Synthesis, characterization and catalytic performance of titania supported VPO catalysts for the ammoxidation of 3-picoline

Article abstract of DOI:10.1016/j.apcata.2010.07.034

Series of vanadium phosphorus oxide (VPO) catalysts supported over titania (anatase) were synthesised with varying contents of VPO (5-50 wt%). These solids were characterised by ICP-OES, TG/DTA, BET, XRD, FTIR (Py-ads) and XPS. The catalytic activity was

Full text of DOI:10.1016/j.apcata.2010.07.034

Applied Catalysis A: General 391 (2011) 52–62
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Synthesis, characterization and catalytic performance of titania supported VPO
catalysts for the ammoxidation of 3-picoline
∗
V.N. Kalevaru , N. Madaan, A. Martin
Leibniz-Institut für Katalyse e.V. an der Universität Rostock, Albert-Einstein-Str. 29a, D-18059 Rostock, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Series of vanadium phosphorus oxide (VPO) catalysts supported over titania (anatase) were synthesised
with varying contents of VPO (5–50 wt%). These solids were characterised by ICP-OES, TG/DTA, BET, XRD,
FTIR (Py-ads) and XPS. The catalytic activity was evaluated for ammoxidation of 3-picoline (3-pic) to
nicotinonitrile (NN) in a fixed bed catalytic reactor. Thermal analysis provided good hints on the phase
Received 26 February 2010
Received in revised form 17 June 2010
Accepted 19 July 2010
◦
transformation of VHP precursor into active VPP phase at around 400 C. BET surface areas and pore
2
2
volumes are found to depend on VPO loading and varied in the range from 70 m /g to 133 m /g. XRD
demonstrates the formation of (VO)2P2O7 (VPP) phase. XPS showed that an average oxidation state of
vanadium around 4.0, which is found to be unaltered in the spent samples. FTIR (Py-ads) revealed the
presence of both Lewis and Brønsted sites with varying proportions, which again depend upon VPO
loading. Correlation of acidic properties (Lewis and Brønsted acid sites) of the catalysts with that of
performance of catalysts was explored. VPO loading has a clear influence on the acidic properties and
thereby catalytic activity and selectivity. Catalytic results showed that the supported catalysts gave better
performance compared to bulk VPO. Yield of NN increased up to 20 wt% VPO loading and then decreased
with further increase in VPO content. Among all catalysts tested, the 20 wt% VPO/TiO2 exhibited the best
performance (X-3-pic = ca. 100% and Y-NN = 83%).
Dedicated to 75th birth anniversary of Prof.
Dr. H. Knözinger, University of Munich,
Germany.
Keywords:
Ammoxidation
3
-Picoline
Nicotinonitrile
VPO catalysts
Acidity characteristics
©
2010 Elsevier B.V. All rights reserved.
1
. Introduction
[5] that the number of deaths caused by niacin deficiency disease
i.e. pellagra) dropped in the U.S. from 7500 to just 70 in the years
(
Ammoxidation refers to the formation of nitriles by partial
from 1929 to 1956. This study clearly shows the importance of
niacin in our day to day life.
oxidation of hydrocarbons in the presence of ammonia [1]. Ammox-
idation of substituted aromatics/hetero aromatics opens the way
for the direct synthesis of fine chemicals and/or intermediates
for fine chemical syntheses [2,3]. Various pharmaceuticals, pesti-
cides, dyestuffs and other speciality products are produced from
substituted aromatic or heteroaromatic nitriles [e.g. [4]]. Ammox-
idation of alkyl pyridines is one such good example for producing
industrially important nitriles. Interestingly, the conversion of
methyl pyridines to their corresponding nitriles is relatively easier,
economic as well as eco-friendly process. Particularly, the ammox-
idation of 3-methylpyridine or 3-picoline (3-pic) to nicotinonitrile
Even though niacin is found naturally in various foodstuffs
(e.g. pork, yeast, wheat etc.), the majority of niacin today is
produced synthetically. The worldwide production amounts to
22,000 ton/a in 1995, which is significantly increased to about
35,000–40,000 ton/a in 2006 and the demand is further increas-
ing in recent times [6]. Lonza AG, Switzerland alone produces
>15,000 ton of niacin and is the world leader in niacin produc-
tion. The very classical method (discovered ca. 100 years ago) for
the production of niacin was based on oxidation of nicotine using
K Cr O7. This process is however unattractive today due to sig-
2
2
(
NN) is frequently investigated mainly due to the commercial
importance of the nitrile for large-scale production of nicotinamide
niacinamide) and nicotinic acid (niacin/vitamin B3), which are
nificantly larger amounts of waste production (i.e. about 9 ton of
waste produced per 1 ton of niacin production). In the direction of
change of feedstocks, attempts were also made to use 3-picoline as
suitable starting material for the production of niacin by means of
liquid phase oxidation with permanganate, chromic acid or nitric
acid. These processes also suffer similar problem of larger waste
(e.g. permanganate process gives 4 ton of waste while chromic acid
gives ca. 2 ton of waste per 1 ton of niacin production) [6]. Another
alternative to produce niacin is oxidation or ammoxidation of 2-
methyl-5-ethyl pyridine (MEP). Ammoxidation of MEP to NN is
being practiced at Lonza AG for the past 40 years [5]. Although this
(
used extensively as feed additives [e.g. [5]]. It is known that human
body cannot produce directly either nicotinic acid or nicotinamide
on its own and hence it is very much dependent on the intake of
these components through foodstuffs. It is also reported elsewhere
∗ Corresponding author. Tel.: +49 381 1281284; fax: +49 381 128151284.
E-mail address: narayana.kalevaru@catalysis.de (V.N. Kalevaru).
0
926-860X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2010.07.034

V.N. Kalevaru et al. / Applied Catalysis A: General 391 (2011) 52–62
53
Table 1
raw material is cheaper, the process is not really very attractive
from ecological point of view due to less carbon efficiency owing to
List of different VPO catalysts prepared and used in the ammoxidation of 3-picoline
to nicotinonitrile.
loss of two carbon atoms as CO during its conversion. On the whole,
2
S. No.
Catalyst
Denotation
it is generally agreed that among different methods available for
niacin/niacinamide production, the 3-picoline is indeed an ideal
feedstock from chemical standpoint and the vapour phase ammox-
idation is the most preferred route. Ammoxidation of 3-picoline to
NN followed by hydrolysis are at present existing commercial pro-
cesses, e.g. commercialized by Reilly Industries, U.S. and Lonza AG,
Switzerland.
Numerous catalytic systems have been developed and applied
for the ammoxidation of alkyl aromatics. Generally, the most
active and selective catalyst systems used for different ammox-
idation reactions are based on V–Ti–O [7], V–Mo–O [8], V–Sb–O
1
2
3
4
5
6
7
8
Bulk VPO
A
B
C
D
E
F
G
H
5 wt% VPO/TiO2
10 wt% VPO/TiO2
15 wt% VPO/TiO2
20 wt% VPO/TiO2
25 wt% VPO/TiO2
50 wt% VPO/TiO2
TiO2
◦
ethanol. The solid precursor thus obtained was oven dried at 120 C
for 24 h.
