Investigation of sol-gel supported palladium catalysts for Heck coupling reactions in o/w-microemulsions

Article abstract of DOI:10.1016/j.molcata.2014.06.016

Sol-gel supported palladium catalysts are investigated for the Heck coupling reaction between styrene and iodo-/bromobenzene to trans-stilbenes in o/w-microemulsions as alternative reaction medium. High conversions and selectivities are obtained with these catalysts and they show better catalytic performance than their commercial analogs Pd@SiO₂ or Pd/C. The influence of the catalyst structure on the activity is investigated in detail showing mass transport limitations that can be optimized by the palladium loading. The catalyst is recyclable >6 times with negligible palladium leaching into the solution. Because of the good recyclability under retention of activity and selectivity, the influence of transport limitations is suppressed and the total catalyst efficiency is increased to more than 2.

Full text of DOI:10.1016/j.molcata.2014.06.016

Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Investigation of sol–gel supported palladium catalysts for Heck
coupling reactions in o/w-microemulsions
I. Volovycha, Y. Kasakaa, M. Schwarzea,∗, Z. Nairoukhb, J. Blumb, M. Fanunc,
D. Avnirb, R. Schomäckera
a Institute of Chemistry, Berlin Institute of Technology, Straße des 17, Juni 124, D-10623 Berlin, Germany
b Institute of Chemistry, The Hebrew University, Jerusalem 91904, Israel
c Center for Colloid and Surface Research, Al-Quds University, East Jerusalem 51000, Palestinian Authority
a r t i c l e i n f o
a b s t r a c t
Article history:
Sol–gel supported palladium catalysts are investigated for the Heck coupling reaction between styrene
and iodo-/bromobenzene to trans-stilbenes in o/w-microemulsions as alternative reaction medium. High
conversions and selectivities are obtained with these catalysts and they show better catalytic performance
than their commercial analogs Pd@SiO2 or Pd/C. The influence of the catalyst structure on the activity is
investigated in detail showing mass transport limitations that can be optimized by the palladium loading.
The catalyst is recyclable >6 times with negligible palladium leaching into the solution. Because of the
good recyclability under retention of activity and selectivity, the influence of transport limitations is
suppressed and the total catalyst efficiency is increased to more than 2.
Received 29 January 2014
Received in revised form 12 June 2014
Accepted 14 June 2014
Available online 23 June 2014
Keywords:
Heck coupling
Microemulsion
Palladium acetate
Sol–gel method
Trans-stilbene
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
can be used in a variety of catalytic reactions including hydro-
genation [11], hydroformylation [12,13], disproportionation [14],
A general trend in catalysis is the immobilization of homoge-
nous catalyst complexes or metal nanoparticles on solid supports
in order to obtain high turnover numbers by facilitating the catalyst
recycling. The recycling becomes more important with increasing
catalyst costs and decreasing catalytic activity. Especially in impor-
tant organic transformations, e.g. coupling reactions, an active
and stable supported catalyst would allow for continuous reac-
tion processes and definitively lead to higher added values. An
often investigated coupling reaction is the Heck reaction and as
supported catalysts highly active zeolite PdCl2-CsX [1] or Pd-NaY
[2], magnetically recyclable palladium nanoparticles [3–5], palla-
dium immobilized on carbon [6] or on different modified silica’s
[7–9] have been investigated. The main criteria for the selection
of support materials are the activity of the prepared catalysts and
the extent of leaching of the active component into the reaction
solution. For example, palladium catalysts immobilized on zeolite
are more active than on silica, but they show higher leaching [10];
therefore, silica is often used as support material. Rhodium and
palladium catalysts immobilized on silica via the sol–gel method
isomerization [15], aldehyde decarbonylation [16] and coupling
reactions [17], and almost all cases the catalyst retained its activity
in consecutive runs.
Not only the structure and morphology of the heterogeneous
catalyst is responsible for its activity but also the choice of the
solvent. Commonly organic solvents are applied as typical reac-
tion medium, but in the recent years the use of aqueous solutions
as solvents has increased [6,18–20] due to the more environmen-
tally friendly properties of water. Unfortunately, reactions with
hydrophobic reactants, as used in Heck reactions, cannot be per-
formed in water due to the low solubility of the reactants. For
this reason the addition of solubilizers is recommended. Typical
solubilizers are surfactants. At concentrations above the criti-
cal micelle concentration (cmc), the surfactants form micelles,
large supramolecular aggregates which consist of hydrophilic head
groups directed to water and hydrophobic cores able to solubilize
the hydrophobic reactants and the resulting products.
In this contribution we perform Heck reactions in a surfactant
system with sol–gel supported catalysts. The mechanism of a het-
erogeneously catalyzed reaction in micellar systems is complicated
due to the inhomogeneous distribution of the reactants within
the micro-heterogeneous reaction medium. Therefore many fac-
tors influence the course of the reaction. From Scheme 1 it can
∗
Corresponding author. Tel.: +49 3031424097; fax: +49 3031421595.
E-mail address: ms@chem.tu-berlin.de (M. Schwarze).
http://dx.doi.org/10.1016/j.molcata.2014.06.016
1381-1169/© 2014 Elsevier B.V. All rights reserved.

