How does palladium coordination affect the electrophilicities of allyl cations? Development of a robust kinetic method for following reactions of [(η3-diarylallyl)Pd(PPh3)2]+ with nucleophiles

Article abstract of DOI:10.1021/om3000357

Kinetics of the reactions of [(η³-1,3-diarylallyl) Pd(PPh₃)₂]⁺ complexes with carbanions, enamines, amines, and triphenylphosphine have been investigated photometrically in dichloromethane, DMSO, and acetonitril

Full text of DOI:10.1021/om3000357

Article
pubs.acs.org/Organometallics
How Does Palladium Coordination Affect the Electrophilicities
of Allyl Cations? Development of a Robust Kinetic Method for
Following Reactions of [(η3-Diarylallyl)Pd(PPh3)2]+ with Nucleophiles
Konstantin Troshin, Peter Mayer, and Herbert Mayr*
Department Chemie, Ludwig-Maximilians-Universitat Munchen, Butenandtstraße 5-13 (Haus F), 81377 Munchen, Germany
̈
̈
̈
S
* Supporting Information
ABSTRACT: Kinetics of the reactions of [(η3-1,3-diarylallyl)-
Pd(PPh3)2]+ complexes with carbanions, enamines, amines,
and triphenylphosphine have been investigated photometri-
cally in dichloromethane, DMSO, and acetonitrile solutions
at 20 °C. Amines were found to react both at palladium
(substitution of PPh3) and at the allyl ligands, causing complex
and poorly reproducible kinetics. Excellent reproducibility of
the second-order rate constants for the attack at the allyl ligand
was achieved when the reactions were studied in the presence
of traces of PPh3 and fumaronitrile. The second-order rate
constants (k2) for the attack of nucleophiles at the allyl ligands were found to follow the correlation log k2 = sN(N + E), where sN
and N are solvent-dependent nucleophile-specific parameters and E is an electrophilicity parameter (J. Am. Chem. Soc. 2001, 123,
9500−9512). While the electrophilicities of the free 1,3-bis(3,5-difluorophenyl)allyl (E = 6.11) and the 1,3-bis(4-
dimethylaminophenyl)allyl cations (E = −7.50) differ by more than 13 orders of magnitude, the electrophilicities of the
corresponding Pd(PPh3)2 complexes were found to be almost independent of the nature of the substituents (E ≈ −14), showing
that Pd(PPh3)2 coordination reduces the electrophilic reactivities of allyl cations by 7−20 orders of magnitude.
of chloride ions, neutral (η1-allyl)palladium species have also
INTRODUCTION
■
been characterized as reactive intermediates.2
The Tsuji−Trost reaction, which often proceeds under mild
conditions with high regio- and stereoselectivity, is a powerful
synthetic tool for asymmetric allylations of various types of
nucleophiles.1 These reactions have been demonstrated to
proceed generally via the cationic (η3-allyl)palladium complexes
(1) (Scheme 1). In certain cases, for example, in the presence
Detailed mechanistic studies on the influence of leaving
groups, ligands, and solvents on the formation of the
allylpalladium species have been reported.3 Investigations of
the reactions of isolated (η3-allyl)palladium complexes 1 with
nucleophiles have shown that the reaction mechanism strongly
depends on the nature of the ligands. Jutand et al. reported the
reactions of [(η3-CH2CHCH2)Pd(PAr3)2]+ (Ar = 4-ClC6H4,
4-MeC6H4)4 and [(η3-PhCHCHCHPh)PdL2]+ (L = PPh3 or
L2 = dppb)5 with morpholine and piperidine to proceed via
selective attack of the amines at the allyl ligands, in agreement
with earlier work by Kuhn and Mayr.6 On the other hand,
Crociani et al.7 and Canovese et al.8 observed that nucleophilic
attack at the allyl ligand may be accompanied by reversible
coordination of the amines to the metal center. Thus, the
formation of pentacoordinated palladium species was found
when iminophosphine ligands were employed,7 while terden-
tate (S−N−S), (N−S−N), and (N−N−N) ligands and
α-diimines were reversibly replaced by amines.8
Scheme 1. Typical Catalytic Cycle for Tsuji−Trost
Allylations
In recent work9 we have studied the kinetics of the reactions
of noncoordinated 1,3-diarylallyl cations with nucleophiles and
reported that the second-order rate constants can be described
by the linear free energy relationship (eq 1), where E is an
Received: January 17, 2012
Published: March 2, 2012
© 2012 American Chemical Society
2416
dx.doi.org/10.1021/om3000357 | Organometallics 2012, 31, 2416−2424

Organometallics
Article
electrophilicity parameter, N is a nucleophilicity parameter, and sN is
a nucleophile-specific sensitivity parameter (previously termed s).10
As shown in Chart 1, their electrophilic reactivities differ by
characterized by X-ray diffraction analysis (details, see below).
Compound 1c-BF4 was synthesized by the reaction of (E)-1,
3-bis(4-dimethylaminophenyl)allylium tetrafluoroborate (5c-BF4)
with Pd(PPh3)4 in dichloromethane at ambient temperature.
Kinetic Investigations. All kinetic measurements were per-
formed under pseudo-first-order conditions using the nucleo-
philes in high excess; the reaction progress was monitored UV−
vis spectroscopically at the absorbance maxima of the allyl com-
plexes 1(a−c). Freshly prepared solutions of electrophiles were
generally required to obtain reproducible results.
Chart 1. Cationic Palladium Complexes 1(a−c) and the
Corresponding Free 1,3-Diarylallyl Cations 5(a−c) with
Their Electrophilicity Parameters E (from ref 9)
Reactions with C-Nucleophiles. Even when the enamines
2a and 2b and the potassium salt of Meldrum’s acid (2d) were
used in high excess (10−100 equiv) over the Pd complex 1a-
BF4 in order to achieve pseudo-first-order kinetics, deviations
from the monoexponential decay of the absorbance at 360 nm
(absorption maximum of 1a) were often observed. These
deviations were explained by the UV absorptions of the η2-Pd
complexes 3 (Scheme 1) and their decomposition products. As
fumaronitrile is a weak nucleophile, which does not react with
1a and has previously been shown to strongly coordinate to
PdL2,14 the kinetics of the reactions of 1a-BF4 with nucleophiles
were investigated in the presence of fumaronitrile. Under these
conditions, the decays of the absorbances at 360 nm were
generally monoexponential, when 1a-BF4 was combined with
more than 10 equiv of various nucleophiles, because the
released Pd fragments form stable complexes with fumaroni-
trile, which do not significantly absorb at 360 nm. The pseudo-
first-order rate constants (kobs), which were obtained by least-
more than 13 orders of magnitude, corresponding to relative
reaction times of 1 s vs 105 to 106 years.
log k (20 °C) = s (N + E)
(1)
2
N
We have now investigated the kinetics of the reactions of the
corresponding [(η3-(1,3-diarylallyl)Pd(PPh3)2]+ complexes
1(a−c) (Chart 1) with a variety of nucleophiles (Table 1) in
Table 1. Nucleophiles 2(a−i) and Their Reactivity
Parameters N and sN in Different Solvents
obst
squares fitting to the function At = A0e−k + C (Figure 1a),
a
Reactivity parameters from ref 12.
order to elucidate how Pd(PPh3)2 coordination affects the elec-
trophilic reactivities of 1,3-diarylallyl cations. While products of
the reactions of (η3-allyl)palladium complexes with most of
these nucleophiles have previously been reported,11 there have
been only a few kinetic investigations.
