Bimetallic-organic framework as a zero-leaching catalyst in the aerobic oxidation of cyclohexene

Article abstract of DOI:10.1002/cctc.201300529

A gallium 2,2′-bipyridine-5,5′-dicarboxylate metal-organic framework (MOF), denoted as COMOC-4, has been synthesized by solvothermal synthesis. This MOF exhibits the same topology as MOF-253. CuCl₂ was incorporated into COMOC-4 by a post-synthetic modification (PSM). The spectroscopic absorption properties of the MOF framework before and after PSM were compared with theoretical data obtained by employing molecular dynamics combined with time-dependent DFT calculations on both the as-synthesized and functionalized linker. The catalytic behavior of the resulting Cu ²⁺@COMOC-4 material was evaluated in the aerobic oxidation of cyclohexene with isobutyraldehyde as a co-oxidant. In addition, the catalytic performance of Cu²⁺@COMOC-4 was compared with that of the commercially available Cu-BTC (BTC=benzene-1,3,5-tricarboxylate) MOF. Cu ²⁺@COMOC-4 exhibits a good cyclohexene conversion and an excellent selectivity towards cyclohexene oxide in comparison to the Cu-based reference catalyst. Furthermore, no leaching of the active Cu sites was observed during at least four consecutive runs. MOF with claws: COMOC-4, a gallium-based metal-organic framework is synthesized, which contains open bipyridine sites on which CuCl₂ is anchored. The catalytic performance of Cu ²⁺@COMOC-4 is evaluated for the aerobic oxidation of cyclohexene. The bimetallic-organic framework shows an excellent selectivity towards cyclohexene oxide and has good stability and reusability. Copyright

Full text of DOI:10.1002/cctc.201300529

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201300529
Bimetallic–Organic Framework as a Zero-Leaching Catalyst
in the Aerobic Oxidation of Cyclohexene
Ying-Ya Liu,[a] Karen Leus,[a] Thomas Bogaerts,[a, b] Karen Hemelsoet,[b] Els Bruneel,[c]
Veronique Van Speybroeck,[b] and Pascal Van Der Voort*[a]
A
gallium 2,2’-bipyridine-5,5’-dicarboxylate metal–organic
al was evaluated in the aerobic oxidation of cyclohexene with
isobutyraldehyde as a co-oxidant. In addition, the catalytic per-
formance of Cu2+@COMOC-4 was compared with that of the
commercially available Cu-BTC (BTC=benzene-1,3,5-tricarboxy-
late) MOF. Cu2+@COMOC-4 exhibits a good cyclohexene con-
version and an excellent selectivity towards cyclohexene oxide
in comparison to the Cu-based reference catalyst. Furthermore,
no leaching of the active Cu sites was observed during at least
four consecutive runs.
framework (MOF), denoted as COMOC-4, has been synthesized
by solvothermal synthesis. This MOF exhibits the same topolo-
gy as MOF-253. CuCl2 was incorporated into COMOC-4 by
a post-synthetic modification (PSM). The spectroscopic absorp-
tion properties of the MOF framework before and after PSM
were compared with theoretical data obtained by employing
molecular dynamics combined with time-dependent DFT calcu-
lations on both the as-synthesized and functionalized linker.
The catalytic behavior of the resulting Cu2+@COMOC-4 materi-
Introduction
Catalyzed liquid-phase oxidation reactions are widely em-
ployed in industrial processes and are becoming increasingly
important for the synthesis of fine chemicals.[1] Among the dif-
ferent oxidation reactions, the epoxidation of olefins plays
a prominent role as epoxides are highly reactive and versatile
intermediates. Although homogeneous catalysts are still often
utilized in industrial processes, there is an increasing interest in
the employment of heterogeneous catalysts as they have a big
advantage in terms of reuse and waste minimization. Metal–or-
ganic frameworks (MOFs) can be considered as potential candi-
dates for use in catalysis.[2] MOFs are 3D crystalline porous ma-
terials that consist of metal nodes connected by multifunction-
al organic linkers. Almost every transition metal ion and many
different organic linkers can be used to obtain a MOF struc-
ture, which makes the plausible metal–ligand combinations
endless.[3] However, only a minor amount of MOFs are practi-
cally usable in catalysis as many of them show limited stability
in typically employed catalytic reaction conditions.[4] By pre-/
post-synthetic modification of the organic linker, complemen-
tary catalytic active sites can be introduced.[5] Very recently we
reported on the post-synthetic modification of V-NH2-MIL-47
with TiO(acac)2 (acac=acetylacetonate). The resulting NH2-MIL-
47[Ti] material exhibited a significantly higher stability and ac-
tivity in the oxidation of cyclohexene compared to nonfunc-
tionalized NH2-MIL-47.[5c]
In this study we looked for another MOF support that is
more rigid in comparison to NH2-MIL-47. Within this context,
the bipyridine-based MOF-253, synthesized by Yaghi and co-
workers,[6] is an excellent candidate to serve as an MOF sup-
port. The beauty of this Al(OH)(bpydc) (bpydc2À =2,2’-bipyri-
dine-5,5’-dicarboxylate) framework lies in the fact that the or-
ganic linkers offer free 2,2’-bipyridine sites, which are often
employed as chelating ligands in coordination chemistry. More
specifically, it can form metal complexes in which the metal is
bound to the two N atoms to make it a bidentate-secured
stable complex. Several attempts have been made to graft sec-
ondary metal sites on the bipyridine site by a post-synthetic
modification approach. For example, Pd2+ and Cu2+ ions have
been incorporated into the MOF-253 framework and were sub-
sequently evaluated for their CO2 uptake.[6] Zou and co-workers
have incorporated RuCl3 into MOF-253, and the product was
examined as a catalyst for the selective oxidation of primary
and secondary alcohols.[5b] In a very recent report of Li et al.,
Cu+ ions were incorporated into MOF-253 to catalyze the
cross-coupling of phenols and alcohols with aryl halides.[7]
In this contribution, we report on the catalytic performance
of a member of the M(OH)(bpydc)series with M=Ga, denoted
as COMOC-4 (COMOC=Center for Ordered Materials, Organo-
[a] Dr. Y.-Y. Liu, Dr. K. Leus, T. Bogaerts, Prof. Dr. P. Van Der Voort
COMOC – Center for Ordered Materials, Organometallics and Catalysis
Department of Inorganic and Physical Chemistry
Ghent University
Krijgslaan 281-S3, 9000 Ghent (Belgium)
Fax: (+32)9-264-49-83
E-mail: pascal.vandervoort@ugent.be
[b] T. Bogaerts, Prof. Dr. K. Hemelsoet, Prof. Dr. V. Van Speybroeck
Center for Molecular Modeling
Ghent University
Technologiepark 903, 9052 Zwijnaarde (Belgium)
[c] Dr. E. Bruneel
SCRiPTs, Department of Inorganic and Physical Chemistry
Ghent University
Krijgslaan 281-S3, 9000 Ghent (Belgium)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201300529.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3657 – 3664 3657

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
metallics and Catalysis, Ghent University), which we published
recently.[8] The MOF structure features an analogous structure
to MOF-253 and is stable in air and water (508C for 24 h).
