Novel approach to rhenium oxide catalysts for selective oxidation of methanol to DMM

Article abstract of DOI:10.1016/j.jcat.2011.01.028

Novel rhenium oxide catalysts have been prepared by impregnation of mesoporous TiO₂-anatase support using amounts of homometallic, such as Re₄O₄(OEt)₁₂, and heterometallic, such as Ta₄O₂(OEt)₁₄(ReO₄)₂, rhenium alkoxide solutions with subsequent thermal treatment. The structure of the produced catalytic centers was investigated by X-ray diffraction, XPS, and IR spectroscopy in order to examine the role of Re in enhancing catalytic activity. It has been demonstrated that rhenium is mostly present as perrhenate ligands attached via an oxo-bridge to an oxide surface. The catalytic activity was studied for the selective oxidation of methanol as model reaction, revealing selectivity strongly dependent on the nature of the molecular precursor: in case of ReO_x/TiO₂, the catalysts are showing selectivity to formaldehyde, and in case of ReO_x/Ta₂O ₅/TiO₂ to DMM in rich conditions.

Full text of DOI:10.1016/j.jcat.2011.01.028

Journal of Catalysis 279 (2011) 310–318
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Novel approach to rhenium oxide catalysts for selective oxidation of methanol
to DMM
Olesya A. Nikonova a, , Mickaël Capron b,c, Ge Fang b,c, Jérémy Faye b,c, Anne-Sophie Mamede b,c,d,
⇑
b,c
b,c,e
a
Louise Jalowiecki-Duhamel , Franck Dumeignil
, Gulaim A. Seisenbaeva
a
Department of Chemistry, SLU, Box 7015, SE-75007 Uppsala, Sweden
Univ. Lille Nord de France, F-59000 Lille, France
CNRS UMR8181, Unité de Catalyse et Chimie du Solide, UCCS, F-59655 Villeneuve d’Ascq, France
Ecole Nationale Supérieure de Chimie de Lille, ENSCL, F-59652 Villeneuve d’Ascq, France
Institut Universitaire de France, Maison des Universités, 10 Boulevard Saint-Michel, 75005 Paris, France
b
c
d
e
a r t i c l e i n f o
a b s t r a c t
Article history:
2
Novel rhenium oxide catalysts have been prepared by impregnation of mesoporous TiO -anatase support
Received 1 December 2010
Revised 25 January 2011
Accepted 26 January 2011
Available online 2 March 2011
4 4 4 2 4 2
using amounts of homometallic, such as Re O (OEt)12, and heterometallic, such as Ta O (OEt)14(ReO ) ,
rhenium alkoxide solutions with subsequent thermal treatment. The structure of the produced catalytic
centers was investigated by X-ray diffraction, XPS, and IR spectroscopy in order to examine the role of Re
in enhancing catalytic activity. It has been demonstrated that rhenium is mostly present as perrhenate
ligands attached via an oxo-bridge to an oxide surface. The catalytic activity was studied for the selective
oxidation of methanol as model reaction, revealing selectivity strongly dependent on the nature of the
Keywords:
Rhenium and tantalum alkoxides
Catalysts
Methanol Conversion
DMM
molecular precursor: in case of ReO
x 2
/TiO , the catalysts are showing selectivity to formaldehyde, and
in case of ReO /Ta /TiO to DMM in rich conditions.
x
2
O
5
2
Ó 2011 Elsevier Inc. All rights reserved.
1
. Introduction
3 3
well-known methyltrioxorhenium (CH ReO , MTO). This catalyst
is particularly useful in oxidation chemistry and shows high activ-
ity for epoxidations, CAH and SiAH oxidations, and can also be ap-
plied in other types of catalytic reactions such as olefin metathesis
and aldehyde olefinations [8–12].
The unique properties of rhenium oxide catalysts due to a vari-
ety of acidic and redox properties make them suitable materials for
different types of catalytic reactions. For instance, Re has shown
2 7
O
high catalytic activity in olefin metathesis and is the only catalyst
that can work efficiently at room temperature [1]. Moreover, this
heterogeneous catalyst can also be used to treat oleate esters when
activated by organic derivatives of tin. However, their use is detri-
x
Recently, it was shown that ReO -containing catalysts provide
high catalytic activity and selectivity on a number of supports in
certain conditions for partial oxidation of methanol to methylal
(DMM) [13–16]. Numerous studies have shown that methanol oxi-
dation is very sensitive to the nature of active sites and can be used
to study the acidic and oxidation properties of catalytic surfaces
[17–19]. The process of methanol oxidation leads to stepwise for-
mental to the regeneration process [2]. Compared to Re
2 7
O on
common -Al dispersed on -Al has shown higher
c
2
O
3, Re
2
O
7
a
2 3
O
catalytic activity in the metathesis of internal as well as terminal
olefins in the liquid phase [3,4]. Rhenium oxide supported on or-
dered mesoporous alumina has even been claimed as a more active
2
mation of formaldehyde (F), formic acid (FA), and finally CO . If the
catalyst also has acidic properties, then the aforementioned prod-
ucts can react through a dehydration/condensation reaction with
methanol to produce dimethylether (DME), dimethoxymethane
(DMM), and finally methylformate (MF) (see Fig. 1).
