Synthesis, catalytic activity, and leaching studies of a heterogeneous Pd-catalyst including an immobilized bis(oxazoline) ligand

Article abstract of DOI:10.1016/j.jcat.2011.10.013

The synthesis and characterization of a novel catalytic system including Pd(OAc)₂ attached to a bis(oxazoline) (=BOX) ligand that is covalently bonded to 3-mercaptopropyl-functionalized silica gel is presented. The catalyst was tested for Suzuki-Miyaura reactions of different aryl halides with phenylboronic acid. The heterogeneity of the catalytic system was investigated using different approaches, indicating that there is virtually no Pd leaching into the reaction solution under the applied reaction conditions. Furthermore, our results show that the catalytic system can be reused multiple times without significant loss of stability or structure.

Full text of DOI:10.1016/j.jcat.2011.10.013

Journal of Catalysis 286 (2012) 30–40
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Synthesis, catalytic activity, and leaching studies of a heterogeneous Pd-catalyst
including an immobilized bis(oxazoline) ligand
H. Gruber-Woelfler a, P.F. Radaschitz a, P.W. Feenstra a, W. Haas b,c, J.G. Khinast a,
⇑
a Institute for Process and Particle Engineering, Graz University of Technology, Inffeldgasse 21a, A-8010 Graz, Austria
b Institute for Electron Microscopy, Graz University of Technology, Steyrergasse 17, A-8010 Graz, Austria
c Christian Doppler Laboratory for Nanocomposite Solar Cells, Graz University of Technology, Graz, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis and characterization of a novel catalytic system including Pd(OAc)2 attached to a bis(oxaz-
oline) (=BOX) ligand that is covalently bonded to 3-mercaptopropyl-functionalized silica gel is presented.
The catalyst was tested for Suzuki–Miyaura reactions of different aryl halides with phenylboronic acid.
The heterogeneity of the catalytic system was investigated using different approaches, indicating that
there is virtually no Pd leaching into the reaction solution under the applied reaction conditions. Further-
more, our results show that the catalytic system can be reused multiple times without significant loss of
stability or structure.
Received 25 May 2011
Revised 12 October 2011
Accepted 15 October 2011
Available online 23 November 2011
Keywords:
Supported Pd-catalyst
Suzuki-Miyaura reaction
3-Mercaptopropyl silica gel
Bis(oxazoline) ligand
Leaching
Ó 2011 Elsevier Inc. All rights reserved.
Proof of heterogeneity
1. Introduction
avoid metal carryover is intricate and costly, reducing the potential
for industrial implementation, especially in the field of pharmaceu-
Catalytic C–C, C–O, and C–N couplings using Pd-catalysts are
widely employed both in academia and in industry, making Pd
the most extensively used metal catalyst for the synthesis of a wide
variety of organic compounds [1–3], including pharmaceuticals
[4,5], heteroarenes [6], liquid crystals [7] as well as natural prod-
ucts prepared by total synthesis [8]. One of the most successful
strategies for C–C bond formation is the Suzuki–Miyaura coupling
[1,2,4,9,10], which involves the reaction of organoboron reagents
with aryl halides [11,12]. Due to the high stability of the organobo-
ron reagents, their ease of preparation and low toxicity, as well as
the importance of the produced biaryls for pharmaceuticals [4,5]
and liquid crystals [13], this reaction has become part of the stan-
dard toolbox of organic chemists. The importance of the reaction is
also reflected by the fact that one of the inventors, Akira Suzuki,
was awarded with the Nobel Prize in Chemistry in 2010.
A plethora of papers and patents claiming new and highly active
catalytic systems for Suzuki–Miyaura couplings has been published.
For recent reviews see [9,14–17]. The main shortcoming of many of
the reported systems is the homogeneous nature of the organo-Pd
complexes or colloidal palladium. Homogeneous catalysts are in
the same phase as the educts and products (except for phase-trans-
fer catalysts). Thus, complete removal from the reaction mixture to
tical synthesis. Furthermore, many homogeneous systems usually
consist of the metal and the ligands in a particular stoichiometric ra-
tio. Thus, the removal of the ligands makes the purification of the
product even more expensive [18].
Immobilization of the catalytic system on solid supports can
mitigate these problems, as it allows the straightforward removal
of the catalyst from the reaction system. The known heterogeneous
Pd-systems include supported Pd-complexes [19–28], supported
Pd-nanoparticles, unsupported Pd-nanoparticles, and encapsulated
Pd-complexes (for examples see [29] and references therein).
Although heterogenization of the catalysts is advantageous, the
majority of heterogeneous catalysts require harsher conditions
than well-tuned homogeneous catalysts, which often results in
leaching of Pd from the support [30,31]. Considering the strict
guidelines placed on Pd levels in products destined for human con-
sumption [32], there is a clear need for methods to prevent Pd
incorporation into organic and pharmaceutical products. These
methods can be broadly divided into scavenging and preventative
approaches, and the latter being dominated by the use of
non-leaching heterogeneous catalysts.
In this study, we report the preparation and testing of a novel
heterogeneous Pd-catalyst. In this system, the Pd-complex is
attached to functionalized silica gel particles via an organic ligand.
Modified silica is by far the most utilized support for immobilizing
Pd-complexes [30,33]. Unfunctionalized mesoporous silica as
⇑
Corresponding author.
E-mail address: khinast@tugraz.at (J.G. Khinast).
0021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.10.013

H. Gruber-Woelfler et al. / Journal of Catalysis 286 (2012) 30–40
31
a NORAN Vantage system with Si(Li)-detector (10 mm2, PlexUs
0.65 m window, and energy resolution 132 eV). Transmission elec-
Pd(OAc)2
N
l
N
N
tron microscopy (TEM) micrographs were acquired using a Tecnai 12
microscope (FEI Company, 120 kV, LaB6 Cathode). FTIR spectra were re-
corded on a Bruker VERTEX 70 using a DLaTGS detector and an ATR
unit: MVP Pro Star, Diamond crystal, resolution: 4 cmꢁ1, 16 scans,
and 4000–600 cmꢁ1. For the IR spectra of the functionalized silica par-
N
O
O
O
O
N
N
n
O
n
O
ticles, a DRIFT unit was used (EasyDiff, spectral resolution: 4 cmꢁ1
,
n
time: 3 min). ICP/OES was carried out using a Varian Vista-MPX CCD
Simultaneous ICP/OES device, the used Pd-standard was 10.000 ppm
Pd in 5% HNO3. Elemental analysis: The amounts of C, H, N, and S were
analyzed using an Elementar Vario El III device with helium as carrier
gas. The nitrogen physisorption analysis was carried out on a Micromer-
itics ASAP 2010. Inner surface area was determined based on the BET
method. The pore size distribution was determined using the BJH
method applied to the adsorption isotherms.
HS
S
S
Pd(OAc)2
radical starter
[SiO]x
[SiO]x
[SiO]x
Scheme 1. Two-step preparation of the heterogeneous Pd-complex Pd(OAc)2–
BOX–MPSG (n = 7).
2.2. Materials
11-Bromo-1-undecene (95%), 1-bromoundecane (98%), diethyl
methylmalonate (P99%), NaH (60% dispersion in mineral oil), thio-
nylchloride (99%), AIBN (2,20-azobis(2-methylpropionitrile) (98%),
ACHN (1,10-azobis(cyclohexanecarbonitrile) (98%), 3-mercaptopro-
pyl silica gel (=MPSG, 1.2 mmol/g loading, 200–400 mesh, 60 Å pore
size, 500 m2/g surface area), Pd(OAc)2 (98%), potassium carbonate
(99%), phenylboronic acid (P97%), 1-bromo-4-methylbenzene
(98%), 4-bromobenzoyl chloride (P97%), 1-bromo-4-(trifluoro-
methyl)benzene (97%), 4-bromoacetophenone (98%), 3-aminopropyl
triethoxysilane, n-dodecane (98%), NMP (1-methyl-2-pyrrolidinone,
99%), DMAc (N,N-dimethylacetamide, 99.8%), 3-aminopropyl silica
support was used by Ying et al. [34]. Shimizu et al. [35], Bedford et
al. [36], and Crudden et al. [23,30,37] have shown that mesoporous
silica can be modified with commercially available thiol ligands,
resulting in recoverable catalysts for cross-coupling reactions, such
as the Suzuki–Miyaura and Heck reactions. However, detailed
mechanistic studies have shown that the catalytic reactions using
these supports likely occur in solution via leached Pd [23,38]. For
example, the group of Jones et al. have reported that catalysis with
Pd-SH-SBA-15 is solely associated with leached metal [39]. Also
the group of Crudden et al. found that small amounts of active
Pd leach from the surface when using mesoporous molecular
sieves modified by mercaptopropyl trimethoxysilane [40].
