One-pot preparation of magnetic N-heterocyclic carbene-functionalized silica nanoparticles for the Suzuki-Miyaura coupling of aryl chlorides: Improved activity and facile catalyst recovery

Article abstract of DOI:10.1039/c0gc00955e

Based on a reverse micelle strategy, we successfully synthesized new magnetic silica nanoparticles functionalized with a bulky N-heterocyclic carbene (N,N′-bis(2,6-diisopropylphenyl)imidazol-2-ylidene, denoted as IPr) precursor through the co-condensation of IPr-bridged organosilane and tetraalkoxysilane in a one-pot reaction. TEM and SEM investigations revealed that the particle sizes of the synthesized materials were uniformly distributed in the range 15-30 nm, and could be tuned by varying the amount of siliceous precursors. FT-IR and XPS characterizations showed that the IPr ligand was successfully incorporated onto magnetic silica nanoparticles. Such materials show good coordination capability toward Pd(acac)₂ (acac = acetylacetonate), leading to a higher loading in comparison with magnetic silica nanoparticles (without functionalization). This Pd-loaded material is active toward the Suzuki-Miyaura couplings of challenging aryl chlorides under relatively mild conditions (at 80°C). An 81% yield for biphenyl was achieved in the presence of 0.32 mol% of Pd within 8 h, using iso-propyl alcohol as a solvent and KO^tBu as a base. The activity of the functionalized nanoparticles is much higher than that of mesoporous silica-based catalysts, as well as a commercial Pd/C catalyst. This catalyst can be easily isolated by using a magnetic field and directly used in the next reaction cycle without significant loss of its activity.

Full text of DOI:10.1039/c0gc00955e

View Online / Journal Homepage / Table of Contents for this issue
Evaluation GOnlyr. Cereaetednwith ACspohse.PeDFm. Copiysrigthtr20y02-2021 Aspose Pty Ltd.
Dynamic Article Links
Cite this: Green Chem., 2011, 13, 1352
www.rsc.org/greenchem
PAPER
One-pot preparation of magnetic N-heterocyclic carbene-functionalized
silica nanoparticles for the Suzuki–Miyaura coupling of aryl chlorides:
improved activity and facile catalyst recovery
Hengquan Yang,*a Yunwei Wang,a Yong Qin,*b Yanzhu Chong,a Qiaozhen Yang,a Guang Li,a
Li Zhanga and Wei Lia
Received 27th December 2010, Accepted 7th March 2011
DOI: 10.1039/c0gc00955e
Based on a reverse micelle strategy, we successfully synthesized new magnetic silica nanoparticles
functionalized with a bulky N-heterocyclic carbene (N,N¢-bis(2,6-diisopropylphenyl)imidazol-
2-ylidene, denoted as IPr) precursor through the co-condensation of IPr-bridged organosilane and
tetraalkoxysilane in a one-pot reaction. TEM and SEM investigations revealed that the particle
sizes of the synthesized materials were uniformly distributed in the range 15–30 nm, and could be
tuned by varying the amount of siliceous precursors. FT-IR and XPS characterizations showed
that the IPr ligand was successfully incorporated onto magnetic silica nanoparticles. Such
materials show good coordination capability toward Pd(acac)2 (acac = acetylacetonate), leading to
a higher loading in comparison with magnetic silica nanoparticles (without functionalization).
This Pd-loaded material is active toward the Suzuki–Miyaura couplings of challenging aryl
chlorides under relatively mild conditions (at 80 ◦C). An 81% yield for biphenyl was achieved in
the presence of 0.32 mol% of Pd within 8 h, using iso-propyl alcohol as a solvent and KOtBu as a
base. The activity of the functionalized nanoparticles is much higher than that of mesoporous
silica-based catalysts, as well as a commercial Pd/C catalyst. This catalyst can be easily isolated by
using a magnetic field and directly used in the next reaction cycle without significant loss of its
activity.
toxicity.3 Moreover, IPr can coordinate other metals such as
Au, Ag, Cu, Au, Ni and Ru, yielding many active catalysts for
Introduction
various reactions.4 Such advantages make metal-IPr complexes
potential catalysts for practical applications, if these complexes
can be immobilized in a proper fashion for recovery and
reuse.
