Kinetic studies of the oxidative addition and transmetallation steps involved in the cross-coupling of alkynyl stannanes with aryl iodides catalysed by η2-(dimethyl fumarate)(iminophosphane)palladium(0) complexes

Article abstract of DOI:10.1002/ejic.200300376

The complexes [Pd(η²-dmfu)(P-N)] {dmfu = dimethyl fumarate; P-N = 2-(PPh₂)C₆H₄-1-CH=NR, R = C ₆H₄OMe-4 (1a), CHMe₂ (2a), C₆H ₃Me₂-2,6 (3a), C₆H₃(CHMe ₂)₂-2,6 (4a)} undergo dynamic processes in solution which consist of a P-N ligand site exchange through initial rupture of the Pd-N bond at lower energy and an olefin dissociation-association at higher energy. According to equilibrium constant values for olefin replacement, the complex [Pd(η²-fn)(P-N)] (fn = fumaronitrile, 1b) has a greater thermodynamic stability than its dmfu analogue 1a. The kinetics of the oxidative addition of ArI (Ar = C₆H₄CF₃-4) to 1a and 2a lead to the products [PdI(Ar)(P-N)] (1c, 2c) and obey the rate law, k _obs = k_1A k_2A[ArI]. The k_1A step involves oxidative addition to a reactive species [Pd(solvent)(P-N)] formed from dmfu dissociation. The k_2A step is better interpreted in terms of oxidative addition to a species [Pd(η²-dmfu)(solvent) (κ¹-P-N)] formed in a pre-equilibrium step from Pd-N bond breaking. The complexes 1c and 2c react with PhC≡CSnBu₃ in the presence of an activated olefin (ol = dmfu, fn) to yield the palladium(0) derivatives [Pd(η²-ol)(P-N)] along with ISnBu₃ and PhC≡CAr. The kinetics of the transmetallation step, which is rate-determining for the overall reaction, obey the rate law: k_obs = k_2T[PhC≡CSnBu₃]. The k_2T values are markedly enhanced in more polar solvents such as CH₃CN and DMF. The solvent effect and the activation parameters suggest an associative S _E2 mechanism with substantial charge separation in the transition state. The kinetic data of the above reactions in various solvents indicate that, for the cross-coupling of PhC≡CSnBu₃ with ArI catalysed by 1a or 2a, the rate-determining step is represented by the oxidative addition and that CH₃CN is the solvent in which the highest rates are observed. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejic.200300376

FULL PAPER
Kinetic Studies of the Oxidative Addition and Transmetallation Steps Involved
in the Cross-Coupling of Alkynyl Stannanes with Aryl Iodides Catalysed by
η2-(Dimethyl fumarate)(iminophosphane)palladium(0) Complexes
Bruno Crociani,*[a] Simonetta Antonaroli,[a] Luciano Canovese,[b] Paolo Uguagliati,[b] and
Fabiano Visentin[b]
Keywords: Palladium / N,P ligands / Kinetics / Cross-coupling
The complexes [Pd(η2-dmfu)(P−N)] {dmfu = dimethyl fumar-
ate; P−N = 2-(PPh2)C6H4−1-CH=NR, R = C6H4OMe-4 (1a),
CHMe2 (2a), C6H3Me2-2,6 (3a), C6H3(CHMe2)2-2,6 (4a)} un-
complexes 1c and 2c react with PhCϵCSnBu3 in the pres-
ence of an activated olefin (ol = dmfu, fn) to yield the palla-
dium(0) derivatives [Pd(η2-ol)(P−N)] along with ISnBu3 and
dergo dynamic processes in solution which consist of a P−N PhCϵCAr. The kinetics of the transmetallation step, which is
ligand site exchange through initial rupture of the Pd−N
bond at lower energy and an olefin dissociation-association
at higher energy. According to equilibrium constant values
rate-determining for the overall reaction, obey the rate law:
kobs = k2T[PhCϵCSnBu3]. The k2T values are markedly en-
hanced in more polar solvents such as CH3CN and DMF. The
for olefin replacement, the complex [Pd(η2-fn)(P−N)] (fn = fu- solvent effect and the activation parameters suggest an asso-
maronitrile, 1b) has a greater thermodynamic stability than
its dmfu analogue 1a. The kinetics of the oxidative addition
of ArI (Ar = C6H4CF3-4) to 1a and 2a lead to the products
ciative SE2 mechanism with substantial charge separation in
the transition state. The kinetic data of the above reactions
in various solvents indicate that, for the cross-coupling of
PhCϵCSnBu3 with ArI catalysed by 1a or 2a, the rate-deter-
[PdI(Ar)(P−N)] (1c, 2c) and obey the rate law, kobs = k1A
+
k2A[ArI]. The k1A step involves oxidative addition to a reac- mining step is represented by the oxidative addition and that
tive species [Pd(solvent)(P−N)] formed from dmfu dissoci- CH3CN is the solvent in which the highest rates are ob-
ation. The k2A step is better interpreted in terms of oxidative served.
addition to a species [Pd(η2-dmfu)(solvent)(κ1-P−N)] formed
in a pre-equilibrium step from Pd−N bond breaking. The
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2004)
Introduction
ligand. In fact, the rate-determining step of the cycle can
either be the oxidative addition or the transmetallation or
The cross-coupling of organo-stannanes with carbon even the reductive elimination.[1c,2,4]
electrophiles catalysed by palladium complexes (Stille reac-
tion) is a powerful synthetic method in organic chemistry.[1]
The generally accepted catalytic cycle involves an initial oxi-
dative addition of the organic electrophile to a palladium(0)
species followed by transmetallation, trans-to-cis isomeris-
ation if trans-diorganopalladium() intermediates are
formed, and eventually reductive elimination of the coupled
product.[1,2] As shown in previous studies,[3] for the com-
plexity of the reactions involved the mechanism of the pro-
cess can be variously affected by the nature of the ligand,
the solvent, the reacting stannane and organic electrophile,
and by the presence of additives such as LiCl or the free
The mechanism described in Scheme 1 for the coupling
of ArX (Ar ϭ pentahalophenyl; X ϭ halide, triflate) with
RSnBu3 (R ϭ vinyl) has been recently proposed.[5]
In this mechanism, two transmetallation pathways may
be present without any prior dissociation (or solvent dis-
placement) of the ligand L in the intermediates trans-
[PdXArL2]. The associative SE2 cyclic step (a) is favoured
in weakly coordinating solvents of low polarity when X is
a good bridging ligand such as a halide. It leads to a highly
reactive cis species which undergoes an immediate reductive
elimination and is strongly retarded by the presence of the
free ligand L. The associative SE2 open step (b) is favoured
when LЈ is a ligand or a polar coordinating solvent both of
which lack bridging capability and when X is a good-leav-
ing and poorly coordinating anionic ligand without bridg-
ing ability.
[a]
`
Dipartimento di Scienze e Tecnologie Chimiche, Universita di
Roma ‘‘Tor Vergata’’,
Via della Ricerca Scientifica, 00133 Roma, Italy
Fax: (internat.) ϩ 39-06-72594328
E-mail: crociani@stc.uniroma2.it
[b]
In contrast, the transmetallation of [{η5-(1-Ph2P)(2,4-
Ph2)C5H2}(CO)3MoPdI(PPh3)] by PhCϵCSnBu3, leading
`
Dipartimento di Chimica, Universita di Venezia,
Dorsoduro 2137, 30123 Venezia, Italy
Fax: (internat.) ϩ 39-041-2348517
E-mail:cano@unive.it
to
[{η5-(1-Ph2P)(2,4-Ph2)C5H2}(CO)3MoPd(CϵCPh)-
732
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejic.200300376
Eur. J. Inorg. Chem. 2004, 732Ϫ742

