Selective oxidation of methanol to form dimethoxymethane and methyl formate over a monolayer V2O5/TiO2 catalyst

Article abstract of DOI:10.1016/j.jcat.2013.10.026

The oxidation of methanol over highly dispersed vanadia supported on TiO₂ (anatase) has been investigated using in situ Fourier transform infrared spectroscopy (FTIR), near ambient pressure X-ray photoelectron spectroscopy (NAP XPS), X-ray absorption near-edge structure (XANES), and a temperature-programmed reaction technique. The data were complemented by kinetic measurements collected in a flow reactor. It was found that dimethoxymethane competes with methyl formate at low temperatures, while the production of formaldehyde is greatly inhibited. Under the reaction conditions, the FTIR spectra show the presence of non-dissociatively adsorbed molecules of methanol, in addition to adsorbed methoxy, dioxymethylene, and formate species. According to the NAP XPS and XANES data, the reaction involves a reversible reduction of V⁵⁺ cations, indicating that the vanadia lattice oxygen participates in the oxidation of methanol via the classical Mars-van Krevelen mechanism. A detailed mechanism for the oxidation of methanol on vanadia catalysts is discussed.

Full text of DOI:10.1016/j.jcat.2013.10.026

Journal of Catalysis 311 (2014) 59–70
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Selective oxidation of methanol to form dimethoxymethane and methyl
formate over a monolayer V2O5/TiO2 catalyst
a
a
a
c
d
V.V. Kaichev a,b, , G.Ya. Popova , Yu.A. Chesalov , A.A. Saraev , D.Y. Zemlyanov , S.A. Beloshapkin ,
⇑
A. Knop-Gericke e, R. Schlögl e, T.V. Andrushkevich a, V.I. Bukhtiyarov a,b
a Boreskov Institute of Catalysis of SB RAS, Lavrentieva Ave. 5, 630090 Novosibirsk, Russia
b Novosibirsk State University, Pirogova St. 2, 630090 Novosibirsk, Russia
c Birck Nanotechnology Center, Purdue University, 1205 West State Street, West Lafayette, IN 47907-2100, USA
d Materials and Surface Science Institute, University of Limerick, Limerick, Ireland
e Fritz-Haber-Institute of MPG, Faradayweg 4-6, D-14195 Berlin, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
The oxidation of methanol over highly dispersed vanadia supported on TiO2 (anatase) has been
investigated using in situ Fourier transform infrared spectroscopy (FTIR), near ambient pressure X-ray
photoelectron spectroscopy (NAP XPS), X-ray absorption near-edge structure (XANES), and a tempera-
ture-programmed reaction technique. The data were complemented by kinetic measurements collected
in a flow reactor. It was found that dimethoxymethane competes with methyl formate at low tempera-
tures, while the production of formaldehyde is greatly inhibited. Under the reaction conditions, the FTIR
spectra show the presence of non-dissociatively adsorbed molecules of methanol, in addition to adsorbed
methoxy, dioxymethylene, and formate species. According to the NAP XPS and XANES data, the reaction
involves a reversible reduction of V5+ cations, indicating that the vanadia lattice oxygen participates in
the oxidation of methanol via the classical Mars–van Krevelen mechanism. A detailed mechanism for
the oxidation of methanol on vanadia catalysts is discussed.
Received 16 July 2013
Revised 30 October 2013
Accepted 31 October 2013
Keywords:
Methanol oxidation
Vanadium pentoxide
Monolayer catalyst
Near ambient pressure X-ray photoelectron
spectroscopy
Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction
mainly used to manufacture formamide, dimethylformamide, and
formic acid. Dimethoxymethane is primarily used as a solvent and
TheoxidationofmethanoloverV2O5/TiO2 catalystsis bothfunda-
mentally and practically interesting. This reaction is a relatively sim-
pleand, nevertheless, maygenerateawiderangeofproducts, suchas
formaldehyde, dimethyl ether, methyl formate, dimethoxymethane,
and formic acid. The selectivity depends on the catalyst, reaction
temperature, conversion, and partial pressures of the reactants
[1–4]. For instance, unsupported and silica-supported crystallites
of vanadium pentoxide demonstrate high selectivity toward formal-
dehyde with yields of 96–98% at the temperatures near 300–400 °C
[5–8], whereas monolayer V2O5/TiO2 catalysts exhibit high activity
and selectivity toward methyl formate and dimethoxymethane at
low temperatures (100–200 °C), with suppressedformation of form-
aldehyde [9–22]. All of the above-mentioned reaction products have
practicalsignificance. Theproductionofformaldehyde viatheoxida-
tion of methanol is one of the most important industrial catalytic
processes. Additionally, vanadia-based catalysts may be used to pro-
duce methyl formate and dimethoxymethane (methylal), both of
which are important and versatile chemicals. Methyl formate is
for the production of perfumes, resins, adhesives, paint strippers,
and protective coatings. Dimethoxymethane is an environmentally
friendly material with extremely low toxicity and might also be used
as a H2 storage material for compact hydrogen generators.
Currently, the industrial production of methyl formate occurs
via the carbonylation of methanol using carbon monoxide over so-
dium methoxide catalysts in the liquid phase. Due to the catalyst’s
extreme sensitivity toward water, dry CO must be used in this pro-
cess. Dimethoxymethane is currently produced via a two-stage
process involving (1) the oxidation of methanol to formaldehyde
over silver or iron-molybdate catalysts followed by (2) condensa-
tion reactions in a methanol–formaldehyde mixture using sulfuric
acid or solid acidic catalysts. The development of efficient one-step
catalytic methanol oxidations in the gas phase to selectively form
methyl formate or dimethoxymethane has practical applications.
The V2O5/TiO2 catalyst is
processes.
a promising candidate for these
The V2O5/TiO2 system has been actively studied over the last
thirty years. It has been found that vanadia can form different
structures on titania surfaces depending on the vanadia content
and preparation techniques [23–30]. With a vanadia content under
10% of monolayer (ML), only an isolated, monomeric structure with
⇑
Corresponding author at: Boreskov Institute of Catalysis of SB RAS, Lavrentieva
Ave. 5, 630090 Novosibirsk, Russia. Fax: +7 383 330 80 56.
E-mail address: vvk@catalysis.ru (V.V. Kaichev).
0021-9517/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jcat.2013.10.026

