A DFT-assisted mechanism for evolution of the ammoxidation of 2-chlorotoluene (2-CLT) to 2-chlorobenzonitrile (2-CLBN) over alumina-supported V2O5 catalyst prepared by a solution combustion method

Article abstract of DOI:10.1016/j.jcat.2016.10.033

Ammoxidation of 2-chlorotoluene (2-CLT) has been carried out over vanadates and supported vanadates in the temperature range 350–450 °C. The catalysts were characterized by XRD, XPS, FT-IR, and Raman and surface area measurement. A maximum yield of 76% of 2-CLBN was obtained at a temperature of 425 °C and a 2-CLT:NH₃:air mole ratio of 1:8:22. DFT computations suggest that the reaction follows a Mars–van Krevelen (MVK) type of redox mechanism. The catalyst is first reduced by ammonia, producing an imine species. The 2-CLT molecule is also adsorbed through a vanadyl oxygen to produce a CH₂[dbnd]C₆H₅ species. This moiety is adsorbed over the catalyst through an imine formed on the surface to produce 2-chlorobenzonitrile. XPS spectra of fresh and spent catalysts confirm reduction of the catalyst during ammoxidation.

Full text of DOI:10.1016/j.jcat.2016.10.033

Journal of Catalysis 345 (2017) 245–257
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
A DFT-assisted mechanism for evolution of the ammoxidation
of 2-chlorotoluene (2-CLT) to 2-chlorobenzonitrile (2-CLBN) over
alumina-supported V O catalyst prepared by a solution combustion
2
5
method
a
a
a
b
a,
⇑
Ritambhara Dwivedi , Prabhakar Sharma , Akrati Sisodiya , Manohar Singh Batra , Rajendra Prasad
a
Molecular and Heterogeneous Catalysis Lab, School of Chemical Sciences, Devi Ahilya University, Indore 452001, India
Post-Graduate Department of Chemistry, Khalsa College, Amritsar, Punjab, India
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Ammoxidation of 2-chlorotoluene (2-CLT) has been carried out over vanadates and supported vanadates
in the temperature range 350–450 °C. The catalysts were characterized by XRD, XPS, FT-IR, and Raman
and surface area measurement. A maximum yield of 76% of 2-CLBN was obtained at a temperature of
Received 2 August 2016
Revised 30 October 2016
Accepted 31 October 2016
Available online 18 December 2016
4
25 °C and a 2-CLT:NH
Mars–van Krevelen (MVK) type of redox mechanism. The catalyst is first reduced by ammonia, producing
an imine species. The 2-CLT molecule is also adsorbed through a vanadyl oxygen to produce a CH @C
species. This moiety is adsorbed over the catalyst through an imine formed on the surface to produce
-chlorobenzonitrile. XPS spectra of fresh and spent catalysts confirm reduction of the catalyst during
ammoxidation.
3
:air mole ratio of 1:8:22. DFT computations suggest that the reaction follows a
2
6 5
H
Keywords:
Ammoxidation
2
2
2
-Chlorooluene(2-CLT)
-Chlorobenzonitrile(2-CLBN)
Al
2
O
3
/V
2
O
5
Ó 2016 Elsevier Inc. All rights reserved.
Density functional theory (DFT)
Gaussian 09W
1
. Introduction
Reaction of an organic compound containing a methyl group
[6,8–11]. The choice of P/V ratio is found to play a key role in
determining the catalytic properties. One-step hydrothermal
synthesis and characterization of VACrAO nanospheres and
their excellent performance in the ammoxidation of 3,4-DCT
and 2,6-DCT have been reported by Xu et al. [12]. Performance
adjacent to a double bond with dioxygen and ammonia over a solid
surface to produce nitriles is called ammoxidation [1,2]. The pro-
cess can be used to produce a wide range of industrially important
nitriles that can be transformed easily to numerous value-added
fine chemicals [1–5]. The process of ammoxidation is a vapor-
phase single-step ecologically and economically efficient route
for synthesis of aliphatic and aromatic nitriles employing cheap
raw materials such as air and ammonia. Water is the only bypro-
duct in the reactions [1,5].
Halogen-substituted nitriles produced by ammoxidation are
useful in the production of different dyestuffs, herbicides,
pharmaceuticals, and pesticides [6–11]. In particular, ammoxida-
tion of 2-chlorotoluene (2-CLT) is an industrially important reaction
for producing 2-chlorobenzonitrile (2-CLBN) in a single step. The
desired product (2-CLBN) is mainly used for synthesis of dye inter-
mediates, medicines, pesticides, and other fine chemicals [6,7,9].
Ammoxidation of halogen-substituted toluenes in the vapor
phase has been studied previously over VAPAO catalysts
2 5 2 3
of Mo-, W-, and La-promoted V O /Al O catalysts for ammoxi-
dation of 2-chlorotoluene to 2-chlorobenzonitrile has been
reported by Babu et al. [7]. Vanadium oxide industrial catalysts
have been used in a variety of chemical reactions such as selec-
tive oxidation, ammoxidation, and oxidative dehydrogenation
(ODH) of alkanes to alkenes [13–15]. There are a few reports
on the density functional theory modeling of ammoxidation
reactions of aliphatic compounds [16–22]. There seems to be
no report of ammoxidation of 2-chlorotolune over alumina-
supported vanadium catalysts prepared by a solution combus-
tion method [23,24] or transition state modeling of the reaction
employing density functional theory (DFT)-level quantum
mechanical computations.
The present problem was therefore undertaken to evaluate the
performance of supported vanadium catalysts prepared by a solution
combustion method and to predict the mechanism of the reaction
based on information collected from DFT computations.
⇑
Corresponding author. Present Address: SSISM, Sandalpur, India.
E-mail address: rjndr.prsd3@gmail.com (R. Prasad).
http://dx.doi.org/10.1016/j.jcat.2016.10.033
021-9517/Ó 2016 Elsevier Inc. All rights reserved.
0

