Effect of the leaving group on the rate and mechanism of the palladium-catalyzed isomerization of cyclic allylic benzoates in allylic substitutions

Article abstract of DOI:10.1002/ejoc.200500708

In chloroform, the reaction of cis-5-phenylcyclohex-2-enyl 4-Z-benzoate (cis-1_z, Z = NO₂, Cl, H, Me, MeO) with Pd⁰ complexes ligated to PPh₃ is reversible and proceeds with isomerization at the allylic position.

Full text of DOI:10.1002/ejoc.200500708

FULL PAPER
DOI: 10.1002/ejoc.200500708
Effect of the Leaving Group on the Rate and Mechanism of the Palladium-
Catalyzed Isomerization of Cyclic Allylic Benzoates in Allylic Substitutions
Christian Amatore,*[a] Anny Jutand,*[a] Laure Mensah,[a] Gilbert Meyer,[a]
Jean-Claude Fiaud,*,[b] and Jean-Yves Legros[b]
Keywords: Palladium / Allylic benzoates / Allylic complexes / Mechanism / Kinetics / Isomerization
In chloroform, the reaction of cis-5-phenylcyclohex-2-enyl 4-
Z-benzoate (cis-1Z, Z = NO2, Cl, H, Me, MeO) with Pd0 com-
plexes ligated to PPh3 is reversible and proceeds with isom-
erization at the allylic position. The rate of isomerization of
cis-1Z to trans-1Z depends on the catalytic precursor:
Pd0(PPh3)4 Ͼ {Pd0(dba)2 + 2PPh3} in agreement with an SN2
mechanism in the rate-determining isomerization of the cat-
ionic (η3-allyl)palladium complexes formed in the oxidative
addition. For a given precursor, the rate of isomerization of
cis-1Z to trans-1Z also depends on the substituent Z, i.e., on
the leaving group. The isomerization rate follows the same
librium constant between the neutral cis-1Z and the cationic
(η3-allyl)palladium complex in DMF. The higher the concen-
tration of the cationic (η3-allyl)palladium complex, the faster
the isomerization of cis-1Z to trans-1Z is. The isomerization of
cis-1Z to trans-1Z and that of the cationic (η3-allyl)palladium
complexes are at the origin of the lack of stereospecificity
observed in catalytic nucleophilic allylic substitutions. These
isomerizations are affected by both the leaving groups and
the Pd0 precursors, which therefore are not “innocent” but
may play an important role in palladium-catalyzed nucleo-
philic substitutions.
–
order as the leaving group properties: 4-NO2–C6H4-CO2
Ͼ 4-Cl–C6H4-CO2– Ͼ C6H5-CO2– Ͼ 4-Me–C6H4-CO2– Ͼ 4-
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2005)
MeO–C6H4-CO2–. The same tendency is found for the equi-
Introduction
According to these mechanisms, catalytic substitutions
on one isomer should proceed either by retention or by in-
version of configuration according to the nature of the nu-
cleophile, and the reactions should be stereospecific. How-
ever, a loss of stereospecificity is often observed. This was
first accounted for by isomerization of the starting cyclic
allylic carboxylates,[4] and then by an isomerization of the
cationic (η3-allyl)palladium by an SN2 mechanism induced
by the palladium(0) catalyst, as already proposed in the acy-
clic series.[5] In this context, we have also established that
the oxidative addition of cyclic allylic carbonates is revers-
ible and proceeds by an isomerization at the allylic carbon,
via the isomerization of the cationic (η3-allyl)palladium by
an SN2 mechanism (Scheme 2, Scheme 3).[5f]
The palladium-catalyzed nucleophilic substitution on al-
lylic carboxylates is a well-documented synthetic reaction.[1]
Investigations of allylic substitutions on substituted cyclic
allylic carboxylates have brought interesting insights into
the mechanism of the successive steps of these catalytic re-
actions (Scheme 1).[2,3]
The oxidative addition of a substituted cyclic allylic car-
boxylate to a palladium(0) complex, which generates a cat-
ionic (η3-allyl)palladium complex, was established to pro-
ceed by an inversion of configuration,[3] whereas the nucleo-
philic attack proceeds: i) by inversion of configuration in
the case of soft or stabilized nucleophiles, NuS, which attack
the allylic ligand with an overall retention of configuration
or ii) by retention of configuration in the case of hard or
nonstabilized nucleophiles, NuH, which attack the palla-
dium center, resulting in an overall inversion of configura-
tion (Scheme 1).[2,3]
Palladium-catalyzed asymmetric allylic substitution may
also depend on the leaving group attached to the allylic moi-
eties.[1d,5d,6]
We wish to report here a mechanistic study of the oxidat-
ive addition of cyclic allylic para-Z-substituted benzoates
(cis-1Z) with Pd0 complexes, with a focus on the effect of
the leaving group on the kinetics, thermodynamics and
stereochemistry of the oxidative addition. To the best of
our knowledge, no kinetic or thermodynamic data on the
oxidative addition dealing with the effect of the leaving
group on the palladium-catalyzed isomerization of allylic
carboxylates are available in the literature.
[a] Ecole Normale Supérieure, Département de Chimie, UMR
CNRS-ENS-UPMC 8640,
24, Rue Lhomond, 75231 Paris Cedex 5, France
Fax: +33-1-44323325
E-mail: Anny.Jutand@ens.fr
[b] ICMO, Bat. 420, Université Paris XI,
91405 Orsay, France
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Eur. J. Org. Chem. 2006, 1185–1192
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1185

C. Amatore, A. Jutand, L. Mensah, G. Meyer, J.-C. Fiaud, J.-Y. Legros
FULL PAPER
Scheme 1.