I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
211
Scheme 1. Mechanism of the sol–gel supported catalysis in microemulsion as reaction medium.
be primarily expected that the structure of the sol–gel immobi-
lized catalyst influences the reaction rate. The structure depends
on the nature of entrapped metal nanoparticles, the gel-building
agent and on the modification of the support surface. Additionally,
the choice of the solvent (surfactant, cosurfactant) and substrate
type is crucial for optimizing the reaction conditions.
In this contribution we apply sol–gel supported palladium cat-
alysts and study the influence of the support materials, catalyst
precursors, surfactants, solvents and substrates on the activity
and selectivity of the Heck reaction of halobenzene and styrene
(Scheme 2). The more thermodynamically stable E-(trans)-stilbene
product is favored (>99%) which can be applied further for the syn-
thesis of fine chemicals, e.g. resveratrol [21,22]. Furthermore, for a
detailed comparison, unsupported palladium catalysts are inves-
tigated for the same reaction. Finally, the heterogeneity of the
supported catalysts, the influence of the pore structure and their
recycling behavior are studied.
methylacrylate, ˛-methylcinnamic acid, methylmethacrylate,
methyl-˛-methyl-cinnamate, palladium on alumina powder
(Pd@Al2O3, 1 wt%), palladium(II) acetate (Pd(OAc)2), poly(4-
vinylpyridine) (PvyPy), potassium carbonate (K2CO3), tetramethyl
orthosilicate (TMOS), trans-stilbene, triethoxyphenylsilane
(PhSi(OEt)3), trimethoxy(octyl)-silane (OcSi(OMe)3), dioctyl
sodium sulfosuccinate (DSS), Triton X-100 (TX-100) and 4,5-
bis(diphenylphosphino)-9,9-dimethylxanthene (Xantphos) were
obtained from Sigma–Aldrich Company Munich/Germany. Trans-
4-bromostilbene and trans-4-chlorostilbene were purchased from
TCI Europe Eschborn/Germany.
2.2. Synthesis of sol–gel supported palladium catalysts
The synthesis is comparable to the procedure reported by
Rozin-Ben Baruch et al. [17]. Entrapment of the catalyst within
hydrophobic silica sol–gel matrix was carried out as follows:
trimethoxy(octyl)silane (2.1 mL, 9.877 mmol) or triethoxyphenyl-
silane (1.612 mL, 6.680 mmol) was hydrolized in 0.4 mL distilled
water and 4.25 mL methanol or ethanol for 24 h. Then separately
a solution of tetramethyl orthosilicate (3.6 mL, 22.961 mmol) was
stirred in 2.4 mL methanol and 2 mL water for 20 min. Both solu-
tions were combined and stirred for another 30 minutes to produce
the hydrophobically modified silane. Palladium(II) acetate (30 mg,
0.134 mmol) or palladium(II) bromide (35 mg, 0.134 mmol) was
dissolved in 4 mL dichloromethane and stirred for 20 minutes.
After this step, the catalyst solution was combined with the solu-
tion of the hydrophobically modified silane and the mixture was
shaken until gelation. Thereafter, the gel was first dried for 4–6 h
at 80 ◦C and then it was dried for another 8–10 h at 80 ◦C and
103 Pa to remove the solvents. The catalyst was washed carefully
with boiling dichloromethane and dried again at 80 ◦C and 103 Pa
for 8–10 h. After the immobilization procedure about 2.2–2.4 g of
black-colored sol–gel supported palladium catalyst was obtained.
Palladium catalyst immobilized on hydrophilic support was
prepared without addition of trimethoxy(octyl)silane or tri-
ethoxyphenylsilane to the hydrolyzed tetramethyl orthosilicate
solution.
2. Experimental
2.1. Chemicals
2-Bromobenzonitrile,
trans-˛-methylstilbene,
2-chlorotoluene,
palladium on
methylcinnamate,
silica powder
(Pd@SiO2, 5 wt% Pd) and palladium(II) bromide (PdBr2)
were purchased from abcr GmbH Co. KG Karlsruhe/
&
Germany. 1-(Tetradecyl)trimethylammonium bromide (TTAB) was
obtained from Alfa Aesar Company Karlsruhe/Germany.
Sodium dodecylsulfate (SDS) was obtained from AppliChem
GmbH Darmstadt/Germany. 30 mol-% 3,3ꢀ,3ꢀꢀ-phosphinidynetris
(benzenesulfonic acid) trisodium salt solution (TPPTS) was
obtained from Celanese Corporation Irving/USA. Cinnamic acid
(CA), 2-ethylhexylacrylate and styrene were purchased from
Fluka. Acrylic acid (AA), polyoxyethylen (23) laurylether (Brij®35)
and palladium on activated carbon (Pd@C, 10 wt% Pd) were
ordered from Merck KGaA Darmstadt/Germany. Acetonitril (ACN),
dimethyl formamide (DMF) and water (HPLC grade) were pur-
chased from Carl Roth GmbH & Co. KG Karlsruhe/Germany. Marlipal
24/60 was provided from Sasol Germany GmbH Marl/Germany.
˛-Methylstyrene, 4-bromoacetophenone, bromobenzene (PhBr),
4-bromobenzylamine, 1-bromo-4-iodobenzene (IPhBr), tert-
butylmethacrylate, 4-chloroacetophenone, chlorobenzene (PhCl),
2-chloro-benzonitrile, 2-chloro-1,3-dimethylbenzene, 1-chloro-2-
nitrobenzene, 4-chlorostyrene, trans-4-chlorostilbene, dodecyltri-
For the immobilization of palladium catalysts in the presence of
phosphine ligands (1:1 ratio) the following modification was done:
palladium acetate (11.8 mg, 0.052 mmol) and Xantphos (30.420 mg,
0.052 mmol) or TPPTS (0.508 g, 0.052 mmol) were mixed with 4 mL
dichloromethane and stirred for 6 h.
methylammonium
bromide
(DTAB),
hexadecyltrimethyl-
2.3. Reaction procedure
ammonium bromide (CTAB), 2-hydroxyethylmethacrylate, IGEPAL
CA-520, iodobenzene (PhI), palladium (1000 mg/L) and phos-
phor (1000 mg/L) ICP standard solutions, methacrylic acid,
A typical Heck reaction was performed in a double walled
glass reactor with reflux condenser at ambient pressure. 153 ␮L

212
I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
Scheme 2. Heck reaction.
(1.34 mmol) alkene (e.g. styrene) and 158 ␮L (1.50 mmol) aryl
halide (e.g. iodobenzene) were dissolved in 46 mL organic sol-
vent or aqueous microemulsion in the presence of 0.276 g
(2.00 mmol) K2CO3 as base and stirred at desired reaction
temperature.1.2–1.4 g of the immobilized catalyst containing
15 mg (0.067 mmol) Pd(OAc)2 on silica was added to the reaction
mixture and the reaction was started. For the Heck reaction with
homogeneous catalyst only 15 mg (0.067 mmol) of palladium (II)
acetate was added. The reaction progress was monitored by mea-
suring the reactant concentrations at different reaction times by
high performance liquid chromatography (HPLC).
The used one phase microemulsion consists of CTAB (1.510 g,
3.3 wt%) as surfactant, 1-propanol (3.020 g, 6.6 wt%) as cosurfactant,
H2O (40.855 g, 89.3 wt%) and organic substrates e.g. iodobenzene
and styrene (2.84 mmol, 0.8 wt%).
The composition of the microemulsion is usually described by
the parameters ˛ (oil ratio), ꢀ (surfactant ratio) and ı (cosurfactant
ratio). For the used microemulsions ˛, ꢀ and ı were 0.9 wt%, 9.9 wt%
and 66.7 wt%, respectively.
the instrument was performed with commercial palladium and
phosphorus standards.
The compositions of samples withdrawn from the reaction mix-
ture were determined by HPLC using an Agilent instrument with
250 × 4 mm chromatographic column Multospher 120 RP18-5␮
from Ziemer Chromatographie Langerwehe/Germany. A mixture of
acetonitrile/water (70 vol%/30 vol%) was used as eluent, the flow-
rate was 1 mL/min, wavelengths were ꢁ = 225 nm and 275 nm,
tmeasurement = 11 min, T = 25 ◦C and injection volume was 10 ␮L.
Specific surface area and pore size distribution of sol–gel
immobilized and commercial palladium catalysts were obtained
by N2 adsorption measurements (N2-BET) using a Micromeritics
Gemini 1325 instrument. Transmission electron microscopy was
performed with a conventional LaB6-TEM Tecnai G220 S-TWIN
instrument (FEI Company, USA) operated at 200 kV and equipped
with EDAX-EDS for identification of elemental compositions.
High resolution XPS measurements were performed with a
Kratos Axis ultra X-ray photoelectron spectrometer for the deter-
mination of the oxidation state of palladium. The spectra were
acquired with monochromatic AlK␣ (1486.7 eV) X-ray source with
0◦ take off angle. The pressure in the test chamber was maintained
at 1.99 × 10−7 Pa during the acquisition process. Data analysis was
performed with Vision processing data reduction software (Kratos
Analytica Ltd. and Casa SPX Software Ltd.).
The size of palladium particles and the structure of the palla-
dium acetate immobilized on PhSiO2 were determined by X-ray
diffraction measurements (XRD) in X-ray apparatus (D8 Bruker
Advance) with KFL Cu 2k X-ray tube as source and Lyux Eye Detec-
tor. The samples were previously dried and pulverized.
Effective diffusion coefficients Deff were determined by a
method used for porous materials in the construction chemistry
[23]. For this purpose the immobilized catalysts were stored in an
aqueous SrCl2 solution for 24 h under N2. Then they were filtered
and placed in a stirred glass reactor with pure water. The increase
in electrical conductivity ꢂ(t) due to the diffusion of the salt ions out
of the pores into the solution was measured at 25 ◦C. The measured
diffusion coefficient D (Eq. (6)) was calculated from the slope m (Eq.
(7)) of the semilogarithmic plot (see Fig. S2b in Supporting Infor-
mation) of the relative conductivity change against time, where r
is radius of the catalyst nanoparticles:
For the calculation of the parameters, Eqs. (1)–(3) were used.
moil
˛ =
ꢀ =
ı =
(1)
(2)
(3)
moil + mH
O
2
msurfactant + mcosurfactant
msurfactant + mcosurfactant + mH O + moil
2
mcosurfactant
msurfactant + mcosurfactant
The conversions of reactants and yields of the products were
calculated from the concentrations determined by HPLC analysis
using the following equations:
co,styrene − ct,styrene
conversion X(t) =
· 100
(4)
(5)
co,styrene
ct,stilbene/Mstilbene
yield Y(t) =
· 100
co,styrene/Mstyrene
The amount of catalyst and ligand that leached from the
immobilized catalyst into the reaction mixture after the coupling
reaction was measured by ICP. The formed trans-stilbene was sep-
arated from the microemulsion by extraction with heptane or
dichloromethane, followed by solvent evaporation.
m · r2
D =
m =
(6)
(7)
ꢃ2
d ln((ꢂ(t) − ꢂend)/(ꢂO − ꢂend))
2.4. Instruments
dt
The dissolution of palladium from the silica support was
carried out by microwave decomposition (p = 20 × 105 Pa,
t = 35 min and T = 200 ◦C) with a CEM Discover SP-D (sample
preparation − digestion) instrument (CEM GmbH, Camp-
Lintfort/Germany). Before this procedure, the catalysts were
ball milled and mixed with 12 mL of a HNO3/HCl/H2SO4 solution
(2eq/6eq/4eq). After the microwave treatment, the solid white
silica was removed by filtration and the residual solution was
analyzed for palladium or/and phosphorus contents using a Varian
715-ES Optical Emission Spectrometer (ICP-OES). Calibration of
3. Results
3.1. Catalyst characterization
The mechanism of the sol–gel process was considerably
described in the literature by Avnir [24], Landau [25] and Schmidt
[26]. In our case palladium precursor was entrapped into a
hydrophobically modified silane matrix, which was formed from a
mixture of the hydrolized hydrophilic tetramethyl orthosilicate and

