Activation of aryl chlorides in water under phase-transfer agent-free and ligand-free Suzuki coupling by heterogeneous palladium supported on hybrid mesoporous carbon

Article abstract of DOI:10.1021/cs501356h

In this study, heterogeneous palladium catalysts supported on ordered mesoporous cobalt oxide-carbon nanocomposites were applied to the water-mediated Suzuki coupling reaction of chlorobenzene and phenylboronic acid and exhibited a high yield of biphenyl (49%) under mild reaction conditions free of phase-transfer agents and ligands. Product yields in the reaction of aryl chlorides containing electron-withdrawing groups attached to their benzene ring can reach approximately 90%. Thiol-functionalized mesoporous silica, which can trap soluble Pd species, was used to confirm the negligible leaching in solution and therefore heterogeneous reaction. This heterogeneous catalyst is stable, showing unobvious activity loss after 10 catalytic runs. Characterization by X-ray diffraction, transmission electron microscopy, X-ray photoelectron spectroscopy, X-ray absorption fine structure analysis, and N₂ sorption techniques revealed intercalated CoO nanoparticles inside a carbon matrix with uniform mesopore sizes (～3 nm), high surface area (～504 m²/g), and large pore volumes (～0.38 cm³/g). Additionally, very small Pd clusters consisting of approximately three atoms and Pd-O bonds formed on the interface between CoO and Pd nanoparticles. The unsaturated coordinative Pd may be responsible for the activation of chlorobenzene in the absence of any additives or ligands. Uniform mesopores and the hydrophobic nature of the carbon support may also facilitate the mass transfer of the reactant molecules and enrichment inside pores. For comparison, the catalytic activities of Pd catalysts supported on pristine mesoporous carbon and carbon embedded with nickel oxide nanoparticles were also tested.

Full text of DOI:10.1021/cs501356h

Research Article
pubs.acs.org/acscatalysis
Activation of Aryl Chlorides in Water under Phase-Transfer Agent-
Free and Ligand-Free Suzuki Coupling by Heterogeneous Palladium
Supported on Hybrid Mesoporous Carbon
Linlin Duan,† Rao Fu,† Zhiguang Xiao,† Qingfei Zhao,† Jian-Qiang Wang,‡ Shangjun Chen,†
and Ying Wan*,†
†Key Laboratory of Resource Chemistry of Ministry of Education, Shanghai Key Laboratory of Rare Earth Functional Materials, and
Department of Chemistry, Shanghai Normal University, Shanghai 200234, P. R. China
‡Shanghai Synchrotron Radiation Facility (SSRF), Shanghai Institute of Applied Physics, Chinese Academy of Sciences, Shanghai,
201204, P. R. China
S
* Supporting Information
ABSTRACT: In this study, heterogeneous palladium catalysts
supported on ordered mesoporous cobalt oxide−carbon
nanocomposites were applied to the water-mediated Suzuki
coupling reaction of chlorobenzene and phenylboronic acid
and exhibited a high yield of biphenyl (49%) under mild
reaction conditions free of phase-transfer agents and ligands.
Product yields in the reaction of aryl chlorides containing
electron-withdrawing groups attached to their benzene ring
can reach approximately 90%. Thiol-functionalized mesopo-
rous silica, which can trap soluble Pd species, was used to
confirm the negligible leaching in solution and therefore
heterogeneous reaction. This heterogeneous catalyst is stable, showing unobvious activity loss after 10 catalytic runs.
Characterization by X-ray diffraction, transmission electron microscopy, X-ray photoelectron spectroscopy, X-ray absorption fine
structure analysis, and N2 sorption techniques revealed intercalated CoO nanoparticles inside a carbon matrix with uniform
mesopore sizes (∼3 nm), high surface area (∼504 m2/g), and large pore volumes (∼0.38 cm3/g). Additionally, very small Pd
clusters consisting of approximately three atoms and Pd−O bonds formed on the interface between CoO and Pd nanoparticles.
The unsaturated coordinative Pd may be responsible for the activation of chlorobenzene in the absence of any additives or
ligands. Uniform mesopores and the hydrophobic nature of the carbon support may also facilitate the mass transfer of the
reactant molecules and enrichment inside pores. For comparison, the catalytic activities of Pd catalysts supported on pristine
mesoporous carbon and carbon embedded with nickel oxide nanoparticles were also tested.
KEYWORDS: Pd cluster, hybrid mesoporous carbon, aryl chloride, Suzuki coupling, water, TBAB-free, ligand-free
the properties of the Pd intermediates in the catalytic cycle has
allowed the reaction to be performed using a wider range of
substrates, with improved selectivity under milder conditions.5
A large amount of studies have focused primarily on designing
electron-rich catalytic sites on homogeneous catalysts by
introducing special and costly ligands, such as sterically
hindered and electron-rich phosphanes and N-heterocyclic
carbenes or palladacycles,6 that facilitate activation of the C−Cl
bond. However, the contamination of the products by Pd
residues even after the purification step poses an acute problem,
particularly for the pharmaceutical industry.7 Ligand-free
heterogeneous Pd-based catalysts have attracted much attention
because they possess great advantages in terms of recyclability.
Such catalysts include metallic Pd nanoparticles (NPs)
1. INTRODUCTION
Palladium-catalyzed Suzuki coupling reactions that lead to the
formation of new C−C bonds are of strategic importance in
synthetic organic chemistry because of the broad functional
group tolerance of this catalytic process, its mild reaction
conditions, and the stability and low toxicity of the boronic
acids employed.1 Aryl bromides, iodides, and triflates have been
extensively investigated for use as substrates in the reaction, as
they are easily activated by active Pd(0) species.2 However, aryl
chlorides, which are more economical and accessible than the
previously mentioned species, have rarely been used as
electrophiles in Pd-catalyzed Suzuki coupling due to the high
activation barriers associated with the difficult oxidative
insertion of Pd(0) species into the Ar−Cl bond.3
The reaction of aryl chlorides in Suzuki coupling has been
made possible due to the improved understanding of the
reaction mechanism and of the methods for overcoming the
rate-determining oxidative insertion step.3a,b,4 Understanding
Received: September 8, 2014
Revised: November 9, 2014
© XXXX American Chemical Society
575
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586