The Pd-catalyzed Suzuki–Miyaura coupling is currently a
fundamental reaction for the construction of C–C bonds and
plays an important role in the synthesis of agrochemical and
pharmaceutical compounds.1 The use of aryl chlorides as
substrates has proven difficult because of the high bonding
energy of C–Cl, but is highly desirable in view of the wide
diversity, ready availability and low cost of aryl chlorides, in
contrast with aryl bromides and iodides.2 It is well-known
that Pd–phosphine systems can efficiently promote the coupling
reactions of aryl chlorides, but suffer from high toxicity and low
stability in air. Recently, the N-heterocyclic carbene complex Pd–
IPr [IPr = N,N¢-bis(2,6-diisopropylphenyl)imidazol-2-ylidene]
has been found not only to have comparable activity with
Pd–phosphine complexes in the Suzuki–Miyaura couplings
of aryl chlorides but also to be stable in air and have low
However, there are few reports on immobilization of the
versatile bis(2,6-diisopropylphenyl)-substituted carbene (IPr)
ligand and further investigations of its catalytic activity in
Suzuki–Miyaura couplings, although the grafting of small
methyl-substituted carbene complexes on insoluble supports has
been extensively reported, which usually exhibit low activity
towards aryl chlorides due to lacking bulky electron-donating
substituents on the carbene ligand.5 In our recent work, we have
successfully incorporated the IPr ligand into the framework
of a cubic mesoporous material and then prepared an active
catalyst toward the Suzuki–Miyaura coupling of aryl chlorides.6
However, this solid catalyst required a relatively long time to
obtain a high yield (24 h for the Suzuki–Miyaura coupling
of chlorobenzene and phenylboronic acid). The limitation of
diffusion in nanopores was believed to be one of the reasons
for low reaction rate because the activity was lower than its
homogeneous counterpart. For this reason, there is still a need
aSchool of Chemistry and Chemical Engineering, Shanxi University,
Taiyuan, 030006, PR China. E-mail: hqyang@sxu.edu.cn;
Fax: +86-351-7011688; Tel: +86-351-7010588
bSchool of Chemistry and Chemical Engineering, Changzhou University,
Changzhou, 213164, PR China
1352 | Green Chem., 2011, 13, 1352–1361
This journal is
The Royal Society of Chemistry 2011
©

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Online
to find a method and strategy to prepare a new solid catalyst so
as to overcome these limitations.
Decreasing particle sizes down to nanoscale may be an
effective way to reduce the diffusion limitations because of the
shortened diffusion lengths. However, when the particle sizes are
less than 100 nm, to isolate these nanoparticles from a reaction
system is relatively difficult and time-consuming when applied
in practice. The usual manipulation processes involving a high-
speed centrifugation are quite inconvenient for the consecutive
processes. Fortunately, magnetically recoverable nanoparticles
such as SiO2-encapuslated Fe3O4 nanocomposites are now
emerging as new, appealing supports for catalyst recovery.7
Fe3O4, usually used as a core, confers the catalyst support
with a magnetic function. The silica part, as a shell, provides
a versatile platform for loading metal catalysts. In the process
of catalyst recycling, the catalyst can be magnetically recovered
inside the reactor vessel by employing an external magnetic field
while draining the liquid products. Because such a magnetic
field-assisted separation can minimize energy consumption and
catalyst loss, and save the time in achieving catalyst recovery, it is
considered a “green” separation approach. Rossi and co-workers
have prepared magnetic nanosilica-supported Pd, Rh and Au,
which demonstrated their excellent activity and recyclability in
alcohol oxidations, as well as hydrogenation.8 Baiker’s group
synthesized a magnetic Pt/SiO2/Fe3O4 catalyst that showed a
high activity and facile recovery in asymmetric hydrogenation.9
The preparation of these magnetically recoverable catalysts
often involves two steps: synthesis of the SiO2-encapuslated
Fe3O4 nanocomposites and subsequent immobilization of the
metal catalyst through a post-modification method (silylation
with surface silanols).10 Compared with the post-modification
method, the co-condensation method, that is, condensation
of a mixture of complex/ligand-bridged trialkoxysilane and
tetraalkoxysilane in a one-pot reaction, is a promising method
for preparing solid catalysts because of the ease of controlling
the complex/ligand loading and the uniform distribution of
the complex/ligand on the solid matrix at a molecular level.
