Evaluation of SILP-Pd catalysts for Heck reactions in a microfluidics-based high throughput flow reactor

Article abstract of DOI:10.1016/j.molcata.2014.08.031

Heck reaction of aryl iodides and methyl acrylate was carried out in an X-Cube reactor in the presence of supported catalysts. Palladium was immobilised by different methods on silica with covalently grafted ionic liquid moieties. Activity and selectivity of the SILP-Pd (supported ionic liquid phase) catalysts were found to depend greatly on the conditions (such as solvent and additives) of immobilisation. By the proper choice of the methodology used during heterogenisation, a selective catalyst, showing stable performance for hours on stream, was obtained.

Full text of DOI:10.1016/j.molcata.2014.08.031

Journal of Molecular Catalysis A: Chemical 395 (2014) 364–372
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Evaluation of SILP-Pd catalysts for Heck reactions in a
microfluidics-based high throughput flow reactor
Béla Urbána, Dávid Srankób, György Sáfránc, László Ürged, Ferenc Darvasd, József Bakosa,
Rita Skoda-Földesa,∗
a University of Pannonia, Institute of Chemistry, Department of Organic Chemistry, PO Box 158, H-8201 Veszprém, Hungary
b Hungarian Academy of Sciences, Centre for Energy Research, Department of Surface Chemistry and Catalysis, PO Box 49, H-1525 Budapest, 114, Hungary
c Hungarian Academy of Sciences, Research Institute for Technical Physics and Materials Science, PO Box 49, H-1525 Budapest, Hungary
d ThalesNano Inc., Záhony u. 7, H-1031 Budapest, Hungary
a r t i c l e i n f o
a b s t r a c t
Heck reaction of aryl iodides and methyl acrylate was carried out in an X-CubeTM reactor in the presence
of supported catalysts. Palladium was immobilised by different methods on silica with covalently grafted
ionic liquid moieties. Activity and selectivity of the SILP-Pd (supported ionic liquid phase) catalysts were
found to depend greatly on the conditions (such as solvent and additives) of immobilisation. By the proper
choice of the methodology used during heterogenisation, a selective catalyst, showing stable performance
for hours on stream, was obtained.
Article history:
Received 30 January 2014
Received in revised form 4 August 2014
Accepted 20 August 2014
Available online 28 August 2014
Keywords:
© 2014 Elsevier B.V. All rights reserved.
Supported ionic liquid
Flow chemistry
Aryl iodides
Palladium catalyst
1. Introduction
Heck reaction using a completely automated microreactor system
[12].
Transition metal catalysed coupling reactions serve as powerful
tools for the construction of new C C bonds [1] and can be used
efficiently in the synthesis of pharmaceuticals [2], fine chemicals
[3] and natural products [4].
The application of heterogeneous catalysis in conjunction
with microreactor technology can provide a cleaner methodology
for coupling reactions that makes purification steps, to remove
catalyst from the product, unnecessary. As a consequence, sev-
eral immobilised palladium catalysts were developed and used
in microflow systems. Monolithic polyionic supports contain-
ing imidazolium- [13–15] or ammonium [16] ions, derivatised
with Pd(0) nanoparticles were incorporated inside flow reac-
tors. Other examples involve palladium species immobilised by
bidentate N,S ligand-functionalised organic monoliths [17] or
commercially available Pd/C [18–20]. However, when using Pd/C
under microflow conditions, either significant catalyst deactivation
was found in the subsequent experiments [19] or the forma-
tion of a considerable amount of side products was observed
[20].
Another significant improvement in the methodology of the
Heck reaction was based on the findings of Kaufmann [21] and
Herrmann [22], revealing that the coupling could be catalysed
by simple palladium salts, such as PdCl2 or Pd(OAc)2, without
additional ligands if it was carried out in phosphonium- [21] or
ammonium [22] ionic liquids instead of conventional organic sol-
vents. The catalysts showed increased activity in ionic liquids
compared to that in DMF, especially at high temperatures. The
One of the most prominent representatives of these processes
with industrial relevance is the Heck–Mizoroki reaction, involv-
ing the formation of substituted alkenes by the coupling of two
sp carbon centres [5]. It is a very reliable reaction and many
non-conventional methodologies have been explored in order to
increase its efficiency [6]. One of the possibilities for the intensifica-
tion of this reaction is the application of the microflow technique [7]
that offers several advantages over conventional methods includ-
ing a better control of reaction parameters, superior heat and mass
transfer rates, possibility of fast optimisation and ease of scale-up
[8]. Heck reactions were carried out in continuous flow microreac-
tors under homogeneous conditions in organic solvents [9], in ionic
liquids [10] or under segmented flow conditions [11]. Buchwald
and Jensen demonstrated the rapid optimisation and scale-up of a
∗
Corresponding author. Tel.: +00 36 88 624719; fax: +00 36 88 624469.
E-mail address: skodane@almos.uni-pannon.hu (R. Skoda-Földes).
http://dx.doi.org/10.1016/j.molcata.2014.08.031
1381-1169/© 2014 Elsevier B.V. All rights reserved.

B. Urbán et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 364–372
365
2. Experimental
efficiency of these systems can be attributed to the stabilising
effect of the ionic liquid on the Pd(0) species, obtained from
the Pd(II) precursors. In some reactions, the formation of Pd-
nanoparticles, stabilised by ammonium ions, was also revealed.
