Dynamic functionalised metallogel: An approach to immobilised catalysis with improved activity

Article abstract of DOI:10.1016/j.molcata.2009.11.001

A metal-organic gel (MOG) route to immobilised catalysts has been developed based on the reversibility of coordination bond formation in a Fe³⁺-carboxylate MOG. The ferric gel was rationally designed based on a bifunctional ligand, 5-1H-benzo[d]imidazole-1,3-dicarboxylic acid (L1), in which the imidazole N atom may further bind to a soft ion like Pd²⁺ more strongly due to the HSAB matching. Palladium(II) was successfully loaded in the gel to yield a supported coordination catalyst. Compared with its homogeneous counterparts, the immobilised gel catalyst exhibited significantly improved catalytic activity under mild conditions as shown in Suzuki-Miyaura C-C coupling, and could be reused for several times.

Full text of DOI:10.1016/j.molcata.2009.11.001

Journal of Molecular Catalysis A: Chemical 317 (2010) 97–103
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Dynamic functionalised metallogel: An approach to immobilised catalysis with
improved activity
Jing Huang, Lisi He, Jianyong Zhang∗, Liuping Chen, Cheng-Yong Su
MOE Laboratory of Bioinorganic and Synthetic Chemistry, School of Chemistry and Chemical Engineering, Sun Yat-Sen University,
135 Xingang Xi Road, Guangzhou 510275, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 25 June 2009
Received in revised form 30 October 2009
Accepted 1 November 2009
Available online 11 November 2009
A metal–organic gel (MOG) route to immobilised catalysts has been developed based on the reversibil-
ity of coordination bond formation in a Fe3+–carboxylate MOG. The ferric gel was rationally designed
based on a bifunctional ligand, 5-1H-benzo[d]imidazole-1,3-dicarboxylic acid (L1), in which the imi-
dazole N atom may further bind to a soft ion like Pd2+ more strongly due to the HSAB matching.
Palladium(II) was successfully loaded in the gel to yield a supported coordination catalyst. Compared
with its homogeneous counterparts, the immobilised gel catalyst exhibited significantly improved cat-
alytic activity under mild conditions as shown in Suzuki–Miyaura C–C coupling, and could be reused for
several times.
Keywords:
Gel
Hybrid material
Palladium
© 2009 Elsevier B.V. All rights reserved.
Coordination polymer
Supported catalysis
1. Introduction
to the following reasons. (a) A large number of organic lig-
ands and metal ions provide a rich pool for active catalyst. (b)
Sol–gel processes have been used to produce a large num-
ber of useful solid materials for optical, magnetic, electrical,
and catalytic applications [1]. The sol–gel process provides a
route to efficient heterogeneous catalysts [2]. Sol–gel catalysts
have many advantages for materials preparation and cataly-
sis, such as preparation under mild conditions, homogeneous
distribution of catalytic active centres, and easy removal from
the reaction medium. Metal–organic gels (MOGs) are a class
of supramolecular metallogels that rely on metal–ligand coor-
dination extending into 3D structures via coordination and/or
other supramolecular interactions, such as hydrogen-bonding,
␲–␲ stacking, van der Waals and hydrophobic interactions [3].
MOGs have received growing attention due to the incorpo-
ration of coordination interactions [3]. MOG-based materials
may have some advantages over the materials prepared by the
sol–gel processes which are normally amorphous and lack of
short-range order in molecular level in spite of some regular-
ity in mesoscale in some cases (e.g. MCM- and SBA-type silicas)
[4].
Short-range ordered structures of MOGs provide highly regulated
metal centres or microenvironment for catalyst immobilisation
[5]. (c) Stable organic ligands and metal salts are usually used as
MOG precursors, which avoid the tedious manipulation in tradi-
tional sol–gel processes. (d) The wet gel itself with catalytically
active centre can be also a highly porous catalyst without fur-
ther processing due to its relatively inert nature compared to
silica gels and others [6]. (e) To be the most important, the
reversibility of coordination bond formation may make MOGs
have some promising properties that traditional supports with
rigid bonds (e.g. organic polymers and inorganic oxides) do
not have.
In spite of the potential application of MOGs in catalysis, only a
few transition metal catalysts have been immobilised by MOGs via
direct incorporation [5,6] or post-modification [7] with catalyti-
cally active metal ions reported by us and others. In comparison,
the latter strategy may offer a versatile entry into transition
metal gel catalysts, however, it is still a challenge to synthe-
sise a metallogel with free donor groups since the donor has
a tendency to coordinate. Herein, we wish to report the syn-
thesis and characterisation of a functionalised Fe3+–carboxylate
MOG based on 5-1H-benzo[d]imidazole-1,3-dicarboxylic acid (L1).
The reversibility of coordination bond formation in the resulting
gel was subsequently employed to immobilise Pd(II) coordination
catalyst.
Compared with traditional sol–gel materials, MOG-based mate-
rials may be of potential interest for catalytic applications due
∗
Corresponding author. Tel.: +86 20 8411 0539; fax: +86 20 8411 5178.