I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
213
Scheme 3. Entrapment of catalyst into sol–gel matrix with R1 = Ph or Oc and R = Et or Me.
immobilized catalysts are in the range 200–300 m2/g. The aver-
age pore size of the immobilized palladium catalyst was ≤2 nm
and pore volume Vpore ≈ 0.176 g/cm3. Only for OcSiO2, Pd(OAc)2@
OcSiO2 and in the presence of phosphine ligands, the surface areas
were significantly decreased. These results can be explained due
to the differences in the formed structures of sol–gel materials
containing octyl moieties [27].
Transmission electron microscopy image of the immobilized
Pd(OAc)2 catalyst is shown in Fig. 1. The catalyst has many
small pores with an average pore diameter of 1–5 nm as it was
already seen from the BET measurements. The entrapped palla-
dium nanoparticles with the metallic fcc structure were identified
hydrophobic triethoxyphenyl- or trimethoxyoctyl-silane according
to the steps shown in Scheme 3.
During the catalyst entrapment, red palladium (II) species
turned in black palladium (0) nanoparticles. Measurements were
done to analyze the oxidation state of the immobilized palla-
dium catalyst. Binding energy peaks at 335 eV and 340.471 eV were
detected, which are characteristic for 3d5/2 and 3d3/2 Pd(0) species.
The resulting palladium acetate loading on support material was
0.8–1.1 wt% and was determined from microwave decomposi-
tion of the catalyst followed by ICP measurements. No leaching
of the palladium was detected after the washing steps with
dichloromethane. From these results the immobilization efficiency
ꢄimmobilization of the catalyst can be estimated to be 1. Specific
surface areas of all catalysts are shown in Table 1. For the hydropho-
bic surface modification essentially phenyl groups were chosen to
provide a good compatibility with the aromatic reactants. Octyl
modified and unmodified catalysts were prepared for comparison
studies. From Table 1 it can be seen that BET surface areas of sol–gel
Table 1
Specific surface areas A of supported palladium catalysts.
Catalyst
Support
A [m2/g]
PhSiO2
PhSiO2
PhSiO2
PhSiO2
OcSiO2
OcSiO2
SiO2
SiO2
PhSiO2
SiO2
225.0
249.7
110.2
66.7
22.8
4.2
338.2
386.5
216.3
254.7
792.6
160.9
Pd(OAc)2
a
Pd(OAc)2
Pd(OAc)2
Pd(OAc)2
b
Pd(OAc)2
PdBr2
Pdc
Pdc
C
Pdc
Al2O3
a
TPPTS as ligand.
Xantphos as ligand.
Commercial catalyst.
b
c
Fig. 1. TEM image of palladium (II) acetate@PhSiO2.

214
I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
Fig. 2. Comparison of (a) catalyst/ligand precursors immobilized on hydropho-
bically modified PhSiO2 surface in coupling of bromobenzene with styrene and
(b) catalyst/substrate interactions (1.34 mmol alkene, 1.5 mmol PhX, 0.276 g K2CO3,
15 mg PdX2@PhSiO2, 46 mL microemulsion, 80 ◦C, AA = acrylic acid).
Fig. 3. Comparison of commercial and sol–gel immobilized catalysts (a) bro-
mobenzene/styrene and (b) iodobenzene/styrene (1.34 mmol alkene, 1.5 mmol PhX,
0.276 g K2CO3, 15 mg PdX2@support, 46 mL microemulsion, 80 ◦C).
PdBr2@PhSiO2 can be explained by a lower reactivity of the bro-
mobenzene.
from the XRD patterns (see Fig. S1 in Supporting Information).
From line shape analysis an average diameter for the Pd particles
of 3.0 0.3 nm was derived with the Scherrer equation.
The commercial catalysts tested in this reaction were very sta-
ble and no metal leaching was detected. However, the coupling of
bromobenzene and styrene with these catalysts was very slow in
comparison to the sol–gel immobilized catalysts (Fig. 3a). These
commercial catalysts are commonly applied for the hydrogenat-
ions of unsaturated substrates and are not suitable for the coupling
reactions.
3.2. Catalytic activity of the sol–gel immobilized palladium
composites
To optimize the reaction conditions, Heck coupling of bro-
mobenzene and styrene to trans-stilbene was performed with
heterogeneous sol–gel immobilized palladium catalysts prepared
from different orthosilicate precursors (Fig. 2a). In comparison to
Pd(OAc)2, the reaction rate with PdBr2 as precursor was slower by a
factor of about 2. No improvement of the catalytic performance was
obtained when the palladium catalysts were modified with water
soluble ligand TPPTS or with Xantphos. The decrease in activity by
using a metal-ligand complex can be explained by hindered access
of reactants or retarded catalyst dynamics due to the large structure
inside the pores of the catalyst. The ligand modified catalysts were
also not very stable and ligand leaching into the solutions occurred.
Catalysts prepared from PdBr2 or Pd(OAc2) as precursor were
also tested for the coupling of more reactive iodobenzene with
styrene or acrylic acid (Fig. 2b). In contrast to bromobenzene,
similar conversions were obtained in the coupling reaction with
iodobenzene regardless the used precursor. The lower reaction
rate for the coupling of bromobenzene with styrene catalyzed by
The same catalysts were tested for the coupling of more reactive
iodobenzene and styrene. The results are shown in Fig. 3b. Here, we
found that the reaction rates with commercial catalysts were much
higher than in the Heck coupling of bromobenzene and styrene.
The reaction rate with Pd@Al2O3 as catalyst was comparable to the
sol–gel immobilized catalyst. Therefore, the commercial catalysts
can be applied for Heck reaction. However, the reaction rate mainly
depends on the type of substrate. Furthermore, their reactivity is
lower compared to the hydrophobically modified sol–gel catalysts.
The strong difference in the behavior of the commercial catalysts
with certain substrates is unexpected and very difficult to predict.
3.3. Effect of the surfactant
As it was shown in Scheme 1, beside the parameters determined
by the preparation of the sol–gel immobilized Pd catalyst also the
selected solvent is important for the total catalytic performance. In

