Oxidative dehydrogenation of ethylbenzene over Sm2O3-V2O5 system

Article abstract of DOI:10.1246/bcsj.75.463

SM₂O₃/V₂O₅ catalysts were synthesized by wet impregnation of samaria with an aqueous solution of ammonium trioxovanadate(V). The characterization of the prepared samples was carried out using EDX, XRD, FTIR and TG-DTA as well as a measurement of the surface area and pore volume. The catalysts were found to be stable up to 800°C. Both amorphous and crystalline vanadate species were detected in the supported system. Vanadia species became anchored to the surface through an interaction with basic hydroxy groups and Lewis acid sites of the surface. Lewis acid and base sites decreased upon vanadia addition along with a concomitant enhancement of Broonsted acid sites. The oxidative dehydrogenation of ethylbenzene was carried out over Sm/V catalysts, which exhibited a styrene selectivity > 80%. The active species present in the supported system was concluded to be amorphous vanadia in tetrahedral coordination. The higher vanadia-loaded systems exhibited constant styrene selectivity, where crystalline tetraoxovanadates were predominantly formed.

Full text of DOI:10.1246/bcsj.75.463

c. Jpn.
© 2002 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 75, 463–471 (2002)
463
e Chemical
ciety of Ja-
n
e Chemical
ciety of Ja-
n
Oxidative Dehydrogenation of Ethylbenzene over Sm2O3–V2O5 System
Oxidative Dehydrogenation of Ethylbenzene
S. Sugunan et al.
Sankaran Sugunan* and Neeroli Kizhakayil Renuka
Department of Applied Chemistry, Cochin University of Science and Technology, Kochi 682 022, India
(Received July 23, 2001)
02
3
1
SJA8
09-2673
237
Sm2O3/V2O5 catalysts were synthesized by wet impregnation of samaria with an aqueous solution of ammonium
trioxovanadate(Ⅴ). The characterization of the prepared samples was carried out using EDX, XRD, FTIR and TG-DTA
as well as a measurement of the surface area and pore volume. The catalysts were found to be stable up to 800 °C. Both
amorphous and crystalline vanadate species were detected in the supported system. Vanadia species became anchored to
the surface through an interaction with basic hydroxy groups and Lewis acid sites of the surface. Lewis acid and base
sites decreased upon vanadia addition along with a concomitant enhancement of Brønsted acid sites. The oxidative de-
hydrogenation of ethylbenzene was carried out over Sm/V catalysts, which exhibited a styrene selectivity > 80%. The
active species present in the supported system was concluded to be amorphous vanadia in tetrahedral coordination. The
higher vanadia-loaded systems exhibited constant styrene selectivity, where crystalline tetraoxovanadates were predomi-
nantly formed.
01
01
.2
Modern technology is searching for efficient catalytic pro-
cesses that either eliminate or minimize the production of un-
wanted chemicals, giving emphasis to selectivity along with
catalytic activity. In recent years, supported vanadia catalysts
have attracted much attention, which boast wide applications
in the area of partial or selective oxidation reactions of indus-
trial importance. A large number of research work has been
devoted to provide insight into the nature and reactivity of sup-
ported vanadia catalysts. The nature of the selected support
material plays a crucial role in determining the active vanadia
species. i.e., its chemical composition, crystallographic modi-
fication, morphology, and textural properties. Other parame-
ters, such as preparation method and catalyst pretreatment,
also influence the physical and chemical properties of the re-
sulting catalysts. A preferential formation of compounds was
observed on basic supports, whereas acidic supports favored
V2O5 crystallites.1,2 Reactions catalyzed by supported vanadia
systems include the oxidation of hydrocarbons,3,4 the ammoxi-
dation of aromatics5 and the selective catalytic reduction of ni-
trogen oxides.6 Earlier studies have established that metal ox-
ides, which are more acidic than vanadia, favor the formation
of acidic products, whereas oxidative dehydrogenation reac-
tions, which produce basic substances, proceed well on sys-
tems of basic oxides when supported with vanadia.7
Even though catalyst systems containing vanadia have been
extensively studied, little attention has been paid to rare earth-
supported vanadia analogues. In the present paper, we report
on the synthesis and characterization of samaria-supported va-
nadia. The effect of vanadia incorporation on the textural
properties of samaria is briefly discussed. The adsorption of
probe molecules and a cyclohexanol decomposition reaction
have been adopted to explore the acid-base property of the sys-
tem. The catalytic behavior of the supported system towards
the oxidative dehydrogenation of ethylbenzene has also been
studied.