Scheme 2.
Scheme 4.
1
It is worthwhile to note that the H NMR spectrum ex-
Results and Discussion
hibited only the signals of cis-1H and trans-1H. No other
signals could be detected, in particular those which would
have characterized the cationic complexes, trans-2+ or cis-
2+, generated in the reaction of cis-1H with the Pd0 com-
plexes. This suggests that for the experimental conditions
indicated above, the relative abundance of the cis-1/trans-1
pair vs. the cationic trans-2+/cis-2+ pair lies in favor of the
Mechanism of the Reaction of cis-5-Phenylcyclohex-2-enyl
Benzoate (cis-1H) with Pd0 Complexes: Influence of the
Catalytic Precursor
Reversibility of the Oxidative Addition and Isomerization
The reaction of cis-1H (40 m) with two precursors of starting compounds. The equilibrium constant between
palladium(0) [Pd0(PPh3)4 (20 m) or Pd0(dba)2 (20 m)] as- trans-1H and cis-1H has been determined: K = [trans-1H]/
sociated with two equiv. of PPh3 (40 m), was monitored [cis-1H] = 1.17 (CDCl3, 25 °C).
1
1
by H NMR spectroscopy in CDCl3. The H NMR signals
The percentage of trans-1H in the mixture trans-1H + cis-
of trans-1H were soon detected, and were found to increase
1H was plotted against time for the two catalytic precursors
with time at the expense of those of cis-1H, until the (Figure 1).
thermodynamic equilibrium was reached: [trans-1H]/[cis-1H]
= 54:46 (Scheme 4).
According to Figure 1, the isomerization cis-1H/trans-1H
was faster when Pd0(PPh3)4 was used as the precursor than
Scheme 3.
1186
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 1185–1192

Studies of the Palladium-Catalyzed Isomerization of Cyclic Allylic Benzoates
FULL PAPER
Figure 1. a) (___ and solid circles): percentage of trans-1H formed in the reaction of cis-1H (40 m) with the Pd0 complex generated from
Pd0(PPh3)4 (20 m) in CDCl3 at 25 °C, as determined by 1H NMR spectroscopy. (– – – and solid triangles): percentage of trans-1H formed
in the reaction of cis-1H (40 m) with the Pd0 complex generated from Pd0(dba)2 (20 m) and PPh3 (40 m) in CDCl3 at 25 °C. b) Same
data as in Figure 1 (a) from 0 to 300 min.
with Pd0(dba)2 + 2PPh3. This situation was already encoun-
tered with a cyclic allylic carbonate (Scheme 2), and was
rationalized by invoking a mechanism involving the isomer-
ization of the cationic complexes by an SN2 mechanism
Scheme 8.
(Scheme 3).[5f] The rate of isomerization must increase when
the Pd0(PPh3)2 concentration is made higher, since this low-
ligated intermediate is directly involved in the bimolecular
rate-determining step. We know from our previous work
that for identical concentrations of Pd0(PPh3)3 [the major
complex generated from Pd0(PPh3)4] and Pd0(dba)(PPh3)2
[the major complex generated from Pd0(dba)2 + 2PPh3], the
Pd0(PPh3)2 concentration in the equilibrium of Scheme 5 is
lower than in the equilibrium of Scheme 6 (KЈ0 ϾK0).[7]
The kinetic law for the overall isomerization process
(Scheme 8) is given in Scheme 9 (x= [cis-1Z]/C0 at t; C0 =
initial concentration of cis-1Z).
Scheme 9.
Scheme 5.
The plot of {ln[x(K + 1) – 1] – lnK} against time was
linear for the catalytic precursors, Pd0(PPh3)4 (Figure 2, a)
and Pd0(dba)2 + 2PPh3 (Figure 2, b), as predicted by the
kinetic law reported in Scheme 9. This validates the kinetic
framework shown in Scheme 8.
Scheme 6.
The kinetics of the isomerization cis-1H/trans-1H were
thus thoroughly investigated to confirm the SN2 mechanism
proposed in Scheme 7.
The value of kЈobs = 2.2×10–3 s–1 for Pd0(PPh3)4 was de-
termined from the slope of the straight line (Figure 2, a).
The rate constant kobs = 1.5×10–4 s–1 for Pd0(dba)(PPh3)2
(Figure 2, b).[8a] Two different values for the observed rate
Kinetics of the Isomerization cis-1H/trans-1H
Since the cationic species are not observable, they are constants were obtained, depending on the catalytic precur-
under steady state conditions. The overall isomerization sor. This indicates that the rate of isomerization depends on
may be represented by the formal equation in Scheme 8 as the concentration of the active Pd0(PPh3)2, which in turn
the results of three successive equilibria (Scheme 7), where depends on the precursor for identical Pd0 overall concen-
k+ and k– are apparent rate constants.
trations.[7] This is again in full agreement with the rate law
Scheme 7.
