A single-step procedure for the preparation of palladium nanoparticles and a phosphine-functionalized support as catalyst for Suzuki cross-coupling reactions

Article abstract of DOI:10.1016/j.jcat.2010.09.028

We here report the preparation of supported palladium nanoparticles (NPs) stabilized by pendant phosphine groups by reacting a palladium complex containing the ligand 2-(diphenylphosphino)benzaldehyde with an amino-functionalized silica surface. The Pd na

Full text of DOI:10.1016/j.jcat.2010.09.028

Journal of Catalysis 276 (2010) 382–389
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
A single-step procedure for the preparation of palladium nanoparticles
and a phosphine-functionalized support as catalyst for Suzuki cross-coupling
reactions
Natália J.S. Costa a, Pedro K. Kiyohara b, Adriano L. Monteiro c, Yannick Coppel d,e, Karine Philippot d,e
,
Liane M. Rossi a,
⇑
a Instituto de Química, Universidade de São Paulo, Av. Prof. Lineu Prestes 748, São Paulo 05508-000, SP, Brazil
b Instituto de Física, Universidade de São Paulo, 05315-970 São Paulo, SP, Brazil
c Laboratory of Molecular Catalysis, Instituto de Química, UFRGS, Av. Bento Gonçalves 9500, Porto Alegre 91501-970, CP 15003, RS, Brazil
d CNRS, LCC (Laboratoire de Chimie de Coordination du CNRS), 205, route de Narbonne, F-31077 Toulouse, France
e Université de Toulouse, UPS, INPT, LCC, F-31077 Toulouse, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 21 July 2010
Revised 29 September 2010
Accepted 29 September 2010
Available online 30 October 2010
We here report the preparation of supported palladium nanoparticles (NPs) stabilized by pendant
phosphine groups by reacting a palladium complex containing the ligand 2-(diphenylphosphino)benzal-
dehyde with an amino-functionalized silica surface. The Pd nanocatalyst is active for Suzuki
cross-coupling reaction avoiding any addition of other sources of phosphine ligands. The Pd intermedi-
ates and Pd NPs were characterized by solid-state nuclear magnetic resonance and transmission electron
microscopy techniques. The synthetic method was also applied to prepare magnetically recoverable Pd
NPs leading to a catalyst that could be reused for up to 10 recycles. In summary, we gathered the advan-
tages of heterogeneous catalysis, magnetic separation and enhanced catalytic activity of palladium pro-
moted by phosphine ligands to synthesize a new catalyst for Suzuki cross-coupling reactions. The Pd NP
catalyst prepared on the phosphine-functionalized support was more active and selective than a similar
Pd NP catalyst prepared on an amino-functionalized support.
Keywords:
Palladium
Iminophosphine
Suzuki reaction
Nanoparticle
Magnetic separation
Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction
ophosphino–palladium complexes in Suzuki reactions has been
investigated [8–11]. As the reactions are typically catalyzed by
The Pd-catalyzed Suzuki cross-coupling reaction has been
shown as an efficient method for the construction of CAC bonds
and plays an important role in pharmaceutical industry and organ-
ic synthesis [1–3]. Various efficient Pd catalyst precursors have
been developed in recent years that allow aryl halides to be effec-
tively coupled with aryl boronic acids under mild reaction condi-
tions [4,5]. Most catalysts for the Suzuki cross-coupling reaction
make use of phosphines. Triarylphosphine was the first class of
phosphine ligands used, and bulky and electron-rich phosphine li-
gands that can highly improve the efficiency and selectivity of such
cross-coupling reactions were introduced later [1]. Iminophos-
phines are bidentate ligands with both hard and soft donor atoms
and are expected to exhibit hemi-labile behavior when coordi-
nated to palladium [6,7]. These ligands can be easily obtained by
the reaction of o-(diphenylphosphino)benzaldehyde with an
amine, while the steric and electronic properties of the N atom
can be modulate by the amine used. The catalytic activity of amin-
homogeneous palladium complexes, the possibility of reuse in suc-
cessive reactions is limited. Alternatively, heterogenized palladium
catalysts have been developed in order to improve catalyst stabil-
ity and handling during workup procedures and recycling [12]. Re-
cycle and reuse of the catalyst were improved by using palladium
catalyst based on polymer-supported dialkylphosphinobiphenyl li-
gands [13]. In this example, the presence of phosphine ligands on
the support is the key factor for the catalytic activity achieved,
since the polymer containing only palladium acetate led to only
traces of products in the Suzuki reaction [13]. However, another
example shows that palladium acetate impregnated in PEG 200
in the absence of any ligand is an active catalyst for Suzuki reaction
[14]. Recent examples show that palladium nanoparticles (NPs) are
active catalysts for cross-coupling reactions in the presence [15] or
absence [16–18] of phosphine ligands.
