Construction of covalent organic framework for catalysis: Pd/COF-LZU1 in Suzuki-Miyaura coupling reaction

Article abstract of DOI:10.1021/ja206846p

Covalent organic frameworks (COFs) are crystalline porous solids with well-defined two- or three-dimensional molecular structures. Although the structural regularity provides this new type of porous material with high potentials in catalysis, no example has been presented so far. Herein, we report the first application of a new COF material, COF-LZU1, for highly efficient catalysis. The easily prepared imine-linked COF-LZU1 possesses a two-dimensional eclipsed layered-sheet structure, making its incorporation with metal ions feasible. Via a simple post-treatment, a Pd(II)-containing COF, Pd/COF-LZU1, was accordingly synthesized, which showed excellent catalytic activity in catalyzing the Suzuki-Miyaura coupling reaction. The superior utility of Pd/COF-LZU1 in catalysis was elucidated by the broad scope of the reactants and the excellent yields (96-98%) of the reaction products, together with the high stability and easy recyclability of the catalyst. We expect that our approach will further boost research on designing and employing functional COF materials for catalysis.

Full text of DOI:10.1021/ja206846p

ARTICLE
pubs.acs.org/JACS
Construction of Covalent Organic Framework for Catalysis:
Pd/COF-LZU1 in SuzukiÀMiyaura Coupling Reaction
San-Yuan Ding,† Jia Gao,† Qiong Wang,†,‡ Yuan Zhang,† Wei-Guo Song,‡ Cheng-Yong Su,§ and Wei Wang*,†
†State Key Laboratory of Applied Organic Chemistry, College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou,
Gansu 730000, China;
‡Beijing National Laboratory for Molecular Sciences (BNLMS); Laboratory for Molecular Nanostructures and Nanotechnology,
Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China;
§School of Chemistry and Chemical Engineering, Sun Yat-Sen University, Guangzhou 510275, China
S Supporting Information
b
ABSTRACT: Covalent organic frameworks (COFs) are crystalline porous solids with
well-defined two- or three-dimensional molecular structures. Although the structural
regularity provides this new type of porous material with high potentials in catalysis, no
example has been presented so far. Herein, we report the first application of a new COF
material, COF-LZU1, for highly efficient catalysis. The easily prepared imine-linked
COF-LZU1 possesses a two-dimensional eclipsed layered-sheet structure, making its
incorporation with metal ions feasible. Via a simple post-treatment, a Pd(II)-containing
COF, Pd/COF-LZU1, was accordingly synthesized, which showed excellent catalytic
activity in catalyzing the SuzukiÀMiyaura coupling reaction. The superior utility of Pd/
COF-LZU1 in catalysis was elucidated by the broad scope of the reactants and the
excellent yields (96À98%) of the reaction products, together with the high stability and easy recyclability of the catalyst.
We expect that our approach will further boost research on designing and employing functional COF materials for catalysis.
’ INTRODUCTION
water, and most of the organic solvents. Another is that the COF
material should incorporate with catalytic active sites. The most
well-known boron-containing COFs linked by boroxine or
boronate-ester groups are unfortunately not stable to moisture,
although they display high thermal stabilities.27À29 In contrast,
the triazole-,15 the imine-,16 and hydrazone-based30 COF mate-
rials, being highly stable in water and most organic solvents, meet
the first requirement as robust catalysts. Moreover, it has been
well demonstrated31 in coordination chemistry that the imine-
type (Schiff base) ligands are versatile in incorporating a variety
of metal ions. This intrigued us to explore the possibility of using
imine-linked metal-ion-incorporated COF materials for catalysis.
However, the only reported example of imine-linked COF, i.e.,
COF-300,16 possesses a three-dimensional diamond structure in
which the nitrogen atoms of the imine-bonds are far away from
each other (the minimum distance is ∼6.6 Å). This arrangement
makes COF-300 less possible to coordinate with metal ions
efficiently. Accordingly, we synthesized a new imine-linked COF
material (denoted as COF-LZU1) with a two-dimensional (2D)
layered-sheet structure (Figure 1). The eclipsed layered-sheet
arrangement renders the distance of eclipsed nitrogen atoms in
adjacent layers as ∼3.7 Å, which makes the synthesized COF-
LZU1 material an ideal scaffold for incorporating a variety of
metal ions.
As a new type of crystalline porous material, covalent organic
frameworks (COFs) are ingeniously constructed with organic
moieties linked by strong covalent bonds through the principles
of reticular chemistry.1À7 Similar to those found in crystalline
inorganic zeolites8 and metal organic frameworks (MOFs),9,10
COF materials possess well-defined and predictable two- or
three-dimensional pore structures.2,3 In comparison with amor-
phous porous materials, these crystalline porous materials have
superior potential in adsorption, in gas storage, and especially
in catalysis because of the structural regularity.11 Indeed, inor-
ganic zeolites have been widely used as catalysts in refining and
petrochemical industries.8,12 Recent research has also demon-
strated the possibility of employing MOF materials in catalysis.13
On the contrary, the catalytic application of COF materials has
never been reported so far. Since the first discovery of COF
materials in 2005,2 research in this area has mainly focused on the
construction of new COF structures with the aim of increasing the
surface area and pore volume,4À7,14À19 together with some rare
examples of applying the synthesized COFs in gas storage20À22
and photoelectricity.23À26 In this work, we report the first ex-
ample of design, synthesis, and application of an imine-linked
COF material for catalysis.