ACS Catalysis
Research Article
supported on silica,8 zeolites,9 carbon nanotubes,10 graphite
oxide and graphene,11 polymers,12 metal organic frame-
works,3c,13 and oxides.14 Among these catalysts, Pd/C has
been found to act as an efficient alternative to the usual
homogeneous conditions.15 Buchecker and co-workers devel-
oped the Pd/C-catalyzed Suzuki reaction in organic solvents
under nitrogen.16 Since then, ligand-free Pd protocols that
catalyze Suzuki cross-coupling with aryl chlorides have been
developed and are attracting an increasing amount of attention.
Sun, Sowa, and co-workers reported the high-yield cross-
coupling reaction of electron-poor aryl chlorides with phenyl-
boronic acid when catalyzed by 5% Pd/C (5 mol %) using the
very polar solvent dimethylacetamide (DMA) with a small
amount of water. Electron-rich, nonactivated aryl chlorides can
also react, albeit with much lower yields.17 Additives used to
enhance the catalytic activity of Pd/C would address this issue.
chlorobenzene in water.24 A further enhancement of the activity
of heterogeneous Pd NP-based catalysts can be attained by
modifying their electron density by incorporating intercalated
transition metal oxide nanoparticles in the carbon matrix. In
this work, we found that such transition metal oxide
nanoparticles, which have high crystallinity and can be
intercalated into carbon mesopore systems,25,26 can signifi-
cantly modify the electron density and dispersion of Pd
nanoparticles through metal oxide interfacial interactions. Very
small Pd clusters of approximately three atoms are formed.
These clusters favor the activation of chlorobenzene in the
absence of TBAB and ligands. The mesoporous Pd/CoO−C
catalyst exhibits a high yield of biphenyl (49%). When
chlorobenzene derivatives with aromatic electron-withdrawing
groups are employed in the reaction, the product yields are as
high as 90%. The heterogeneous nature of the catalysis was
shown by the negligible activity loss in the presence of a
trapping agent of thiol-functionalized mesoporous silica added
at an SH/Pd ratio of approximately 35. This catalyst is stable,
showing negligible metal leaching and activity loss after 10
catalytic runs. The reusability and good activity of the Pd/
CoO−C catalyst and its ability to function with additives (such
as TBAB) or ligands offer good opportunities for practical
applications in coupling reactions and hydrogenation.
Lysen and Kohler reported exceptional enhancement of the
́
̈
catalytic activity of Pd(II)/C by tetrabutylammonium bromide
((C4H9)4N+Br−, TBAB) in water.18 TBAB has been proven to
be an efficient additive for activating aryl chlorides using a
variety of supported Pd nanoparticle catalysts, while the long-
chain surfactant cetyltrimethylammonium bromide (CTAB) is
not.19,20 This result implies that the positive effect of TBAB
cannot be solely attributed to being a phase-transfer agent. A
relatively high temperature (100−140 °C) in water may favor
the precipitation of colloidal Pd nanoparticles that can be
stabilized by R4N+X−.7,21 Even at a relatively mild conditions
(65 °C) in water, bromide ions enhance the dissolution of
active Pd species and stabilize these species in solution. The
leaching Pd species can redeposit on the solid at the end of the
reaction, and noticeably, this process shows a minor effect on
metal distribution in some cases.19,20 Tetrabutylammonium
ions also provide activating counterions for the boronate and
palladate species and work as a phase-transfer agent. As a result,
the possibility of leaching catalytically active palladium colloid
species into solution cannot be ruled out, and the enhancement
of TBAB to the catalyst activity implies that the reaction takes
place in the solution phase.20
2. EXPERIMENTAL SECTION
2.1. Catalyst Preparation. The ordered mesoporous
oxide−carbon composite supports were synthesized using
Co(NO3)2·6H2O or Ni(NO3)2·6H2O as metal sources, low-
polymerized phenolic resins (preparation described in the
Supporting Information, SI) as the carbon source, and triblock
copolymer P123 as a structure-directing agent. In a typical
synthesis of mesoporous CoO−C, 0.5 g of triblock copolymer
P123 was dissolved in 10.0 g of ethanol while stirring at 40 °C
until the solution was clear. Then, 0.125 g of Co(NO3)2·6H2O
was added to the solution, which was stirred for an additional
30 min. Then, 5.0 g of 20 wt % phenolic resin ethanolic
solution was added, and stirring continued for another 10 min.
The mixture was transferred into multiple dishes. These dishes
were then placed in a hood to evaporate the ethanol at room
temperature for 5−8 h and were subsequently placed in an
oven to thermopolymerize the Bakelite at 80 °C for 24 h. The
as-made transparent films were scraped from the dishes, ground
into fine powders, placed in a quartz boat, and put in a tubular
oven. The powders were calcined at 600 °C in high-purity
nitrogen (99.99 vol %) for 6 h to remove the triblock
copolymer templates, carbonize the matrix, and decompose the
transition oxide. The transition metal oxide content in the
catalyst was determined by inductively coupled plasma atomic
emission spectroscopy (ICP-AES). The synthesis is easily
duplicated. The syntheses of the mesoporous NiO−C and
carbon (MC) materials were performed as described by the
references in the SI.25,27 Supported palladium catalysts were
prepared by isochoric impregnation. In a typical procedure, 1.7
g of an aqueous solution of PdCl2 (1.1 wt %) was impregnated
with 0.4 g of dry mesoporous carrier. The mixture was placed in
a hood overnight. The catalysts were then dried at 100 °C
under vacuum for 8 h, reduced at 200 °C in forming gas (10 vol
% H2 in nitrogen) for 3 h, and used as fresh supported catalysts
(Pd/CoO−C, Pd/NiO−C, and Pd/MC).
A heterogeneous, surfactant-free and ligand-free Pd-catalyzed
carbon−carbon cross-coupling reaction system that utilizes less
expensive chlorobenzene compounds in an aqueous environ-
ment offers the environmental benefits of using water instead of
dipolar aprotic organic solvents. In addition, this system offers
the advantages of easy catalyst recycling and negligible
contamination of the products by Pd residues. Although
homogeneous Pd-containing catalysts electron-enriched by the
introduction of special ligands can activate aryl chlorides, the
activity of heterogeneous Pd NP-based catalysts can also be
enhanced by modifying their electron density via support
effects. However, approaches for significantly improving
heterogeneous Pd NP catalysts are still rare; the electron
density within the metal has previously been difficult to modify
on the nanoscale.12,22 This problem is much more serious with
the presence of the conventional activated carbon supports.
The functionalization of the support, which is an efficient
method for modifying the electronic chemistry of the supported
metal, is always limited by the poor controllability of activated
carbon due to its inert surface.23 In addition, many examples of
significant Pd leaching have been reported.22a This drawback
greatly limits the applications of carbon-supported Pd catalysts.
Recently, we found that embedding inert silica into the
mesoporous carbon matrix serves to stabilize the Pd nano-
particles. The 3.5 nm Pd NPs facilitate the activation of
2.2. Characterization. X-ray diffraction (XRD) measure-
ments were obtained on a Rigaku Dmax-3C diffractometer
using Cu Kα radiation (40 kV, 30 mA, λ = 0.15408 nm). The d
576
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586

ACS Catalysis
Research Article
spacing values were calculated by the formula d = λ/2 sin θ, and
the unit cell parameters were calculated from the formula a0 =
2d10/√3. The metallic Pd sizes were estimated according to the
Scherrer formula: size = 0.89λ/β cos θ on the basis of the (111)
diffraction peak in the wide-angle XRD patterns. CO
chemisorption was conducted on a Quantachrome CHEM-
BET-3000 system by pulsing CO on the supported Pd catalyst.
Palladium nanoparticle sizes were calculated from CO pulse
chemisorption data assuming that all palladium particles were
spherical and that the adsorption stoichiometry was one CO
molecule per Pd atom.28 CO adsorption was assumed to be
complete after three successive peaks showed the same peak
areas. High-resolution scanning electron microscopy (HRSEM)
images were taken on a JEOL JSM 6380LV microscope
operated at 25 kV. Transmission electron microscopy (TEM)
experiments were conducted on a JEM 2100 microscope
operated at 200 kV. The samples for TEM measurements were
suspended in ethanol and supported on a holey carbon film on
a Cu grid. Energy-dispersive X-ray spectroscopy (EDX) was
performed on a Philips EDAX instrument. X-ray photoelectron
spectroscopy (XPS) measurements were performed on a
PerkinElmer PHI 5000CESCA system with a base pressure of
10−9 Torr. The metal ion concentrations were quantified using
ICP-AES (VISTA-MPX). The X-ray absorption spectroscopy
data were collected on Beamline BL14W1 at the Shanghai
Synchrotron Radiation Facility (SSRF). The electron storage
ring was operated at 3.5 GeV, and a double Si(311) crystal
monochromator was employed for energy selection. XAS data
of the samples were acquired in fluorescence mode using a 32-
element Ge solid-state detector. XAS data of standard samples,
including Pd foil and PdO, were also measured under similar
conditions for comparison. XAS data analysis was performed
using the Ifeffit software package.29 Standard procedures were
followed to analyze the XAS data. First, the raw absorption
spectrum in the pre-edge region was fitted to a straight line, and
the background above the edge was fitted with a cubic spline.
The EXAFS function, χ(E), was obtained by subtracting the
post-edge background from the overall absorption and then
normalized with respect to the edge jump step. The normalized
χ(E) was transformed from energy space to k space, where k is
the photoelectron wave vector. The χ(k) data were multiplied
by k2 as a weighting factor to compensate for the damping of
EXAFS oscillations in the high k region. The range of the k
space used for Fourier transformation (FT) is 3.2−12 Å, and
the range of the r space used for the curve fitting of EXAFS data
is 1−3 Å. The backscattering amplitude and the phase shift
were obtained by theoretical calculation using FEFF code
0.03 g of supported palladium catalyst containing 0.01 mmol
metallic Pd and 1.0 mmol of chlorobenzene, and the mixture
was heated to 80 °C. Finally, 1.2 mmol of phenylboronic acid
was added to start the coupling reaction. The products of the
reaction were centrifuged, extracted with ethyl acetate or
toluene, and analyzed. At different time intervals, hot filtration
was used to determine the degree of Pd leaching and
chlorobenzene conversion.31 The catalyst was separated and
submitted to continuous solid−liquid extraction with ethyl
acetate or toluene (5 mL) using micro-Soxhlet equipment.
Then, the solid catalyst was washed with a copious amount of
water and dried at 80 °C overnight under vacuum for reuse.
The extracted solution was mixed with the reaction solution,
and the mixture was extracted with ethyl acetate (10 mL). The
solids obtained were weighed and analyzed. The recovered
material accounted for more than 95% of the starting material.
About 98% carbon balance can be established between
chlorobenzene and biphenyl with the error of 5%. The coupling
reaction was repeated three times, yielding the same initial rates
and total conversion within 10%. To identify and analyze the
products, GC−MS equipped with a JW DB-5 capillary column
was used. Standard samples of each chemical were also analyzed
for comparison. The analysis was repeated at least three times
for all tests, and the experimental errors were within 5%.
Catalytic results are presented in terms of the absolute yield of
biphenyl, the conversion of chlorobenzene, the initial rate
(mmol of reacted chlorobenzene per mmol of Pd per hour)
calculated at the beginning of the reaction and the turnover
number (TON, mmol of reacted chlorobenzene/total Pd
concentration).32 The products were also isolated and purified
by column chromatography. The isolated components were
identified by 1H NMR analysis on a Bruker DRX 400
spectrometer, using tetramethylsilane as the internal standard.
The aqueous solution was collected to determine the Pd
leaching concentration. In reactions in which homogeneous
palladium acetate was used as the (pre)catalyst, 2 mmol of
TBAB was added before the preheating step, and the reaction
was initiated by adding 0.01 mmol of palladium acetate in
water.
For the soluble Pd trapping experiments, we followed the
method reported by Jones and co-workers using thiol-group-
modified mesoporous silica (SH-SBA-15).33 The preparation
and characterization of SH-SBA-15 are described in the SI
(Figures S1−S5). In standard experiments, 0.15 g of SH-SBA-
15 was added to the reaction flask before the addition of the
reaction solution (with a molar ratio of SH/Pd = ∼35:1).
For the recycling study, the Suzuki reaction was performed
with chlorobenzene and phenyl boronic acid, maintaining the
same reaction conditions as described above except using
recovered catalyst. After each completion of the reaction, the
catalyst was recovered by hot filtration and thoroughly washed
with ethyl acetate followed by a copious amount of water to
remove the organic and inorganic impurities from the used
catalyst. The recovered catalyst was dried under vacuum at 80
°C overnight, weighed, and reused in the next run. Several
parallel reactions were performed to ensure that the catalyst
amount was the same in each run. The palladium content of the
solid was measured at the first and the last reuses.
(version 6.0).30 An S0 value of 0.7 was obtained from fitting
2
PdO. From the analysis, structural parameters such as the
coordination number (N), bond distance (R), Debye−Waller
factor (σ2), and inner potential shift (ΔE0) were calculated for
the Pd component of each sample. N2 adsorption−desorption
isotherms were measured at 77 K with a Micromeritics Tristar
II 3020 analyzer. The Brunauer−Emmett−Teller (BET)
method was utilized to calculate the specific surface areas
(SBET) of the samples. By using the Barrett−Joyner−Halenda
(BJH) model, the pore volumes and pore size distributions
were derived from the adsorption branches of isotherms.
2.3. Catalytic Reactions. Suzuki cross-coupling reactions
were conducted in a 50 mL round-bottomed flask under
refluxing conditions with water as the solvent. Typically, 2
mmol of sodium carbonate (K2CO3) was dissolved in 10 mL of
water and placed in the reactor. To this solution were added
3. RESULTS AND DISCUSSION
3.1. Composition and Structure of Supported Palla-
dium Mesoporous Catalysts. The syntheses of mesoporous
carriers used a metal-containing lyotropic liquid-crystal self-
577
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586