Different properties of catalysts make therefore the methanol
oxidation reaction also suitable as a surface probe, as proposed by
[19], for the characterization of structural (crystallographic plane
reactivity) and chemical (acid–base and redox) properties of oxide
catalysts. The most interesting product for industry is DMM, which
possesses a high degree of chemical stability while the absence of
carbon–carbon bonds makes this compound a good candidate as a
fuel additive. Industrially, methylal is produced by two-stage
2 7 2 3
and selective catalyst than Re O /c-Al O for alkene metathesis [5].
Nevertheless, even if this family of catalysts is quite active, they
generally have a problem of application in catalytic selective oxida-
tion reactions due to Re oxide sublimation under reaction condi-
tions [6]. However, it was discovered that the rhenium oxide/
zeolite catalyst was very active in the selective oxidation of ben-
zene to phenol [7]. The most efficient catalyst based on Re is the
⇑
Corresponding author.
E-mail addresses: olesya.nikonova@kemi.slu.se (O.A. Nikonova), gulaim.seisen-
baeva@kemi.slu.se (G.A. Seisenbaeva).
0
021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.01.028

O.A. Nikonova et al. / Journal of Catalysis 279 (2011) 310–318
311
Fig. 1. Scheme of oxidation/dehydration of methanol.
processes: methanol oxidation to formaldehyde on silver and ferric
2.2.2. Catalysts
Preparation of catalysts was carried out in two steps: the former
involving impregnation of the TiO support with a calculated
amount of alkoxide solutions and the latter with thermal treat-
ment of the catalysts at 300 °C in air for 2 h with an initial heating
rate of 5 °C/min. In both steps, we had, as we expected, some losses
molybdate catalysts and dehydrative condensation of the formalde-
hyde with methanol catalyzed by sulfuric acid. Thus, DMM produc-
tion patents are rather scarce because of the difficulty of obtaining a
high yield in DMM in one step. Different catalytic systems have
2
been studied in oxidation of methanol into DMM such as VO
x
ATiO
2
2
4
À
[
20], V
2
O5 À TiO
2
=SO
[21], Mo-supported catalysts [22] and
of the Re presumably as Re
and also of the organic part of the alkoxide complexes.
To obtain 1 wt% (S1) and 2 wt% (S2) of ReO /TiO , the calculated
volumes of Re (OEt)12 alkoxide solution (0.032 mmol and
0.064 mmol) were used to impregnate 2.97 g and 2.94 g of pre-
treated TiO support, respectively, followed by stirring of the mix-
ture until complete infusion of alkoxide solutions into TiO . The
2 7
O , due to the volatility of this oxide,
FeMo-based catalysts [23,24]. Among these catalysts, it has been
shown that ReO -based catalysts provide high activity and selectiv-
x
3
2
ity toward DMM in certain conditions [13,25] e.g. low conversion of
methanol. It should be noted that these catalysts have the disad-
vantage of difficult preparation and poor stability under catalytic
reaction.
4 4
O
2
2
In this work, we report the application of Re homo- and het-
erometallic alkoxide complexes as a new approach for the prep-
aration of Re oxide-based catalysts. The catalysts have been
prepared in one step at room temperature with subsequent cal-
cination at 300 °C in air, resulting in the formation of crystalline
phases based on Re, mainly in the oxidation state + VII, as mono-
metallic oxide product and as a new bimetallic one based on Ta
and Re. The latter bimetallic catalysts show an increased activity
and selectivity in directing oxidation of methanol to DMM due
to a synergetic effect between the rhenium oxide centers and
the tantalum oxide phase.
samples were then dried in vacuum and left overnight to dry in
air at room temperature. The next step of preparation was calcina-
tion of catalysts at 300 °C in air for 2 h with an initial heating rate
5 °C/min. The mass losses were not significant and observed for S1
and S2 as 2 wt%.
The desired amount of Ta
(0.064 mmol and 0.641 mmol) was used accordingly to prepare
1 wt% (S3) and 10 wt% (S4) of ReO /Ta /TiO catalysts. The ap-
4 2 4 2
O (OEt)14(ReO ) alkoxide solution
3
2
O
5
2
plied procedures of impregnation with alkoxide solutions (II)
were the same as described earlier. After thermal treatment of
catalysts, the weight losses were observed as 5 wt% for both
samples.
To provide an experimental reference to the catalytic activity of
2
. Experimental
.1. Synthesis of precursors
Re (OEt)12 (I) has been obtained by the anodic dissolution of
S3 and S4 catalysts, a material with 7 wt% TaO
thesized using Ta (OEt)10 as the starting material. The aforemen-
tioned techniques were used for the preparation of this catalyst
S5). The calculated amount of (III) 1.9 mmol was impregnated
on 4.65 g of pretreated TiO . After the thermal treatment, the
x 2
/TiO (S5) was syn-
2
2
(
4 4
O
rhenium metal in ethanol in an electrochemical cell without subdi-
vision into cathode and anode space, supplied with a reflux con-
denser and water-cooling, as described in [26].