Preparation of TiO2 supported VPO catalysts (2nd step): Desired
amounts of VHP precursor and TiO (anatase) powders were mixed
together in a mortar by means of physical mixing for about 15 min,
i.e. until the colour of solid mixture was perfectly uniform. Using
the same procedure, various catalysts over a wide range of VPO con-
[
9], VOX/Nb O5 [10], VOX/Al O [11], VOX/ZrO2 [12], Fe–Sb–O pro-
2 2 3
moted by further transition metals [13] and also some zeolite based
catalysts [14] are reported to be active and selective. However, the
majority of catalysts employed for different ammoxidation reac-
tions contain vanadium as the key component. Favoured supports
are titanium oxide (anatase) [15], zirconium oxide [16], tin oxide
2
[
17] mixed supports such as titanium–tin oxide [18]. In recent
◦
tents were prepared. Finally the catalysts were calcined at 450 C,
times, metal fluorides are also used as effective supports for vanadia
catalysts [19–21].
3
h, 1% O /N . VPO loading was varied from 5 wt% to 50 wt% and the
2 2
P/V ratio was kept constant at 0.95. The list of catalysts prepared and
their denotation is given in Table 1. It should be noted that uncal-
cined TiO2 support was used as it is in preparing various catalysts.
However, the weight loss of TiO2 during calcination is consid-
ered in calculations and accordingly the weight of TiO2 is taken.
In addition, one should also note that the TiO2 support obtained is
a commercial sample supplied by Kronos, Germany, which is pre-
pared using sulphate precursor and hence this support material
contains certain amounts of residual sulphates in it. Removal of
such sulphates is also observed during TG/DTA measurements and
is also discussed below in a separate section. The detailed proce-
dure for the preparation of these catalysts is described elsewhere
Furthermore, polyoxometallates [22] and certain metal oxides
such as vanadium phosphorous oxides [e.g. [23]], vanadium oxide
combined with antimony oxide [24,25] are also used as effective
catalytic systems but they are applied mainly in the form of bulk
materials. In most cases, VPOs are used in their bulk form mainly
for butane to maleic anhydride reaction [e.g. [26]]. On the other
hand, investigations on the synthesis and application of supported
VPO catalysts are rare, probably due to the difficulty of their syn-
thesis because of the insoluble nature of VPO precursors in known
solvents. Nonetheless, preparation of supported VPOs by means of
solid–solid wetting is possible, which seems to be a suitable option.
With this background, it is intended to extend such synthesis pro-
cedure further to various VPO solids and check its influence on the
ammoxidation activity. Thus, the goal of the present investigation is
to synthesize, characterize and evaluate the catalytic performance
[
27,28].
2.2. Characterization of catalysts
of TiO supported VPO catalysts towards selective ammoxidation of
2
3
-picoline to nicotinonitrile. The objective is also to develop highly
ICP-OES (Optima 3000XL, Perkin-Elmer) using a microwave
pressure digestion (MDS 200; CEM) with hydrofluoric acid and
aquaregia at 9 bar was used to analyse the chemical composition of
fresh and spent catalysts. All the samples were analysed twice and
the results presented here are the average values.
active and selective catalysts in the direction of maximizing the
yield of nicotinonitrile due its high industrial importance.
2
. Experimental
Thermogravimetric analysis (TGA) was carried out using simul-
taneous TG and DTA instrument (Netzsch GmbH, model: STA 409).
About 20 mg of the powdered sample was placed in a platinum
sample holder of 8 mm dia and 10 mm length. The experiments
2.1. Catalyst preparation
◦
Preparation of supported VPOs involves two steps. First step
were carried out at a uniform heating rate of 10 C/min from room
◦
deals with the preparation of VPO precursor (VOHPO ·0.5 H O
temperature to 900 C in 1% O /N atmosphere and at a flow rate of
4
2
2
2
(
VHP)) by an organic route, while the 2nd step concerns the
50 ml/min. In addition, thermal analysis coupled with mass spec-
trometer were also carried out using Setaram-Sensys instrument
preparation of TiO2 (anatase) supported VPO catalysts using the
precursor prepared in the 1st step. The samples were then oven
dried and calcined under appropriate conditions to obtain active
◦
◦
in the temperature range from 20 C to 820 C in 1% O /N atmo-
2
2
sphere. The typical ion current curves of the different samples are
shown in the figures for comparison. The purpose is to determine
various components evolved during such thermal analysis in the
present VPO catalysts.
(
VO) P O7 phase. The details of the two-step preparation proce-
2 2
dure are described below.
Preparation of VPO precursor (1st step): 52.5 g sample of V O5
2
(
UH: 2862; Gesellschaft für Elektrochemie, Nürnberg) was sus-
The surface areas (BET-SA) and pore size distribution of the cat-
pended in a mixture of 315 ml of 2-butanol (purity > 99%, Merck)
and 210 ml of benzyl alcohol (purity > 99%, Merck) at room tem-
perature. The suspension was stirred continuously under reflux for
alysts were determined on Gemini III – 2375 (Micromeritics, U.S.)
◦
by N -physisorption at −196 C. Prior to the measurements, the
2
◦
known amount of catalyst was evacuated for 2 h at 150 C to remove
3
h, then cooled to room temperature and kept under stirring at
physically adsorbed water.
room temperature for overnight. Then 63.2 g of 85% o-H PO4 was
The X-ray diffraction (XRD) patterns were obtained on a X-ray
3
slowly added, and the solution was again heated and maintained
under reflux with constant stirring for another 2 h. The resulting
slurry was cooled to room temperature, filtered, and washed with
diffractometer STADI-P (Stoe, Darmstadt, Germany) using Ni-
filtered CuK␣ radiation (ꢀ = 1.5418 A˚ ). The crystalline phases were
identified by referring to the ASTM data files.

5
4
V.N. Kalevaru et al. / Applied Catalysis A: General 391 (2011) 52–62
Fig. 1. TGA and DTA curves of bulk and TiO2 supported VPO catalysts with different VPO contents (A, bulk VPO; B, 5% VPO/TiO2; E, 20% VPO/TiO2; H, pure TiO2).
For the characterization of surface acidity, pyridine was
3. Results and discussion
adsorbed on the activated catalysts at room temperature until satu-
ration. However, before adsorption experiments the catalysts were
activated in airflow at 300 C for 30 min to remove moisture from
3.1. Elemental analysis by ICP-OES
◦
the samples and after such activation the samples were cooled
down to room temperature. Spectra were recorded using a Bruker
After preparing the catalysts, the actual contents of all cata-
lyst components were estimated using ICP-OES technique. Such
ICP analysis results of various VPO/TiO2 catalysts are shown in
Table 2. The contents of vanadium, phosphorous and titanium
estimated from ICP are in good agreement with those of nomi-
nal values. In addition, the contents of all these components were
also checked again in the spent samples and found that they are
quite comparable with those of fresh samples. This fact suggests
that there is no loss of catalyst components during the course
of the reaction. A slight loss of phosphorous could be noticed
particularly in the spent 5% VPO/TiO2 sample, which is however
negligible.
