The influence of support on ammoxidation of 3-picoline over vanadia catalyst

Article abstract of DOI:10.1016/j.molcata.2004.04.046

The influence of high content of MoO₃ (32 mol%) in V ₂O₅-MoO₃-P₂O₅ catalyst system supported on alumina, silica, HZSM-5 and modified clay was investigated for ammoxidation of 3-picoline. Un

Full text of DOI:10.1016/j.molcata.2004.04.046

Journal of Molecular Catalysis A: Chemical 223 (2004) 211–215
The influence of support on ammoxidation of 3-picoline
over vanadia catalyst
Shyam Kishore Roy, P. Dutta, L.N. Nandi, S.N. Yadav, T.K. Mondal,
∗
S.C. Ray, Swapan Mitra, P. Samuel
Central Fuel Research Institute, Dhanbad 828108, India
Received 8 June 2003; received in revised form 4 April 2004; accepted 12 April 2004
Available online 22 September 2004
Abstract
The influence of high content of MoO
clay was investigated for ammoxidation of 3-picoline. Unsupported VMPO catalyst and supported catalysts were prepared by incipient wet
impregnation method keeping the mole ratio of V –MoO –P (1.00:3.27:0.14) constant. FT-IR spectra and X-ray diffractograms showed
3 2 5 3 2 5
(32 mol%) in V O –MoO –P O catalyst system supported on alumina, silica, HZSM-5 and modified
2
O
5
3
2 5
O
interaction among the active oxides and the support. The major phases identified were Mo
4
V
6
O
25, Mo
6
V
9
O
40 and MoV
2
O
8
. TG–DTA also
◦
showed the exothermic peaks indicating the formation of new compounds above 400 C. The physico-chemical properties of the catalysts
were correlated with their activity and selectivity for ammoxidation of 3-picoline. The unsupported catalyst showed 96.0% selectivity whereas
VMPO/HZSM-5 showed the highest selectivity (91.3%) amongst the supported catalysts.
©
2004 Elsevier B.V. All rights reserved.
Keywords: Ammoxidation; 3-Picoline; Nicotinonitrile; VMPO; ZSM-5; Modified clay
1
. Introduction
activity and selectivity of V2O –MoO3 catalysts vary signif-
5
icantly with MoO3 content. The maximum selectivity in ben-
zene oxidation for industrial production of maleic anhydride
has been reported [9] at around 30 mol% MoO3. The role of
high content of MoO3 in VMPO catalyst for ammoxidation
has not yet been studied. Earlier, we had reported vanadia cat-
alysts with low content of MoO3 for ammoxidation of alkyl
pyridines over alumina support [10,11].
Recently vanadium modified ZSM-5 zeolite has been re-
ported for oxidation and ammoxidation of 3-picoline with
62.6% yield of nicotinonitrile at 88% conversion [12]. It has
been observed that oxidation properties of zeolite are con-
fined to dehydroxylated surfaces of decationated zeolites and
the oxygen chemisorption occurs on anion and cation lattice
vacancies [13]. Modified clay has also been used as support
for SCR of NO with NH3 [14].
Supported vanadium oxide catalysts are good catalyst sys-
tems for selective oxidation and ammoxidation [1,2] of or-
ganic compounds as well as for selective catalytic reduction
(
SCR) of NO with ammonia [3]. The catalytic properties of
vanadium oxides are highly influenced by the nature of sup-
port, type of promoter and the method of preparation. The
structure and the composition of the materials used as sup-
ports influence the activity and selectivity of the active phase
to a marked degree [4–6]. Alumina, silica, titania and zirco-
nia are the most common oxide supports which are generally
used in oxidation and ammoxidation processes and have been
widely investigated. Molybdenum, chromium and phospho-
rous are generally incorporated with vanadium oxide because
they improve the activity and selectivity of the catalyst due
to change in the phase composition of the catalyst [7,8]. The
This paper deals with physico-chemical studies of
V2O –MoO3–P2O catalyst system with high content of
5
5
MoO3 (32 mol%) supported over alumina, silica, HZSM-
5 and modified clay for ammoxidation of 3-picoline to
∗
Corresponding author.
1
381-1169/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2004.04.046

2
12
S.K. Roy et al. / Journal of Molecular Catalysis A: Chemical 223 (2004) 211–215
◦
◦
nicotinonitrile. Nicotinamide and niacin are important phar-
maceutical ingredients which can be produced either by
multi-step route of oxidation or by single step catalytic hy-
dration of nicotinonitrile [15] obtained in a single step am-
moxidation reaction of 3-picoline. We report here for the first
time the use of modified clay as support for ammoxidation
reaction as well as the effect of higher concentration of MoO3
in the system.
condition under constant heating at 10 C/min up to 600 C.