Magnetic separation has emerged as a robust, highly efficient
and fast separation tool with many advantages compared with
product/catalyst isolation by means of other chemical or physical
procedures, such as liquid–liquid extraction, chromatography, dis-
tillation, filtration or centrifugation. The catalyst immobilized on
⇑
Corresponding author. Fax: +55 11 38155579.
E-mail address: lrossi@iq.usp.br (L.M. Rossi).
0021-9517/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2010.09.028

N.J.S. Costa et al. / Journal of Catalysis 276 (2010) 382–389
383
superparamagnetic nanomaterials can be easily separated from the
products due to a strong interaction between the magnetic NPs and
an external applied magnetic field, and it can be easily redispersed
in the absence of the magnetic field due to the absence of ‘‘mag-
netic memory”. This is of special importance for washing steps
(with the catalysts inside the reactor) and for redispersion of the
catalyst for successive reactions. Many authors described palla-
dium complexes and palladium NPs supported on magnetic NPs
as catalysts for Suzuki reactions showing excellent catalytic activ-
ity [19–25]. As for example, a palladium(II)–phosphine complex
supported on silica-coated magnetic nanoparticles [25] was re-
ported as a highly active and recyclable catalyst for the Suzuki–
Miyaura cross-coupling reaction.
In this work, we gathered the advantages of heterogeneous
catalysis, magnetic separation and enhanced catalytic activity of
palladium in the presence of phosphine ligands to synthesize a
new catalyst for Suzuki cross-coupling reaction. In a single-step
procedure, Pd NPs are formed and phosphine ligands were cova-
lently attached to conventional and magnetic support surfaces.
As a result, the Pd NPs stabilized by the phosphine pendant groups
lead to an active heterogeneous catalyst system to coupling
reactions.
The amino-modified catalyst support Fe3O4@SiO2–NH2
(300 mg) was added to the Pd(II) complex stock solution, and the
suspension was refluxed for 24 h. The resulting solid (Fe3O4@-
SiO2–iminophosphine–Pd) was recovered magnetically, washed
three times with toluene and dried at room temperature.
As a reference, another catalyst was prepared in similar condi-
tions by reacting the amino-functionalized support (Fe3O4@SiO2–
NH2) with Pd(OAc)2 giving rise to a catalyst system without phos-
phine ligands.
2.4. General conditions for Suzuki–Miyaura reactions
The cross-coupling reactions were carried out in a sealed
Schlenk flask containing 0.2 mmol of aryl halide, 0.24 mmol of
phenylboronic acid, 0.4 mmol of base, 0.001 mmol of palladium
(0.5 mol%) and 3 mL of solvent. The reactions were conducted un-
der magnetic stirring (700 rpm) for 2 h at 100 °C. The temperature
was maintained using an oil bath and a hot-stirring plate con-
nected to a digital controller (ETS-D4 IKA). The catalyst was sepa-
rated with
a permanent magnet, and the products were
characterized by GC-MS and quantified by GC – FID.
The magnetically recovered catalyst was washed three times
with ethanol and dried in vacuum. For recycling experiments,
new portions of substrates and base were added. The leaching of
palladium in hot solutions after removing the catalyst magnetically
was measured by ICP-AES.
2. Materials and methods
2.1. Methods
The transmission electron microscopy (TEM) and energy disper-
sive X-ray spectroscopy (EDS) analysis were accomplished using a
Philips CM 200 Microscope operating at a 200 kV. Mass spectra
were obtained on a GC-MS Shimadzu 14B/QP5050A with a 30-m
column DB5. Gas chromatography analyses were performed on a
Shimadzu GC-2010 with a 30-m column Rtx-5. The metal was
quantified by inductively coupled plasma atomic emission spec-
troscopy (ICP-AES) with the Spectro Ciros CCD. Nuclear magnetic
resonance (NMR) experiments in solution were recorded on a Bru-
ker DPX300 spectrometer. Solid-state NMR experiments were re-
corded on a Bruker Avance 400 spectrometer equipped with a 4-
mm probe. Samples were spun between 8 and 12 kHz at the magic
angle using ZrO2 rotors. 13C{1H} and 31P{1H}-CPMAS spectra were
recorded with a recycle delay of 5 s and contact times of 2 and
3 ms, respectively. All chemical shifts for 1H and 13C are relative
to tetramethylsilane (TMS) and 31P chemical shifts were referenced
to an external 85% H3PO4 sample.