2
weight loss was about 4 wt%, which corresponds only with the re-
lease of the organic part of the precursor (III).
Ta
(OEt)10 with Re
27,28].
Ta (OEt)10 (III) (tantalum (V) ethoxide, 99+% Alfa Aesar) was
used as a precursor for the preparation of the Ta /TiO catalyst
to compare the catalytic activity with the Re catalysts.
4
O
2
(OEt)14(ReO
4
)
2
(II) was obtained by the interaction of
in dry toluene as it has been described in
Ta
[
2
2
O
7
2.3. Catalytic activity
2
Catalytic activity measurements were made using a fixed bed
reactor described elsewhere [29]. Two different conditions in
terms of methanol concentration were studied applying 7% meth-
anol (poor condition) and 40% methanol (rich condition) diluted by
air. After having passed through the catalytic bed, the reaction
2
O
5
2
2
2
.2. Preparation of catalysts
.2.1. Support
products were analyzed by online
l-gas chromatography (SRA)
equipped with Poraplot U and 5 Å molecular sieve columns and
TCD detectors. All the lines carrying methanol or reaction products
were warmed up to 80 °C by heating tapes (Horst) to avoid
condensation.
2
The TiO -anatase support (titanium (IV) oxide catalyst support,
Alfa Aesar) was calcined before use at 250 °C overnight in air and
then transferred to a dessicator to avoid any traces of humidity.

3
12
O.A. Nikonova et al. / Journal of Catalysis 279 (2011) 310–318
Table 1
Results of EDX analysis of the catalysts.
Composition
Catalyst
Ti
Ta
Re
wt.%
at.%
wt.%
at.%
wt.%
at.%
S1
S2
S3
S4
S5
99.01 ± 2.08
98.08 ± 2.05
61.51 ± 1.29
62.57 ± 1.31
85.99 ± 1.81
99.74 ± 2.09
99.51 ± 2.09
85.81 ± 1.80
86.25 ± 1.81
95.86 ± 2.01
0.99 ± 0.03
1.88 ± 0.05
3.47 ± 0.10
9.75 ± 0.27
0.26 ± 0.01
0.49 ± 0.01
1.25 ± 0.04
2.94 ± 0.08
34.99 ± 0.98
28.97 ± 0.81
14.01 ± 0.39
12.94 ± 0.36
10.80 ± 0.30
4.14 ± 0.12
2.4. Catalysts characterization
was used as internal reference. The spectrometer BE scale was
initially calibrated against the Ag 3d5/2 (368.2 eV) level. Pressure
À10
The specific surface areas of the catalysts were measured on a
was in the 10
torr range during the experiments. Simulation
Quanta chrom – Quanta sorb apparatus using the single-point
BET method, for the experimental details see Appendix Table A1.
Solid state FT-IR measurements in KBr pellets were recorded
with a Perkin Elmer FT-IR spectrometer Spectrum-100. To establish
the exact peak positions, the fitting of the peaks was carried out
using Origin software.
of the experimental photo peaks was carried out using CasaXPS
software. Quantification took into account a nonlinear Shirley
[30] background subtraction. The decomposition of the doublet
Re 4f7/2ARe 4f5/2 was performed considering a spin–orbit split-
ting of 2.4 eV.
SEM micrographs of the samples were obtained using a HIT-
ACHI TM-1000 scanning electron microscope equipped with an
Oxford Instruments EDX detector. To obtain reliable statistics in
the analysis, the data for each point were taken as average of 10
single measurements. The mapping across the analyzed surface
3. Results
3.1. SEM
was made by moving over the sample with steps of 8
l
m.
After thermal treatment, all catalyst samples were investigated
by SEM. To confirm the composition of all samples, the EDX anal-
ysis was performed on different parts of the catalysts/TiO2 cylin-
ders. First, the surfaces of the cylinders were investigated (see
Table 1). The results corresponded in average to expected contents
except for S5, where a higher content of Ta compared to the calcu-
lated one was detected on the surface.
To be able to understand this phenomenon, a crossover cut of
catalysts/TiO2 cylinders was made followed up by EDX analysis
(see Fig. 2). In case of S1 and its analog S2, it was found that the
ReOx species were not uniformly distributed in the volume of
The X-ray powder data collection was carried out at room tem-
perature using a Bruker APEX II diffractometer (Mo K
graphite-monochromator).
HRTEM investigations were performed on a JEOL JEM-2100F
transmission electron microscope.
a radiation,
The XPS analyses were performed using a Kratos Analytical
DLD
AXIS Ultra
spectrometer. A monochromated aluminum source
(
Al Ka = 1486.6 eV) was used for excitation. The X-ray beam
diameter was approximately 1 mm. The analyzer was operated
at constant pass energy of 40 eV using an analysis area of
approximately 700
applied to compensate for the charging effects that occurred
l
m  300
l
m. Charge compensation was
TiO cylinders. However, in the case of S3 and S4, better homoge-
neity was observed in the distribution of TaReOx species in the
2
during the analysis. The C 1s (285.0 eV) binding energy (BE)
whole volume of the TiO2 cylinders, while in the case of S5, the
Fig. 2. SEM images: (a) surface of the S2; (b) crossover cut of S4 and (c) crossover cut of S5.
