Ti-functionalized NH2-MIL-47: An effective and stable epoxidation catalyst

Article abstract of DOI:10.1016/j.cattod.2012.09.037

In this paper, we describe the post-functionalization of a V-containing Metal-organic framework with TiO(acac)₂ to create a bimetallic oxidation catalyst. The catalytic performance of this V/Ti-MOF was examined for the oxidation of cyclohexene using molecular oxygen as oxidant in combination with cyclohexanecarboxaldehyde as co-oxidant. A significantly higher cyclohexene conversion was observed for the bimetallic catalyst compared to the non-functionalized material. Moreover, the catalyst could be recycled at least 3 times without loss of activity and stability. No detectable leaching of V or Ti was noted. Electron paramagnetic resonance measurements were performed to monitor the fraction of V-ions in the catalyst in the +IV valence state. A reduction of this fraction by ～17% after oxidation catalysis is observed, in agreement with the generally accepted mechanism for this type of reaction.

Full text of DOI:10.1016/j.cattod.2012.09.037

GModel
CATTOD-8284; No. of Pages9
ARTICLE IN PRESS
Catalysis Today xxx (2012) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Ti-functionalized NH2-MIL-47: An effective and stable epoxidation catalyst
Karen Leusa, Gauthier Vanhaelewynb, Thomas Bogaertsa,c, Ying-Ya Liua, Dolores Esquivela,
Freddy Callensb, Guy B. Marind, Veronique Van Speybroeckc, Henk Vrielinckb,∗
Pascal Van Der Voorta,∗∗
,
a Department of Inorganic and Physical Chemistry, Center for Ordered Materials, Organometallics and Catalysis (COMOC), Ghent University, Krijgslaan 281-S3, 9000 Ghent, Belgium
b Department of Solid State Sciences, Ghent University, Electron Magnetic Resonance (EMR) Research Group, Krijgslaan 281-S1, 9000 Ghent, Belgium
c Center for Molecular Modeling, Ghent University, Technologiepark 903, 9052 Zwijnaarde, Belgium
d Laboratory for Chemical Technology, Ghent University, Krijgslaan 281-S5, 9000 Ghent, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
In this paper, we describe the post-functionalization of a V-containing Metal-organic framework with
TiO(acac)2 to create a bimetallic oxidation catalyst. The catalytic performance of this V/Ti-MOF was
examined for the oxidation of cyclohexene using molecular oxygen as oxidant in combination with cyclo-
hexanecarboxaldehyde as co-oxidant. A significantly higher cyclohexene conversion was observed for the
bimetallic catalyst compared to the non-functionalized material. Moreover, the catalyst could be recycled
at least 3 times without loss of activity and stability. No detectable leaching of V or Ti was noted. Electron
paramagnetic resonance measurements were performed to monitor the fraction of V-ions in the catalyst
in the +IV valence state. A reduction of this fraction by ∼17% after oxidation catalysis is observed, in
agreement with the generally accepted mechanism for this type of reaction.
Received 13 July 2012
Received in revised form
14 September 2012
Accepted 18 September 2012
Available online xxx
Keywords:
Metal-organic frameworks
Vanadium
© 2012 Elsevier B.V. All rights reserved.
Titanium
Oxidation catalysis
1. Introduction
the catalytic activity in comparison with the parent framework.
Alternatively, the catalytic activity can be influenced by the intro-
Metal-organic frameworks (MOFs) are a highly versatile class
of ordered porous materials constructed of metal nodes or metal
clusters linked together by organic ligands. This recent class of
crystalline materials has been extensively studied for their poten-
tial use in various applications, e.g. in gas storage and separations
[1–3], catalysis [4–8] and luminescence [9,10]. MOFs possess many
intriguing features including very high surface areas, well defined
pore properties (pore size, shape and volume) and have easily
tailorable structures and chemical functionalities. Several reports
have appeared on the introduction of an additional functionality in
the linker, by pre-functionalization with, e.g., NH2, OH, CH3 and
NO2 groups [11–15]. In the field of catalysis, it has been shown
that the introduction of these additional functional groups can gen-
erate a significant alteration in the catalytic performance compared
to the parent framework. In a recent report of Vermoortele et al., a
UiO-66-X (X = H, NH2, CH3, OCH3, F, Cl, Br, NO2)series with different
functional groups was tested for the citronellal cyclization [16]. This
study demonstrated that the presence of NO2 groups enhanced
duction of additional framework active sites [17]. Maksimchuk et al.
reported the encapsulation of polyoxometalates into the nanocages
of the MIL-101 framework [18]. The resulting material demon-
strated a good catalytic activity for the epoxidation of various
alkenes. Alkordi et al. encapsulated metalloporphyrins into zeolite-
like MOFs which enhanced the catalytic activity for cyclohexane
oxidation [19]. A third approach is to bring new catalytically active
sites into the framework by a post-modification of the embedded
functional groups [20]. Ingleson et al. reported the two steps mod-
ification of IRMOF-3 to an imine grafted VO(acac)2 complex [21].
However, the resulting catalyst showed a rather low catalytic activ-
ity and stability for the oxidation of cyclohexene. The group of
Battacharjee reported a more straightforward one step function-
alization of the IRMOF-3 with Mn(acac)2 [22]. Although a good
catalytic and stability performance was obtained, the yield of the
grafted complex was rather low (8%).
In our previous studies, we demonstrated that vanadium
based MOFs (MIL-47 and COMOC-3), with saturated vanadium
sites have a remarkable catalytic activity in the liquid phase
epoxidation of cyclohexene with tert-Butyl hydroperoxide (TBHP)
as oxidant [23–25]. However, both the experimental results and
theoretical calculations showed that the catalytic cycle starts with
the breaking of at least one V-carboxylate bond to coordinate
the peroxide. The structural defects created during this process
∗
Corresponding author. Tel.: +32 9 264 43 56; fax: +32 9 264 49 96.
Corresponding author. Tel.: +32 9 264 44 42; fax: +32 9 264 49 83.
E-mail addresses: Henk.Vrielinck@ugent.be (H. Vrielinck),
∗∗
Pascal.Vandervoort@ugent.be (P. Van Der Voort).