60
V.V. Kaichev et al. / Journal of Catalysis 311 (2014) 59–70
a tetrahedral coordination exists under dehydrated conditions
[25,29]. Polymeric structures such as chains and ribbons of VOx
units with an octahedral coordination appear at the vanadia con-
centration above 20% of monolayer [29]. These structures may
change under the reaction conditions via reduction/oxidation or
hydration/dehydration processes; however, these changes are con-
fined to the surface, and the migration of cations to or from the
sub-surface region does not occur at ambient temperatures [25].
The incorporation of V4+ cations into the crystal lattice of TiO2
has been observed at temperatures above 600 °C during the ana-
tase-to-rutile transformation of V2O5/TiO2 catalysts [31]. The
monolayer coverage of polymerized vanadia species on different
oxide supports was measured by Raman spectroscopy and was
found to be approximately 7–8 vanadium atom/nm2 for TiO2
[27]. When the vanadium content exceeds what is necessary for
the ideal monolayer, V2O5 crystallites are favorable. The surface
vanadia species demonstrate higher catalytic activity and selectiv-
ity in many catalytic reactions compared with the V2O5 crystallites
[32,33]. The unique properties of the surface vanadia species are
related to the strong vanadia–support interaction.
feed consisted of methanol, oxygen, and helium in the 1:1:23 M
ratio. The concentrations of reactants and products were deter-
mined with an on-line gas chromatograph (GC) equipped with
thermal conductivity and flame ionization detectors. Methanol,
dimethoxymethane, methyl formate, formaldehyde, formic acid,
water, and CO2 were analyzed with a Porapak T column, while
the CO, oxygen, and nitrogen contents were analyzed with a NaA
molecular sieve column. All gas lines from the reactor to the sam-
pling valve were maintained at 120 °C to prevent the condensation
of the reactants and products. The selectivity, Si, was calculated as
the molar concentration of the carbon appurtenant, ci, of the
product i divided by the sum of the concentrations of all products,
Si = ci/Rcj  100%. The carbon balance was at least 97 2% for all
GC measurements.
2.3. Catalyst characterization
The catalysts were characterized by XRD, N2 adsorption, ele-
mental analysis, and FTIR spectroscopy. Powder XRD measure-
ments were performed with
a Siemens D500 diffractometer
This paper reports the catalytic properties of highly dispersed
vanadia supported on TiO2 (anatase) in the selective gas-phase
methanol oxidation. The selectivity and reaction rates depended
on the reaction temperature. Dimethoxymethane is the main reac-
tion product below 120 °C with selectivities up to 95%. An increase
in the selectivity toward methyl formate was observed above
120 °C. At 140–150 °C, methyl formate became the main reaction
product with a selectivity of 80–85%. To provide further insight
into the reaction mechanism, near ambient pressure X-ray photo-
electron spectroscopy (NAP XPS), X-ray absorption near-edge
structure (XANES), Fourier transform infrared spectroscopy (FTIR),
X-ray diffraction (XRD), and temperature-programmed reaction
(TPR) were used. The results of these experiments led to the con-
clusion that the lattice oxygen is involved in the oxidation of meth-
anol via the classical Mars–van Krevelen mechanism.
using monochromatic CuK radiation. The 2h scan covered a range
a
of 5–70°. The specific surface area was calculated with the Bru-
nauer–Emmett–Teller (BET) method using nitrogen adsorption iso-
therms measured at liquid nitrogen temperatures with an
automatic Micromeritics ASAP 2400 sorptometer. The elemental
analysis was performed with a Baird ICP atomic emission spec-
trometer. The IR spectra were recorded using a BOMEM MB-102
FTIR spectrometer. The IR samples were pellets composed of the
catalysts or the TiO2 powder pressed with CsI (2 mg of the sample
and 500 mg of CsI).
2.4. In situ characterization
The chemical states of the catalysts were examined in situ using
NAP XPS and XANES techniques. The experiments were performed
at the synchrotron facility, BESSY II, Berlin, using the ISISS (Innova-
tive Station for In Situ Spectroscopy) beamline. The experimental
system was described in detail elsewhere [35]. The key feature of
this system is the capability for in situ measurements of photo-
emission spectra under mbar pressures. The high brilliance of the
synchrotron radiation combined with a short travel length of the
photoelectrons through a high-pressure zone in the gas cell al-
lowed us to obtain high-quality spectra at pressures up to 10 mbar
under flow conditions. Powder samples were pressed into thin self-
supporting pellets and mounted on a sapphire holder between the
SiC and stainless steel plates. The heating was performed from the
rear with a near-infrared semiconductor laser (k = 808 nm). The
sample temperature was measured with a K-type thermocouple
pressed directly against the rear of the sample. The flows of meth-
anol and oxygen through the gas cell were regulated separately
with calibrated mass-flow controllers. The total flow rate was
approximately 3 sccm. During the experiments, the total pressure
in the gas cell was constant 0.25 mbar. The XANES spectra were col-
lectedin the total electronyield mode. The V 2p3/2, Ti 2p3/2, C 1s, and O
1s core-level spectra were measured with a photon energy of 730 eV.
The intensity of the core-level spectra was normalized to the ring
current. The charge correction was performed by setting the Ti 2p3/
2 at 459.0 eV. FitXPS software was used for curve fitting. The core-le-
vel spectra were curve fitted after a Shirley-type background sub-
traction assuming a Lorentzian/Gaussian line shape.
2. Experimental
2.1. Catalyst preparation
A two-step procedure was used to prepare the monolayer
V2O5/TiO2 catalyst. The TiO2 support (anatase, 350 m2/g) was
impregnated with an aqueous solution of vanadyl oxalate via the
incipient-wetness impregnation method and was subsequently
dried at 110 °C for 12 h; the final calcination occurred in a flow
of air for 4 h at 400 °C. This catalyst was referred to as ‘‘fresh’’.
The fresh catalyst (20 wt% V2O5 and 80 wt% TiO2) was subse-
quently treated in a 10% aqueous solution of nitric acid at room
temperature. After the washing process, the catalyst (12.5 wt%
V2O5 and 87.5 wt% TiO2) was calcined in a flow of air for 4 h at
400 °C. This catalyst was referred to as ‘‘washed’’. The morphology
of this catalyst corresponds to the structure of a so-called mono-
layer catalyst.
2.2. Catalytic tests
The steady-state activity of each catalyst was tested at atmo-
spheric pressure in a differential reactor with a flow-circulating
configuration [34]. The reactor was constructed from Pyrex glass
tubing with a 12 mm inner diameter and a 50 mm length. A coaxial
thermocouple pocket with a 4 mm outer diameter was fitted in the
catalyst bed to control the temperature. The reactor was placed in-
side of an oven. The temperature was controlled within 0.5 °C
with a K-type thermocouple. A fraction of the catalyst powder con-
taining grains in the size range of 0.25–0.50 mm was used. The
For TPR, a sample was heated at 10 K/min from 50 to 200 °C in
the reaction mixture. The gas-phase products were analyzed using
an on-line quadruple mass-spectrometer (Prizma QMS-200, Bal-
zers) connected directly to the gas cell through a leak valve.
To identify the reaction intermediates involved in the oxidation
of methanol, FTIR spectra were obtained in situ with the same

