Silylated pyrrolidones via diastereoselective Pd-catalysed intramolecular allylic alkylations

Article abstract of DOI:10.1016/S0040-4039(01)01249-7

A new palladium-catalysed allylic alkylation affording silylated 3-vinyl-pyrrolidones has been developed. The method relies upon the interaction between a stabilised acetamide enolate anion and a silicon-containing nitrogen-tethered η³-allyl-pa

Full text of DOI:10.1016/S0040-4039(01)01249-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 6287–6289
Silylated pyrrolidones via diastereoselective Pd-catalysed
intramolecular allylic alkylations
Giovanni Poli,a,* Giuliano Giambastiani,a,† Max Malacriab,* and Serge Thorimbertb
aLaboratoire de Chimie des Organoe´le´ments, UMR 7611 CNRS, Universite´ Pierre et Marie Curie, Tour 44-45, 4, Place Jussieu,
Boˆıte 183, F-75252, Paris Cedex 05, France
bLaboratoire de Chimie Organique de Synthe`se, UMR 7611 CNRS, Universite´ Pierre et Marie Curie, Tour 44-54,
4, Place Jussieu, Boˆıte 229, F-75252, Paris Cedex 05, France
Received 5 July 2001; accepted 10 July 2001
Abstract—A new palladium-catalysed allylic alkylation affording silylated 3-vinyl-pyrrolidones has been developed. The method
relies upon the interaction between a stabilised acetamide enolate anion and a silicon-containing nitrogen-tethered h3-allyl-palla-
dium moiety. Two variants have been studied, involving location of the silicon atom on either olefinic carbon atom of the
cyclisation precursor. In both cases 5-exo-trig ring closure was the only cyclisation process observed. These results contrast with
related b-ketoester cyclisations, were competitive 7-endo-trig is observed. © 2001 Elsevier Science Ltd. All rights reserved.
We (G.P and G.G.) recently reported a new palladium-
catalysed intramolecular allylic alkylation process1
based on the interaction between a stabilized acetamide
enolate anion and a juxtaposed nitrogen-tethered h3-
allyl-palladium appendage. This totally diastereoselec-
tive reaction constantly favoured a 5-exo-trig mode of
were the first to report that Me3Si-substituted h3-allyl-
palladium complexes direct the addition of nucleophiles
exclusively to the distal position (relative to silicon
atom), thereby leading to the corresponding vinylsi-
lanes. Such a regioselectivity may be accounted for in
terms of (a) steric factors, (b) charge distribution of the
allyl complex and (c) stability of the newly formed
olefin–Pd(0) complex.3
cyclisation,
thereby
producing
3,4-disubstituted
pyrrolidones (Scheme 1).
On the other hand, it is known that a trialkylsilyl group
can sensibly modify the reactivity of h3-allyl-palladium
complexes. Thus, for example, Hirao and co-workers2
In this context, we (M.M. and S.T.) have developed a
palladium-catalysed cyclisation reaction, wherein 2-tri-
ethylsilyl-1,4-diacetoxy-but-2-ene is converted into a
silyl-substituted cyclopentene.4 In this global annula-
tion process the role of silicon is crucial and twofold. In
fact, in the former CꢀC bond formation the bulky
trialkylsilyl donor group regiodirects the ionisation of
the initial bis-allylic system, whereas in the latter one it
favours the syn configuration of the h3-allylpalladium
complex, so as to allow an unusual 5-endo-trig process
to take place (Scheme 2).
Scheme 1. Reagents and conditions: Pd2(dba)3, (0.05 equiv.),
PPh3, (0.5 equiv.), BSA (1.2 equiv.), AcOK (0.1 equiv.), THF
reflux, 12 h. EWG: CO2Me, COMe, CN, SO2Ph, PO(OEt)2.
Given the above background we decided in a joint
project to study the effect of silicon-substitution on
the amide-based cyclisation. In particular, we were
intrigued to verify if such modifications could still
guarantee cyclisations, and, in the positive case, to
determine the 5-exo-trig versus 7-endo-trig preference.
In fact, such a competition in the cyclisation mode has
been previously observed by Tsuji and co-workers in
pioneering studies on intramolecular palladium-
catalysed cyclisations (Scheme 3).5
Keywords: palladium and compounds; allylation; silicon and com-
pounds.
* Corresponding authors. Tel.: +33-1-44275572; fax: +33-1-44277567;
e-mail: poli@ccr.jussieu.fr
† Present address: Dipartimento di Chimica Industriale e Organica,
Universita` degli Studi di Milano, Via Venezian 21, I-20133, Milano,
Italy.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01249-7