0920-5861/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cattod.2012.09.037
Please cite this article in press as: K. Leus, et al., Ti-functionalized NH2-MIL-47: An effective and stable epoxidation catalyst, Catal. Today (2012),
http://dx.doi.org/10.1016/j.cattod.2012.09.037

GModel
CATTOD-8284; No. of Pages9
ARTICLE IN PRESS
2
K. Leus et al. / Catalysis Today xxx (2012) xxx–xxx
contribute to the catalytic performance. Hence, a small amount of
leaching was observed in the start of the catalysis. Our objective is
to prevent leaching while maintaining or enhancing the catalytic
performance. Therefore, we opted for a milder system using O2
as oxidant in combination with cyclohexanecarboxaldehyde as a
co-oxidant. To the best of our knowledge, a Ti grafted V-MOF has
not been reported so far. Nevertheless, bimetallic Ti–V catalysts
are reported to have a synergistic effect in oxidation catalysis
[26].
Helium was used as carrier gas, and the flow rate was programmed
as 0.8 mL/min.
2.2. Synthesis of NH2-MIL-47 [Ti]
In a first step NH2-MIL-47 was prepared via a microwave syn-
thesis route. Typically 1.7 mmol VCl3 was mixed with 1.7 mmol
2-aminoterephthalic acid in 3.02 mL·H2O. The reaction was carried
out at 150 ◦C for 20 min. Afterwards, an extraction in dimethyl-
formamide (DMF) was performed at 125 ◦C for 90 min to remove
unreacted linker.
In a second step TiO(acac)2 was grafted onto the V-MOF. Differ-
ent temperatures and reaction times have been applied as shown
in Table S. 1 of the Supporting Information. In each experiment
0.0785 g TiO(acac)2 was dissolved in 30 mL of dry toluene. After
stirring the solution for 4 h at 55 ◦C, 0.15 g NH2-MIL-47 was added
to the solution. Subsequently this mixture was stirred at a certain
temperature (55 ◦C, 70 ◦C and 90 ◦C) for 20 or 40 h under an inert
argon atmosphere. Hereafter, the MOF was filtered on a combined
nylon-membrane filter and washed several times with acetone
to remove unreacted TiO(acac)2. Afterwards the solid was dried
overnight under vacuum.
In this contribution, we report a bimetallic V/Ti-MOF via
post-functionalization of the V-MOF, NH2-MIL-47 with a Titany-
lacetylacetonate complex (denoted as NH2-MIL-47 [Ti] hereafter)
(see Fig. 1). The obtained NH2-MIL-47 [Ti] was examined in
the oxidation of cyclohexene and was compared with the non-
functionalized V-MOF as well as with the homogeneous TiO(acac)2
catalyst. Additionally regeneration and stability tests were carried
out. In our previous study of the catalytic activity of MIL-47 in the
oxidation of cyclohexene with TBHP as oxidant, we used X-band
(9.8 GHz) Electron Paramagnetic Resonance (EPR) measurements
at room temperature (RT) to monitor the evolution in the fraction
of V-ions in the paramagnetic +IV valence state [24]. We observed
a reduction of the V+IV concentration by about 20% in the first
few hours of reaction, as a result of V+IV → V+V oxidation, con-
firming a radical parallel reaction pathway suggested by density
functional theory (DFT) calculations. Similar EPR experiments are
performed in this study in order to gain insight in the catalytic reac-
tion mechanisms. In view of the absence of hyperfine (HF) structure
in the spectra, the X-band measurements are complemented with
Q-band experiments (34 GHz) and spectral simulations (Easyspin
[27] library in Matlab®) in order to corroborate the assignment of
2.3. Catalytic setup
In a typical catalytic test, a schlenk flask was loaded with 40.0 mL
of acetonitrile as solvent, 7.0 mL cyclohexene (substrate) and 8.4 mL
1,2,4-trichlorobenzene used as internal standard. Oxygen (99.9%
pure) was applied as the oxidant in combination with the co-
oxidant cyclohexanecarboxaldehyde (7.6 mL). A constant oxygen
flow of 7 mL/min was bubbled through the solution by means of
a mass flow controller. Blanc reactions were carried out in the
absence of catalyst which showed no conversion of cyclohexene.
All the catalytic tests were carried out at a temperature of 40 ◦C
in a schlenk flask equipped with a liquid condenser coupled with
recirculating cooling liquid at −4 ◦C to prevent evaporation of the
reaction mixture. Aliquots were gradually taken out of the mix-
ture, diluted with 500 ␮l ethylacetate, and subsequently analyzed
by GC–FID.
the spectrum to V+IV
.
2. Materials and methods
2.1. General procedures
All chemicals were bought from Sigma–Aldrich and used with-
out further purification. X-Ray Fluoresence (XRF) measurements
were performed on a NEX CG from Rigaku using a Mo X-ray source.
X-ray powder diffraction (XRPD) patterns were collected on a ARL
X’TRA X-ray diffractometer with Cu-K␣ radiation of 0.15418 nm
wavelength and a solid state detector. The Cross Polarization Magic
Angle Spinning Nuclear Magnetic Resonance (13C CP/MAS NMR)
spectra were recorded at 100.6 MHz on a Bruker AVANCE-400 WB
spectrometer at RT. The samples were spun at 13 kHz. An overall
10,000 free induction decays were accumulated with 4 s of recycle
time. Chemical shifts were measured relative to a tetramethylsilane
standard. The 1H NMR spectrum was recorded on a Bruker AVANCE
DRX500 NMR spectrometer (500 MHz). Diffuse Reflectance Infrared
Fourier Transform Spectroscopy (DRIFTS) measurements were
recorded on a Thermo Nicolet 6700 FT-IR spectrometer equipped
with a N2 cooled MCT-A (mercury–cadmium–tellurium) detector.
