Comparative reactivity of palladium(0) complexes generated in situ in mixtures of triphenylphosphine or tri-2-furylphosphine and Pd(dba)2

Article abstract of DOI:10.1021/om971064u

For PPh₃, mixtures of Pd(dba)₂ and nTFP (TFP = tri-2-furylphosphine, n ≥ 2) in DMF and THF (S) lead to the formation of Pd(dba)(TFP)₂, SPd(TFP)₃ in equilibrium with SPd-(TFP)₂. The substitution of dba by the phosphine in Pd(dba)L₂ to form SPdL₃ is easier for L = TPF than for L = PPh₃. The less ligated complex SPd(TFP)₂ is the reactive species in the oxidative addition with phenyl iodide. In THF, {Pd(dba)₂ + nTFP}, a mixture often used as a catalyst promoter in several synthetic organic reactions, is found to be less reactive than {Pd(dba)₂ + nPPh₃} for small values of n (n = 2 or 4) whereas it is more reactive for higher values of n (n > 6). Conversely, in DMF, {Pd(dba)₂ + nTFP} is always found to be more reactive than {Pd(dba)₂ + nPPh₃} whatever n (n ≥ 2).

Full text of DOI:10.1021/om971064u

2958
Organometallics 1998, 17, 2958-2964
Com p a r a tive Rea ctivity of P a lla d iu m (0) Com p lexes
Gen er a ted in Situ in Mixtu r es of Tr ip h en ylp h osp h in e or
Tr i-2-fu r ylp h osp h in e a n d P d (d ba )2
Christian Amatore,*,† Anny J utand,*,† Gilbert Meyer,† Hamid Atmani,‡
Fouad Khalil,§ and Fouad Ouazzani Chahdi§
De´partement de Chimie, URA 1679, Ecole Normale Supe´rieure, 24 Rue Lhomond,
75231 Paris Cedex 5, France, Faculte´ des Sciences, Laboratoire de Chimie Physique Applique´e,
Universite´ Moulay Ismail, BP 4010, Beni M’Hamed, Meknes, Morocco, and Faculte´ des
Sciences et Techniques, Laboratoire de Chimie Organique, Universite´ Sidi Mohamed Ben
Abdellah, BP 2202, Fes Saiss, Morocco
Received December 3, 1997
For PPh3, mixtures of Pd(dba)2 and nTFP (TFP ) tri-2-furylphosphine, n g 2) in DMF
and THF (S) lead to the formation of Pd(dba)(TFP)2, SPd(TFP)3 in equilibrium with SPd-
(TFP)2. The substitution of dba by the phosphine in Pd(dba)L2 to form SPdL3 is easier for
L ) TPF than for L ) PPh3. The less ligated complex SPd(TFP)2 is the reactive species in
the oxidative addition with phenyl iodide. In THF, {Pd(dba)2 + nTFP}, a mixture often
used as a catalyst promoter in several synthetic organic reactions, is found to be less reactive
than {Pd(dba)2 + nPPh3} for small values of n (n ) 2 or 4) whereas it is more reactive for
higher values of n (n > 6). Conversely, in DMF, {Pd(dba)2 + nTFP} is always found to be
more reactive than {Pd(dba)2 + nPPh3} whatever n (n g 2).
In tr od u ction
addition with aryl halide. In DMF and THF (noted S
hereafter), the following mechanism has been estab-
lished for monodentate ligands L such as triarylphos-
phines (Scheme 1).2,5
Among the various precursors of palladium(0) com-
plexes, mixtures of Pd(0)(dba)2 (dba ) trans,trans-
dibenzylidenacetone) and phosphine ligands afford ef-
ficient palladium(0) catalysts.1-5 We have previously
reported a mechanistic investigation concerning the
effective palladium(0) catalytic species generated in situ
in a mixture of Pd(dba)2 with triphenylphosphine,2 para-
substituted triphenylphosphines,5 and bidentate phos-
phine ligands.4 As a result, it has been shown that dba
is not a labile ligand, since a stable complex Pd(dba)L2
is always formed as the major complex. This complex
is involved in an equilibrium with a less ligated complex
PdL2, which is the more reactive species in the oxidative
Sch em e 1
Pd(0)(dba)2 + 2L f Pd(0)(dba)L2 + dba
(1)
(2)
Pd(0)(dba)L2 + S U SPd(0)L2 + dba
[SPdL2][dba]
K1 )
[Pd(dba)L2]
Pd(0)(dba)L2 + L U SPd(0)L3 + dba
[SPdL3][dba]
K0 )
(3)
[Pd(dba)L2][L]
* To whom all correspondence should be addressed.
† Ecole Normale Supe´rieure.
SPd(0)L3 U SPd(0)L2 + L
[SPdL2][L]
‡ Universite´ Moulay Ismail.
§ Universite´ Sidi Mohamed Ben Abdellah.
(1) For the use of Pd(dba)2 with bidentate ligands, see ref 10 in ref
2 of this paper.
(2) Amatore, C.; J utand, A.; Khalil, F.; M′Barki, M. A.; Mottier, L.
Organometallics 1993, 12, 3168.
(3) For the use of Pd(dba)2 with monodentate ligands, see ref 11 in
ref 2 of this paper.
