Memory and dynamics in Pd-catalyzed allylic alkylation with P,N-ligands

Article abstract of DOI:10.1016/j.tetasy.2010.03.031

The memory effect, known to exert a strong influence on selectivity in some applications of the Pd-catalyzed allylic alkylation, has been investigated for a catalytic system based on a bidentate P,N-ligand. Although this system might be expected to show s

Full text of DOI:10.1016/j.tetasy.2010.03.031

Tetrahedron: Asymmetry 21 (2010) 1585–1592
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Memory and dynamics in Pd-catalyzed allylic alkylation with P,N-ligands
Charlotte Johansson a, Guy C. Lloyd-Jones b, Per-Ola Norrby a,
*
a University of Gothenburg, Department of Chemistry, Kemigården 4, SE-412 96 Göteborg, Sweden
b School of Chemistry, University of Bristol, Cantock’s Close, Bristol, BS8 1TS, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
The memory effect, known to exert a strong influence on selectivity in some applications of the Pd-cat-
alyzed allylic alkylation, has been investigated for a catalytic system based on a bidentate P,N-ligand.
Although this system might be expected to show strong memory effects due to the difference in the trans
influence of the two donor atoms (P > N), isotopic labeling revealed an almost complete absence of regio-
Received 10 March 2010
Accepted 29 March 2010
Available online 4 May 2010
retention. Modeling of dynamic processes in the intermediate (g
3-allyl)Pd complex allowed this obser-
Dedicated to Professor Henri Kagan on the
occasion of his 80th birthday
vation to be rationalized in terms of anion-assisted apparent rotation. The study has allowed seemingly
conflicting reports on the behavior of (
g
3-allyl)Pd systems to be unified by a computational model.
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
regiochemistry,10 or double bond geometry2,11 of the substrate.
The underlying mechanism of the memory effect has been studied,
and been found to result from the fact that different substrates
Palladium-assisted allylic alkylation, often termed as the
Tsuji–Trost reaction, is a well-explored process and is frequently
applied in organic synthesis.1 In the Tsuji–Trost reaction, isomeric
starting materials give rise to the same set of products, since the
yield different isomeric forms of the (g
3-allyl)Pd intermediate.
When all such isomers are in rapid equilibrium and nucleophilic
attack is slow (Curtin–Hammett conditions), there is no memory
effect, and isomeric substrates do indeed give identical product
reaction proceeds, in principle, via
a common intermediate
(Scheme 1). The product distribution is influenced by the allyl sub-
stituents and by the ligands (L) that are used. Although there is a
strong, substrate-dependent preference for certain motifs, for
example, a conjugated E-configured double bond in the product,
significant selectivity can still be achieved via the appropriate
matching of substrates with ligands.2,3 In particular, the asymmet-
ric allylic alkylation (AAA) has been very successful,4 especially for
distributions. However, if the dynamic processes in (g
3-allyl)Pd
complexes12 are slow, relative to the nucleophilic attack,13 then
the reactions can display substantial regio- and stereo-retention,
that is, powerful memory effects (Scheme 2).11,13,14
When the intermediate is an unsymmetrically 1,3-disubstituted
allyl, the dynamics occur in two mirror image manifolds that can
crossover, possibly via Pd(0)-catalysis15,16 (Scheme 2). However,
in most cases this process is either very slow or undetectable rela-
tive to the net catalytic flux within the individual manifolds. Pairs
of substrates related by the simultaneous inversion of both the
allylic and alkenyl configurations (e.g., R,E- and S,Z-) belong to
the same manifold, and the enantiomeric pair to the other. The fast
dynamic equilibration within a manifold and the very slow cross-
over into the complementary manifold was recently utilized as
part of an enantioconvergent procedure, where both enantiomers
of a racemic starting material selectively, and without separation,
furnished one enantiomer of an intermediate in a total synthesis
of the natural products pyranicin and pyragonicin.17,18
substrates that proceed via symmetric 1,3-disubstituted (
g
3-al-
lyl)Pd intermediates.5–7
R
X
R
Nu
R
Pd0
2L
Nu-
R
R
Pd
L
L
X
Nu
Scheme 1. Pd-assisted allylic alkylation.
The isomerization process illustrated in Scheme 2 is an g3
process, sometimes referred to as ‘ ’, or ‘syn–anti’, isomeriza-
tion. Compared to the initial (
3-allyl)Pd complex, the ( 1-allyl)Pd
–
g1
In some cases, it has been found that the product distribution is
dependent on the substrate used, even when two substrates should
proceed through a common intermediate, as outlined in Scheme 1.
This phenomenon is generally termed a ‘memory effect’,8 and can
be seen as a propensity for the product to retain the chirality,9
p–r–p
g
g
isomer has an additional coordination site available for the coordi-
nation of an additional ligand, either the counter-ion or a ligand
from solution, during or after the
mation of (
1-allyl)Pd intermediates and the rate of syn–anti isom-
g g
3- to 1-isomerization. The for-
g
erization is therefore accelerated by ligands in solution, in
* Corresponding author. Tel.: +46 31 7869034; fax: +46 31 7723840.
E-mail address: pon@chem.gu.se (P.-O. Norrby).