Surface area and pore volume of supports and catalysts were
determined by surface area Analyser (Micromeritics – ASAP
2010) by nitrogen adsorption at liquid nitrogen temperature
(77 K).
Surface acidity of the catalysts was determined by NH3-
TPD (Micromeritics Auto Chem 2910). Catalyst sample
(0.1 g) was placed in a pulse reactor and calcined at 500 C in
◦
a flow of helium for 2 h. The temperature of the reactor was
◦
brought down to 100 C. Then a series of 0.5 ml NH3 gas
2
. Experimental
pulses were continued until no more uptake of NH3 was ob-
served. The system was flushed with helium for 30 min, then
the adsorbed ammonia was desorbed at a programmed heat-
2
.1. Preparation of catalyst
◦
ing rate of 10 C/min. The desorbed ammonia was detected
Materials used: vanadyl oxalate (prepared by refluxing
by a thermal conductivity detector.
V2O , Loba chemie, G.R.Grade, in water with oxalic acid),
5
ammonium heptamolybdate (AnalaR, BDH), orthophospho-
2
ric acid (85%), ␥-alumina (SA = 160 m /g), silica (SA
=
3. Results and discussion
2
200 m /g), ZSM-5 (SiO2:Al2O3 = 400) UCIL, Baroda.
HZSM-5 was prepared by ammonium ion exchange. Ben-
tonite clay was modified with group-IV metal salt.
FT-IR spectra of VMPO catalyst system, unsupported and
supported over alumina, silica, HZSM-5 and modified clay
are shown in Fig. 1. It is evident from all the spectra that
most of the bands for all the catalysts are in the range of
Supported and unsupported catalysts in the weight ra-
tio of V2O :MoO3:P2O :support (1.00:2.57:0.11:1.91) were
5
5
−
1
prepared in aqueous medium by incipient wet impregna-
tion method. A mixed solution of the required amount of
vanadyloxalate and orthophosphoric acid and another solu-
tion of ammonium heptamolybdate were separately heated
over steam bath and the former was added to the latter fol-
lowed by addition of support. This was mixed thoroughly and
refluxed over steam bath till the mass reduced to about half.
400–1500 cm . V O stretching band in pure and crystalline
−
1
−1
V2O is generally at 1020 cm , the shifting of 1020 cm
5
band towards lower frequency is in good agreement with
the observation that when MoO3 is added to V2O the ab-
5
sorption band weakened and shifted to lower wave number
−
1
between 974 and 1014 cm . The shift from 1020 band to
−
1
1014 cm is observed only in unsupported catalyst which in-
dicates the presence of V O stretching frequency. The shift-
ing of 1020 cm band towards lower frequency is due to
◦
The impregnated catalyst was oven dried at 120 C for 15 h
◦
◦
−1
and calcined first at 300 C for 3 h and then at 450 C for 15 h.
The catalyst was pelleted and sized (−6 + 14 BS mesh).
introduction of Mo into the lattice of V2O resulting in the
5
−1
stretching of metal oxygen bond. The band at 989 cm is
observed in the catalysts supported on zeolite, alumina and
also in unsupported VMPO catalyst. This band has further
2
.2. Activity studies
−
1
−1
Ammoxidationreactionwascarriedoutusingadownflow,
shifted to 977 cm in modified clay and 974 cm in silica
due to metal oxygen multiple bond which has the frequency
fixed bed, pyrex glass reactor of 20 mm i.d. The reaction mix-
ture (3-picoline and water) was fed from the top using a sy-
ringepump. Ammoniagasandairwerefedthroughcalibrated
flow meters. The reaction was carried out in the temperature
−
1
similar to that of MoO3 (995 cm band), which suggests
that intermediate compounds with doubly bonded oxygen ap-
pear as oxygen coordinated to Mo atoms and not to V atoms
[16]. The multiple bonds may be due to the formation of
◦
range 375–475 C and contact time 0.7–1.5 s. The product
was collected at the bottom using ice cold water and was
analysed by gas chromatograph with carbowax 20 M column
at 160 C temperature and TCD detector.
Mo V9O40 and Mo4V O and MoV2O8 intermediate com-
6 6 25
pounds which is evident from the phases identified by XRD.