Figure 1. (a) Exponential decay of the absorbance At (λ = 360 nm) and
(b) linear correlation between the pseudo-first-order rate constants kobs
and the nucleophile concentration for the reaction between the enamine
2b and 1a-BF4 ((6.2−6.3) × 10−5 M) in the presence of fumaronitrile
((5.7−5.8) × 10−4 M) in dichloromethane at 20 °C.
RESULTS AND DISCUSSION
■
Synthesis of the Complexes 1(a−c). The cationic Pd
complexes 1a and 1b were synthesized and isolated as tetra-
fluoroborate salts (1a-BF4 and 1b-BF4) by treatment of the
corresponding dimeric allylpalladium chlorides (5(a,b)-PdCl) with
PPh3 and NaBF4, as described in ref 13. The triflate 1b-OTf was
obtained analogously by using NaOTf instead of NaBF4 and
were found to be independent of the concentration of
fumaronitrile. As shown in Figure 1b for the reaction of 1a-
BF4 with the enamine 2b, plots of the pseudo-first-order rate
constants vs the concentrations of the nucleophiles were linear
with the second-order rate constants (k2) as slopes.
2417
dx.doi.org/10.1021/om3000357 | Organometallics 2012, 31, 2416−2424

Organometallics
Article
Reactions with Triphenylphosphine. The reaction of
triphenylphosphine (2e) with 1a-BF4 in CH3CN yielded a
2:1 mixture of (E)-(1,3-diphenylallyl)triphenylphosphonium
tetrafluoroborate (6) and (Z)-(1,3-diphenylprop-1-en-1-yl)-
triphenylphosphonium tetrafluoroborate (6′) in an overall yield
of 54% (Scheme 2).
40% of the absorbance at 360 nm decayed by a very fast process,
which was followed by a slower secondary reaction; the decay due
to the first process increased with increasing concentration of the
amine. Fitting the decays for the secondary reactions to the
exponential function At = A0e−k t + C yielded the kobs values, which
obs
are plotted against the amine concentrations (Figure 2b). Though
the kobs values are poorly reproducible, it is obvious that at low
concentrations of amines these values increase with increasing
amine concentrations, but do not exceed 7 × 10−3 s−1 even at high
concentrations of amines, indicating changes in the reaction
mechanism.
Scheme 2
A rationalization for these observations can be derived from the
UV spectra described in Figure 3. When a solution of 1a-BF4 in
As an independently synthesized sample of the allylphos-
phonium salt 6 was observed to isomerize to a 4:1 mixture15 of
6 and 6′ in the presence of 4 mol % of Pd(PPh3)4 in acetonitrile
at rt, we assume that the reaction of 1a-BF4 with PPh3 proceeds
via nucleophilic attack of triphenylphosphine at the allyl ligand
to yield 6, which subsequently undergoes a partial Pd-catalyzed
isomerization with formation of a mixture of 6 and 6′.
The decay of the 360 nm absorbance during the reactions
of 1a-BF4 with PPh3 in dichloromethane, acetonitrile, and
DMSO was found to be monoexponential, and the linear
kobs vs [PPh3] plots yielded the second-order rate constants
Figure 3. UV−vis evidence for the reversible ligand exchange in the
reaction of 1a-BF4 with benzylamine (2g).
k2PPh (CH2Cl2) = 0.366 M−1 s−1, k2PPh (CH3CN) = 0.695 M−1 s−1,
3
3
and k2PPh (DMSO)= 0.232 M−1 s−1 at 20 °C (for details see the
3
acetonitrile (black dotted curve) is combined with an excess of
benzylamine (2g), an immediate change of the UV−vis spectrum is
observed (red dashed curve). Addition of trifluoroacetic acid imme-
diately after recording the red curve restored the initial absorption
spectrum (blue curve, slightly lower than the initial black curve
because of dilution), which remained stable at room temperature.
We, therefore, assume that at high concentrations of amines a
rapid reversible exchange of Ph3P by amines yields the complex 1a′-
BF4, which reverts to 1a-BF4 when treated with CF3COOH
(Scheme 3). An analogous ligand exchange has previously been
Supporting Information).
Reactions with Amines. Even when fumaronitrile was
present, the reactions of the allylpalladium complex 1a-BF4 with
excess amines often did not follow a first-order rate law. As shown
for the reaction of 1a-BF4 with benzylamine (2g) in Figure 2a, up to
Scheme 3. Reversible Exchange of PPh3 in 1a by Amines
reported by Canovese et al.8 As the amine complexes 1a′ can be
assumed to be significantly less electrophilic than 1a,16 the limitation
of kobs at high amine concentrations (Figure 2) can be explained.
Regeneration of 1a-BF4 by addition of CF3COOH (as shown
in Figure 3) occurred only when CF3COOH was added imme-
diately after the formation of 1a′-BF4. Acidification 15 min after the
addition of benzylamine did not restore the original spectrum
(Supplementary Figure S4 of the Supporting Information),
indicating the occurrence of an irreversible process, i.e., attack of
the amine at the allyl ligand of 1a or 1a′.
The formation of 1a′ by ligand exchange, as depicted in
Scheme 3, has also been confirmed by a cation-mode ESI mass
spectrum17 of the reaction mixture recorded immediately after
addition of benzylamine to the acetonitrile solution of 1a-BF4
(Figure 4).
On the basis of these observations, the reactions of the
[(η3-allyl)Pd(PPh3)2]+ complexes with amines can be described
by the mechanism depicted in Scheme 4.