E-mail addresses: zhjyong@mail.sysu.edu.cn, bittxia@yahoo.com (J. Zhang).
1381-1169/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2009.11.001

98
J. Huang et al. / Journal of Molecular Catalysis A: Chemical 317 (2010) 97–103
2. Experimental
2.1. Methods
1710 (vs, C O), 1674 (vs), 1449 (w), 1383 (m), 1242 (s, CN), 1175
(w), 749 (w), 677 (w); EI MS (m/z (%)): 282 (3) [M+].
Scanning electron micrographs were recorded by using a
Quanta 400 Thermal FE Environment Scanning Electron Micro-
scope or a JSM-6330F Field Emission Scanning Electron Microscope.
Transmission electron micrographs were recorded on a JEOL
JEM-2010HR microscope. Nitrogen adsorption and desorption
isotherms were measured at 77 K on a BELSORP-max system. Pd
elemental analyses were performed by inductively coupled plasma
atomic emission spectroscopy using a TJA IRIS HR ICP instru-
ment. To prepare the samples for ICP-AES, a wet ash method with
HClO4–H3PO4 was used to digest the xerogel. The xerogel was dried
in vacuum prior to analysis and a sample of 4.0 mg was added in
HClO4–H3PO4 (v:v = 5:2, 2 mL) and heated until no yellow fume
produced. The mixture was filtered to remove some flotage and
diluted with HClO4–H3PO4 to a volume of 25 mL. For a list of the
materials used and their symbols, see Table 1.
2.4. Preparation of G1
L1 (11.3 mg, 0.04 mmol) in DMF (2 mL) and Fe(NO3)3·9H2O
(24.2 mg, 0.06 mmol) in DMF–H2O (2 mL, v:v = 1:1) were mixed
together, and then left to stand. A brown gel was obtained after
about 2 days. IR (KBr): ꢀ = 3422 (s, O–H), 2980 (m), 2780 (m), 1620
(m), 1572 (w), 1383 (vs), 1239 (w), 1217 (w), 1158 (w), 757 (w),
713 (w).
2.5. Preparation of G1–Pd
Method a: A solution of Pd(COD)Cl2 (5.7 mg, 0.02 mmol) in DMF
(2 mL) was allowed to stand above a sample of G1. After diffusion
for ca. 3 days, the supernatant turned from yellow to colourless
and was decanted, and the resultant brown gel was washed with
MeCN, yielding a gel, G1–Pd. IR (KBr): ꢀ = 3430 (s, O–H), 1620 (m),
1572 (m), 1440 (m), 1383 (ws), 1111 (w), 713 (w), 631 (w), 472 (w);
elemental analysis found (%) for the xerogel of G1–Pd: C 21.16, H
5.23, N 11.60.
Method b: A solution of L1 (11.3 mg, 0.04 mmol) and Pd(COD)Cl2
(5.7 mg, 0.02 mmol) in DMF (2 mL) were mixed with a solution of
Fe(NO3)3·9H2O (24.2 mg, 0.06 mmol) in DMF–H2O (2 mL, v:v = 1:1),
and then left to stand. A brown gel was obtained after about 4–5
days.
2.2. Preparation of
2-(3,5-dimethylphenyl)-1H-benzo[d]imidazole [8]
Anewmethodwasdeveloped:3,5-dimethylbenzoicacid (0.76 g,
5 mmol), 1,3-phenylenediamine (0.625 g, 5.8 mmol), polyphospho-
ric acid (5 mL) and phosphoric acid (3 mL) were heated at 190 ◦C
for 15 h. The dark green reaction mixture was cooled down to
about 100 ◦C, and poured into icy water (100 mL) while stirring
to get a black solid. The solid was dissolved in 100 mL of water
and the pH was adjusted to 12 with a NaOH solution (2 mol/L). Fil-
tered and washed with ethyl acetate to get a white solid (0.705 g,
yield 63%).
2.6. Preparation of G2–Pd
A Pd(II)-functionalised gel (G2–Pd) was prepared by plac-
ing an acetonitrile solution (2 mL) containing excess Pd(COD)Cl2
(14.3 mg, 0.05 mol) on top of Fe–BTC gel G2 (prepared by 60.5 mg
of Fe(NO3)3·9H2O, and 21.2 mg of H3BTC in 1 mL of ethanol, each
[3g]). After diffusion for 10 days, the supernatant was decanted and
the gel washed several times with CH3CN until the washings were
colourless. IR (KBr): ꢀ = 3425 (s), 1630 (s), 1450 (s), 1380 (vs), 758
(w), 712 (m), 623 (w), 461 (w); elemental analysis found (%) for
the xerogel of G2–Pd: C 24.60, H 4.46, N 1.72 and for the aerogel of
G2–Pd: C 27.06, H 3.99, N 1.46.