2
46
R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
2
. Experimental
2.1. Preparation of catalyst
All the chemicals employed were of analytical grade and used
as such without further purification. Aluminum nitrate (LOBA Che-
mie Pvt. Ltd.) as Al precursor, ammonium metavanadate
Sigma–Aldrich) as vanadium precursor, and citric acid (Merck)
as a fuel were used as received. To prepare alumina-supported
(AV) in molar ratio 90:10, a solution of 16.55 g of aluminum
nitrate in 100 ml of water was mixed with another solution pre-
pared by mixing 0.64 g of NH VO in 50 ml of water. The final solu-
2 3
O
(
2 5
V O
4
4
tion was mixed with 17.19 g of citric acid and fired in a microwave
oven for 2 min. The material swelled into a green-colored gel. The
gel was powdered and calcined in a muffle furnace at 500 °C for
4
h. A dark-green-colored residue was obtained, which was ground
in a motor pestle to a fine powder. A series of catalysts were pre-
pared by changing the vanadium content from 5 to 25%. A similar
procedure was adopted for preparation of ZrV
TV). The catalysts were designated as MV , where M = Al/Zr/Ti,
V = vanadium, and X is the vanadium percentage in the catalyst.
2 7 2 7
O (ZV) and TiV O
Fig. 2. XRD patterns of AlV20 catalyst at different calcination temperatures.
(
X
2.2. Catalyst characterization
XRD measurements of catalysts were made using a Rigaku X-
ray powder diffractometer using Cu K
a radiation with wavelength
1
.5406 Å as a source. The diffractometer was equipped with a gra-
phite crystal monochromator (for the diffracted beam) and a scin-
tillation counter as detector. The powder XRD data were collected
ꢀ1
in the 2h range from 20° to 80° with a scan rate of 0.5° min
.
Table 1
BET surface area of a few supported vanadia catalysts.
V
2
O
5
loading (wt.%)
BET surface area
Average particle size
(in nm)
2
ꢀ1
(
m g )
AV10 (500)
AV15 (500)
AV20 (500)
AV20 (600)
AV20 (700)
ZV10 (500)
TV10 (500)
91.2651
66.9221
57.2605
35.8401
25.2386
25.8695
14.5708
7.35
7.85
8.23
12.15
19.85
15.00
36.73
Fig. 3. Raman spectrum of AlV20 catalyst calcined at 500 °C.
ꢀ1
Raman records in the range 100–4000 cm were made on a
Labram HR800 micro Raman spectrometer using a 488 nm wave-
length Ar+ laser source at energy 2.53 eV and using Labspec soft-
ware. Surface area was measured using an ASAP 2020 V3.04 H
2
surface area analyzer in a N physisorption apparatus equipped
with a thermal conductivity detector. The pore size distribution
was calculated using the BJH method. The FT-IR spectra of the sam-
ꢀ1
ples were recorded in the range 400–4000 cm using a Bruker
VERTEX 70.
2.3. Catalytic activity
The activity of the catalysts for vapor-phase ammoxidation of 2-
CLT to 2-CLBN was carried out at atmospheric pressure in a fixed-
bed, vertical downflow glass reactor placed inside a tubular fur-
nace. The temperature was regulated with the help of energy reg-
ulators and metered by a thermocouple inserted in the reactor.
Fresh catalyst (1.0 g) was charged into the center of the reactor
in such a way that the catalyst was sandwiched between two lay-
ers of inert glass beads. The reactant 2-chlorotoluene (2-CLT) was
2 5 2 3 2 5
Fig. 1. XRD patterns of V O /Al O catalysts with different V O loadings calcined at
5
00 °C: (@) V
2
O
5
, (Ó)
c
-Al
2
O
3
.
3
fed using a pressure-equalizing funnel. Air and NH were taken

R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
247
X 2 5
Fig. 4. Comparison of FT-IR spectra of AlV catalysts and pure V O calcined at 500 °C.
decrease in surface area with increasing amount of vanadia can
be attributed to the blocking of pores by V [25,26]. Pore size
and pore volume of AV10 catalysts were found to be 10.64 nm
2 5
O
3
ꢀ1
and 0.24 cm g , respectively. The corresponding values for
3
ꢀ1
AV20 were found to be 5.45 nm and 0.07 cm g . The decrease
in the pore size and pore volume with increase in vanadia content
2 5
can be attributed to partial blocking of pores by V O [25].
3.2. XRD results
2 5 2 3
The X-ray diffraction patterns of V O /Al O , catalysts having
different V
2
O
5
loadings and calcination temperatures are shown
up to
in Figs. 1 and 2. The absence of characteristic peaks of V
2 5
O
the 15 wt.% loading of vanadia suggests a highly dispersed state
of vanadia. This result is in agreement with the theoretical mono-
layer capacity [25,26]. However, characteristic crystalline peaks of
V
2
O
5
were detected at both 20 and 25 wt.% V
peaks at 2h values of 67.21°, 45.28°, and 37.58° can be assigned
to -alumina (JCPDS card No. 10-0425) and sharp peaks appeared
at 2h values of 20.27°, 26.09°, and 34.35° can be ascribed to crys-
talline V [15,27,28]. Some weak peaks at 2h values of 25.386°,
7.79°, and 28.68° can be assigned to AlVO (JCPDS card No. 39-
76). XRD patterns of the catalysts with 15 wt.% V seem mostly
2 5
O loadings. The
Fig. 5. FT-IR spectra of AlV20 catalyst at different calcination temperature.
c
from compressed gas cylinders. The flow rates of these gases were
measured using a manometer. Experiments were carried out in the
temperature range 350–450 °C. Reaction products were cooled
through an ice-cooled condenser and collected in an ice-cooled
receiver. Product analysis was carried out with the help of a gas
chromatograph filled with a flame ionization detector and a capil-
lary column.
2 5
O
2
2
4
2 5
O
to be in amorphous form. The extent of crystalline structure
increases with increased vanadia concentration above 15 wt.%. At
2
0 and 25 wt.% V
port the idea that bulk V
deposit coverage may be greater than one monolayer [28]. When
coverage exceeds the monolayer, the formation of bulk (V ) in
crystallites occurs and uniform V dispersion decreases rapidly
[14]. The average particle size of V /Al in different catalysts
2
O
5
loadings, the V
2 5
O crystal peaks clearly sup-
2
O
5
crystallites were generated and V
2 5
O
3
. Results and discussion
2 5
O
2 5
O
3.1. BET surface area
2
O
5
2 3
O
was calculated with the Debye-Scherrer equation and is shown
in Table 1. Particle size increased with increased calcination
temperature.
The BET surface areas of the various catalysts samples obtained
by N
2
physisorption at ꢀ197 °C are listed in Table 1. A consistent

