Mechanistic and kinetic investigation on the formation of palladacyclopentadiene complexes. A novel interpretation involving a bimolecular self-reaction of a monoalkyne intermediate

Article abstract of DOI:10.1021/om8002398

The stoichiometric reaction between the complex [Pd(η²dmfu) (BiPy)] (dmfu = dimethylfumarate; BiPy = 2,2′-bipyridine) and the deactivated alkynes dmbd (dimethyl-2-butynedioate) and pna (methyl (4-nitrophenyl)propynoate), providing the respective palladacyclopentadienes, was investigated. The mechanism leading to the palladacyclopentadiene derivative involves a bimolecular self-rearrangement of the monoalkyne intermediate [Pd(η²-alk)(BiPy)] (alk = dmbd, pna), followed by the customary attack of the free alkyne on the intermediate [Pd(η²-alk)(BiPy)] itself and on the elusive and highly reactive naked palladium [Pd(BiPy)(0)] formed. The alkyne pna proved to be less effective in the displacement of dmfu than dmbd. The reaction under stoichiometric equimolar conditions of the latter with [Pd(η²-dmfu)(BiPy)] allows the direct determination of the bimolecular self-reaction rate constant k _c and consequently the assessment of all the rate constants involved in the overall mechanistic network.

Full text of DOI:10.1021/om8002398

4050
Organometallics 2008, 27, 4050–4055
Mechanistic and Kinetic Investigation on the Formation of
Palladacyclopentadiene Complexes. A Novel Interpretation Involving
a Bimolecular Self-Reaction of a Monoalkyne Intermediate
A. Holuigue,† J. M. Ernsting,† F. Visentin,‡ C. Levi,‡ L. Canovese,*,‡ and C. J. Elsevier*,†
Van’t Hoff Institute for Molecular Sciences, Nieuwe Achtergracht 166, 1018WV Amsterdam,
The Netherlands, and Dipartimento di Chimica, UniVersita` Ca’ Foscari, Calle Larga S. Marta 2137,
30123 Venezia, Italy
ReceiVed March 14, 2008
The stoichiometric reaction between the complex [Pd(η2-dmfu)(BiPy)] (dmfu ) dimethylfumarate;
BiPy ) 2,2′-bipyridine) and the deactivated alkynes dmbd (dimethyl-2-butynedioate) and pna (methyl
(4-nitrophenyl)propynoate), providing the respective palladacyclopentadienes, was investigated. The
mechanism leading to the palladacyclopentadiene derivative involves a bimolecular self-rearrangement
of the monoalkyne intermediate [Pd(η2-alk)(BiPy)] (alk ) dmbd, pna), followed by the customary attack
of the free alkyne on the intermediate [Pd(η2-alk)(BiPy)] itself and on the elusive and highly reactive
“naked palladium” [Pd(BiPy)(0)] formed. The alkyne pna proved to be less effective in the displacement
of dmfu than dmbd. The reaction under stoichiometric equimolar conditions of the latter with [Pd(η2-
dmfu)(BiPy)] allows the direct determination of the bimolecular self-reaction rate constant kc and
consequently the assessment of all the rate constants involved in the overall mechanistic network.
synthesis of conjugated dienes from two molecules of an alkyne
via a metallacyclopentadiene is limited to only a few transition
metals, such as titanium,6 zirconium,7 iridium,8 and palladium.9
Introduction
The stereospecific synthesis of conjugated dienes is of
considerable importance since many natural products and
bioactive compounds contain the 1,3-diene unit or even multiple
unsaturations with higher degrees of conjugation. For these
reasons, transition-metal-mediated conversion of alkynes into
polyenes and, more precisely, the development of selective
methods for the synthesis of 1,3-dienes have been research areas
of interest for many years now. This had led to the development
of catalysts based on cobalt,1 ruthenium,2 nickel,3 and
palladium3b,4 for the double addition of diazo compounds to
alkynes or the direct coupling of a stereodefined alkenyl-metallic
compound with a stereodefined vinyl electrophile such as a vinyl
halide. Other stereoselective methods of diene synthesis based
ondiynereductionwithazinc/copperreagentorasodium-mercury
amalgam were also described.5
However, owing to the complexity of the catalytic cycle,
which is often complicated by the presence of several byprod-
ucts, few exhaustive mechanistic studies have been carried out.
Particularly, the formation of the key intermediate metallacy-
clopentadiene species has been studied in detail only once.9g
Nevertheless, dentification of the species involved and their
reactivity at the very first step of the catalytic cycle seem
somehow crucial for the understanding of the whole process.
For that reason measurable rates of reaction and mild conditions
are necessary for a viable approach so that all possible information
could be gathered from the system under investigation. In this
respect we have found that the complex [Pd(η2-dmfu)(BiPy)] (dmfu
) dimethylfumarate; BiPy ) 2,2′-bipyridine) and the alkynes dmbd
(dimethyl-2-butynedioate) and pna (methyl (4-nitrophenyl)propy-
noate) seem to be tailored for a detailed stoichiometric investigation
In addition to these approaches, some catalytic processes
involving metallacyclopentadiene species as intermediates have
been published. However, to the best of our knowledge, the
(6) (a) Johnson, E. S.; Balaich, G. J.; Fanwick, P. E.; Rothwell, I. P.
