Palladium(II)-catalyzed isomerization of (Z)-1,4-diacetoxy-2-butene: Solvent effects

Article abstract of DOI:10.1002/ejoc.200700246

The isomerization of (Z)-1,4-diacetoxy-2-butene (1) catalyzed by PdCl ₂(MeCN)₂ was studied in THF and DMF. The reaction occurs more rapidly in THF than in DMF, but in both solvents it did not proceed to complete consumption of the substrate and led to a mixture of 1, (E)-1,4-diacetoxy-2-butene (2), and 1,2-diacetoxy-3-butene (3). The formation of 2 is more favored in DMF than in THF. The reactivity of 1 and the solvent effect differ strongly from those previously obtained with Pd(PPh ₃)₄ as the catalyst. Interpretations are provided for the crucial role of the nature of both solvent and intermediates on the course of the isomerizations. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejoc.200700246

FULL PAPER
DOI: 10.1002/ejoc.200700246
Palladium(II)-Catalyzed Isomerization of (Z)-1,4-Diacetoxy-2-Butene: Solvent
Effects
[a]
[a]
[a]
Anna Maria Zawisza, Sandrine Bouquillon, and Jacques Muzart*
Keywords: Palladium / Catalysis / Allylic diacetate / Isomerization / Solvent effects
The isomerization of (Z)-1,4-diacetoxy-2-butene (1) cata-
lyzed by PdCl (MeCN) was studied in THF and DMF. The
the solvent effect differ strongly from those previously ob-
tained with Pd(PPh as the catalyst. Interpretations are pro-
2
2
3 4
)
reaction occurs more rapidly in THF than in DMF, but in both
solvents it did not proceed to complete consumption of the
substrate and led to a mixture of 1, (E)-1,4-diacetoxy-2-but-
ene (2), and 1,2-diacetoxy-3-butene (3). The formation of 2 is
more favored in DMF than in THF. The reactivity of 1 and
vided for the crucial role of the nature of both solvent and
intermediates on the course of the isomerizations.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
Introduction
The [3,3] sigmatropic rearrangement of allylic acetates
through an 1,3-acetoxonium ion type intermediate is usu-
II
[3,5–7]
The 1,3-transposition of allylic acetates in the presence of
catalytic amounts of palladium complexes is a well known
ally admitted with Pd catalysts,
especially when
PdCl (RCN) is used with solvents such as THF, PhMe,
2
2
[
1,2]
[
8–11]
reaction
that has been extensively used in synthetic or-
and CH Cl .
Another mechanism was proposed for
2
2
ganic chemistry. The mechanism of this reaction depends
on the oxidation state of palladium.
when PdCl in AcOH or Pd(OAc) alone was used; this
2
2
II
mechanism involves the addition of Pd and acetate across
the double bond followed by the regeneration of the catalyst
from β-elimination of the acetate, specifically, an acetoxy-
3
An η -allylpalladium intermediate is usually admitted
0
[3]
from Pd catalysis, but we have demonstrated the influ-
ence of the polarity of the solvent on the mechanism of the
[
9,12]
palladation–deacetoxypalladation sequence.
It was also
0
Pd -catalyzed isomerization of (Z)-1,4-diacetoxy-2-butene
envisaged that the isomerization of linalyl acetate could im-
ply the cleavage of the C–OAc bond to afford an intermedi-
ate having an allylic cation moiety and the 6,7-C=C bond
4]
(
1).[ In the presence of Pd(PPh ) , 1 was selectively iso-
3
4
merized to (E)-1,4-diacetoxy-2-butene (2) in THF while
both 2 and 1,2-diacetoxy-3-butene (3) were simultaneously
obtained in DMF (Scheme 1). A set of experiments led to
the conclusion that the involvements of an η -allylpalla-
dium in the former solvent and of a cationic η -allylpalla-
dium in the latter are the key intermediates that are derived
from 1.
–
[13]
coordinated to (AcOPdCl ) .
To the best of our knowl-
2
II
edge, the Pd -catalyzed isomerization of 1,2-diacetoxy-3-
ene and 1,4-diacetoxy-2-ene type compounds has only been
reported from 6,7-diacetoxy-8-tetradecene
1
3
[
11]
and 1-substi-
With these last
[
14]
tuted-1,4-diacetoxy-4-phenyl-2-butenes.
substrates, the acetate transfer occurs only from the ben-
zylic position and leads to the complete rearrangement of
the substrate into the corresponding 1-substituted-1,2-di-
[
14]
acetoxy-4-phenyl-3-butenes.