This method is currently popular for preparing various hybrid
materials and PMO-type (periodic mesoporous organosilica)
catalysts.11 However, to our knowledge, so far there has been no
report on the preparation and catalytic applications of magnetic
organic–inorganic hybrid silica nanoparticles through the co-
condensation of bridged organosilane.
Scheme 1 A structural description of the magnetic IPr-functionalized
silica nanoparticle (MSN-IPr). Note: The CF3SO3 was exchanged by
Cl- during the material’s synthesis.
-
precursor with hydrolysable alkoxysilyl groups. We followed a
modified route for the introduction of two trialkoxysilyl groups
onto the terminals of the IPr precursor. The detailed synthesis
of the IPr precursor-bridged organosilica was reported in our
recent publication.6
As aforementioned, controlling the particle sizes on a nano-
scale is an important way to improve the reaction rate. A reverse
micelle system was used to synthesize the nanocomposites
because the reverse micelle could control the particle sizes
through the restrictions of the micelle size.12 As shown in
Scheme 2, the synthesis of magnetic silica nanoparticles func-
tionalized with the IPr precursor began with the construction
of a reverse micelle system (water-in-oil). A small amount of
water, along with a given amount of Fe(II)/Fe(III), added into
a large amount of xylene under vigorous stirring conditions
Driven by the unique properties of magnetically recoverable
nanoparticles and the potential applications of IPr in catalysis,
herein, we synthesized new magnetic organic–inorganic hybrid
silica nanoparticles functionalized with IPr (denoted as MSN-
IPr, as shown in Scheme 1) by co-condensation of IPr-bridged
organosilane and tetraalkoxysilane in reverse micelles through a
single step. After coordination with Pd(II), these magnetic func-
tionalized silica nanoparticles exhibit a high activity and facile
recyclability in the Suzuki–Miyaura coupling of challenging aryl
chlorides.
Scheme 2 A schematic demonstration of the synthesis of magnetic IPr-
functionalized silica nanoparticle (MSN-IPr): (a) reverse micelle (water-
in-oil); (b) Fe(II) and Fe(III) are hydrolyzed in the core of the reverse core;
(c) TEOS and IPr-bridged organosilane were hydrolyzed under the basic
conditions; (d) hydrolyzed silicon moieties are co-condensed around the
magnetite core, forming the MSN-IPr.
Results and discussion
Synthesis of magnetic silica nanoparticles (MSN-IPr)
In order to incorporate the IPr precursor into silica through a co-
condensation process, we needed to synthesise the IPr carbene
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1352–1361 | 1353
©

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Online
led to a reverse micelle solution. Due to the thermodynamic
driving forces, water and the dissolved ferrous salts formed
the cores of the reverse micelles. In the second step, a given
amount of aqueous hydrazine (N2H4·H2O) was added into the
system for the hydrolysis of ferrous salts. Iron oxide cores were
thus yielded in the reverse micelles under heating conditions.
Lastly, a mixture of tetraethyl orthosilicate (TEOS) and IPr-
bridged silane as siliceous precursors were added into the reverse
micelle system. Due to the presence of base and water, siliceous
precursors began hydrolyzing at the water/oil interfaces and co-
condensation of siliceous precursors around the Fe3O4 core then
occurred. The isolation and eventual drying gave the magnetic
silica nanoparticles (MSN-IPr). By increasing the amounts
of tetraethyl orthosilicate (TEOS) and IPr-bridged silane (the
molar ratio remained constant), magnetic functionzalized silica
nanoparticles with large particle sizes were obtained (denoted
as MSNL-IPr).