CuCl2, which shows a good binding affinity towards the bipyri-
dine sites, was grafted on the COMOC-4 framework by post-
synthetic modification (PSM). The spectroscopic properties of
the MOF before and after PSM were elucidated by using ab ini-
tio simulations. In particular, molecular dynamics (MD) compu-
tations were performed on a model linker with and without Cu
coordination to simulate the flexibility of the structures. Subse-
quently, time-dependent density functional theory (TD-DFT)
was applied on snapshots extracted from the MD runs to com-
pute an average UV/Vis spectrum. This methodology has been
shown to be successful previously.[9] The newly synthesized
Cu2+@COMOC-4 was extensively evaluated as a bimetallic cata-
lyst in the aerobic epoxidation of cyclohexene with an alde-
hyde as co-oxidant.[10] Additionally, regenerability and stability
tests were performed. Finally, the catalytic performance of
Cu2+@COMOC-4 was compared with another Cu-based refer-
ence MOF, namely, Cu-BTC (BTC=benzene-1,3,5-tricarboxylate).
into the COMOC-4 framework, the Bragg diffraction angles in
COMOC-4 and Cu2+@COMOC-4 are essentially identical, which
confirm that the COMOC-4 crystalline structure is preserved.
After Cu incorporation, the intensity of the reflections de-
creased. The main difference is seen in the diffraction peak at
6.58, which is related to the (101) planes that are parallel to
the linkers (both directions are equivalent because of symme-
try). The incorporation of CuCl2 will induce slight changes in
the shape and angle of the linkers. This results in several new
Bragg reflections close to the original ones that are then
merged together into one broader peak of lower intensity as
observed in the XRD pattern.
The UV/Vis absorption spectra of the synthesized materials
are presented in Figure 2 (top). The H2bpydc ligand displays
one absorption band centered at 299 nm, which arises from
a p–p* transition in the aromatic rings. If the carboxylate linker
is coordinated to Ga ions to form the COMOC-4 framework,
the absorption spectrum exhibits a redshift of ꢀ8 nm. In addi-
tion, the absorption band reveals a shoulder at 330 nm. For
comparison, Cu(Me2bpydc)Cl2 was synthesized. As observed
from the absorption spectra, the band that corresponds to the
organic ligand is redshifted to 317 nm, whereas a shoulder
peak appears at 334 nm and a broad band at 420 nm is ob-
served, which indicates the metal-to-ligand coordination. Cu2+
Results and Discussion
The powder XRD pattern of COMOC-4 (Figure S1, Supporting
Information) reveals that this framework is isostructural with
DUT-5[11] (Al(OH)(bpdc), bpdc2À =biphenyl-4,4’-dicarboxylate) as
well as with MOF-253.[6] The structure is indexed with ortho-
rhombic unit cell parameters of a=21.98(24), b=7.302(8), and
c=17.470(24) ꢁ. The COMOC-4 framework is constructed of in-
finite chains of octahedral GaO4(OH)2 units, in which each Ga3+
ion is bound to four bpydc2À ligands and two m2-trans hydrox-
ide anions (Figure 1). This is a common coordination motif that
has already been observed in a series of M3+ carboxylate
frameworks (M=Al, Fe, V, Ga, and In).[12] The GaO4(OH)2 chains
are aligned parallel to the crystallographic b axis, and the hy-
droxide and carboxylate moieties alternate on either side of
the chains, which are further linked to each other to form a 3D
open framework (Figure S2). After the incorporation of CuCl2
Figure 2. UV/Vis absorption spectra of suspended Cu2+@COMOC-4 and
COMOC-4 in MeOH solution compared to Cu(Me2bpydc)Cl2 and Me2bpydc
dissolved in MeOH (top); Solid-state UV/Vis spectra calculated from diffuse
reflection spectra (bottom).
Figure 1. Representative structure of Cu2+@COMOC-4. View along the 1D
pore system. The structure model was generated based on the crystal
structure of DUT-5 with 100% CuCl2 occupancy.[11]
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3657 – 3664 3658

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
@COMOC-4 exhibits two bands at 317 and 334 nm, which are
in agreement with the absorption bands of Cu(Me2bpydc)Cl2.
