Journal of Molecular Catalysis A: Chemical 423 (2016) 248–255

Contents lists available at ScienceDirect

Journal of Molecular Catalysis A: Chemical

journal homepage: www.elsevier.com/locate/molcata

Editor’s choice paper

Photo-epoxidation of cyclohexene, cyclooctene and 1-octene with

molecular oxygen catalyzed by dichloro
dioxo-(4,4 -dicarboxylato-2,2 -bipyridine) molybdenum(VI) grafted on
mesoporous TiO2
Henry Martínez a,∗ , María F. Cáceres b , Fernando Martínez a , Edgar A. Páez-Mozo a ,
Sabine Valange c,∗ , Nelson J. Castellanos d , Daniel Molina e , Joël Barrault c ,
Henri Arzoumanian f
a
Centro de Investigaciones en Catálisis-CICAT, Universidad Industrial de Santander, Escuela de Química, Km 2 vía El Refugio, Piedecuesta, Santander,
Colombia
b
Laboratório de Química Atmosférica (LQA), Universidade Católica do Río de Janeiro, Rua Marquês de Sao Vicente, 225, Gávea-Rio de Janeiro, RJ, Brazil
c
Institut de Chimie des Milieux et Matériaux de Poitiers (IC2MP), Université de Poitiers, CNRS, ENSIP, B1, 1 rue M. Doré, F-86073 Poitiers, France
d
Estado Sólido y Catálisis Ambiental (ESCA), Universidad Nacional de Colombia, Carrera 30 No. 45-03, Bogotá, Colombia
e
Laboratorio de Resonancia Magnética Nuclear, Universidad Industrial de Santander, Km 2 vía El Refugio, Piedecuesta, Colombia
f
Chirosciences UMR CNRS 7313 Institut des Sciences Moléculaires de Marseille (iSm2), Aix-Marseille Université, Campus St Jérôme, 13397 Marseille, France

a r t i c l e i n f o a b s t r a c t

Article history: The photo-assisted oxygen transfer to cyclohexene, cyclooctene and 1-octene was investigated
Received 12 November 2015 using molecular O2 as a primary oxidant with a dichloro-dioxo-(4,4 -dicarboxylato-2,2 -bipyridine)
Received in revised form 31 May 2016 molybdenum(VI) complex grafted on the surface of mesoporous TiO2 supports exhibiting different tex-
Accepted 4 July 2016
tural properties. Such mesoporous titania samples were prepared following a hydrothermal treatment
Available online 6 July 2016
(HT) and through an original procedure under supercritical CO2 (SC) conditions. Non porous TiO2 Degussa
P-25 was also used for comparison purposes. The porosity of the titania supports, as well as the amount
Keywords:
and dispersion of the covalently anchored dioxo Mo(VI) complex were shown to strongly influence the
Dioxomolybdenum complex
Mesoporous TiO2
resulting photo-catalytic properties. A particular attention was also devoted to the stability of the grafted
Photocatalysis catalysts by performing long-term epoxidation experiments. The new hybrid Mo(VI) -derivatized meso-
Selective oxidation porous SC TiO2 catalyst proved to be the most efficient in the photo-assisted oxygen transfer to various
Epoxidation alkenes. Additionally, the high stability of such catalyst was ascertained by recycling studies.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction chemicals, pharmaceuticals, agrochemical products, epoxy paints

Traditional oxidation reactions based on stoichiometric oxi-
dants tend to produce waste side products generating a negative
environmental impact [1]. In this respect, the use of environmen-
tally friendly oxidants such as molecular oxygen is a more attractive
and challenging option in order to generate oxygenates without
formation of by-products [2–4].
Epoxidation of alkenes is a suitable route prime of importance
for the preparation of relevant industrial intermediates [5]. Epox-
ides are versatile and reactive compounds for the synthesis of fine
∗ Corresponding authors.
E-mail addresses: henry.martinez1@correo.uis.edu.co (H. Martínez),
sabine.valange@univ-poitiers.fr (S. Valange).
and dyes [6,7]. However, the production of epoxide compounds is
still a big issue because several side reactions can take place, such
as oxidation in the allylic positions, ring opening of the epoxides
or complete oxidation to CO2 [8]. Furthermore, the epoxidation of
terminal olefins such as 1-octene is difficult and therefore requires
long time reaction [9]. Traditionally, epoxides are prepared by
partial oxidation of alkenes using “exotic” oxidants like iodosyl-
benzene, hypochlorite or organic hydroperoxides [10]. Hydrogen
peroxide or t-butyl hydroperoxide (TBHP) are known as efficient
oxidants [11] with the assistance of homogeneous transition metal-
based catalysts [12–14]. The epoxidation of ethylene and propylene
was recently reported with molecular oxygen or air as a primary
oxidant with a heterogeneous silver catalyst [15,16]. However, the
reaction conditions are rather severe and risky. Thus, the develop-
ment of catalytic systems that use molecular oxygen or air in mild