The unexpected results obtained from 1 at 70–72 °C un-
0
der Pd catalysis led us to examine the isomerization of 1,
II
2
, and 3 under Pd catalysis; this study is here reported.
Scheme 1.
Results
[
a] Unité Mixte de Recherche 6519 “Réactions Sélectives et Appli-
cations”, CNRS-Université de Reims Champagne-Ardenne,
UFR Sciences,
The reaction of 1 at 70–72 °C in THF in the presence of
.05 equiv. of PdCl (MeCN) (4) led to the formation of 2
0
2
2
Boîte no. 44, B. P. 1039, 51687 Reims cedex 2, France
Fax: +33-3-26-05-31-66
and 3. No products other than 1, 2, and 3 were detected
1
from monitoring the course of the reaction by GC and H
E-mail: jacques.muzart@univ-reims.fr
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
NMR spectroscopy. After heating for 6 h, the diacetate pro-
portions evolved very moderately (Figure 1, Table S1). The
Eur. J. Org. Chem. 2007, 3901–3904
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3901

A. M. Zawisza, S. Bouquillon, J. Muzart
FULL PAPER
addition of a new batch of 4 after 24 h heating led only to
To slow down the isomerization of 1 in THF, this reac-
a slight increase in the conversion of 1. Switching from tion was repeated at room temperature (Figure 3, Table S4).
THF as solvent to DMF slowed down the reaction, but
increased the conversion of 1 (Figure 1, Table S2). In both
solvents, it was impossible to reach the complete transfor-
mation of 1.
Figure 3. Evolution over time of the diacetate proportions obtained
from the PdCl (MeCN) -catalyzed reaction of (Z)-1,4-diacetoxy-2-
2 2
butene (1) at room temperature in THF.
Figure 1. Evolution over time of the diacetate proportions obtained
from the PdCl (MeCN) -catalyzed reaction of (Z)-1,4-diacetoxy-2-
Discussion
2
2
butene (1) at 70–72 °C in either THF or DMF.
From comparison of the results depicted in Figures 1 and
3
with those previously obtained in the presence of
Diacetates 2 and 3 were also subjected to PdCl2-
MeCN)2 catalysis under the same conditions (Figure 2,
[4]
Pd(PPh ) , it appears that the efficiency of the Pd-cata-
lyzed isomerization of 1 depends dramatically on the oxi-
3
4
(
Table S3). Isomerizations of 2 into 3 and 3 into 2 without
production of 1 were obtained; they occurred more rapidly
in THF than in DMF. A deposit of palladium was observed
in DMF especially with 3 as the substrate. This catalyst
degradation did not allow more than 20% conversion of 3
in DMF to be obtained.
dation state of the catalyst. Indeed, the complete conversion
of
1 was obtained with Pd(PPh3)4 but not with
PdCl (MeCN) . Moreover, the solvent effect on the reactiv-
2
2
ity and selectivity are quite different.
The (Z/E) isomerization of double bonds under Pd ca-
II
[15–19]
talysis is well-documented.
termediate formation of an η -allyl(hydrido)palladium(IV)
For the mechanism, the in-
3
[
18]
complex was envisaged but subsequent studies ruled out
[
19,20]
such a possibility.
Spencer et al. suggested either the
formation of a carbocation from the interaction of the C=C
bond with PdCl (MeCN) , followed by rotation around the
2
2
+
ClPdC–C bond and regeneration of the catalyst, or the
trans addition of PdCl to the C=C bond followed by rota-
2
[
19]
tion and syn β-elimination of PdCl2.
PdCl (RCN)2 forms olefin complexes,
It is known that
and theoretical
[
21]
2
studies have shown that (1) the coordination of CH =CH
2
2
II
to Pd results in the lowering of the C=C torsional bar-
[
17]
rier
and (2) ethylene coordinated to a Pd atom is posi-
[
22]
tively charged.
It is not possible to distinguish between
the two mechanistic possibilities proposed by Spencer. Nev-
ertheless, our reaction was carried out in the absence of
[
23]
added chloride,
and Sen and Lai have shown that the
II
II
Pd -catalyzed migration of C=C bonds occurs via the Pd –
C–C intermediate.