Figure 2. Complex behavior of a mixture of the allylpalladium complex
1a-BF4 and benzylamine (2g) in acetonitrile at 20 °C. (a) Typical
decay of the absorbance at λ = 360 nm ([1a-BF4] = 1.0 × 10−4 M,
[2g] = 1.5 × 10−2 M, black line). kobs values were achieved by fitting
the slow parts of the absorbance decays to the monoexponential
obst
function At = A0e−k + C (red dashed line, the fit range is shown in
bold). (b) Dependence of kobs on the amine concentration.
2418
dx.doi.org/10.1021/om3000357 | Organometallics 2012, 31, 2416−2424

Organometallics
Article
Scheme 4, pp S3, S4 of the Supporting Information), addition
of a small amount of PPh3 to a mixture of 1a and 1a′ restored
the original spectrum of 1a (Supplementary Figure S5 of the
Supporting Information). As a consequence, the presence of a
small amount of PPh3 in a mixture of 1a, 1a′, and amines shifts
the 1a ⇌ 1a′ equilibrium toward 1a, and the nonexponential
decay of the absorbance at 360 nm observed in the absence of
PPh3 (Figure 5a) is replaced by the monoexponential decay
Figure 4. Cation-mode ESI mass spectrum recorded immediately after
addition of benzylamine (12 mM) to a 0.2 mM solution of 1a-BF4 in
acetonitrile. Insets: comparison between experimental (black) and
calculated (red) isotopic patterns for peaks at m/z = 668 and 823. R =
(η3-(1,3-diphenylallyl)).
Scheme 4. Mechanism for Amination of Allylpalladium
Complexes
Figure 5. Decays of the absorbances at 360 nm during the reactions of
1a-BF4 with benzylamine 2g (7.5 × 10−2 M) in the presence of
fumaronitrile (5.8 × 10−4 M): (a) without PPh3; (b) in the presence of
1.4 × 10−3 M PPh3 (acetonitrile, 20 °C).
The amine attacks the palladium complex 1a either fast and
reversibly at Pd to give 1a′ or slowly and irreversibly at the allyl
ligand to give 3a. As the attack of amines at the allyl ligand of
1a′ is significantly slower than that at 1a, this reaction is
neglected in Scheme 4. In the case of low amine concentrations,
the complex 1a is the predominant species in the equilibrium
1a ⇌ 1a′, and the overall reaction rate can be described by eq 2
(n = 1).
shown in Figure 5b. Analogous behavior was found for re-
actions of diethylamine with allylpalladium complexes contain-
ing diimine ligands.8b
As discussed above for enamines, all reactions were studied
in the presence of fumaronitrile in order to coordinate Pd
fragments that are released during these reactions (the whole
reaction can then be described with Scheme 4, where L is
fumaronitrile). Since the observed rate constants were found to
be independent of the fumaronitrile concentration (Supple-
mentary Figure S6 of the Supporting Information), one can
conclude that fumaronitrile is not involved in the rate-
determining step.
As triphenylphosphine may also add to the allyl ligand of 1
(vide supra Reactions with Triphenylphosphine), the rate
constants (kobs) measured for reactions of 1 with amines in the
presence of PPh3 are the sum of the rate constants for the
reactions of 1 with amines and PPh3 (eq 3).
d[1a]
dt
n
−
= k [1a][amine]
2
(2)
At high amine concentrations, [1a′] becomes comparable to
[1a], and because of the low reactivity of 1a′, the reaction order
n in amine concentration is lowered to 0 ≤ n ≤ 1. Apart from
poor reproducibility, a further complication for evaluating
kinetics under these conditions is the different absorption co-
efficients of 1a and 1a′ at 360 nm. These problems stimulated us
to search for a more efficient method of determining the rates of
attack of amines at the allyl ligands of [(η3-allyl)Pd(PPh3)2]+
complexes.
PPh
2
amine
3
k
= k
1ψ
+ k
[PPh ]
3
obs
(3)
Four-Component-System for Determining the Rates
of the Reactions of [(η3-allyl)Pd(PPh3)2]+ with Amines.
The exchange of PPh3 by N ligands upon combination of 1a
with amines was recognized by the change of the UV spectrum
of 1a (black dashed graph vs red graph in Figure 3). In line with
the small value of the equilibrium constant KE (≈ 4 × 10−3,
As illustrated by the open symbols in Figure 6, the rate con-
stants kobs determined at a constant amine concentration
increased with increasing [PPh3]. While the strong increase of
kobs in the left part of Figure 6 is due to the shift of the
equilibrium 1a ⇌ 1a′ toward the more electrophilic complex 1a,
2419
dx.doi.org/10.1021/om3000357 | Organometallics 2012, 31, 2416−2424

Organometallics
Article
responsible for the poor reproducibility of the kinetic data
obtained in the absence of triphenylphosphine, is indicated by
the change of the UV−vis spectra of solutions of 1a-BF4 in
acetonitrile within a few hours. Addition of 3 equiv of tri-
phenylphosphine to such aged solutions fully regenerates the
original UV−vis spectrum of 1a and, therefore, makes reliable
kinetic measurements possible.
As the phosphonium ions 6 and 6′ do not react with
benzylamine in the absence of Pd complexes, we can exclude
that (E)-N-benzyl-1,3-diphenylprop-2-en-1-amine (7), which is
the only isolable product of the reaction of 1a-BF4 with ben-
zylamine in the presence of PPh3 and fumaronitrile, is produced
via initial attack of PPh3 at the allyl ligand of 1a and subsequent
noncatalyzed reaction of 6/6′ with the amine. Since the phos-
phonium salts 6 and 6′ do react with benzylamine in the
presence of palladium complexes, 6 and 6′ are not among the
reaction products, even when high concentrations of PPh3 are
present in the reaction of 1a-BF4 with benzylamine.
Figure 6. Plot of kobs vs [PPh3] for the reaction between 1a (1.1 ×
10−4 M) and benzylamine 2g (7.52 × 10−2 M) in the presence of
fumaronitrile (5.82 × 10−4 M) and variable concentrations of PPh3
(acetonitrile, 20 °C).
the linear increase of kobs in the right part of Figure 6 reflects the
attack of PPh3 at the allyl ligand of 1a, which was independently
determined (see above and in the Supporting Information). As
illustrated by the lower plot in Figure 6, subtraction of the
contribution of phosphine from kobs yields a plateau for [PPh3] >
1.5 mM, where ka1mψpinlaeteaureflects the attack of benzylamine at the
allyl ligand of complex 1a.
Addition of triphenylphosphine has an even stronger effect
on the reaction of 1a-BF4 with piperidine (2i) than on the re-
action with benzylamine discussed above. Supplementary
Figure S7 of the Supporting Information as well as Figure 8
The value of k2amine can therefore be derived from the slope of
a plot of kamine
1ψplateauvs [amine] (Figure 7).