particular by anions such as halides or carboxylates.12 In addition,
0957-4166/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2010.03.031

1586
C. Johansson et al. / Tetrahedron: Asymmetry 21 (2010) 1585–1592
R
R
R
R
R
R
R
ple allyl rotation has been considered improbable, since PdII com-
plexes have a strong preference for the square planar geometry,
but (
1-allyl)Pd could potentially undergo positional isomerization
(Scheme 3, left). Likewise, ligand dissociation has been demon-
strated,19 and the remaining ligand could then shift position via a
T-shaped transition state or intermediate (Scheme 3, center). Final-
ly, the observation that the process is strongly accelerated by chlo-
ride ions has led to the proposal of rapid Berry pseudorotation in a
pentacoordinate complex as the major contributor to the apparent
rotation (Scheme 3, right).20
–Nu
Pd
R
X
R'
X
R'
Nu R'
g
‡
‡
R'
η3−η1
R'
R'
R'
R'
R'
R'
Pd
–Nu
Nu
R
R'
Pd
R
Returning to the asymmetric allylic alkylation (AAA), the reac-
tion is generally employed with substrates that generate a sym-
R'
η3−η1
X
R
Nu
R
Pd
metric 1,3-disubstituted (g
3-allyl)Pd complex. In such systems,
the enantioselectivity arises from the ability of the ligand to direct
the nucleophile to one specific allyl terminus. This can be achieved,
for example, by directing the nucleophile via hydrogen bond-
ing21,22 or by employing an electronically dissymmetric ligand
with a strong trans effect difference, like the highly effective P,N-li-
gands.23–26 In the latter case, it is expected that during the gener-
R'
–Nu
Pd
X
R
Nu
very slow,
possibly
via:
Pd
g g
3-allyl intermediate from the 2-alkenyl precursor,
R
R'
R
ation of the
the strong trans influence of the phosphorous will result in the
expulsion of the leaving group (X) when the alkene is orientated
such that X is trans to P. Ionization, followed by nucleophilic attack,
without isomerization, would then lead to a strong region-reten-
tion as the nucleophile would, again, be directed by the trans influ-
ence of P. When the stereoselectivity is dependent on an induced
Pd
X
X
R'
R
R
R'
R'
R'
R
R'
Pd
X
X
regioselectivity, as is the case with
g g
3-cycloalkenyl or 3-1,
R
3-diphenylallyl complexes, regio-retention will have a deleterious
effect on the stereoselectivity, albeit only at higher conversions if
there is a significant degree of kinetic resolution of the substrate.10
Despite this, the P,N-systems give fast and selective reactions, and
thus do not seem to suffer from memory effects. In stark contrast,
we have previously shown11,27 that P,Cl systems afford as much as
40% regio-retention,à despite there being a much smaller difference
in the trans effect and substantially slower reactions where one
would expect the dynamics outlined in Scheme 3 to facilitate full
equilibration and negate any opportunity for memory.
Scheme 2. Two dynamic manifolds of (g
3-allyl)Pd isomerization.
there is an apparent rotation or ‘syn–syn, anti–anti’ isomerization,
which exchanges the position of the two ligands (L) in the square
plane, relative to the
g
3-allyl moiety. Several underlying processes
have been proposed for this observed exchange (Scheme 3). A sim-
*
To address this conundrum, we have undertaken a combined
experimental and computational study of memory effects in cata-
lytic systems with P,N-ligands. We want to ensure that our results
are not complicated by kinetic resolution of enantiomeric sub-
strates,10 and have therefore elected to work with a non-chiral
P,N-ligand, Ph2P–CH2CH2–NMe2 5 which can be viewed as a chi-
meric combination of two frequently employed non-chiral palla-
dium ligands, dppe and TMEDA.
Pd
*
X
A
B
Pd
X
A
B
B
A
A
*
*
Pd
*
A
B
Pd
*
A
‡
Pd
A
2. Results and discussion
X
‡
*
Pd
The above-mentioned analysis centers around the trans effect,
and we first wanted to verify the expectation that this would favor
nucleophilic attack trans to phosphorus, a key aspect of the
mechanism for asymmetric induction by chiral P,N-ligands.28
We addressed this question computationally by calculating the
approximate transition states for attack by a sodium dimethyl mal-
Pd
A
A
A
B
*
Pd
B
X
*
*
onate nucleophile at the two termini of a simple (g
3-allyl)Pd com-
Pd
Pd
Pd
plex bearing the ligand 5 that was employed experimentally
(Fig. 1). As expected, the attack trans to the phosphorus is favored,
by ca. 7 kJ/mol. At room temperature, this energy difference corre-
sponds to about an order of magnitude in reactivity difference.
Experimental studies were then carried out with labeled acyclic
allylic substrates 1a–c, synthesized from acryloyl chloride, and the
B
A
A
*
Pd
X
*
*
B
X
Pd
A
B
B
à
Scheme 3. Three mechanisms for apparent rotation in (
g
3-allyl)Pd complexes: left
We have used a definition of stereo- and regio-retention similar to that of
hand side, g3 1-isomerization; center, ligand (B) dissociation allowing access to a
–g
enantioselectivity, that is, the excess of the isomer showing retention. Thus, if the
ratio between isomers A and B is expected to be 1 in the absence of memory effect,
and B is favored by memory, the retention is 100% Ã (1 À A/B)/(1 + A/B).
T-shaped complex; right hand side, ligand (X, e.g., counter-ion) association allowing
Berry pseudorotation.

C. Johansson et al. / Tetrahedron: Asymmetry 21 (2010) 1585–1592
1587
Table 1
Results from the allylic alkylation of 1a–c and 2a–c by NaC(Me)(CO2Et)2 using Pd
complexes of ligand 5
Entry
Substrate
Pre-catalyst
Retention§ (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1a
1b
1c
2a
2b
2c
1a
1b
1c
2a
2b
2c
2a
2b
2c
5 + Pd2dba3
5 + Pd2dba3
5 + Pd2dba3
5 + Pd2dba3
5 + Pd2dba3
5 + Pd2dba3
6
6
6
6
6
6
5 + 7
5 + 7
5 + 7
À2a
0a
À2a
2
0
3
14
17
8
10
5
6
9
1
7
Figure 1. Preferred nucleophilic attack at a (g
3-allyl)Pd complex with a P,N-ligand.
a
The memory effect should be a positive number. Small negative numbers arise
from minor errors in the integration of 1H NMR signals.
corresponding labeled cyclic substrates 2a–c, synthesized from
2-cyclohexenone, as depicted in Scheme 4.
dba is not an innocent spectator ligand, although the magnitude of
the effect is small.
The partial regio-retention in the reaction when using 6 or 7 as
pre-catalysts presumably arises from the rate of the dynamic pro-
RCOCl
O
LiAlD4
Et2O
OCOR
OH
Pyridine
CH2Cl2
cesses (Scheme 3), such as g3 1-isomerization and ligand rota-
–g
D
D
D
D
1a-c
Cl
3
tion via Berry pseudorotation induced by CF3COOÀ anion
coordination, being only slightly slower than nucleophilic attack
of the malonate on the allyl.
O
OH
OCOR
NaBD4/
RCOCl
D
D
CeCl3∗7H2O
Pyridine
2.1. Isomerization/rearrangement of starting allyls
MeOH
CH2Cl2
4
2a-c
The deuterium labeling strategy (Scheme 4) is based on the iso-
tope acting simply as a marker for the carbon that originally bore
the leaving group ‘X’. It is thus essential that the substrate does
not undergo isomerization prior to reaction. In one of the experi-
ments, substrate 2c was recovered and it was found to have
undergone complete equilibration; under such circumstances, no
regio-retention would be detected irrespective of whether the
alkylation reaction was proceeding with regio-retention or not.
