Pd-loaded superparamagnetic mesoporous NiFe2O4 as a highly active and magnetically separable catalyst for Suzuki and Heck reactions

Article abstract of DOI:10.1016/j.molcata.2010.12.009

A superparamagnetic solid catalyst has been successfully synthesized by loading palladium nanoparticles into the pore network of a mesoporous NiFe ₂O₄ support (Pd/NF300), which was used as a magnetically separable and highly active catalyst for Suzuki and Heck C-C coupling reactions. Various techniques were employed to characterize the synthesized NiFe ₂O₄ supports and Pd-loaded catalysts. The Pd/NF300 catalyst showed high activity for both Suzuki and Heck reactions even under a very low Pd using amount (0.08 mol% Pd based on aryl halide). Moreover, the catalyst could be magnetically separated, recycled, and showed a very slight reduction in catalytic activity after five cycles. A synergetic catalytic effect between the well dispersed Pd⁰ and the basic mesoporous structure of magnetic supports has been proposed to understand its greatly enhanced catalytic activities.

Full text of DOI:10.1016/j.molcata.2010.12.009

Journal of Molecular Catalysis A: Chemical 336 (2011) 51–57
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Pd-loaded superparamagnetic mesoporous NiFe2O4 as a highly active and
magnetically separable catalyst for Suzuki and Heck reactions
Zhe Gaoa, Yingjun Fengb, Fangming Cuia, Zile Huaa, Jian Zhoua, Yan Zhua, Jianlin Shia,∗
a State Key Laboratory of High Performance Ceramics and Superfine Microstructure, Shanghai Institute of Ceramics, Chinese Academy of Sciences, No. 1295, Dingxi Rd., Shanghai
200050, PR China
b Key Laboratory for Ultrafine Materials of Ministry of Education, School of Materials Science and Engineering, East China University of Science and Technology, Shanghai 200237, PR
China
a r t i c l e i n f o
a b s t r a c t
Article history:
A superparamagnetic solid catalyst has been successfully synthesized by loading palladium nanoparticles
into the pore network of a mesoporous NiFe2O4 support (Pd/NF300), which was used as a magnetically
separable and highly active catalyst for Suzuki and Heck C–C coupling reactions. Various techniques
were employed to characterize the synthesized NiFe2O4 supports and Pd-loaded catalysts. The Pd/NF300
catalyst showed high activity for both Suzuki and Heck reactions even under a very low Pd using amount
(0.08 mol% Pd based on aryl halide). Moreover, the catalyst could be magnetically separated, recycled, and
showed a very slight reduction in catalytic activity after five cycles. A synergetic catalytic effect between
the well dispersed Pd0 and the basic mesoporous structure of magnetic supports has been proposed to
understand its greatly enhanced catalytic activities.
Received 29 October 2010
Received in revised form
15 December 2010
Accepted 16 December 2010
Available online 24 December 2010
Keywords:
Palladium
Mesoporous materials
Magnetic separation
Heterogeneous catalyst
C–C coupling reaction
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
lites [20–22], mesoporous silica [23,24], and metal oxides [25,26],
have been adopted for Pd-based heterogeneous catalysts, in which
Suzuki and Heck coupling reactions catalyzed by palladium
metal or palladium complexes have been well-established for the
formation of C–C bonds due to their high tolerance against most sol-
vents, high selectivity and moderate toxicity [1–6]. Much progress
has been made by using various Pd complexes as homogeneous cat-
alysts [7–9]. The catalytic properties of such Pd-based catalysts can
be tuned by various ligands, such as phosphines [10], dibenzylide-
neacetone (dba) [11] and carbenes [12]. Carefully designed ligands
would lead to catalysts with higher catalytic activities, prolonged
lifetimes, and higher stability to the specific reactions. However, the
relatively high costs of the palladium complexes, and the difficulties
in removing the catalysts and separating the products continu-
ously from the reaction systems, are the practical limitations to the
industrial application of Pd-based homogeneous catalysts. More-
over, serious environmental and economic problems should also be
addressed before their applications in industry [13]. In this regard,
heterogeneous Pd-based catalysts will be a promising alternative
to some of these problems, owing to their air-stability, recov-
erability, reusability and non-residual property [14]. Up to date,
various solid supports, such as clays [15,16], carbon [17–19], zeo-
Pd complexes or nanoparticles are involved as active species. One
fantastic design of monolith type material supports bearing both
palladium nanoparticles and a hierarchical porosity that combines
a macroporous structure (for good mass transport) with microp-
orous domains (for large specific surface area) could be a significant
improvement. Unfortunately, synthesis of these supports required
multiple steps [27–30].
In recent years, magnetic materials have emerged as viable
alternatives to conventional heterogeneous supports [31–36]. The
magnetic separation technology offers many advantages over con-
ventional filtration and other purification methods. For example,
in the recycling process, the catalysts can be simply and efficiently
recovered from reaction media with the external magnetic field.
