD3) d 4.4 (1H,
m, CH±OH), 3.85 (1H, d, Jˆ4.6 Hz, OH), 3.6 (3H, s),
2.42 (4H, m), 2.1 (3H, s), 2.0±1.8 (2H, m), 1.75±1.45
(4H, m), 1.30±1.1 (2H, m). 13C NMR (100 MHz,
CD3COCD3) d 208.8, 174.1, 75.4 (CH±OH), 57.7 (CH±
CO2), 51.9, 44.1, 40.1, 35.9, 35.1, 30.2, 29.9, 23.3. MS
m/z (rel. int., %) 200 (2), 182 (5), 178 (10), 171 (36), 168
(10), 153 (11), 139 (27), 125 (11), 121 (10), 111 (39), 109
(19), 93 (50), 83 (38), 81 (21), 79 (14), 71 (17), 69 (10), 67
(16), 58 (12), 55 (29), 43 (100), (CI) 246 (M2NH41), 229
(M2H). IR nmax (thin ®lm) 3492 (broad, OH), 1740±1734
(s, CvO). Calculated for C12H20O4: C, 63.14; H, 8.83.
Found: C, 62.87; H, 8.96.
7.3.10. (anti,syn) 2-Hydroxy-5-(5-hydroxy-pentyl)-cyclo-
pentanecarboxylic acid methyl ester 14f.15 15 mL of 2N
aqueous hydrochloride were added at 08C (ice bath) to a
solution of silyl ether 9e in 10 mL of acetone. The mixture
was stirred for 2 h at rt before extraction with chloroform,
washing with aqueous NaHCO3 and drying over MgSO4.
The crude product was ¯ash-chromatographied over silica
gel with ethyl ether following with CHCl3. Colourless liquid
(0.8 g, 69%), d.e.$95%. 1H NMR (400 MHz, C6D6/TMS) d
4.57 (1H, broad s, OH), 4.02 (1H, m, CH±OH), 3.2 (1H,
broad s, OH), 3.15 (2H, t, Jˆ6.2 Hz), 3.1 (3H, s), 2.36 (1H,
m), 1.9 (1H, dd, Jtransˆ9.8, Jcisˆ5.2, CH±CO2), 1.7 (1H, m),
1.55±0.6 (11H, m). 13C NMR (100 MHz, C6D6) d 176.7,
74.8 (CH±OH), 63.1, 56.6 (CH±CO2), 51.4, 40.4, 35.6,
34.1, 32.9, 29.4, 28.1, 26.1. MS m/z (rel. int., %) 182 (5),
180 (7), 173 (7), 155 (13), 152 (30), 141 (47), 135 (22), 123
(25), 115 (43), 111 (78), 110 (52), 97 (23), 95 (93), 87(23),
83 (100), 79 (29), 71 (31), 67 (54), 55 (72), 41 (37), (CI) 248
(M2NH41), 231 (M2H). IR nmax (thin ®lm) 3400 (broad s,
OH), 1725 (CvO). Calculated for C12H22O4: C, 62.58; H,
9.63. Found: C, 62.27; H, 9.89.
7.3.13. (anti,syn) 2-(tert-Butyl-dimethyl-silanyloxy)-5-(2-
[1,3]dioxolan-2-yl-ethyl)-cyclopentanecarboxylic
acid
methyl ester 9g. A solution of 1-bromo 3-(1,3-dioxo-
lane)-n-propane (2.26 g, 12.5 mmol, 2.5 equiv.) in 10 mL
of THF was stirred at 308C with magnesium (0.48 g,
20 mmol, 2 equiv.) for 45 min, then added to 2-silyloxy-
cyclopentene 7d according to typical procedure. The
crude product obtained was puri®ed over silica gel (hexane/
diethyl etherˆ9:1). Colourless liquid (1.09 g, 61%),
1
d.e.$95%. H NMR (200 MHz, CDCl3/TMS) d 4.83 (1H,
broad t), 4.44 (1H, dt, Jcisˆ5.6, 3.6 Hz, CH±OSi), 3.97±
3.81 (4H, m), 3.65 (3H, s), 2.44 (1H, m), 2.37 (1H, dd,
Jtransˆ9.3, Jcisˆ5.6 Hz, CH±CO2), 2.04±1.6 (6H, m),
1.40±1.10 (2H, m), 0.84 (9H, s), 0.03 (3H, s), 0.00 (3H,
s). 13C NMR (50 MHz, CDCl3) d 172.5, 104.5, 75.6 (CH±
OSi), 64.7, 57.3 (CH±CO2), 51.1, 38.6, 34.8, 32.4, 29.7,
28.9, 25.6, 17.8, 24.7, 25.3. MS m/z (rel. int., %) 301
(M257, 32), 269 (32), 151 (23), 105 (15), 99 (15), 89
(60), 79 (14), 75 (34), 73 (100), 59 (15), 45 (15), (CI) 359
(M2H). IR nmax (thin ®lm) 1745 (s, CvO), 1065 (O±Si).
Calculated for C18H34O5Si: C, 60.30; H, 9.56. Found: C,
60.67; H, 9.33.