K2 )
) K1/K0 (4)
[SPdL3]
SPd(0)L2 + ArX f ArPdXL2 + S
k
(5)
As previously established,2,5 the overall reactivity of
{Pd(dba)2 + nL} in the oxidative addition with an aryl
halide depends (i) on the value of the intrinsic rate
constant k of the elementary step of the oxidative
addition of the reactive species SPd(0)L2 (eq 5) (more
basic the phosphine, higher k) and (ii) on the concentra-
tion of the active species itself, viz. SPd(0)L2. This
concentration is controlled by the values of the two
equilibrium constants K1 and K0 (eqs 2-4). Since the
three species SPd(0)L3, Pd(0)(dba)L2, and SPd(0)L2 are
interconnected by three equilibria (eqs 2-4), only two
equilibrium constants such as K1 and K0 are required
(4) Amatore, C.; Broeker, G.; J utand, A.; Khalil, F. J . Am. Chem.
Soc. 1997, 119, 5176.
(5) Amatore, C.; J utand, A.; Meyer, G. Inorg. Chim. Acta 1998, 273,
76.
(6) For the use of Pd(dba)2 + nTPF, see: (a) Farina, V.; Baker, S.
R.; Begnini, D.; Sapino, C., J r. Tetrahedron Lett. 1988, 29, 5739. (b)
Farina, V.; Baker, S. R.; Sapino, C., J r. Tetrahedron Lett. 1988, 29,
6043. (c) Obora, Y.; Tsuji, Y.; Kobayashi, M.; Kawamura, T. J . Org.
Chem. 1995, 60, 4647. (d) Traumer, F.; Le Floch, P.; Lefour, J . M.;
Ricard, L.; Mathey, F. Synthesis 1995, 717. (e) Castro, J .; Balme, G.;
Gore, J . J . Chem. Res. 1995, 12, 504. (f) Moreno-Manas, M.; Perez,
M.; Pleixats, R. J . Org. Chem. 1996, 61, 2346. (g) Rottaelander, M.;
Palmer, N.; Knochel, P. Synlett 1996, 6, 573. (h) Stevenson, T. M.;
Prasad, A. S. B.; Citineni, J . R.; Knochel, P. Tetrahedron Lett. 1996,
37, 8375. (i) Tamao, K.; Ohno, S.; Yamagushi, S. J . Chem. Soc., Chem.
Commun. 1996, 1873.
S0276-7333(97)01064-9 CCC: $15.00 © 1998 American Chemical Society
Publication on Web 06/09/1998

Reactivity of Palladium(0) Complexes
Organometallics, Vol. 17, No. 14, 1998 2959
Ta ble 1. 31P NMR Ch em ica l Sh iftsa of th e
P a lla d iu m (0) Com p lex Gen er a ted in Situ in
Mixtu r es of P d (d ba )2 a n d 2TF P
(n g 2) in DMF a n d THF a s In vestiga ted by 31P
NMR Sp ectr oscop y. In DMF, a set of 2 broad signals
δ1 and δ2 of equal magnitude were observed in the 31P
NMR spectrum of a mixture of Pd(dba)2 + 2TFP while
the signal δ0 of the free ligand was not detected (Table
1), eq 6. As already established for PPh3,2,8 the presence
Pd(0)(dba)(TFP)2
solvent
TFP δ0
δ1
δ2
PhPdI(TFP)2 δ3
DMF
THF
-76.19
-75.24
-36.26
-36.54
-33.71
-33.84
-28.36
-27.83
Pd(0)(dba)2 + 2TFP f Pd(0)(dba)(TFP)2 + dba (6)
31P NMR chemical shifts of the complex PhPdI(TFP)2 resulting
from the oxidative addition with PhI. In ppm vs H3PO4 as an
external reference.
a
of two different phosphorus atoms is due to a monoli-
gation of the ligand dba in the complex Pd(dba)(TFP)2.
However, the signals were too broad (∆ν1/2 ) 85 Hz) to
allow determination of the coupling constant J PP. In
to express the concentration of SPd(0)L2 comparatively
to that of SPd(0)L3 and Pd(0)(dba)L2. Depending on the
basic properties of the phosphines, evolution of the
concentration and of the intrinsic reactivity of SPdL2
(k in eq 5) could be antagonist, leading to nonlinear
correlation of the reactivity with the basicity of the
phosphine, as established for para-substituted tri-
phenylphosphines.5
Catalytic reactions involve successive elemental steps
(oxidative addition, ligand exchange, reductive elimina-
tion, etc.) whose respective rates depend on the nature
of the ligand with, as a consequence, a possible inversion
of the rate-determining step. In palladium(0)-catalyzed
reactions involving aryl halides, it is generally admitted
that the first step of the catalytic cycle is oxidative
addition of a palladium(0) complex with the aryl halide.
It is, therefore, of interest to compare the reactivity of
the oxidative addition of the palladium(0) complexes
generated in situ in mixtures of Pd(dba)2 with phos-
phines ligands as a function of the ligand. In 1988, a
new ligand tri-2-furylphosphine (TFP) was tested by
the presence of 10 equiv of phenyl iodide, a new signal
at δ3 appeared instead of δ1 and δ2 (Table 1). This signal
characterizes the trans complex PhPdI(TPF)2 formed in
the oxidative addition (eq 7).9
Pd(0)(dba)(TFP)2 + PhI f PhPdI(TPF)2 + dba (7)
In the mixtures of Pd(dba)2 + nTFP (n > 4), the two
signals at δ1 and δ2 disappeared and a broad signal was
detected at δ4 (e.g., n ) 8, δ4 ) -64.46 ppm, ∆ν1/2
)
6
Farina6a in reactions catalyzed by mixtures of Pd(dba)2
202 Hz). Despite the excess of ligand, the signal of the
free phosphine at δ0 was never observed. This means
that addition of more than 2 equiv of TFP to Pd(dba)2
affords a palladium(0) complex involved in an equilib-
rium with the ligand (eqs 8 and 9). In the presence of
7
or Pd2(dba)3 and phosphine ligands. More efficient
catalytic systems were very often observed when switch-
ing from triphenylphosphine to tri-2-furylphosphine.6,7
We wish, therefore, to report our investigation on the
comparative reactivity of {Pd(dba)2 + nTFP} with that
of {Pd(dba)2 + nPPh3}2 in oxidative addition with
phenyl iodide. In DMF, {Pd(dba)2 + nTFP} is always
more reactive than {Pd(dba)2 + nPPh3} whatever n (n
g 2). In THF, {Pd(dba)2 + nTFP} is more reactive than
{Pd(dba)2 + nPPh3} when n > 6 whereas it is less
reactive when n < 6. This may have important conse-
quences on the kinetics of the overall catalytic reaction.