This can be considered as a green technology that avoids the con-
sequences brought about by filtration steps. Magnetically separable
Pd-based catalysts have been reported in recent years [37–40].
However, in most cases, specific functionalization procedures were
inevitably needed to anchor the active Pd species onto the mag-
netic support surfaces. Due to the relatively low surface area of the
magnetic supports ever reported [37–40], high Pd using amounts
(about 1–10 mol% Pd based on aryl halide) were necessary to induce
effective coupling when using these catalysts, which clearly made
these particulate supports unattractive because of the high cost of
Pd metal. In order to solve this problem, it is necessary to greatly
∗
Corresponding author. Tel.: +86 21 52412714; fax: +86 21 52413122.
E-mail addresses: jlshi@sunm.shcnc.ac.cn, jlshi@mail.sic.ac.cn (J. Shi).
1381-1169/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2010.12.009

52
Z. Gao et al. / Journal of Molecular Catalysis A: Chemical 336 (2011) 51–57
increase the active surface area of the catalyst. Mesoporous materi-
als possess extraordinarily high surface areas, tunable pore sizes in
nanoscale and well-defined pore size distributions, therefore they
are the suitable supports to load noble metal nanoparticles of high
active surface areas.
the amount of CO2 desorbed was detected by a gas chromatogra-
phy.
2.3. Catalyzed Suzuki cross-coupling reactions
Recently, our group has developed a simple template-free strat-
egy to prepare a high surface area superparamagnetic mesoporous
spinel ferrite by controlled thermal decomposition of their metal
oxalate precursors [41]. Herein, we report the successful loading
of Pd0 species onto this magnetic mesoporous material, which is
used as a highly efficient, environmentally friendly, magnetically
separable and reusable catalyst for Suzuki and Heck reactions. A
synergistic catalytic reaction mechanism is discussed and proposed
accounting for the extremely high catalytic activity.
All reactants and solvents were used without further purifi-
cation or drying. The reaction parameters (solvent, reaction time
and base concentration) were investigated to find the optimal con-
ditions, which were determined to be as follows. Typically, aryl
halide (5 mmol), phenylboronic acid (7.5 mmol), potassium car-
bonate (5.5 mmol), 10 mL 1-methyl-2-pyrrolidinone (NMP), 4 mL
H2O and Pd-containing catalyst (0.004 mmol based on Pd, i.e.,
0.08 mol% Pd based on aryl halide) were added to a 25 mL round-
bottom flask equipped with a magnetic stirring bar. The reaction
mixture was stirred at 80 ◦C for 1–12 h. After the reaction, the Pd-
containing catalyst was separated with a magnet, washed with
dichloromethane for removing the adsorbed organic substrates and
dried at room temperature. The pure product was isolated and
confirmed with gas chromatographic mass spectrometry (GC–MS)
(Agilent, 6890N/5973N).
2. Experimental details
2.1. Preparation of Pd0/magnetic NiFe2O4
Mesoporous NiFe2O4-300 (NF300) and NiFe2O4-700 (NF700)
were synthesized according to our earlier publication 41. A typical
preparation procedure: 10 mmol NiCl2·6H2O, 20 mmol FeSO4·7H2O
and 5 mmol AOT were dissolved in 80 mL water at 80 ◦C in oil bath
with stirring for at least 1 h; then 30 mL 1 M oxalic acid solution
was slowly dropped into the former solution. Precipitates yielded
during the process. The precipitated products were quickly cooled
down by ice water, then separated by centrifugation, subsequently
washed with deionized water and ethanol several times, and dried
at 60 ◦C overnight. Mesoporous samples were obtained by calcin-
ing the precursor NiFe2(C2O4)3·6H2O at 300 and 700 ◦C for 60 min,
respectively. The heating rate during calcinations should be con-
2.4. Catalyzed Heck coupling reactions
The optimal reaction parameters (solvent, reaction time and
base concentration) for Heck reactions were also determined as
follows. Typically, aryl halide (5 mmol), styrene (7.5 mmol), tri-
ethylamine (5.5 mmol), and Pd-containing catalyst (0.004 mmol Pd,
i.e., 0.08 mol% Pd based on aryl halide) were added to a 25 mL
round-bottom flask equipped with a magnetic stirring bar. 10 mL of
N,N-dimethylformamide (DMF) were added and the reaction mix-
ture was stirred at 80–140 ◦C. The separation and product isolation
procedures were the same as that adopted for the Suzuki coupling
reactions.
trolled to be as low as 1 ◦C min−1
.
Then 1 g of these materials were suspended in 100 mL of
aqueous ammonium tetrachloropalladate (II) (0.07 g) solution and
stirred at 25 ◦C for 12 h. The solid catalysts were collected by
an external magnet, washed thoroughly with excess of deionized
water and dried at 100 ◦C overnight. Then the catalysts (1 g) were
reduced under flowing 5% H2/Ar at 200 ◦C for 3 h to give air-stable
black powder products. The temperature program: heating from
room temperature to 200 ◦C at 1 ◦C min−1, maintain the tempera-
ture for 3 h, then natural cooling to room temperature. The reducing
gas was always on until the end.
