Size-dependent catalytic activity of supported palladium nanoparticles for aerobic oxidation of alcohols

Article abstract of DOI:10.1002/adsc.200700350

Silica-alumina (SiO₂-Al₂O₃)-supported palladium catalysts prepared by adsorption of the tetrachloropalladate anion (PdCl₄^2-) followed by calcination and reduction with either hexanol or hydrogen were studied for the aerobic oxidation of alcohols. The mean size of the Pd particles over the SiO₂-Al₂O ₃ support was found to depend on the Si/Al ratio, and a decrease in the Si/Al ratio resulted in a decrease in the mean size of the Pd nanoparticles. By changing the Si/Al ratio, we obtained supported Pd nanoparticles with mean sizes ranging from 2.2 to 10 nm. The interaction between the Pd precursor and the support was proposed to play a key role in tuning the mean size of the Pd nanoparticles. The Pd/SiO₂-Al₂O₃ catalyst with an appropriate mean size of Pd particles could catalyze the aerobic oxidation of various alcohols to the corresponding carbonyl compounds, and this catalyst was particularly efficient for the solvent-free conversion of benzyl alcohol. The intrinsic turnover frequency per surface Pd atom depended significantly on the mean size of Pd particles and showed a maximum at a medium mean size (3.6-4.3 nm), revealing that the aerobic oxidation of benzyl alcohol catalyzed by the supported Pd nanoparticles was structure-sensitive.

Full text of DOI:10.1002/adsc.200700350

FULL PAPERS
DOI: 10.1002/adsc.200700350
Size-Dependent Catalytic Activity of Supported Palladium
Nanoparticles for Aerobic Oxidation of Alcohols
Jing Chen,a Qinghong Zhang,a,* Ye Wang,a and Huilin Wana,*
a
State Key Laboratory of Physical Chemistry of Solid Surfaces and Department of Chemistry, College of Chemistry and
Chemical Engineering, Xiamen University, Xiamen 361005, Peopleꢀs Republic of China
Fax : (+86)-592-218-3047; e-mail: zhangqh@xmu.edu.cn or hlwan@xmu.edu.cn
Received: July 17, 2007; Revised: December 29, 2007; Published online: February 11, 2008
Abstract: Silica-alumina (SiO2-Al2O3)-supported pal- size of Pd particles could catalyze the aerobic oxida-
ladium catalysts prepared by adsorption of the tetra- tion of various alcohols to the corresponding carbon-
chloropalladate anion (PdCl42À) followed by calcina- yl compounds, and this catalyst was particularly effi-
tion and reduction with either hexanol or hydrogen cient for the solvent-free conversion of benzyl alco-
were studied for the aerobic oxidation of alcohols. hol. The intrinsic turnover frequency per surface Pd
The mean size of the Pd particles over the SiO2- atom depended significantly on the mean size of Pd
Al2O3 support was found to depend on the Si/Al particles and showed a maximum at a medium mean
ratio, and a decrease in the Si/Al ratio resulted in a size (3.6–4.3 nm), revealing that the aerobic oxida-
decrease in the mean size of the Pd nanoparticles. tion of benzyl alcohol catalyzed by the supported Pd
By changing the Si/Al ratio, we obtained supported nanoparticles was structure-sensitive.
Pd nanoparticles with mean sizes ranging from 2.2 to
10 nm. The interaction between the Pd precursor
and the support was proposed to play a key role in Keywords: aerobic oxidation; alcohols; heterogene-
tuning the mean size of the Pd nanoparticles. The ous catalysis; palladium; solvent-free reactions;
Pd/SiO2-Al2O3 catalyst with an appropriate mean structure-activity relationships
Introduction
ing performances for the aerobic oxidation of alco-
hols. Some of these catalysts were heterogenized ana-
The selective oxidation of alcohols to the correspond- logues of homogeneous Pd complexes. For example,
ing carbonyl compounds is one of the most essential the Pd(II) acetate-pyridine complex supported on hy-
transformations in the synthesis of fine chemicals and drotalcite was found to be effective for the aerobic
intermediates.[1] Currently, many such transformations oxidation of various alcohols in the presence of pyri-
are still being carried out using stoichiometric oxi- dine in toluene as solvent, and the catalyst could be
dants such as dichromate and permanganate. Howev- reused several times.[4] A 1,10-phenanthroline ligand-
er, these stoichiometric oxidants are expensive and/or preserved Pd cluster (ca. 3 nm in size) immobilized
toxic, and have the problem of producing a large on TiO2 catalyzed the oxidation of primary alcohols
amount of wastes. From the viewpoints of green by oxygen in acetic acid.[5] A bipyridylamide-bound
chemistry, there is an urgent need to develop a cata- ionic Pd species located inside the mesoporous chan-
lytic oxidation process using oxygen or air as the oxi- nels of SBA-15 could catalyze the aerobic oxidation
dant in place of the stoichiometric metal oxidants. A of alcohols in the presence of K2CO3 in toluene,[6] but
number of homogeneous catalysts including palladiu- it was also pointed out that the derived Pd nanoparti-
m(II) salts or complexes has been investigated for the cles may also account for the actual activity.
aerobic oxidation of alcohols.[2] As compared with the
Simple heterogeneous Pd catalysts without any ad-
homogeneous catalysts, heterogeneous catalysts gen- ditives and organic ligands have attracted particular
erally possess advantages in product isolation and cat- interest for the aerobic oxidation of alcohols in recent
alyst recycling operations, and have received much at- years. Mori et al.[7] reported an efficient alcohol oxi-
tention for the aerobic oxidation of alcohols in recent dation catalyst based on a hydroxyapatite-immobi-
years.[3]
lized Pd(II) catalyst. Benzylic alcohols could even be
Among various heterogeneous catalysts reported, oxidized with this catalyst under solvent-free condi-
palladium-based catalysts showed particularly promis- tions, and the turnover frequency (TOF) for the oxi-
Adv. Synth. Catal. 2008, 350, 453 – 464
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
453