Nitrogen adsorption experiments were measured at −196 ◦C using
a Belsorp mini II gas analyzer. The reaction products were iden-
tified with a TRACE GC × GC (Thermo, Interscience), coupled to a
TEMPUS TOF-MS detector (Thermo, Interscience). The first column
consists of a dimethyl polysiloxane package and has a length of
50 m, with an internal diameter of 0.25 mm, whereas the second
column has a length of 2 m with an internal diameter of 0.15 mm.
The package of the latter is a 50% phenyl polysilphenylene-siloxane.
Helium was used as carrier gas with a constant flow (1.8 mL/min).
An ultra-fast gas chromatograph (GC) equipped with a flame ioniza-
tion detector (FID) and a 5% diphenyl/95% poly-dimethylsiloxane
column with 10 m length and 0.10 mm internal diameter was used
to follow the conversions of the products during the catalytic tests.
2.4. EPR measurements
For quantitative EPR analysis, spectra of dry powders, contained
in quartz tubes with an outer diameter of 4 mm, filled to a height
of approximately 6 mm (30–35 mg), were recorded at RT using a
Bruker ESP300E X-band spectrometer (ER 4102-ST rectangular cav-
ity). In the bottom part of the EPR cavity an open quartz tube was
fixed, on which the sample tubes rested, in order to ensure repro-
ducible positioning. The microwave frequency was measured using
a HP 5350 B microwave frequency counter. The magnetic field was
modulated at 100 kHz with a peak-to-peak amplitude of 0.1 mT and
the microwave power was set to 2 mW, avoiding saturation. A broad
magnetic field range was swept (350 mT) around the free elec-
tron g-value (2.0023). All spectra were normalized to a microwave
frequency of 9.77 GHz for comparison and corrected for the back-
ground spectrum of the empty cavity. For comparison of intensities,
the spectra were divided by the sample mass. In order to resolve
the anisotropy, selected powders were also measured in Q-band
(≈34 GHz) at RT on a Bruker Elexsys E500 spectrometer equipped
with a Pendulum CNT-90XL frequency counter (spectra normalized
to 33.98 GHz). These measurements were performed on powders
contained in 2 mm outer diameter quartz tubes filled to a height of
about 2 mm.
Please cite this article in press as: K. Leus, et al., Ti-functionalized NH2-MIL-47: An effective and stable epoxidation catalyst, Catal. Today (2012),
http://dx.doi.org/10.1016/j.cattod.2012.09.037

GModel
CATTOD-8284; No. of Pages9
ARTICLE IN PRESS
K. Leus et al. / Catalysis Today xxx (2012) xxx–xxx
3
Fig. 1. Schematic representation of the post-functionalization of NH2-MIL-47 with a Titanylacetylacetonatecomplex (the carbon atoms are shown in gray, the oxygen atoms
are depicted in red and the N, Ti and V atoms are shown in blue, black and yellow, respectively). (For interpretation of the references to color in the artwork, the reader is
referred to the web version of the article.)
2.5. Computational details
1.25 mmol Ti/g, which indicates that almost 30% of the –NH2
groups on the organic linker are modified. It should be noted
that this loading is much higher compared to the earlier reported
one step functionalization of Mn(acac)2 on IRMOF-3, which was
approximately 0.23 mmol Mn/g [22]. Furthermore, the crystallinity
of all the obtained Ti-grafted materials was verified by XRPD mea-
surements (see Fig. S.1 Supporting Information). As can be seen
from this figure, the XRPD pattern of each functionalized material
presents the pure phase of the non-functionalized NH2-MIL-47.
This explicitly shows that the framework integrity of the parent
MOF was well preserved during the post-functionalization.
Ab initio calculations are performed with the gaussian09 pro-
gram [28]. All models were optimized with a B3LYP [29,30]
functional using a 6–31 + G(d) Pople basis set. Vibrational frequency
calculations were done at the same level of theory. To compare the
calculated frequencies with the experimental DRIFTS spectrum a
scale factor of 0.9648 was imposed, as proposed by Merrick et al.
[31].
3. Experimental results and discussion
3.1. Characterization of the functionalized materials
3.1.2. DRIFTS, 13C CP/MAS NMR and 1H NMR measurements
DRIFTS, 13C CP/MAS and 1H NMR measurements were car-
ried out to verify if the TiO(acac)2 complex was actually grafted
onto the V-MOF. We selected the sample NH2-MIL-47 [Ti], 70 ◦C,
20 h, for a detailed characterization and subsequent catalytic stud-
ies. In Fig. 2 the 13C CP/MAS NMR spectra of NH2-MIL-47 and
3.1.1. XRPD analysis and determination of the Ti loading
By means of XRF the Ti loading was determined of the func-
tionalized MOF materials (see Table S. 1). Approximately the same
loading was obtained in each experiment, with an average of
Fig. 2. 13C CP/MAS NMR spectra of NH2-MIL-47 and NH2-MIL-47 [Ti]. (INSET): detailed NMR spectrum of the signal at 148 ppm and its deconvolution.
Please cite this article in press as: K. Leus, et al., Ti-functionalized NH2-MIL-47: An effective and stable epoxidation catalyst, Catal. Today (2012),
http://dx.doi.org/10.1016/j.cattod.2012.09.037

GModel
CATTOD-8284; No. of Pages9
ARTICLE IN PRESS
4
K. Leus et al. / Catalysis Today xxx (2012) xxx–xxx
Fig. 4. Model used in the calculation of the IR vibrations after post-functionalization
with the TiO(acac)2 moiety (the C atoms are shown in gray, N atoms in blue, and
the O and Ti atoms are depicted in red and black, respectively). (For interpretation
of the references to color in the artwork, the reader is referred to the web version
of the article.)