As in the cases of other solid catalysts, two basic but important
issues should be taken into account in order to design and screen
suitable COF candidates for catalytic applications. One is that the
COF catalyst must show high stability to thermal treatments,
Received: July 22, 2011
Published: October 25, 2011
r
2011 American Chemical Society
19816
dx.doi.org/10.1021/ja206846p J. Am. Chem. Soc. 2011, 133, 19816–19822
|

Journal of the American Chemical Society
ARTICLE
Figure 1. Construction of COF-LZU1 and Pd/COF-LZU1. Schematic representation for the synthesis of COF-LZU1 and Pd/COF-LZU1 materials
(a). Proposed structures of COF-LZU1 (b) and Pd/COF-LZU1 (c, d) possessing regular microporous channels (diameter of ∼1.8 nm), simulated with
a 2D eclipsed layered-sheet arrangement. C: blue, N: red, and brown spheres represent the incorporated Pd(OAc)2. H atoms are omitted for clarity.
As an example, the palladium(II)-coordinated COF material
(denoted as Pd/COF-LZU1) was facilely synthesized via a
simple treatment of COF-LZU1 with Pd(OAc)2 at room tem-
perature. The structural preservation and robust Pd(II)-incor-
poration of the synthesized Pd/COF-LZU1 was verified by
powder X-ray diffraction (PXRD), solid-state NMR spectrosco-
py, and X-ray photoelectron spectroscopy (XPS). We then
applied the Pd/COF-LZU1 material in catalyzing the SuzukiÀ
Miyaura coupling reaction, an important reaction for the forma-
tion of CÀC bonds.32 The excellent catalytic activity of Pd/
COF-LZU1 was elucidated by the broad scope of the reactants
and the excellent yields (96À98%) of the reaction products,
together with the high stability and easy recyclability of the
catalyst. Incomparison with that of zeolitesor MOFs, the superior
activity of Pd/COF-LZU1 material should be attributed to its
unique structure, which offers efficient access to the active sites
and fast diffusion for the bulky products.
room temperature. The resulting solid was isolated by centrifugation, washed
with dichloromethane using Soxhlet extraction for 24 h, and then dried
at 80 °C under vacuum for 12 h to yield Pd/COF-LZU1 as a brown-colored
powder (166 mg, 79% yield). The Pd content was 7.13% as determined by
ICP. Anal. Cald for 0.077 nPd(OAc)2@(C6H4N)n: C 70.57; H 4.16;
3
N 13.05. Found: C 66.04; H 3.97; N 12.51. IR (powder, cmÀ1) 3339,
2862, 1696, 1617, 1495, 1411, 1250, 1147, 968, 881, 836, 730, 685. PXRD
[2θ (relative intensity)] 4.68 (100), 8.18 (6), 9.35 (4).
General Procedure for the SuzukiÀMiyaura Coupling
Reactions. In a typical run for catalytic activity test of Pd/COF-
LZU1, aryl halide (1.0 mmol), phenylboronic acid (183 mg, 1.5 mmol),
potassium carbonate (276 mg, 2.0 mmol), and Pd/COF-LZU1 (7.5 mg,
0.5 mol %) were added to 4 mL of p-xylene. The reaction mixture was
stirred at 150 °C under reflux at ambient atmosphere. After the reaction
was completed (monitored by TLC), the mixture was centrifugated, and
the solid was washed with dichloromethane (3 Â 5 mL). The combined
organic phase was washed with water (20 mL) to remove potassium
carbonate. The organic phase was then evaporated under reduce pressure
to leave the crude products which were further purified by column
chromatography over silica gel to obtain the desired products. In a recycle
test, p-nitrobromobenzene (202 mg, 1.0 mmol), phenylboronic acid
(183 mg, 1.5 mmol), potassium carbonate (276 mg, 2.0 mmol), and Pd/
COF-LZU1 (15 mg, 1.0mol %) in4 mLof p-xylene were used. After each
cycle, the catalyst was dried and then used directly without any further
treatment.
’ EXPERIMENTAL SECTION
Synthesis of COF-LZU1. 1,3,5-Triformylbenzene 1 (48 mg,
0.30 mmol) and 1,4-diaminobenzene 2 (48 mg, 0.45 mmol) were weighed
into a vial and dissolved in 3.0 mL of 1,4-dioxane. The mixture was
transferred into a glass ampule (volume ∼20 mL with a body length of
18 cm and a neck length of 9 cm), and to the mixture was added 0.6 mL
of 3.0 mol/L aqueous acetic acid. The glass ampule was flash frozen in a
liquid nitrogen bath, evacuated to an internal pressure of 19 mbar and
flame-sealed, reducing the total length by ∼10 cm. Upon warming to
room temperature, the suspension was placed in an oven at 120 °C and
left undisturbed for 3 days, yielding a yellow solid along the tube which
was isolated by centrifugation, washed with N,N-dimethylformamide
(3 Â 10 mL) and tetrahydrofuran (3 Â 10 mL), and dried at 80 °C under
vacuum for 12 h to yield COF-LZU1 as a yellow-colored powder
(72 mg, 90% yield). Anal. Cald for (C6H4N)n: C 80.00; H 4.44; N 15.55.