Oxidative Addition of Alkynyl Stannanes
FULL PAPER
Scheme 1. Catalytic cycle for the cross-coupling reaction with pal-
ladium complexes containing monodentate ligands L
Scheme 2. Alternative catalytic cycle for the cross-coupling reaction
with the system [PdCl(η3-C3H5)]2/LϪLЈ (1:2 molar ratio)
[Pd(η2-dmfu)(PϪN)] molar ratio Ն 1. In a mechanistic
investigation based mainly on IR and NMR spectroscopic
data,[13] we have reported that the cross-coupling of
PhCϵCSnBu3 with IC6H4CF3-4 proceeds through the
classical cycle involving initial oxidative addition of the aryl
iodide to [Pd(η2-dmfu)(PϪN)] {PϪN ϭ 2-(PPh2)C6H4Ϫ1-
(PPh3)] in dimethylformamide (DMF), was reported to pro-
ceed through
a
reactive species [{η5-(1-Ph2P)(2,4-
Ph2)C5H2}(CO)3MoPdI(DMF)],formed in a displacement
equilibrium of the ligand PPh3 by the solvent when a low
initial concentration of the starting complex was used.[6]
On the other hand, the complex trans-[PdIPh(AsPh3)2]
was also found to be in equilibrium with the species
[PdIPh(solvent)(AsPh3)] (solvent ϭ DMF) upon release of
AsPh3 in dimethylformamide,[7] in agreement with previous
kinetic studies.[1b][3e] In a recent report, however, the con-
centration of [PdXR(solvent)(AsPh3)] (X ϭ Cl, I; R ϭ Ph,
C6Cl2F3) was found to be negligible in the presence of free
arsane, indicating that under catalytic conditions the pre-
dominant transmetallation step involves the associative re-
action of the stannane with the complex [PdXR-
(AsPh3)2].[8]
CHϭNR,
R ϭ
C6H4OMe-4, CHMe2}, or [Pd(η2-
dmfu)(PϪN)2] formed from the reaction of [Pd(η2-
dmfu)(PϪN)] with one equivalent of the iminophosphane
PϪN, followed by trans-metallation of the product
[PdI(C6H4CF3-4)(PϪN)] and by fast reductive elimination
of the alkyne PhCϵCC6H4CF3-4 to regenerate the starting
palladium(0) compound. With [Pd(η2-dmfu)(PϪN)] as the
catalyst, the oxidative addition is the rate-determining step,
while for [Pd(η2-dmfu)(PϪN)2] the oxidative addition and
transmetallation steps proceed at comparable rates.
In the coupling of pentahaloaryl triflates (ArX) with
CH2ϭCHSnBu3 catalysed by [PdXAr(dppe)] [dppe ϭ 1,2-
bis(diphenylphosphanyl)ethane] the main intermediates,
[PdAr(CHϭCH2)(dppe)] and [Pd(η2-CH2ϭCHAr)(dppe)],
have been spectroscopically detected and characterised.[9]
The system [PdCl(η3-C3H5)]2/LϪLЈ (1:2 molar ratio)
{LϪLЈ ϭ 2-(PPh2)C6H4Ϫ1-CHϭNC2H4Ph}, was found to
have a high catalytic efficiency in the coupling of 4-(trifluor-
omethyl)iodobenzene with PhCϵCSnBu3.[10] From the de-
tection of the labile complexes [PdI(SnBu3)(LϪLЈ)] and
[Pd(CϵCPh)(SnBu3)(LϪLЈ)] in the 31P NMR spectra of the
reaction mixture, an alternative catalytic cycle was proposed
We report herein the results of a kinetic investigation of
the oxidative addition of IC6H4CF3-4 to [Pd(η2-
dmfu)(PϪN)] and the transmetallation of [PdI(C6H4CF3-
4)(PϪN)] by PhCϵCSnBu3. The studies have been carried
out using UV/Vis spectroscopic techniques in order to get
a better understanding of the mechanism and of the factors
which affect these fundamental steps of the catalytic cycle.
Results and Discussion
The structures of the palladium complexes used in the
involving the initial oxidative addition of PhCϵCSnBu3 to kinetic study are shown in Scheme 3.
an in situ generated palladium(0) species, Pd(LϪLЈ)
According to the X-ray structural analyses of the imino-
(Scheme 2).
phosphane derivatives [Pd(η2-fn)(PϪN)] (fn ϭ fumaronitr-
The
zero-valent
complexes
[Pd(η2-dmfu)(PϪN)] ile, 1b)[14] and [PdI(Ph)(PϪN)] {PϪN ϭ 2-(PPh2)C6H4Ϫ1-
{dmfu ϭ dimethyl fumarate; PϪN ϭ 2-(PPh2)C6H4Ϫ1- CHϭNR; R ϭ Me, Et},[15] the olefin carbons are almost
CHϭNR, R ϭ alkyl and aryl group},[11] are quite active coplanar with the NϪPdϪP coordination plane for the pal-
catalysts (or catalyst precursors) in the coupling of organo- ladium(0) complexes, while in the palladium() complexes
stannanes with aryl halides.[12] The catalytic efficiency is re- the iodide ligand is trans to the phosphorus atom.
tained for a prolonged period of time and increases con-
In order to clarify the intimate mechanism governing the
siderably on going from alkyl to aryl N-substituents, and cross-coupling reaction between IC6H4CF3-4 and
also in the presence of the free iminophosphane for a PϪN/ PhCϵCSnBu3 catalysed by the complexes 1a and 2a,[12,13]
Eur. J. Inorg. Chem. 2004, 732Ϫ742
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
733

B. Crociani, S. Antonaroli, L. Canovese, P. Uguagliati, F. Visentin
FULL PAPER
dative addition by the organic halide, thereby explaining its
lack of reactivity.
The equilibrium constant in CHCl3 is one-order of mag-
nitude higher than in THF. We have confirmed the ob-
served solvent effect by comparing the corresponding KE
values for the 2-(iminomethyl)pyridine complex [Pd(η2-
dmfu)(C5H4N-2-CHϭNC6H4OMe-4)]. In THF, a value of
731Ϯ32 was obtained whereas in CHCl3 a value of
4000Ϯ400 was previously determined.[16] We surmise that
such an effect predominantly stems from a different stabilis-
ation, by solvation, of the olefins in the two solvents exam-
ined even though they have comparable dielectric constants.
Solution Behaviour of the Palladium(0) and Palladium(II
)
Complexes
It is known that palladium(0) olefin complexes with P,N-
chelating ligands undergo various dynamic processes in
solution such as olefin rotation around its bond axis to the
central metal, olefin dissociation, PdϪN bond rupture and
exchange between the free and coordinated olefin.[11,18] The
dissociative processes lead to unsaturated (or solvent-coor-
dinated) species which may be of importance in the mecha-
nism of the oxidative addition.[19] For instance, olefin dis-
sociation leads to a species of the type Pd(PϪN), which can
Scheme 3. Structures of the (iminophosphane)palladium complexes
[bornyl ϭ endo-(1R)-1,7,7-trimethylbicyclo[2.2.1]hept-2-yl]
we have chosen to break the problem down into separate be generated in situ also from the reaction of a mixture
studies, i.e. the oxidative addition and the transmetallation/ of [PdCl(η3-C3H5)]2/LϪLЈ (1:2 molar ratio) {LϪLЈ ϭ 2-
reductive elimination reactions. Preliminary studies on the (PPh2)C6H4Ϫ1-CHϭNC2H4Ph} with NaCH(CO2Me)2.[10]
relative stability of the palladium(0) olefin complexes 1a Such
a
species undergoes oxidative addition with
and 1b, and on the solution behaviour of the palladium(0) PhCϵCSnBu3 or IC6H4CF3-4 to give the products
and the palladium() complexes of Scheme 3 have been also [Pd(CϵCPh)(SnBu3)(PϪN)] or [PdI(SnBu3)(PϪN)]. We
1
carried out.
have therefore examined the H NMR spectra of the com-
plexes 1a and 2a at 25 °C in CDCl3 and [D8]toluene and at
variable-temperatures in [D8]toluene (Table 1).
Relative Stability of Palladium(0) Olefin Complexes from
Olefin Exchange Equilibria
As can be seen, the signals of the imino proton appear at
lower frequencies (δ ϭ 0.6Ϫ0.7 ppm) in the aromatic sol-
vent whereas those of the olefin protons appear at higher
frequencies (δ ϭ 0.4Ϫ0.7 ppm) relative to the correspond-
ing signals in CDCl3. The proton resonances of 1a are fairly
sharp in CD3Cl but broader in [D8]toluene indicating the
occurrence of dynamic processes at increasing rates in the
latter solvent. At higher temperatures in [D8]toluene, the
olefin Hcis and Htrans resonances (∆ν ϭ 260 Hz) coalesce at
56 °C, from which a ∆G϶ value of 63.4 kJ·molϪ1 can be
evaluated.[20] The olefin methyl signals also coalesce at ca.
35 °C but in this case, however, a more precise measurement
of the coalescence temperature was prevented by overlap-
Previous results on the oxidative addition of IC6H4CF3-
4 to [Pd(η2-ol)(PϪN)] (ol ϭ dmfu, fn) involved in the cata-
lytic cycle of the cross-coupling reaction[12,13] have shown
that only the dmfu complexes display appreciable reactivity.
Thus, we have tried to attribute this feature to a lower
thermodynamic stability of such species with respect to
their fn analogues by studying the equilibrium 1 of olefin
exchange in THF and CHCl3 by UV/Vis spectrophotome-
try [Equation (1)].
(1)
ping with the iminophosphane OCH3 singlet (at δ ϭ
1
The resultant KE values of 8520Ϯ300 in CHCl3 and 3.31 ppm in the spectrum at 25 °C). The H NMR spectra
640Ϯ20 in THF indicate that the position of equilibrium is of 2a at 25 °C are characterised by sharp olefin signals in
strongly shifted to the right. The much greater stability of the solvents examined. Upon heating the [D8]toluene solu-
complex 1b is largely due to the greater π-accepting proper- tion, the Hcis and Htrans signals (∆ν ϭ 129 Hz) coalesce at
ties of fumaronitrile which increase the strength of the pal- 75 °C (∆G϶ ϭ 69.3 kJ·molϪ1). At approximately the same
ladium-olefin bond as already observed for the olefin ex- temperature the olefin methyl resonances are also in co-
change equilibria in similar zero-valent complexes [Pd(η2- alescence but overlap largely with the CHMe2 septet (at δ ϭ
ol)(LϪLЈ)] {LϪLЈ ϭ 2-(iminomethyl)pyridine[16] or 2- 3.22 ppm in the spectrum at 25 °C) of the iminophosphane.
(methylthio)pyridine[17]}. A reduced electron density on the It is interesting to point out that the isopropyl Me groups
metal in 1b will render the complex less susceptible to oxi- remain diastereotopic at the various temperatures exam-
734
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 732Ϫ742