Fig. 3. DRIFTS spectra of NH2-MIL-47 (shown in red) and NH2-MIL-47 [Ti], 70 ◦C,
20 h(shown in black) INSET: detailed DRIFTS spectra of both materials in the region
3550–3300 cm−1. The spectra were normalized to the vibration at 800 cm−1 which
corresponds to the aromatic C H out of plane bend mode. (For interpretation of the
references to color in the artwork, the reader is referred to the web version of the
article.)
into account to calculate the spectra. Therefore, in both models
the carboxylic acid group and its coordination with vanadium was
not included since it would not contribute any relevant infor-
mation to the calculated spectrum. The absorbance due to the
vibrations at 1338 cm−1(C N stretch) and the doublet at 3494 cm−1
and 3386 cm−1(NH2 asymmetric and symmetric stretch) decrease
in comparison to the non-functionalized V-MOF [35,36]. Also the
skeletal vibration bands that contain NH2 rocking at 1050 cm−1
and 1100 cm−1 become less intense due to a partial modification
of the amine groups. The presence of the TiO(acac)2 modification
NH2-MIL-47 [Ti] are presented. The 13C CP/MAS NMR spectrum
of NH2-MIL-47 gave several signals at 118, 125, 131 and 146 ppm
associated with different sp2 carbons of the organic linker and one
peak at 171 ppm assigned to the carbonyl groups of the amino
terephthalate linker [32]. In the non-functionalized NH2-MIL-47 an
additional signal at 162 ppm confirmed the incomplete removal of
solvent (DMF) used in the synthesis of the material. After function-
alization, NH2-MIL-47[Ti] showed new bands which corroborate
the presence of the grafted TiO(acac)2 complex on NH2-MIL-47. An
intense signal appeared at ca. 148 ppm. As can be seen from Fig. 2
(inset), this signal could be deconvoluted into 3 peaks: the signal at
146 ppm corresponds to the C N bond (non-functionalized mate-
rial), whereas the signal at 147 and 149 ppm can be associated to
the C N and C N groups of the bimetallic MOF, respectively [32].
This proves the success of the post-functionalization. Moreover,
the NH2-MIL-47 [Ti] showed several peaks in the aromatic region
(120–130 ppm) and a new peak at ca. 191 ppm, which correspond
to the C C and C O groups of the titanylacetylacetonate complex,
respectively [32]. The 1H NMR spectrum (see Fig. S4) of the digested
NH2-MIL-47 [Ti] shows peaks at 2.06, 1.94 and 1.85 ppm due to the
CH3 groups further confirming the presence of acac and imine-
functionalized acac ligands grafted onto NH2-MIL-47.
can be seen by the vibrations at 735 cm−1(C
C H twisting) and
1026 cm−1(C CO C wagging). The C N vibration cannot be seen
since it overlaps with the vibrations from the parent MOF between
1600 cm−1 and 1700 cm−1. The new peak that appears at 1710 cm−1
could not be assigned with this model. This vibration could be
due to free C O vibrations, which are the consequence of some
TiO(acac)2 complexes that break during the post-functionalization.
This would result in a ketone substituent which is not coordinated
with titanium on the MOF.
3.1.3. Nitrogen sorption measurement
In Fig. 5 the nitrogen adsorption isotherms of NH2-MIL-47 and
NH2-MIL-47 [Ti], 70 ◦C, 20 h are compared. The Langmuir surface
area of the non-functionalized and post-functionalized material is
650 m2/g and 190 m2/g respectively. Despite the high loading of Ti
(approximately 30%), the post-functionalized MOF retains a partial
The DRIFTS spectra of the NH2-MIL-47 and NH2-MIL-47[Ti],
70 ◦C, 20 h are shown in Fig. 3 (the DRIFTS spectra of the other
materials are presented in Fig. S.2). It is clear that the charac-
teristic vibrations of the parent MOF are still present. However,
after the post-functionalization a decrease in intensity of some
vibrations is observed, pointing toward a decrease in concentra-
tion of the -NH2 groups. The typical vibrations of the benzene
linker at 1510–1450 cm−1 (aromatic ring stretch), 1225–950 cm−1
(aromatic C H in plane bend) and 900–670 cm−1 (aromatic C
H
out of plane bend) [33] remain unchanged. Furthermore, the sym-
metric and asymmetric CO2 stretching vibrations in the region
1463–1415 cm−1and 1616–1597 cm−1 are still present after graft-
ing [34].
The principal changes in the spectrum before and after post-
functionalization can be connected to the amine moiety. To obtain
more insight in these changes, molecular modeling calculations
were applied. The model of the catalyst was built by isolating
the aromatic part of the linker with a TiO(acac)2 substituent, as
shown in Fig. 4. For comparison, the NH2-MIL-47 was modeled as
aniline. Only the relevant parts of the structure have been taken
Fig. 5. Nitrogen adsorption isotherms of NH2-MIL-47 and NH2-MIL-47 [Ti].
Please cite this article in press as: K. Leus, et al., Ti-functionalized NH2-MIL-47: An effective and stable epoxidation catalyst, Catal. Today (2012),
http://dx.doi.org/10.1016/j.cattod.2012.09.037

GModel
CATTOD-8284; No. of Pages9
ARTICLE IN PRESS
K. Leus et al. / Catalysis Today xxx (2012) xxx–xxx
5
Fig. 6. Cyclohexene conversion (ꢀ) and the yield of cyclohexene oxide (᭹) and 2-cyclohexene-1-one (ꢁ) for NH2-MIL-47 and NH2-MIL-47 [Ti] using acetonitrile as solvent
and oxygen as oxidant in combination with the co-oxidant cyclohexanecarboxaldehyde at a temperature of 40 ◦C. In both experiments a V loading of 0.42 mmol was applied.
porosity. In addition, the calculated pore volume of the parent and
post-modified MOF is 0.30 mL/g and 0.10 mL/g, respectively.
conversion is observed for the 2nd and the 3rd run (30% and 24%).
This is due to the V leaching in each run as can be noted from
Table 1. In each run a small amount of V-leaching is observed
(1,8%, 2.5% and 3.5%) which gives rise to additional defects in the
framework structure. Extra defects present in the structure can
significantly influence the catalytic performance in the additional
3.2. Evaluation of the catalytic performance in the oxidation of
cyclohexene
The catalytic performance of NH2-MIL-47 and NH2-MIL-47 [Ti],
70 ◦C, 20 h was evaluated in the oxidation of cyclohexene. To make
a fair comparison of both materials, the catalyst loading was chosen
such that the number of V-sites was equal in the two experiments.