ACS Catalysis
Research Article
Figure 1. XPS spectra in the (A) Co 2p and (B) Pd 3d regions for the mesoporous carrier, fresh supported Pd catalyst, and used Pd catalyst.
assembly route with hydrated nitrate salt as a metal source,
soluble phenolic resin as a carbon source, and triblock
copolymer as a structure-directing agent.25,26 The wide-angle
XRD patterns for the mesoporous CoO−C carrier display
crystalline peaks that can be matched to the series of Bragg
reflections corresponding to the pure cubic rock salt structure
of CoO (Fm3m, JCPDS: 65-2901, SI Figure S6). The oxide
particle size, estimated from the Scherrer formula, is
approximately 5 nm (the size perpendicular to the 111
plane), and the Co content, estimated from ICP-AES analysis,
is approximately 5.8 wt %. The XPS spectrum of the CoO−C
carrier in the Co 2p region (Figure 1A) shows peaks at 780.7
eV due to Co 2p3/2 (with a shakeup satellite at 786.4 eV) and at
796.7 eV due to Co 2p1/2 (with a satellite peak at 802.7 eV).34
The Co2+ cations in CoO are high-spin d8 and in octahedral
lattice sites. The intense satellite structure has been proposed to
result from the charge-transfer band structure found in late 3d
transition metal monoxides with partially filled eg character and
is commonly used to identify cobalt-containing oxides as the
rock salt monoxide.35 Therefore, this cobalt oxide is consistent
with the presence of Co2+ in the high-spin state. The absence of
a feature at 778.1 eV indicates the absence of Co metal
impurity.35b The formation of low chemical state transition
metal oxide in a mesoporous oxide−carbon composite has also
been confirmed for MnO−C.26 It should be mentioned that the
peak intensities of the CoO−C carrier are rather low,
demonstrating that the surface and near-surface concentrations
of Co are far below the concentration in the bulk material.
These results indicate that the transition metal oxide nano-
particles are embedded in the carbon pore system, similar to
the structure of mesoporous MnO−C, with a low chemical
state and coordinative unsaturation.26 The NiO−C carrier
exhibits several diffraction peaks, which can be indexed to the
(111), (200), and (220) planes of face-centered cubic (fcc)
NiO (Fm3m, JCPDS: 65-2901, SI Figure S6), as reported in the
literature.25 The NiO particle size is calculated to be 9.0 nm,
and the content is 7.2 wt %. The diffused diffraction peaks at
approximately 20° for all mesoporous carriers can be attributed
to the amorphous carbonaceous materials.
the involvement of CoO inside the carbon matrix shows a
significant effect on the Pd chemical state. The Pd 3d3/2 and
3d5/2 peaks for Pd/CoO−C can be fitted using a model
consisting of two peaks, where the peaks corresponding to
Pd(0) are at 340.9 and 335.7 eV, and the peaks of Pd−O are at
342.7 and 337.7 eV.37 At the same time, the features in the Co
2p region for cobalt monoxide can be maintained. The strong
electronic interactions of the catalyst with the support are
thought to be the predominant cause of the effect on the
electronic structure of the active particle.38 The electronic
changes are often specifically attributed to interface sites
between the particle and the support.39 The formation of a thin
Pd−O layer at the interface between Pd and transition oxide
Fe3O4 support has been investigated by Libuda and co-
workers.38 The interface oxide layer, which is stabilized by the
support, is capable of oxygen exchange with both the metallic
Pd surface and Fe3O4, showing complex and low kinetics,
respectively. Therefore, the interaction between Pd and low-
oxidation-state CoO may lead to the formation of a PdO layer
at the Pd/CoO−C interface.
The K-edge XANES was used to probe the electronic
transition from the 1s to 5p Pd orbitals and is sensitive to the
chemical state of the sample. The pre-edge line in the Pd K-
edge XANES spectrum of the Pd/NiO−C catalyst shows a
strong resemblance to the reference spectrum of the Pd foil,
indicating that the Pd is mainly in its metallic state (Figure 2A).
The intensity of the metallic peak at ∼24.39 keV, however,
shows a slight decrease. In addition, a small feature at ∼24.37
keV may indicate the presence of Pd−O species. These changes
become more pronounced in Pd/CoO−C. Although the pre-
edge line does not display a distinct shift to higher energies, the
intensity of the metallic peak decreases significantly. The
improvement in the feature at ∼24.37 keV is more significant,
implying a high content of Pd−O species. These phenomena
suggest the maintenance of the metallic features in Pd/NiO−C
and Pd/CoO−C and also the appearance of the Pd−O
component accompanied by a concomitant suppression of the
Pd0 contribution, in particular, for Pd/CoO−C.
Comparing the Fourier transformed k2-weighted Pd K-edge
EXAFS spectrum for Pd/NiO−C with the Pd foil reference
reveals that the Pd nanoparticles are almost completely
metallic, in good agreement with the XANES and XPS results
(Figure 2B). The relative intensity of the feature at ∼2.5 Å
After incorporation of the Pd species, the XPS spectrum
(Figure 1B) of Pd/NiO−C shows characteristic binding
energies of Pd 3d3/2 and 3d5/2 in metallic Pd species, similar
to Pd supported on pristine mesoporous carbon.36 In addition,
578
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586