3. Results and discussion
3.1. Catalyst characterization
Palladium acetate was reacted with 2-(diphenylphos-
phino)benzaldehyde (oPCHO) in order to prepare the catalyst pre-
cursor [Pd(OAc)2(oPCHO)2] 1. Palladium is known to form a
square–planar complex with oPCHO [31], which can be character-
ized by the disappearance of the free ligand chemical shift at
À10.8 ppm and the appearance of new positive peaks in the 31P
NMR. The appearance of 31P NMR peaks at 17.5 and 24.2 ppm,
which were attributed to the trans and cis complex [Pd(OAc)2 (oP-
CHO)2], respectively [32], attested to the formation of the palla-
dium precursor 1. This palladium complex 1 was further used for
the immobilization of a palladium–iminophosphine complex tak-
ing profit of the presence of free aldehyde groups in the coordi-
nated ligands that can react with amino groups grafted on a
silica surface through the formation of Schiff bases [33,34]. There-
fore, we expect the complex [Pd(OAc)2(oPCHO)2] 1 to react with
amino groups grafted on the surface of the silica support forming
an iminophosphino–palladium complex 2. However, under reflux
conditions, this complex decomposes into supported Pd NPs and
thus leaving the support surface functionalized with pendant
phosphine groups, as shown in Scheme 1. The proposed intermedi-
ates on Scheme 1, the phosphine-functionalized support and sup-
ported Pd NPs, were studied by solid-state NMR techniques using
non-magnetic silica spheres (SiO2) synthesized by the same micro-
emulsion methodology used for the preparation of the magnetic
support. The same steps were followed to prepare the palladium
catalyst 1 and intermediates 2 and 3 using this non-magnetic silica
support functionalized with APTES. The amino-functionalized
support (SiO2–NH2) and the solids containing palladium obtained
before and after reflux were analyzed by 13C and 31P cross-polari-
zation technique combined with magic angle spinning (CPMAS) so-
lid-state NMR experiments, as shown in Figs. 1 and 2. The 13C{1H}
CPMAS NMR spectrum of the amino-functionalized support SiO2–
NH2 (Fig. 1a) exhibits peaks at 10.2 (CH2Si groups), around
2.2. Synthesis of the catalyst support
The catalyst support comprised of silica-coated magnetic NPs
was synthesized by a reverse microemulsion method, and the silica
surfaces were modified with 3-(aminopropyl)-triethoxysilane
(APTES) to give the amino-functionalized solid (Fe3O4@SiO2–NH2)
as previously reported by us [26–30].
2.3. Synthesis of the supported catalyst
Palladium acetate (12.6 mg, 0.056 mmol) and 2-(diphenylphos-
phino)benzaldehyde (oPCHO) (32.6 mg, 0.112 mmol) were reacted
in 40 mL of dry toluene under N2 in a Schlenk flask. The solution
was stirred for 30 min at room temperature and immediately used
for the next step of the preparation of the supported catalyst. o-
PCHO – 31P NMR (121 MHz, H3PO4): d (ppm) À10.80 (s). 13C
NMR (300 MHz, CDCl3): d (ppm) 191.7, 141.4, 136.2, 133.9, 132.6,
132.3, 132.1, 130.9, 129.4, 129.1, 128.9. [Pd(OAc)2 (oPCHO)2]: 31P
NMR (121 MHz, H3PO4): d (ppm) 24.2 (s), 17.5 (s).
23.0 ppm (CH2 groups) and at 42.7 ppm (CH2 group in
a-position

384
N.J.S. Costa et al. / Journal of Catalysis 276 (2010) 382–389
Fig. 1. 13C MAS NMR spectra of (a) silica amino-functionalized (SiO2–NH2), (b)
SiO2–iminophosphine–Pd(II) and (c) SiO2–iminophosphine–Pd(0).
Fig. 2. 31P MAS NMR spectra of (a) SiO2–iminophosphine–Pd(II) and (b) SiO2–
iminophosphine–Pd(0).
two other new broad peaks at 168.3 and 178.6 ppm that could
be attributed to C@N bond formation with at least two different
geometries or environments. Another new peak at 47.3 ppm agrees
with CH2 in
a-position of the imine function. The signal corre-
sponding to the carbonyl group of the ligand o-PCHO at
191.7 ppm is not present in 13C{1H} CPMAS NMR spectrum of the
solid 2, which is another strong evidence of the formation of
C@N bond. The 31P{1H} CPMAS NMR spectrum shown in Fig. 2a
exhibits a peak at 28.8 ppm, which characterize the palladium
complex 1, since the free phosphine ligand peak appears at
À10 ppm in solution 31P NMR. The chemical shift at 28.8 ppm is
close to the one found for related tertiary phosphine palladium
complexes [35]. The 13C{1H} CPMAS NMR spectrum of the solid iso-
lated after 24 h of reflux (Fig. 1c) exhibits the same signals than be-
fore the reflux step. However, the peaks of the iminophosphine are
now significantly broader than before. This broadening could result
from an increase of the structure, geometry or environment heter-
ogeneity of the iminophosphine at the surface of the NPs in 3. Pos-
sibly also, bulk magnetic susceptibility (BMS) effects and metal–
ligand electron nuclear interactions, due to the presence of Pd
NPs, could contribute to the broadening, as previously observed
[36–41]. The 31P{1H} CPMAS NMR spectrum (Fig. 2b) exhibits a
predominant peak at 32.6 ppm showing that the palladium com-
plex was not totally dissociated or most probably that there is an
Scheme 1. Step-by-step synthesis of the Fe3O4@SiO2–iminophosphine–Pd catalyst.
of the NH2 function), corresponding to the propylamine chain [35].