J. Am. Chem. Soc. 1997, 119, 11086. (b) Yamaguchi, S.; Jin, R.-Z.; Tamao,
K.; Sato, F. J. Org. Chem. 1998, 63, 10060. (c) Yamamoto, Y.; Ohno, T.;
Itoh, K. Chem. Commun. 1999, 1543. (d) Ryan, J.; Micalizio, G. C. J. Am.
Chem. Soc. 2006, 128, 2764.
(7) (a) Buchwald, S. L.; Nielsen, R. B. J. Am. Chem. Soc. 1989, 111,
2870. (b) Takahashi, T.; Kotora, M.; Kasai, K.; Suzuki, N. Organometallics
1994, 13, 4183. (c) Takahashi, T.; Xi, Z.; Yamazaki, A.; Liu, Y.; Nakajima,
K.; Kotora, M. J. Am. Chem. Soc. 1998, 120, 2870.
* Corresponding author. E-mail: elsevier@science.uva.nl.
† Van’t Hoff Institute for Molecular Sciences.
‡ Universita` Ca’ Foscari.
(1) (a) Wakatsuki, Y.; Aoki, K.; Yamazaki, H. J. Am. Chem. Soc. 1974,
96, 5284. (b) Wakatsuki, Y.; Aoki, K.; Yamazaki, H. J. Am. Chem. Soc.
1979, 101, 1123.
(2) (a) Trost, B. M.; Indolese, A. F.; Mu¨ller, T. J. J.; Treptow, B. J. Am.
¨
Chem. Soc. 1995, 117, 615. (b) Le Paih, J.; De´rien, S.; Ozdemir, I.; Dixneuf,
P. H. J. Am. Chem. Soc. 2000, 122, 7400. (c) Morita, R.; Shirakawa, E.;
Tsuchimoto, T.; Kawakami, Y. Org. Biomol. Chem. 2005, 3, 1263.
(3) (a) Negishi, E.-I.; Takahashi, T.; Baba, S.; Van Horn, D. E.; Okukado,
N. J. Am. Chem. Soc. 1987, 109, 2393. (b) Shirakawa, E.; Takahashi, G.;
Tsuchimoto, T.; Kawakami, Y. Chem. Commun. 2001, 2688.
(4) (a) Tellier, F.; Sauveˆtre, R.; Normant, J. F. J. Organomet. Chem.
1986, 303, 309. (b) Stille, J. K.; Groh, B. L. J. Am. Chem. Soc. 1987, 109,
2393. (c) Li, J.-H.; Liang, Y.; Xie, Y.-X. J. Org. Chem. 2004, 69, 8125.
(d) Hattori, H.; Katsukawa, M.; Kobayashi, Y. Tetrahedron Lett. 2005, 46,
5871.
(8) O’Connor, J. M.; Hiibner, K.; Merwin, R.; Gantzel, P. K.; Fong,
B. S.; Adams, M.; Rheingold, A. L. J. Am. Chem. Soc. 1997, 119, 3631.
(9) (a) Moseley, K.; Maitlis, P. M. Chem. Commun. 1971, 1604. (b)
Suzuki, H.; Itoh, K.; Ishii, Y.; Simon, K.; Ibers, J. A. J. Am. Chem. Soc.
1976, 98, 8494. (c) Van Belzen, R.; Hoffmann, H.; Elsevier, C. J. Angew.
Chem., Int. Ed. Engl. 1997, 36, 1743. (d) Belzen, R. V.; Klein, R. A.;
Kooijman, H.; Veldman, N.; Spek, A. L.; Elsevier, C. J. Organometallics
1998, 17, 1812. (e) Shirakawa, E.; Yoshida, H.; Nakao, Y.; Hiyama, T.
J. Am. Chem. Soc. 1999, 121, 4290. (f) Yoshida, H.; Shirakawa, E.; Nakao,
Y.; Honda, Y.; Hiyama, T. Bull. Chem. Soc. Jpn. 2001, 74, 637. (g)
Canovese, L.; Visentin, F.; Chessa, G.; Uguagliati, P.; Levi, C.; Dolmella,
A. Organometallics 2005, 24, 5537, and references therein.
(5) Solladie´, G.; Stone, G. B.; Andre´s, J.-M.; Urbano, A. Tetrahedron
Lett. 1993, 34, 2835.
10.1021/om8002398 CCC: $40.75
2008 American Chemical Society
Publication on Web 07/23/2008

Formation of Palladacyclopentadiene Complexes
Organometallics, Vol. 27, No. 16, 2008 4051
Scheme 1
Scheme 2
of the reaction yielding the palladacyclopentadienyl derivatives
according to Scheme 1.