ACS Catalysis
Research Article
theoretical contributions: one originating from Pd in a metallic
environment (Pd−Pd) and a second component corresponding
to Pd−O (Table 1). A Pd−Pd coordination number of 6.8
1.1 and a bond distance of 2.73 0.01 Å were obtained for Pd/
NiO−C. The under-coordination relative to the bulk metal can
be explained by the formation of Pd−O, which decreases the
coordination number; the Pd−O coordination number in bulk
PdO is 4.3. Indeed, a small Pd−O contribution appears in Pd/
NiO−C, with a coordination number of 1.9. The first nearest
neighbor Pd−Pd coordination number dramatically decreases
to 2.1
0.8 in Pd/CoO−C, while the Pd−O coordination
0.7. These phenomena may be
number increases to 3.6
related to the formation of very small three-atom Pd clusters, as
well as to the formation of Pd−O.41 The Pd−Pd distance is
reduced to 2.71 0.01 Å. The shorter length of the Pd−Pd
bond, reduced by approximately 0.07 nm compared with bulk
Pd (2.78
0.01 Å), also supports the formation of finely
dispersed Pd clusters.42
The Pd−O contribution, however, cannot be simply
attributed to the formation of Pd−O bonds in a palladium
oxide phase.40b,43 The XANES spectrum does not exhibit the
white-line feature characteristic of PdO, and the Pd−O−Pd
coordination in palladium oxide is insignificant in the FT-
EXAFS spectra of both Pd/CoO−C and Pd/NiO−C. In
addition, the Pd−O bond in the Pd/CoO−C (0.210 nm) and
Pd/NiO−C (0.208 nm) samples is longer than the covalent
Pd−O bond (0.202 nm) in PdO. A previous study showed that
long metal−O interaction distances between atoms within
nanoparticles in contact with O from an underlying oxide
support or from a zeolite were observed under H2
treatment.39a,41 The preferential formation of Pd−O at the
nanoparticle−oxide interface rather than on the outer particle
surface has also been observed on Pd/Fe3O4 by Libuda and co-
workers.38 As a result, we assume that the highly dispersed Pdx
clusters are intercalated with the CoO-containing support and
strongly interact with transition metal oxide to form Pd−O
bonds instead of palladium oxides.
HRSEM images for the NiO−C carrier show ordered
mesopore arrays and the confinement of nanoparticles
approximately ∼9.0 nm in diameter within the pore walls
(Figure 3A). To confirm that the particles are embedded inside
the matrix, dark-field SEM images were taken from the same
domain (Figure 3B). Many well-dispersed, uniform nano-
particles are observed, indicating the intercalation of the oxide
nanoparticles. Particle aggregation outside the carbon meso-
structure can be excluded, indicating the restriction of phase
separation between NiO and support upon heating. After the
Pd nanoparticles are loaded, the hexagonally arranged pores
Figure 2. (A) Pd K-edge XANES spectra and (B) Fourier transform of
Pd K-edge EXAFS spectra for the Pd/Co−C and Pd/NiO−C samples
and Pd foil and PdO references.
(phase uncorrected) is slightly smaller than the reference peak,
hinting at the presence of an overlapping additional scattering
pair at a shorter distance than Pd−Pd.40 A more pronounced
reduction in the intensity is observed for Pd/CoO−C. In
addition, none of the characteristic features of Pd oxide at ∼1.5
and 2.9 Å (phase uncorrected) is detected. Therefore, the
second component can be attributed to a Pd−O signal with a
low Z scatterer.40 The EXAFS data were then fitted with two
Table 1. Fit Parameters of Pd K-Edge EXAFS Spectra for the Pd/CoO−C and Pd/NiO−C Samples and the Pd and PdO
References
a
b
c
d
sample
shell
N
R (Å)
σ2 (×10−3 Å2)
ΔE0 (eV)
Pd/CoO−C
Pd−O
Pd−Pd
Pd−O
Pd−Pd
Pd−Pd
Pd−O
Pd−Pd
Pd−Pd
3.6 0.7
2.1 0.8
1.9 0.5
6.8 1.1
12.4 1.4
4.3 0.3
4.6 0.6
7.7 1.1
2.10 0.01
2.71 0.01
2.08 0.01
2.73 0.01
2.78 0.01
2.02 0.01
3.03 0.01
3.42 0.01
5.2 2.6
4.2 2.3
6.0
2.1
−6.6
3.0
Pd/NiO−C
5.7 1.0
8.4 1.3
1.3 0.7
4.4 0.8
4.4 0.8
−7.6
−7.0
−3.2
−7.2
−7.2
Pd
PdO
a
b
c
d
Coordination number. Distance between absorber and backscatterer atoms. Debye−Waller factor. Inner potential correction.
579
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586

ACS Catalysis
Research Article
Figure 3. (A) HRSEM, (B) dark-field HRSEM, (C,G,H) TEM, and (E,F) HRTEM images and (D) EDX pattern for the mesoporous carriers and
supported palladium catalysts: (A,B) NiO−C; (C−E) fresh Pd/NiO−C; (F,G) fresh Pd/CoO−C; and (H) used catalyst Pd/CoO−C-10run. The
images (A) and (B) are taken from the same domain, and (B) is a dark-field SEM image.
and well-dispersed oxide nanoparticles with a size of
approximately 9.0 nm are retained (Figure 3C). The EDX
pattern reveals the coexistence of Ni and Pd, although the
particles of each metal cannot be exclusively identified (Figure
3D). Figure 3E shows a representative particle with a size of 5
nm, revealing the characteristic fcc Pd{111} lattice fringe at
0.22 nm. TEM images for the Pd/CoO−C catalyst show
oriented and hexagonally arranged pores, indicating an ordered
2-D hexagonal mesostructure (Figure 3G). Dark dots are highly
dispersed inside the whole mesoporous matrix. No large
particles can be found, implying the absence of any highly
aggregated oxide or palladium particles. The particle shown in
Figure 3F, which is approximately 5 nm in size with a d spacing
value of 0.33 nm, can be attributed to the {110} lattice of
CoO.44 The resolution of the Pd nanoparticles is difficult to
determine, possibly due to the extremely small dimensions of
the Pd clusters. In fact, a rather diffused diffraction peak at
approximately 40° is detected in the wide-angle XRD pattern
that overlaps with the CoO peaks, implying the existence of
relatively small palladium nanoparticles (SI Figure S7).
CO chemisorption measurements were used to calculate the
size of the metallic Pd particles (Table 2). The palladium
particle sizes corresponding to the Pd/CoC−C < Pd/MC <
Table 2. Structural and Textural Parameters of the Supports,
Freshly Supported Palladium Catalysts after Reduction in
Forming Gas, and the Catalyst after 10 Catalytic Runs in the
Suzuki Reaction
a
dPd
a0
(nm)
SBET
Vt
Dp
catalyst
CoO−C
Pd/CoO−C
Pd/CoO−C-10run
NiO−C
(nm)
(m2/g)
(cm3/g) (nm)
9.3
9.3
9.3
10.1
9.9
9.5
9.5
591
504
499
521
446
515
418
0.43
0.38
0.37
0.42
0.42
0.36
0.32
3.3
3.0
3.0
4.1
4.1
3.0
3.0
1.3
1.5
b
Pd/NiO−C
MC
5.4
1.9
Pd/MC
a
b
Particle size, calculated from the CO chemisorption. Pd/CoO−C
catalyst after 10 catalytic runs.
580
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586

ACS Catalysis
Research Article
Pd/NiO−C series are 1.3, 1.9, and 5.4 nm, respectively, in good
agreement with TEM imaging and XAFS results.
Representative diffraction peaks in the small-angle range
attributed to the two-dimensional (2-D) hexagonal meso-
structure are observed for all of the mesoporous carriers and
catalysts (Figure 4), especially one strong (100) diffraction
are stable, and the highly ordered mesostructure is retained
after impregnation with metallic solution and the subsequent
drying and reduction steps.45 The sorption isotherms of the
supported palladium catalysts show typical type-IV curves with
significant nitrogen uptake at moderate pressures, indicative of
uniform mesopores (Figure 5). The supported palladium
catalysts over different carriers of CoO−C, NiO−C, and
pristine mesoporous carbon show similar BET surface areas
(∼500 m2/g) and uniform pore sizes (∼3 nm) (Table 2). A
comparison of the isotherms of the mesoporous carriers (SI
Figure S8) shows that the incorporation of Pd produces minor
effects on the pore size, pore volume, and surface area.
3.2. Suzuki Coupling Reactions of Chlorobenzene and
Phenylboronic Acid in Water. Organic reactions in water
have been investigated for use in the chemical and
pharmaceutical industries, as they offer the possibility of
providing environmentally benign reaction conditions by
reducing the burden of organic solvent disposal. The ability
to perform the coupling reaction under mild aqueous
conditions is highly sought after. The mesoporous supported
palladium catalyst Pd/CoO−C was used as a heterogeneous
catalyst in Suzuki coupling reactions with water as the solvent
and in air at 80 °C for the synthesis of biphenyls (Figure 6).
The catalyst exhibited a relatively high conversion of
chlorobenzene (49%, 8 h) and selectivity (>98%) to biphenyl,
indicating its ability to activate aryl chlorides under the organic
solvent-free, ligand-free, and phase-transfer catalyst-free con-
ditions imposed. The initial reaction rate and TON value were
calculated to be approximately 11 mmol·mmol(Pd)−1 h−1 and
49, respectively. The controlled reactions included the
homocoupling of either chlorobenzene or phenylboronic acid
under the same conditions. Undetectable and less than 4%
biphenyl can be found in the coupling reactions of
chlorobenzene or phenylboronic acid, respectively. As a result,
Figure 4. Small-angle patterns for (A) the mesoporous carriers and
(B) the supported Pd catalysts. Pd/CoO−C-10run denotes the
heterogeneous palladium catalyst after being used to catalyze 10
Suzuki reaction runs.
peak at 2θ = 0.5−1.5° and two weak peaks (110 and 200) at
higher angles. The cell parameters for the supported palladium
catalysts are similar to the mesoporous carriers. These results
further indicate that the mesoporous oxide−carbon composites
Figure 5. (A) Nitrogen sorption isotherms and (B) pore size distribution curves of the freshly mesoporous supported palladium catalysts and the
used Pd/CoO−C-10run catalyst after 10 catalytic runs. The isotherms of Pd/CoO−C-10run are offset vertically by minus 50 cm3/g.
581
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586