However, the absorption band that corresponds to the blue
color is not visible because of the rather high background
signal of the spectra recorded from the suspension. Diffuse re-
flection spectra (DRS) have also been measured from powder
samples to get better resolution in the visible region. COMOC-
4 has a broad, weak absorption band in the range of 430–
500 nm, whereas after incorporation of Cu2+ cations, a well-
distinguished absorption band centered at 460 nm was ob-
served (Figure 2 bottom).
The model used for the empty linker is shown in Figure 3.
This is the most stable conformation.[8] To allow CuCl2 to coor-
dinate to this ligand, both N atoms should be twisted to the
same side as shown in Figure 3. Both models were optimized
Figure 4. a) Comparison between the calculated UV/Vis spectra of the pro-
tected Me2bpydc linker and with Cu coordinated to the linker. The most im-
portant features of the experimental spectrum are indicated. b) Visualization
of the orbitals involved in the metal-to-ligand electron transfer that corre-
sponds to the visible excitation in the Cu-modified linker. The corresponding
participation of the linker (in dark grey) and CuCl2 (in light grey) fragments
are also given.
tal: 334 nm) is also present and is a consequence of Cu2+ en-
capsulation. The broad peak around 450 nm is also visible in
the ab initio spectrum, though this is shifted to a higher wave-
length as compared to the measured result. To confirm that
these changes are actually a result of metal coordination, one
can take a closer look at the molecular orbitals involved in
these excitations. The dominant orbitals involved in the excita-
tion around 450 nm are displayed in Figure 4b and full details
can be found in Figure S3. The orbitals and corresponding par-
ticipation of the different fragments (Figure 4b) indicate that
the coordinated metal contributes to the electronic transition.
We can thus conclude that the observed changes in the UV/Vis
spectrum are a result of the coordination of the CuCl2 complex.
The two bands at 338 and 450 nm can be employed to reveal
the successful incorporation of CuCl2 in the MOF framework,
which was clearly observed in the absorption spectra (334 nm)
and UV/Vis DRS spectra (460 nm) of Cu2+@COMOC-4 UV/Vis
(Figure 2 bottom).
Figure 3. Optimized structure of the protected linker Me2bpydc and with
coordinated CuCl2.
by using the B3LYP/6-311+g(d) methodology to find their
most stable conformations. UV/Vis spectra were calculated by
using the computational method described below, which ac-
counts for the flexibility of the structure. The main features
from the experimental spectra are represented in the calculat-
ed dynamic spectra of Cu2+ coordinated to the bipyridine
moiety compared with the as-synthesized linker. There is
a slight shift to higher wavelengths, from 309 to 315 nm, but
the shift is smaller in comparison to the experimental spec-
trum (299–317 nm; Figure 4a). The main absorption band for
the empty linker (309 nm) is consistent with that shown before
with static calculations.[8] The shoulder at 338 nm (experimen-
COMOC-4 maintains a permanent microporosity after the re-
moval of the guest molecules, as demonstrated by a type I N2
sorption isotherm (Figure 5), and exhibits a Langmuir surface
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3657 – 3664 3659

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
work decomposition. The residue of 26.1 wt% (calculated:
27.6%) is Ga2O3 and CuO. Calculated values are based on the
results of elemental analysis.
The aerobic epoxidation of cyclohexene in the presence of
an aldehyde as co-oxidant was used in this study. Such liquid-
phase oxidation of cycloalkenes (or cycloalkanes) has been in-
tensively studied by using MOF-based materials as catalysts.
With variable active sites on MOF catalysts, and the use of dif-
ferent oxidants, this reaction can lead to different product dis-
tributions.[13] Kholdeeva and co-workers[13f] reported Cr- and
Fe-MIL-101 materials as catalysts for the solvent-free selective
oxidation of cyclohexane with O2 and/or tert-butylhydroperox-
ide (TBHP) as oxidant. The substrate conversion was in the
range of 9–36% within 8 h, the major product formed can be
cyclohexyl hydroperoxide or cyclohexanone. This strongly de-
pends on the nature of the active metal. Kleist and co-worker-
s[13a] investigated the aerobic epoxidation of olefins ((E)-stilene
and styrene) in a basic solvent (dimethylformamide, DMF), cat-
alyzed by a Co-based MOF (STA-12(Co)). Different selectivities
were obtained that depended on the substrates. The selectivity
in styrene epoxidation was low because of substrate oligomeri-
zation. However, (E)-stilene was epoxidized with high selectivi-
ties between 80–90%. Garcia and co-workers[13b] have reported
the aerobic oxidation of cycloalkenes catalyzed by an Fe-based
MOF in the presence of N-hydroxyphthalimide, and the cycloal-
kenes were mainly converted to allylic oxidation products.
