Mechanism of the oxidative carbonylation of terminal alkynes at the ≡C-H bond in solutions of palladium complexes

Article abstract of DOI:10.1134/S0023158407020073

The formation mechanism of the active catalyst in the oxidative carbonylation of terminal alkynes at the ≡C-H bond has been investigated for the catalytic system Pd(OAc)₂-PPh₃-p-benzoquinone (Q)-MeOH. It has been demonstrated by NMR spectroscopy, X-ray crystallography, and kinetic measurements that the catalytically active palladium is in the oxidation state 0 and is bound into complexes stabilized by p-benzoquinone (PdL₂Q, where L = PPh₃). A mechanism is suggested for the catalytic process, which includes the formation of the complex PdL₂Q, the oxidative addition of the alkyne to this complex at the ≡C-H bond, the insertion of CO into the Pd-C bond, and steps in which hydride hydrogen is intramolecularly transferred to the p-quinone.

Full text of DOI:10.1134/S0023158407020073

ISSN 0023-1584, Kinetics and Catalysis, 2007, Vol. 48, No. 2, pp. 228–244. © MAIK “Nauka /Interperiodica” (Russia), 2007.
Original Russian Text © V.R. Khabibulin, A.V. Kulik, I.V. Oshanina, L.G. Bruk, O.N. Temkin, V.M. Nosova, Yu.A. Ustynyuk, V.K. Bel’skii, A.I. Stash, K.A. Lysenko, M.Yu. Antipin,
2007, published in Kinetika i Kataliz, 2007, Vol. 48, No. 2, pp. 244–260.
Mechanism of the Oxidative Carbonylation of Terminal Alkynes
at the ∫C–H Bond in Solutions of Palladium Complexes
V. R. Khabibulina, A. V. Kulika, I. V. Oshaninaa, L. G. Bruka, O. N. Temkina, V. M. Nosovab,
Yu. A. Ustynyukb, V. K. Bel’skiic, A. I. Stashc, K. A. Lysenkod, and M. Yu. Antipind
a Lomonosov State Academy of Fine Chemical Technology, Moscow, 117571 Russia
e-mail: alkulik@gmail.com
b Moscow State University, Moscow, 119899 Russia
c Karpov Research Institute of Physical Chemistry, Moscow, 103064 Russia
d Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences, Moscow, 117813 Russia
Received January 11, 2006
Abstract—The formation mechanism of the active catalyst in the oxidative carbonylation of terminal alkynes
at the ≡C–H bond has been investigated for the catalytic system Pd(OAc)2–PPh3–p-benzoquinone (Q)–MeOH.
It has been demonstrated by NMR spectroscopy, X-ray crystallography, and kinetic measurements that the cat-
alytically active palladium is in the oxidation state 0 and is bound into complexes stabilized by p-benzoquinone
(PdL2Q, where L = PPh3). A mechanism is suggested for the catalytic process, which includes the formation of
the complex PdL2Q, the oxidative addition of the alkyne to this complex at the ≡C–H bond, the insertion of CO
into the Pd–C bond, and steps in which hydride hydrogen is intramolecularly transferred to the p-quinone.
DOI: 10.1134/S0023158407020073
The oxidative carbonylations of terminal alkynes at
the ≡ë–ç bond (I), which yield esters of arylpropiolic
and alkylpropiolic acids, are of considerable theoretical
and applied interest, since these esters have great syn-
thetic potential [1].
Comparatively selective catalytic systems for reac-
tion (I) were reported for the first time by Tsuji et al.
[2]. The most effective system is PdCl2–CuCl2–
NaOAc, on which the following reactions proceed with
70% alkyne selectivity:
RC≡CH + CO + MeOH + 2NaOAc + 2CuCl2
(II)
PdCl2
RC≡CH + CO + MeOH + Ox
RC≡CC(O)OMe + 2NaCl + 2CuCl + 2AcOH.
RC≡CC(O)OMe + Red + 2H+
.
(I)
It was found earlier [3] that this catalytic system is
multifunctional and reaction (II) is autocatalytic. Cu(I)
chloride, which results from CuCl2 reduction, is the
component that catalyzes the palladium alkynyl com-
plex formation step in mechanism (III):
Various oxidizers (Ox) can be used in this reaction.
In the particular case of Ox = quinone, the oxidizer
reduction product is Red + 2H+ = hydroquinone.
PdCl2
RC≡CH + CuCl
RC≡CCu –CuCl RC≡CPdCl
–HCl
2CuCl2
RC≡CC(O)PdCl MeOH HPdCl
PdCl2 + 2CuCl + HCl
CO
(III)
RC≡CC(O)OMe
Mechanism (III) was confirmed by the synthesis and
reactivity studies of Cu(I), Ag(I), and Hg(II) alkynyl
complexes in reaction (II), by kinetic studies [3, 4], and
by determination of the kinetic isotope effect [5]. Fur-
thermore, these studies made it possible to carry out a
new catalytic reaction using oxygen as the oxidizer and
Pd(II), Cu(I), and Cu(II) as catalysts:
RC≡CH + CO + MeOH + 1/2O2
(IV)
PdCl2, CuCl, CuCl2
RC≡CCOOMe + H2O.
This reaction, which does not yield any acid and,
therefore, does not need a base for acid neutralization,
can provide a basis for small-scale production of esters
of alkynylcarboxylic acids (productivity, ~40 g l–1 h–1;
228

MECHANISM OF THE OXIDATIVE CARBONYLATION OF TERMINAL ALKYNES
229
Table 1. Solvent effects on MEPA synthesis in the
Pd(OAc)2–PPh3–p-quinone–MeOH–solvent systems
(I) involving this system as the catalyst and phenylacet-
ylene as the substrate.
Initial MEPA PA-to-MEPA
p-Quinone
Solvent
CH3CN
formation rate, conversion
PRELIMINARY EXPERIMENT: CHOOSING THE
COMPOSITION OF THE CATALYTIC SYSTEM
mol l–1 h–1
selectivity, %
p-Benzo-
quinone
0.03
0.05
0.05
0.05
0.032
0.042
0.065
0.054
0.36
0.40
0.40
0.40
0.25
51
56
47
82
74
65
51
59
80
81
85
85
72
In our preliminary experiments, we studied how the
oxidative carbonylation of phenylacetylene (PA) into
methyl ester of phenylpropiolic acid (MEPA) depends
on the natures of the ligand, of the solvent, and of the
p-quinone.
CH3COCH3
CHCl3
1,4-dioxane
THF
A p-quinone must be present in the Pd(OAc)2–
MeOH system for reaction (I) to occur. Other oxidizers,
for example, Cu(II) and Fe(III) compounds, do not
ensure the synthesis of MEPA. Besides p-benzo-
quinone, 2,3,5,6-tetrachloro-p-benzoquinore (chlo-
ranil) was tested in the synthesis. The rate and MEPA
selectivity of the process were found to depend strongly
on the quinone (Table 1). Therefore, the quinone plays
a significant role in the process.
C6H6
CH3OH
CH3CN
CH3COCH3
CHCl3
Chloranil
1,4-dioxane
THF
Phosphine admixtures, which serve as efficient
ligands in the Pd(OAc)2–MeOH system, cause different
effects depending on their nature (Table 2). The greatest
speedup of the reaction is observed with PPh3, and this
effect is stronger for the chloranyl-containing system. It
is essential that PPh3 does not decrease the selectivity
of the reaction. Likewise, P(cyclo-C6H11)3 added to the
p-benzoquinone-containing system speeds up the reac-
tion significantly without causing a decrease in the
CH3OH
Note: T = 31°C, P = 1 atm, [Pd] = 0.006 mol/l, [L] = 0.010 mol/l,
CO
[p-quinone] = 0.1 mol/l, and [PA] = 0.1 mol/l.
0
desired-product selectivity, 75–85%, depending on the
alkyne) [7].
For reaction (I), Tsuji et al. [2] suggested the funda- selectivity. For the chloranil-containing system,
mentally different catalytic system Pd(OAc)2–p-benzo- P(cyclo-C6H11)3 affords a greater increase in the reac-
quinone (Q), which seems simpler at first glance. The tion rate but diminishes the selectivity of the reaction.
subject of this publication is the mechanism of reaction P(n-C3H7)3 exerts the opposite effect. Diphosphines
Table 2. Ligand effects on MEPA synthesis in the Pd(OAc)2–p-quinone–MeOH systems
Initial MEPA formation rate,
mol l–1 h–1
PA-to-MEPA
conversion selectivity, %
Quinone
Ligand
p-Benzoquinone
PPh3
0.065
0.01
0.05
0*
0
51
44
69
25
0
P(n-C3H7)3
P(cyclo-C6H11)3
PPh2(2-Py)
Ph2PCH2CH2PPh2
Ph2PCH2CH2CH2PPh2
o-phenanthroline
PPh3
0
14
0
0
Chloranil
0.25
0
72
82
54
40
0
P(n-C3H7)3
P(cyclo-C6H11)3
PPh2(2-Py)
0.14
0
Ph2PCH2CH2CH2PPh2
o-phenanthroline
0
0
0
Note: T = 31°C, P = 1 atm, [Pd] = 0.006 mol/l, [L] = 0.010 mol/l, [p-quinone] = 0.1 mol/l, and [PA] = 0.1 mol/l.
CO
0
* The processes with a zero initial rate have an induction period.
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