2.3. Preparation of 5-1H-benzo[d]imidazole-1,3-dicarboxylic acid
(L1)
2-(3,5-Dimethylphenyl)-1H-benzo[d]imidazole
(0.305 g,
1.37 mmol) was dissolved in a mixture of tert-butanol (8 mL)
and water (8 mL). A hot solution of KMnO4 (2.0 g) in H2O (20 mL)
was added at 80 ◦C in four portions during the course of 4 h. The
reaction was carried out for 28 h. Subsequently a Na2SO3 solution
was added to quench excess KMnO4. After stirring for 30 min
more, the reaction mixture was filtered to remove MnO2 and the
filtrate was adjusted to acidic using concentrated H2SO4 to obtain
a white solid. The solid was recrystallised from boiling water to
get the final product (0.201 g, yield 52%). L1 is soluble in DMF,
DMSO, boiling H2O, or alkaline solutions, but insoluble in alcohols,
acetone, THF, chloroform, diethyl ether, and ethyl acetate. 1H NMR
(300 MHz, DMSO-d6, ı): 8.62 (s, 2H; Ha), 8.55 (s, 1H; Hb), 8.34 (s,
2H; Hc), 7.58 (s, 2H; Hd); IR (KBr): ꢀ = 3410 (vs, O–H), 2930 (m),
2.7. Preparation of aerogels, ag1–Pd and ag2–Pd
The solvent in the wet gel was extracted with liquid CO2 in a
high pressure Soxhlet extractor. The extraction time was 8 h for
G1–Pd or 4 h for G2–Pd.
Table 1
List of the materials used and their symbols.
Symbol
Material
Symbol
H3BTC
Material
L1
G1
g1
G1–Pd
g1–Pd
ag1–Pd
Gel synthesised from L1 and Fe(NO3)3·9H2O in DMF–H2O
Xerogel of G1
G2
g2
G2–Pd
g2–Pd
ag2–Pd
Gel synthesised from H3BTC and Fe(NO3)3·9H2O in EtOH
Xerogel of G2
G1 modified with Pd2+
G2 modified with Pd2+
Xerogel of G1–Pd
Aerogel of G1–Pd
Xerogel of G2–Pd
Aerogel of G2–Pd

J. Huang et al. / Journal of Molecular Catalysis A: Chemical 317 (2010) 97–103
99
2.8. General procedure for the coupling reactions of aryl/pyridyl
halides with boronic acids
A typical procedure is given below on the reaction expressed
in entry 1 of Table 1. Iodobenzene (102.0 mg, 0.50 mmol), phenyl-
boronic acid (91.4 mg, 0.75 mmol), Na2CO3 (159.0 mg, 1.5 mmol),
MeOH (5 mL) and G1–Pd (0.5 mol% of Pd(II)), were introduced to
a flask under ambient atmosphere. The mixture was stirred with
a magnetic stirrer bar at 60 ◦C. H2O (10 mL) was added into the
resultant mixture and extracted three times with Et2O (10 mL). The
ethereal extract was evaporated to dryness and analysed by GC/MS
and NMR spectroscopy. For recovering, the gel was centrifuged
after reaction.
Fig. 2. Schematic representation of use of dynamic metallogel, G1, as a support for
catalyst.
Data for 4-(3,5-dimethylphenyl)pyridine: Pale yellow oil. 1H NMR
3
the imidazole N atom is favourable [10]. If Pd2+ is added to G1,
Fe–N(imidazole) bond may cleave and new Pd–N bond may be
formed instead due to the reversibility of coordination bond forma-
tion [11] (Fig. 2). In addition Pd2+ may be more strongly anchored in
imidazole-functionalised carboxylate gel network (e.g. G1) than in
a “pure” carboxylate gel (e.g. Fe3+–benzenetricarboxylate gel, G2,
see below). Such a new strategy may broaden the scope of available
gels for catalyst immobilisation, together with other strategies of
incorporation of catalytically active units directly in the framework
of metallogels (Fig. 3a) [5,6], and of post-modification of metallo-
gels with free groups available for coordination (Fig. 3b) [7].
To prove our hypothesis, two Pd(II)-functionalised gels were
prepared via a post-modification method. G1 was modified by
placing a DMF or MeCN solution containing Pd(COD)Cl2 in with
the molar ratio of Pd:L1 = 1:2 on top of the gel G1. After diffu-
sion, a brown Pd(II) gel, G1–Pd, was obtained. For comparison,
a Pd(II)-functionalised gel (G2–Pd) was prepared by topping a
Fe–tricarboxylate gel (G2, prepared by Fe(NO3)3·9H2O and 1,3,5-
benzenetricarboxylic acid in ethanol) with an MeCN solution
containing excess Pd(COD)Cl2. It is worth mentioning that G1–Pd
could also be prepared via direct gelation by mixing a DMF solution
of L1 and Pd(COD)Cl2 and a DMF/H2O solution of Fe(NO3)3·9H2O
with prolonged gelation time (4–5 days). Although the latter
method gave a more homogeneous gel catalyst, all the samples
mentioned below were prepared by the former method in order
to give a blueprint of the MOG route to supported catalyst.