ACS Catalysis
Research Article
excluded.7 Soluble Pd colloids and complexes may contribute
to the activity.51,52 In fact, Sowa and co-workers postulated that
the oxidative addition step produces an Ar−Pd(II)−X species,
which desorbs from the charcoal support.21a After the
transmetalation step, the soluble palladium species precipitate
onto the support, and the system cools down. This interesting
observation has been described as the “boomerang effect” or
dissolution−redeposition mechanism.19,34b,53,54 Kohler and co-
̈
workers investigated the dissolution−redeposition mechanism
over a solid-supported Pd nanoparticles during Suzuki coupling
reaction in water even under a mild condition of 65 °C.19
A
substantial amount of Pd leaching (up to 50%) from Pd/Al2O3
can be detected during the reaction which is enhanced by
TBAB,19 and the aryl bromide conversion is highly related to
the Pd leaching. Dissolved Pd is deposited onto the support at
the end of the reaction.54
The hot filtration test has been considered a strong test for
assessing the presence of soluble active palladium. In this
procedure, solid catalysts are filtered out of the reaction, and
the filtrate is monitored for continued activity.55 When Pd/
CoO−C is hot filtered, after approximately 44% conversion, the
filtrate loses its activity, which is proof of a heterogeneous
catalysis (Figure 7). However, this interpretation shows certain
Figure 6. Time conversion plots for the Suzuki reaction of (A)
chlorobenzene (1 mmol) and (B) bromobenzene (1 mmol),
phenylboronic acid (1.2 mmol), and sodium carbonate (2 mmol) in
water at 80 °C over the supported palladium catalysts (0.01 mmol Pd
for mesoporous catalyst and 0.017 mmol Pd for the activated carbon
supported catalyst).
most biphenyl is originated from the cross-coupling, although
we cannot completely exclude the reductive homocoupling of
chlorobenzene and oxidative homocoupling of phenylboronic
acid. A conversion plateau occurs within 8 h. The reused Pd/
CoO−C catalyst after hot infiltration exhibits an initial reaction
rate of 10.5 mmol·mmol(Pd)−1 h−1, similar to the fresh one.
Thus, the product inhibition may be responsible for the
suppression of the reaction. When bromobenzene was used as a
reactant, approximately 96% of it was converted to product
within 6 h, and the initial reaction rate increased to 75 mmol·
mmol(Pd)−1 h−1. The different initial reaction rates imply the
difficult activation of the C−Cl bond.
Figure 7. Plots of bromobenzene conversion with either palladium
acetate or Pd/CoO−C as catalyst: (a) Pd/CoO−C activity without
any poisoning agent, (b) cessation of Pd/CoO−C activity upon hot
filtration of the catalyst after 30 min reaction, (c) Pd/CoO−C activity
in the presence of mercapto-functionalized mesoporous silica SH-SBA-
15 trapping agent with a SH:Pd of 35, (d) control reaction using
palladium acetate as catalyst in the presence of TBAB without any
poison, and (e) cessation of palladium acetate activity by adding SH-
SBA-15 at start of reaction. Reaction conditions: 0.01 mmol Pd, 1
mmol aryl bromide, 1.2 mmol phenylboronic acid, 2 mmol K2CO3, 10
mL water, 80 °C.
The high activity for the Suzuki coupling reaction of aryl
borides or aryl iodides has also been observed for other Pd-
containing heterogeneous catalysts, such as Pd/MCM-41,46
Pd/MgO,47 Pd/PdO/Cu,48 and Pd/activated carbon.49 There
have only been a few reports using Pd(0)-based heterogeneous
catalysts to catalyze the Suzuki coupling reaction of
chlorobenzene, and almost all cases employed organic solvents,
such as DMF and DMA, and the additive of TBAB; in addition,
sometimes harsh conditions, such as temperatures as high as
200 °C (SI Table S1), were employed.13b,17,50 The promoting
role of TBAB is distinct. A much lower conversion of
chlorobenzene has been shown to be obtained in the absence
of TBAB.50c As mentioned above, the stabilization of Pd
colloids and complexes by TBAB cannot be completely
limitations, especially for palladium-catalyzed reactions, when
other tests are not used to support the conclusions.43,51 The
(local) concentration of Pd leached may be continuously low,
and Heck and Suzuki reactions can be catalyzed by extremely
small amounts of palladium (as low as 0.00001%).56
Furthermore, Pd species can first dissolve in solution as a
function of time correlated to conversion and then redeposit on
the solid as mentioned above.19,54 These phenomena can lead
to the incorrect conclusion that no active soluble catalytic
species were present before filtration.51 Therefore, other tests
are needed to verify whether heterogeneous catalysis is indeed
occurring.33
582
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586