The oxidative dehydrogenation of ethylbenzene is a reaction
of much industrial importance owing to the commercial appli-
cations of styrene. Industrially, styrene is prepared by the
alkylation of benzene using ethylene. Generally used catalyst
systems for the preparation of styrene from ethylbenzene com-
prise SiO2-Al2O3,8,9 V–Mg–O,10 and SnO2–P2O5 systems.11
Among the various systems screened, V–Mg–O systems were
found to be highly selective for the preparation of styrene.10
Experimental
1. Catalyst Preparation. Sm(NO3)3 (purity 99.9%) and
NH4VO3 (purity 99.5%) were obtained from Indian Rare earth
Ltd., Udyogamandal, Kerala, and S. D. Fine Chem. Ltd., Boisar,
respectively. Samarium oxide was prepared through precipitation
by a hydroxide method using a 1:1 ammonia solution.12 Support-
ed vanadia catalysts were prepared by the conventional wet im-
pregnation method (excess solvent technique), as reported by
Kanta Rao and co-workers.13 Catalysts with 0, 3, 7, 11, and 15
percentages of vanadia were prepared and designated as S, V3,
V7, V11, and V15, respectively. The catalysts were sieved so as to
obtain particles of mesh size < 100 microns. Prior to each exper-
iment, the catalyst was activated in a current of dry air at 500 °C.
2. Catalyst Characterisation.
2.1. Analytical Data.
The chemical composition of the supported systems was deter-
mined by using energy-dispersive X-ray (EDX) analysis (Ste-
reoscan 440 Cambridge, UK). Identification of different phases
was done using a Rigaku D-Max C X-ray Diffractometer, with Ni-
filtered Cu Kα radiation (λ = 1.5404 Å). The mean crystallite
size of the sample was determined from the broadening of the X-
ray diffraction peak, following the Scherrer method. The FTIR
spectra of the powder samples were recorded over the range
4000–400 cm−1 by the KBr disk method using a Shimadzu DR
8001 instrument. A Micromeritics Flowprep-060 instrument was
used to determine the specific surface area by the BET method
with nitrogen gas used as an adsorbate. The pore volume and
pore-size distribution of the catalysts were measured by mercury

464 Bull. Chem. Soc. Jpn., 75, No. 3 (2002)
Oxidative Dehydrogenation of Ethylbenzene
Table 1. Physico Chemical Characteristics of the Supported Systems
Percentage
of vanadia
Crystallite
size/nm
Surface
Pore volume
/cm3 g−1
Limiting amount
Catalyst
area/m2 g−1
adsorbed/mmol m−2
TCNQ
5.50
5.26
3.84
2.28
2.25
Chloranil
4.37
4.26
3.32
1.91
Sm2O3
V3
V7
V11
V15
0
3.37
6.85
11.27
14.59
19.94
9.39
9.63
8.61
6.51
20.78
17.19
21.91
25.54
25.48
1.41
0.89
0.64
0.67
0.59
1.87
porosimetry using Quantachrome, Auto scan-92 porosimetry
(USA). The thermal stability of the samples was established by a
TG-DTA analysis by using a Shimadzu TGA-50 instrument with a
heating rate of 10 °C/min in a nitrogen atmosphere.
2.2. Acid Base Properties.
2.2.1 Electron Donating
Properties For determining the basicity of the system, adsorp-
tion studies of electron acceptors (EA) were performed using the
following probe molecules: 7,7,8,8-tetracyanoquinodimethane
(TCNQ) [Merk-Schuchandt], 2,3,5,6-tetrachoro benzoquinone
(chloranil) [Sisco Research Laboratories Pvt. Ltd.], and p-dini-
trobenzene (PDNB) [Merk-Schuchandt] with electron affinity val-
ues of 2.84, 2.40, and 1.77 eV, respectively. The following proce-
dure was adopted for adsorption studies. Previously activated
samples were stirred with a solution of EA in acetonitrile for 4 h
in an airtight stirrer. The amount of the electron acceptor ad-
sorbed on the catalyst surface was estimated from the difference in
the concentration of EA in solution before and after adsorption,
which was measured by means of a Shimadzu 160-A spectropho-
tometer (λmax of the electron acceptors in acetonitrile are 393.5,
288, and 262 nm for TCNQ, chloranil, and PDNB, respectively).