V.V. Kaichev et al. / Journal of Catalysis 311 (2014) 59–70
61
BOMEM MB-102 spectrometer. The spectrometer was operated in
the transmission mode using a specially designed quartz cell-reac-
tor with CaF2 windows. The cell volume was 1.5 cm3. The catalyst
was pressed into a thin self-supporting pellet (ꢀ15 mg/cm2) and
placed into the cell-reactor. The FTIR experiments were performed
at atmospheric pressure using a feed of 2 vol% CH3OH in an air flow
at 50 sccm. The spectra were recorded within 4000–1100 cmÀ1 at a
resolution of 4 cmÀ1 at 50–180 °C. Before exposure to the reaction
mixture, the sample was treated in a flow of air for 60 min at
300 °C. Subsequently, the cell-reactor and the catalyst sample were
cooled to the desired temperature, and the air flow was replaced by
the methanol/air mixture flow.
deformation of the V–O–V bridges, respectively. The broad band
at approximately 600 cmÀ1 was assigned to the bending vibrations
[39,40]. The peaks at 1015 and 856 cmÀ1 in the FTIR spectrum of
the fresh catalyst (Fig. 2a) confirmed the presence of the V2O5 crys-
talline phase. However, based on a comparison of the FTIR spectra
of the fresh and washed catalysts, the band at 946 cmÀ1 should be
assigned to another species. Following the treatment in the nitric
acid solution, the intensities of the bands at 1015 and 856 cmÀ1
substantially decreased, whereas the band at 946 cmÀ1 did not
change (Fig. 2a). Therefore, this band might be assigned to the
V–O–Ti vibration of the surface vanadate species, which is consis-
tent with previous reports [36–38]. Note that VOx species becomes
hydrated at these conditions, and this effect leads to the significant
red shift of the V@O stretching vibration. A spectrum of the washed
catalyst dehydrated at 200 °C is shown in Fig. 2b. The spectrum
3. Results
exhibits the characteristic V@O vibration at ꢀ1030 cmÀ1
.
Hence, the washed catalyst was covered with approximately
1 ML of the highly dispersed vanadia species. Typically, this cata-
lyst is referred to as a monolayer catalyst [25]. According to the Ra-
man spectroscopy data [41], the polymeric vanadia was the
dominant species; the amount of isolated monomeric and dimeric
vanadia species was negligible.
3.1. Catalyst characterization
Fig. 1 presents the XRD patterns of the fresh and washed V2O5/
TiO2 catalysts and of the TiO2 support. All peaks observed in the
TiO2 XRD pattern can be assigned to a pure tetragonal anatase
phase (JCPDS File, No. 21-1272). XRD signatures of both anatase
and V2O5 were observed from the fresh V2O5/TiO2 catalyst. The
vanadia coverage was estimated from the specific surface area;
the vanadia content is approximately 1.5 ML (Table 1). The washed
catalyst does not exhibit the XRD reflections of the vanadia crystal-
line phase because the nitric acid selectively dissolved the V2O5
crystallites to leave only highly dispersed vanadia species on the
titania surface [30]. The amount of insoluble vanadia was approx-
imately 1 ML (Table 1). Therefore, the washed catalyst contains
vanadium oxide in only a highly dispersed state.
The presence of the highly dispersed vanadia species was con-
firmed by FTIR for both the fresh and washed V2O5/TiO2 catalysts.
Fig. 2 displays the FTIR spectra of the fresh and washed catalysts,
the TiO2 (anatase) support, and bulk V2O5. The anatase phase of
TiO2 exhibited broad, intense absorption bands at 260–360 and
400–850 cmÀ1, consistent with the literature [36–38]. The spec-
trum of the fresh catalyst was a superposition of the anatase spec-
trum with weak but well-resolved bands at 800–1030 cmÀ1. After
the subtraction of the anatase spectrum, these features appeared as
a sharp band at 1015 cmÀ1 and broader bands at approximately
946 and 856 cmÀ1. Bulk V2O5 was characterized by the peaks at
1021, 828, 605, and 480 cmÀ1 [36,37,39]. Therefore, the peaks at
1021 and 828 cmÀ1 were assigned to the V@O stretch and the
3.2. Catalytic performance
The oxidation of methanol using molecular oxygen was exam-
ined using the washed V2O5/TiO2 catalysts between 70 and
150 °C. The methanol/oxygen molar ratio was maintained at 1:1.
Dimethoxymethane (H2C(OCH3)2), methyl formate (HCOOCH3),
formaldehyde (HCHO), formic acid (HCOOH), carbon oxides (CO
and CO2), and water were detected as products. No dimethyl ether
or any other organic compounds were detected over the entire
temperature range. The conversion and selectivity for each product
at each temperature, catalyst loading, and feed flow-rate are pre-
sented in Table 2. Pure TiO2 (anatase) did not catalyze any measur-
able methanol conversion up to 200 °C. This result agrees with the
literature data, which indicate that TiO2 provides a high level of
methanol conversion only above 300 °C to form CO and CO2 as
the main products [42].
Table 2 also illustrates the product distribution, which depends
strongly on the reaction temperature. Between 70 and 100 °C, dim-
ethoxymethane is the major product, with a selectivity of 80–95%.
Subsequently, the dimethoxymethane selectivity decreases with
the reaction temperature and the reaction shifts toward methyl
formate production. Between 140 and 150 °C, methyl formate be-
comes the main reaction product. The formation of small amounts
of formaldehyde, formic acid, CO, and CO2 was also observed. The
conversion of methanol also influenced the selectivity (Table 2).
Comparison of the data at 120 °C (highlighted with bold in Table 2)
indicated that, at higher methanol conversions, the selectivity to-
ward dimethoxymethane was lower and the selectivity for methyl
formate was higher. For this reason, the oxidation of methanol was
examined at a constant methanol conversion.
Fig. 3 demonstrates the effects of the reaction temperature on
the rate and the selectivity at 50% methanol conversion. The rates
of methyl formate, formaldehyde, formic acid, and COx formation
increased with the temperature, while the dimethoxymethane
yield had a maximum at 130 °C. At low temperatures, dim-
ethoxymethane was the main reaction product. Above 120 °C, the
dimethoxymethane selectivity decreased and the methyl formate
selectivity increased. Methyl formate was the main reaction prod-
uct between 140 and 150 °C. As shown by Busca et al. [9–11], the
methyl formate selectivity decreases above 170 °C and carbon
monoxide becomes the main reaction product at 190–200 °C.
Fig. 1. XRD patterns obtained from TiO2 anatase (1), fresh (2) and washed (3) V2O5/
TiO2 catalysts. The marked features are from V2O5.

62
V.V. Kaichev et al. / Journal of Catalysis 311 (2014) 59–70
Table 1
Characteristics of the fresh and washed V2O5/TiO2 catalysts.
Catalyst
V2O5 content, wt%
SBET, m2/g
V2O5 surface densitya, V atom/nm2
V2O5 surface coverageb, ML
Fresh
Washed
20
12.5
111
115
11.9
7.3
1.5
0.9
a
Refers to the BET area.
b
1 ML corresponds to 7.9 V atom/nm2 [27].
Fig. 2. (a) FTIR spectra of TiO2 anatase (1), fresh (2) and washed (3) V2O5/TiO2 catalysts, bulk V2O5 (4) after calcination at 400 °C in a flow of air, as well as difference spectra
(2-1 and 3-1), which were obtained via direct subtraction of spectrum 1 from spectra 2 and 3, respectively. The samples were pellets composed of the catalysts or the TiO2
powder pressed with CsI. (b) FTIR spectra of thin self-supported pellets of the washed catalyst obtained at 25 °C (1) and 200 °C (2) in air.
Table 2
Catalytic properties of the washed V2O5/TiO2 catalyst during the oxidation of methanola.
Temp, °C
Catalyst loading, g
Feed flow, l/h
Conversion, %
Selectivity, %
DMMb
MFc
Fd
FAe
CO
CO2
70
100
120
130
140
100
120
140
100
120
130
140
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
7.5
7.5
7.5
7.5
2.5
2.5
2.5
2.5
2.5
4.5
4.5
4.5
2.5
2.5
2.5
2.5
10.0
17.0
30.1
44.0
60.0
10.0
20.5
45.0
45.0
59.7
73.5
85.0
95
90
75
43
3
94
80
8.2
80
57
15
1
5.0
10.0
19.4
46.7
83.9
6.0
17.0
78.5
20.0
32.8
74.5
82.5
0
0
2.5
4.0
3.6
0
3
5.0
0
2.0
2.5
3.0
0
0
3.1
2.3
2.5
0
0
3.1
0
5.1
4.0
3.0
0
0
0
3.0
5.5
0
0
4.1
0
0
0
0
1.0
1.5
0
0
1.1
0
3.1
4.0
8.5
0
0.0
2.0
a
b
c
The molar methanol/oxygen ratio was 1:1.
DDM is dimethoxymethane.
MF is methyl formate.
F is formaldehyde.
FA is formic acid.
d
e
3.3. In situ FTIR study
of the adsorbed surface species. At temperatures up to 120 °C,
the methoxy species (–OCH3) is the most abundant reaction inter-
mediate. This conclusion was drawn from the observation of two
intense bands at 2931 and 2825 cmÀ1, which corresponded to
Fig. 4 presents the FTIR spectra acquired under the steady-state
conditions during the oxidation of methanol at different tempera-
tures. A reaction mixture of 2 vol% CH3OH in air was passed
through the IR cell-reactor loaded with the washed V2O5/TiO2 cat-
alyst. The spectrum from the catalyst before exposure to the reac-
tion mixture and the spectrum from gas-phase methanol were
subtracted from the raw IR spectra to identify the contributions
the symmetric stretching ms(CH3) mode and the Fermi resonance
of 2ds(CH3) of the adsorbed methoxy species, respectively
[9,13,43–49]. The bands at 2955 and 2847 cmÀ1 could be assigned
to the same vibration modes of non-dissociatively adsorbed
methanol [9]. The shoulder at 2970 cmÀ1 can be attributed to the