Oxidative Addition of Alkynyl Stannanes
FULL PAPER
1
Table 1. Selected H NMR spectroscopic data of the palladium complexes
Complex
Solvent
Iminophosphane protons[a]
Olefin protons[a]
CHcis
[b]
[c]
NϭCH
NϪCH
CH3
CHtrans
CH3
1a
[D8]toluene
7.54 d
[2.8]
3.31 s
4.85 br. s
4.20 br. s
3.30 br. s,
3.20 br. s
[11]
CDCl3
8.11 d
[4.0]
3.81 s
4.23 d
(9.9)
3.55 dd
(9.9)[9.9]
3.32 s,
3.14 s
2a
3a
4a
[D8]toluene
CDCl3
7.54 d
[3.6]
3.22 spt
(6.4)
1.31 d, 1.28 d
(6.4) (6.4)
4.65 dd
(10.0) [2.8]
4.32 dd
(10.0) [10.0]
3.52 s,
3.12 s
8.24
[3.2]
3.65 spt[d]
1.30 d, 1.29 d
(6.3) (6.3)
4.13 dd
(10.4) [2.4]
3.88 dd
(10.4) [10.4]
3.65 s,
3.14 s
[D8]toluene
CDCl3
7.42 d
[3.0]
2.14 s, 1.91 s
4.82 dd
(10.2) [2.6]
3.94 dd
(10.4) [10.4]
3.23 s,
3.17 s
8.06 d
[3.2]
2.16 s, 1.90 s
4.24 dd
(10.0) [2.4]
3.37 dd
(10.2) [10.2]
3.18 s
[D8]toluene
7.73 d
[3.2]
1.45 d, 1.19 d,
(6.8) (6.8)
0.92 d, 0.59 d
(6.8) (6.8)
4.82 dd
(10.0) [3.0]
4.05 dd
(10.0) [10.0]
3.15 s,
2.97 s
CDCl3
8.06 d
[2.1]
1.45 d, 1.09 d,
(6.8) (6.8)
0.98 d, 0.56 d
(6.8) (6.8)
4.25 dd
(10.4) [2.5]
3.49 dd
(10.4) [10.4]
3.24 s,
2.93 s
5b[e]
5b[f]
[D8]toluene
[D8]toluene
7.92 s
8.12 s
3.63 m
4.41 m
0.92 s, 0.72 s,
0.21 s
3.16 dd
(9.5) [3.8]
2.82 dd
(9.5) [9.5]
mk
mk
2.66 dd
(9.7) [9.7]
1c
2c
CD3CN
8.31 s
7.54 s
3.80 s
[D8]toluene
5.93 spt
(6.8)
1.05 d
(6.8)
[a]
[b]
[c]
At 25 °C. Coupling constants JH,H in round brackets, JP,H in square brackets. br ϭ broad, mk ϭ masked.
cis to phosphorus.
[d]
[e]
[f]
trans to phosphorus. Overlapping with the olefin CH3 signal. Major diastereoisomer, see text. Minor diastereoisomer.
ined. Even at 75 °C, they are detected as two sharp doublets stereoisomers are distinct and sharp, although the molar
at δ ϭ 1.27 and 1.24 ppm, respectively. ratio between the major and minor isomer decreases to
The fluxional behaviour of complexes 1a and 2a may be 3.6:1. These findings indicate that the observed dynamic
explained either by olefin rotation, by olefin dissociation- process does not involve olefin dissociation-association
association or by PϪN ligand site exchange through initial (which would cause interconversion of the diastereoisomers
rupture of the PdϪN bond. In order to ascertain which is and eventually coalescence of their signals). However, the
the actual process operating in solution, we have examined olefin dissociation-association process does occur at a lower
the variable-temperature 1H NMR spectra of the complexes rate (on the NMR time-scale) as indicated by the change in
3a and 4a which contain N-aryl substituents with increasing the isomer molar ratio when the temperature is increased.
1
steric requirements, as well as those of complex 5b which
The H NMR spectra of 3a and 4a at 25 °C are charac-
contains a chiral iminophosphane ligand. We have also terised by the non-equivalence of the 2,6 groups on the N-
compared the olefin 13C NMR signals of 1a and 2a with aryl moieties: for 3a, two singlets can be detected for the
those of 3a and 4a (Table 2).
protons of the 2,6-Me groups (the singlet at δ ϭ 2.14 ppm
1
The H NMR spectrum of 5b at 25 °C shows the pres- partially overlapping with the solvent signals) whereas for
ence of two diastereoisomers in a molar ratio of 6.0:1. The 4a, four doublets can be detected for the Me protons and
olefin Hcis and Htrans signals of the major diastereoisomer two septets (at δ ϭ 3.38 and 2.83 ppm, respectively) for the
(∆ν ϭ 138 Hz) coalesce at 73 °C (∆G϶ ϭ 68.7 kJ·molϪ1). CH protons of the 2,6-CHMe2 groups. This is clearly a
At this temperature, the other resonances of the two dia- consequence of the asymmetric nature of the complexes and
Eur. J. Inorg. Chem. 2004, 732Ϫ742
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
735

B. Crociani, S. Antonaroli, L. Canovese, P. Uguagliati, F. Visentin
Table 2. Selected 13C NMR spectroscopic data of the palladium(0) complexes
FULL PAPER
Complex
Iminophosphane carbons[a]
Olefin carbons[a]
CϭO
NϭCH
NϪC
148.7
137.5
CH3
CH3
HCϭCH
1a[11]
2a
164.6
55.5
173.8, 173.5
174.2, 174.0
50.4Ϫ50.0[b]
50.4, 50.3
50.4Ϫ50.0[b]
161.8
23.4, 22.1
48.2[c], 48.0 d[d]
(32)
3a
167.3
166.9
151.9
150.5
18.6, 18.1
173.8, 173.3
50.6
50.2[c], 49.6 d[d]
(30)
4a
24.2, 23.1,
22.6, 22.0
174.0, 172.6
50.5, 50.4
50.3[c], 49.8 d[d]
(31)
[a]
[b]
[c]
[d]
In CDCl3 at 25 °C. Coupling constants JP,H in round brackets. Overlapping signals. cis to phosphorus. trans to phosphorus.
of the restricted rotation around the NϪaryl bond. When
the [D8]toluene solution of 4a is warmed up, all the olefin
and the CHMe2 signals progressively broaden but at 95 °C
(the highest temperature explored) only the coalescence of
the olefin methyl signals (∆ν ϭ 70 Hz) can be observed
(∆G϶ ϭ 75.3 kJ·molϪ1). In contrast, for the solution of 3a,
the olefin methyl signals (∆ν ϭ 23 Hz) coalesce at 53 °C
(∆G϶ ϭ 69.4 kJ·molϪ1), the methyl signals of the N-
C6H3Me2-2,6 substituent (∆ν ϭ 91 Hz) coalesce at 72 °C
(∆G϶ ϭ 69.7 kJ·molϪ1) and eventually the olefin Hcis and
H
trans signals (∆ν ϭ 351 Hz) coalesce at 93 °C (∆G϶ ϭ 70.0
kJ·molϪ1). The close values of the activation free energies
indicate that the same dynamic process is operating which
interconverts simultaneously the two olefin methyl groups,
the two methyl groups of the N-C6H3Me2-2,6 substituent
and the olefin Hcis and Htrans protons. Such a process can-
not be the olefin rotation because it would lead to intercon-
version of only the olefin protons. The dynamic behaviour
of 3a, as well as that of the other complexes examined, is
better rationalised by a process involving an initial solvent-
assisted rupture of the PdϪN bond, followed by rotation
around the NϪaryl bond in the P-monodentate iminophos-
phane and ligand site exchange, as depicted in Scheme 4.
Scheme 4. Proposed mechanism for the dynamic behaviour of com-
plex 3a (R1 ϭ R2 ϭ CO2Me; S ϭ solvent)
As can be seen by comparing structure A and the (ro- 1a, 3a, and 4a since they resonate in the narrow range of
tated) structure B, this process brings about the simul- 50.4Ϫ49.6 ppm. In other words, this means that the extent
taneous interconversions R1 o R2, Me1 o Me2, and H1 of π back-donation in the PdϪη2-dmfu bond and the elec-
o H2.
tron donor properties of the PϪN ligands (in essence, their
It is worth noting that the same mechanism applied to bond strengths to the central metal) are comparable in this
complex 2a leads to the observed interconversions R1 o R2 closely related series of complexes.
and H1 o H2, but not to Me1 o Me2 because the Me
The different rate of the dynamic process observed for
groups of the N-CHMe2 substituent cannot interconvert by the complexes 1a and 2a (1a Ͼ 2a) is essentially due to
a simple rotation around the NϪC bond in the P-monod- electronic factors since the N-C6H4OMe-4 and N-CHMe2
entate iminophosphane. For such an interconversion, an in- groupings of the PϪN ligands have similar steric require-
version of configuration at the NϪCisopropyl carbon atom ments. From the data in Table 2 it appears that the olefin
would also be required.
CϭC carbons of 2a are shielded by ca. 2 ppm relative to
According to this mechanism, the decreasing rate of the the corresponding carbons of 1a, indicating a certain in-
dynamic process in the complexes with N-aryl groups (1a Ͼ crease of π back-donation in the PdϪη2-dmfu bond. This
3a Ͼ 4a) can be explained by the increasing steric hindrance is clearly related to the electron-releasing isopropyl group
exerted by the 2,6-substituents on the N-aryl group towards which increases the σ donor properties of the imino nitro-
the incoming solvent molecule. Inspection of the 13C NMR gen and, therefore, the strength of the PdϪN bond in 2a.
spectroscopic data in Table 2 shows that the olefin CϭC The imino ligands have been found to act more as σ-donors
carbons have comparable electron density in the complexes than as π-acceptors to palladium.[21] Conversely, the for-
736
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 732Ϫ742