[23]. In imidazolium based ionic liquids with halide counter ions,
Pd(OAc)2 was found to be converted to catalytically active Pd-N-
heterocyclic carbene complexes [24]. Beside simple ammonium-,
phosphonium-, and imidazolium salts, a wide range of task-specific
ionic liquids were prepared and used efficiently as reaction medium
[25,26]. Ionic liquids can also be used to confine the catalyt-
ically active species on a solid support (SILP: supported ionic
liquid phase). Pd(OAc)2 was immobilised in the pores of silica gel
[27] and reversed phase silica gel [28] with an adsorbed imidaz-
olium salt to give excellent results with TONs up to 1600,000 in
dodecane and water as solvent, respectively. Heck reaction was
carried out under solvent-free conditions with a Pd catalyst sup-
ported on 1,1,3,3-tetramethylguanidinium-modified SBA-15. The
catalyst could be recycled five times without loss of activity [29].
A palladium—N-heterocyclic carbene complex was grafted on sil-
ica via the trialkoxysilyl side chain of the imidazolium moiety
resulting in an efficient catalyst with TONs up to 36,600 [30]. Aero-
gels containing palladium metal nanoparticles were prepared by a
sol–gel method starting form Pd(OAc)2, dissolved in imidazolium
[31,32] or ammonium [33] ionic liquids and (EtO)4Si. A detailed
investigation of palladium catalysts obtained with polystyrene-
grafted imidazolium ionic liquids revealed that initial Pd–NHC
supported complexes are only precatalysts and the active species
were shown to be released into the medium as soluble nanoclusters
[34].
2.1. Preparation of the supported ionic liquid phase (2a) for
CAT-1–CAT-4
A mixture of 1.66 mmol (400 ␮l) 3-chloropropyl-triethoxysilane
6 ml toluene and 1 g silica (Kieselgel 60 (0.040–0.063 mm), Merck,
pre-treated by heating for 6 h at 250 ◦C) was heated under argon
at 120 ◦C for 24 h. Then the solid material was filtered, washed
with 2 ml toluene, 2 ml dichloromethane and 2 ml methanol
and dried in vacuo. (1a, Anal. Found: C, 5.15; H, 1.01.) Then
it was suspended in 5 ml toluene and, under stirring, 5 mmol
(397 ␮l) 1-methylimidazole was added dropwise in an inert
atmosphere. The mixture was heated at 80 ◦C for 24 h. Then
the solid material was filtered, washed with 2 ml toluene, 2 ml
dichloromethane and 2 ml methanol and dried in vacuo for 8 h
(2a, Anal. Found: C, 7.66; H, 1.18; N, 1.58.). Ionic liquid loading:
0.74 mmol/g (based on the weight increase of the support mea-
sured after heating the material to constant weight at 150 ◦C in
vacuo)
2.2. Preparation of the supported ionic liquid phase (2b) for
CAT-5 and CAT-6
A mixture of 1.66 mmol (400 ␮l) 3-chloropropyl-triethoxysilane
6 ml MeOH and 1 g silica (Kieselgel 60 (0.040–0.063 mm), Merck,
pre-treated by heating for 6 h at 250 ◦C) was heated under argon
at 80 ◦C for 24 h. Then the solid material was filtered, washed with
2 ml toluene, 2 ml dichloromethane and 2 ml methanol and dried
in vacuo. Then it was suspended in 5 ml MeOH and, under stir-
ring, 5 mmol (397 ␮l) 1-methylimidazole was added dropwise in
an inert atmosphere. The mixture was heated at 80 ◦C for 24 h.
Then the solid material was filtered, washed with 2 ml toluene,
2 ml dichloromethane and 2 ml methanol and dried in vacuo for 8 h.
Ionic liquid loading: 0.36 mmol/g (based on the weight increase of
the support measured after heating the material to constant weight
at 150 ◦C in vacuo).
Although the SILP-Pd catalysts offer the possibility to be used in
fixed bed reactors, there are only a few examples for Heck reactions
carried out under flow conditions. As an example, a continuous
flow Heck coupling, based on catalysts immobilised onto mono-
lithic SILP material, was performed in near-critical EtOH as solvent
[13]. In spite of the fact that a great variety of silica supported SILP-
palladium catalysts, using diverse palladium precursors, have been
applied efficiently in Heck reactions [25,26], they were tested only
in batch reactors.
The main goal of the present work was the evaluation of some
SILP-palladium catalysts, prepared by known techniques [34–36],
in Heck reactions carried out in a microfluidics-based flow reac-
tor (XCubeTM) [37]. X-CubeTM is ideal for a rapid optimisation
of reaction conditions [38] or a quick test of several catalysts
[39]. It can be equipped with pre-packed, replaceable stainless
steel cartridges (CatCartTM) that hold the immobilised catalyst.
A number of palladium-catalysed reactions, such as Sonogashira
reaction [39], aminocarbonylation [40] and double carbonylation
[38] had been carried out efficiently in the flow reactor. It should
be mentioned however, that the Pd/C catalysed Heck reaction of 4-
iodobenzonitrile and butyl acrylate led to a considerable amount of
dehalogenated and homocoupling products in the X-Cube reactor
[20].
It should be mentioned that in case of supported catalysts, the
use of prolonged reaction times is often beneficial for an effective
catalyst recirculation because of the Pd dissolution/reprecipitation
mechanism reported by Köhler [41] during a detailed investigation
of Pd/C catalysed Heck coupling. According to this report, after the
total conversion of the substrate, the palladium-content of the solu-
tion was found to decrease due to the precipitation of catalytically
active palladium nanoparticles to the support. This ensured effec-
tive catalyst recirculation with retained activity. In the XCubeTM
reactor, residence time of the reaction mixture is limited due to the
capacity of the pumps (min. flow rate 0.1 ml/min) and the length
of the column. So the main goal of the present work was to find
a catalyst that showed stable performance for several hours on
stream.
