Stereoselective homogeneous catalytic arylation of methyl methacrylate: Experimental and computational study

Article abstract of DOI:10.1016/j.molcata.2012.01.007

Catalytic systems trans-[PdCl₂(DEA)₂]/DEA and trans-[PdCl₂(DEA)₂]/[DEA][HAc], used in the model reaction of methyl methacrylate with iodobenzene, 4-iodoanisole, and bromobenzene, provide homogeneous catalysis, good regioselectivity and excellent stereoselectivity. The major product of the regioselective reaction is internal olefin. In all examined cases the only stereoisomer of the internal olefin methyl 3-phenyl-2-methylpropenoate is the E-isomer, whereas the only stereoisomer of the double arylated reaction product methyl 2-benzyl-3- phenylpropenoate is the Z-isomer. A DFT study, which investigates mechanistic aspects of migratory insertion, β-hydride elimination and reductive elimination of this phosphine-free Heck reaction, is in agreement with our experimental findings.

Full text of DOI:10.1016/j.molcata.2012.01.007

Journal of Molecular Catalysis A: Chemical 356 (2012) 144–151
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Stereoselective homogeneous catalytic arylation of methyl methacrylate:
Experimental and computational study
∗
Zorica D. Petrovi c´ , Vladimir P. Petrovi c´ , Du sˇ ica Simijonovi c´ , Svetlana Markovi c´
Faculty of Science, University of Kragujevac, P.O. Box 60, 34000 Kragujevac, Serbia
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 26 October 2011
Received in revised form
Catalytic systems trans-[PdCl (DEA) ]/DEA and trans-[PdCl (DEA) ]/[DEA][HAc], used in the model reac-
2
2
2
2
tion of methyl methacrylate with iodobenzene, 4-iodoanisole, and bromobenzene, provide homogeneous
catalysis, good regioselectivity and excellent stereoselectivity. The major product of the regioselective
reaction is internal olefin. In all examined cases the only stereoisomer of the internal olefin methyl
2
0 December 2011
Accepted 9 January 2012
Available online 16 January 2012
3
-phenyl-2-methylpropenoate is the E-isomer, whereas the only stereoisomer of the double arylated
reaction product methyl 2-benzyl-3-phenylpropenoate is the Z-isomer. A DFT study, which investi-
gates mechanistic aspects of migratory insertion, ␤-hydride elimination and reductive elimination of
this phosphine-free Heck reaction, is in agreement with our experimental findings.
Keywords:
Homogeneous catalysis
Migratory insertion
©
2012 Elsevier B.V. All rights reserved.
␤
-Hydride elimination
Diethanolamine palladium complex
Density functional theory
1
. Introduction
anionic, cationic, or neutral, attracts attention of both experimental
and theoretical chemists [23–37].
Among palladium-catalyzed reactions, the arylation and viny-
The preactivation reaction of a phosphine-free Heck reac-
tion, where trans-dichlorobis(diethanolamine-N)palladium(II)
lation of alkenes, known as the Heck reaction, is one of the most
important carbon–carbon bond forming processes [1–7]. This reac-
tion has major impact on organic chemistry due to its significant
synthetic versatility and possibility for countless complex organic
molecules to be prepared [8–10]. Over the last two decades an
impressive number of applications have been developed at the
laboratory and industry. The Heck reaction becomes more inter-
esting when the used alkene is disubstituted, what allows different
regio- and stereoselectivity. To achieve synthesis of one regio- or
stereoisomer of a trisubstituted alkene with high yield is still a
challenge in modern synthetic organic chemistry. One of the most
important and difficult targets is the control of the configuration of
the double bond, the E- and Z-selectivity.
A lot of work has been devoted to the elucidation of the
Heck reaction pathways where phosphines are used as ligands
11–18], whereas the results regarding non-phosphine ligands are
scarce [19–22]. Mechanistically, the Heck reaction is a multistep
organic reaction, whereby at least one intermediate contains ␴
carbon–metal bond. The Heck reaction mechanism, which can be
complex (trans-[PdCl (DEA) ]) was used as
a
precatalyst
2
2
[30–32], has recently been elaborated. It was established that
the catalytically active DEA–Pd(0)–Cl complex was obtained
in the preactivation process [30–32]. The next step in the
Heck catalytic cycle, oxidative addition of aryl halide to
Pd(0) species, is considered a key step in the mechanism of
Pd-mediated cross-coupling reactions, and usually involves
coordinatively unsaturated Pd(0) complexes [11–16,34].
The reaction further proceeds through the migratory inser-
tion and ␤-hydride transfer/reductive elimination steps
[38–45].