7.3.11. (anti, syn) 2-(tert-Butyl-dimethyl-silanyloxy)-5-[3-
(2-methyl-[1,3]dioxolan-2-yl)-propyl]-yclopentanecar-
boxylic acid methyl ester 9f. A solution of 1-chloro 4-(1,3-
dioxolane)-n-pentane (2.05 g, 12.5 mmol, 2.5 equiv.),
prepared following the literature,10 in 10 mL of THF was
stirred with magnesium (0.48 g, 20 mmol, 2 equiv.) at 35±
408C for 3 h then added (at rt) to 2-silyloxycyclopentene 7d
according to typical procedure. The crude product obtained
was puri®ed over silica gel (hexane/ethyl acetateˆ7:3).

V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
2167
7.3.14. (1R,2S,5S) 2-Hydroxy-5-butyl-cyclopentane-1-
carboxylic acid methyl ester 14c. The cyclopentanol was
prepared and puri®ed according to the same procedures
(Copper(I) mediated addition of n-butylmagnesium bro-
mide and desilylation) used for the racemic product 7b
and gave identical spectral data described for racemic
product 14c. It was puri®ed by silica gel chromatography
(hexane/diethyl etherˆ4:1). Colourless liquid (0.72 g,
72%), e.e.21 ˆ50%. [a]D20ˆ129.2 (cˆ1.085, CHCl3).
int., %) 386 (1), 357 (63), 223 (11), 181 (10), 151 (13), 147
(21), 115 (45), 75 (20), 73 (100), (CI) 401 (M2H). IR nmax
(thin ®lm) 1682 (CvC), 1081±1044 (broad, O±Si).
7.4.4. (anti,syn) 2-(tert-Butyl-diphenyl-silanyloxy)-5-n-
butyl-cyclohexanecarboxylic acid ethyl ester 12c. Typical
procedure from 4c yielded the crude silyl ketene acetal 11c
which collapsed to ester 12c as a single diastereoisomer
after chromatography on silica gel (hexane/CH2Cl2ˆ95:5).
Colourless visquous oil crystalized as white needles after
1
one week at rt (1.81 g, 78%), d.e.$95%, mp 51±538C. H
7.4. Conjugate addition reactions to 6-membered ring
cycloalkenols
NMR (200 MHz, C6D6/TMS) d 7.66±7.36 (10H, m, Ar),
4.38 (1H, m, CH±OSi), 4.02±3.89 (2H, ABX3 System),
2.3 (1H, m, CH±n-Bu), 2.15 (1H, dd, Jtransˆ10.22, Jcisˆ
3.06 Hz, CH±CO2), 2.0±1.75 (12H, m), 1.6 (3H, t, Jˆ
7.1 Hz), 1.55 (9H, s), 0.9 (3H, broad t). 13C NMR
(50 MHz, CDCl3) d 173.12, 136.04 (Ar), 136.0 (Ar),
134.8 (Ar), 134.0 (Ar), 129.5 (Ar), 127.36 (Ar), 70.12
(CH±OSi), 59.82, 54.13 (CH±CO2), 34.32, 32.7, 32.4
(CH±n-Bu), 29.9, 28.9, 27.0, 22.9, 19.4, 19.3, 14.0. MS
m/z (rel. int., %) 409 (M257, 98), 227 (46), 199 (100),
183 (29), 135 (22), 81 (19), (CI) 467 (M2H). IR nmax
(thin ®lm) 3071±3049 (Ar), 1740 (CvO). Calculated for
C29H42O3Si: C, 74.63; H, 9.07. Found: C, 74.68; H, 8.88.
Conjugate addition reactions were performed according to
the typical procedure described for 5-membered ring ana-
logues and using 2.9 equiv. of Grignard reagent at 2108C.
7.4.1. (Z) (anti)1-Butyl-3-(tert-butyl-dimethyl-silanyloxy)-
2-[ethoxy-(trimethyl-silanyloxy)-methylene]-cyclohexane
11b. Typical procedure was used from 4b, affording the title
compound without puri®cation in order to prevent from
subsequent degradation (b-elimination). Colourless liquid
1
solidi®ed at 2188C (2.15 g, quant. yield), d.e.$95%. H
NMR (200 MHz, CDCl3) d 4.8 (1H, m, CH±OSi), 3.7
(2H, ABX3 System), 2.37 (1H, m), 1.9±1.7 (4H, m), 1.6±
1.1 (8H, m), 1.2 (3H, t, Jˆ7.1 Hz), 0.9 (9H, s), 0.9 (3H,
broad t), 0.2 (9H, s), 0.05 (3H, s), 0.0 (3H, s). 13C NMR
7.4.5. (anti,syn) 2-(tert-Butyl-dimethyl-silanyloxy)-5-butyl-
cyclohexanecarboxylic acid ethyl ester 12b. Typical
hydrolysis procedure. A solution of ca. 5 mmol of crude
silyl ketene acetal 11b was stirred from 08C to rt in 15 mL
of `wet' THF with glacial acetic acid (420 mL, 7.5 mmol,
1.5 equiv.) over a period of 3 h. The mixture was then
treated with 10 mL of water and with saturated aqueous
NaHCO3 until pH 7. Extraction with ethyl ether, washing
with brine and drying over MgSO4 afforded the crude
product, which was distilled under reduced pressure afford-
ing pure ester. Colourless liquid (1.57 g, 92%), d.e.$95%,
(50 MHz, CDCl3)
d
146.6 (CvC(OSiMe3)), 106.4
(CvC(OSiMe3)), 67.5 (CH±OSi), 65.4, 35.2, 34.7 (CH±
n-Bu), 33.1, 31.9, 30.7, 25.6, 22.8, 20.10, 17.8, 14.9, 13.8
(CH3), 0.0, 24.9, 25.2. MS m/z (rel. int., %) 29 (M, 2), 165
(31), 147 (34), 115 (22), 107 (22), 73 (100), (CI) 415
(M2H). IR nmax (thin ®lm) 1675 (CvC), 1084 (broad,
O±Si).
