Magnetically recoverable AuPd nanoparticles prepared by a coordination capture method as a reusable catalyst for green oxidation of benzyl alcohol

Article abstract of DOI:10.1039/c3cy00261f

AuPd nanoparticles (AuPd NPs) have received considerable attention because of their importance as catalysts in selective aerobic oxidations; however, improved synthetic methods for making them reusable catalysts are necessary. Here, we present the prepara

Full text of DOI:10.1039/c3cy00261f

Catalysis
Science & Technology
View Article Online
View Journal | View Issue
PAPER
Magnetically recoverable AuPd nanoparticles prepared
by a coordination capture method as a reusable
catalyst for green oxidation of benzyl alcohol†
Cite this: Catal. Sci. Technol., 2013, 3,
2993
a
b
a
*
Tiago A. G. Silva, Richard Landers and Liane M. Rossi
AuPd nanoparticles (AuPd NPs) have received considerable attention because of their importance as catalysts in
selective aerobic oxidations; however, improved synthetic methods for making them reusable catalysts are
necessary. Here, we present the preparation of magnetically recoverable AuPd NPs by immobilizing pre-synthesized
PVA-stabilized AuPd NPs onto magnetic supports. The immobilization method is based on the functionalization of
the support with strong coordinating ligands, such as amino and thiol, for the attachment of AuPd NPs containing
weak coordinating groups. We were able to increase the metal loading by a coordination capture method with
functionalized supports, but the ligands grafted on the support surface affected the catalytic activity of AuPd NPs.
Moreover, the amino groups provided stabilization in solution, which made the catalyst separation after reaction
difficult. Finally, we obtained highly active and recyclable catalysts by removing the functional groups by a thermal
treatment. Negligible metal leaching and catalyst recycling in successive reactions are important features of the
new catalytic system.
Received 16th April 2013,
Accepted 26th June 2013
DOI: 10.1039/c3cy00261f
www.rsc.org/catalysis
Introduction
to benzaldehyde and esters in the presence of a base. The
requirement for a strongly alkaline medium for the oxidation
Aldehydes are valuable chemicals both as very active interme-
diates in organic synthesis and as high value components for
the perfume industry. They can be prepared by the oxidation
of primary alcohols, an important laboratory and commercial
procedure,1,2 traditionally performed using stoichiometric
oxidants that are major sources of toxic wastes in organic syn-
thesis. Replacement of stoichiometric oxidation methods by
heterogeneously catalyzed selective aerobic oxidation methods
prompted the need for the development of more efficient
solid catalysts.3 In recent years, the unexpectedly high activity
and selectivity of Au nanoparticle (Au NP) catalysts in different
reactions, for example oxidation reactions, have stimulated
intensive research into the use of Au NPs.4 In contrast, there
are few examples of gold-catalysed homogeneous oxidation
reactions.5 Rossi and Prati and co-workers6–8 were the first to
demonstrate that alcohols, in particular diols and sugars, can
be oxidized by Au NP catalysts to give various monoacids, as
long as a base is present. Aerobic oxidation and oxidative
esterification of benzyl alcohol are model reactions catalyzed
by supported Au NPs that also resulted in higher conversions
reaction to occur at acceptable rates may represent a consider-
able limitation to the use of Au NP catalysts, as side reactions
can lower the catalyst selectivity (keto–enol equilibration,
Cannizzaro reaction, oxidative decarbonylation), or the cata-
lyst/catalyst support can be attacked in basic media.9,10 We
recently prepared Au NPs supported on a silica-coated mag-
netic support, which exhibited clean, fast and efficient separa-
tion after chemical reactions by means of the application of
an external magnetic field.11 The catalyst was highly active in
the oxidation of alcohols and could be successfully separated
from the reaction products; however, the Au NPs suffered the
consequences of the use of a base, such as support corrosion
due to the action of the base, and the catalysts grew into
larger (∼80 nm) and less active or inactive particles. The ini-
tially very active catalyst lost activity gradually when used in
successive reactions.
In order to avoid the problems mentioned above caused
by the use of a base, we chose to adhere to the literature that
says that the combination of Au with a second metal from
the platinum group can avoid the use of a base and still give
high conversion and selectivity in the oxidation of alcohols.