Figure 8. Correlation between the first-order rate constants for the
decay of the absorbance at λ = 360 nm (absorption maximum of 1a)
during the reaction of 1a ((4.5−7.9) × 10−4 M) with piperidine (2i)
in the absence (open symbols) and presence of PPh3 (3.7 × 10−4 to
9.70 × 10−3 M, filled symbols) and the amine concentration
(acetonitrile, 20 °C).
Figure 7. Correlation between the first-order rate constants for the
decays of the absorbance at λ = 360 nm (absorption maximum of 1a)
during the reaction of 1a ((1.4−1.5) × 10−4 M) with benzylamine
(2g) in the absence (open symbols) and presence of PPh3 (3.2 × 10−4
to 5.75 × 10−3 M, filled symbols) and the amine concentration
(acetonitrile, 20 °C).
illustrate that the problem of low reproducibility of the kinetics
in the absence of PPh3 (Figure 8, open symbols) is solved when
the reactions are studied in the presence of triphenylphosphine
(Figure 8, filled symbols).
amine
Figure 7 clearly shows that k
correlates linearly with
1ψplateau
Similar results were found for the reactions of the complexes
1(a−c) with a variety of other amines. Only in rare cases were
deviations from the second-order rate law, i.e., first-order with
respect to 1 and first-order with respect to nucleophile 2,
observed when PPh3 was present. Thus, for reasons that
presently are not exactly known,18 the reaction of morpholine
(2h) with 1a-BF4 was found to be of mixed first- and second-
order in amine concentration.
Determination of Electrophilicity Parameters E for
1(a−c). Using the four-component system developed above
(additives of PPh3 and fumaronitrile to mixtures of 1 and 2),
rate constants for the reactions of the allylpalladium complexes
1(a−c) with a variety of C-, N-, and P-nucleophiles have been
determined (Table 2).
the concentration of amines, also in the concentration range
in which the poorly reproducible values of kobs measured
in the absence of PPh3 do not increase with increasing
[amine].
It is noteworthy that completely different batches of all
components were used for the kinetic measurements shown in
Figure 7, which indicates excellent reproducibility of the
individual rate constants kamine
1ψplateau and contrasts the scatter of
kobs determined in the absence of PPh3. Obviously, side
reactions, which have different weight with different batches of
reagents and are responsible for the scatter of the rate constants
determined in the absence of triphenylphosphine, can
completely be suppressed using the four-component kinetic
method. When PPh3 was present, reproducible kinetic data
could also be obtained with solutions of 1a-BF4 that have been
stored for several hours. Decomposition of 1a-BF4, which is
Figure 9 shows a plot of (log k2/sN) vs N (for the N and sN
values see Table 1) for the reactions of complex 1a with
nucleophiles 2. As required by eq 1, the slope of 1.0 was enforced
2420
dx.doi.org/10.1021/om3000357 | Organometallics 2012, 31, 2416−2424

Organometallics
Article
Table 2. Second-Order Rate Constants (in M−1 s−1) for the
Reactions of (η3-Diarylallyl)palladium Complexes 1(a−c)
with the Nucleophiles 2(a−i) at 20 °C (solvents are specified
in Table 1) and the Resulting Electrophilicity Parameters E
a
a
complex
E
Nu
k2
kcalc
k2/kcalc
1a
−14.14
2a
2b
2d
2e
2f
1.99 × 10−1
2.51 × 10−1
5.66
6.32 × 10−1
1.33
0.79
1.44
2.36
0.28
0.62
1.00
0.82
1.35
1.05
4.71
0.46
1.00
1.54
0.84
1.25
1.01
0.72
8.15
bc
,
1.49
c
3.66 × 10−1
d
4.52
7.30
d
2g
2h
2i
1.26
1.26
cde
, ,
1.07 × 10
2.05 × 102
1.31 × 10
1.52 × 102
2.51 × 10−1
6.32 × 10−1
7.30
d
1b
1c
−14.21
−14.46
2a
2d
2f
2.63 × 10−1
bc
,
Figure 9. Plot of (log k2)/sN vs N for the reactions of complex 1a with
the nucleophiles 2(a−i) (solvents are specified in Table 1). Open
points correspond to the rate constants that were not used for the
determination of E.
2.98
3.39
1.13
d
d
2g
2i
1.13
d
2.10 × 102
1.37 × 102
2.44
2c
2f
2.05
Comparison of the Allylpalladium Complexes 1(a−c)
with the Free Diarylallyl Cations 5(a−c). The structural
data listed in Table 3, which are based on X-ray diffraction anal-
ysis or quantum chemical calculations, show significant differ-
ences between the palladium complexes 1a and 1b and the
corresponding free cations 5a and 5b. While the bond lengths
in the allylic fragments of the complexes 1(a,b) and the free
cations 5(a,b) are almost identical, the bonds between the
terminal allyl and the aromatic carbons are significantly longer
in the complexes 1(a,b) than in the free cations. Conjugative
stabilization of the positive charge by the aryl rings is thus in-
dicated to be much more important in the free diarylallyl
cations 5(a,b) than in the palladium complexes 1(a,b). The
smaller conjugative interaction between the aromatic rings and
the allyl fragments in the complexes 1(a,b) is also reflected by
the torsional angles of approximately 20° for C2−C3−C4−C5
and C2−C1−C10−C11, while they are close to 0° in the free
cations 5(a−c). The planar configuration of the allyl fragment
in the free cations 5(a,b) is only slightly disturbed in the Pd
complexes 1(a,b), as shown by the torsional angles C1−C2−
C3−C4 and C3−C2−C1−C10, which deviate by less than 5°
from planarity.