As a matter of fact, type 2 complexes, which are produced
by the equilibrium displacement of an alkene by an activated
alkyne, represent a peculiar class of substrates that can ac-
cumulate as unreactive species, which, depending on their
electronic and steric characteristics, react with an additional
molecule of the alkyne to give palladacyclopentadiene deriva-
tives or react with the palladacyclopentadiene itself to yield
mellitate derivatives.10 The understanding of the initial equilibria
in the formation of palladacycles is also of crucial importance
for a better understanding of the intimate mechanism governing
the first step of other catalytic cycles, such as the formation of
functional dienes from coupling reactions between alkynes,
organic halides, and organotin reagents.9c,d
the equilibrium constant by direct titration. Moreover, the
reactions subsequent to the equilibrium displacement in Scheme
2 need to be quenched; otherwise no equilibrium concentrations
could be measured with confidence. Some of us have determined
previously the coordinating ability order among deactivated
alkenes bonded to Pd(0) complexes,11 which was often con-
firmed later.12 Therefore, the low coordinating capability of
dmfu is not surprising since this alkene is one of the less
coordinating among the deactivated ones (maleic anhydride and
fumaronitrile being ca. 7900 and 4400 times more effective than
dmfu, respectively). We have thus determined the equilibrium
constant of the displacement of fn from the complex [Pd(η2-
Results and Discussion
Reaction between [Pd(η2-dmfu)(BiPy)] and Dimethyl-2-
butynedioate (dmdb). When an equimolar amount of dmbd is
added to a solution of [Pd(η2-dmfu)(BiPy)] ([[Pd(η2-dmfu)-
(BiPy)]] ≈ 1 × 10-2 mol dm-3), the formation of the
cyclopalladate [Pd(BiPy)(C4(COOMe)4)] together with unre-
acted [Pd(η2-dmfu)(BiPy)] is observed by 1H NMR. The same
reaction carried out in the presence of dmbd in 2-fold (or higher)
excess ([[Pd(η2-dmfu)(BiPy)]] ≈ 1 × 10-2 mol dm-3; [dmbd]
1
fn)(BiPy)] by titration with dmbd at 213 K, monitored by H
NMR. Under these conditions K became easily accessible (K
) 0.16 ( 0.01) because of the coordinating ability of fn and
the slow rates of subsequent reactions. Figure 1 displays the
nonlinear regression fit based on the following relevant relation-
ships (fn ) fumaronitrile):
)
(2-3)
×
10-2 mol dm-3
)
yields the complex
[Pd(BiPy)(C4(COOMe)4)] and free dmfu (or free dmfu and
dmbd in stoichiometric excess). A reasonable reaction scheme
that takes all the experimental observations into account is
summarized in Scheme 2.
2
2
[
(
)
]
[
(
)
]
(
)
(
)
Pd η -fn BiPy + dmbd T Pd η -dmbd BiPy + fn; K
1
2a
However, Scheme 2 is of considerably complexity. Only a
stepwise approach can be adopted in order to solve such a
mechanistic network, since multiparametric analysis of kinetic
data does not warrant reliable equilibrium and rate constants
because of their high correlation. In this respect we first tried
to determine the value of KE by titration of an [Pd(0)(η2-alkene)]
derivative with the alkyne dmbd.
Determination of KE. It was already stated that addition of
an equimolar amount of dmbd to a solution of [Pd(η2-
dmfu)(BiPy)] at RT yields instantaneously the reaction products.
Apparently, the displacement of dmfu by dmbd is energetically
favored and therefore quantitative; the use of a Pd(0) derivative
bearing a more coordinating alkene is recommended in order
to contrast efficiently the electrophilicity of dmbd and determine
2
[
]
K ) 2a
⁄
1 - 2a
dmbd 0 - 2a
1
2
3
[
]
eq
(([
]
[
])(
eq
[
]))
eq
0
1 0 ) 1 × 10-2 mol dm-3
[ ]
dmbd 0 ) in the range (1-6.2) × 10-2 mol dm-3
[
]
The KE value related to the equilibrium displacement of dmfu
by dmbd (equilibrium A in Scheme 2) was then calculated by
multiplying K by 4400. The ensuing value (KE ≈ 700) was taken
as a reasonable equilibrium constant for the displacement of
dmfu by dmbd at 213 K. The value of KE ≈ 70 estimated at
298 K justifies the almost quantitative displacement of dmfu
(11) Canovese, L.; Visentin, F.; Uguagliati, P.; Crociani, B. J. Chem.
Soc., Dalton Trans. 1996, 1921.
(12) Canovese, L.; Chessa, G.; Visentin, F.; Uguagliati, P. Coord. Chem.
ReV. 2004, 248, 945, and references therein.
(10) Canovese, L.; Visentin, F.; Chessa, G.; Santo, C.; Levi, C.;
Uguagliati, P. Inorg. Chem. Commun. 2006, 9, 388.

4052 Organometallics, Vol. 27, No. 16, 2008
Holuigue et al.
Figure 1. Nonlinear regression fit of the equilibrium concentrations
determined by 1H NMR technique at 213 K for the reaction [Pd(η2-
fn)(BiPy)] + dmbd T [Pd(η2-dmbd)(BiPy)] + fn.