Oxidative Addition of Alkynyl Stannanes
FULL PAPER
mally zero-valent complexes [Pd(η2-ol)(NϪNЈ)] (ol ϭ acti-
vated olefin, NϪNЈ ϭ α-diimine) and [Pd(η2-ol)(PϪN)] are
better described as PdII-cyclopropane complexes on the
basis of spectroscopic and crystallographic data.[11,14,21]
The 1H spectra of 1c (in CD3CN) and 2c (in [D8]toluene)
at 25 °C show that these palladium() complexes exist as
single geometrical isomers in solution. No isomerisation or
other dynamic process was observed in the variable-tem-
perature 1H NMR spectra of 2c in [D8]toluene in the range
25Ϫ90 °C.
As suggested by electrical conductivity measurements in
acetonitrile, the complexes 1c and 2c undergo the solvolytic
equilibrium 2 [Equation (2)].
Figure 1. Plots of kobs versus the aryl iodide concentration for the
oxidative addition of ArI to complex 1a in different solvents at 25
°C in the presence of dmfu ([1a]0 ϭ 1·10Ϫ4 ; [dmfu]ϭ 3·10Ϫ3 )
(2)
The cationic complexes [Pd(CH3CN)(Ar)(PϪN)]ϩ have
Ϫ
Ϫ
been isolated as BF4 or CF3SO3 salts upon addition of
AgBF4 or AgCF3SO3 to the solution (see Exp. Sect.). The
equilibrium constants KS were determined by conductivity
and checked by UV/Vis experiments at 25 °C (see Exp. Sect.).
From the measured values [KS ϭ (6.9Ϯ0.9)ϫ10Ϫ6 for 1c;
KS ϭ (2.1Ϯ0.5)ϫ10Ϫ5 for 2c] it appears that more than 20%
and 35% of the cationic species [Pd(CH3CN)(Ar){2-
(PPh2)C6H4Ϫ1-CHϭNC6H4OMe-4}]ϩ and [Pd(CH3CN)-
(Ar){2-(PPh2)C6H4Ϫ1-CHϭNCHMe2}]ϩ, respectively, are
present in acetonitrile solution at 25 °C under kinetic con-
ditions ([1c]0 and [2c]0 ϭ 1·10Ϫ4 ). Conductivity measure-
ments in other solvents (CHCl3, THF, toluene, DMF) sug-
gest that only in CH3CN does the solvolytic equilibrium (2)
take place.
The ensuing kobs values display a linear dependence on
[ArI] in any solvent. As can be seen in Figure 1, a concen-
tration independent (k1A) and a concentration dependent
(k2A) path can always be detected so that the observed rate
constants are represented by the following relationship:
kobs ϭ k1A ϩ k2A[ArI]
The values of k1A and k2A in the solvents examined are
reported in Table 3.
Table 3. Kinetic constants for the oxidative addition of ArI to 1a
in different solvents at 25 °C in the presence of dmfu ([1a]0
ϭ
1·10Ϫ4 ; [dmfu]ϭ 3·10Ϫ3 )
Solvent
k1A/ sϪ1
k2A/ Ϫ1sϪ1
(2.6 Ϯ 0.2)ϫ10Ϫ5
(4.81 Ϯ 0.12)ϫ10Ϫ5
(7.1 Ϯ 0.4)ϫ10Ϫ5
(8.7 Ϯ 1.1)ϫ10Ϫ5
(8.3 Ϯ 0.5)ϫ10Ϫ4
(2.9 Ϯ 0.3)ϫ10Ϫ4
Oxidative Addition Reactions
DMF
THF
CHCl3
Toluene
CH3CN
(2.11 Ϯ 0.02)ϫ10Ϫ3
(1.09 Ϯ 0.06)ϫ10Ϫ3
(6.7 Ϯ 0.2)ϫ10Ϫ3
(3.8 Ϯ 0.7)ϫ10Ϫ3
The oxidative addition of IC6H4CF3-4 (ArI) to the com-
plexes 1a or 2a proceeds according to Equation (3).
(3)
1
The lowest k1A and k2A values were observed in DMF,
while the highest k1A value was observed in CH3CN and
the highest k2A value in toluene. Such a reactivity clearly
does not depend on the difference in dielectric constants,
indicating that the transition states for both steps do not
involve a substantial separation of electric charge. Further-
more, a radical mechanism can be ruled out since no di-
iodo derivative, which represents a typical by-product in the
radical reactions, could be detected.
The kinetic constants were not appreciably influenced by
different concentrations of free dmfu, as reported in
Table 4.
From the kinetic data at different temperatures (Table 5),
the activation parameters were determined using a re-par-
ameterised Eyring equation.[22]
The progress of the reaction was monitored by H and
31P NMR spectroscopy at 25 °C in various solvents
(CDCl3, [D8]THF, [D7]DMF, and CD3CN) using 1a or 2a/
ArI molar ratios in the range of 1:5Ϫ1:10. The spectra at
different times showed the presence of the reactants and
products of Equation (3). No signal attributable to inter-
mediates or side-products was detected. Under the same ex-
perimental conditions, no reaction was observed to occur
between the fumaronitrile derivative 1b and ArI.
For kinetic measurements, the oxidative addition to 1a
was monitored in different solvents and at different tem-
peratures (in the range 15Ϫ45 °C) in the presence of an
excess of ArI ([ArI] Ն 10 ϫ [1a]0). The rate of any single
reaction obeys the mono-exponential law:
Dt ϭ Dϱ ϩ (D0 Ϫ Dϱ) exp(Ϫ kobs t)
The negative activation entropies for the k1A step suggest
where Dt, D0, and Dϱ represent the absorbance of the sys- an essentially associative process. Negative ∆S# values were
tem at time t, at time t ϭ 0 and at the end of the reaction also obtained for the k2A step, but these data should be
respectively, and kobs is the pseudo-first-order rate constant. regarded with caution (see further discussion).
Eur. J. Inorg. Chem. 2004, 732Ϫ742
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
737