To confirm if the rearrangement of allylic substrates occurs under
these conditions, a series of experiments were performed in which
the reaction was assembled without the addition of the nucleo-
phile (Table 2). After 65 h, the reaction was quenched and the mix-
ture was analyzed by 1H NMR spectroscopy. All substrates 1a–c,
2b–c, except for cyclohexenyl acetate 2a, led to full regio-isotopo-
meric equilibration of the allyl moiety, when using Pd2(dba)3 as a
Scheme 4. Preparation of labeled substrates to test for memory effects in allylic
alkylation.
Initial allylic alkylations were performed using 5 mol % of ligand
5 and 2.5 mol % of Pd2(dba)3 to give a 1:1 ligand/Pd0 pre-catalyst;
the reactions were run at room temperature for 20 h under
nitrogen atmosphere, (Chart 1). Sodium diethyl methylmalonate
was used as the nucleophile. No significant regio-retention was
observed for any of the substrates (Table 1, entries 1–6).
NMe2
NMe2
Ph2P
Ph2P
5
5
+
+
Pd2dba3
Pd
O
O
O
O
Table 2
Results from the isomerization experiments
F3C
CF3
BF4
Entry
Substrate
Catalyst
Isomerized allyl
Pd
Pd
7
Ph2P
NMe2
6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
1a
1b
1c
2a
2b
2c
2a
2b
2c
2a
2b
2c
2a
2b
2c
2a
2b
2c
5 + Pd2dba3
5 + Pd2dba3
5 + Pd2dba3
5 + Pd2dba3
5 + Pd2dba3
5 + Pd2dba3
6
Yes
Yes
Yes
No
Yes
Yes
No
No
No
No
No
No
No
No
No
‘Slight’
‘Slight’
‘Slight’
Chart 1. Pre-catalyst complexes.
Amatore and Jutand have shown that dba (E-dibenzylidene ace-
tone) undergoes equilibrium complexation with [Pd0(PPh3)2] to
generate [Pd0(dba)(PPh3)2].29 Since the pre-catalyst liberates dba,
it is possible that that an analogous coordination affects our
results. To exclude this possibility, we prepared two alternative
6
6
6 + 0.05 equiv NuÀ
6 + 0.05 equiv NuÀ
6 + 0.05 equiv NuÀ
5 + 7
pre-catalyst complexes, [(
g
3-allyl)PdÁ5]BF4 6 and [(
g
3-allyl)PdO-
COCF3]2 7, Chart 1.
5 + 7
5 + 7
The amount of catalyst was reduced to 2 mol % in order to min-
imize the amount of diethyl allylmethylmalonate 8 formed from
the pre-catalyst activation event. The reaction showed regio-reten-
tion for all the substrates (entries 7–15 in Table 1) confirming that
5 + 7 + 0.1 equiv NuÀ
5 + 7 + 0.1 equiv NuÀ
5 + 7 + 0.1 equiv NuÀ

1588
C. Johansson et al. / Tetrahedron: Asymmetry 21 (2010) 1585–1592
palladium source (Table 1, entries 1–6), but no equilibration of any
of the allyls was observed when using pre-catalysts 6 or 7 (Table 2,
entries 7–9 and 13–15). This suggested that Pd(0) was responsible
for the equilibration (via reversible ionization of the leaving
group). To test this, we conducted another series of experiments
using 6 or 7, but this time added 5 and 10 mol %, respectively, of
sodium diethyl methylmalonate in order to alkylate the pre-cata-
lyst and generate Pd(0). However, after 20 h, no equilibration
was observed when using 6 (Table 1, entries 10–12) and only a
minor extent (5–10%) of equilibration of all the allyls was observed
when using 7 (Table 1, entries 16–18), that is, there is no sufficient
equilibration to explain the low regio-retention in the reaction. By
contrast, the small amount of alkylation product 9 formed in the
reaction when using 7 showed no measurable regioselectivity for
any of the substrates. These results strongly suggest that it is not
equilibration of the allylic substrate prior to its alkylation that is
responsible for the low regio-retention in the reaction when using
7 as a pre-catalyst.
aid in the resolution of the long-standing controversy regarding
this process, we initiated a computational study of plausible dy-
namic processes in cationic (
P,N-ligand (Scheme 6).
g
3-allyl)Pd complexes with a model
‡
*
Pd
*
*
P
Pd
Pd
P
P
N
N
N
*
‡
Pd
P
N
*
*
2.2. Isomerization mechanism
Pd
Pd
‡
P
N
N
P
We became interested in the mechanism for the equilibration of
2b and 2c with their regio-isotopomers. The isomerization of allyls
has been the subject of several investigations, but mainly concerns
the stereochemical aspects of the process.30–32
*
‡
‡
*
‡
In order to examine whether the return of the leaving group
was internal or external, a crossover experiment was performed.
Cyclohexenyl carbonate and allyl benzoate were mixed, and then
a catalyst generated from 5 mol % of 5 and 2.5 mol % of Pd2(dba)3
was added. A detailed analysis of the reaction mixture by GC–MS
showed that neither crossover product (cyclohexenyl benzoate
and allyl carbonate) was detectable in the reaction mixture after
1 h or indeed after 70 h reaction time; strongly pointing to an
intramolecular or internal ion pair isomerization mechanism. A
plausible explanation for this observation is that benzoate and car-
bonate groups can stabilize the 1,3-oxo cyclohexenyl cation in the
transition state, thus leading to a Pd(II)-catalyzed 1,3-migration,
similar to the Overman rearrangement33 (Scheme 5). This process
would be expected to be less favorable for acetates, as observed
in the isomerization/rearrangement experiment, where the acetate
did not return to the allyl (entry 4 in Table 2).
*
*
Pd
Pd
Pd
Pd
P
N
P
N
N
P
N
P
Scheme 6. Plausible dynamic processes in
(g
3-allyl)Pd complexes with
P,N-ligands.