The aim of this experimental and DFT study is to investigate
possible ways of migratory insertion, ␤-hydride and reductive
elimination of phosphine-free Heck reaction, where a disubstituted
alkene methyl methacrylate is used as a substrate. In the Heck
reaction of aryl halides with methyl methacrylate, biologically
active and useful trisubstituted olefins with broad applications in
pharmaceutical industry (for example ␣-methylcinnamates) can
be obtained. These compounds show interesting physico-chemical
properties, too. In order to provide sustainable green homogeneous
catalysis, beside DEA as solvent, the ionic liquid (IL) diethanolam-
monium acetate ([DEA][HAc]) was also used as reaction
medium.
[
∗ Corresponding author. Tel.: +381 34 336223; fax: +381 34 335040.
E-mail address: zorica@kg.ac.rs (Z.D. Petrovi c´ ).
1
381-1169/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2012.01.007

Z.D. Petrovi ´c et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 144–151
145
Scheme 1. ␤-hydride elimination in 1,1-disubstituted olefins.
2
. Experimental
vibrational frequencies) for equilibrium structures, or first-order
saddle points (one imaginary vibrational frequency) for transition
state structures, by frequency calculations. The intrinsic reaction
coordinates (IRCs), from the transition states down to the two lower
energy structures, were traced using the IRC routine in Gaussian,
in order to verify that each saddle point is linked with two putative
minima. The natural bond orbital analysis (Gaussian NBO version)
was performed for all structures.
2.1. Materials
The reactions were monitored and analyzed with GC chro-
1
matography and
H NMR spectroscopy. The GC analyses
were performed on an Agilent 6890N (G 1530N) instrument
Serial CN10702033), with capillary apolar column. The
NMR spectra were run in CDCl3 on Varian Gemini
00 MHz spectrometer. For column chromatography silica gel 60
Merck, particle size 0.063–0.200 mm) was used. The compounds
(
#
1
H
a
2
(
3. Results and discussion
PdCl , diethanolamine, aryl iodides, bromobenzene and methyl
methacrylate were obtained from Aldrich Chemical Co.
2
In our previous studies we established that PdCl2 upon heat-
ing in DEA or [DEA][HAc] produced an orange solution, from
which a yellow-orange substance crystallized [30,46,52]. The
NMR and IR spectra, as well as elemental analysis, showed that
the structure of the obtained crystal substance corresponds to
that of trans-[PdCl (DEA) ] complex. The crystal structure of
2.2. General procedure for the Heck reaction
Corresponding aryl halide (1.05 mmol), methyl methacry-
2
2
late (1 mmol), DEA (1.0 ml), or ([DEA][HAc]) (1.0 ml), and PdCl
2
trans-[PdCl (DEA) ] was provided in [52], whereas its electronic
◦
2
2
(
1.5 mol%) were placed in 25-ml flask, and stirred at 110 C
structure was presented in [31]. The trans-[PdCl (DEA) ] complex
2
2
for 12 and 24 h. The mixture was cooled to room temperature.
The reaction products (E-methyl 3-phenyl-2-methylpropenoate
appeared to be very efficient in the Heck coupling reaction of dif-
ferent aryl halides with nonsubstituted acrylates, in the presence
of strong and weak base [30,34], and ILs [46].
In order to test catalytic efficiency of this complex in the
Heck reaction with a 1,1-disubstituted olefin, a model reaction
with methyl methacrylate was performed. The reaction conditions
(
a), methyl 2-benzylpropenoate (b) and Z-methyl 2-benzyl-3-
phenylpropenoate (c)) were separated from the reaction mixture
by extraction and decantation with diethyl ether/hexane (5:1).
The combined organic layer was washed with water and brine,
dried with Na SO , and evaporated under the reduced pressure.
2
4
provided in situ formation of trans-[PdCl (DEA) ]. We were partic-
1
2
2
The reactions were monitored and analyzed with H NMR and
GC. It appeared that the orange trans-[PdCl (DEA) ] complex was
ularly focused on the impact of the complex, base (which also acts
as reaction medium), and reaction time on the regio- and stere-
oselectivity of such reaction. It is known that the choice of base
can affect the products distribution and rate of the Heck reaction.
The combinations of K CO3 and DMF as solvent, NaOAc and Bu N
2
2
formed after ten minutes of heating the reaction mixture, in
both reaction media – DEA or [DEA][HAc] [30,46]. The reaction
products were purified by column chromatography (silica gel;
pethrolether/ethylacetate (8:1) for the mixture of a and b, and with
pethrolether/ethylacetate (2:1) for c). After the extraction of the
product of the first-time reaction, the solvent residue of the Heck
reaction was washed with C H5OH (3 ml) and Et O (3 ml × 3 ml),
2
3
in DMAc, and bulky tertiary amine Cy NMe in dioxane have been
2
used in various earlier attempts to carry out efficient synthesis of
stereo- and regiodefined trisubstituted olefins [47,53,54]. In our
homogeneous catalytic protocol DEA and [DEA][HAc] played role of
reaction media, bases, and precatalyst-precursors. Iodobenzene, 4-
iodoanisole, and bromobenzene were used for arylation of methyl
methacrylate.