2
48
R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
ꢀ
1
1
97, and 145 cm can be ascribed to bulk V
2
O
5
[29,30]. The band
ꢀ1
at 145 cm can be assigned to the vibrations of AOAVAOAVAOA
atoms and that at 197 cm can be ascribed to the (V O ) stretch-
2 2 n
ing mode. The two peaks located at 281 and 407 cm are assigned
ꢀ
1
ꢀ1
to the bending vibration of the V@O bonds. The peaks at 701 and
ꢀ
1
9
95 cm are assigned to the doubly coordinated oxygen (VAOAV)
stretching mode and terminal oxygen (V@O) stretching mode,
respectively [14,29,30–33].
FT-IR spectra of Al
loadings are shown in Fig. 4. Bulk V
stretching and bending vibrations of the V@O bond at 1022 and
2
O
3
2
/V O
5
catalysts with different vanadia
2 5
O
displays well-resolved
ꢀ
1
8
34 cm , respectively. The former band corresponds to the V@O
stretching vibration, and the latter corresponds to the VAOAV
bending vibration [25,28,34,35]. The occurrence of these character-
istic bands of V O species confirms the presence of well-defined
2 5
vanadia crystallites on catalysts loaded up to 20% of vanadia [15].
Similar conclusions can be drawn from the XRD records of the cat-
alysts. Vibrations of AlAOH and AlAO bonds are observed in the
ꢀ1
range 3000–3600 and 1632 cm , corresponding to the stretching
and bending vibration of hydrogen bonding OH groups, respec-
tively [25,36]. Disappearance of the band around 3400–
ꢀ
1
3
600 cm at temperatures higher than 500 °C can be attributed
to loss of hydroxyl groups (Fig. 5).
3.4. XPS results
The full-scan XPS spectrum of fresh AlV20 catalyst is repro-
duced in Fig. 6A and assignments are shown in Table 2. The spec-
trum suggests the presence of V, Al, O, and C. The C peak appears
due to surface contamination. The presence of the elements O, V,
and Al is also confirmed by EDAX analysis reproduced in the Sup-
plementary Information (Fig. S1). The XPS spectra of fresh and
spent catalysts in the region 510–540 eV are presented in
Fig. 6B and C, respectively, which provides information on the
change in oxidation state in the spent catalyst. The V2p core level
2 5
peaks of V O can be seen in the range 515–524 eV [13,37–47]. The
broad peak of the V2p3/2 core level displays more than one compo-
nent in both fresh and spent catalysts [42–44]. A careful scrutiny of
+
5
+4
Fig. 6A and B suggests the presence of V and V in the fresh cat-
+
4
+3
+5
alyst and of V and V along with traces of V in the spent cata-
lyst [42–44]. The binding energy of vanadium (2p) is shifted
slightly to lower levels of 515.22 eV, which means the oxidation
state of the vanadium is changed to the mixed vanadium state of
V (V/IV, III). The rest of the assignments are shown in Table 2.
3
3
.5. Catalyst activity
.5.1. Comparative performance of catalysts
In the present work, ammoxidation of 2-CLT in the vapor phase
was studied over titanium vanadate, zirconium vanadate, and
alumina-supported vanadium oxide and results are presented in
Table 3. The performance of ZV20 was found to be better than that
of ZV10. On the other hand, the performance of TV10 was found to
be better than that of TV20. The yield of 2-CLBN was higher for
AlV20 than for AlV10. The overall performance of AlV20 was found
to be better than that of all the six catalysts listed in Table 3. The
2 5 2 3
effect of V O loading on Al O support on the performance of
Fig. 6. X-ray photoelectron spectra of the AlV20 catalyst: (A) full scan of fresh
AlV20 catalyst; (B) O1s1/2, V2p3/2 core levels (fresh catalyst); (C) O1s1/2, V2p3/2 core
levels (spent catalyst).
AlV catalyst was studied in detail and results are graphically shown
in Fig. 7. The conversion of 2CLT, the yield of 2-CLBN, and its selec-
tivity increased up to 20% loading. Although selectivity increased
slightly with further loading, yield of 2-CLBN and conversion of
2-CLT decreased. This can be attributed to formation of a crys-
3.3. Vibrational spectroscopy
2 5
talline V O phase at higher loading. Babu et al. [7] reported the
The Raman spectrum of the AlV20 catalyst calcined at 500 °C is
reproduced in Fig. 3. Bands at 995, 866, 701, 528, 479, 407, 283,
influence of promoters on the structural and catalytic functionali-
ties of V O /Al O catalysts for the ammoxidation of ortho-
2
5
2 3