I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
215
Table 2
Structures and characteristic properties of the surfactants.
Surfactant
Structure
Type
cmc [gL−1] [43,44]
dmicelle [nm] [43,44]
O
SDS
Na+
N+
Anionic
2.250
2.06
O S O
O
Br
CTAB
Cationic
0.330
0.325
5.39 (4.3) [45]
H
O
n
O
Triton TX-100
Nonionic
8.68
N+
Br
TTAB
Cationic
1.344
3.84 [45]
O
H
n
O
O
Igepal CA-520
Nonionic
O
O
DSS
Anionic
Cationic
1.267 [46]
O
O S O
O
Na+
N+
Br
DTAB
4.330
3.34 [45]
OH
6
n
O
Marlipal 24/60
Nonionic
Nonionic
0.012
0.016
n=6-7
OH
Brij®35
7.94
23
O
Section 3.2 we studied the influence of the Pd precursor and sil-
ica support, and we could find optimized conditions with respect
to the catalyst synthesis. For the solubilization of the hydropho-
bic reactants in an aqueous environment, surfactants can be used.
The interactions between the silica support and formed micellar
aggregates will also affect the reaction rate. Therefore, Heck reac-
tion was investigated in aqueous microemulsions with different
anionic, nonionic and cationic surfactants. Their structures, criti-
cal micelle concentrations (cmc) and the average diameter of the
micelles (dmicelle) measured in aqueous-micellar solutions are given
in the Table 2.
All microemulsions were transparent and homogeneous except
microemulsions containing DSS or Igepal CA-520 as surfactant.
As can be seen from Fig. 4, a broad variety of surfactants can
be applied for the formulation of the reaction medium. Full con-
version was obtained after 6–7 h with cationic CTAB, nonionic
Triton TX-100 and anionic SDS as surfactants. The conversion
increases with increasing C-chain length, decreasing HLB-value
(hydrophilic–lipophilic balance) and decreasing cmc, if homologue
series of cationic surfactants CTAB, TTAB and DTAB were used:
Fig. 4. Influence of HLB values of surfactants on the conversion of the Heck
reaction (1.34 mmol styrene, 1.5 mmol bromobenzene, 0.276 g K2CO3, 15 mg
Pd(OAc)2@PhSiO2, 46 mL microemulsion, 80 ◦C).
C
16 > >C14 ≈ C12. The most hydrophobic surfactant CTAB is preferred
because of the more hydrophobic micelle cores (dmicelle) and largest

216
I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
Table 3
Substrate variationa
.
Entry
X
Y
W
Product
Reaction time t (min)
Yield Y (%)
1
2
3
4
5
6
7
8
9
Cl
Br
I
I
Br
I
Br
I
Br
I
Br
I
Br
I
H
H
H
Br
H
H
H
H
H
H
H
H
H
H
H
H
-COOH
-COOH
-COOH
-COOH
-COOMe
-COOMe
-COOCH2-CH(Et)(Bu)
-COOCH2-CH(Et)(Bu)
Cinnamic acid
Cinnamic acid
Cinnamic acid
429
396
359
513
372
407
359
451
389
477
339
435
387
422
407
485
327
555
480
398
458
433
393
477
367
444
446
419
448
371
415
377
318
70
76
98
77
45
72
64
82
64
62
46b
54b
45
89
50
70
62
4-Bromocinnamic acid
Trans-methylcinnamate
Trans-methylcinnamate
2-Ethylhexylcinnamate
2-Ethylhexylcinnamate
␣-Methylcinnamic acid
␣-Methylcinnamic acid
Methyl-(E)-␣-methylcinnamate
Methyl-(E)-␣-methylcinnamate
(E)-tButyl-2-methyl-3-phenylacrylate
(E)-tButyl-2-methyl-3-phenylacrylate
2-Hydroxyethyl-(E)-␣-methylcinnamate
2-Hydroxyethyl-(E)-␣-methylcinnamate
Trans-stilbene
␣-Me-COOH
␣-Me-COOH
␣-Me-COOMe
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
␣-Me-COOMe
␣-Me-COOtBu
␣-Me-COOtBu
Br
I
Cl
Br
I
␣-Me-COO(CH2)2-OH
␣-Me-COO(CH2)2-OH
Ph-
Ph-
Ph-
Ph-
Ph-
Ph-
Ph-
Ph-
Ph-
Ph-
Ph-
Ph-
Ph-
H
H
H
Br
Trans-stilbene
Trans-stilbene
100
94
89
57
41
59
44
55
65
60
0
20
50
92
32b
65b
I
4-Trans-bromostilbene
4-[(1E)-2-(phenyl)ethenyl]-benzenemethanamine
2-Trans-cyanostilbene
2-Trans-cyanostilbene
Trans-2-methylstilbene
2,6-Trans-dimethylstilbene
4-Trans-acetylstilbene
4-Trans-acetylstilbene
Distyrylbenzene
Br
Cl
Br
Cl
Cl
Cl
Br
Cl
Br
Br
I
Ortho-NH2-CH2-Ph-
Ortho-CN-Ph-
Ortho-CN-Ph-
Ortho-Me-Ph-
Ortho-Me-Ph-
MeOCPh-
MeOCPh-
Styrene
Styrene
H
Distyrylbenzene
Para-Cl-Ph-
Para-Cl-Ph-
␣-Me-Ph-
␣-Me-Ph-
4-Trans-chlorostilbene
4-Trans-chlorostilbene
␣-Methylstilbene
H
H
H
Br
I
␣-Methylstilbene
a
Reaction conditions: 1.5 mmol olefine, 1.34 mmol alkene or styrene, 2 mmol K2CO3, 1.25 wt% Pd(OAc)2@ PhSiO2 (0.067 mmol Pd(OAc)2), 46 mL microemulsion (3.3 wt%
CTAB, 89.3 wt% H2O, 6.6 wt% propanol), 80 ◦C.
b
1:1 product: terminal olefine
storage capacity for the substrates. Predictions cannot be done for
the behavior of the other not homologue surfactants in this reaction
because of their different structures. There is also no dependence
on the type of the surfactant and the reaction rate.
para-substituted substrates (e.g. para-iodobromobenzene)
are applied in the reaction (Table Entry 21–29) and is
3
very small with the ortho-substituted substrates (e.g. ortho-
methylchlorobenzene). The reaction rate depends not only on
the position of the substituent, but also on the substituent type.
Substituents with electron withdrawing groups (e.g. cyano or
acetyl group) are more active than these with electron donating
groups (e.g. methoxy or methyl group).
The nonaromatic acrylic acid shows good reactivity with
iodobenzene and its reactivity is comparable to aromatic styrene.
The more sterically hindered methacrylic acid and derivates were
less reactive (Table 3 entry 5–16) and side-products were formed.
The arylation of medically interesting 1,1-substituated olefine like
methacrylate or ˛-methylstyrene can afford different products in
dependency of the direction of the ˇ-H-elimination: the main
desired product ˛-methylcinnamate with E- or Z-stereochemistry
or a by-product ˛-benzylacrylate with a terminal double bond. The
product and by-product ratio was 1:1.
3.4. Substrates
To extend the scope of the protocol, we applied the palladium(II)
acetate immobilized on hydrophobically modified silica catalyst to
screen for a wide variety of substrates. A screening of aryl halides
Y-Ph-X shows that this catalytic system is suitable for all kinds of
aryl halide. It results in the expected ranking with higher reactiv-
ity of iodobenzene than bromobenzene, 1-bromo-4-iodobenzene
(with 4-bromo-trans-stilbene as product) and lowest reactivity of
less active chlorobenzene as can be seen from Table 3. These results
are typical for the reaction of aryl halide with styrene (Entry 17–20),
acrylic acid (Entry 1–4) and methacrylic acid (Entry 9–10).
Because of the mechanistic aspects of Heck coupling, trans-
products are favored. The reactivity of styrene is higher than that
of 4-chlorostyrene and ˛-methylstyrene (with terminal olefins
as by products 1:1) as shown in Table 3 (Entry 30–33) for the
reaction with iodo- and bromobenzene. Also different aromatic
ortho- and para- substituted aryl halides Y-Ph-X react with
styrene to trans-stilbene derivates. The reactivity decreases if the
All substrates tested in Heck coupling reaction in aqueous
microemulsions show similar reactivities as in commonly used
organic solvents, e.g. dimethylformamide or acetonitrile, at higher
temperatures. For the following investigations only the reactions
of styrene and iodo- or bromobenzene to trans-stilbene were con-
sidered.