230
KHABIBULIN et al.
and nitrogen-containing ligands do not produce a favor-
able effect: with these ligands, PA is not converted into
MEPA. Further studies were carried out with triphe-
nylphosphine (hereafter, L = PPh3).
(3) Mechanisms involving the formation of a key
Pd(II) ethynyl hydrido complex at the stage of the C–H
oxidative addition of the alkyne to Pd(0):
Pd(0) + RC≡CH
RC≡CPdH
MeOH
(VII)
The effect of the nature of the solvent on the process
was again studied for the p-benzoquinone- and chlo-
ranil-containing systems (Table 1). In all runs involving
a solvent, PA is completely converted in 1–2 h. Note
again that the same solvent causes different effects in
systems with different p-quinones. For the system con-
taining p-benzoquinone, the initial reaction rate is lower
in a mixed solvent (MeOH : solvent = 1 : 1 vol/vol) than
in pure methanol. For the chloranil-containing system,
the initial MEPA formation rate is higher in a mixed
solvent than in methanol. Most solvents (except
CH3CN and CHCl3 in the p-benzoquinone-containing
system) enhance the selectivity of the process. The
greatest increase in selectivity is achieved by adding
1,4-dioxane or tetrahydrofuran (THF). The strong
dependence of reaction (I) on the kind of p-quinone and
the fact that a solvent added to methanol produces dif-
ferent effects in systems containing different p-quino-
nes suggest that the catalytically active complexes in
these similar systems contain a p-quinone or that this
p-quinone is involved in reaction steps preceding the
rate-limiting step.
CO
HPdC(O)C≡CR –HPdH RC≡CC(O)OMe.
A first-group MEPA formation mechanism was sug-
gested by Heck [9] in 1972. No second-group mecha-
nisms have been reported. Palladium(II) forms alkynyl
complexes unreadily (much less readily than Cu(I) or
Ag(I); see above). However, some Pd(II) alkynyl com-
plexes do exist, so there is no reason to rule out the for-
mation of such a complex by the electrophilic substitu-
tion of palladium for a hydrogen atom in the alkyne.
There is only very little information concerning the
possibility of the C–H oxidative addition of 1-alkynes
to Pd(0) [10]. For Ni(0) and Rh(0), the possibility of
this step is beyond any doubt [11].
DISCRIMINATION OF HYPOTHETICAL
MECHANISMS
The above hypotheses can be verified and discrimi-
nated by experiments aimed at gaining the following
information:
(1) the state of the catalyst during the oxidative car-
bonylation of alkynes (the oxidation state of palladium
and the composition and structure of palladium com-
plexes);
(2) the kinetic isotope effect (KIE) arising from the
replacement of ëç3éç with ëç3OD or ëD3OD (if a
first-group mechanism takes place, KIE will be far
above unity);
HYPOTHETIC MECHANISMS
OF MEPA SYNTHESIS
IN THE Pd(OAc)2–PPh3–p-QUINONE–MeOH
SYSTEMS
(3) kinetics observed while varying the concentra-
tions of the reactants and of the components of the cat-
alytic system.
When constructing hypotheses, we took into consid-
eration the sets of mechanisms obtained by using a for-
malized procedure involving the ChemNet program [5,
8]. All hypothetic mechanisms can be divided into the
following three groups.
State of the Catalyst
The IR spectrum of the reaction solution in the
mixed solvent MeOH–CHCl3 changes mainly owing to
MEPA bands (CO group, 1717 cm–1; C≡C bond,
2200 cm–1) appearing during the process.
(1) Mechanisms involving a key alkoxycarbonyl
(methoxycarbonyl) intermediate:
PdX2 + MeOH + CO–HX XPdC(O)OMe
1
By contrast, the changes in the H and 31P NMR
RC≡CH
31
spectra are very sophisticated. For example, some P
XPdC(R)=CHC(O)OMe –HPdX RC≡CC(O)OMe.
NMR spectra show ten to twenty time-variable phos-
phorus resonances. Table 3 lists the multicomponent
systems studied by NMR and their designations.
(V)
(2) Mechanisms involving the formation of a key
Pd(II) alkynyl complex at the stage of the electrophilic
substitution of palladium(II) for hydrogen in the
alkyne:
In order to identify the intermediates, we investi-
gated less complicated, model systems, including sys-
tems into which reactants were introduced one after
another. NMR spectra were recorded during the reac-
tion. Table 4 lists chemical shifts (δ) and spin–spin cou-
pling constants (J) for the components of the catalytic
systems and for the complexes obtained by independent
synthesis.
PdX2 + RC≡CH –HX XPdC≡CR
(VI)
CO
MeOH
–HPdX
XPdC(O)C≡CR
RC≡CC(O)OMe.
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

MECHANISM OF THE OXIDATIVE CARBONYLATION OF TERMINAL ALKYNES
231
Table 3. Multicomponent systems studied by 1H, 13C, and 31P NMR spectroscopy
Designation
System
Three-component
1.1
1.2
1.3
1.4
1.5
1.6
[PPh3–Q–CD3OD], 25°C
[PPh3–Q–CDCl3], –25°C
[Pd(OAc)2–Q–CD3OD], 23°C
[Pd(OAc)2–PPh3–CD3OD], 25°C
[Pd(OAc)2–2PPh3–CDCl3], –25°C
[Pd(PPh3)4–10Q–CDCl3], –25°C
Four-component
Five-component
2.1
2.2
2.3
[Pd(OAc)2–PPh3–Q–CH3OD], 25°C
[Pd(OAc)2–PPh3–Q–CD3OD], 27°C
[Pd(PPh3)4–Q–CO–CD3OD], 30°C
3.1
3.2
[Pd(OAc)2–PPh3–Q–CO–CH3OD], 30°C
[Pd(OAc)2–PPh3–Q–CO–CD3OD], 30°C
Six-component
4.1
4.2
[Pd(OAc)2–PPh3–Q–CO–CH3OH–CDCl3], –25°C
[Pd(PPh3)4–Q–CO–PhC≡CH–CH3OH–CDCl3], –25°C
Seven-component
5.1
[Pd(OAc)2–PPh3–Q–CO–PhC≡CH–CH3OH–CDCl3], –25°C
Systems into which the reactants were introduced in sequence
[Pd(OAc)2–Q–CD3OD] + [PPh3], 23°C
[Pd(OAc)2–PPh3–CD3OD] + [Q], 23°C
[Pd(OAc)2–PPh3–Q–CD3OD] + [CO], 24°C
[Pd(OAc)2–PPh3–CD3OD] + [Q] + [CO], 26°C
[Pd(OAc)2–PPh3–CD3OD] + [Q–CO–PhC≡CH], 27°C
[Pd(OAc)2–CDCl3] + [PPh3] + [Q], –25°C
[Pd(OAc)2–3PPh3–CDCl3] + [CH3OH] + [Q] + [CO], –25°C
[PPh3–Q–CDCl3] + [CH3OH] + [CO] + [PhC≡CH], –25°C
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
Note: The components that were mixed or added together are bracketed.
Q = p-benzoquinone.
The actual component ratios in the reactions studied their derivatives, including the products resulting from
their interaction with PPh3 and palladium acetate.
by NMR were derived from total integrated intensities
in the following three 1H NMR spectral regions:
Palladium acetate (I) is known to be readily reduc-
ible with triphenylphosphine (II) [12]; carbon monox-
ide [13–16]; and, probably, methanol, all of which are
present in the catalytic system. Indeed, we observed
triphenylphosphine oxide (III) signals in nearly all
reaction systems, although the amount of III did not
exceed 4% of the amount of triphenylphosphine
involved in the reaction (see below). The 31P chemical
shift for O=PPh3 depends strongly on the solvent, the
concentration of the compound, and temperature. The
true position of the triphenylphosphine oxide signal
was determined against 13C satellites (1JP–C = 102–
104 Hz) in the 31P NMR spectrum and from character-
istic chemical shifts and constants in the 13C NMR
(1) 1–2.3 ppm, where the signals from the OAc
groups of various palladium complexes and the signal
from the OAc– anion are observed;
(2) 7–7.9 ppm, the aromatic range accommodating
the signals from the phenyl groups of triphenylphos-
phine, both free and coordinated to palladium in vari-
ous complexes, and from the phenyl groups of the other
PPh3-containing compounds; integration in this range
allows the total amount of triphenylphosphine involved
in the reaction to be determined;
(3) 5.2–7.1 ppm, the range accommodating the sig-
nals from p-benzoquinone, hydroquinone, and all of spectrum.
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