In the Heck reaction with methacrylates two possible pathways
for double-bond formation (␤-hydride elimination) are possible:
the first pathway yields the expected trisubstituted (internal)
E/Z olefin A, and the other one leads to the formation of 1,1-
disubstituted (terminal) olefin B (Scheme 1) [55].
2
2
and dried under the reduced pressure. After this treatment, the
catalyst/DEA or catalyst/[DEA][HAc] system can be reused directly
without further purification [46].
The reaction products were analyzed and characterized on the
1
basis of their H NMR spectroscopic data, and by comparing these
data to the literature data [47,48] and spectra of the commercially
available compounds. The 1H NMR spectrum of the mixture of a
and b, as well as that of c, are presented in Supplementary Data.
In our model reaction the mixtures of three olefins (a, b, and c,
Fig. 1) were observed, whereby the internal olefin a was the major
product of the regioselective reaction (Table 1). Taking into account
that the olefin b suffers further Heck reaction yielding the diary-
lated olefin c, it can be concluded that the best regioselectivity was
achieved in the reaction with 4-iodoanisole in DEA or [DEA][HAc],
after 24 h (a:b + c = 5:2.5). The regioselectivity of the reactions with
iodobenzene and bromobenzene was somewhat lower. The yield
of the diarylated product c was the highest in the reaction with
iodobenzene, after prolongation of reaction time from 12 h to 24 h.
It should be pointed out that in all cases the only stereoisomer of
the internal olefin a is the E-isomer, whereas the only stereoisomer
of the double arylated product c is the Z-isomer. The stereoselec-
2
.3. Computational methods
The geometrical parameters of all stationary points and transi-
tion states were optimized with Gaussian09 [49], using the CPCM
model, and diethanolamine as solvent (ε = 25.75). All calculations
were performed using the M06 functional [50]. This hybrid meta
functional is a method with good accuracy “across-the-board” for
transition metals, main group thermochemistry, medium-range
correlation energy, and barrier heights [50]. The triple split valence
basis set 6-311 + G(d,p) was used for C, H, O, N, and Cl, whereas
LANL2DZ + ECP [51] was employed for the Pd and I centers. All
calculated structures were confirmed to be local minima (all real

1
46
Z.D. Petrovi ´c et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 144–151
Fig. 1. Optimized geometries of the Heck reaction products a, b, and c.
tivity was determined on the basis of the 1H NMR spectra of the
isolated products.
area in methyl methacrylate (double bond), while the LUMO map
depicts the most electron deficient area in 1 (Pd atom). The NBO
charges on Pd, and C2 and C3 of methyl methacrylate amount 0.17,
−0.30, and −0.12, respectively. These facts support possible coordi-
nation of methyl methacrylate to 1. The obtained complex 3 is trans
coordinated (Scheme 2). As known in literature [18] trans complex 3
isomerizes into the less stable (by 3 kJ/mol), but catalytically active,
cis complex 3.
One can suppose that the Pd N coordinative bond in 2 is par-
ticularly weak, due to the trans effect [56–58]. This assumption is
supported by the NBO analysis of 2, which shows that there is a
very weak donation of density from the lone pair on N (sp3 orbital)
to two formally empty p orbitals of Pd. As a consequence, the neu-
tral DEA ligand can be easily substituted with methyl methacrylate
[38,39], thus forming cis-3.
The catalytic system IL Pd remains unchanged during the
homogeneous catalytic reaction (orange color of the reaction mix-
ture, no palladium black) [46]. It seems that the nature of the
used IL increases the stability of the palladium catalyst, extend-
ing its lifetime. Catalyst and products of this reaction can be easily
separated via simple extraction and decantation from the reac-
tion mixture. Although it is clear that the used catalytic systems
trans-[PdCl (DEA) ]/DEA and trans-[PdCl (DEA) ]/[DEA][HAc] are
2
2
2
2
effective in the Heck homogeneous catalysis, little is known about
mechanistic details of such reactions.
This fact provoked us to investigate possible mechanistic path-
ways for migratory insertion, ␤-hydride and reductive elimination,
using DFT method. The mechanisms of these relevant steps in
the Heck reaction of methyl methacrylate with all aryl sub-
strates used in our experiments (iodobenzene, 4-iodoanisole, and
bromobenzene) are mutually very similar. For this reason, we
here present the mechanism of the reaction with iodobenzene,
whereas the other two reactions are presented in Supplementary
Data.
cis-3 exhibits square planar coordination. The NBO analysis of
this complex reveals that Pd is spd2 hybridized. It forms two cova-
lent (with C1 and iodine) and two coordinative (C2 C3 ␲ orbital
and chlorine) bonds. Low occupancy in ␴ Pd C1 (1.81) and ␴ Pd
I
(1.88) orbitals is a consequence of mutual delocalization of each
*
bonding orbital into the adjacent ␴ antibonding orbital. Pd pos-
sesses four lone electron pairs in the d orbitals, where one of them
shows a low occupancy of 1.86. This electron pair delocalizes into
the ␲* antibonding C2 C3 orbital, whereas the C2 C3 ␲ orbital
delocalizes into the formally empty, almost pure p orbital of Pd.