Figure 2. Linear regression plot of kobs vs [dmbd]0 for the reaction
of complex 2a with dmbd.
when an equimolar amount of dmbd was added to a solution of
[Pd(η2-dmfu)(BiPy)] at 298 K.13,14
tively. The ensuing value for the rate constant was kc ) 44 (
5 mol-1 dm3 s-1 and was independent of the concentration of
the substrates.
Determination of kc. An interpretation of Scheme 2 suggests
another route to the analytical solution of the rate constant
network. Addition of an equimolar amount of dmbd to a solution
of [Pd(η2-dmfu)(BiPy)] would yield almost quantitatively 2a
(path A in Scheme 2);15 two molecules of 2a would react with
each other to give the palladacyclopentadiene 3a and the so-
called naked palladium Pd(0) (path B), which reacts very rapidly
with the free dmfu to give the starting complex 1 (path C).
Owing to the stoichiometry of the reaction, such an equimolar
reaction would therefore yield an equimolar mixture of com-
Determination of k2. The value of k2 was also measured by
UV-vis technique under pseudo-first-order conditions ([dmbd]0
g 10[1]0). The excess of dmbd shifts the equilibrium mixture
well over to the right and drives the reaction to completion.
Therefore, at the end of the reaction only 3a, excess dmbd, and
free dmfu have been detected in solution. The nonlinear
regression process was based on the model reported below with
the same symbols as above:
1
plexes 1 and 3a (it is noteworthy that H NMR experiments
[
]
[ ] (
2a 0 ) 1 0 from the equilibrium mixture
completely shifted to the right
carried out under equimolar conditions, albeit too fast to be
analyzed kinetically, yielded the mixture we expected; see
Experimental Section). Therefore, we have determined the kc
value from three independent measurements carried out by
UV-vis technique at different concentrations of 1 and dmbd
(ratio 1:1; [1]0 ) [dmbd]0 ) 4 × 10-4, 2 × 10-4, 1 × 10-4
mol dm-3) by nonlinear regression of the integrated form of
the second-order differential equation model given below, with
kc and ε2a as the parameters to be optimized:
)
[
]
[
]
[
]
[
]
(
)
Pd(0) ) 2a 0 - 2a - 3a mass balance
2
]
[
]
[
]
[
[
][
]
-d 2a ⁄ dt ) kobs 2a + 2kc 2a -km Pd(0) dmbd 0
2
]
[
]
[
]
[
d 3a ⁄ dt ) kobs 2a + kc 2a
[
]
[
]
Dt)ε2a 2a + ε3a 3a
[
2
]
[
]
[
] [
]
k
(
)k dmbd 0; initial conditions 2a ) 2a 0, 3a ) 0
)
obs
[ ]
[ ]
[
]
[
]
(
)
1 0 ) 1 + 2a + 3a mass balance
The regression analysis yields the values for kobs at four
2
]
different dmbd concentrations, with kc ) 44 and km ) 10 000
mol-1 dm3 s-1 16 held constant during the refinement process.
The linear regression of the kobs vs [dmbd] yielding k2 ) 0.79
( 0.05 mol-1 dm3 s-1 is reported in Figure 2.
[
]
[
]
[
-d 2a ⁄ dt ) d Pd fin⁄dt ) 2kc 2a
[
]
[
]
Dt ) ε2a 2a + εfin Pd fin
[ ]
[
]
[
]
1 + 3a ) Pd fin;ε ) ε + ꢀ
(
)
3a
fin
1
Reaction between [Pd(η2-dmfu)(BiPy)] and Methyl (4-Nitro-
phenyl)propynoate (pna). In analogy with the previous study
carried out with dmbd, we tried to determine the reactivity of
methyl (4-nitrophenyl)propynoate toward palladium(0) dmfu
derivatives. Unfortunately, the exchange equilibrium reaction
between dmfu and pna
Dt represents the optical density at time t, and ε1, ε2a, and ε3a
are the molar extinction coefficients of 1, 2a, and 3a respec-
(13) In the case of exchange between alkenes in R-diimine derivatives
of Pd(0) a value of ∆H0 )-3.5 ( 0.5 kcal mol-1 was determined.11 From
that value under the reasonable hypothesis that the slope of the Van’t Hoff
relationship does not change considerably on going from the exchange
between two alkenes to the exchange of an alkene with an alkyne, the value
of KE298K was calculated from the expression ln KE298K ) ln KE213K-(∆H0/
R)(213-298)/(213 × 298); KE ≈ 66. We consider that approach reliable
since we have also studied the RT exchange between maleic anhydride
and dmdb in pyridylthioether derivatives of Pd(0)9g and the exchange
constant between [Pd(η2-fn)(neoc)] and dmbd.14 The equilibrium constants
determined in those cases when multiplied by 7900 and 4400 give the values
of 110 and 54, respectively, which compare quite well with the value of 66
estimated in the present work. As a matter of fact, such values represent
the equilibrium constants for the exchange between dmfu and dmdb in
complexes bearing different ancillary ligands. Not surprisingly the KE value
is almost independent of the nature of the ancillary ligand.