7.4.2. (Z) (anti) 1-Ethyl-3-(tert-butyl-dimethyl-silanyl-
oxy)-2-[ethoxy-(trimethyl-silanyloxy)-methylene]-cyclo-
hexane 11d. The same procedure described for the
preparation of 11b was used, affording the title compound
without puri®cation in order to prevent from subsequent
degradation (b-elimination). Colourless liquid (2.0 g,
1
bp 96±978C/0.04 mmHg. H NMR (400 MHz, CDCl3) d
4.24 (1H, m, CH±OSi), 4.1 (2H, q, Jˆ7.0 Hz), 2.06±1.25
(14H, m), 1.23 (3H, t, Jˆ7.0 Hz), 0.8 (9H, s), 0.8 (3H, broad
t), 0.00 (3H, s), 20.06 (3H, s). 13C NMR (100 MHz, CDCl3)
d 173.9, 69.7 (CH±OSi), 60.4, 53.4 (CH±CO2), 35.4, 34.3,
32.5 (CH±n-Bu), 31.3, 29.4, 26.3, 23.6, 19.8, 18.6, 14.9,
14.7. MS m/z (CI) 343 (M2H1). IR nmax (thin ®lm) 1744
(CvO), 1042 (O±Si). Calculated for C19H38O3Si: C, 66.61;
H, 11.18. Found: C, 66.55; H, 11.24.
1
quant. yield), d.e.$95%. H NMR (200 MHz, CDCl3) d
4.6 (1H, m, CH±OSi), 3.75 (2H, ABX3 System), 2.05±1.2
(1H, m, CH±Et), 1.7±1.5 (4H, m), 1.4±1.1 (4H, m), 1.05
(3H, t, Jˆ7.1 Hz), 0.7 (9H, s), 0.7 (3H, broad t), 0.0 (9H, s),
20.15 (3H, s), 20.2 (3H, s). 13C NMR (50 MHz, CDCl3) d
147.6 (CvC(OSiMe3)), 107.0 (CvC(OSiMe3)), 68.4
(CH±OSi), 66.4, 37.7 (CH±Et), 36.1, 32.6, 27.0, 26.6,
21.0, 18.8, 15.6, 13.9, 0.9, 23.9, 24.2.
7.4.6. (anti,syn) 2-Hydroxy-5-n-butyl-cyclohexanecar-
boxylic acid ethyl ester 15b and (syn,anti) 2-Hydroxy-
5-butyl-cyclohexanecarboxylic acid ethyl ester 16b.
Typical desilylation procedure. The silyl ether 12b
(1.57 g, 4.6 mmol, 1 equiv.) was stirred in 15 mL of THF
with n-Bu4NF, 3H2O (3.15 g, 10 mmol, 2.2 equiv.) at rt over
a period of 4 h before hydrolysis with 20 mL of water and
extraction with diethyl ether. Then, the combined organic
layers were washed with brine and saturated aqueous
CuSO4, and dried over MgSO4. Removal of the solvents
in vacuo afforded a mixture of two diastereoisomers in a
7.4.3. (Z) (anti) 1-Isopropyl-3-(tert-butyl-dimethyl-silanyl-
oxy)-2-[ethoxy-(trimethyl-silanyloxy)-methylene]-cyclo-
hexane 11e. The same procedure described for the
preparation of 11b was used, affording the title compound
without puri®cation in order to prevent from subsequent
degradation (b-elimination). Colourless liquid (1.98 g,
1
99%), d.e.$95%. H NMR (200 MHz, CDCl3) d 4.6 (1H,
1
ca. 95:5 ratio according to H NMR. Those two isomers
were separated over silica gel (hexane/etherˆ9:1) giving
m, CH±OSi), 3.6 (2H, ABX3 System), 2.0±1.1 (8H, m), 1.0
(3H, t, Jˆ7.1 Hz), 0.8 (3H, s), 0.7 (9H, s), 0.6 (3H, s), 0.0
(9H, s), 20.15 (3H, s), 20.2 (3H, s). 13C NMR (50 MHz,
CDCl3) d 146.6 (CvC(OSiMe3)), 105.5 (CvC(OSiMe3)),
67.7 (CH±OSi), 65.4, 45.4, 35.2, 28.2, 27.9, 25.6, 22.4,
20.8, 20.1, 17.8, 14.6, 1.7, 0.0, 23.8, 24.9. MS m/z (rel.
the alcohols in 70% overall yield.