http://dx.doi.org/10.1016/j.molcata.2016.07.006
1381-1169/© 2016 Elsevier B.V. All rights reserved.
 H. Martínez et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 248–255
conditions remains a challenging target [17]. In previous studies
[18–21], we have reported on the selective production of organic
oxygenated compounds by Oxygen Atom Transfer (OAT) reactions,
using O2 as the primary oxidant, with a dioxoMo(VI) complex
grafted on a commercial TiO2 (MoO2 L/TiO2 ) and UV–vis irradiation.
Here, the photo-assisted oxygen transfer to cyclohexene,
cyclooctene and 1-octene was investigated on new grounds with
a dichloro-dioxo-(4,4 -dicarboxylato-2,2 -bipyridine)-Mo(VI) com-
plex (labelled as Mo(VI) Cl2 O2 Bipy) grafted on the surface of various
mesoporous titania powders. Two different TiO2 synthesis routes
were explored with the aim of highlighting the influence of the
porosity on the dispersion of the Mo(VI) Cl2 O2 Bipy complex, and
hence, on the resulting photo-catalytic properties. High surface
area mesoporous TiO2 supports with anatase walls were prepared
following a CTAB-assisted hydrothermal process, as well as by a
sol-gel route under supercritical carbon dioxide (SC CO2 ). As a non-
toxic and environmentally benign green solvent, SC CO2 exhibits
many unique features and great versatility in materials process-
ing and synthesis. Such characteristics allow for a better control of
the nanoarchitecture and the textural properties of the resulting
materials [22].
Several complementary techniques, including XRD, nitrogen
sorption isotherm, Raman, ATR-FTIR, UV–vis and 13 C CP-MAS NMR
spectroscopies were used to characterize the parent mesoporous
TiO2 supports and the Mo(VI) Cl2 O2 Bipy grafted counterparts.
Finally, the stability of these hybrid catalysts was evaluated after
long-term photo-epoxidation experiments (56 h).
2. Experimental
2.1. Materials
All chemical reagents were purchased and used without fur-
ther purification unless otherwise specified. All solvents were
thoroughly degassed prior to use. Acetonitrile was distilled and
kept under argon. Cetyltrimethylammonium bromide (CTAB) and
HCl (37%) were purchased from Acros, while ethanol (absolute
for analysis) and NaOH were obtained from Aldrich. Titanium
oxide (Degussa-P25) with a specific area of 50 m2 g−1 and a crys-
talline structure composed of 80% anatase and 20% rutile was
used for comparison. All supports were dehydrated at 60 ◦ C under
5 × 10−4 mbar for 48 h prior to the anchoring procedure.
2.2. Catalysts synthesis
2.2.1. CTAB-assisted hydrothermal route
The hydrothermal preparation route of the mesoporous TiO2
was based on the following procedure. An aqueous solution of tita-
nium sulphate (1.1402 g) was first prepared and added to a CTAB
ethanolic solution under stirring. The molar ratio of Ti(SO4 )2 :CTAB
is 1:0.12. The resulting mixture was then aged at room tempera-
ture for 12 h before being transferred to an autoclave at 100 ◦ C for
72 h. After the hydrothermal treatment, the solid was recovered by
centrifugation, washed with water and ethanol, dried at 120 ◦ C and
then calcined at 400 ◦ C in air for 6 h with a heating rate of 2 ◦ C min−1
[23]. The TiO2 sample was labelled as TiO2 (HT-400).
2.2.2. SC CO2 -assisted sol-gel route
This procedure consists in the elaboration of mesoporous TiO2
by a sol-gel route under supercritical CO2 through hydrolysis-
condensation of titanium tetraisopropoxide at 150 ◦ C under
30 MPa. Briefly, titanium tetraisopropoxide was introduced into the
reactor prior to the injection of CO2 . As soon as the desired reac-
tion was reached, an excess of water was added under supercritical
conditions. The alcohol formed through hydrolysis as well as the
249
excess of precursor were separated from the reaction medium dur-
ing the CO2 expansion. The TiO2 sample synthesized under SC CO2
was referred to as TiO2 (SC-150).
2.2.3. Mo(VI) Cl2 O2 Bipy grafting procedure
The covalent grafting of the dichloro-dioxo-(4,4 -dicarboxylato-
2,2 -bipyridine)-Mo(VI) complex on the surface of both dehydrated
titania supports was carried out according to the following
procedure. Hexamethyldisilazane was added to a benzene sus-
pension of TiO2 , stirred 24 h at room temperature, filtered,
washed thoroughly with benzene and dried under vacuum. A
benzene suspension of freshly prepared trimethylsilylated TiO2
was then added to a benzene suspension of 2,2 -bipyridyl-4,4 -
dicarboxylic acid, stirred 72 h at room temperature, filtered
and thoroughly washed with benzene. Finally, a THF suspen-
sion of MoO2 Cl2 was added to a benzene suspension of the
previously obtained solid, stirred 4 h at room temperature and
the excess of solvent evaporated. The solid product was then
washed with acetonitrile and dried at room temperature [20]. The
dioxo Mo(VI) grafted titania samples were designated as follows:
Mo(VI) Cl2 O2 Bipy/TiO2 (HT-400), Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150) and
Mo(VI) Cl2 O2 Bipy/TiO2 (P-25), where the label in brackets stands for
the titania support prepared under hydrothermal treatment, super-
critical conditions and the reference Degussa TiO2 , respectively.
2.3. Catalysts characterization
The TiO2 supports and the grafted catalysts were character-
ized by powder X-ray diffraction (XRD) using a Bruker AXS D8
Advance DaVinci geometry with monochromatized Cu K␣ radia-
tion (␭ = 1.5418 Å) at 40 kV and 30 mA. The diffraction patterns were
recorded in the 2 value range of 20–70◦ (with a step size of 0.01◦
and a step time of 0.4 s) for wide-angle analysis and at 2 = 2–10◦
for low-angle analysis. Nitrogen adsorption/desorption isotherms
at 77 K were measured on a Micromeritics Tristar 3000 apparatus
at −196 ◦ C. Before analysis, the samples were degassed under vac-
uum at 150 ◦ C for 12 h. The specific surface area was determined
from the linear part (0–0.23 P/Po ) of the BET plot. The total pore
volume was measured from the isotherms at P/Po = 0.95 and the
mean pore diameter was determined by the BJH method applied