As expected, the nitrogen atom could be dissociated,19 and the
resulting complex isomerized through a T-shaped transition state;
however, the (
state for bond rotation, not a stable structure. Since a transition
state cannot itself go through another process, the (
1-allyl)Pd
g
1-allyl)Pd complex turned out to be a transition
g
complex cannot engage in the apparent rotation process that leads
to positional isomerization. Attempts to locate such a reaction path
instead yielded a skewed (
unsymmetrical rotation of the
g
3-allyl)Pd complex corresponding to
g
3-allyl. This 3-allyl structure is
g
in fact the lowest energy path for apparent rotation in the cationic
complex, with a free energy barrier of 113 kJ/mol.
R
O
O
R
R
O
O
O
O
Pd
Pd
Pd
P
P
N
N
P
P
Cl
N
Cl
Pd0
Pd0
Pd
Pd
P
N
P
Cl
Cl
R
‡
Cl
Cl
P
O
O
δ+
Pd
Cl
Pd
Cl
R
R
Pd
Pd
‡
N
N
P
N
N
N
O
O
O
O
PdII
PdII
PdII
‡
‡
Pd
Pd
Pd
Scheme 5. Plausible Pd(0)- and Pd(II)-assisted pathways for isomerization of allyl
Cl
Cl
P
substrates.
N
Cl
P
N
Pd
N
Pd
P
P
P
Cl
N
Cl
From the very minor memory effects observed in the experi-
mental studies, it is clear that the chemical information encoded
in the initially formed (
nucleophilic attack, presumably through apparent rotation. To
g
3-allyl)Pd complex is scrambled before
Scheme 7. Chloride-assisted dynamic processes in (
P,N-ligands.
g
3-allyl)Pd complexes with

C. Johansson et al. / Tetrahedron: Asymmetry 21 (2010) 1585–1592
1589
We also investigated the effect of added anionic ligands mod-
eled by a chloride anion (Scheme 7). There are no ground state
structures corresponding to trigonal bipyramids. All the stable
structures are square planar, with the last ligand loosely associated
outside the coordination plane, but not coordinated to Pd. This is,
in some ways, expected behavior, since there are no low-lying
orbitals on Pd that are able to accept electron pairs from outside
the coordination plane. An X-ray structure of a presumed pentaco-
or in other words, that the initial ionization in the catalytic cycle is
reversible. With two phosphine ligands (bidentate or bis-mono-
dentate), the equilibrium has been shown to lie on the side of
the Pd0/allyl carboxylate side.35 If the re-entry is fast relative to
the nucleophilic attack, as has been implied with bidentate phos-
phines,22,30 Curtin–Hammett considerations would apply, and thus
the geometries and energies of Pd-allyl intermediates would lose
their ability to cause a memory effect.
ordinate (
g
3-allyl)Pd complex with phenanthroline and chloride20
The full free energy surface of these processes is shown in
Scheme 8. As can be seen, chloride coordination is strongly favored,
which is in agreement with the experimental observations of a
tight ion pair.32 However, the energy gain on association may be
overestimated due to approximations in the computational model.
In any case, it is clear that isomerization barriers are very low in
the presence of anionic ligands such as chloride. Even the small
amount of comparatively weak acetate ligand present in the cata-
lytic reactions will be able to catalyze the apparent rotation, which
is in perfect agreement with the observation of a very low memory
effect in the experimental system. The lowest energy states corre-
spond to a complex with dissociated nitrogen ligand at À39 kJ/mol,
reminiscent of a previously determined X-ray structure,20 and an
does in fact corresponds to the computed structures seen here; on
close inspection, it is clear that one nitrogen is rigidly held by the
ligand framework, such that it is at a distance that corresponds to a
non-bonded interaction, not coordination. The transition states
corresponding to Berry pseudorotation exist, but the structures
connected by these paths are not pentacoordinate. Thus, these
18-electron transition states correspond to the 10-electron struc-
ture in the SN2 reaction; they are in fact transition states for ligand
exchange. This is in qualitative agreement with the view on bond-
ing in transition metal complexes reported by Landis and
Weinhold.34
The (g
1-allyl)Pd complexes are stable, square planar structures
when coordinated with a chloride ligand (Scheme7). This is in agree-
ment with the observation that syn–anti isomerization is strongly
accelerated by chloride, but also shows that this process cannot be
responsible for the apparent rotation, since no positional isomeriza-
tion is possible in this square planar complex. Thus, the only chlo-
ride-accelerated process corresponding to an overall apparent
rotation is the series of pseudo-rotation-like ligand replacement
processes shown in the bottom of Scheme 7.
(
g
1-allyl)Pd complex at À42 kJ/mol, in good agreement with
observations by Amatore and Jutand in systems with phosphine
ligands.36
The Pd0/allyl chloride complex at 32 kJ/mol is worthy of note.
From the energies, it is clear that formation of this complex can
be neglected compared to other chloride-catalyzed processes.
However, this is not necessarily true for carboxylates. The relative
leaving group ability of chloride and acetate is correlated with their
pKaH values, differing by ca. 12. This corresponds to a stability dif-
ference of ca. 70 kJ/mol at ambient temperature. The difference
In addition to the dynamic processes within the Pd-allyl mani-
fold, it has also been shown that the leaving group can re-enter,30
‡
‡
136
‡
Pd
Pd
P
125
P
P
N
N
Pd
P
N
113
‡
103
N
Pd
Pd
85
Pd
P
Cl
N
P
N
‡
‡
32
Pd
Pd
Pd
N
Cl
Cl
Cl
P
N
N
P
P
Pd
4
2
Cl
0
N
Cl
Cl
P
-9
Pd
Cl
P
-20
-34
-20
-42
P
N
‡
-32
Pd
-39
Pd
P
N
Cl
N
P
Cl
Pd
Pd
N
Pd
Cl
P
P
Cl
N
N
Scheme 8. Free energies of isomerization processes in Pd-allyl complexes.

1590
C. Johansson et al. / Tetrahedron: Asymmetry 21 (2010) 1585–1592
should be smaller in this system, since the comparison is being
made between a covalently attached ligand with a coordinated,
not free, leaving group; nonetheless, it is still plausible that the
Pd0/allyl carboxylate complex could become significant, or even
the resting state.35
by warming with a heat-gun. Column chromatography was per-
formed using Silica Gel 60, 0.04–0.06 mm.
Products were identified by NMR spectroscopy. 1H and 13C NMR
spectra were recorded on a 400 MHz Varian instrument and a JEOL
Eclipse 400 instrument, operating at 400 MHz for 1H nuclei. Sam-
ples were dissolved in CDCl3 and chemical shifts are given in (parts
per million) ppm relative to residual CHCl3 (7.27 ppm). Substrates
were synthesized using modified literature procedures.