These bands are more intense in zeolite based catalyst than
other supported and unsupported catalysts. The sharp bands
◦
− −1
1
2
.3. Catalyst characterization
are observed at 894 cm in alumina 872 cm in zeolite,
8
− −1
1
83 cm in silica and 894 cm in unsupported catalyst. A
−1
The phase composition of calcined catalysts was deter-
small peak is also observed at 817 cm in the supported
−1
mined by XRD recorded in D-8 ADVANCE diffractometer
Bruker AXS, Germany) using Cu K␣ radiation at 40 kV
catalyst. The bands in the region of 800–900 cm are at-
tributed to the vibration of polymeric chains M O M O in
which oxygen atoms are in coordinated polyhedra. The bands
(
and 40 mA in parallel beam geometry. Infrared spectra were
recorded in Perkin-Elmer FT-IR-2000 Spectrometer using
−1
which are in the region of 1003–1362 cm may be due to
the P O stretching vibration related to phosphate, pyrophos-
phates and metal phosphates and weak bands between 409
−
1
KBr pellet in the range of 4000–400 cm . TG–DTA of the
catalyst samples was done using Netzsch Thermal Analyser
Model STA 409C. The experiment was conducted in ambient
−
1
and 733 cm present in all the catalysts may be due to the

S.K. Roy et al. / Journal of Molecular Catalysis A: Chemical 223 (2004) 211–215
213
Fig. 2. XRD pattern of (a) VMPO, (b) VMPO/SiO2, (c) VMPO/Al2O3,
d1) VMPO/modified clay before reaction, (d2) VMPO/modified clay after
(
reaction, (e) VMPO/ZSM.
helps in the formation of intermediate phases. This is in good
agreement with the observations of Satsuma et al. [9] and
Munch and Pierron [19]. It seems that incorporation of low
concentration of oxide of phosphorus in the catalyst sys-
tem also helps in the formation of intermediate phases like
phosphates and pyrophosphates of vanadium as is evident
from FT-IR and XRD studies. Phosphorus oxide generally
increases the selectivity and life of the catalyst.
Fig. 1. FT-IR spectra of (a) VMPO, (b) VMPO/SiO2, (c) VMPO/Al2O3, (d)
VMPO/modified clay, (e) VMPO/ZSM-5.
TG–DTA graph of VMPO catalyst supported on HZSM-5
and modified clay is shown in Fig. 3. In addition to the en-
dothermic peaks due to loss of water, exothermic peaks are
presence of V O and P O stretching vibration of phosphates
of vanadium [17,18].
◦
observed between 280–440 and 440–500 C in zeolite sup-
X-ray powder diffraction patterns of supported and
unsupported VMPO catalysts are presented in Fig. 2.
The various crystalline phases identified are MoO3,
ported catalyst. This probably indicates that initially decom-
position of the precursor salts takes place and then resulting
oxides interact to form new compounds. For the catalyst sup-
ported on modified clay, exothermic peak between 165 and
V O126H2O, MoV2O8, Mo4V O , Mo V9O40, AlVO3,
5
6
25
6
◦
V(PO3)3, MoO0.03 VP1.1O4.84 with some other minor
unidentified phases in varying intensities as well as the broad
phases due to support materials. The “d” values obtained
are in agreement with JCPDS Values of the above phases.
320 C also shows the decomposition of the salts after that
weak exothermic peaks are observed indicating the weak in-
◦
teraction. However, the catalyst is stable up to 600 C.
The concept that the lattice oxygen of reducible metal
oxides would serve as a more versatile and functional ox-
idizing agent than molecular oxygen [20] play the crucial
factor in selecting the ammoxidation catalyst because the
reaction is governed by both oxidation and reduction pro-
cess. To have an insight into the promotive action of MoO3
The active intermediate phases are MoV2O8, Mo4V O ,
6
25
Mo V9O40 which are present in all the catalysts. The X-ray
6
diffractogram of VMPO on zeolite support exhibits a number
of sharp peaks as compared to alumina/silica/modified clay
supported catalysts. This may be due to specific physico-
chemical properties of HZSM-5. The introduction of high
on V2O , FT-IR spectra of different catalyst samples are
5
−1
content of MoO3 (32 mol%) into the lattice of V2O forms a
shown in Fig. 1. It is evident that 1020 cm characteris-
5
◦
4+
solid solution after calcining up to 500 C, it generates V
due to promoting effect of hexavalent molybdenum, which
tic band of V2O is shifted to lower frequencies by adding
5
MoO3, this may be explained as that the promotive action

2
14
S.K. Roy et al. / Journal of Molecular Catalysis A: Chemical 223 (2004) 211–215
action of the metal oxide forming intermediate compounds
showing better activity and selectivity for ammoxidation of
3-picoline.