2.5. Recycling procedures
In the Suzuki recycling experiment, 4-bromoacetophenone
(20 mmol), phenylboronic acid (30 mmol), potassium carbonate
(22 mmol), 40 mL NMP, 16 mL H2O and Pd catalyst (0.016 mmol)
were added into the reactor and stirred at 80 ◦C. After 3 h, the reac-
tion was completed. The catalyst was separated with a magnet and
washed with dichloromethane several times. Then the catalyst was
dried under vacuum and reused in the next run. For Heck recycling
experiment, iodobenzene (20 mmol), styrene (30 mmol), triethy-
lamine (22 mmol), 40 mL DMF and Pd catalyst (0.016 mmol) were
added in the reactor and stirred at 80 ◦C for 4 h. Then the separa-
tion and following procedure were the same as those for Suzuki
recycling reaction.
2.2. Material characterization
X-ray diffraction (XRD) patterns were recorded on a Rigaku
D/Max 2200PC diffractometer using Cu Ka radiation at 40 kV and
40 mA. The N2 sorption measurements were performed using
Micromeritics Tristar 3000 at 77 K for mesoporosity, and the
mesoporous specific surface area and the pore size distribution
were calculated using the Brunauer–Emmett–Teller (BET) and
Barrett–Joyner–Halenda (BJH) methods, respectively. Transmission
electron microscopy (TEM) images were obtained on a JEOL-2010
electron microscope operated at 200 kV. A vibrating-sample mag-
netometer (PPMS Model 6000 Quantum Design, San Diego, USA)
was used to study the magnetic properties. X-ray photoelectron
spectroscopy (XPS) signals were collected on a VG Micro MK II
instrument using monochromatic Mg Ka X-rays at 1253.6 eV oper-
ated at 200 W. The basicities of the supports were determined using
the temperature-programmed desorption of carbon dioxide (CO2-
TPD) as follows. The supports (50 mg) were treated by heating at
450 ◦C under N2 and then exposed to the probe gas stream at room
temperature. Physically adsorbed CO2 was removed by flushing
N2 at room temperature for 1 h. Then the TPD analysis was run
under N2 flow at a heating rate of 10 ◦C min−1 up to 500 ◦C and
3. Results
3.1. Characterization of the supports and catalysts
Fig. 1 gives the CO2-TPD spectra for NF300 and NF700 supports.
Both of them show broadened desorption peaks at temperatures
between 50 and 400 ◦C, which can be assigned to bicarbonates
formed on Brønsted OH groups. The CO2-TPD results suggest rela-
tively weak basicity for both supports. Apparently, the desorption
peak of NF300 is clearly located at a higher temperature (132 ◦C)
than that (80 ◦C) of NF700, indicating that NF300 possesses rela-
tively higher basic strength. On the other hand, the overall amount
of CO2 desorbed from NF300 is greatly larger than that of NF700,
revealing its remarkably higher amount of basic sites/Brønsted OH
groups per unit weight, which is due to its much higher surface area
(301.6 m2 g−1) of NF300 than that (27.7 m2 g−1) of NF700.

Z. Gao et al. / Journal of Molecular Catalysis A: Chemical 336 (2011) 51–57
53
bears larger surface area, larger pore volume, smaller BJH pore size,
lower crystallinity than NF700. Also, NF300 may have some resid-
ual organic groups, and higher amount of surface/lattice defects
than NF700. Through wet impregnation with aqueous (NH4)2PdCl4
solution and subsequent hydrogen reduction, both supports were
successfully loaded with Pd0, and the colors of two powders turned
into black after Pd loading. The weight percentages of palladium
loaded in NF300 and NF700 were found to be 2.1% and 1.0% by
inductively coupled plasma-atomic emission spectroscopy (ICP-
AES) analysis.
The TEM image of Pd/NF300 shown in Fig. 2a demonstrates that
the Pd/NF300 material is still mesoporous after loading. No discrete
large palladium particles can be found on support surface, and most
of the palladium particles are believed to have been loaded within
the pore channel due to the growth/agglomeration inhibition effect
of the mesoporous frameworks of the support. Comparatively, we
can find that part of palladium nanoparticles have agglomerated
into bigger particles and dispersed onto the external surface of
NF700 support (Fig. 2c). This may be resulted from lower surface
area and pore volume of NF700, which cannot accommodate the
large amount of palladium loading. After five repeated Heck reac-
tion cycles (Fig. 2b and d), we did not observe significant change in
the morphology of the catalysts, but some Pd particles might have
aggregated onto the surfaces of the matrices. This may take place
through the leaching/re-deposition processes during the catalytic
reactions.