to the adsorption branch. Raman analysis of both supports and
catalysts was carried out on a LabRAM HR800UV HORIBA Jobin
Yvon spectrometer with a BXFM microscope, equipped with a CCD
detector and an internal HeNe laser (632.81 nm) and an external
Ar laser (514.532 nm). All grafted Mo(VI) Cl2 O2 Bipy/TiO2 catalysts
were characterized by attenuated total reflectance (ATR FT-IR)
spectroscopy on a Perkin Elmer spectrometer and 13 C CP-MAS
NMR spectroscopy on a Bruker Avance-400 MHz NMR spectrome-
ter equipped with a 4 mm magic angle spinning (MAS) probe head
for solids. The CP technique was applied during Magic Angle Spin-
ning (MAS) of the rotor at 10 kHz. A ramped 1H-pulse starting at
100% power and decreasing until 50% was used during contact
time (2 ms) in order to circumvent Hartmann–Hahn mismatches.
In order to obtain a good signal-to-noise ratio in 13 C CPMAS exper-
iment, 2048 scans were accumulated using a delay of 2 s. The
13 C chemical shifts were referenced to tetramethylsilane and cal-
ibrated with the adamantane signal set at 38.48 ppm [24]. Diffuse
reflectance spectra (200–700 nm) were measured on a Cary 5000
Varian spectrophotometer. The amount of grafted complex was
determined by thermo gravimetric analyses (TGA) under N2 atmo-
sphere between 30 and 900 ◦ C with a 10 ◦ C min−1 heating ramp
using a SDT-Q600 apparatus. Molybdenum elemental analysis was
carried out with a Thermo S4 atomic absorption spectrophotometer
(AAS), after acid digestion of the samples.
250 H. Martínez et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 248–255
Fig. 1. Wide and small-angle (inset) XRD patterns obtained for (a) TiO2 (HT-400), (b)
Mo(VI) Cl2 O2 Bipy/TiO2 (HT-400), (c) TiO2 (SC-150) and (d) Mo(VI) Cl2 O2 Bipy/TiO2 (SC-
150).
2.4. Photocatalytic experiments
The photocatalytic-assisted alkene epoxidation was carried out
by using a 15 mL batch microreactor (ACEGLASS) equipped with
a mercury lamp (UV PenRay, ␭ ≥ 360 nm) at a constant temper-
ature of 19 ◦ C. All substrates dissolved in CH3 CN (10−2 mol L−1 )
were thoroughly deoxygenated by bubbling N2 for several hours
before the addition of the catalyst (15 mg). The reaction system was
allowed to equilibrate during 1 h under O2 atmosphere in the dark.
The photo-reaction was performed for 5 h under O2 and UV–vis
light. Then the reaction system was maintained in the dark with
O2 during 12 h. The illumination-darkness cycle was repeated four
more times. In order to evaluate the stability of the catalyst, the
powder was separated from the reaction medium at the end of the
photo-oxidation process and washed with acetonitrile. Then the
recovered catalyst was introduced into a new CH3 CN-based olefin
solution prior to the photo-epoxidation reaction. The illumination-
darkness cycle was repeated five times. Long term photo-oxidation
reactions were also carried out under O2 atmosphere and UV–vis
illumination during 35 h and 56 h. The catalyst reuse was evaluated
by performing six consecutive oxidation runs with the same cata-
lyst, each time using a fresh alkene solution. All photo-catalytic
experiments were performed with the MoCl2 O2 Bipy/TiO2 (P-25)
catalyst used for comparison purposes.
The reaction products were analysed by using a gas chromatog-
raphy (GC-HP-6890) equipped with a flame ionization detector and
a HP-INNOWAX column (30 m × 0.32 mm × 0.25 ␮m). The quantifi-
cation of the products was performed using benzene as internal
standard.
3. Results and discussion
3.1. Catalysts characterization
Wide-angle X-ray diffraction analyses indicated that anatase
is the main phase for both HT-400 and SC-150 TiO2 supports
(JCPDS No. 21-1272), along with traces of rutile in the case of the
mesoporous titania prepared under hydrothermal heating. Fig. 1
confirmed the presence of the (101), (004), (200), (105) and (211)
diffraction peaks characteristics of anatase [25] for the parent
supports, as well as for both TiO2 grafted with the dichloro-dioxo-
(4,4 -dicarboxylato-2,2 -bipyridine)-Mo(VI) complex. The thermal
stability of the TiO2 mesostructures prepared under hydrothermal
or supercritical conditions were clearly evidenced by the low-angle
Fig. 2. N2 adsorption-desorption isotherms for the mesoporous (a) TiO2 (SC-150)
and (b) TiO2 (HT-400). Insets showed the corresponding pore size distribution
obtained from the adsorption loop.
XRD patterns (inset). The diffractograms were similar and exhib-
ited a sharp diffraction peak located at 2 = 3◦ . The TiO2 sample
synthesized through the SC CO2 -assisted sol-gel route displayed
however a broadening coupled to a reduction in the intensity of
this low-angle peak. After the Mo(VI) Cl2 O2 Bipy grafting procedure,
the samples still exhibited a single strong XRD peak indicating that
the TiO2 mesoporous structure was maintained. Moreover no shift
of this diffraction peak was observed after the functionalization
step, thereby confirming that no change occurred in the lattice
parameter and in the wall thickness of the titania supports.
The N2 adsorption-desorption isotherms of the titania supports
and their corresponding pore size distributions are shown in Fig. 2.
According to the IUPAC classification, these isotherms are of type
IV and are attributed to mesoporous materials. The shape of the
isotherms also indicated that the TiO2 (HT-400) sample contains
larger mesopores than the support prepared under supercritical
conditions. A wider pore size distribution was indeed observed
for the TiO2 (HT-400) titania (Fig. 2b) compared with that of the
TiO2 (SC-150) sample (Fig. 2a). The specific surface areas, mean
pore diameters and pore volumes of the TiO2 samples before and
after the Mo(VI) Cl2 O2 Bipy grafting procedure are summarized in
Table 1. The corresponding data for the reference P-25 TiO2 and
the Mo(VI) Cl2 O2 Bipy/TiO2 (P-25) counterpart are also provided for
comparison. It appeared that the TiO2 support prepared by the sol-
gel route under supercritical CO2 exhibited a higher BET surface
area coupled to a higher pore volume with respect to the meso-
porous titania obtained after hydrothermal treatment. As expected