The resulting gels were subjected to elemental analyses. The
weight percentage of Pd atoms in the xerogel of G1–Pd (g1–Pd)
was determined to be 3.75% by inductively coupled plasma spec-
troscopic analysis, higher than that in g2–Pd (2.28%) even if only
a limited amount of Pd(COD)Cl2 was used for G1 and excess for
G2. The theoretical Pd value would be 5.3% if 100% of the used
Pd(II) was immobilised. The results show that the entrapment of
Pd was more efficient in G1 than in G2 as expected, resulting in
a concentration decrease of the supernatant from 10−2 mol L−1 to
2.4 × 10−3 mol L−1 after standing for 3 days. The surface analysis by
XPS revealed the presence of Fe and Pd in both the power xero-
gel samples of G1–Pd and G2–Pd. Both the XPS spectra showed a
(300 MHz, DMSO-d6): ı = 8.57 (d, 2H, JH–H = 5.5 Hz), 7.63 (d, 2H,
3JH–H = 6.0 Hz), 7.36 (s, 2H), 7.07 (s, 1H), 2.33 (s, 6H).
Data for 3-(3,5-dimethylphenyl)pyridine: Pale yellow oil. 1H NMR
3
(300 MHz, DMSO-d6): ı = 8.82 (s, 1H), 8.52 (d, 1H, JH–H = 4.7 Hz),
3
8.00 (d, 1H, JH–H = 8.0 Hz), 7.43 (m, 1H), 7.30 (s, 2H), 7.04 (s, 1H),
2.34 (s, 6H).
3. Results and discussion
3.1. Gelation study and Pd(II) loading
The ligand 5-1H-benzo[d]imidazole-1,3-dicarboxylic acid (L1,
Table 1) was designed with three functional groups in which
two carboxylic groups are necessary to coordinate to Fe3+
to form a gel network [7], and the imidazole group coordi-
nates to Fe3+ less strongly according to the HSAB matching. L1
was synthesised through oxidation of 2-(3,5-dimethylphenyl)-
1H-benzo[d]imidazole by KMnO4. DMF, DMSO, and other mixed
solvents were tested as solvent for gelation and gels were success-
fully obtained in DMF–H2O (referred to as G1), DMF or DMF–MeOH,
judged by turning the sample upside down (Fig. 1). The time of met-
allogel formation increased either in DMF or in DMF–MeOH (about
9 days) compared with in DMF–H2O (2 days). The gel obtained in
DMF/H2O had visually better mechanical strength. Metal–ligand
ratios also played an important role in the gel formation. Gels could
be obtained in different L1/Fe ratios from 1:1, 2:3 to 1:2, while only
a weak flowing gel was obtained when the ratio was increased to
2:1. The colour of the metallogels turned gradually from orange to
brown with increased L1/Fe ratios. The gel with L1/Fe ratio of 1:2
was transparent, and the other gels were turbid. When the gel G1
was heated, no gel–sol transition was observed indicating a robust
gel network was formed.
In G1, the hard metal ion Fe3+ strongly coordinates to car-
boxylate groups to form the gel network as shown previously [4],
however, Fe3+ may also coordinate to the imidazole group less
strongly [9]. In contrast, binding of a soft metal ion like Pd2+ via
Fig. 3. Immobilisation of transition metal catalysts by metal–organic gels via direct
incorporation (a) or post-modification (b).
Fig. 1. A photographic image of metallogel G1.

100
J. Huang et al. / Journal of Molecular Catalysis A: Chemical 317 (2010) 97–103
Table 2
Comparison of ꢁꢀ for the gels/xerogels/aerogels.
ꢀasym(CO2)/cm−1
ꢀsym(CO2)/cm−1
ꢁꢀ/cm−1
L1
g1
1710
1620
1620
1703
1630
1630
1630
1449
1440
1440
1450
1450
1450
1450
261
180
180
253
180
180
180
g1–Pd
H3BTC
g2
g2–Pd
ag2–Pd
signal at 712 eV in the Fe 2p3/2 region, and analysis in the Pd 3d3/2
and 3d5/2 regions showed two signals at 338 eV and 343 eV, respec-
tively, confirming attachment of the Pd(II) ions to the gel network.
The Pd/Fe ratio was calculated to be about 1:2 for g1–Pd.
FT-IR spectra confirmed the presence of carboxylate in the metal
coordination sphere for the gels and the Pd(II) gels (Table 2). In the
gels, the bands due to the coordinated carboxylates are shifted to
lower frequencies with respect to the free ligands. For the xerogels,
strong bands at around 1380 cm−1 are due to NO3−. The difference
between ꢀasym(CO2) and ꢀsym(CO2) of around 180 cm−1 suggests
that the carboxylate ligands are bridging–chelating. The FT-IR spec-
tra of the Pd-loaded gel, G1–Pd and G2–Pd, generally displayed
similar bands to that of G1 and G2, respectively.