Found: C 76.34; H 4.61; N 14.20. IR (powder, cmÀ1) 3382, 2864, 1694,
1618, 1496, 1444, 1250, 1146, 968, 880, 832, 730, 685. PXRD [2θ
(relative intensity)] 4.70 (100), 8.05 (8), 9.47 (3), 12.45 (2), 26.22 (3).
Synthesis of Pd/COF-LZU1. Palladium acetate (45 mg, 0.2 mmol)
was dissolved in 10 mL of dichloromethane, and then COF-LZU1
(165 mg) was added. The mixture was kept stirring for 24 h at
Other detailed experimental descriptions can be found in the
Supporting Information [SI].
’ RESULTS AND DISCUSSION
As shown in Figure 1a, the imine-linked COF-LZU1 material
could be reproducibly synthesized by heating the suspension of
1,3,5-triformylbenzene 1 and 1,4-diaminobenzene 2 in a mixture
of 1,4-dioxane and aqueous acetic acid within a flame-sealed glass
ampule (see SI for details). The synthesized COF-LZU1 ma-
terial is insoluble in water or common organic solvents (such as
N,N-dimethylformamide, tetrahydrofuran, dimethyl sulfoxide,
acetone, trichloromethane, etc.). The Fourier transform infrared
(FT-IR) spectrum of COF-LZU1 shows a strong CdN stretch
at 1618 cmÀ1 (Figure S1, SI), indicating the formation of
19817
dx.doi.org/10.1021/ja206846p |J. Am. Chem. Soc. 2011, 133, 19816–19822

Journal of the American Chemical Society
ARTICLE
Figure 2. Left: (a) Observed PXRD pattern (black, the major reflections are assigned) and refined modeling profile (red) of COF-LZU1. (b) Difference
plot between the observed and refined PXRD patterns. (c) Simulated PXRD pattern for an eclipsed structure. (d) Simulated PXRD pattern for a
staggered structure. Right: (e,f) Eclipsed structure with the unit cell parameters indicated. (g,h) Staggered structure with the unit cell parameters
indicated. C: blue, N: red, and H atoms are omitted for clarity. For the staggered structure, alternating gray layers with undefined atoms are presented for
better visuality.
imine bonds. Meanwhile, the aldehyde (1695 cmÀ1) and amino
(3415 cmÀ1) bands of COF-LZU1 were greatly attenuated in
comparison with those of the starting materials 1 and 2
(Figure S1, SI), which offers further evidence for the forma-
tion of imine bonds via the condensation of aldehyde 1 and
primary amine 2. The residual signals at these wavenumbers
correspond to the terminal aldehyde and amino groups at the
edges of the COF-LZU1 material, respectively.
low-energy minima is very close to the c value derived from
Pawley refinement (3.7 Å). In comparison, the staggered model
gives the layer distance of 3.3 Å at low-energy minima. Therefore,
the structure of COF-LZU1 should preferably be determined as
the eclipsed layered-sheet arrangement shown in Figure 1b. The
layer structure of COF-LZU1 was also observed through scan-
ning electron microscopy (SEM) and transmission electron micro-
scopy (TEM). The SEM image of COF-LZU1 (Figure 3a) shows
the layered-sheet morphology and the boundaries of the TEM
images (Figure 3c and Figure S19, SI) show the layered-sheet
arrangement of COF-LZU1. The formation of the eclipsed
structure originates from the strong tendency for the hexagonal
units to form coplanar aggregates which could stabilize the πÀπ
stacking interactions between adjacent layers (distance of ∼3.7 Å).
The atomic-level construction of COF-LZU1 was eventually
verified by solid-state NMR spectroscopy. Figure 4a shows the
13C cross-polarization magic-angle spinning (CP/MAS) NMR
spectrum recorded for COF-LZU1. The 13C NMR peak at
∼157 ppm corresponds to the carbon atom of the CdN bond, the
formation of which36 is characteristic for the condensation reaction of
aldehyde 1 and primary amine 2. The signals at ∼122, 130, 137, and
148 ppm can be assigned to the carbon atoms of the phenyl groups,
while the minor peak at ∼191 ppm is ascribed to the carbon atoms of
terminal aldehyde groups in the COF-LZU1 network.
As mentioned above, being highly selective and sensitive
toward the incorporation of various metal ions, the imine-type
ligand is one of the most versatile ligands studied in coordination
chemistry.31 Moreover, the imine-based metal complexes have
also been used as homogeneous catalysts.37 Recently, amorphous
triazine-based frameworks were successfully applied as hetero-
geneous supports for noble metal catalysts.38À40 We were thus
intrigued to explore the unprecedented possibility of utilizing
COF materials in catalysis. Through a simple post-treatment of
COF-LZU1 with palladium acetate (Figure 1a), a Pd(II)-
containing COF-LZU1 material, Pd/COF-LZU1, was facilely pre-
pared (see SI for details). A comparison of the PXRD patterns
(Figure 5) of COF-LZU1 and Pd/COF-LZU1 revealed that the
crystal structure of COF-LZU1 was well preserved after the
Powder X-ray diffraction (PXRD) analysis with Cu Kα radi-
ation was used to determine the crystalline structure of COF-
LZU1. The PXRD pattern (Figure 2a, black) indicates that
COF-LZU1 is a microcrystalline material with a long-range
structure depicted in Figure 1b (see details in e and f of Figure 2).