A bimetallic catalyst, i.e. 1% Au–0.1% Pd/Al2O3, was used for
the transformation of 6-hydroxyhexanoic acid to adipic acid
in the absence of a base.12 Hutchings and co-workers13 have
shown that the alloying of gold with palladium leads to a
twenty-five-fold enhancement in activity in the oxidation of
a Departamento de Química Fundamental, Instituto de Química,
Universidade de São Paulo, Av. Prof. Lineu Prestes 748, São Paulo, 05508-000,
SP, Brazil. E-mail: lrossi@iq.usp.br; Fax: +55 11 38155579
b Instituto de Física, UNICAMP, Campinas, 13083-970, SP, Brazil
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3cy00261f
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 2993–2999 | 2993

View Article Online
Paper
Catalysis Science & Technology
alcohols compared to monometallic catalysts with the simul-
taneous retention of selectivity. The oxidation of alcohols by
AuPd NPs has received attention as an excellent alternative to
the monometallic catalysts.14–17
and polyvinyl alcohol (PVA) (6.0 mg) with 1.65 mL NaBH4
(0.1 mol L−1). The reaction was stirred for 30 min.
Synthesis of supported AuPd NPs
In this paper we report the immobilization of pre-
synthesized PVA-stabilized AuPd NPs on magnetically respon-
sive materials. In order to explore the potential utilization of
magnetic separation in the recovery of metal nanoparticles
from their reaction mixture after liquid phase catalysis, the
ability to attach them to the surfaces of a magnetic material
is an important issue. Methods to immobilize nanoparticles
on different surfaces are continuously being developed.18 The
metal nanoparticle–support linkage can occur on the basis of
electrostatic interactions or covalent bonding. In both cases,
nanoparticles and/or the support surface must be modified
with complementary ligands to achieve the appropriate link-
age between nanoparticles and the support. Ligands with var-
ious linkage capabilities can be readily introduced either on
the surface of the support (for example, by means of organo-
alkoxysilanes in silica) or into the metal NPs. Gold NPs can
be well supported on silica by an electrostatic assembly of
The catalyst support (500 mg) was added to 60 mL of water
and the colloidal AuPd NPs solution prepared as above. The
mixture was stirred at 25 °C for 2 h. The solid was then mag-
netically collected from the solution and washed with hot dis-
tilled water (2 × 25 mL) and ethanol (2 × 25 mL). The final
solid was air-dried for 10 h. Samples of the prepared catalysts
were calcined at 500 °C for 2 h.
Characterization
High-resolution transmission electron microscopy (HRTEM)
was performed at the Brazilian Nanotechnology National Labo-
ratory (LNNano, Campinas, SP) using a Jeol-3010 ARP micro-
scope. Samples for TEM observations were prepared by placing
a drop containing the nanoparticles on a carbon-coated copper
grid. The histograms of the gold nanoparticle size distribution,
assuming spherical shape, were obtained from the measure-
ment of the diameters of about 300 particles found in arbi-
trarily chosen areas of enlarged micrographs. The samples
were observed in different regions of the Cu grid.
The XPS spectra were obtained at the Instituto de Física
(IF-UNICAMP) with a VSW HA-100 spherical analyzer using
an aluminum anode (AlKα line, hν = 1486.6 eV) X-ray source.
The high-resolution spectra were measured with a constant
analyzer pass energy of 44 eV, which produces a FWHM line
width of 1.7 eV for the Au 4f7/2 line. The powdered samples
were pressed into pellets and fixed onto a stainless steel sam-
ple holder with a double-faced tape and analyzed without fur-
ther preparation. To correct for charging effects, the spectra
were shifted so that the Si 2p binding energy in SiO2 was
103.5 eV. Curve fitting was performed by using Gaussian line
shapes, and a Shirley-type background was subtracted from
the data. The Pd/Au molar ratios were calculated using the
Pd(3d5/2) and Au(4d5/2) XPS peak areas corrected using the
corresponding Scofield cross-section values of 9.48 and 11.74,
respectively.23
amino-modified silica spheres and citrate-coated Au NPs.19
A
modification of this method is based on the functionalization
of the support with strong coordinating ligands and the
impregnation of NPs synthesized with weak coordinating
groups (for example, poly(vinyl-pyrrolidone) (PVP) or polyvinyl
alcohol (PVA)), called the coordination capture method. The
metal NPs are captured due to the affinity of ligands present
in the functionalized support for the metal NPs. Khatri et al.20
studied the deposition of ionic liquid-stabilized Au NPs on
3-mercaptopropyl-functionalized silica. The authors observed
that the weakly coordinated ionic liquid molecules could be,
to some extent, replaced with the thiol groups of the function-
alized support. The Au–S chemisorption linkage resulted in
strong immobilization of the Au NPs. This method can, in
principle, be applied for the preparation of supported metal
NPs for catalytic applications, but the effect of the functional
groups grafted to the support on the catalytic performance
has not been described.