Chart 2 shows that the tremendous differences in the elec-
trophilicities of the differently substituted free allyl cations
5(a−c) are leveled in the corresponding palladium complexes
1(a−c), all of which have similar electrophilic reactivities. Com-
parison of the two columns in Chart 2 shows that Pd(PPh3)2
coordination of the highly reactive allyl cation 5b reduces the
electrophilic reactivity by 20 orders of magnitude, while the
reactivity of the stabilized allyl cation 5c is only reduced by 7
orders of magnitude.
f
5.67
4.54
f
2g
2i
7.79 × 10−1
7.71 × 10−1
9.25 × 10
ce f
, ,
6.67 × 10
a
The E parameters for 1(a−c) result from the least-squares
minimization of Δ2 = ∑(log k2 − sN(N + E))2, which uses the
second-order rate constants k2 (this table) and the N and sN
parameters of the nucleophiles 2(a−i) listed in Table 1. E values
with more decimals than given in this table were used for the
calculation of kcalc by eq 1. The use of E parameters given in this table
b
leads to slightly deviating results. The kinetic data are only of
c
moderate quality. This value was not used for the determination of E.
d
e
Four-component kinetic method used. Mixed first- and second-
f
order in nucleophile. The rate constant was determined in the
presence of PPh3. Fumaronitrile must not be added, as it causes
immediate release of the free (E)-1,3-bis(4-dimethylaminophenyl)-
allylium ion (5c).
in the drawn correlation line. The electrophilicity parameters
E (Table 2) of the allylpalladium complexes 1(a−c) were
calculated by minimizing the sum of the squared deviations Δ2 =
∑(log k2 − sN(N + E))2 with the nonlinear solver “What’s
Best!”19 using the rate constants listed in Table 2 and the N and
sN values derived from the reactions of corresponding nucleo-
philes with benzhydrylium ions (Table 1). Because of the lower
accuracy of the second-order rate constants for 2d and 2h,
these k2 values were not used for the determination of the E
parameters. Triphenylphosphine was also not included in the
correlation in order to have comparable reference nucleophiles
for all three allylpalladium complexes. The proximity of these
points to the correlation line implies that the calculated E
values would not be significantly changed if these rate constants
were included in the correlation.
Jutand et al. reported small differences between the rate
constants of the reactions of [(η3-(CH2CHCH2)Pd(PAr3)2]+
and [(PhCHCHCHPh)Pd(PPh3)2]+ with morpholine in
DMF.4,5 This observation may a priori be explained by a com-
pensation of electronic acceleration and steric retardation of the
nucleophilic attack by the phenyl substituents. As the data in
Chart 2 show that electronic substituent effects do not modify
the electrophilicities of allylpalladium complexes, we can con-
clude that steric effects also play a minor role. In line with this
conclusion the rate constants for the reactions of piperidine
(2i) with 1(a−c) (205, 210, and 66.7 M−1 s−1, Table 2) are
It is remarkable that the second-order rate constants of
reactions of 1(a−c) with C-, N-, and P-nucleophiles performed
in three different solvents can be predicted within a factor of 5
by using eq 1. This observation indicates that the N and sN
values of the nucleophiles, determined on the basis of their
reactions with benzhydrylium ions and therefore incorporating
the solvent effects on the stabilities of the diarylmethyl cations
(as the E values were defined as solvent-independent),10 can be
applied for the relatively accurate prediction of the reactivities
of these nucleophiles even toward electrophiles of widely differ-
ent nature.
2421
dx.doi.org/10.1021/om3000357 | Organometallics 2012, 31, 2416−2424

Organometallics
Article
a
Table 3. Geometrical Parameters of the (η3-Diarylallyl)palladium Complexes 1(a,b) and Free Diarylallyl Cations 5(a−c)
Pd complexes
free cations
b
c
d
d
e
d
1a
1b
5a
5b
5c
5c
Bond Lengths (Å)
rC1C10
rC3C4
rC1C2
rC2C3
1.484
1.473
1.397
1.391
1.466
1.476
1.403
1.393
1.425
1.428
1.428
1.390
1.390
1.391
1.413
1.386
1.374
1.415
1.415
1.392
1.392
1.425
1.391
1.391
Torsional Angles (deg)
C2C3C4C5
C2C1C10C11
C1C2C3C4
C3C2C1C10
−13.14
27.48
−19.54
16.09
0.00
−0.008
0.00
3.13
0.00
0.00
−1.69
174.37
−177.96
0.00
−177.02
−175.14
−174.94
179.32
180.00
0.00
180.00
−180
179.99
180.00
a
b
c
d
e
For the numbering of atoms see Figure 10. X-ray data from ref 13. X-ray data from this work. B3LYP/6-31G(d,p) data from ref 9. X-ray data
from ref 9.
can derive that nucleophiles with N > 9 are suitable substrates
for Tsuji−Trost reactions via [(η3-allyl)Pd(PPh3)2]+ complexes.
Weaker nucleophiles may react, however, if another reaction
mechanism is followed or other ligands are used, e.g., in
reactions with indoles21 (N ≈ 5−7), which have been
suggested21a to proceed via the precoordination of the indole
to the metal.
CONCLUSIONS
■
Contrary to previous assumptions,5 secondary and primary
amines were found to attack not only the allyl ligand but also
the central metal in reactions with [(η3-allyl)Pd(PPh3)2]+ com-
plexes. Since the products of the latter reaction, in which one
PPh3 ligand is reversibly replaced by an amine molecule, are
considerably less electrophilic than the Pd(PPh3)2-coordinated
species, complex kinetics were encountered. Small amounts of
PPh3 were shown to suppress the ligand exchange, giving rise to
the high reproducibility of the kinetic measurements of these
reactions.
Substitution of the previously reported12−16 nucleophile-
specific parameters N and sN and the second-order rate con-
stants k2 for the reactions of 1(a−c) with various nucleophiles
in the correlation log k2 = sN(N + E) (eq 1) allowed us to
derive the electrophilicities E ≈ −14 for the [(η3-diarylallyl)-
Pd(PPh3)2]+ almost independent of the nature of the sub-
stituents and to compare them with the reactivities of other
important classes of electrophiles22 (Figure 11).
Figure 10. X-ray structure of complex 1b-OTf. The triflate anion and
acetone (cosolvent for crystallization), which were found in the unit
cell, are omitted in this figure and are depicted in the Supporting
Information.
Chart 2. Electrophilicities of Complexes 1(a−c) and the
Corresponding Free 1,3-Diarylallyl Cations 5(a−c)
EXPERIMENTAL SECTION
■
Materials for Kinetic Measurements. Dichloromethane (p.a.
grade) was subsequently treated with concentrated sulfuric acid, water,
10% NaHCO3 solution, and again water. After predrying with
anhydrous CaCl2, it was freshly distilled over CaH2. Acetonitrile
(HPLC grade) and DMSO (99.7% purity) were used as received.
1
NMR Spectroscopy. In the H and 13C NMR spectra chemical
shifts are expressed in δ (ppm) and refer to CD2Cl2 (δH 5.32, δC
54.00), DMSO (δH 2.50, δC 39.50), and CD3CN (δH 1.94, δC 1.39) as
internal standards. The coupling constants are given in Hz. Abbre-
viations used are s (singlet), d (doublet), t (triplet), q (quartet), m
(multiplet). In the case of 13C NMR spectra, these abbreviations refer
to the multiplicity in hydrogen decoupled spectra, and the hydrogen
multiplicity (based on DEPT or HSQC experiments) is shown as CH3,
CH2, CH, or C to avoid ambiguity.
similar to that for the reaction of piperidine with the [(η3-1-
phenylallyl)Pd(PPh3)2]+ complex6 (30.9 M−1 s−1).20
Since the reactivities of the allylpalladium complexes [(η3-RCH-
CHCHR)PdL2]+ are obviously not related to the reactivities of the
free allyl cations [RCHCHCHR]+, the designation of the former
species as “palladium-stabilized allyl cations” may be misleading.