Eur. J. Org. Chem. 2006, 1185–1192
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1187

C. Amatore, A. Jutand, L. Mensah, G. Meyer, J.-C. Fiaud, J.-Y. Legros
FULL PAPER
could be measured in CDCl3. The absorbance of
Pd0(dba)(PPh3)2 at λmax= 400 nm[9] decreased and stabi-
lized at different values upon successive additions of cis-1H
(n equiv.) (Figure 3), as in an equilibrium (Scheme 10). The
equilibrium constant K0K1 = C0x2(1+x)/[(n – x)(1 – x)] was
determined by plotting x2(1 + x)/(n – x) vs. (1 – x) (see
Figure S1, Supporting Information) [(1 – x) = molar frac-
tion of Pd0(dba)(PPh3)2 in the equilibrium, (1 – x) =
(Dequil – Dlim)/(D0 – Dlim) with Dequil = absorbance of
Figure 2. a) Kinetics of the isomerization of cis-1H to trans-1H cata- Pd0(dba)(PPh3)2 at the equilibrium in the presence of
lyzed by the Pd0 complex generated from Pd0(PPh3)4 (20 m) in
n equiv. of cis-1H, Dlim = residual absorbance, D0 = initial
CDCl3 at 25 °C. Plot of ln[x(K + 1) – 1] – lnK against time; x =
absorbance]. K0K1 was determined to be 7.2×10–6  (DMF,
[cis-1Z]/C0 at t, C0 = initial concentration of cis-1H = 40 m, K =
25 °C).
equilibrium constant between cis-1H and trans-1H determined from
the data in part a of of Figure 1. b) Same as inpart a of Figure 2
for the kinetics of isomerization cis-1H (C0 = 40 m) to trans-1H
catalyzed by the Pd0 complex generated from Pd0(dba)2 (20 m)
and PPh3 (40 m) in CDCl3 at 25 °C.
given in Scheme 9 and the SN2 mechanism proposed in
Scheme 7.
Scheme 10.
From the experimental values of kЈobs and kobs, one de-
duces that kЈobs/kobs = [Pd0(PPh3)2]equil6/[Pd0(PPh3)2]equil5
=
14, which gives the ratio of the concentration of [Pd0-
(PPh3)2]equil6 and [Pd0(PPh3)2]equil5 in their equilibria with
their respective precursors in CDCl3 (Scheme 5 and
Scheme 6). The value of the equilibrium constant K0
(Scheme 5) is not known in CDCl3, but has been deter-
mined independently in DMF at 25 °C (K0 = 1.5×10–5 )
with [Pd0(PPh3)2]equil5 = 1.5×10–5 .[8b] As far as Pd0-
(PPh4)4 is concerned, the equilibrium constant KЈ0
(Scheme 6) could not be determined independently. How-
ever, in a previous work, we established that, for identical
concentrations of PhI and the Pd0 precursor, the oxidative
addition was 8.6 times faster with Pd0(PPh4)4 than with
Pd0(dba)(PPh3)2 in DMF at 20 °C,[7] thus KЈ0/K0 = 8.6
Figure 3. UV spectrum of a solution of Pd(dba)2 (1 m) and PPh3
(2 m) in the presence of n equiv. of cis-1H at 25 °C. Dequil = ab-
sorbance of Pd(dba)(PPh3)2 at equilibrium in the presence of
n equiv. of cis-1H; Dlim = residual absorbance observed after ad-
dition of PhI (5 m)].
(DMF, 20 °C). This gives a ratio of [Pd0(PPh3)2]equil6
/
[Pd0(PPh3)2]equil5 = 8.6 (DMF, 20 °C) which is not far from
the value of 13, obtained above in this work in CDCl3 at
25 °C. The values of k+ and k– have been estimated from
the estimated value of k+ + k– associated with the value of
K = k+/k– = 1.17 determined in this work.[8c]
The same procedure was used to determine the overall
equilibrium constant KЈ0K1 for the formation of the cat-
ionic complex trans-2+ by reaction of cis-1H with Pd0-
(PPh3)3 generated from Pd0(PPh3)4 (1 m) in DMF
(Scheme 11 and Figures S2a and S2b).
This kinetic study establishes that the oxidative addition
of the cyclic allylic benzoate (cis-1H) with Pd0 complexes
ligated by PPh3 is reversible and proceeds with isomeriza-
tion at the allylic position via isomerization of the cationic
complexes 2+ by an SN2 mechanism (Scheme 7).
Thermodynamics of the Oxidative Addition
The thermodynamics of the overall equilibrium
(Scheme 10) in which the cationic complex trans-2+ was
formed by reaction of cis-1H with Pd0(dba)(PPh3)2 gener-
ated from Pd0(dba)2 (1 m) and 2 equiv. of PPh3 (2 m)
was investigated by UV spectroscopy in DMF at 25 °C.
DMF was used as solvent instead of CHCl3 because it al-
lowed the observation of cationic complexes 2+ by conduc-
tivity measurements (free ions in DMF,[10] vide infra in Fig-
ure S6 in the Supporting Information), whereas ion pairs
Scheme 11.
Mechanism of the Reaction of cis-5-Phenylcyclohex-2-enyl
4-Z-Substituted Benzoate (cis-1Z) with Pd0 Complexes:
Influence of the Leaving Group
The effect of the leaving group on the kinetics of the
are produced in THF[11] or CHCl3, and no conductivity palladium-catalyzed isomerization of cis cyclic allylic para-
1188
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 1185–1192

Studies of the Palladium-Catalyzed Isomerization of Cyclic Allylic Benzoates
FULL PAPER
Z-substituted benzoates (cis-1Z) has been investigated for Z
= NO2, Cl, Me and OMe, with the two precursors Pd0-
(dba)2 + 2PPh3 and Pd0(PPh3)4 in CDCl3, under the same
experimental conditions as for Z = H (Scheme 4). para-Z-
Substituted benzoates are good leaving groups and conse-
quently poor nucleophiles when Z is a good electron-with-
drawing group. The kinetics of the isomerization of cis-1Z
1
to trans-1Z was followed by H NMR, as above for cis-1H
[Figure 4 for Pd0(dba)2 + 2PPh3 and Figure 5 for Pd0-
(PPh3)4].