There is a strong donation of density from the lone pair of Cl (sp3
orbital) to the formally empty p orbital of Pd.
3
.1. DFT study of the model reaction
As previously postulated [34] the oxidative addition of iodoben-
zene to the catalytically active DEA–Pd(0)–Cl complex yields two
possible intermediates (1 and 2 in Scheme 2).
The next reaction step is the formation of the new C1 C2 bond,
with simultaneous cleavage of the ␴ Pd C1 bond. This step occurs
via transition state 4-TS (Scheme 2), which requires activation
energy of 45.4 kJ/mol, and yields the complex 5, the final interme-
diate in the migratory insertion step. The NBO analysis of 5 reveals
that Pd forms only one covalent ␴ bond (Pd C3), and three coordi-
native bonds (with Cl, I, and ␲ C1 C4 orbital). It is worth pointing
out that coordinative interaction of Pd with ␲ C1 C4 bond is par-
ticularly weak.
One can assume that migratory insertion begins with olefin
coordination to 1 or 2. Let us consider the HOMO map of methyl
methacrylate and LUMO map of 1 (Fig. 2), as well as their NBO
charges. The HOMO map delineates the most electron sufficient
Table 1
Palladium-catalyzed arylation of methyl methacrylate.
Entry
Aryl halidea
Solvent
Time (h)
Products ratiob a:b:c
1
2
3
4
5
6
7
8
9
C6H5I
C6H5I
C6H5I
C6H5I
CH3OC6H4I
CH3OC6H4I
CH3OC6H4I
CH3OC6H4I
C6H5Br
DEA
DEA
[DEA][HAc]
[DEA][HAc]
DEA
12
24
12
24
12
24
12
24
12
24
12
24
2.5:1:1
4:1:2
5 can undergo ␤-hydrogen elimination in two possible ways:
one is ␤-elimination of H1, and the other is ␤-elimination of H2
3.5:1:1
5.5:1:2
2.5:1:0.8
5:1:1.5
3:1:1
5:1:1.5
2.5:1:1
3:1:1
(Scheme 2). Bearing in mind the reaction conditions (increased
temperature), and the fact that the coordinative interaction
between Pd and ␲ C1 C4 orbital is very weak, it is reasonable to
expect that 5 can isomerize into agostic complexes, which are char-
acterized by the presence of hypovalent three-center two-electron
bonds [59,60]. Our investigations show that the structures of the
agostic complexes 7 and 11 are appropriate for ␤-hydride elimina-
tion [18,61,62]. These complexes are formed from 5 via transition
states 6-TS and 10-TS (Scheme 2), and require the activation ener-
gies of 29.4 and 26.5 kJ/mol, respectively. In 7 Pd forms three
covalent bonds – with C3, Cl and I, whereas in 11 Pd forms just two
DEA
[DEA][HAc]
[DEA][HAc]
DEA
DEA
[DEA][HAc]
[DEA][HAc]
10
11
12
C6H5Br
C6H5Br
C6H5Br
3:1:1
3.5:1:1
a
About 85% of bromobenzene and 4-iodoanisole, and about 95% of iodobenzene
were converted to the reaction products.
b
The products ratio determined by GC and their isolated amounts.

Z.D. Petrovi ´c et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 144–151
147
Scheme 2. Possible pathways for migratory insertion, ␤-hydride elimination and reductive elimination of the catalytic arylation of methyl methacrylate with iodobenzene.
covalent (with C3 and I), and one coordinative (with Cl) bonds. H1
in 7 and H2 in 11 are directed toward Pd, thus enabling the forma-
Our numerous attempts to reveal a reaction path for possible
isomerization between 7 and 11 were unsuccessful. In addition,
we were not able to locate a transition state for the formation of
the agostic complex whose ␤-hydride elimination would lead to
the Z product. We suppose that the isomerization is hindered by
the presence of the methyl group. On the other hand, the structure
of this complex was optimized (Fig. S1). The complex is apparently
significantly strained, and consequently, its free energy is by 10.5
and 8.8 kJ/mol higher than those of 7 and 11. These results are in
accord with the experimentally determined stereoselectivity of the
reaction.
tion of three-center two-electron C
H Pd bonds. The existence of
the C2 H1 Pd three-center two-electron bond is confirmed by the
NBO analysis [63] of 7. Namely, ␴ C2 H1 bond delocalizes into the
3
formally empty sp orbital of Pd. As a consequence, the occupancy
in the formally empty orbital of Pd equals 0.10, and the C2 H1 bond
in 7 is by 0.08 A˚ longer than the corresponding bond in 5. Similarly,
there is C5 H2 Pd three-center two-electron bond in isomer 11.