Jing Chen et al.
FULL PAPERS
dation of 1-phenylethanol reached 9800 hÀ1, markedly ence in the Pd nanoparticles-catalyzed aerobic oxida-
higher than both the heterogeneous and the homoge- tion of alcohols. The preparation of Pd nanoparticles
neous catalysts reported previously. An induction with controllable mean sizes is still a challenging task.
period was observed during the oxidation of 1-phenyl- Generally, Pd nanoparticles can be prepared by re-
ethanol over this catalyst, and it was demonstrated duction of ionic or oxidic Pd species with a proper re-
that the monomeric Pd(II) species were transformed ductant such as hydrogen or alcohol. The presence of
to Pd nanoparticles (3–4 nm) during the reaction, a stabilizer such as linear polymers, surfactants or het-
which were proposed to be the genuine active spe- erogeneous supports may prevent the nanoparticles
cies.[7b] Grunwaldt et al.[8] recently clarified that metal- from aggregation.[23] For the reduction with alcohols,
lic Pd is more active for the aerobic oxidation of previous studies showed that the kind of alcohols
benzyl alcohol than the oxidic Pd species using in situ used for reduction could control the mean size of Pd
X-ray absorption spectroscopy. Although
a
few nanoparticles finally obtained.[7b,23] Very recently, we
papers report that palladium oxide can act as the succeeded in preparing silica-alumina (SiO2-Al2O3)-
active phase for alcohol oxidation,[9,10] most of recent supported Pd nanoparticles with tunable mean sizes.
studies have shown that the supported Pd nanoparti- We found that, by varying the ratio of Si/Al in the
cles are excellent catalysts for the selective oxidation oxide support, catalysts containing Pd nanoparticles
of alcohols by oxygen or air.[11–15] For example, Pd/ with mean sizes of 2.2 to 10 nm could be prepared
MgO containing Pd particles with a mean size of after reduction. In this paper, we report the character-
6.9 nm was reported to be efficient for the aerobic ox- izations of the Pd nanoparticles supported on the
idation of various alcohols in trifluorotoluene sol- SiO2-Al2O3 mixed oxides and the size-dependent cata-
vent.[11] Pd nanoparticles with a mean diameter of ca. lytic activity of the prepared Pd nanoparticles in the
9 nm supported on an amphiphilic resin catalyzed the aerobic oxidation of benzyl alcohol.
aerobic oxidation of benzylic and allylic alcohols in
water medium.[12] Palladium nanoparticles entrapped
in aluminum hydroxide with sizes of ca. 2–3 nm or Results and Discussion
supported on aluminum oxide with diameters of ca.
5 nm were efficient catalysts for the aerobic oxidation Characterization of SiO2-Al2O3-Supported Pd
of alcohols.[13,14] Pd nanoparticles located in zeolite Nanoparticles
NaX or mesoporous silica SBA-15 could catalyze the
solvent-free aerobic oxidation of several kinds of al- Figure 1 shows the XRDpatterns of the SiO -Al2O3
2
cohols.[15,16] A Pd-Au/TiO2 catalyst could give excel- mixed oxides. Al2O3 without SiO2 exhibited three
lent TOFs for the aerobic oxidation of various alco- peaks at 2q of ca. 37.5, 46.0 and 66.78, which could be
hols, and it was suggested that the Au-Pd nanocrystals assigned to g-Al2O3, while SiO2 alone showed a broad
made of an Au-rich core with a Pd-rich shell account- diffraction peak at 2q of 22–238, assignable to the
ed for the superior catalytic performances of this cata- amorphous silica. The diffraction peaks belonging to
lyst.[17]
It is well known that catalytic performances can be
g-Al2O3 were observable for the SiO2-Al2O3 mixed
very sensitive to the size of the active phase.[18] Many
Pd nanoparticles-catalyzed organic reactions such as
hydrogenation, Heck and Suzuki couplings, and vinyl
acetate synthesis are structure sensitive.[19–22] In other
words, the conversion rate per surface Pd atom in
these reactions changes with the size of Pd particles.
Although many papers have been contributed to the
Pd nanoparticles-catalyzed oxidation of alcohols,[11–17]
studies on the effect of Pd particle size are unexpect-
edly scarce.[7b,15] Mori et al.[7b] compared the catalytic
activities of two Pd/hydroxyapatite catalysts contain-
ing Pd nanoclusters with mean diameters of 3.8 and
7.8 nm for the aerobic oxidation of 1-phenylethanol
and benzhydrol, and found that the smaller Pd nano-
clusters (3.8 nm) gave a higher TOF than the larger
Pd nanoclusters.
Further studies using catalysts containing Pd nano-
particles prepared on other general supports and with
mean sizes variable over a larger range are undoubt-
Figure 1. XRDpatterns of SiO -Al2O3 samples with differ-
2
edly helpful in better understanding the size-depend- ent Si/Al ratios along with SiO2 and Al2O3.
454
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 453 – 464

Size-Dependent Catalytic Activity of Supported Palladium Nanoparticles
FULL PAPERS
oxides with Si/Al ratios of 1/1 and 1/2, suggesting that
these samples were comprised of g-Al2O3 together
with amorphous silica. On the other hand, the SiO2-
Al2O3 mixed oxide with an Si/Al ratio of 2/1 was only
comprised of an amorphous silica phase, and Al2O3
might be dispersed on SiO2 or also exist in amorphous
phase. After loading of palladium, the XRDpatterns
did not change, and no additional diffraction peaks
assignable to any palladium compounds such as PdO
or metallic Pd could be observed before or after re-
duction because the content of Pd in the samples we
prepared using the adsorption method was low
(Table 1).
Because the change in Si/Al ratio may affect the
acidity of the catalyst and may thus change the cata-
lytic behavior, we have investigated the acidic proper-
ties of some typical Pd/SiO2-Al2O3 catalysts after H2
reduction by NH3-TPDmeasurements. As shown in
Figure 2, the samples using SiO2 and SiO2-Al2O3 with
lower Al contents (Si/Al=2/1 and 1/1) as the supports
only exhibited a single NH3 desorption peak at 455–
476 K, which could be attributed to the desorption
from weak Lewis acid sites. On the other hand, in ad-
dition to this lower temperature peak, a broad de-
Figure 2. NH3-TPDprofiles of samples after H reduction.
2
(a) 0.55 wt% Pd/SiO2, (b) 0.52 wt% Pd/SiO2-Al2O3 (Si/Al=
2/1), (c) 0.30 wt% Pd/SiO2-Al2O3 (Si/Al=1/1), (d) 0.52 wt%
Pd/SiO2-Al2O3 (Si/Al=1/2), (e) 0.53 wt% Pd/Al2O3.
sorption peak at ~660 K was also observed for the after the reduction with hexanol. The binding ener-
samples with higher Al contents, i.e., Pd/Al2O3 and gies of Pd 3d5/2 for the samples before reduction (but
Pd/SiO2-Al2O3 (Si/Al=1/2). Thus, acid sites with after calcination) were observed at ca. 336.3 eV,
medium strength appeared on these samples. As com- which can be assigned to Pd(II) in palladium oxide or
pared with the pure Al2O3, the doping of SiO2 into halide.[26] After the reduction with hexanol under re-
Al2O3 with an Si/Al ratio of 1/2 increased the concen- fluxing conditions for 6 h, the binding energy of Pd
trations of both weak- and medium-strength acid 3d5/2 in each sample decreased to ca. 335.1 eV, corre-
sites. It is known that the formation of composite sponding to metallic palladium species.[26] The reduc-
metal oxides may enhance the acidity.[24,25]
tion with H2 at 573 K for 30 min also shifted the bind-
The oxidation state of palladium before and after ing energy of Pd 3d5/2 in each sample to ca. 335.1 eV,
reduction was determined by XPS. Figure 3 shows the confirming that Pd(II) was also reduced to Pd(0) in
spectra of Pd 3d for some typical samples before and this case. We also recorded the XPS spectra of the
Table 1. Content, particles size and dispersion of palladium in SiO2-Al2O3-supported Pd catalysts.
Sample[a]
Pd content [wt%]
Mean size of Pd [nm]
Pd dispersion[b]
Pd dispersion[c]
Pd/SiO2-hexanol
0.55
0.52
0.30
0.52
0.53
0.55
0.52
0.30
0.30
0.52
0.39
0.56
0.53
5.7
5.1
4.2
3.2
2.6
10
4.3
3.6
3.1
3.1
2.9
2.7
2.2
0.20
0.22
0.27
0.35
0.43
0.11
0.26
0. 31
0.36
0.36
0.39
0.42
0.51
n.d.[d]
n.d.
n.d.
n.d.
n.d.
0.04
0.27
0.34
0.39
0.37
0.43
0.48
0.69
Pd/SiO2-Al2O3-hexanol (2/1)
Pd/SiO2-Al2O3-hexanol (1/1)
Pd/SiO2-Al2O3-hexanol (1/2)
Pd/Al2O3-hexanol
Pd/SiO2-H2
Pd/SiO2-Al2O3-H2 (2/1)
Pd/SiO2-Al2O3-H2 (1/1)
Pd/SiO2-Al2O3-H2 (1/2)
Pd/SiO2-Al2O3-H2 (1/2)
Pd/SiO2-Al2O3-H2 (1/4)
Pd/SiO2-Al2O3-H2 (1/4)
Pd/Al2O3-H2
[a]
The ratio in the parenthesis is the Si/Al ratio; hexanol and H2 denote the reductant used for the preparation.
Estimated from the mean size of Pd particles by 1.12/(mean size of Pd) (nm).
Evaluated from CO chemisorption measurements.
[b]
[c]
[d]
Not detectable.
Adv. Synth. Catal. 2008, 350, 453 – 464
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
455