Fig. 6 depicts the cyclohexene conversion and the detailed prod-
uct distribution of both catalysts in the first run. Several features
can be deduced from this figure. (1) A cyclohexene conversion of
approximately 14.1% and 25% is observed respectively for the non-
functionalized and bimetallic material after 6 h of reaction. This
clearly shows that the bimetallic MOF exhibits a significantly higher
catalytic performance than the (monometallic) V-MOF due to the
presence of the extra active sites. (2) Moreover, faster conversion
is observed for the bimetallic MOF. After 2 h of reaction, 10% of
cyclohexene is already converted, whereas NH2-MIL-47 shows only
6% of conversion. (3) Finally, it can be noted that both materials
show only 2 products: cyclohexene oxide and the radical product 2-
cyclohexene-1-one, which are produced in almost equal amounts.
In Section 3.5 reaction mechanisms are proposed that explain this
product distribution.
Additionally, the homogenous catalyst TiO(acac)2 has been
tested using the same experimental conditions as for the MOF
materials. The same amount of Ti has been used as in the
post-functionalized material. The conversion pattern of this homo-
geneous catalyst is depicted in Fig. 7. After 6 h of reaction a
cyclohexene conversion of 25% is reached, which is practically the
same as for the Ti/V-MOF. It is interesting to note that the product
distribution observed for the homogeneous catalyst is very similar
as for the NH2-MIL-47 and NH2-MIL-47 [Ti]: almost equal percent-
ages of the cyclohexene oxide and the ketone are observed.
Fig. 7. Cyclohexene conversion (ꢀ) and the yield of cyclohexene oxide (᭹) and 2-
cyclohexene-1-one(ꢁ) for TiO(acac)2 using acetonitrile as solvent and oxygen as
oxidant in combination with the co-oxidant cyclohexanecarboxaldehyde at a tem-
perature of 40 ◦C. The same Ti loading was applied as for the NH2-MIL-47 [Ti]
(0.22 mmol Ti).
3.3. Stability and regenerability of the catalysts
To test the regenerability of the V-MOFs, two additional runs
have been executed on both materials. In Fig. 8, the cyclohexene
conversion is shown after each run for the two catalysts, whereas
in Table 1 the TON, TOF and leaching percentage are presented
after each run for both materials. As can be seen from Fig. 8, the
cyclohexene conversion for the post-functionalized material is
slightly enhanced after the first run which is due to the generation
of the extra catalytic V+V sites (see Section 3.4). However, for
the non-functionalized V-MOF a significantly higher cyclohexene
Fig. 8. Cyclohexene conversion for the executed additional runs on NH2-MIL-47 and
NH2-MIL-47 [Ti].
Please cite this article in press as: K. Leus, et al., Ti-functionalized NH2-MIL-47: An effective and stable epoxidation catalyst, Catal. Today (2012),
http://dx.doi.org/10.1016/j.cattod.2012.09.037

GModel
CATTOD-8284; No. of Pages9
ARTICLE IN PRESS
6
K. Leus et al. / Catalysis Today xxx (2012) xxx–xxx
(b)
(a)
0.3
105
Relative intensity (%)
100
95
90
85
80
75
0.2
0.1
0.0
-0.1
-0.2
-0.3
Prior catalytic reaction
120 min catalytic reaction
240 min catalytic reaction
70
65
310 320 330 340 350 360 370 380 390 400
0
20 40 60 80 100 120 140 160 180 200 220 240
Catalytic reaction time (min)
= 9.77 GHz
Magnetic field (mT)
Fig. 9. (a) X-band EPR spectra at RT of the NH2-MIL-47 [Ti] samples before and after catalysis; (b) Relative V+IV EPR signal intensity with respect to the signal of NH2-MIL-47
[Ti] prior to catalysis. The error (1ꢀ) is estimated at 4% (see text). (For interpretation of the references to color in the artwork, the reader is referred to the web version of the
article.)
runs [24]. The NH2-MIL-47 [Ti] did not show any detectable Ti or
V leaching, which clearly demonstrates that the original MOF after
post-functionalization becomes more stable. The exact reason for
the enhanced leaching stability is currently unclear. Both shielding
effects and electronic effects may influence the leaching behavior
and would require advanced modeling techniques [37]. This falls
beyond the scope of this current paper.
MIL-47
NH2-MIL-47
NH2-MIL-47 [Ti]
Comparison of the XRPD pattern of NH2-MIL-47 [Ti] before
catalysis and after each run (see Fig. S.3) clearly shows that the
framework integrity of the MOF is well preserved.
3.4. EPR measurements
The aim of the EPR measurements is gaining insight in the
reaction mechanisms by monitoring the fraction of paramagnetic
V+IV ions in the catalyst. Fig. 9 a shows the RT X-band spectra of
NH2-MIL-47 [Ti] dry powder samples before and after 2 and 4 h
of catalysis. The spectra of the samples prior to and after cataly-
sis appear as single broad lines exhibiting widths of about 9.2 mT
before and 10.4 mT after catalysis. No 51V hyperfine (HF) structure
is resolved in the spectra. The line position in the X-band spectra
is practically the same for all samples and already suggests a signal
assignation to V+IV (g = 1.96 < ge = 2.0023) [38]. Furthermore, spec-
tra were recorded at higher microwave frequency (Q-band), which
partially resolves the anisotropy of the (V+IV = O) O4 complexes. In
Fig. 10 the RT Q-band EPR spectrum of NH2-MIL-47 [Ti] is compared
with that of NH2-MIL-47 and MIL-47. Spectrum simulations show
that in all three samples the paramagnetic species have (approx-
1100 1125 1150 1175 1200 1225 1250 1275 1300 1325 1350 1375
= 33.98 GHz
Magnetic field (mT)
Fig. 10. Q-band (33.98 GHz) EPR spectrum of NH2-MIL-47 [Ti] compared with the
spectra of non-functionalized NH2-MIL-47 and of MIL-47 at RT. (For interpretation
of the references to color in the artwork, the reader is referred to the web version
of the article.)
was resolved in the spectra, most probably as a result of interactions
between spins in these paramagnetically concentrated samples.