R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
249
Table 2
Results of XPS studies of AlV20 catalyst.
BE (eV)
FWHM (eV)
Core line
Assignment
References
[35–43]
5
5
5
5
5
5
5
7
1
5
5
17.78
18.20a
16.23
1.9
2.0
2.4
2.3
2.5
3.6
3.9
2.2
2.4
3.1
2.8
V2p3/2
V5+
V2p3/2
V4+
[35–43]
[36]
a
16.56
15.22a
23.03
22.97a
3.26
V2p3/2
V2p1/2
V3+
—
Al2p1/2
Al2s1/2
O1s1/2
Al3+
—
[40]
[40]
[35–40]
17.5
2
ꢀ
29.75
O
29.01a
a
Spent catalyst.
chlorotoluene. A maximum conversion of 2-CLT of 76% was
obtained for the La -promoted V /Al catalyst.
O
2 3
2
O
5
2 3
O
The active AlV20 catalyst calcined at different temperatures
was evaluated for its performance in the ammoxidation of 2-CLT
and the results are shown in Fig. 8. The results suggest that there
was not much variation in activity. However, a marginal decrease
in 2-CLT conversion and 2-CLBN yield was noticed with increase
in catalyst calcination temperature from 500 to 700 °C. The XRD
patterns of these catalysts are shown in Fig. 2. These results sug-
gest that the crystallinity of the metal oxide increased when
AlV20 was exposed to high temperatures. The marginal variation
in ammoxidation activity might be related to the difference in crys-
tallinity of the catalyst.
Fig. 7. Effect of vanadium loading on the performance of AlV catalysts in the
ꢀ1
ꢀ1
ammoxidation of 2-CLT to 2-CLBN: 2-CLT = 2 ml h
,
3
NH = 33 ml h ,
ꢀ
1
air = 120 ml h , reaction temperature = 400 °C, catalyst weight = 2g.
3.5.2. Effect of reaction temperature
The effect of reaction temperature on the performance of the
AlV20 catalyst was studied in the temperature range 340–460 °C
as shown in Fig. 9. With increasing temperature, the yield of 2-
CLBN and conversion of 2-CLT increased and a maximum was
obtained at 425 °C. However, selectivity to 2-CLBN decreased with
temperature up to 450 °C. Under these conditions, the maximum
selectivity to 2-CLBN of 91% was obtained at 350 °C.
3.5.3. Effect of air and ammonia flow rates
The effect of air flow rate on the performance of AlV20 catalyst
Fig. 8. Effect of different calcination temperatures of AlV20 catalyst on the
in the ammoxidation of 2-CLT at a constant temperature of 425 °C
ꢀ
1
ꢀ1
ammoxidation of 2-CLT to 2-CLBN: 2-CLT = 2 ml h
air = 120 ml h , reaction temperature = 400 °C, catalyst weight = 2g.
,
3
NH = 33 ml h ,
ꢀ1
and ammonia flow rate of 33 ml h is shown in Fig. 10. The best
ꢀ1
ꢀ1
results were obtained at an air flow rate of 120 ml h . Higher oxy-
gen content favors the oxidation of ammonia, and under these con-
ditions the V@NH species are transformed into VAO electrophilic
ones [48]. Similarly, the effect of ammonia flow rate on the perfor-
mance of the catalyst at a constant temperature of 425 °C and air
3.5.4. Effect of 2-CLT flow rate
2-CLT can also be a limiting reagent in 2-CLT ammoxidation
over the AlV20 catalyst, and it seems that the 2-CLT:oxygen and
2-CLT:ammonia ratios have a significant influence on the catalyst
selectivity. The effect of 2-CLT on the performance of AlV20 cata-
lyst was studied at constant flow rates of air and ammonia at a
ꢀ1
flow rate of 120 ml h was studied and results are graphically
ꢀ
1
shown in Fig. 11. An ammonia flow rate of 50 ml h was found
to be optimum for the reaction.
Table 3
2 7 2 7 2 5
Comparative performance of TiV O , ZrV O , and alumina-supported V O in the ammoxidation of 2-CLT to 2-CLBN.
S. no.
Catalyst
Catalyst amount (g)
2CLT conversion (%)
Product distribution (%)
% Selectivity
2CLT
2CLBN
Others
1
2
3
4
5
6
ZV10
ZV20
AlV10
AlV20
TV10
TV20
1
1
1
1
1
1
17.50
16.00
45.00
56.41
19.40
8.00
82.50
84.00
55.00
43.59
80.60
92.00
6.03
9.00
28.11
41.65
6.23
5.00
11.47
7.00
16.89
14.76
13.17
3.00
34.45
56.25
62.46
73.83
32.11
62.50
Notes: 2-CLT = 2 ml hꢀ1, NH3 = 20 ml h , air = 50 ml h , reaction temperature = 400 °C.
ꢀ1
ꢀ1