ACS Catalysis
Research Article
Comparisons of the activities of soluble palladium acetate
and immobilized Pd nanoparticles on mesoporous CoO−C are
displayed in Figure 7. The reaction using soluble palladium
acetate in the presence of TBAB begins rapidly, with an initial
reaction rate (20 mmol·mmol(Pd)−1 h−1) much higher than
that of Pd/CoO−C. Subsequently, we investigated the use of
thiol-functionalized mesoporous silicate SH-SBA-15 to distin-
guish homogeneous and heterogeneous contributions to
catalysis. SH-SBA-15 has been postulated to be an effective
and selective poison of homogeneous palladium species and
can therefore be used as a test for the heterogeneity of
palladium catalysts developed for the Heck and Suzuki
reactions.33 The SH/Pd ratio is approximately 35, with a
sufficiently high local sulfur concentration arising from the
bound thiols such that captured palladium is rapidly over-
coordinated by neighboring thiols before it can be released back
into solution.33 In our study, the complete quenching of
catalysis by palladium acetate that is observed upon the
addition of the poison is similar to literature reports,
demonstrating the successful capture of soluble palladium
species by SH-SBA-15.33 In contrast, an indistinct change in the
conversion plot in the presence of thiol groups is detected in
Pd/CoO−C catalysis, excluding the possibility that the activity
is due to soluble Pd in liquid phase (Figure 7).57
ported Pd catalyst, in which there are no intercalated transition
oxide nanoparticles, shows a reduced reaction rate of 4.8 mmol·
mmol(Pd)−1 h−1 and product yield of 30% after 8 h in the
coupling reaction of chlorobenzene and phenylboronic acid, as
well as a reduced reaction rate of 37 mmol·mmol(Pd)−1 h−1
and product yield of 66% after 4 h in the coupling reaction of
bromobenzene. A further decrease in catalytic performance is
detected for the activated carbon-supported Pd catalyst. This
catalyst possesses a high surface area of 1043 m2/g. The pore
size distribution is wide but shows a high proportion of
mesopores with a most probable distribution at 4.0 nm (SI
Figure S9). The reaction rates are as low as 2.1 and 11 mmol·
mmol(Pd)−1 h−1 in the coupling of chlorobenzene and
bromobenzene, respectively.
3.3. Stability of the Catalyst. The time plots for the
catalytic conversion of bromobenzene over the recovered Pd/
CoO−C catalyst are similar to the time plots obtained over the
fresh catalyst (data not shown here). We also tested the
conversion of bromobenzene after 1 h in successive runs; the
compiled data are shown in Figure 8. No obvious changes are
Different aryl halides were also used as substrates (Table 3).
A high yield of approximately 90% can be achieved for the
Table 3. Catalytic Performance in the Coupling Reaction for
a
Aryl Chlorides and Bromides over the Pd/CoO−C Catalyst
entry
X
R
t (h)
yield (%)
1
2
3
4
5
6
7
8
Cl
Cl
Cl
Cl
Cl
Br
Br
Br
H
8
49
94
88
29
19
96
89
97
CN
8
NO2
CH3
CH3O
H
8
Figure 8. Reusability of the Pd/CoO−C catalyst upon recycling it in
the water-mediated Suzuki reaction of bromobenzene and phenyl-
boronic acid.
24
24
4
OH
8
COOH
8
detected in the catalytic performance after 10 runs,
demonstrating that the reused catalyst maintains the reaction
rate and conversion. The reaction mother liquors after hot
separation of the catalyst after each run were collected and
mixed. The Pd content in the final mixture is below the
detection limit of ICP-AES analysis, implying that negligible
metal leaching occurs during repeated reactions on the same
catalyst. These results suggest that the Pd/CoO−C catalyst is
stable and reusable.
The XPS spectrum, N2 sorption isotherms, small-angle XRD,
CO chemisorption, and TEM images for the catalyst after 10
runs are shown. The Pd 3d5/2 region can be qualitatively
decomposed into two regions: a low BE component (335.0−
335.1 eV) related to metallic Pd and a high BE region
(approximately 336.6−336.7 eV) due to Pd atoms in thin oxide
layers (Figure 1B).37 Compared with the fresh catalyst, the
signal at the high BE region is slightly stronger, implying an
improvement in the Pd−O bond formation during the reaction,
possibly due to the interaction between Pd atoms and the O
atom in CoO. This phenomenon implies the involvement of
coordinated Pd species such as Pd−O in the Suzuki reaction.
The TEM image shows the stripe-like pore arrays of the
catalyst, containing highly dispersed nanoparticles (Figure 3H).
a
All Suzuki reactions were performed using 0.03 g of supported
palladium catalyst containing 1 mol % of Pd, 1 mmol aryl chlorides or
bromide, phenylboronic acid (1.2 mmol), and K2CO3 (2 mmol) in 10
mL of water under air at 80 °C.
reaction between phenylboronic acid and aryl chlorides
containing electron-withdrawing groups attached to their
benzene ring. When electron-donating groups are present, the
yields drop significantly but remain at approximately 20%. The
product yields are approximately 90% over the Pd/CoO−C
catalyst in the coupling reactions of phenylboronic acid and aryl
bromides, regardless of the presence of electron-withdrawing or
electron-donating groups.
The catalytic performance in the Suzuki coupling reaction of
supported palladium catalysts with different pore surface and
textural properties was also tested and is described in Figure 6.
The ordered mesoporous Pd/NiO−C catalyst displays a lower
product yield within 10 h (42%) but a similar initial reaction
rate (10 mmol·mmol(Pd)−1 h−1) in the Suzuki coupling
reaction of chlorobenzene; this catalyst performs similarly to
Pd/CoO−C in the conversion of bromobenzene. By
comparison, the pristine ordered mesoporous carbon-sup-
583
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586

ACS Catalysis
Research Article
High amounts of aggregation of large nanoparticles cannot be
found, demonstrating that the heterogeneous catalyst is stable,
and the reaction conditions have only minor effects on the
catalyst structure. The particle size for Pd estimated by CO
chemisorption analysis is approximately 1.5 nm, similar to the
size measured for the fresh catalyst. The XRD pattern and N2
sorption isotherms further confirm the stability of the catalyst
(Figure 4B and Figure 5). The used catalyst exhibits similar
structural and textual properties to the fresh one, indicating that
the ordered 2-D mesostructure, high surface area, and uniform
mesopores of the catalyst are all retained.
into the solution due to the large enough mesopores, similar to
Pd supported on activated carbon19 and ordered mesoporous
silica SBA-15.33 An obvious decrease for bromobenzene
conversion with addition of SH-SBA-15 suggests the
heterogeneous reaction over the Pd clusters supported on
CoO−C. However, an assumption that the Pd leaching is
restricted inside the pores cannot be completely ruled out. The
dissolution and redeposition completely occur inside the pores,
and after the reaction, the catalyst can be entirely recovered. A
thorough study should be undertaken.
The Pd/CoO−C catalyst possesses the highest product
yields and initial reaction rates in the coupling reaction of both
bromobenzene and chlorobenzene. Pd/NiO−C exhibits a
much higher aryl bromide conversion activity and initial
reaction rate, as well as a moderately higher aryl chloride
conversion activity and initial reaction rate when compared to
Pd/MC. When the conversion of chlorobenzene over the
exposed surface Pd atom (turnover frequency, TOF) was
considered, Pd/NiO−C with a larger particle size of about 5.4
nm (surface atoms/total atoms faction of about 0.255) exhibits
a much higher TOF value than Pd/CoO−C (atomic fraction
exposed >0.9). This result may also reflect that the surface
coverage of product may inhibit the reaction. The slow
reduction elimination step of the product may hinder the
reaction rate.
The interface interaction between the catalyst metal and the
support oxides through the formation of Pd−O bonds may also
facilitate the stabilization of the Pd nanoparticles on carbon,
thereby restricting the leaching and aggregation of the metal.
Finally, the uniform mesopores and the hydrophobic feature
of the carbon support may also contribute to the high activity of
the catalyst by assisting in the transportation of reactants and
products during the course of the reaction. In fact, the activated
carbon-supported Pd catalyst shows a much lower initial
reaction rate and product yield than the mesoporous carbon
catalysts.
4. DISCUSSION
In general, the Suzuki reaction is thought to proceed through
three steps: oxidative addition, transmetalation, and reductive
elimination.58 The decreased reactivity of aryl bromides and
chlorides in the Pd-catalyzed reaction compared to that of
iodobenzene can be attributed to their more difficult oxidative
addition to Pd(0). The activation of iodobenzene undergoes an
associative or direct mechanism requiring either single or
double coordination to Pd.59 The oxidative addition of
bromoarenes appears to occur by a combination of irreversible
reactions of the bromoarene with the two-coordinate Pd(0)
species and dissociation of the ligand. By comparison, the
activation of chlorobenzene requires ligand loss on Pd.59
Dissociation of the bidentate ligands to singly coordinated
ligands occurs on the active catalyst.60 Therefore, the
dissociation of bidentate ligands to a reactive ligand loss
intermediate prior to reaction with chloroarene is expected to
cleave the less reactive C−Cl bond.
The contribution of soluble Pd species in liquid phase to the
performance of the catalyst can be almost entirely excluded.
Therefore, the coupling reaction occurs inside the mesopores
which have high adsorption capacities for organic substances
from water.61 The enrichment of substrate inside the pores and
the enhancement of accessibility of substrates to active Pd sites
can improve the reaction to some extent.20 XAFS data reveal
that a fraction of the Pd atoms in the nanoparticles are
coordinative unsaturated after immobilization on a transition
metal oxide containing carbon support. This result is supported
by the decrease in the metallic Pd−Pd scattering intensity and
the increase in the intensity of the Pd−O XANES absorption
peak, which is accompanied by an increase of the additional
Pd−O component (Figure 2A). In addition, a decrease in the
Pd−Pd coordination number is observed in parallel with the
appearance of Pd−O bonding. The Pd−O contribution can be
attributed to the interface Pd−O bonds between Pdx
nanoclusters and the transition metal oxide rather than to a
distinct palladium oxide phase, as evidenced by XPS and XAFS
results. The driving force for the stabilization of the Pd−O
bond can be attributed to the interfacial interaction between Pd
and CoO or NiO. This phenomenon is more distinct for the
Pd/CoO−C catalyst. The higher fraction of Pd−O in Pd/
CoO−C than in Pd/NiO-C may be related to the smaller
particle size of the former catalyst, which should lead to a much
higher interface atom/surface atom interaction ratio. The
unsaturated coordination Pd may be responsible for the
activation of chlorobenzene, as was observed previously for
Pd on the USY zeolite, on which strong acid sites stabilize
surface defect atoms and tune the coordination between Pd and
O in zeolite.9,41 The cleavage of the C−Cl bond is promoted by
Pd atoms that have low coordination numbers. In the present
case, the leaching Pd species, if any, are assumed to be diffused
5. CONCLUSIONS
The interfacial interaction between palladium and CoO can
generate small Pdx clusters with low coordination that are
stabilized by the coordination of Pd and O. These clusters can
activate chlorobenzene in the water-mediated Suzuki coupling
reaction performed in the absence of TBAB and ligands. The
transition metal oxide nanoparticles are intercalated into a
mesoporous carbon matrix and can catalyze the production of
biphenyl in a yield of 49% over Pd/CoO−C. The product
yields of the reaction of chlorobenzene derivatives containing
electron-withdrawing groups attached to their benzene ring can
reach approximately 90%. Negligible Pd leaching is detected
either during the reaction, confirmed by thiol-functionalized
mesoporous silica trapping, or after reaction, confirmed by hot
filtration. The heterogeneous catalyst can be reused with no
observable decrease in the initial reaction rate. The TBAB-free
and ligand-free conditions, the use of pure water as solvent, and
the reusability of the Pd/CoO−C catalyst tested here make this
catalyst an attractive material for use in practical applications.
ASSOCIATED CONTENT
■
S
* Supporting Information
The following file is available free of charge on the ACS
Publications website at DOI: 10.1021/cs501356h.
584
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586