−1
IFS 66 spectrometer (2 cm resolution, 100 scans) equipped with
a heatable and evacuable IR cell with CaF2 windows, which is con-
nected to a gas dosing/evacuation system. For these experiments,
the catalyst powder was pressed into self-supporting discs (50 mg,
Ø 20 mm). Difference spectra were obtained by subtracting the
spectrum of the activated catalyst at room temperature.
The X-ray photoelectron spectroscopic (XPS) measurements
were done with a VG ESCALAB 220iXL unit using MgK␣ radiation
(
E = 1253.6 eV) at a base pressure of the UHV chamber.
3.2. Thermal analysis
2.3. Catalytic tests
The thermogravimetric analysis (TGA) coupled with mass spec-
Ammoxidation runs were performed in a fixed bed, tubular
trometer of some selected VPO/TiO2 catalysts, bulk VPO and pure
support are presented in Fig. 1(a). It is evident from Fig. 1(a) that
there are three distinct stages of weight loss of catalysts. These
glass reactor at atmospheric pressure. In a typical experiment, 3 g
of catalyst particles (1–1.25 mm size) mixed with equal amount
of glass beads (1.7–2.0 mm) were loaded into the glass reactor
◦
◦
◦
are (i) from room temperature to 200 C (ii) from 200 C to 500 C
(
300 mm long and 12 mm i.d.) and heated in an electrical furnace.
◦
◦
(
iii) above 500 C. The weight loss of the catalysts below 200 C is
Air, NH3 and N2 (for dilution) supplied were commercially avail-
able gases from compressed gas cylinders. The flow rates of these
gases were controlled using mass flow controllers. The 3-picoline
and water were injected by means of HPLC pumps (LC-9A HPLC,
Shimadzu) and were vaporized in a preheating zone provided on
the top of the catalyst bed. The reaction was carried out in the range
clearly due to desorption of physically adsorbed water. It can also
be seen from the figure that the bulk VPO contains small amount
of physically adsorbed water (<2%) compared to pure titania and
also the catalysts prepared using titania. Such low water content
in bulk VPO is expected because the VPO precursor is prepared
◦
of 325–360 C and at a mole ratio of 3-pic:H O:NH :air = 1:9:6:30
2
3
(
3-pic = 10 mmol/h); cat: 3 g (4.0 ml); catalyst dilution (cat:glass
beads) = 1:2 (by wt.). Two thermocouples were positioned one at
the centre of the catalyst bed to indicate reaction temperature and
the other one was attached to furnace through temperature indi-
cator cum controller to monitor the temperature of the reactor.
The product stream was collected for every half-an-hour under
steady state conditions and analysed by gas chromatograph (GC
Table 2
Estimation of V, P, Ti contents from ICP-OES in VPO/TiO2 catalysts with different
VPO loadings.
Catalyst
V
P
Ti
Fresh
Used
Fresh
Used
Fresh
Used
5
1
1
% VPO/TiO2
0% VPO/TiO2
5% VPO/TiO2
1.6
3.0
4.7
6.4
8.6
1.4
2.9
4.7
6.4
8.9
0.9
1.6
2.6
3.8
5.3
0.7
1.5
2.7
3.8
5.3
55.4
48.1
45.4
42.0
42.6
52.6
46.8
45.6
42.0
40.0
1
7A, Shimadzu, Japan) equipped with FID module using a capillary
column (FS-SE-54). The total oxidation products (COx) were con-
tinuously monitored by on line non-dispersive infrared analyser
(
20% VPO/TiO2
5% VPO/TiO2
2
Infralyt 40 E, Germany).

V.N. Kalevaru et al. / Applied Catalysis A: General 391 (2011) 52–62
55
Table 3
through an organic route and hence the water content in such sam-
Surface areas and pore volumes of fresh and spent VPO/TiO2 catalysts.
ple would be naturally less. On the other hand, pure TiO2 support
contains the maximum amount of water (i.e. 12–13%), which on
evolution at around 120 C exhibit an intense endothermic peak.
a
b
a
b
Catalyst
BET-SA
(m2/g)
BET-SA
(m2/g)
PV
PV
◦
(cm3/g)
(cm3/g)
The evolution of water at this temperature is also further con-
firmed by mass spectrometer, which is coupled with TG analysis
Pure TiO2 (anatase)
5% VPO/TiO2
144
133
111
92
70
24
123
108
–
78
50
12
0.261
0.230
0.206
0.175
0.163
0.071
0.296
0.271
–
0.230
0.111
0.030
1
2
5
0% VPO/TiO2
0% VPO/TiO2
0% VPO/TiO2
(
see Fig. 1(a)). In addition, the amount of water that is lost in this
temperature range is decreasing with increasing VPO loading and
hence the intensity of the endothermic peaks of the catalysts is also
decreasing.
Bulk VPO
PV, pore volume.
◦
The second stage of weight loss occurred at around 400 C,
a
Fresh catalysts.
Spent catalysts.
where the second endothermic effect can clearly be seen. This effect
is undoubtedly due to the phase transformation of VHP into VPP and
also partly to the evolution of alcohols trapped in the layers of VHP
precursor. This is possible because benzyl alcohol and 2-butanol are
used as reducing agents during preparation of catalyst samples and
hence certain amounts of such alcohols are trapped in the layered
structure of VHP precursor even after oven drying. The alcohols
trapped in the layers have to go out during thermal treatment in
the form of CO2 as long as the oxygen is present in the calcinating
atmosphere. However, O2 concentration in the calcinating atmo-
sphere is deliberately limited to avoid over oxidation of vanadium
species to V+ oxidation state. On the other hand, if the calcination
is performed in an inert atmosphere (e.g. in N2 or He), then the
trapped alcohols during their evolution must use the lattice oxy-
gen of VPO structure and go out again in the form of CO2 or simply
coke and poison the catalyst, as a result the catalyst gets reduced.
b
3.3. BET surface areas and pore volumes
BET surface areas and pore volumes of fresh and spent VPO/TiO2
catalysts with different VPO loadings are given in Table 3. For the
sake of better comparison, the surface areas of the pure support
(TiO ) and bulk VPO are also presented in Table 3. The catalysts
2
showed a wide range of surface areas and pore volumes depend-
ing upon the content of VPO in the samples. The specific surface
areas of the supported VPO catalysts are appreciably lower than the
5
pure support. Pure TiO (anatase) has exhibited the highest surface
2
2
3
area (144 m /g) and pore volume (0.261 cm /g), while the pure VPO
2
3
showed the lowest (24 m /g and 0.071 cm /g). The surface area of
pure TiO2 is however decreased considerably after addition of VPO
[29]. VPO loading had a clear impact on the surface areas and pore
volumes of the samples. The surface areas are observed to decrease
+3
Such situation further leads to the generation of more V species,
which in turn reduces the catalytic activity. Therefore calcinating
atmosphere must be carefully chosen in order to obtain the desired
phase composition. The temperature required for such transforma-
tion again depends on the VPO content of the solid. Therefore, the
peak maxima during such transformation varied in the range from
2
2
from 133 m /g to 70 m /g with increase in VPO loading from 5 wt%
to 50 wt%. Spent catalysts showed considerably lower surface areas
compared to their corresponding fresh samples. Such decrease in
surface areas might be due to possible structural changes occurred
during the course of reaction under the influence of reactant feed
mixture and severe reaction conditions. Some hints on structural
changes inthe spentsamples are alsoobtainedfromchanges inpore
size distribution and FTIR analysis, which are however discussed
below separately.