Here, ␴ C5 H2 orbital donates density to the formally empty sp3
orbital of Pd, thus increasing its occupancy (0.10), and elongating
the C6 H2 bond by 0.07 A˚ in comparison to the corresponding bond
in 5.
In the further course of the reaction, the ␤ hydrogens (H1 and
H2) undergo transfers from the corresponding carbon atoms (C2
Fig. 2. HOMO map of methyl methacrylate (left) and LUMO map of 1. The regions where the values of the HOMO and LUMO are greatest are indicated in blue. In the grayscale
image, the darker regions (around double bond and palladium) depict the greatest values of the HOMO and LUMO. (For interpretation of the references to color in this figure
legend, the reader is referred to the web version of this article.)

1
48
Z.D. Petrovi ´c et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 144–151
4
6
9
-TS. It can be concluded that relatively low activation energies for
-TS–12-TS enable both reaction paths, leading to the formation of
and 13, and further of a and b.
Due to the trans effect, the coordinative interactions with ␲
orbitals are particularly weak, implying that 9 and 13 are sus-
ceptible to ligand substitution. As DEA is available in the reaction
mixture, it coordinates to Pd, thus forming complex 14 (Fig. 4) and
liberating the products a and b (Fig. 1). Our investigation shows
that a is more stable than b (by 15.7 kJ/mol for the reactions with
iodobenzene and bromobenzene, and by 22.0 for the reaction with
4
-iodoanisole). The relative stabilities of the products a and b are
in agreement with the found product ratios. Namely, the regiose-
lectivity is higher in the case of the reaction with 4-iodoanisole in
comparison to that with iodobenzene and bromobenzene (Table 1).
In addition, Table 1 shows that the ratio a:b + c increases with the
prolonged reaction time. These facts, as well as relatively low and
very similar activation energies needed for the formation of a and b,
indicate that the investigated reaction is thermodynamically con-
trolled. Our finding is different from that of Ambrogio et al., who
have investigated the Heck reaction of aryl iodides and bromides
with allyl ethers, and found that the reaction is kinetically con-
trolled [36]. The difference in reaction mechanisms is probably due
to the reaction conditions, as well as different chemical nature of
the used substrates.
Fig. 3. The free energy profile of the migratory insertion and ␤-hydride elimination
steps in the arylation of methyl methacrylate with iodobenzene.
in 7 and C5 in 11) to the Pd center (Scheme 2). These steps of the
reaction proceed via 8-TS and 12-TS transition states, and require
the activation energies of 28.9 and 26.7 kJ/mol, respectively. In both
transition states, C H bond is completely broken, while new Pd
H
bond is being established. In addition, a new ␲ bond between car-
bon atoms (C2 C3 in 8-TS and C3 C5 in 12-TS) is being formed.
The products of this reaction step are intermediates 9 and 13. In
these intermediates, Pd forms covalent bonds with I and H, and
coordinative bonds with Cl and ␲ orbital (C2 C3 in 9, and C3 C5
in 13).
The free energy profile of the migratory insertion and ␤-hydride
elimination steps is depicted in Fig. 3. The diagram shows that the
rate determining step is the formation of the new C1 C2 bond,
which occurs via 4-TS. Other two pairs of transition states (6-TS
and 10-TS; 8-TS and 12-TS) require mutually very similar activation
energies, which are noticeable lower than that for the formation of
The structure of 14 is very similar to those of the complex
intermediates whose reductive elimination has already been inves-
tigated [30–32]. Thus, one can suppose that 14 will also undergo
reductive elimination. Since the most electron sufficient area in
14 involves iodine (Fig. 4), it is reasonable to expect that HI will
be liberated, and catalytically active Pd(0) complex will be recov-
ered. This assumption was confirmed by revealing the transition
state 15-TS (Fig. 4), which requires activation energy of 75.6 kJ/mol.
Fig. 4. Optimized structures of 14 (left), and 15-TS (right). The HOMO of 14 is depicted.
Fig. 5. The results of the IRC calculations for 4-TS (left), 8-TS (middle), and 12-TS (right).

Z.D. Petrovi ´c et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 144–151
149
Scheme 3. Full catalytic cycle for the Heck reaction performed with PdCl2 and methyl methacrylate in DEA.