Pd(0)(dba)(TFP)2 + TFP U
TFP
SPd(0)(TFP)3 + dba
K0
(8)
(9)
TFP
SPd(0)(TPF)3 U SPd(0)(TPF)2 + TFP
K2
phenyl iodide, the signals δ3 of PhPdI(TFP)2 and δ0 of
the free ligand were detected. We noticed that in
mixtures of Pd(dba)2 + 4TFP, the two signals of Pd-
(dba)(TFP)2 were no longer observed, but surprisingly,
no other signals could be detected in the 31P NMR
spectrum. However, after addition of 10 equiv of PhI
into the NMR tube, the signals δ3 of PhPdI(TFP)2 and
δ0 of free TFP were observed, suggesting that the signal
of the palladium(0) formed from Pd(dba)2 + 4TFP was
so broad that it could not be detected.
Resu lts a n d Discu ssion
Id en tifica tion of th e P a lla d iu m (0) Com p lexes
F or m ed in Situ in a Mixtu r e of P d (d ba )2 + n TF P
(7) For the use of Pd2(dba)3 + nTPF, see: (a) Farina, V.; Hauck, S.
I. Synlett 1991, 3, 157. (b) Farina, V.; Krishnan, B. J . Am. Chem. Soc.
1991, 113, 9585. (c) Farina, V.; Firestone, R. A. Tetrahedron 1993, 49,
803. (d) Farina, V.; Krishna, B.; Marshall, D. R.; Roth, G. P. J . Org.
Chem. 1993, 58, 5434. (e) Siesel, D. A.; Staley, S. W. Tetrahedron Lett.
1993, 34, 3679. (f) Siesel, D. A.; Staley, S. W. J . Org. Chem. 1993, 58,
7870. (g) Badone, D.; Cardamone, R.; Guzzi, U. Tetrahedron Lett. 1994,
35, 5477. (h) Busacca, C. A.; Swestock, J .; J ohnson, R. E.; Bailey, T.
R.; Musza, L.; Rodger, C. A. J . Org. Chem. 1994, 59, 7553. (i) Yang,
Y.; Wong, H. N. C. Tetrahedron Lett. 1994, 50, 9583. (j) Casson, S.;
Kocienski, P.; Reid, G.; Smith, N.; Street, J . M.; Webster, M. Synthesis
1994, 1301. (k) Wagner, R. W.; J ohnson, T. E.; Li, F.; Lindsey, J . S. J .
Org. Chem. 1995, 60, 5266. (l) Caldirola, P.; Chowdury, R.; J ohansson,
A. M.; Hacksell, U. Organometallics 1995, 14, 3897. (m) Koo, S.;
Liebeskind, L. S. J . Am. Chem. Soc. 1995, 117, 3389. (n) Goux, C.;
Massacret, M.; Lhoste, P.; Sinou, D. Organometallics 1995, 14, 4585.
(o) Bao, Z.; Chan, W. K.; Yu, L. J . Am. Chem. Soc. 1995, 117, 12426.
(p) Attwood, M. R.; Raynham, T. M.; Smyth, D. G.; Stephenson, G. R.
Tetrahedron Lett. 1996, 37, 2731. (q) Wolfe, J . P.; Rennels, R. A.;
Buchwald, S. L. Tetrahedron 1996, 52, 7525. (r) Fournier-Nguefack,
C.; Lhoste, P.; Sinou, D. Tetrahedron 1997, 53, 4353.
The same investigation was undertaken in THF
(Table 1) but will not be discussed here since the 31P
NMR spectra were very similar to those reported by
Farina for mixtures of Pd2(dba)3 and 4TFP.7b
(8) Herrmann, W. A.; Thiel, W. R.; Brossmer, C.; O¨ fele, K.; Prier-
meier, T.; Scherer, W. J . Organomet. Chem. 1993, 461, 51.
(9) (a) The complex PhPdI(TPF)2 was also slowly formed by addition
of 2 equiv of TFP to a solution of PhPdI(PPh3)2 according to the
reaction: PhPdI(PPh3)2 + 2TFP f PhPdI(TPF)2 + 2PPh3 (eq I). (b)
Substitution of TFP by PPh3 in PhPdI(TPF)2 to form PhPdI(PPh3)2 is
faster than in eq I,9a suggesting that PPh3 is a better ligand than TPF
for palladium(II), i.e., PPh3 is more basic than TFP.7b

2960 Organometallics, Vol. 17, No. 14, 1998
Amatore et al.