V.V. Kaichev et al. / Journal of Catalysis 311 (2014) 59–70
63
Fig. 3. Effect of reaction temperature on the selectivity (right side) and the rate (10À9 mol mÀ2 sÀ1) of formation (left side) of dimethoxymethane (DMM), methyl formate
(MF), formaldehyde (F), formic acid (FA), and COx. Conversion of methanol was 50%.
Fig. 4. FTIR difference spectra of the washed V2O5/TiO2 catalyst obtained during the oxidation of methanol at different temperatures: 1 – 50 °C, 2 – 100 °C, 3 – 120 °C, 4 –
140 °C, 5 – 160 °C, 6 – 180 °C.
asymmetric stretching mode
m
as(CH3) of the methyl group in the
The methanol adsorption on the TiO2 support was also investi-
gated between 100 and 170 °C. Fig. 5 displays the typical FTIR spec-
trum of the adsorbed species on the pure TiO2 (anatase) after the
methanol adsorption at 100 °C. The intense peaks at 2960, 2925,
2831, and 1462 cmÀ1 correspond to the CH3 stretching and bend-
ing modes of the adsorbed molecular methanol and methoxy spe-
cies [9,13]. No peaks characteristic of the adsorbed formaldehyde,
dioxymethylene, and formate species were detected between 100
and 170 °C. These results agree with the low activity of TiO2 during
the oxidation of methanol.
methoxy species. The symmetric and asymmetric deformation
vibrational modes d(CH3) at 1433 and 1447 cmÀ1, as well as the
rocking vibration mode
q
(CH3) at 1150 cmÀ1 detected in the low
frequency region, also corresponded to the methoxy species [9].
In addition, bands assigned to the surface dioxymethylene
(CH2O2) species [46] were detected at 2923 and 2884 cmÀ1. The
spectrum obtained at 50 °C displays an intense band at
1627 cmÀ1 attributed to the bending mode of the adsorbed water
molecules. Simultaneously,
3400 cmÀ1 (not shown) corresponding to the stretching vibration
(OH) of adsorbed water and adsorbed molecular methanol. When
a strong band was observed at
m
3.4. In situ NAP XPS, XANES, and TPR study
the temperature increased from 50 to 120 °C, no significant
changes were observed in the intensities of the vibrational bands
attributed to the surface methoxy and dioxymethylene species.
Above 140 °C, the peaks at 1550, 1360, and 1655 cmÀ1 began to
dominate in the spectrum. The two former peaks were characteris-
In situ NAP XPS and XANES techniques could be used under
mbar pressures [35,50–52]. The comparison of the spectroscopic
data with the kinetics measurements performed in the flow reactor
at atmospheric pressure (Table 2 and Fig. 3) is only possible if there
is no pressure gap. The TPR technique was applied to verify this
statement. The washed V2O5/TiO2 catalyst was heated inside the
in situ XPS reaction gas cell in the equimolar CH3OH/O2 mixture
and in pure methanol at a total pressure of 0.25 mbar; the product
tic of the
species, and the feature at 1655 cmÀ1 was due to the stretching
vibration (C@O) of adsorbed methyl formate [9,13,49]. The vibra-
tional frequencies are summarized in Table 3.
mas(O–C–O) and ms(O–C–O) modes of a bidentate formate
m

64
V.V. Kaichev et al. / Journal of Catalysis 311 (2014) 59–70
Table 3
IR bands and assignments of the surface species formed during the oxidation of methanol on the V2O5/TiO2 catalyst.
Frequency, cmÀ1
Type of surface species
Type of vibration
Rocking (CH3) mode
Symmetric stretching
Symmetric deformation d(CH3) mode
Asymmetric deformation d(CH3) mode
Ref.
1150
1360
1433
1447
1550
1655
2825
2847
2884
2923
2931
2955
2970
Methoxy species
Bidentate formate species
Methoxy species
Methoxy species
Bidentate formate species
Methyl formate
q
[9]
[9,13,49]
[9]
m
s(O–C–O) mode
[9]
Asymmetric stretching
m
as(O–C–O) mode
[9,13,49]
[9,13,49]
[9,13,43–49]
[9]
[46]
[46]
[9,13,43–49]
[9]
[9]
Stretching (C@O) mode
m
Methoxy species
Fermi resonance of 2ds(CH3)
Fermi resonance of 2ds(CH3)
Symmetric stretching
Asymmetric stretching
Symmetric stretching
Symmetric stretching
Asymmetric stretching
Molecular methanol
Dioxymethylene species
Dioxymethylene species
Methoxy species
Molecular methanol
Methoxy species
m
m
m
s(CH2) mode
as(CH2) mode
s(CH3) mode
ms(CH3) mode
m
as(CH3) mode
Fig. 5. FTIR difference spectra of TiO2 (anatase) obtained during the adsorption of methanol at 100 °C.
distribution was monitored with a differentially pumped mass-
spectrometer (Fig. 6). The catalyst was pretreated in 0.25 mbar of
flowing O2 at 300 °C for 30 min.
V 2p3/2 and V L2,3-edges XAS spectra, along with the Ti 2p3/2
core-level and Ti L2,3-edges XAS spectra, are presented in Figs. 7
and 8, respectively. Before exposure to the reaction mixture, the
catalyst was pretreated in 0.25 mbar of flowing O2 for 30 min at
300 °C inside the in situ XPS reaction gas cell. This treatment gen-
erated the fully oxidized vanadium species: only a narrow single
peak at 517.7 eV corresponding to the V5+ state was observed un-
der the oxygen atmosphere (Fig. 7a). A new V 2p3/2 peak at 516.4–
516.5 eV appears in the reaction mixture, pointing to the partial
reduction of V5+ to V4+, even at room temperature. It should be
noted that in highly cited studies [53–57], the vanadium ions in
the vanadium oxides, such as V2O5, V2O4, and V2O3, are character-
ized by the V 2p3/2 binding energies in the ranges of 516.9–517.2,
515.7–516.2, and 515.2–515.7 eV, respectively. However, the latest
studies have reported higher binding energy values of V 2p3/2 at
517.3–517.7 and 516.2–516.5 eV for bulk and supported V2O5, as
well as for V2O4, respectively [20,58–64]. These numbers agree
with our data.
The detection of formaldehyde and CO with a mass-spectrome-
ter was hindered due to overlap in the methanol fragmentation
pattern. However, the presence of formic acid, methyl formate,
and dimethoxymethane among the products was unambiguously
proven by the mass-spectrometric signals with m/z of 45, 60,
and 75, respectively. Within the reaction mixture (Fig. 6a), the
yield of dimethoxymethane displayed a maximum at 95 °C and
subsequently decreased. The methyl formate yield reached its
maximum at approximately 130 °C. The formic acid exhibited stea-
dy growth above 130 °C. The maximum for dimethoxymethane
was also detected at 95 °C in the experiment using pure methanol
(Fig. 6b); however, in the absence of O2, the methyl formate forma-
tion was suppressed and formic acid formation was facilitated
above 130 °C. Therefore, the catalytic behavior under mbar pres-
sure follows the same trends as observed under atmospheric pres-
sure (Table 2, Fig. 3); the results of the in situ NAP XPS and XANES
study are therefore fully applicable.
Fig. 7b displays the V L2,3-edges XAS spectra obtained during the
same experiment. The transition energy and the line shape are sen-
sitive to the chemical environment and can be used as a fingerprint
for the vanadium oxidation state [65,66]. Because the selection
rules for the electron transition during photon absorption require
The NAP XPS and XANES spectra were measured during heating
the V2O5/TiO2 catalyst in the step-wise manner in both the equi-
molar CH3OH/O2 mixture and pure methanol. The main NAP XPS
results for the washed catalyst are summarized in Table 4. The
D
l = 1, the peaks at 518–519 and 524–525 eV in the XANES

V.V. Kaichev et al. / Journal of Catalysis 311 (2014) 59–70
65
Fig. 6. TPR spectra measured in an equimolar CH3OH/O2 mixture (a) and in methanol (b) with a total pressure of 0.25 mbar.
Table 4
Binding energies and FWHMs of the Ti 2p3/2 and V 2p3/2 peaks shown in Figs. 7a, 8a, and 9; the relative intensity of the different components (%) are shown in parentheses.
Mixture
T, °C
FWHM of Ti 2p3/2, eV
V 2p3/2, eV (%)
V5+
V4+
V3+
O2
300
100
150
200
50
100
150
200
1.12
1.05
1.07
1.09
1.10
1.12
1.13
1.14
517.66 (100)
517.73 (51)
517.68 (59)
517.63 (67)
–
–
–
–
–1
–
–
–
–
CH3OH + O2
CH3OH + O2
CH3OH + O2
CH3OH
CH3OH
CH3OH
516.50 (49)
516.43 (41)
516.40 (33)
516.60 (41)
516.60 (31)
516.60 (28)
516.50 (32)
515.54 (59)
515.42 (69)
515.45 (72)
515.47 (68)
CH3OH
Fig. 7. Normalized V 2p3/2 core-level spectra (a) and V L2,3-edges XAS spectra (b) of the washed V2O5/TiO2 catalyst measured in 0.25 mbar O2 at 300 °C (1), as well as in the
equimolar CH3OH/O2 mixture with a total pressure of 0.25 mbar at 100, 150, and 200 °C (2, 3, and 4); 2-1 is the difference XAS spectrum obtained via the direct subtraction of
spectrum 1 multiplied on the factor 0.5 from spectrum 2, respectively.