Experimental
Synthesis of magnetite and silica-coated magnetite NPs
Catalytic experiments
The synthesis of oleic acid-coated Fe3O4 NPs and the core–shell
Fe3O4@SiO2 composite has been reported elsewhere.21 The
solid was modified with amino groups by a reaction with
3-aminopropyltriethoxysilane (APTES) and with thiol groups by
a reaction with 3-mercaptopropyltrimethoxysilane in dry toluene
under N2, to give the amino-functionalized (Fe3O4@SiO2-NH2)
and thiol-functionalized (Fe3O4@SiO2-SH) supports. The final
solid was washed with toluene and dried at 100 °C for 20 h.
The oxidation reactions were performed using a modified
Fischer–Porter 100 mL glass reactor. In a typical solventless
reaction, the glass reactor was loaded with the supported
AuPd-catalyst (75 mg) and benzyl alcohol (9.6 mmol). The
reactor, immersed in an oil bath, was loaded with O2 to the
desired pressure. The temperature was maintained by an oil
bath placed on a hot stirring plate connected to a digital con-
troller (ETS-D5 IKA). The reactions were conducted under
magnetic stirring, using a Teflon-coated magnetic stir bar,
for the desired time. The catalyst was magnetically recovered
by placing a permanent magnet on the reactor wall. The
products were collected using a syringe and analyzed by gas
chromatography (GC) using p-xylene as the standard. The
Synthesis of AuPd NPs
Gold–palladium NPs were synthesized as described by
Hutchings and co-workers22 by the reduction of 40 mL of a solu-
tion containing HAuCl4 (16.5 μmol Au), PdCl2 (16.5 μmol Pd)
2994
|
Catal. Sci. Technol., 2013, 3, 2993–2999
This journal is © The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
Table 1 Metal loading in magnetically responsive supports
isolated catalyst could be reused when a new amount of sub-
strate was added.
Support
Au (wt%) Pd (wt%) Au : Pda Au + Pd (mmol g−1
)
Fe3O4
Fe3O4@SiO2
Fe3O4@SiO2-NH2 0.57
Fe3O4@SiO2-SH 0.51
0.65
0.20
0.34
0.10
0.30
0.27
1 : 1
∼1 : 1
1 : 1
0.065
0.020
0.057
0.051
Results and discussion
Catalyst preparation method and characterization
1 : 1
a
PVA-stabilized AuPd NPs were synthesized via NaBH4 reduc-
tion of aqueous solutions containing a 1 : 1 molar ratio of
Au(III) and Pd(II) metal salts in the presence of PVA. This poly-
mer is known to bind weakly to metal nanoparticles, such that
the reactants can access the nanoparticle surface in the pres-
ence of the stabilizer. Then, the AuPd NPs were immobilized
on magnetically responsive supports, comprised of silica-
coated magnetite NPs,21 to give an additional property to the
catalyst so that it can be easily recovered magnetically. This
special support can be attracted to an external magnetic field;
however, it will be able to carry the metal nanoparticles only if
they are firmly attached to the support surface. In order to
improve the catalyst–support interaction, the magnetic NPs
were coated with a layer of silica and then functionalized with
strong coordinating ligands, so the PVA-stabilized AuPd NPs
were immobilized on the magnetically responsive material by
a coordination capture method (Fig. 1). Silica was chosen
because it is a very versatile support material, and displays
some advantageous properties, such as easy functionalization
of the surface silanol groups, which can react with various
organo-trialkoxysilanes to attach organic groups to silica sur-
face. The presence of such organic groups on the support sur-
faces has been shown to positively impact catalyst stability by
improving metal–support interactions,21 catalyst metal load-
ing by impregnation,24 and catalyst activity by tuning the size
of metal NPs and their activity in hydrogenation reactions.25
However, their influence on catalysts for selective aerobic oxi-
dations has not been established.
Molar ratio.
the uptake of the metal from the colloidal solution and
allowed the preparation of catalysts with ca. 1 wt% AuPd.
The size and size distribution of the bimetallic AuPd
NPs were assessed by the analysis of TEM micrographs such
as those presented in Fig. 2. The TEM image of the
Fe3O4@SiO2-NH2–AuPd NPs shown in Fig. 2A revealed the
morphology of the catalyst support comprised of a core–shell
nanostructure of magnetite coated with silica and AuPd NPs
of 5.1 1.2 nm in size decorating the support surface. The
composition of the bimetallic NPs analyzed by FAAS revealed
molar ratios close to 1 : 1 Au : Pd (Table 1); however it is not
an easy task to discern whether AuPd NPs are random alloys
or core–shell nanostructures. According to the methodology
chosen for the solution synthesis of the bimetallic AuPd NPs,
we can expect to obtain AuPd alloy NPs.26 Analysis of the
HRTEM micrographs of supported AuPd NPs, such as those
presented in Fig. 2B, revealed a lattice spacing of 2.31 Å,
which can be attributed to the AuPd(111) plane. This value is
between the lattice spacing of Au(111) at 2.36 Å and Pd(111)
at 2.23 Å. Monometallic planes were not found in the many
NPs analyzed, suggesting that segregation of metals did not
occur.