Table 2
IR spectra of catalysts (comparison with precursors Re
4
O
4
(OEt)12 (I) and Ta
4
O
2
(OEt)14 (ReO
4
)
2
(II)).
À1
Vibrations (cm
)
S1
S2
S3
S4
S5
(I)
(II)
d(ReAOATi)
855 w br
874 w br
910 m 922 s br
832 s br
870 m br
911 m br
m
m
(ReAOARe)
(Re@O) (asymmetric
878 w br
920 s
737 w br
589 w br 610 m 666 s br
921 s
908 s 934 m
911 m 936 s 1021 m
806 s 876 m 891 w
478 w 521 m 563 m
À
4
] )
stretching of [ReO
(TaAOARe)
(TaAOATi)
(TaAOATa)
(TaAO)
m
m
m
m
860 m br
857 s br
645 s br
949 w
608 w br 652 m br

O.A. Nikonova et al. / Journal of Catalysis 279 (2011) 310–318
313
4 4
ation. When Re O (OEt)12 is crystallized, it is sensitive to oxidation,
but stable to hydrolysis [26]. The islands of rhenium-rich materials
apparently result from the thermal decomposition of the initially
formed crystals.
The tantalum–rhenium ethoxide (II) is highly soluble and easily
forms glass-like solids on drying, which explains uniform coating
[
27,28]. It should also be mentioned that we did not have any sig-
nificant rhenium loss during preparation for the catalyst S4 accord-
ing to the EDS analysis. The tantalum ethoxide (III) is known to be
sensitive not only to hydrolysis, but also to ether elimination
[
31,32], easily forming dense oxide coatings [33]. The latter hin-
ders the deeper diffusion of this precursor.
3.2. FT-IR
IR spectroscopy was used to determine the chemical groups
present in the catalyst. Since TiO
band between 400 cm and 880 cm , all spectra were normal-
2
has in the IR spectra a big broad
À1
À1
2
ized and the TiO phase signal was subtracted to reveal the peaks
corresponding to loaded catalysts. After the fitting procedure, the
positions of the peaks corresponding to different ReAO and TaAO
vibrations were identified (see Table 2).
À
To be able to distinguish differences between Re
2
O
7
and ReO4
bands, the spectrum of the crystalline rhenium heptoxide (Sig-
ma–Aldrich, rhenium (VII) oxide, P99.9% metals basis), partially
hydrolyzed to perrhenic acid during the preparation of the sample,
À1
À1
was recorded (see Fig. 3c). The bands at 870 cm , 975 cm , and
À1
1
004 cm could be assigned to Re
2
O
7
according to [34], while
À1
À1
À1
À
the bands at 903 cm , 924 cm , and 942 cm belong to ReO4
group in compliance with [35].
À1
À1
In case of S1 and S2, the vibrations at 855 cm and 832 cm
,
respectively, are close to the ReAOAH stretching modes reported
in [35]. Since the samples of the catalysts did not contain OH-
bonds or perrhenic acid, those peaks could be assigned to the Re-
À1
À1
OATi vibrations. Peaks at 870 cm and 874 cm for samples S2
and S1, respectively, should be attributed to the ReAOARe stretch-
À1
ing mode, which is in good correlation with the peak at 878 cm
À1
observed for the starting material Re
for Re (see Fig. 3).
The Re@O stretching modes for the Re
plex are considerably different from those for S1 as well as S2,
4 4
O (OEt)12 (I) and at 870 cm
2 7
O
4 4
O (OEt)12 alkoxide com-
À1
À1
À1
where the vibrations at 910 cm , 922 cm , and 911 cm could
-
4
be assigned to vibrations of the perrhenate group [ReO ] according
to [35] (see Fig. 3). The samples S1 and S2 contain a mixture of two
À
different species, Re
2
O
5
and ReO , according to the IR spectra (see
4
Fig. 3) and XPS results, which can be explained by the properties of
the Re (OEt)12 complex. The alkoxide precursor (I) contains rhe-
4 4
O
nium in the oxidation state + V. On its thermal decomposition, it is
then partially oxidized into Re (VI–VII).
For the catalyst S5, three peaks have been detected correspond-
À1
ing to different types of vibrations: the first one at 645 cm to
TaAOATa and two others to TaAO, while the peaks observed for
the precursor Ta
4
O
2
(OEt)14(ReO
4
)
2
(II) are situated at the range
À1
4
78–563 cm (see Fig. 4). It should be mentioned that the appear-
À1
ance of the peak at 949 cm indicates the presence of suboxides
36]. Almost the same position of peaks of the TaAOATa vibrations
4 4 2 7 4
Fig. 3. IR spectra of (a) S1, (b) Re O (OEt)12 (I), (c) mixture of Re O and HReO .