3. Conclusion
The expected regio-retention in allylic alkylations with deliber-
ately simple P,N-ligands could be detected experimentally, but
only to a very low extent, and not when using a pre-catalyst con-
taining dba. The presence of coordinating anions, such as the leav-
ing group, suppresses the memory effect. This is a strong indication
that the apparent rotation is fast compared to nucleophilic attack
4.1.1. Allyl alcohol-1-D2 337
At first, LiAlD4 (262 mg, 6.25 mmol) was added to dry Et2O
(15 ml) at 0 °C. The solution was flushed with N2 and then cooled
to 0 °C in an ice-bath. Acryloyl chloride (813
added dropwise. The reaction mixture was stirred at 0 °C for 2 h.
The reaction was then quenched with H2O (260 l), NaOH (15%,
260 l), and H2O (790 l), and then stirred at room temperature
ll, 10 mmol) was
for the intermediate (g
3-allyl)Pd complex, and that the apparent
l
l
l
rotation is catalyzed by anionic ligands in solution, in good agree-
ment with the literature data.12
for 30 min. The white precipitate was filtered off and washed with
Et2O (2–3 ml). The solution was dried with MgSO4, filtered through
a glass filter funnel, and the solvent was distilled off, yielding the
crude product as a yellow oil. The crude product was used directly
without further purification. 1H NMR (400 MHz, CDCl3): d (ppm)
5.99 (m, 1H), 5.27 (m, 1H), 5.14 (m, 1H).
The computational study has identified intermediates and
interconversion paths relevant to the observations. In general, dy-
namic processes are slow in cationic complexes with non-coordi-
nating counter-ions. The lowest barriers for syn–anti exchange
and apparent rotation are ca. 103 kJ/mol and 113 kJ/mol, respec-
tively. These two processes are independent, with no mechanistic
connection. The apparent rotation is most reminiscent of a bar-
rier-less ligand dissociation followed by an SN2-like ligand substi-
tution, P replacing N or vice versa, at one coordination site in the
square plane.
4.1.2. Cyclohexenol-1-D 438
Cyclohexenone (2.42 ml, 25 mmol) was added to a solution of
CeCl3Á7H2O (9.3 g, 25 mmol) in methanol (20 ml). After cooling to
0 °C, NaBD4 (1.0 g, 25 mmol) was added in portions over a period
of 5 min. The solution was stirred at 0 °C for 20 min. The reaction
was quenched with HCl (1 M, 30 ml), and the product was ex-
tracted with Et2O (3 Â 20 ml), washed with brine (15 ml), then
dried over MgSO4, and yielded the pure product after evaporation
of the solvent, as a clear oil (2.2 g, 90%). 1H NMR (400 MHz, CDCl3):
d (ppm) 5.84 (m, 1H), 5.75 (m, 1H), 1.54–2.19 (m, 6H).
In the presence of a coordinating anion (here modeled by chlo-
ride), dynamic processes are strongly accelerated, sufficient to fully
scramble the regiochemical memory in the (g
3-allyl)Pd complex.
The most likely process for the apparent rotation is an associative
ligand replacement, with transition states highly reminiscent of
Berry pseudorotation, but connecting square planar structures
with loosely associated fifth ligands, not true pentacoordinated
complexes. syn–anti Exchange via square planar (g
1-allyl)Pd com-
4.2. Synthesis of allylic acetates11
4.2.1. Compound 1a
plexes is also a very fast process, but again without a mechanistic
connection to the apparent rotation. Finally, nucleophilic attack of
the leaving group, leading to reversibility of the initial ionization
step, can also be a favored process. Even with the chloride model
Acetyl chloride (570 ll, 8.0 mmol) and pyridine (645 ll, dry,
8.0 mmol) were added in portions to a cooled solution of 3
(0.36 g, 6 mmol) in CH2Cl2 (15 ml, dry) at 0 °C. A white precipitate
formed almost immediately. The reaction mixture was allowed to
reach room temperature during 20 h. The reaction mixture was
cooled again and the precipitate formed was filtered off and
washed with CH2Cl2. The solution was washed with ice-water
(10 ml), ice-cooled brine (15 ml), dried with Na2SO4, and yielded
a yellow oil after careful evaporation. The residue was distilled,
by using a Hickmann apparatus, collecting the product at 101–
104 °C (222 mg, 2.2 mmol, 35%). 1H NMR (400 MHz, CDCl3): d
(ppm) 5.86–5.95 (m, 1H), 5.28–5.35 (m, 1H), 5.23 (m, 1H), 4.59
(m, 2H) absent when D, 2.08 (s, 3H).
used here, the (g
2-chloropropene)Pd0 complex is only 74 kJ/mol
above the most stable PdII complex, a relatively low energy com-
pared to the barriers of >100 kJ/mol for the dynamic processes in
the free, cationic (g
3-allyl)Pd complexes. For carboxylates that
can form strong C–O bonds, the energy difference is expected to
be much smaller, may be even disappearing.35
The reversibility of the initial ionization could also help ratio-
nalize the starting material racemization observed. However, as
evidenced by lack of leaving group exchange in crossover experi-
ments, other pathways must also be active for this process. We
speculate that a PdII-catalyzed Overman rearrangement could
rationalize our observations (Scheme 5).
4.2.2. Compound 2a
4. Experimental
4.1. General
Acetyl chloride (1.10 ml, 15 mmol) and pyridine (1.20 ml, dry,
15 mmol) were added in portions to a cooled solution of 4
(0.99 g, 10 mmol) in dry CH2Cl2 (15 ml), at 0 °C. A white precipitate
formed almost immediately. The reaction mixture was allowed to
reach room temperature during 20 h, and the reaction was then
quenched with NH4Cl (aq satd, 20 ml). The product was extracted
with CH2Cl2 (2 Â 25 ml), washed with aq. saturated NaHCO3
(25 ml), brine (25 ml), and then dried with MgSO4, yielding the
crude product as a yellow oil after evaporation of solvent. The
crude product was purified by column chromatography (EtOAc/
pentane, 1:9), yielding the product as a clear oil (0.85 g, 6.0 mmol,
60%). 1H NMR (400 MHz, CDCl3): d (ppm) 5.95 (m, 1H), 5.68 (m,
1H), 5.24 (m, 1H) absent when D, 1.55–2.10 (m, 9H).