Figure 5. Percentage of trans-1Z formed in the reaction of cis-1Z
(40 m) with the Pd0 complex generated from Pd0(PPh3)4 (20 m)
1
in CDCl3 at 25 °C, as determined by H NMR spectroscopy. (᭹)
Z = Cl; (o) Z = H; (᭿) Z = Me; (᭡) Z = OMe.
Figure 4. Percentage of trans-1Z formed in the reaction of cis-1Z
(40 m) with the Pd0 complex generated from Pd0(dba)2 (20 m)
1
and PPh3 (40 m) in CDCl3 at 25 °C, as determined by H NMR
spectroscopy. (+) Z = NO2; (᭹) Z = Cl; (᭿) Z = Me; (᭡) Z = OMe.
Figure 6. a) Kinetics of the isomerization cis-1OMe to trans-1OMe
catalyzed by the Pd0 complex generated from Pd0(PPh3)4 (20 m)
in CDCl3 at 25 °C. Plot of ln[x(K + 1) – 1] – lnK against time (x =
[cis-1OMe]/C0 at t, C0 = initial concentration of cis-1OMe = 40 m,
K = equilibrium constant between cis-1OMe and trans-1OMe).
The observed rate constants kobs (Scheme 9) have been
determined as for cis-1H. For example, the determination of
kobs for the isomerization of cis-1OMe (40 m) to trans-1OMe
catalyzed by Pd0(PPh3)4 (20 m) is shown in Figure 6. The
other kinetic curves are shown in the Supporting Infor-
The values of the equilibrium constants K0K1 and KЈ0K1
mation (Figures S3 and S4). The values of the rate con- (Scheme 10, Scheme 11) have also been determined by UV
stants kobs and equilibrium constants K are collected in spectroscopy in DMF as above for cis-1H.[12] When trying
Table 1. The values of k+ and k– have been estimated.[8d]
to determine the equilibrium constant K0K1 for cis-1OMe
Table 1. Kinetic and thermodynamic data for the isomerization of cis-1Z to trans-1Z catalyzed by Pd0 complexes in CDCl3 at 25 °C (for
the definition of K and kobs, see Scheme 8 and Scheme 9). Thermodynamic data for the formation of the cationic trans- and cis-2+
complexes in DMF at 25 °C (for the definition of K0K1 and KЈ0K1, see Scheme 10 and Scheme 11).
cis-1Z
K
4-Z (σ)[a]
1
2
3
4
5
OMe (–0.268)
Me (–0.17)
H (0)
Cl (0.227)
NO2 (0.778)
0.8 0.1
0.9 0.1
1.2 0.2
1.3 0.1
0.8 0.1
Precursor
L = PPh3
4–Z
kobs (s–1)
K0K1 (M)
6
7
8
9
Pd0(dba)L2
Pd0(dba)L2
Pd0(dba)L2
Pd0(dba)L2
Pd0(dba)L2
OMe
Me
H
Cl
NO2
n.d.[b]
n.d.[b,c]
6.5( 0.1)×10–5
1.7( 0.1)×10–4
1.7( 0.1)×10–4
4.5( 0.1)×10–3
1.5×10–5
7.9( 0.5) ×10–6
n.d.[b]
10
1.0
kobs (s–1)
KЈ0K1 (M)
1
Pd0L4
Pd0L4
Pd0L4
Pd0L4
Pd0L4
OMe
Me
H
Cl
NO2
1.0( 0.1)×10–3
1.0( 0.1)×10–3
2.2( 0.1)×10–3
3.8( 0.1)×10–3
n.d.[b]
n.d.[b]
12
13
14
15
n.d.[b]
2.4×10–5
3.1×10–4
n.d.[b]
[a] σ: para Hammett constant. [b] N.d.: not determined. [c] K0K4 was determined instead (see text).
Eur. J. Org. Chem. 2006, 1185–1192
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1189

C. Amatore, A. Jutand, L. Mensah, G. Meyer, J.-C. Fiaud, J.-Y. Legros
FULL PAPER
(Figure S5a), the plot of x2(1+x)/(n – x) vs. (1 – x) was not mechanism induced by Pd0(PPh3)2 (Scheme 7), as estab-
linear as expected (vide supra for cis-1H). This suggests that lished by Bäckvall et al. in a related series of cationic com-
the equilibrium did not involve the formation of the cat- plexes (CO2Me instead of Ph substituted on the cycle).[5e]
ionic complex trans-2+ but presumably the neutral interme-
For a given precursor, one sees that the isomerization
diate complex (η2-cis-1OMe)Pd0(PPh3)2 (Scheme 12). We of cis-1Z to trans-1Z is faster when Z is a better electron
have already reported kinetic and structural evidences for withdrawing group in the order NO2 Ͼ Cl Ͼ H Ͼ Me
the formation of such intermediate neutral complexes in Ͼ OMe (compare kobs in entries 6–10, entries 11–15 in
acyclic allylic acetates.[13]
Table 1). In other words, the better the leaving group
–
4-Z-C6H4-CO2 , the faster the isomerization is: 4-NO2–
C6H4-CO2– Ͼ 4-Cl–C6H4-CO2– Ͼ C6H5-CO2– Ͼ 4-Me–C6H4-
–
CO2– Ͼ 4-MeO–C6H4-CO2 . This tendency is correlated
with the evolution of the equilibrium constant of the forma-
tion of the cationic complex with Z (see K0K1 for
Pd0(dba)(PPh3)2 and KЈ0K1 for Pd0(PPh3)4 in Table 1, de-
termined in DMF).[14] Indeed, the better the leaving group,
the higher the concentration of the cationic complex in its
equilibrium with the neutral allylic benzoates, and thus the
Scheme 12.