232
KHABIBULIN et al.
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

MECHANISM OF THE OXIDATIVE CARBONYLATION OF TERMINAL ALKYNES
233
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

234
KHABIBULIN et al.
nal from triphenylphosphine oxide (1–2%, 30.98 ppm)
+
and several weak signals from unidentified compounds
(~3%) are present in the spectrum along with the sig-
nals due to VI. Unreacted triphenylphosphine (16–
26%, –6.37 ppm) was also detected in system 6.7.
Complex VI was synthesized using a standard proce-
PPh3
Ph3P Pd OAc
PPh3
OAc–
Fig. 1. Presumable structure of the complex forming in sys-
tem 1.4.
1
dure [21] and was characterized by H, 13C, and 31P
NMR, so its identification based on NMR data for cat-
alytic systems in CDCl3 was not a particular problem.
In the 1H NMR spectrum of VI, the methyl signal of the
acetoxy group (0.806 ppm) is split into a triplet because
of the spin–spin coupling of the protons with the phos-
Two phosphorus signals not assignable to palladium
complexes (21.7 and 16.9 ppm) appeared in the NMR
spectra of all systems containing triphenylphosphine,
excess p-benzoquinone, and methanol (Table 3, sys-
tems 2.1–2.3, 3.1, 3.2, 6.1–6.5). As was demonstrated
in earlier works [17–19], the interaction between triph-
enylphosphine and p-benzoquinone is a 1,4-addition
reaction yielding (2,5-dihydroxyphenyl)phosphonium
betaine (IV); however, no NMR data for IV were
reported in those works. We found that the main reac-
tion product in the simplest system 1.1 (PPh3–Q–
CD3OD, +25°C) is indeed compound IV, which is
phorus nuclei of the two phosphine ligands (5JH–P
=
0.51 Hz).
In CD3OD (system 1.4, PPh3: Pd ~ 3.5), the domi-
nant complex at 23–27°C is VII, which likely forms
from complex VI. The 31P NMR spectrum of this sys-
tem shows two broad signals (halfwidth, ~10 Hz; total
amount of phosphorus nuclei, 90%), one at 30.81 ppm
(triplet, 1P) and the other at 29.32 ppm (doublet, 2P),
both arising from the spin system Aï2 (JP–P = 18.1 Hz),
as well as signals from free triphenylphosphine (4.2%,
broad singlet at –4.47 ppm) and triphenylphosphine
1
readily identifiable by H NMR. The spectrum of IV
exhibits signals from three nonequivalent protons at
5
3
6.13 ppm (4JHH= 3.18 Hz, JHH = 0.38 Hz, JêH
=
1
15.02 Hz), 6.60 ppm (3JHH = 9.04 Hz, 5JHH = 0.38 Hz,
oxide (3.8%, 33.66 ppm). The H NMR spectrum
4JêH = 7.23 Hz), and 7.01 ppm (3JHH = 9.04 Hz, 4JHH
=
exhibits broad and smooth signals from phenyl protons
(7.08–7.55 ppm) and two signals from acetate groups
(a broad one at 1.87 ppm and the other at 0.29 ppm),
and the ratio of these signals is 45 : 3 : 3. These data
suggest that the molecule of complex VII contains
three phosphine ligands and two acetoxy groups, one of
which is coordinated to the palladium atom.
3.18 Hz, 5JêH = 0.71 Hz). In the 31P NMR spectrum of
the reaction mixture, compound IV gives rise to a reso-
nance at 21.7 ppm. In the reaction mixture containing
excess benzoquinone, IV is converted into another
compound (V), which manifests itself as a 31P reso-
nance at 16.9 ppm. The structure of V is being deter-
mined now.
An AX2 spin system in the 31P NMR spectrum was
also observed for the planar cationic complex
[(PPh3)3PdH]+ and for the complex resulting from the
interaction between palladium diacetate and triphe-
nylphosphine in a highly polar medium (80%
CF3COOH in water) [22]. It was assumed that the latter
is also a cationic complex ([(PPh3)3PdII(ï)]+ or
[(PPh3)3PdII(S)]2+, where X = Oac or éëéëF3 and S is
probably an ç2é molecule) [22]:
The nature of the complexes resulting from the
interaction between palladium diacetate and triphe-
nylphosphine was found to depend strongly on the sol-
vent and on the PPh3 : Pd ratio.
Amatore et al. [12, 20] studied the interaction
between Pd(OAc)2 and a tenfold excess of PPh3 in THF
and dimethylformamide as solvents. It was demon-
strated that the complex Pd0(PPh3)(OAc)– initially
results from the reaction between Pd(OAc)2 and 2 mol
of PPh3 in these solvents and that this complex reacts
with excess triphenylphosphine to yield the complex
Pd0(PPh3)3(OAc)– (which has a pyramidal structure). It
was noted that, immediately after mixing Pd(OAc)2with
10 mol PPh3 in THF, signals from the complex
PdII(PPh3)2(OÄÒ)2 (VI) (14.48 ppm) and triphenylphos-
phine oxide are also observed in the 31P NMR spectrum.
80%, CF3COOH ‚ H2O
Pd(OAc)2 + PPh3
[(PPh3)3PdIIX]+
or [(PPh3)3PdII(S)]2+.
A planar Pd(II) complex with three triphenylphos-
phine ligands and one acetoxy group in the coordina-
tion sphere of palladium can also form in system 1.4.
The second OAc group is likely an anion (Fig. 1).
The signals of complex VII are observed in the 1H
and 31P NMR spectra of systems 6.2, 6.4, and 6.5
immediately after mixing Pd(OAc)2 with PPh3(L : Pd =
5–8). For all four systems in methanol-D4, complex VII
is the dominant palladium-containing product before
quinone and CO are introduced. Besides the signals of
VII, triphenylphosphine oxide, and unreacted triphe-
nylphosphine, two unidentified signals at 16.3 and
The interaction between Pd(OAc)2 and PPh3 in
CDCl3 at 25°C (Table 3, system 1.5: 2 mol PPh3; sys-
tem 6.7: 3 mol PPh3) yields compound VI (70–80%,
1
singlet at 14.92 ppm in the 31P NMR spectrum). A sig-
1
Hereafter, when considering a compound present in the reaction
mixture, we present the percentage of its phosphorus nuclei in the
total number of phosphorus nuclei. From these data, one can
readily derive the molar concentration of this compound in the
reaction mixture.
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