As bimolecular reactions at room temperature take place only
within acceptable times, when k2 (20 °C) > 10−5 M−1 s−1, one
(E)-5,5′-(3-Chloroprop-1-ene-1,3-diyl)bis(1,3-difluoroben-
zene) (5b-Cl). Prepared from (E)-1,3-bis(3,5-difluorophenyl)prop-2-
en-1-ol (4.58 g, 16.2 mmol) and thionyl chloride (2.71 g, 22.8 mmol)
2422
dx.doi.org/10.1021/om3000357 | Organometallics 2012, 31, 2416−2424

Organometallics
Article
113.8 (t, JCP = 7.3 Hz, CH), 129.3−129.4 (m, CH), 129.6 (d, JCP
=
22.1 Hz, C), 131.5 (s, CH), 134.1−134.2 (m, CH), 139.4−139.6
(m, C), 163.2 ppm (dd, JCF = 248, 13.0 Hz, C). 31P NMR (CD2Cl2,
162 MHz): δ 27.5 ppm. 19F NMR (CD2Cl2, 376 MHz): δ −150.63
to −150.62 (m, 3 F, BF4), −150.58 to −150.57 (m, 1 F, BF4), −110.32
to −110.28 ppm (m, 4 F, FAr). HRMS (ESI+): calcd 895.1492
(C51H39F4P2106Pd+), found 895.1498.
(E)-[(η3-1,3-Bis(3,5-difluorophenyl)allyl)Pd(PPh3)2]+TfO− (1b-OTf).
Analogous to 1b-BF4. From 5b-PdCl (186 mg, 0.228 mmol), PPh3
(242 mg, 0.923 mmol), and sodium triflate (393 mg, 2.28 mmol):
205 mg (0.196 mmol, 43%); yellow needles; mp 206.5−208.5 °C
(dec). To obtain crystals suitable for X-ray diffraction analysis, 1b-OTf
was recrystallized from acetone−Et2O.
(E)-[(η3-1,3-Bis(4-(dimethylamino)phenyl)allyl)Pd-
−
(PPh3)2]+BF4 (1c-BF4). Pd(PPh3)4 (871 mg, 0.754 mmol) and 5c-
BF4 (268 mg, 0.732 mmol) were mixed in the solid state followed by
addition of dichloromethane (10 mL). The resulting solution was
stirred for 10 min at rt; then pentane (ca. 50 mL) was added. After 2 h
of cooling (4 °C), the precipitate was filtered, washed with pentane,
and dried in vacuo to yield 1c-BF4: 476 mg (0.477 mmol, 65%); red
1
powder; mp 162.0−164.0 °C (dec). H NMR (CD2Cl2, 400 MHz):
δ 2.91 (s, 12 H, 2 × NMe2), 5.48−5.53 (m, 2 H, ArCHCHCHAr),
3
6.11 (t, JHH = 12.4 Hz, 1 H, ArCHCHCHAr), 6.26−6.28 (m, 4 H,
HAr), 6.71−6.73 (m, 4 H, HAr), 7.00−7.04 (m, 12 H, o-PPh3), 7.13−
7.17 (m, 12 H, m-PPh3), 7.32−7.36 ppm (m, 6 H, p-PPh3). 13C NMR
(CD2Cl2, 101 MHz): δ 40.5 (s, CH3), 98.7 (t, JCP = 11.7 Hz, CH),
101.3 (t, JCP = ca. 6 Hz, CH), 112.4 (s, CH), 122.0 (t, JCP = 4.1 Hz,
C), 129.1−129.2 (m, CH), 130.0 (s, CH), 130.9 (s, CH), 131.2 (d,
JCP = 19.3 Hz, C), 134.0−134.1 (m, CH), 151.4 ppm (s, C). HRMS
(ESI, positive): calcd 909.2713 (C55H53N2P2106Pd+), found 909.2721.
Reaction of 1a-BF4 with Triphenylphosphine (2e). Triphenyl-
phosphine (553 mg, 2.11 mmol) was added to a solution of 1a-BF4
(315 mg, 0.346 mmol) in acetonitrile (10 mL). After 1 h of stirring
(rt), the precipitate was filtered off, and the filtrate was dried in vacuo.
Diethyl ether (50 mL) was added to the solid residue, and the result-
ing slurry was placed in an ultrasonic bath for 20 min. The remaining
solid was filtered and recrystallized from CH2Cl2/Et2O, yielding a 2:1
mixture of (E)-(1,3-diphenylallyl)triphenylphosphonium tetrafluorobo-
rate (6) and (Z)-(1,3-diphenylprop-1-en-1-yl)triphenylphosphonium
tetrafluoroborate (6′). The 1H and 13C NMR spectra of 6′ were
obtained comparing the spectra of the mixture with those of isolated 6.9
Figure 11. Comparison of the electrophilicities of the allylpalladium
complexes 1(a−c) and electron-deficient π-systems of similar
reactivity.22
in dichloromethane (20 mL) using the procedure from ref 23: 3.82 g
(12.7 mmol, 78%); colorless oil. 1H NMR (CDCl3, 599 MHz): δ 5.54
3
3
(d, JHH = 7.7 Hz, 1 H, ArCHCHCH(Cl)Ar), 6.41 (dd, JHH = 15.6,
3
7.7 Hz, 1 H, ArCHCHCH(Cl)Ar), 6.57 (d, JHH = 15.6 Hz, 1 H,
ArCHCHCH(Cl)Ar), 6.71−6.81 (m, 2 H, HAr), 6.89−6.93 (m, 2 H,
HAr), 6.98−7.02 ppm (m, 2 H, HAr). 13C NMR (CDCl3, 151 MHz): δ
61.2 (t, JCF = 2.2 Hz, CH), 103.4 (s, CH), 103.8 (t, JCF = 25.5 Hz,
CH), 103.9 (t, JCF = 2.9 Hz, CH), 104.1 (t, JCF = 25.2 Hz, CH), 109.6
(dd, JCF = 20.3, 5.3 Hz, CH), 110.5 (dd, JCF = 20.8, 5.6 Hz, CH), 138.7
(t, JCF = 9.5 Hz, C), 143.3 (t, JCF = 8.7 Hz, C), 163.0 (dd, JCF = 251,
12.0 Hz, C), 163.2 ppm (dd, JCF = 249, 13.6 Hz, C).
1
3
4
6′: H NMR (CD2Cl2, 400 MHz): δ 3.62 (dd, JHH = 7.5, JHP
=
3.0 Hz, 2 H, CH2), 6.93−7.02 (m, 3 H, CH, HAr), 7.05−7.10 (m,
2 H, HAr, overlapped with HAr of 6), 7.26−7.36 (m, 5 H, HAr over-
lapped with HAr of 6), 7.38−7.45 (m, 7 H, o-PPh3, HAr, overlapped
with HAr of 6), 7.62−7.70 (m, 6 H, m-PPh3, overlapped with m-PPh3
of 6), 7.83−7.88 ppm (m, 3 H, p-PPh3 overlapped with p-PPh3 of 6).