Table 2
The energetics of the catalytic cycle for the reaction performed with PdCl2 and methyl methacrylate in DEA.
f1 [30,31]
ꢀG = 584.8 kJ/mol
Ea = 111.8 kJ/mol
[
[
DEA − Pd(0) − CI]− + PhI → 1 + DEA
DEA − Pd(0) − CI]− + PhI → 2
f2 [34]
ꢀG = − 169.0 kJ/mol
Ea = 17.1 kJ/mol
ꢀ
G = − 194.5 kJ/mol
Ea = 50.7 kJ/mol
1
2
+ methyl methacrylate → 5
f3
f4
ꢀG = − 32.9 kJ/mol
+ methyl methacrylate → 5 + DEA
ꢀ
G = − 53.6 kJ/mol
5
5
→ 9
ꢀG = 29.8 kJ/mol
Ea = 29.4 kJ/mol
→ 13
ꢀ
G = 30.6 kJ/mol
Ea = 26.5 kJ/mol
9
+ DEA → 14 + a
f5
f6
ꢀG = 28.0 kJ/mol
1
3 + DEA → 14 + b
14 + DEA → [DEA − Pd(0) − CI]− + [HDEA][I]
ꢀ
G = 43.0 kJ/mol
ꢀG = 40.0 kJ/mol
Ea = 75.6 kJ/mol

1
50
Z.D. Petrovi ´c et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 144–151
In 15-TS the Pd H and Pd I bonds are being cleaved, and the
new H I bond is being formed. It is worth pointing out that the
electronic pair from the Pd H bond remains on Pd, implying that
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.molcata.2012.01.007.
Pd(II) is reduced to Pd(0). In this way, the catalytically active
−
[
DEA–Pd(0)–Cl] complex is recovered. The liberated HI is captured
by DEA, yielding the salt [DEAH][I].
References
The Pd(0) complex undergoes oxidative addition [34], and
further reaction with b (as it has a terminal double bond), con-
forming the same mechanism as other 1,1-disubstituted alkenes,
and yielding c (Fig. 1). Mechanistic pathways for migratory inser-
tion, ␤-hydride elimination, and reductive elimination with b as a
substrate are examined, and presented in Supplementary Data.
The results of the IRC calculations for transition states 4-TS, 8-TS,
and 12-TS are presented in Fig. 5.
[1] R. Heck., Acc. Chem. Res. 12 (1979) 146–151.
[2] D.A. Alonso, C. Nájera, M.C. Pácheco, Org. Lett. 2 (2000) 1823–1826.
[
[
3] I.P. Beletskaya, A.V. Cheprakov, Chem. Rev. 100 (2000) 3009–3066.
4] C.S. Consorti, M.L. Zanini, S. Leal, G. Ebeling, J. Dupont, Org. Lett. 5 (2003)
983–986.
[5] C.M. Jin, B. Twamley, J.M. Shreeve, Organometallics 24 (2005) 3020–3023.
[6] R. Wang, B. Twamley, J.M. Shreeve, J. Org. Chem. 71 (2006) 426–429.
[7] X. Cui, J. Li, Z.P. Zhang, Y. Fu, L. Liu, Q.X. Guo, J. Org. Chem. 72 (2007) 9342–9345.
[8] J.G. De Vries, Can. J. Chem. 79 (2001) 1086–1092.
This paper examines migratory insertion, ␤-hydride and reduc-
tive elimination in the Heck reaction of methyl methacrylate and
some aryl halides. Taking into account our results on the preacti-
vation reaction [30–32], and oxidative addition [34], now we are
able to present the full catalytic cycle for the reaction performed
with PdCl2 and methyl methacrylate in DEA (Scheme 3). Table 2
provides the energetics of the catalytic cycle.
The preactivation step is pronouncedly endothermic, and
requires the highest activation energy in the overall Heck reac-
tion. As for the catalytic cycle, oxidative addition and migratory
insertion are exothermic, whereas ␤-hydride elimination, ligand
substitution, and reductive elimination (catalyst recovery) are
endothermic. The overall catalytic cycle is slightly exothermic, and
the highest activation barrier is needed for the reductive elimina-
tion step.
[9] B.M. Trost, in: Atta-Ur-Rahman (Ed.), Advances in Natural Product Chemistry,
Harwood Academic Publishers, Harwood, 1992, p. 194.
10] B.M. Trost, M.L. Crawley, Chem. Rev. 103 (2003) 2921–2943.
11] M. Ahlquist, P.-O. Norrby, Organometallics 26 (2007) 550–553.
[
[
[12] Z. Li, Y. Fu, Q.X. Guo, L. Liu, Organometallics 27 (2008) 4043–4049.
[
[
13] Y.L. Huang, C.M. Weng, F.E. Hong, Chem. Eng. J. 14 (2008) 4426–4434.
14] M. Ahlquist, P. Fristup, D. Tanner, P.-O. Norrby, Organometallics 25 (2006)
2066–2073.
[15] L.J. Goossen, D. Koley, H. Hermann, W. Thiel, Chem. Commun. (2004)
141–2143.
2
[
[
16] R.J. Deeth, A. Smith, M.J. Brown, J. Am. Chem. Soc. 126 (2004) 7144–7151.
17] H.M. Senn, T. Ziegler, Organometallics 23 (2004) 2980–2988.
[18] P. Surawatanawong, Y. Fan, M.B. Hall, J. Organomet. Chem. 693 (2008)
1552–1563.