Fig. 2 A SEM image of MSN-IPr. The bar is 100 nm.
Characterization
The TEM images of the magnetic silica nanoparticles are
displayed in Fig. 1. In Fig. 1A, MSN-IPr materials consist
of spherical particles. The particle size distribution is relatively
narrow and mainly centered in the range 15–25 nm. Increasing
the amounts of TEOS and IPr-bridged silane leads to an increase
in particle size. As shown in Fig. 1B, the particles of MSNL-
IPr are also broadly spherical in shape, but their sizes undergo
an increase. Most of them mainly range from 20–30 nm in
size. The results confirm that the particle sizes of magnetic
functionalized silica nanoparticles are tunable by variation of
the amount of siliceous precursors. The SEM image of MSN-
IPr is presented in Fig. 2. In agreement with the TEM results, the
SEM image also shows that MSN-IPr consists of densely packed
individual spherical nanoparticles. A reverse micelle strategy is
believed to be a crucial factor for achieving uniformly spherical
nanoparticles.
Fig. 3 N2 adsorption/desorption isotherms of MSN-IPr and MSNL-
IPr. MSN-IPr, offset vertically by 100.
particle sizes increase, which is in good agreement with the TEM
observations.
Fig. 4 shows the wide angle XPD pattern of MSN-IPr. A
broad peak at ca. 2q = 23◦ is observed, which is assigned to
the amorphous silica. A series of characteristic peaks including
(200), (331), (440), (511) and (533) reflections of Fe3O4 are also
clearly identified. As expected, MSN-IPr has magnetic proper-
ties that can be inferred from the magnetic field dependence of
the magnetization displayed in Fig. 5A. The magnetization curve
of MSN-IPr measured at room temperature exhibits moderate
magnetism with a saturation magnetization of ca. 14 emu g-1.
Such a magnetization is sufficient for performing a magnetic
separation. As shown in Fig. 5B, the catalyst was easily isolated
magnetically by placing a magnet beside the reactor wall. MSN-
IPr could re-disperse into the reaction medium when the external
magnet was removed (a negligible magnetic hysteresis loop at
room temperature).
Fig. 1 TEM images of MSN-IPr (A) and MSNL-IPr (B). The bar is
50 nm.
Fig. 3 shows the N2 sorption isotherms and pore size
distribution plots of MSN-IPr and MSNL-IPr. The two ma-
terials both show type III isotherms with a capillary condensa-
tion/evaporation step at high relative pressure (P/P0 = 0.9–
1.0), indicating the presence of large pores on the magnetic
silica nanoparticles. These large pores probably arise from the
apertures between nanoparticles. These findings are in agree-
ment with the SEM observations that honeycomb networks
are present throughout the materials, as shown Fig. 2. The
determined specific surface area (BET) of MSN-IPr is 265 m2 g-1,
while the specific surface area (BET) of MSNL-IPr is 206 m2 g-1.
The decrease in the surface area probably suggests that the
The FT-IR spectrum of MSN-IPr is shown in Fig. 6. The
spectrum exhibits peaks at 2970, 1540, 1465 and 1095 cm-1. The
peaks at 2970 and 1465 cm-1 correspond to C–H stretching and
C–H bending vibrations, respectively. The peak at 1540 cm-1
is related to the characteristic vibrations of the imidazole ring
1354 | Green Chem., 2011, 13, 1352–1361
This journal is
The Royal Society of Chemistry 2011
©

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Online
Table 1 The change of Pd loadings on MSN-IPr with the amount of
Pd(acac)2 added
Materials
MS
2%
MSN-IPr
2% 3%
0.68 1.01 1.34 2.58
0.63 0.78 0.86 0.90
Pd(acac)2 wt%a
4%
5%
Theoretical Pd contents (wt%)b 0.68
Determined Pd contents (%)c
The molar ratio of IPr/Pdd
0.40
—
5.7
4.6
4.1
3.6
a The weight of added Pd(acac)2 with respect to MSN-IPr. b Theoretical
Pd contents on the assumption that all Pd(acac)2 molecules were
completely introduced onto MSN-IPr. c The determined Pd contents
on MSN-IPr by ICP-AES. d The actual molar ratio of IPr/Pd on the
Pd(c)/MSN-IPr.