13C NMR (CD2Cl2, 101 MHz): δ 38.0 (d, JCP = 14.4 Hz, CH2), 117.4
(d, JCP = 88.9 Hz, C), 124.5 (d, JCP = 77.9 Hz, C), 127.7 (s, CH),
129.0 (s, CH), 129.6 (s, CH), 130.4 (d, JCP = 2.5 Hz, CH), 130.08−
130.13 (m, overlapped with resonance of 6, CH), 130.7 (m, over-
lapped with resonance of 6, CH), 130.9 (d, JCP = 12 Hz, CH), 131.5
(d, JCP = 9.5 Hz, C), 135.1 (d, JCP = 10.1 Hz, CH), 136.1 (d, JCP = 3.0
Hz, CH), 137.2 (d, JCP = 1.8 Hz, C), 157.4 ppm (d, JCP = 10.1 Hz,
CH). 31P NMR (CD2Cl2, 162 MHz): δ 24.5 ppm.
[Pd(1,3-bis(3,5-difluorophenyl)-π-allyl)(μ-Cl)]2 (5b-PdCl). Pre-
pared from palladium(II) chloride (400 mg, 2.26 mmol), 5b-Cl (2.70 g,
8.98 mmol), and lithium chloride (405 mg, 9.55 mmol) following
the procedure described in ref 24: 834 mg (1.02 mmol, 90%); yellow
1
powder; mp 247.8−249.2 °C (dec). H NMR (DMSO, 400 MHz):
δ 5.20 (d, 3JHH = 11.8 Hz, 4 H, ArCHCHCHAr), 7.09 (t, J = 11.8 Hz,
2 H, ArCHCHCHAr), 7.17−7.34 (m, 4 H, HAr), 7.40−7.64 ppm (m,
8 H, HAr). 13C NMR (DMSO, 101 MHz): δ 80.7 (s, CH), 103.4 (t,
JCF = 26.0 Hz, CH), 109.5 (s, CH), 111.4−111.6 (m, CH), 141.5
(t, JCF = 10.1 Hz, C), 162.7 ppm (dd, JCF = 245.1, 13.5 Hz, C). 19F NMR
(DMSO, 376 MHz): δ −110.7 to −110.4 ppm (m). HRMS (ESI+,
acetonitrile−water as eluents): calcd 453.0201 (C19H15F4N2106Pd+,
[C6H3F2CHCHCHC6H3F2Pd(CH3CN)2]+), found 453.0202.
Isomerization of 6 in the Presence of Pd(PPh3)4. Pd(PPh3)4
(8.4 mg, 7.3 μmol) and 6 (104 mg, 0.192 mmol) were mixed in the
solid state, and acetonitrile (10 mL) was added. After 15 h of stirring
(ambient temperature), the reaction mixture was freed from the
1
solvent, yielding a 4:1 mixture of 6 and 6′ (based on H NMR).
Reaction of 1a-BF4 with Benzylamine (2g) in the Presence of
Triphenylphosphine and Fumaronitrile. A solution containing
benzylamine (151 mg, 1.41 mmol), triphenylphosphine (185 mg,
0.706 mmol), and fumaronitrile (17 mg, 0.22 mmol) in acetonitrile
(5 mL) was added to a solution of 1a-BF4 (199 mg, 0.218 mmol) in
the same solvent (5 mL). After 1 h of stirring (rt), the precipitate was
filtered off and the filtrate was dried in vacuo. (E)-N-Benzyl-1,
3-diphenylprop-2-en-1-amine (7) was found to be the only allylic
product in the crude mixture (1H NMR). Column chromatography
(MPLC, silica gel, isohexane−EtOAc = 93:7, flow rate 25 mL/min)
−
(E)-[(η3-1,3-Bis(3,5-difluorophenyl)allyl)Pd(PPh3)2]+BF4 (1b-BF4).
Synthesized according to ref 17 from 5b-PdCl (420 mg, 0.516 mmol),
triphenylphosphine (542 mg, 2.07 mmol), and NaBF4 (568 mg,
5.17 mmol): 605 mg (0.615 mmol, 60%); yellow powder; mp 170−
1
172 °C (dec). H NMR (CD2Cl2, 400 MHz): δ 5.45−5.51 (m, 2 H,
ArCHCHCHAr), 6.38−6.51 (m, 7 H, ArCHCHCHAr, HAr), 7.11−
7.16 (m, 12 H, o-PPh3), 7.19−7.23 (m, 12 H, m-PPh3), 7.34−7.38
ppm (m, 6 H, p-PPh3). 13C NMR (CD2Cl2, 101 MHz): δ 91.1 (t, JCP
=
15.0 Hz, CH), 103.7 (t, JCF = 25.6 Hz, CH), 111.3−111.9 (m, CH),
2423
dx.doi.org/10.1021/om3000357 | Organometallics 2012, 31, 2416−2424

Organometallics
Article
afforded 7 (35.4 mg, 0.118 mmol, 54%), the 1H and 13C NMR spectra
of which agreed with literature data.24,25
(c) Amines: Faller, J. W.; Wilt, J. C. Org. Lett. 2005, 7, 633−636. For
stochiometric reactions see, e.g., refs 5 and 6.
Isomerization of 6 in the Presence of Benzylamine (2g). The
phosphonium salt 6 (29.6 mg, 54.6 μmol) was dissolved in CD3CN
(0.5 mL). Then a solution of benzylamine (43.6 mg, 0.407 mmol) in
the same solvent (0.2 mL) was added. 1H and 31P NMR spectra of the
reaction mixture taken 2 and 5 min after mixing, respectively, showed
resonances corresponding to a mixture of 6, 6′, and benzylamine.
After 24 h, no traces of 7 were found, and the ratio between 6 and 6′
(ca. 1:5) remained unchanged.