We present a Pd-catalyst that is attached to a tethered bis(oxaz-
oline) (=BOX) ligand that is covalently bonded to functionalized sil-
ica gel. In particular, we report here the covalent immobilization
of 2,20-(1-methyl-11-dodecenylidene)bis(4,5-dihydrooxazole) on
3-mercaptopropyl-functionalized silica gel (MPSG). The metalation
of the immobilized BOX ligand with Pd(OAc)2 (Scheme 1) yielded
the stable and active catalyst Pd(OAc)2–BOX–MPSG. Similar immo-
bilized BOX–M complexes (M = Pd, Cu, Zn, Mg, Rh) have been re-
ported for a large variety of enantioselective reactions [41,42].
We studied the performance of the Pd(OAc)2–BOX–MPSG system
using Suzuki–Miyaura reactions as a test system. Our results show
that the complexation of the Pd-compound to the ligand leads to a
significant decrease of metal leaching into the solution in compar-
ison with other heterogeneous Pd systems.
gel (1 mmol/g loading, 40–63 lm particle size, 60 Å pore size,
550 m2/g surface area), N-methyl morpholine (P98%), pyridine
(99%), poly(vinylpyridine) (2% cross-linked with divinyl benzene),
QuadraPureÒ TU (400–600
lm particle size), and (3-mercaptopro-
pyl)trimethoxysilane (95%) were purchased from Aldrich. The silica
gel 60 was purchased from Machery Nagel (0.063–0.2 mm). DMF
(>99.8%), CH2Cl2 (P99.5%), toluene (P99.5%), and ethanol (99.8%)
were purchased from Roth. To prepare degassed and anhydrous
THF, CH2Cl2 and toluene, the solvents were purified and degassed
using the manual solvent purification system Pure Solve (Innovative
Technology). All other chemicalsand solvents were used withoutfur-
ther treatment. All manipulations of air- and/or moisture-sensitive
materials were performed under argon using a dual vacuum/argon
line and standard Schlenk techniques.
2.3. Preparation of the heterogeneous Pd-catalyst
2. Materials and methods
2.3.1. Synthesis of the BOX ligand
The tethered bis(oxazoline)ligands, i.e., 2,20-(1-methyl-11-dode-
cenylidene)bis(4,5-dihydrooxazole) = BOX and 2,20-(tridecane-2,2-
diyl)bis(4,5-dihydrooxazole) = BOX 2 were synthesized following
the synthetic pathway described by Hara et al. [43] 2,20-(1-
methyl-11-dodecenylidene)bis(4,5-dihydrooxazole), BOX: 1H d
(CDCl3) 5.80 (1H,m), 4.96 (2H, 2 dd, J = 1.63, 17.29 Hz), 4.28 (4H, t,
J = 9.28 Hz), 3.87 (4H, t, J = 9.52 Hz), 2.01 (2H,m), 1.90 (2H,m),
1.48 (3H,s), 2.24 (14H,m) 13C d (CDCl3) 169.5, 139.4, 114.2, 67.9,
54.4, 42, 36.6, 33.9, 29.9–29.0 (6Cs), 24.3, 21.5, 11,2 IR: mmax
(cmꢁ1) 3076, 2925, 2854, 1655, 1481, 1458, 1351, 1265, 1194,
1139, 1086, 982, 954, 913, 735. 2,20-(tridecane-2,2-diyl)bis(4,5-
dihydrooxazole), BOX 2: 1H d (CDCl3) 4.20 (4H, t, J = 10 Hz), 3.87
(4H, t, J = 9.52 Hz), 2.01 (2H,m), 1.90 (2H,m), 1.48 (3H,s), 2.24
(14H,m) 13C d (CDCl3) 169.5, 67.9, 54.4, 42, 36.6, 33.9, 29.9–29.0
2.1. General
1H and 13C NMR spectra were recorded on a Varian Inova 500 MHz
spectrometer at 499.82 MHz and 125.69 MHz, respectively. Chemi-
cal shifts are listed in delta (d/ppm), employing the residual protio
solvent resonance as internal standard (d = 7.24 ppm in 1H spectra
and 77.0 ppm in 13C spectra for CHCl3). GC analysis was carried out
on a Perkin–Elmer Clarus 500 using an Optima-5 MS capillary col-
umn (Machery-Nagel, 30.0 m ꢀ 320
lm ꢀ 0.25
lm nominal) with
a flame ionization detector and nitrogen as carrier gas. GC/MS
analysis: CI mass spectra were recorded with an Agilent 5973N
mass selective detector on a Agilent Technologies 6890N gas
chromatograph, column: HP-5MS (Agilent 19091S-433), Capillary:
30.0 m ꢀ 250
l
m ꢀ 0.25
l
m nominal, carrier gas: helium. SEM/
(6Cs), 24.3, 21.5, 14.3, 11.2 IR:
1351, 1265, 1194, 1137, 1091, 982, 955, 914, 735.
m
max (cmꢁ1) 2925, 2854, 1655, 1458,
EDX was carried out with a FEI Quanta 600 FEG-ESEM system and

32
H. Gruber-Woelfler et al. / Journal of Catalysis 286 (2012) 30–40
2.3.2. Immobilization of the BOX ligand on 3-mercaptopropyl-
functionalized silica gel
4-Bromobenzamide 3-propyltriethoxysilane was prepared also
by following the synthesis presented by Crudden et al. [23]: 2 g
(9.1 mmol) of 4-bromobenzoyl chloride was mixed with 1 ml
(9.1 mmol) N-methyl morpholine in 50 ml THF. This mixture was
cooled to ꢁ10 °C, and then 3-aminopropyl triethoxysilane (2 ml,
9 mmol) in 10 ml THF was added dropwise to the cooled solution.
After warming this mixture to room temperature, the solid was re-
moved by filtration, and the filtrate was concentrated under vac-
uum yielding a beige solid. This solid was mixed under argon
with pyridine (0.9 ml, 11.5 mmol) and was added to a suspension
of silica gel 60 (1 g) in anhydrous toluene (3 ml). After refluxing
this mixture for 24 h, the suspension was filtered, and the resulting
solid was Soxhlet extracted in CH2Cl2 for 24 h. After drying under
vacuum at room temperature, we obtained 1.2 g of a beige powder
(BrPhCONH-propyl silica gel II). Elemental analysis: 1.28%
(0.91 mmol) N, 14.64% (12.2 mmol) C, and 1.54% (15.4 mmol) H.
For the actual three-phase tests, 4-bromoacetophenone (86 mg,
0.5 mmol, 1 mol equiv.), phenylboronic acid (134 mg, 1.1 mmol,
2.22 mol equiv.), and K2CO3 (173 mg, 1.25 mmol, 2.5 mol equiv.)
were stirred in 5 ml toluene in the presence of n-dodecane (72 mg,
(=MPSG, according to Sigma 1.2 mmol thiol groups per gram sil-
ica gel): The silica gel and catalytic amounts of AIBN (2,20-azobis(2-
methylpropionitrile) or ACHN (1,10-azobis(cyclohexanecarbonitri-
le) were added to a solution of the BOX ligand (1 mol equivalent
(=equiv.) referring to the thiol groups) in degassed anhydrous
CH2Cl2 (or toluene when ACHN was used) under argon. The sus-
pension was continuously stirred and heated under reflux for
18 h. To remove the non-immobilized ligand, the particles were fil-
tered, washed, and sonicated five times for 1 min in CH2Cl2/tolu-
ene. Then the material was dried in vacuum at room
temperature. After that, the amount of immobilized ligand was
determined by elemental analysis.