2
[
(
)
]
(
)
Pd η -dmfu BiPy + pna T
1
2
[
(
)
]
(
)
Pd η -pna BiPy + dmfu K *
(
)
E
2b
is far from complete, and even at low temperature (213 K in
CDCl3) a number of other species are detected in solution.
Apparently the palladium(0) alkyne complex 2b, which is
produced by the addition of pna to the starting complex [Pd(η2-
dmfu)(BiPy)], is a very reactive species, and the addition
(necessarily at RT) of the titrant pna into the NMR tube induces
(14) Canovese, L.; Visentin, F.; Santo, C. J. Organomet. Chem. 2007,
692, 4187.
(15) With KE ) 70 the formation of complex 2 from an equimolecular
addition of dmbd to 1 is about 90% of the initial concentration of [1]0 (ꢀ
) [1]0(K-(KE)1/2)/(KE-1)).
(16) The km value used during the refinement process is an arbitrarily
high number related to a very fast reaction, which can be varied from 1000
to 10 000 without affecting the ensuing kobs value.

Formation of Palladacyclopentadiene Complexes
Organometallics, Vol. 27, No. 16, 2008 4053
Table 1
k2*
kc*
kc*
[pna]0
(random and
negligible)
(mol-1
dm3 s-1
)
(mol-1 dm3 s-1
(average)
)
(mol dm-3
)
1.48 × 10-2
1.98 × 10-2
2.91 × 10-2
negligible
negligible
negligible
97 ( 1.5
109 ( 1.8
105 ( 1.3
103 ( 6
the complex [Pd(η2-pna)(BiPy)] as the sole starting material.18
Under these conditions the reaction scheme becomes consider-
ably simplified even though the independent determination of
kc* and k2* is not possible since the equimolar reaction between
[[Pd(η2-dmfu)(BiPy)] and pna is complicated by the presence
of several species and side reactions. The new conditions can
be summarized as follows:
(1) 2b + pna f 3bt (k2*)19 ([3bt] ) [3b] + [3b′])
(2) 2 2b f 3bt + Pd(0) (kc*)
(3) Pd(0) + pna f 2b (km)
Figure 3. Nonlinear regression fit of the equilibrium determined
by UV-vis technique at 298 K for the reaction [Pd(η2-dmfu)(neoc)]
+ pna T [Pd(η2-pna)(neoc)] + dmfu.
Therefore nonlinear regression analysis of the kinetics
performed by UV-vis technique based on the model:
the onset of the reaction with the formation of the following
species:
(1) the complex [Pd(η2-pna)(BiPy)] (2b);
(2) the unreacted [Pd(η2-dmfu)(BiPy)] (1);
(3) the unreacted pna;
[
]
[ ] (
2a 0 ) 1 0 from the equilibrium mixture
completely shifted to the right
)
[
]
[
]
[
]
[
]
(
)
Pd(0) ) 2b 0 - 2b - 3bt mass balance
(4) the free dmfu;
(5) the symmetric (3b) and the nonsymmetric (3b′) palladium
cyclopentadienyl derivatives;
(
[
]
[
]
[
]
)
3bt ) 3b + 3b ′
2
[
] [
]
[
]
[
][
]
-d 2b] ⁄ dt ) k * pna 0 2b + 2k * 2b - km Pd(0) pna 0
[
2
c
(6) the symmetric (4b) and the nonsymmetric mellitate (4b′).
2
]
[
]
[
] [
]
[
d 3bt ⁄ dt ) k2* pna 0 2b + kc* 2b
[
]
[
]
Dt)ε2b 2b + ε3bt 3bt
[
]
[
]
[
] [
]
k
(
)k dmbd 0; initial conditions: 2b ) 2b 0, 3bt ) 0
)
2
obs
yields the results summarized in Table 1.
Formation of the Symmetric and Unsymmetric
Mellitate from Methyl (4-Nitrophenyl)propynoate. The stable
palladacyclopentadiene (3b/3b′) does not react with one further
alkyne molecule to give the mellitate at 298 K, but the finally
observed ratio of cyclotrimers (4b/4b′) is the same as the
observed ratio of the regioisomers (3b/3b′).We may therefore
assume that, in the present case, the mellitates (4b/4b′) are also
arising from the reaction of 3b′/3b with [Pd(η2-pna)(BiPy)]
according to Scheme 3.
The conservation of the observed ratio of regioisomers could
imply that the insertion of the third molecule of methyl (4-
nitrophenyl)propynoate originates from [Pd(η2-pna)(BiPy)],
which regioselectively inserts in the palladium-carbon bond
of 3b′.
In order to evaluate the equilibrium constant KE*, we took
several independent H NMR spectra of the reaction mixture
1
obtained upon subsequent additions of pna and estimated only
the concentrations of the species directly involved in the
equilibrium and calculated therefrom the KE* values. The
ensuing KE* values were badly determined, thus we decided to
follow a different approach to a better evaluation of the
equilibrium constant. Thus, we titrated spectrophotometrically
at RT a solution of the complex [Pd(η2-dmfu)(neoc)] (neoc )
2,2′-dimethyl-o-phenanthroline) ([[Pd(η2-dmfu)(neoc)]]0 ) 1 ×
10-4 mol dm-3) with successive weighed amounts of solid pna.