[
can be found for S3 and S4 (see Fig. 4c and d). As expected, the
vibrations for TaAOARe stretching modes could be observed at
À1
À1
TaO
x
species were deposited closely to the surface with an impreg-
737 cm and 860 cm for S3 and S4, respectively, which are in
good correlation with bands for the staring material Ta (OE-
t)14(ReO (II) (see Fig. 4b). It should be noted that while for S1
nation depth of about 300
vy element contrast).
l
m (see Fig. 2c – bright area due to hea-
4 2
O
4 2
)
The observed difference in the deposition of the material can be
explained in view of the different reactivity of the alkoxide precur-
sors in relation to crystallization and hydrolysis/polycondensation.
The rhenium (V) ethoxide (I) cluster is known to be easily crys-
tallized with higher crystal growth rate compared to that of nucle-
and S2, both ReAOARe and Re@O vibrations were detected, for
S3 and S4 only the last one at 920 cm was observed, at the same
position as for the starting material, which can be explained by the
preservation of the perrhenate unit already present in the precur-
4 2 4 2
sor Ta O (OEt)14(ReO ) .
À1

3
14
O.A. Nikonova et al. / Journal of Catalysis 279 (2011) 310–318
4 2 4 2
Fig. 4. IR spectra of (a) S5, (b) Ta O (OEt)14(ReO ) (II), (c) S3, (d) S4.
The sample S5 showed an important quantity of 14 at.% tanta-
lum at the surface. As expected, the BE corresponding to Ta 4d5/2
exhibited typically at 230.6 eV corresponding to Ta atoms in the
oxidation state + V [[38] and Ref. therein].
Table 3
Results of XPS analysis (unit: at.%).
S1
0.2
99.8
S2
0.1
99.9
S3
2.7
91.6
5.7
S4
8.6
62.2
29.2
S5
Re
Ti
Ta
Concerning two other samples S3 and S4, it seems that the
85.6
14.4
catalysts have the TaReO
equal to 3.3 and 2.1, respectively. The thickness of the TaReO
x
phases at the surface with Ta/Re ratios
layer
x
increased as a higher quantity of each element (i.e. Ta and Re) was
observed, correlating with the decrease in the Ti content analyzed
by XPS (see Table 3). This indicates a better dispersion of the active
3.3. XPS
In order to confirm the aforementioned hypothesis, the outer
2
phase on the surface of the TiO support ((Re + Ta)/Ti = 0.6 and 0.1
atomic layers up to 50 Å were scanned using X-ray photoelectron
spectroscopy. Table 3 presents the results obtained for the series.
XPS analysis exhibits almost the same composition for samples
S1 and S2. The results obtained by XPS and EDS analysis showed a
significant difference in content of the surface of the crystallites
and the bulk samples. It could be confirmed that the external
for S4 and S3, respectively). The oxidation state of Re in these two
samples is mainly + VII with a binding energy of the Re 4f7/2 equal
to 46.3 eV corresponding to ReO4 [37], a minor contribution of
À
Re + VI is clearly evidenced for the sample S3 by a peak at
3
44.4 eV corresponding to ReO . To prove the presence of Re + VI
is not straightforward on the S4 sample. The Ti 2p and O 1s pho-
5
0 Å of the TiO
2
crystallites were poorer in rhenium than the vol-
topeaks occurred in the typical areas for TiO
530.2 eV, respectively.
2
at 459.0 eV and
ume studied by EDS. The XPS spectra show that the rhenium atoms
are present in oxidation state + V or +VI for outer shells of these
two samples. The principal Re 4f7/2 component, located at
3.4. XPD and HRTEM
4
4.1 eV for S1 and 43.7 eV for S2 respectively, could be referred
to Re or ReO according to [37]. The slight shift could be ex-
2
O
5
3
2
All samples were represented mainly by anatase form of TiO in
plained by the lower content of Re in the S2 sample. The peak posi-
tions for the Ti 2p and O 1s were at 458.8 and 530.1 eV,
respectively.
the X-ray powder patterns. Only some extra peaks identified in
Fig. 5 could be referred to catalytic phases in the samples (see
Fig. 5).

O.A. Nikonova et al. / Journal of Catalysis 279 (2011) 310–318
315
Fig. 6. HRTEM images of S1 and S4 samples.
4
0% of methanol, respectively, diluted by air with an approximate
flow of 40 ml/min. It was found that the conversions increased
with the reaction temperature for each sample. Table 4 shows
the catalytic results for all catalysts at the higher and lower meth-
anol conversion. In each case, the quantity of oxygen was enough
and there were no limitations in performing oxidation reaction
caused by low oxygen concentration.
Fig. 5. Results of XPD of (a) S1 and S2, (b) S3 and S4, (c) S5.
Furthermore, the XPD results have indicated that the grain size
of crystallites of the catalysts decreased with the loading of cata-
lysts onto the support. From the TEM images, it can be seen that
particles are present as agglomerates. Fig. 6 displays representative
TEM images for S1 and S4 samples. The SAED image analysis shows
polycrystallinity of the samples, which made it difficult to deter-
mine the structure of the catalyst phase (see Fig. 6).
It could be seen that the activities for the two samples, S1 and
S2, were comparable revealing methanol conversion of 14% and
1
1% in the poor condition and 6.5% and 4% in the rich condition,
respectively. In this case, the selectivity for both catalysts was
mainly toward F at rich conditions (see Table 4). It should also be
noted that the CO molecules were present in a non-negligible
x
quantity especially in the poor methanol condition (see Table 4).