2.3.3. Metalation
The silica gel particles including the immobilized BOX-ligand
were stirred at room temperature in a solution of Pd(OAc)2
(1.5 mol equiv. to the immobilized BOX-ligand) in CH2Cl2 for 2 h.
In order to assure that no Pd is adsorbed on the surface, the parti-
cles were repeatedly sonicated for several minutes and washed
with CH2Cl2. The Pd-amount was determined using ICP/OES.
0.5 mmol, 1 mol equiv.), BrPhCONH-propyl silicagel
I or II
(500 mg), and Pd(OAc)2-BOX-MPSG (2 mol%) at 115 °C for 24 h.
Then the supernatant was analyzed by GC, and the solid was sepa-
rated by filtration, washed with ethanol, and further extracted with
CH2Cl2. The solid was hydrolyzed with KOH in ethanol/water (1.7 g
KOH in 10 ml of EtOH plus 5 ml of H2O) at 90 °C for 3 days. The
resulting mixture was neutralized with aqueous HCl, extracted with
CH2Cl2, followed by ethyl acetate, concentrated, and the resulting
mixture was analyzed by 1H and 13C NMR. Further experimentswere
carried out with 4.4 mol equiv. boronic acid (and 4.3 mol equiv.
K2CO3) in order to ensure that the reaction is not starved of boronic
acid. The amounts of the other materials were kept constant.
2.4. General procedure for the Suzuki–Miyaura reactions
2.4.1. Suzuki–Miyaura reactions
The aryl halide (1 mol equiv.) and the phenylboronic acid
(1.7 mol equiv.) were dissolved in the reaction solvents (see Tables
4 and 7). Then the base (K2CO3, 1 mol equiv. to the boronic acid),
the catalyst (0.025–5 mol% with respect to the aryl halide), and n-
dodecane (internal standard, 1 mol equiv.) were added and the reac-
tion mixture was heated to the appropriate reaction temperature
(Tables 4 and 7). The reaction progress was monitored using GC
analysis. The products were characterized using GC/MS analysis.
In the case of the homogeneous (hom in Table 4) reaction,
5 mol% Pd(OAc)2 (with respect to aryl halide) was used and
1 mol equiv. (referring to the catalyst) BOX ligand.
3. Results and discussion
3.1. Synthesis and characterization of the heterogeneous Pd-catalyst
For the ‘‘in situ’’ experiments the BOX-MPSG particles were
added to the reactants (amount according to 5 mol% BOX referring
to the aryl halide) and 5 mol% Pd(OAc)2 (with respect to aryl ha-
lide) was added to the reaction mixture.
Analogous to the procedure reported by Hara et al., who
grafted the tethered BOX ligand (2,20-(1-methyl-11-dodecenylid-
ene)bis(4,5-dihydrooxazole)) on a H-terminated Si(111)-wafer fol-
lowed by complexation with Pd(OAc)2 [43], the preparation of our
novel catalytic system was also carried out in two steps (Scheme 1).
First, the BOX ligand was immobilized on 3-mercaptopropyl-
functionalized silicagel (MPSG)particles usingAIBN or ACHN as rad-
ical initiator [20,44–46]. Then the immobilized ligand (BOX-MPSG)
was metalated with Pd(OAc)2 at room temperature in CH2Cl2 to give
the complex Pd(OAc)2-BOX-MPSG. The attachment of the ligand and
the metal was confirmed using ATR-IR spectroscopy. The spectrum
of BOX-MPSG (Fig. 1b) showed the characteristic C@N stretching
at ꢂ1660 cmꢁ1 and the C–H stretching modes of the alkyl chain at
ꢂ2940 and 2850 cmꢁ1, similar to the non-immobilized BOX ligand
(Fig. 1a). Furthermore, the spectra of Pd-BOX-MPSG showed three
additional peaks at 1490, 1510, and 1600 cmꢁ1 (Fig. 1c). According
to the IR data reported by Hara et al., we associated these peaks to
the signals of the acetato ligands on the Pd atom.
2.4.2. Recycling studies
Recycling studies were carried out using the general Suzuki–
Miyaura reactions procedure. After the reaction, the catalyst was
separated from the reaction mixture by filtration. The catalyst
was washed with CH2Cl2 and H2O to remove residual salts and
organics. Before reuse, the catalyst was dried under vacuum at
room temperature, and the amounts of necessary reactants were
recalculated for a 0.75 mol% Pd reaction.
The same cleaning and drying procedure was carried out before
the used catalyst was analyzed by nitrogen physisorption.
2.4.3. Three-phase tests
The three-phase tests were carried out in analogy to the proce-
dure published by Crudden et al. [23,37]. For the immobilization of
the aryl halide, 4-bromobenzoyl chloride (219 mg, 1 mmol) was
dissolved in dry THF in a round-bottom flask. 3-Aminopropyl-func-
tionalized silica gel (1 g functional silica gel with 1 mmol amino
The heterogeneous catalyst was additionally characterized
using elemental analysis (Table 1), SEM/EDX (Table 2), and ICP/
OES (Table 3). The results indicate that the mol ratio of S/BOX is
about 4:1, and the ratio of S/Pd is 4:1. Thus, the BOX/Pd ratio is
1:1, and we assume that all of the Pd forms a complex with the
BOX ligand.1 The nitrogen physisorption analysis of the material
groups) and pyridine (80 ll, 1 mmol) were added under an argon
atmosphere. The resulting mixture was stirred at 40 °C for 12 h,
then filtered and washed with HCl, followed by K2CO3, distilled
water, ethanol, and large amounts of CH2Cl2 to give (after drying
under vacuum at room temperature) 0.98 g of BrPhCONH-propyl
silica gel I. Elemental analysis: 1.64% (1.17 mmol) N, 15.04%
(12.5 mmol) C, and 1.70% (17.0 mmol) H.
1
We tried to quantify the amount of available thiol groups with a spectrophoto-
metric method based on a thiol-exchange reaction with 2,20-pyridyl thione according
to Lindner et al. [47]. However, the presence of Pd influenced the UV-signal which led
to implausible results.

H. Gruber-Woelfler et al. / Journal of Catalysis 286 (2012) 30–40
33
3.2. Suzuki reactions
The activity of Pd(OAc)2-BOX-MPSG was tested for Suzuki–
Miyaura reactions (Scheme 2) under various conditions (Table 4).
The initial experiments were carried out with the homogeneous
catalytic system, i.e., the BOX ligand and Pd(OAc)2, using two
different aryl halides (1-bromo-4-methylbenzene and 1-bromo-
4-(trifluoromethyl)benzene) as substrates. Bases, solvents, and
reaction temperatures were varied in the experiments (Table 4,
entry 1–5 and 8). The results show that the nature of the initial
precatalyst, the reaction solvent, and the base [48] is critical for
the outcome of the reaction. For both aryl halides, the highest
yields were obtained with toluene and K2CO3 at a reaction temper-
ature of 80 °C. Furthermore, the results indicate that the catalyst
also works in aqueous systems at 25 °C. However, the yield was
very low (entry 3 in Table 4).
In addition to the homogeneous catalytic system, we also stud-
ied the activity of the ‘‘in situ’’ immobilized catalyst (as reported,
e.g., by Kuriyama et al. [49] as well as by Yang et al. [50]). For these
experiments, the BOX ligand immobilized on MPSG (=BOX-MPSG)
was used. The metalation of the ligand was carried out during the
Suzuki reaction, i.e., in the presence of the aryl halides, the boronic
acid, and the base. The yields using this catalytic system (Table 4,
entry 6 and 9 and Fig. 2) are definitely higher than the yields ob-
tained with the homogeneous catalyst.