The number of paramagnetic centers in the samples was estimated
by comparing the signal intensity with that of a VO(acac)2 grafted
silica sample with known V+IV loading. It was found to be compa-
rable to the number of V-sites in the catalyst, indicating that at RT
EPR provides information on the majority of the metal V-sites, and
not only on a minority fraction, e.g. corresponding to defects.
The intensity of the RT X-band spectra was evaluated by double
integration of the EPR spectra over a range of 80.0 mT centered on
the broad EPR line, in order to avoid errors due to non-perfect back-
ground correction in the broader range recorded. The data points in
Fig. 9b represent the relative intensity with respect to the signal of
NH2-MIL-47 [Ti] prior to catalytic reaction. The error (one standard
deviation) is estimated at 4%, combining the effects of variations in
consecutive measurements (sample positioning) and day-by-day
changes in the spectrometer sensitivity. One observes that within
imately) axial symmetry. The g (along the V O bond direction)
||
and g values obtained through fitting of the spectra are listed in
⊥
Table 2. These values are typical for V+IV species [38] and are in
very good agreement with earlier EPR reports on V-MOFs [39]. It is
worth noting that both in the work of Meilikhov et al. [39] and in
our own previous investigations of V-MOFs [24,25], no HF structure
Table 1
TON, TOF and leaching percentage after each run for both V-based catalysts.
Sample
TONa
TOFb (h−1
)
Leaching (%)
NH2-MIL-47
Run 1
Run 2
17.7
44.6
63.9
6.0
14.1
10.4
1.8
2.5
3.5
Table 2
Best fit principal g values for the V+IV centers in MIL-47, NH2-MIL-47 and NH2-MIL-47
[Ti] at RT (see Fig. 10).
Run 3
NH2-MIL-47 [Ti]
Run 1
Run 2
39.5
44.8
49.5
7.5
12.1
1.1
0
0
0
T
Sample
g
⊥
g
||
Run 3
RT
MIL-47
NH2-MIL-47
NH2-MIL-47 [Ti]
1.972
1.970
1.970
1.941
1.940
1.940
a
TON was calculated after 6 h of reaction.
TOF was calculated after 30 min of reaction.
b
Please cite this article in press as: K. Leus, et al., Ti-functionalized NH2-MIL-47: An effective and stable epoxidation catalyst, Catal. Today (2012),
http://dx.doi.org/10.1016/j.cattod.2012.09.037

GModel
CATTOD-8284; No. of Pages9
ARTICLE IN PRESS
K. Leus et al. / Catalysis Today xxx (2012) xxx–xxx
7
Scheme 1. General mechanism of the epoxidation with O2 and aldehydes, most termination steps where omitted for clarity.
the first 2 h of catalysis, the spectrum intensity decreases about
17%. Thereafter it remains stable or decreases more slowly. A very
similar behavior was observed for MIL-47 in the oxidation of cyclo-
hexene with TBHP/decane as oxidant [24]. Hence, we propose a
similar interpretation: in the first few hours of catalysis a fraction of
the V+IV centers is oxidized to diamagnetic V+V centers. Mechanis-
tic considerations in Section 3.5 suggest a radical parallel pathway,
involving a partial oxidation of the V-metal sites.
presence of metal-oxo species could explain the decrease of the
V+IV signal intensity in the EPR measurements since the generation
of vanadium-oxo species is accompanied by an oxidation to V+V
,
which corresponds to what was already reported for the V-MIL-47
in the cyclohexene oxidation using TBHP as oxidant [24].
Another argument in favor for the mechanism shown in
Scheme 1 is the presence of cyclohexanone in the product mixture.
This is due to the formation of the peroxide radical ((II) in Scheme 1),
followed by the formation of peroxide by a hydrogen abstraction.
Afterwards, the obtained peroxide can react with another peroxide
to form the cyclohexanone and oxygen (see Scheme 2).
3.5. Reaction mechanisms
This type of reaction generally exhibits a high (>80%) selectivity
toward the epoxide when it is catalyzed by complexes where the
metal can only have a low oxidation state (maximum II or III) [44]. A
MOF containing the same metals as those mostly used as homoge-
neous catalysts (Ni, Cu, Co) for this reaction gave similar results [41].
In the present system however, the selectivity toward the epoxide
is significantly lower. It is known that the presence of carboxylic
acid radicals and peroxo radicals ((I) and (II) in Scheme 1) can lead
to oxidation in the allylic position [40,41]. It is likely that these
intermediates are stabilized by titanium and vanadium in the cat-
alyst, leading to a decreased selectivity. Furthermore, the presence
of M O species and peroxide moieties can lead to the formation
Based on the above described EPR and catalytic results a
plausible reaction mechanism is proposed. The mechanism of
oxidation using molecular oxygen and a sacrificial co-oxidant has
been the subject of many studies. A radical mechanism involving
a peracid radical as primary oxidizing species has been generally
accepted (Scheme 1) [40]. The role of the catalyst is considered to
be the activation of molecular oxygen by the generation of radicals.
However, a metal-oxygen moiety can be present as alternative,
direct oxidizing species in the metal-catalyzed system. This oxi-
dation pathway is generally accepted for late transition metals
[41,42]. Nevertheless, for metals like titanium and vanadium,
this pathway is less likely as described in Sheldon et al. [43]. The
Scheme 2. formation of cyclohexanone, which was observed in the product mixture.
Please cite this article in press as: K. Leus, et al., Ti-functionalized NH2-MIL-47: An effective and stable epoxidation catalyst, Catal. Today (2012),
http://dx.doi.org/10.1016/j.cattod.2012.09.037

GModel
CATTOD-8284; No. of Pages9
ARTICLE IN PRESS
8
K. Leus et al. / Catalysis Today xxx (2012) xxx–xxx
Scheme 3. Pathway to cyclohexenol formation.
of cyclohexenol as shown in Scheme 3. Cyclohexenol can then be
oxidized further to 2-cyclohexene-1-one, this conversion was also
shown by Murahashi et al. [44].