Such a neutral complex might accumulate if its ioniza- faster the isomerization of cis-1Z to trans-1Z.
tion to trans-2+ were slow. To check this hypothesis, the
reaction of Pd0(dba)(PPh3)2 with cis-1OMe in DMF was
monitored by conductivity measurements to detect any cat-
Conclusions
ionic complex.[10] A low and constant conductivity of
2.5 µS·cm–1 (compared to the solvent residual conductivity
In chloroform, the reactions of cyclic allylic para-Z-sub-
of 1.5 µS·cm–1) was observed after addition of cis-1OMe stituted benzoates cis-1Z with Pd0 complexes generated
(32 m) to Pd0(dba)(PPh3)2 (1 m) up to 4000 s. This indi- from the precursors Pd0(PPh3)4 or [Pd0(dba)2 + 2PPh3] are
cates that cationic complex trans-2+ was not generated in reversible and proceed with isomerization at the allylic posi-
high concentration (vide infra for the cationic complex gen- tion to give trans-1Z. This isomerization is a consequence
+
erated from cis-1NO2). Assuming that the investigated equi- of the isomerization of the cationic (η3-allyl)Pd(PPh3)2
librium stopped at the level of (η2-cis-1OMe)Pd0(PPh3)2 complex through an SN2 mechanism induced by the species
(Scheme 12), the expression of the equilibrium constant Pd0(PPh3)2. The rate of isomerization depends on the cata-
would become K0K4 = x(1+x)/[(n – x)(1 – x)]. The plot of lytic precursor {Pd0(PPh3)4 Ͼ [Pd0(dba)2 + 2PPh3]} in
x(1+x)/(n – x) vs. (1 – x) was indeed linear (Figure S5b) in agreement with the SN2 mechanism, since the concentration
agreement with the equation in Scheme 12. The value of of Pd0(PPh3)2 is about 10 times less in the second case than
K0K4 = 0.025 (DMF, 25 °C) was determined from the corre- in the first, for an identical overall catalyst concentration.
sponding slope. The fact that the formation of the cationic The para-substituent Z on the phenyl group of the benzoate
complex trans-2+ from (η2-cis-1OMe)Pd0(PPh3)2 was ex- has been modified to examine the effect of the leaving group
tremely slow explains why the isomerization cis-1OMe to on the rate of the isomerization process. The rate of isomer-
trans-1OMe in the presence of Pd0(dba)(PPh3)2 was the slow- ization follows the same order as the leaving group proper-
–
est one observed (Figure 4).[14]
ties: 4-NO2–C6H4–CO2– Ͼ 4-Cl–C6H4–CO2– Ͼ C6H5–CO2
–
–
When considering cis-1NO2, which possesses a better leav- Ͼ 4-Me–C6H4-CO2 Ͼ 4-MeO–C6H4-CO2 . From the de-
ing group than cis-1OMe (4-NO2–C6H4-CO2– Ͼ 4-MeO– termination of the equilibrium constant between the neutral
–
C6H4-CO2 ), the conductivity of the solution did increase cis-1Z and the cationic complex in DMF, it also comes out
over time to reach a final conductivity of 16 µS·cm–1 by that under identical experimental conditions, the higher the
reacting cis-1NO2 (32 m) with Pd0(dba)(PPh3)2 (1 m) in concentration of the cationic complex, the faster the isom-
DMF at 25 °C (Figure S6). This experiment showed that erization of cis-1Z to trans-1Z.
the cationic complex 2+ was indeed formed from cis-1NO2
,
In a real catalytic reaction, a competition may occur at
which was more reactive than cis-1OMe in both the complex- the level of the cationic (η3-allyl)palladium complex be-
ation and the ionization steps. tween the re-entrance of the leaving group and the attack
From the values of kobs collected in Table 1, one observes of the nucleophile. This would be accentuated for poor nu-
that for a given Z, the isomerization of cis-1Z to trans-1Z is cleophiles. This competition should be more in favor of the
always faster when catalyzed by Pd0(PPh3)4 than by attack of the leaving group, viz. in favor of the isomeriza-
Pd0(dba)2 + 2PPh3, in chloroform. These data confirm that: tion, as the catalytic reaction proceeds because the concen-
i) the oxidative addition in which the cationic complexes tration of the leaving group increases while that of the nu-
are formed is reversible, and ii) the isomerization process cleophile decreases during the course of the catalytic reac-
depends on the structure of the Pd0 precursor i.e., on the tion. Therefore, the leaving group cannot be considered as
concentration of the active Pd0(PPh3)2 complex. This shows a “spectator and innocent” species because of the reversibil-
that the overall isomerization proceeds through the revers- ity of the oxidative addition and its ensuing influence on
ible isomerization of the cationic complexes via an SN2 the rate of the isomerization processes. The structure of the
1190
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 1185–1192

Studies of the Palladium-Catalyzed Isomerization of Cyclic Allylic Benzoates
FULL PAPER
= 1713 cm–1. HRMS C19H17ClO2: calcd. 312.0912; found 312.0895.