The adsorption isotherms were of the Langmuir type. From the
Langmuir adsorption isotherms, the limiting amount of the elec-
tron acceptor adsorbed was obtained.
2.2.2 Temperature-Programmed Desorption (TPD) Stud-
ies. About 0.5 g of a pelletised sample was activated at 300 °C
inside the reactor under nitrogen flow for 2 h. After cooling to
room temperature, the sample was allowed to adsorb a definite
amount of ammonia in the absence of a carrier gas, and the system
was allowed to attain equilibrium. The excess and physisorbed
ammonia was flushed out by a current of nitrogen, and under a
controlled temperature program the amount of ammonia leached
out was determined volumetrically.
2.2.3 Cyclohexanol Decomposition Reaction.
The reac-
Fig. 1. X-ray diffraction patterns of catalysts (a) Sm2O3; (b)
V3; (c) V7; (d) V11; (e) V15; (f) V2O5.
tion was performed in a gas-phase silica reactor and the products
were analysed by a GC fitted with a 6ꢀ × (1/8 ) stainless-steel col-
umn packed with 5% NPQSB + H3PO4 on anachrom A 80/100
mesh.
2.3. Catalytic Activity Study.
Oxidative dehydrogenation
supported systems. The elemental composition of the support-
ed system determined by an EDX analysis, is presented in Ta-
ble 1. As revealed by the XRD patterns (Fig. 1), the intensity
of the peaks became lowered, suggesting a decrease in the
crystallinity of samaria as a result of vanadia incorporation,
i.e., the interaction between V2O5 and the support hinders the
transition of samaria from the amorphous phase to the crystal-
line phase. Prominent peaks due to crystalline samaria ap-
peared at 2θ values of 28.0, 32.5, 48.0, and 55.5°. Additional
peaks appeared at 2θ values of 18.6, 24.6, 31.0, 33.2, and
49.0°, which were identified to be due to samarium tetraoxo-
vanadate. This agrees with the reports by Corma et al. that ba-
sic oxides preferably form compounds with vanadia.4 The rel-
of ethylbenzene was carried out in the vapor phase. Before each
experiment, the catalyst was activated in a current of dry air at 500
°C, and then brought to the reaction temperature in the presence of
a nitrogen flow. Ethylbenzene was then introduced into the carrier
stream from a saturator containing liquid ethylbenzene. The mass
balance was noted each time and liquid products were analyzed
using a Shimadzu GC-15A gas chromatograph fitted with a xylene
master capillary column and FID.
Results and Discussion
1. Catalyst Characterization.
Vanadia impregnation imparted a pale-yellow colour to the
1.1. Analytical Data.

S. Sugunan et al.
Bull. Chem. Soc. Jpn., 75, No. 3 (2002) 465
ative intensity of the peaks supports a gradual increase of the
samarium tetraoxovanadate percentage along with an increase
in the vanadia content. The characteristic peaks of crystalline
vanadia were not observed in the supported catalysts, even at
higher loadings, due to the basic nature of samaria. The aver-
age crystallite size calculated by employing Scherrer’s formula
is given in Table 1.