after the dioxo-Mo(VI) complex immobilization step, a decrease of
adsorbed N2 , surface area, pore volume and of the pore diameter
was observed for all grafted samples when compared to the parent
TiO2 supports.
Fig. 3 shows the Raman spectra of bare TiO2 supports and
their Mo(VI) Cl2 O2 Bipy/TiO2 analogous systems. Both titania sup-
ports exhibited the characteristic peaks of the anatase phase at
145 cm−1 (symmetric stretching vibration of O Ti O), 390 cm−1
(symmetric bending vibration of O Ti O), 509 cm−1 (antisymmet-
ric bending vibration of O-Ti-O) and 632 cm−1 , corresponding to
the Eg(1) , B1g(1) , (A1g + B1g(2) ) and Eg(2) modes, respectively [26,27].
For the Mo(VI) Cl2 O2 Bipy/TiO2 (HT and SC) systems, all the vibra-
tion bands for anatase can still be found (inset). Additionally, the
Raman spectra of the grafted samples showed the vibrational bands
related to the symmetric stretching of the bipyridyl ligand at 1631,
1561, 1505, 1429 and 1301 cm−1 , as well as the ring breathing
 H. Martínez et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 248–255
Table 1
Textural properties, density of TiO2 surface hydroxyl groups and content of grafted MoCl2 O2 Bipy complex for the titania supports before and after immobilization.
Solid S BET (m2 g−1 ) Vp (cm3 g−1 )
TiO2 (P-25) 50 0.08
Mo(VI) Cl2 O2 Bipy/TiO2 (P-25) 47 0.07
TiO2 (HT-400) 92 0.26
Mo(VI) Cl2 O2 Bipy/TiO2 (HT-400) 84 0.20
TiO2 (SC-150) 150 0.32
Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150) 126 0.18
a
Determined by TGA.
b
Mo(VI) -dioxo complex concentration determined by Mo elemental analysis.
Fig. 3. Raman spectra of the mesoporous anatase TiO2 supports and the correspond-
ing MoCl2 O2 Bipy/TiO2 systems.
at 1025 cm−1 . The characteristic Mo = O symmetric vibration was
observed at 949 cm−1 [12]. All these bands confirmed the presence
of the Mo(VI) Cl2 O2 Bipy moiety in both TiO2 supports. Moreover we
can also observed that the Raman band intensity of the immobi-
lized dioxo-Mo(VI) complex was higher for TiO2 (SC-150) than for
TiO2 (HT-400), thereby suggesting that the titania prepared through
the supercritical route contained an increased amount of grafted
Mo(VI) Cl2 O2 Bipy than that of the hydrothermally heated one.
The amount of the dioxomolybdenum(VI) complex covalently
anchored on all TiO2 supports was determined by Mo elemental
analysis (Table 1). The use of atomic absorption spectrophotom-
etry was indeed preferred in order to get an accurate value of
the grafted Mo-dioxo active centers. The determination of the
loading of the Mo(VI) complex through C, H and N elemental anal-
ysis before and after the grafting procedure was also performed
but proved to be not well suited for a precise evaluation of the
Mo content. The percentage of carbon, hydrogen and nitrogen
determined for all Mo(VI) Cl2 O2 Bipy/TiO2 samples was systemati-
cally higher than that of the stoichiometric amount corresponding
to the dichloro-dioxo-(4,4 -dicarboxylato-2,2 -bipyridine)-Mo(VI)
complex, suggesting the presence of solvent molecules trapped
within the pores of the TiO2 supports.
Important differences were observed regarding the concentra-
tion of the grafted Mo(VI) complex for all catalysts. While this
amount was in the same range for the TiO2 (HT-400) and P-25 sup-
ports, the mesoporous TiO2 synthesized under supercritical CO2
was shown to exhibit a far increased content of covalently anchored
dioxo-Mo(VI) complex as high as 0.43 mmol g−1 (Table 1). Such dif-
ference could be due to the higher surface density of hydroxyl
groups exhibited by the TiO2 (SC-150) support with respect to the
titania prepared under hydrothermal treatment (TiO2 (HT-400)).
The anchoring of the Mo(VI) Cl2 O2 Bipy complex on both meso-
porous TiO2 was further confirmed by FTIR and the band
251
Mean Øp (nm) OH (mmol g−1 )a MoCl2 O2 Bipy (mmol g−1 )b
– 0.37 –
– 0.08 0.14
11.3 1.31 –
10.2 0.92 0.19
7.5 3.51 –
6.1 2.53 0.48
Fig. 4. Total reflectance (ATR-FTIR) spectra of a) reference TiO2 (P-25),
(b) Mo(VI) Cl2 O2 Bipy/TiO2 (P-25), c) Mo(VI) Cl2 O2 Bipy/TiO2 (HT-400) and d)
Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150).
assignments are provided in Table S1 (ESI). The asymmetric and
symmetric stretching vibrations of the cis MoO2 group were
observed at 940 cm−1 and 918 cm−1 , respectively (Fig. 4) [18].
The stretching vibration bands related to the carboxylate lig-
and were observed at 1717 cm−1 (C O), 1271 cm−1 (asymmetric
COO− ) and 1125 cm−1 (symmetric COO− ) [19]. The Ti COO− link-
age was identified by the 1396 cm−1 asymmetric and 1368 cm−1
symmetric scissor vibration modes [20,28]. The IR spectra of
the Mo(VI) Cl2 O2 Bipy/TiO2 samples clearly showed that the dioxo
molybdenum complex was covalently bonded to the mesoporous
TiO2 supports.
Diffuse reflectance UV–vis analysis provided additional evi-
dence for the successful anchoring of the MoCl2 O2 Bipy complex
on the surface of the two mesoporous titania supports and of the
reference P-25 TiO2 . The comparison of the DR UV–vis spectra of
the titania before and after the dioxo-MoVI immobilization indi-
cated a new absorption band at ca. 560 nm (Fig. 5), related to the
presence of the Mo = O group (n → ␲* transition) [21]. The inten-
sity of this band was higher for the TiO2 (SC-150) sample than for
TiO2 (HT-400), in accordance with the higher amount of complex
immobilized on the surface of the TiO2 synthesized under super-
critical CO2 (Table 1).
Fig. 6 shows the 13 C NMR spectra of the mesoporous
and P-25 Mo(VI) Cl2 O2 Bipy/TiO2 samples. The occurrence of the
C( O) O Ti4+ group was observed at ␦ ≈ 167 ppm, while the peaks
corresponding to the pyridinic ligand [29] were detected between
160 and 120 ppm.
All the spectroscopic characterizations of the functional-
ized samples clearly showed that the dichloro-dioxo-(4,4 -
dicarboxylato-2,2 -bipyridine) molybdenum(VI) complex was duly
immobilized on the surface of the mesoporous titania supports.
252 H. Martínez et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 248–255