◦
◦
3
66 C (5% VPO) to 417 C (pure VPO). Bulk VPO solid displayed the
◦
maximum weight loss at around 450 C and then remained more
or less constant. The theoretical weight loss for the transforma-
tion of VHP into VPP is around 10.5%. However, the VPOs showed
slightly higher weight loss than theoretical, which is undeniably
due to removal of additional trapped alcohols in the layers of pre-
cursor. As expected, pure TiO2 did not show any such effect in this
temperature range.
The pore volume distribution curves of fresh VPO/TiO catalysts
are depicted in Fig. 3(a). The pure support has a total pore vol-
ume of 0.261 cm /g and a total pore area of 144 m /g. The pore
size distribution of these catalysts was made by the application
of N2 adsorption method. The pure support has shown bi-modal
pore volume distributions in the pore diameter range from 25 A˚
to 60 A˚ . It has shown a very narrow volume distribution at 25 A˚ ,
and a broad and prominent distribution around 60 A˚ ; major part
of the contribution to the pore volume of the sample comes from
the pores located in this range [30]. However, after the addition of
2
◦
3
2
The third stage of weight loss is occurred at around 600 C, which
is purely due to removal of residual sulphates from the TiO2 sup-
port. The endothermic effect due to such removal is shown only
by the VPO/TiO2 catalysts and the support but not by the bulk
VPO solid due to absence of sulphates in it. The peak maxima of
temperature is again depended on the content of VPO and var-
◦
◦
ied in the range from 607 C to 631 C. The peak intensity appears
to be proportional to the content of TiO . In other words, the
2
higher the VPO content the lower the TiO2 content and hence the
lower the intensity of such endothermic effect. Interestingly, all
the effects observed from these samples are only endothermic and
no exothermic effects are noticed. It is worthwhile to mention
VPO to TiO2 support, the narrow pore volume distribution, which
originally appeared at around 25 A˚ in pure support is disappeared,
as a result the surface area decreased. This observation points to
the fact that the addition of VPO has an influence on the pore struc-
ture of resulting catalysts. All the supported VPO catalysts showed
uni-modal pore volume distribution. The intensity of pore volume
distribution curve is decreasing with increasing in VPO content of
the solids. This result is in line with the changes in surface areas.
The pore size distribution patterns of spent VPO/TiO2 solids are
illustrated in Fig. 3(b). It is evident that the pore size distribution
exhibited by the spent samples is considerably different from their
corresponding fresh samples. In addition, an appreciable shift in
the pore diameter towards higher pore range is also noticed irre-
spective of VPO content. Some pores are vanished and some new
pores are also formed. This observation clearly suggests that the
reaction conditions have brought about drastic changes in their
pore structures and thereby surface areas and total pore volumes
here that the removal of H O, alcohols and sulphates at differ-
2
ent stages/temperatures is also confirmed by mass spectra coupled
with TG analysis. The mass spectra obtained from such analysis
are shown in Fig. 2. Curves (i)–(vi) of Fig. 2 illustrate the evolu-
tion of various components such as H O, benzyl alcohol, ethanol,
2
2
-butanol, CO2 and SO . It can also be seen from Fig. 2 (curve-i)
2
that the evolution of H O is occurring in two different stages, i.e. at
2
◦ ◦
200 C and at ca. 450 C. As discussed above, the former one is due
<
to removal of physically adsorbed water and the latter one is due to
phase transformation from VHP to VPP. The results of TG–MS fur-
ther confirm the evolution of various other components at different
temperatures, which however depend upon the composition of the
catalysts.

5
6
V.N. Kalevaru et al. / Applied Catalysis A: General 391 (2011) 52–62
Fig. 2. Mass spectra obtained during thermal analysis over bulk and TiO2 supported VPO catalysts (A, bulk VPO; B, 5% VPO/TiO2; E, 20% VPO/TiO2; H, pure TiO2).
(
see Fig. 3(b)). Interestingly, pure titania support after the catalytic
3.4. XRD and FTIR
tests displayed tri-modal pore volume distribution curve against
bi-modal distribution in the fresh sample. The pores are located
in the pore diameter range of ca. 40 A˚ , a narrow distribution at
8
spent VPO/TiO2 solids also showed similar such tri-modal distri-
bution like support but with some variations in the intensity of
pore volume distribution curves.
XRD patterns of fresh samples confirm the formation of VPP
phase in the catalysts. However, the crystalline VPP phase can
only be seen in the samples with high VPO loading (>15 wt% VPO).
This means the VPP phase present in the low loading VPO cat-
alysts seems to be X-ray amorphous in nature. However, FTIR
spectra of fresh solids provided good hints on the formation of VPP
5 A˚ and a prominent broad distribution at 155 A˚ . Furthermore,

V.N. Kalevaru et al. / Applied Catalysis A: General 391 (2011) 52–62
57
Fig. 3. (a) Pore size distribution of fresh bulk and TiO2 supported VPO catalysts (A,
bulk VPO; B, 5% VPO/TiO2; D, 15% VPO/TiO2; E, 20% VPO/TiO2; G, 50% VPO/TiO2; H,
pure TiO2). (b) Pore size distribution of spent bulk and TiO2 supported VPO catalysts
(
A, bulk VPO; B, 5% VPO/TiO2; D, 15% VPO/TiO2; E, 20% VPO/TiO2; G, 50% VPO/TiO2;
H, pure TiO2).
phase even in low VPO-containing solids. No significant changes in
the XRD patterns of spent samples compared to fresh ones could
be noticed. The intensity of XRD reflections corresponds to VPP
phase is increasing with increase in VPO contents, as expected.
XRD patterns of both fresh and spent samples can be viewed from
supplementary material.
FTIR spectra of fresh and spent bulk and supported VPO solids
are given in Fig. 4(a) and (b). FTIR spectra further confirm the for-
−
1
mation of VPP phase in these solids. The band appeared at 634 cm
−1
corresponds to ␦-OPO and the one at 744 cm is due to ꢁs P–O–P
−1
can be attributed to ꢁ V+4
O
vibrations. The band at 968 cm
species, which provides hints on the formation of VPP. Other bands
−1
−1
−1
appeared at 1095 cm , 1142 cm and 1241 cm are assigned to
Fig. 4. (a) FTIR spectra of fresh bulk and TiO2 supported VPO catalysts (A, bulk VPO;
B, 5% VPO/TiO2; D, 15% VPO/TiO2; F, 25% VPO/TiO2; H, pure TiO2). (b) FTIR spectra of
spent bulk and TiO2 supported VPO catalysts (A, bulk VPO; B, 5% VPO/TiO2; D, 15%
VPO/TiO2; E, 20% VPO/TiO2; F, 25% VPO/TiO2; H, pure TiO2).
symmetric (ꢁs PO ) and asymmetric (ꢁas PO ) stretching vibrations
3
3
of phosphate groups (Fig. 4(a)). On the whole, FTIR gives good sup-
porting evidence to the results obtained from XRD on the formation
of active VPP phase.