2
50
R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
Fig. 9. Effect of reaction temperature on the ammoxidation of 2-CLT to 2-CLBN over
ꢀ
1
ꢀ1
ꢀ1
AlV20 catalyst: 2-CLT = 2 ml h
weight = 2g.
,
NH
3
= 33 ml h
,
air = 120 ml h
,
catalyst
Fig. 11. Effect of ammonia flow rate on the ammoxidation of 2-CLT to 2-CLBN over
AlV20 catalyst: 2-CLT = 2 ml h , air = 120 ml h , reaction temperature = 425 °C,
ꢀ1
ꢀ1
catalyst weight = 2g.
Fig. 12. Effect of 2-CLT flow rate on the ammoxidation of 2-CLT to 2-CLBN over
ꢀ
1
ꢀ1
Fig. 10. Effect of air flow rate on the ammoxidation of 2-CLT to 2-CLBN over AlV20
AlV20 catalyst: NH
catalyst weight = 2g.
3
= 50 ml h , air = 120 ml h , reaction temperature = 424 °C,
ꢀ
1
ꢀ1
3
catalyst 2-CLT = 2 ml h , NH = 33 ml h , reaction temperature = 400 °C, catalyst
weight = 2g.
constant temperature of 425 °C. The conversion of 2-CLT and yield
was at a maximum at a 2-CLT flow rate of 2 ml h and decreased
4. Transition state modeling and mechanism of reaction
ꢀ1
above that. However, 2-CLBN selectivity increased with increasing
All the DFT calculations for the current problem were per-
formed using the Gaussian 09 W suite [49,50]. The B3LYP function-
als were used to describe electron exchange and correlation [51],
while the 631-G basis set was used to find the optimized geome-
tries of reactants, products, and the transition state. For transition
state calculation the QST2 keyword was used [52]. This option
requires the reactant and product structures as input. It generates
an approximate geometry for the transition state that is midway
between the reactants and products, followed by automatic opti-
mization of the starting structure to the first-order saddle point.
At many places the optimization of the transition state was
2
2
2
-CLT feed rate, as depicted in Fig. 12. The maximum conversion of
-CLT, corresponding to 96%, was obtained at a 2-CLT flow of
ꢀ1
ml h . When very low contact times of 2-CLT are employed,
the selectivity of 2-CLBN reached 92%. Thus, a proper feed compo-
sition provides a correct equilibrium between the active nitrogen
and oxygen species.
Comparative performance of various catalysts used in ammox-
idation of 2-CLT is shown in Table 4. An inspection of Table 4
reveals that the performance of AlV20, used in the present study,
is better than that of other catalysts used in previous studies.

R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
251
Fig. 13. Four-coordinated alumina-supported vanadium oxide: (A) monomeric and
B) polymeric.
(
Fig. 15. FT-IR spectrum of AlV20 catalyst after passing ammonia at 425 °C.
achieved directly using the TS option. Vibrational frequencies were
calculated for the optimized geometries to identify the nature of
the reactant or product (no imaginary frequency) and TS structure
generating unique mechanistic insights that are difficult to access
experimentally [52]. The enthalpy and the activation energies were
calculated at the reaction temperature as described by Ochterski
(
one imaginary frequency). DFT calculations permit accurate deter-
[
53]. The adsorption energy (Eads) was calculated as [54,55]
mination of the energetic parameters of elementary reaction steps
at the atomic level, and have been adopted as a powerful tool for
Eads ¼ EðadsorbateꢀsubstrateÞ ꢀ ðEadsorbate þ EsubstrateÞ;
2 5
Fig. 14. Optimized geometry of catalyst (polymeric V O ).
Table 4
Comparative performance of various catalysts used in ammoxidation of 2-CLT.
S. no.
Catalyst
2-CLT conversion %
2-CLBN yield %
2-CLBN selectivity
Reference
1
2
3
4
5
6
7
8
b-VOPO
4
85
41
62
50
70.0
41.0
62.0
50.0
58.0
72.8
47.8
76.0
82.0
100
100
100
99.0
85.6
54.8
79.0
[9]
[7]
[7]
[7]
[7]
[11]
[11]
10 wt.%V
2
O
5
/Al
-Al
/Al
/Al
10 wt.% VPO/SiO
10 wt.% V-O/SiO
20 wt.% V /Al
2
O
3
La
MO
WO
2
O
3
-V
-V
-V
2
O
5
2
O
3
3
2
O
5
2
O
O
3
3
2
O
5
2
3
58
2
85.00
87.3
96
2
2
O
5
2 3
O

2
52
R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
where E(adsorbate–substrate) is the total energy of the adsorbate–sub-
strate system in the equilibrium state, and Esubstrate and Eadsorbate
are the energies of the isolated substrate and adsorbate, respec-
tively. By this definition, a negative value of Eads corresponds to
an exothermic process [54].
In the present calculation, we take four-coordinated alumina-
supported polymeric vanadium oxide [45,56,57]. The monomeric
and polymeric forms of the catalyst are shown in Fig. 13. In the
polymeric form, we can distinguish three kinds of oxygen atoms,
namely, terminal oxygen O1 (V@O), bridging oxygen O2 (VAOAV),
and bridging to support oxygen O3 (VAOAM) [45,56,57]. The opti-
mized geometry of the support less the vanadium catalyst is shown
in Fig. 14. Divalancy of oxygen is fulfilled by hydrogen atoms. The
V@O and VAOH bond lengths were found to be 1.68 and 1.76 Å,
respectively. The bridging VAO bond length was computed to be
1.77 Å. Most of the mechanisms suggested for ammoxidation of
organic compounds over vanadium catalysts are based on the
Mars–van Krevelen redox mechanism [16–21]. Ammoxidation of
2-CLT involves activation of ammonia by terminal V@O groups to
form imido species (V@NH), followed by 2-CLT activation. Overall,
the vanadyl oxygen O(1) on the V O surface is slightly more reac-
2
5
tive in CAH bond activation than is the O(2) atom [20,49,52].
4.1. Activation of ammonia
Fine details of the mechanism were obtained by modeling the
transition state for the reaction under the QST2 option of the
Gaussian-09 suite. We start our computation with the activation
of ammonia and the formation of the V@NH group. Passing ammo-
nia over the catalyst produced water, confirming the reaction V@O
Fig. 16. Activation of ammonia over catalyst and formation of imido group: (A) first H abstraction; (B) second H abstraction.