MECHANISM OF THE OXIDATIVE CARBONYLATION OF TERMINAL ALKYNES
235
O2q
C4q
C5q
C6q
C3q
O1h
O1q
H1h
C2q
C1q
C1h
H1m
C2h
O1m
C1m
P2
C3h
Pd2
Pd1
P1
Fig. 2. Molecular structure of the complex [Pd(PPh ) Q] –H Q · MeOH (X).
3 2
2
2
31
17.9 ppm (total intensity, 3%) are observed in the P
Complexes between Pd(0) and p-benzoquinone
have long been known, but hypotheses as to the struc-
ture of Pd(0)–PPh3 compounds have been based only
on indirect data [24, 25]. There is no information con-
cerning compounds between palladium and chloranil.
For identification of palladium complexes with
p-quinones and triphenylphosphine in catalytic solu-
tions, it was necessary to obtain, by independent syn-
thesis, complexes that could form in the [Pd(OAc)2–
PPh3–Q–MeOH] and [Pd(PPh3)4–Q–MeOH] systems.
Earlier, we synthesized several palladium complexes
with p-benzoquinone, triphenylphosphine, and p-hyd-
roquinone (IX) and studied their structures by X-ray
NMR spectrum. The signal at 16.3 ppm is likely due to
complex VI.
In all the [Pd(OAc)2 + PPh3+ methanol-D4] systems
examined, including the systems containing benzo-
quinone and CO (1.4, 6.2, 6.4, and 6.5), the amount of
triphenylphosphine oxide does not exceed 4% of the
total concentration of phosphorus-containing com-
pounds (according to 31P NMR data); therefore, triphe-
nylphosphine is not the main Pd(II) reducer. In sys-
tem 1.5, the main product of the interaction between
palladium diacetate and triphenylphosphine in CDCl3
is complex VI. Therefore, palladium is not reduced to
any significant extent because there is no nucleophile crystallography [26]. In the study reported here, we
(OAc–, MeOH, OMe–) necessary for PR3 oxidation characterized these complexes in detail by NMR spec-
with palladium.
troscopy.
The red complex [Pd(PPh3)2Q]2–H2Q · MeOH (X),
p-Benzoquinone (VIII) does not form any complex
with palladium diacetate in methanol (NMR data for
system 1.3, IR spectroscopic data), chloroform (IR
spectroscopic data), or acetone (IR spectroscopic data).
As far as we know, the literature contains no informa-
tion concerning complexation between Pd(II) and
p-benzoquinone. Only one complex between Pd(I) and
p-benzoquinone has been reported [23], and its struc-
ture has not been determined.
where ç2Q is hydroquinone (Table 4, Fig. 2), was
obtained by reacting Pd(PPh3)4 with hydroquinone in a
methanol solution in air at room temperature. The
p-benzoquinone that appears in this compound results
from the oxidation of p-hydroquinone with atmo-
spheric oxygen. According to X-ray diffraction data,
the structure of X is composed of two Pd(PPh3)2Q mol-
Table 5. NMR parameters for the ABX spin systems of the phenyl carbon atoms in complex X
Spin system ABX
Spin–spin coupling constant, Hz
X
A
B
13C nucleus
δ, ppm
31P(13C), ppb
31P(12C), ppb
2JAB
JAX
JBX
ipso
133.02
132.15*
133.73
128.13
0
14
2
30.45
35.1*
30.45
30.45
33.95, 1J
34.7*
–13.5, 2J
9.7, 3J
0.55, 3J
0.65*
–0.5, 4J
0.01, 5J
ortho
meta
0
0
0
0
Note: CDCl , 27°C, Bruker AM-360, 90.568 MHz.
3
* CDCl , –25°C, Bruker DPX-300, 75.468 MHz.
3
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

236
KHABIBULIN et al.
meta-
orto-
*
*
para-
10000
9980
9960
9940 H, Hz
10 Hz
ipso-
*
134.0 133.5 133.0 132.5 132.0 131.5 131.0 130.5 130.0 129.5 129.0 128.5 128.0 δ, ppm
13
Fig. 3. C NMR spectrum of complex X in the phenyl range. The signals of triphenylphosphine oxide are starred. The arrows point
to the 9th and the 14th lines in the X-part of the ABX spin system of the ipso carbon atom.
ecules linked by a p-hydroquinone bridge through The signal intensity ratio Ph : Q = 30H : 4H is consis-
tent with the structure of complex X. Furthermore, the
hydrogen bonding (Fig. 2). Pure complex X dissolved
in CDCl3 at a low temperature in an argon atmosphere
1H NMR shows a p-hydroquinone signal at 6.58 ppm.
gives rise to a 31P resonance at 31.17 ppm (92.6%). The
For X and other palladium complexes containing
two triphenylphosphine ligands, the 13C NMR spectral
region pertaining to the phenyl carbon nuclei of these
ligands deserves special attention (Fig. 3).
1H NMR signal of X occurs at 7.14–7.32 ppm. The sig-
nal from the protons of coordinated p-benzoquinone
31
'
'
(A2A2 part of the A2A2 XX' spin system (X, X' = P))
The 13C NMR spectrum displays signals from all
four nonequivalent types of carbon atoms, which are
the X-parts of ABX spectra (A, B = 31P). These spectra
arise from isotopomers each containing a single 13C
isotope (this is possible because the natural abundance
of this isotope is as low as 1.1%). As a consequence, the
is a triplet occurring at 5.35 ppm with 1.98-Hz splitting.
13
phosphorus atoms are nonequivalent. The C/12C iso-
O1
tope effect on 31P estimated by analysis of the AB part
of the spin system in the phosphorus spectrum is 14
2 ppb (Table 5). The profile of the ipso-13C multiplet is
the most informative. An analysis of this multiplet
allowed us to determine JAX and JÇX. As was expected,
these parameters are very different. Therefore, the pro-
files of the multiplets from the phenyl carbon nuclei in
triphenylphosphine complexes of palladium and other
transition metals provide a reliable analytical means for
determining the number of phosphine ligands coordi-
nated to the metal atom.
C6
C1
C4
C2
C3
C5
O2
Pd2
O3
P2
P1
Pd1
C12
C7
C8
C9
C11
C10
O4
The dissolution of complex X in acetone at room
temperature yields black crystals of the compound
[Pd(PPh3)Q]2 · 1.5MeOH (XI), in which the Pd(PPh3)frag-
ments are linked by two p-benzoquinone bridges (Fig. 4).
Fig. 4. Molecular structure of the complex [Pd(PPh )Q] ·
3
2
1.5MeOH (XI).
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

MECHANISM OF THE OXIDATIVE CARBONYLATION OF TERMINAL ALKYNES
237
1
+
+
This structure of complex XI is confirmed by H,
Ac
O
Pd Pd
Me
O
Pd Pd
13C, and P NMR data. The H NMR spectrum of XI
(spin system AA'MM'XX', where X and X' are phos-
phorus nuclei) shows two signals with different multi-
plicities from the protons of the CH groups of p-benzo-
quinone coordinated to palladium. These signals occur
at 4.34 ppm (4H) and 4.02 ppm (4H). Therefore, the
asymmetric structure characteristic of crystals of XI
persists in the solution. The 13C NMR spectrum of XI
shows characteristic split signals of aromatic carbon
atoms, indicating that the molecule contains two triph-
enylphosphine ligands (see above) characterized by a
spin–spin coupling constant of JP…P = 20–50 Hz,
although these ligands are coordinated to different pal-
ladium atoms.
31
1
L
L
L
L
L
L
O
OAc
O
OAc
Ac
Me
XIII
XIV
Fig. 5. Presumable structures of the complexes forming in
systems 6.1 and 6.2.
from three nonequivalent phosphorus nuclei forming an
AMX spin system, namely, a doublet at 35.5 ppm (A),
a doublet at 28.4 ppm (M), and a singlet at 23.2 ppm
(X) with spin–spin coupling constants of J(PA–På) =
20.4 Hz, J(PA–Pï) = 1.02 Hz, and J(På–Pï) = 0.46 Hz.
The small values of JÄï and JÇï indicate the formation
of a dinuclear complex containing Pd(PPh3)2 and
Pd(PPh3)ï fragments, where X is likely an OAc group.
The signals of this complex are missing in the spectra
of the similar systems 6.6 (CDCl3) and 6.7 (CDCl3 +
CH3OH before CO blowing) at –25°ë. Apparently,
under these conditions, CDCl3 prevents the formation
of this complex even if methanol is added to the system.
The above results suggest that the complexes form-
ing in systems 6.1 and 6.2 are Pd(II) complexes with
structures ïIII and ïIV (Fig. 5).
When the reactants are combined successively (sys-
tems 6.1, 6.2, and 6.4), the amount of complex ïIII is
25–40% (as is estimated from the total integrated
intensity of the signals from the three phosphine
ligands). When the four reactants are combined at once
(systems 2.2 and 6.3), the amount of complex ïIII is
only 9–12%.
Complex X forms in a number of model systems. In
system 1.6 (êd(PPh3)4–p-benzoquinone (tenfold
excess)–CDCl3, –25°C), it accounts for ~48% of the
total organophosphorus compounds content (according
to 31P NMR data). Furthermore, the 31P NMR spectrum
indicates the presence of triphenylphosphine oxide
(12.4%) and of the complex Pd(PPh3)2Cl2 (XII) (signal
at 23.89 ppm, 6.3%) in this system. The spectrum
exhibits a broad signal at 21.0 ppm (30.3%), which is
likely due to the triphenylphosphine ligands in
Pd(PPh3)4. The 31P NMR signal of this complex in ben-
zene occurs at 19.3 ppm. This assignment is confirmed
1
by intensity analysis of the H NMR signals of these
compounds in the same system.
The 31P NMR spectrum of the model system
[Pd(OAc)2–Q–PPh3–CD3OD], which does not contain
PA or CO, shows signals from complexes X (systems 6.1–
6.5) and XI (systems 6.1 and 6.3). The 1H NMR spec-
trum of this system shows a triplet at 5.434 ppm due to
complex X and two multiplets at 4.02 and 4.34 ppm due
to complex XI. The concentration of the red complex X
in methanol (CD3OD, 25°ë) is 3–14%, depending on
the system. The concentration of XI in the methanolic
solutions does not exceed 3%.
The above results together with relevant data
reported in the literature suggest that, even before CO
is introduced into the reactor, Pd(II) in the reaction
solution is converted into Pd(0)–Q complexes like ï
and ïI and into a complex like ïIII or ïIV.
For systems 6.1, 6.2, and 6.4 (before passing CO),
In view of the data concerning the synthesis of com-
irrespective of the order in which triphenylphosphine and plex VI, the sequence of reactions in the Pd(OAc)2–
p-benzoquinone were added to Pd(OAc)2 (PPh3 : Pd = PPh3–Q–methanol system can be represented as fol-
5−8), the 31P NMR spectrum shows narrow signals lows:
Pd2L3X3+
L2PdQ
MeOH, Q
(3)
2L
(1)
Q
(2)
Pd(OAc)2
Pd(OAc)2L2
VI
XIII, XIV
X or XI
excess L
(4)
Q
PdL3(OAc)+OAc–
(5)
VII
Scheme.
Dimeric, presumably Pd(II), complexes XIII and result from the reduction of Pd(II) in the dimeric cat-
XIV coexist with Pd(0) complexes, which most likely ionic complexes by methanol (reaction (3)). The
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