As we have already stated, the peculiar steric hindrance induced
by the methyl groups of the ligand neoc stabilizes the monoalkyne
derivative [Pd(η2-dmbd)(neoc)] and allows the equilibrium
constant determination.13,14,17 Figure 3 shows the regression
analysis of the titration curve based on the same relationships
of the equilibrium determination previously reported.
Conclusions
The results emerging from this kinetic study rationalize the
reactivity of the monoalkyne derivatives of Pd(0) and indicate
(18) The value KE* ) 0.236 obviously represents a rough estimate of
the equilibrium constant. However, as was already stated, the nature of the
ancillary ligands hardly affects the equilibrium position since their electronic
and steric characteristics equally act in stabilizing (or destabilizing) both
the entering and the leaving groups. It is noteworthy that with KE* ) 0.236,
[1]0 ) 1 × 10-4 mol dm-3 and [pna]0 ) 1.48 × 10-2 mol dm-3, and the
degree of advancement of the reaction is ꢀ ) 0.97.
We took this equilibrium constant (K* ) 0.236 ( 0.004) as
a reasonable indication of the KE*, thus, in the presence of a
strong excess of pna over Pd(0) complex ([pna]0 ) 148-291
× [[Pd(η2-dmfu)(BiPy)]]0) it is possible to rule out the species
[Pd(η2-dmfu)(BiPy)] from the mechanistic network and consider
(19) Formation of the symmetric 3b′′ species was never detected.
(17) Unfortunately, the determination of the equilibrium constant in the
case of the direct exchange between dmfu and dmbd is not possible since
an extensive decomposition takes place upon addition of the titrant dmbd
to a solution of the complex [Pd(η2-dmfu)(neoc)] (ref 14).

4054 Organometallics, Vol. 27, No. 16, 2008
Holuigue et al.
CH3), 3.26 (s, 3H, neoc-CH3), 3.84 (s, 3H, COOCH3), 7.59 (d, 2H,
Ha, J ) 8.9 Hz), 7.67 (d, 1H, H3, J ) 8.3 Hz), 7.77 (d, 1H, H3′,
J ) 8.3 Hz), 7.83 (s, 2H, H5, H5′), 8.18 (d, 2H, Hb, J ) 8.9 Hz),
8.29 (d, 1H, H4, J ) 8.3 Hz), 9.31 (d, 1H, H4, J ) 8.3 Hz).
Pallada-2,5-bis(carbomethoxy)-3,4-bis(4-nitrophenyl)cyclopen-
tadienebipyridine (3b) and Pallada-2,4-bis(carbomethoxy)-3,5-
bis(4-nitrophenyl)cyclopentadienebipyridine (3b′). MS (FAB+)
m/z: found 673.0568 (M + H): calcd (C30H23N4O8Pd) 673.0563.
a novel mechanism for the formation of palladacyclopentadienyl
complexes. The peculiar reactivity of the complex [Pd(η2-
dmbd)(BiPy)], which not only reacts with one further molecule
of dmbd but also yields the palladacyclopentadiene species by
the more efficient bimolecular self-reaction, to the best of our
knowledge implies a completely new mechanism that has never
been proposed before. The presence of the elusive “palladium
naked” Pd(0) species, which was already hypothesized else-
where,10 was also confirmed as a key intermediate in the
mechanistic network depicted in Scheme 2. Moreover a first
hint of the alkyne reactivity order was also provided by the
marked difference between the equilibrium constants related to
the displacement reactions between the alkene dmfu and the
alkynes dmbd and pna. It is quite clear that the coordinating
capability of pna is less strong than that of dmbd, and therefore
the monoalkyne species [Pd(η2-pna)(BiPy)] is less stable and
consequently more reactive than its counterpart [Pd(η2-dmbd-
)(BiPy)]. This difference in reactivity is reflected by the
difference between kc and kc*, the latter being more than twice
higher than the former. The formation of the symmetric and
unsymmetric mellitate can also be traced back to the reactivity
of the complex [Pd(η2-pna)(BiPy)], which reacts with the
symmetric and unsymmetric palladacyclopentadiene species to
give both mellitate compounds and the unsaturated palladium
species according to the mechanism already suggested in another
paper.10
1
3
3b: H NMR (CDCl3, 300.13 MHz): δ 8.79 (d, 2H, H6, J )
3
5.4), 8.09 (m, 4H, H3 and H4), 7.97 (d, 4H, Hm, J ) 8.1), 7.60
(pst, 2H, H5), 7.17 (d, 4H, Ho, 3J ) 8.1), 3.46 (s, 6H, 2 × OCH3).
13C NMR (CDCl3, 125.70 MHz): δ 174.1 (CO), 156.4 (Cꢁ), 155.5
(C2), 155.2 (Cp), 151.6 (C6), 146.3 (CR), 146.2 (Ci), 139.6 (C4),
129.7 (Co), 126.8 (C5), 123.2 (Cm), 122.4 (C3), 51.2 (OCH3).