Jing Chen et al.
FULL PAPERS
Figure 4. Diffuse reflectance UV-vis spectra for samples
before reduction together with PdO. (a) 0.55 wt% Pd/SiO2,
(b) 0.52 wt% Pd/SiO2-Al2O3 (Si/Al=2/1), (c) 0.30 wt% Pd/
SiO2-Al2O3 (Si/Al=1/1), (d) 0.53 wt% Pd/Al2O3.
Figure 3. Pd 3d XPS spectra of Pd/SiO2-Al2O3 before and
after reductions with hexanol. Before reductions: (a) 0.53
wt% Pd/Al2O3, (b) 0.52 wt% Pd/SiO2-Al2O3 (Si/Al=1/2),
(c) 0.30 wt% Pd/SiO2-Al2O3 (Si/Al=1/1), (d) 0.52 wt% Pd/
SiO2-Al2O3 (Si/Al=2/1), (e) 0.55 wt% Pd/SiO2; after reduc-
tion: (f) 0.53 wt% Pd/Al2O3, (g) 0.52 wt% Pd/SiO2-Al2O3
(Si/Al=1/2), (h) 0.30 wt% Pd/SiO2-Al2O3 (Si/Al=/1), (i)
0.52 wt% Pd/SiO2-Al2O3 (Si/Al=2/1), (j) 0.55 wt% Pd/SiO2.
Pd(II) species attached to Al2O3 via oxygen.[27–29] For
the Pd/SiO2-Al2O3 with an Si/Al ratio of 1, a shoulder
at ca. 280 nm assignable to small PdO nanoclusters
could be observed in addition to the bands at ca. 220
and 420 nm (curve c). The band at ca. 280 nm became
more intense for the sample with a higher Si/Al ratio
samples reduced with H2 at 573 K for 10 min. We (2/1) (curve b), and moreover, with increasing the Si/
found that the binding energies of Pd 3d5/2 for the Pd/ Al ratio, the broad absorption at 340–600 nm became
SiO2 and Pd/SiO2-Al2O3 (Si/Al=2/1) samples were intense, suggesting the gradual aggregation of dis-
shifted to ca. 335.1 eV while those for the Pd/SiO2- persed Pd(II) species to PdO nanoclusters and further
Al2O3 (Si/Al=1 and 1/2) and Pd/Al2O3 samples still to larger PdO particles with increasing the Si/Al ratio
remained at ca. 336.3 eV. This observation suggests in the SiO2-Al2O3 support.[27] These observations
that the palladium species on the first two samples allow us to speculate that the interaction between the
have already been reduced to Pd(0) after 10 min of precursor Pd(II) species and the support before calci-
reduction, whereas those on the latter three samples nation is weaker in the sample with a higher Si/Al
remained mostly in the Pd(II) state. Thus, it is likely ratio. In other words, the interaction between the ad-
that the nature of Pd species or the interaction be- sorbed Pd(II) species and Al2O3 was the strongest,
tween the Pd species and the support before reduc- leading to highly dispersed Pd(II) species after calci-
tion may be different for the samples with different nation, whereas aggregated PdO particles were
Si/Al ratios.
To gain insight into the nature of the palladium action between the adsorbed Pd(II) and SiO2.
species before reduction, we have performed diffuse The size of metallic Pd particles formed after re-
mainly obtained on SiO2 because of the weakest inter-
reflectance UV-vis spectroscopic studies for some typ- duction with either hexanol or H2 was determined by
ical samples after calcination. From the UV-vis spec- TEM observations. Typical TEM micrographs for
tra shown in Figure 4, we could see that the Pd/SiO2 some representative samples reduced with hexanol
(curve a) showed very broad absorption in the visible and H2 are shown in Figure 5 and Figure 6, respective-
region without a defined structure. This was quite ly. Particle size distributions derived from the TEM
similar to that for the reference compound PdO, sug- micrographs by counting ca. 150–200 particles in each
gesting that PdO particles were formed on the Pd/ case are also plotted in Figure 5 and Figure 6. It is of
SiO2 sample. On the other hand, the Pd/Al2O3 interest that, for both series of samples, the centers of
showed an intense absorption band at ca. 220 nm particle size distributions shift to smaller sizes with
along with a weak band at ca. 420 nm (curve d). decreasing Si/Al ratio. In other words, smaller Pd par-
These bands could be assigned to the highly dispersed ticles were obtained over the samples with lower Si/
456
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 453 – 464

Size-Dependent Catalytic Activity of Supported Palladium Nanoparticles
FULL PAPERS
Figure 5. TEM micrographs and Pd particle size distribu-
tions for the samples reduced with hexanol. (a) 0.55 wt%
Pd/SiO2, (b) 0.52 wt%Pd/SiO2-Al2O3 (Si/Al=2/1), (c) 0.30
wt% Pd/SiO2-Al2O3 (Si/Al=1/1), (d) 0.52 wt% Pd/SiO2-
Al2O3 (Si/Al=1/2), (e) 0.53 wt% Pd/Al2O3.
Figure 6. TEM micrographs and Pd particle size distribu-
tions for the samples reduced with H2. (a) 0.55 wt% Pd/
SiO2, (b) 0.52 wt% Pd/SiO2-Al2O3 (Si/Al=2/1), (c) 0.30
wt% Pd/SiO2-Al2O3 (Si/Al=1/1), (d) 0.52 wt%Pd/SiO2-
Al2O3 (Si/Al=1/2), (e) 0.53 wt% Pd/Al2O3.
Al ratios. The mean sizes of Pd particles in the two for adsorption. As can be seen from Table 1, the var-
series of samples were calculated and are summarized iation in the content of Pd in such a range did not
in Table 1. We have also prepared the Pd/SiO2-Al2O3- alter the mean size of Pd particles significantly. Thus,
H2 (reduced with H2) samples with the same Si/Al the Si/Al ratio was a key factor in determining the
ratio (1/2 or 1/4) but different Pd contents (ca. 0.3–0.6 size of Pd nanoparticles. By changing the Si/Al ratio
2À
wt%) by changing the concentration of PdCl4 used of the support, we could obtain supported Pd nano-
Adv. Synth. Catal. 2008, 350, 453 – 464
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
457