238
KHABIBULIN et al.
P4
C80
O9
H5O
O5
C79
C73
C74
C78
C75
O10
Pd2
C85
C77
C76
P2
O2
O1
Pd1
O7
O4
O3
C81
H4O
C83
C82
O6
O8
C84A
H823
P1
Fig. 6. Molecular structure of the complex [Pd(COOCH )(OAc)(PPh ) ] H Q · ëH OH (XVI). The disordered methanol molecule
3
3 2 2
2
3
is not shown.
quinone involved in steps (2) and (3) plays a significant phosphine-free complex results from the replacement
role: it favors the binding of part of the PPh3 into of the triphenylphosphine ligands by carbon monoxide
betaines IV and V in systems 6.1, 6.2, and 6.4 and the in X. The amount of complex XV in systems 3.2, 6.3,
formation of the dimeric complex; as a ligand, it stabi- and 6.4 is 14 to 60%. Complex XV is stable in solution.
lizes Pd(0). In systems 6.1 and 6.2, 31–51% of the PPh3 After systems 6.3 and 6.4 were stored in the cold for a
31
reacts with quinone to yield betaines. Complex VII short time, their P NMR spectra indicated the same
forms via reaction (4) in methanol, a polar solvent, in amounts of the complex.
the presence of excess phosphine. It also turns into
dimeric complexes upon addition of Q (scheme, reac-
tion (5)).
The concentration of hydrogen ions (acid) in the
systems examined is very low. The only process that
can cause the acid concentration in the reaction sys-
tem to exceed the acid concentration in the initial
methanolic solution is palladium reduction involving
PPh3 and methanol, CO and H2O, or CO and metha-
nol, for example,
The formation of palladium methoxide complexes
from the cationic dimer is beyond question, since meth-
anol is oxidized to formaldehyde and then formic acid
even by palladium chloride via the formation of palla-
dium methoxide [27].
Pd(OAc)2 + CO + 2MeOH
(VIII)
Contrary to the reports on Pd(II) reduction with
PPh3 in THF and DMF [12, 20], PPh3 in methanol
apparently does not play any significant role in Pd(II)
reduction at 25°C.
Pd(0) + O=C(OMe)2 + 2AcOH.
The mechanism of reaction (VIII) includes the for-
mation of an intermediate methoxycarbonyl complex,
whose decomposition yields dimethyl carbonate and
Pd(0) [15, 16, 28]. The methoxycarbonyl complex can
also be an intermediate in the process examined (see the
first-group hypotheses). The dimeric methoxycarbonyl
As CO is passed through the reactor (systems 6.3
and 6.4), the signals from complexes X and XI disap-
pear. A signal at 19.3 ppm appears in the 31P NMR
spectrum. The carbon signals corresponding to this sig-
nal (X-multiplets of the ortho and meta carbon atoms of
ABX spin systems; see above) indicate that the new
complex (XV) contains a Pd(PPh3)2 fragment and its
composition can be, e.g., (ëé)2Pd(PPh3)2 or (êPh3)Pd–
Pd(PPh3). This complex does not contain quinone as a
complex
[Pd(COOMe)(OAc)(PPh3)2]2C6H4(OH)2
(ïVI) was isolated from the Pd(OAc)2–PPh3–Q–
MeOH system treated with CO (Fig. 6; see Experimen-
tal).
The compound XVI was characterized by X-ray
ligand. A Pd(0) benzoquinone complex containing no crystallography and H, 13C, and 31P NMR spectros-
1
1
phosphine ligands presumably forms along with com- copy. The H NMR spectrum of XVI (Table 4) shows
1
plex XV, as is indicated by a H singlet at 5.34 ppm signals from p-hydroquinone (6.624 ppm, 4H, CH=;
from coordinated p-benzoquinone. It is likely that the 9.146 ppm, 2H, OH). As deduced from the integrated
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

MECHANISM OF THE OXIDATIVE CARBONYLATION OF TERMINAL ALKYNES
239
Table 6. Effects of acid and base admixtures on the param-
eters of reaction (I) with the catalytic system Pd(OAc)2–
PPh3–Q–MeOH
Thus, CO brings about the formation of complex
XVI and its decomposition in the presence of a
p-quinone—another route for the formation of Pd(0)
complexes:
Initial MEPA
formation rate,
mol l–1 h–1
PA-to-MEPA
conversion
MeOH
PPh3, Q
n*
pH
6.0
Pd(II)
(PPh3)nPdQ
selectivity, %
PPh3, CO
MeOH
CO
Reference system
MeOH, Q
–
0.07
51
XVI
(PPh3)2Pd(CO)2 XV
QPd(CO)m
Admixture of CH3COOH
1
2
3
6.0
5.9
6.0
0.12
47
45
46
(presumably)
0.10
Likely Mechanisms of Reaction (I) (Ox = O)
0.08
The concentration of hydrogen ions in the system
and its effect on the parameters of the process are the
major discriminating factors governing the possibility
of Pd(0) oxidation by a p-quinone (the oxidation poten-
tial of p-quinones decreases greatly as pH is raised to
5–6) and the realizability of first- and second-group
hypotheses. pH measurements during the processes
demonstrated that, at the early stages of the reaction,
the solution acidity does not rise and pH remains invari-
able (~6). Investigation of the effects of acetic acid and
sodium acetate admixtures on the pH value of the reac-
tion solution and on the process parameters demon-
strated that the system as a whole has the properties of
a buffer and all its components can react with acetic
acid (Table 6). The pH of the system remains almost
invariable as 1 to 4 mol of CH3COOH per mole of pal-
ladium is added. The process parameters also vary only
slightly (Table 6). Only addition of 2 mol of the strong
acid CF3COOH per mole of palladium reduces pH to
0.4 and completely terminates the carbonylation reac-
Admixture of CF3COOH
0.00
2
0.4
0
Admixture of CH3COONa
1
2
4
6
7.0
7.3
7.5
7.7
7.7
0.07
0.12
0.13
0.10
0.09
55
53
61
59
62
10
* Number of moles of an acid or a base per mole of Pd(II) at
[PPh ] : [Pd] = 2.
3
Σ
line intensities in the proton spectrum, the complex
contains two acetoxy groups (0.894 ppm, 6H) and four
triphenylphosphine ligands (7.62–7.71 ppm, 24H,
ortho protons; 7.32–7.44 ppm, 36H, meta and p pro-
tons) per p-hydroquinone molecule. The 13C NMR
spectrum of XVI exhibits signals from all these func- tion. These data are consistent with the third-group
mechanisms, since the acetic acid resulting from palla-
dium(II) acetate reduction is neutralized through reac-
tions with the components of the catalytic system.
Likewise, adding up to 10 mol of a base
(CH3COONa) per mole of palladium causes only slight
changes in pH and in the initial rate and selectivity of
the process (Table 6). As pH is changed from 6 to 7.7,
the process parameters do not change significantly.
These results are most consistent with the third-group
tional groups, and their intensities are consistent with
the proton spectrum.
According to X-ray structure determination data
(Fig. 6; see Experimental), complex XVI is a dimer in
which two Pd(COOMe)(OAc)(PPh3)2 fragments are p-
hydroquinone-bridged through hydrogen bonds
between the hydroxyl hydrogen atoms of p-hydro-
quinone and the oxygen atoms of the acetate groups.
Table 7. Effect of the nature of the initial palladium compound on the parameters of reaction (I)
Quasi-steady-state
PA-to-MEPA conversion
selectivity, %
Catalyst precursor
Pd(PPh3)2Cl2
Reaction time, min
reaction rate, mol l–1 h–1
0.000
0.043
0.070
0.080
0.042
0.031
120
200
127
104
184
219
Traces
58
Pd(OAc)2 + 2PPh3
Pd(PPh3)4
54
[Pd(PPh3)2Q]2H2Q · MeOH
[Pd(PPh3)Q]2 · 1.5MeOH + 2PPh3
Pd(dba)2 + 2PPh3
63
68
72
Note: T = 31°C, P = 1 ata, [Pd] = 0.006 mol/l, [PA] = 0.1 mol/l, [Q] : [Pd] = 17.
CO
0
dba = dibenzylideneacetone.
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