All reactions were performed in oven-dried glassware and un-
der a nitrogen atmosphere, except for the synthesis of cyclohex-
enol. Solvents were distilled prior to use. THF was distilled from
Na/benzophenone, CH2Cl2, and pyridine from CaH2, under a nitro-
gen atmosphere. Et2O was purchased as anhydrous and stored over
molecular sieves (4 Å). Commercially available reagents were used
as delivered unless stated otherwise. TLC was performed using alu-
mina plates coated with Silica Gel 60, F254. The plates were visual-
ized by using a mixture of 2% KMnO4 and 4% K2CO3 (aq) followed

C. Johansson et al. / Tetrahedron: Asymmetry 21 (2010) 1585–1592
1591
4.3. General procedure for the synthesis of allylic benzoates
Pyridine (970 l, 12 mmol) and benzoyl chloride (1.40 ml,
4.4.3. Compound [(g
3-allyl)PdOCOCF3]2 741
At first, Pd(dba)2 (57.5 mg, 0.10 mmol) was dissolved in dry THF
(0.8 ml) and carefully evacuated, then refilled with N2 three times.
Allyl trifluoroacetate (16.95 mg, 0.11 mmol) was dissolved in dry
acetonitrile (0.2 ml) and added to the deep-red Pd(dba)2 solution.
The solution became light green after 5 min., and was stirred at
room temperature for a further 25 min. The mixture was filtered
through a glass filter funnel and the solvents removed in vacuo.
The residue was triturated with 10% acetonitrile in H2O
(8 Â 1 ml). The extract was filtered and evaporation of the solvent
yielded the product as a yellow solid (45 mg, 0.045 mmol, 90%). 1H
NMR (400 MHz, CDCl3): d (ppm) 5.61 (m, 1H), 4.14 (m, 2H), 3.07
(m, 2H).
l
12 mmol) were added in portions to a solution of allylic alcohol
1 or 2 (10 mmol) in dry CH2Cl2 (15 ml) at 0 °C. The solution was al-
lowed to reach room temperature and was stirred for 20 h. The
reaction mixture was filtered through basic alumina into a separa-
tion funnel and washed with H2O (20 ml) and brine (20 ml), then
dried with MgSO4. After evaporation of the solvent, the crude prod-
uct was purified by column chromatography (EtOAc/pentane, 1:4),
yielding the product as a clear oil (1.2 g, 67%, for cyclohexenyl ben-
zoate). 1H NMR (400 MHz, CDCl3): 1b d (ppm) 8.08 (m, 2H), 7.56
(m, 1H), 7.45 (m, 2H), 5.98–6.08 (m, 1H), 5.38–5.98 (m, 1H), 5.29
(m, 1H), 4,84 (m, 2H) absent when D. 2b d (ppm) 8.08 (m, 2H),
7.55 (m, 1H), 7.42 (m, 2H), 6.01 (m, 1H), 5.83 (m, 1H), 5.51 (m,
1H) absent when D, 1.55–2.18 (m, 6H).
4.5. General procedure for the allylic alkylation11
At first, NaH (40 mg, 1 mmol, 60% suspension in oil), was dis-
solved in dry THF (2 ml). Next the flask was evacuated and refilled
with N2. Diethyl methylmalonate (170 ll, 1 mmol) was added
4.4. General procedure for the synthesis of allylic carbonates39
dropwise to the suspension, which turned clear after the addition.
The solution was stirred at room temperature for 10 min.
Pyridine (806
cyclohexenol (0.78 g, 8 mmol) in CH2Cl2 (15 ml). Thereafter methyl
chloroformate (770 l, 10 mmol) was added dropwise, and a white
ll, 10 mmol) was added to a cooled solution of
Either Pd2(dba)3 (11.5 mg, 0.0125 mmol) or
7
(6.5 mg,
l
0.0125 mmol) was dissolved in dry THF (1 ml) and the flask was
carefully evacuated and refilled with N2 three times. Next, 5
(6.2 ll, 0.0250 mmol) was added. The solution was stirred at room
precipitate started to form. The stirred solution was allowed to
reach room temperature during 17 h. H2O (15 ml) was added and
the product was extracted with CH2Cl2 (3 Â 20 ml). The organic
phase was washed with 1 M HCl (2 Â 20 ml), H2O (15 ml), and
brine (15 ml). Thereafter the reaction was dried with MgSO4, and
after evaporation of the solvent, the crude product was purified
by column chromatography (EtOAc/pentane, 1:9), yielding the
product as a clear oil (0.45 g, 2.9 mmol, 36%). 1H NMR (400 MHz,
CDCl3): 1c d (ppm) 5.88–6.00 (m, 1H), d 5.36 (m, 1H), d 5.28 (m,
1H), d 4.63 (m, 1H) absent when D, 3.80 (s, 3H). 2c d (ppm) 5.97
(m, 1H), 5.75 (m, 1H), 5.11 (m, 1H) absent when D, 3.76 (m, 3H),
1.56–2.12 (m, 6H).
temperature for 10 min. Allylic substrate (0.50 mmol) was dis-
solved in dry THF (1 ml) and added to the Pd-solution, which
turned green after a few minutes.
Alternatively, 6 (0.01 mmol, 5 mg) in CH2Cl2 (0.1 ml) was added
to the allylic substrate (0.50 mmol) in THF (2 ml).
Thereafter the malonate solution was added to the mixture con-
taining allylic substrate and Pd catalyst and was stirred at room
temperature for 20 h. The reaction mixture was diluted with Et2O
(5 ml), and washed with 1 M HCl (5 ml), H2O (5 ml), and brine
(5 ml). Thereafter it was dried with Na2SO4 and the product mix-
ture was obtained as a yellow oil, after evaporation of solvents.
The crude mixture was analyzed without further purification. 1H
NMR (400 MHz, CDCl3): 8 d (ppm) 5.69 (m, 1H), 5.10 (m, 2H),
4.19 (q, 4H), 2.60 (m, 2H), 1.24 (t, 6H). 9 d (ppm) 5.78 (m, 1H),
5.54 (m, 1H), 4.19 (q, 4H), 3.05 (m, 1H), 1.50–2.20 (m, 6H), 1.32
(s, 3H), 1.45 (t, 6H).