[8] M.G. Goesten, J. Juan-Alcaniz, E.V. Ramos-Fernandez, K.B.S.S. Gupta, E. Stavitski,
H. van Bekkum, J. Gascon, F. Kapteijn, Journal of Catalysis 281 (2011) 177–187.
[9] W.Q. Kan, B. Liu, J. Yang, Y.Y. Liu, J.F. Ma, Crystal Growth and Design 12 (2012)
2288–2298.
[10] J. Rocha, L.D. Carlos, F.A.A. Paz, D. Ananias, Chemical Society Reviews 40 (2011)
926–940.
[11] H. Liu, Y.G. Zhao, Z.J. Zhang, N. Nijem, Y.J. Chabal, H.P. Zeng, J. Li, Advanced
Functional Materials 21 (2011) 4754–4762.
4. Conclusions
12 Y.G. Zhao, H.H. Wu, T.J. Emge, Q.H. Gong, N. Nijem, Y.J. Chabal, L.Z. Kong, D.C.
Langreth, H. Liu, H.P. Zeng, J. Li, Chemistry – A European Journal 17 (2011)
5101–5109.
[13] T.K. Prasad, D.H. Hong, M.P. Suh, Chemistry – A European Journal 16 (2010)
14043–14050.
[14] J.P.S. Mowat, S.R. Miler, J.M. Griffin, V.R. Seymour, S.E. Ashbrook, S.P. Thompson,
D. Fairen-Jimenez, A.M. Banu, T. Duren, P.A. Wright, Inorganic Chemistry 50
(2011) 10844–10858.
[15] C. Zlotea, D. Phanon, M. Mazaj, D. Heurtaux, V. Guillerm, C. Serre, P. Horcajada,
T. Devic, E. Magnier, F. Cuevas, G. Ferey, P.L. Llewellyn, M. Latroche, Dalton
Transactions 40 (2011) 4879–4881.
We succeeded in the post-functionalization of NH2-MIL-47 with
TiO(acac)2.13C CP/MAS NMR, 1H NMR and DRIFTS measurements
clearly established the effectiveness of the grafting procedure,
whereas XRPD measurements proved the stability of the bimetal-
lic MOF during the grafting process. The obtained NH2-MIL-47 [Ti]
exhibits a significantly higher cyclohexene conversion compared
to the non-functionalized material, probably mainly due to the
extra active sites. Furthermore, no leaching of V or Ti was detected.
Regenerability tests have shown that only a slight increase in
cyclohexene conversion is observed during the first 3 runs. EPR
measurements indicate that about 17% of the V+IV sites are oxi-
dized to V+V in the first two hours of catalysis. In agreement with
the latter, the general mechanism of the epoxidation with O2 and
aldehydes shows the generation of vanadium-oxo species which is
[16] F. Vermoortele, M. Vandichel, B. Van de Voorde, R. Ameloot, M. Waroquier,
V. Van Speybroeck, D. De Vos, Angewandte Chemie International Edition 51
(2012) 4887–4890.
[17] S.M. Cohen, Chemical Reviews 112 (2012) 970–1000.
[18] N.V. Maksimchuk, K.A. Kovalenko, S.S. Arzumanov, Y.A. Chesalov, M.S. Mel-
gunov, A.G. Stepanov, V.P. Fedin, O.A. Kholdeeva, Inorganic Chemistry 49 (2010)
2920–2930.
[19] M.H. Alkordi, Y.L. Liu, R.W. Larsen, J.F. Eubank, M. Eddaoudi, Journal of the
American Chemical Society 130 (2008) 12639–12641.
accompanied by a partial oxidation of the V centers toward V+V
.
[20] O.V. Zalomaeva, K.A. Kovalenko, Y.A. Chesalov, M.S. Mel’gunov, V.I. Zaikovskii,
V.V. Kaichev, A.B. Sorokin, O.A. Kholdeeva, V.P. Fedin, Dalton Transactions 40
(2011) 1441–1444.
Acknowledgments
[21] M.J. Ingleson, J.P. Barrio, J.B. Guilbaud, Y.Z. Khimyak, M.J. Rosseinsky, Chemical
Communications (23) (2008) 2680–2682.
[22] S. Bhattacharjee, D.A. Yang, W.S. Ahn, Chemical Communications 47 (2011)
3637–3639.
[23] K. Leus, I. Muylaert, M. Vandichel, G.B. Marin, M. Waroquier, P. Van der Voort,
Chemical Communications 46 (2010) 5085–5087.
[24] K. Leus, M. Vandichel, Y.Y. Liu, I. Muylaert, J. Musschoot, S. Pyl, H. Vrielinck, F.
Callens, G.B. Marin, C. Detavernier, P.V. Wiper, Y.Z. Khimyak, M. Waroquier, V.
Van Speybroeck, P. Van der Voort, Journal of Catalysis 285 (2012) 196–207.
[25] Y.Y. Liu, K. Leus, M. Grzywa, D. Weinberger, K. Strubbe, H. Vrielinck, R. Van
Deun, D. Volkmer, V. Van Speybroeck, P. Van der Voort, European Journal of
Inorganic Chemistry 16 (2012) 2819–2827.
[26] I. Muylaert, J. Musschoot, K. Leus, J. Dendooven, C. Detavernier, P. Van der Voort,
European Journal of Inorganic Chemistry (2) (2012) 251–260.
[27] S. Stoll, A. Schweiger, Journal of Magnetic Resonance 178 (2006) 42–55.
[28] M.J.T. Frisch, G.W. Schlegel, H.B. Scuseria, G.E. Robb, M.A. Cheeseman, J.R. Scal-
mani, G. Barone, V. Mennucci, B. Petersson, G.A. Nakatsuji, H. Caricato, M. Li,
X. Hratchian, H.P. Izmaylov, A.F. Bloino, J. Zheng, G. Sonnenberg, J.L. Hada, M.
Ehara, M. Toyota, K. Fukuda, R. Hasegawa, J. Ishida, M. Nakajima, T. Honda,
Y. Kitao, O. Nakai, H. Vreven, T. Montgomery Jr., J.A. Peralta, J.E. Ogliaro, F.
Bearpark, M. Heyd, J.J. Brothers, E. Kudin, K.N. Staroverov, V.N. Kobayashi, R.