2.3. Preparation of the supported ionic liquid phase (2c) for
CAT-7
A mixture of 0.83 mmol (200 ␮l) 3-chloropropyl-triethoxysilane
6 ml toluene and 1 g silica (Kieselgel 60 (0.040–0.063 mm), Merck,
pre-treated by heating for 6 h at 250 ◦C) was heated under argon
at 70 ◦C for 24 h. Then the solid material was filtered, washed with
2 ml toluene, 2 ml dichloromethane and 2 ml methanol and dried
in vacuo. Then it was suspended in 5 ml toluene and, under stir-
ring, 2.5 mmol (198 ␮l) 1-methylimidazole was added dropwise in
an inert atmosphere. The mixture was heated at 70 ◦C for 24 h.
Then the solid material was filtered, washed with 2 ml toluene,
2 ml dichloromethane and 2 ml methanol and dried in vacuo for 8 h.
Ionic liquid loading: 0.21 mmol/g (based on the weight increase of
the support measured after heating the material to constant weight
at 150 ◦C in vacuo).
2.4. Preparation of supported catalyst CAT-1 [35]
1.16 mmol (330 mg) [BMIM][PF6] was dissolved in 20 ml THF,
then 0.58 mmol (130 mg) Pd(OAc)2 was added and the mixture
was stirred under argon until homogeneous. Two grams of 2a
was added and the mixture was stirred for 40 min. The solvent
was removed in vacuo and the catalyst was dried at 60 ◦C for 5 h.
Palladium content of the catalyst: 0.23 mmol/g (determined by
ICP).

366
B. Urbán et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 364–372
2.5. Preparation of supported catalyst CAT-2 [34]
Reaction mixtures were analysed by gas chromatography
(Hewlett Packard 5890) and GC–MS (Hewlett Packard 5971A
GC–MSD, HP-1 column). Conversions and selectivities of the reac-
tions were determined by GC.
Under inert conditions, 3.12 mmol (350 mg) KOtBu was added
with stirring to a mixture containing 2 g of 2a and 20 ml THF. The
resulting mixture was stirred for 10 min, and after the addition
of 1.56 mmol (350 mg) Pd(OAc)2, it was stirred at 50 ◦C for 3 h.
Then the solid material was filtered, washed with 20 ml THF, 20 ml
methanol and 20 ml dichloromethane and dried in vacuo for 5 h.
Palladium content of the catalyst: 0.25 mmol/g (determined by ICP).
The products 5a–5g and 6b–6d were identified on the basis of
their MS spectra.
(5a) MS(m/z/rel. int.): 162(M+)/175; 131/100; 103/55; 77/21;
51/10.
(5b) MS(m/z/rel. int.): 207(M+)/39; 176/100; 130/46; 118/23;
102/73; 76/44; 51/42.
(5c) MS(m/z/rel. int.): (240/62; 242/62) (M+); 211/100; 209/100;
183/22; 181/22; 130/12; 102/88; 75/27; 51/17.
(5d) MS(m/z/rel. int.): (230/7; 232/5) (M+); 197/33; 195/100;
173/6; 171/15; 136/18; 99/13.
2.6. Preparation of supported catalysts CAT-3 and CAT-5[36]
Under inert conditions, 2 g of 2a (for CAT-3) or 2b (for CAT-
5) was stirred in 20 ml ethanol for 10 min. 1.56 mmol (350 mg)
Pd(OAc)2 was added and the mixture was stirred at room tem-
perature for 24 h. Then the solid material was filtered, washed
with 20 ml ethanol and 20 ml diethylether and dried in vacuo for
5 h. Palladium content of the catalyst: 0.66 mmol/g (CAT-3) and
0.63 mmol/g (CAT-5) (determined by ICP).
(5e) MS(m/z/rel. int.): 190(M+)/57; 175/18; 159/100; 131/21;
115/50; 91/46; 77/21; 51/36.
(5f) MS(m/z/rel. int.): 192(M+)/58; 161/100; 133/33; 118/16;
89/36; 77/20; 63/30; 51/18.
(5g) MS(m/z/rel. int.): 178(M+)/50; 147/100; 119/36; 91/43;
65/43; 63/29.
(6b) MS(m/z/rel. int.): 244(M+)/68; 214/20; 152/100; 151/85;
139/43; 63/35; 51/25.
2.7. Preparation of supported catalysts CAT-4, CAT-6 and CAT-7
(6c) MS(m/z/rel. int.): (310/50; 312/100; 314/50) (M+); 152/69;
126/6; 76/19.
Under inert conditions, 3.12 mmol (350 mg) KOtBu was added
with stirring to a mixture containing 2 g of 2a (for CAT-4), 2b (for
CAT-6) or 2c (for CAT-7) and 20 ml ethanol. The resulting mixture
was stirred for 10 min, and after the addition of 1.56 mmol (350 mg)
Pd(OAc)2, it was stirred at room temperature for 24 h. Then the
solid material was filtered, washed with 20 ml ethanol and 20 ml
diethylether and dried in vacuo for 5 h. Palladium content of the cat-
alyst: 0.61 mmol/g (CAT-4), 0.70 mmol/g (CAT-6) and 0.63 mmol/g
(CAT-7) (determined by ICP).