Fig. 1. Temperature-programmed desorption profiles of CO2 on NF300 and NF700
samples by using the same amount (50 mg) of the samples.
Magnetic mesoporous NF300 sample was used as catalyst sup-
port in this study. As a contrast, the palladium particles were also
loaded on NF700 as catalyst (Pd/NF700). As mentioned in Ref. [41],
these two supports have the same composition. However, NF300
Fig. 2. TEM images of the Pd/NF300 (a and b) and Pd/NF700 (c and d) catalysts before reaction and after five cycles of Heck reactions.

54
Z. Gao et al. / Journal of Molecular Catalysis A: Chemical 336 (2011) 51–57
Fig. 5. Nitrogen adsorption–desorption isotherms and corresponding BJH desorp-
tion pore size distribution plots (inset) of the Pd/NF300 and Pd/NF700 catalysts.
Fig. 3. XRD patterns of the catalysts Pd/NF300 (a), Pd/NF700 (b).
there is a significant decrease in the BET surface area of the mate-
rial (120.5 m2 g−1) compared to the support without Pd loading
(301.6 m2 g–1), due to the Pd deposition within the pore and/or on
the surface, as expected. The Pd/NF300 catalyst has a sharp pore
size distribution centered at 3.1 nm (Fig. 5 inset), which is a little
larger than that of the support (centered at 2.5 nm). Such a larger
average pore size might be caused by the filling of Pd nanoparticles
in those mesopores which are smaller than, e.g., 2.5 nm. As far as
known, it is the first report that Pd0 was successfully loaded onto
a single magnetic material which possesses mesoporous structure
itself. When the Pd species were loaded on NF700, its surface area of
The powder X-ray diffraction patterns of the Pd/NF300 and
Pd/NF700 catalysts are shown in Fig. 3. The characteristic XRD peaks
of the fcc crystal structure of metallic palladium ((2 0 0), (2 2 0) and
(3 1 1)) cannot be clearly seen from the patterns because of the rel-
atively low Pd content. But broadened (1 1 1) diffraction peaks still
can be identified in both patterns. The diffraction peak at 39.86◦
indicates that Pd particles are in metallic state. Moreover, there
were no Bragg reflections due to any crystalline PdO phase within
the detection limit.
Furthermore, the materials were characterized by X-ray photo-
electron spectroscopy to ascertain the oxidation state of Pd species.
The XPS spectrum of the Pd/NF300 catalyst (see Fig. 4) show typ-
Pd/NF700 decreased to 18.8 m2 g−1 compared to that (27.7 m2 g−1
)
of the NF700 support. The Pd/NF700 sample shows a broadened
pore size distribution centered at 22.1 nm.
ical Pd0 absorptions at 335.75 and 341.00 eV for 3d5/2 and 3d3/2
,
respectively, with a spin-orbit separation of 5.25 eV. No peaks for
Pd2+ were observed. This absorption doublet typically seen in Pd0
containing catalysts confirms that most of the Pd2+ ions have been
reduced to Pd0 by H2, which are well dispersed in the pore net-
work of the magnetic mesoporous support. The XPS spectrum of
Pd/NF700 shows almost the same result.
After loading palladium particles, the Pd/NF300 catalyst still
possesses mesoporous structure, as can be confirmed by the typ-
ical Langmuir IV type N2 sorption isotherms (Fig. 5). However,
The magnetization curves of the catalysts were measured at
room temperature, as shown in Fig. 6. Neither of the curves
reaches saturation point. The intensities of magnetization at 10 kOe
for Pd/NF300 and Pd/NF700 are 3.92 emu g−1 and 31.36 emu g−1
respectively, which are a little lower than those of the corre-
sponding supports (4.00 emu g−1 for NF300 and 32.08 emu g−1 for
NF700). This is mostly due to the Pd0 loading on mesoporous
NiFe2O4. For Pd/NF300, there is no apparent hysteresis detected,
and the coercive force is as low as 25 Oe, suggesting a superparam-
agnetic feature of the catalyst. The magnetic properties of these
Fig. 4. XPS spectrums of the catalysts Pd/NF300 (a), Pd/NF700 (b).
Fig. 6. Magnetization curves of the Pd/NF300 (a), Pd/N700 (b) catalysts.

Z. Gao et al. / Journal of Molecular Catalysis A: Chemical 336 (2011) 51–57
55
Scheme 1. Suzuki reaction of aryl halides and phenylboronic acid.
Scheme 2. Heck reaction of aryl halides and styrene.
Table 2
catalysts are useful for their efficient recovery and recycling in
liquid-phase reactions.
The influence of the solvent on catalytic activity for the coupling of iodobenzene
and styrene with catalyst Pd/NF300.