The weak peaks at 17.3 and 60.7 ppm may correspond to CH3 and
CH2O groups of remaining ethoxysilane. The 13C{1H} CPMAS NMR
spectrum of the solid 2 obtained after reaction with the complex
[Pd(OAc)2 (oPCHO)2] 1 (Fig. 1b) shows an additional signal at
129.3 ppm corresponding to the presence of phenyl groups and

N.J.S. Costa et al. / Journal of Catalysis 276 (2010) 382–389
385
interaction of the phosphine ligand with the Pd NPs surface (shift
of 3.8 ppm). The signals at À7.0 and 1.5 ppm could be attributed
to phosphine ligands bound directly to the silica NPs with different
geometries or environments. There are also minor peaks between 5
and 25 ppm that have not been identified at this time.
The preparation of the palladium catalyst supported on the
magnetic support was performed following the same steps of
Scheme 1. The solid support Fe3O4@SiO2–NH2 was reacted with
complex 1 followed by reflux to give Fe3O4@SiO2–iminophos-
phine–Pd 3. The content of Pd in the catalyst 3 was found to be
1.3 wt.% (determined by ICP-AES analysis). Similarly, a palladium
catalyst was prepared without phosphine ligands using the same
support and Pd(OAc)2 and named Fe3O4@SiO2–NH2–Pd 4. The cat-
alyst 4 contains 0.95 wt.% of Pd (determined by ICP-AES analysis),
and it was used for comparative studies in Suzuki cross-coupling
reactions.
The TEM images of Fe3O4@SiO2–iminophosphine–Pd 3 and
Fe3O4@SiO2–NH2Pd 4 (Fig. 3) show the morphological features of
the support, which is comprised of silica spheres of about 40 nm
containing one or more magnetic NPs of about 10 nm inside each
particle, and the formation of palladium NPs of about 3 nm on
the support surface. The Pd nanoparticles mean diameters deter-
mined by size distribution histograms fitted to Gaussian function
were 2.9 0.6 and 3.8 1.2 nm for iminophosphine and amino-
functionalized supports, respectively. The smaller size of the nano-
particles formed in the presence of phosphine groups indicates an
influence of the ligands on the nucleation and growth process of
the palladium nanoparticles, as reported elsewhere for different
amines [30].
3.2. Catalytic activity
The catalytic activity of the so-prepared catalysts was tested in
Suzuki reactions of aryl halides with phenyl boronic acid. The CAC
coupling reaction is more difficult when there is an electron donor
group as substituent for the aryl halide. The conditions for the
reaction were thus determined using the coupling of the deacti-
vated aryl bromide 4-bromoanisole and phenylboronic acid as a
model reaction. The Suzuki coupling reactions product and by-
products were depicted in Table 1. In a typical Suzuki experiment,
0.2 mmol of the aryl bromide was taken in 3 mL of solvent. To this
were added 1.2 equiv. of phenyl boronic acid and 3 equiv. of base.
The amount of catalyst used in each reaction was 8 mg, corre-
sponding to 0.5 mol%. After 2 h of heating, the reaction mixture
was cooled down to room temperature and then a magnet was
used to separate the catalyst from the reaction mixture. The prod-
ucts were analyzed by GC-MS. Table 1 summarizes the Suzuki cou-
pling reaction results with 4-bromoanisole using different bases
and solvents. All reactions give 100% selectivity to the biaryl cou-
pling product, and the best catalytic conditions for the reaction
were obtained at 100 °C with KOH as base and toluene as solvent
(Table 1, entry 3). An important feature of our catalytic system is
that it does not need any additional phosphine ligand to carry
out the reactions, because the pendant phosphine ligands present
on the support are probably sufficient for the stability of the cata-
lytic sites. The formation of byphenyl, a product of homocoupling
of phenylboronic acid, is less than 3% of the final product.
In order to exclude any activity related to the catalyst support,
the reaction of phenylboronic acid and 4-bromobenzaldehyde, a
much more activated aryl halide, was carried out in the same con-
dition of the model reaction (Table 1, entry 3), without metal cat-
alyst, in the presence and absence of the support (Fe3O4@SiO2–
NH2). It is worth to mention that the biaryl coupling product was
not formed, but a small amount of the biphenyl product was ob-
served in both reactions.
The catalyst separation at the end of the reaction was per-
formed exclusively by using a permanent magnet to recover the
catalyst in the reactor wall, taking profit of the composition of
the catalyst support that consists of silica-coated magnetite NPs.