Major diastereoisomer (®rst eluted): 15b. Colourless liquid
(0.75 g), d.e.$95%. Rfˆ0.38 (hexane/etherˆ1:1). 1H NMR

2168
V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
(400 MHz, CDCl3/TMS) d 4.18 (2H, m), 4.07 (1H, m, CH±
OH), 3.32 (1H, broad s, OH), 2.20 (1H, dd, Jtransˆ11.2, Jcisˆ
1.9 Hz, CH±CO2), 2.0 (1H, m, CH±n-Bu), 1.9±1.7 (4H, m),
1.46 (1H, m), 1.35±1.20 (6H, m), 1.25 (3H, X3 part of a
ABX3 system, JA±Xˆ7.1 Hz), 0.9 (1H, m, H5), 0.85 (3H,
broad t, Jˆ6.7 Hz). 13C NMR (100 MHz, CDCl3) d 177.0,
67.4 (CH±OH), 61.1, 53.7 (CH±CO2), 34.7, 34.0 (CH±n-
Bu), 32.2, 31.2, 29.0, 23.5, 19.9, 14.8, 14.6. MS m/z (rel.
int., %) 228 (M, 3), 210 (6), 200 (28), 157 (100), 143 (69),
137 (16), 129 (50), 125 (51), 115 (18), 111 (21), 81 (43), 69
(30), 57 (18), 55 (44), 41 (36), (CI) 229 (M2H1). IR nmax
(thin ®lm) 3522 (broad, OH), 1708 (CvO). Calculated for
C13H24O3: C, 68.38; H, 10.59. Found: C, 68.14; H, 10.77.
7.4.8. (anti, syn) 2-Hydroxy-5-isopropyl-cyclohexanecar-
boxylic acid ethyl ester 15e and (syn, anti) 2-Hydroxy-5-
isopropyl-cyclohexanecarboxylic acid ethyl ester 16e.
1
The two diastereoisomers (ca. 92:8 ratio according to H
NMR) were obtained by the same procedure described
before and were separated over silica gel with hexane/
etherˆ97:3 (major diastereoisomer) then 95:5 (minor
diastereoisomer) affording the alcohols in 64% overall yield.
Major diastereoisomer (®rst eluted): 15e. Colourless liquid
(0.630 g), d.e.$95%. Rfˆ0.42 (hexane/etherˆ1:1). 1H
NMR (400 MHz, CDCl3/TMS) d 4.18 (2H, m), 4.12 (1H,
m, CH±OH), 3.50 (1H, broad s, OH), 2.42 (1H, dd, Jtrans
ˆ
11.6, Jcisˆ2.03 Hz, CH±CO2), 2.02 (1H, m, CH±i-Pr), 1.9
(1H, m), 1.83±1.63 (2H, m), 1.55 (2H, m), 1.38±1.30 (1H,
m), 1.29 (3H, t, Jˆ7.1 Hz), 1.03 (1H, m), 0.94 (3H, d,
Jˆ6.9 Hz), 0.83 (3H, d, Jˆ6.9 Hz). 13C NMR (100 MHz,
CDCl3) d 177.3 (CO2), 67.6 (CH±OH), 61.2, 51.4 (CH±
CO2), 39.5 (CH±i-Pr), 32.3, 29.9, 24.5, 21.7, 20.0, 16.5,
14.8. MS m/z (rel. int., %) 214 (M, 1), 171 (28), 143 (62),
130 (16), 125 (100), 123 (23), 114 (21), 108 (11), 97 (55), 86
(19), 83 (16), 82 (51), 79 (56), 69 (62), 67 (39), 55 (52), 43
(52), 41 (61), 29 (38). IR nmax (thin ®lm) 3479 (broad, OH),
1711 (CvO). Calculated for C12H22O3: C, 67.26; H, 10.35.
Found: C, 67.18; H, 10.20.
Minor diastereoisomer: 16b. Colourless liquid (0.04 g),
d.e.$95%. Rfˆ0.35 (hexane/etherˆ1:1). 1H NMR (400
MHz, CDCl3/TMS) d 4.17 (2H, ABX3 system), 3.96 (1H,
dt, Jtransˆ9.8, 4.5 Hz, CH±OH), 3.03 (1H, broad s, OH),
2.40 (1H, dd, Jcisˆ9.8, Jtransˆ4.6 Hz, CH±CO2), 2.19 (1H,
m, CH±n-Bu), 2.00 (1H, m), 1.7 (1H, m), 1.6±1.2 (6H, m),
1.30 (3H, X3 part of a ABX3 system, JA±Xˆ7.2 Hz), 1.1 (2H,
m), 0.8 (3H, broad t, Jˆ7.1 Hz). 13C NMR (100 MHz,
CDCl3) d 175.5, 67.1 (CH±OH), 61.0, 55.6 (CH±CO2),
36.8 (CH±n-Bu), 33.6, 30.7, 28.3, 23.2, 19.8, 14.9, 14.6.
IR and MS spectra were identical to those of the major
diastereoisomer.
Minor diastereoisomer:16e. Colourless liquid (0.055 g),
d.e.$95%. Rfˆ0.39 (hexane/etherˆ1:1). 1H NMR (400
MHz, CDCl3/TMS) d 4.12±4.02 (3H, m), 2.63 (1H, dd,
Jcisˆ5.6, Jtransˆ3.6, CH±CO2), 2.39 (1H, broad s, OH),
1.88 (1H, m, CH±i-Pr), 1.7±1.4 (7H, m), 1.20 (3H, t, Jˆ
7.1 Hz), 0.83 (6H, pseudo t). 13C NMR (100 MHz, CDCl3) d
174.9, 67.9 (CH±OH), 60.8, 51.9 (CH±CO2), 42.3 (CH±i-
Pr), 31.4, 29.6, 27.1, 22.2, 21.8, 20.4, 14.8. IR and MS
spectra were identical to those of the major diastereoisomer.