R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
253
Fig. 16 (continued)
Fig. 17. PES diagram for ammonia activation.

2
54
R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
+
NH
3
? V@NH + H
2
O. The presence of the imido group was con-
length to the VAOX bond. The transition imaginary frequency is
ꢀ1
firmed by recording the IR spectrum of the catalyst, reproduced
in Fig. 15. The presence of the Mo@NH group has also been con-
firmed by Pudar et al. [17,18] and Goddard et al. [16,19] in the
ammoxidation of propylene over bismuth molybdates. A similar
view has been suggested for the ammoxidation of toluene [58].
Optimized geometries for adsorbed ammonia over catalyst,
transition state, and product resulting from the first H abstraction
are shown in Fig. 16A. The adsorption of ammonia is found to be
m
= 1441i cm . The geometries for second H abstraction from
the imido group are shown in Fig. 16B. The activation energy for
ꢀ1
second H abstraction was calculated to be 27.10 kcal mol . Here
the OAH distance is 1.25 Å and the corresponding NAH distance
is 1.23 Å; the VAOH distance elongates from 1.77 to 2.05 Å. The
ꢀ1
transition state imaginary frequency is
m = 1423i cm . The PES
diagram for ammonia activation is shown in Fig. 17.
ꢀ1
exothermic with a heat of adsorption of ꢀ22.93 kcal mol , which
4.2. Activation of 2-chlorotoluene (2-CLT)
ꢀ1
is very close to the corresponding value of ꢀ24.5 kcal mol for this
step over a bismuth molybdate catalyst [16]. The activation energy
For the activation of 2-CLT, the 2-CLT molecule is absorbed
ꢀ1
for the first H abstraction was calculated to be 56.48 kcal mol
.
through a vanadyl oxygen to produce a CH @C H species. This
2
6
5
ꢀ
1
Pudar et al. reported 44.2 kcal mol for the first H abstraction
transfer from the NH to an Mo@O group on the same molybde-
num site. In the first H transfer of NH through the V@O bond,
the OAH distance is 1.11 Å and the respective NAH distance is
.57 Å, indicating that the hydrogen transfer is almost complete.
The VAO bond elongates from 1.59 to 1.72 Å and is very near in
step was exothermic and the energy calculated was
ꢀ
1
3
ꢀ22.61 kcal mol . This CH @C H moiety is adsorbed over the
2
6
5
3
catalyst through an imine formed on the surface to produce
2-chlorobenzonitrile. Activation of 2-CLT is depicted in Fig. 18.
Our computations suggest that this process is completed in four
successive steps [16,19,58].
1
3
1
Fig. 18. PES for activation of 2-chlorotoluene (2-CLT): (A) second H abstraction from V ; (B) second H abstraction from V . All energies in the figure are in Hartree (or
equivalent atomic units).

R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
255
Fig. 18 (continued)
Table 5
The geometrical parameters and the imaginary frequencies for transition states.
Transition state
Bond length in Å
Imaginary frequency
VAO
NAH
CAH
OAH
Ts1
Ts2
Ts3
Ts4
Ts2
Ts3
Ts4
1.65
1.98
1.60
1.78
1.67
2.03
1.78
–
–
1.79
–
–
1.98
1.59
–
2.60
2.84
–
1.78
1.10
1.68
1.73
2.45
1.27
1.97
ꢀ1244.00
ꢀ1346.43
ꢀ1355.31
ꢀ979.98
0
0
0
0
0
0
ꢀ1235.12
ꢀ1277.00
ꢀ916.64
1.67
–
2.06
1
Step 1: Abstraction of first hydrogen: The first H abstraction from
form V AOH—see Fig. 18B (III). The activation energies were found
3
3
ꢀ1
the CH
3
3
group was attempted from V vanadyl oxygen (V @O) to
CH moiety is attached to
to be 33.51 and 68.72 kcal mol , respectively. This suggests that
the preferred route for second H abstraction should be from the
V OH group.
Step 3: Abstraction of third hydrogen: There remain two more H
atoms to be abstracted, one from the CH moiety and another from
the NH moiety, to produce nitrile by two paths. Since NH is a
weaker bond than CH, we decided that the third H was abstracted
form V AOH, and the remaining C
H
6 5
2
3
the V@NH group in Fig. 18A (III). The activation energy for this step
ꢀ1
was found to be 79.38 kcal mol
.
Step 2: Abstraction of second hydrogen: We found two distinct
pathways for abstraction of the second methyl hydrogen: (a) from
3
3
1
2
V AOH group to form V AOH —see Fig. 18A (V); (b) from V @O to