[
19] L.J. Goossen, D. Koley, H. Hermann, W. Thiel, Organometallics 24 (2005)
398–2410.
2
[20] O. Esposito, M.P.G. Pedro, K.A. Lewis, F.S. Caddick, G.N. Cloke, P.B. Hitchcock,
Organometallics 27 (2008) 6411–6418.
21] M.T. Lee, M.H. Lee, H.C. Hu., Organometallics 26 (2007) 1317–1324.
[
[
22] J.C. Green, B.J. Herbert, J.R. Lonsdale, J. Organomet. Chem. 690 (2005)
6
054–6067.
23] A.J. Carmichael, M.J. Earle, J.D. Holbrey, P.B. McCormac, K.R. Seddon, Org. Lett.
(1999) 997–1000.
24] C. Ye, J.C. Xiao, B. Twamley, A.D. LaLonde, M.G. Norton, J.M. Shreeve, Eur. J. Org.
Chem. (2007) 5095–5100.
[
[
4
. Conclusion
1
Catalytic systems trans-[PdCl (DEA) ]/DEA and trans-
2
2
[
[
25] L. Wang, H. Li, P. Li, Tetrahedron 65 (2009) 364–368.
26] L. Xu, W. Chen, J. Xiao, Organometallics 19 (2000) 1123–1127.
[
PdCl (DEA) ]/[DEA][HAc], used in the reaction of methyl
2 2
methacrylate with aryl halides, provide sustainable homogeneous
catalysis, good regioselectivity and excellent stereoselectivity.
In all cases the only stereoisomer of the internal olefin methyl
[27] I. Pryjomska-Ray, A.M. Trzeciak, J.J. Ziółkowski, J. Mol. Catal. A: Chem. 257
(2006) 3–8.
[
28] S.T. Henriksen, P.-O. Norrby, P. Kaukoranta, P.G. Andersson, J. Am. Chem. Soc.
30 (2008) 10414–10421.
29] P. Surawatanawong, M.B. Hall, Organometallics 27 (2008) 6222–6232.
1
3
-phenyl-2-methylpropenoate is the E-isomer, whereas the only
stereoisomer of the double arylated reaction product methyl
-benzyl-3-phenylpropenoate is the Z-isomer. Our DFT study is
[
[30] Z.D. Petrovi c´ , S. Markovi c´ , D. Simijonovi c´ , V.P. Petrovi c´ , Monatsh. Chem. 140
(
2009) 371–374.
2
[
31] S. Markovi c´ , Z.D. Petrovi c´ , V.P. Petrovi c´ , Monatsh. Chem. 140 (2009) 171–175.
[
32] V.P. Petrovi c´ , S. Markovi c´ , Z.D. Petrovi c´ , Monatsh. Chem. 142 (2011) 141–144.
in agreement with this experimental finding. Namely, possible
isomerization of 5 to an intermediate whose transformation
would lead to Z-methyl 3-phenyl-2-methylpropenoate would be
hindered by the presence of the methyl group. There is similar
steric hindrance that prevents the formation of E-methyl 2-benzyl-
[33] X. Cui, Z. Li, Z.C. Tao, Y. Xu, J. Li, L. Liu, Q.X. Guo, Org. Lett. 8 (2006) 2467–2470.
[
[
[
[
34] Z.D. Petrovi c´ , V.P. Petrovi c´ , D. Simijonovi c´ , S. Markovi c´ , J. Organomet. Chem.
694 (2009) 3852–3858.
35] R.J. Deeth, A. Smith, K.K.M. Hii, J.M. Brown, Tetrahedron Lett. 39 (1998)
3229–3232.
36] I. Ambrogio, G. Fabrizi, S. Cacchi, S.T. Henriksen, P. Fristrup, D. Tanner, P.-O.
Norrby, Organometallics 27 (2008) 3187–3195.
3
-phenylpropenoate (possible isomerization to an appropriate
intermediate would be hindered by the presence of the benzyl
group). In both cases the geometries of these intermediates,
suitable for ␤-hydride elimination, would be extremely strained.
Our experiments show that, in all examined cases, the internal
olefin a is the major product of the regioselective reaction. This
finding is in agreement with our DFT investigation. Namely, the
activation barriers for the formation of a and b are relatively low
and mutually very similar, and their influence to the products dis-
tribution is poor. On the other hand, a is thermodynamically more
stable than b, and thus, a is the preferred reaction product.
Our anionic Heck protocol allows efficient separation of the cat-
alyst and reaction products from the reaction mixture (via simple
extraction and decantation), in contrast to many other homoge-
neous catalyses, where it is a significant problem.
37] C. Backtorp, P.-O. Norrby, Dalton Trans. 40 (2011) 11308–11314.
[38] W. Cabri, I. Candiani, Acc. Chem. Res. 28 (1995) 2–7.
[
[
[
39] E.G. Samsel, J.R. Norton, J. Am. Chem. Soc. 106 (1984) 5505–5512.