More recently, Xamena, Corma, and co-workers[13e] have report-
ed that MOFs with Cu2+ centers linked to four N atoms from
azaheterocyclic compounds are active catalysts for the aerobic
oxidation of activated alkanes. Furthermore, a tandem reaction
was designed that used a Cu-MOF combined with silylated Ti-
MCM-41 as a solid catalyst, in which the Cu-MOF first catalyzed
cumene oxidation to form cumene hydroperoxide as the major
product, and the intermediate hydroperoxide together with si-
lylated Ti-MCM-41 further catalyzed 1-octene to obtain
1-octene oxide. However, at high temperatures (908C) the
presence of Cu-MOF will catalyze the 1-octene at the allylic po-
sition. Therefore, to increase the selectivity to the epoxide
product, 1-octene and the Cu-MOF were kept in separate reac-
tors. In the present work, we further explore the catalytic activ-
ity of Cu-MOFs in the epoxidation of alkenes by the Mukaiya-
ma system.[14]
Figure 5. N2 adsorption (solid symbols) and desorption (open symbols) iso-
therms of COMOC-4 and Cu2+@COMOC-4 at 77 K.
area of 920 m2 gÀ1. After the incorporation of CuCl2, with a Cu/
Ga ratio of 0.4:1, 0.4Cu2+@COMOC-4 reveals a reduced Lang-
muir surface area of 630 m2 gÀ1. The thermal stability of
COMOC-4 and 0.4Cu2+@COMOC-4 has been examined by ther-
mogravimetric analysis (TGA; Figure 6). For COMOC-4, the first
Figure 6. TGA curves of COMOC-4 and Cu2+@COMOC-4 measured in an air
flow.
mass loss in the TGA profile corresponds to the elimination of
water from the pores (ꢀ6.5 wt%). COMOC-4 is thermally
stable up to 3008C, above which a further weight loss of
62 wt% before 5608C is indicative of the decomposition of the
framework. The final residue (observed: 29.6 wt%, calculated
26.6 wt%) is Ga2O3. The incorporation of Cu2+ decreases the
thermal stability of the MOF framework; a similar behavior was
observed in Ru@MOF-253.[5b] The TGA profile of Cu2+
@COMOC-4 depicts the as-synthesized sample, the TGA curve
shows a consistent weight loss of 11.5 wt% of solvent (water,
methanol) release in the initial stage, and the delayed release
(up to 2008C) can be assigned to weakly bound methanol mol-
ecules, which are presumably held by weak forces (hydrogen
bonds and van der Waals interactions) within the channels of
the main framework as well as water molecules that are more
firmly bound to the CuCl2. A second major weight loss
(62.5 wt%) occurs at 2308C, which is attributed to the frame-
All our catalytic tests on the Cu-based MOFs were performed
under identical reaction conditions to allow a fair comparison
(see Table 1). Cu-BTC, a copper trimesate Cu3(BTC)2(H2O)3,
known as HKUST-1, and commercially available as Basolite
C300, was applied as a reference catalyst. This material forms
face-centered cubic crystals that contain an intersecting 3D
system of large square-shaped pores (9ꢂ9 ꢁ).
The TON and TOF values of the two Cu-based MOFs are de-
picted as a function of the reaction time in Figure 7. Cu2+
@COMOC-4 and Cu-BTC show a good catalytic performance as
evidenced by the linear increase in the TON. Nevertheless,
Cu2+@COMOC-4 has a much higher TON value than Cu-BTC.
After 7 h of catalysis, the TON value of Cu-BTC is 75, whereas
Cu2+@COMOC-4 exhibits a TON value of almost 138. Moreover,
it can be seen that both catalysts reach a plateau in their TOF
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3657 – 3664 3660

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Table 1. Results of the epoxidation of cyclohexene catalyzed by Cu-
based compounds.
Catalyst
Conv. Sel. TOF Leaching Byproduct
[%]
[%] [hÀ1
]
[%]
selectivity [%]
3[c] 4[c] 5[c]
Cu2+@COMOC-4[a] 1st run 49.0 89.0 21.5
2nd run 46.2 87.0 19.7
0
0
0
0
4.3 2.6 3.9
5.3 3.1 4.5
5.6 3.1 3.4
5.1 2.7 2.9
6.9 8.5 7.0
3
4
rd run 41.5 87.8 18.9
th run 41.6 89.2 18.0
Cu-BTC
COMOC-4[b]
41.7 77.5 11.1 13.2
11.4 52.5 2.7
Scheme 1. Oxidation of cyclohexene (1) towards the main reaction products:
a) epoxidation to cyclohexene oxide (2), b) allylic oxidation to 2-cyclohex-
ene-1-one (3) and 2-cyclohexen-1-ol (4), c) consecutive ring opening to
cyclohexane-1,2-diol (5).
0
–
–
–
[a] Reaction conditions: Cu2+@COMOC-4 (0.19 g in the first run, 0.2 mmol
Cu sites), chloroform (40 mL), substrate (7 mL), isobutyraldehyde
(11.4 mL), O2 (7.7 mLminÀ1), T=313 K, t=7 h. [b] Based on Cu2+
@COMOC-4, an equal amount of Ga sites (0.65 mmol) was used, and the
TOF was calculated based on the number of moles of Cu sites. [c] Refer
to Scheme 1 for product distributions.
ing of 13.2% was detected after catalysis, which clearly dem-
onstrates that the catalytic activity of Cu-BTC is mainly a result
of homogeneous catalysis.
Additionally, the recyclability and stability of Cu2+@COMOC-
4 was evaluated. In total, four successive runs were performed
on Cu2+@COMOC-4. The results of the consecutive runs are
presented in Table 1 and Figure S6. Although there is a slight
reduction in the cyclohexene conversion during the additional
runs, the TOF value remains fairly constant. This observation
demonstrates the good recyclability of Cu2+@COMOC-4. The
slight decrease in the observed cyclohexene conversion is
probably because of a small loss of catalyst after each run.