F igu r e 1. UV spectrum performed in DMF in a 1 mm path length cell at 20 °C. (a) Pd(dba)2 (0.1 mmol dm-3) + n equiv
of PPh3: n ) 2 (+), 4 (3), 6 (O), 10 (4), 20 (0), 50 ([), 100 (b). (b) Pd(dba)2 (1 mmol dm-3) + n equiv of TFP: n ) 2 (+),
4 (3), 6 (O), 10 (4), 20 (0), 50 ([).
Id en tifica tion of th e P a lla d iu m (0) Com p lexes
F or m ed in Situ in a Mixtu r e of P d (d ba )2 + n TF P
(n g 2) in DMF a s In vestiga ted by UV Sp ectr os-
cop y. As already reported,5 the complex Pd(dba)(PPh3)2
(1 mmol dm-3 in DMF), quantitatively formed in a
mixture of Pd(dba)2 + 2PPh3, was characterized by UV
spectroscopy with its absorption band at λmax ) 396 nm
(Figure 1a). Addition of increasing amounts of PPh3
resulted in a decay of the adsorption band of Pd(dba)-
(PPh3)2 due to the equilibrium with Pd(0)(PPh3)3 (eq 10)
(Figure 1a).2,5 The complex Pd(dba)(TFP)2 (1 mmol
Pd(0)(dba)(PPh3)2 + PPh3 a
PPh3
SPd(0)(PPh3)3 + dba
K0
(10)
dm-3 in DMF), quantitatively generated in mixtures of
Pd(dba)2 + 2TFP, was characterized by an adsorption
band at ca. 390 nm, less well-defined than that of Pd-
(dba)(PPh3)2 (compare Figures 1a and b). Addition of
increasing amounts of TFP resulted in a decay of the
absorbance due to the equilibrium with Pd(0)(TFP)3 (eq
8). It was qualitatively observed that for the same
amount of added phosphine, the relative concentration
of the complex Pd(0)(dba)L2 was smaller when L ) TFP.
This suggests that Pd(0)L3 is more stable vis a` vis Pd-
F igu r e 2. Cyclic voltammetry of a mixture of Pd(dba)2 (2
mmol dm-3) and n equiv of TFP (n ) 2, 3, 4) at a stationary
gold disk electrode (i.d. ) 0.5 mm) with a scan rate of 0.2
V s-1 at 20 °C: (a) in DMF (containing n-Bu4NBF4, 0.3 mol
dm-3), (b) in THF (containing n-Bu4NBF4, 0.3 mol dm-3).
TFP
(0)(dba)L2 for TFP than for PPh3. In other words, K0
PPh3
PPh3
> K0
. In DMF, the equilibrium constant K0
)
0.16 ( 0.02 had been calculated from UV data.5 Al-
though the calculation was less precise for TFP, the
value of the equilibrium constant K0
Pd(0)(dba)(TFP)2 + TFP U
TFP
could be deter-
TFP
SPd(0)(TFP)3 + dba
K0
(8)
mined at λ ) 410 nm, where C0 is the initial concentra-
tion of Pd(dba)2, n the number of equivalents of TFP
relative to Pd(dba)2, and x ) (D0 - Deq)/(D0 - D∞) (D0 )
initial absorbance of Pd(0)(dba)(PPh3)2, Deq ) absor-
bance at the equilibrium position, D∞: absorbance when
at equil:
[Pd(dba)(TFP)2] ) C0(1 - x); [TFP] ) C0(n - 2 - x)
[SPd(0)(TFP)3] ) C0x; [dba] ) C0(1 + x)
the equilibrium is totally shifted to its right-hand side).
[SPd(0)(TFP)3][dba]
TFP
TFP
K0
) 0.40 ( 0.05. Thus, the substitution of dba by
K0
)
[Pd(dba)(TFP)2][TFP]
the phosphine in Pd(dba)L2 to form SPdL3 is easier for
L ) TPF than for L ) PPh3. These results confirm what
we have previously reported: dba is more easily sub-
stituted by the phosphine in Pd(dba)L2 to form PdL3
when the phosphine is more electron deficient.5
x(1 + x)
(1 - x)(n - 2 - x)
TFP
K0
)
Id en tifica tion of th e P a lla d iu m (0) Com p lexes
F or m ed in Situ in a Mixtu r e of P d (d ba )2 + n TF P
(n g 2) in DMF a n d THF a s In vestiga ted by Cyclic
Volta m m etr y. As already reported for mixtures of Pd-

Reactivity of Palladium(0) Complexes
Organometallics, Vol. 17, No. 14, 1998 2961
Ta ble 2. Oxid a tion P ea k P oten tia ls (Ep )a of th e
P a lla d iu m (0) Com p lexes Gen er a ted in Situ in
Mixtu r es of P d (d ba )2 a n d n TF P (n > 2)
Sch em e 2
–dba + S
+dba – S
+L
–L
[SPdL2][dba]
[Pd(dba)L2]
Pd(dba)L2
SPdL2
SPdL3 K1
=
=
=
Pd(0)(TFP)3
O1
Pd(0)(dba)(TFP)2
[SPdL3][dba]
[Pd(dba)L2][L]
Ep (V vs SCE)
Ep (V vs SCE)
k
+PhI
solvent
O2
K0
K2
DMF
THF
+0.34
+0.38
+0.72
+0.77
PhPdIL2
[SPdL2][L]
[SPdL3]
= K1/K0
a
Determined at 0.2 V s-1 at a stationary gold disk electrode
(i.d. ) 0.5 mm) in solvent containing n-Bu4NBF4 (0.3 mol dm-3),
20 °C; [Pd(dba)2] ) 2 mmol dm-3
coexisted (compare Figure 2 in this work with Figure 3
of ref 2). This suggests that K0
.