The FTIR spectra of the supported systems are provided in
Fig. 2. The spectra of the different vanadia-loaded systems
were similar, except that the intensity of the bands due to vana-
dia was low in lower vanadia-loaded systems. For a compari-
son, the spectra of a V15 sample along with pure samaria and
V2O5 are given in Fig. 3. The characteristic band at 1020 cm−1
due to V2O5 is not observed in the spectrum of the supported
system thus confirming the results of an X-ray diffraction
study. However, micro-crystals of V2O5 may be present,
which cannot be detected by the X-ray diffraction pattern and
an IR spectral analysis. The weak band at 1062.7 cm−1 is at-
tributed to the VwO stretching mode arising from the surface
vanadium(ꢁ) oxide species. The probability of dioxo struc-
tures in the surface species is ruled out, which when present
exhibits several combination bands.14 Surface vanadia was
identified as being an isolated species, since the polyvanadates
exhibit VwO stretching frequencies in the 1000–950 cm−1 re-
gion as suggested by Frederickson et al.15 The characteristic
peaks due to vanadia are more clear in the DRIFT spectra giv-
en in Fig. 4. A tetrahedral geometry was assigned to isolated
monooxo species, because it has been well established that iso-
lated monooxo species exist as tetrahedral vanadium(ꢁ) oxide
units [(OwV(O–M)3 type, where “M” is the support metal ion]
3−
in the supported systems.16–18 The VO4 entity that results
from tetraoxovanadate justifies the presence of the broad band
observed in the range 700–900 cm−1.19 A broad band typical
for surface hydroxy groups was detected at around 3500 cm−1.
In the case of pure samaria, the corresponding broad band ap-
peared at 3462 cm−1, whereas for vanadia-loaded systems the
above-mentioned band was located at 3448 cm−1. Hence, it is
plain that, while comparing with Sm2O3, there is a shift of the
band to lower frequency for the supported system. The IR
band at higher frequency has been assigned to the most basic
hydroxy group, and the decrease in frequency of the surface
hydroxy group has been associated with increasing acidity.20
Hence, the aforementioned shift indicates the participation of
basic hydroxy groups in the bond formation with vanadia. In
addition, this leads us to the conclusion that vanadia incorpora-
tion induces an enhancement of the Brønsted acidity.
A thermal-analysis curve exhibited two distinct weight loss-
es, as confirmed by the DTG pattern. The initial weight loss in
the 100–200 °C region corresponds to a loss of physisorbed
water. A further weight loss, which occurs in the 350–450 °C
range, accounts for the loss of surface hydroxy group due to a
high percentage of rare earth oxide in supported samples. Be-
sides, tetragonal tetraoxovanadate formation also takes place
in this region, as reported by Oliveira and co-workers.21 The
DTA profile for the supported system showed two endothermic
peaks, supporting the above observation. Vanadia loss takes
place only after 800 °C, indicating a high stability of the sup-
ported system.
Fig. 2. FTIR spectra of the supported vanadia systems (a)
V3; (b) V7; (c) V11; (d) V15.
Fig. 4. Diffuse reflectance infrared (DR-IR) spectrum of V15
system.
Fig. 3. FTIR spectra of (a) V2O5 ;(b) V15 system ;(c) Sm2O3.

466 Bull. Chem. Soc. Jpn., 75, No. 3 (2002)
Oxidative Dehydrogenation of Ethylbenzene
From Table 1, it is apparent that there are no marked varia-
tions in the surface area, indicating that the impregnation pro-
cedure didn’t appreciably affect the morphology of samaria.
Still, there is an initial marginal loss, which can be accounted
for by the reduction of pore radii of bigger pores, as evident
from the pore-volume distribution curves. Further, vanadia ad-
dition prevents the agglomeration of particles, which leads to
an enhancement of the surface area.
2. Acid Base Properties. 2.1. Ammonia TPD Studies.
Ammonia can be adsorbed on an oxide surface through hydro-
gen bonds or through dipolar interaction yielding the total
acidity of the system (both Brønsted and Lewis type). Figure 5
shows acid-strength distribution profiles of pure samaria and
supported analogues. The area of the peaks corresponds to the
acid amount in that particular temperature range. As evident
from the figure, pure samaria possesses a small amount of
weak acid sites and an appreciable amount of strong acid sites.
Vanadia impregnation increased the amount of ammonia des-
orbed at a lower temperature region, whereas a considerable
reduction is noticed concerning the amount of ammonia des-
orbed at higher temperatures. It is generally accepted that
evacuation of metal oxide surfaces at 400 °C removes most of
the adsorbed probe molecules from the Brønsted acid sites.22,23
In the present case, the amount of ammonia desorbed up to 300
°C is taken as a measure of the Brønsted acidity. The amount
of ammonia desorbed at higher temperatures is considered to
be due to ammonia bound on Lewis acid sites.