Jing Chen et al.
FULL PAPERS
particles with mean sizes ranging from 5.7 to 2.6 nm could be evaluated using the ratio of CO chemisorbed
and from 10 to 2.2 nm for the series of samples re- to all of the Pd atoms in the catalyst. The dispersion
duced with hexanol and H2, respectively.
of Pd evaluated from CO chemisorption (Table 1)
It is expected that the size of a metal particle is de- also increased with decreasing Si/Al ratio, and the
termined mainly by the rate of nucleation and the value was in good agreement with that estimated
rate of nuclei growth during the reduction.[30] The in- from the mean size of Pd particles, except for the Pd/
teraction between the Pd species and the support SiO2 sample.
would affect the mobility of the species and the rate
of nuclei growth. Therefore, it is understandable that
the weaker interaction between the PdO particles and Aerobic Oxidation of Alcohols Catalyzed by SiO2-
SiO2 leads to the formation of larger Pd particles for Al2O3-Supported Pd Nanoparticles
Pd/SiO2, while smaller Pd particles have been formed
over Al2O3 because of the stronger interaction be- Catalytic performances of the two series of Pd/SiO2-
tween the highly dispersed Pd(II) species and Al2O3. Al2O3 samples for the solvent-free aerobic oxidation
Medium mean sizes of Pd particles have been ob- of benzyl alcohol are shown in Table 2. We confirmed
tained over the Pd/SiO2-Al2O3 samples, probably be- that no conversion of benzyl alcohol occurred over
cause the use of SiO2-Al2O3 mixed oxide as the sup- the support alone, i.e., SiO2, Al2O3 or SiO2-Al2O3
port could create a medium interaction between the under the reaction conditions shown in Table 2. More-
Pd species and the support.
over, the SiO2-Al2O3-supported Pd(II) or PdO sam-
The dispersion of Pd, i.e., the fraction of surface Pd ples without reduction were also inactive under the
atoms in all of the Pd atoms in each sample can be es- same reaction conditions (343 K) although benzyl al-
timated from the size of Pd particles. By assuming cohol could be converted by the unreduced catalysts
spherical particles, we could calculate the dispersion at temperatures >ca. 358 K, where Pd(II) or PdO
of Pd using the diameter of Pd particles by the follow- became reducible to Pd(0) by benzyl alcohol during
ing equation: dispersion=1.12/diameter (nm).[31] the reaction. Therefore, the catalytic activities shown
Values of the dispersion of Pd thus calculated are in Table 2 should arise mainly from the supported Pd
summarized in Table 1. We also performed CO chem- nanoparticles.
isorption measurements for the evaluation of the dis-
Table 2 shows that, for both series of catalysts, the
persion of Pd. However, for the samples reduced with Pd/SiO2-Al2O3 catalysts with appropriate Si/Al ratios
hexanol, CO chemisorption was hard to detect proba- give remarkably better catalytic performances than
bly because the surface Pd sites were occupied by either Pd/SiO2 or Pd/Al2O3. A further inspection of
hexanol molecules, which were difficult to remove catalytic results for the two series of catalysts in-
during the pre-treatment used for CO chemisorption formed us that the mean diameter of Pd rather than
(evacuation at 373 K). On the other hand, CO chemi- the Si/Al ratio likely determined the catalytic activity
sorption could be obtained for the samples reduced because the catalysts with similar mean sizes of Pd
by H2. By assuming a chemisorption stoichiometry of particles but not similar Si/Al ratios exhibited compa-
CO to surface Pd atom of 1, the dispersion of Pd rable benzyl alcohol conversions. Moreover, the com-
Table 2. Catalytic activities of Pd/SiO2-Al2O3 for solvent-free aerobic oxidation of benzyl alcohol.[a]
Entry Catalyst[b]
Pd content [wt%] Mean size of Pd [nm] Conversion [%] Aldehyde selectivity [%]
1
2
3
4
5
6
7
8
Pd/SiO2-hexanol
0.55
Pd/SiO2-Al2O3-hexanol (2/1) 0.52
Pd/SiO2-Al2O3-hexanol (1/1) 0.30
Pd/SiO2-Al2O3-hexanol (1/2) 0.52
5.7
5.1
4.2
3.2
2.6
10
4.3
3.6
3.1
3.1
2.9
2.7
2.2
7.8
37
75
81
64
2.8
73
97
87
89
63
81
42
83[c]
99
97
98
99
98
98
98
>99
99
Pd/Al2O3-hexanol
Pd/SiO2-H2
0.53
0.55
0.52
0.30
0.30
0.52
0.39
0.56
0.53
Pd/SiO2-Al2O3-H2 (2/1)
Pd/SiO2-Al2O3-H2 (1/1)
Pd/SiO2-Al2O3-H2 (1/2)
Pd/SiO2-Al2O3-H2 (1/2)
Pd/SiO2-Al2O3-H2 (1/4)
Pd/SiO2-Al2O3-H2 (1/4)
Pd/Al2O3-H2
9
10
11
12
13
98
97
98
[a]
[b]
[c]
Reaction conditions: catalyst, 0.1 g; benzyl alcohol, 48.5 mmol; O2, 3 mLminÀ1; temperature, 343 K; time, 10 h.
The ratio in the parenthesis is the Si/Al ratio; hexanol and H2 denote the reductant used for the preparation.
The remaining product was toluene.
458
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 453 – 464