Finally, we obtained yields >90% for the Suzuki reactions when
the heterogeneous catalytic system was used (Table 4). For these
experiments, Pd(OAc)2 was attached to the immobilized ligand in
a separate step as described in Scheme 1 before the catalytic sys-
tem was used for the Suzuki reactions. With this heterogeneous
system, 90% yield could be achieved within 4 h; after 24 h yields
up to 97% were obtained (Table 4, entries 12, 13 and 20) when
1-bromo-4-methylbenzene was the aryl halide substrate. In addi-
tion to that, almost quantitative yields could be achieved after
4 h using 4-bromoacetophenone as starting material (Table 4, en-
try 21). Comparison of these data with the results reported for
other supported Pd-catalysts used for Suzuki couplings2 shows that
the activity of our novel system is in the same range.
A reason for the significantly lower yields of the reactions with
the homogeneous catalytic system might be the multiple sites of
interaction of the palladium(II)acetate with the BOX-ligand. We as-
sume that the low performance can be attributed to the interac-
tions of the Pd-complex with the terminal double bond of the
BOX ligand. Since in the heterogeneous case the BOX ligand is
covalently attached to the solid support, Pd(OAc)2 can only (beside
the possible complexation of Pd with the remaining SH-groups of
the solid support) form a complex with the nitrogen atoms of the
bis(oxazoline) functionality. This assumption is also supported by
the fact that experiments with the analogous ligand without termi-
nal double bond (i.e., 2,20-(tridecane-2,2-diyl)bis(4,5-dihydrooxaz-
ole) = BOX 2, entry 22, Table 4) showed higher yields than the
experiments with the BOX ligand. Another reason for the low activ-
ity of the homogeneous catalyst can also be self-deactivation of
Pd(OAc)2, as it was reported, e.g., by Richardson and Jones [39]
and many others.
Fig. 1. Parts of the IR spectra of (a) the BOX ligand (b) BOX–MPSG and (c) Pd(OAc)2–
BOX–MPSG. Shown are the regions of the C–H stretching (2950–2830 cmꢁ1) and the
peaks in the range of 1670–1430 cmꢁ1
.
Table 1
Results of the elemental analysis.
Sample
Element
Wt%
Loading (mmol/g)
3-Mercaptopropyl silica gel (MPSG)
N
S
N
S
N
S
<0.1
4.6
1.2
4.6
1.2
4.5
1.4
0.8
1.4
0.8
1.4
BOX-MPSG
Pd(OAc)2-BOX-MPSG
Table 2
Results of the SEM/EDX measurements of Pd(OAc)2-
BOX-MPSG.
Element
Wt%
At%
Si
S
Pd
55.7
10.6
33.7
75.4
12.6
12.0
Table 3
Results of the ICP/OES measurements.
The lower activity of the in situ immobilized catalyst in compar-
ison with the heterogeneous catalyst can be attributed to the fact
that the thiol groups that are not attached to the BOX ligand act
as an effective and selective poison [4,23,39] of the homogeneous
Sample
Element
Loading
(mg/g)
Loading
(mmol/g)
3-Mercaptopropyl silica gel (MPSG)
Pd(OAc)2-BOX-MPSG
Si
S
Si
S
325.0
40.8
310.0
39
11.6
1.3
11.1
1.2
Pd
34
0.3
2
(e.g., Crudden et al., Suzuki coupling of bromoacetophenone with phenyl boronic
acid using 1% SBA-SH-Pd, 97% yield after 8 h at 80 °C in H2O [37], Corma et al., 100 mg
zeolite-supported PdCl2 for Suzuki coupling of 1 mmol PhBr with 1.5 mmol phenyl
boronic acid, up to 89% after 24 h at 50 °C in ethanol [51], Richardson and Jones,
Suzuki coupling of bromoacetophenone with phenyl boronic acid using 1% Pd-SH-
SBA-15, 94% conversion after 6 h in DMF, 80% conversion after 6 h in H2O [39]).
showed that the BET surface area of the material is 192 m2/g, and the
average pore diameter is 47 Å (Table 6).

34
H. Gruber-Woelfler et al. / Journal of Catalysis 286 (2012) 30–40
Table 4
Results of the Suzuki–Miyaura reactions.
Entry
R
Cat[a]
Cat (mol%)
Solvent
Base
T (°C)
Yield[b] (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
CH3
CF3
CF3
CF3
CF3
Pd(OAc)2
hom
hom
hom
hom
in situ
het
hom
in situ
het
het
het
het
het
het
het
5
5
5
5
5
5
5
5
5
5
2
2
Toluene
H2O
H2O/MeOH
THF
K2CO3
NaOH
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
80
80
25
60
80
60
80
80
80
80
60
100
115
80
80
80
80
80
80
80
80
80
80
21
–
4
21
25
43
85
26
56
90
27
94
97
28
30
37
56
56
79
90
97
54
41
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
CF3
CF3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
COCH3
CH3
CH3
2
0.025
0.075
0.125
0.25
0.75
1.25
2
2
5
5
het
het
het
het
het
BOX 2, Pd(OAc)2
MPSG, Pd(OAc)2
BOX = 2,20-(1-methyl-11-dodecenylidene)bis(4,5-dihydrooxazole), BOX 2 = 2,20-(tridecane-2,2-diyl)bis(4,5-dihydrooxazole). [a] hom = Homogeneous system with BOX and
Pd(OAc)2, in situ = BOX immobilized on 3-mercaptopropyl silica gel, metalation with Pd(OAc)2 during the Suzuki reactions (in situ), het = heterogeneous system, immobi-
lization of BOX and metalation ex-situ. [b] Yields determined after 4 h with GC using n-dodecane as internal standard.
Pd-species that is added to the starting reaction mixture. This poi-
soning effect is also illustrated by the results of entry 23 in Table 4.
When homogeneous Pd(OAc)2 was used in the presence of 3-mer-
captopropyl silica gel particles only 41% yield could be achieved
(entry 23, Table 4). This result, i.e., the lower yield obtained with
the Pd(OAc)2-MPSG system in comparison with the novel
Pd(OAc)2–BOX–MPSG system is very important, since it clearly
shows the positive influence of the BOX ligand on the reaction.
Interestingly, although the thiol groups should lead to a deactiva-
tion of Pd(OAc)2, the yield of the Pd(OAc)2-MPSG experiment was
higher than the yield that was obtained with homogeneous
Pd(OAc)2 (without BOX, entry 1 in Table 4). As mentioned before,
a reason for that might be the self-deactivation of the homoge-
neous Pd(OAc)2 [39].
Finally, we tested the influence of the temperature and the cat-
alyst loading on the yield obtained with Pd(OAc)2–BOX–MPSG. The
results of entries 10, 11, 12, and 13 in Table 4 demonstrate that an
increasing reaction temperature led to increasing yields (when tol-
uene was used as solvent). In addition, the data presented in Fig. 3
and Table 4 show that the yield increased also with increasing cat-
alyst loading. The lowest tested catalyst loading was 0.025 mol%.
However, only low yields (28% after 4 h, Table 4, entry 14) were ob-
served with this small catalyst amount.
Br
B(OH)2
0.025-5 mol%
Pd cat
R
+
base
R
3.3. Heterogeneity of the reaction
Scheme 2. Suzuki–Miyaura couplings (R = CH3, CF3, COCH3).
For the synthesis of active pharmaceutical intermediates, metal
contamination of the product is a serious issue [15,32]. Leaching of
the catalyst into the product would implicate a time-consuming
and costly cleaning step, which would make the whole process
more expensive [52]. Furthermore, leaching of the active metal
also leads to a decreasing turn over frequency (TOF) of the catalyst.