Moreover, the selectivity towards the epoxide stays fairly con-
stant in the additional runs. Besides cyclohexane-1,2 diol,
which is the result of epoxide opening owing to the presence
of trace amounts of water adsorbed on the hydrophilic MOFs,
2-cyclohexen-1-ol and 2-cyclohexen-1-one are the observed
byproducts during each catalytic test. The formation of these
byproducts is a result of the allylic oxidation of cyclohex-
ene.[10c,15] No leaching of Cu and Ga species was observed
during these successive runs, which demonstrates the stability
and regenerability of the catalyst. The powder XRD patterns of
the Cu2+@COMOC-4 catalyst before and after each consecutive
run are presented in Figure 8. It can be seen that the structural
integrity of the framework is well preserved during these four
following runs.
Figure 7. Catalytic activity expressed as TON (top) and TOF (bottom) for
Cu2+@COMOC-4 (1) and Cu-BTC (2). Reaction conditions: 0.2 mmol Cu sites,
50 mmol cyclohexene, 100 mmol isobutyraldehyde, 40 mL chloroform,
T=408C, O2 flow: 7.7 mLminÀ1
.
The Cu-BTC catalyst, which also contains unsaturated Cu
sites, exhibits a similar product distribution and cyclohexene
conversion (41.7% for Cu-BTC and 49% for Cu2+@COMOC-4),
however, a difference in selectivity was detected between the
catalysts. In contrast to Cu2+@COMOC-4, which has an average
selectivity of 89% towards the epoxide, Cu-BTC exhibits
a lower selectivity of 77.5% towards cyclohexene oxide, which
results from the formation of a larger amount of byproducts
(2-cyclohexene-1-one (3), 2-cyclohexen-1-ol (4), and cyclohex-
ane-1,2-diol (5)). This could be because of the different struc-
ture of the Cu-MOF in comparison to Cu2+@COMOC-4. The Cu
paddlewheel units in Cu-BTC contain unsaturated Cu sites that
favor the binding of water. The presence of this adsorbed
water can play a prominent role in the ring opening of the ep-
oxide.[16] Moreover, the paddlewheel structure has been shown
to have a catalytic influence on epoxide ring-opening reac-
tions.[16b]
value after nearly 1 h of catalysis (Figure 7 bottom). The TOF
value of Cu-BTC is approximately 12 hÀ1, whereas for Cu2+
@COMOC-4, a significantly higher TOF value of nearly 22 hÀ1 is
noted, which demonstrates that Cu2+@COMOC-4 converts cy-
clohexene much faster in comparison to the reference catalyst.
As presented in Table 1, both Cu-based catalysts gave cyclo-
hexene oxide as the predominant product (Scheme 1, path-
way a). Cu2+@COMOC-4 shows a remarkable catalytic activity
to afford 49% of cyclohexene conversion after 7 h of catalysis
with a selectivity of 89% towards the epoxide in the first run
(Figure S4). No leaching of Cu and Ga sites was detected
during the first run, which indicates that the catalysis occurs
truly heterogeneously. In contrast to Cu2+@COMOC-4, for Cu-
BTC, although the XRD pattern indicates no obvious changes
in crystallinity after catalysis (Figure S5), a rather high Cu leach-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3657 – 3664 3661

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Scheme 3. The alkylperoxy radicals generated in the main catalytic cycle
lead to the formation of byproducts (pathway b) in addition to cyclohexene
oxide (pathway a). Adapted from Ref. [15a].
bilize the acylperoxy radical. Afterwards, the unstable carboxyl
radical decomposes to form an alkyl radical, which can in turn
be oxidized to an alkylperoxy radical. This species is less selec-
tive towards the formation of epoxides and leads to the forma-
tion of byproducts through the allylic oxidation pathway
(Scheme 3, pathway b).
Figure 8. XRD patterns of Cu2+@COMOC-4 before and after each catalytic
run. (The peak at 32.98 is because of the background of the silicon sample
holder)
Conclusions
A Ga-based metal–organic framework (MOF), denoted as
COMOC-4, has been successfully synthesized and character-
ized. In a second step CuCl2 was incorporated into COMOC-4
by a post-synthetic modification approach. The successful in-
corporation was verified with the aid of ab initio techniques.
The catalytic performance of the resulting Cu2+@COMOC-4
material was investigated for the aerobic oxidation of cyclo-
hexene in the presence of the co-oxidant isobutyraldehyde. In
comparison to Cu-BTC (BTC=benzene-1,3,5-tricarboxylate),
Cu2+@COMOC-4 shows the best catalytic performance in
terms of selectivity towards cyclohexene oxide. Furthermore,
no leaching of either Ga or Cu species was detected over four
successive runs, which indicates the good stability and reusa-
bility of the catalyst.
The reaction mechanism for the transition-metal-catalyzed
aerobic oxidation of alkenes in the presence of an aldehyde as
a co-reagent is widely known in the literature as the Mukaiya-
ma–Yamada epoxidation reaction (Scheme 2).[14] In this study,
Experimental Section
General
The 2,2’-bipyridine-5,5’-dicarboxylic acid (H2bpydc) ligand was pre-
pared according to a procedure published elsewhere.[18] Commer-
cially available spectroscopic grade methanol was applied for the
spectroscopic studies. All the other starting materials (analytical
grade) were bought and used without further purification. Cu-BTC
(BasoliteTM C300, Sigma–Aldrich) as well as the as-synthesized
porous compounds were activated at 1208C under vacuum for 3 h
prior to use.
Scheme 2. Main mechanism to catalyze cyclohexene epoxidation in the
presence of O2 and isobutyraldehyde.
the autoxidation of the aldehyde plays a critical role in the cat-
alytic process. Without the aldehyde, the Cu2+-catalyzed
alkene reaction is more in favor of the allylic reaction path-
way.[13e,17] The co-reactant isobutyraldehyde is transformed
in situ into an acylperoxy radical, which is the predominant ox-
idizing species and facilitates the oxygen transfer to the olefin.