There is no much difference in the FTIR spectra of the spent
samples compared to fresh ones except a new band appeared at
(i.e. more likely (NH ) (VO) (P O7) ) in addition to the major VPP
4
2
3
2
2
phase [31]. Formation of such phase is expected from the reaction
of vanadium species with that of ammonia during the course of the
reaction.
−1
1
420 cm
(Fig. 4(b)). This means FTIR spectra of spent samples
gave further hints on the formation of NH -containing VPO phase
4

5
8
V.N. Kalevaru et al. / Applied Catalysis A: General 391 (2011) 52–62
of supported catalysts designates a good and effective interaction
between the support and active phase.
3.6. Pyridine desorption measurements (FTIR)
After evaluating the influence of VPO loading on the concen-
tration of both LS and BS, further studies were focused on the
desorption behaviour of adsorbed pyridine over a wide range of
◦
◦
temperatures starting from 25 C to 400 C. For this purpose, three
different solids such as the lowest VPO loading (5% VPO), the best
VPO loading (20% VPO) and the pure TiO2 support were selected
and analysed. Comparison of pyridine desorption spectra of 5% and
2
0%VPO/TiO2 and pure TiO2 are portrayed in Fig. 6. The purpose
of this study is to check the strength of acid sites as a function of
temperature. All the three samples exhibited intense bands corre-
spond to Lewis and Brønsted sites. However, the intensity of the
bands is found to decrease gradually with increase in desorption
temperature. It is noteworthy that these bands are still present
◦
Fig. 5. Acidity measurements by FTIR pyridine adsorption at 25 C on bulk and
TiO2 supported VPO solids (A, bulk VPO; B, 5% VPO/TiO2; D, 15% VPO/TiO2; E, 20%
VPO/TiO2; F, 25% VPO/TiO2; H, pure TiO2).
◦
even at 400 C indicating their higher strength and good thermal
stability. From the intensity of the bands, it can be assumed that
the VPO catalyst with low loading (5%) is more acidic than the one
with high VPO loading (20%). The higher acidity of low VPO catalyst
seems to be arising from the support. Catalytic results showed that
too much acidity enhanced total oxidation that leads to the forma-
tion of higher amounts of COx and hence reduces the selectivity
of desired nitrile. As mentioned earlier, Fig. 6 gives good evidence
for the higher acidity of pure support and also the existence of
such acidic sites of support even at temperatures higher than the
reaction temperature.
3
.5. Acidity measurements by pyridine adsorption (FTIR)
For characterizing the surface acidity, pyridine used as a probe
molecule was adsorbed at room temperature. Usually, the bands
appeared at around 1540–1548 cm
characteristic bands of Brønsted (PyH+) and Lewis (L-Py) acid sites,
respectively. Furthermore, the bands correspond to hydrogen-
bonded pyridine (hb-Py) are also expected in the similar range
−1
−1
and 1445–1460 cm are
−1
−1
and 1590–1600 cm ) and the bands due to
(
1440–1447 cm
physically adsorbed pyridine (ph-Py) at 1439 cm and 1580 cm
e.g. [32,33]]. It should be noted that the intensity of the band is
−
1
−1
3.7. X-ray photoelectron spectroscopy (XPS)
[
proportional to the concentration of acid sites. The FTIR spectra
of present VPO/TiO2 solids obtained after pyridine adsorption are
illustrated in Fig. 5. The pyridine adsorption was carried out at two
XPS analysis provides information on the oxidation state of
vanadium and concentration of surface elements. XP spectra of
fresh VPO/TiO2 catalysts with varying VPO loadings are given in
Fig. 7(a) and (b). In Fig. 7(a), O 1s and V2p1/2 and V2p3/2 pho-
toelectron peaks are presented while in Fig. 7(b), P2p peaks are
shown. It is known from the literature [34,35], that the position
◦
different temperatures such as room temperature and at 200 C. At
−
1
both the temperatures, the catalysts showed bands at 1446 cm
−1
and 1608 cm that correspond to Lewis sites and the other band
−1
+5
appeared at 1540 cm is due to Brønsted sites. It can also be seen
from the figure that all catalysts contain both Lewis and Brønsted
sites in different proportions depending upon VPO content. It s also
obvious from Fig. 5 that the supported VPO catalysts exhibit intense
bands compared to their parent materials such as bulk VPO and
of V component in VPO solids lies at 518 eV while the binding
+4
energy for V would be 516.9 eV [36,37]. Due to such difference
in binding energy values of V+5 and V components, their con-
tributions can be well separated. On the other hand, if any V+3
species are present in the samples, then their contributions fur-
ther shift the position of the peak to lower binding energy values
and would exhibit a bigger FWHM than V+4 peak [34]. However,
+4
pure TiO . This means the supported VPOs are more acidic than the
2
parent samples. Such enhancement in the acidity characteristics
Fig. 6. Comparison of pyridine desorption spectra (FTIR) of 5% and 20% VPO/TiO2 solids and pure TiO2 support (B, 5% VPO/TiO2; E, 20% VPO/TiO2; H, pure TiO2).

V.N. Kalevaru et al. / Applied Catalysis A: General 391 (2011) 52–62
59
Table 4
Variation of P/Ti, V/Ti and P/V surface ratios (determined from XPS) in VPO/TiO2
catalysts with different VPO loadings.
Catalysts
P/Ti
V/Ti
P/V
5
1
2
% VPO/TiO2
5% VPO/TiO2
0% VPO/TiO2
0
0.17
0.26
0.49
–
–
0.27
0.43
–
1.0
0.9
1.3
Bulk VPO
3
3
.8. Catalytic results
.8.1. Influence of VPO loading on the catalytic performance
The influence of VPO loading on the catalytic performance of
VPO/TiO solids is carried out at two different temperatures such as
2
◦ ◦
3
25 C and 360 C and presented in Fig. 8(a) and (b). For better com-
parison, the catalytic results obtained on bulk VPO sample (i.e. 100%
VPO (un-supported)) are also included in both the figures. Fig. 8(a)
and (b) demonstrate that the supported VPO catalysts displayed
better performance than the pure VPO solid. In addition, the content
of VPO has shown a strong influence on the activity and selectiv-
Fig. 7. (a) XP spectra for fresh VPO/TiO2 catalysts with varying VPO loadings (show-
ing O 1s and V2p peaks) (A, bulk VPO; B, 5% VPO/TiO2; D, 15% VPO/TiO2; F, 25%
VPO/TiO2). (b) XP spectra for fresh VPO/TiO2 catalysts with varying VPO load-
ings (showing P2p peaks) (A, bulk VPO; B, 5% VPO/TiO2; D, 15% VPO/TiO2; F, 25%
VPO/TiO2).