(6d) MS(m/z/rel. int.): (290/75; 292/100; 294/42; 296/10) (M+);
259/4; 257/11; 255/11; 224/7; 222/41; 220/61; 184/9; 150/16;
110/14.
2.10. Characterisation of CAT-3 and CAT-4
Surface concentrations of Pd were determined by X-ray photo-
electron spectroscopy (XPS) performed by a KRATOS XSAM 800 XPS
machine equipped with an atmospheric reaction chamber. Al K␣
characteristic X-ray line, 40 eV pass energy and FAT mode were
applied for recording the XPS lines of Pd 3d, C 1s, O 1s and Si 2p.
Si 2p binding energy at 103.3 eV was used as reference for charge
compensation.
Transmission electron microscope (TEM) investigations were
carried out by a JEOL 3010 high resolution TEM operating at 300 kV,
with a point resolution of 0.17 nm. The microscope was equipped
with a GATAN Tridiem energy filter used for electron energy loss
spectroscopy (EELS) elemental mapping. The samples were sus-
pended in ethanol and drop-dried on carbon-coated microgrids for
the measurements of the microstructure of the catalyst particles
and their distribution over the support.
2.8. General procedure for the Heck reaction (batch mode)
A
mixture of 10 mg supported catalyst, 0.2 mmol (22 ␮l)
iodobenzene, 0.4 mmol (36 ␮l) methyl acrylate, 0.68 mmol (100 ␮l)
Et3N and 1 ml DMF was stirred in an inert atmosphere at 100 ◦C for
2 h. The solution was removed by a syringe and the catalyst was
reused without purification.
Reaction mixtures were analysed by gas chromatography
(Hewlett Packard 5890) and GC–MS (Hewlett Packard 5971A
GC–MSD, HP-1 column). Conversions and selectivities of the reac-
tions were determined by GC.
2.9. General procedure for the Heck reaction in the flow reactor
The reactions were carried out in the microfluidics-based flow
reactor (XCubeTM) in continuous flow mode. The catalysts were
placed into 70 × 8 mm CatCartTM cartridges (approx. 0.5 g, see
Table 1). The catalyst bed was washed continuously with the DMF
solutions of aryl halide (0.05 M or 0.5 M), methyl acrylate (0.1 M or
1 M), and Et3N (0.17 M or 1.7 M) with 0.1 ml/min flow rate (res-
idence time on the catalyst bed: 6 min). The temperature was
adjusted at 100 ◦C.
3. Results and discussion
3.1. Preparation of the supported catalysts
Palladium was immobilised by different methods on silica with
covalently grafted ionic liquid moieties. The support was prepared
in two steps (Scheme 1): chloropropylated silica (1) was obtained
by the reaction of silica and 3-chloropropyl-triethoxysilane, and
was used as an alkylating agent to form the imidazolium moieties
of 2 in the reaction with 1-methylimidazole. Elemental analysis
and the decrease in the BET surface of the support (to 400.9 m2/g
from the original 466.7 m2/g in case of 2a) proved the incorporation
of the ionic liquid into the silica material. Another support with
lower ionic liquid loading (2b) was obtained using MeOH as solvent.
The preparation of SILPs with higher IL loadings was attempted at
higher temperatures in solvents with higher boiling point (such as
cymene). However, in these cases the removal of solvent was found
to be extremely difficult and IL loadings could not be increased.
Table 1
Palladium-content of the supported catalysts.
Catalyst
Support
Pd content (mmol/g)
CAT-1
CAT-2
CAT-3
CAT-4
CAT-5
CAT-6
CAT-7
2a
2a
2a
2a
2b
2b
2c
0.23
0.25
0.66
0.61
0.63
0.70
0.63

B. Urbán et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 364–372
367
Pd(OAc)2
CAT-1
CAT-2
[BMIM][PF6]
CH3
N
+
CH3
N
N
Cl-
Cl
Pd(OAc)2
silica
+
EtO
EtO Si
EtO
N
toluene
toluene
or MeOH
or MeOH
KOtBu, THF
Si
O O
Pd(OAc)2
EtOH
Cl
CAT-3 (on 2a)
CAT-5 (on 2b)
OEt
Si
O O
OEt
1a, 1b (toluene)
1c (MeOH)
CAT-4 (on 2a)
CAT-6 (on 2b)
CAT-7 (on 2c)
Pd(OAc)2
2a, 2b (toluene)
2c (MeOH)
KOtBu, EtOH
Scheme 1. Preparation of supported palladium catalysts.
R
R
[Pd]
CO2CH3
+
I
CO2CH3
Et3N, DMF
3a R=H
4
5a R=H
3b R=4-NO2
5b R=4-NO2
3c
3d
3e
R=3-Br
5c
5d
5e
R=3-Br
R=2-Cl, 4-Cl
R=2-Cl, 4-Cl
R=3-CH3, 4-CH3
R=3-CH3, 4-CH3
3f R=4-OCH3
3g R=4-OH
5f R=4-OCH3
5g R=4-OH
Scheme 2. Heck reaction of aryl iodides (3a–g) with methyl acrylate (4).