Initially, an aqueous solution containing pre-synthesized
AuPd NPs was added to non-functionalized and functional-
ized supports in order to evaluate the impregnation effi-
ciency. After 2 h, the solids were isolated magnetically and
the metal loading on the supports was measured by FAAS, as
shown in Table 1. The non-functionalized support was loaded
with 0.3 wt% AuPd, and the functionalized solids improved
Fig. 1 Preparation of supported AuPd NP catalysts by a coordination capture
Fig. 2 (A) TEM image of AuPd NPs supported on Fe3O4@SiO2-NH2. (B) HRTEM
method.
micrograph of one AuPd NP with a lattice spacing of 2.31 Å.
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 2993–2999 | 2995

View Article Online
Paper
Catalysis Science & Technology
with increasing temperature. Decreasing the oxygen pressure
from 6 to 2 bar O2 had a strong influence on the rate of the
reaction, which dropped from 92 to 48% conversion, and
consequently the selectivity towards benzaldehyde increased.
In view of this, reactions were carried out at a temperature of
100 °C and 6 bar in order to increase conversion to the
desired benzaldehyde product (Table 2, entry 3). The benzyl
alcohol conversion curve and the product distribution over
time are shown in Fig. 3. Benzaldehyde was the major product
of benzyl alcohol oxidation with the AuPd NP catalyst, but
after 80% conversion, the selectivity for benzaldehyde decreased
with the formation of benzoic acid.
Scheme 1 Reaction pathway for the oxidation of benzyl alcohol.
Benzyl alcohol oxidation
Influence of the catalyst support
The catalytic oxidation of benzyl alcohol was used to probe
the activity, recovery and recyclability of the supported AuPd
NP catalysts. The possible products of the oxidation of benzyl
alcohol are summarized in Scheme 1.
The effects of reaction temperature and oxygen pressure
were investigated, and the results obtained with the
Fe3O4@SiO2-NH2–AuPd catalyst are shown in Table 2. Increas-
ing the temperature from 60 to 100 °C showed a marked
effect on the rate of the reaction, which changed from about
4% to 92% conversion, while only a minor effect on the selec-
tivity to benzaldehyde was seen, decreasing from 100 to 90%
As mentioned before, functionalized supports improved the
uptake of metal NPs from their colloidal solution (the coordi-
nation capture method) and allowed the preparation of cata-
lysts loaded with three times more metal (Au + Pd) than the
non-functionalized support. In principle, we realized that the
functional groups grafted on the support surface allowed a
more efficient adsorption of metal NPs, probably due to a
stronger interaction of the AuPd NPs with the functionalized
support than with the silanol groups of the non-functionalized
support. At this point, we questioned if those functional groups
grafted on the support could influence the catalytic properties
of supported metal NPs. We decided to investigate how differ-
ent ligands might affect the catalytic activity of supported AuPd
NPs, again using benzyl alcohol oxidation to probe the activity,
recovery and recyclability of the supported AuPd NP catalysts.
We observed that the presence of functional groups had a
major influence on the stability and reusability of the catalysts.
The catalysts prepared with functionalized supports containing
–NH2 groups (Fe3O4@SiO2-NH2–AuPd) exhibited higher activity
and stability towards catalyst recycling (Table 3, entries 1–5)
than the functionalized support containing –SH groups
(Fe3O4@SiO2-SH–AuPd) and the non-functionalized catalyst.
The catalyst prepared with the non-functionalized support
Fe3O4@SiO2–AuPd was less active (Table 3, entries 8 and 9)
and exhibited metal leaching during recycling.
Table 2 Oxidation of benzyl alcohol by Fe3O4@SiO2-NH2–AuPd
Selectivityc (%)
Entrya
P O2 (bar) T (°C) Conv.b (%)
I
II
III
IV
1
2
3
4
5
6
6
6
4
2
60
80
100
100
100
3.8
17.3
92.1
82.5
48.5
—
—
<1
<1
<1
100
—
—
97.1 1.9 1.0
90.1 7.2 1.3
95.5 2.0 1.9
96.8 1.8 1.0
a
Reaction conditions: benzyl alcohol (9.6 mmol), 75 mg catalyst
(4.3 × 10−3 mmol Au + Pd), 2.5 h. b Determined by GC. c I refers to toluene,
II refers to benzaldehyde, III refers to benzoic acid and IV refers to
benzyl benzoate.