one should note that V O5 based catalysts exhibit somewhat lower
2
binding energy values compared to VPO catalysts for similar oxi-
dation states of vanadium (e.g. V+4 and V ). Our samples showed
the binding energy values at around 517.1 eV that mainly corre-
spond to V+ species of active VPP phase in VPO/TiO2 solids. This
+5
4
+4
value is comparable with that of the one reported above for V oxi-
dation state in literature. No significant shift in the peak position
with increase in VPO loading and hence no considerable change
in the vanadium valence in all the catalysts. However, the inten-
+4
sity of peak corresponds to V species is increasing with increase
in VPO content of the samples, as expected. In addition, bulk VPO
solid displayed higher B.E. (531.2 eV) for O 1s peak than the sup-
ported VPOs (530.2 (± 1) eV). Such difference in B.E. value between
bulk and supported VPOs might be due to the presence of different
nature of oxygen present in the pure VPO sample and TiO2 sup-
ported VPOs. Fig. 7(b) shows that there is no considerable shift in
the P2p photoelectron peak position in all the samples and hence
no change in binding energy values, which remained more or less
constant at around 134.4 eV.
Additionally, variation of P/Ti, V/Ti and P/V surface ratios esti-
mated from XPS in VPO/TiO2 catalysts with different VPO loadings
are given in Table 4. The P/Ti and V/Ti ratios are observed to
increase with increase in VPO content of the samples. The sur-
face P/V ratios of supported VPOs are found to be comparable with
that of theoretical value (P/V = 0.95). However, an enrichment of
phosphorous at the surface can clearly be seen in case of bulk VPO
solid (P/V = 1.3). Enrichment of phosphorous in VPO solids is a com-
mon phenomenon and such enrichment helps stabilizing reduced
vanadium species and also limits the over oxidation [e.g. [38]].
Fig. 8. (a) Catalytic performance of bulk and supported VPO solids with different
VPO loadings at a temperature of 325 C (mole ratio of 3-pic:H2O:NH3:air = 1:9:6:30
◦
(
(
3-pic = 10 mmol/h); cat: 3 g (4.0 ml); dilution (cat:glass beads) = 1:2 (by wt.)).
b) Catalytic performance of bulk and supported VPO solids with different VPO
◦
loadings at a temperature of 360 C (mole ratio of 3-pic:H2O:NH3:air = 1:9:6:30
(
3-pic = 10 mmol/h); cat: 3 g (4.0 ml); dilution (cat:glass beads) = 1:2 (by wt.)).

6
0
V.N. Kalevaru et al. / Applied Catalysis A: General 391 (2011) 52–62
ity behaviour of the catalysts. In every case, the catalyst with the
lowest VPO loading displayed the highest conversion of 3-picoline
and vice versa. At low VPO contents, not only the conversion of 3-
picoline is high but also the total oxidation is also considerably high,
which in turn reduces the selectivity of desired product. It is evident
from Fig. 8(a) that the conversion of 3-picoline decreased from 81%
to 48% with increase in VPO loading from 5 wt% to 50 wt%. On the
other hand, bulk VPO showed very low conversion (X-3-pic = 27%)
compared to supported catalysts. NN is the major product, while
CO, CO2 are the by-products. However, pyridine and isonicotinoni-
trile are also observed in small amounts as side products depending
upon the catalyst applied. The highest selectivity of desired product
(
NN) is obtained on 20 wt% VPO/TiO2 catalyst (Fig. 8(b)). Interest-
ingly, the same tendency is also observed when the reaction is
◦
carried out even at 360 C. At this temperature, the catalyst with
the lowest VPO loading displayed almost total conversion of 3-pic
(
X = ca. 100%) but the selectivity of target product obtained is rela-
tively low (S-NN = 70%). The selectivity of NN is observed to increase
from 70% to 86% with rise in VPO content from 5 wt% to 20 wt% and
then decreased with further increase in VPO loading. The best opti-
mum seems to be 20 wt% VPO, where the highest conversion of
Fig. 9. Effect of temperature on the catalytic performance of 20% VPO/TiO2 catalyst
mole ratio of 3-pic:H2O:NH3:air = 1:9:6:30 (3-pic = 10 mmol/h); cat: 3 g (4.0 ml);
dilution (cat:glass beads) = 1:2 (by wt.)).
(
3
-picoline and yield of NN (X = 98% and Y = 83%) could be obtained
butane to MA. The authors claim that the use of silica enhanced
the selectivity of MA but reduces the activity. On the other hand,
titania improved the activity but reduces the selectivity. Such dif-
ferences in the catalytic properties are assigned to the nature of
interaction between active VPO phase and the support, which in
turn modifies the reducibility of the catalysts. They also found
a strong interaction with reducible supports such as titania and
zirconia, which remarkably enhances the activity due to higher
amount of active oxygen species. However, the reverse is true in
case of non-reducible supports (e.g. silica), where the interaction
of active phase with the support is relatively weak. In addition,
Bueno et al. [45] and Li et al. [46] also report the poor perfor-
mance of supported VPOs compared to bulk VPO in the oxidation
of butane. On the other hand, Nie et al. [42,47] observed a signifi-
cant improvement in MA selectivity when Al-MCM-41 supported
VPOs are used catalysts again for butane oxidation. Our approach
is completely different from all the above-mentioned reports in
many ways. For instance, our method of catalyst preparation, P/V
ratio of the catalyst, calcination conditions, calcination atmosphere,
nature of reaction etc. are totally different and hence our results
cannot be truly compared with that of literature reports where
supported VPOs are used. To the best of our knowledge, articles
on the application of supported VPOs for ammoxidation reactions
in general and 3-picoline in particular are very rare. The superior
performance of the present supported VPOs was also further con-
firmed by preparing similar type of VPO catalysts with the same
procedure and composition and applied them for another indus-
trially important ammoxidation reaction such as ammoxidation
of 2,6-dichlorotoluene to 2,6-dichlorobenzonitrile and found again
improved performance of supported VPOs compared to their parent
bulk VPO. These results on the superior performance of supported
VPOs are also reported in various publications [27,28,30,31,48].
(
Fig. 8(b)). Such increase in selectivity of NN can be attributed to the
changes in acidic characteristics of the catalysts when VPO contents
are changed. The correlation of acidity against activity is however
discussed below in a separate section. It appears that when the VPO
loading is increased beyond a certain level, the catalytic behaviour
of the catalysts seems to be moving in the direction of bulk VPO
and thereby reduced activity of the catalysts. At a reaction tem-
◦
perature of 360 C, the bulk VPO solid gave only 48% conversion of
3
-picoline with 83% selectivity of NN. Among all catalysts tested,
bulk VPO showed the poorest performance in terms of yield of NN.