Such a catalytic support shows excellent magnetic properties to
render magnetic separation, as previously reported [28]. The mate-
rial responds to an external magnetic field due to the high satura-
tion magnetization at room temperature (MS = 69 emu gÀ1 of
Fe3O4 at 70 kOe) and it does not remain magnetized when the
magnetic field is removed (negligible coercivity field at room tem-
perature) [28]. These magnetic properties allow a fast separation of
the catalyst from the multi-component reaction media and the
easy redispersion for washing steps or reuse in successive reac-
tions. All this procedures can be done with the great advantage
of having the catalyst inside the reactor, which minimizes the
Fig. 3. TEM images of fresh Pd NP-catalysts: (a) with phosphine, Fe3O4@SiO2–
iminophosphine–Pd, and (b) without phosphine, Fe3O4@SiO2–NH2–Pd. Inset: his-
tograms illustrating the Pd nanoparticle size distribution.

386
N.J.S. Costa et al. / Journal of Catalysis 276 (2010) 382–389
Table 1
Suzuki cross-coupling reactions catalyzed by Fe3O4@SiO2–iminophosphine–Pd 3a.
biaryl
coupling
product
OMe
+
+
Br
OMe
biphenyl
homocoupling
by-product
B(OH)2
+
reduction
by-product
OMe
Entry
Temp. (°C)
Base
Solvent
Biaryl coupling productb (%)
Biphenyl homocoupling productc (%)
1
2
3
4
5
6
7
8
100
100
100
100
100
100
80
KOH
KOH
KOH
Et3N
Na2CO3
K2CO3
KOH
DMF
5
1
2
0
0
0
0
3
1
Dioxane
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
97
>99
0
3
39
48
13
60
KOH
a
Reaction conditions: 0.2 mmol of 4-bromoanisole, 0.24 mmol of phenylboronic acid, 0.5 mol% of Fe3O4@SiO2–iminophosphine–Pd catalyst (8 mg, 1 lmol of Pd), 0.4 mmol
of base and 3 mL of solvent in inert atmosphere, 2 h. No reduction by-products were observed.
b
100% selectivity with respect to 4-bromoanisole.
Based on the initial amount of phenylboronic acid.
c
Table 2
Recyclability of the Pd catalysts in Suzuki reactiona.
Catalyst
(0.5 mol%)
Conv.
(%)
Biphenylb Selectivity to
(%)
Selectivity to
biarylc (%)
anisolec (%)
Fe3O4@SiO2–iminophosphine–Pd
1
2
3
4
5
>99
>99
69
60
32
0
0
1
3
2
0
0
0
0
0
100
100
100
100
100
Fe3O4@SiO2–NH2Pd
1
2
3
88
96
44
3
8
3
4
5
0
96
95
100
a
Reaction conditions: 0.2 mmol of 4-bromoanisole, 0.24 mmol of phenylboronic
acid, 0.5 mol% of catalyst (1 mol of Pd), 0.4 mmol of KOH and 3 mL of toluene. The
l
reaction was carried out at 100 °C for 2 h in air atmosphere.
b
Based on the initial amount of phenylboronic acid.
Based on the 4-bromoanisole conversion.
c
Fig. 4. Kinetic profiles for Suzuki cross-coupling reactions catalyzed by Fe3O4@-
SiO2–iminophosphine–Pd 3: (d) reference reaction and (N) reaction with Hg (300
equiv) added after 5 min. Reaction conditions: 1 mmol of 4-bromoanisole, 1.2 mmol
of phenylboronic acid, 0.5 mol% of Fe3O4@SiO2–iminophosphine–Pd catalyst
(40 mg, 5 lmol of Pd), 2.0 mmol of KOH and 15 mL of toluene. The reaction was
carried out at 100 °C for 2 h in air atmosphere. No reduction product was observed.
use of auxiliary substances, the use of solvents, catalyst losses, cat-
alyst oxidation and the time consumption in laborious filtration
procedures.
The performance of the Pd catalyst supported on phosphine-
functionalized solid (Fe3O4@SiO2–iminophosphine–Pd 3) was com-
pared with another Pd catalyst supported on amino-functionalized
solid (Fe3O4@SiO2–NH2–Pd 4) in the coupling reaction of phenylbo-
ronic acid with 4-bromoanisole. The results are summarized in Ta-
ble 2. The catalyst 3 was more active and selective and could be
reused five times, with >99% conversion and selectivity in the
two first reactions. After the third reuse, the catalyst lost activity
gradually. The catalyst 4 is less selective (the reduction product
anisole was observed) and less active when compared with cata-
lyst 3. The presence of pendant phosphine groups on the catalyst
support seems to play a special role in the catalytic activity of
the Pd NPs, resulting in higher selectivity, conversion and reusabil-
ity. Phosphine groups are softer donors and would be better stabi-
lizers for the catalytic active palladium species than amine groups,
and thus might explain why phosphine-containing catalysts out-
perform those without phosphine.