Fig. 5. UV–vis diffuse reflectance spectra of (a) Mo(VI) Cl2 O2 Bipy/TiO2 (P-25) (b)
Mo(VI) Cl2 O2 Bipy/TiO2 (HT-400), (c) Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150) and of the corre-
sponding TiO2 supports (grey lines).
Fig. 7. Cyclohexene photo-conversion over the various Mo(VI) O2 Cl2 Bipy/TiO2 sys-
tems.
Table 2
Cyclohexene photo-oxidation with various fresh and used Mo(VI) Cl2 O2 Bipy/TiO2
catalysts. Yield to cyclohexene oxide and selectivities to cyclohexene oxide, cyclo-
hexenone and 2-cyclohexen-1-ol.

Fresha and usedb catalysts Yield (%)c Selectivity (%)

Mo(VI) Cl2 O2 Bipy/TiO2 P-25 fresh 50.6 70 17 13

Mo(VI) Cl2 O2 Bipy/TiO2 P-25 used 46.2 68 19 13
Mo(VI) Cl2 O2 Bipy/TiO2 HT-400 fresh 51.8 74 15 11
Mo(VI) Cl2 O2 Bipy/TiO2 HT-400 used 41.9 70 18 12
Mo(VI) Cl2 O2 Bipy/TiO2 SC-150 fresh 54.7 83 10 7
Mo(VI) Cl2 O2 Bipy/TiO2 SC-150 used 50.6 79 13 8
a
Measured after 56 h of reaction.
b
Fresh: new catalyst after 4 reaction cycles (56 h).
c
Used: catalyst filtered after 4 reaction cycles (56 h), then employed with a new
reaction solution during 4 reaction cycles (56 h).