I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
217
Table 4
Weisz-modulus for the Pd(OAc)2 catalysts immobilized on silica.
Material
Deff
,
[cm2 s−1
]
r0 [mol l−1 ·s−1
]
Weisz-Modulus ꢅ
styrene
Pd(OAc)2@SiO2
6.0 × 10−10
1.2 × 10−9
7.5 × 10−7
2.3 × 10−6
3.0 × 10−6
4
3
5
Pd(OAc)2@PhSiO2
Pd(OAc)2@OcSiO2 3.6 × 10−9
In Eq. (8), Lcat is radius of the catalyst nanoparticles, n is
reaction order, r is the reaction rate, ꢆcat is density of the cata-
lyst, c0 is initial styrene concentration and Deff,styrene is effective
diffusion coefficient of styrene. For ꢅ ꢁ 1 no diffusion limita-
tion influences the reaction. For our discussion the measured
diffusion coefficients Deff,tracer of SrCl2 were transferred in the
effective diffusion coefficients of styrene Deff,styrene according to
the ratio of the molecular diffusion coefficients at ambient tem-
perature in water (DM,tracer = SrCl2 = 1.4 × 10−4 cm2 s−1 [31] and
DM,styrene = styrene = 1 × 10−5 cm2 s−1 [32]):
Fig. 5. Influence of the hydrophobicity of the sol–gel support on the cat-
alytic activity (1.34 mmol styrene, 1.5 mmol bromobenzene, 0.276 g K2CO3, 15 mg
Pd(OAc)2@support, 46 mL microemulsion, 80 ◦C).
ε
Deff,styrene = DM,styrene
·
(9)
ꢇ
Deff,tracer
ε
ꢇ
=
(10)
3.5. Effect of surface modification
DM,tracer
As shown in Section 3.2, Figs. 2 and 3, various commercial and
synthesized palladium catalysts were tested for the Heck coupling.
Pd(OAc)2 immobilized on PhSiO2 was the choice of catalyst for
further investigations, because of its high stability and reactivity.
The influence of different silica support materials on the reaction
rate was also studied. In addition to the nonmodified sol–gel pre-
pared from tetramethyl orthosilicate alone, two hydrophobically
modified sol–gel matrices were prepared using octyl and phenyl
substituted alkoxysilane precursor.
In Eq. (10), ␧ is the porosity of the immobilized catalyst and ꢇ is
the tortuosity.
The porosity of Pd(OAc)2@PhSiO2 can be calculated from the
ratio of pore volume Vpore (BET measurement) to the volume of
catalyst resulting in ε = 33.4%. The tortuosity ꢇ describes the way
of the diffusion in the pores of the catalyst and can be determined
from Equation 10 (ꢇ = 4.689). The pore diameter is 1–2 nm.
The initial reaction rates using different catalysts were deter-
mined according to:
The adsorption of the aromatic compounds used in the reac-
tion was stronger if catalysts immobilized on hydrophobic support
were used. The inverse adsorption effect was already shown in
our previous publication about the enantioselective hydrogena-
tion of itaconic acid and derivates [28]. The hydrogenation of
hydrophilic nonaromatic substrates is favored using Rh/BPPM
catalysts immobilized on hydrophilically modified sol–gel mate-
rials. Whereas the reaction is much slower when performed
using hydrophobically modified surfaces. The transport processes
of the reactants in the catalyst pores can be described by the
diffusion coefficients Deff,tracer which were calculated from dif-
fusion measurements with SrCl2 salts as tracers [29]. From the
diffusion coefficients Deff,tracer it is also clear that the adsorp-
tion of the substrates on hydrophilic surfaces is smaller than on
hydrophobically modified surfaces (see Fig. S2a and b in Suppor-
ting Information). The determined diffusion coefficients Deff,tracer
show the following ranking: SiO2 (8.1 × 10−9 cm2 s−1) < OcSiO2
(1.5 × 10−8 cm2 s−1) ꢁ PhSiO2 (4.7 × 10−8 cm2 s−1).
dX
r0 = c0
·
(11)
dt
The average particle size of 200 ␮m and ꢆcat of 1.9 g/cm3 were
considered for all calculations (Table 4).
In our case the Weisz modulus was ꢅ ≈ 3 for the heteroge-
neous reaction with mostly applied Pd(OAc)2@PhSiO2, indicating
a substantial limitation of the reaction rate by the diffusion of the
reactants inside the pores of the catalyst. This result is also sup-
ported by the comparison of the temperature dependency of the
heterogeneous and the homogeneous catalysts as will be discussed
in Section 3.6. The pore efficiency ꢄpore can be calculated from the
Weisz-Prater criterion ꢅ for various reaction orders [33]: ꢄpore = 1
(for ꢅ < 1, no diffusion limitations) and ꢄpore = ꢅ−1 (for ꢅ ≥ 1, dif-
fusion limited reaction). From the correlation for ꢅ ≥ 1, for the
applied Pd(OAc)2@PhSiO2 catalysts a value of about 40% is obtained
for ␩pore. By decreasing the Pd(OAc)2 loading on silica material
from 1.2% to 0.5%, the specific productivity of the palladium was
increased by a factor of 2 from 2.5 × 10−4 to 5 × 10−4 mol/gPd·min
(see Fig.S3 in Supporting Information).
As shown in Fig. 5, the hydrophobicity of the sol–gel affects the
reaction rate and conversion of the palladium catalyzed coupling of
bromobenzene and styrene. The best results were obtained when
phenyl moiety was introduced, followed by octyl modified catalyst.
The reaction was the slowest when hydrophilic modification was
applied.
3.6. Kinetics of the reaction
The coupling of styrene with bromobenzene or iodobenzene
is known to follow a second order reaction rate [34,35]. In our
studies the reaction rate also increases linearly with the con-
centrations of both reactants. The distinction has to be drawn
between homogeneously (in situ formed not supported palladium
(0) nanoparticles) and heterogeneously catalyzed Heck reaction
(Pd(OAc)2 immobilized on silica by a sol–gel method). In compar-
ison to the homogeneously catalyzed reaction, the reaction with
sol–gel immobilized catalyst (1.2% Pd(OAc)2 on phenyl modified
silica) is slower at the beginning as can be seen from the mea-
surements at 80 ◦C shown in Fig. 6a, but finally proceed to reach
To examine the influence of the diffusion of the reactants inside
the catalyst pores on the reaction rate, the ratio of the reac-
tion time constant and the diffusion time constant are estimated
and expressed as the Weisz modulus ꢅ. This is a characteristic
parameter to test for mass transfer limitations on heterogeneously
catalyzed reactions and can be calculated using the Weisz–Prater
criterion [15,30]:
n + 1
r · ꢆcat
ꢅ = Lc2at
·
·
(8)
2
co · Deff,styrene