Entry
Temp. (◦C)
Solvent
Yield (%)
97
94
21
48
0
3.2. Catalytic activities
1
2
3
4
5
80
80
80
80
66
DMF
NMP
Toluene
CH3CN
THF
The catalytic activities of all the synthesized supports (NF300,
NF700) and Pd-loaded catalysts (Pd/NF300, Pd/NF700) were tested
for both Suzuki (Scheme 1) and Heck (Scheme 2) reactions. The
results are summarized in Table 1. The reaction conditions, namely
the use of various bases, additives, solvents and the ratio of the
reacting components, were tested to make sure that the reactions
were carried out under the optimized conditions (see Section 2). As
a typical example for Suzuki reaction, bromobenzene was selected
as a non-activated aryl halide. In general, non-activated aryl halides
are less active than activated aryl halides. The reaction between
iodobenzene and styrene was chosen as the typical Heck reaction.
The influence of the solvent on catalytic activity was studied in the
coupling of iodobenzene and styrene with catalyst Pd/NF300. The
results are summarized in Table 2. Among all the solvents, DMF
was the best solvent for this catalyst. NMP gave lower yields. The
results obtained in the less polar solvents such as toluene, CH3CN
and THF, were not satisfactory.
Reaction conditions: iodobenzene = 5 mmol; styrene = 7.5 mmol; Et3N = 5.5 mmol;
solvent 10 mL; time = 4 h; 0.08 mol% Pd. Yield = mol of product/mol of reactant con-
verted.
proceed at 80 ◦C with relatively high product yields under a low
Pd using amount of 0.08 mol% based on aryl halide. These largely
enhanced catalytic activities are believed to be resulted from the
synergetic catalytic effect, which will be discussed later. So we
choose Pd/NF300 as the appropriate catalyst for the other Suzuki
and Heck reactions.
In further studies, Suzuki reactions of a wide range of other
aryl halides with phenylboronic acid were conducted. The results
are shown in Table 3. The Suzuki reaction of iodobenzene with
phenylboronic acid could reach the complete conversion within
1 h. Various aryl bromides and chlorides were used in the reac-
tions under the optimized conditions, and relatively high yields
were obtained, though longer reaction periods were needed. Both
electron-deficient and electron-rich substrates have coupled very
efficiently under the catalysis.
The NF300 and NF700 supports themselves do not possess any
catalytic activity for both Suzuki and Heck reactions as shown in
Table 1. When Pd/NF700 was used as catalyst for the same two reac-
tions, it shows lower catalytic activities. However, when Pd/NF300
was used as a catalyst, both Suzuki (bromobenzene with phenyl-
boronic acid) and Heck (iodobenzene with styrene) reactions can
The Pd/NF300 catalyst was also applied to catalyze Heck
carbon–carbon coupling reactions of various aryl halides. The reac-
tions were conveniently carried out in air at temperatures between
80 and 140 ◦C. Table 4 gives the catalytic reaction results. Aryl
bromides and chlorides are difficult to be activated because of
Table 1
Yield and reaction time of typical Suzuki and Heck reactions catalyzed by the meso-
porous supports (NF300 and NF700) themselves and Pd-loaded catalysts (Pd/NF300
and Pd/NF700).
Entry
R
X
Catalyst
Yield (%)
Time (h)
12
12
12
12
4
4
4
4
Table 3
Suzuki reactions of various aryl halides with phenylboronic acid at 80 ◦C using
Pd/NF300 as catalyst.
1a
2a
3a
4a
5b
6b
7b
8b
H
H
H
H
H
H
H
H
Br
Br
Br
Br
I
I
I
I
NF300
NF700
Pd/NF300
Pd/NF700
NF300
NF700
Pd/NF300
Pd/NF700
0
0
97
13
0
0
Entry
R
X
Yield (%)
Time (h)
1
2
3
4
5
6
CH3
OCH3
NO2
COCH3
H
Br
Br
Br
Br
I
95
67
99
100
100
19
12
12
9
3
1
97
9
a
Suzuki reaction: bromobenzene = 5 mmol; phenylboronic acid = 7.5 mmol;
K2CO3 = 10 mmol; NMP 10 mL, H2O 4 mL; 0.08 mol% Pd; 80 ◦C. Yield = mol of prod-
uct/mol of reactant converted.
H
Cl
24
Reaction conditions: aryl halide = 5 mmol; phenylboronic acid = 7.5 mmol;
K2CO3 = 10 mmol; NMP 10 mL, H2O 4 mL; 0.08 mol% Pd. Yield = mol of product/mol
of reactant converted.
b
Heck reaction: iodobenzene = 5 mmol; styrene = 7.5 mmol; Et3N = 5.5 mmol;
DMF 10 mL; 0.08 mol% Pd; 80 ◦C. Yield = mol of product/mol of reactant converted.

56
Z. Gao et al. / Journal of Molecular Catalysis A: Chemical 336 (2011) 51–57
Table 4
reusability and stabilities for both the Suzuki and Heck reactions.
Especially for the Suzuki reaction, almost no deactivation can be
found even after five cycles.
Heck reactions of various aryl halides with styrene using Pd/NF300 as catalysts.