240
KHABIBULIN et al.
form of the catalyst. Therefore, the p-quinone does not
oxidize Pd(0) in the process examined. The presence of
complex X (48%) in the Pd(PPh3)4–10Q–CDCl3 solu-
tion (system 1.6) is further evidence that excess p-
quinone does not cause Pd(0) oxidation, apparently
because of the nearly complete absence of hydrogen
ions.
Comparing the catalytic activities of the initial com-
pounds (precursors) containing Pd(II) or Pd(0) suggests
that the compounds tested show similar catalytic prop-
erties in reaction (I) (Table 7). The most plausible
explanation for this fact is that all precursors but
Pd(PPh3)2Cl2 turn into the same or similar compounds.
These compounds are most likely Pd(0) p-quinone
complexes containing a PdLnQ fragment. Since reduc-
ing agents (CO, MeOH, PPh3) are present in the system
and the p-quinones have a low oxidation potential, the
reduction of Pd(II) to Pd(0) is much more probable than
the reverse process.
C18
C10
Cl6
C6
P1
C200
C20A
C10A
O3
O11
C13
Cl5
Pd1
C7B
O2
C1B
C14A
Cl2
C4B
P2
C12
C5
The above results suggest that the third-group mech-
anisms are the most likely. We have not detected any
hydrido complexes of palladium in the catalytic solu-
tion. It is possible that, after phenylacetylene (XVII) is
coordinated to palladium, its proton is rapidly trans-
ferred intramolecularly to the quinone coordinated to
the same metal atom, involving or not involving an
intermediate hydride, to yield the alkynyl–p-hydro-
Fig. 7. Molecular structure of the complex
(PPh ) ClPdC Cl O (OMe) (XVIII).The solvate molecule
3 2
6
2 2
of 2,5-dichloro-3,6-dimethoxy-1,4-benzoquinone is not
shown.
quinolate
palladium(II)
derivative
Pd(O–
C6H4OH)(C≡CPh).
mechanisms, since the first- and second-group mecha-
nisms imply that the reaction rate depends significantly
on the proton concentration (they involve reversible
and, probably, quasi-equilibrium steps in which H+ is
released). The termination of the process by CF3COOH
in a third-group mechanism can be caused by the
decomposition of the palladium alkynyl complex
resulting from the oxidative addition of the alkyne to
Pd(0) at the C–H bond.
The complex (PPh3)2ClPdC6Cl2O2(OMe) (XVIII),
which was isolated from the Pd(OAc)2–PPh3–MeOH–
CO–chloranil system, provides further evidence that
Pd(0) is present in the reaction system examined. The
most likely route for the formation of XVIII is the oxi-
dative addition of chloranil to the Pd(0) complex at the
C–Cl bond followed by the substitution of a methoxy
group for a chlorine atom in the chloranil molecule.
The
composition
and
structure
of
the
(PPh3)2ClPdC6Cl2O2(OMe) complex were determined
by X-ray crystallography (Fig. 7; see Experimental).
It is possible that the weak dependence of the pro-
cess parameters on the acidity of the reaction solution
is evidence that the p-quinone protonation step does not
play any significant role in the oxidation of the reduced
As was noted above, the first-group hypotheses can
be checked by KIE measurements. The kinetic experi-
ments carried out for the Pd(OAc)2–PPh3–chloranil–
CH3OH (CH3OD)–chloroform (95%) system demon-
strated that KIEH/D is close to unity (Table 8). This
result rules out the first-group hypotheses.
Table 8. MEPA formation rates in the system Pd(OAc)2–
chloranil–PPh3–CH3OH/CH3OD–chloroform (95%)
The totality of our data allow us to exclude from
consideration the hypotheses involving a Pd(II) com-
pound as an active species, the steps involving protons,
and Pd(0) reoxidation with p-benzoquinone (because of
the insufficient proton concentration) and to suggest a
reaction mechanism consistent with the third-group
hypotheses.
The following two schemes of the catalytic cycle
can be suggested:
(1) Hydridoalkynyl mechanism (Fig. 8, outer cycle).
PA adds oxidatively to the Pd(0)–PdL2Q intermedi-
ate to yield an organopalladium hydride for which two
Reaction rate,
mol l–1 h–1
Standard
deviation
Alcohol
CH3OH
Average
0.22
0.25
0.21
0.21
0.20
0.17
0.20
0.19
0.20
0.02
CH3OD
0.19
0.01
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

MECHANISM OF THE OXIDATIVE CARBONYLATION OF TERMINAL ALKYNES
241
Pd(II)
L, CO
Q
QH2
PhC≡CH
L2PdQ
Q, L
1
PhC≡CH
QH2
Q
L2Pd(HQ)C≡CPh
L2Pd(H)(QH)
2
L
H Pd(Q)(C≡CPh)
H2Pd(Q)L
L
L
CO,
MeOH
PhC≡CCOOMe
L
PhC≡CCOOMe
CO
H Pd(Q)(COC≡CPh)
L
MeOH
Fig. 8. MEPA synthesis mechanism including the oxidative addition of PA to Pd(0) at the ≡ë–ç bond. Q = p-quinone and L = PPh .
3
further conversion pathways are possible, namely, the show similar catalytic activities are evidence that
oxidation of the hydrido ligand by the p-quinone, yield- LnPdQ or (CO)mPdQ palladium(0) complexes partici-
ing palladium hydroquinolate (Fig. 8, mechanism 2), pate in reaction (I) as catalysts. The role of p-quinone
and the insertion of carbon monoxide into the palla- in this system is not only to oxidize the hydrido com-
dium–alkynyl σ-bond. The alcoholysis of the Pd–C(O) plexes of palladium. It is likely that p-quinone is
bond in the propioloyl derivative of palladium results in involved in the reaction steps yielding the active Pd(0)
the decomposition of the organopalladium intermediate complex and, as a ligand, in the alkyne activation step,
into molecules of the product (MEPA, XIX) and the followed by the intramolecular transfer of hydride
dihydrido p-quinone complex (Q)LPdH2. The hydrido hydrogen or a proton from the alkyne to the oxygen
ligands are oxidized by the coordinated p-quinone mol- atoms of coordinated Q.
ecule, and the phosphine ligands and p-quinone present
in the system ensure the regeneration of the L2PdQ
complex.
EXPERIMENTAL
(2) Hydroquinolate–alkynyl mechanism (Fig. 8,
inner cycle).
Mechanism 2 does not imply the formation of a
The PA and MEPA concentrations in catalytic solu-
tions were determined using an LKhM-80 gas chro-
matograph (thermal-conductivity detector; column
packed with Porapak P, l = 1 m, d = 3 mm; carrier gas,
helium; column, injector, and detector temperatures,
230, 250, and 240°ë, respectively). Isobutyl benzoate
or ethyl benzoate was used as the standard. The gas in
contact with the solution was analyzed on an LKhM-
8MD gas chromatograph (thermal-conductivity detec-
tor; column with activated carbon AR-3, l = 3 m, d =
3 mm; carrier gas, argon; column and detector temper-
ature, 160°ë; injector temperature, 200°ë).
hydrido complex from PA. In this mechanism, the pro-
ton of the coordinated alkyne is intramolecularly trans-
ferred to the coordinated p-quinone molecule, resulting
in Pd(0) oxidation and the formation of a hydroquino-
late derivative. This is followed by CO insertion into
the Pd–alkynyl σ-bond and, next, the alcoholysis of the
Pd–C(O) bond yielding a molecule of the product
(MEPA) and the hydroquinolate complex L2(HQ)PdH.
The complex L2(HQ)PdH is reduced to L2Pd0 via the
reductive elimination of p-hydroquinone, and the latter
complex reacts with p-quinone to turn into L2PdQ.
Infrared spectra were recorded on a Specord M82
spectrophotometer.
The 1H, 1H{31P}, 13C{1H}, and 31P{1H} NMR spec-
tra of compounds I–XIX in CDCl3 (–25°C) were
Both mechanisms are in agreement with our data,
and discrimination between them will be the subject of
forthcoming studies.
Our data concerning the oxidative carbonylation of obtained on a Bruker DPX 300 spectrometer operating
PA suggest that the reaction system possesses the prop- at 300.13, 75.468, and 121.495 MHz, respectively.
erties of a buffer at pH 6–7.7. The findings that Chemical shifts for phosphorus nuclei were measured
KIEH/D ≈ 1 is close to unity and that the initial system relative to the signal from external 85% ç3êé4 in ç2é.
based on Pd(OAc)2 and the complexes PdLn, X, and XI Reactions were carried out in an Ar or N2 atmosphere.
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