The most intense peak with d spacing of 18.78 Å is attributed to
the (100) diffraction, while other diffraction peaks with d spacings
of 10.98, 9.33, 7.11, and 3.45 Å are attributed to the (110), (200),
(210), and (001) facets, respectively. Furthermore, no diffraction
peaks from the starting material 1 or 2 could be observed (Figure
S8, SI), indicating the sole formation of crystalline COF-LZU1
material. The lattice modeling and Pawley refinement (see SI for
details) of COF-LZU1 were conducted with the Materials Studio
(ver. 4.4) suite of programs.33 The most probable structure of
COF-LZU1 was simulated, analogous to that of COF-5,2 as the
eclipsed layered-sheets using the space group of P6/m with the
optimized parameters of a = b = 22.040 ((0.228) Å and c = 3.729
((0.040) Å (e and f of Figure 2). To our delight, the calculated
PXRD pattern (Figure 2c) of this hexagonal unit cell matched the
experimental profile in peak positions and relative intensities
quite well, with wRp of 4.70% and Rp of 3.68%, respectively.
An alternative staggered model was also predicted (Figure 2g).
However, the simulated PXRD pattern (Figure 2d) did not
match the observed data.34 In order to verify the preferred
eclipsed-structure of COF-LZU1, we further applied the force-
field-based molecular mechanics calculations33 to compare the
total energies of both models against the stacking distances35 (see
SI for details). The eclipsed model showed a distinct energetic
preference to the staggered model (Table S2 and Figure S10, SI).
Meanwhile, the layer distance (3.6 Å) for the eclipsed structure at
19818
dx.doi.org/10.1021/ja206846p |J. Am. Chem. Soc. 2011, 133, 19816–19822

Journal of the American Chemical Society
ARTICLE
Figure 3. SEM images of COF-LZU1 (a) and Pd/COF-LZU1 (b), TEM images of COF-LZU1 (c), Pd/COF-LZU1 (d), and Pd/COF-LZU1
after the fourth cycle use (e). The magnified pictures are shown in the SI for better visuality.
Figure 5. PXRD patterns of COF-LZU1 (black), Pd/COF-LZU1
(blue), Pd/COF-LZU1 after the first cycle (green) and after the fourth
cycle (red).
Figure 4. 13C CP/MAS NMR spectra of COF-LZU1 (a, black), Pd/
COF-LZU1 (b, blue), Pd/COF-LZU1 after the first cycle use (c, green)
and after the fourth cycle use (d, red). Asterisks denote spinning
sidebands. The assignments of 13C chemical shifts of COF-LZU1 were
indicated in the chemical structure. The minor signals at ∼179 and 23 ppm
(b) were assigned to the acetate groups of the incorporated Pd(OAc)2.
The signals at ∼43, 26, and 21 ppm (d) originated from the solvent.
shifted by 11 ppm. This confirms the strong incorporation of
Pd(OAc)2 with COF-LZU1. As shown in c and d of Figure 3, the
layer structure of COF-LZU1 also remained unchanged after the
treatment with palladium acetate. The well-dispersed palladium
species (black dots)42 could also be clearly observed in the TEM
images (Figures 3d and S20, SI). The Pd content in Pd/COF-
LZU1 is ∼7.1 ( 0.5 wt % as determined by ICP analysis,
corresponding to ∼0.5 Pd atom per unit cell (Figure 1c).
We then performed X-ray photoelectron spectroscopy (XPS)
measurements to further investigate the incorporation of palla-
dium within COF-LZU1. As shown in Figure 6 (black), the
binding energy (BE) of Pd3d5/2 in Pd/COF-LZU1 is 337.7 eV,
indicating that the Pd species is present in +2 state.43 On
the other hand, the BE of 337.7 eV for the Pd(II) species in
treatment with palladium acetate.41 The 13C CP/MAS NMR
spectrum of Pd/COF-LZU1 (Figure 4b) is almost identical to
that of COF-LZU1 in the phenyl areas, indicating also the
structural preservation of COF-LZU1 after the post-treatment.
Two additional minor peaks at ∼179 and 23 ppm are assigned
to the carbonyl and methyl groups of the incorporated
Pd(OAc)2. In comparison with those of free Pd(OAc)2 at
∼190 ppm (Figure S27, SI), the 13C chemical shift for the
carbonyl group in the incorporated Pd(OAc)2 is high-field
19819
dx.doi.org/10.1021/ja206846p |J. Am. Chem. Soc. 2011, 133, 19816–19822

Journal of the American Chemical Society
ARTICLE
Figure 7. Nitrogen adsorptionÀdesorption isotherms of COF-LZU1
(cycles) and Pd/COF-LZU1 (triangles). (Inset) Pore size distribution
of COF-LZU1 (cycles) and Pd/COF-LZU1 (triangles). Adsorption
and desorption points are represented by filled and empty symbols,
respectively.