The aerogels were prepared by supercritical CO2 drying to
remove the solvent in the gel matrix. The aerogels were subjected
to scanning electron micrograph (SEM) and SEM images of the
aerogels of G1–Pd (Fig. 4a) and G2–Pd (Fig. 4b) revealed each an
interconnected macroporous network of globular nanometer-sized
particles of diameter ca. 100 nm. Transmission electron microscopy
(TEM) images of the xerogel of G1–Pd confirmed a continuous
Fig. 5. Nitrogen adsorption–desorption isotherm of the aerogels of G1–Pd and
G2–Pd.
amorphous network with connected particles (Fig. 4c and d).
Adsorption isotherm of both the aerogels performed using N2 gas
as the adsorbate at liquid nitrogen temperature exhibited a type
IV adsorption branch with a combination of H1- and H3-type hys-
teresis loops (Fig. 5) [12]. Detailed BET analyses showed that the
aerogel of G1–Pd has a BET surface area of 38.8 m2 g−1 and total
pore volume of 0.13 cm3 g−1. In contrast, the aerogel of G2–Pd is
highly porous with the BET surface area of 482 m2 g−1 and total
pore volume of 0.36 cm3 g−1. In both samples, the rise in the N2
uptake at P/P0 > 0.8 in the adsorption isotherms may arise in part
Fig. 4. (a and b) SEM images of the aerogels of G1–Pd (a) and G2–Pd (b), and (c and d) TEM images of the xerogel of G1–Pd.

J. Huang et al. / Journal of Molecular Catalysis A: Chemical 317 (2010) 97–103
101
tion (Table 3). The coupling reactions were examined in different
solvents, such as MeOH, DMF, and H2O at 60 ◦C under aerobic
conditions with 0.5 mol% catalyst loading. An excellent yield of
91% after 15 min was achieved in MeOH in the reaction of bro-
mobenzene and phenylboronic acid, whereas poor yields were
achieved in DMF (9% after 15 min, 18% after 5 h) or H2O (10%
after 15 min). Thus MeOH was chosen as solvent in the follow-
ing reactions. G1–Pd efficiently catalysed the coupling reactions
of aryl halides with phenylboronic acids. The coupling reactions
of iodobenzene with phenylboronic acid gave nearly quantitative
yield catalysed by the wet gel of G1–Pd within 0.5 h. The wet gel
of G1–Pd also showed excellent activity towards aryl bromides. 4-
Bromoacetophenone could couple with boronic acid quantitatively
after 0.5 h. The reactions of bromobenzene and 4-bromoanisole
with phenylboronic acid afforded yields of 91% after 15 min and
72% after 0.5 h, respectively. As expected, aryl chlorides are stub-
born and G1–Pd showed relatively modest activity towards them.
Chlorobenzene, 4-chloroacetophenone or 4-chloroanisole did not
give attractive results catalysed by the wet gel or xerogel, and a best
yield of 27% was achieved for chlorobenzene after 3.0 h in DMF at
110 ◦C.
G1–Pd was also shown to be an effective catalyst in the cou-
pling reactions of bromopyridines and 3,5-dimethylphenylboronic
acid/3,5-difluorophenylboronic acid (Table 4). However, G1–Pd
had no activity in the coupling of aryl halides (iodobenzene or
4-bromoacetophenone) with 4-pyridylboronic acid in the tested
reactions. The catalytic activity of the xerogel of G1–Pd was also
examined. As expected, the xerogel generally showed lower activ-
ity compared to the wet gel probably due to the lower surface area
and decreased substrate accessibility of catalytic sites.
Fig. 6. TG curve of the aerogels of G1–Pd and G2–Pd under nitrogen flow.
from textural interparticular porosity associated with the meso-
and macro-structures of the samples. The reduced microporosity
for the aerogel of G1–Pd may result from the lower concentration of
the gelator, which ledtoclosureofmicropores inthe dryingprocess.
The aerogel of G2–Pd contained both micropores and mesopores,
which microporous nature of the network was confirmed by the
steep uptake of nitrogen at very low relative pressures [13].
The thermal stability of the aerogels was investigated with ther-
mogravimetric analysis (TGA) under nitrogen (Fig. 6). From room
temperature to 267 ◦C, an observed 23.7% weight loss for the aero-
gel of G1–Pd may be due to physisorbed or chemisorbed water
or ethanol and a weight loss of 56.4% above 267 ◦C is accounted
for the framework decomposition. The aerogel of G2–Pd showed
similar thermal behaviour, and the TG analysis showed a weight
loss of 29.1% from room temperature to 246 ◦C and the framework
collapsed from 246 ◦C to 500 ◦C.
To study its origin of the excellent catalytic activity of G1–Pd,
coupling of phenylboronic acid and 4-bromoanisole were per-
formed as a control reaction (Fig. 7) due to its importance in
comparing highly active palladium catalysts [14]. In contrast to the
good yield of 72% after 0.5 h catalysed by the gel of G1–Pd, no cat-
alytic activity was observed in the absence of palladium (no Pd,
G1 itself as catalyst), and only poor activity was achieved in the
presence of only Pd(COD)Cl2 (1% of Pd) and 2 equiv. of benzimida-
zole (17% yield after 2 h), or in the presence of Pd(COD)Cl2 and L1
only with no gelation (12% yield after 5 h). These results confirm
that the observed good catalytic behaviour is unique to the immo-
3.2. Catalytic activity
The catalytic activity of G1–Pd arising from its palladium(II)
centre was evidenced in the Suzuki–Miyaura cross-coupling reac-
Table 3
Suzuki–Miyaura cross-coupling of aryl halides with boronic acids.a
.