On the contrary, when the quantity of methanol (rich condition)
was increased, these molecules almost disappeared (see Fig. 7).
For S3 and S4 samples, the main product at each temperature
was DMM (see Fig. 8) in rich condition. It is worth noting that the
3
.5. Catalytic activity
The methanol oxidation reaction on all the catalysts was per-
formed at 275 °C under poor and rich conditions with 7% and

3
16
O.A. Nikonova et al. / Journal of Catalysis 279 (2011) 310–318
Table 4
Catalytic results.
Sample
MeOH vol (%)
7
Max MeOH conversion in (%)
MeOH conversion (mol)
S(F)
S(FM)
17
S(DME)
S(DMM)
S(CO)
13
S(CO
1
2
)
TiO
S1
S2
S3
S4
2
6
4
0.42
1.6
41
49
42
44
4
4
4
4
4
0
7
20
17
52
7
0
14
6.5
0.98
2.6
41
43
17
11
28
26
7
0
11
4
0.77
1.6
39
37
15
13
30
30
15
2
1
1
7
0
37
15
2.59
6
40
15
17
17
15
12
26
3
2
1
7
0
43
23
3.01
9.2
39
6
19
13
15
11
1
67
21
1
5
1
This knowledge permitted to evaluate the redox and acidic
properties of S3 and S4. To be able to understand different proper-
ties of ReO /TiO and ReTaO /TiO catalysts, a Ta O /TiO catalyst
x 2 x 2 2 5 2
was obtained. The Ta atoms in S5 effectively add acidic property
to the support (main product is DME) without changing drastically
the methanol conversion, which did not exceed 6% in rich metha-
nol condition (see Fig. 8c).
4
. Discussion
SEM investigation demonstrated that the distribution of the cat-
alysts occurred in different ways. In fact, for S1 and S2, the distri-
bution of the precursor of ReO species, the Re (OEt)12 alkoxide,
resulted in a difficulty in spreading them evenly in the whole vol-
ume of TiO cylinders. On impregnation of TiO cylinders, the spe-
cific behavior of the precursor caused, in case of S5, the formation
of a dense TaO layer with a thickness of about 300 m. However,
in the case of S3 and S4, another trend resulting in better distribu-
tion through the volume of TiO cylinders was observed. These
x
4 4
O
2
2
x
l
2
phenomena could be explained by the nature and properties of
the starting materials as described earlier.
The BET results did not show significant increases in the specific
surface areas of the catalysts (the average value for all samples was
2
2
1
5 m /g) in comparison with the support (14 m /g), but at the
same time, it did not decrease the observable surface area for the
TiO . This can be explained by the formation of a relatively thin
2
layer of catalysts on the surface of the substrate hindering closing
or filling of the original mesopores, but not creating considerable
volumes of an additional porous phase.
The FT-IR results demonstrated that all samples contained the
À
perrhenate group [ReO
the catalytic O
4
] . According to IR spectra investigations,
3
2
ReAO group was connected to the TiO support
through the oxo-bridge in the case of S1 and S2 samples. In the case
of S3 and S4, the perrhenate group was connected through oxo-
2 5 2
bridges to Ta atoms. Since in the Ta O /TiO catalyst, the Ta atoms
were connected to the support through oxo-bridges, it is plausible
to suggest that in the catalysts S3 and S4, the same type of connec-
Fig. 7. Selectivity of S1 (a) and S2 (b) function of temperature in rich methanol
condition.
x
tion is present. This would mean that the ReO species were in turn
situated on the surface of the coating, which would explain the
selectivity of these catalysts toward DMM caused by the presence
of more acidic TaO species compared to the TiO [39].
x 2
According to XPS data for S1 and S2, the Re atoms were present
mainly in the oxidation state + V or +VI, while for S3 and S4
samples, they were in the oxidation state + VII. Moreover, it should
selectivity increased with temperature up to 250 °C and then
decreased. In order to verify the catalyst survival at the end of
the experiment, the decrease in the temperature to 200 °C
demonstrated an increase in DMM selectivity. The other detected
x
products were MF, DME, and F, with almost no trace of CO as evi-
denced in rich condition, while the opposite tendency was ob-
served in poor condition (see Table 4). It should also be noted
that in case of 7% methanol conversions, the selectivity of S3 and
S4 was mainly toward formaldehyde and no DMM formation
occurred.
x
be mentioned that ReO species were reduced onto the surface of
the support. These phenomena resulted in the appearance of extra
peaks corresponding to rhenium in oxidation state + VI in the S3
sample. Thus, the reduced rhenium oxide species help in keeping
a good activity and selectivity of our catalysts.