The first catalyst (CAT-1) was prepared by impregnation of
2a with a solution of Pd(OAc)2 in [BMIM][PF6], according to the
method used by Hagiwara [28] and Jin [35] for the immobilisa-
tion of Pd on reversed phase silica gel [28] and hierarchical MFI
zeolite [35]. CAT-2 was obtained by the addition of Pd(OAc)2 to
the imidazolium modified support (2a) using KOtBu as the base to
facilitate deprotonation of the C2 H of the imidazolium ring and
formation of Pd-NHC complexes [34]. CAT-3 was prepared from
Pd(OAc)2 and 2a in ethanol to reduce Pd(II) to Pd(0) [36]. Immo-
bilisation of palladium in the presence of KOtBu and in ethanol
as solvent led to CAT-4. CAT-5–CAT-7 were prepared by the same
methods, but by the use of SILPs with lower ionic liquid loading (2b
and 2c). The palladium-content of the catalysts was determined by
ICP (Table 1). Immobilisation was least effective in case of CAT-1
and CAT-2, where barely one third of the original amount of pal-
ladium could be incorporated into the supported catalysts. There
is no significant difference between the palladium-content of the
catalysts obtained from SILP phases with higher (2a) and lower (2b
or 2c) ionic liquid loadings.
carried out in water, that is not miscible with [BMIM][PF6]. Under
our conditions, no conversion of iodobenzene was observed in
water, and similarly disappointing results were obtained in other
solvents (THF, 1,4-dioxane or toluene).
Catalysts CAT-2–CAT-4 and CAT-7 showed good performance
and a relatively small loss of activity upon reuse. Conversion of
iodobenzene in the presence of the catalyst with lower ionic liquid
loading (CAT-7) was found to be only a little less than with CAT-
4. CAT-5 and CAT-6, prepared from the SILP phase obtained by the
immobilisation of the ionic liquid in methanol (2b), exhibited lower
activity in spite of the almost similar palladium-content compared
to CAT-3, CAT-4 and CAT-7. This shows that probably some trace
amounts of solvent may remain in the SILP phase in spite of heating
it in vacuo that alters catalytic activity.
Because of the low residence time of the substrate solution in the
microreactor column, during the present investigation relatively
100
90
80
3.2. Heck reaction of iodobenzene (3a) in batch mode
70
CAT-1
CAT-2
CAT-3
CAT-4
CAT-5
CAT-6
CAT-7
60
50
40
30
20
10
0
First, in order to choose the most active catalysts, the perfor-
mance of CAT-1–CAT-7 was compared in the Heck reaction of
iodobenzene (3a, Scheme 2) and methyl acrylate (4) carried out
in batch mode. Although the main goal of the present work was
the evaluation of the catalysts under microflow conditions, their
recyclability in batch reactions was also investigated (Fig. 1).
CAT-1 was shown to be the least active among the catalysts.
Moreover, a great loss of activity was observed upon recirculation. It
should be mentioned that similar catalysts were used by Jin [35] and
Hagiwara [28] in Suzuki and Heck coupling reactions, respectively,
1
2
3
4
5
Run
Fig. 1. Activity of catalysts CAT-1–CAT-7 in the Heck reaction of iodobenzene (3a)
and methyl acrylate (4) (batch, 100 ◦C, 2 h).

368
B. Urbán et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 364–372
Fig. 3. Activity of catalysts CAT-2–CAT-4 in the Heck reaction of iodobenzene (3a)
and methyl acrylate (4) (X-CubeTM, 100 ◦C, flow rate: 0.1 ml/min, concentration of
3a: 0.05 M).
Fig. 2. Conversion of iodobenzene (3a) in the Heck reaction with methyl acrylate
(4) in the presence of CAT-2–CAT-4 (batch, 100 ◦C).
low reaction time (2 h) was used to compare the activity of the
catalysts under batch conditions. Moreover, after 2 h there is only
a little change in the conversion of iodobenzene (Fig. 2) in the
presence of catalysts CAT-2–CAT-4. At the same time it should be
mentioned that for an optimal recycling, longer reaction times are
more favourable to ensure reprecipitation of palladium on the sup-
port [41]. However, optimisation of the catalysts in batch mode was
not the goal of the present work, the efficiency of similar catalysts
in batch reactions had been proved before [34–36].
3.3. Heck reaction of iodobenzene (3a) in continuous mode
Fig. 4. Activity of CAT-4 in the Heck reaction of a more concentrated solution of
iodobenzene (3a) and methyl acrylate (4) (X-CubeTM, 100 ◦C, flow rate: 0.1 ml/min,
concentration of 3a: 0.5 M).
On the basis of the batch experiments, catalysts CAT-2–CAT-4
were chosen for further investigations in the X-CubeTM reactor.
The reactions were carried out in the microfluidics-based flow
reactor in continuous flow mode. CatCartTM cartridges were loaded
with the catalysts and the catalyst bed was washed continuously
with the DMF solutions of the reactants. At the end of the reactor
column, the reaction mixtures were collected and were analysed
by GC.
After an initial good performance, CAT-2 showed a continuously
decreasing activity (Fig. 3). A drop of iodobenzene conversion was
observed in the first hour using CAT-3 but after that the catalyst
was found to give stable results. Both CAT-3 and CAT-4 retained
their activity for several hours on stream (the measurements were
carried out for 9.5 h and 13.5 h, respectively). CAT-4 showed the
higher activity.
Palladium-content of the collected reaction mixtures was deter-
mined by ICP. The average loss of palladium was between 0.33 and
0.46% of the original load per hour (Table 2).
CAT-4 (fresh catalyst loading) was also tested using a more
concentrated solution of reactants (Fig. 4, 0.5 M iodobenzene (3a)
instead of 0.05 M). In this case, the catalyst showed somewhat
less stable performance but approximately same loss of palladium
(Table 2) and higher TOF value (an average of 5.95 h−1 instead of
0.86 h−1 for the 0.05 M iodobenzene solution).