Figure 6. XPS spectra of free Pd(OAc)2 (blue), Pd/Phen (red), and
Pd/COF-LZU1 (black).
Pd/COF-LZU1 shifted negatively by 0.7 eV in comparison with
that of 338.4 eV for free Pd(OAc)2 (Figure 6, blue). This
negative shift indicates the strong coordination of Pd(OAc)2
with the imine groups of COF-LZU1: the imine groups further
donate electrons to Pd(II) which makes the Pd species less
electron-deficient. In order to further confirm the imine groups
between the adjacent layers are indeed incorporating with palla-
dium acetate, a model complex Pd/Phen (structure shown in
Figure 6) was synthesized from 1,10-phenanthroline (Phen)
and Pd(OAc)2 (see SI for details). As shown in Figure 6 (red),
the XPS spectrum of the coordination complex Pd/Phen
[with the BE of 337.8 eV for the Pd(II) species] also shows a
negative shift of 0.6 eV in comparison with that of free
Pd(OAc)2. Moreover, in the 13C CP/MAS NMR spectrum
of Pd/Phen (Figure S28, SI), the carbonyl group (177 ppm)
of the incorporated Pd(OAc)2 also shows a high-field shift by
13 ppm, similar to the case of Pd/COF-LZU1. Therefore,
Pd(OAc)2 is most probably located in between the adjacent
layers of COF-LZU1 and efficiently coordinated with two
nitrogen atoms (distance of ∼3.7 Å) from different layers
(Figure 1).
The porosity and surface areas of COF-LZU1 and Pd/COF-
LZU1 were measured by nitrogen adsorptionÀdesorption anal-
ysis at 77 K (Figure 7). The application of the BrunauerÀ
EmmettÀTeller (BET) model resulted in the surface areas
of 410 and 146 m2 gÀ1 for COF-LZU1 (Figure S11, SI) and
Pd/COF-LZU1 (Figure S13, SI), respectively. The total pore
volumes were calculated as 0.54 cm3 gÀ1 (P/P0 = 0.99) and
0.19 cm3 gÀ1 (P/P0 = 0.98) for COF-LZU1 and Pd/COF-LZU1,
respectively. The pore size distributions of COF-LZU1 and Pd/
COF-LZU1 were calculated using nonlocal density functional
theory (NLDFT). As shown in Figure 7 (inset), the pore size
distribution plots of both COF-LZU1 and Pd/COF-LZU1
revealed a narrow pore width distribution at ∼1.2 nm. In
comparison with the case of COF-LZU1, the large decrease in
the surface area and the slight change in the pore size distribution
for Pd/COF-LZU1 indicate that the presence of palladium
acetate would not block the cavities of Pd/COF-LZU1 but
could enhance the weight significantly. The thermogravimetric
analysis (Figure S15, SI) identified that both COF-LZU1 and
Pd/COF-LZU1 possess good thermal stabilities (up to 310 and
270 °C, respectively), which meets the demands for potential
applications in catalysis.
The catalytic activity of Pd/COF-LZU1 catalyst was exam-
ined in one of the representative Pd-catalyzed reactions, i.e., the
SuzukiÀMiyaura coupling reaction. The SuzukiÀMiyaura cou-
pling reaction has been widely applied in homogeneous media
for the facile formation of CÀC bonds.32 However, its potential
application in industry is still limited due to the difficulty in
separating and recycling the expensive Pd-catalysts from the
product mixture. Using recyclable palladium catalysts is therefore
a promising solution to these problems.44,45 For the purpose of
comparison, the SuzukiÀMiyaura coupling reaction catalyzed by
Pd/COF-LZU1 was conducted under conditions similar to
those for the Pd-containing MOF,46 and the results are summar-
ized in Table 1. Catalyzed by 0.5 mol % of Pd/COF-LZU1, the
substituted aryl iodides with either an electron-donating (entry 1)
or an electron-withdrawing (entry 2) group afforded the cross-
coupling products in excellent yields (96À97%). The less
active bromobenzenes (entries 3À8) also gave excellent yields
(96À98%) within 2.5À4 h at 150 °C. Furthermore, when
0.1 mol % catalyst was used (entry 10), the reaction also worked
well but with a longer reaction time (5 h). All these results clearly
indicate that Pd/COF-LZU1 possesses high catalytic activity in
catalyzing the SuzukiÀMiyaura coupling reaction. In comparison
with the Pd(II)-containing MOF46 (entry 9), Pd/COF-LZU1
required less catalyst-loading and shorter reaction time and
showed higher reaction yield (entry 8).