Entry
Catalyst
X, R
GC yield/% (time/h)
1
2
3
G1–Pd (wet gel)
G1–Pd (xerogel)
G1–Pd (wet gel)
X = I, R = H
X = I, R = H
X = Br, R = H
93 (0.25), >99 (0.5)
92 (0.25), >99 (0.5)
91 (0.25)b
4
5
6
7
8
9
10
11
12
13
14
15
G1–Pd (wet gel)
G1–Pd (wet gel)
G1–Pd (xerogel)
G1–Pd (wet gel)
G1–Pd (xerogel)
G1–Pd (wet gel)
G1–Pd (xerogel)
G1 (no Pd)
X = Br, R = COCH3
X = Br, R = OCH3
X = Br, R = OCH3
X = Cl, R = H
95 (0.25), >99 (0.5)
54 (0.25), 72 (0.5), 74 (2.0)
29 (0.25), 57 (0.5), 69 (1.0), 71 (2.0)
20 (0.25), 26 (0.5), 29 (1.0), 46 (11)c
8 (0.5), 10 (2.0)
X = Cl, R = H
X = Cl, R = COCH3
X = Cl, R = COCH3
X = I, R = H
X = Br, R = OCH3
X = Br, R = OCH3
X = Br, R = OCH3
X = Br, R = OCH3
14 (1.0), 18 (3.0)c
6 (1.0), 12 (3.0)c
Trace (2.0)
G2–Pd (wet gel)
10 (0.25), 12 (0.5), 19 (1.0), 30 (2.0), 44 (3.0)
16 (1.0), 24 (2.0), 26 (3.0)
7 (0.25), 15 (1.0), 17 (2.0)
12 (0.5)b
Pd(COD)Cl2 (1 mol%)
Pd(COD)Cl2 (1 mol%) + 2 equiv. benzimidazole
Pd(COD)Cl2 + L1
a
Reaction conditions: 0.5 mmol of aryl halide, 0.75 mmol of PhB(OH)2, 0.5 mol% of Pd(II) of the catalyst, 1.5 mmol of Na2CO3, 6 mL of CH3OH at 60 ◦C under ambient
atmosphere.
b
Yield remained unchanged till 5 h.
c
Performed in DMF at 110 ◦C.

102
J. Huang et al. / Journal of Molecular Catalysis A: Chemical 317 (2010) 97–103
Table 4
Suzuki–Miyaura cross-coupling catalysed by 0.5 mol% of Pd(II) of G1–Pd.
.
Entry
Catalyst
PyBr
R
GC yield/% (time/h)
1
2
3
4
5
6
7
8
9
Wet gel
Xerogel
Wet gel
Xerogel
Wet gel
Xerogel
Wet gel
Wet gel
Xerogel
Xerogel
4-Bromopyridine
4-Bromopyridine
3-Bromopyridine
3-Bromopyridine
2-Bromopyridine
2-Bromopyridine
4-Bromopyridine
3-Bromopyridine
3-Bromopyridine
2-Bromopyridine
R = Me
R = Me
R = Me
R = Me
R = Me
R = Me
R = F
R = F
R = F
R = F
70 (0.25), 97 (0.5), >99 (1.0)
56 (0.25), 82 (0.5), 91 (0.75), >99 (1.25)
95 (0.25), >99 (0.5)
63 (0.25), 71 (1.0), 77 (2.0)
39 (0.25), 42 (0.5), 67 (1.0), 68 (2.0), 76 (3.0), 86 (6.0)
10 (0.25), 30 (1.0), 32 (2.5)
58 (3.0)
13 (1.0), 39 (3.0), 77 (7.0), 93 (27)
2 (0.5), 4 (4.0), 8 (5.0)
3 (0.5), 10 (1.0), 16 (2.0)
10
Table 5
bilisation of Pd(II) in the gel matrix of G1–Pd. In fact, solid-phase
catalysts are commonly difficult to retain or raise their activities
compared to their homogeneous counterparts, even if a better per-
formance of immobilised catalysts than homogeneous catalysts has
been reportedin alimitednumberofcases, for example, some solid-
supported Pd catalysts show higher activities than homogeneous
catalysts [15]. The higher activity of G1–Pd is probably due to the
higher stability and confined environment or cooperativity result-
ing from the entrapment of catalytic centres in the supporting gel
network.