Normand, J. Raghavachari, K. Rendell, A. Burant, J.C. Iyengar, S.S. Tomasi, J.
Cossi, M. Rega, N. Millam, N.J. Klene, M. Knox, J.E. Cross, J.B. Bakken, V. Adamo,
C. Jaramillo, J. Gomperts, R. Stratmann, R.E. Yazyev, O. Austin, A.J. Cammi, R.
Pomelli, C. Ochterski, J.W. Martin, R.L. Morokuma, K. Zakrzewski, V.G. Voth,
G.A. Salvador, P. Dannenberg, J.J. Dapprich, S. Daniels, A.D. Farkas, O. Foresman,
J.B. Ortiz, J.V. Cioslowski, J. Fox, D.J. Gaussian 09, {R}evision {A}. 02, Gaussian,
Inc., Wallingford CT, 2009.
K.L. is grateful to the Long Term Structural Methusalem Grant
No. 01M00409 funding by the Flemish Government. Furthermore,
this research is co-funded by Ghent University, GOA Grant No.
01G00710, the Research Foundation Flanders (FWO-Vlaanderen)
through Grant No. G.0700.08, BELSPO in the frame of IAP 6/27 and
the European Research Council (FP7(2007–2013) ERC Grant No.
240483). Computational resources (Stevin Supercomputer Infra-
structure) and services were provided by Ghent University. The
authors would like to thank the research group FQM-346 and the
SCAI from the University of Córdoba (Spain) for the 13C CP/MAS
NMR measurements. The NMRSTR research group of Ghent Univer-
sity is acknowledged for performing the 1H NMR measurements.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.cattod.
2012.09.037.
References
[29] A.D. Becke, Journal of Chemical Physics 98 (1993) 5648–5652.
[30] C.T. Lee, W.T. Yang, R.G. Parr, Physical Review B 37 (1988) 785–789.
31 J.P. Merrick, D. Moran, L. Radom, Journal of Physical Chemistry A 111 (2007)
11683–11700.
32 E. Pretsch, P. Bühlmann, C. Affolter, A. Herrera, R. Martínez, Determinaci on
estructural de compuestos org anicos, Elsevier Masson (2001).
[33] J. Coates (Ed.), Interpretation of Infrared Spectra, A Practical Approach, 2000.
[34] T. Loiseau, C. Serre, C. Huguenard, G. Fink, F. Taulelle, M. Henry, T. Bataille, G.
Ferey, Chemistry – A European Journal 10 (2004) 1373–1382.
[35] J. Gascon, U. Aktay, M.D. Hernandez-Alonso, G.P.M. van Klink, F. Kapteijn, Jour-
nal of Catalysis 261 (2009) 75–87.
[1] J.R. Li, J. Sculley, H.C. Zhou, Chemical Reviews 112 (2012) 869–932.
[2] J. Sculley, D.Q. Yuan, H.C. Zhou, Energy and Environmental Science 4 (2011)
2721–2735.
[3] J.R. Li, Y.G. Ma, M.C. McCarthy, J. Sculley, J.M. Yu, H.K. Jeong, P.B. Balbuena, H.C.
Zhou, Coordination Chemistry Reviews 255 (2011) 1791–1823.
[4] D. Farrusseng, S. Aguado, C. Pinel, Angewandte Chemie International Edition
48 (2009) 7502–7513.
[5] J. Lee, O.K. Farha, J. Roberts, K.A. Scheidt, S.T. Nguyen, J.T. Hupp, Chemical Society
Reviews 38 (2009) 1450–1459.
[6] Y. Liu, W.M. Xuan, Y. Cui, Advanced Materials 22 (2010) 4112–4135.
[7] A. Corma, H. Garcia, F.X.L. Xamena, Chemical Reviews 110 (2010) 4606–4655.
Please cite this article in press as: K. Leus, et al., Ti-functionalized NH2-MIL-47: An effective and stable epoxidation catalyst, Catal. Today (2012),
http://dx.doi.org/10.1016/j.cattod.2012.09.037

GModel
CATTOD-8284; No. of Pages9
ARTICLE IN PRESS
K. Leus et al. / Catalysis Today xxx (2012) xxx–xxx
9
[36] M. Kandiah, M.H. Nilsen, S. Usseglio, S. Jakobsen, U. Olsbye, M. Tilset, C.
Larabi, E.A. Quadrelli, F. Bonino, K.P. Lillerud, Chemistry of Materials 22 (2010)
6632–6640.
[37] J.M. Ho, C.J. Easton, M.L. Coote, Journal of the American Chemical Society 132
(2010) 5515–5521.
[38] R. Pilbrow (Ed.), Transition Ion Electron Paramagnetic Resonance, Clarendon
Press, Oxford, 1990.
[39] M. Meilikhov, K. Yusenko, A. Torrisi, B. Jee, C. Mellot-Draznieks, A.
Poppl, R.A. Fischer, Angewandte Chemie International Edition 49 (2010)
6212–6215.
[40] B.B. Wentzel, P.L. Alsters, M.C. Feiters, R.J.M. Nolte, Journal of Organic Chemistry
69 (2004) 3453–3464.
[41] E. Angelescu, O.D. Pavel, R. Ionescu, R. Birjega, M. Badea, R. Zavoianu, Journal of
Molecular Catalysis A Chemical 352 (2012) 21–30.
[42] P. Mastrorilli, C.F. Nobile, G.P. Suranna, L. Lopez, Tetrahedron 51 (1995)
7943–7950.
[43] G.J. ten Brink, I.W.C.E. Arends, R.A. Sheldon, Chemical Reviews 104 (2004)
4105–4123.
[44] N. Komiya, T. Naota, Y. Oda, S.I. Murahashi, Journal of Molecular Catalysis A
Chemical 117 (1997) 21–37.
Please cite this article in press as: K. Leus, et al., Ti-functionalized NH2-MIL-47: An effective and stable epoxidation catalyst, Catal. Today (2012),
http://dx.doi.org/10.1016/j.cattod.2012.09.037