218
I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
Fig. 6. Comparison of homogeneous and heterogeneous catalyzed Heck reactions
in microemulsion (a) Xstyrene-t-curve (80 ◦C), (b) Arrhenius plot (1.34 mmol styrene,
1.5 mmol PhBr, 0.276 g K2CO3, 15 mg Pd(OAc)2@PhSiO2 or 15 mg homogeneous
Pd(OAc)2, 46 mL microemulsion).
a higher conversion than the homogeneous catalyst. This experi-
ment was repeated for temperatures ranging from 50 to 90 ◦C (for
both catalysts).
Activation energies EA were determined for the coupling of bro-
mobenzene and styrene using homogeneous and heterogeneous
catalysts (Fig. 6b) according to Eq. (12):
EA
ln(k) = −
+ ln(k
)
(12)
∞
R · T
Fig. 7. Recycling of sol–gel immobilized catalyst for (a) iodobenzene, styrene,
Pd(OAc)2@PhSiO2; (b) bromobenzene, styrene, Pd(OAc)2@OcSiO2; (c) bromoben-
zene, styrene, Pd(OAc)2@PhSiO2 (1.34 mmol activated alkenes, 1.5 mmol aryl
halides, 0.276 g K2CO3, Pd(OAc)2@R1SiO2, 46 mL microemulsion, 80 ◦C).
E
E
A, heterogeneous = 60.4 kJ mol−1
A, homogeneous = 74.5 kJ mol−1
The activation energy of the heterogeneous reaction is lower
than that of the homogeneous reaction, because of the diffusion
limitation that is much less temperature sensitive than the chemi-
cal reaction itself.
catalysts, done in the kinetically controlled regime, the impact of
intercalation was calculated to be ꢄintercalation = 0.81.
It is known that for porous heterogeneous catalysts pore dif-
fusion limitations can lead to lower rates in comparison to the
homogenously catalyzed reaction. By adjusting the catalyst size
(smaller particles) or temperature (lower temperatures) it is possi-
ble to suppress the diffusion limitation effect because the reaction
rate is no longer limited to the rate for the transport of reactants.
In this case, the catalyst efficiency is comparable to the homoge-
nous catalyst. For supported metal catalysts, in which a metal is the
active material, it is possible that the reaction rate remains lower
even if diffusion limitation is eliminated because some of the metal
can be intercalated and not be available for the reactants. By com-
parison of the reaction rates for the supported and unsupported
3.7. Catalyst recycling
The recycling experiments were carried out for the cou-
pling of bromobenzene or iodobenzene with styrene to produce
trans-stilbene. Palladium(II)acetate catalysts immobilized on
hydrophobic octyl or phenyl modified silica supports could be
recycled 6 times without any significant decrease in activity (Fig. 7).
Only small amount of palladium leaching into the solution was
detected by ICP measurements after each reaction as can be seen in
Table 3 and no leaching was detected after second and the following
recycling steps.

I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
219
Table 5
Catalyst leaching into the reaction solution.
d,e
e
Catalyst precursor
Support
gPd(OAc) /gsupport
[%]
Leaching after reaction mPd(OAc)
solution/m
[%]
in
Pd(OAc) catalyst
2
2
2
Pd(OAc)2
PdBr2
Pd(OAc)2
PhSiO2
PhSiO2
PhSiO2
PhSiO2
OcSiO2
SiO2
SiO2
Charcoal
Al2O3
1.2
1.2
1.2
1.2
1.2
1.2
5.0
10.0
1.0
0.010–0.012
0.031–0.077
0.062 (5.475 ligand)
0.068 (7.800 ligand)
0.013
0.013
0
0
0
a
b
Pd(OAc)2
Pd(OAc)2
Pd(OAc)2
Pdc
Pdc
Pdc
a
TPPTS.
Xantphos.
b
c
Commercial catalyst.
Determined by microwave decomposition.
Palladium and ligand amount determined from ICP measurements.
d
e
Different tests were performed to check if the active catalyst
earlier, but no conversions were obtained for the less water-
soluble substrates bromo- or chlorobenzene [6]. Heck reaction
of iodobenzene and styrene in aqueous-micellar solutions of dif-
ferent surfactants with palladium catalyst and PPh3 ligand [41]
or Suzuki coupling with Pd/C catalyst could also be performed
[20]. The addition of small amount of alcohol as a cosurfac-
tant and the formation of aqueous microemulsion is desirable,
for example for a Heck reaction in a microemulsion of TX-
100/n-heptane/butanol/water/propylene glycol mixture [35]. A
comparison of the reaction of styrene and bromobenzene in dif-
ferent reaction media with Pd(OAc)2@PhSiO2 is given in Table 6,
including some examples from literature.
To clearly show the micellar effect it is common to compare
the surfactant system with pure water. Here, the reaction in water
was also possible, but the substrates could not be completely sol-
ubilized. A small increase in solubility was achieved by adding
the hydrophilic surfactant SDS and much higher solubilization was
achieved if the hydrophobic surfactant CTAB was added. The full
solubility could only be obtained by using an aqueous microemul-
sion with a short chain alcohol as cosurfactant. Under complete
dissolution of the reactants full conversion at 80 ◦C is obtained in
420 min. We tested the same reaction in different organic solvents
instead of microemulsions and activity was much lower. Within the
same reaction time, the conversion was only about 30%. As shown
by other groups, in organic solvents higher temperatures (about
130–150 ◦C) are needed to obtain higher conversions.
is indeed heterogeneous. First the hot filtration of the catalyst
after obtaining a conversion of about 50% was done. Thereafter,
the reaction was continued with the residual solution. No fur-
ther conversion was observed in the reaction as a proof that no
homogeneous catalyst was present in the reaction mixture. Also a
poisoning test was done with 300 eq of highly cross linked poly(4-
vinylpyridine) [10,36,37]. This substance can bind to homogeneous
Pd(II) species and poisons it, resulting in a decrease in reaction
rate. The reaction with heterogeneous catalyst (Pd(OAc)2@PhSiO2)
showed no decrease in reactivity upon the addition of poly(4-
vinylpyridine). Also the analysis of the reaction mixtures after the
reactions (Table 5) showed only a small amount of leaching of Pd
into the solution. From these experiments we conclude that no
homogeneous catalyst is involved in the Heck coupling with sol–gel
immobilized catalyst in aqueous microemulsion.
From these results the recycling efficiency of the sol–gel immo-
bilized catalyst ꢄrecycling can be estimated. For mostly applied
Pd(OAc)2@PhSiO2 only 0.01% palladium acetate leaching after the
first reaction and no further metal leaching was determined after-
wards. That’s why the recycling efficiency ꢄrecycling of the catalyst
can be estimated to be 0.999.
4. Discussion
Heck coupling of iodobenzene and styrene catalyzed by sol–gel
immobilized palladium (II) acetate represents a good alternative to
the commonly used homogeneous, e.g. Pd(dba)2/phosphoramidite,
catalyzed coupling of iodobenzene and styrene in dimethyl-
formamide at 80 ◦C (TOF = 315/h) [38]. Also heterogeneously
catalyzed reactions, e.g. Heck coupling of styrene and 4-
bromofluorobenzene catalyzed by palladium immobilized on
zeolite at 140 ◦C in dimethyl acetamide (TOF[Pd(0)]-NaY = 20/h
and TOF[Pd(OAc)2]-NaY = 2000/h) [2], or reactions under biphasic
conditions, e.g. PdCl2(bipy) catalyzed coupling of iodobenzene
and styrene in n-octane/p-xylene/ethylene glycol at 150 ◦C
(TOF = 227/h) [39], were already reported by other research groups.
The reactions in conventional solvents at high temperatures, which
were already reported in the literature, had higher turnover
frequencies. For example, Heck coupling catalyzed by Pd/C in
NMP at 150 ◦C [40] had TOF > 18,000/h. By lowering the reac-
tion temperature the Turnover Frequency would decrease below
During the reaction the homogeneous Pd(II) species are reduced
to the black active Pd(0) in the presence of surfactant.
The use of the sol–gel immobilized palladium catalyst in the
Heck reaction with different substrates is preferable because of the
higher activities in comparison to the commercial catalysts. The
catalysts can be reused more than 6 times without loss in activ-
ity and with only a minor leaching of the palladium species into
the reaction solution and can indeed be classified as a heteroge-
neous catalyst. The separation process (Scheme 4) of the catalyst
and product after the reaction is very simple: (a) filtration and reuse
of solid catalyst, (b) product separation through extraction with
heptane and (c) separation and reuse of surfactant from water, e.g.
by micellar enhanced ultra filtration (MEUF) or adsorption on silica
or granite sand.
At last the overall efficiency ꢄ of porous Pd(OAc)2 catalyst immo-
bilized on PhSiO2 was estimated from the different efficiencies,
which were calculated above by multiplication of all factors:
◦
TOF80 C = 8–140/h. On the other hand, a more environmentally
ꢄ = ꢄimmobilisation · ꢄintercalation · ꢄpore · (> Ncycle) × ꢄ(>N)
(13)
friendly process management on the example of Heck coupling in
microemulsion with >90% of water at lower temperatures could be
shown. In addition to this, the catalyst and product recycling was
enabled.
The reaction of iodobenzene and styrene with Pd/C in differ-
ent aqueous-micellar solutions of cationic surfactants was reported
recycling
Á = 1·0.81·0.4·(>6)·(>0.999) = (≥1.92)
In Eq. (13), ꢄimmobilization and ꢄrecycling are palladium acetate
leaching fractions into the solution after the catalyst synthesis
and after N recycling experiments and were determined from