Size-Dependent Catalytic Activity of Supported Palladium Nanoparticles
FULL PAPERS
parison of the trend in Table 2 (catalytic activity) with vent-free aerobic oxidation of benzyl alcohol. The re-
that in Figure 2 (NH3-TPDresult) for the samples sults in Table 2 reveal that the catalysts containing Pd
with different Si/Al ratios strongly suggests that there nanoparticles with mean sizes of 2.7–4.3 nm exhibit
is no correlation between the catalytic activity and higher catalytic performances for the solvent-free
the acidity of catalyst. Therefore, we speculate that aerobic oxidation of benzyl alcohol.
the support in our case mainly exerts effects on the
The 0.30 wt% Pd/SiO2-Al2O3-H2 (Si/Al=1/1, re-
mean size of the supported Pd nanoparticles, and it is duced with H2) catalyst, which possesses a mean size
reasonable to consider that the variation in catalytic of Pd particles of 3.6 nm, has been examined for the
behaviors with changing Si/Al ratio mainly arises aerobic oxidation of various alcohols. Table 3 shows
from the change in the mean size of Pd nanoparticles. that the catalyst could catalyze the solvent-free aero-
The two series of catalysts exhibited a similar tenden- bic oxidation of various substituted benzyl alcohols
cy in the change of catalytic performance with the and 1-phenylethanol selectively to the corresponding
mean size of Pd particles; both the larger and smaller carbonyl compounds (entries 2–7). The present cata-
Pd nanoparticles show lower performances in the sol- lyst was also effective for the oxidation of allylic alco-
Table 3. Aerobic oxidation of various alcohols over the Pd/SiO2-Al2O3-H2 (Si/Al=1).[a]
Entry
Alcohol
Temperature [K]
Time [h]
Conversion [%]
Carbonyl compound selectivity [%]
99
1
353
24
99
2
3
353
423
24
24
91
54
100
100
4
423
24
26
100
5
6
7
423
423
423
24
24
24
60
52
82
100
100
90[b]
8
373
353
373
24
24
24
93
96
80
100
100
100
9
10
11
373
373
24
24
83
75
100
100
12
[a]
Reaction conditions: catalyst, 0.1 g (Pd, 2.82 mmol); O2, 3 mLminÀ1; entries 1–7: solvent-free, substrate 50 mmol; en-
tries 8–12: solvent trifluorotoluene, substrate 2 mmol.
The remaining product was benzaldehyde.
[b]
Adv. Synth. Catal. 2008, 350, 453 – 464
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
459

Jing Chen et al.
FULL PAPERS
hol (entry 9) and non-activated alcohols such as cyclo- Structure-Sensitivity of Pd Nanoparticles-Catalyzed
octanol, 2-octanol and 1-octanol (entries 10–12) selec- Oxidation of Benzyl Alcohol
tively to the corresponding aldehydes and ketones.
Furthermore, the oxidation of heterocyclic alcohols, To elucidate the effect of Pd particle size on intrinsic
i.e., 2-thiophenemethanol proceeded efficiently over catalytic reactivities of Pd sites for the solvent-free
the present catalyst (entry 8), whereas such heterocy- aerobic oxidation of benzyl alcohol, we measured the
clic alcohols are generally difficult to convert by ho- conversion of benzyl alcohol at the initial reaction
mogeneous metal complex catalysts due to the strong stage (conversion <20%) at 353 K and evaluated the
coordination of the substrate to the metal center. An- initial conversion rate (Table 4). From the initial con-
other specific point of the present Pd/SiO2-Al2O3 cat- version rate and the dispersion of Pd shown in
alyst is that all the substituted benzyl alcohols investi- Table 1, we calculated the intrinsic TOF, i.e., moles of
gated here exhibit lower reactivities than benzyl alco- benzyl alcohol converted per mole of surface Pd per
hol. However, for many homogeneous Pd complex second. For the catalysts reduced with H2, because Pd
catalysts,[32,33] it has been reported that the reactivities dispersion had been evaluated by both the size of Pd
of substituted benzyl alcohols with electron-donating particles and CO chemisorption, we obtained two in-
substituents are higher than that of benzyl alcohol, trinsic TOFs for this series of samples. To make a
while the presence of electron-withdrawing substitu- better comparison, we plotted the TOFs against the
ents decreases the reactivity. The same phenomenon mean size of Pd nanoparticles in Figure 8. It becomes
has also been reported for some heterogeneous cata- clear that the intrinsic TOFs depend significantly on
lysts such as Ru(OH)x/Al2O3.[34] The unique observa- the mean size of Pd particles. Moreover, because all
tion in our system may indicate that the reaction the data points in Figure 8 can be fitted in one curve,
mechanism or the rate-determining step for our it is reasonable to consider that the mean size of Pd
system is different from that for the homogeneous particles is the essential factor to determine the intrin-
catalytic systems or the Ru(OH)x/Al2O3.
sic TOF, whereas the Si/Al ratio and the reagent (H2
We have carried out recycling uses of the 0.30 wt% or hexanol) used for reduction influence the activity
Pd/SiO2-Al2O3-H2 (Si/Al=1/1) catalyst for the sol- via changing the mean size of Pd particles. Further-
vent-free aerobic oxidation of benzyl alcohol at more, the results in Table 4 and Figure 8 also indicate
353 K. As shown in Figure 7, the conversion and se- that the content of Pd should not be a critical factor
lectivity did not undergo significant changes during in determining the intrinsic TOF. Figure 8 reveals that
the repeated use for 5 cycles. Thus, the present cata- the catalysts with both larger and smaller mean sizes
lyst could be used recyclably.
of Pd particles exhibit lower intrinsic TOFs and there
is an optimum mean size (3.6–4.3 nm) of Pd particles
for the oxidation of benzyl alcohol. Although the re-
quirement of an appropriate size of metal particles is
known for several metal nanoparticles-catalyzed reac-
tions,[18,35] our present observation is quite significant
for Pd nanoparticles-catalyzed organic reactions be-
cause smaller Pd particles containing more coordin-
ately unsaturated Pd sites are generally believed to be
more active.[7b,19–22]
The unique observation in Figure 8 may be under-
stood by thinking about the possible reaction mecha-
nism. The reaction mechanism for the homogeneous
Pd complex-catalyzed alcohol oxidation has been in-
vestigated extensively.[36,37] The ligand-coordinated
Pd(II) species are proposed for the conversion of al-
cohol via Pd-alcoholate intermediate, which under-
goes b-hydride elimination to give carbonyl com-
pounds, and the generated Pd(0) should be oxidized
to Pd(II) quickly to avoid decomposition of the ho-
mogeneous catalysts.[36,37] The b-hydride elimination,
oxidation of Pd(0) to Pd(II), and the dissolution of O2
have all been suggested to limit the rate of oxidation
of alcohols depending on the systems and the reaction
conditions. The mechanism for the aerobic oxidation
of alcohols catalyzed by Pd- or Ru-based heterogene-
ous catalysts has also been discussed in several stud-
Figure 7. Recycling uses of 0.30 wt% Pd/SiO2-Al2O3-H2 (Si/
Al=1/1) for the aerobic oxidation of benzyl alcohol. Reac-
tion conditions: catalyst, 0.1 g (Pd, 2.82 mmol); benzyl alco-
hol, 48.5 mmol; O2, 3 mLminÀ1; temperature, 353 K; reac-
tion time, 4 h.
460
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 453 – 464