7.4.7. (anti, syn) 2-Hydroxy-5-ethyl-cyclohexanecarboxylic
acid ethyl ester 15d and (syn, anti) 2-Hydroxy-5-ethyl-
cyclohexanecarboxylic acid ethyl ester 16d. The two
1
diastereoisomers (ca. 88:12 ratio according to H NMR)
were obtained by the same procedure described before
and were separated over silica gel (hexane/etherˆ95:5),
affording the alcohols in 84% overall yield.
Major diastereoisomer (®rst eluted): 15d. Colourless liquid
crystalized after 2 weeks at 2188C (0.74 g), d.e.$95%.
Rfˆ0.36 (hexane/etherˆ1:1). H NMR (400 MHz, CDCl3/
7.4.9. (anti, syn) 2-Hydroxy-5-(5-hydroxy-pentyl)-cyclo-
hexanecarboxylic acid ethyl ester 15f. Compound 11f
was prepared from 4b following the typical procedure
described above (see also 9e for the preparation of the
organometallic reagent). The compound 15f was prepared
from the crude silylketene acetal 11f by hydrolysis with
acetic acid in THF and fully deprotected with acetone/HCl
(see 14f). Flash chromatography using hexane/diethyl
etherˆ1:1 afforded pure diol as a single diastereoisomer.
Colourless liquid (0.86 g, 67%), d.e.$95%. 1H NMR
(400 MHz, C6D6/TMS) d 4.07 (1H, m, CH±OH), 3.27±
2.96 (5H, m), 2.37±2.17 (1H, m), 1.92 (1H, dd, Jˆ10.9,
2.1 Hz, CH±CO2), 1.78 (1H, m), 1.64±1.0 (13H, m), 0.9
(4H, m). 13C NMR (100 MHz, C6D6) d 177.0 (CO2), 75.8,
64.2 (CH2OH), 62.7, 57.1, 40.0, 39.5, 36.2, 33.3, 31.7, 29.9,
22.9, 21.7, 18.5. MS m/z (CI) 276 (M2NH41), 259 (M2H).
IR nmax (thin ®lm) 3550 (broad, OH), 1718 (CvO). Calcu-
lated for C14H26O4: C, 65.09; H, 10.14. Found: C, 64.73; H,
10.42.
1
TMS) d 4.10 (2H, m), 4.01 (1H, m, CH±OH), 3.30 (1H,
broad s, OH), 2.14 (1H, dd, Jtransˆ11.2, Jcisˆ2.0 Hz, CH±
CO2), 1.90±1.60 (4H, m), 1.43 (1H, m), 1.35±1.23 (2H, m),
1.20 (3H, t, Jˆ7.1 Hz), 1.05 (1H, m), 0.85 (1H, m), 0.80
(3H, broad t, Jˆ7.5 Hz). 13C NMR (100 MHz, CDCl3) d
177.0, 67.3 (CH±OH), 61.2, 53.3 (CH±CO2), 35.4 (CH±
Et), 32.3, 30.5, 27.6, 19.8, 14.8, 11.2. MS m/z (rel. int.,
%) 200 (M, 2), 172 (27), 155 (12), 143 (60), 137 (10),
129 (100), 125 (41), 115 (17), 109 (38), 101 (64), 69 (30),
97 (33), 83 (29), 79 (31), 69 (27), 55 (45), 41 (35), 29 (38).
IR nmax (thin ®lm) 3511 (broad, OH), 1707 (CvO). Calcu-
lated for C11H20O3: C, 65.97; H, 10.07. Found: C, 65.46; H,
10.23.
Minor diastereoisomer: 16d Colourless liquid (0.1 g),
d.e.$95%. Rfˆ0.35 (hexane/etherˆ1:1). 1H NMR (400
MHz, CDCl3/TMS) d 4.12 (2H, ABX3 system, 3.9 (1H, dt,
Jcisˆ9.8, 4.5 Hz, CH±OH), 3.0 (1H, broad s, OH), 2.35 (1H,
dd, Jcisˆ9.8, Jtransˆ4.5 Hz, CH±CO2), 2.05 (1H, m, CH±
Et), 1.95 (1H, m), 1.70 (1H, m), 1.55±1.0 (5H, m), 1.2 (3H,
t, Jˆ7.1 Hz), 0.80 (1H, m), 0.78 (3H, broad t, Jˆ7.5 Hz).
13C NMR (100 MHz, CDCl3) d 175.5, 67.4 (CH±OH), 61.0,
55.5 (CH±CO2), 38.8 (CH±Et), 33.5, 27.6, 21.6, 19.6, 14.8,
13.1. IR and MS spectra were identical to those of the major
diastereoisomer.