Cyclohexene is a good substrate to investigate whether the ox-
idizing species prefers allylic oxidation or epoxidation.[15] The
acylperoxy radicals preferentially react with the double bond
of the alkenes to yield the epoxide (Scheme 3). The latter path-
way is the major reaction pathway observed during our cata-
lytic tests, which indicates that the Cu2+ active sites mainly sta-
Synthesis
Synthesis of the Cu complex: [Cu(Me2bpydc)]Cl2(H2O)1.5
The dimethyl-(2,2’-bipyridine)-5,5’-dicarboxylate (Me2bpydc) ligand
was synthesized according to the procedure described by Gunyar
et al.[19] In a second step, Me2bpydc (0.03 g) and CuCl2·2H2O
(0.019 g) were mixed in a Pyrex tube with methanol (5 mL). The
Pyrex tube was subsequently heated to 1208C and kept at this
temperature overnight. The blue powder was collected by filtra-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3657 – 3664 3662

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
tion, washed with acetone, and dried under vacuum. Elemental
analysis calcd (%) for CuCl2 (CH3)2(C12H6O4N2)·1.5H2O (433.73): C
41.49, H 2.70, N 6.24; found: C 41.24, H 2.90, N 6.77.
path length quartz cuvette to record the spectrum. The UV/Vis
DRS experiments were performed by using a Hitachi U-3000 UV/
VIS Spectrophotometer with a diffuse reflectance accessory (inte-
grated sphere) for spectrophotometric measurements in the range
of 350–800 nm. The spectra were converted by using the Kubelka–
Munk function. X-ray fluorescence spectrometry (XRF) measure-
ments were performed by using a Rigaku NexCG, Energy Disper-
sive X-ray Fluorescence (EDXRF) instrument.
Synthesis of COMOC-4 (Ga(OH)(bpydc))
The synthesis of COMOC-4 was optimized at the gram scale based
on our earlier reported synthesis procedure.[8] Ga(NO3)3·H2O (1.2 g,
4.4 mmol) and H2bpydc (1.2 g, 5 mmol) were added to DMF
(120 mL) in a 250 mL Schlenk flask equipped with a magnetic stir-
rer. In the first instance, the mixture was heated to 1108C and kept
at this temperature for 0.5 h. Afterwards the mixture was further
heated to 1508C and held at this temperature for 48 h with gentle
stirring. An orange powder was collected over a membrane filter
and washed thoroughly with DMF, methanol, and acetone. For the
removal of unreacted linker from the pores, the solid product was
suspended in DMF (0.5 g per 50 mL DMF), heated at 808C for 2 h,
collected by filtration, washed with DMF and acetone, and dried
under vacuum. To ensure the complete exclusion of the organic
species encapsulated within the pores of the open framework,
a Soxhlet extraction in methanol was performed over 48 h at
1208C. Afterwards the COMOC-4 material was dried under vacuum
overnight at RT. As a result of the presence of the ÀOH moiety on
the Ga building unit, the 1D channels are highly hydrophilic. For
this reason, the activated sample was stored under an inert atmo-
sphere. The yield was 33% based on the Ga source. IR (KBr pellet):
n˜ =3379 (br), 1619 (s), 1595 (s), 1421 (s), 1394 (s), 1158 (w), 1050
(w), 847 (w), 775(m), 705 (w), 600 (w), 479 cmÀ1 (w); elemental
analysis calcd (%) for Ga(OH)(C12H6N2O4)·2.7H2O (377.56): C 38.36, H
3.27, N 7.46; found: C 38.26, H 3.12, N 7.17.
Computational methodology
MD simulations were performed on the linker with and without Cu
coordination in a vacuum box of 20ꢂ20ꢂ20 ꢁ by using the CP2K
package.[20] All DFT calculations were performed by using the Gaus-
sian plane waves (GPW) method,[21] with a DZVP basis set, GTH
pseudopotentials,[22] and the BLYP functional. MD runs were con-
ducted by using the canonical (NVT) ensemble at 300 K with
a time step of 1 fs. A chain of five Nosꢃ–Hoover thermostats was
used to control the temperature. The system was first allowed to
equilibrate after which a simulation of 20 ps was used for analysis.
The dynamic UV/Vis spectra were obtained by taking 100 snap-
shots from the simulation on which vertical TD-DFT calculations
were performed. This methodology has previously proven valuable
for the simulation of absorption spectra.[9,23] An average optical
spectrum was then obtained. The influence of the methanol sol-
vent was included with a polarizable continuum model (PCM). All
TD-DFT calculations were performed by using the Gaussian 09[24]
program using the B3LYP[25] functional and a 6-311+G(g) Pople
basis set. We previously demonstrated that the B3LYP functional is
very efficient for these types of systems.[8] The effect of relativistic
contributions was found to be small, and full details are given in
the Supporting Information.
Grafting the CuCl2 complex onto the COMOC-4 framework
Cu2+@COMOC-4 was prepared by stirring Ga(OH)(bpydc) (0.5 g)
and CuCl2·2H2O (0.1 g) in absolute methanol (30 mL) at 508C for
6 h. The green MOF powder was collected by filtration and was
stirred in pure methanol (20 mL) for 6 h followed by filtration. This
procedure was repeated twice to guarantee the complete removal
of physisorbed CuCl2 salts.