C19H17ClO2 (312): calcd. C 72.96, H 5.48; found C 73.32, H 5.65.
catalytic precursor is not innocent either, since its choice
determines the concentration of the key intermediate Pd0L2
and therefore affects the rate of the competitive isomeriza-
tion. This appears to be a likely origin of the lack of stereo-
specificity observed in catalytic reactions. A pertinent
choice of the leaving group and the catalytic precursor
should therefore allow optimization of the overall competi-
tion and therefore improve the stereospecificity of the nucle-
ophilic substitution.
cis-5-Phenylcyclohex-2-enyl 4-Nitrobenzoate was obtained in 84%
yield as pale yellow crystals. m.p. 85 °C. 1H NMR (250 MHz,
CDCl3): δ = 2.03 (td, J = 12 and 10 Hz, 1 H), 2.15–2.55 (m, 3 H),
2.95–3.15 (m, 1 H), 5.7–5.85 (m, 2 H), 5.95–6.05 (m, 1 H), 7.1–7.4
(m, 5 H), 8.18 (d, J = 9 Hz, 2 H), 8.28 (d, J = 9 Hz, 2 H). 13C
NMR (62.9 MHz, CDCl3): δ = 164.4, 150.5, 144.9, 135.9, 131.3,
130.7, 128.6, 126.8, 126.6, 126.4, 123.5, 72.6, 38.8, 35.1, 33.2. IR:
ν
= 1720 cm–1. HRMS C19H17NO4: Calcd. 323.1152; found:
˜
CO
323.1163. C19H17NO4 (323): calcd. C 70.58, H 5.30; found C 70.42;
H 5.35.
Experimental Section
Typical Procedure for UV Experiments: From a stock solution of
DMF (10 mL) containing Pd0(dba)2 (5.7 mg, 10 µmol) and PPh3
(5.4 mg, 20 µmol), a 300 µL aliquot was transferred under Ar to
the thermostatted UV cell (1 mm length) and a UV measurement
was performed. It was followed by successive additions of known
amounts of cis-1Z (see n in Figure 3) from a stock solution (0.3 
in DMF). The UV measurement was performed immediately after
hand-shaking the cell. From a stock solution of DMF (10 mL)con-
taining Pd0(PPh3)4 (11.5 mg, 10 µmol), a 300 µL aliquot was trans-
ferred under Ar to the thermostatted UV cell (1 mm length) and
a UV measurement was performed. It was followed by successive
additions of known amounts of cis-1Z (see n in Figure 3) from a
stock solution (0.3  in DMF). The UV measurement was per-
formed immediately after hand-shaking the cell.
General: 1H and 13C NMR spectra were recorded with a Bruker
AC-250 MHz spectrometer with tetramethylsilane as an internal
standard. UV spectra were recorded with an mc2 Safas Monaco
spectrometer. Conductivity measurements were performed with a
Tacussel CDM210 conductivity meter (cell constant: 1 cm–1). All
experiments were performed under Ar.
Chemicals: DMF was distilled from calcium hydride under vacuum
and kept under argon. CDCl3, dba, and PPh3 were commercial.
Pd0(dba)2,.[15] cis-5-phenylcyclohex-2-enol,[15] and cis-5-phenylcy-
clohex-2-enyl benzoate[16] were prepared as described in the litera-
ture.
General Procedure for the Synthesis of cis-5-Phenylcyclohex-2-enyl
4-Z-Benzoate: cis-5-Phenylcyclohex-2-enol (3 mmol) was treated
with a 4-Z-benzoyl chloride (1.1 equiv.) in the presence of DMAP
(0.1 equiv.) and Et3N (1.2 equiv.) in Et2O (10 mL). The reaction
mixture was stirred at room temperature overnight, diluted with
20 mL of diethyl ether, and washed successively with 1  aqueous
HCl and saturated Na2CO3. The organic phase was dried with
MgSO4 and concentrated. The crude product was purified by flash
chromatography (silica, heptane/ethyl acetate: 8:2) to give the cor-
responding 4-substituted benzoate.
Typical Procedure for Conductivity Measurements: To a thermostat-
ted cell connected to a Schlenk line and containing a solution of
Pd0(dba)2 (5.7 mg, 10 µmol) and PPh3 (5.3 mg, 20 µmol) in DMF
(10 mL) was added cis-1OMe (98.4 mg, 320 µmol). The conductivity
was recorded over time using a computerized home-made program.
In another experiment, cis-1NO2 (103.4 mg, 320 µmol) was added
to the solution of the Pd0 complex, and the measurement was re-
peated.
Typical Procedure for the Kinetics of Isomerization of cis-1Z to
1
cis-5-Phenylcyclohex-2-enyl 4-Methoxybenzoate was obtained in
90% yield as white crystals. m.p. 68 °C. 1H NMR (250 MHz,
CDCl3): δ = 1.98 (td, J = 12, 10 Hz, 1 H), 2.15–2.55 (m, 3 H),
2.95–3.15 (m, 1 H), 3.86 (s, 3 H), 5.7–5.85 (m, 2 H), 5.95–6.05 (m,
1 H), 6.91 (d, J = 9 Hz, 2 H), 7.15–7.4 (m, 5 H), 7.99 (d, J = 9 Hz,
2 H) ppm. 13C NMR (62.9 MHz, CDCl3): δ = 166.2, 163.4, 145.3,
131.8, 130.5, 128.7, 127.5, 126.9, 126.6, 123.0, 113.7, 71.3, 55.6,
trans-1Z, as Followed by H NMR Experiments: In an NMR tube
containing Pd0(dba)2 (5.7 mg, 10 µmol), PPh3 (5.3 mg, 20 µmol),
cis-1H 5.6 mg (20 µmol) was added, followed by of CDCl3 (0.5 mL).