We conducted the control experiments to confirm that the
reaction was indeed catalyzed by Pd/COF-LZU1 instead of free
Pd(OAc)2. In the absence of reactants, the mixture of Pd/COF-
LZU1, p-xylene and sodium carbonate was heated and stirred for
7 h at 150 °C. The filtrate from the hot mixture was then used to
catalyze the reaction of p-nitrobromobenzene and phenylboronic
acid. No conversion could be observed in this case and the ICP
analysis showed that less than 0.08% of Pd content (corres-
ponding to 0.0057 wt % Pd content in Pd/COF-LZU1) had
leached out into the filtrate. In the presence of reactants, the
catalyst was removed from the hot reaction mixture after ∼50%
conversion of reactants, and the isolated solution did not exhibit any
further reactivity under the same reaction conditions. Moreover, the
ICP analysis showed that essentially no Pd (<0.007% Pd content)
had leached out into the isolated solution. These studies identified
thatPd(OAc)2 is strongly coordinated within COF-LZU1 under the
19820
dx.doi.org/10.1021/ja206846p |J. Am. Chem. Soc. 2011, 133, 19816–19822

Journal of the American Chemical Society
ARTICLE
Table 1. Catalytic Activity Test of Pd/COF-LZU1 in the
SuzukiÀMiyaura Coupling Reaction
Table 2. Recycle Test of Pd/COF-LZU1 in the
SuzukiÀMiyaura Cross-Coupling Reaction of
p-Nitrobromobenzene and Phenylboronic Acid
entry
time (h)
yield (%)a
fresh
3
3
3
3
3
97
96
97
97
97
cycle 1
cycle 2
cycle 3
cycle 4
a Isolated yield.
by the moderate activity and relatively poor recyclability of Pd(II)-
containingMOFcatalyst inthe SuzukiÀMiyaura reaction.46 In the
case of zeolite catalysts, a general problem12 is that the structural
topology and small pore/channel size (normally <0.8 nm) would
cause severe diffusion problems, restricting their catalytic applications
for bulky systems. Moreover, further decrease of the void space
makes the postmodified zeolites less applicable for catalysis, espe-
cially for fine chemical catalysis.
a Unless otherwise noted, the reaction conditions are: aryl halide
(1.0 mmol), phenylboronic acid (1.5 mmol), K2CO3 (2.0 mmol), and
Pd/COF-LZU1 (0.5 mol %), 4 mL of p-xylene, 150 °C. b Isolated yield.
c Result from ref 46 (2.5 mol % Pd/MOF was used). d 0.1 mol % Pd/
COF-LZU1 was used.
’ CONCLUSION
In conclusion, for the purpose of catalytic applications, we
have successfully constructed an imine-based COF material,
COF-LZU1, from simple starting materials. The two-dimen-
sional layered-sheet structure together with the eclipsed imine
bonds render the COF-LZU1 material as an ideal scaffold for
coordinating metal ions. Accordingly, the robust incorporation of
Pd(OAc)2 into COF-LZU1 was realized with a simple post-
treatment and verified by spectroscopic analyses. The synthe-
sized Pd/COF-LZU1 material was further applied to catalyze the
SuzukiÀMiyaura coupling reaction, an important reaction for the
formation of CÀC bonds. The superior activity of Pd/COF-
LZU1 catalyst was testified by the broad scope of the reactants,
the excellent yields of the reaction products, and the high stability
and easy recyclability of the catalyst. In comparison with other
crystalline porous materials (zeolites and MOFs), the unique
structure of Pd/COF-LZU1 provides efficient access to the
catalytic sites and fast mass-transport of the reactants/products,
which are responsible for its superior activity in catalyzing the
SuzukiÀMiyaura coupling reaction.
applied reaction conditions and that the trace amount of Pd(OAc)2
leaching out could not catalyze the SuzukiÀMiyaura reaction.
The recycle use of Pd/COF-LZU1 catalyst (1.0 mol %) was
further examined in the reaction of p-nitrobromobenzene with
phenylboronic acid (Table 2). The results demonstrated that
Pd/COF-LZU1 catalyst could be reused at least for four times
without loss of catalytic activity and selectivity. The XPS spectra
of the recycled Pd/COF-LZU1 catalyst (Figure S25, SI) show
that the Pd species was still present in +2 state. The TEM images
of the recycled Pd/COF-LZU1 catalyst (Figures 3e and S21-
[SI]) indicate that the Pd species were still highly dispersed. The
comparison of the 13C CP/MAS NMR spectra (Figure 4aÀd) and
the PXRDpatterns(Figure5) ofthePd/COF-LZU1catalysts after
different reaction runs indicates that the crystallinity and the
atomic-level structure of Pd/COF-LZU1 was almost maintained,
although the intensity of the PXRD pattern was gradually de-
creased. The decrease of structural regularity should be due to the
repetitive exposure of Pd/COF-LZU1 catalyst to the relatively
harsh conditions of the SuzukiÀMiyaura reaction (temperature of
150 °C and sodium carbonate as the base).
We infer that the superior activity of Pd/COF-LZU1 catalyst
should be attributed to its unique structure. The eclipsed
layered-sheet arrangement of COF-LZU1 offers a robust scaffold
for Pd(OAc)2 incorporation. The distance (∼3.7 Å) of eclipsed
nitrogen atoms in adjacent layers falls into the ideal requirement
for strong coordination of Pd(OAc)2. Moreover, the regular
channels with a diameter of ∼1.8 nm provide efficient access to
the active sites together with fast diffusion for the bulky products.