1
3b′: H NMR (CDCl3, 300.13 MHz): δ 8.79 (d, 1H, H6), 8.15
3
3
(d, 2H, Hm or Hm′, J ) 9.0), 8.13 (d, 2H, Hm′ or Hm, J ) 8.7),
8.15-8.12 (m, 2H, H4 and H6′), 8.08-7.95 (m, 2H, H3 and H5′),
3
7.60 (br, 1H, H5), 7.58 (d, 2H, Ho or Ho′, J ) 9.0), 8.13 (d, 2H,
Ho′ or Ho, 3J ) 8.7), 7.16 (m, 2H, H3′ and H4′), 3.44 (s, 3H,
COOCH3), 3.18 (s, 3H, COOCH3).
Experimental Section
NMR and UV-Vis Spectra and Elemental Analysis. The 1H
NMR spectra were recorded on a Bruker 300 Avance spectrometer.
UV-vis spectra were taken on a Perkin-Elmer Lambda 40
spectrophotometer equipped with a Perkin-Elmer PTP6 (Peltier
temperature programmer) apparatus. Elemental analyses for new
palladacycles are not provided in this paper, but will appear in a
forthcoming paper for a number of very similar compounds that
are part of this series.
Data Analysis. Mathematical and statistical analysis of equilib-
rium and kinetic data was carried out by a nonlinear regression of
locally adapted algorithms written under Scientist environment.
Synthesis of Complexes. The synthesis of the complexes [Pd(η2-
dmfu)(BiPy)], [Pd(η2-fn)(BiPy)],20 [Pd(η2-dmfu)(neoc)],14 [Pd-
(BiPy)(C4(COOMe)4)],9d pallada-2,5-bis(carbomethoxy)-3,4-bis(4-
nitrophenyl)cyclopentadienebipyridine (3b),21 pallada-2,4-bis(car-
bomethoxy)-3,5-bis(4-nitrophenyl)cyclopentadienebipyridine (3b′),21
and pna22 was carried out according to published procedures. Dmbd,
CD2Cl2, CDCl3, and CHCl3 were commercial grade reagents and were
used as purchased.
The symmetric (4b) and unsymmetric (4b′) mellitate from methyl
(4-nitrophenyl)propynoate were isolated as a mixture of products.
1
They were however identified from H and 13C NMR.
1,3,5-Tris(carbomethoxy)-2,4,6-tris(4-nitrophenyl)benzene (4b)
and 1,2,4-Tris(carbomethoxy)-3,5,6-tris(4-nitrophenyl)benzene (4b′).
MS (FAB+) m/z: found 616.1237 (M + H): calcd (C30H22N3O12)
616.1203.
4b: 1H NMR (CDCl3, 300.13 MHz): δ 8.32-8.05 (m, 6H, Har),
7.57-7.19 (m, 6H, Har), 3.26 (s, 9H, OCH3). 13C NMR (CDCl3,
75.48 MHz): δ 166.57 (CO), 148.24 (Cp), 142.97 (Ci), 137.1 (C),
137.0 (C), 130.8 (CH), 123.7 (CH), 52.82 (OCH3).
4b′: 1H NMR (CDCl3, 300.13 MHz): δ 8.32-8.05 (m, 6H, Har),
7.57-7.19 (m, 6H, Har), 3.58 (s, 3H, OCH3), 3.55 (s, 3H, OCH3),
3.21 (s, 3H, OCH3). 13C NMR (CDCl3, 75.48 MHz): δ 166.74 (CO),
166.64 (CO), 166.46 (CO), 148.15 (Cp), 147.70 (Cp), 147.62 (Cp),
143.33 (2 × Ci), 142.97 (Ci), 139.2 (C), 138.3 (C), 137.1 (C), 135.1
(C), 134.6 (C), 133.0 (C), 130.1 (4 × CH), 129.9 (2 × CH), 124.0
(2 × CH), 123.5 (4 × CH), 53.26 (2 × OCH3), 52.42 (OCH3).
NMR Studies. All reactions were preliminarily carried out by
1H NMR technique.
[Pd(η2-pna)(neoc)]. To 0.05 g (0.25 mmol) of neocuproine
dissolved in freshly distilled CH2Cl2 (20 mL) were added 0.12 g
(0.116 mmol) of Pd2dba3 · CHCl3 and 0.490 g (0.24 mmol) of pna
under inert atmosphere (N2). The obtained orange solution was
stirred for 20 min. Activated charcoal was then added, and the
resulting mixture was eventually filtered off on a Celite filter. The
resulting clear orange solution was concentrated under reduced
pressure, and the title complex was obtained as orange microcrystals
upon addition of diethyl ether (0.0839 g, 0.081 mmol, yield 70%).