220
I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
Table 6
Effect of reaction media on the Heck coupling of bromobenzene and styrene.
Solvent
Catalyst
Base
T [◦C]
X [%]
t [min]
Ref.
MeOHa
Pd(OAc)2@PhSiO2
Pd(OAc)2@PhSiO2
Pd(OAc)2@PhSiO2
PdCl2@zeolite
Pd@hydroxyapatite
Pd(OAc)2@PhSiO2
Pd(OAc)2@PhSiO2
Pd(OAc)2@PhSiO2
Pd(OAc)2@PhSiO2
Pd@polystyrene
Pd(OAc)2@PhSiO2
Pd(OAc)2@PhSiO2
Pd@C
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
KOH
K2CO3
K2CO3
K2CO3
K2CO3
NEt3
80
80
80
153
130
80
80
80
80
90
10
31
30
74
94
17
2
45
88
10
100
50
438
380
433
MeOHa,b
DMFa
DMF
150
[1]
[47]
NMP
1440
1335
1334
420
420
1320
420
360
240
420
120
DMFa,b
DMFa,c
ACN/H2O (1:1)a
H2Oa
H2Oe
[48]
10 g/L CTAB/H2Oa
70 g/L SDS/H2Oa
CTAB/H2Oe
CTAB/propanol/H2Oa,d
TX-100/H2O/[BMIM]PF6
80
80
80
80
60
100
100
[6]
Pd(OAc)2@PhSiO2
PdCl2
e
100
[49]
a
Reaction conditions: 1.5 mmol bromobenzene, 1.34 mmol styrene, 2 mmol K2CO3, 1.25 wt% Pd(OAc)2@PhSiO2 (0.067 mmol Pd(OAc)2), V = 46 mL, 80 ◦C.
TPPTS.
Xantphos.
3.3 wt% CTAB, 89.3 wt% H2O, 6.6 wt% propanol.
Iodobenzene and styrene as reactants.
b
c
d
e
Scheme 4. Product and catalyst separation.
ICP measurements, Ncycle is the number of recycling experiments,
the catalyst:
ꢄintercalation is the efficiency of the immobilized catalysts compared
to the homogeneous catalysts and can be estimated from the com-
parison of reaction rates for homogeneously and heterogeneously
catalyzed reaction and ꢄpore is the pore efficiency calculated from
the Weisz–Prater criterion ꢅ for the second order reaction. For
ꢅ = 3, the pore efficiency was 40%. Because no deactivation and
leaching of the catalyst after sixth recycling step was seen, further
application of the catalyst in much more recycling steps can be pre-
dicted. That’s why the number of the recycling steps was estimated
to be >6.
nproduct/nPd(OAc)
TON
t
2
TOF =
=
t
(1.34 mmol)/0.0668 mmol Pd(OAc)2
20.1
6h
3.34
h
=
=
=
(14)
6 h
For the most often used Pd(OAc)2@PhSiO2, the TON is 20 and TOF
is 3.3/h for the reaction of iodobenzene and styrene. For the homo-
geneous Pd(OAc)2 catalyzed reaction the same results would be
obtained. TOF should be >500/h (TON > 1000) for industrially usable
catalysts [42]. In Fig. 7 we have shown that Pd(OAc)2@PhSiO2 could
be recycled more than 6 times without loss in activity. Considering
these results the TON would be >501 after 25 runs, which is use-
ful for industrial application. The reactant concentrations can be
increased up to 5 times, which would also increase the Turnover
Number to TON = 100 after one run up to TON > 1000 for 10 runs.
Catalyst efficiencies ꢄ ≥ 1 are typical for very stable and active
catalysts and demonstrate that palladium acetate immobilized on
hydrophobically modified silica by sol–gel method is a good alter-
native to homogeneous one. In comparison to these results, the
catalysts with high metal leaching and low reactivity would show
smaller section efficiencies and the complete efficiency of the cat-
alyst would be ꢁ1.
Of course is the immobilization of homogeneous catalysts dis-
advantageous in comparison to homogeneous catalysts because of
diffusion limitations and decrease in reactivity. But on the other
side the possibility to recycle the catalysts more than 6 times
increases the efficiency of sol–gel immobilized catalysts enor-
mously and reduces the cost of the complete process.
5. Conclusion
Heck reaction in aqueous microemulsions with Pd(OAc)2 or
PdBr2 catalysts immobilized on modified silica support mate-
rials is a alternative to the conventional organic solvents. One
phase microemulsions (water/propanol/CTAB) with the consis-
tence ˛ = 0.9%, ꢀ = 9.9% and ı = 66.7% as reaction media allows a
better solubility of hydrophobic substrates in water. To improve
The Turnover Number (TON) and Turnover frequency (TOF)
provides the information about the activity and efficiency of