In palladium-catalyzed coupling reactions, it is increasingly
found that leached soluble species are the true catalytic entities

N.J.S. Costa et al. / Journal of Catalysis 276 (2010) 382–389
387
Fig. 6. TEM image of Fe3O4@SiO2–iminophosphine–Pd catalyst isolated after ten
successive Suzuki reactions using an alternative procedure (the phenylboronic acid
was previously reacted with KOH and then the catalyst was added). Inset:
histogram illustrating the Pd nanoparticle size distribution.
Fig. 5. TEM image of the Fe3O4@SiO2-iminophosphine-Pd catalyst isolated after five
Table 3
successive Suzuki reactions. Inset: EDS analysis.
Recyclability of the Fe3O4@SiO2–iminophosphine–Pd 3 catalyst in Suzuki reaction
using as strategy the reaction of phenylboronic acid and base before the catalyst
addition to the reaction medium.a
of heterogeneous catalyst [12]. Therefore, poisoning experiments
with Hg(0) were performed by adding 300 equiv. of mercury to a
typical Suzuki reaction catalyzed by the Fe3O4@SiO2–iminophos-
phine–Pd 3. The addition of Hg(0) at 5 min of reaction (4 % conver-
sion) immediately quenches the activity of catalyst 3 as shown in
Fig. 4. This result is consistent with the presence of Pd(0) active
species; however, the Hg(0) test does not distinguish catalyses
by the supported Pd NPs or leached Pd(0) species.
Cycle
Conv. (%)
Biphenylb (%)
Selectivity to biarylc (%)
1
2
3
4
5
6
7
8
9
10
56
63
49
52
55
57
47
34
42
25
4
2
1
2
2
1
1
1
1
1
100
100
100
100
100
100
100
100
100
100
One of the key factors to be investigated is the possible leaching
of active metal into the reaction mixture, thereby leading to cata-
lytic activity decrease under recycling. To address this possibility,
the leaching of palladium to still hot solutions of the first reaction
of both catalysts isolated after magnetic separation was analyzed
by ICP-AES. The results have shown that only 0.2 ppm (0.56 wt.%
of the initial metal present in the support) of palladium leached
from Fe3O4@SiO2–iminophosphine–Pd 3 and undetectable amount
(<0.01 ppm) of palladium leached from Fe3O4@SiO2–NH2Pd 4. This
result demonstrates the high affinity palladium-support, the good
performance of the magnetic separation in the present catalytic
system and, as a first conclusion, indicates that the catalyst deacti-
vation under reuse should not be directly related to metal losses.
Another key factor to be considered is the stability of the catalyst
support when submitted to the reaction conditions, especially with
the use of strong base in the Suzuki reaction. The TEM image and
the corresponding EDS analysis of the catalyst 3 after five recycles,
shown in Fig. 5, provided evidences of changes in the catalyst
structure under recycling. The core–shell morphology of the
Fe3O4@SiO2 catalyst support is no longer observed in the spent cat-
alyst. The EDS analysis confirmed that palladium is still present.
Therefore, it is reasonable to conclude that excess of base etched
the silica layer and, consequently, caused leaching and aggregation
of the Pd NPs.
a
Reaction conditions: 0.24 mmol of phenylboronic acid, 0.24 mmol of KOH,
0.2 mmol of 4-bromoanisole, 0.5 mol% of Fe3O4@SiO2–iminophosphine–Pd catalyst
(8 mg, 1 mol of Pd) and 3 mL of toluene. The reaction was carried out at 100 °C for
l
2 h in air atmosphere. No reduction product was observed.
b
Based on the initial amount of phenylboronic acid.
Based on the 4-bromoanisole conversion.
c
free base with the catalyst. Solutions of phenylboronic acid and
KOH (1:1 M ratio) in toluene were mixed under magnetic stirring
for 1 h, and then added to the reaction medium containing 4-
bromoanisole and catalyst 3. The TEM image shown in Fig. 6 con-
firms the preservation of the support core–shell morphology and a
slightly increase in the Pd nanoparticles size to 3.4 0.7 nm in the
spent catalyst. The catalytic performance of catalyst 3 using this
alternative method in successive Suzuki reactions is reported in
Table 3. The improved stability of the catalyst 3 allowed ten suc-
cessive recycles, while keeping similar conversion and selectivity
for seven successive runs. The Pd leaching in the recycling process
is negligible (0.04 ppm, which corresponds to only 0.1 wt.% of the
initial metal present in the support). The small amount of biphenyl
observed in the product distribution in Table 3 is related to the
reaction temperature and the presence of the Pd catalyst, since
In order to avoid the silica shell etching, an alternative experi-
mental condition was considered to minimize the contact of the

388
N.J.S. Costa et al. / Journal of Catalysis 276 (2010) 382–389
Table 4
Selected examples for the Suzuki cross-coupling reaction with phenylboronic acid catalyzed by Fe3O4@SiO2–iminophosphine–Pd 3a.