ACS Catalysis
Preparation of carbon precursors; preparation and
Research Article
(15) Ravat, V.; Nongwe, I.; Meijboom, R.; Bepete, G.; Coville, N. J. J.
Catal. 2013, 305, 36−45.
characterization of pristine mesoporous silica SBA-15
and thiol-functionalized mesoporous silica SH-SBA-15,
including XPS spectra, elemental analysis, TG curves in
air, FT-IR spectra, XRD patterns, and N2 sorption
isotherms; XRD patterns and N2 sorption isotherms of
carriers and supported Pd catalysts; and Suzuki coupling
of aryl chlorides and phenylboronic acid catalyzed by
different palladium catalysts (PDF)
(16) Marck, G.; Villiger, A.; Buchecker, R. Tetrahedron. Lett. 1994,
35, 3277−3280.
(17) LeBlond, C. R.; Andrews, A. T.; Sun, Y.; Sowa, J. R. Org. Lett.
2001, 3, 1555−1557.
(18) (a) Lysen
́
, M.; Kohler, K. Synlett 2005, 11, 1671−1674.
̈
(b) Lysen
́
, M.; Kohler, K. Synthesis 2006, 4, 692−698.
̈
(19) Soomro, S. S.; Ansari, F. L.; Chatziapostolou, K.; Kohler, K. J.
̈
Catal. 2010, 273, 138−146.
(20) Soomro, S. S.; Rohlich, C.; Kohler, K. Adv. Synth. Catal. 2011,
̈
̈
353, 767−775.
AUTHOR INFORMATION
Corresponding Author
*Tel: 86-21-6432-2516. Fax: 86-21-6432-2511. E-mail: ywan@
shnu.edu.cn.
■
(21) (a) Conlon, D. A.; Pipik, B.; Ferdinand, S.; LeBlond, C. R.;
Sowa, J. R.; Izzo, B.; Collins, P.; Ho, G.-J.; Williams, J. M.; Shi, Y.-J.;
Sun, Y. Adv. Synth. Catal. 2003, 345, 931−935. (b) Webb, J. D.;
MacQuarrie, S.; McEleney, K.; Crudden, C. M. J. Catal. 2007, 252,
97−109.
(22) (a) Fihri, A.; Bouhrara, M.; Nekoueishahraki, B.; Basset, J.-M.;
Polshettiwar, V. Chem. Soc. Rev. 2011, 40, 5181−5203. (b) Li, X.-H.;
Baar, M.; Blechert, S.; Antonietti, M. Sci. Rep. 2013, 3, 1743.
(23) Stenehjem, E. D.; Ziatdinov, V. R.; Stack, T. D. P.; Chidsey, C.
E. D. J. Am. Chem. Soc. 2013, 135, 1110−1116.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by State Key Basic Research Program
of China (2013CB934102), the NSF of China (21322308 and
21173149), the Ministry of Education of China (PCSIRT-
IRT1269 and 20123127110004), the International Joint
Laboratory on Resource Chemistry (IJLRC), and the Shanghai
Sci. & Tech. and Edu. Committee (11JC1409200, DZL123 and
S30406).
(24) Wan, Y.; Wang, H.; Zhao, Q.; Klingstedt, M.; Terasaki, O.;
Zhao, D. J. Am. Chem. Soc. 2009, 131, 4541−4550.
(25) Wang, W.; Wang, H.-Y.; Wei, W.; Xiao, Z.-G.; Wan, Y. Chem.
Eur. J. 2011, 17, 13461−13472.
(26) Kong, L.; Wei, W.; Zhao, Q.; Wang, J.-Q.; Wan, Y. ACS Catal.
2012, 2, 2577−2586.
(27) Meng, Y.; Gu, D.; Zhang, F.; Shi, Y.; Yang, H.; Li, Z.; Yu, C.; Tu,
B.; Zhao, D. Angew. Chem., Int. Ed. 2005, 44, 7053−7059.
(28) Wang, S. Y.; Moon, S. H.; Vannice, M. A. J. Catal. 1981, 71,
167−174.
REFERENCES
■
(1) Yin, L.; Liebscher, J. Chem. Rev. 2007, 107, 133−173.
(2) (a) Walvoord, R. R.; Kozlowski, M. C. J. Org. Chem. 2013, 78,
8859−8864. (b) Balanta, A.; Godard, C.; Claver, C. Chem. Soc. Rev.
2011, 40, 4973−4985.
(29) Newville, M. J. Synchrotron. Radiat. 2001, 8, 322−324.
(30) Rehr, J. J.; Albers, R. C. Rev. Mod. Phys. 2000, 72, 621−654.
(31) (a) Corma, A.; Das, D.; García, H.; Leyva, A. J. Catal. 2005, 229,
́
́
, A. Tetrahedron 2007, 63,
322−331. (b) Polshettiwar, V.; Molnar
6949−6976.
(3) (a) Grushin, V. V.; Alper, H. Chem. Rev. 1994, 94, 1047−1062.
(b) Littke, A. F.; Fu, G. C. Angew. Chem., Int. Ed. 2002, 41, 4176−
4211. (c) Yuan, B.; Pan, Y.; Li, Y.; Yin, B.; Jiang, H. Angew. Chem., Int.
Ed. 2010, 49, 4054−4058. (d) Jin, M.-J.; Lee, D.-H. Angew. Chem., Int.
Ed. 2010, 49, 1119−1122. (e) Sau, S. C.; Santra, S.; Sen, T. K.;
Mandal, S. K.; Koley, D. Chem. Commun. 2012, 48, 555−557.
(4) Bedford, R. B.; Cazin, C. S. J.; Holder, D. Coord. Chem. Rev. 2004,
248, 2283−2321.
(32) Della Pina, C.; Falletta, E.; Rossi, M.; Sacco, A. J. Catal. 2009,
263, 92−97.
(33) Richardson, J. M.; Jones, C. W. J. Catal. 2007, 251, 80−93.
(34) (a) Kim, K. S. Phys. Rev. B 1975, 11, 2177−2185. (b) Brundle,
C. R.; Chuang, T. J.; Rice, D. W. Surf. Sci. 1976, 60, 286−300. (c) Tan,
B. J.; Klabunde, K. J.; Sherwood, P. M. A. J. Am. Chem. Soc. 1991, 113,
855−861.
(35) (a) Shen, Z. X.; Allen, J. W.; Lindberg, P. A. P.; Dessau, D. S.;
Wells, B. O.; Borg, A.; Ellis, W.; Kang, J. S.; Oh, S. J.; Lindau, I.; Spicer,
W. E. Phys. Rev. B 1990, 42, 1817−1828. (b) Ghosh, M.;
Sampathkumaran, E. V.; Rao, C. N. R. Chem. Mater. 2005, 17,
2348−2352.
(36) Zhao, Y.; Jia, L.; Medrano, J. A.; Ross, J. R. H.; Lefferts, L. ACS
Catal. 2013, 3, 2341−2352.
(37) Semagina, N.; Renken, A.; Laub, D.; Kiwi-Minsker, L. J. Catal.
2007, 246, 308−314.
(38) Schalow, T.; Laurin, M.; Brandt, B.; Schauermann, S.; Guimond,
S.; Kuhlenbeck, H.; Starr, D. E.; Shaikhutdinov, S. K.; Libuda, J.;
Freund, H.-J. Angew. Chem., Int. Ed. 2005, 44, 7601−7605.
(39) (a) Koningsberger, D. C.; Gates, B. C. Catal. Lett. 1992, 14,
(5) Barnard, C. Platinum Met. Rev. 2008, 52, 38−45.
(6) (a) Barder, T. E.; Walker, S. D.; Martinelli, J. R.; Buchwald, S. L.
J. Am. Chem. Soc. 2005, 127, 4685−4696. (b) Gstottmayr, C. W. K.;
̈
Bohm, V. P. W.; Herdtweck, E.; Grosche, M.; Herrmann, W. A. Angew.
̈
Chem., Int. Ed. 2002, 41, 1363−1365. (c) Dupont, J.; Consorti, C. S.;
Spencer, J. Chem. Rev. 2005, 105, 2527−2572.
(7) Felpin, F.-X.; Ayad, T.; Mitra, S. Eur. J. Org. Chem. 2006, 2006,
2679−2690.
(8) (a) Jana, S.; Haldar, S.; Koner, S. Tetrahedron. Lett. 2009, 50,
4820−4823. (b) Ungureanu, S.; Deleuze, H.; Babot, O.; Achard, M. F.;
Sanchez, C.; Popa, M. I.; Backov, R. Appl. Catal., A 2010, 390, 51−58.
(9) Okumura, K.; Tomiyama, T.; Okuda, S.; Yoshida, H.; Niwa, M. J.
Catal. 2010, 273, 156−166.
(10) Zhang, P.-P.; Zhang, X.-X.; Sun, H.-X.; Liu, R.-H.; Wang, B.;
Lin, Y.-H. Tetrahedron. Lett. 2009, 50, 4455−4458.
(11) Scheuermann, G. M.; Rumi, L.; Steurer, P.; Bannwarth, W.;
271−277. (b) Hayek, K.; Kramer, R.; Paal
162, 1−15.
́
, Z. Appl. Catal., A 1997,
(40) (a) Okumura, K.; Niwa, M. J. Phys. Chem. B 2000, 104, 9670−
9675. (b) Paredis, K.; Ono, L. K.; Behafarid, F.; Zhang, Z.; Yang, J. C.;
Frenkel, A. I.; Cuenya, B. R. J. Am. Chem. Soc. 2011, 133, 13455−
13464.
Mulhaupt, . J. Am. Chem. Soc. 2009, 131, 8262−8270.
̈
(12) Gallon, B. J.; Kojima, R. W.; Kaner, R. B.; Diaconescu, P. L.
Angew. Chem., Int. Ed. 2007, 46, 7251−7254.
́
(13) (a) Llabres i Xamena, F. X.; Abad, A.; Corma, A.; Garcia, H. J.
(41) Okumura, K.; Matsui, H.; Tomiyama, T.; Sanada, T.; Honma,
Catal. 2007, 250, 294−298. (b) Saha, D.; Sen, R.; Maity, T.; Koner, S.
Langmuir 2013, 29, 3140−3151.
T.; Hirayama, S.; Niwa, M. ChemPhysChem 2009, 10, 3265−3272.
(42) Yudanov, I. V.; Metzner, M.; Genest, A.; Rosch, N. J. Phys.
(14) (a) Choudary, B. M.; Madhi, S.; Chowdari, N. S.; Kantam, M.
L.; Sreedhar, B. J. Am. Chem. Soc. 2002, 124, 14127−14136.
(b) Lakshmi Kantam, M.; Roy, S.; Roy, M.; Sreedhar, B.; Choudary,
B. M. Adv. Synth. Catal. 2005, 347, 2002−2008.
̈
Chem. C 2008, 112, 20269−20275.
(43) Keating, J.; Sankar, G.; Hyde, T. I.; Kohara, S.; Ohara, K. Phys.
Chem. Chem. Phys. 2013, 15, 8555−8565.
585
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586