A gradual decrease in the amount of Lewis acidity along
with an increase of the vanadia percentage is evident from the
TPD curves. According to Kantcheva et al., sites responsible
for Lewis-type acidity are coordinately unsaturated cations of
the support.24 Hence, it is obvious that Lewis acid sites result-
ing from the exposed cations of samaria are being utilized for
the formation of bonds with vanadia, as reported by Das et al.16
Besides, the figure reveals that the incorporation of V2O5 re-
markably increases the concentration of Brønsted acid sites,
which is consistent with earlier reports.25,14 This is also in
agreement with the conclusions derived from the FTIR spectra,
which predicted an increase of acidic hydroxy groups. Turek
et al. proposed a crowding effect of the supported species as a
necessary condition for the generation of Brønsted acidity at
higher vanadia loadings.26 They attributed Brønsted acid sites
at high vanadia loadings to an increase in the surface density of
molecularly dispersed species. According to them, the sim-
plest conceivable model of the Brønsted acid site that can be
proposed here consists of two surface species sharing a com-
mon H+ located between two oxygen atoms belonging to two
different V–O–M groups. Brønsted acidity is created indepen-
dent of the structure of H+ connected surface metal oxide
units, if they are in a privileged configuration to each other.
However, from the figure, it is clear that no considerable
change in the amount of Brønsted acid sites is found as the
weight percentage of vanadia is increased from 3 to 15. This
implies that the crowding effect might have been reached by
the addition of 3 wt% vanadia, itself. Further, vanadia addition
may be primarily contributing towards samarium tetraoxovan-
adate.
2.2. Electron-Donating property—Measure of Lewis
Basicity. The utility of electron acceptor adsorption for
studying the electron-donating property has been well estab-
lished,27–29 which provides useful information regarding the
Lewis basicity of the catalysts. In this method, the strength
and distribution of basic sites are followed by the adsorption of
electron acceptors (EA). The electron-donating capacity de-
pends on the nature of the donor sites and the electron affinity
of the electron acceptors used. The extent of electron transfer
decreases along with a decrease in the electron affinity value of
the electron acceptor. An electron acceptor with a high elec-
tron affinity value become adsorbed on both weak and strong
donor sites, while the adsorption of EA with low electron affin-
ity occurs only at strong basic sites.
For supported systems, PDNB adsorption was too negligi-
ble, indicating the absence of very strong basic sites. Electron-
donor adsorption imparted characteristic coloration to the cata-
lyst surface, suggesting the formation of radical anions. On
pure Sm2O3, TCNQ adsorption developed a bluish-green color
and chloranil developed a pale-pink color. In the case of sup-
ported catalysts, which were pale-yellow colored, the colors
generated were pale-green and pale-grey, respectively. The
limiting amount of electron acceptors, which is a measure of
the Lewis basicity, can be determined from the Langmuir plot
(Fig. 6). Since TCNQ is a stronger EA, leading to adsorption
on both weak and strong donor sites, the limiting amount ad-
sorbed was higher in the case of TCNQ adsorption, while the
limiting amount adsorbed for chloranil was low (Table 1). A
survey of the limiting amount adsorbed clearly indicates that
the basicity shows a decreasing trend with a progressive addi-
tion of vanadia. A similar observation as a result of vanadia
addition was reported by Le Bars and co-workers.30 It was ac-
cepted by earlier workers that the factors responsible for the
electron-donating capacity are surface hydroxy ions and sur-
face O2− centers. At higher activation temperatures, trapped
electron centers also function as electron donor sites.31,32
However, it has been reported that, electron-defect centers are
created only at an activation temperature above 500 °C.33 Thus
the present case, a major contribution towards basic sites arises
from surface hydroxide ions resulting from the addition of va-
nadia. A decreasing amount of the electron acceptor adsorbed
Fig. 5. Temperature programmed desorption curves of am-
monia over Sm/V catalysts.