TFP
PPh3
TFP
> K0
and K1
(dba)2 + 2PPh3,2 cyclic voltammetry performed on a
solution of Pd(dba)2 (2 mmol dm-3) and 2 equiv of TFP,
in DMF or THF (containing n-Bu4NBF4, 0.3 mol dm-3),
exhibited two oxidation peaks O1 and O2 (Figure 2).
Peak O2 characterizes the complex Pd(dba)(TFP)2,
whereas the plateau-shaped one O1 characterizes Pd-
PPh3 14
> K1
.
Com p a r a t ive R ea ct ivit y of t h e P a lla d iu m (0)
Com p lexes F or m ed in Situ in Mixtu r es of P d (d ba )2
+ n TF P a n d P d (d ba )2 + n P P h 3 (n g 2). Kin etics
of th e Oxid a tive Ad d ition w ith P h en yl Iod id e. As
already reported, the reactivity of palladium(0) com-
plexes in oxidative additions is easily monitored by
amperometry performed at a rotating disk electrode.2,4,5
However, this method could not be used for the ligand
TFP, due to a systematic pollution of the rotating disk
electrode. Thus, the kinetics of the oxidative addition
of the palladium(0) complexes generated in mixtures of
Pd(dba)2 + nTFP (n ) 2, 4, and 10) with PhI was
monitored by performing cyclic voltammetry at a steady
electrode as a function of time (the electrode being
polished just before each voltammogram was per-
formed). The oxidative addition was investigated under
stoichiometric conditions to achieve reasonable reaction
times compatible with the duration between two volta-
mmograms (ca. 1 min). The oxidation currents being
proportional to the concentration of the palladium(0)
complexes, the ratio i/i0 ) [Pd]/[Pd]0 (i ) oxidation peak
current of O1 at t, i0 ) initial oxidation peak current of
O1) was plotted as a function of time (Figure 3a) for
different values of n and for both TFP and PPh3 to allow
comparison. The respective values of the half-reaction
times t1/2 are collected in Table 3.
(TFP)2 involved in an equilibrium (CE mechanism10
with Pd(dba)(TFP)2 as for PPh3, eq 11.2
)
Pd(0)(dba)(TFP)2 + S U
O2
TFP
SPd(0)(TFP)2 + dba
K1
(11)
O1
In the presence of more than 2 equiv of TFP per Pd-
(dba)2, the magnitude of peak O1 increased at the
expense of that of O2 (Figure 2), showing that the
complex Pd(TFP)3 was formed at the expense of Pd(dba)-
(TFP)2 according to equilibrium 8. The two complexes
Pd(0)(dba)(TFP)2 + TFP U
O2
TFP
SPd(0)(TFP)3
+ dba
K0
(8)
O1
SPd(0)(TFP)3 and SPd(0)(TFP)2 are oxidized at the same
As expected, the higher n, the slower the oxidative
addition, for both ligands and both solvents (Figure 3a
and Table 3). However, for identical values of n, one
observes that in THF the oxidative addition is faster
for PPh3 when n ) 2 or 4 whereas at higher n (n ) 10)
the reaction is faster for TPF. Therefore, depending on
the ratio n ) L/Pd, the order of reactivity can be
reversed (Figure 3). In DMF, the oxidative addition is
always faster for TFP than for PPh3, whatever the value
of n (n g 2) (Table 3).
wave O1 because they are involved in a fast equilibrium
(eq 9) (CE mechanism10).2,11 A comparison with PPh3
2
evidences that palladium(0) complexes ligated by TFP
are less easily oxidized than those ligated by PPh3
(Table 2). This indicates that palladium(0) complexes
are more electron rich when ligated by PPh3. In other
words, PPh3 is more electron rich than TFP.7b,9b,12,13
A comparison with PPh3 also shows that for the same
ligand ratio, i.e., same value of n and for the same scan
rate (0.2 V s-1), the ratio iox(O1)/iox(O2) is always higher
for TFP than for PPh3. As soon as 4 equiv of TFP per
Pd(dba)2 was added, peak O1 was detected alone,
whereas for 4 equiv of PPh3, both peaks O1 and O2 still
In the previous paper concerning PPh3,2 the mecha-
nism of the oxidative addition was established and the
reactive species in the oxidative addition identified as
SPd(0)(PPh3)2 (Scheme 2).
When the concentration of Pd(dba)2 and PhI are the
same and when the ligands L and dba are in excess,
the kinetic law is expressed as eq 12, where C0 is the
(10) Bard, A. J .; Faulkner, L. R. Electrochemical Methods; Wiley:
New York, 1980; p 443.
(11) Amatore, C.; J utand, A.; Medeiros, M. J .; Mottier, L. J .
Electroanal. Chem. 1997, 422, 125.
(12) Although to our knowledge, the pK of the tri-2-furylphosphine
is not reported, it is admitted that this ligand is less basic than
PPh3.7b,9b
1/x ) kappC0t + 1
(12)
(13) We noticed that in THF, the oxidation peak O′1 assigned to the
nonsolvated palladium(0) complex Pd(PPh3)3 in equilibrium with the
solvated complex2
initial concentration of palladium(0), x ) i/i0 ) [Pd]/
[Pd]0, and kapp is the apparent rate constant of the
overall reaction. Plotting the variation of 1/x as a
function of time for 10 equiv of TFP per palladium and
different concentrations of dba results in a series of
straight lines that go through 1 (Figure 4a). This shows
that the reaction order in PhI and Pd(0) is 1, and kapp
PPH3
+ TFP h
KS
Pd(PPh2)3
(THF)Pd(0)(PPh3)3
O1
O′1
was not detected for the palladium(0) complex ligated by TFP
(compare Figure 2b here with Figure 3c in ref 2). This means that
TFP
PPh3
KS
. KS
and also suggests that the palladium(0) complex
ligated by TFP is less electron rich than that ligated by PPh3.