Table 3 Oxidation of benzyl alcohol by bimetallic AuPd supported on Fe3O4@SiO2
with different functional groups
Selectivityc (%)
Functional
Entrya groups
Conversionb
Cycle (%)
I
II
III
IV
1.9
1.6
1.5
2.2
1.8
2.3
1
2
3
4
5
6
7
8
9
NH2
NH2
NH2
NH2
NH2
SH
SH
—
—
1
2
3
4
5
1
2
1
2
62.7
49.3
41.4
28.4
33.8
73.6
13.7
1.0
<1
<1
<1
<1
<1
85.6 12.6
86.6 11.8
85.1 13.4
74.8 23.0
80.8 17.5
5.8 69.8 22.1
28.1 71.9 <1
<1
<1
<1
70.7 29.2 <1
44.5 55.5 <1
1.0
a
b
Fig. 3 Effect of the reaction time on the conversion and selectivity in the solvent-
free oxidation of benzyl alcohol (at 100 °C and 6 bar O2) over the Au–Pd magnetic
catalyst.
Reaction conditions: benzyl alcohol (9.6 mmol), 75 mg catalyst, 2.5 h.
Determined by GC. c I refers to toluene, II refers to benzaldehyde, III
refers to benzoic acid and IV refers to benzyl benzoate.
2996
|
Catal. Sci. Technol., 2013, 3, 2993–2999
This journal is © The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
and strong vibrational bands of Si–OH, Si–O, and Si–O–Si typi-
cal of amorphous silica that overlay the features corresponding
to ν(CN) at ∼1630 cm−1. We also tried to follow the strong
absorption expected for imine groups at ∼400 cm−1 by UV-Vis
spectrometry, but again the observation in hindered by the
absorption of the Fe3O4@SiO2 support. A set of experiments
using silica (without iron oxide on it) clearly showed that
amino-functionalized silica reacts with benzaldehyde to give a
yellow solid with a strong UV-Vis absorption at 400 cm−1,
confirming the hypothesis of a reaction of free amino groups
with the reaction product (benzaldehyde) is plausible. In order
to reinforce the presence of amino groups as the main reason
for stabilization and difficult separation of the catalysts after
the reaction, these groups were removed by calcination. Ele-
mental CHN analysis of the pristine Fe3O4@SiO2-NH2–AuPd
catalyst (0.34 %N, 2.85 %C, 1.53 %H) and of the calcined
Fe3O4@SiO2-NH2–AuPd(C) catalyst (0.34 %N, 2.85 %C, 1.53 %H)
revealed a marked decrease in the organic portion of the cata-
lyst upon calcination, which suggests that most of the amino
groups were removed. Consequently, the metal content deter-
mined by FAAS (Au = 0.78 wt% and Pd = 0.37 wt%) increased
to 0.074 mmol (Au + Pd) g−1, but the 1 : 1 Au : Pd molar ratio
was maintained.
The calcined Fe3O4@SiO2-NH2–AuPd(C) catalyst was
used in the oxidation of benzyl alcohol under similar condi-
tions. The results are summarized in Table 4, together
with the catalytic results for Fe3O4@SiO2–AuPd(C) and
Fe3O4@SiO2-SH–AuPd(C) after calcination. The first important
observation is that, after the removal of –NH2 groups, the cata-
lyst does not react with the reaction product benzaldehyde, as
before, which resulted in fast and highly efficient catalyst recov-
ery. Moreover, the calcined catalyst Fe3O4@SiO2-NH2–AuPd(C)
was highly active and selective (88% benzaldehyde at 85%
conversion), and was recyclable for up to five successive reac-
tions without deactivation or changes in selectivity (Table 4,
entries 1–5). The Fe3O4@SiO2-SH–AuPd(C) catalyst was also
highly active but was deactivated in the second reaction (Table 4,
entries 6 and 7), while the Fe3O4@SiO2–AuPd(C) catalyst did
not show any activity after calcination (Table 4, entries 8 and 9).
Fig. 4 Schematic representation of stabilizer exchange by the addition of
1-octanethiol.
Moreover, differences in the catalytic behavior of
Fe3O4@SiO2-SH–AuPd and Fe3O4@SiO2-NH2–AuPd were noted
in the conversion of benzyl alcohol under similar reaction
conditions. The Fe3O4@SiO2-NH2–AuPd catalyst exhibited
62.7% conversion of benzyl alcohol in the first reaction and
then a steady decrease in activity in the subsequent recycles,
although the selectivity was maintained (Table 3, entries 1–5).