B. Crociani, S. Antonaroli, L. Canovese, P. Uguagliati, F. Visentin
FULL PAPER
approach and with reasonable approximations,[23] the rate
law:
d[1a]/dt ϭ Ϫ[1a](k1A ϩ k2A[ArI])
Table 4. Kinetic constants for the oxidative addition of ArI to 1a
in CHCl3 at 25 °C in the presence of variable concentrations of
added dmfu ([1a]0 ϭ 1·10Ϫ4 )
can be obtained which is identical to the experimental
results irrespective of the nature of the reactive intermedi-
[dmfu]/
k1A/ sϪ1
k2A/ Ϫ1sϪ1
0
(5.4Ϯ0.3)ϫ10Ϫ5
(5.9Ϯ0.1)ϫ10Ϫ5
(7.1Ϯ0.4)ϫ10Ϫ5
(7.4Ϯ0.2)ϫ10Ϫ5
(1.24Ϯ0.05)ϫ10Ϫ3 ate. However, a comparison of the ∆G϶ value of 63.4
1·10Ϫ3
3·10Ϫ3
5·10Ϫ3
(1.20Ϯ0.02)ϫ10Ϫ3
(1.09Ϯ0.06)ϫ10Ϫ3
(1.08Ϯ0.04)ϫ10Ϫ3
kJ·molϪ1 {calculated at the coalescence temperature of 56
°C for the dynamic process of 1a in [D8]toluene, in which
the intermediate [Pd(η2-dmfu)(solvent)(κ1-PϪN)] is
formed} and the ∆G϶ value of 95.5 kJ·molϪ1 (calculated
at the same temperature from the activation parameters in
Table 5 for step k1A in toluene) rules out any intermediacy
of [Pd(η2-dmfu)(solvent)(κ1-PϪN)]. In the solvolytic path-
way k1A, a higher energy process is involved which is well
represented by the solvent-assisted dissociation of the η2-
bound dmfu to form the intermediate [Pd(solvent)(PϪN)]
whose enhanced reactivity toward oxidative additions[10,13]
is essentially due to the increased electron density on the
central metal.
Table 5. Kinetic constants and activation parameters (∆H϶/
kJ·molϪ1; ∆S϶/J·molϪ1KϪ1) for the oxidative addition of ArI to 1a
at variable temperature in different solvents in the presence of dmfu
([1a]0 ϭ 1·10Ϫ4 ; [dmfu] ϭ 3·10Ϫ3 )
Solvent
CH3CN
T / °C
k1A/sϪ1
k2A/Ϫ1sϪ1
15
25
35
45
(3.0Ϯ0.1) ϫ 10Ϫ4
(8.3Ϯ0.5) ϫ 10Ϫ4
(2.34Ϯ0.04) ϫ 10Ϫ3
(5.2Ϯ0.2) ϫ 10Ϫ3
(9Ϯ1) ϫ 10Ϫ4
(3.8Ϯ0.7) ϫ 10Ϫ3
(4.8Ϯ0.5) ϫ 10Ϫ3
(1.6Ϯ0.4) ϫ 10Ϫ2
In order to get a better understanding of the second-or-
der step k2A, a kinetic study of the oxidative addition of ArI
to the N-CHMe2 complex 2a was carried out in CH3CN at
45 °C in the presence of dmfu ([2a]0 ϭ 1·10Ϫ4 ; [dmfu] ϭ
3·10Ϫ3 ). The observed rate law is identical to that found
∆H϶ ϭ 71Ϯ21
∆S϶ ϭ Ϫ63Ϯ8
∆H϶ ϭ 59Ϯ12
∆S϶ ϭ Ϫ92Ϯ 38
Toluene
15
25
35
45
(5.2Ϯ0.4) ϫ 10Ϫ5
(1.57Ϯ0.04) ϫ 10Ϫ4
(8.3Ϯ0.8) ϫ 10Ϫ4
(1.39Ϯ0.03) ϫ 10Ϫ3
(2.26Ϯ0.05) ϫ 10Ϫ3
(5.97Ϯ0.05) ϫ 10Ϫ3
(1.06Ϯ0.09) ϫ 10Ϫ2
(3.87Ϯ0.17) ϫ 10Ϫ2
for 1a, with k1A ϭ (2.71Ϯ0.03)ϫ10Ϫ4 sϪ1 and k2A
ϭ
(6.3Ϯ0.4)ϫ10Ϫ4 Ϫ1sϪ1. If compared with the correspond-

ing rate constants for the reaction of 1a under the same
experimental conditions [k1A ϭ (5.2Ϯ0.2)ϫ10Ϫ3 sϪ1 and
∆H϶ ϭ 83Ϯ11
∆S϶ ϭ Ϫ38Ϯ4
∆H϶ ϭ 67Ϯ2
∆S϶ ϭ Ϫ61Ϯ5
k2A ϭ (1.6Ϯ0.4)ϫ10Ϫ2 Ϫ1sϪ1] one can see that for 2a the

k1A value has decreased by one order of magnitude, while
the k2A value has decreased by two orders of magnitude.
These changes are mainly related to electronic factors since
the steric requirements of the N-substituents in these com-
plexes are comparable. Thus, the decrease in k1A can be
explained by the presence of a stronger PdϪη2-dmfu bond
in 2a due to the greater extent of π back-donation (see the
13C NMR spectroscopic data in Table 2) which makes the
olefin dissociation step more difficult. However, the marked
decrease in k2A is in contrast with the k2A step, being a
direct electrophilic attack on the palladium center of the
complexes [Pd(η2-dmfu)(PϪN)]. The presence of a more
electron-donating iminophosphane in 2a increases the elec-
tron density on the central metal and, therefore, would fa-
vour such attack as reported for other palladium(0) com-
plexes when the donor properties of the ancillary ligands
are increased.[3f,24] The changes in the k2A values can be
rationalised if the oxidative addition of ArI in the path k2A
does not occur on the complexes [Pd(η2-dmfu)(PϪN)] but
on the intermediates [Pd(η2-dmfu)(solvent)(κ1-PϪN)]
formed in a pre-equilibrium step (K).
On the basis of the above experimental data, the follow-
ing mechanism can be proposed in which the reaction pro-
ceeds through two parallel pathways to yield the final prod-
uct 1c.
In the step k1A, the interaction with the solvent produces
a highly reactive intermediate which rapidly reacts with ArI
to give 1c. To a first approximation, step k2A may be
conceived as a bimolecular attack by the electrophile ArI
on the starting complex 1a. According to the solution be-
haviour of the complexes [Pd(η2-dmfu)(PϪN)] and to lit-
erature data on the mechanism of oxidative addition of
iodobenzene to [Pd(η2-dba)(LϪLЈ)] (dba ϭ dibenzylidene-
acetone; LϪLЈ ϭ N, P ligand of the type aminophosphane
or oxazolinophosphane),[19] the reactive intermediate may
be formulated as a [Pd(solvent)(PϪN)] species, resulting
from the dissociation of the η2-bound dmfu, or as a [Pd(η2-
dmfu)(solvent)(κ1-PϪN)] species, containing a P-mono-
dentate iminophosphane and resulting from the PdϪN
bond rupture (see Scheme 4). Under the steady state
738
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 732Ϫ742

Oxidative Addition of Alkynyl Stannanes
FULL PAPER
In this case, the observed rate law becomes:
kobs ϭ k1A ϩ k2AЈ·K[ArI]
The measured k2A term can now be represented by the
product k2AЈ·K. Although the K constant cannot be exper-
imentally determined due to the low, undetectable concen-
tration of [Pd(η2-dmfu)(solvent)(κ1-PϪN)] in the NMR
solutions, its value is expected to decrease on going from
1a to 2a due to the presence of a stronger PdϪN bond in
the latter complex. In contrast, it is reasonable to assume
that the kinetic constants k2AЈ are not particularly affected
by the different electronic properties of the N-substituents
in the intermediates with a P-monodentate iminophos-
phane.
Figure 2. Plots of kobs versus the stannane concentration for the
transmetallation of 1c with PhCϵCSnBu3 in different solvents at
25 °C in the presence of fn ([1c]0 ϭ 1·10Ϫ4 ; [fn]ϭ 2·10Ϫ4 )
Transmetallation and Reductive Elimination Reactions
The complexes 1c and 2c react with the stannane
PhCϵCSnBu3 in the presence of an activated olefin accord-
ing to Equation (4).
As reported in Figure 2 for the reactions of 1c, the kobs
values at 25 °C depend on the stannane concentration in a
linear manner so that the observed rate law can be rep-
resented by the relationship:
kobs ϭ k2T[PhCϵCSnBu3]
and the corresponding mechanism may be formulated as
follows:
As reported for palladium complexes with chelating
bidentate
ligands,[3d,9,10]
the
transmetallation
of
[PdI(Ar)(PϪN)] by PhCϵCSnBu3 leads to a labile inter-
mediate [Pd(CϵCPh)(Ar)(PϪN)] which undergoes re-
ductive elimination of PhCϵCAr and η2-coordination of
the olefin to give the product [Pd(η2-ol)(PϪN)]. In the IR
1
spectra (in THF) and in the H, 119Sn and 31P spectra (in
[D8]THF) of the reaction mixtures at different times, only
the starting materials and the final products were ob-
served.[13] No side-product or intermediate of the type
[Pd(CϵCPh)(Ar)(PϪN)] was detected during the course of
the reaction, indicating that the reductive elimination and
η2-coordination of the olefin are faster than the initial
transmetallation step.
A kinetic study of reaction (4) was carried out under
pseudo-first-order conditions using UV/Vis spectroscopy in
different solvents and at different temperatures with an in-
itial concentration [1c]0 or [2c]0 ϭ 1·10Ϫ4 . Preliminary
investigations in which the nature and the concentration of
the activated olefin were changed (ol ϭ dmfu, fn; [ol] in the
range 2·10Ϫ4 to 6·10Ϫ4 ) led to the conclusion that the
reaction rates are influenced neither by the different π-ac-
cepting properties nor by the concentration of the olefin.
These findings along with the previous spectroscopic re-
sults[13] imply that i) transmetallation is the rate-determin-
ing step of the overall reaction (4) and ii) it is possible to
In all the solvents examined, the rate law does not display
a solvolytic path, no intercept being detected in the plots in
Figure 2. The measured k2T values are listed in Table 6.
Table 6. Kinetic constants k2T for the transmetallation of 1c with
PhCϵCSnBu3 in different solvents at 25 °C in the presence of fn
([1c]0 ϭ 1·10Ϫ4 ; [fn]ϭ 2·10Ϫ4 )
Solvent
k2T/ Ϫ1sϪ1
THF
0.0107Ϯ0.0006
0.10Ϯ0.01
0.59Ϯ0.01
31Ϯ1
Toluene
CHCl3
CH3CN[a]
DMF
48Ϯ2
[a]
In the presence of an excess of NEt4I ([NEt4I] ϭ 1·10Ϫ3 ).
As can be seen, a marked solvent effect is present with
choose the more π-accepting olefin fn that will impart a the highest k2T values being observed in the more polar
higher stability to the zero-valent products [Pd(η2- solvents CH3CN and DMF. The kinetic measurements in
fn)(PϪN)] in order to asses unequivocally the mechanistic CH3CN were carried out in the presence of a large excess
path without interference from side-reactions involving de- of NEt4I in order to shift the position of equilibrium (2)
composition.
completely to the left. In the absence of NEt4I and with all
Irrespective of solvent and temperature, each reaction ob- other things being equal, the reaction in CH3CN proceeds
eys the mono-exponential law:
Dt ϭ Dϱ ϩ (D0 Ϫ Dϱ)·exp(Ϫkobs·t)
at a higher rate indicating that the cationic complex
[Pd(Ar)(CH3CN){2-(PPh2)C6H4Ϫ1-CHϭNC6H4OMe-4}]ϩ
Eur. J. Inorg. Chem. 2004, 732Ϫ742
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
739