(12) For N and sN of (a) enamines 2a−c: Kempf, B.; Hampel, N.;
Ofial, A. R.; Mayr, H. Chem.−Eur. J 2003, 9, 2209−2218. (b)
Carbanion 2d: Lucius, R.; Loos, R.; Mayr, H. Angew. Chem., Int. Ed.
2002, 41, 91−95. (c) Phosphine 2e: Kempf, B.; Mayr, H. Chem.
Eur. J. 2005, 11, 917−927. (d) Amine 2f: Phan, T. B.; Nolte, C.;
Kobayashi, S.; Ofial, A. R.; Mayr, H. J. Am. Chem. Soc. 2009, 131,
11392−11401. (e) Amines 2g−i: Kanzian, T.; Nigst, T. A.; Maier, A.;
Pichl, S.; Mayr, H. Eur. J. Org. Chem. 2009, 6379−6385.
(13) Amatore, C.; Bahsoun, A. A.; Jutand, A.; Mensah, L.; Meyer, G.;
Ricard, L. Organometallics 2005, 24, 1569−1577.
Reaction of 6 with Benzylamine in the Presence of
Pd(PPh3)4. The phosphonium salt 6 (29.3 mg, 0.0540 mmol) and
Pd(PPh3)4 (7.3 mg, 6.1 μmol) were dissolved in CD3CN (0.5 mL).
Then a solution of benzylamine (39.6 mg, 370 μmol) in the same
(14) Antonaroli, S.; Crociani, B. J. Organomet. Chem. 1998, 560,
137−146.
1
solvent (0.2 mL) was added. The H NMR spectrum of the reaction
(15) This ratio was found after 15 h of reaction time and probably
does not correspond to the equilibrium mixture.
mixture taken 20 min after mixing showed resonances corresponding
to a mixture of 7, 6, and 6′ in a ratio of ca. 1:0.3:0.6. After a reaction
time of 6 h, the 7:6:6′ ratio reached 1:0.04:0.13.
(16) Åkermark, B.; Åkermark, G.; Hegedus, L. S.; Zetterberg, K.
J. Am. Chem. Soc. 1981, 103, 3037−3040.
(17) For a detailed description of the method used, see: Putau, A.;
Koszinowski, K. Organometallics 2010, 29, 3593−3601, and the
Supporting Information.
(18) An explanation for the higher reaction order in amine in a
similar reaction was proposed in ref 7.
(19) The least-squares optimization was performed using the com-
puter program What’sBest! 7.0 (May 25, 2004), Lindo Systems Inc.,
2004.
(20) Even though the reactions of 1(a−c) with 2i were performed in
acetonitrile and that of (η3-1-phenylallyl)Pd(PPh3)2 in dichloro-
methane, a direct comparison is possible, because amines were
shown to have almost the same reactivity in these solvents: Ammer, J.;
Baidya, M.; Kobayashi, S.; Mayr, H. J. Phys. Org. Chem. 2010, 23,
1029−1035.
(21) (a) Mo, H.; Bao, W. Adv. Synth. Catal. 2009, 351, 2845−2849.
(b) Ma, S.; Yu, S.; Peng, Z.; Guo, H. J. Org. Chem. 2006, 71, 9865−
9868. (c) Usui, I.; Schmidt, S.; Keller, M.; Breit, B. Org. Lett. 2008, 10,
1207−1210.
(22) For a database of reactivity parameters E, N, and sN: www.cup.
lmu.de/oc/mayr/DBintro.html.
ASSOCIATED CONTENT
* Supporting Information
■
S
Details of kinetic and ESI experiments and of the determination
of the equilibrium constant KE, supplementary figures, copies of
the NMR spectra of new compounds, and crystallographic data
in CIF format for 1b-OTf. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: herbert.mayr@cup.uni-muenchen.de.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge financial support by the Deutsche
Forschungsgemeinschaft (SFB 749) and the Fonds der
Chemischen Industrie. We thank Aliaksei Putau for perform-
ing of ESI experiments and Dr. Armin R. Ofial, Dr. Markus
Horn, and Johannes Ammer for their help in preparation of
the manuscript.
(23) Nolte, C.; Mayr, H. Eur. J. Org. Chem. 2010, 1435−1439.
(24) Hayashi, T.; Yamamoto, A.; Yoshihiko, I.; Nishioka, E.; Miura,
H.; Yanagi, K. J. Am. Chem. Soc. 1989, 111, 6301−6311.
(25) Breeden, S.; Wills, M. J. Org. Chem. 1999, 64, 9735−9738.
REFERENCES
■
(1) (a) Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96, 395−
422. (b) Trost, B. M. Chem. Pharm. Bull. 2002, 50, 1−14. (c) Trost,
B. M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921−2943.
(2) Jutand, A. Appl. Organomet. Chem. 2004, 18, 574−582, and
references therein.
(3) Jutand, A. Eur. J. Inorg. Chem. 2003, 2017−2040.
́
(4) Cantat, T.; Genin, E.; Giroud, C.; Meyer, G.; Jutand, A.
J. Organomet. Chem. 2003, 687, 365−376.
(5) Amatore, C.; Genin, E.; Jutand, A.; Mensah, L. Organometallics
́
2007, 26, 1875−1880 , and references therein.
(6) Kuhn, O.; Mayr, H. Angew. Chem., Int. Ed. 1999, 38, 343−346.
(7) Crociani, B.; Antonaroli, S.; Bandoli, G.; Canovese, L.; Visentin, F.;
Uguagliati, P. Organometallics 1999, 18, 1137−1147, and references
therein.
(8) (a) Canovese, L.; Visentin, F.; Chessa, G.; Niero, A.; Uguagliati,
P. Inorg. Chim. Acta 1999, 293, 44−52. (b) Crociani, B.; Antonaroli, S.;
Di Bianca, F.; Canovese, L.; Visentin, F.; Uguagliati, P. J. Chem. Soc.,
Dalton Trans. 1994, 1145−1151.
(9) Troshin, K.; Schindele, C.; Mayr, H. J. Org. Chem. 2011, 76,
9391−9408.
(10) Mayr, H. Angew. Chem., Int. Ed. 2011, 50, 3612−3618.
(11) For catalytic reactions see e.g.: (a) Enamines: Zhao, X.; Lui, D.;
Xie, F.; Zhang, W. Tetrahedron 2009, 65, 512−517. (b) Carbanions:
Ross, J.; Chen, W.; Xu, L.; Xiao, J. Organometallics 2001, 20, 138−142.
2424
dx.doi.org/10.1021/om3000357 | Organometallics 2012, 31, 2416−2424