4.4.1. 2-(Diphenylphosphino)-N,N-dimethylethaneamine 540
Potassium tert-butoxide (2.47 g, 22 mmol) was dissolved in dry
THF (30 ml) and the solution was flushed with N2. Diphenylphos-
phine (1.73 ml, 10 mmol) was added and the orange-red solution
was stirred at room temperature for 30 min. Then 2-chloro-N,N-
dimethylethylamine hydrochloride (1.44 g, 10 mmol) was added
in one portion and the reaction mixture was refluxed for 20 h.
The solvent was evaporated and 10% HCl (25 ml) was added. The
acidic phase was washed with Et2O (25 ml), and then made alka-
line with 5 M NaOH. The product was extracted with Et2O
(3 Â 25 ml), the organic phase was washed with brine (20 ml),
dried with Na2SO4, and the crude product was obtained as a cloudy
oil after evaporation. The oil was dissolved in a small amount of
Et2O and filtered through a column of basic alumina, yielding the
product as a clear oil (0.73 g, 2.8 mmol, 28%). 1H NMR (400 MHz,
CDCl3): d (ppm) 7.40–7.48 (m, 4H), 7.29–7.38 (m, 6H), 2.36–246
(m, 2H), 2.21–2.28 (m, 8H).
5. Computational details
All DFT calculations were performed in the Jaguar program–
using the B3LYP functional42–44 in conjunction with the LACVP basis
*
set.45 Solvent effects were represented by the PBF continuum mod-
el46,47 with default parameters for THF (solvent = tetrahydrofuran).
Vibrational contributions to the free energy were calculated in the
gas phase. No standard state corrections have been applied; energies
are appropriate for 1 M concentration of all solvated species. Each
order of magnitude reduction in concentration corresponds to a free
energy increase of ca. 6 kJ/mol at room temperature.
Transition states for attack of nucleophiles on cationic (g
3-al-
4.4.2. Compound [(
g
3-allyl)PdÁ5]BF4 63
lyl)Pd complexes should be located for the solvated species.48
However, we have found vibrational contributions to be unreliable
for structures optimized with continuum solvation. In addition, the
potential energy surface for nucleophilic attack on Pd-allyls in sol-
vent is extremely flat, making the location of true transition states
very challenging. We have therefore followed an earlier proce-
dure22,49 and located approximate transition states by relaxed
driving of the coordinate for the forming bond, with full optimiza-
tion in solvent at each scan point. The full experimental ligand was
At first, [(
g
3-allyl)PdCl]2 (18.3 mg, 0.05 mmol) in CH2Cl2 (1 ml)
was added to AgBF4 (20.4 mg, 0.105 mmol), under N2. A white pre-
cipitate was formed almost immediately. The solution was stirred
at 20 °C for 15 min. Next, 5 (25.7 mg, 0.10 mmol) was added,
whereupon the solution turned brown and was stirred for 1 h.
The reaction mixture was filtered through a 2 cm column of Celite
in a Pasteur pipette and thereafter the solvent was removed under
reduced pressure, yielding the crude product as a beige crystalline
solid (44.5 mg, 0.91 mmol, 91%). 1H NMR (400 MHz, CDCl3): d
(ppm) 7.50–7.70 (m, 10H), 5.85 (m, 1H), 4.84 (m, 1H), 4.02 (m,
1H), 3.90 (m, 1H), 2.98–3.11 (m, 5H).
–
Jaguar v. 7.6 from Schrödinger, Inc., see: http://www.schrodinger.com.

1592
C. Johansson et al. / Tetrahedron: Asymmetry 21 (2010) 1585–1592
12. Vrieze, K. In Dynamic Nuclear Magnetic Resonance Spectroscopy; Jackman, L. M.,
Cotton, F. A., Eds.; Academic Press: New York, 1975.
employed, and the nucleophile was represented by dimethyl mal-
onate ion chelating a sodium atom, which was further solvated by
two dimethyl ether molecules as THF models.50 The energies in
solvent were compared directly, with the assumption that vibra-
tional contributions and systematic errors in the DFT functional
cancel for the very similar structures used in the comparison.22
13. Krämer, K.; Kazmaier, U. J. Org. Chem. 2006, 71, 8950–8953.
14. Faller, J. W.; Sarantopoulos, N. Organometallics 2004, 23, 2179–2185.
15. Granberg, K. L.; Bäckvall, J.-E. J. Am. Chem. Soc. 1992, 114, 6858–6863.
16. Martin, J. T.; Oslob, J. D.; Åkermark, B.; Norrby, P.-O. Acta Chem. Scand. 1995, 49,
888–893.
17. Strand, D.; Norrby, P.-O.; Rein, T. J. Org. Chem. 2006, 71, 1879–1891.
18. Pedersen, T. M.; Hansen, E. L.; Kane, J.; Rein, T.; Helquist, P.; Norrby, P.-O.;
Tanner, D. J. Am. Chem. Soc. 2001, 123, 9738–9743.
19. Gogoll, A.; Örnebro, J.; Grennberg, H.; Bäckvall, J.-E. J. Am. Chem. Soc. 1994, 116,
3631–3632.
20. Hansson, S.; Norrby, P.-O.; Sjögren, M.; Åkermark, B.; Cucciolito, M. E.;
Giordano, F.; Vitagliano, A. Organometallics 1993, 12, 4940–4948.
21. Hayashi, T.; Yamamoto, A.; Ito, Y.; Nishioka, E.; Miura, H.; Yanagi, K. J. Am.
Chem. Soc. 1989, 111, 6301–6311.
22. Butts, C. P.; Filali, E.; Lloyd-Jones, G. C.; Norrby, P.-O.; Sale, D. A.; Schramm, Y. J.
Am. Chem. Soc. 2009, 131, 9945–9957.
23. von Matt, P.; Pfaltz, A. Angew. Chem., Int. Ed. Engl. 1993, 32, 566–569.
24. Sprinz, J.; Helmchen, G. Tetrahedron Lett. 1993, 34, 1769–1773.
25. Allen, J. V.; Coote, S. J.; Dawson, G. J.; Frost, C. G.; Martin, C. J.; Williams, J. M. J. J.
Chem. Soc., Perkin Trans. 1 1994, 2065–2072.
26. Goldfuss, B.; Kazmeier, U. Tetrahedron 2000, 56, 6493–6496.
27. Svensen, N.; Fristrup, P.; Tanner, D.; Norrby, P.-O. Adv. Synth. Catal. 2007, 349,
2631–2640.