Thus, the true catalytically active species (i.e., homogeneous or
heterogeneous) has been studied systematically for many catalytic
systems over the last decades. Several methods have been devel-
oped to distinguish between soluble vs. insoluble catalysts, and de-
tailed reviews have been presented by Phan et al. [17] and Lamblin
et al. [15]. Some of these methods were also used by us in order to
investigate whether the solid catalyst is really heterogeneous or
not. The results of these experiments will be presented in the fol-
lowing. However, already the data presented in Table 4 indicate
that the catalytic system does not behave like homogeneous
Pd(OAc)2. Our results show that an increasing amount of catalyst
leads to higher activity (Fig. 3). These data contrast the findings re-
ported, e.g., by Richardson and Jones that a higher loading of
homogeneous Pd(OAc)2 leads to lower conversions because of
self-deactivation of the homogeneous catalyst [39].
Fig. 2. Comparison of the obtained yields of the Suzuki reactions using the
homogeneous, the in situ, and the heterogeneous catalytic system. The reactions
were carried out with 5 mol% catalyst, 1-bromo-4-methylbenzene, phenylboronic
acid, K2CO3, and toluene at 80 °C. Yields were determined with GC using n-
dodecane as internal standard. (Curves are drawn to aid the eye.)

H. Gruber-Woelfler et al. / Journal of Catalysis 286 (2012) 30–40
35
In fact, our catalytic system yielded similar results as the data
presented by Richardson and Jones for Heck reactions with Pd-
SH-SBA-15 [39]. They also reported an increasing catalyst activity
with increasing catalyst loading, as well as higher activities at
higher reaction temperatures. This temperature dependency was
also found by us (Table 4, entries 10–13). The decrease of activity
with decreasing temperature was attributed by Richardson and
Jones either due to effects inhibiting the reaction on the surface
or to temperature effects on the leaching of the immobilized Pd
from supported thiols. We believe that the increasing activity with
increasing temperature could be attributed to a standard Arrhenius
behavior of a rate-determining reaction step.
The increasing activity at higher catalyst loadings was ex-
plained by Richardson and Jones by one of the two possibilities:
(i) Pd(0) formed during the Heck cycle cannot aggregate due to
its binding to an immobilized organic surface, and thus, increased
activity is simply associated with more metal present or (ii) the
amount of leached palladium from the surface is low, and thus,
the actual concentration of active catalytic species in solution is
below levels that promote self-deactivation. In order to evaluate
the leaching behavior of the presented Pd(OAc)2-BOX-MPSG sys-
tem and to compare it with other systems reported in the litera-
ture, several tests were carried out by us, which will be
presented in the following sections.
Fig. 3. Influence of the catalyst loading on the yield of the Suzuki–Miyaura reaction
of 1-bromo-4-methylbenzene and phenylboronic acid carried out with Pd(OAc)2–
BOX–MPSG in toluene at 80 °C. Yields were determined with GC using n-dodecane
as internal standard. (Curves are drawn to aid the eye.)
3.3.1. Hot filtration test
During this test, the catalytically active particles were removed
from the reaction by filtration after 50 min using a hot frit, and the
filtrate was monitored for continued activity. The results show
(Fig. 4) that after removal of the catalyst particles, the reaction
did not proceed, indicating that no catalytically active Pd remained
in the filtrate. This fact is in contrast to the findings reported by
Richardson and Jones [39], Crudden et al. [23,37], as well as by
Davis et al. [38], who also used silica-supported Pd-catalysts for
cross-coupling reactions. Their systems continued to react after
the catalyst was removed by filtration. However, it is known that
the hot filtration test alone cannot prove that the reaction occurs
heterogeneously, because the leached Pd can re-deposit so quickly
that it may not be detected by a filtration test [15,17]. Furthermore,
disruptions, such as hot filtration, may deactivate an existing active
dissolved metal, leading to the incorrect conclusion that there are
no active homogeneous catalytic species before the filtration
[17,39,53]. Therefore, other tests were carried out to investigate
the heterogeneity of the Pd(OAc)2-BOX-MPSG system.
Fig. 4. Hot filtration test for the Pd(OAc)2–BOX–MPSG catalyzed Suzuki–Miyaura
reaction of 1-bromo-4-methylbenzene and phenylboronic acid with K2CO3, toluene
at 115 °C. In case a, the catalytic active particles were filtered off after 50 min using
a hot frit.
Rebek et al. [55,56], is based on covalently immobilizing one of
the reaction partners, in this case an aryl halide, and examining
its reaction with a soluble reagent, i.e., the boronic acid. The sup-
ported catalyst represents the third phase. If the catalyst remains
immobilized, it will be incapable of accessing the supported aryl
halide. If homogeneous Pd is released, the covalently immobilized
substrate will be converted into the product. To ensure that an ac-
tive catalyst is present, a soluble aryl halide, i.e., 4-bromoacetophe-
none, is added. After reaction, the soluble portion is analyzed after
filtration while the amide is hydrolyzed from the support and iso-
lated as carboxylic acid (Scheme 5).
3.3.2. ICP/OES analysis
Potential Pd leaching into the reaction mixture was also ana-
lyzed with ICP/OES analysis. For this purpose, samples were taken
through a syringe filter (Whatman Puradisc 4, 4 mm diameter,
0.45 lm, PTFE) during standard heterogeneous Suzuki–Miyaura
reactions (reaction temperature: 115 °C), the solvent was evapo-
rated, and the residue was dissolved in HNO3. The analysis of these
samples with ICP/OES showed that the Pd concentration in the
reaction solution was less than the detection limit (i.e., 50 ppb),
which corresponds to less than 0.11% of the starting Pd-amount.
The same result was obtained when the complete reaction mixture
of a standard heterogeneous Suzuki–Miyaura reaction was filtered,
the solvent was evaporated, and the residue was dissolved in
HNO3. Both findings indicate that virtually no Pd leaches from
the surface into the solution.
To carry out the three-phase test, we immobilized 4-bromoben-
zoyl chloride on 3-aminopropyl silica gel according to Scheme 3.
Since Crudden et al. reported that they found free amines after
the immobilization of 4-bromobenzoyl chloride (Scheme 3), we
also prepared another supported reagent (BrPhCONH-propyl silica-
gel II) according to Scheme 4. Both supported reagents were then
used for Suzuki reactions together with 4-bromoacetophenone as
free available aryl halide (Scheme 5). The reactions were carried
out with an excess of boronic acid (2.2 mol equiv. and 4.4 mol
equiv., respectively) in order to make sure that the reaction is
not starved of boronic acid [23]. The analysis of the supernatant
with GC indicated that 4-bromoacetophenone was successfully
converted (up to 90% conversion after 24 h). However, in all exper-
iments, the biphenyl carboxylic acid could not be detected using 1H
and 13C NMR spectroscopy after the amide was hydrolized from
the support. Thus, this test is another important indicator that
3.3.3. Three-phase test
The three-phase test was carried out in analogy to the proce-
dure reported by Crudden et al. [37]. This test is considered to be
a highly useful proof for the presence of catalytically active homo-
geneous metal species [17,37,54]. The test, first developed by

36
H. Gruber-Woelfler et al. / Journal of Catalysis 286 (2012) 30–40
Cl
O
O
N
NH2
pyridine
[SiO]x
+
[SiO]x
H
THF, 40°C,
16 h
Br
Br
Scheme 3. Immobilization of 4-bromobenzoyl chloride on 3-aminopropyl silica gel
to prepare the solid phase reagent I for the three-phase test.
the catalyst is (at least mostly) heterogeneous under the applied
conditions (reactions carried out in toluene, with 2 mol% catalyst,
K2CO3 as the base, reaction temperature: 115 °C). When we carried
out the same experiment using Pd/C as catalyst, which is known to
leach from the surface [23,31], the NMR spectra showed both, 4-
bromo-benzoic acid as well as biphenyl-4-carboxylic acid.