2
56
R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
Scheme 1. Possible reaction path for ammoxidation of 2CLT to 2CLBN over alumina-supported vanadium.
ꢀ
from the NH group. Based on the two distinct paths mentioned for
second H abstraction above (Step 2a, 2b), third abstraction is
expected to follow two paths:
two pathways was computed to be 60.19 and 58.93 kcal mol 1
respectively.
,
PES diagrams for paths a and b of Step 3 are shown in
Fig. 18A and B. The first H abstraction, associated with the highest
activation energy, seems to be the rate-determining step. The over-
all possible reaction path is presented in Scheme 1. Structure coor-
dinates of reactant, product, and transition state models obtained
from geometry optimization of ammoxidation of 2-CLT are shown
in the Supplementary Information (Figs. S2 and S3). The geometric
parameters and the imaginary frequencies for the four transition
states (four-H abstraction) are presented in Table 5.
(
a) When second H abstraction is from the V3 AOH group Since
1
V AO is the only vacant site, the third H abstraction is
1
1
expected to be from the V @O group, converting V @O to a
1
V AOH group, as in Fig. 18A (VII). This step is associated
with an activation energy of 33.51 kcal mol
ꢀ1
.
1
(
b) When second H abstraction is from the V @O group In this
3
case third H abstraction is expected from the V AOH group
3
to form V AOH
2
(Fig. 18B (V)). The associated activation
ꢀ1
energy was computed to be 51.24 kcal mol
.
5
. Conclusions
Step 4: Abstraction of fourth hydrogen: For fourth H abstraction,
the a and b pathways of Step 3 converge to the same products
and are energetically comparable. The activation energy for the
Ammoxidation of 2-chlorotoluene has been carried out over an
alumina-supported
2
V O
5
catalyst prepared by
a
solution
combustion method. A maximum yield of 2-chlorobenzonitrile of