66
V.V. Kaichev et al. / Journal of Catalysis 311 (2014) 59–70
Fig. 8. Normalized Ti 2p3/2 core-level spectra (a) and Ti L2,3-edge XAS spectra (b) of the washed V2O5/TiO2 catalyst measured in 0.25 mbar O2 at 300 °C (1), as well as in the
equimolar CH3OH/O2 mixture with a total pressure of 0.25 mbar at 100, 150, and 200 °C (2, 3, and 4), respectively.
spectra presented in Fig. 7b can be roughly assigned to the elec-
tronic excitations from the spin–orbit split levels 2p3/2 (L3-edge)
and 2p1/2 (L2-edge) into the empty or partially occupied vanadium
3d orbitals. The L3-edge spectrum of the fully oxidized vanadia ob-
tained in pure oxygen exhibits a well-defined fine structure with at
least four distinct resonances. The observed line shape is typical of
bulk V2O5 [66,67]. The position of the main resonance at 518.9 eV
is very similar to the value reported for bulk V2O5 [66].
The spectra reveal a well-defined fine structure with at least seven
distinct resonances. The main near-edge structure of the spectra
was assigned to the 2p63d0 ? 2p5c3d1 dipole transition, where c
represents the 2p core hole [78]. Similar to vanadium, the 2p
spin–orbit coupling splits the initial state into 2p3/2 and 2p1/2
,
resulting in two L-edge features denoted as L3 and L2, respectively.
Both the L3 and L2 features were further split into the t2g and e2g
components because of the low symmetry of the ligand field. The
L3–e2g feature also splits into a doublet (peak D and E in Fig. 8b) be-
cause of slight distortions in the TiO6 octahedra among the titania
polymorphs resulting from the configurational deformation pre-
dicted by the Jahn–Teller theorem [79]. Detailed descriptions of
the Ti L-edges spectra can be found in the literature [80,81]. The
different polymorphs of TiO2 could be distinguished via the relative
intensities of the D/E doublet [79,82,83]. For example, in the ana-
tase structure, the intensity of the D peak was substantially stron-
ger than the intensity of the E peak, whereas it would have been
the opposite in the rutile structure. In this study, all of spectra
shown in Fig. 8b are identical; their shape corresponds to the pre-
viously reported Ti L-edges spectra of anatase [79,82,83], meaning
that, during the oxidation of methanol, the Ti4+ cations were octa-
hedrally coordinated in the TiO2 anatase phase.
The results presented above clearly indicate that, during the
oxidation of methanol, the partial reduction of V5+ to V4+ occurs
but the titanium cations remain in the Ti4+ state. The curve fitting
of the V 2p3/2 spectra (Fig. 7a) indicated that the relative contribu-
tion of different vanadium oxidation states depended on the tem-
perature; the quantitative data are presented in Table 4. Increasing
the temperature from 100 to 200 °C caused more than a 15% de-
crease in the V4+ fraction. The catalyst underwent complete reduc-
tion from V5+ to V4+ and V3+ after exposure to methanol at
0.25 mbar, even at room temperature. This conclusion arose from
the analysis of the V 2p3/2 spectra of the washed V2O5/TiO2 cata-
lysts presented in Fig. 9. In the methanol atmosphere, the typical
V 2p3/2 spectrum of the fully oxidized V5+ catalyst exhibiting a sin-
gle peak transformed into a spectrum with two peaks: one at
516.5–516.6 and the other at 515.4–515.5 eV, which can be attrib-
uted to V4+ and V3+, respectively. The fraction of the V3+ state grows
slightly when the temperature increases (Table 4). The following
treatment in oxygen at 300 °C led again to the full oxidation of
the vanadium to V5+. These data indicate that supported vanadium
Under the reaction conditions, the V L2,3-edge shifts toward
lower energies and the line shape changes. To identify the fine
structure, the difference XAS spectra were analyzed. Because only
half of the V5+ cations were reduced to V4+ under the reaction con-
ditions according to the NAP XPS measurements (Table 4), a corre-
sponding difference spectrum was obtained by subtracting the
spectrum of the fully oxidized vanadia multiplied by 0.5 from the
spectrum of the partially reduced vanadia (Fig. 7b). The difference
L3-edge spectrum was shifted toward lower energies and has a dif-
ferent line shape. Only two distinct resonances were observed in
the spectrum: the main resonance at 517.8 eV and a well-distin-
guished shoulder at 515.9 eV. A similar line shape has been re-
ported for bulk VO2 [66,68,69]. Additionally, according to the
correlation between the oxidation state and the L3-edge peak posi-
tion of the binary vanadium oxides, this resonance corresponds to
the V4+ state [70,71]. Chen et al. [70] reported that the V L3-edge
position shifted by 0.7 eV for every ionic charge within the range
from 515.5 eV (metallic vanadium) to 519.0 eV (V2O5). In the cur-
rent study, the shift was approximately 0.8 eV, corresponding to
the partial reduction of V5+ to V4+. Therefore, the XANES data
proved that the partial reduction of V5+ to V4+ occurred under these
reaction conditions.
No changes were detected in the Ti 2p core-level and Ti L-edges
XAS spectra under the reaction conditions from room temperature
to 200 °C. These data indicate that titanium in the support remains
in the Ti4+ state. The narrow Ti 2p3/2 peaks with the full width at
half maximum (FWHM) of approximately 1.1 eV were at
459.0 eV (Table 4), which is typical for bulk TiO2 (Fig. 8a). Accord-
ing to the literature data [72–77], pure TiO2 is characterized by the
Ti 2p3/2 binding energy in the range of 458.7–459.2 eV, while the
binding energy for the Ti3+ species is between 456.2 and
457.4 eV. The X-ray absorption spectra are displayed in Fig. 8b.
The spectra were normalized to the same maximum peak height.

V.V. Kaichev et al. / Journal of Catalysis 311 (2014) 59–70
67
during XPS measurements. The third component at ca. 287 eV
was assigned to dioxymethylene. In agreement with the FTIR data,
the amount of the methoxy species decreased with increasing tem-
perature. A similar trend was observed for the V4+ state (Table 4),
indicating that the formation of the methoxy species on the van-
adia surface was accompanied by the partial reduction of V5+ to
V4+
.
The C 1s spectra obtained in pure methanol (Fig. 10b) revealed
additionally a weak peak at 288.9 eV, which was assigned to the
surface formate species, and a sharp peak at 287.6–287.8 eV, which
could be assigned mainly to gas-phase methanol [4] (the pressure
was high enough to observe the gas-phase signals). The FWHM of
the peak at 287.6–287.8 eV was about 0.65–0.85 eV, whereas the
FWHM of the other peaks is in the range of 1.5–1.8 eV. According
to our previous studies, such sharp peaks are usually observed
for molecules in the gas phase [35,51]. In this case, a significant
fraction of vanadium was in the V3+ state (Table 4). We conclude
that the formation of the formate species on the vanadia surface
causes the partial reduction of V4+ to V3+; however, the presence
of oxygen led to a fast re-oxidation of V3+ to V5+ (Fig. 10a).
4. Discussion
Fig. 9. Normalized V 2p3/2 core-level spectra of the washed V2O5/TiO2 catalyst
measured in 0.25 mbar O2 at 300 °C (1), as well as in 0.25 mbar CH3OH at 50, 100,
150, and 200 °C (2, 3, 4, and 5), respectively.
The catalytic and spectroscopic data obtained in this study clar-
ified the detailed mechanism for the oxidation of methanol over
vanadia–titania catalysts proposed by Busca et al. [9–13]. A review
of this mechanism has been previously published [84]. The com-
bined data from the in situ surface sensitive techniques, such as
NAP XPS and XANES, with FTIR, TPR, and kinetics measurements
allowed us to include the variations of vanadium’s chemical state
in the reaction mechanism. The proposed scheme is illustrated in
Fig. 11. The active sites are assigned to the highly dispersed van-
adia species VOx on TiO2. The polymeric vanadium oxide species
prevail in the monolayer V2O5/TiO2 catalyst. These species consist
of a terminal V@O bond with at least one V–O–Ti and one or two
bridging V–O–V bonds.
could undergo reversible oxidation and reduction during the oxi-
dation of methanol.
Fig. 10 displays the C1s core-level spectra that were measured
in situ concurrent with the V 2p3/2 and Ti 2p3/2 spectra. The oxygen
treatment fully removed carbon-contained species from the cata-
lyst surface. In the reaction mixture and in pure methanol, the cat-
alyst surface was covered with different carbon-contained species.
In the reaction mixture (Fig. 10a), three components of the C 1s
spectrum could be distinguished at 285.3–285.4, 286.1–286.2,
and 286.9–287.2 eV. The most intense component at ca. 286 eV
corresponded to the methoxy species [4]. The weaker component
at ca. 285 eV was assigned to C–C–O coupling products and/or to
hydrocarbon impurities, which can accumulate on the surface
Under the mild conditions, dimethoxymethane and methyl for-
mate are the main products formed over the V2O5/TiO2 catalyst.
The vibration bands at 2955 and 2847 cmÀ1 at low reaction tem-
peratures (Fig. 4) indicate that the first step of the catalytic cycle
Fig. 10. C 1s core-level spectra of the washed V2O5/TiO2 catalyst measured in 0.25 mbar O2 at 300 °C (1), as well as in an equimolar CH3OH/O2 mixture (a) and pure CH3OH (b)
at 100, 150, and 200 °C (2, 3, and 4), respectively; a total pressure was 0.25 mbar; all the spectra were normalized to the V 2p integral intensity.