Entry
R
X
Yield (%)
41
30
53
98
96
9
Temp. (◦C)
Time (h)
1
2
3
4
H
Br
Br
Br
Br
Br
Cl
140
140
140
120
120
140
24
24
24
12
12
24
CH3
OCH3
NO2
COCH3
H
4. Discussion
5
6a
The postulated mechanisms for homogeneous Suzuki and Heck
reactions catalyzed by Pd species have been well described in
many organic chemistry textbooks. But for heterogeneous reac-
tions, mechanisms are not yet fully understood and findings in this
field are sometimes contradictory. Some authors insisted that the
reactions take place at the surface of solid Pd as truly heteroge-
neous reactions [42,43]. On the other side, a variety of experimental
results clearly demonstrated that the leached palladium from the
solid support was the real active species serving as the homoge-
neous catalyst for Suzuki and Heck reactions [44]. In the present
study, we have actually observed the presence of Pd leaching in
our experiments. No matter which mechanism prove to be true,
the initial oxidative addition step, which has been considered to be
rate-determining [45], would always take place at the surface of
solid Pd.
Reaction conditions: aryl halide = 5 mmol; styrene = 7.5 mmol; Et3N = 5.5 mmol;
DMF 10 mL; 0.08 mol% Pd. Yield = mol of product/mol of reactant converted.
a
0.4 mol% Pd was used, PPh4Br was used as co-catalyst, the ratio of P/Pd is 10:1.
the relatively strong C–Br and C–Cl bonding. As a result, the cou-
pling reactions of aryl bromides and chlorides with different olefins
require higher reaction temperatures and prolonged reaction
time.
The effect of electron-withdrawing substituents was observed
in these two reactions. Aryl halides with electron-donating groups
are less active than the ones with electron-withdrawing groups
in the same para positions. The conversion rate of aryl bromide is
dependent on the activation brought by the substitute in the para
position, which is similar to homogeneous catalysts. Nevertheless,
the Heck reactions of a variety of aryl bromides with styrene can
still proceed to give the corresponding products in relatively high
yields under the catalysis of Pd/NF300.
The catalytic activities of the heterogeneous magnetic catalysts
reported in literature [37–40] and in this work, both for Suzuki reac-
tions of bromobenzene with phenylboronic acid and Heck reactions
of bromobenzene and iodobenzene with styrene, are compared. It
is clear that the type of catalyst support plays an important role
in determining the catalytic activity. Pd/NF300 reported in this
work shows remarkably higher catalytic activity than the mag-
netic heterogeneous catalysts for Suzuki and Heck reactions ever
reported.
The enhanced catalytic activity of Pd/NF300 may be ascribed
to a synergetic catalytic effect between the highly dispersed Pd
nanoparticles and basic mesoporous NiFe2O4 support, as discussed
in the following section.
The surface area of the catalyst will surely affect the cat-
alytic activity. Homogeneous loading of Pd nanoparticles in the
pore network is greatly helpful to achieve a high Pd0 dispersion.
As the activity of the catalysts mainly depends on the property
and concentration of the active palladium species, the increased
Pd0 dispersivity would lead to the greatly increased concentra-
tion of active species in the reaction system, and consequently
the enhanced catalytic activity. Therefore the catalyst activity was
greatly enhanced with the Pd catalyst highly dispersed in/on high
surface area mesoporous materials.
3.3. Catalyst reusability
On the other hand, the CO2-TPD result shows that the meso-
porous NiFe2O4 supports possess a certain degree of basicity,
especially for NF300 with higher specific surface area. This implies
that there exists a considerable amount of Brønsted basic sites on
the pore surface of mesoporous materials. Some authors proposed
that for Heck and Suzuki reactions, palladium on basic supports
may exhibit an increased activity compared to that of acidic sup-
ports, depending on the basicity of the supports [26,46]. The higher
basicity of the support would lead to higher electronic density
on the palladium surface, which, in turn, facilitates the oxida-
tive addition. Therefore it is inferred that the basic supports in
the present study should act as an electron donor for palladium,
resulting in higher electronic density on the palladium surface and
subsequently promoting the oxidative addition, as illustrated in
Scheme 3, while acidic supports act as an electron acceptor [47]. In
a homogeneous catalytic mechanism, the formation of the interme-
diates ArPdIIX in the reaction solutions will promote the following
reactions between the intermediates and phenylboronic acid for
Suzuki reaction or styrene for Heck reaction, giving the final reac-
tion products. As the oxidative addition step is the rate-determining
step [45], therefore the basic support plays an important synergis-
tic role in accelerating the catalytic reactions through promoting
the formation of intermediate active species of ArPdIIX.