Fig. 5. Effect of catalyst recycling on the yield. All reactions were carried out with
Pd(OAc)2–BOX–MPSG (0.75 mol%), 1-bromo-4-methylbenzene, phenylboronic acid,
K2CO3, and toluene at 80 °C. Yields were determined with GC using n-dodecane as
internal standard.
3.3.4. Catalyst reusability
In addition to the total leaching and overall catalytic activity,
reusability is a critical feature for supported catalysts [23]. There-
fore, we used the same catalytic particles after filtration from the
reaction mixture for several reactions under the same conditions.
In contrast to the other heterogeneity tests, these experiments
were carried out at 80 °C. Since the other tests indicated that high-
er reaction temperatures do not lead to increased Pd-leaching into
the reaction solution, it is assumed that the lower reaction temper-
ature only influences the obtained yields and not the leaching
behavior. The results shown in Fig. 5 and Table 5 indicate that
there is no apparent loss of catalytic activity even after 10 runs,
and the obtained yields of the reusability experiments are within
the experimental deviations.
Table 5
Effect of catalyst recycling on yield and turn over frequency (TOF).
Entry
Yield (%)a
TOF (hꢁ1
)
1st use
45
37
37
38
45
37
38
42
42
14
12
12
12
14
12
12
13
13
Recycle 1
Recycle 2
Recycle 3
Recycle 4
Recycle 5
Recycle 6
Recycle 8
Recycle 10
a
Yields determined after 4 h with GC using n-dodecane as internal standard. All
3.3.5. Catalyst texture
reactions were carried out with 0.75 mol% Pd(OAc)2–BOX–MPSG, 1-bromo-4-
methylbenzene, phenylboronic acid, K2CO3, and toluene at 80 °C.
Additionally to assessing the loss of Pd, it is crucial to determine
whether the catalyst support is affected by the reaction conditions
[23]. Therefore, we analyzed the supported catalyst by nitrogen
adsorption analysis before and after 24 h of reaction (Fig. 6). The
results indicate that although the BET surface area and the pore
volume decrease, the mesoporous structure of the solid support
is maintained. All samples showed the expected type IV isotherms
that are typical for mesoporous materials [57]. Thus, the used solid
support shows a higher stability than other mesoporous silica
Br
Br
OEt
MeN
O
OEt
H
+
EtO Si
OEt
NH2
EtO Si
OEt
N
O
Cl
O
O
EtO
Br
O
N
H
Si
silica-gel
OEt
EtO Si
OEt
O
H
N
pyridine
reflux
Br
O
Scheme 4. Second way of immobilizing 4-bromobenzoyl chloride on silica gel.
H3C
O
HO
O
HO
O
H3C
O
B(OH)2
O
Pd-cat
N
H
+
+
+
+
[SiO]x
K2CO3, toluene,
115°C, then hydrolysis
Br
Ph
Br
Ph
Br
Solution phase
Solid phasere agent
Scheme 5. Reaction scheme of the three-phase test.

H. Gruber-Woelfler et al. / Journal of Catalysis 286 (2012) 30–40
37
Fig. 7. Effect of the solvent on the reaction progress of Suzuki–Miyaura reactions of
1-bromo-4-methylbenzene and phenylboronic acid carried out with 2 mol%
Pd(OAc)2–BOX–MPSG.
Fig. 6. Nitrogen adsorption analysis of the solid support (MPSG), the immobilized
catalyst before (Pd–BOX–MPSG) and after use (Pd–BOX–MPSG after reaction).
a trap of leached Pd (see below and [38,39,52,61,62]), did not lead
to a catalyst poisoning, i.e., in all solvents, the reaction progress
was almost the same as with no poison (Table 7). Thus, we can con-
clude that highly polar solvents and elevated reaction tempera-
tures do not lead to an extended Pd leaching.
Table 6
Effect of catalyst recycling on BET surface area, total pore volume, and average pore
diameter.
Entry
BET
Total pore volume
(cm3/g)
Average pore
diameter (Å)
(m2/g)
3.3.7. Catalyst poisoning
MPSG
365
192
112
0.41
0.23
0.15
45
47
55
Another method that was used to assess the homogeneity/het-
erogeneity of our catalytic system was catalyst poisoning. By addi-
tion of materials that are known to act as catalyst poisons, soluble
reactive Pd can be removed from the solution. The poisons that
were used by us are shown in Fig. 8. The first compound was 3-
mercaptopropyl-functionalized silica gel (MPSG), as thiol-modified
surfaces are known to be very effective Pd-scavengers [4,39,40]. In
addition, poly(4-vinylpyridine) (PVPy, 2% cross-linked) was used
because pyridines are known to bind strongly to Pd(II) [38,63],
and several groups applied insoluble PVPy as a Pd trap to confirm
the presence of leached, soluble Pd [38,39,52,61,62]. Furthermore,
Quadrapure TU, an organic polymer resin (polystyrene) with a
thiourea functionality, and 3-aminopropyl silica gel were studied
as poisons. Since all these compounds are solids, the addition of
them can lead to a decrease of reaction progress by a limited mass
transfer of the heterogeneous Pd-catalyst to the reactants. There-
fore, two homogenous compounds, pyridine and 3-mercaptopro-
pyl-trimethoxysilane, were also utilized as poisons in order to
exclude this effect. The results of the experiments are shown in
Fig. 9 and Table 8.
The results presented in Fig. 9 and Table 8 show that all poisons
except for PVPy and pyridine led to a deactivation of the catalyst,
which would indicate that Pd leaches from the immobilized BOX
ligand and is therefore quenched by the used poisons. The results
also show that the addition of (340 mol equiv.) pyridine did not
lead to a deactivation of the catalyst. Similar results were found
for Heck reactions by the Davis group when homogeneous
Pd(OAc)2 and Pd-SiO2(Cab) were the catalysts [38]. Also, Jones
et al. reported that the addition of pyridine to a Heck reaction car-
ried out with a homogeneous Pd(II)-SCS-N pincer complex had
only little effect on the reaction rate in comparison with a reaction
under normal conditions [62]. These findings would indicate that
the activity of the Pd(OAc)2-BOX-MPSG catalyst arrives from lea-
ched, homogeneous Pd-species and is therefore contradictory to
all the other tests that indicated that the catalytic system does
not show metal leaching, at least no Pd traces could be detected
in the reaction solutions.
Pd(OAc)2-BOX-MPSG
Pd(OAc)2-BOX-MPSG
after 1st use
materials, such as SBA-15 or MCM-41 [4,23,29]. These materials
lost their mesoporous properties after the materials were used at
the applied reaction conditions (80 °C, 20:1 DMF:H2O, K2CO3 in
[4]).
3.3.6. Effect of the solvent on the activity and stability of the catalyst
Several groups reported that the nature of the reaction solvent
had a dramatic effect on catalytic activity and the extent of Pd
leaching [38,51,58,59]. For example, Ji et al. reported that the cat-
alytic activity of supported Pd-catalysts for Heck reactions was
much lower in toluene than in DMF. In order to investigate
whether the presented catalyst shows a similar behavior, addi-
tional experiments were carried out with NMP (1-methyl-2-
pyrrolidinone), DMAc (N,N-dimethylacetamide), and DMF (N,
N-dimethylformamide) as solvents. The results are presented in
Fig. 7 and Table 7.
As can be seen, the polar solvents do not lead to an increase of
activity of the catalytic system. In fact, the obtained yields for 4-
methylbiphenyl were lower when DMAc, NMP, and DMF were
used even when the reactions were carried out at higher tempera-
tures. Additionally, significant by-product was noticed. In all
experiments with toluene, this by-product formation was not
observed.
Although the influence of the solvent on the catalytic activity is
not completely understood [38], it has been shown by other groups
that the Pd leaching is more extensive in highly polar solvents (as
DMAc, NMP, and DMF) than in non-polar toluene [38,51,60].