Entry
1
Substrate
Conv. (%)
70
Biphenylc (%)
1
Selectivity to biaryld (%)
100
Br
Br
2b
3b
54
93
3
3
100
100
O
Br
Br
Cl
4b
>99
–
–
100
O
O
5
0
–
6
7
0
–
–
NO2
Cl
>99
0.5
100
Br
8
9
53
92
1
1
100
100
MeO
Br
OMe
Br
10
11
>99
22
1
1
100
100
Br
N
Br
a
Reaction conditions: 0.2 mmol of substrate, 0.24 mmol of phenylboronic acid, 0.5 mol% of Fe3O4@SiO2–iminophosphine–Pd catalyst (8 mg, 1 lmol of Pd), 0.4 mmol of
KOH and 3 mL of toluene. The reaction was carried out at 100 °C for 2 h in inert atmosphere. No reduction product was observed.
b
Reaction carried out in air atmosphere.
Based on the initial amount of phenylboronic acid.
Based on the aryl halide conversion.
c
d
no biphenyl product was observed in the neutralization step when
KOH and phenylboronic acid are reacted before addition of 4-
bromoanisole.
tion using 4-bromoanisol displays good performance in air atmo-
sphere, for others substrates the presence of oxygen decreased
the conversion rate (entry 1 and 2 of Table 4).
The catalytic activity of Fe3O4@SiO2–iminophosphine–Pd 3 in
the presence of a series of different substrates was studied (Table
4). As expected for iminophosphine ligands, good conversion and
selectivity were obtained, except for aryl chlorides. The lack of
reduction by-products in the reactions makes more attractive the
catalyst proposed. It is important to notice that although the reac-
4. Conclusion
The complex [Pd(OAc)2 (oPCHO)2] reacts with amino groups
grafted on a silica support to form an anchored iminophosphine–

N.J.S. Costa et al. / Journal of Catalysis 276 (2010) 382–389
[11] S.M. Nobre, A.L. Monteiro, J. Mol. Catal. A: Chem. 313 (2009) 65.
389
palladium complex. Under reflux, this iminophosphine–palladium
complex decomposes into supported Pd NPs leaving the support
surface functionalized with pendant phosphine groups. This new
heterogeneous palladium catalyst presented good activity and high
selectivity in Suzuki cross-coupling reaction without the need of
addition of other sources of phosphine ligands. The supported cat-
alyst developed in this study also has the advantage to be com-
pletely recoverable with the simple application of an external
magnetic field, while keeping unaltered its efficiency for a second
run. Catalyst deactivation caused by silica etching and Pd NPs
aggregation was observed in the successive runs. However, follow-
ing an alternative method consisting in adding the base to the aryl-
boronic acid solution before the contact with the catalyst
minimizes such problems and the catalyst could be reused for up
to 10 recycles, while keeping similar catalytic activity for seven
successive reactions. Poisoning experiments with Hg(0) immedi-
ately quenches the activity of the PdNPs catalyst, which is consis-
tent with the presence of Pd(0) active species.
[12] N.T.S. Phan, M. Van Der Sluys, C.W. Jones, Adv. Synth. Catal. 348 (2006) 609.
[13] C.A. Parrish, S.L. Buchwald, J. Org. Chem. 66 (2001) 3820.
[14] T.M. Razler, Y. Hsiao, F. Qian, R. Fu, R.K. Khan, W. Doubleday, J. Org. Chem. 74
(2009) 1381.
[15] B. Tamami, S. Ghasemi, J. Mol. Catal. A: Chem. 322 (2010) 98.
[16] T. Borkowski, A.M. Trzeciak, W. Bukowski, A. Bukowska, W. Tylus, L. Kepinski,
App. Catal. A: Gen. 378 (2010).
[17] S. Ogaasawara, S. Kato, J. Am. Chem. Soc. 132 (2010) 4608.
[18] S. Fujii, S. Matsuzawa, Y. Nakamura, A. Ohtaka, T. Teratani, K. Akamatsu, T.
Tsuruoka, H. Nawafune, Langmuir 26 (9) (2010) 6230.
[19] A. Taber, J.B. Kirn, J.Y. Jung, W.S. Ahn, M.J. Jin, Synlett (2009) 2477.
[20] G.H. Lv, W.P. Mai, R. Jin, L.X. Gao, Synlett (2008) 1418.
[21] A.J. Amali, R.K. Rana, Green Chem. 11 (2009) 1781.
[22] U. Laska, C.G. Frost, G.J. Price, P.K. Plucinski, J. Catal. 268 (2009) 318.
[23] D. Rosario-Amorin, X. Wang, M. Gaboyard, R. Clérac, S. Nlate, K. Heuzé, Chem. –
Eur. J. 15 (2009) 12636.