242
KHABIBULIN et al.
Table 9. Crystal data and X-ray structure determination pa-
rameters for complex XVI
X-ray structure determination of complexes XVI
and XVIII. The X-ray diffraction experiment for
[Pd(COOCH3)(OAc)(PPh3)2]2H2Q · ëH3OH (XVI) was
carried out on a CAD-4 automated diffractometer
(β-filter, λMo – Kα, λ = 0.71073 Å, T = 293(2) K,
θ/2θ scan mode). The crystal disintegrated slowly dur-
ing the experiment, and this caused a 36% change in the
reference reflection intensities.
Formula
Molecular weight
System
Space group
a, Å
C87H80O11P4Pd2
1638.19
Triclinic
P-1
10.594(2)
14.119(3)
26.705(5)
85.57(3)
89.40(3)
76.82(3)
3877.5(13)
2
The intensity losses caused by crystal breakdown
were compensated for by recovering integrated intensi-
ties from reflection profiles [29]. Crystal data and the
structure determination parameters for complex XVI
are presented in Table 9.
b, Å
c, Å
α, deg
β, deg
γ, deg
V, Å3
The structure was solved by direct methods using
the SHELXS97 program [30] and was refined using the
SHELXL program [31]. The positions and thermal
parameters of the non-hydrogen atoms were refined
isotropically and then anisotropically using full-matrix
least squares. The hydrogen atoms were placed in their
geometrically calculated positions and were included in
the refinement in the riding model.
Z
dcalc, g/cm3
1.403
µ, cm–1
0.607
Total number of reflections
measured (Rint)
14428 (Rint = 0.0208)
In the dimeric noncentrosymmetric complex, two
Number of reflections with I > 2σ(I)
Number of refined parameters
R1 (I > 2σ(I))
5922
metal
complex
fragments
of
composition
Pd(C(O)OCH3)(OAc)(PPh3)2 are linked by a hydro-
quinone bridge (Fig. 6). p-Hydroquinone is hydrogen-
bonded with the carbonyl oxygen atoms of the acetate
groups coordinated to the metal atom. The distances
between oxygen atoms involved in hydrogen bonding
are similar (O5–O2 2.679(4) Å, O4–O6 2.632(4) Å).
At the same time, H–O–H angles differ significantly
(O4–H40–O6 157.15(1)°, O2–H50–O5 129.95(2)°).
Note that one proton of p-hydroquinone (H40) is nearly
completely shared between an oxygen atom of p-hyd-
roquinone (H40–O4 1.260(1) Å) and an oxygen atom of
the acetate group (H40–O6, 1.425(2) Å). This is not the
case for the other hydrogen bond (O5–H50 0.674(1) Å,
O2–H50 2.195(2) Å). However, the sharing of the pro-
ton exerts only a slight effect on the C–O bond length
in p-hydroquinone (in the case of proton sharing, C76–
O6 1.375(1) Å; in the case of the normal hydrogen
bond, C73–O5 1.366(1) Å) and a stronger effect on the
C=O bond of the acetate ligand. The length of this bond
is 1.275(2) Å in the ligand that shares the proton and
1.245(2) Å in the ligand that does not share the proton.
775
0.0677
0.1831
wR2 (for all reflections)
Kinetic experiments were carried out in a closed,
thermostated glass reactor fitted with a precessing
mechanical stirrer. The weighed components of the cat-
alytic system were charged into the reactor, and metha-
nol and, if necessary, another solvent were added. The
reactor was thermostated, carbon monoxide was passed
through the reaction system, and phenylacetylene was
injected using a microsyringe. During the reaction, we
measured the volume of the gas consumed and sampled
the gas and the solution for GC analysis.
Synthesis of complex XVI. A mixture of Pd(OAc)2
(0.15 mmol) and PPh3 (0.31 mmol) in methanol (4 ml)
was stirred for 5 min. This resulted in a light green pre-
cipitate. p-Benzoquinone (0.573 mmol) was added to
the precipitate, and CO was passed through the mixture
for 10–15 min while stirring. This yielded a light yel-
low solution. The solution was filtered, and hexane
(0.5 ml) was added. The resulting solution was cooled
to +5°ë. Yellow crystals formed in 48 h.
In the aromatic ring of the bridging p-hydroquinone
molecule, all C–C bond lengths are between 1.389 and
1.391 Å, within the range typical of aromatic rings. The
metal atom is in a square-planar coordination environ-
ment, which is typical of Pd(II) complexes. The phos-
phine ligands are trans, and the Pd–P bond lengths lie
in the range 2.327(4)–2.236(4) Å. The aceto group is
coordinated to Pd through one oxygen atom, and the
O3–Pd2 bond length is 2.085(8) Å. The methoxycarbo-
nyl group is trans to the aceto group, and the length of
the Pd2–C85 σ-bond is 1.966(2) Å. The angles between
the Pd–C bond and the carbonyl and ester oxygen
Synthesis of complex XVIII. A mixture of
Pd(OAc)2 (0.114 mmol) and PPh3 (0.23 mmol) in
methanol (4 ml) was stirred for 5 min. This resulted in
a light green precipitate. CO was passed through the
system for ~10–15 min while stirring. This yielded a
homogeneous light yellow solution. Chloranil
(0.55 mmol) was added to the solution. The resulting
solution was stirred for 15–20 min and was then cooled atoms are O9–C85–Pd2 125.2(1)° and O10–C85–Pd2
to +5°ë. Yellow crystals resulted in 2 months.
113.4(1)°, respectively. In one of the two metal com-
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