6288
G. Poli et al. / Tetrahedron Letters 42 (2001) 6287–6289
aminoalcohol derived from amine addition to the less
substituted p-allyl terminus. At this stage the synthesis
needed N-malonylation of the major isomer 8b fol-
lowed by O-derivatisation. Surprisingly, N-malonyla-
tion of this aminoalcohol took place with competitive
O-malonylation. Ironically enough, treatment of the
same aminoalcohol with acetic anhydride gave exclu-
sively nitrogen acetylation! The problem was solved by
temporary protection of the hydroxyl function as TMS
derivative before the N-malonylation step. The desired
precursor 4 was eventually obtained in nearly quantita-
tive yield by treating allylic alcohol 9 with methyl
chloroformate (Scheme 4).
Scheme 2. Reagents and conditions: (i) NaCHE2 (E=
CO2Me), Pd(PPh3)4 (5%), THF (85%); (ii) NaH, 5%
Pd(PPh3)4, THF, 60°C, 74%.
After some preliminary and deceiving cyclisation exper-
iments we found that treatment of the sodium enolate
of 3 with 5% Pd(OAc)2 and 10% diphenylphosphi-
noethane (dppe), in DMF at 100°C for 30 min, gave
pyrrolidone 10 cleanly in a completely regio- and
stereoselective way (Scheme 5). More interestingly, sub-
mission of the silylated derivative 4 to the same reac-
tion conditions as before, gave pyrrolidone 11 as a
single isomer.
Scheme 3. Reagents and conditions: Pd(OAc)2 (5–10%), PPh3,
THF reflux.
The above experiments reveal the intrinsic and total
preference of these substrates for a 5-exo-trig process
over the alternative 7-endo-trig one, even when CꢀC
bond formation involves reaction at the h3-allyl termi-
nus carrying the bulky trialkylsilyl group. Not unex-
pectedly, on passing from the former to the latter
cyclisation an important decrease in yield is observed
(90 versus 52%). Determination of the relative configu-
ration of the cyclised products turned out to be
difficult. However, thermodynamic control via NaOAc
promoted equilibration has been constantly observed in
our laboratory in a number of analogous reactions.
Therefore, trans stereochemistry has been assigned to
pyrrolidone 10. Similarly, compound 11 is expected to
be the most stable diastereoisomer.6 Although for the
time being the lack of knowledge of several parameters
does not allow us to be more specific, some fundamen-
Accordingly, the synthesis of the two silylated cyclisa-
tion precursors 3 and 4 has been undertaken (Scheme
4). The first cyclisation precursor 3 was obtained in
three steps from the silylated allylic alcohol 5.4a Chlori-
nation of the alcohol function gave the corresponding
chloride 6, which was immediately treated with benzyl-
amine to give the secondary allylic amine 7. Malonyla-
tion under standard conditions gave uneventfully the
desired precursor 3.
The second cyclisation precursor 4 was obtained from
the same key vinylsilane 5 as used in the previous
synthetic sequence. Palladium-catalysed allylic amina-
tion of 5 in the presence of benzylamine afforded the
regioisomeric secondary amines 8a and 8b in a 24:76
ratio, the major isomer being a stereohomogeneous
Scheme 4. Reagents and conditions: (i) PPh3, CCl4; (ii) PhCH2NH2, MeCN (63% from 5); (iii) MeO2CCH2COCl, NEt3, CH2Cl2
(81%); (iv) PhCH2NH2, NEt3, Pd(OAc)2 (5%), dppe (10%), THF, rt (68%); (v) Me3SiCl, NEt3, THF, rt; (vi) MeO2CCH2COCl,
NEt3, CH2Cl2; (vii) HCl 10% until pH 2–3, THF, rt (82% from 8b); (viii) MeO2CCl, pyridine, CH2Cl2, rt (98%).