1H NMR was performed vs. time up to equilibrium. To an NMR
tube containing Pd0(PPh3)4 (11.5 mg, 10 µmol) was added cis-1Cl
(6.3 mg, 20 µmol) followed by CDCl3 (0.5 mL). 1H NMR was per-
formed vs. time up to equilibrium.
39.1, 35.4, 33.6 ppm. IR: ν
= 1705 cm–1. HRMS C20H20O3:
˜
CO
Supporting Information: Graphs for the determination of equilib-
rium and rate constants.
calcd. 308.1407; found 308.1416. C20H20O3 (308): calcd. C 77.90,
H 6.54; found C 77.93, H 6.59.
cis-5-Phenylcyclohex-2-enyl 4-Methylbenzoate was obtained in 92%
yield as white crystals. m.p. 53 °C. H NMR (250 MHz, CDCl3): δ
Acknowledgments
1
= 1.98 (td, J = 12, 10 Hz, 1 H), 2.15–2.55 (m, 3 H), 2.40 (s, 3 H),
2.95–3.15 (m, 1 H), 5.7–5.85 (m, 2 H), 5.95–6.05 (m, 1 H), 7.05–
7.4 (m, 7 H), 7.93 (d, J = 8 Hz, 2 H) ppm. 13C NMR (62.9 MHz,
CDCl3): δ = 166.5, 145.3, 143.7, 130.5, 129.8, 129.2, 128.7, 127.9,
This work has been supported in part by the Centre National de
la Recherche Scientifique (UMR CNRS-ENS-UPMC 8640) and
the Ministère de la Recherche (Ecole Normale Supérieure). We
thank Johnson Matthey for a generous loan of sodium tetrachlo-
ropalladate.
127.4, 126.9, 126.6, 71.4, 39.1, 35.4, 33.6, 21.8 ppm. IR: ν
=
˜
CO
1708 cm–1. HRMS C20H20O2: calcd. 292.1458; found 292.1467.
C20H20O2 (292): calcd. C 82.16, H 6.89; found C 82.32, H 6.91.
[1] a) S. A. Godleski, in: Comprehensive Organic Synthesis, vol. 4
(Eds: B. M. Trost, I. Pflemming, M. F Semmelhack) Perga-
mon, Oxford, 1991; b) C. G. Frost, J. Howard, J. M. J. Wil-
liams, Tetrahedron: Asymmetry 1992, 3, 1089; c) J. Tsuji, Palla-
dium Reagents and catalysts, John Wiley & Sons, Chichester,
1996, p. 290; d) B. M. Trost, D. L. Van Kanken, Chem. Rev.
1996, 96, 395; e) G. Consiglio, R. Waymouth, Chem. Rev. 1989,
89, 257; f) T. Hayashi, Pure Appl. Chem. 1988, 60, 7; g) M.
cis-5-Phenylcyclohex-2-enyl 4-Chlorobenzoate was obtained in 98%
yield as white crystals. m.p. 74 °C. H NMR (250 MHz, CDCl3): δ
= 1.99 (td, J = 12, 10 Hz, 1 H), 2.15–2.55 (m, 3 H), 2.95–3.15 (m,
1 H), 5.7–5.85 (m, 2 H), 5.95–6.05 (m, 1 H), 7.1–7.45 (m, 5 H),
7.40 (d, J = 8 Hz, 2 H), 7.96 (d, J = 8 Hz, 2 H) ppm. 13C NMR
(62.9 MHz, CDCl3): δ = 165.4, 145.0, 139.3, 131.1, 130.8, 129.0,
1
128.7, 128.6, 126.9, 126.8, 126.5, 71.8, 38.9, 35.2, 33.4 ppm. IR: ν
˜
CO
Eur. J. Org. Chem. 2006, 1185–1192
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1191

C. Amatore, A. Jutand, L. Mensah, G. Meyer, J.-C. Fiaud, J.-Y. Legros
FULL PAPER
Sawamura, Y. Ito, Chem. Rev. 1992, 92, 857; h) A. Pfaltz, Acc.
Chem. Res.1993, 26, 339; i) P. S. Pregosin, R. Salzman, Coord.
Chem. Rev. 1996, 155, 36; j) G. Helmchen, A. Pfaltz, Acc.
Chem. Res. 2000, 33, 336.
the resulting straight line. The value of the concentration of
Pd0L2 (L = PPh3) in the equilibrium in Scheme 5 could be esti-
mated from the value of K0, considering the overall Pd0 con-
centration (C0 = 20 m) used in the NMR study. Thus
[Pd0L2]equil5 = 1.5×10–5  (DMF, 25 °C, Scheme 5); c) The val-
ues of k+ and k– could be estimated using the concentration of
[Pd0L2]equil5 determined in DMF at 25 °C, although the solvent
is slightly different (CDCl3 instead of DMF), by using the
value of k+ + k– = kobs/[Pd0L2]equil5 associated to K = k+/k– =
1.17 (determined in CDCl3 at 25 °C, vide supra). One esti-
mated: k+ = 5.4 –1s–1 and k– = 4.6 –1s–1 (25 °C); d) For Z =
Me, k+ = 2.1 0.1 –1s–1 and k– = 2.3 0.1 –1s–1. For Z = Cl,
k+ = 6.3 0.1 –1s–1 and k– = 4.9 0.1 –1s–1. For Z = NO2,
k+ = 133 1 –1s–1 and k– = 166 1 –1s–1.