B. Crociani, S. Antonaroli, L. Canovese, P. Uguagliati, F. Visentin
FULL PAPER
is a more reactive species than the neutral complex 1c. Un- (ii) a parallel path the rate of which is first-order with re-
fortunately, the reactivity of [Pd(Ar)(CH3CN){2- spect to ArI and involves the species [Pd(η2-dmfu)(sol-
(PPh2)C6H4Ϫ1-CHϭNC6H4OMe-4}]ϩ (as CF3SO3 salt) vent)(κ1-PϪN)] formed in a pre-equilibrium dissociation of
Ϫ
could not be determined because extensive decomposition the PdϪN bond.
occurred in the reaction mixture in the absence of the IϪ
ions.
The reaction of [PdI(Ar)(PϪN)] with PhCϵCSnBu3 in
the presence of activated olefins yields the complexes
From kinetic measurements at variable temperature, the [Pd(η2-ol)(PϪN)]. For this reaction, the rate-determining
activation parameters shown in Table 7 were evaluated.
transmetallation step obeys a rate law which is first-order
with respect to the stannane. The kinetic data suggest an
associative SE2 mechanism via an open transition state with
substantial charge separation in highly polar solvents such
as CH3CN and DMF (path b of Scheme 1).
Table 7. Kinetic constants k2T and activation parameters for the
transmetallation of 1c with PhCϵCSnBu3 in different solvents
Solvent
T / °C k2T/ Ϫ1sϪ1
∆H϶/ kJ·molϪ1 ∆S϶/ J·molϪ1KϪ1
From the present study it clearly emerges that the oxidat-
ive addition represents the rate-determining step in the
cross-coupling of PhCϵCSnBu3 with ArI catalysed by
[Pd(η2-dmfu)(PϪN)]. The nature of the solvent is of great
importance for increasing the overall rate of the catalysed
reaction. In this respect, CH3CN (a solvent with good coor-
dinating properties and a high dielectric constant) proves to
be the best suited medium since it enhances the rates of
both the oxidative addition and transmetallation steps in-
volved in the Stille process. Moreover, in the case of trans-
metallation in CH3CN, the absence of added iodide allows
the formation of a significant amount of the cationic species
[Pd(CH3CN)(Ar)(PϪN)]ϩ which increases the apparent
rate of the reaction without leading to secondary decompo-
sition products.
[a]
CHCl3
15
25
35
45
0.334Ϯ0.006
0.59Ϯ0.01
0.80Ϯ0.02
1.00Ϯ0.03
31Ϯ3
23Ϯ1
Ϫ146Ϯ12
Ϫ139Ϯ9
CH3CN[a][b]
15
25
35
45
27Ϯ1
55Ϯ1
68Ϯ4
108Ϯ4
[a] In each kinetic run: [1c]0 ϭ 1·10Ϫ4 ; [fn] ϭ 2·10Ϫ4 . [b] In each
kinetic run: [NEt4I] ϭ 1·10Ϫ3 .
The markedly negative ∆S϶ values clearly point to an
associative SE2 mechanism for the transmetallation step in
both solvents. Because complex 1c has an iodide ligand of
good bridging ability, a cyclic four-center transition state
may be operative in solvents of low polarity (toluene, THF,
CHCl3).[5] However, the strong acceleration observed in
highly polar solvents (CH3CN, DMF) is very much in
favour of an open transition state with substantial charge
separation.[5]
Experimental Section
1H, 13C{1H}, and 31P{1H} NMR spectra were recorded on a
Bruker AM400 spectrometer operating at 400.13, 100.61, and
161.98 MHz, respectively, using solutions of the palladium com-
plexes with concentrations in the range of 1·10Ϫ2 to 2·10Ϫ2 .
Chemical shifts are reported in ppm relative to TMS (1H, 13C) or
In the case of complex 2c, the kinetic study of reaction
(4) was carried out at 25 °C in CHCl3 and CH3CN under
the same experimental conditions as for 1c. In CH3CN, an external 85% H3PO4 (31P). For the 1H NMR spectra at variable
excess of NEt4I was added ([NEt4I] ϭ 1·10Ϫ3 ). The ob-
temperature, the probe temperature was controlled by a standard
unit calibrated with a methanol reference. Coalescence tempera-
tures were determined to within an accuracy of 3 K. The estimated
served kinetic law is identical to that found for 1c, with
k
2T ϭ 0.32Ϯ0.01 Ϫ1sϪ1 in CHCl3 and k2T ϭ 27Ϯ4 Ϫ1sϪ1
error in the calculated ∆G϶ values is 0.6Ϫ0.8 kJ·molϪ1. IR spectra
were carried out using a PerkinϪElmer 983 spectrophotometer.
UV/Vis spectra at variable temperature were recorded on a Peltier
equipped PerkinϪElmer lambda 40 spectrophotometer. Conduc-
tivities were measured with CDM83 and Metrohm 660 conductim-
eters at 25 °C. All solvents were purified and dried before use ac-
cording to standard methods.[25] The reactions were carried out
in CH3CN. When compared with the corresponding values
of 1c in Table 6, one can see that the transmetallation step
proceeds at comparable rates for both complexes. Presum-
ably, the lower electrophilic character of the palladium
center in 2c, due to the greater electron-donating properties
of the N-CHMe2 substituent, is compensated for by the in-
creased lability of the iodide ligand as can be inferred from under N2 unless otherwise stated. The compounds PhCϵCSnBu3,
dimethyl fumarate, fumaronitrile, AgBF4, and AgCF3SO3 were
commercial samples and were used without further purification.
The aryl iodide IC6H4CF3-4 was distilled at reduced pressure be-
fore use. The iminophosphane 2-(PPh2)C6H4Ϫ1-CHϭNR [R ϭ
CHMe2, C6H4OMe-4, C6H3Me2-2,6, C6H3(CHMe2)2-2,6, bornyl]
and the complexes [Pd(η2-dmfu)(PϪN)] (1a, 2a), [Pd(η2-fn)(PϪN)]
(1b, 2b, 5b), [PdI(C6H4CF3-4)(PϪN)] (1c, 2c) were prepared by
published methods.[11,13,26] The complex Pd2(dba)3·CHCl3 was syn-
thesised as described in the literature.[27]
the difference in the KS constants of the equilibrium 2 for
these complexes.
Conclusion
From the solution behaviour of the complexes [Pd(η2-
dmfu)(PϪN)] and kinetic data for the oxidative addition
of ArI (Ar ϭ C6H4CF3-4), a mechanism can be proposed
consisting of (i) a solvolytic path the rate of which is zero-
order with respect to ArI and involves the reactive species
Preparations
[Pd(η2-dmfu)(P؊N)] (3a, 4a): The required iminophosphane
[Pd(solvent)(PϪN)] formed in the olefin dissociation and (1.0 mmol) and dmfu (173 mg, 1.2 mmol) were added to a stirred
740  2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjic.org
Eur. J. Inorg. Chem. 2004, 732Ϫ742