3.4. Heck reaction of aryl iodides (3b–3f) in continuous mode
Catalysts CAT-3 and CAT-4 were tested in the Heck reaction
of other aryl halides (3b–3f) with methyl acrylate (Scheme 2,
Figs. 5, 6 and 8). From time to time, the stable performance of the
catalysts was checked by repeated experiments with iodobenzene
(3a).
CAT-3 was found to convert aryl iodides 3b–3f to Heck prod-
ucts 5b–5f with moderate to good yields (Fig. 5). In the reactions
of 3c and 3d, selective substitution of iodine took place, in accor-
dance with the fact that bromobenzene could not be coupled with
methyl acrylate under the present conditions. Aryl iodides with
electron withdrawing substituents were expected to react more
readily because of the more facile oxidative addition of these sub-
strates to palladium. In fact, conversion of 3b–3d was indeed higher
than that of iodobenzene (3a) (Fig. 6) but Heck coupling was accom-
panied by homocoupling, leading to compounds 6b–6d (Fig. 7) in
considerable amounts. Homocoupling was the prevailing reaction
in case of 3b with a 55% average yield of 6b, possibly due to the
presence of the strongly electron withdrawing nitro group in the
para position. The side reaction was less pronounced with 3c (aver-
age yield of 6c: 11%), and in case of 3d only 5% of the homocoupling
Table 2
Supported catalysts used in continuous mode.
Catalyst
Pd content (mmol/g)
Weighta (mg)
Pd content in CatCart (mg)
Time on stream (h)
Average Pd leachingb (%)
CAT-2c
CAT-3c
CAT-4c
CAT-4d
0.25
0.66
0.61
0.61
544
511
480
530
14.4
35.7
31.0
34.3
3
9.5
13.5
10.5
0.33
0.46
0.42
0.43
a
Weight of supported catalyst in CatCart.
b
c
(total leached palladium (mg)/Pd content in CatCart (mg)/time on stream (h)) × 100.
100 ◦C, flow rate: 0.1 ml/min, concentration of 3a: 0.05 M.
d
100 ◦C, flow rate: 0.1 ml/min, concentration of 3a: 0.5 M.

B. Urbán et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 364–372
369
100
90
80
70
60
50
40
30
20
10
3a
3b
3c
3d
3e
3f
3g
0
1
2
Time on stream (h)
Fig. 5. Yield of Heck products (5a–5f) in the presence of CAT-3 (X-CubeTM, 100 ◦C,
flow rate: 0.1 ml/min, concentration of aryl halides: 0.05 M).
Fig. 8. Heck reaction of aryl iodides (3a–3e) and methyl acrylate (4) in the presence
of CAT-4 (X-CubeTM, 100 ◦C, flow rate: 0.1 ml/min, concentration of aryl halides:
0.05 M).
On the other hand, selective reactions leading to coupling prod-
ucts 5b–5f were observed in all cases (Fig. 8) in the presence of
CAT-4. No products arising from either hydrodehalogenation or
homocoupling could be detected. At the same time no conversion
of 3g was observed under identical conditions.
3.5. Characterisation of CAT-3 and CAT-4
X-ray photoelectron spectroscopy (XPS) analysis was performed
on the as-synthesised as well as on the spent catalysts (after 5 cycles
in batch) (Figs. 9-12, Table 3). The Pd 3d regions of the catalysts are
shown in Figs. 10 and 12. The two bands at around 335.0 eV and
340.3 eV were assigned to Pdo (3d5/2 and 3d3/2, respectively), while
the bands at around 337.0 eV and 342.3 eV were attributed to Pd(II)
(3d5/2 and 3d3/2, respectively). These values are very near the ones
generally reported [42,43]. The surface atomic ratio of Pd, C and Si
were calculated using sensitivity factors given by the manufacturer
and are presented in the table. In order to prove the presence of
Pd(II) content, we in situ oxidised the spent CAT-4 sample after cat-
alytic cycle at 300 ◦C for 60 min by air. After the oxidation the Pd(II)
content increased 20 At% from 10 At%. This phenomenon distinctly
proves the Pd(II) contents of the samples.
Fig. 6. Conversion of aryl iodides (3a–3f) in the Heck reaction with methyl acrylate
(4) in the presence of CAT-3 (X-CubeTM, 100 ◦C, flow rate: 0.1 ml/min, concentration
of aryl halides: 0.05 M).
R
R
6b R=4-NO2
6c R=3-Br
6d R=2-Cl, 4-Cl
Fig. 7. Homocoupling products 6b–6d obtained in the presence of CAT-3.
Similar Pd0 content are typical for all of the samples accord-
ing the analysis of their XPS data. Upon the catalytic reaction, the
Pd/Si ratios as well as the Pd/C ratios were altered. While higher
Pd enrichment on the surface was observed in the sample CAT-4
after the catalytic reaction with higher Pd/C ratio, smaller Pd/Si and
Pd/C ratios were described for the spent CAT-3 catalyst. The distinct
changes of the surface structure can be explained as the effect of
the different preparation methods.
product 6d was formed, probably because of the close proximity
of the 2-Cl moiety. Selective conversion of substrates with electron
donating groups (3e and 3f) into the Heck products was observed
resulting in the formation of 5e and 5f in moderate yields. Even
5g could be produced, although in low yield, starting from 3g. It
should be mentioned that no hydrodehalogenation reaction of the
substrates could be detected.