[24] D. Yuan, Q. Zhang, J. Dou, Catal. Commun. 11 (2010) 606.
[25] S. Shylesh, L. Wang, W.R. Thiel, Adv. Synth. Catal. 352 (2010) 425.
[26] M.J. Jacinto, P.K. Kiyohara, S.H. Masunaga, R.F. Jardim, L.M. Rossi, Appl. Catal. A:
Gen. 338 (2008).
[27] M.J. Jacinto, R. Landers, L.M. Rossi, Catal. Commun. 10 (2009) 1971.
[28] M.J. Jacinto, O. Santos, R.F. Jardim, R. Landers, L.M. Rossi, Appl. Catal. A: Gen.
360 (2009).
[29] M.J. Jacinto, O. Santos, R. Landers, P.K. Kiyohara, L.M. Rossi, Appl. Catal. B:
Environ. 90 (2009).
Acknowledgments
[30] L.M. Rossi, I.M. Nangoi, N.J.S. Costa, Inorg. Chem. 48 (2009) 4640.
[31] H. Mahalakshmi, P.P. Phadnis, V.K. Jain, E.R.T. Tiekink, Indian J. Chem. Sect. A:
Inorg., Bio-inorg., Phys., Theor., Anal. Chem. 43 (2004) 2287.
[32] J.J.M. de Pater, B.J. Deelman, C.J. Elsevier, G. van Koten, J. Mol. Catal. A: Chem.
258 (2006) 334.
The authors are grateful to the Brazilian agencies FAPESP,
CAPES, CNPq and CNPq/INCT-Catálise for financial support and fel-
lowships. We also thank the support from the International Coop-
eration Programs PICS-CNPq Brazil–France (Grant 490678/2008-4)
and CAPES-COFECUB (Grant 695/10).
[33] A. Scrivanti, U. Matteoli, V. Beghetto, S. Antonaroli, B. Crociani, Tetrahedron 58
(2002) 6881.
[34] S. Antonaroli, B. Crociani, J. Organomet. Chem. 560 (1998) 137.
[35] R.A. Komoroski, A.J. Magistro, P.P. Nicholas, Inorg. Chem. 25 (1986) 3917.
[36] R.H. Terrill, T.A. Postlethwaite, C.H. Chen, C.D. Poon, A. Terzis, A. Chen, J.E.
Hutchison, M.R. Clark, G. Wignall, J.D. Londono, R. Superfine, M. Falvo, C.S.
Johnson Jr., E.T. Samulski, R.W. Murray, J. Am. Chem. Soc. 117 (1995)
12537.
References
[1] R. Martin, S.L. Buchwald, Acc. Chem. Res. 41 (2008) 1461.
[2] A. Suzuki, J. Organomet. Chem. 576 (1999) 147.
[3] N. Miyaura, Top. Curr. Chem. 219 (2002) 11.
[37] A. Badia, W. Gao, S. Singh, L. Demers, L. Cuccia, L. Reven, Langmuir 12 (1996)
1262.
[4] S. Kotha, K. Lahiri, D. Kashinath, Tetrahedron 58 (2002) 9633.
[5] F. Bellina, A. Carpita, R. Rossi, Synthesis-Stuttgart (2004) 2419.
[6] Z.Q. Weng, S.H. Teo, T.S.A. Hor, Acc. Chem. Res. 40 (2007) 676.
[7] M.A. Garralda, CR Chim. 8 (2005) 1413.
[8] B. Crociani, S. Antonaroli, A. Marini, U. Matteoli, A. Scrivanti, Dalton Trans.
(2006) 2698.
[9] A. Scrivanti, M. Bertoldini, U. Matteoli, V. Beghetto, S. Antonaroli, A. Marini, B.
Crociani, J. Mol. Catal. A: Chem. 235 (2005) 12.
[10] A. Scrivanti, U. Matteoli, V. Beghetto, S. Antonaroli, R. Scarpelli, B. Crociani, J.
Mol. Catal. A: Chem. 170 (2001) 51.
[38] A. Badia, L. Cuccia, L. Demers, F. Morin, R.B. Lennox, J. Am. Chem. Soc. 119
(1997) 2682.
[39] M.J. Hostetler, J.E. Wingate, C.J. Zhong, J.E. Harris, R.W. Vachet, M.R. Clark, J.D.
Londono, S.J. Green, J.J. Stokes, G.D. Wignall, G.L. Glish, M.D. Porter, N.D. Evans,
R.W. Murray, Langmuir 14 (1998) 17.
[40] C. Pan, K. Pelzer, K. Philippot, B. Chaudret, F. Dassenoy, P. Lecante, M.J.
Casanove, J. Am. Chem. Soc. 123 (2001) 7584.
[41] E. Ramirez, S. Jansat, K. Philippot, P. Lecante, M. Gomez, A.M. Masdeu-Bulto, B.
Chaudret, J. Organomet. Chem. 689 (2004) 4601.