In the case of MOF catalysts, the metal active sites, being also
responsible for structural preservation, are surrounded by bulky
ligands. Not only would the access to these active sites be limited,
but also the involvement of these active sites in catalysis could
lose the structural integrity, the case of which has been evidenced
In this contribution, not only is the first example of utilizing
COF material for catalysis demonstrated but also the toler-
ance of COF-LZU1 in relatively harsh reaction conditions is
verified. It is therefore believed that the COF-LZU1 scaffold
could act as a stable, robust, and versatile “supramolecular
ligand” and host a variety of metal ions for catalyzing a wide
scope of reactions, even at harsh reaction conditions. Also, we
expect that our approach will boost the research on employing
functional COF materials for catalysis, the further progress of
which may eventually bring COF materials into industrial
applications.
’ ASSOCIATED CONTENT
S
Supporting Information. Detailed synthetic procedures,
b
FT-IR spectra, PXRD patterns, modeling details and atomic co-
ordinates, gas adsorption, TGA traces, SEM images, TEM images,
19821
dx.doi.org/10.1021/ja206846p |J. Am. Chem. Soc. 2011, 133, 19816–19822

Journal of the American Chemical Society
ARTICLE
XPS spectra, and 13C CP/MAS NMR spectra. This material is
available free of charge via the Internet at http://pubs.acs.org.
(29) Lanni, L. M.; Tilford, R. W.; Bharathy, M.; Lavigne, J. J. J. Am.
Chem. Soc. 2011, 133, 13975.
(30) Uribe-Romo, F. J.; Doonan, C. J.; Furukawa, H.; Oisaki, K.;
Yaghi, O. M. J. Am. Chem. Soc. 2011, 133, 11478.
(31) Singh, G.; Singh, P. A.; Sen, A. K.; Singh, K.; Dubey, S. N.;
Handa, R. N.; Choi, J. H. Synth. React. Inorg. Met.-Org. Chem. 2002,
32, 171.
(32) Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457.
(33) Materials Studio Release Notes v.4.4 (Accelrys Software, San
Diego, 2008).
(34) As addressed in a recent work (ref 30), Rietveld refinement,
which can determine precisely the framework topology, is not applic-
able in the cases of COF materials due to the low resolution of the
PXRD patterns. Pawley refinement, which is being used for COF
systems, could not offer the unique assignment for the packing issue
based on the unit cell and the space group only. Accordingly, although
the simulated PXRD pattern via Pawley refinement does not match the
observed data, the staggered-packing structure cannot be absolutely
ruled out.
’ AUTHOR INFORMATION
Corresponding Author
wang_wei@lzu.edu.cn
’ ACKNOWLEDGMENT
This work was supported by the National Natural Science
Foundation of China (Nos. 20933009 and 20972064), the Key
Grant Project of Chinese Ministry of Education (No. 309028),
and the 111 Project. We are grateful to the anonymous reviewers
for their valuable comments and suggestions.
’ REFERENCES
(1) Yaghi, O. M.; O’Keeffe, M.; Ockwig, N. W.; Chae, H. K.;
Eddaoudi, M.; Kim, J. Nature 2003, 423, 705.
(2) C^otꢀe, A. P.; Benin, A. I.; Ockwig, N. W.; O’Keeffe, M.; Matzger,
A. J.; Yaghi, O. M. Science 2005, 310, 1166.
(35) Lukose, B.; Kuc, A.; Heine, T. Chem.—Eur. J. 2011, 17, 2388.
(36) Pandey, P.; Katsoulidis, A. P.; Eryazici, I.; Wu, Y. Y.; Kanatzidis,
M. G.; Nguyen, S. T. Chem. Mater. 2010, 22, 4974.
(37) Martell, A. E.; Perutka, J.; Kong, D. Coord. Chem. Rev. 2001,
216, 55.
(38) Palkovits, R.; Antonietti, M.; Kuhn, P.; Thomas, A.; Sch€uth, F.
Angew. Chem., Int. Ed. 2009, 48, 6909.
(3) El-Kaderi, H. M.; Hunt, J. R.; Mendoza-Cortes, J. L.; C^otꢀe, A. P.;
Taylor, R. E.; O’Keeffe, M.; Yaghi, O. M. Science 2007, 316, 268.
(4) C^otꢀe, A. P.; El-Kaderi, H. M.; Furukawa, H.; Hunt, J. R.; Yaghi,
O. M. J. Am. Chem. Soc. 2007, 129, 12914.
(39) Chan-Thaw, C. E.; Villa, A.; Katekomol, P.; Su, D.; Thomas, A.;
Prati, L. Nano Lett. 2010, 10, 537.
(5) Hunt, J. R.; Doonan, C. J.; LeVangie, J. D.; C^otꢀe, A. P.; Yaghi,
O. M. J. Am. Chem. Soc. 2008, 130, 11872.
(40) Chan-Thaw, C. E.; Villa, A.; Prati, L.; Thomas, A. Chem.—Eur.
J. 2011, 17, 1052.
(6) Tilford, R. W.; Gemmill, W. R.; zur Loye, H. C.; Lavigne, J. J.
Chem. Mater. 2006, 18, 5296.
(7) Tilford, R. W.; Mugavero, S. J.; Pellechia, P. J.; Lavigne, J. J. Adv.
Mater. 2008, 20, 2741.