Fig. 9. Wide scan XPS spectra of Pd catalyst CAT-3 (as synthesized (left), spent (right)).

370
B. Urbán et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 364–372
Fig. 10. Pd 3d XPS spectra of Pd catalyst CAT-3 (as synthesized (left), spent (right)).
Fig. 11. Wide scan XPS spectra of Pd catalyst CAT-4 (as synthesized (left), spent (right)).
Fig. 12. Pd 3d XPS spectra of Pd catalyst CAT-4 (as synthesized (left), spent (right)).
Table 3
XPS Pd 3d5/2 binding energies (eV), Pd(II)/Pd0, Pd/Si and Pd/C surface atomic ratios of as prepared and spent Pd catalysts.
Position of the peak Pd 3d5/2
Pd0 (eV)
Ratio of the atomic concentrations
Sample
Pd(II) (eV)
Pd(II)/Pd0
Pd/Si
Pd/C
CAT-3
CAT-3 (spent)
CAT-4
334.9
335.0
334.9
335.1
335.0
337.0
337.3
337.2
337.4
337.1
1.29 × 10−1
1.16 × 10−1
1.56 × 10−1
1.17 × 10−1
2.75 × 10−1
1.38 × 10−1
9.45 × 10−2
8.93 × 10−2
1.43 × 10−1
8.57 × 10−2
3.15 × 10−1
2.36 × 10−1
2.05 × 10−1
3.14 × 10−1
7.25 × 10−1
CAT-4 (spent)
CAT-4 (spent)a
Oxidised, 300 ◦C, 60 min.
a
TEM images of both CAT-3 and CAT-4 shows the presence of
some non-supported Pd particles (Figs. 13 and 14) in the fresh
catalyst. Some aggregation of Pd nanoparticles can be observed in
TEM images of the used catalysts, with a less even distribution of
nanoparticles in case of spent CAT-3.
The above measurements do not show a great difference in the
structure of CAT-3 and CAT-4.
At the same time, the exact nature of the catalytic system in case
of these heterogeneous catalysts is still not clear. When a supported
palladium source is used, the catalyst may work in a heterogeneous

B. Urbán et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 364–372
371
Fig. 13. TEM images of Pd catalyst CAT-3 (as synthesized (left), spent (right)).
Fig. 14. TEM images of Pd catalyst CAT-4 (as synthesized (left), spent (right)).
manner, or the support may act only as a reservoir for a more active
Acknowledgments
soluble form of Pd [29]. Moreover, the reaction may take place with
the involvement of a number of species, e.g. palladium-carbene
complexes supported on the silica gel, palladium nanoparticles
bound to the surface or leached into the reaction mixture, soluble
palladium complexes formed from the precursors, etc. [34]. From
the difference in the conversion of 3g with very low reactivity, a
more facile oxidative addition may be deduced in case of CAT-3.
This may lead to an enhanced concentration of aryl-palladium(II)
complexes especially in case of the most reactive substrates with
electron withdrawing groups (3b–3d) and as result, an increased
rate of homocoupling [44].
The support of the Hungarian National Science Foundation
(OTKA K105632) and projects TÁMOP 4.2.2.A-11/1/KONV-2012-
0071 and TÁMOP-4.1.1.C-12/1/KONV-2012-0017 realised with the
support of the Hungarian Government and the European Union,
with the co-funding of the European Social Fund, is acknowl-
edged.
References
[1] G.P. McGlacken, I.J.S. Fairlamb, Eur. J. Org. Chem. (2009) 4011–4029.
[2] J. Magano, J.R. Dunetz, Chem. Rev. 111 (2011) 2177–2250.
[3] J.G. de Vries, Can. J. Chem. 79 (2001) 1086–1092.
[4] R. Forke, K.K. Gruner, K.E. Knott, S. Auschill, S. Agarwal, R. Martin, M. Böhl, S.
Richter, G. Tsiavaliaris, R. Fedorov, D.J. Manstein, H.O. Gutzeit, H.J. Knölker, Pure
Appl. Chem. 82 (2010) 1975–1991.
[5] A. de Meijere, F. Diederich (Eds.), Metal-Catalyzed Cross-Coupling Reactions,
vols. 1–2, Wiley-VCH, Weinheim, 2004.
[6] F. Alonso, I.P. Beletskaya, M. Yus, Tetrahedron 61 (2005) 11771–11835.
[7] B.K. Singh, N. Kaval, S. Tomar, E. Van der Eycken, V.S. Parmar, Org. Proc. Res.
Dev. 12 (2008) 468–474.
[8] K.D. Nagy, K.F. Jensen, Chem. Today 29 (2011) 18–21.
[9] M. Viviano, T.N. Glasnov, B. Reichart, G. Tekautz, C.O. Kappe, Org. Proc. Res. Dev.
15 (2011) 858–870.
[10] S. Liu, T. Fukuyama, I. Ryu, Org. Proc. Res. Dev. 8 (2004) 477–481.
[11] B. Ahmed-Omer, D.A. Barrow, T. Wirth, Tetrahedron Lett. 50 (2009) 3352–3355.
[12] J.P. McMullen, M.T. Stone, S.L. Buchwald, K.F. Jensen, Angew. Chem. Int. Ed. 49
(2010) 7076–7080.