Oxidative Addition of Alkynyl Stannanes
FULL PAPER
suspension of Pd2(dba)3·CHCl3 (1.035 g, 1 mmol) in dry acetone
(30 mL). Stirring was continued for 2 h and the solution was then
evaporated to dryness at reduced pressure. The solid residue was
extracted with CH2Cl2 (20 mL) in the presence of activated char-
coal. After filtration through celite, the clear red-brown solution
was concentrated to a small volume (ca. 3 mL) and diluted with
Et2O to precipitate the products as yellow-orange solids. The com-
plexes were purified by further precipitation from a toluene/n-hex-
ane mixture.
JH,H ϭ 6.4 Hz, 6 H, CH3) ppm. 31P NMR (CDCl3, 25 °C): δ ϭ
32.70 (s) ppm.
Determination of the Equilibrium Constant KE: To a thermo-
statically controlled (25 °C) solution of the complex under study
(50 mL, ca. 1·10Ϫ4 ) in the appropriate solvent were added suc-
cessive micro-aliquots of the designated olefin and the absorbance
changes were recorded in the range of 500Ϫ280 nm at 25 °C. The
resultant array of absorbance data vs. the total concentration of
added olefin was treated according to a published mathematical/
statistical model.[16,17]
3a: Yield 83% (535 mg). C33H32NO4PPd (644.0): calcd. C 61.54, H
5.01, N 2.17; found C 61.56, H 5.10, N 2.12%. IR (Nujol): ν˜ ϭ
Determination of the Equilibrium Constant KS: Several solutions of
the complexes 1c and 2c at different concentrations (concentration
intervals: 2.3·10Ϫ5 to 2.4·10Ϫ3 , 1c; 2.5·10Ϫ5 to 2.6·10Ϫ3 , 2c) in
freshly distilled CH3CN were prepared and thermostatically con-
trolled at 25 °C. The specific conductivities were determined for
each solution and the related molar conductivities were treated ac-
cording to the method of Fuoss and Shedlovsky.[28]
1
1670 vs (CϭO), 1615 ms (CϭN) cmϪ1. H NMR (CDCl3, 25 °C):
δ ϭ 8.06 (d, JP,H ϭ 3.2 Hz, 1 H, NϭCH), 7.6Ϫ6.9 (m, 17 H, C6H3
ϩ C6H4 ϩ C6H5), 4.24 (dd, JH,H ϭ 10.0, JP,H ϭ 2.4 Hz, 1 H, ϭ
CH cis to P), 3.77 (dd, JH,H ϭ JP,H ϭ 10.2 Hz, 1 H, ϭCH trans to
P), 3.18 (s, 6 H, OCH3), 2.16 (s, 3 H, CH3), 1.90 (s, 3 H, CH3)
ppm. 31P NMR (CDCl3, 25 °C): δ ϭ 19.69 (s) ppm.
4a: Yield 80% (560 mg). C37H40NO4PPd (700.1): calcd. C 63.47, H
5.76, N 2.00; found C 63.20, H 5.85, N 1.95%. IR (Nujol): ν˜ ϭ
A UV/Vis spectroscopic check was carried out as follows: the molar
extinction coefficients (εk) of the starting complex (1c or 2c) were
determined from their solutions in the presence of a strong excess
of NEt4I, and those (εs) of the cationic species
[Pd(CH3CN)(Ar)(PϪN)]ϩ were determined from the solutions of
their CF3SO3 or BF4 salts synthesised independently. The con-
sistency of the conductivity determined KS values and the reliability
of the results was thus checked by a single spectrophotometric
experiment by measuring the absorbance of a 1·10Ϫ4  solution of
1c (or, 2c) in CH3CN. The ensuing absorbance values confirm, in
both cases, the independently determined KS constants.
1
1685 vs (CϭO), 1622 ms (CϭN) cmϪ1. H NMR (CDCl3, 25 °C):
δ ϭ 8.06 (d, JP,H ϭ 2.1 Hz, 1 H, NϭCH), 7.6Ϫ6.9 (m, 17 H, C6H3
ϩ C6H4 ϩ C6H5), 4.25 (dd, JH,H ϭ 10.4, JP,H ϭ 2.5 Hz, 1 H, ϭ
CH cis to P), 3.49 (dd, JH,H ϭ JP,H ϭ 10.4 Hz, 1 H, ϭCH trans to
P), 3.37 (spt, JH,H ϭ 6.8 Hz, 1 H, CHMe2), 3.24 (s, 3 H, OCH3),
2.93 (s, 3 H, OCH3), 2.64 (spt, JH,H ϭ 6.8 Hz, 1 H, CHMe2), 1.45
(d, JH,H ϭ 6.8 Hz, 3 H, CH3), 1.09 (d, JH,H ϭ 6.8 Hz, 3 H, CH3),
0.98 (d, JH,H ϭ 6.8 Hz, 3 H, CH3), 0.56 (d, JH,H ϭ 6.8 Hz, 3 H,
CH3) ppm. 31P NMR (CDCl3, 25 °C): δ ϭ 22.36 (s) ppm.
Ϫ
Ϫ
[Pd(CH3CN)(C6H4CF3-4)[2-(PPh2)C6H4؊1-CH؍NC6H4OMe-
4)]CF3SO3: AgCF3SO3 (31 mg, 0.12 mmol) dissolved in CH3CN
(1 mL) was added to a solution of 1c (93 mg, 0.12 mmol) in a
CH2Cl2/CH3CN mixture (7:1 v/v, 8 mL). After standing for 2 h in
the dark, the insoluble AgI was removed by filtration and the re-
sultant solution was evaporated to dryness at reduced pressure. The
solid residue was redissolved in CH2Cl2 (10 mL). Addition of acti-
vated charcoal and filtration gave a clear solution which was con-
centrated to small volume (ca. 2 mL), then diluted with Et2O to
precipitate the product as a yellow microcrystalline solid. Yield
Kinetic Measurements: The kinetics of the oxidative addition and
transmetallation reactions were followed by UV/Vis spectroscopy
upon addition of known micro-aliquots of a pre-thermostatically
controlled solution of the appropriate reagent (ArI, or
PhCϵCSnBu3 and olefin) to a thermostatically controlled solution
in the appropriate solvent of the complex under study {[Pd(η2-
dmfu)(PϪN)] or [PdI(Ar)(PϪN)} at the suitable temperature. The
concentration of the reactant ensured in any case pseudo-first-or-
der conditions and the reactions were followed by recording spec-
tral changes in the suitable wavelength range or at a fixed wave-
length. Mathematical and statistical data analysis was carried out
on a PC equipped with a locally adapted non-linear regression pro-
cedure.
69% (69 mg). ΛM (1·10Ϫ3 , CH3CN) ϭ 134.4 ohmϪ1cm2molϪ1
.
C36H29F6N2O4PPdS (837.0): calcd. C 51.65, H 3.49, N 3.35; found
C 51.81, H 3.55, N 3.45%. IR (CHCl3): ν˜ ϭ 2289 ms (CϵN), 1636
1
m (CϭN) cmϪ1. H NMR (CDCl3, 25 °C): δ ϭ 8.33 (s, 1 H, Nϭ
CH), 7.9Ϫ7.7 (m, 2 H, o-disubstituted C6H4), 7.59 (m, 1 H, o-
disubstituted C6H4), 7.5Ϫ7.2 (m, 15 H, o-disubstituted C6H4 ϩ p-
disubstituted C6H4 ϩ C6H5), 7.0Ϫ6.9 (m, 4 H, meta protons of p-
disubstituted C6H4), 3.84 (s, 3 H, OCH3), 2.30 (s, 3 H, CH3CN)
ppm. 31P NMR (CDCl3, 25 °C): δ ϭ 34.74 (s) ppm.
[1]
[1a] J. K. Stille, Angew. Chem. Int. Ed. Engl. 1986, 25, 508Ϫ524.
[1b]
[1c]
T. N. Mitchell, Synthesis 1992, 803Ϫ815.
V. Farina, in:
Comprehensive Organometallic Chemistry II (Eds: E. W. Abel,
F. G. A. Stone, G. Wilkinson), Pergamon, Oxford, 1995, 12,
chapter 3.4, p. 161Ϫ240. [1d]V. Farina, G. P. Roth, Adv. Meta-
[Pd(CH3CN)(C6H4CF3-4)[2-(PPh2)C6H4؊1-CH؍NCHMe2)]BF4:
AgBF4 (49 mg, 0.25 mmol) was added to a solution of 2c (177 mg,
0.25 mmol) in CH3CN (10 mL). The mixture was worked up as
described above for the preparation of the analogous cationic com-
plex to give the product as a yellow microcrystalline solid. Yield
[1e]
lorg. Chem. 1996, 5, 1Ϫ53.
T. N. Mitchell, in: Metal-Cata-
lyzed Cross-Coupling Reactions (Eds. F. Diederich, P. J. Stang),
Wiley-VCH, Weinheim, 1998, p. 167Ϫ202.
[1f]
V. Farina, V.
Krishnamurthy, V. J. Scott, The Stille Reaction, John Wiley &
Sons, New York, 1998.
[2]
[2a] C. Amatore, A. Jutand, A. Suarez, J. Am. Chem. Soc. 1993,
73% (130 mg). ΛM (1·10Ϫ3 , CH3CN) ϭ 147.0 ohmϪ1cm2molϪ1
.
[2b]
115, 9531Ϫ9541.
C. Mateo, D. J. Cardenas, C. Fernandez-
C31H29BF7N2PPd (710.7): calcd. C 52.38, H 4.11, N 3.94; found C
52.15, H 4.02, N 3.86%. IR (CHCl3): ν˜ ϭ 2316 mw, 2289 mw
(CϵN), 1634 ms (CϭN) cmϪ1. 1H NMR (CDCl3, 25 °C): δ ϭ 8.36
(s, 1 H, NϭCH), 7.90 (m, 1 H, o-disubstituted C6H4), 7.78 (m, 1
H, o-disubstituted C6H4), 7.56 (m, 1 H, o-disubstituted C6H4),
7.5Ϫ7.2 (m, 13 H, o-disubstituted C6H4 ϩ p-disubstituted C6H4 ϩ
C6H5), 6.95 (m, 2 H, meta protons of C6H4CF3-4), 4.52 (spt,
JH,H ϭ 6.4 Hz, 1 H, CHMe2), 2.30 (s, 3 H, CH3CN), 1.37 (d,
Rivas, A. M. Echavarren, Chem. Eur. J. 1996, 2, 1596Ϫ1606.
[3] [3a]
A. Gillie, J. K. Stille, J. Am. Chem. Soc. 1980, 102,
[3b]
4933Ϫ4941.
K. Tatsumi, R. Hoffmann, A. Yamamoto, J.
[3c]
K. Stille, Bull. Chem. Soc. Jpn. 1981, 54, 1857Ϫ1867.
P. J.
Stang, M. H. Kowalski, M. D. Schiavelli, D. J. Longford, J.
Am. Chem. Soc. 1989, 111, 3347Ϫ3356. [3d] J. M. Brown, N. A.
[3e]
Cooley, Organometallics 1990, 9, 353Ϫ359.
Krishnan, J. Am. Chem. Soc. 1991, 113, 9585Ϫ9595.
V. Farina, B.
[3f]
V.
Eur. J. Inorg. Chem. 2004, 732Ϫ742
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
741