Determination of the Equilibrium Constant K. The equilibrium
constant K for the reaction:
2
2
[
(
)
]
[
(
)
]
(
)
(
)
Pd η -fn BiPy + dmbd T Pd η -dmbd BiPy + fn
was determined by adding known aliquots of dmbd ([dmbd] )
0-0.06 mol dm-3) to a CD2Cl2 solution of the complex [Pd(η2-
fn)(BiPy)] ([[Pd(η2-fn)(BiPy)]] ) 1 × 10-2 mol dm-3) at 213 K
and recording the signal at 9.05 ppm of the H6pyr proton of the
complex [Pd(η2-dmbd)(BiPy)].
Formation of Palladacyclopentadiene Derivatives. The cy-
clometalation reaction between complex 1 and dmbd carried out
(20) Crociani, B.; Di Bianca, F.; Uguagliati, P.; Canovese, L.; Berton,
A. J. Chem. Soc., Dalton Trans. 1991, 71.
(21) Holuigue, A.; Sirlin, C.; Pfeffer, M.; Goubitz, K.; Fraanje, J.;
Elsevier, C. J. Inorg. Chim. Acta 2006, 359, 1773.
(22) Eckert, T. Ipaktschi, J. Synth. Commun. 1998, 28, 327.
IR (KBr pellets), cm-1: νC-NO2 856; νNO2 1334, 1509; νCdN 1586;
νCdO 1682. H NMR (CDCl3, 298 K, δ ppm): 2.84 (s, 3H, neoc-
1

Formation of Palladacyclopentadiene Complexes
Organometallics, Vol. 27, No. 16, 2008 4055
Scheme 3. Proposed Reaction Path for the Formation of 4b/4b′
× 10-3 mol dm-3. The resulting absorbance spectra were recorded
in the 300-500 nm wavelength interval. The equilibrium constant
was calculated from the absorbance value vs pna concentration taken
at λ ) 435 nm.
under second-order conditions ([1]0: [dmbd]0 ) 1:1, 1:2) was
investigated by dissolving the alkene complex 1 in 0.8 mL of
CD2Cl2 ([1]0 ≈ 1 × 10-2 mol dm-3) and adding the appropriate
aliquot of a mother solution of dmbd ([dmbd]0 ) 0.4 mol dm-3
)
at 213 K. In both cases the complete conversion of 1 into 2a was
observed. Gradually increasing the temperature led to the reaction
products, which were unreacted 1 and 3a in the case of molar 1:1
ratio and 3a and free dmfu in the case of 1:2 molar ratio. The
reactions were followed up to half or total completion by monitoring
the signals of the disappearance of 2a and of the contemporary
appearance of those of the palladacyclopentadiene species 3a. When
the alkyne under study was pna, the cyclometalation reaction was
investigated by dissolving the alkene complex 1 in 0.6 mL of CDCl3
([1]0 ≈ 3 × 10-2 mol dm-3). The alkyne was added at 298 K as
a solid in order to obtain the concentration of ∼9 × 10-2 mol dm-3
in solution. The progress of the reaction toward the products 3b,
3b′ and 4b, 4b′ (in traces) was followed by monitoring the signals
of the disappearance of 1 and 2b and the concomitant appearance
of those belonging to 3b, 3b′, 4b, and 4b′.
Determination of kc and kc*. The rate constant kc was calculated
from three independent determinations at three different concentra-
tions of 1 and dmbd under equimolar second-order conditions ([1]0:
[dmbd]0 ) 1:1 ) 4 × 10-4, 2 × 10-4, 1 × 10-4 mol dm-3) at λ
) 420 nm in CHCl3 (stored on silver foil) at 298 K.
In the case of the reaction between 1 and pna, the alkyne was
added as a solid to 3 mL of a CHCl3 solution of complex 1 ([1]0
≈ 1 × 10-4 mol dm-3) in order to realize a concentration in the
range (1-3) × 10-2 mol dm-3, and the absorbance change was
monitored in the 400-600 nm wavelength interval at different
times.
Determination of k2. To 3 mL of a CHCl3 solution of 1 ([1]0 ≈
1 × 10-4 mol dm-3) placed in the thermostated (298 K) cell
compartment of the UV-vis spectrophotometer were added mi-
croaliquots of a concentrated solution of dmbd. The absorbance
change was monitored in the 300-530 nm wavelength interval at
different times. The rate constant k2 was eventually calculated from
the linear regression analysis of the ensuing kobs values (kobs ) k2
[dmbd]0) obtained from each kinetic experiment performed under
pseudo-first-orderconditions([dmbd]0)30-60[[Pd(η2-dmfu)(BiPy)]]0).
UV-Vis Kinetic Studies. Determination of the Equilibrium
Constant K*. The equilibrium constant K* at 298 K for the reaction
2
2
[
(
)
]
[
(
)
]
(
)
(
)
Pd η -dmfu neoc + pna T Pd η -pna neoc + dmfu
was determined by adding to a 50 mL solution of the complex
[Pd(η2-dmfu)(neoc)] ([[Pd(η2-dmfu)(neoc)]]0 ) 1 × 10-4 mol
dm-3) in CHCl3 increasing amounts of pna as a weighed solid in
order to establish a pna concentration in the range 1 × 10-4 to 5
OM8002398