O.A. Nikonova et al. / Journal of Catalysis 279 (2011) 310–318
317
one needed to process to the dehydration/condensation reaction
producing DMM. Nevertheless, the presence of MF seems to be
proof that the redox/acidic properties were revealed simulta-
neously in these catalysts. Increase in the quantity of methanol
in the feed, resulting in the increase in probability of higher meth-
anol adsorption on the surface of the catalysts, favors condensation
of formaldehyde with methanol producing DMM. Table 4 shows
that DMM selectivity reaches more than 50% for both S3 and S4
samples. At the same time, the selectivity toward F decreases dras-
tically, which seems to be in agreement with our previous assump-
tions. The mixed oxidation numbers + VI and +VII for rhenium in
the S3 and S4 samples increased drastically the DMM selectivity
under rich methanol condition. It seems that the assumption
claimed by Iwasawa et al. [15] is true but not sufficient. The fact
that the two oxidation states are present in the catalysts permits
the formation of formaldehyde, but then the adsorption energy
should be strong enough to keep the methoxy group at the surface
of the catalyst during its reaction with two methanol molecules,
which leads to the formation of DMM.
5
. Conclusions
The application of alkoxide molecular precursors permits the
creation of new rhenium and chemically connected rhenium–tan-
talum catalysts on mesoporous TiO supports. The materials were
2
obtained in ambient atmosphere with subsequent calcination at a
relatively low temperature of 300 °C for 2 h leading to crystalline
products. The results of FT-IR and XPS investigations showed that
Re is present mainly in the oxidation state + VII as a perrhenate
group for samples S3 and S4. The reduced rhenium oxide sites of
the samples help to keep the activity during and at the end of
the catalytic reaction of the produced catalysts. The small values
of the BET surface areas do not influence the catalytic properties
of the samples. The selectivity of samples S1 and S2 was mostly to-
ward formaldehyde, but in the case of S3 and S4, the product of the
methanol oxidation was DMM. The simultaneous presence of the
chemically connected oxidative (perrhenate) and acidic (tantalum
oxide) components in the catalyst apparently permit one-step pro-
duction of DMM.
Acknowledgments
The authors express their gratitude to Prof. Vadim Kessler for
fruitful discussions. We are very grateful to the Faculty of Natural
Resources SLU for the support in purchasing the Hitachi TM-1000
scanning electron microscope. As well, we are grateful for possibil-
ities to use TEM at Berzelii center EXSELENT on Porous Materials at
Department of Materials and Environmental Chemistry at Stock-
holm University. Authors want to thank M. Sundberg, C.-W. Tai,
A. Beaurain, M. Trentesaux, and G. Tesquet for their technical help.
Appendix A. Supplementary material
Fig. 8. Selectivity of S3 (a), S4 (b), and S5 (c) function of temperature in rich
methanol condition.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jcat.2011.01.028.
In terms of selectivity, the Re oxide species supported on TiO
2
supports (S1 and S2) appeared to have a good balance between re-
dox and acidic properties in order to form DMM and MF, which
were not present at all or were formed in very small quantities
References
[1] J.C. Mol, Catal. Today 51 (1999) 289.
[
[
2] J.C. Mol, Top. Catal. 27 (2004) 97.
3] P.C. Bakala, E. Briot, Y. Millot, J.-Y. Piquemal, J.-M. Bregeault, J. Catal. 258 (2008)
2
for the TiO support alone. According to Table 4, the methanol con-
version decreased with the rhenium loading in poor and rich meth-
anol conditions for S3 and S4. In terms of selectivity, in the poor
methanol condition, the main product was formaldehyde (SF
61.
[4] M. Onaka, T. Oikawa, Chem. Lett. 31 (2002) 850.
[
[
5] T. Oikawa, T. Oosohi, T. Tanaka, T. Yamamoto, M. Onaka, Micropor. Mesopor.
Mater. 74 (2004) 93.
6] S. Albonetti, F. Cavani, F. Trifiro, Catal. Rev. Sci. Eng. 38 (1996) 413.
ꢀ
40%), meaning that the redox properties were dominant under
this condition or that the adsorption energy was lower than the
[7] T. Kusakari, T. Sasaki, Y. Iwasawa, Chem. Commun. (2004) 992.

3
18
O.A. Nikonova et al. / Journal of Catalysis 279 (2011) 310–318
[
8] R. Buffon, A. Auroux, F. Lefebvre, M. Leconte, A. Choplin, J.-M.A. Basset, J. Mol.
Catal. 76 (1992) 287.
9] K.R. Jain, F.E. Kühn, J. Organom. Chem. 692 (2007) 5532.
[23] J. Gornay, X. Secordel, M. Capron, G. Tesquet, P. Fongarland, E. Payen
J.-L. Dubois, F. Dumeignil, Oil and Gas Sci. Tech. – Rev. IFP Energies Novellas
65 (2010) 751.
[
[
[
10] A. Salameh, A. Baudoin, S. Daravong, V. Boehm, M. Roeper, J.-M. Basset
C. Coperet, J. Catal. 253 (2008) 180.
11] A. Salameh, J. Joubert, A. Baudouin, W. Lukens, F. Delbecq, P. Sautet, J.M. Basset,
C. Coperet, Angew. Chem. Int. Ed. 18 (2007) 3870.
[24] J. Gornay, X. Sécordel, B. de Ménorval, S. Cristol, P. Fongarland, M. Capron
L. Duhamel, E. Payen, J.-L. Dubois, F. Dumeignil, Green Chem. 12 (2010) 1722.
[25] X. Secordel, E. Berrier, M. Capron, S. Cristol, J.-F. Paul, M. Fournier, E. Payen,
Catal. Today 155 (2010) 177.