28. (a) Alcock, N. W.; Brown, J. M.; Hulmes, D. I. Tetrahedron: Asymmetry 1993, 4,
743–756; (b) Helmchen, G.; Pfaltz, A. Acc. Chem. Res. 2000, 33, 336–345.
29. Amatore, C.; Jutand, A. Coord. Chem. Rev. 1998, 178–180, 511–528.
30. Amatore, C.; Jutand, A.; Meyer, G.; Mottier, L. Chem. Eur. J. 1999, 5, 466–473.
31. Amatore, C.; Gamez, S.; Jutand, A.; Meyer, G.; Moreno-Mañas, M.; Morral, L.;
Pleixats, R. Chem. Eur. J. 2000, 6, 3372–3376.
Dynamic processes in (g
3-allyl)Pd complexes were studied
using a smaller model system, with H2P-CH2CH2-NH2 used as a
model for the real ligand. This approach will underestimate steric
repulsion in the system, but on the other hand, B3LYP is known
to overestimate non-bonded repulsive interactions due to the lack
of an appropriate treatment of dispersive interactions. The effect of
an added chloride was investigated. Chloride is present in small
amounts in many catalytic systems and is known to be a highly
efficient catalyst for dynamic processes in (g
3-allyl)Pd com-
plexes,20 but could also be considered a simplified model for other
anions like carboxylates that are used as leaving groups and there-
fore always present. For these complexes, stationary points were
determined in gas phase and validated by frequency calculation.
Energies in solvent were determined by single point continuum
calculations at the optimized geometries. Final free energies were
obtained by adding the vibrational contributions (including zero
point energies) to the calculated free energies in solvent. This ap-
proach is known to exaggerate the entropic contributions and thus
favors dissociated species, just like the dispersion error in DFT, but
this effect will to some extent be compensated by the choice of
small model system.
32. Moreno-Mañas, M.; Morral, L.; Pleixats, R. J. Org. Chem. 1998, 63, 6160–6166.
33. Overman, L. E.; Knoll, F. M. Tetrahedron Lett. 1979, 4, 321–324.
34. Weinhold, F.; Landis, C. Valency and Bonding: A Natural Bond Orbital Donor–
Acceptor Perspective; Cambridge University Press, 2005.
35. Evans, L. A.; Fey, N.; Harvey, J. N.; Hose, D.; Lloyd-Jones, G. C.; Murray, P.;
Orpen, A. G.; Osborne, R.; Owen-Smith, G. J. J.; Purdie, M. J. Am. Chem. Soc. 2008,
130, 14471–14473.
36. Cantat, T.; Gélin, E.; Giroud, C.; Meyer, G.; Jutand, A. J. Organomet. Chem. 2003,
687, 365–376.
37. Schuetz, R. D.; Millard, W. J. Org. Chem. 1959, 24, 297–300.
38. Luche, J. L. J. Am. Chem. Soc. 1978, 100, 2226–2227.
Acknowledgments
Support from the Swedish Research Council is gratefully
acknowledged. G.C.L.-J. thanks the Royal Society for a Wolfson Re-
search Merit Award.
´
39. Fairlamb, I. J. S.; Lloyd-Jones, G. C.; Vyskocil, S.; Kocovsky, P. Chem. Eur. J. 2002,
8, 4443–4453.
40. Habtemariam, A.; Watchman, B.; Potter, B. S.; Palmer, R.; Parsons, S.; Parkin, A.;
Sadler, P. J. J. Chem. Soc., Dalton Trans. 2001, 1306–1318.
41. Vitagliano, A.; Åkermark, B.; Hansson, S. Organometallics 1991, 10, 2592–2599.
42. Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
43. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789.
44. Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994,
98, 11623–11627.
45. The LACVP basis set uses 6-31G for the lighter elements, and the Hay-Wadt
small-core ECP with accompanying basis set for Pd: Hay, P. J.; Wadt, W. R. J.
Chem. Phys. 1985, 82, 299–310.
46. Tannor, D. J.; Marten, B.; Murphy, R.; Friesner, R. A.; Sitkoff, D.; Nicholls, A.;
Ringnalda, M.; Goddard, W. A., III; Honig, B. J. Am. Chem. Soc. 1994, 116, 11875–
11882.
47. Marten, B.; Kim, K.; Cortis, C.; Friesner, R. A.; Murphy, R. B.; Ringnalda, M. N.;
Sitkoff, D.; Honig, B. J. Phys. Chem. 1996, 100, 11775–11788.
48. Hagelin, H.; Åkermark, B.; Norrby, P.-O. Chem. Eur. J. 1999, 5, 902–909.
49. Fristrup, P.; Ahlquist, M.; Tanner, D.; Norrby, P.-O. J. Phys. Chem. 2008, 112,
12862–12867.
References
1. Tsuji, J. Palladium Reagents and Catalysts: Innovations in Organic Synthesis;
Wiley: Chichester, UK, 1997.
2. Åkermark, B.; Hansson, S.; Vitagliano, A. J. Am. Chem. Soc. 1990, 112, 4587–
4588.
3. Sjögren, M.; Hansson, S.; Norrby, P.-O.; Åkermark, B.; Cucciolito, M. E.;
Vitagliano, A. Organometallics 1992, 11, 3954–3964.
4. Trost, B. M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921–2943.
5. Pfaltz, A.; Lautens, M. In Comprehensive Asymmetric Catalysis; Jacobsen, E. N.,
Pfaltz, A., Yamamoto, H., Eds.; Springer: Heidelberg, 1999; pp 833–886.
6. Trost, B. M.; Lee, C. In Catalytic Asymmetric Synthesis; Ojima, I., Ed., 2nd ed.;
Wiley-VCH: New York, 2000; pp 593–649.
7. Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96, 395–422.
8. Fiaud, J. C.; Malleron, J. L. Tetrahedron Lett. 1981, 22, 1399–1402.
9. Trost, B. M.; Bunt, R. C. J. Am. Chem. Soc. 1996, 118, 235–236.
10. Lloyd-Jones, G. C.; Stephen, S. C. Chem. Eur. J. 1998, 4, 2539–2549.
11. Fristrup, P.; Jensen, T.; Hoppe, J.; Norrby, P.-O. Chem. Eur. J. 2006, 12, 5352–
5360.
*
*
50. Kleimark, J.; Hedström, A.; Larsson, P.-F.; Johansson, C.; Norrby, P.-O.
ChemCatChem 2009, 1, 152–161.