ACS Catalysis
Research Article
(44) Liu, J. F.; He, Y.; Chen, W.; Zhang, G. Q.; Zeng, Y. W.;
Kikegawa, T.; Jiang, J. Z. J. Phys. Chem. C 2007, 111, 2−5.
(45) Sakamoto, Y.; Fukuoka, A.; Higuchi, T.; Shimomura, N.;
Inagaki, S.; Ichikawa, M. J. Phys. Chem. B 2004, 108, 853−858.
(46) Gannouni, A.; Rozanska, X.; Albela, B.; Saïd Zina, M.; Delbecq,
F.; Bonneviot, L.; Ghorbel, A. J. Catal. 2012, 289, 227−237.
(47) Dumbre, D. K.; Yadav, P. N.; Bhargava, S. K.; Choudhary, V. R.
J. Catal. 2013, 301, 134−140.
(48) Korzec, M.; Bartczak, P.; Niemczyk, A.; Szade, J.; Kapkowski,
M.; Zenderowska, P.; Balin, K.; Lelątko, J.; Polanski, J. J. Catal. 2014,
313, 1−8.
(49) Narayanan, R.; El-Sayed, M. A. J. Catal. 2005, 234, 348−355.
(50) (a) Kohler, K.; Heidenreich, R. G.; Krauter, J. G. E.; Pietsch, J.
̈
Chem.Eur. J. 2002, 8, 622−631. (b) Bohm, V. P. W.; Herrmann, W.
̈
A. Chem.Eur. J. 2000, 6, 1017−1025. (c) Herrmann, W. A.; Bohm,
̈
V. P. W. J. Organomet. Chem. 1999, 572, 141−145. (d) Jana, S.; Haldar,
S.; Koner, S. Tetrahedron Lett. 2009, 50, 4820−4823. (e) Yang, H.;
Han, X.; Ma, Z.; Wang, R.; Liu, J.; Ji, X. Green Chem. 2010, 12, 441−
̀
451. (f) Calo, V.; Nacci, A.; Monopoli, A.; Montingelli, F. J. Org. Chem.
2005, 70, 6040−6044. (g) Zhang, L.; Su, Z.; Jiang, F.; Zhou, Y.; Xu,
W.; Hong, M. Tetrahedron. 2013, 69, 9237−9244. (h) Artok, L.; Bulut,
H. Tetrahedron Lett. 2004, 45, 3881−3884. (i) Bhunia, S.; Sen, R.;
Koner, S. Inorg. Chim. Acta 2010, 363, 3993−3999. (j) Islam, M.;
Mondal, P.; Roy, A. S.; Tuhina, K. Transition Met. Chem. 2010, 35,
491. (k) Tyrrell, E.; Whiteman, L.; Williams, N. J. Organomet. Chem.
2011, 696, 3465−3472. (l) Arvela, R. K.; Leadbeater, N. E. Org. Lett.
2005, 7, 2101−2104.
(51) Kohler, K.; Kleist, W.; Prockl, S. S. Inorg. Chem. 2007, 46,
̈
̈
1876−1883.
(52) Reimann, S.; Stotzel, J.; Frahm, R.; Kleist, W.; Grunwaldt, J.-D.;
̈
Baiker, A. J. Am. Chem. Soc. 2011, 133, 3921−3930.
(53) (a) Davies, I. W.; Matty, L.; Hughes, D. L.; Reider, P. J. J. Am.
Chem. Soc. 2001, 123, 10139−10140. (b) Lipshutz, B. H.; Tasler, S.;
Chrisman, W.; Spliethoff, B.; Tesche, B. J. Org. Chem. 2003, 68, 1177−
1189.
(54) (a) MacQuarrie, S.; Horton, J. H.; Barnes, J.; McEleney, K.;
Loock, H.-P.; Crudden, C. M. Angew. Chem., Int. Ed. 2008, 47, 3279−
3282. (b) Kohler, K.; Heidenreich, R. G.; Soomro, S. S.; Prockl, S. S.
̈
̈
Adv. Synth. Catal. 2008, 350, 2930−2936.
(55) Sheldon, R. A.; Wallau, M.; Arends, I. W. C. E.; Schuchardt, U.
Acc. Chem. Res. 1998, 31, 485−493.
(56) (a) Phan, N. T. S.; Van Der Sluys, M.; Jones, C. W. Adv. Synth.
Catal. 2006, 348, 609−679. (b) de Vries, A. H. M.; Mulders, J. M. C.
A.; Mommers, J. H. M.; Henderickx, H. J. W.; de Vries, J. G. Org. Lett.
2003, 5, 3285−3288.
(57) Ellis, P. J.; Fairlamb, I. J. S.; Hackett, S. F. J.; Wilson, K.; Lee, A.
F. Angew. Chem., Int. Ed. 2010, 49, 1820−1824.
(58) Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457−2483.
(59) Barrios-Landeros, F.; Carrow, B. P.; Hartwig, J. F. J. Am. Chem.
Soc. 2009, 131, 8141−8154.
(60) Stambuli, J. P.; Buhl, M.; Hartwig, J. F. J. Am. Chem. Soc. 2002,
̈
124, 9346−9347.
(61) Zhuang, X.; Wan, Y.; Feng, C.; Shen, Y.; Zhao, D. Chem. Mater.
2009, 21, 706−716.
586
DOI: 10.1021/cs501356h
ACS Catal. 2015, 5, 575−586