B. Crociani, S. Antonaroli, L. Canovese, P. Uguagliati, F. Visentin
FULL PAPER
[19]
Farina, B. Krishnan, D. R. Marshall, G. P. Roth, J. Org. Chem.
C. Amatore, A. Fuxa, A. Jutand, Chem. Eur. J. 2000, 6,
[3g]
1993, 58, 5434Ϫ5444.
Soc. 1995, 117, 11598Ϫ11599.
J. Louie, J. F. Hartwig, J. Am. Chem.
1474Ϫ1482.
[3h]
[20]
´
C. Amatore, E. Carre, A.
The activation free energies have been estimated from the co-
Jutand, H. Tanaka, Q. Ren, S. Torii, Chem. Eur. J. 1996, 2,
alescence temperature (Tc) using the equation ∆G϶
ϭ
[3i]
957Ϫ966.
C. Amatore, G. Broeker, A. Jutand, F. Khalil, J.
ϪRTcln[πh∆ν/(21/2kTc)]. The palladium(0) species are fairly
stable towards decomposition when the temperature is in-
creased. A slight decomposition with formation of free dmfu
is observed at the higher temperatures. The coalescence tem-
peratures are, however, scarcely affected because the exchange
between free and coordinated olefin is slow. Addition of free
dmfu to the solution of 1a at 56 °C actually caused only a
slight broadening of the methyl signals of the free and coordi-
nated olefin.
[3j]
Am. Chem. Soc. 1997, 119, 5176Ϫ5185.
J. F. Hartwig, An-
gew. Chem. Int. Ed. 1998, 37, 2046Ϫ2067.
E. Negishi, T. Takahashi, S. Baba, D. E. Van Horn, N. Oku-
[4]
kado, J. Am. Chem. Soc. 1987, 109, 2393Ϫ2401.
[5] [5a]
A. L. Casado, P. Espinet, J. Am. Chem. Soc. 1998, 120,
[5b]
8978Ϫ8995.
A. L. Casado, P. Espinet, A. M. Gallego, J.
Am. Chem. Soc. 2000, 122, 11771Ϫ11782.
[6]
A. Ricci, F. Angelucci, M. Bassetti, C. Lo Sterzo, J. Am. Chem.
Soc. 2002, 124, 1060Ϫ1071.
[21]
R. van Asselt, C. J. Elsevier, W. J. J. Smeets, A. L. Spek, Inorg.
Chem. 1994, 33, 1521Ϫ1531.
[7] [7a]
C. Amatore, A. Bucaille, A. Fuxa, A. Jutand, G. Meyer, A.
[22] [22a] D. M. Himmeblau, Process Analysis by Statistical Methods,
John Wiley & Sons, New York, 1970. [22b]P. Uguagliati, R. A.
Michelin, U. Belluco, R. Ros, J. Organomet. Chem. 1979, 169,
[7b]
N. Ntepe, Chem. Eur. J. 2001, 7, 2134Ϫ2142.
A. A. Bahsoun, A. Jutand, G. Meyer, A. Ndedi Ntepe, L.
C. Amatore,
Ricard, J. Am. Chem. Soc. 2003, 125, 4212Ϫ4222.
J. A. Casares, P. Espinet, G. Salas, Chem. Eur. J. 2002, 8,
4844Ϫ4853.
115Ϫ122.
[8]
[23]
The steady-state approach leads to the equation: kobs ϭ k1A
k2A[ArI] Ϫ k1A
Ϫ1A/(kϪ1A ϩ k3A[ArI]) for [Pd(η2-dmfu)(solv-
ent)(κ1-PϪN)] as intermediate, or to the equation: kobs ϭ k1A
ϩ k2A[ArI] Ϫ k1A Ϫ1A[dmfu]/(kϪ1A[dmfu] ϩ k3A[ArI]) for
ϩ
k
[9]
`
A. L. Casado, P. Espinet, A. M. Gallego, J. M. Martınez-Ilard-
uya, Chem. Commun. 2001, 339Ϫ340.
k
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[Pd(solvent)(PϪN)] as intermediate. These equations become
the observed rate law: kobs ϭ k1A ϩ k2A[ArI] when the reason-
able approximations (kϪ1A ϩ k3A[ArI]) ϾϾ k1AkϪ1A or
(kϪ1A[dmfu] ϩ k3A[ArI]) ϾϾ k1AkϪ1A[dmfu] are operative. On
the other hand, the pre-equilibrium (K1) condition:
E. Shirakawa, T. Hiyama, J. Organomet. Chem. 1999, 576,
169Ϫ178, and references cited therein.
S. Antonaroli, B. Crociani, J. Organomet. Chem. 1998, 560,
137Ϫ146.
A. Scrivanti, U. Matteoli, V. Beghetto, S. Antonaroli, B. Croci-
ani, Tetrahedron 2002, 58, 6881Ϫ6886.
B. Crociani, S. Antonaroli, V. Beghetto, U. Matteoli, A. Scriv-
anti, Dalton Trans. 2003, 2194Ϫ 2202.
G. Bandoli, A. Dolmella, L. Crociani, S. Antonaroli, B. Croci-
ani, Transition Met. Chem. 2000, 25, 17Ϫ25.
L. Crociani, G. Bandoli, A. Dolmella, M. Basato, B. Corain,
Eur. J. Inorg. Chem. 1998, 1811Ϫ1820.
L. Canovese, F. Visentin, P. Uguagliati, B. Crociani, J. Chem.
Soc., Dalton Trans. 1996, 1921Ϫ1925.
L. Canovese, F. Visentin, G. Chessa, P. Uguagliati, A. Dol-
.
was never taken into account since for both intermediates it
leads to rate laws which do not correspond to the experimental
one. As an example, for [Pd(η2-dmfu)(solvent)(κ1-PϪN)] as the
intermediate the rate law should become: kobs
(k2Aϩk3AK1)[ArI]/(1ϩK1).
ϭ
[24]
[25]
[26]
[27]
M. Huser, M.-T. Youinou, J. A. Osborn, Angew. Chem. Int. Ed.
Engl. 1989, 28, 1386Ϫ1388.
W. L. Armarengo, D. D. Perrin, Purification of Laboratory
Chemicals, 3rd ed., Pergamon, New York, 1988.
B. Crociani, S. Antonaroli, L. Canovese, F. Visentin, P. Uguag-
liati, Inorg. Chim. Acta 2001, 315, 172Ϫ182.
mella, J. Organomet. Chem. 2000, 601, 1Ϫ15.
[18] [18a]
´
B. Jedlicka, R. E. Rülke, W. Weissensteiner, R. Fernandez-
´
´
´
Galan, F. A. Jalon, B. R. Manzano, J. Rodrıguez-de la Fuente,
N. Valdman, H. Kooijman, A. L. Spek, J. Organomet. Chem.
[18b]
´
´
´
1966, 516, 97Ϫ110.
R. Fernandez-Galan, F. A. Jalon, B.
T. Ukai, H. Kawazura, Y. Ishii, J. J. Bonnet, J. A. Ibers, J.
Organomet. Chem. 1974, 65, 253Ϫ266.
´
R. Manzano, J. Rodrıguez-de la Fuente, M. Vrahani, B. Jed-
[28] [28a]
licka, W. Weissensteiner, Organometallics 1997, 16, 3758Ϫ3768.
R. M. Fuoss, T. Shedlovsky, J. Am. Chem. Soc. 1949, 120,
[18c]
[28b]
K. Selvakumar, M. Valentini, M. Wörle, P. S. Pregosin, A.
1496Ϫ1498.
7, 260Ϫ277.
J. Barthel, Angew. Chem. Int. Ed. Engl. 1968,
[28c]
[18d]
Albinati, Organometallics 1999, 18, 1207Ϫ1215.
F. A. Ja-
´
B. Crociani, F. Di Bianca, P. Uguagliati, L.
´
´
lon, B. R. Manzano, F. Gomez-de la Torre, A. M. Lopez-Ag-
Canovese, A. Berton, J. Chem. Soc., Dalton Trans. 1991,
71Ϫ79.
´
´
enjo, A. M. Rodrıguez, W. Weissensteiner, T. Sturm, J. Mahıa,
M. Maestro, J. Chem. Soc., Dalton Trans. 2001, 2417Ϫ2424.
Received June 18, 2003
[18e]
M. Zehnder, M. Neuburger, S. Schaffner, M. Jufer, D. A.
Early View Article
Plattner, Eur. J. Inorg. Chem. 2002, 1511Ϫ1517.
Published Online January 2, 2004
742
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 732Ϫ742