7.4.10. (anti,syn) 2-Hydroxy-5-(4-oxo-pentyl)-cyclohexane-
carboxylic acid ethyl ester 15g. Compound 11g was ®rst
prepared from 4b following the typical procedure described
above (see also 9f for the preparation of the organometallic
reagent). Then the crude adduct (silylketene acetal) 11g was
stirred with acetic acid in THF and removal of the protective
groups was performed with acetone/HCl/KIO4 (see 14g)

V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
2169
Flash-chromatography (hexane/diethyl etherˆ3:7) ®nally
afforded pure ketoalcohol as a single diastereoisomer.
Colourless liquid (0.83 g, 65%), d.e.$95%. 1H NMR
(400 MHz, CD3COCD3/TMS) d 4.42 (1H, m, CH±OH),
3.85 (2H, m), 3.07 (1H, s, OH), 2.47±2.37 (4H, m), 2.08
(3H, s), 1.93±1.16 (10H, m), 0.85 (3H, t, Jˆ7.1 Hz). 13C
NMR (100 MHz, CD3COCD3) d 208.7 (CO), 177.2 (CO2),
73.9, 61.0, 52.2, 44.0, 41.2, 35.9, 35.1, 35.0, 30.5, 29.9,
23.3, 15.0. MS m/z (CI) 274 (M2NH41). IR nmax (thin
®lm) 3479 (broad, OH), 1711±1694 (s, CvO). Calculated
for C14H24O4: C, 65.60; H, 9.44. Found: C, 65.47; H, 9.67.
diethyl etherˆ4/1) 17b. Colourless liquid (1.10 g). 1H NMR
(200 MHz, CDCl3/TMS) d 6.23 (1H, qd, 3Jˆ7.0, 4Jˆ
1.0 Hz), 4.45 (1H, m), 4.23 (2H, ABX3 system), 2.83 (1H,
3
broad s, OH), 1.97 (3H, d, Jˆ7.0 Hz), 1.33 (6H, m). 13C
NMR (50 MHz, CDCl3) d 167.3, 136.3, 134.5, 68.1, 59.9,
22.3, 14.8, 13.8. IR and MS spectra were identical to those
of the minor diastereoisomer. Calculated for C8H14O3: C,
60.74; H, 8.92. Found: C, 60.44; H, 9.12.
7.5.2. (E)-2-(1-Methyl-pentyl)-but-2-enoic acid ethyl
ester 18. Alkene 18 was obtained when copper(I) mediated
conjugate addition of n-butyl magnesium bromide was
performed with 17b. It was puri®ed by a kugelrohr distil-
lation. Colourless liquid. (0.35 g, 66%), bp 50±658C/
7.4.11. (anti, syn) 2-(tert-Butyl-dimethyl-silanyloxy)-5-(3-
oxo-propyl)-cyclohexanecarboxylic acid ethyl ester 12h.
The silylketene acetal 11h was prepared from 4b following
the typical procedure described above (see also 9f for the
preparation of the organometallic reagent). The crude
silylketene acetal 11h was directly treated according to
the procedure previously described (e.g. HCl/Dioxan/
KIO4, see 14g), to give aldehyde 12h as a single diastereo-
isomer with a high level of purity without further puri®-
cation in order to prevent from degradation. Colourless
liquid (1.02 g, 96%), d.e.$95%. 1H NMR (200 MHz,
CDCl3/TMS) d 9.68 (1H, broad t, Jˆ1.6 Hz), 4.24 (1H,
m, CH±OH), 4.11±3.96 (2H, ABX3 system), 2.42±2.34
(2H, m), 2.10±1.16 (10H, m), 1.19 (3H, broad t, Jˆ
7.0 Hz), 0.80 (9H, s), 20.04 (3H, s), 20.1 (3H, s). 13C
NMR (50 MHz, CDCl3) d 202.7, 172.6, 68.8, 59.9, 54.3,
41.4, 33.3, 31.5, 30.3, 27.1, 25.6, 18.8, 17.8, 14.0, 24.6,
25.5. MS m/z (CI) 360 (M2NH41), 343 (M2H). IR nmax
(thin ®lm) 3510 (broad, OH), 1740±1710 (broad s, CvO).
1
0.1 mmHg. H NMR (200 MHz, CDCl3/TMS) d 7.16 (1H,
q, Jˆ7.1 Hz), 4.15 (2H, broad q, Jˆ7.2 Hz), 2.68 (1H, m),
1.76 (3H, d, Jˆ7.1 Hz), 1.51±1.21 (6H, m), 1.16 (3H, d, Jˆ
10.4 Hz), 0.92±0.81 (6H, m). 13C NMR (50 MHz, CDCl3) d
167.7, 137.7, 136.0, 59.9, 34.6, 32.3, 30.3, 22.7, 19.2, 14.2,
13.9, 13.8. MS m/z (rel. int., %) 198 (M, 18), 183 (13), 153
(29), 141 (29), 127 (14), 116 (100), 113 (32), 101 (47), 95
(57), 88 (22), 83 (19), 69 (44), 67 (27), 55 (33), 43 (33), 41
(39). IR nmax (thin ®lm) 1730±1713 (CvO1CvC). Calcu-
lated for C12H22O2: C, 72.68; H, 11.18. Found: C, 73.03; H,
10.88.