I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 393 (2014) 210–221
221
the reactivity of the sol–gel immobilized catalysts the application of
hydrophobically modified surfaces (PhSiO2) is preferred, because of
the attractive interactions between the hydrophobic aromatic sub-
strates, e.g. iodobenzene, and the palladium catalyst immobilized
on phenyl modified silica. The synthesized catalysts are more active
in comparison to commercial catalysts (e.g. Pd@SiO2 or Pd@C) and
can be used in a variety of Heck reactions. These immobilized cat-
alysts could be recycled more than 6 times without loss in activity.
No homogeneous species were detected in the reaction mixture, as
proved by the heterogeneity tests. Due to the high porosity and the
strongly branched pore systems of the sol–gel materials the diffu-
sion of small reactant molecules inside the pore systems was not
very high. To obtain a good utilization of the active noble metal
immobilized in the support material the loading should be limited
to an amount that still guarantees a high effectiveness factor of the
composite catalysts (see Fig. S3 in Supporting Information). The
influence of the mass transport limitations on the reaction rate
can be quantified from the comparison of the activation energies
EA of the coupling reaction between bromobenzene and styrene
with sol–gel immobilized catalyst and with homogeneous cata-
lyst. It was also showed that the Weisz-modulus was ꢅ > 1. With
a well adjusted Pd(OAc)2 loading in the range of 0.5% a highly effi-
cient catalyst was obtained for the reaction tested here that also
show stable recycling behavior. The immobilization process of the
palladium catalysts was disadvantageous in comparison to homo-
geneous catalysts. But on the other side the possibility of catalyst
recycling more than 6 times compensate this disadvantage and a
catalyst efficiency ꢄ ꢂ 1 could be arrived, which is characteristic for
very stable and active heterogeneous catalysts.
[7] C.M. Crudden, M. Sateesh, R. Lewis, C. Kl, J. Am. Chem. Soc. 127 (2005)
10045–10050.
[8] V. Polshettiwar, Á. Molnár, Tetrahedron 63 (2007) 6949–6976.
[9] S. Jana, B. Dutta, R. Bera, S. Koner, Inorg. Chem. 47 (2008) 5512–5520.
[10] Y. Ji, S. Jain, R.J. Davis, J. Phys. Chem. B 109 (2005) 17232–17238.
[11] F. Gelman, D. Avnir, H. Schumann, J. Blum, J. Mol. Catal. A: Chem. 146 (1999)
123–128.
[12] Z. Nairoukh, J. Blum, J. Mol. Catal. A: Chem. 358 (2012) 129–133.
[13] K. Hamza, H. Schumann, J. Blum, Eur. J. Org. Chem. (2009) 1502–1505.
[14] T. Yosef, R. Schomäcker, M. Schwarze, M. Fanun, F. Gelman, J. Mol. Catal. A:
Chem. 351 (2011) 46–51.
[15] D. Meltzer, D. Avnir, M. Fanun, V. Gutkin, I. Popov, R. Schomäcker, M. Schwarze,
J. Blum, J. Mol. Catal. A: Chem. 335 (2011) 8–13.
[16] S. Dahoah, Z. Nairoukh, M. Fanun, M. Schwarze, R. Schomäcker, J. Blum, J. Mol.
Catal. A: Chem. 380 (2013) 90–93.
[17] A. Rosin-Ben Baruch, D. Tsvelikhovsky, M. Schwarze, R. Schomäcker, M. Fanun,
J. Blum, J. Mol. Catal. A: Chem. 323 (2010) 65–69.
[18] B.H. Lipshutz, S. Ghorai, Ald. Acta 41 (2008) 59–72.
[19] S. Bhattacharya, A. Srivastava, S. Sengupta, Tetrahedron Lett. 46 (2005)
3557–3560.
[20] A. Arcadi, G. Cerichelli, M. Chiarini, M. Correa, D. Zorzan, Eur. J. Org. Chem. 2003
(2003) 4080–4086.
[21] L. Botella, C. Nájera, Tetrahedron 60 (2004) 5563–5570.
[22] C.H. Kee, et al., Org. Biomol. Chem. 8 (2010) 5646–5660.
[23] A. Buchwald, C. Kaps, GDCh-Monographie, 1999.
[24] D. Avnir, Acc. Chem. Res. 28 (1995) 328–334.
[25] G. Ertl, H. Knözinger, F. Schüth, Handbook of Heterogeneous Catalysis, Wiley-
VCH, Weinheim, 2008.
[26] H.K. Schmidt, Chem. Unserer Zeit 3 (2001) 176–184.
[27] J. Parka, M. Komarnenia, S. Lima, W.T. Heoa, N.H. Choia, J. Mater. Res. 15 (2000)
1437–1440.
[28] I. Volovych, M. Schwarze, T. Hamerla, J. Blum, R. Schomäcker, J. Mol. Catal. A:
Chem. 366 (2013) 359–367.
[29] W. Jost, K. Hauffe, Ber. Der Bunsenges. Phys. Chem. 66 (1962) 77.
[30] P.B. Weisz, C.D. Prater, Adv. Catal. 6 (1954) 143–196.
[31] M. Polestra, H.S. Harned, J. Am. Chem. Soc. 75 (1953) 4168–4169.
[32] E.S. Meadows, T.J. Crowley, C.D. Immanuel, Ind. Eng. Chem. Res. 42 (2003)
555–567.
[33] G. Ertl, H. Knözinger, F. Schüth, Handbook of Heterogeneous Catalysis, Wiley-
VCH, 2006.
[34] S.V. Jagtap, R.M. Deshpande, React. Kinet. Mech. Cat. 106 (2012) 457–473.
[35] J.-Z. Jiang, C. Cai, J. Colloid Interface Sci. 299 (2006) 938–943.
[36] W.J. Sommer, K. Yu, J.S. Sears, Y. Ji, X. Zheng, R.J. Davis, C.D. Sherrill, C.W. Jones,
M. Weck, Organometallics 24 (2005) 4351–4361.
[37] J.M. Richardson, C.W. Jones, Adv. Synth. Catal. 348 (2006) 1207–1216.
[38] D.L. Dodds, M.D.K. Boele, G.P.F. van Strijdonck, J.G. de Vries, P.W.N.M. van
Leeuwen, P.C.J. Kamer, Eur. J. Inorg. Chem. (2012) 1660–1671.
[39] S.V. Jagtap, R.M. Deshpande, Catal. Today 131 (2008) 353–359.
[40] L. Yin, J. Liebscher, Chem. Rev. 107 (2007) 133–173.
[41] G. Lu, C. Cai, Colloids Surf. A 355 (2010) 193–196.
[42] A. Behr, Applied Homogeneous Catalysis, Wiley-VCH, Weinheim, 2008.
[43] M. Schwarze, J.S. Milano-Brusco, V. Strempel, T. Hamerla, S. Wille, C. Fischer,
W. Baumann, W. Arlt, R. Schomäcker, RSC Adv. 1 (2011) 474–483.
[44] T. Hamerla, Untersuchungen zur Hydrierung von prochiralen Olefinen mit
Rhodium-Katalysatoren in Tensidsystemen, Diploma thesis, 2010.
Acknowledgment
We gratefully acknowledge the financial support of this
study by Deutsche Forschungsgemeinschaft (DFG) through grant
SCHO687/8-2.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.molcata.2014.06.016.
References
[45] K. Weidemaier, H.L. Tavernier, M.D. Fayer, J. Phys. Chem.
9352–9361.
B 101 (1997)
[1] A. Corma, H. Garcıꢀa, A. Leyva, A. Primo, Appl. Catal. A 247 (2003) 41–49.
[2] L. Djakovitch, K. Koehler, J. Am. Chem. Soc. 123 (2001) 5990–5999.
[3] P. Li, L. Wang, L. Zhang, G. Wang, Adv. Synth. Catal. 354 (2012) 1307–1318.
[4] C.W. Lim, I.S. Lee, Nano Today 5 (2010) 412–434.
[46] A. Chatterjee, S.P. Moulik, S.K. Sanyal, B.K. Mishra, P.M. Puri, J. Phys. Chem. B
105 (2001) 12823–12831.
[47] K. Mori, K. Yamaguchi, T. Hara, T. Mizugaki, K. Ebitani, K. Kaneda, J. Am. Chem.
Soc. 124 (2002) 11572–11573.
[5] Q. Du, W. Zhang, H. Ma, J. Zheng, B. Zhou, Y. Li, Tetrahedron 68 (2012)
3577–3584.
[48] A. Ohtaka, T. Yamaguchi, T. Teratani, O. Shimomura, R. Nomura, Molecules 16
(2011) 9067–9076.
[6] M.M. Shinde, S.S. Bhagwat, J. Dispers. Sci. Technol. 33 (2012) 117–122.
[49] G. Zhang, H. Zhou, J. Hu, M. Liu, Y. Kuang, Green Chem. 11 (2009) 1428–1432.