40] K.J. Cavell, Coord. Chem. Rev. 155 (1996) 209–243.
41] D.L. Thorn, R. Hoffmann, J. Am. Chem. Soc. 100 (1978) 2079–2090.
[42] B.L. Lin, L. Liu, Y. Fu, S.W. Luo, Q. Chen, Q.X. Guo, Organometallics 23 (2004)
2114–2123.
[43] W. Cabri, I. Candiani, A. Bedeschi, S. Penco, J. Org. Chem. 57 (1992) 1481–1486.
[44] K. Albert, P. Gisdakis, N. Rꢀsch, Organometallics 17 (1998) 1608–1616.
[45] A. Sundermann, O. Uzan, J.M.L. Martin, Chem. Eng. J. 7 (2001) 1703–1711.
[46] Z.D. Petrovi c´ , D. Simijonovi c´ , V.P. Petrovi c´ , S. Markovi c´ , J. Mol. Catal. A: Chem.
327 (2010) 45–50.
[
[
47] A.F. Littke, G.C. Fu, J. Am. Chem. Soc. 123 (2001) 6989–7000.
48] T.R. Hoye, M.J. Kurth, J. Org. Chem. 45 (1980) 3549–3554.
[49] M.J. Frisch, W.G. Trucks, B.H. Schlegel, E.G. Scuseria, A.M. Robb, R.J. Cheeseman,
G. Scalmani, V. Barone, B. Mennucci, A.G. Petersson, H. Nakatsuji, M. Caricato,
X. Li, P.H. Hratchian, F.A. Izmaylov, J. Bloino, G. Zheng, L.J. Sonnenberg, M. Hada,
M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, A.J. Montgomery Jr., A.J. Montgomery Jr., E.J. Peralta,
F. Ogliaro, M. Bearpark, J.J. Heyd, E. Brothers, N.K. Kudin, N.V. Staroverov, R.
Kobayashi, J. Normand, K. Raghavachari, A. Rendell, C.J. Burant, S.S. Iyengar, J.
Tomasi, M. Cossi, N. Rega, M.J. Millam, M. Klene, E.J. Knox, B.J. Cross, V. Bakken,
C. Adamo, J. Jaramillo, R. Gomperts, E.R. Stratmann, O. Yazyev, J.A. Austin, R.
Cammi, C. Pomelli, W.J. Ochterski, L.R. Martin, K. Morokuma, G.V. Zakrzewski,
A.G. Voth, P. Salvador, J.J. Dannenberg, S. Dapprich, D.A. Daniels, O. Farkas, B.J.
Acknowledgment
This work is supported by theMinistry of Science and Environ-
ment of Serbia, project No 172016.

Z.D. Petrovi ´c et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 144–151
151
Foresman, V.J. Ortiz, J. Cioslowski, J.D. Fox, Gaussian 09, Rev A. 1 Gaussian Inc.,
Wallingford, 2009.
[57] J.V. Quagliano, L. Schubert, Chem. Rev. 50 (1952) 201–260.
[58] F.R. Hartley, Chem. Soc. Rev. 2 (1973) 163–179.
[50] Y. Zhao, E.N. Schultz, G.D. Truhlar, J. Chem. Theory Comput. 2 (2006) 364–382.
[51] J.P. Hay, R.W. Wadt, J. Chem. Phys. 82 (1985) 270–283.
[
52] Z.D. Petrovi c´ , M.I. Djuran, F.W. Heinemann, S. Rajkovi c´ , S. Trifunovi c´ , Bioorg.
Chem. 34 (2006) 225–234.
[59] S. Moncho, G. Ujaque, A. Lled1s, P. Espinet, Chem. Eur. J. 14 (2008) 8986–8994.
[60] J.P. Stambuli, C.D. Incarvito, M. Bühl, J.F. Hartwig, J. Am. Chem. 126 (2004)
1184–1194.
[61] S.T. Henriksen, D. Tanner, S. Cacchi, P.-O. Norrby, Organometallics 28 (2009)
[53] M. Beller, T.H. Riermeier, Eur. J. Inorg. Chem. (1998) 29–35.
[54] I. Kondolff, H. Doucet, M. Santelli, Tetrahedron Lett. 44 (2003) 8487–8491.
[55] D. Morales-Morales, C. Grause, K. Kasaoka, R. Redon, R.E. Cramer, C.M. Jensen,
Inorg. Chim. Acta 300–302 (2000) 958–963.
6201–6205.
[62] K.K. (Mimi) Hii, T.D.W. Claridge, J.M. Brown, Helv. Chim. Acta 84 (2001)
3043–3056.
[63] F. Weinhold, C.R. Landis, Valency and Bonding
A Natural Bond Orbital
[56] C. Bäcktorp, P.-O. Norrby, J. Mol. Catal. A: Chem. 328 (2010) 108–113.
Donor–acceptor Perspective, Cambridge University Press, New York, 2005.