7.5.3. (Z)-2-[1-(tert-Butyl-dimethyl-silanyloxy)-ethyl]-but-
2-enoic acid ethyl ester 19. The title alcohol was silylated
in the same way described for the preparation of 8d. Colour-
less liquid (0.355 g, 90%), bp 58±598C/0.1 mmHg. 1H
3
NMR (200 MHz, CDCl3/TMS) d 6.30 (1H, qd, Jˆ7.3,
4Jˆ1 Hz), 4.61 (1H, m), 4.28±4.14 (2H, ABX3 system),
3
4
1.96 (3H, dd, Jˆ7.3, Jˆ1.1 Hz), 1.31 (3H, broad t, Jˆ
7.0 Hz), 1.24 (3H, d, Jˆ6.1 Hz), 0.87 (9H, s), 0.1 (3H, s),
0.02 (3H, s). 13C NMR (50 MHz, CDCl3) d 167.4, 137.5,
133.4, 68.0, 59.9, 25.7, 24.6, 18.0, 15.0, 14.1. MS m/z (rel.
int., %) 257 (M215, 3), 215 (92), 187 (41), 169 (100), 119
(14), 103 (24), 95 (28), 75 (96), 67 (29), 57 (196), 41 (21).
IR nmax (thin ®lm) 1718 (s, CvO), 1654 (CvC), 1082 (O±
Si). Calculated for C14H28O3Si: C, 61.72; H, 10.36. Found:
C, 61.55; H, 10.49.
7.5. Syntheses of acyclic analogues
7.5.1. 2-(1-Hydroxy-ethyl)-but-2-enoic acid ethyl ester,
mixture of E and Z isomers 17a and 17b.17 To a suspen-
sion of cuprous iodide (192 mg, 1 mmol, 0.1 equiv.) in
10 mL of dry THF, a solution of methyllithium±LiBr in
diethyl ether (1.45N) (0.69 mL, 1 mmol, 0.1 equiv.) was
slowly added and stirred at 2308C for 20 min. Then, a
solution of HMPA (3.4 mL, 19.6 mmol, 1.9 equiv.) in
20 mL of dry THF was added prior to slow addition of a
solution of diisobutylaluminumhydride (DIBAL-H) in tolu-
ene (1N) (15 mL, 15 mmol, 1.5 equiv.). The mixture was
stirred for 2 h at 2308C, ethyl but-2-ynoate (1.12 g,
10 mmol, 1 equiv.) added and allowed to warm up to
2208C. After 5 h at this temperature, acetaldehyde
(2.23 mL; 40 mmol, 4 equiv.) in 20 mL of THF was ®nally
added and the temperature gently warmed to rt overnight.
Hydrolysis with 10 mL of 1N HCl, addition of diethyl ether
and washing successively with 3£10 mL of 1N HCl, 10 mL
of aqueous saturated NaHCO3, 2£10 mL of brine afforded
an organic layer which was dried over MgSO4 before
concentration in vacuo. Then, chromatography over silica
gel succeeded in separating the two isomers as pure
products in 72% overall yield. (E)-isomer (®rst eluted
with hexane/diethyl etherˆ9:1) 17a. Colourless liquid (30
7.5.4. 2-(1-Hydroxy-ethyl)-3-methyl-heptanoic acid ethyl
ester, mixture of isomers 20. The conjugate addition reac-
tion of O-silylated Z-butenoate 19 (0.27 g, 1 mmol,
1 equiv.) was conducted according to typical procedure
(magnesium di-n-butylcuprate addition). Then, the crude
product was desilylated as previously described affording
a mixture of three diastereoisomers in a ca. 5:3:2 ratio
which were unseparable over silica gel. Colourless liquid
(0.180 g, 83%).
Major diastereoisomer: 1H NMR (200 MHz, CDCl3/TMS) d
4.30±4.05 (3H, m), 1.85 (1H, dd, 3Jˆ7.3, 17 Hz), 1.35±1.26
(8H, m), 0.92±0.87 (12H, m). 13C NMR (50 MHz, CDCl3) d
175.3, 65.8, 59.9, 55.7, 37.3, 35.7, 32.1, 25.6, 19.8, 15.2,
14.3, 14.0. MS m/z (CI) 234 (M2NH41), 217 (M2H). IR
nmax (thin ®lm) 3600 (broad, OH), 1743 (CvO).
1
mg). H NMR (200 MHz, CDCl3/TMS) d 7.12 (1H, q, Jˆ
7.0 Hz), 4.16 (3H, m), 1.82 (3H, d, Jˆ7.0 Hz), 1.25 (6H, m).
13C NMR (50 MHz, CDCl3) d 166.8, 136.1, 133.9, 69.3,
60.2, 21.9, 14.7, 13.8. MS m/z (CI) 176 (M2NH41), 159
(M2H). IR nmax (thin ®lm) 3417 (broad, OH), 1718 (s,
CvO), 1636 (CvC). (Z)-isomer (second eluted with hexane/
Acknowledgements
We would like to thank Drs I. Beaudet and A. Guigant for

2170
V. Dambrin et al. / Tetrahedron 57 (2001) 2155±2170
reported by Sommer, L. H. In Stereochemistry, Mechanism
and Silicon; McGraw-Hill, New York, 1965.
fruitful discussions and help, Drs J.-Y. Le Questel and B.