+
[24]
Consequently, the first proposal of
Spencer seems to be the most adequate to explain the 1 Ǟ
2 isomerization (Scheme 2).
Figure 2. PdCl
2 2
(MeCN) -catalyzed isomerization of (E)-1,4-diacet-
oxy-2-butene (2) and 1,2-diacetoxy-3-butene (3) at 70–72 °C.
3902
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 3901–3904

II
Pd -Catalyzed Isomerization of (Z)-1,4-Diacetoxy-2-Butene: Solvent Effects
FULL PAPER
Scheme 4.
Scheme 2.
tially at the (E) double bond with PdCl (MeCN) to afford
2
2
The other reaction pathway from 1 and 4 is the forma-
tion of 1,3-acetoxonium ion B (Scheme 3). According to the
(3E,5Z)-2-acetoxy-hepta-3,5-diene, whereas the reaction
with Pd(PPh ) involves the (Z) double bond and leads to
[
1,2]
3 4
literature,
tates with 100% chirality transfer,
particularly the allylic transposition of ace-
[
3]
(
3E,5E)-2-acetoxy-hepta-3,5-diene. The observations of
[10,11,25]
B is produced via
II
Golding led us to suspect that the interaction of Pd species
with the C=C bond of 2 and 3 is easier than with the (Z)
double bond of 1; this preference would lead to a reaction
occurring almost exclusively with 2 and/or 3 when the rela-
tive concentration of 1 decreases to a given level. From the
results depicted in Figure 1, this level depends on the sol-
vent. This could be due to the influence of the nature of
the intramolecular trans addition of the carbonyl of one
ester moiety to the Pd-coordinated double bond.
[
26]
the solvent on the conformation of the substrates.
The preference for reactions from 2 and/or 3 rather than
from 1 can be, in part, also explained by comparing the
steric interactions developed in acetoxonium ions B and C,
as well as their formation and their reactive pathways
Scheme 3.
(Scheme 5). All substituents of C are quasiequatorial while
In THF, the isomerization reactions catalyzed by 4 lead B possesses a quasiaxial substituent. Consequently, the for-
to the thermodynamic equilibrium between 2 and 3 (2/3 ≈ mation of C that is produced from 2 and 3 and leads to
1
.7, Figures 1 and 2, Tables S1 and S3) already observed both 2 and 3, is likely easier than that of B that is produced
0
[4]
under Pd catalysis. In contrast, the 2/3 ratio becomes ap- from 1 and 3 but evolves to only 3.
parently stable for a value of ca. 2.3 (Figure 1) or 2.6 (Fig-
ure 2) in DMF. This strong difference between the final 2/3
ratios obtained in THF and DMF could be due to the sol-
vent effect on the equilibria[
26,27]
or, more probably, to the
destruction in DMF of the catalyst with time. Indeed, the
addition of a new batch of 4 to the 2.3 mixture of 2 and 3
in DMF decreased the 2/3 ratio (Table S2). The Pd -cata-
Scheme 5.
II
lyzed 2 Ǟ 3 and 3 Ǟ 2 isomerizations are due to [3,3] sig-
The solvent also impacts the rate of the conversion of 1,
matropic rearrangements involving acetoxonium ion C and the selectivity of this process as exemplified in Figures 1
Scheme 4). Acetoxonium ion B is also produced from 3, (Table S1 and S2) and 4 (Table S5), respectively. For a con-
(
but given the absence of the formation of 1 from 3 (Fig- version of 16%, the 2/3 ratios were ca. 2 and 6 in THF and
ure 2, Table S3), its only reactive pathway is the back reac- DMF, respectively (Figure 4). As expected, the concurrent
tion leading to 3 (Scheme 4).
2 Ǟ 3 isomerization led to the decrease in these differences
π
The palladium oxidation state depending reactivity of 1, with the reaction time. These results show that in a polar
namely complete transformation with Pd(PPh3)4[ but not solvent such as DMF,
4]
[29]
in which separated ions should
with PdCl (MeCN) , is difficult to rationalize because both be produced, the PdCl (MeCN) -catalyzed reaction of 1
[
30]
2
2
2
2
2
catalysts involve the formation of an η -palladium complex involves mainly the (Z) Ǟ (E) isomerization of the C=C
as the first step. The only partial conversion of 1 under bond, specifically, a reaction via carbocation A as depicted
II
Pd catalysis is not justified by the mechanisms depicted in in Scheme 2 rather than via acetoxonium ion B (Scheme 3).