Fig. 6. 13 C CP-MAS NMR spectra recorded at 10 kHz of (a) Mo(VI) Cl2 O2 Bipy/TiO2 (P-
25) (b) Mo(VI) Cl2 O2 Bipy/TiO2 (HT-400) and (c) Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150).
3.2. Photo-catalytic oxidation of cyclohexene
Blank photo-oxidation experiments for bare TiO2 under O2
atmosphere and UV irradiation were first performed so as to con-
firm that there was no significant alkene conversion (less than 1%).
Moreover no activity was observed for the anchored catalytic sys-
tems in the dark.
Results presented in Fig. 7 showed the cyclohexene
photo-conversion in the presence of the various dioxo-
molybdenum(VI) anchored-TiO2 catalysts. It appeared that
there was a strong influence of the hybrid Mo(VI) -TiO2 based
catalysts on the cyclohexene conversion at the beginning of
the reaction (300 min), whereas a quasi-linear increase of
conversion with reaction time was observed. All the cata-
lysts were active in cyclohexene oxidation in the following
order: Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150) > Mo(VI) Cl2 O2 Bipy/TiO2 (HT-
400) > Mo(VI) Cl2 O2 Bipy/TiO2 (P-25). The dioxo-Mo(VI) immobilized
on the TiO2 SC-150 sample proved to be the most active with
respect to the other catalysts. Such behavior could be related to
an increased amount of exposed dioxo-Mo(VI) species over the
surface of the mesoporous TiO2 prepared under supercritical CO2
(Table 1).
The results presented in Table 2 indicated that no allylic
products were observed. In each case, the selectivity to the
epoxide product was always higher than 70%, thereby indicat-
ing that the OAT is the predominant process. An effect of the
TiO2 porosity and hence on the dispersion of the dioxo-Mo(VI)
species was also noticed since the selectivity to the epoxide
values were 70, 74 and 83% for Mo(VI) Cl2 O2 Bipy/TiO2 (P-25),
Mo(VI) Cl2 O2 Bipy/TiO2 (HT-400) and Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150),
respectively. Recycling studies were also performed over 56 h of
reaction and they provided evidence of the stability of the catalysts.
Interestingly the yield to cyclohexene oxide and the selectivity val-
ues were quite similar for the fresh and used catalysts based on
the TiO2 (SC-150) and P-25 supports, while a significant drop of
yield was observed for the Mo(VI) O2 Cl2 Bipy/TiO2 (HT-400) system
after 56 h of reaction. Again such data confirmed that the TiO2
synthesized in supercritical CO2 medium exhibited appropriate
characteristics for acting as support for the immobilization of the
dioxo-Mo(VI) complex.
Fig. 8a shows the evolution of the moles of cyclohexene oxide
per mole of dioxo-Mo(VI) complex as a function of irradiation time
for the various Mo(VI) Cl2 O2 Bipy/TiO2 systems. When compared
with the commercial TiO2 P-25 matrix, a significant increase was
obtained with the mesoporous titania supports, in accordance with
their higher surface area that promotes the absorption-desorption
reaction between the substrate and the product. One can also note
the higher ability of the complex immobilized onto the mesoporous
TiO2 -SC support to transfer oxygen atoms to cyclohexene.
The stability of the catalyst based on the mesoporous TiO2
support synthesized under SC medium during the epoxidation of
 H. Martínez et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 248–255
Fig. 8. (a) Evolution of cyclohexene oxide under molecular oxygen atmosphere and
UV–vis light during 35 h with continuous irradiation and (b) evolution of cyclohex-
ene oxide during 4 reaction cycles (56 h) for the fresh and used Mo(VI) Cl2 O2 Bipy/TiO2
SC-150 catalyst.
cyclohexene is presented in Fig. 8b. The Mo(VI) Cl2 O2 Bipy/TiO2 (SC-
150) catalyst was used in a second reaction cycle during long-term
photo-oxidation. Interestingly, the catalytic activity is perfectly
maintained after 56 h of reaction with almost no reduction in the
yield to cyclohexene oxide, which is a strong indication of the sta-
bility of the covalently anchored dioxo molybdenum complex on
the TiO2 elaborated under supercritical CO2 .
However, when the dioxo-Mo(VI) complex was immobilized on
the surface of the titania prepared under hydrothermal conditions
(Mo(VI) Cl2 O2 Bipy/TiO2 (HT-400) system), a significant drop of activ-
ity was observed (not shown). Such behavior was related to a
partial leaching of the Mo(VI) Cl2 O2 Bipy complex as ascertained by
Mo elemental analysis in solution. The amount of Mo leached was
found to be 6.5 wt%, which prompted us to carry out the photo-
oxidation with the supernatant solution under irradiation and O2
atmosphere. The presence of the epoxide was observed in the
supernatant solution of the recovered Mo(VI) Cl2 O2 Bipy/TiO2 (HT-
400) sample, while no photo-oxidation reaction occurred under
these conditions for the analogous Mo(VI) Cl2 O2 Bipy/TiO2 (P-25) and
Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150) catalysts (ESI, Fig. S1). Evidence of
Mo(VI) species in the supernatant solution of the catalyst based on
the TiO2 -HT sample was confirmed by UV–vis analysis. The UV–vis
spectrum of this solution (after solid filtration) showed a strong