(8) Cejka, J.; Corma, A.; Zones, S. Zeolites and Catalysis: Synthesis,
Reactions and Applications; Wiley-VCH: New York, 2010.
(9) Eddaoudi, M.; Moler, D. B.; Li, H. L.; Chen, B. L.; Reineke,
T. M.; O’Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2001, 34, 319.
(10) James, S. L. Chem. Soc. Rev. 2003, 32, 276.
(11) Thomas, A. Angew. Chem., Int. Ed. 2010, 49, 8328.
(12) Corma, A. Chem. Rev. 1997, 97, 2373.
(41) The PXRD patterns of Pd/COF-LZU1 and COF-LZU1
showed subtle differences only. Therefore, the incorporation of ∼0.5
Pd(OAc)2 per unit cell does not affect the layered structure of COF-
LZU1 but may slightly change the initial layer space (3.7 Å) of COF-
LZU1.
ꢁ
(42) As overwhelmingly demonstrated by the XPS and solid-
state NMR studies, the palladium species incorporated in Pd/COF-
LZU1 is Pd(OAc)2 with the Pd in +2 state. However, the black dots
observed in the TEM images of Pd/COF-LZU1 (d and e of Figure 3
and Figures S20 and S21(SI)) are attributed to the Pd(0) nanopar-
ticles formed upon the electron-beam induced reduction of the
original Pd(II) species during the TEM measurements. This phe-
nomenon has previously been noticed and explained by other
groups. For example, Wahl et al. observed the formation of palla-
dium nanoparticle arrays from K2PdCl4 which was initiated by
the electron beam in the TEM. Wahl, R.; Mertig, M.; Raff, J.;
Selenska-Pobell, S.; Pompe, W. Adv. Mater. 2001, 13, 736. Stark et al.
found the metalization of Pd(OAc)2 films induced by electron-beam.
Stark, T. J.; Mayer, T. M.; Griffis, D. P.; Russell, P. E. J. Vac. Sci. Technol.,
B 1991, 9, 3475. J. Vac. Sci. Technol., B 1992, 10, 2685. We also found
that the black dots were absent in the TEM images of Pd(OAc)2-free
Phen sample (Figure S22, SI), but showed up in those of Pd(OAc)2-
coordinated Pd/Phen sample (Figure S23, SI) and of pure Pd(OAc)2
sample (Figure S24, SI). These observations further verified the origin of
Pd(0) species.
(13) Corma, A.; Garcia, H.; Xamena, F. X. L. Chem. Rev. 2010,
110, 4606.
(14) Mastalerz, M. Angew. Chem., Int. Ed. 2008, 47, 445.
(15) Kuhn, P.; Antonietti, M.; Thomas, A. Angew. Chem., Int. Ed.
2008, 47, 3450.
(16) Uribe-Romo, F. J.; Hunt, J. R.; Furukawa, H.; Klock, C.;
O’Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2009, 131, 4570.
(17) Campbell, N. L.; Clowes, R.; Ritchie, L. K.; Cooper, A. I. Chem.
Mater. 2009, 21, 204.
(18) Dogru, M.; Sonnauer, A.; Gavryushin, A.; Knochel, P.; Thomas,
B. Chem. Commun. 2011, 47, 1707.
(19) Colson, J. W.; Woll, A. R.; Mukherjee, A.; Levendorf, M. P.;
Spitler, E. L.; Shields, V. B.; Spencer, M. G.; Park, J.; Dichtel, W. R.
Science 2011, 332, 228.
(20) Furukawa, H.; Yaghi, O. M. J. Am. Chem. Soc. 2009, 131,
8875.
(21) Doonan, C. J.; Tranchemontagne, D. J.; Glover, T. G.; Hunt,
J. R.; Yaghi, O. M. Nature Chem. 2010, 2, 235.
(22) Assfour, B.; Seifert, G. Microporous Mesoporous Mater. 2010,
133, 59.
(43) Thomas, A. C. Photoelectron and Auger Spectroscopy; Plenum:
New York, 1975.
(44) Magano, J.; Dunetz, J. R. Chem. Rev. 2011, 111, 2177.
(45) Arpad, M. Chem. Rev. 2011, 111, 2251.
(23) Wan, S.; Guo, J.; Kim, J.; Ihee, H.; Jiang, D. L. Angew. Chem., Int.
Ed. 2008, 47, 8826.
(46) Xamena, F.; Abad, A.; Corma, A.; Garcia, H. J. Catal. 2007,
250, 294.
(24) Wan, S.; Guo, J.; Kim, J.; Ihee, H.; Jiang, D. L. Angew. Chem., Int.
Ed. 2009, 48, 5439.
(25) Spitler, E. L.; Dichtel, W. R. Nature Chem. 2010, 2, 672.
(26) Spitler, E. L.; Giovino, M. R.; White, S. L.; Dichtel, W. R. Chem.
Sci. 2011, 2, 1588.
(27) Li, Y. W.; Yang, R. T. AlChE J. 2008, 54, 269.
(28) Jiang, J. X.; Cooper, A. I. Top. Curr. Chem. 2010, 293, 1.
19822
dx.doi.org/10.1021/ja206846p |J. Am. Chem. Soc. 2011, 133, 19816–19822