Another important issue concerning the application of a het-
erogeneous catalyst is its reusability and stability under reaction
conditions. To gain insight into this issue, catalyst recycling
experiments were carried out using a Suzuki reaction of 4-
bromoacetophenone and phenylboronic acid, and a Heck reaction
of iodobenzene and styrene. Before reuse, the magnetic catalyst
can be easily and efficiently recovered from the reaction mixture
with an external magnet, washed with dichloromethane and then
dried. The results are shown in Table 5. The amounts of Pd leaching
into solution for the Suzuki and Heck reactions were detected by
checking the Pd loading amount before and after each reaction cycle
through ICP. The loss of Pd amount for the Heck reaction were cal-
culated to be: 1.5 wt.%, 1.1 wt.%, 2.0 wt.%, 0.9 wt.%, 1.3 wt.% of total
Pd content for each run, respectively. Even though a small amount
of Pd loss could be detected, the catalyst still showed relatively high
Table 5
Recycling performance of the Pd/NF300 catalyst.
Run
Suzuki reactiona
Heck reactionb
Yield
100
99
100
99
TON
TOF
Yield
TON
TOF
1st
1250
1237
1237
1225
1223
417
413
413
408
408
97
95
95
93
94
1212
1204
1204
1203
1202
303
301
301
301
300
2nd
3rd
4th
5th
In a word, the basicity of the support surface, together with its
large surface area, synergistically promote the oxidative addition
of the aryl halide on palladium atoms, through the electron donat-
ing effect from the palladium surface with the assistance of the
basic support, whichever mechanism, homogeneous catalysis or
true heterogeneous catalysis, proved to be true.
99
a
Reaction conditions: 4-BrPhCOCH3, PhB(OH)2, K2CO3, NMP + H2O, 0.08 mol% Pd,
T = 80 ◦C and t = 3 h.
b
Reaction conditions: PhI, styrene, Et3N, DMF, 0.08 mol% Pd, T = 80 ◦C and t = 4 h.

Z. Gao et al. / Journal of Molecular Catalysis A: Chemical 336 (2011) 51–57
57
[4] F. Bellina, A. Carpita, R. Rossi, Synthesis (2004) 2419–2440.
[5] S. Kotha, K. Lahiri, D. Kashinath, Tetrahedron 58 (2002) 9633–9695.
[6] A.F. Littke, C.Y. Dai, G.C. Fu, J. Am. Chem. Soc. 122 (2000) 4020–4028.
[7] D.S. McGuinness, K.J. Cavell, Organometallics 19 (2000) 741–748.
[8] A. Schnyder, A.F. Indolese, M. Studer, H.U. Blaser, Angew. Chem., Int. Ed. 41
(2002) 3668–3671.
[9] W.A. Herrmann, V.P.W. Bohm, C.P. Reisinger, J. Organomet. Chem. 576 (1999)
23–41.
[10] C.A. Fleckenstein, H. Plenio, Chem. Soc. Rev. 39 (2010) 694–711.
[11] C. Amatore, A. Jutand, Coord. Chem. Rev. 178 (1998) 511–528.
[12] N. Marion, S.P. Nolan, Acc. Chem. Res. 41 (2008) 1440–1449.
[13] A. Zapf, M. Beller, Top. Catal. 19 (2002) 101–109.
[14] L.X. Yin, J. Liebscher, Chem. Rev. 107 (2007) 133–173.
[15] K. Shimizu, R. Maruyama, S. Komai, T. Kodama, Y. Kitayama, J. Catal. 227 (2004)
202–209.
[16] B.M. Choudary, S. Madhi, N.S. Chowdari, M.L. Kantam, B. Sreedhar, J. Am. Chem.
Soc. 124 (2002) 14127–14136.
[17] G. Marck, A. Villiger, R. Buchecker, Tetrahedron Lett. 35 (1994) 3277–3280.
[18] F.X. Felpin, T. Ayad, S. Mitra, Eur. J. Org. Chem. 12 (2006) 2679–2690.
[19] K. Kohler, R.G. Heidenreich, J.G.E. Krauter, M. Pietsch, Chem. Eur. J. 8 (2002)
622–631.
[20] L. Djakovitch, K. Koehler, J. Am. Chem. Soc. 123 (2001) 5990–5999.
[21] L. Artok, H. Bulut, Tetrahedron Lett. 4 (2004) 3881–3884.
[22] A. Corma, H. Garcia, A. Leyva, Appl. Catal. A: Gen. 236 (2002) 179–185.
[23] L. Li, J.L. Shi, J.N. Yan, Chem. Commun. 17 (2004) 1990–1991.
[24] K. Shimizu, S. Koizumi, T. Hatamachi, H. Yoshida, S. Komai, T. Kodama, Y.
Kitayama, J. Catal. 228 (2004) 141–151.
Scheme 3. Schematic drawings of the synergetic catalytic effect between the highly
dispersed Pd0 and surface basicity of the mesoporous support for the C–C cou-
pling reactions: the oxidative addition step, which has been recognized as the
rate-determining step [45], is promoted by the electron donating effect of basic sup-
port surface and the high dispersion of Pd0 in the pore network of the mesoporous
support.