absorption band at 560 nm (ESI, Fig. S2), corresponding to the pres-
ence of the Mo = O group [30]. In order to provide some explanations
253
Fig. 9. Cyclohexene oxidation with the different Mo(VI) Cl2 O2 Bipy/TiO2 catalysts.
Time evolution of the mol of epoxide per mol of dioxo-Mo(VI) species in successive
cycles of light under N2 and O2 in darkness.
for such differences of behavior, silylation of both mesoporous TiO2
supports was performed. The aim was to determine the amount
of reactive OH groups [31] able to serve as active sites for the
anchoring of the Mo(VI) complex. The number of reactive OH of the
hydrothermally treated titania support (TiO2 (HT-400)) was found
to be around 1/3 of that of the mesoporous TiO2 (SC-150). Leach-
ing of Mo(VI) species may therefore be due to a relative amount
of complex physisorbed instead of anchored on the mesoporous
TiO2 (HT-400) support.
The photo-oxidation of cyclohexene was also performed under
inert nitrogen atmosphere and UV–vis light irradiation over the
three Mo(VI) Cl2 O2 Bipy/TiO2 systems, to show the real effect of the
anchored dioxo-Mo entity in stoichiometric conditions. As shown
in Fig. 9, Mo(VI) Cl2 O2 Bipy/TiO2 (P-25), Mo(VI) Cl2 O2 Bipy/TiO2 (HT-
400) and Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150) catalysts undergo an
oxygen atom transfer to cyclohexene during the first illumina-
tion cycle under N2 atmosphere. The amount of cyclohexene oxide
produced corresponded to the transfer of one oxygen atom from
the MoO2 unit (stoichiometric amount of the OAT reaction), which
resulted in the formation of the reduced anchored MoO(IV) species,
as depicted in Scheme 1.
Then light was tuned off and the dinitrogen evacuated and
replaced by O2 in order to reoxidize the reduced MoO(IV) species.
During this period no cyclohexene epoxidation was observed since
no OAT occurred. The dioxygen was further evacuated, replaced by
N2 and the light turned back on. After this stage, all Mo(VI) Cl2 O2 Bipy
anchored systems exhibited immediately an oxidation reaction.
The OAT process was twice more important than in the first irra-
diation cycle, thereby indicating that two oxygen atoms per Mo
center had been transferred. After reoxidation with O2 in the dark,
2 mol of cyclohexene oxide were indeed produced during the sec-
ond and third cycles of illumination, which corresponds to 2 oxygen
atoms transferred from the MoO(O2 ) unit [32–34], in agreement
with Scheme 1.
3.3. Photo-oxidation of cyclooctene and 1-octene
Remarkably, we also found that cyclooctene epoxidation over
Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150) under O2 atmosphere and UV–vis
light during 35 h of reaction led to cyclooctene oxide with 100%
yield (Fig. 10). Photo-oxidation of 1-octene with such catalyst
resulted in 98% of selectivity to the corresponding 1-octene oxide,
indicating 100% OAT process. In general, cyclic olefins exhibit a
254 H. Martínez et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 248–255
Scheme 1. Stoichiometric oxygen transfer from the MoO2 (VI) entity to cyclohexene under inert N2 atmosphere and UV–vis irradiation, as well as O2 activation by the
Mo(VI) Cl2 O2 Bipy/TiO2 system under mild conditions.
Fig. 10. Selectivity and conversion values for the photo-oxidation of various alkenes
over Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150) after 35 h of reaction.
higher reactivity than that of linear olefins. As can be observed in
Fig. 10, higher conversion values were indeed obtained in the case
of cyclohexene and cyclooctene with respect to 1-octene.
The evolution of the mol of epoxides (cyclooctene oxide, cyclo-
hexene oxide and 1-octene oxide) per mol of dioxo-Mo(VI) species
as a function of irradiation time is presented in Fig. 11 for the
Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150) catalyst. It can be seen that the
activity of the mesoporous TiO2 -SC support modified with the
dioxomolybdenum(VI) complex was maintained during 35 h in the
photo-epoxidation of the various olefins in the presence of O2 as
O-donor and light.
In order to assess the stability and reusability of the dioxo-
Mo(VI) modified mesoporous TiO2 prepared under SC CO2 during
the photo-oxidation of cyclooctene, a series of six successive runs
was performed. The solid catalyst was separated from the reaction
Fig. 11. Epoxides production after 35 h of reaction under O2 and UV–vis light with
the Mo(VI) Cl2 O2 Bipy/TiO2 (SC-150) catalyst.
medium by filtration after a catalytic run for 6 h, then washed with
acetone and finally dried before being subjected to another cycle of
reaction. Remarkably, the activity of the Mo(VI) Cl2 O2 Bipy/TiO2 (SC-
150) catalyst was preserved after the six cyclooctene epoxidation
cycles during 6 h under O2 atmosphere and light. The yield to
cyclooctane epoxide was perfectly maintained as well as the con-
version values after six reaction cycles (ESI, Table S2).
4. Conclusions
In this paper, it was demonstrated that dichloro-dioxo-[4,4 -
dicarboxylato-2,2 -bipyridine]-Mo(VI) complexes anchored on the
surface of various mesoporous TiO2 were highly active, selec-
tive and stable for the photo-assisted epoxidation of alkenes with