Surfaceacidityofthesupports, supportedandunsupported
catalysts has been determined by NH3-TPD, which are pre-
sented in Table 1. It is evident that surface acidity is de-
creased after loading VMPO on the supports. VMPO catalyst
has very low surface acidity (0.007 mmol/g) giving highest
yield of nicotinonitrile and conversion of 3-picoline. It is ob-
served that supported catalysts have higher surface acidity
than VMPO. It may be inferred that due to higher acidity,
ammonia may neutralize the acid sites during ammoxidation
reaction thereby lowering the active centers responsible for
the reaction. This in turn results in decreasing the yield and
conversion using supported VMPO catalysts. However, re-
garding the role of MoO3 in the catalyst system it is reported
[9] that acid sites do not seem to be responsible for improve-
ment of catalytic activity because the type, strength and sur-
face concentration of acid sites are not varied by addition of
MoO3.
In the supported oxide catalyst, the oxide exists generally
in the form of a layer or molecular dispersion and is max-
imally influenced by the support [21,22]. Oxide layers also
appear to be quite thermally stable. Ease of formation and
thermal stability of monolayers has been related to the ratio
of the charge on the support cation to the sum of the cation
and oxide ion radii [23]. The stabilization of the monolayer
species on the surface of the support as a result of interac-
tion can be viewed either in terms of minimization of surface
free energy or in terms of the formation of new chemical
compounds.
Fig. 3. TG/DTA curves of VMPO/modified clay and VMPO/ZSM-5 cata-
lysts.
of MoO3 on V2O is due to weakening of V O stretching
5
Ammoxidation reaction of 3-picoline was carried out in
bond by doping of MoO3. This may be caused by increase
◦
of V4 resulted from the substitution of V with Mo
+
5+
6+
the temperature range of 375–475 C using 8–12 g of cata-
lyst. The performance of all the supported and unsupported
catalysts is given in Table 2. It is evident that unsupported
VMPO catalyst is highly active and selective (96.0% yield
with 98.0% conversion). Amongst the supported catalysts
VMPO supported on zeolite and modified clay gave higher
yield of nicotinonitrile (81.3% and 80.5%) with higher con-
version of 3-picoline as compared to alumina and silica
supported catalysts (71.1% and 72.8%). Fig. 4 shows the
effect of reaction temperature on conversion of 3-picoline
and selectivity to nicotinonitrile with different supported and
unsupported catalysts. Both the conversion and selectivity
and the high concentration of Mo with V2O in the catalyst
system leading to the formation of intermediate phases like
5
Mo4V O , Mo V9O40 and MoV2O8 in the catalyst. This
6
25
6
has also been observed by Munch and Pierron [19]. Surface
area, pore volume and surface acidity data of supports and
catalysts are presented in Table 1. Surface area and pore vol-
ume of supported VMPO catalysts decrease drastically due
to high concentration of metal oxide on the support. How-
ever, zeolite supported catalyst has the highest surface area
2
3
(
116.8 m /g) and pore volume (0.08 cm /g) due to micro-
porous structure of zeolite. This may help for better inter-
Table 1
Physico-chemical properties of supports/catalysts
2
3
Supports/catalysts
Surface area (m /g)
Pore volume (cm /g)
Surface acidity (mmol/gcat)
Alumina
Silica
HZSM-5 (SiO2/Al2O3 = 400)
Modified clay
VMPO
VMPO/SiO2
VMPO/HZSM-5
VMPO/modified clay
160.00
200.00
400.00
183.8
–
14.60
116.8
5.2
0.7438
–
0.11
–
0.432
–
0.462
0.506
0.007
0.219
0.116
0.151
–
0.0471
0.0797
0.0269

S.K. Roy et al. / Journal of Molecular Catalysis A: Chemical 223 (2004) 211–215
215
Table 2
Ammoxidation of 3-picoline to nicotinonitrile over different catalysts
Catalysts
Conversion of 3-picoline (wt.%)
Yield (wt.%)
Pyridine
Nicotinonitrile
Others
VMPO
VMPO/Al2O3
VMPO/SiO2
VMPO/HZSM-5
VMPO/modified clay
98.0
75.2
79.6
87.7
87.6
2.0
4.1
2.2
1.9
7.1
96.0
71.1
72.8
81.3
80.5
0.0
0.0
4.6
4.5
0.0
◦
Reaction temperature 425 C, contact time 1.1–1.3 s, mole ratio of 3-picoline:NH3:air = 1:9.7:18, 3-picolone:water = 1:3 (vol.%).
itrile (96.0%), the selectivity of the supported catalyst is also
sufficiently high >90% and further more, the supported cat-
alyst is stable with the added features.