[2] a) J.-E. Bäckvall, R. E. Nordberg, J. Am. Chem. Soc. 1981, 103,
4959; b) B. Åkermark, A. Jutand, J. Organomet. Chem. 1981,
217, C41; c) J.-E. Bäckvall, R. E. Nordberg, K. Zetterberg, B.
Åkermark, Organometallics 1983, 2, 1625.
[3] J. C. Fiaud, J. Y. Legros, J. Org. Chem. 1987, 52, 1907.
[4] a) B. M. Trost, T. R. Verhoeven, J. F. Fortunak, S. M. McEl-
vain, Tetrahedron Lett. 1979, 2301; b) B. M. Trost, T. R. Verho-
even, J. Am. Chem. Soc. 1980, 102, 4730.
[5] For postulated and established SN2 mechanism, see: a) T. Taka-
hashi, Y. Jinbo, K. Kitamura, J. Tsuji, Tetrahedron Lett. 1984,
25, 5921; b) P. B. Mackenzie, J. Whelan, B. Bosnich, J. Am.
Chem. Soc. 1985, 107, 2046; c) H. Kurosawa, S. Ogoshi, N.
Chatani, Y. Kawasaki, S. Murai, I. Ikeda, Chem. Lett. 1990,
1745; d) J.-E. Bäckvall, K. L. Granberg, A. Heumann, Isr. J.
Chem. 1991, 31, 17; e) K. L. Granberg, J.-E. Bäckvall, J. Am.
Chem. Soc. 1992, 114, 6858; f) C. Amatore, S. Gamez, A. Jut-
and, G. Meyer, M. Moreno-Mañas, L. Morral, R. Pleixats,
Chem. Eur. J. 2000, 6, 3372.
[6] a) J. C. Fiaud, J. Y. Legros, J. Org. Chem. 1990, 55, 4840; b) K.
Hiroi, J. Abe, Tetrahedron Lett. 1990, 31, 3623; c) K. Hiroi, M.
Haraguchi, Y. Masuda, J. Abe, Chem. Lett. 1992, 2409; d) J. Y.
Legros, J. C. Fiaud, Tetrahedron 1994, 50, 465.
[9]
C. Amatore, A. Jutand, G. Meyer, Inorg. Chim. Acta 1998, 273,
76.
[10]
For the use of conductivity measurements for the determi-
nation of mechanisms, see: A. Jutand, Eur. J. Inorg. Chem.
2003, 2017.
C. Amatore, A. Jutand, G. Meyer, L. Mottier, Chem. Eur. J.
1999, 5, 466.
For the effect of leaving group on the equilibrium constant, see
also: N. Agenet, C. Amatore, S. Gamez, H. Gérardin, A. Jut-
and, G. Meyer, C. Orthwein, ARKIVOC, 2002, 92. http://
www.arkat-usa.org.
[11]
[12]
[7] a) C. Amatore, A. Jutand, F. Khalil, M. A. MЈBarki, L. Mott-
ier, Organometallics 1993, 12, 3168; b) C. Amatore, A. Jutand,
Coord. Chem. Rev. 1998, 178–180, 511; c) The formation of
the complex Pd(dba)(PPh3)2 is quantitative in the reaction of
Pd(dba)2 with 2 equiv. of PPh3.[7a,7b]
[13] a) C. Amatore, S. Gamez, A. Jutand, Chem. Eur. J. 2001, 7,
1273; b) C. Amatore, A. A. Bahsoun, A. Jutand, L. Mensah,
G. Meyer, L. Ricard, Organometallics 2005, 24, 1569.
[14] DMF was used as solvent instead of CHCl3 because it allows
the observation of cationic complexes 2+ by conductivity mea-
surements (see Figure S6), whereas ions pairs are produced in
CHCl3.
[15] Y. Takahashi, Ts. Ito, Y. Ishii, J. Chem. Soc., Chem. Commun.
1970, 1065.
[16] a) F. Gauchet, M. Julia, H. Mestdagh, C. Rolando, Bull. Soc.
Chim. Fr. 1987, 1036; b) E. Dunkelblum, R. Levene, J. Klein,
Tetrahedron 1972, 28, 1009.
[8] a) The straight line does not go through zero due to an S-
shaped curve at short time (see Figure 1, b); b) The value of
K0 was determined in DMF at 25 °C by performing UV spec-
troscopy on the complex Pd0(dba)(PPh3)2 generated from
Pd0(dba)2 and 2 equiv. PPh3 in DMF. The absorbance D of
Pd0(dba)(PPh3)2 at λ = 400 nm was measured for different con-
centrations C0 of Pd0(dba)2 in the range 2×10–5–10–3 . The
value of D = εlC0 – εlK0 was plotted vs. C0. K0 = 1.5×10–5
could be determined from the ratio of slope and intercept of

Received: September 19, 2005
Published Online: December 13, 2005
1192
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 1185–1192