4. Conclusions
The results show that reaction conditions (solvents, addi-
tives) applied during the heterogenisation of palladium catalysts
may affect noticeably the effectiveness of the immobilised cat-
alysts considering either activity or selectivity. With the use of
ethanol as solvent and in the presence of a strong base, a cata-
lyst, showing stable performance for several hours on stream and
resulting in selective Heck coupling, could be obtained. CAT-4 was
proved to be a proper catalyst for the use in continuous mode
in the X-CubeTM reactor for a rapid test of Heck reactions of aryl
iodides.

372
B. Urbán et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 364–372
[13] N. Karbass, V. Sans, E. Garcia-Verdugo, M.I. Burguete, S.V. Luis, Chem. Commun.
(2006) 3095–3097.
[14] V. Sans, F. Gelat, N. Karbass, M.I. Burguete, E. García-Verdugo, S.V. Luis, Adv.
Synth. Catal. 352 (2010) 3013–3021.
[29] X. Ma, Y. Zhou, J. Zhang, A. Zhu, T. Jiang, B. Han, Green Chem. 10 (2008) 59–66.
[30] B. Karimi, D. Enders, Org. Lett. 8 (2006) 1237–1240.
[31] K. Anderson, S. Cortin˜as Fernández, C. Hardacre, P.C. Marr, Inorg. Chem. Com-
mun. 7 (2004) 73–76.
[15] N. Nikbin, M. Ladlow, S.V. Ley, Org. Proc. Res. Dev. 11 (2007) 458–462.
[16] K. Mennecke, A. Kirschning, Beilstein J. Org. Chem. 5 (2009), http://dx.doi.
org/10.3762/bjoc.5.21.
[17] R.C. Jones, A.J. Canty, J.A. Deverell, M.G. Gardiner, R.M. Guijt, T. Rodemann, J.A.
Smith, V.A. Tolhurst, Tetrahedron 65 (2009) 7474–7481.
[18] D.A. Snyder, C. Noti, P.H. Seeberger, F. Schael, T. Bieber, G. Rimmel, W. Ehrfeld,
Helv. Chim. Acta 88 (2005) 1–9.
[32] S. Volland, M. Gruit, T. Régnier, L. Viau, O. Lavastre, A. Vioux, New J. Chem. 33
(2009) 2015–2023.
[33] V. Polshettiwar, P. Hesemann, J.J.E. Moreau, Tetrahedron 63 (2007) 6784–6790.
[34] M.I. Burguete, E. García-Verdugo, I. Garcia-Villar, F. Gelat, P. Licence, S.V. Luis,
V. Sans, J. Catal. 269 (2010) 150–160.
[35] M.J. Jin, A. Tather, H.J. Kang, M. Choi, R. Ryoo, Green Chem. 11 (2009) 309–313.
[36] Y. Kume, K. Qiao, D. Tomida, C. Yokoyama, Catal. Commun. 9 (2008) 369–375.
[37] ꢀhttp://www.thalesnano.comꢁ.
[19] X. Fan, M. Gonzalez Manchon, K. Wilson, S. Tennison, A. Kozynchenko, A.A.
Lapkin, P.K. Plucinski, J. Catal. 267 (2009) 114–120.
[38] J. Balogh, Á. Kuik, L. Ürge, F. Darvas, J. Bakos, R. Skoda-Földes, J. Mol. Catal. A:
Chem. 302 (2009) 76–79.
[20] T.N. Glasnov, S. Findenig, C.O. Kappe, Chem. Eur. J. 15 (2009) 1001–1011.
[21] D. Kaufmann, M. Nouroozian, H. Henze, Synlett (1996) 1091–1092.
[22] V.P.W. Böhm, W.A. Herrmann, Chem. Eur. J. 6 (2000) 1017–1025.
[23] V. Calò, A. Nacci, A. Monopoli, J. Organomet. Chem. 690 (2005) 5458–5466.
[24] L. Xu, W. Chen, J. Xiao, Organometallics 19 (2000) 1123–1127.
[25] F. Bellina, C. Chiappe, Molecules 15 (2010) 2211–2245.
[26] C.I.M. Santos, J.F.B. Barata, M.A.F. Faustino, C. Lodeiro, M.G.P.M.S. Neves, RSC
Adv. 3 (2013) 19219–19238.
[27] H. Hagiwara, Y. Sugawara, K. Isobe, T. Hoshi, T. Suzuki, Org. Lett. 6 (2004)
2325–2328.
[28] H. Hagiwara, Y. Sugawara, T. Hoshi, T. Suzuki, Chem. Commun. (2005)
2942–2944.
[39] B. Borcsek, G. Bene, G. Szirbik, G. Dormán, R. Jones, L. Ürge, F. Darvas, ARKIVOC
v (2012) 186–195.
[40] C. Csajági, B. Borcsek, K. Niesz, I. Kovács, Z. Székelyhidi, Z. Bajkó, L. Ürge, F.
Darvas, Org. Lett. 10 (2008) 1589–1592.
[41] R.G. Heidenreich, J.G.E. Krauter, J. Pietsch, K. Köhler, J. Mol. Catal. A: Chem.
182–183 (2002) 499–509.
[42] M. Brun, A. Berthet, J.C. Bertolini, J. Electron, Spectrosc. Relat. Phenom. 104
(1999) 55–60.
[43] C. Pavia, E. Ballerini, L.A. Bivona, F. Giacalone, C. Aprile, L. Vaccaro, M. Grut-
tadauria, Adv. Synth. Catal. 335 (2013) 2007–2018.
[44] B.S. Lane, D. Sames, Org. Lett. 6 (2004) 2897–2900.