2962 Organometallics, Vol. 17, No. 14, 1998
Amatore et al.
F igu r e 3. (a) Kinetics of the oxidative addition of PhI (2 mmol dm-3) with the palladium(0) complexes generated in situ
from mixtures of Pd(dba)2 (2 mmol dm-3) + n equiv of TFP (s) (n ) 2, 4, 10) and from Pd(dba)2 (2 mmol dm-3) + n equiv
of PPh3 (- - -) (n ) 2, 4, 10), monitored by cyclic voltammetry at a stationary gold disk electrode (i.d. ) 0.5 mm) with a scan
rate of 0.5 V s-1 in THF (containing n-Bu4NBF4, 0.3 mol dm-3) at 20 °C. Plot of i/i0 ) [Pd]/[Pd]0 (i ) oxidation peak current
PPh3
TFP
of O1 at t, i0 ) initial oxidation peak current) as a function of time. (b) Plot of the ratio t1/2
THF.
/t1/2
as a function of n, in
2
TFP
PPh3
Ta ble 3. Deter m in a tion of th e Ha lf-Rea ction Tim e
t1/2 of th e Oxid a tive Ad d ition of P h I w ith th e
P a lla d iu m (0) Com p lexes Gen er a ted in Situ in
Mixtu r es of P d (d ba )2 a n d n L (n g 2) a t 20 °Ca
for PPh3 (Table 4). In both solvents, K0
> K0
but the effect is more important in DMF.
In the absence of any added dba, the oxidative
addition, when performed under stoichiometric condi-
tions, follows the kinetic laws, eqs 15 and 16. In THF,
t1/2 (s)
2L
4L
10L
For n ) 2: 1/x + 1/2 ln x ) 1/2kLK1 t + 1 (15)
L
solvent
PPh3
TFP
PPh3
TFP
PPh3
TFP
DMF
THF
195
86
110
143
280
120
190
170
1650
1800
640
690
L
For n ) 4: 1/x + (1/2 ln x)/(1 + K0 ) )
L
L
(kLK1 t)/(2(K0 + 1)) + 1 (16)
[PhI] ) [Pd(dba)2] ) 2 mmol dm-3
.
a
when n ) 2, a plot of 1/x + 1/2 ln x as a function of time
was calculated from the slopes of the regression lines.
The reaction went slower and slower when the concen-
tration of dba was increased. This demonstrates that
the reactive species in the oxidative addition is involved
in an equilibrium with dba, as postulated in Scheme 2.
The apparent rate constant kapp is expressed as eq
132
gave a straight line (Figure 5) which goes through 1 (eq
15). The value of kLK1 has been calculated from the
L
TFP
slope (kTFPK1
) 1.08 × 10-2 s-1 in THF) and agrees
perfectly with the value calculated above.
L
When n ) 4, the kinetic law depends on K0 (eq 16),
which has already been determined as reported above
(Table 4). Using this value, one plots the variation of
1/x + (1/2 ln x)/(1 + K0L) as a function of time (eq 16).
k
kapp
)
(13)
The value of kLK1 is calculated from the slope of the
L
1 + [dba]/K1 + [TFP]/K2
straight line (Figure 5). kTFPK1
) 1.06 × 10-2 s-1 in
TFP
THF, which again compares excellently with the above
or
TFP
determination. In THF, the values of kTFPK1
calcu-
lated by three different kinetic laws are very similar
and, thus, confirm the mechanism proposed in Scheme
2. Similar analysis of the kinetics of the oxidative
addition performed in DMF allows the determination
[dba]
1
1
1
kK2
)
+
+
(14)
{
}
kapp[TFP]
k[TFP]
kK1[TFP]
When 1/kapp[TFP] was plotted as a function of [dba]/
of kTFPK1
and comparison with kPPh K1
(Table 4).
TFP
PPh3
3
[TFP] (eq 14), a straight line was obtained (Figure 4b)
whatever the TFP concentration, showing that 1/k[TFP]
is always very small compared to 1/kK2 and to experi-
mental uncertainty, i.e., k > 50 M-1 s-1. In THF, kK2
) (4.2 ( 0.1) × 10-2 s-1 was calculated from the
intercept whereas kK1 ) (1.07 ( 0.01) × 10-2 s-1 was
calculated from the slope. Then, K0 ) K1/K2 ) 0.26 (
0.01. The same calculation was made in DMF (Table
4) and the results compared to those already reported
F in a l Discu ssion a n d Con clu sion
As recalled in the Introduction, the overall reactivity
of {Pd(dba)2 + nL} in the oxidative addition is deter-
mined by the intrinsic reactivity of the reactive species
Pd(0)L2 in the elementary step of the oxidative addition
(i.e., k in eq 5) and also by its concentration which is
controlled by K1 and K0. K0 (eq 3) plays a special role

Reactivity of Palladium(0) Complexes
Organometallics, Vol. 17, No. 14, 1998 2963
F igu r e 4. Kinetics of the oxidative addition of PhI (2 mmol dm-3) with the palladium(0) complex generated in situ from
a mixture of Pd(dba)2 (2 mmol dm-3) + 10 equiv of TFP in THF (containing n-Bu4NBF4, 0.3 mol dm-3) as a function of the
dba concentration. (a) Plot of i0/i ) [Pd]0/[Pd] as a function of time and the amount of added dba: (b) 0, (O) 10, (+) 20
equiv of dba, 20 °C. (b) Plot of 1/kapp[TFP] as a function of [dba][TFP], F ) 0.996 (see eq 14).