68
V.V. Kaichev et al. / Journal of Catalysis 311 (2014) 59–70
Fig. 11. Proposed mechanism for the oxidation of methanol over highly dispersed vanadia supported on TiO2.
is the adsorption of methanol. Methanol chemically adsorbs by
donating the electron density from the oxygen atom to a surface
V5+ cation. Methanol adsorption via the hydrogen bonding with
the surface hydroxyl groups or the lattice oxygen atoms could
not be ruled out as well. Subsequently, the adsorbed methanol dis-
sociates to form the surface methoxy and hydroxyl groups (stage 1
in Fig. 11), which was confirmed by FTIR (Fig. 4). The partial reduc-
tion of the vanadium cations from the V5+ to V4+ state was detected
by NAP XPS and XANES (Fig. 7). The methoxy species may be de-
scribed as a monodentate structure with a covalent-like bond to
the vanadium cation [9,44]. The main question in this model is
the position of oxygen atom, which takes a proton and transforms
to the hydroxyl group. Kilos et al. [85] studied the oxidation of eth-
anol over VOx/Al2O3 and suggested that ethanol dissociates over
vanadia to form adsorbed ethoxide species and OH groups that
bond with adjacent vanadium atoms. In contrast, Beck et al. [86]
studied the oxidation of ethanol over supported VOx and suggested
that the hydroxyl group bonds with cations of the support. Accord-
ing to theoretical studies of the oxidation of methanol over VOx
supported on titania, the hydroxyl group bonds with a titanium
atom [87,88]. We suppose that in our monolayer catalyst, which
consisted of polymeric VOx species, the methoxy group probably
coordinated to a vanadium cation, whereas H atom could be trans-
ferred to oxygen of the V–O–Ti bond to form the Ti–OH group
(stage 1 in Fig. 11).
Due to the further loss of the hydrogen atom, the methoxy spe-
cies transforms into molecularly adsorbed formaldehyde, and this
event is accompanied by the partial reduction of the second vana-
dium cation to V4+ (stage 2 in Fig. 11). In light of the recent inves-
tigations [86,87,89], we believe that the formation of formaldehyde
occurs via the transfer of an H atom of a methoxy group to the O
atom of the V@O group. The hydroxyl groups recombine to gener-
ate water (stage 3 in Fig. 11).
Depending on the temperature, formaldehyde can desorb as a
product or undergo further transformations to form dioxymeth-
ylene species (stage 4 in Fig. 11). The dioxymethylene species
were detected on the V2O5/TiO2 catalyst by FTIR spectroscopy
(Fig. 4). Because only a minor amount of gas-phase formaldehyde
was detected by gas chromatography at 130 °C (Table 2) and a
large amount of the surface dioxymethylene species was
observed during the formaldehyde oxidation on the same cata-
lyst at 70 °C [46,90], the formation of the dioxymethylene species
from the adsorbed formaldehyde species is likely a rapid process.
The reaction of the dioxymethylene species with the methoxy
species or methanol generates dimethoxymethane, which is the
main reaction product at low temperatures (Table 2). This reaction
most likely proceeds through the nucleophilic attack to the
dioxymethylene carbon atom by the oxygen atoms from an adja-
cent methoxy species or from an adsorbed methanol molecule
[91]. This proposal is consistent with the high selectivity observed
toward dimethoxymethane at temperatures up to 120 °C. At this
temperature, the surface was covered with the dioxymethylene
species, molecularly adsorbed methanol, and the methoxy species
(Fig. 4).
At higher temperatures, the dioxymethylene is oxidized to gen-
erate the formate species (stage 5 in Fig. 11). The formation of the
formate species is accompanied by the reduction of vanadium cat-
ions to the V3+ state observed by XPS under the methanol atmo-
sphere (Fig. 8). In the methanol/oxygen mixture, the V3+ cation
was not detected because the re-oxidation of V3+ to form V4+
and/or V5+ is rapid in the presence of gas-phase oxygen. The for-
mate species may transform into formic acid and carbon monoxide
[9] or may react with methanol to generate methyl formate. As
shown by FTIR, the concentration of the dioxymethylene species
decreases above 120 °C, while the intensities of the stretching
vibrations for the formate species (1550 and 1360 cmÀ1) and the
intensity of the adsorbed methyl formate band at 1655 cmÀ1 in-
crease (Fig. 4). In good agreement with this mechanism, the
V2O5/TiO2 catalysts demonstrate high selectivity during the oxida-
tion of formaldehyde to generated formic acid under similar condi-
tions [30,46,90]. The formate species is stable in helium up to
180 °C, but readily decomposes in oxygen at 100–150 °C, produc-
ing formic acid [46]. A similar behavior was observed during the
in situ XPS study. The formate species were identified in the C 1s
spectra measured in methanol (Fig. 10b), whereas the presence
of oxygen in the reaction mixture leads a decreased concentration
of the adsorbed formate species below the XPS detection limit
(Fig. 10a). As a result, the formation of formic acid was observed
in the TPR experiments (Fig. 6).