The Fe3O4@SiO2-SH–AuPd catalyst exhibited 73.6% conver-
sion of benzyl alcohol in the first reaction, then a sudden
decrease in activity in the subsequent reaction cycle (Table 3,
entries 6 and 7). This catalyst with SH-groups formed more
toluene than the others, which suggests the disproportion-
ation of benzyl alcohol. The presence of thiol groups, known
to poison gold catalyst surfaces, may have been responsible
for blocking the activity of supported AuPd NPs upon recycling.
A similar effect was obtained by the addition of 5 equiv. of
1-octanethiol to our catalyst during the oxidation of benzyl
alcohol, which irreversibly reduced the catalytic activity (Fig. 4).
Major differences were observed in the product distri-
bution when comparing Fe3O4@SiO2-NH2–AuPd and
Fe3O4@SiO2-SH–AuPd: (i) benzaldehyde was the main prod-
uct, but a higher selectivity of 85.6% was achieved with
the amino-functionalized catalyst compared to the thiol-
functionalized catalyst (69.8%); (ii) benzoic acid was formed
by both catalysts, but this product was formed in larger
amounts by the thiol-functionalized catalyst (22%) than by
the amino-functionalized catalyst (12.6%); (iii) toluene forma-
tion was negligible with the amino-functionalized catalyst,
but reached 6% to 28% with the thiol-functionalized catalyst.
All these observations regarding conversion and selectivity
suggest that the amino-functionalized catalyst is more selec-
tive for the desired benzaldehyde product and produces fewer
unwanted by-products, for example toluene. However, the
decrease in activity in successive reaction cycles revealed the
need for further studies on this system. It is worth mention-
ing that catalyst recovery by placing a permanent magnet
in the reactor wall was slower and less efficient for the
Fe3O4@SiO2-NH2–AuPd catalyst after the first reaction than for
Fe3O4@SiO2-SH–AuPd, indicating changes at the material surface
that make it more stable in solution after the reaction. A plausi-
ble reaction that can take place on Fe3O4@SiO2-NH2–AuPd, but
not Fe3O4@SiO2-SH–AuPd, is the formation of an aldimine by
the reaction of NH2 groups grafted onto the catalyst support
with the benzaldehyde formed during the oxidation of benzyl
alcohol. Unfortunately, we could not follow the formation of
–CN bonds by FT-IR because Fe3O4@SiO2 contains broad
Table 4 Oxidation of benzyl alcohol by bimetallic AuPd supported catalysts
after calcination
Selectivityc (%)
Entrya Catalyst Cycle Conversionb (%)
I
II
III
IV
2.6
2.5
2.9
2.8
2.8
4.7
2.5
d
1
2
3
4
5
6
7
8
9
1
2
3
4
5
1
2
1
2
85.3
73.5
72.1
59.2
80.4
86.7
39.6
2.5
<1
88.4 8.3
d
d
d
d
e
e
f
1.0 92.6 3.9
<1
2.5 91.0 3.7
94.9 1.6
<1
92.6 3.9
1.2 77.8 16.3
3.0 91.6 2.9
<1
<1
19.0 81.0 <1
7.3 92.7 <1
f
3.6
a
b
Reaction conditions: benzyl alcohol (9.6 mmol), 75 mg catalyst, 2.5 h.
Determined by GC. I refers to toluene, II refers to benzaldehyde,
c
III refers to benzoic acid and IV refers to benzyl benzoate. d Fe3O4@SiO2-
NH2–AuPd(C). e Fe3O4@SiO2-SH–AuPd(C). f Fe3O4@SiO2–AuPd(C).
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 2993–2999 | 2997

View Article Online
Paper
We anticipated that thermal treatment would not change
Catalysis Science & Technology
revealed the contribution of each metal. The Pd(3d5/2) and
the Au(4d5/2) peaks of the calcined Fe3O4@SiO2-NH2–AuPd(C)
catalyst (Fig. 5B) exhibit a significant chemical shift to higher
binding energies when compared with the uncalcined catalyst
(Fig. 5A); this can be attributed to the oxidation of Au and Pd
species upon calcination. However, after the calcined catalyst
was reused in five successive reactions, the XPS peaks shifted
back to lower binding energies, suggesting the reduction of
Au and Pd species (Fig. 5C). Another important feature in the
XPS spectra in Fig. 5A and B is the enrichment of Pd on the
surface of the AuPd NPs. The Pd/Au molar ratio increased
from 0.72 to 1.78 after calcination. Additionally, the Pd/Au
molar ratio of the reused catalyst was maintained at 1.58
(Fig. 5C). The XPS results indicate that Pd surface segregation
occurs during calcination, producing a Pd-rich shell on the
AuPd NPs, which is in agreement with the literature,27 and
the catalyst retained the Pd surface enrichment upon
recycling. The XPS analysis and the catalyst recycling results
corroborate the high stability of the catalyst upon recycling.