Catalysis
The oxidation of cyclohexene was performed in a 100 mL glass re-
actor equipped with a reflux condenser with recirculating cooling
at À48C. In a typical catalytic test, the reactor was loaded with
Cu2+@COMOC-4 (0.19 g, 0.2 mmol Cu active sites), cyclohexene
(7 mL, 5 mmol), isobutyraldehyde (11.4 mL), chloroform (40 mL),
and 1,2,4-trichlorobenzene (9 mL) employed as an internal stan-
dard. The molar ratio of cyclohexene/co-oxidant (isobutyraldehyde)
was 1:2. The O2 flow rate was set to 7.7 mLminÀ1 by using a mass
flow controller. All the catalytic tests were performed at 408C.
Blank reactions at this temperature showed no formation of oxida-
tion products. During the catalytic tests, aliquots were gradually
taken out of the mixture, diluted with ethyl acetate (500 mL), and
subsequently analyzed by GC with flame ionization detection (FID).
The reaction products were identified by using a TRACE GCꢂGC
(Thermo, Interscience) coupled to a TEMPUS TOF-MS detector
(Thermo, Interscience). The first column consists of a dimethyl
polysiloxane packing and has a length of 50 m with an internal di-
ameter of 0.25 mm, and the second column has a length of 2 m
with an internal diameter of 0.15 mm. The packing of the latter is
a 50% phenyl polysilphenylene siloxane. He was used as the carrier
gas with a constant flow (1.8 mLminÀ1).
Characterization
Powder XRD and TGA
Powder XRD patterns were recorded by using a Thermo Scientific
ARL X’Tra diffractometer, operated at 40 kV, 40 mA using CuKa radi-
ation (l=1.5406 ꢁ). TGA data were obtained by using a Netzsch
STA 449 F3 Jupiter-Simultaneous TG-DSC analyzer with a heating
rate of 108CminÀ1 in air. N2 sorption measurements were per-
formed by using a Belsorp II, Bell Japan, Inc. All the samples were
activated under vacuum at 1208C for 3 h prior to analysis.
Spectroscopic characterization
FTIR spectra were recorded in the region of 400–4000 cmÀ1 by
using a Bruker EQUINOX 55 FTIR spectrometer. UV/Vis absorption
spectra were collected by using a Perkin–Elmer Lambda 950 UV/Vis
spectrometer in the range of 260–900 nm. The spectra were re-
corded by using fine suspensions of powder samples (COMOC-4
and Cu2+@COMOC-4) in methanol. In a typical measurement, 4 mg
of powder sample was suspended in 3 mL of methanol in an ultra-
sonic bath for 5 min. The suspension was transferred to a 10 mm
All the fresh catalysts were activated under vacuum at 1208C for
3 h prior to catalysis. After each catalytic run, the catalyst was re-
covered by filtration, washed with acetone, and dried at RT over-
night under vacuum to reuse it in another run. To investigate the
recyclability of the Cu2+@COMOC-4 catalyst, four consecutive runs
were performed. Moreover, to examine the heterogeneity of the
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3657 – 3664 3663

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
A. Bhaumik, Dalton Trans. 2011, 40, 12510–12518; c) N. Komiya, T.
Naota, Y. Oda, S.-I. Murahashi, J. Mol. Catal. A 1997, 117, 21–37.
[11] I. Senkovska, F. Hoffmann, M. Froba, J. Getzschmann, W. Bohlmann, S.
Kaskel, Microporous Mesoporous Mater. 2009, 122, 93–98.
catalyst, the filtrate, after the removal of the catalyst, was analyzed
by XRF to quantify the possibly leached Cu and Ga species.
[12] a) T. Loiseau, C. Serre, C. Huguenard, G. Fink, F. Taulelle, M. Henry, T. Ba-
taille, G. Ferey, Chem. Eur. J. 2004, 10, 1373–1382; b) A. J. Jacobson, M.
Vougo-Zanda, J. Huang, E. Anokhina, X. Q. Wang, Inorg. Chem. 2008, 47,
11535–11542.
[13] a) M. J. Beier, W. Kleist, M. T. Wharmby, R. Kissner, B. Kimmerle, P. A.
Wright, J.-D. Grunwaldt, A. Baiker, Chem. Eur. J. 2012, 18, 887–898; b) A.
Dhakshinamoorthy, M. Alvaro, H. Garcia, J. Catal. 2012, 289, 259–265;
c) N. V. Maksimchuk, K. A. Kovalenko, V. P. Fedin, O. A. Kholdeeva, Chem.
Commun. 2012, 48, 6812–6814; d) A. M. Balu, C. S. K. Lin, H. Liu, Y. Li, C.
Vargas, R. Luque, Appl. Catal. A 2013, 455, 261–266; e) I. Luz, A. Leon,
M. Boronat, F. X. Llabres i Xamena, A. Corma, Catal. Sci. Technol. 2013, 3,
371–379; f) I. Y. Skobelev, A. B. Sorokin, K. A. Kovalenko, V. P. Fedin, O. A.
Kholdeeva, J. Catal. 2013, 298, 61–69.
Acknowledgements
The authors acknowledge the financial support from the Ghent
University BOF Grant (Nr. 01P02911T), GOA Grant (01G00710)
and the Long Term Structural Methusalem grant nr. 01M00409
Funding by the Flemish Government. Furthermore, this research
is co-funded by BELSPO in the frame of IAP 7/05 and the Europe-
an Research Council (FP7(2007-2013) ERC Grant (Nr. 240483).
Computational resources and services were provided by Ghent
University.
[14] T. Takai, T. Yamada, T. Mukaiyama, Chem. Lett. 1991, 9, 1499–1502.
[15] a) B. B. Wentzel, P. L. Alsters, M. C. Feiters, R. J. M. Nolte, J. Org. Chem.
2004, 69, 3453–3464; b) W. W. Nam, H. J. Kim, S. H. Kim, R. Y. N. Ho, J. S.