At low VPO loadings, the formation of total oxidation products is
significantly much higher (Y-COx = >25%) compared to higher VPO
loadings. This is probably due to exposure of bare acidic surface of
the support, which favours the total oxidation. Interestingly, the
Y-COx is decreased from over 25% (5 wt% VPO) to around 10% (on
2
0 wt% VPO) with simultaneous increase in NN selectivity. In other
words, NN selectivity is increasing at the expense of COx forma-
tion with rise in VPO loading to a better optimum. Moreover, the
increase in VPO loading beyond 20 wt% has again caused detri-
mental effect on the conversion of 3-picoline, which is decreased
considerably from 100% (at 5 wt%) to 48% when pure VPO is applied.
This fact suggests that an increase in VPO loading beyond 20 wt%
is not at all effective in improving the performance. On the whole,
it can be stated that 20 wt% VPO is found to exhibit better perfor-
mance in terms of both activity and selectivity compared to others.
Our results clearly showed that the performance of supported
VPOs is much superior to their corresponding bulk VPO. It is also
not very clear from the literature on the potential application of
supported VPOs and their influence on the catalytic activity and
selectivity. Some groups reported that the supported VPOs show
inferior performance compared to bulk VPO while the other claim
the reverse. However, such investigations are mainly devoted to
study the oxidation of butane to maleic anhydride (MA) but not
for ammoxidation reactions, in general. Some researchers [39–41]
propose that the usage of support can result in support–oxide inter-
action, which in turn either hinder the formation of VPP phase or
bring about some modifications in phase composition. As a result,
the supported VPO catalysts might show poor performance com-
pared to bulk VPO. In contrast, Nie et al. [42] suggest that the
supported VPOs offer various potential advantages over the unsup-
ported ones such as better heat transfer characteristics, high surface
area, better mechanical strength, and controllable catalyst textures.
Overbeek et al. [39,43,44] prepared VPO catalysts using vari-
ous supports and tested their performance towards oxidation of
3.8.2. Effect of temperature on the catalytic performance of
20%VPO/TiO catalyst
2
Influence of reaction temperature on the activity and selectiv-
ity behaviour of VPO/TiO2 catalysts is shown in Fig. 9. It is evident
from the figure that the temperature has a highly pronounced pro-
motional effect on the selective ammoxidation of 3-picoline over
20%VPO/TiO2 (anatase) catalyst. The conversion of 3-picoline is
observed to increase continuously from 38% to almost 100% with
◦
◦
rise in temperature from 300 C to 360 C. The selectivity of NN is
also increased initially and then remained more or less constant
at around 80%. At the same time, the selectivity of total oxidation

V.N. Kalevaru et al. / Applied Catalysis A: General 391 (2011) 52–62
61
lowed by H-abstraction. In conclusion, it can be said that the acidity
properties are important and the concentration of both LS and BS
obtained on 20% VPO solid seems to be optimum in achieving supe-
rior performance.
4. Conclusions
XRD and FTIR confirm the formation of vanadyl pyrophosphate
phase in the calcined catalysts. BET surface areas and pore volumes
are found to depend on the content of VPO in the total catalyst.
Spent catalysts showed different pore volume distributions com-
pared to their corresponding fresh ones. The present VPO/TiO2
catalysts contain both BS and LS in different proportions. The acidic
site distribution is also observed to depend on the concentration of
VPO in the system. Pyridine desorption studies indicate that the
acidic sites present in the catalysts are quite strong and remain
◦
stable even at 400 C. VPO loading on titania has shown a strong
influence on the catalytic performance. Supported VPOs exhibited
superior performance compared to bulk VPO. Acidity characteris-
tics appear to play a key role on the catalytic properties. Among
different catalysts tested, 20 wt% VPO/TiO2 has displayed the best
performance in terms of higher yield of nitrile compared to others.
This catalyst gave a conversion of 3-picoline close to 100% and the
yield of NN is over 80%.
Fig. 10. Correlation of acidity per unit area with that of catalytic activity and selec-
tivity of VPO/TiO2 catalysts with varying VPO loadings.
products is also influenced considerably by increase in reaction
temperature. The selectivity of CO decreased initially and then
remained stable, while the selectivity of CO2 has been increased
continuously with rise in temperature. The maximum yield of
◦
Acknowledgements
nicotinonitrile and conversion of 3-picoline observed at 360 C is
8
3% and 98%, respectively. At the same time, the yield of CO is also
2
◦
Thanks are due to Dr. U. Bentrup, LIKAT for FTIR measurements
and Dr. J. Radnik, LIKAT for XPS measurements.
increased from 2% to 10% with increase in temperature from 300 C
to 360 C.
◦
Appendix A. Supplementary data
3
.8.3. Correlation of acidity with that of activity over VPO/TiO2
catalysts
The correlation of activity and selectivity behaviour of different
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.apcata.2010.07.034.
VPO/TiO2 solids with that of acidity characteristics is illustrated
in Fig. 10. It is evident that the amount of VPO in the total cata-
lyst has a clear influence on LS and BS integral intensity per unit
surface area of the samples. The integral intensity values of BS are
decreasing with increasing in VPO contents of the catalysts. On the
other hand, somewhat different trend is observed concerning the
integral intensity values of LS, which are increasing up to 20 wt%
VPO and then decrease. It is also noteworthy that the low VPO
loading catalysts contain more BS than LS. Interestingly the ratio
of LS to BS is also increased up to 20 wt% VPO loading and then
decreased. Alternatively, if we carefully observe the catalytic per-
formance of these solids, the best performance is obtained over
References
[
1] V.N. Kalevaru, B. David Raju, S.K. Masthan, V.V. Rao, P. Kanta Rao, R. Subrahma-
nian, A. Martin, J. Mol. Catal. A: Chem. 223 (2004) 321–328.
[2] A. Martin, B. Lücke, Catal. Today 57 (2000) 61–70.
[3] K.V. Narayana, B. David Raju, S.K. Khaja Masthan, Catal. Lett. 84 (2002) 27–30.
[
4] D.K. Lee, T.S. Chang, D.H. Cho, S.B. Hong, C.H. Shin, Kor. J. Chem. Eng. 20 (2003)
79–283.
2
[5] R. Chuck, Appl. Catal. A: Gen. 280 (2005) 75–82.
[6] R. Chuck, A.G. Lonza, Proceedings of 1st Int. IUPAC Conf. on Green Sustainable
Chemistry (2006 September).
[
[
7] M. Sanati, A. Andersson, J. Mol. Catal. A 59 (1990) 233–255.
8] W.S. Chen, H.B. Chiany, M.D. Lee, J. Chin, Chem. Eng. 25 (1994) 45.
2
0 wt% VPO/TiO2 solid. Comparing these two aspects such as best
[9] A. Andersson, S.L.T. Andersson, G. Centi, R.K. Grasselli, M. Sanati, F. Trifiro, Appl.
Catal. A: Gen. 43 (1994) 43–57.
10] K.V.R. Chary, Ch. Praveen Kumar, A. Murali, A. Tripathi, A. Clearfield, J. Mol.
Catal. A: Chem. 216 (2004) 139–146.
performance vs acidity characteristics, it seems more probable that
the acidity of the catalysts control the activity and selectivity to a
greater extent. However, one should also note that too much acid-
ity also leads to high total oxidation and at the same time too less
acidity reduces the activity, as has been evidenced from the present
catalytic results. Comparing these two types of acid sites (BS and
LS), our results gave indications that LS appears to be more impor-
tant than BS for better selectivity and thereby better yield of nitrile.