5. Conclusions
In conclusion, magnetically separable palladium-loaded meso-
porous NiFe2O4 catalysts (Pd/NF300 and Pd/NF700) have been
successfully prepared. Compared to other magnetic Pd-loaded cat-
alysts reported previously in literature, Pd/NF300 shows greatly
enhanced catalytic activities for both Suzuki and Heck reactions
in air using various aryl halides as reactants, under a very low
Pd using amount of 0.08 mol% based on aryl halide. Superparam-
agnetic Mesoporous NF300 has been found to be a distinguished
support possessing apparent basicity and well-defined mesoporous
structure for high Pd dispersion, synergistically accelerating the
catalytic reactions through promoting the oxidative addition step.
All these properties lead to an excellent catalyst Pd/NF300 with
high catalytic activity towards Suzuki and Heck reactions, excellent
reusability and magnetic separability.
[25] K. Kaneda, M. Higuchi, T. Imanaka, J. Mol. Catal. 63 (1990) L33–L36.
[26] A. Cwik, Z. Hell, F. Figueras, Adv. Synth. Catal. 348 (2006) 523–530.
[27] A. Desforges, R. Backov, H. Deleuze, O. Mondain-Monval, Adv. Funct. Mater. 15
(2005) 1689–1695.
[28] A. Desforges, H. Deleuze, O. Mondain-Monval, R. Backov, Ind. Eng. Chem. Res.
44 (2005) 8521–8529.
[29] S. Ungureanu, H. Deleuze, C. Sanchez, M.I. Popa, R. Backov, Chem. Mater. 20
(2008) 6494–6500.
[30] S. Ungureanu, H. Deleuze, O. Babot, M.F. Achard, C. Sanchez, M.I. Popa, R. Backov,
Appl. Catal. A: Gen. 390 (2010) 51–58.
[31] R. Abu-Reziq, H. Alper, D.S. Wang, M.L. Post, J. Am. Chem. Soc. 128 (2006)
5279–5282.
[32] Z.F. Wang, W. Yang, N.Y. He, Prog. Chem. 21 (2009) 2053–2059.
[33] S. Shylesh, J. Schweizer, S. Demeshko, V. Schunemann, S. Ernst, W.R. Thiel, Adv.
Synth. Catal. 351 (2009) 1789–1795.
[34] L.M. Rossi, L.L.R. Vono, F.P. Silva, P.K. Kiyohara, E.L. Duarte, J.R. Matos, Appl.
Catal. A: Gen. 330 (2007) 139–144.
[35] S. Shylesh, V. Schunemann, W.R. Thiel, Angew. Chem. Int. Ed. 49 (2010)
3428–3459.
[36] R. Abu-Reziq, D. Wang, M. Post, H. Alper, Chem. Mat. 20 (2008) 2544–2550.
[37] B. Baruwati, D. Guin, S.V. Manorama, Org. Lett. 9 (2007) 5377–5380.
[38] H. Yoon, S. Ko, J. Jang, Chem. Commun. 14 (2007) 1468–1470.
[39] Y.H. Zhu, S.C. Peng, A. Emi, S. Zhenshun, R.A. Monalisa, Kemp, Adv. Synth. Catal.
349 (2007) 1917–1922.
[40] A.J. Amali, R.K. Rana, Green Chem. 11 (2009) 1781–1786.
[41] Z. Gao, F.M. Cui, S.Z. Zeng, L.M. Guo, J.L. Shi, Micropor. Mesopor. Mater. 132
(2010) 188–195.
[42] R.L. Augustine, S.T. Oleary, J. Mol. Catal. A: Chem. 95 (1995) 277–285.
[43] C.R. LeBlond, A.T. Andrews, Y.K. Sun, J.R. Sowa, Org. Lett. 3 (2001) 1555–1557.
[44] N.T.S. Phan, M. Van Der Sluys, C.W. Jones, Adv. Synth. Catal. 348 (2006) 609–679.
[45] J.P. Corbet, G. Mignani, Chem. Rev. 106 (2006) 2651–2710.
[46] M.L. Kantam, K.B.S. Kumar, P. Srinivas, B. Sreedhar, Adv. Synth. Catal. 349 (2007)
1141–1149.
Acknowledgments
The authors gratefully acknowledge the financial support from
National Natural Science Foundation of China (grant no. 20633090,
20703055 and 50872140), and Shanghai Nano-Science Projects
(grant no. 0852nm03900).
References
[1] R.F. Heck, J.P. Nolley, J. Org. Chem. 37 (1972) 2320–2322.
[2] N. Miyaura, A. Suzuki, J. Chem. Soc. Chem. Commun. 19 (1979) 866–867.
[3] V. Farina, Adv. Synth. Catal. 346 (2004) 1553–1582.
[47] A.Y. Stakheev, L.M. Kustov, Appl. Catal. A: Gen. 188 (1999) 3–35.