Valentine, Inorg. Chem. 1996, 35, 1045–1049.
Keywords: ab initio calculations · copper · epoxidation ·
heterogeneous catalysis · metal–organic framework
[16] a) H. L. Bao, J. Zhou, Z. Wang, Y. L. Guo, T. P. You, K. L. Ding, J. Am. Chem.
Soc. 2008, 130, 10116–10127; b) K. Tanaka, K. Otani, T. Murase, S. Nishi-
hote, Z. Urbanczyk-Lipkowska, Bull. Chem. Soc. Jpn. 2012, 85, 709–714;
c) I. Vilotijevic, T. F. Jamison, Science 2007, 317, 1189–1192.
[17] Y. Fu, D. Sun, M. Qin, R. Huang, Z. Li, RSC Adv. 2012, 2, 3309–3314.
[18] K. C. Szeto, K. O. Kongshaug, S. Jakobsen, M. Tilset, K. P. Lillerud, Dalton
Trans. 2008, 2054–2060.
[19] A. Gunyar, D. Betz, M. Drees, E. Herdtweck, F. E. Kuhn, J. Mol. Catal. A
2010, 331, 117–124.
[20] http://cp2k.berlios.de/.
[1] a) A. Corma, J. Catal. 2003, 216, 298–312; b) J. H. Clark, D. J. Macquarrie,
Org. Process Res. Dev. 1997, 1, 149–162.
[2] a) J. R. Long, O. M. Yaghi, Chem. Soc. Rev. 2009, 38, 1213–1214; b) K. K.
Tanabe, S. M. Cohen, Chem. Soc. Rev. 2011, 40, 498–519; c) J. L. C. Row-
sell, O. M. Yaghi, Microporous Mesoporous Mater. 2004, 73, 3–14.
[3] a) M. Eddaoudi, D. B. Moler, H. L. Li, B. L. Chen, T. M. Reineke, M.
O’Keeffe, O. M. Yaghi, Acc. Chem. Res. 2001, 34, 319–330; b) O. M. Yaghi,
H. L. Li, C. Davis, D. Richardson, T. L. Groy, Acc. Chem. Res. 1998, 31,
474–484.
[21] G. Lippert, J. Hutter, M. Parrinello, Mol. Phys. 1997, 92, 477–487.
[22] a) S. Goedecker, M. Teter, J. Hutter, Phys. Rev. B 1996, 54, 1703–1710;
b) C. Hartwigsen, S. Goedecker, J. Hutter, Phys. Rev. B 1998, 58, 3641–
3662.
[4] a) A. Dhakshinamoorthy, M. Alvaro, H. Garcia, Catal. Sci. Technol. 2011, 1,
856–867; b) V. Van Speybroeck, K. Leus, I. Muylaert, M. Vandichel, G. B.
Marin, M. Waroquier, P. Van der Voort, Chem. Commun. 2010, 46, 5085–
5087; c) A. Corma, H. Garcia, F. X. Llabres i Xamena, Chem. Rev. 2010,
110, 4606–4655.
[5] a) L. Ma, J. M. Falkowski, C. Abney, W. Lin, Nat. Chem. 2010, 2, 838–846;
b) F. Carson, S. Agrawal, M. Gustafsson, A. Bartoszewicz, F. Moraga, X.
Zou, B. Martin-Matute, Chem. Eur. J. 2012, 18, 15337–15344; c) K. Leus,
G. Vanhaelewyn, T. Bogaerts, Y.-Y. Liu, D. Esquivel, F. Callens, G. B. Marin,
V. Van Speybroeck, H. Vrielinck, P. Van Der Voort, Catal. Today, 2013, 208,
97–105.
[6] E. D. Bloch, D. Britt, C. Lee, C. J. Doonan, F. J. Uribe-Romo, H. Furukawa,
J. R. Long, O. M. Yaghi, J. Am. Chem. Soc. 2010, 132, 14382–14384.
[7] M. Wang, B. Yuan, T. Ma, H. Jiang, Y. Li, RSC Adv. 2012, 2, 5528–5530.
[8] Y.-Y. Liu, R. Decadt, T. Bogaerts, K. Hemelsoet, A. M. Kaczmarek, D. Poel-
man, M. Waroquier, V. Van Speybroeck, R. Van Deun, P. Van Der Voort, J.
Phys. Chem. C 2013, 117, 11302–11310.
[9] a) T. De Meyer, K. Hemelsoet, L. Van der Schueren, E. Pauwels, K. De
Clerck, V. Van Speybroeck, Chem. Eur. J. 2012, 18, 8120–8129; b) V.
Barone, J. Bloino, S. Monti, A. Pedone, G. Prampolini, Phys. Chem. Chem.
Phys. 2010, 12, 10550–10561; c) K. Hemelsoet, Q. Qian, T. De Meyer, K.
De Wispelaere, B. De Sterck, B. M. Weckhuysen, M. Waroquier, V. Van
Speybroeck, Chem. Eur. J. 2013, doi: 20.1002/chem.201301965.
[10] a) B. Monteiro, S. Gago, S. S. Balula, A. A. Valente, I. S. GonÅalves, M. Pil-
linger, J. Mol. Catal. A 2009, 312, 23–30; b) M. Nandi, P. Roy, H. Uyama,
[23] C. Adamo, D. Jacquemin, Chem. Soc. Rev. 2013, 42, 845–856.
[24] Gaussian 09, Revision B.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢄ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[25] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652; b) C. T. Lee, W. T.
Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
Received: July 3, 2013
Revised: September 10, 2013
Published online on October 17, 2013
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3657 – 3664 3664