Size-Dependent Catalytic Activity of Supported Palladium Nanoparticles
Table 4. Intrinsic turnover frequencies of Pd/SiO2-Al2O3-catalyzed solvent-free aerobic oxidation of benzyl alcohol.[a]
FULL PAPERS
Catalyst[b]
Pd content [wt%] Mean size of Pd [nm] Initial conversion
rate[c] [mol gÀ1 sÀ1]
TOF[d] [sÀ1] TOF[e] [sÀ1]
Pd/SiO2-hexanol
0.55
5.7
5.1
4.2
3.2
2.6
10
4.3
3.6
3.1
3.1
2.9
2.7
2.2
0.26
0.68
1.82
2.12
1.78
0.094
3.22
2.15
1.14
2.30
0.62
1.39
1.45
0.26
0.64
2.42
1.24
0.83
0.16
2.53
2.45
1.12
1.30
0.44
0.64
0.57
-
-
-
-
Pd/SiO2-Al2O3-hexanol (2/1) 0.52
Pd/SiO2-Al2O3-hexanol (1/1) 0.30
Pd/SiO2-Al2O3-hexanol (1/2) 0.52
Pd/Al2O3-hexanol
Pd/SiO2-H2
Pd/SiO2-Al2O3-H2 (2/1)
Pd/SiO2-Al2O3-H2 (1/1)
Pd/SiO2-Al2O3-H2 (1/2)
Pd/SiO2-Al2O3-H2 (1/2)
Pd/SiO2-Al2O3-H2 (1/4)
Pd/SiO2-Al2O3-H2 (1/4)
Pd/Al2O3-H2
0.53
0.55
0.52
0.30
0.30
0.52
0.39
0.56
0.53
-
0.46
2.45
2.25
1.03
1.28
0.40
0.55
0.42
[a]
Reaction conditions: catalyst, 0.1 g; benzyl alcohol, 48.5 mmol; O2, 3 mLminÀ1; temperature, 353 K.
The ratio in the parenthesis is the Si/Al ratio; hexanol and H2 denote the reductant used for preparation.
Evaluated from the benzyl alcohol conversion at the initial reaction stage (conversion <20%).
Calculated from the initial conversion rate and the Pd dispersion evaluated from the mean size of Pd particles.
Calculated from the initial conversion rate and the Pd dispersion evaluated from CO chemisorption.
[b]
[c]
[d]
[e]
urements; Grunwaldt et al.[8] have clarified that the
metallic Pd over Al2O3 support is much more active
(over 50 times) for the aerobic oxidation of benzyl al-
cohol than the oxidic Pd species. Our present study
also suggests that Pd(0) nanoparticles are the active
species because the Pd/SiO2-Al2O3 catalysts without
reduction are inactive under our reaction conditions.
To further confirm that the reaction proceeds steadily
over the Pd(0) nanoparticles, we have measured the
O2 consumption in the oxidation of benzyl alcohol at
343 K under the reaction conditions shown in Table 2.
After 10 h of reaction, the amounts of O2 consumed
were 25 and 19 mmol over the 0.30 wt% Pd/SiO2-
Al2O3-H2 (Si/Al=1/1) and the 0.52 wt% Pd/SiO2-
Al2O3-H2 (Si/Al=1/2) catalysts, respectively. The
molar ratio of O2 consumption to benzaldehyde yield
was calculated to be ~0.5 over either catalyst, suggest-
ing that the oxidation of benzyl alcohol by O2 to ben-
zaldehyde and water proceeded stoichiometrically
over the present Pd(0)-based catalysts. Furthermore,
we have carried out XPS measurements for the recov-
ered 0.30 wt% Pd/SiO2-Al2O3-H2 (Si/Al=1/1) catalyst
after reaction. The result showed that the binding
energy of Pd 3d5/2 remained at 335.1 eV, confirming
Figure 8. Dependence of the intrinsic turnover frequency on
the mean size of Pd particles. Symbols: ( ) catalysts reduced
&
*
with hexanol, ( ) catalysts reduced with H2, and TOF based
*
on Pd dispersion estimated by mean size of Pd, ( ) catalysts
reduced with H2, and TOF based on Pd dispersion evaluated
by CO chemisorption. See Table 4 for reaction conditions.
ies.[7b,11,34] It is generally accepted that the Pd- or Ru- that palladium species were kept in the Pd(0) state
alcoholate also acts as the intermediate in these heter- after the reaction.
ogeneous catalytic systems. Nevertheless, for the het-
It is reasonable to speculate that the aerobic oxida-
erogeneous Pd-catalyzed oxidation of alcohols, Pd(0) tion of benzyl alcohol over our catalyst also proceeds
has been proposed as the active phase in many sys- through a Pd-alcoholate intermediate. By consulting
tems.[7,8,11–16] Solid evidence has been obtained to sup- the reaction mechanism proposed by Mori et al.,[7b]
port this in some research papers. For examples, Mori we speculate that, in the first step, benzyl alcohol may
et al.[7b] have confirmed that Pd species over hydrox- be chemisorbed via the interaction of the delocalized
yapatite are transformed to Pd(0) nanoparticles after p-bond of benzene ring with the terrace Pd atoms of
the oxidation of 1-phenylethanol via EXAFS meas- a Pd nanoparticle. Subsequently, the Pd-alcoholate in-
Adv. Synth. Catal. 2008, 350, 453 – 464
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
461