Schemes 2 and 3. The slower PdCl -induced rearrangement The equilibrium between carbocations A and AЈ is indepen-
2
–
of (Z)-allylic acetates relative to the corresponding (E) iso- dent of a reaction with Cl (Scheme 2) whereas the B Ǟ 3
[
2,9,10b,28]
could be invoked, but is not a sufficient pathway requires Cl– (Scheme 3). In DMF, the chloride
mers
reason. A possible explanation is issued from reports of anion is more solvated than in THF; consequently, it is less
[
3,7,28]
Golding et al.
According to these authors, the isomer- prone to react with the PdClL moiety, hence the increase in
ization of (E,Z)-4-acetoxy-hepta-2,5-diene occurs preferen- the (Z) Ǟ (E) isomerization compared to the 1,3-transposi-
Eur. J. Org. Chem. 2007, 3901–3904
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3903

A. M. Zawisza, S. Bouquillon, J. Muzart
FULL PAPER
tion. This solvent effect on the selectivity differs strongly Acknowledgments
0
[4]
from the one previously observed under Pd catalysis
We are grateful to CNRS for a temporary position to A. M. Z.
since (1) in THF, Pd(PPh ) leads to the exclusive or almost
3
4
exclusive isomerization of 1 into 2, while PdCl (MeCN)2
2
[
1] J. Tsuji, Palladium Reagents and Catalysts, Wiley, Chichester,
995, p. 399–402.
2] L. E. Overman, Angew. Chem. Int. Ed. Engl. 1984, 23, 579–
86.
The differences in the conversion rates with the nature of [3] E. Curzon, B. T. Golding, C. Pierpoint, B. W. Waters, J. Or-
affords a ca. 2:1 mixture of 2 and 3 and (2) in DMF, both
PdCl (MeCN) and Pd(PPh ) lead to 2 and 3 from 1, but
1
2
2
3 4
[
II
the formation of 2 is more privileged by the Pd catalyst.
5
the solvent (Figure 1) can also be explained by the differ-
ences in the solvation of the ionic intermediates that are
more stabilized in DMF than in THF, and, consequently,
they should evolve more slowly towards the products.
ganomet. Chem. 1984, 262, 263–269.
[
[
[
[
4] S. Bouquillon, J. Muzart, Eur. J. Org. Chem. 2001, 3301–3305.
5] P. M. Henry, J. Chem. Soc. Chem. Commun. 1971, 328–329.
6] P. M. Henry, J. Am. Chem. Soc. 1972, 94, 5200–5206.
7] M. M. L. Crilley, B. T. Golding, C. Pierpoint, J. Chem. Soc.
Perkin Trans. 1 1988, 2061–2067.
[
[
[
8] L. E. Overman, F. M. Knoll, Tetrahedron Lett. 1979, 20, 321–
324.
9] A. C. Oehlschlager, P. Mishra, S. Dhami, Can. J. Chem. 1984,
62, 791–797.
10] a) S. Saito, S. Hamano, H. Moriyama, K. Okada, T. Moriwake,
Tetrahedron Lett. 1988, 29, 1157–1160; b) S. Saito, A. Kuroda,
H. Matsunaga, S. Ikeda, Tetrahedron 1996, 52, 13919–13932.
11] X. Ariza, N. Fernández, J. Garcia, M. López, L. Montserrat,
J. Ortiz, Synthesis 2004, 128–134.
12] W. Kitching, Z. Rappoport, S. Winstein, W. G. Young, J. Am.
Chem. Soc. 1966, 88, 2054–2055.
[
[
[
[
13] K. Dunne, F. J. McQuillin, J. Chem. Soc. (C) 1970, 2200–2203.
14] X. Ariza, J. Garcia, Y. Georges, M. Vicente, Org. Lett. 2006,
8, 4501–4504.
[
[
15] P. M. Henry, J. Am. Chem. Soc. 1972, 94, 7316–7322.
16] a) J. Muzart, A. Riahi, J. P. Pète, J. Organomet. Chem. 1985,
Figure 4. Influence of the solvent and the conversion on the ratio
of diacetates 2 and 3 obtained from 1.
2
80, 269–279; b) R. K. Boeckman, S. W. Goldstein, M. A. Wal-
ters, J. Am. Chem. Soc. 1988, 110, 8250–8252.