R. Dwivedi et al. / Journal of Catalysis 345 (2017) 245–257
257
7
6% was obtained with a selectivity of 79.76%. DFT level computa-
tions were performed to find the path with minimum activation
energy. Computations suggest that NH and 2-CLT molecules
adsorbed and produce imine and C CH species, respectively.
The C CH species is then adsorbed onto the imido species to
[26] A.L. Ahmad, B. Koohestani, S. Bhatia, S.B. Ooi, Int. J. Appl. Ceram. Technol.
2012) 1–11.
(
[
27] J. Yanzi, L. Dingjianyi, Y. Qinghe, L. Dadong, S. Shuling, N. Hong, China Pet.
Process. Petrochem. Technol. 15 (2013) 1–9.
3
H
6 5
2
[28] N.R. Shiju, M. Anilkumar, S.P. Mirajkar, C.S. Gopinath, B.S. Rao, C.V.
Satyanarayana, J. Catal. 230 (2005) 484–492.
H
6 5
2
[
[
29] K. Routray, K.R.S.K. Reddy, G. Deo, Appl. Catal. A 265 (2004) 103–113.
30] I.E. Wachs, Dalton Trans. 42 (2013) 11762–11769.
produce nitrile in three successive dehydrogenation steps. Among
these, the first H abstraction from the methyl group had a maxi-
mum activation energy of 79.38 kcal mol ; this seems to be the
rate-determining step.
[31] N. Magg, B. Immaraporn, B.J. Giorgi, T. Schroeder, M. Bäumer, J. Döbler, Z. Wu,
E. Kondratenko, M. Cherian, M. Baerns, P.C. Stair, J. Sauer, F.H.J. Chim, J. Catal.
ꢀ1
226 (2004) 88–100.
[
[
[
[
32] C. Sanchez, J. Livage, G. Lucazeau, J. Raman Spectrosc. 12 (1982).
33] C. Prameela, K. Srinivasarao, Int. J. Chem. Technol. Res. 6 (2014) 1746–1748.
34] K. Sivaranjani, A. Verma, C.S. Gopinath, Green Chem. 14 (2012) 461–471.
35] V. Murgia, E. Sham, J.C. Gottifredi, E.M. Farfantorres, Latin Am. Appl. Res. 34
Acknowledgments
(
2004) 75–82.
We are thankful to UGC-DAE-CSR, Indore, for the XRD, XPS, FT-
IR, and Raman recordings. The group is also thankful to IIT
MADRAS for SEM and EDAX recordings.
[36] A.M. Elfadly, A.M. Badawi, F.Z. Yehia, Y.A. Mohamed, A.M. Betiha, A.M. Rabie,
Egypt J. Pet. 22 (2013) 373–380.
37] G. Silversmit, D. Depla, H. Poelman, G.B. Marin, R.D. Gryse, J. Electron.
Spectrosc. Relat. Phenom. 135 (2004) 167–175.
[
[
[
38] M. Demeter, M. Neumann, W. Reichelt, Surf. Sci. 454–456 (2000) 41–44.
39] L.E. Briand, O.P. Tkachenko, M. Guraya, X. Gao, I.E. Wachs, W. Gru1nert, J. Phys.
Chem. B 108 (2004) 4823–4830.
Appendix A. Supplementary material
[
[
[
40] Q.H. Wua, A. Thissena, W. Jaegermanna, M. Liub, Appl. Surf. Sci. 236 (2004)
473–478.
41] B.M. Reddya, B. Chowdhury, E.P. Reddya, A. Fernández, Appl. Catal. A 213
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2016.10.033.
(
2001) 279–288.
42] J. Li, J. Xu, Huang, J. Cryst. Eng. Commun. 16 (2014) 375–384.
References
[43] Z. Quan-yao, C. Wen, X. Qing, M. Li-qiang, J. Wuhan, Univ. Technol. Mater. Sci.
Ed. 18 (2003) 27–29.
[
44] T. Rostamzadeh, S. Adireddy, X. Zhang, B. Koplitz, D.B. Chrisey, J. B. Wiley
Chem. NanoMater. 2 (2015) 54–60.
45] V.V. Kaichev, G.Y. Popova, Y.A. Chesalov, A.A. Saraev, S.A. Zemlyanov, K.A.
Beloshapkin, R. Schlögl, T.V. Andrushkevich, V.I. Bukhtiyarov, J. Catal. 311
[
[
1] A. Martin, V.N. Kalevaru, Chem. Cat. Chem. 2 (2010) 1504–1522.
2] A. Teimouri, B. Najari, A.N. Chermahini, H. Salavatia, M.F. Najafabadic, RSC Adv.
[
4
(2014) 37679–37686.
[
3] R.K. Grasselli, in: G. Ertl, H. Knçzinger, F. Schth, J. Weitkamp (Eds.), Tenhover
in Handbook of Heterogeneous Catalysis, Wiley-VCH, Weinheim, 2008, pp.
(
2014) 59–70.
[46] G. Gao, R. Meszaros, M. Peng, L. Wondraczek, Opt. Express 19 (2011) 12–18.
[47] I. Rajiha, A.L. Nuaimy, Iraqi J. Vet. Med. 36 (1) (2012) 14–21.
[48] G. Centi, S. Perathoner, J. Catal. 142 (1993) 84–89.
[
3
489–3517.
[
[
[
4] A. Martin, B. Lucke, Catal. Today 57 (2000) 61–70.
5] A. Martin, V.N. Kalevaru, B. Lucke, J. Sans, Green Chem. 4 (2002) 481–485.
6] A. Martin, B. Luck, H. Seeboth, G. Ladwig, E. Fischer, React. Kinet. Catal. Lett. 38
49] P. Sharma, R. Dwivedi, R. Dixit, M.S. Batra, R. Prasad, RSC Adv. 5 (2015) 39635–
39642.
(
1989) 33–38.
[
50] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M.
Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y.
Honda, O. Kitao, V.H.T. Nakai, J.A. Montgomery Jr, J.E. Peralta, F. Ogliaro, M.
Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, T. Keith, R.
Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J.
Tomasi, M. Cossi, N. Rega, J.M. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken,
C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R.
Cammi, C. Pomelli, J.W. Ochterski, R.L. Martin, K. Morokuma, V.G. Zakrzewski,
G.A. Voth, P. Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, O. Farkas, J.B.
Foresman, J.V. Ortiz, J. Cioslowski, D.J. Fox, Gaussian 09 References, Gaussian
Inc, Wallingford CT, Revision B.01, 2010.
[
7] B.H. Babu, K.T.V. Rao, M. Surenda, P.S.S. Prasad, N. Lingaiah, React. Kinet. Mech.
Cat. (2014).
[
[
8] N. Dropka, V.N. Kalevaru, A. Martin, D. Linke, B. Lucke, J. Catal. 240 (2006) 8–17.
9] A. Martin, B. Liicke, G. Wolf, Catal. Lett. 33 (1995) 349–355.
[
[
[
10] V.N. Kalevaru, B. Lucke, A. Martin, Catal. Today 142 (2009) 158–164.
11] X.G. Yong, H. Chi, Indian J. Chem. Technol. 14 (2007) 371–375.
12] L. Xu, Y. Zhang, Y. Deng, Y.Z. Mo, G. Cheng, C. Huang, Mater. Res. Bull. 48 (2013)
3
620–3624.
13] J. Mendialdua, R. Casanova, Y. Barbaux, J. Electron. Spectrosc. Relat. Phenom.
1 (1995) 249–261.
14] Y. Jeon, S.W. Row, A. Dorigtov, S.D. Lee, K. Oh, Y.G. Shul, Korean J. Chem. Eng. 30
8) (2013) 1566–1570.
15] F. Klose, T. Wolf, H. Lorenz, A.S. Morgenstern, Y. Suchorski, M. Piorkowska, H.
Weiss, J. Catal. 247 (2007) 176–193.
16] Y.H. Jang, W.A. Goddard, J. Phys. Chem. B 106 (2002) 5997–6013.
17] S. Pudar, J. Oxgaard, W.A. Goddard, Top. Catal. 50 (2008) 2–18.
18] S. Pudar, J. Oxgaard, W.A. Goddard, J. Phys. Chem. C 114 (2010) 15678–15694.
19] W.A. Goddard, K. Chenoweth, S. Pudar, A.C.T. Duin, M. Cheng, Top. Catal. 54
[
[
[
7
(
[
51] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785–789.
[
52] J.B. Foresman, A. Frisch, Exploring chemistry with electronic structure
methods, second ed., Gaussian Inc, Pittsburgh, PA, 1996.
53] Thermochemistry in Gaussian, J.W. Ochterski, <www.gaussian.com/g_
whitepap/thermo.htm>.
[
[
[
[
[
[
[
54] H. Fu, Z.P. Liu, Z.H. Li, W. Nin, Wang, K.N. Fan, J. Am. Chem. Soc. 128 (2006)
(
2011) 659–668.
1
1114–11123.
55] N.H. Nguyen, T.H. Tran, M.T. Nguyen, M.C. Le, Int. J. Quantum Chem. 110
2010) 2653–2670.
[
[
20] K. Muthukumar, J. Yu, Y. Xu, V.V. Guliants, Top. Catal. 54 (2011) 605–613.
21] E. Rojas, M. Calatayud, M.A. Bañares, M.O.G. Pérez, J. Phys. Chem. C 116 (2012)
(
9
132–9141.
[
56] T. Kim, I.E. Wachs, J. Catal. 255 (2008) 197–205.
[
[
[
22] J. Yu, Y. Xu, V.V. Guliants, Top. Catal. 57 (2014) 1145–1151.
23] K. Sivaranjani, C.S. Gopinath, J. Mater. Chem. 21 (2011) 2639–2647.
24] S.T. Aruna, A.S. Mukasyan, Curr. Opin. Solid State Mater. Sci. 12 (3–4) (2008)
[
57] A.L. Ahmad, S.R.A. Shukor, C.P. Leo, J. Nanosci. Nanotechnol. 6 (2006) 3910–
3914.
[
58] M. Sanati, A. Andersson, Ind. Eng. Chem. Res. 30 (1991) 312–320.
4
4–50.
[
25] E.P. Reddy, R.S. Varma, J. Catal. 221 (2004) 93–101.