As an important feature of G1–Pd, its dynamic support prop-
erty arising from the introduction of benzimidazole group was
evidenced in the comparison of the catalytic activity of G1–Pd and
G2–Pd (no benzimidazole group, “purely” carboxylate) (Fig. 7). In
the coupling reaction of 4-bromoanisole with phenylboronic acid,
a lower modest yield of 44% was achieved after 4 h catalysed by
the gel of G2–Pd. Thus the introduction of benzimidazyl group
is believed to play an important role in the improvement of cat-
alytic activity of G1–Pd. The results therefore agree well with our
hypothesis that G1 is dynamic.
Suzuki–Miyaura coupling of 4-bromopyridine with 3,5-difluorophenylboronic acid
using G1–Pd as recovered catalyst.
Entry
Run
Time/h
GC yield/%
1
2
3
4
1st
3
3
3
58
53
45
26
2nd
3rd
4th
16
was readily recovered by simple filtration due to their heteroge-
neous nature. As an example, the gel could be reused for three runs
with yields of 58%, 53%, and 45%, respectively, in the reaction of 4-
bromopyridine and 3,5-difluorophenylboronic acid. The catalytic
activity was gradually diminished in the first three runs, and there-
after the activity dropped considerably (26% yield, after 16 h) in the
fourth run. This trend was also observed for the xerogel of G1–Pd,
G2–Pd and other reactions (e.g. the coupling of 4-bromopyridine
and 3,5-dimethylphenylboronic acid), which was also reported for
some other supported Pd catalysts [16]. It was to note that no palla-
dium black was formed and collapsing of gels did not appear in the
reactions. The reactivity loss is attributed to slow continuous leach-
ing of the Pd active component resulting from destruction of the gel
catalyst by the reaction medium. Therefore stable MOG systems
need to be investigated for future gel catalysts.
Interestingly, G1–Pd could be reused for several times in con-
trast to its homogeneous counterparts (Table 5). The gel catalyst
4. Conclusions
In summary, we have described the anchoring of palladium(II)
to a dynamic Fe–carboxylate network by a metal–organic gel
methodology and their use as efficient and recoverable catalysts
for Suzuki–Miyaura coupling of aryl/pyridyl bromides with boronic
acids under mild conditions. The approach may theoretically work
for other soft metal catalysts such as Pt, Rh, and Ir as well. Interest-
ingly, the supported Pd2+ gel catalyst showed enhanced catalytic
activity in the Suzuki coupling under mild conditions, compared
with its homogeneous analogue. This behaviour is observed for
metallogel for the first time, to the best of our knowledge. It may
offer a new approach to active supported catalysts. The recyclability
of the catalyst was performed in the reactions of 4-bromopyridine
and 3,5-difluorophenylboronic acid or 3,5-dimethylphenylboronic
acid. The catalyst could be reused for several times.
As a novel strategy, the present method paves a way to employ
the unique dynamic property of MOGs to immobilise transition
metal catalysts. The Fe–N(imidazole) bond in the Fe–carboxylate
gel, G1, may break to form new Pd–N bond in the presence of
Fig. 7. Suzuki–Miyaura coupling of 4-bromoanisole and phenylboronic acid (cata-
lyst: circle, G1–Pd (with 0.5 mol% of Pd); square, G2–Pd (with 1 mol% Pd); triangle,
L1/Pd(COD)Cl2 = 2:1 (with 0.5 mol% of Pd); star, G1 or G2).

J. Huang et al. / Journal of Molecular Catalysis A: Chemical 317 (2010) 97–103
103
Pd2+. This bond cleavage and generation arises from their different
binding ability with imidazole. Different from direct functionalisa-
tion with organic ligand or catalytically active metal complex via
sol–gel method or other methods [17], the reversibility of coor-
dination bond formation in MOGs may offer a new chance for
immobilisation of transition metal complexes. Future work is to
develop chemically, mechanically and thermally stable MOG in
various reaction media for heterogeneous catalysts. This approach
may also be interesting in preparation of various metal-containing
materials by incorporation of metal complexes into a networked
matrix.
(c) L. Applegarth, N. Clark, A.C. Richardson, A.D.M. Parker, I. Radosavljevic-
Evans, A.E. Goeta, J.A.K. Howard, J.W. Steed, Chem. Commun. (2005) 5423;
(d) Q. Liu, Y. Wang, W. Li, L. Wu, Langmuir 23 (2007) 8217;
(e) W.L. Leong, A.Y.-Y. Tam, S.K. Batabyal, L.W. Koh, S. Kasapis, V.W.-W. Yam,
J.J. Vittal, Chem. Commun. (2008) 3628;
(f) W. Lu, Y.-C. Law, J. Han, S.S.-Y. Chui, D.-L. Ma, N. Zhu, C.-M. Che, Chem. Asian
J. 3 (2008) 59;
(g) Q. Wei, S.L. James, Chem. Commun. (2005) 1555;
(h) J. Zhang, X. Xu, S.L. James, Chem. Commun. (2006) 4218.
[4] J.Y. Ying, C.P. Mehnert, M.S. Wong, Angew. Chem. Int. Ed. 38 (1999) 56.
[5] Y.-R. Liu, L. He, J. Zhang, X. Wang, C.-Y. Su, Chem. Mater. 21 (2009) 557.