Jing Chen et al.
FULL PAPERS
termediate may be formed through the interaction of propriate Si/Al ratios showed significantly higher
À
the O H bond of the alcohol with another type of Pd benzyl alcohol conversions than both the Pd/SiO2 and
site, followed by the b-hydride elimination and the the Pd/Al2O3. The mean size of Pd particles played a
formation of benzaldehyde. The coordinately unsatu- key role in the aerobic oxidation of benzyl alcohol.
rated edge and corner Pd atoms on the same Pd With changing the mean size of Pd particles from 2.2
nanoparticle probably work for the formation of Pd- to 10 nm, the intrinsic turnover frequency per surface
alcoholate and the b-hydride elimination. Mori Pd atom for benzyl alcohol conversion showed a max-
et al.[7b] argued that the b-hydride elimination over imum at a medium size of 3.6–4.3 nm, revealing that
the edge or corner Pd atoms was the rate-limiting the reaction was structure-sensitive. The existence of
step, and thus, these Pd sites played key roles in the optimum particle size implies that an appropriate
oxidation of alcohols. On the other hand, through in ratio of the edge and corner Pd atoms to the terrace
situ ATR-IR spectroscopic studies combined with se- Pd atoms is required for the conversion of benzyl al-
lective site blocking by co-adsorbed CO and isotope cohol. It is likely that not only the b-hydride elimina-
labeling, Ferri et al.[38] proposed that the dehydrogen- tion of the Pd-alcoholate intermediate but the chemi-
ation of alcohol to carbonyl product could occur vir- sorption of alcohol on Pd nanoparticles also deter-
tually on all atoms on the Pd/Al2O3 catalyst. It is mines the rate of benzyl alcohol oxidation over our
known that the ratio of coordinately unsaturated Pd catalyst.
atoms (edge and corner Pd atoms) to terrace Pd
atoms increases as the size of Pd particles decreas-
es.[39] Our observation that a medium size of Pd parti-
cles is beneficial to benzyl alcohol oxidation implies
Experimental Section
that there may exist an appropriate ratio of these two
types of sites for the catalytic aerobic oxidation of
benzyl alcohol. This allows us to further consider that
not only the b-hydride elimination of Pd-alcoholate
but the chemisorption of alcohol also play a crucial
role in the aerobic oxidation of benzyl alcohol over
our catalyst. It should be remembered that the substi-
tuted benzyl alcohols with either electron-donating or
electron-withdrawing substituents in the benzene ring
were all less reactive than benzyl alcohol over our
catalyst (Table 3). This is also a quite unique observa-
tion because the benzyl alcohols with an electron-do-
nating substituent, which can facilitate the b-hydride
elimination, show higher reactivities in many homoge-
neous systems.[32,33] We speculate that the lower reac-
Catalyst Preparation
SiO2-Al2O3 mixed oxides with different Si/Al molar ratios
were synthesized by a co-precipitation method. In this
method, the aqueous solution of Na2SiO3 was added to that
of Al(NO3)3, and the Si/Al ratio in the final sample was
A
varied by changing the relative amounts of the two aqueous
solutions. The pH value of the mixed solution was adjusted
to 9.0 by aqueous NH3 solution or diluted aqueous HNO3
solution. The obtained suspension was further stirred for
1 h, and then the precipitate was filtered, washed thoroughly
with deionized water, and dried at 393 K overnight. The
dried powder was calcined at 773 K for 6 h to obtain the
SiO2-Al2O3 mixed oxide.
The Pd/SiO2-Al2O3 catalysts were prepared by an adsorp-
tivity of the substrates with electron-donating sub- tion method described in our previous paper for the prepa-
ration of Pd/Al2O3-ads.[14] The SiO2-Al2O3 mixed oxide pre-
stituents in our case may stem from the lower ability
2À
pared above was added into an aqueous solution of PdCl4
of chemisorption of these substrates on the Pd sites
with the pH value adjusted. To allow the adsorption of
because of the steric hindrance. This further supports
2À
PdCl42À species onto the support, we adjusted the pH of the
the speculation that both the coordinately unsaturated
PdCl4 solution to 1.2–1.7 by dilute HCl aqueous solution.
sites and the terrace sites are required for obtaining a
high intrinsic TOF for benzyl alcohol conversion.
+
À
ꢀ À
The surface would be transformed to =Al OH2 or Si
+
OH2 because the points of zero charge for Al2O3 and SiO2
were ca. 8–9 and 2–3, respectively.[40–42] The adsorption of
2À
PdCl4 could occur on the surfaces via coulombic interac-
Conclusions
tions. The content of Pd in the final catalyst could be regu-
lated to some extent by changing the concentration of
PdCl42À. After the adsorption, the powdery solid was recov-
ered by filtration followed by thorough washing with a large
amount of deionized water until no ClÀ could be detected in
the filtrate by AgNO3 aqueous solution. The sample was
subsequently dried at 393 K overnight and calcined at 773 K
in air for 3 h. The reduction of thus obtained Pd(II)/SiO2-
Al2O3 or PdO/SiO2-Al2O3 was carried out by two methods,
i.e., (i) hexanol reduction under refluxing conditions (at
~428 K) for 6 h and (ii) H2 reduction at 573 K for 0.5 h.
These two series of catalysts were denoted as Pd/SiO2-
We have succeeded in preparing SiO2-Al2O3-support-
2À
ed Pd nanoparticles by adsorption of PdCl4 ions
onto the support followed by calcination and reduc-
tion with either hexanol or H2. The mean size of the
Pd nanoparticles was tunable from 2.2 to 10 nm by
changing the Si/Al ratio and the reductant. The sup-
ported Pd nanoparticles could catalyze the aerobic
oxidation of various alcohols, and were particularly
efficient for benzyl alcohol oxidation under solvent-
free conditions. The Pd/SiO2-Al2O3 catalysts with ap- Al2O3-hexanol and Pd/SiO2-Al2O3-H2, respectively.
462
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 453 – 464

Size-Dependent Catalytic Activity of Supported Palladium Nanoparticles
FULL PAPERS
(grants 2003CB615803 and 2005CB221408), the Key Scientif-
ic Project of Fujian Province of China (No. 2005HZ01–3)
and the Program for New Century Excellent Talents in
Fujian Province (grant to Q.Z.).
Catalytic Reaction
The catalytic oxidation of alcohols by O2 was carried out
using a batch-type reaction vessel with a reflux condenser.
Typically, the powdery catalyst (typically 0.1 g containing Pd
of ca. 2.8–5.2 mmol) was added into the alcohol (benzyl alco-
hol, 48.5 mmol) placed in the reaction vessel, and the mix-
ture was then heated to the reaction temperature with stir-
ring. Then, an O2 flow was bubbled into the mixture to start
the reaction. After the reaction, the catalyst was separated
by centrifugation, and the liquid products were analyzed by
a gas chromatograph (Shimazu GC-14 B) after the addition
of an internal standard.
References
[1] R. A. Sheldon, J. K. Kochi, Metal-catalyzed oxidations
of organic compounds, Academic Press, New York,
1981.
[2] For recent reviews on homogeneous aerobic oxidation
of alcohols, see: a) R. A. Sheldon, I. W. C. E. Arends,
G.-J. ten Brink, A. Dijksman, Acc. Chem. Res. 2002, 35,
774; b) J. Muzart, Tetrahedron 2003, 59, 5789; c) B. M.
Stoltz, Chem. Lett. 2004, 33, 362; d) S. S. Stahl, Angew.
Chem. 2004, 116, 3480; Angew. Chem. Int. Ed. 2004, 43,
3400; e) M. Schultz, S. S. Hamilton, D. R. Jensen, M. S.
Sigman, J. Org. Chem. 2005, 70, 3343; f) M. S. Sigman,
D. R. Jensen, Acc. Chem. Res. 2006, 39, 221.
[3] For recent reviews on heterogeneous aerobic oxidation
of alcohols, see: a) M. Besson, P. Gallezot, Catal. Today
2000, 57, 127; b) T. Mallat, A. Baiker, Chem. Rev. 2004,
104, 3037; c) B.-Z. Zhan, A. Thompson, Tetrahedron
2004, 60, 2917; d) K. Kaneda, K. Ebitani, T. Mizugaki,
K. Mori, Bull. Chem. Soc. Jpn. 2006, 79, 981.
[4] a) T. Nishimura, N. Kakiuchi, M. Inoue, S. Uemura,
Chem. Commun. 2000, 1245; b) N. Kakiuchi, Y. Maeda,
T. Nishimura, S. Uemura, J. Org. Chem. 2001, 66, 6620.
[5] K. Ebitani, Y. Fujie, K. Kaneda, Langmuir 1999, 15,
3557.
[6] B. Karimi, S. Abedi, J. H. Clark, V. Budarin, Angew.
Chem. 2006, 118, 4894; Angew. Chem. Int. Ed. 2006, 45,
4776.
[7] a) K. Mori, K. Yamaguchi, T. Hara, T. Mizugaki, K.
Ebitani, K. Kaneda, J. Am. Chem. Soc. 2002, 124,
11572; b) K. Mori, T. Hara, T. Mizugaki, K. Ebitani, K.
Kaneda, J. Am. Chem. Soc. 2004, 126, 10657.
[8] J.-D. Grunwaldt, M. Caravati, A. Baiker, J. Phys.
Chem. B 2006, 110, 25586.
Catalyst Characterization
The Pd content in each catalyst was determined by induc-
tively coupled plasma (ICP) optical emission spectrometry
using an Agilent ICP-MS 4500–300. Powder X-ray diffrac-
tion (XRD) patterns were recorded on a Panalytical X’Pert
Pro Super X-ray diffractometer with Cu Ka radiation
(40 kV, 30 mA). NH3 temperature-programmed desorption
(NH3-TPD) was performed on a Micromeritics AutoChem
II 2920 instrument. The adsorption of NH3 was performed
at 393 K in an NH3-He (10 vol% NH3) mixture for 1 h, and
the remaining or weakly adsorbed NH3 was purged by high
purity He. TPDwas performed in an He flow by raising the
temperature to 973 K at a rate of 10 KminÀ1. The desorbed
NH3 was detected with a mass spectrometer (ThermoStar
GSD301 T2) by monitoring the signal of m/e=16.
Diffuse reflectance UV-Vis spectra were recorded on a
Varian Cary-5000 spectrometer equipped with a diffuse-re-
flectance accessory. The spectra were collected at 200–
800 nm with BaSO4 as a reference. X-ray photoelectron
spectroscopic (XPS) measurements were performed with a
Multilab 2000 system (Thermo Electron Co.) using Al-Ka
radiation (1846.6 eV) as X-ray source to investigate the oxi-
dation state of Pd before and after hexanol or H2 reduction.
Transmission electron microscopy (TEM) was taken on a
Phillips Analytical FEI Tecnai 30 electron microscope oper-
ated at an acceleration voltage of 300 kV. Samples for TEM
measurements were suspended in ethanol and ultrasonically
dispersed. Drops of suspensions were applied on a copper
grid coated with carbon. Chemisorption of CO was per-
formed using a Micromeritics ASAP2010C. The sample
after reduction with hexanol or H2 was purged with high
purity He (99.999%) at 373 K for 30 min and then was evac-
uated for 20 min. After the temperature had decreased to
308 K under vacuum, CO was introduced, and the first iso-
therm (total CO uptake) was measured. After the first iso-
therm, the sample was evacuated for ~10 min, and then the
second isotherm (reversible CO uptake) was measured.
Using the difference between the total and the reversible
adsorbed CO, we obtained the amount of the chemisorbed
CO.
[9] T. L. Stuchinskaya, I. V. Kozhevnikov, Catal. Commun.
2003, 4, 417.
[10] K. Wada, K. Yano, T. Kondo, T. Mitsudo, Catal. Today
2006, 117, 242.
[11] U. R. Pillai, E. Sahle-Demessie, Green Chem. 2004, 6,
161.
[12] Y. Uozumi, R. Nakao, Angew. Chem. 2003, 115, 204;
Angew. Chem. Int. Ed. 2003, 42, 194.
[13] M. S. Kwon, N. Kim, C. M. Park, J. S. Lee, K. Y. Kang,
J. Park, Org. Lett. 2005, 7, 1077.
[14] H. Wu, Q. Zhang, Y. Wang, Adv. Synth. Catal. 2005,
347, 1356.
[15] F. Li, Q. Zhang, Y. Wang, Appl. Catal. A: Gen. 2008,
334, 217.
[16] C. Li, Q. Zhang, Y. Wang, H. Wan, Catal. Lett. 2008,
120, 126.
[17] 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, G. J. Hutchings, Science 2006, 311,
362.
Acknowledgements
This workwas supported by the National Natural Science
Foundation of China (grants 20625310, 20773099 and
20433030), the National Basic Research Program of China
[18] A. T. Bell, Science 2003, 299, 1688.
Adv. Synth. Catal. 2008, 350, 453 – 464
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
463