0.34 mmol g-1. The determined C content is higher with respect
to the theoretical value, which is calculated according to the
determined N contents. These results can be explained by the
fact that a certain amount of sodium dodecylbenzenesulfonate
is present on the surface of MSN-IPr (an inorganic oxide
surface usually has a high adsorption capacity toward anionic
surfactants).
Fig. 4 The wide angle XRD pattern of MSN-IPr.
To probe the coordination capacity of IPr carbene incorpo-
rated on MSN-IPr, we chose Pd(acac)2 (acac = acetylacetonate)
to coordinate with the hybrid materials because the Pd–IPr
complex formed was reported to be stable in air and catalytically
active in the Suzuki–Miyaura reaction of aryl chlorides. The
amount of Pd(acac)2 was tuned from 2 wt% to 5 wt% (with
respect to the weight of MSN-IPr). Magnetic silica nanoparticles
(MSN) without incorporation of the IPr ligand were also
synthesized for comparison. The coordination was conducted
in a 1,4-dioxane solution under the refluxing conditions. ICP-
AES was employed to determine the Pd loadings on the solid
materials. These results are summarized in Table 1. In the cases
of using 2 wt% Pd(acac)2, the Pd loading on MSN (without
incorporation of IPr) was determined to be 0.40 wt%. This
value is lower than the theoretical value (0.68 wt%), indicating
that a portion of the Pd(acac)2 molecules were not captured by
MSN. While for MSN-IPr (with the incorporation of IPr), the Pd
loading was determined to increase up to 0.63 wt%, which was
nearly equal to the theoretical value (0.68 wt%) and much higher
than the case of MSN. The significantly improved Pd loading
demonstrates the role of the IPr ligand in the coordination of
Pd. Increasing the amount of Pd(acac)2 led to an increase in the
determined Pd loading on MSN-IPr. When 4 wt% of Pd(acac)2
was added, the determined Pd loading on the solid materials was
0.86 wt%. However, further increasing the Pd(acac)2 amount up
to 5 wt%, did not result in an apparent increase in the determined
Pd loading. The results suggest that 4 wt% of Pd(acac)2 is the
most suitable for obtaining high Pd loading. On this occasion,
the ratio of IPr/Pd was estimated to be ca. 4. The loading
lower than the theoretical values was explained by the fact that
a portion of the IPr ligands incorporated on MSN-IPr were
inaccessible to the relatively bulky Pd complex.
Fig. 5 A hysteresis loop for MSN-IPr at 298 K (A) and an image
showing isolation of the magnetic solid by employing an external field
(B).
Fig. 6 The FT-IR spectrum of MSN-IPr.
of the IPr precursor. The presence of the characteristic peak
indicates that the IPr precursor was successfully introduced onto
the solid materials. The strong peak at 1095 cm-1 is attributable
to the stretching vibrations of Si–O–Si, which were formed
through the hydrolysis–condensation of the siliceous precursors.
The quantitative determination of the IPr incorporated onto
the magnetic silica nanoparticles was made with elemental
analysis. The C and N contents of MSN-IPr are 0.96 wt%
and 19.73 wt%, respectively. Based on the results of the N
contents, it is estimated that the IPr loading on MSN-IPr is
XPS spectroscopy was employed to investigate the elemental
compositions and coordination states of Pd species on the MSN-
IPr surface. In Fig. 7, an XPS elemental survey scan of surface
elementals of MSN-IPr derived with Pd(acac)2 reveals that Si,
O, Fe, C, N, Cl and Pd elements are present in the solid material.
-
The S element (CF3SO3 of the IPr precursor) was not found
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1352–1361 | 1355
©