As far as we know, this is the first magnetically recover-
able AuPd NP catalyst reported, even though other examples
of catalysts for the selective oxidation of alcohols that are
also magnetically retrievable can be found in the literature,
for example those based on monometallic AuNPs.11,24,28,29
The main advantage of AuPd catalysts is their high catalytic
activity and selectivity in the absence of a base, which increases
the catalyst's life span.
the magnetic response of the support, which was magneti-
cally recovered before and after calcination using the same
permanent magnet. The size and size distribution of the
bimetallic AuPd NPs in the calcined catalyst before use,
Fe3O4@SiO2-NH2–AuPd(C), and the calcined catalyst after
five reuses in successive oxidations of benzyl alcohol,
Fe3O4@SiO2-NH2–AuPd(C)-R5 were assessed by the analysis of
TEM micrographs (Fig. S2 and S3, ESI†). The core–shell nano-
structure of the catalyst support Fe3O4@SiO2 was maintained
after calcination and the mean size of the AuPd NPs increased
from 5.1
1.2 nm (before calcination – Fig. S1, ESI†) to
6.8 nm (σ = 0.38) (after calcination – Fig. S2, ESI†) and to
9.3 nm (σ = 0.45) (after five reuses – Fig. S3, ESI†). Additionally,
the recycled catalyst showed signs of support agglomeration
and the AuPd NPs grew at a moderate rate when compared
with the monometallic Au NPs reported previously.11
Possible changes in the catalysts after thermal treatment
and after recycling were investigated by X-ray photoelectron
spectroscopy (XPS). The Pd(3d)/Au(4d) XPS spectra for
the pristine Fe3O4@SiO2-NH2–AuPd catalyst, the calcined
Fe3O4@SiO2-NH2–AuPd(C) catalyst before use and the cal-
cined catalyst after five reuses in successive oxidations of
benzyl alcohol, Fe3O4@SiO2-NH2–AuPd(C)-R5, are shown in
Fig. 5. Analysis of the Pd(3d5/2) spectra was complicated
by the overlap of the Pd(3d5/2) doublet and the Au(4d5/2)
component. However, deconvolution of the XPS spectra
Conclusions
Pre-synthesized AuPd NPs stabilized by polyvinyl alcohol
(PVA) were used as precursors to prepare well-defined cata-
lysts, and silica-coated magnetite was used as a magnetically
recoverable support. The main issue regarding the preparation
of magnetically recoverable metal nanoparticle is obtaining a
strong metal–support interaction to avoid nanoparticles coa-
lescence, segregation and metal leaching during liquid phase
reactions, while maintaining the catalytic properties of the
selected nanoparticles. In order to attach the NPs to the sup-
port and improve metal–support interactions, we used a coor-
dination capture method based on the functionalization of
the support surface with ligands, such as amino and thiol
groups, to provide coordination sites and keep the nanoparti-
cles attached to the support. However, the ligands grafted
onto the support surface affected the catalytic activity and
selectivity of the supported AuPd NPs. Additionally, the amino
groups provided stabilization in solution, but made it difficult
to separate the catalyst after the reaction. We then demon-
strated that catalytic activity and catalyst stability towards
recycling could be improved upon removal of the functional
groups by thermal treatment.
Acknowledgements
Fig. 5 Au(4d) and Pd(3d) X-ray photoelectron spectra of the Fe3O4@SiO2-NH2–AuPd
catalyst after different heat treatments: (A) uncalcined, (B) calcined at 500 °C and
(C) the calcined catalyst after five reuses in successive oxidations of benzyl alcohol.
The authors are grateful to the Brazilian agencies FAPESP
and CNPq for financial support and indebted to the Brazilian
2998
|
Catal. Sci. Technol., 2013, 3, 2993–2999
This journal is © The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
Nanotechnology National Laboratory (LNNano, Campinas,
Brazil) for HRTEM images and EDS analysis. The authors also
acknowledge the Instituto Nacional de Ciência e Tecnologia
de Catálise em Sistemas Moleculares e Nanoestruturados
(INCT-CMN).
14 P. Miedziak, M. Sankar, N. Dimitratos, J. A. Lopez-Sanchez,
A. F. Carley, D. W. Knight, S. H. Taylor, C. J. Kiely and
G. J. Hutchings, Catal. Today, 2011, 164, 315–319.
15 P. J. Miedziak, Q. He, J. K. Edwards, S. H. Taylor,
D. W. Knight, B. Tarbit, C. J. Kiely and G. J. Hutchings,
Catal. Today, 2011, 163, 47–54.