ˇ
[
17] R. Ponec, R. Re rˇ icha, J. Organomet. Chem. 1988, 341, 549–
57.
18] R. G. F. Giles, V. R. Lee Son, M. V. Sargent, Aust. J. Chem.
990, 43, 777–781.
[19] J. Yu, M. J. Gaunt, J. B. Spencer, J. Org. Chem. 2002, 67, 4627–
629.
5
[
1
Experimental Section
4
General: THF was distilled from sodium/benzophenone under an
[20] In fact, such a possibility was previously eliminated by Henry
but under different experimental conditions.[
15]
argon atmosphere. Diacetates 1, 2, and 3 and PdCl
2
(MeCN)
2
were
prepared as previously reported.[
4,31]
[21] C. Moberg, L. Sutin, I. Csöregh, A. Heumann, Organometal-
lics 1990, 9, 974–980.
NMR spectra were obtained
with a Bruker AC 250 spectrometer. GC analysis were performed
with an HP6890 apparatus equipped with DB1 capillary column
[
22] S. Strömberg, M. Svensson, K. Zetterberg, Organometallics
997, 16, 3165–3168.
1
(length: 25 m, diameter: 0.32 m) and an HP 3395 integrator.
[
23] Even in the presence of chloride, cis chloropalladation of al-
kenes is more important than trans chloropalladation: P. M.
Henry, J. Org. Chem. 1972, 37, 2443–2447. For an explanation
General Isomerization Procedure and Determination of the Acetate
Ratios: To a solution of 1, 2, or 3 (700 mg, 4.07 mmol) in dry THF
or DMF (15 mL) under an atmosphere of argon was added
PdCl (MeCN) (52.8 mg, 0.05 equiv.), and the mixture was warmed
2 2
in a bath heated at 70–72 °C. Samples were removed from the reac-
2 2
of trans chloropalladation with PdCl (PhCN) in the absence
of added chloride, see W. T. Wipke, G. L. Goeke, J. Am. Chem.
Soc. 1974, 96, 4244–4249.
[24] a) A. Sen, T.-W. Lai, Inorg. Chem. 1981, 20, 4036–4038; b) A.
tion mixture at the times mentioned in the tables, filtered through
Sen, T.-W. Lai, Inorg. Chem. 1984, 23, 3257–3258.
1
Celite, and then analyzed by GC. H NMR spectroscopic analyses
[25] B. M. Trost, C. B. Lee, J. Am. Chem. Soc. 2001, 123, 3687–
3696.
were carried out after evaporation of the solvent followed by dissol-
ution of the residues in CDCl
3
.
[26] C. Reichardt, Solvents and Solvent Effects in Organic Chemis-
try, 2nd ed., VCH, Weinheim, 1988, p. 339–413.
GC analysis: Gas: nitrogen (4 bar); injector and detector tempera-
tures: 250 °C; oven temperature: 70 °C for 5 min, then 2 °C/min till
[
27] C. Reichardt, Org. Process Res. Dev. 2007, 11, 105–113.
[28] B. T. Golding, C. Pierpoint, R. Aneja, J. Chem. Soc. Chem.
Commun. 1981, 1030–1031.
2
4
00 °C; internal standard: nonane; retention times, nonane:
.67 min, 3: 10.08 min, 1: 16.12 min, 2: 17.08 min.
[29] For scales of solvent polarity, see:a) ref.[26]; b) J. March, Ad-
vanced Organic Chemistry, 4th ed., John Wiley & Sons, New
York, 1992, pp. 360–362.
1H NMR spectroscopic analysis: The ratios between the different
acetates were established from the respective integrations of the
CH of 1 (δ = 4.62 ppm) and 2 (δ = 4.58 ppm), and the CH proton
2
[30] C. Amatore, A. Jutand, G. Meyer, L. Mottier, Chem. Eur. J.
1999, 5, 466–473.
of 3 (δ = 5.74 ppm); the precision of the integrations was estimated
to beϮ3%.
[31] L. S. Hegedus, O. P. Anderson, K. Zetterberg, G. Allen, K. Sii-
rala-Hansen, D. J. Olsen, A. B. Packard, Inorg. Chem. 1977, 16,
1887–1894.
Supporting Information (see footnote on the first page of this arti-
Received: March 20, 2007
cle): Tables S1, S2, S3, S4, and S5.
Published Online: June 18, 2007
3904
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 3901–3904