 H. Martínez et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 248–255
molecular oxygen as a primary oxidant and UV–vis light. The activ-
ity may be related to the density of the titania surface hydroxyl
groups that favored the dispersion of the Mo(VI) complex on the
surface of the supports and hence the catalytic properties. In this
respect, the mesoporous TiO2 prepared under supercritical CO2
conditions exhibited the highest content and dispersion of the
grafted Mo(VI) Cl2 O2 Bipy complex. Such mesoporous TiO2 support
was shown to possess the most appropriate textural characteristics
and proved after the dioxo-Mo(VI) immobilization step to be the
most efficient in terms of activity and selectivity to the formation
of various epoxides. Moreover such catalyst was stable under the
reaction conditions as demonstrated by several reuses even after
long-term photo-oxidation experiments.
Acknowledgements
This work was financially supported by the Universidad Indus-
trial de Santander, Colombia (project DIEF 9313 and sabbatical year
of FM), HM and NJC were supported by COLCIENCIAS, Colombia
(Fondo Apoyo a Doctorados Nacionales). FM was financially sup-
ported during sabbatical year by the Region Poitou-Charentes, the
CNRS and the University of Poitiers, France (contract 09/RPC-R-
055). The authors are also very much indebted to Jean-Christophe
Ruiz from CEA Bagnols sur Cèze (France) for fruitful discussions
regarding the TiO2 synthesized under supercritical CO2 .
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.molcata.2016.07.
006.
References
[1] B. Puértolas, A.K. Hill, T. García, B. Solsona, L. Torrente-Murcino, Catal. Today
248 (2015) 115–127.
[2] X.L. Wei, X.-H. Lu, X.-T. Ma, C. Peng, H.-Z. Jiang, D. Zhou, Q.-H. Xia, Catal.
Commun. 61 (2015) 48–52.
[3] Y. Yan, X. Tong, K. Wang, X. Bai, Catal. Commun. 43 (2014) 112–115.
[4] A. Feliczak-Guzik, A. Wawrzynczak, I. Nowak, Microporous Mesoporous
Mater. 202 (2015) 80–89.
255
[5] J.W. Brown, Q.T. Nguyen, T. Otto, N.N. Jarenwattananon, S. Glöggler, L.-S.
Bouchard, Catal. Commun. 59 (2015) 50–54.
[6] K. Shimizu, T. Kaneko, T. Fujishima, T. Kodama, H. Yoshida, Y. Kitayama, App.
Catal. A: Gen. 225 (2002) 185–191.
[7] S. Ouidri, C. Guillard, V. Caps, H. Khalaf, App. Clay Sci. 48 (2010) 431–437.
[8] M.S. Hamdy, G. Mul, Catal. Sci. Technol. 2 (2012) 1894–1900.
[9] N.S. Patil, B.S. Uphade, P. Jana, S.K. Bharagava, V.R. Choudhary, J. Catal. 223
(2004) 236–239.
[10] T. Ohno, K. Nakabeya, M. Matsumara, J. Catal. 176 (1998) 76–81.
[11] S. Huber, M. Cokoja, F.E. Kühn, J. Organomet. Chem. 751 (2014) 25–32.
[12] M. Bagherzadeh, M. Zare, T. Salemnoush, S. Ozkar, S. Akbayrak, Appl. Catal. A
475 (2014) 55–62.
[13] C. Muller, N. Grover, M. Cokoja, F.E. Kühn, Adv. Inorg. Chem. 65 (2013) 33–83.
[14] T.R. Amarante, A.C. Gomez, P. Neves, F.A. Almeida-Paz, A.A. Valente, M.
Pillinger, I.S. Gonçalves, Inorg. Chem. Commun. 32 (2013) 59–63.
[15] S.A. Hauser, M. Cokoja, F.E. Kühn, Catal. Sci. Technol. 3 (2013) 552–561.
[16] Y. Li, X. Zhou, S. Chen, R. Luo, J. Jiang, Z. Liang, H. Ji, RSC Adv. 5 (2015)
30014–30020.
[17] S.M. Islam, K. Ghosh, R.A. Molla, A.S. Roy, N. Salam, M.A. Iqubal, J. Organomet.
Chem. 774 (2014) 61–69.
[18] C.A. Páez, N.J. Castellanos, F. Martinez, F. Ziarelli, G. Agrifoglio, E. Páez-Mozo,
H. Arzoumanian, Catal. Today 133 (2008) 619–624.
[19] C.A. Páez, O. Lozada, N.J. Castellanos, F. Martinez, F. Ziarelli, G. Agrifoglio, E.
Páez-Mozo, H. Arzoumanian, J. Mol. Catal. A 299 (2009) 53–59.
[20] N.J. Castellanos, F. Martinez, F. Lynen, S. Biswas, P. Van Der Voort, H.
Arzoumanian, Trans. Met. Chem. 38 (2012) 119–127.
[21] H. Arzoumanian, N.J. Castellanos, F. Martinez, E.A. Paez-Mozo, F. Ziarelli, Eur. J.
Inorg. Chem. (2010) 1633–1641.
[22] X. Zhang, S. Heinonen, E. Levänen, RSC Adv. 4 (2014) 61137–61152.
[23] T. Peng, D. Zhao, K. Dai, W. Shi, K. Hirao, J. Phys. Chem. B 109 (2005)
4947–4952.
[24] N. Preda, L. Mihut, M. Baibarac, I. Baltog, M. Husanu, C. Bucur, T. Velula, Rom. J.
Phys. 54 (7–8) (2000) 667–675.
[25] K.H. Leong, P. Monash, S. Ibrahim, P. Saravanan, Sol. Energy 101 (2014)
321–332.
[26] G.R. Hearne, J. Zhao, A.M. Dawe, V. Pischedda, M. Maaza, M.K. Nieuwoudt, P.
Kibasomba, Phys. Rev. 70 (2004) 134102–134110.
[27] F. Tian, Y. Zhang, J. Zhang, C. Pan, J. Phys. Chem. C 116 (2012) 7515–7519.
[28] N.L. Alpert, W.E. Keiser, H.A. Szymanski, IR; Theory and Practice of Infrarred
Spectroscopy, Plenum Press, New York, 1970, pp 209, 261 and 268.
[29] V. Amani, A. Abedi, S. Ghabeshi, H. Reza, S.M. Hosseini, N. Safari, Polyhedron
79 (2014) 104–115.
[30] H. Arzoumanian, G. Agrifoglio, M.V. Caparelli, R. Atencio, A. Briceño, A.
Alvarez-Larena, Inorg. Chim. Acta 359 (2006) 81–89.
[31] M.C. Capel-Sanchez, J.M. Campos-Martin, J.L.G. Fierro, Catal. Today 158 (2010)
103–108.
[32] J.M. Berg, R.H. Holm, J. Am. Chem. Soc. 107 (1985) 925–932.
[33] C. Lorber, M.R. Plutino, L.I. Elding, E. Nordlander, J. Chem. Soc. Dalton Trans. 2
(1997) 3997–4003.
[34] J. Cipot-Wechsler, D. Covelli, J.M. Pratorius, N. Hearns, O.V. Zenkina, E.C. Keske,
R. Wang, P. Kennepohl, C.M. Crudden, Organometallics 31 (2012) 7306–7315.