Jing Chen et al.
FULL PAPERS
[19] B. Veisz, Z. Kirꢂly, L. Tꢃth, B. Pꢄcz, Chem. Mater.
2002, 14, 2882.
[20] Y. Li, E. Boone, M. A. El-sayed, Langmuir 2002, 18,
4921.
[21] J. L. Bars, U. Specht, J. S. Bradley, D. G. Blackmond,
Langmuir 1999, 15, 7621.
[22] Y.-F. Han, D. Kumar, D. W. Goodman, J. Catal. 2005,
230, 353.
[32] G.-J. ten Brink, I. W. C. E. Arends, M. Hoogenraad, G.
Verspui, R. A. Sheldon, Adv. Synth. Catal. 2003, 345,
497.
[33] J. A. Mueller, C. P. Goller, M. S. Sigman, J. Am. Chem.
Soc. 2004, 126, 9724.
[34] a) K. Yamaguchi, N. Mizuno, Angew. Chem. 2002, 114,
4720; Angew. Chem. Int. Ed. 2002, 41, 4538; b) K. Ya-
maguchi, N. Mizuno, Chem. Eur. J. 2003, 9, 4353.
[35] a) M. Valden, X. Lai, D. W. Goodman, Science 1998,
281, 1647; b) M. Haruta, Chem. Rec. 2003, 3, 75.
[36] a) B. A. Steinhoff, S. R. Fix, S. S. Stahl, J. Am. Chem.
Soc. 2002, 124, 766; b) B. A. Steinhoff, S. S. Stahl, J.
Am. Chem. Soc. 2006, 128, 4348.
[23] T. Teranishi, M. Miyake, Chem. Mater. 1998, 10, 594.
[24] K. Tanabe, Solid Acids and Bases, Kodansha, Tokyo,
1970.
[25] A. M. Venezia, V. La Parola, B. Pawelec, J. L. G.
Fierro, Appl. Catal. A: 2004, 264, 43.
[26] J. F. Moulder, W. F. Stickle, P. E. Sobol, K. D. Bomben, [37] a) J. A. Muller, C. P. Goller, M. S. Sigman, J. Am.
Handbookof X-ray Photoelectron Spectroscopy , Physi-
cal Electronics, Inc., Eden Prairie, MN. 1995.
[27] A. Rakai, D. Tessier, F. Bozon-Verduraz, New J. Chem.
1992, 16, 8.
[28] A. N. Pestryakov, V. V. Lunin, S. Fuentes, N. Bogdan-
chikova, A. Barrera, Chem. Phys. Lett. 2003, 367, 102.
[29] L. S. F. Feio, C. E. Hori, S. Damyanova, F. B. Noronha,
W. H. Cassinelli, C. M. P. Marques, J. M. C. Bueno,
Appl. Catal. A: 2007, 316, 107.
Chem. Soc. 2004, 126, 9724; b) M. J. Schultz, R. S.
Adler, W. Zierkiewicz, T. Privalov, M. S. Sigman, J.
Am. Chem. Soc. 2005, 127, 8499.
[38] D. Ferri, C. Mondelli, F. Krumeich, A. Baiker, J. Phys.
Chem. B 2006, 110, 22982.
[39] R. van Hardeveld, F. Hartog, Surf. Sci. 1969, 15, 189.
[40] N. Santhanam, T. A. Conforti, W. Spieker, J. R. Regal-
buto, Catal. Today 1994, 21, 141.
[41] J. A. Schwarz, C. Contescu, A. Contescu, Chem. Rev.
[30] S. T. Homeyer, W. M. H. Sachtler, J. Catal. 1989, 117,
91.
[31] N. Mahata, V. Vishwanathan, J. Catal. 2000, 196, 262.
1995, 95, 477.
[42] M. Trueba, S. P. Trasatti, Eur. J. Inorg. Chem. 2005,
3393.
464
asc.wiley-vch.de
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 453 – 464