Acknowledgements
The authors are grateful to Mr. N.B. Sarkar, for provid-
ing TG–DTA analysis of catalyst samples. The authors thank
Director, CFRI for his permission to publish the paper.
References
[
[
[
1] A. Beilenski, J. Haber, Catal. Rev. Sci. Eng. 19 (1979) 1.
2] A. Anderson, J.O. Bovin, P. Walter, J. Catal. 98 (1986) 204.
3] I.E. Wachs, G. Deo, B.M. Weckhuysen, A. Andrein, M.A. Vuurman,
M. de Boer, M.D. Amiridis, J. Catal. 161 (1996) 211.
4] M. Akimoto, E. Echigoya, J. Catal. 29 (1973) 191.
5] D. Vanhove, M. Blanchard, J. Catal. 37 (1976) 6.
6] K.V.R. Chari, C.P. Kumar, K.R. Reddy, T. Bhaskar, T. Rajiah, Catal.
Commun. 3 (2002) 7.
Fig. 4. Effect of reaction temperature on ammoxidation of 3-picoline with
different catalysts (1). VPMO (2). VPMO/ZSM-5 (3).VPMO/SiO2 (4).
VPMO/Al2O3 (5). VPMO/modified clay.
[
[
[
◦
increase on increasing the reaction temperature up to 450 C
and then decreases. The maximum yield obtained with ze-
◦
olite supported catalyst was 91.3% at 450 C and 80.5%
[7] A. Beilenski, T. Camra, M. Nazbar, J. Catal. 57 (1979) 326.
◦
[8] M. Ai, Bull. Chem. Soc. Jpn. 44 (1971) 761.
with supported on modified clay at 425 C. Supported cat-
[9] A. Satsuma, A. Hattori, K. Mizuatani, A. Faruta, A. Miyamoto, T.
Hattori, Y. Murakami, J. Phys. Chem. 93 (1989) 1484.
alysts showed low conversion and yield at lower tempera-
ture unlike unsupported catalyst. The increase of selectivity
of the product at higher temperature with supported catalyst
indicates that the formation of active phases as intermedi-
ate compounds occur during the reaction at higher temper-
ature. The formation of active phases is due to segregation
[
10] S.K. Roy, Ph.D. Thesis, Indian School of Mines, Dhanbad, India,
984.
[11] Indian Patent Nos. 1106051972 and 1380521972.
1
[
12] R. Ramachandra Rao, S.J. Kulkarni, M. Subramanyam, A.V. Rama
Rao, Zeolites 16 (1996) 254.
[
[
[
13] R.P. Porter, W.K. Hall, J. Catal. 5 (1968) 366.
14] L.S. Cheng, R.T. Yang, N. Chen, J. Catal. 164 (1996) 70.
15] S.K. Roy, S.C. Ray, P.K. Roy, J. Indian Chem. Soc. 57 (1980) 195.
[16] F.Y. Robb, W.S. Glausinger, P. Courtine, J. Solid State Chem. 30
1979) 171.
17] B. Manohar, B.M. Reddy, J. Chem. Technol. Biotechnol. 71 (1998)
41.
18] G. Genti, C. Galassi, I. Manenti, A. Riva, F. Trifiro, Stud. Surf. Sci.
Catal. 16 (1983) 543.
in V2O –MoO3 system in the course of oxidation during the
5
reaction [7]. The interaction of the active oxide components
with zeolite and modified clay seems to be stronger than that
of alumina and silica due to more active sites which may be
the reason for better activity and selectivity of these catalysts
for ammoxidation of 3-picoline.
Thus, it may be concluded that the increase in the ac-
tivity of the catalyst may be attributed to the combined ef-
fect of physico-chemical properties of the support and qual-
itative as well as quantitative change of the surface due to
interaction among the active oxides of the catalysts, result-
ing in compound formation between V and Mo oxides. Al-
though, unsupported catalyst gave higher yield of nicotinon-
(
[
[
1
[19] R. Munch, E. Pierron, J. Catal. 3 (1964) 406.
20] W.K. Lewis, Ind. Eng. Chem. 41 (1949) 1227.
21] I.E. Wachs, R.Y. Saleh, S.S. Chan, C.C. Chersich, Appl. Catal. 15
[
[
(1985) 339.
[
[
22] G.C. Bond, S. Flamerz, Appl. Catal. 33 (1987) 219.
23] F. Roozeboom, T. Fransen, P. Mars, P.J. Gellings, Z. Anorg, Allg.
Chem. 449 (1979) 25.