If we check the concentration of these two types of acid sites on our
best catalyst (20 wt% VPO), LS is slightly higher than BS. In all other
samples, the integral intensity values of BS are always higher than
LS. It is also noticed that wherever BS is dominating LS, the con-
version is high but the selectivity is low. In contrast, when LS is
dominating BS, the selectivity of desired product is high and at the
same time the conversion is also maintained high as a result there
is a considerable increase in the net yield of target product. It is
also known that LS are necessary for initial adsorption of aromatic
substrate on the catalyst surface and then forming ␲-complexation
with the catalyst surface through ␲-electrons of aromatic ring fol-
[
[11] M. Niwa, H. Ando, H. Murakami, J. Catal. 49 (1977) 92–96.
[
12] K.V.R. Chary, G. Kishan, K.V. Narayana, T. Bhaskar, J. Chem. Res. (S) (1998)
14–315.
13] M.D. Allen, G.-J. Hutchings, M. Bowker, Appl. Catal. A: Gen. 217 (2001) 33–39.
3
[
[14] K. Beschmann, S. Fuchs, T. Hahn, Chem. Eng. Technnol. 70 (1998) 1436–1439.
15] K.V. Narayana, A. Venugopal, K.S.R. Rao, V.V. Rao, S.K. Masthan, P.K. Rao, Appl.
Catal. A: Gen. 150 (1997) 269–278.
16] M. Sanati, A. Andersson, L.R. Wallenberg, B. Rebenstorf, Appl. Catal. A: Gen. 196
(1993) 51–72.
[
[
[17] A. Andersson, J. Catal. 69 (1981) 465–474.
18] J.C. Otamiri, S.L.T. Andersson, A. Andersson, Appl. Catal. A: Gen. 65 (1990)
59–174.
[
1
[19] K.V. Narayana, S.K. Masthan, V. Venkat Rao, P. Kanta Rao, J. Chem. Res. (S) (1997)
328–329.
20] K.V. Narayana, S.K. Masthan, V. Venkat Rao, B. David Raju, P. Kanta Rao, Catal.
Commun. 3 (2002) 173–178.
21] K.V. Narayana, S.K. Masthan, V. Venkat Rao, B. David Raju, P. Kanta Rao, Catal.
Lett. 84 (2002) 27–30.
22] K.M. Reddy, N. Lingaiah, P. Nagaraju, P.S. Sai Prasad, I. Suryanarayana, Catal.
Lett. 122 (2008) 314–320.
[
[
[
[
23] Y. Zhang, A. Martin, H. Berndt, B. Lücke, M. Meisel, J. Mol. Catal. A: Chem. 118
(1997) 205–214.

6
2
V.N. Kalevaru et al. / Applied Catalysis A: General 391 (2011) 52–62
[
[
24] G. Centi, P. Mazzoli, S. Perathoner, Appl. Catal. A: Gen. 165 (1997) 273– 290.
25] M.O. Guerrero-Pérez, J.L. Rivas-Cortés, J.A. Delgado-Oyague, J.L.G. Fierro, M.A.
Ba n˜ ares, Catal. Today 139 (2008) 202–208.
[37] L. Rihko-Struckmann, Y. Ye, L. Chalakov, Y. Suchorski, H. Weiss, K. Sundmacher,
Catal. Lett. 109 (2005) 89.
[38] J. Guan, H. Xu, S. Jing, S. Wu, Y. Ma, Y. Shao, Q. Kan, Catal. Commun. 10 (2008)
276–280.
[39] R.A. Overbeek, A.R.C.J. Pekelharing, A.J. van Dillen, J.W. Geus, Appl. Catal. A: Gen.
135 (1996) 231–248.
[40] N.T. Do, M. Baerns, Appl. Catal. 45 (1988) 1–7.
[41] N.T. Do, M. Baerns, Appl. Catal. 45 (1988) 9–26.
[42] W.Y. Nie, X.S. Wang, W.J. Ji, Q.J. Yan, Y. Chen, C.T. Au, Catal. Lett. 76 (2001)
201–206.
[
[
26] M.W. Forkner, WO 2008030714 A1, Huntsman Petrochemical Corporation, USA
(
2008).
27] A. Martin, K.V. Narayana, B. Lücke, Van Deynse Dirk, M. Belmans, F. Boers, WO
3/101939 A2, Tessenderlo Chemie SA, Belgium (2003).
0
[
[
28] N. Dropka, V.N. Kalevaru, A. Martin, D. Linke, B. Lücke, J. Catal. 240 (2006) 8–17.
29] K.V. Narayana, A. Martin, U. Bentrup, B. Lücke, J. Sans, Appl. Catal. A: Gen. 270
(
2004) 57–64.
[
[
30] V.N. Kalevaru, B. Luecke, A. Martin, Catal. Today 142 (2009) 158–164.
31] K.V. Narayana, A. Martin, B. Luecke, M. Belmans, F. Boers, Z. Anorg. Allg. Chem.
[43] R.A. Overbeek, P.R. Warringa, M.J.D. Crombag, L.M. Visser, A.J. van Dillen, J.W.
Geus, Appl. Catal. A 135 (1996) 209–230.
6
31 (2005) 25–30.
[44] R.A. Overbeek, E.J. Bosma, D.W.H. de Blauw, A.J. van Dillen, H.G. Bruil, J.W. Geus,
Appl. Catal. A 163 (1997) 19–144.
[
32] R. Buzzoni, S. Bordiga, G. Ricchiardi, C. Lamberti, A. Zecchina, G. Bellussi, Lang-
muir 12 (1996) 930.
33] G. Busca, Phys. Chem. Chem. Phys. 1 (1999) 723.
[45] J.M.C. Bueno, G.K. Bethke, M.C. Kung, H.H. Kung, Catal. Today 43 (1998)
101–110.
[
[
34] Y. Suchorski, B. Munder, S. Becker, L. Rihko-Struckmann, K. Sundmacher, H.
Weiss, Appl. Surf. Sci. 253 (2007) 5904–5909.
[46] X.-K. Li, W.-J. Ji, J. Zhao, Z.-B. Zhing, C.-T. Au, J. Catal. 238 (2006) 232–
241.
[
[
35] S. Albonetti, F. Cavani, F. Triffiro, P. Venturoli, G. Calestani, M. Lopez Granados,
J.L. Fierro, J. Catal. 160 (1996) 52.
36] Y. Suchorski, L. Rihko-Struckmann, F. Klose, Y. Ye, M. Alandjiyska, K. Sund-
macher, H. Weiss, Appl. Surf. Sci. 249 (2005) 231.
[47] W.Y. Nie, Z.Y. Wang, W.J. Ji, Y. Chen, C.T. Au, Appl. Catal. A: Gen. 244 (2003)
265–272.
[48] A. Martin, V.N. Kalevaru, Q. Smejkal, Catal. Today (in press)
doi:10.1016/j.cattod.2010.01.037.