However, our catalytic system shows a different behavior: ICP/
OES analysis of samples taken through a syringe filter from the
reactions that were carried out in DMAc (150 °C), NMP (160 °C),
and DMF (150 °C) after 2 h and 24 h did not show any Pd. Further-
more, the addition of 100 mol equiv. poly(4-vinylpyridine) (PVPy,
2% cross-linked), a common Pd-scavenger that is known to act as
Another result that supports the heterogeneous catalytic mech-
anism is the fact that when 100 mol equiv. PVPy/Pd were used as a

38
H. Gruber-Woelfler et al. / Journal of Catalysis 286 (2012) 30–40
Table 7
Effect of the reaction solvent on reaction progress and catalyst poisoning with PVPy of Suzuki–Miyaura reactions of 1-bromo-4-methylbenzene and
phenylboronic acid carried out with 2 mol% Pd(OAc)2-BOX-MPSG using K2CO3 as base.
Solvent
Reaction temperature
Yield after 2 h (%)
Yield after 4 h (%)
Yield after 24 h (%)
Toluene
DMAc
NMP
Toluene
DMAc
NMP
DMF
110
110
110
110
150
160
150
110
150
160
150
43
15
–
95
17
–
96
23
–
43
56
21
34
43
56
23
22
95
68
55
45
65
60
59
47
98
75
62
75
88
70
64
74
Toluene, with PVPya
DMAc, with PVPya
NMP, with PVPya
DMF, with PVPya
a
Addition of 100 mol equiv. PVPy (referring to Pd) after 2 h.
OMe
Si
SH
NH2
MeO
SH
[SiO]x
[SiO]x
OMe
3-Mercaptopropyl-silicagel
(MPSG)
3-Aminopropyl-silicagel
(NH2-Prop-SG)
3-Mercaptopropyl-trimethoxysilane
(MPTMOS)
*
S
*
n
N
NH2
H
PS
Quadrapure TU
N
N
Polyvinylpyridine
(PVPy)
Pyridine
Fig. 8. Names and chemical structures of the used poisons (PS = polystyrene).
Table 8
Influence and used amounts of the selective catalyst poisons on the reaction progress
of Suzuki–Miyaura reactions of 1-bromo-4-methylbenzene and phenylboronic acid
carried out with 2 mol% Pd(OAc)2–BOX–MPSG in toluene at 115 °C using K2CO3 as
base. Poisons were added after 2 h.
Poison
Mol equiv.
to Pd
Conversion after
2 h (%)
Conversion after
24 h (%)
No
PVPy
Pyridine
MPSG
MPSG
MPTMOS
NH2-prop-SG
Quadrupure TU
–
100
340
40
400
40
43
43
42
45
45
46
43
42
96
88
93
45
42
50
44
44
40
45
experiment, 1 g PVPy was added to a reaction with 25 mg of func-
tionalized silica gel including 0.8 mg Pd) can also result in a lower
catalytic activity simply by decreasing the accessibility of the reac-
tants to the catalytic sites in the thick slurry multiphase media,
reducing diffusion constants and thus creating mass-transfer lim-
ited conditions.
In summary, the finding that the activity of our novel catalyst is
decreased by selective Pd-poisons is contrary to the results of the
other tests that indicated that the activity of the catalytic system
can be attributed to a real heterogeneous catalytic system.
Fig. 9. Influence of selective catalyst poisons on the reaction progress of Suzuki–
Miyaura reactions of 1-bromo-4-methylbenzene and phenylboronic acid carried
out with 2 mol% Pd(OAc)2-BOX-MPSG in toluene at 115 °C using K2CO3 as base.
poison, the reaction progress was almost the same as with no poi-
son (as already discussed above). However, it is known that due to
the equilibrium on–off coordination of the Pd with the pyridine
sites and the non-porous nature of the resin, a large excess of pyr-
idine units are required to over-coordinate Pd [39]. Thus, other
groups reported that the addition of higher amounts of PVPy was
necessary in order to quench the reaction (e.g., Richardson and
Jones added up to 350 equiv. PVPy to Pd-SH-SBA 15 [39], and Davis
et al. added up to 700 equiv. [38]). Indeed, when we added 500 mol
equiv. PVPy, the reaction also stopped to proceed. However, we be-
lieve that the addition of such large amounts of solids (in this
3.3.8. Transmission electron microscopy (TEM)
Finally, we analyzed the catalytic system using TEM. The bright-
field TEM images (Fig. 10) show that after the catalyst was used

H. Gruber-Woelfler et al. / Journal of Catalysis 286 (2012) 30–40
39
Fig. 10. Bright-field TEM pictures of (a) 3-mercaptopropyl silica gel (MPSG) (b) Pd(OAc)2–BOX–MPSG before use and (c) Pd(OAc)2–BOX–MPSG after 1st use.
[3] S. Kotha, K. Lahiri, D. Kashinath, Tetrahedron 58 (2002) 9633.
once in toluene, nanoparticles with a size of ꢂ5 nm were formed
[4] B.W. Glasspoole, J.D. Webb, C.M. Crudden, Journal of Catalysis 265 (2009) 148.
[5] J. Magano, J.R. Dunetz, Chem. Rev. 111 (2011) 2177.
[6] V.F. Slagt, A.ü.H.M. de Vries, J.G. de Vries, R.M. Kellogg, Org. Process Res. Dev.
14 (2009) 30.
(Fig. 10c).
In combination with our poisoning studies, this finding stands
in contrast to the conclusion by Richardson and Jones [39] who re-
ported that poisoning by SH-SBA-15 implies that the active soluble
catalytic species is not from nanoparticle surfaces but rather from
molecular and dimeric Pd. This theory is also supported by the
work of other groups [13,64–66]. One possible explanation for
the behavior of our catalytic system, i.e., poisoning by mercapto-
functionalized silica gel and nanoparticle formation, could be that
small amounts of Pd leach from the immobilized BOX ligand dur-
ing the reaction. If a poison is present, this Pd is over-coordinated
by the poison. Otherwise, Pd nanoparticles are formed, and these
particles are trapped by the solid support. Thus, the support would
act as the source of the active component and at the same time as
sink for Pd, and therefore, constituting a virtually leaching-free
system. However, further studies are needed to confirm this theory
and to fully understand the reaction mechanism of the investigated
catalytic system.
[7] M. Hird, G.M. Gray, K.L. Toyne, Mol. Cryst. Liq. Cryst. 206 (1991) 187.
[8] K.C. Nicolaou, P.G. Bulger, D. Sarlah, Angew. Chem. Int. Ed. 44 (2005) 4442.
[9] L. Yin, J. Liebscher, Chem. Rev. 107 (2007) 133.
[10] F.X. Felpin, T. Ayad, S. Mitra, Europ J. Org. Chem. (2006) 2679.
[11] N. Miyaura, A. Suzuki, Chem. Rev. 95 (1995) 2457.
[12] A. Suzuki, Journal of Organometallic Chemistry 653 (2002) 83.
[13] S.P. Stanforth, Tetrahedron 54 (1998) 263.
[14] V.F. Slagt, A.H.M. de Vries, J.G. de Vries, R.M. Kellogg, Org. Process Res. Dev. 14
(2009) 30.
[15] M. Lamblin, L. Nassan-Hardy, J.-C. Hierso, E. Fouquet, F.-X. Felpin, Adv. Synth.
Cat. 352 (2010) 33.
[16] J.P. Corbet, G. Mignani, Chem. Rev. 106 (2006) 2651.
[17] N.T.S. Phan, M. Van Der Sluys, C.W. Jones, Adv. Synth. Cat. 348 (2006) 609.
[18] C. Luo, Y. Zhang, Y. Wang, Journal of Molecular Catalysis A: Chemical 229
(2005) 7.
[19] R. Sayah, K. Glegola, E. Framery, V. Dufaud, Adv. Synth. Catal. 349 (2007) 373.
[20] C. Baleizao, A. Corma, H. Garcia, A. Leyva, Chem. Commun. (2003) 606.
[21] S. Schweizer, J.M. Becht, C. Le Drian, Org. Lett. 9 (2007) 3777.