Ta ble 4. Deter m in a tion of th e Equ ilibr iu m Con sta n ts a n d Kin etic P a r a m eter s for P a lla d iu m (0) Com p lexes
a
Gen er a ted in Situ in Mixtu r es of P d (d ba )2 a n d n L (n g 2) a n d In volved in th e Oxid a tive Ad d ition w ith
P h I, Accor d in g to Sch em e 2
102 × kLK1 (s-1
)
102 × kLK2 (s-1
)
K0
L
L
L
b
b
b
solvent
DMF
PPh3
TFP
PPh3
TFP
PPh3
TFP
0.63 ( 0.01
1.13 ( 0.01
4.6 ( 0.4
1.74 ( 0.1
0.14 ( 0.03
0.16 ( 0.02c,d
0.23 ( 0.03
0.32 ( 0.03
0.40 ( 0.05d
0.26 ( 0.01
THF
1.10 ( 0.01
1.07 ( 0.01
4.75 ( 0.1
4.2 ( 0.1
a
b
d
[Pd(dba)2] ) 2 mmol dm-3, 20 °C. Reference 2. c Reference 5. Calculated from UV data (see text).
the usual catalytic conditions where L is not in great
excess, Pd(0)(dba)L2 dominates over SPd(0)L3 so that
it is reasonable to use the pair K1, K0 to describe the
system while keeping in mind that K2 ) K1/K0.
When compared to PPh3, TFP is less electron rich,
and consequently, the intrinsic reactivity (viz. the rate
constant) of SPd(TFP)2 in the oxidative addition (eq 5)
3
should be less than that of SPd(PPh3)2 with kTFP < kPPh
.
But from the data collected in Table 4, one observes that
TFP
PPh3
in DMF and THF, K0
> K0
. Thus, two antagonist
effects might be involved.
In THF, at low n (n ) 2), the overall reactivity of {Pd-
(dba)2 + nL} (viz. the rate constant times the concentra-
tion) is controlled by kLK1 (eq 15) and the complex
L
ligated by PPh3 is more reactive than that ligated by
L
TFP (Figure 3a, Table 3). K1 cannot be calculated.
However, from the cyclic voltammetry,14 we are inclined
to conclude that K1
F igu r e 5. Kinetics of the oxidative addition of PhI (2 mmol
dm-3) with the palladium(0) complex generated in situ from
mixtures of Pd(dba)2 (2 mmol dm-3) + n equiv of TFP in
THF (containing n-Bu4NBF4, 0.3 mol dm-3). i0/i ) [Pd]0/
[Pd]; n ) (O) 2 (see eq 15), (b) 4 (see eq 16), 20 °C.
PPh3
< K1TFP, i.e., the concentration
of Pd(0)(PPh3)2 is less than that of Pd(0)(TFP)2 but Pd-
(0)(PPh3)2 is more reactive than Pd(0)(TFP)2. As a
result of these two antagonist effects, kPPh K1
is
PPh3
3
found to be slightly superior to kTFPK1TFP, which means
since it reflects the ratio of the only species SPd(0)L3
and Pd(0)(dba)L2 present at macroscopic concentration.
Either K1 (eq 2) or K2 (eq 4) can be used to describe the
relative concentration of SPd(0)L2, the true reactive
species present at trace amounts. Ideally, K1 (i.e., SPd-
(0)L2 vs Pd(0)(dba)L2) should be used when K0[L]/[dba]
, 1 (i.e., when Pd(dba)L2 is the major species), and K2
(i.e., SPd(0)L2 vs SPd(0)L3) should be used when K0[L]/
[dba] . 1 (i.e., when SPdL3 is the major species). Under
that kPPh is considerably much higher than kTFP and
3
thus k is more determinant than K1. In other words,
(14) Although the magnitude of the peak current does not represent
the true concentration of the species but a dynamic concentration,2,10
TFP
PPh3
due to equilibrium 8, one concludes that either K0
> K0
(which
was established from UV spectroscopy) or that equilibrium 8 is more
easily shifted to Pd(0)(TFP)3 than it is to Pd(0)(PPh3)3 when the species
is continuously oxidized in the diffusion layer or both. The same
conclusions are drawn for equilibrium 11 and K1.

2964 Organometallics, Vol. 17, No. 14, 1998
Amatore et al.
the relative rate of the intrinsic oxidative addition (k)
is more important than is the ability of the solvent to
displace dba from Pd(dba)L2 to form SPdL2.
Thus, in this solvent, both the oxidative addition and
the nucleophilic attack should be enhanced when using
TFP instead of PPh3.
At higher values of n (n > 2), equilibrium 3 (K2) has
to be considered and consequently K0 with on the one
Exp er im en ta l Section
and on the other hand kPPh > k
,
PPh3
TFP
TFP
31P NMR spectra were recorded in DMF on a Bruker
spectrometer (101 MHz) using H3PO4 as an external reference.
UV spectra were recorded on a DU 7400 Beckman spectro-
photometer. Cyclic voltammetry was performed with a home-
made potentiostat and a waveform generator (Tacussel GSTP4).
The cyclic voltammograms were recorded on a Nicolet oscil-
loscope.
Rea gen ts. All experiments were performed under argon.