Illien for AM-1 calculations and G. Nourisson for the
Â
9. Villieras, J.; Rambaud, M.; Graff, M. Synth. Commun. 1985,
Á
recording of the mass spectra. We are grateful to Ministere
de l'Enseignement Superieur et de la Recherche for
®nancial support.
15, 569±580.
10. Normant, H.; Feugeas, C. Bull. Soc. Chim. Fr. 1963, 1141±
1143.
Â
11. Related cyanocuprate was prepared according to previously
reported procedure: (a) Knochel, P.; Singer, R. D. Chem. Rev.
1993, 93, 2117±2188. (b) Knochel, P.; Yeh, M. C. P.; Berk,
S. C.; Talbert, J. J. Org. Chem. 1988, 53, 2390±2392.
12. When desilylation step14 was performed from OTBDPS
derivatives, a signi®cant epimerization of C1 occured while
desilylation was far from complete.
References
1. For papers of related interest see: (a) Normant, J. F. Synthesis
1972, 63±80. (b) Posner, G. H. Org. React. 1972, 19, 1±113.
(c) Lipshutz, B. H. Synthesis 1987, 325±341. (d) Kozlowski,
J. A. In Comprehensive Organic Synthesis, Trost, B. M.,
Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 4 (Chapter
1.4).
13. The synthesis of ethyl cyclohexylidene trimethylsilylketene
acetal was ®rst reported by Ainsworth, C.; Chen, F.; Yu-Neng,
K. J. Organomet. Chem. 1972, 46, 59±71.
14. Corey, E. J.; Snider, B. B. J. Am. Chem. Soc. 1972, 94, 2549±
2550.
2. See inter alia: (a) Rossiter, B. E.; Swingle, N. M. Chem. Rev.
1992, 92, 771±806. (b) Alexakis, A. Organocopper
ReagentsÐA Practical Approach; Oxford University Press:
New York, 1994 (Chapter 8). (c) Krause, N.; Gerold, A.
Angew. Chem., Int. Ed. Engl. 1997, 36, 186±204.
15. The diols 14f and 15f were obtained after treatment with 2N
aqueous HCl in acetone for 2 h at rt.
16. Removal of the dioxolane moiety was performed with potas-
sium periodate at re¯ux of a mixture of dioxane and 0.1N
aqueous HCl.
È
3. For a recent review concerning the synthesis of iridoõds see:
Nagia, A.; Prasuna, G.; Bheema Rao, P. Tetrahedron 1997, 53,
17. (a) Tsuda, T.; Hayashi, T.; Satomi, H.; Kawamoto, T.;
Saegusa, T. J. Org. Chem. 1986, 51, 537±540. (b) Tsuda,
T.; Yoshida, T.; Kawamoto, T.; Saegusa, T. J. Org. Chem.
1987, 52, 1624±1627. (c) Tsuda, T.; Yoshida, T.; Saegusa,
T. J. Org. Chem. 1988, 53, 1037±1040.
14507±14545.
4. (a) Amri, H.; Villieras, J. Tetrahedron Lett. 1987, 28, 5521±
Â
5524. (b) Amri, H.; Rambaud, M.; Villieras, J. Tetrahedron
Â
1990, 46, 3535±3546.
5. (a) Amri, H.; Rambaud, M.; Villieras, J. J. Organomet. Chem.
Â
18. Cocu, F. G.; Wolczunovicz, G.; Bors, L.; Posternak, T. Helv.
Chim. Acta 1970, 53, 739±742.
Â
1986, 308, C27±C32. (b) Dambrin, V.; Villieras, M.; Moreau,
Â
C.; Amri, H.; Toupet, L.; Villieras, J. Tetrahedron Lett. 1996,
37, 6323±6326.
19. (a) Johnson, F.; Malhotra, S. K. J. Am. Chem. Soc. 1965, 87,
5492±5495. (b) Johnson, F. Chem. Rev. 1968, 68, 375±413.
(c) Chow, Y. L.; Colon, C. J.; Tam, J. N. S. Can. J. Chem.
1968, 46, 2821±2825.
6. Concerning the use of TMS Halide in conjugate addition
reactions see for instance (a) Corey, E. J.; Boaz, N. W.
Tetrahedron Lett. 1985, 26, 6019±6022. (b) Taylor, R. J. K.
Synthesis 1985, 364±392. (c) Alexakis, A.; Berlan, J.; Besace,
Y. Tetrahedron Lett. 1986, 27, 1047±1050.
20. (a) Klibanov, A. M. Acc. Chem. Res. 1990, 23, 114±117.
(b) Yamane, T.; Takahashi, M.; Ogasawara, K. Synthesis
1995, 444±448.
7. 7a and b were obtained starting from the corresponding
trichloroacetimidates, following the procedure previously
described: Armstrong, A.; Backenridge, I.; Jackson, R. F. W.;
Kirk, J. M. Tetrahedron Lett. 1988, 29, 2483±2486.
8. Typical silylation involving imidazole in DMF was used, as
21. Chataignier, I.; Lebreton, J.; Durant, D.; Guingant, A.;
Â
Villieras, J. Tetrahedron Lett. 1998, 39, 1759±1962.
22. Graff, M.; Al Dilaimi, A.; Seguineau, P.; Rambaud, M.;
Â
Villieras, J. Tetrahedron Lett. 1986, 27, 1577±1578.