[
[
[
12] R. Buffon, A. Choplin, M. Leconte, J.-M. Basset, J. Mol. Catal. 72 (1992) L7.
13] Y. Yuan, T. Shido, Y. Iwasawa, Chem. Commun. 15 (2000) 1421.
14] A.S.Y. Chang, W. Chen, H. Wang, J.E. Rowe, T.E. Madey, J. Phys. Chem. B 108
[26] O.A. Nikonova, K. Jansson, V.G. Kessler, M. Sundberg, A.I. Baranov
A.V. Shevelkov, D.V. Drobot, G.A. Seisenbaeva, Inorg. Chem. 47 (2008) 1295.
[27] O.A. Nikonova, V.G. Kessler, D.V. Drobot, P.A. Shcheglov, G.A. Seisenbaeva,
Polyhedron 26 (2007) 862.
(
2004) 14643.
[
[
[
[
[
[
15] Y. Yuan, H. Liu, H. Imoto, T. Shido, Y. Iwasawa, J. Catal. 195 (2000) 51.
16] Y. Yuan, Y. Iwasawa, J. Phys. Chem. B 106 (2002) 4441.
17] L.E. Briand, W.E. Farneth, I.E. Wachs, Catal. Today 62 (2000) 219.
18] Y. Matsuoka, M. Nywa, Y. Murakami, J. Phys. Chem. 94 (1990) 1477.
19] J.-M. Tatibouët, Appl. Catal. A: Gen. 148 (1997) 213.
[28] O.A. Nikonova, G.A. Seisenbaeva, V.G. Kessler, P.A. Shcheglov, D.V. Drobot,
Russ. J. Inorg. Chem. 52 (2007) 1687.
[29] S. Royer, X. Sécordel, M. Brandhorst, F. Dumeignil, S. Cristol, C. Dujardin
M. Capron, E. Payen, J.-L. Dubois, Chem. Commun. 7 (2007) 865.
[30] D.A. Shirley, Phys. Rev. B 5 (1972) 4709.
[31] D.C. Bradley, B.N. Chakravarti, W.J. Wardlaw, J. Chem. Soc. (1956) 4439.
[32] V.G. Kessler, N.Y. Turova, A.I. Yanovsky, A.I. Belokon’, Y.T. Struchkov, Russ.
J. Inorg. Chem. 36 (1991) 1662.
[33] P. Werndrup, M. Verdenelli, F. Chassagneux, S. Parola, V.G. Kessler, J. Mater.
Chem. 14 (2004) 344.
20] H. Guo, D. Li, D. Jiang, W. Li, Y. Su, Catal. Commun. 11 (2010) 396;
Y. Fu, J. Shen, Chem. Commun. 21 (2007) 2172;
J. Liu, Y. Fu, Q. Sun, J. Shen, Micropor. Mesopor. Mater. 116 (2008) 614;
J. Liu, Q. Sun, Y. Fu, J. Shen, J. Colloid Interface Sci. 335 (2009) 216.
21] H. Zhao, S. Bennici, J. Shen, A. Auroux, J. Catal. 272 (2010) 176.
22] M. Brandhorst, S. Cristol, M. Capron, C. Dujardin, H. Vezil, G. Lebourdon
E. Payen, Catal. Today 113 (2006) 34;
[
[
[34] I.R. Beattie, T.R. Gilson, P.J. Jones, Inorg. Chem. 35 (1996) 1301.
[35] M.R. Mohammed, W.F. Sherman, J. Phys. C: Solid State Phys. 14 (1981) 4121.
[36] N. Dharmaraj, H.C. Park, C.H. Kim, P. Viswanathamurthi, H.Y. Kim, Mater. Res.
Bull. 41 (2006) 612.
M. Fournier, A. Aouissi, C. Rocchiccioli-Deltcheff, J. Chem. Soc. Chem. Commun.
3
(1994) 307;
C. Rocchiccioli-Deltcheff, A. Aouissi, M.M. Bettahar, S. Launay, M. Fournier,
J. Catal. 164 (1996) 16;
[37] A. Naor, N. Eliaz, L. Burstein, E. Gileadi, Electrochem. Solid-State Lett. 13 (2010)
D91.
C. Rocchiccioli-Deltcheff, A. Aouissi, S. Launay, M. Fournier, J. Mol. Catal.
A: Chem. 114 (1–3) (1996) 331;
H. Liu, E. Iglesia, J. Phys. Chem. B 107 (39) (2003) 10840;
H. Liu, E. Iglesia, J. Catal. 223 (1) (2004) 161.
[38] S.F. Ho, S. Contarini, J.W. Rabalais, J. Phys. Chem. 91 (1987) 4779.
[39] Y. Chen, J.L.G. Fierro, T. Tanaka, I.E. Wachs, J. Phys. Chem. B 107 (2003)
5243.