DMF (Acros) was distilled over calcium hydride under vacuum
and stored under an argon atmosphere. THF (Acros) was
distilled from sodium benzophenone. The solvent was trans-
ferred to the cells according to standard Schlenk procedures.
The tri-2-furylphosphine was commercial grade (Strem Chemi-
cal) and used without further purification. Phenyl iodide
(Acros) was filtered on neutral alumina and stored under
argon. The bis(dibenzylidenacetone)palladium(0) was synthe-
sized according to published procedures.15
3
hand K0
< K0
i.e., still two antagonist effects. The concentration of
the reactive species Pd(0)L2 is modulated by the values
of n and K0. Even if Pd(0)(TFP)2 is less reactive than
Pd(0)(PPh3)2, its concentration is considerably higher
than that of Pd(0)(PPh3)2 when n is large, resulting in
a higher overall reactivity. Due to these two antagonist
effects, an inversion of the reactivity order occurs when
n increases. This was observed when switching from n
PPh3
) 4 to n ) 10 (Table 3, Figure 3a). The ratio t1/2
/
TFP
t1/2
was plotted versus n (Figure 3b). Its value is 1
(same overall reactivity for both ligands) when n ≈ 6.
Thus, when n > 6, the catalytic system with TPF is more
reactive than that with PPh3, whereas the reverse is
observed for n < 6.
31P NMR Exp er im en ts. A 3.7 mg (0.0064 mmol) amount
of Pd(dba)2 was introduced into an NMR tube containing 0.5
mL of DMF or THF and 0.05 mL of acetone-d6, followed by
the suitable amount of TFP (2 or 4 equiv relative to Pd(dba)2).
In another experiment, phenyl iodide (10 equiv) was added to
the mixture of Pd(dba)2 and nTFP (n ) 2 or 4).
UV Exp er im en ts. The UV spectroscopy was performed in
DMF or THF in a thermostated 1 mm path length cell on
mixtures of Pd(dba)2 (2 mmol dm-3) and the suitable amount
of TFP (4-100 mmol dm-3).
In DMF, the system {Pd(dba)2 + nTFP} is found to
be more reactive than {Pd(dba)2 + nPPh3}, whatever n
(n g 2). In this solvent when n ) 2, kTFPK1
>
TFP
kPPh K1
(Table 4). Since kTFP < kPPh , this means that
PPh3
3
3
TFP
the contribution of K1
which controls the concentra-
tion of Pd(0)(TFP)2 is more important than that of kTFP
(the reverse effect was observed in THF, see above). For
TFP
PPh3
n > 2, K0 is involved with K0
> K0
. Thus, as for
TFP
K1TFP, the contribution of K0
is more important than
that of kTFP. Despite the fact that TFP is less basic than
PPh3, the overall oxidative addition is faster with TFP
because the concentration of SPd(TFP)2 is always higher
than that of SPd(PPh3)2, whatever n (n g 2).
Electr och em ica l Exp er im en ts. These studies were car-
ried out in a three-electrode cell connected to a Schlenk line.
The cell was equipped with a double envelope to have a
constant temperature of 20 °C (Lauda RC20 thermostat). The
working electrode consisted of a gold disk of 0.5 mm diameter.
The counter electrode was a platinum wire of ca. 1 cm2
apparent surface area. The reference was a saturated calomel
electrode separated from the solution by a bridge filled with a
solution of n-Bu4NBF4 (0.3 mol dm-3) in 3 mL of DMF or THF.
A 12 mL amount of DMF or THF containing the same
concentration of supporting electrolyte was poured into the cell.
A 13.6 mg amount of Pd(dba)2 (0.024 mmol) was then added,
followed by the suitable amount of TFP. Cyclic voltammetry
In practice, in THF, one generally uses a mixture of
Pd(dba)2 + nL with n ) 2 or 4.6a,b,d,e,g,h From the results
reported above, the mixtures {Pd(dba)2 + nTFP} (n )
2 or 4) are slightly less reactive than {Pd(dba)2
+
nPPh3} (n ) 2 or 4) in oxidative additions. Thus, the
fact that the rate of a palladium-catalyzed reaction
involving an oxidative addition with ArX is considerably
enhanced when switching from PPh3 to TFP6a,b,e,g does
not originate from the oxidative addition. As a conse-
quence, in this case, this step cannot be the rate-
determining step of the catalytic cycle.7b The higher
rate probably originates from a further step, for ex-
ample, the nucleophilic attack (transmetalation) on
arylpalladium(II) complexes which are more electron
deficient when ligated by TFP than by PPh3.
was performed at a scan rate of 0.2 V s-1
measurements were performed by cyclic voltammetry at a scan
rate of 0.5 V s-1 in the presence of 2.7 µL (0.024 mmol) of
phenyl iodide.
. The kinetic
Ack n ow led gm en t. This work has been supported
in part by the Centre National de la Recherche Scien-
tifique (CNRS, URA 1679, “Processus d’Activation Mo-
le´culaire”), the Ministe`re de l′Enseignement Supe´rieur
et de la Recherche (Ecole Normale Supe´rieure), and a
cooperation between CNRS (France) and CNCPRST
(Morocco), No. 4542.
In DMF,6c,i the results differ since the oxidative
addition is always faster with TFP than with PPh3.
(15) (a) Takahashi, Y.; Ito, T.; Sakai, S.; Ishii, Y. J . Chem. Soc.,
Chem. Commun. 1970, 1065. (b) Rettig, M. F.; Maitlis, P. M. Inorg.
Synth. 1977, 17, 134.
OM971064U