G. Poli et al. / Tetrahedron Letters 42 (2001) 6287–6289
6289
In summary, this new investigation showed that the
formal introduction of a trialkylsilyl group into either
position of the double bond in the precursor still per-
mits the cyclisation to take place. As in the previous
studies only 5-exo-trig cyclisations are observed,
thereby producing 3-vinyl-pyrrolidones silylated at
strategic positions. In both the cases studied the reac-
tions were completely diastereoselective. Further mod-
ifications of the obtained pyrrolidones via the rich
chemistry of allyl- and vinyl-silanes9 is planned in
future studies.
Acknowledgements
Scheme 5. Reagents and conditions: (i) Pd(OAc)2 (5%), dppe
(10%), NaH, DMF, 100°C (90% 3“10, 52% 4“11).
We gratefully acknowledge MURST and CNR (Italy)
as well as the University Pierre et Marie Curie and
CNRS (France) for funding of this research. This work
was supported by the European COST D-12 action
‘Organic Transformations: Selective Processes and
Asymmetric Catalysis’ (D12/0011/98) and a bilateral
CNR–CNRS cooperation.
tal stereochemical speculations can already be put for-
ward. In accord with what has been previously
observed, we assume that the relevant reactive confor-
mations associated with the latter cyclisation are syn-1
and/or syn-2 (Scheme 6), wherein the bulky silicon
atom occupies the syn position7 and the rest of the
chain is forced to occupy the anti position. Although
the two approaches lead to different diastereoisomers,
deprotonation after CꢀC bond formation is expected to
drive the equilibrium toward the most stable
diastereoisomer. In addition, the constant and exclusive
formation of pyrrolidone structures from these amides
suggests that, contrary to the related Tsuji’s b-
ketoesters cyclisations, the only plausible 7-endo-trig
type approach syn-3 has to be highly disfavoured with
respect to syn-1 and/or syn-2. Indeed, inspection of
models reveals that syn-3 approach needs an appropri-
ate C–C–X–C dihedral angle value. Although such a
requirement may be easily met with b-ketoesters (X=
CH2), in the case of amidoesters (X=NCH2Ph) the flat
nature of the amide bond is expected to force C–C–X–
References
1. (a) Giambastiani, G.; Pacini, B.; Porcelloni, M.; Poli, G. J.
Org. Chem. 1998, 63, 804–807; (b) Poli, G.; Giambastiani,
G.; Pacini, B. Tetrahedron Lett. 2001, 42, 5179–5182.
2. (a) Hirao, T.; Enda, J.; Ohshiro, Y.; Agewa, T. Tetra-
hedron Lett. 1981, 32, 3079–3080; (b) Inami, H.; Ito, T.;
Urabe, H.; Sato, F. Tetrahedron Lett. 1993, 34, 5919–5922.
3. Trost, B. M.; Self, C. R. J. Org. Chem. 1984, 49, 468–473.
4. (a) Thorimbert, S.; Malacria, M. Tetrahedron Lett. 1996,
37, 8483–8486; (b) Thorimbert, S.; Malacria, M. Tetra-
hedron Lett. 1998, 39, 9659–9660; (c) Thorimbert, S.;
Malacria, M.; Humilie`re, D. Synlett 1998, 1255–1257.
5. Tsuji, J.; Kobayashi, Y.; Kataoka, H.; Takahashi, T.
Tetrahedron Lett. 1980, 21, 1475–1478.
C dihedral angle values close to 0 or 360°, thereby
.
6. Despite a careful spectroscopic analysis (1H NMR, 13C
NMR, APT, APT-JMOD, COSY (H–H), COSY (H–C)
NOE diff.) of compound 11 a clear-cut assignment of the
relative configuration could not be obtained. However, we
believe that intramolecular CꢁO···SiEt3 coordination might
be responsible for the exclusive formation of the isomer
featuring a cis-(CO2Me/R3Si) relationship.
disfavouring the syn-3 approach and its related 7-endo-
trig cyclisation.8
7. (a) Poli, G.; Scolastico, C. Chemtracts Org. Chem. 1999,
12, 822–836; (b) Canovese, L.; Visentin, F.; Uguagliati, P.;
Lucchini, V.; Bandoli, G. Inorg. Chim. Acta 1998, 277,
247–252; (c) Crociani, B.; Antonaroli, S.; Paci, M.; Dibi-
anca, F.; Canovese, L. Organometallics 1997, 16, 384–391;
,
(d) Sjo¨rgen, M. P. T.; Hansson, S.; Akermark, B.;
Vitagliano, A. Organometallics 1994, 13, 1963–1971.
8. The formal C–X–C angle enlargement on passing from a
Csp3 to an N-amidic atom-type may also have a crucial
role in the different behaviour of the two cyclisation
precursors. Molecular modelling calculations are under
way in order to quantify and validate the above concepts.
9. (a) Colvin, E. W. Silicon in Organic Synthesis; Butter-
worth, 1981; (b) Colvin, E. W. Silicon Reagents in Organic
Synthesis; Academic Press, 1988.
Scheme 6. Relevant reactive conformations in the palladium-
catalysed intramolecular allylic alkylation of silylated ami-
doester 4, and comparison with b-ketoester cyclisation.