16 H. Zhang, Y. Xie, Z. Sun, R. Tao, C. Huang, Y. Zhao and
Z. Liu, Langmuir, 2010, 27, 1152–1157.
17 K. Kaizuka, H. Miyamura and S. Kobayashi, J. Am. Chem.
Soc., 2010, 132, 15096–15098.
18 C.-J. Jia and F. Schueth, Phys. Chem. Chem. Phys., 2011, 13,
2457–2487.
19 F. Osterloh, H. Hiramatsu, R. Porter and T. Guo, Langmuir,
2004, 20, 5553–5558.
Notes and references
1 R. A. Sheldon and J. K. Kochi, Metal-catalyzed oxidations of
organic compounds: mechanistic principles and synthetic
methodology including biochemical processes, Academic Press,
New York, 1981.
2 G. J. ten Brink, I. W. C. E. Arends and R. A. Sheldon, Science,
2000, 287, 1636–1639.
20 O. P. Khatri, K. Adachi, K. Murase, K.-i. Okazaki,
T. Torimoto, N. Tanaka, S. Kuwabata and H. Sugimura,
Langmuir, 2008, 24, 7785–7792.
3 C. P. Vinod, K. Wilson and A. F. Lee, J. Chem. Technol.
Biotechnol., 2011, 86, 161–171.
4 A. S. K. Hashmi and G. J. Hutchings, Angew. Chem., Int. Ed.,
2006, 45, 7896–7936.
21 M. J. Jacinto, P. K. Kiyohara, S. H. Masunaga, R. F. Jardim
and L. M. Rossi, Appl. Catal., A, 2008, 338, 52–57.
22 J. A. Lopez-Sanchez, N. Dimitratos, P. Miedziak, E. Ntainjua,
J. K. Edwards, D. Morgan, A. F. Carley, R. Tiruvalam,
C. J. Kiely and G. J. Hutchings, Phys. Chem. Chem. Phys.,
2008, 10, 1921–1930.
23 J. H. Scofield, J. Electron Spectrosc. Relat. Phenom., 1976, 8,
129–137.
24 R. L. Oliveira, D. Zanchet, P. K. Kiyohara and L. M. Rossi,
Chem.–Eur. J., 2011, 17, 4626–4631.
5 A. S. K. Hashmi, C. Lothschütz, M. Ackermann, R. Doepp,
S. Anantharaman, B. Marchetti, H. Bertagnolli and
F. Rominger, Chem.–Eur. J., 2010, 16, 8012–8019.
6 F. Porta, L. Prati, M. Rossi, S. Coluccia and G. Martra, Catal.
Today, 2000, 61, 165–172.
7 L. Prati and G. Martra, Gold Bull., 1999, 32, 96–101.
8 L. Prati and M. Rossi, J. Catal., 1998, 176, 552–560.
9 S. Carrettin, P. McMorn, P. Johnston, K. Griffin, C. J. Kiely
and G. J. Hutchings, Phys. Chem. Chem. Phys., 2003, 5,
1329–1336.
25 L. M. Rossi, I. M. Nangoi and N. l. J. S. Costa, Inorg. Chem.,
2009, 48, 4640–4642.
10 S. Biella, L. Prati and M. Rossi, Inorg. Chim. Acta, 2003, 349,
253–257.
11 R. L. Oliveira, P. K. Kiyohara and L. M. Rossi, Green Chem.,
2010, 12, 144–149.
26 N. Dimitratos, J. A. Lopez-Sanchez, J. M. Anthonykutty,
G. Brett, A. F. Carley, R. C. Tiruvalam, A. A. Herzing,
C. J. Kiely, D. W. Knight and G. J. Hutchings, Phys. Chem.
Chem. Phys., 2009, 11, 4952–4961.
12 N. Herron, S. Schwarz and J. D. Druliner, WO Pat., 16298-A1,
2002.
13 D. I. Enache, J. K. Edwards, P. Landon, B. Solsona-Espriu,
A. F. Carley, A. A. Herzing, M. Watanabe, C. J. Kiely,
D. W. Knight and G. J. Hutchings, Science, 2006, 311,
362–365.
27 A. A. Herzing, A. F. Carley, J. K. Edwards, G. J. Hutchings
and C. J. Kiely, Chem. Mater., 2008, 20, 1492–1501.
28 F. Mi, X. Chen, Y. Ma, S. Yin, F. Yuan and H. Zhang, Chem.
Commun., 2011, 47, 12804–12806.
29 H. W. Chen, A. Murugadoss, T. S. A. Hor and H. Sakurai,
Molecules, 2010, 16, 149–161.
This journal is © The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 2993–2999 | 2999