V.V. Kaichev et al. / Journal of Catalysis 311 (2014) 59–70
69
Re-oxidation of the vanadium active sites by dioxygen com-
pletes the catalytic cycle (stage 6 in Fig. 11). Although it is difficult
to explain how O2 molecules transforms into lattice O2À species on
monolayer V2O5/TiO2 catalysts, the re-oxidation of V3+ and V4+ to
V5+ during the oxidation of methanol in the presence of O2 was
clearly observed by XPS (Figs. 7 and 8, Table 4). Most probably,
the rapid re-oxidation of reduced vanadia can occur via the adsorp-
tion of O2 at a vacancy site to form a peroxide species; one of two
oxygen atoms eliminates the vacancy, while the other oxygen atom
migrates across the support surface until it reaches and eliminates
a secondary vacancy [92]. The re-oxidation of the catalyst surface
should restore its adsorption properties because methanol adsorbs
mainly on the V5+ sites. Therefore, the fully oxidized vanadia
should be more active than the partially reduced species. This
hypothesis is confirmed because the methanol conversion on the
V–Ti oxide catalyst at 175 °C gradually decreases from 100% to
approximately 30% when the molar methanol/oxygen ratio
increases from 0.5 to 2.7 [9,11]. This also agrees with the data re-
ported by Wang and Madix [93], who have shown using TPRS that
the fully oxidized surface in a model monolayer V2O5/TiO2(110)
catalyst has a much higher activity in the oxidation of methanol
to formaldehyde compared with the reduced surface with vana-
accompanied by the reduction of the vanadium to V3+. The cata-
lytic cycle is completed by re-oxidizing the vanadium cations to
the V5+ state with molecular oxygen from the gas phase.
Acknowledgments
This work was partially supported by the Integrated Infrastruc-
ture Initiative I3 in FP6 R II 3 CT-2004-506008 (BESSY IA-SFS
Access Program) and by the RFBR (Research Project No. 13-03-
00128). The authors thank Drs. M. Hävecker, E. Kleimenov, D.
Tescher, S. Zafeiratos for their assistance in carrying out ambient
pressure XPS and XANES measurements. The authors are also
grateful to the staff of BESSY for their support during the beamline
operation. The authors also thank Dr. L.M. Plyasova for XRD
measurements.
References
[1] J.M. Tatibouët, Appl. Catal. A 148 (1997) 213.
[2] P. Forzatti, E. Tronconi, A.S. Elmi, G. Busca, Appl. Catal. A 157 (1997) 387.
[3] M. Badlani, I.E. Wachs, Catal. Lett. 75 (2001) 137.
[4] Y. Romanyshyn, S. Guimond, H. Kuhlenbeck, S. Kaya, R.P. Blum, H. Niehus, S.
Shaikhutdinov, V. Simic-Milosevic, N. Nilius, H.-J. Freund, M.V. Ganduglia-
Pirovano, R. Fortrie, J. Döbler, J. Sauer, Top. Catal. 50 (2008) 106.
[5] J.M. Tatibuet, J.E. Germain, C.R. Acad. Sci. C 289 (1979) 305.
[6] M. Baltes, K. Cassiers, P.v.D. Voort, B.M. Weckhuysen, R.A. Schoonheydt, E.F.
Vansant, J. Catal. 197 (2001) 160.
[7] C. Hess, I.J. Drake, J.D. Hoefelmeyer, T.D. Tilley, A.T. Bell, Catal. Lett. 105 (2005)
1.
[8] N. Koivikko, T. Laitinen, S. Ojala, S. Pitkäaho, A. Kucherov, R.L. Keiski, Appl.
Catal. B 103 (2001) 72.
[9] G. Busca, A.S. Elmi, P. Forzatti, J. Phys. Chem. 91 (1987) 5263.
[10] P. Forzatti, E. Tronconi, G. Busca, P. Tittarelli, Catal. Today 1987 (1987) 209.
[11] E. Tronconi, A.S. Elmi, N. Ferlazzo, P. Forzatti, G. Busca, Ind. Eng. Chem. Res. 26
(1987) 1269.
[12] A.S. Elmi, E. Tronconi, C. Cristiani, J.P.G. Martin, P. Forzatti, G. Busca, Ind. Eng.
Chem. Res. 28 (1989) 387.
dium oxidation states of V3+ and V4+
.
The presented data follow the proposal that the selective oxida-
tion of methanol on the monolayer V2O5/TiO2 catalyst occurs at the
redox Vn+ sites via the Mars–van Krevelen mechanism [94], which
involves the lattice oxygen species (see the proposed scheme in
Fig. 11). Final proof of this mechanism requires the comparison
of the methanol oxidation rate over the V2O5/TiO2 catalyst in the
presence and in the absence of oxygen [95], which is the subject
of a future work.
[13] G. Busca, J. Mol. Catal. 50 (1989) 241.
5. Conclusions
[14] Y. Fu, J. Shen, Chem. Commun. (2007) 2172.
[15] Q. Sun, Y. Fu, J. Liu, A. Auroux, J. Shen, Appl. Catal. A 334 (2008) 26.
[16] J. Liu, Y. Fu, Q. Sun, J. Shen, Microporous Mesoporous Mater. 116 (2008) 614.
[17] J. Liu, Q. Sun, Y. Fu, J. Shen, J. Colloid Interface Sci. 335 (2009) 216.
[18] H. Zhao, S. Bennici, J. Shen, A. Auroux, J. Mol. Catal. A 309 (2009) 28.
[19] H. Zhao, S. Bennici, J. Shen, A. Auroux, J. Catal. 272 (2010) 176.
[20] H. Zhao, S. Bennici, J. Cai, J. Shen, A. Auroux, Catal. Today 152 (2010) 70.
[21] Q. Sun, J. Liu, J. Cai, Y. Fu, J. Shen, Catal. Commun. 11 (2010) 47.
[22] H. Guo, D. Li, D. Jiang, W. Li, Y. Sun, Catal. Commun. 11 (2010) 396.
[23] G.C. Bond, J.P. Zurita, S. Flamerz, P.J. Gellings, H. Bosch, J.G. van Ommen, B.J.
Kip, Appl. Catal. 22 (1986) 361.
The highly dispersed vanadia species on TiO2 are active and
selective catalysts for the oxidation of methanol to methyl formate
and dimethoxymethane under mild conditions. At low tempera-
tures (70–100 °C), dimethoxymethane is the main reaction product
with the 88–95% selectivity, while at higher temperatures (150 °C),
methyl formate becomes the main reaction product with the 85%
selectivity. The production of formaldehyde is greatly inhibited at
temperatures below 150 °C.
[24] D.A. Bulushev, L. Kiwi-Minsker, F. Rainone, A. Renken, J. Catal. 205 (2002) 115.
[25] G.C. Bond, S.F. Tahir, Appl. Catal. A 71 (1991) 1.
[26] G. Deo, I.E. Wachs, J. Catal. 146 (1994) 323.
Combining the catalytic measurements with in situ FTIR, NAP
XPS, XANES, and TPR provided additional insight into the mecha-
nism for the oxidation of methanol. The FTIR data revealed the
presence of molecularly adsorbed methanol, in addition to the ad-
sorbed methoxy, dioxymethylene, and formate species. NAP XPS
and XANES demonstrated that the reaction involves the reversible
reduction of V5+ cations, proving that the lattice oxygen partici-
pates in the oxidation of methanol via the classical Mars–van Krev-
elen mechanism, which consists of the reduction of the catalyst
surface by methanol and its subsequent re-oxidation by oxygen
from the gas phase. The reaction begins with the molecular chemi-
sorption of methanol on the V5+ cations. The adsorbed molecules
dissociate to form the methoxy species and hydroxyl groups,
which is accompanied by the reduction of the vanadium cations
from the V5+ to V4+ state. The oxidation (H abstraction and electron
transfer) of the methoxy species leads to the formation of adsorbed
formaldehyde, which can desorb as the product or transform into
the surface dioxymethylene species. The reaction of the dioxym-
ethylene species with either the methoxy species or methanol gen-
erates dimethoxymethane, which is the main product at low
temperatures. At higher temperatures, dioxymethylene transforms
into the formate species, which may also react with methanol to
furnish methyl formate. The formation of the formate species is
[27] I.E. Wachs, Catal. Today 27 (1996) 437.
[28] I.E. Wachs, B.M. Weckhuysen, Appl. Catal. A 157 (1997) 67.
[29] B. Grzybowska-Swierkosz, Appl. Catal. A 157 (1997) 263.
[30] G.Y. Popova, T.V. Andrushkevich, E.V. Semionova, Y.A. Chesalov, L.S. Dovlitova,
V.A. Rogov, V.N. Parmon, J. Mol. Catal. A 283 (2007) 146.
[31] M.V. Martinez-Huerta, J.L.G. Fierro, M.A. Banares, Catal. Commun. 11 (2009)
15.
[32] I.E. Wachs, R.Y. Saleh, S.S. Chan, C.C. Chersich, Appl. Catal. 15 (1985) 339.
[33] E.V. Kondratenko, M. Baerns, Appl. Catal. A 222 (2001) 133.
[34] E.V. Ovchinnikova, T.V. Andrushkevich, G.Y. Popova, V.D. Meshcheryakov, V.A.
Chumachenko, Chem. Eng. J. 154 (2009) 60.
[35] A. Knop-Gericke, E. Kleimenov, M. Hävecker, R. Blume, D. Teschner, S.
Zafeiratos, R. Schlögl, V.I. Bukhtiyarov, V.V. Kaichev, I.P. Prosvirin, A.I.
Nizovskii, H. Bluhm, A. Barinov, P. Dudin, M. Kiskinova, Adv. Catal. 52 (2009)
213.
[36] G.C. Bond, A. Sarkany, G.D. Parfitt, J. Catal. 57 (1979) 476.
[37] B.M. Reddy, I. Ganesh, B. Chowdhury, Catal. Today 49 (1999) 115.
[38] G. Busca, A. Zecchina, Catal. Today 20 (1994) 61.
[39] B.P. Barbero, L.E. Cadus, Appl. Catal. A 244 (2003) 235.
[40] V. Dimitrov, Y. Dimitriev, A. Montenero, J. Non-Crystal. Solids 244 (1994) 235.
[41] E.V. Danilevich, G.Y. Popova, T.V. Andrushkevich, Y.A. Chesalov, V.V. Kaichev,
A.A. Saraev, L.M. Plyasova, Stud. Surf. Sci. Catal. 175 (2010) 463.
[42] F. Roozeboom, P.D. Cordingley, P.J. Gellings, J. Catal. 68 (1981) 464.
[43] G. Busca, V. Lorenzelli, J. Catal. 66 (1980) 155.
[44] G. Busca, P.F. Rossi, V. Lorenzelli, M. Benoissa, J. Travert, J.C. Lavalley, J. Phys.
Chem. 89 (1985) 5433.
[45] L.J. Burcham, I.E. Wachs, Catal. Today 49 (1999) 467.
[46] G.Y. Popova, Y.A. Chesalov, T.V. Andrushkevich, E.S. Stoyanov, Kinet. Catal. 41
(2000) 546.