[22] J. Lemo, K. Heuze, D. Astruc, Org. Lett. 7 (2005) 2253.
[23] J.D. Webb, S. MacQuarrie, K. McEleney, C.M. Crudden, Journal of Catalysis 252
(2007) 97.
[24] H. Qiu, S.M. Sarkar, D.H. Lee, M.J. Jin, Green Chem. 10 (2008) 37.
[25] A. Corma, D. Das, H. Garcia, A. Leyva, Journal of Catalysis 229 (2005) 322.
[26] A. Garcia-Bernabe, C. Tzschucke, W. Bannwarth, R. Haag, Adv. Synth. Catal. 347
(2005) 1389.
[27] M.-J. Jin, D.-H. Lee, Angew. Chem. 122 (2010) 1137.
[28] G. Fan, J. Huang, Z. Li, T. Li, G. Li, Journal of Molecular Catalysis A: Chemical 267
(2007) 34.
[29] D.H. Lee, M. Choi, B.W. Yu, R. Ryoo, A. Taher, S. Hossain, M.J. Jin, Adv. Synth.
Catal. 351 (2009) 2912.
[30] S. MacQuarrie, B. Nohair, J.H. Horton, S. Kaliaguine, C.M. Crudden, The Journal
of Physical Chemistry C 114 (2010) 57.
[31] T.N. Glasnov, S. Findenig, C.O. Kappe, Chem. Eur. J. 15 (2009) 1001.
[32] C.E. Garret, K. Prasad, Adv. Synth. Cat. 346 (2004) 889.
[33] V. Polshettiwar, A. Molnar, Tetrahedron 63 (2007) 6949.
[34] C.P. Mehnert, D.W. Weaver, J.Y. Ying, J. Am. Chem. Soc. 120 (1998) 12289.
[35] K. Shimizu, S. Koizumi, T. Hatamachi, H. Yoshida, S. Komai, T. Kodama, Y.
Kitayama, Journal of Catalysis 228 (2004) 141.
4. Conclusions
A novel catalytic system including a Pd-complex immobilized
via a bis(oxazoline) (=BOX) ligand on 3-mercaptopropyl silica gel
is presented. The catalytic system was tested for heterogeneous
Suzuki–Miyaura reactions of different aryl halides with phenylbo-
ronic acid. Using several approaches to test the heterogeneity of
the catalytic system, we could show that there is virtually no Pd
leaching into the reaction solution under the applied reaction con-
ditions. Furthermore, our results indicate that the catalyst is stable
and can be reused for at least 10 times. Such stable heterogeneous
catalysts would be a key to implement continuous processes in the
preparation of pharmaceutical and fine chemical intermediates.
However, the addition of catalyst poisons led to a quenching of
the catalytic activity, which indicates the presence of leached
homogeneus Pd. Additionally, TEM pictures of the used catalytic
material indicate nanoparticles on the surface. Therefore, further
studies are needed to fully understand the reaction mechanism
of the presented catalytic system.
[36] R.B. Bedford, U.G. Singh, R.I. Walton, R.T. Williams, S.A. Davis, Chem. Mater. 17
(2005) 701.
[37] C.M. Crudden, M. Sateesh, R. Lewis, J. Am. Chem. Soc. 127 (2005) 10045.
[38] Y. Ji, S. Jain, R.J. Davis, J. Phys. Chem. B 109 (2005) 17232.
[39] J.M. Richardson, C.W. Jones, Journal of Catalysis 251 (2007) 80.
[40] C.M. Crudden, K. McEleney, S. MacQuarrie, M. Sateesh, J.D. Webb, Pure Appl.
Chem. 79 (2007) 247.
[41] R.J. Clarke, I.J. Shannon, Chem. Commun. (2001) 1936.
[42] J.M. Fraile, J.I. Garcia, J.A. Mayoral, Coordination Chemistry Reviews 252 (2008)
624.
Acknowledgments
[43] K. Hara, S. Tayama, H. Kano, T. Masuda, S. Takakusagi, T. Kondo, K. Uosaki, M.
Sawamura, Chem. Commun. (2007) 4280.
[44] C. Baleizao, B. Gigante, D. Das, M. Alvaro, H. Garcia, A. Corma, J. Catal. 223
(2004) 106.
[45] C. Baleizao, A. Corma, H. Garcia, A. Leyva, J. Org. Chem 69 (2004) 439.
[46] W.J. Sommer, K. Yu, J.S. Sears, Y. Ji, X. Zheng, R.J. Davis, C.D. Sherrill, C.W. Jones,
M. Week, Organometallics 24 (2005) 4351.
We kindly acknowledge the financial support of the Austrian
Science Foundation (Project No. 19410 and Elise Richter Project
No. V171-N19).
[47] R. Nogueira, M. Lämmerhofer, N.M. Maier, W. Lindner, Analytica Chimica Acta
533 (2005) 179.
[48] C. Amatore, A. Jutand, G. Le Duc, Chemistry A European Journal 17 (2011)
2492.
[49] M. Kuriyama, R. Shimazawa, R. Shirai, Tetrahedron 63 (2007) 9393.
[50] C. Yang, H.M. Lee, S.P. Nolan, Org. Lett. 3 (2001) 1511.
[51] A. Corma, H. Garcia, A. Leyva, Applied Catalysis A: General 236 (2002) 179.
References
[1] H.U. Blaser, A. Indolese, A. Schnyder, H. Steiner, M. Studer, J. Mol. Cat. A: Chem.
173 (2001) 3.
[2] G.A. Grasa, M.S. Viciu, J. Huang, C. Zhang, M.L. Trudell, S.P. Nolan, Organomet.
21 (2002) 2866.

40
H. Gruber-Woelfler et al. / Journal of Catalysis 286 (2012) 30–40
[52] M. Weck, C.W. Jones, Inorg. Chem. 46 (2007) 1865.
[53] B.H. Lipshutz, S. Tasler, W. Chrisman, B. Spliethoff, B. Tesche, J. Org. Chem. 68
(2002) 1177.
[54] I.W. Davies, L. Matty, D.L. Hughes, P.J. Reider, J. Am. Chem. Soc. 123 (2001)
10139.
[55] J. Rebek, F. Gavina, J. Am. Chem. Soc. 96 (1974) 7112.
[56] J. Rebek, D. Brown, S. Zimmerman, J. Am. Chem. Soc. 97 (1975) 454.
[57] S.J. Gregg, K.S. Sing, Adsorption, Surface Area and Porosity, Academic Press,
London, 1982.
[60] M. Dams, L. Drijkoningen, B. Pauwels, G. Van Tendeloo, D.E. De Vos, P.A. Jacobs,
Journal of Catalysis209 (2002) 225.
[61] J.S. Chen, A.N. Vasiliev, A.P. Panarello, J.G. Khinast, Appl. Cat. A: General 325
(2007) 76.
[62] K. Yu, W. Sommer, J. Richardson, M. Weck, C. Jones, Adv. Synth. Catal. 347
(2005) 161.
[63] F.R. Hartley, Coordination Chemistry Reviews 35 (1981) 143.
[64] A.F. Schmidt, V.V. Smirnov, Journal of Molecular Catalysis A: Chemical 203
(2003) 75.
[58] K. Köhler, R.G. Heidenreich, J.G.E. Krauter, J. Pietsch, Chemistry - A European
Journal 8 (2002) 622.
[65] A. Alimardanov, L. de Vondervoort, A.H. de Vries, J.G. deVries, Adv. Synth. Catal.
346 (2004) 1812.
[59] F. Zhao, B.M. Bhanage, M. Shirai, M. Arai, Journal of Molecular Catalysis A:
Chemical 142 (1999) 383.
[66] S.G. Fiddy, J. Evans, T. Neisius, M.A. Newton, N. Tsoureas, A.A.D. Tulloch, A.A.
Danopoulos, Chem. Eur. J. 13 (2007) 3652.