MECHANISM OF THE OXIDATIVE CARBONYLATION OF TERMINAL ALKYNES
243
Table 10. Crystal data and X-ray structure determination
parameters for complex XVIII
The most significant structural distinction between
the metal complex moieties is that the C=O bond
lengths in their acetate ligands differ by 0.044 Å. The
greatest bond angle difference is 2.7°, which is
observed for the angle between the Pd–C bond and the
C–O–C bond in the methoxycarbonyl ligand.
X-ray structure determination for the complex
(PPh3)2ClPdC6Cl2O2(OMe) (XVIII) was carried out
according to a standard procedure [33] on a Bruker
AXS SMART 100 diffractometer fitted with a CCD
detector (λMo; graphite monochromator; 110 K; ω-
scan mode; 2θmax = 60°; 0.3° increments; frame mea-
surement time, 15 s). The structure was solved by direct
methods using the SHELXS97 program [30] and was
refined by anisotropic full-matrix least squares using
the SHELXL97 program [31] (the H atoms were
located geometrically and were fixed in their positions
with UH = 0.08 Å2). Crystal data and structure determi-
nation parameters for complex XVIII are presented in
Table 10.
Formula
Molecular weight
System
Space group
a, Å
C47H33Cl4O5P2Pd
987.87
Monoclinic
P2(1)/c
12.382(2)
10.745(19)
32.206(6)
90.00
b, Å
c, Å
α, deg
β, deg
γ, deg
V, Å3
92.770(5)
90.00 (3)
4279.6(13)
4
Z
dcalc, g/cm3
1.533
µ, cm–1
0.805
According to X-ray diffraction data, XVIII (Tables 10,
11) is a tetracoordinated palladium(II) complex. The
metal atom is in a square-planar coordination environ-
ment, which is typical of Pd(II) complexes. The coordi-
nation sphere of the metal atom consists of two trans
triphenylphosphine ligands, a chlorine atom, and
2,5-dichloro-3-methoxy-1,4-benzoquinonyl in the
trans position relative to the chlorine atom. The average
plane of the 2,5-dichloro-3-methoxy-1,4-benzo-
quinonyl ring and the plane passing through the phos-
phorus and chlorine atoms bonded to palladium make
an angle close to 90°. The methyl fragment of the meth-
oxy group of 2,5-dichloro-3-methoxy-1,4-benzo-
quinonyl is partially disordered. The C–C bonds in
Total number of reflections
34176 (Rint = 0.0948)
measured (Rint)
Number of reflections with I > 2σ(I)
Number of refined parameters
R1 (I > 2σ(I))
3905
531
0.0645
0.1484
wR2 (for all reflections)
Table 11. Selected bond lengths (d) and bond angles (ω) for
XVIII
Bond
d, Å
Angle
ω, deg
Pd1–P2
Pd1–P1
Pd1–Cl2
P2–C12
P2–C5
2.307(2)
2.314(2)
2.361(2)
1.809(7)
1.814(7)
1.822(7)
C7B–O11–C200
P2–Pd1–P1
120.4(110)
175.55(7)
88.54(7)
177.8(2)
88.54(7)
coordinated
2,5-dichloro-3-methoxy-1,4-benzo-
quinonyl have lengths usual for the p-hydroquinone
series. The double bonds are C20A–C7B 1.317(1) Å
and C14A–C13 1.323(1) Å. The lengths of the ordinary
bonds are between C13–C10A 1.521(1) Å and C20–
C10 1.490(1) Å. The length of the Pd1–C13 σ-bond is
1.996(8) Å.
P2–Pd1–Cl2
C13–Pd1–Cl2
P2–Pd1–Cl2
P2–C4B
ACKNOWLEDGMENTS
plex fragments, the coordinated methoxycarbonyl
group is partially disordered. A heavily disordered
methanol molecule is also present in the crystal struc-
ture.
This work was supported by the Russian Foundation
for Basic Research, grant nos. 01-03-32883, 04-03-
33014, 05-03-08134, and 05-03-33151.
Comparing this structure with the structure of trans-
acetatomethoxycarbonyl(bistriphenylphosphine)palla-
dium [32], which has no p-hydroquinone bridge, sug-
gests that the most important structural features of the
metal complex moiety, the mutual arrangement and the
coordination modes of the ligands, the geometry of the
environment of the metal atom, and the bond lengths
and bond angles are invariable and are independent of
the synthetic route, the crystallization procedure, and
the presence of a supramolecular environment.
REFERENCES
1. Yashina, O.G. and Vereshchagin, L.I., Usp. Khim., 1978,
vol. 47, p. 557.
2. Tsuji, J., Takahashi, M., and Takahashi, T., Tetrahedron
Lett., 1980, vol. 21, no. 9, p. 849.
3. Zung, T.T., Bruk, L.G., and Temkin, O.N., Mendeleev
Commun., 1994, no. 2, p. 2.
4. Zung, T.T., Bruk, L.G., and Temkin, O.N., Izv. Akad.
Nauk, Ser. Khim., 1993, no. 42, p. 1730.
KINETICS AND CATALYSIS Vol. 48 No. 2 2007

244
KHABIBULIN et al.
5. Bruk, L.G., Gorodsky, S.N., Valdes-Perez, R.E., and 21. Dzhemilev, U.M., Popod’ko, N.R., and Kozlova, E.V.,
Temkin, O.N., J. Mol. Catal. A: Chem., 1998, no. 130,
Metallokompleksnyi kataliz v organicheskom sinteze:
alitsiklicheskie soedineniya (Metal Complex Catalysis
in Organic Synthesis: Alicyclic Compounds), Moscow:
Khimiya, 1999.
p. 29.
6. Bruk, L.G. and Temkin, O.N., Inorg. Chim. Acta, 1998,
vol. 280, p. 202.
22. Zudin, V.N., Chinakov, V.D., Nekipelov, V.M., Rogov, V.A.,
Likholobov, V.A., andYermakov, Yu.I., J. Mol. Catal. A:
Chem., 1989, no. 52, p. 27.
23. Shitova, N.B., Kuznetsova, L.N., Yurchenko, E.N.,
Ovsyannikova, I.A., and Matveev, K.I., Izv. Akad. Nauk
SSSR, Ser. Khim., 1973.
24. Takahashi, S. and Hagihara, N., J. Chem. Soc. Jpn., Pure
Chem. Sect., 1967, no. 88, p. 1306.
25. Milani, B., Anzilutti, A., Vicentini, L., Sessanta o Santi, A.,
Zangrando, E., Geremia, S., and Mestroni, G., Organo-
metallics, 1997, no. 16, p. 5064.
26. Kulik, A.V., Bruk, L.G., Temkin, O.N., Khabibulin, V.R.,
Belsky, V.K., and Zavodnik, V.E., Mendeleev Commun.,
2002, p. 47.
27. Moiseev, I.I., π-Kompleksy v zhidkofaznom okislenii ole-
finov (π-Complexes in Liquid-Phase Alkene Oxidation),
Moscow: Nauka, 1970.
28. Zhir-Lebed’, L.N., Kuz’mina, L.G., Struchkov, Yu.T.,
Temkin, O.N., and Golodov, V.A., Koord. Khim., 1978,
vol. 4, p. 1046.
7. RF Patent 2051896.
8. Shalgunov, S.I., Zeigarnik, A.V., Bruk, L.G., and
Temkin, O.N., Izv. Akad. Nauk, Ser. Khim., 1999, no. 10,
p. 1891.
9. Heck, R.F., J. Am. Chem. Soc., 1972, vol. 94, no. 8,
p. 2712.
10. Chukhadzhyan, G.A., Evoyan, Z.K., and Melkonyan, L.N.,
Zh. Obshch. Khim., 1975, vol. 45, p. 1114.
11. Nesmeyanov, A.N. and Kocheshkov, K.A., Metody ele-
mentoorganicheskoi khimii (Methods of Organoelement
Chemistry), Moscow: Nauka, 1978.
12. Amatore, C., Carre, E., Jutand, A., and M’Barki, M.A.,
Organometallics, 1995, vol. 14, no. 4, p. 1818.
13. Moiseev, I.I., Stromnova, T.A., Vargaftik, M.N.,
Mazo, G.Ja., Kuz’mina, L.G., and Struchkov, Y.T.,
J. Chem. Soc., Chem. Commun., 1978, p. 27.
14. Chernysheva, T.V., Stromnova, T.V., Vargaftik, M.N, and
Moiseev, I.I., Izv. Akad. Nauk, Ser. Khim., 1996, no. 10,
p. 2456.
15. Rivetti, F. and Romano, U., J. Organomet. Chem., 1978,
29. Strel’tsov, V.A. and Zavodnik, V.E., Kristallografiya,
vol. 154, p. 323.
1989, vol. 34, no. 6, p. 1369.
16. Rivetti, F. and Romano, U., J. Organomet. Chem., 1979,
30. Sheldrick, G.M., SHELXS97: Program for the Solution
of Crystal Structures, Göttingen: Univ. of Göttingen,
1997.
31. Sheldrick, G.M., SHELXL97: Program for Refinement
of Crystal Structures, Göttingen: Univ. of Göttingen,
1997.
vol. 174, p. 221.
17. Arshad, M., Beg, A., and Siddiqui, M.S., Tetrahedron,
1966, vol. 22, no. 7, p. 2203.
18. Ramirez, F. and Dershowitz, S., J. Am. Chem. Soc.,
1956, vol. 78, p. 5614.
19. Kutyrev, A.A. and Moskva, V.V., Usp. Khim., 1987,
32. Del Piero, G. and Cesari, M., Acta Crystallogr., Sect. B,
vol. 56, no. 11.
1979, no. 35, p. 2411.
20. Amatore, C. and Jutand, A., Acc. Chem. Res., 2000, 33. SMART (Control) and SAINT (Integration) Software,
vol. 33, p. 314.
Version 5.0, Madison, WI: Bruker AXS, 1997.
KINETICS AND CATALYSIS Vol. 48 No. 2 2007