70
V.V. Kaichev et al. / Journal of Catalysis 311 (2014) 59–70
[47] L.J. Burcham, L.E. Briand, I.E. Wachs, Langmuir 17 (2001) 6164.
[48] L.J. Burcham, M. Badlani, I.E. Wachs, J. Catal. 203 (2001) 104.
[49] V. Lochar, J. Machek, J. Tichy, Appl. Catal. A 228 (2002) 95.
[50] A. Knop-Gericke, M. Hävecker, Th. Schedel-Niedrig, R. Schlögl, Top. Catal. 10
(2000) 187.
[51] V.I. Bukhtiyarov, V.V. Kaichev, I.P. Prosvirin, Top. Catal. 32 (2005) 3.
[52] V.V. Kaichev, A.V. Miller, I.P. Prosvirin, V.I. Bukhtiyarov, Surf. Sci. 606 (2012)
420.
[53] R.J. Colton, A.M. Guzman, J.W. Rabalais, J. Appl. Phys. 49 (1978) 409.
[54] G.A. Sawatzky, D. Post, Phys. Rev. B 20 (1979) 1546.
[55] J. Mendialdua, R. Casanova, Y. Barbaux, J. Electron Spectrosc. Relat. Phenom. 71
(1995) 249.
[56] M. Demeter, M. Neumann, W. Reichelt, Surf. Sci. 454–456 (2000) 41.
[57] G. Silversmit, D. Depla, H. Poelman, G.B. Marin, R.D. Cryse, J. Electron Spectrosc.
Relat. Phenom. 135 (2004) 167.
[58] M. Sambi, G. Sangiovanni, G. Granozzi, F. Parmigiani, Phys. Rev. B 55 (1997)
7580.
[59] M.E. Harlin, V.M. Niemi, A.O.I. Krause, J. Catal. 195 (2000) 67.
[60] K.V.R. Chary, G. Kishan, C.P. Kumar, G.V. Sagar, J.W. Niemantsverdriet, Appl.
Catal. A 245 (2003) 303.
[61] L.K. Boudali, A. Ghorbel, P. Grage, F. Figueras, Appl. Catal. B 59 (2005) 105.
[62] J. Haber, Catal. Today 142 (2009) 100.
[63] Z. Wu, F. Dong, Y. Liu, H. Wang, Catal. Commun. 11 (2009) 82.
[64] Q. Tang, Y. Chen, Y. Yang, J. Mol. Catal. A 315 (2010) 43.
[65] J. Biener, M. Bäumer, R.J. Madix, P. Liu, E.J. Nelson, T. Kendelewicz, G.E. Brown
Jr., Surf. Sci. 441 (1999) 1.
[72] C.M. Greenlief, J.M. White, C.S. Ko, R.J. Gorte, J. Phys. Chem. 89 (1985) 5025.
[73] A.Y. Stakheev, E.S. Shpiro, J. Apijok, J. Phys. Chem. 97 (1993) 5668.
[74] R. Sanjines, H. Tang, H. Berger, F. Gozzo, G. Margaritondo, F. Levy, J. Appl. Phys.
75 (1994) 2945.
[75] Z. Luan, E.M. Maes, P.A.W.v.d. Heide, D. Zhao, R.S. Czernuszewicz, L. Kevan,
Chem. Mater. 11 (1999) 3680.
[76] Y. Hasegawa, A. Ayame, Catal. Today 71 (2001) 177.
[77] P. Finetti, F. Sedona, G.A. Rizzi, U. Mick, F. Sutara, M. Svec, V. Matolin, K.
Schierbaum, G. Granozzi, J. Phys. Chem. C 111 (2007) 869.
[78] C. Heiliger, F. Heyroth, F. Syrowatka, H.S. Leipner, I. Maznichenko, K. Kokko, W.
Hergert, I. Mertig, Phys. Rev. B 73 (2006) 045129.
[79] R. Brydson, H. Sauer, W. Engel, J.M. Thomas, E. Zeitler, N. Kosugi, H. Kuroda, J.
Phys.: Condens. Matter 1 (1989) 797.
[80] J.G. Chen, Surf. Sci. Rep. 30 (1997) 1.
[81] H.C. Choi, H.-J. Ahn, Y.M. Jung, M.K. Lee, H.J. Shin, S.B. Kim, Y.-E. Sung, Appl.
Spectrosc. 58 (2004) 598.
[82] J.P. Crocombette, F. Jollet, J. Phys.: Condens. Matter 6 (1994) 10811.
[83] E. Stoyanov, F. Langenhorst, G. Steinle-Neumann, Am. Miner. 92 (2007) 577.
[84] G. Centi, F. Cavani, F. Trifiro, Selective Oxidation by Heterogeneous Catalysis,
Kluwer Academic/Plenum Publishers, New York, 2001. p. 143.
[85] B. Kilos, A.T. Bell, E. Iglesia, J. Phys. Chem. C 113 (2009) 2830.
[86] B. Beck, M. Harth, N.G. Hamilton, C. Carrero, J.J. Uhlrich, A. Trunschke, S.
Shaikhutdinov, H. Schubert, H.-J. Freund, R. Shlögl, J. Sauer, R. Schomäcker, J.
Catal. 296 (2012) 120.
[87] A. Goodrow, A.T. Bell, J. Phys. Chem. C 112 (2008) 13204.
[88] V. Shapovalov, T. Fievez, A.T. Bell, J. Phys. Chem. C 116 (2012) 18728.
[89] R.Z. Khaliullin, A.T. Bell, J. Phys. Chem. B 106 (2002) 7832.
[90] G.Y. Popova, Y.A. Chesalov, T.V. Andrushkevich, I.I. Zakharov, E.S. Stoyanov, J.
Mol. Catal. A 158 (2000) 345.
[66] R. Zimmermann, R. Claessen, F. Reinert, P. Steiner, S. Hüfner, J. Phys.: Condens.
Matter 10 (1998) 5697.
[67] M. Hävecker, A. Knop-Gericke, R.W. Mayer, M. Fait, H. Bluhm, R. Schlögl, J.
Electron Spectrosc. Relat. Phenom. 125 (2001) 79.
[68] M. Abbate, F.M.F.d. Groot, J.C. Fuggle, Y.J. Ma, C.T. Chen, F. Sette, A. Fujimori, Y.
Ueda, K. Kosuge, Phys. Rev. B 43 (1991) 7263.
[91] Y.C. Liu, G.L. Griffin, S.S. Chan, I.E. Wachs, J. Catal. 94 (1985) 108.
[92] A. Goodrow, A.T. Bell, J. Phys. Chem. C 111 (2007) 14753.
[93] Q. Wang, R.J. Madix, Surf. Sci. 496 (2002) 51.
[69] D. Ruzmetov, S.D. Senanayake, S. Ramanathan, Phys. Rev. B 75 (2007) 195102.
[70] J.G. Chen, C.M. Kirn, B. Frühberger, B.D. DeVries, M.S. Touvelle, Surf. Sci. 321
(1995) 145.
[94] P. Mars, D.W. van Krevelen, Chem. Eng. Sci. 3 (1954) 41.
[95] G.Y. Popova, T.V. Andrushkevich, Y.A. Chesalov, V.N. Parmon, J. Mol. Catal. A
268 (2007) 251.
[71] D.S. Su, V. Roddatis, M. Willinger, G. Weinberg, E. Kitzelmann, R. Schlögl, H.
Knözinger, Catal. Lett. 74 (2001) 169.