[6] (a) B. Xing, M.-F. Choi, B. Xu, Chem. Eur. J. 8 (2002) 5028;
(b) Y.M.A. Yamada, Y. Maeda, Y. Uozumi, Org. Lett. 8 (2006) 4259;
(c) T. Tu, W. Assenmacher, H. Peterlik, R. Weisbarth, M. Nieger, K.H. Dötz,
Angew. Chem. Int. Ed. 46 (2007) 6368;
(d) T. Tu, W. Assenmacher, H. Peterlik, G. Schnakenburg, K.H. Dötz, Angew.
Chem. Int. Ed. 47 (2008) 7127.
[7] J. Zhang, X. Wang, L. He, L. Chen, C.-Y. Su, S.L. James, New J. Chem. 33 (2009)
1070.
[8] J.C. Lewis, A.M. Berman, R.G. Bergman, J.A. Ellman, J. Am. Chem. Soc. 130 (2008)
2493.
[9] (a) M.A. Martinez-Lorente, J.-P. Tuchagues, V. Petrouleas, J.M. Savariault, R.
Poinsot, M. Drillon, Inorg. Chem. 30 (1991) 3587;
Acknowledgements
We gratefully acknowledge the SEM (SRF for ROCS), the NSF of
China (Grants 20525310, 20773167, J0730420) and the 973 Pro-
gram of China (Grant 2007CB815302) for support.
(b) D. Lee, S.J. Lippard, Inorg. Chem. 41 (2002) 2704.
[10] S.U. Milinkovic, T.N. Parac, M.I. Djuran, N.M. Kostic, J. Chem. Soc., Dalton Trans.
(1997) 2771–2776.
References
[11] A. Langner, S.L. Tait, N. Lin, C. Rajadurai, M. Ruben, K. Kern, Proc. Natl. Acad. Sci.
U.S.A. 104 (2007) 17927.
[12] S.J. Gregg, K.S.N. Sing, Adsorption, Surface Area and Porosity, 2nd ed., Academic
Press, 1982.
[13] After we submitted our manuscript, a report was published to describe the
porosity of G2 aerogel J.M.R. Lohe, M. Rose, S. Kaskel, Chem. Commun. (2009)
6056.
[14] J. Dupont, C.S. Consorti, J. Spencer, Chem. Rev. 105 (2005) 2527.
[15] (a) D. Choudhary, S. Paul, R. Gupta, J.H. Clark, Green Chem. 8 (2006) 479;
(b) B. Karimi, A. Zamani, J.H. Clark, Organometallics 24 (2005) 4695;
(c) V. Farina, Adv. Synth. Catal. 346 (2004) 1553.
[1] (a) G. Øye, W.R. Glomm, T. Vrålstad, S. Volden, H. Magnusson, M. Stöcker, J.
Sjöblom, Adv. Colloids Interface Sci. 123–126 (2006) 17;
(b) L.L. Hench, J.K. West, Chem. Rev. 90 (1990) 33;
(c) B. Dunn, J.I. Zink (Eds.), Acc. Chem. Res. 40 (2007) 729;
(d) C. Moreno-Castilla, F.J. Maldonado-Hódar, Carbon 43 (2005) 455;
(e) M. Pagliaro, R. Ciriminna, G. Palmisano, Chem. Soc. Rev. 36 (2007) 932.
[2] (a) M.L. Toebes, J.A. van Dillen, K.P. de Jong, J. Mol. Catal. A: Chem. 173 (2001)
75;
(b) Z. Lu, E. Lindner, H.A. Mayer, Chem. Rev. 102 (2002) 3543;
(c) J.J.E. Moreau, M.W.C. Man, Coord. Chem. Rev. 178–180 (1998) 1073;
(d) M.T. Reetz, Adv. Mater. 9 (1997) 943;
[16] N.T.S. Phan, M.V.D. Sluys, C.W. Jones, Adv. Synth. Catal. 348 (2006) 609.
[17] (a) L. Yin, J. Liebscher, Chem. Rev. 107 (2007) 133;
(b) N. Gurbuz, I. Özdemir, B. C¸ etinkaya, T. Sec¸ kin, Appl. Organometal. Chem. 17
(2003) 776;
(e) S. Benyahya, F. Monnier, M. Taillefer, M.W.C. Man, C. Bied, F. Ouazzani, Adv.
Synth. Catal. 350 (2008) 2205;
(f) J.W. Kriesel, T.D. Tilley, Adv. Mater. 13 (2001) 1645.
[3] (a) F. Fages, Angew. Chem. Int. Ed. 45 (2006) 1680;
(b) G. De Paoli, Z. Dzolic, F. Rizzo, L. De Cola, F. Vögtle, W.M. Müller, G. Richardt,
M. Zinic, Adv. Funct. Mater. 17 (2007) 821;
(c) M. Trilla, G. Borja, R. Pleixats, M.W.C. Man, C. Bied, J.J.E. Moreau, Adv. Synth.
Catal. 350 (2008) 2566.