S. Sugunan et al.
Bull. Chem. Soc. Jpn., 75, No. 3 (2002) 467
brings about a dramatic improvement in the selectivity of de-
hydration products, which were catalyzed by Brønsted acid
sites, as reported by Bezouhanava et al.35 An enhancement of
the Brønsted acidity by vanadia incorporation is apparent from
these data. Further, the reduced selectivity of cyclohexanone
on a supported system indicated a decrease in both the acid and
base sites which take part in dehydrogenation. These observa-
tions lend support to a proposed decrease of Lewis acid and
base sites with a concomitant increase of Brønsted acid sites,
as concluded from the adsorption of probe molecules. More-
over, from the results presented in Table 2, it is clear that there
is no remarkable change in the selectivity of dehydration prod-
ucts as the composition of vanadia varies from 3 to 15 wt%,
thus substantiating the role of vanadia in vanadate formation at
higher concentrations.
3. Catalytic Activity Studies. The effect of vanadia on
the catalytic performance towards the oxidative dehydrogena-
tion (ODH) of ethylbenzene is given in Table 3. Along with
styrene, small amounts of benzene, toluene, and carbon oxides
are also detected among the products. The conversion of ethyl-
benzene showed a gradation with the vanadia content in the
system. The selectivity towards styrene was too poor for pure
vanadia, which catalyzed almost the complete oxidation of eth-
ylbenzene, as evident from the greater percentage of carbon
oxides. From the ODH data, it is obvious that selective oxida-
tion activity is greatly influenced by the supporting vanadia.
The styrene selectivity was almost constant at a higher vanadia
percentage, and was found to be independent of the percentage
conversion. The selectivity of toluene and benzene were de-
creased with the concentration of vanadia in the supported cat-
alysts. The reaction was also performed over pure SmVO4.
Pure SmVO4 exhibited negligible activity due to an extremely
small surface area (9.26 m2/g), though the system was fairly
selective.
Fig. 6. Langmuir adsorption isotherms of chloranil over
Sm/V systems.
accounts for the decrease in the surface hydroxide ions as a re-
sult of vanadia addition. Supporting evidence for this conclu-
sion is obtained from the FTIR spectra of the supported sys-
tems, where a shift for the hydroxy band position was noticed.
2.3. Cyclohexanol Decomposition Reaction. The influ-
ence of vanadia loading on the acid-base properties have been
confirmed through studies on the decomposition of cyclohex-
anol over the supported system (Table 2). The amphoteric
character of the alcohol permits its interaction with acidic and
basic sites, leading to the formation of cyclohexanol and cyclo-
hexanone. Both acid and base sites participate in dehydroge-
nation, whereas dehydration takes place with the intervention
of acid sites. Hence, the dehydrogenation rate/selectivity of
cyclohexanone is proportional to both the acidity and basicity,
whereas the dehydration rate/selectivity of cyclohexene is
found to be proportional to the acidity.34 Vanadia addition
Table 2. Cyclohexanol Decomposition Data over Sm/V Catalyst System
Catalyst Conversion/% Product distribution/%
MCP Cyclohexene Cyclohexanone
Selectivity/%
CwC CwO
26.81 73.19
Sm2O3
V3
40.85
38.61
39.59
42.40
43.07
—
10.95
33.73
34.09
35.76
36.34
29.90
3.69
3.38
3.21
3.10
1.19
2.12
3.43
3.63
90.44
91.46
92.42
92.80
9.56
8.54
7.58
7.20
V7
V11
V15
Reaction conditions: catalyst = 2.5 g; reaction temp = 300 °C; feed rate = 6 mL/h,
time on stream = 1 h.
Table 3. Oxidative Dehydrogenation of Ethylbenzene over Sm/V System
Catalyst
Conversion/%
Selectivity/%
Styrene
5.2
77.0
86.7
87.3
86.6
Toluene
Benzene
3.0
3.1
3.0
1.3
C-oxides
90.2
16.0
9.6
10.1
V2O5
V3
V7
V11
V15
35.8
8.1
13.8
14.9
16.6
1.6
3.9
2.7
1.3
1.2
0.9
10.4
Reaction conditions: reaction temperature: 475 °C; feed rate: 6 mL/h; catalyst: 2 g;
time on stream: 1 h; air flow rate: 20 mL/min.

468 Bull. Chem. Soc. Jpn., 75, No. 3 (2002)
Oxidative Dehydrogenation of Ethylbenzene
3.1. Selective Oxidation Activity of the Supported Sys-
tem. The selective oxidation activity of the supported vana-
dia system is evident from Table 3. The behavior of vanadia-
based systems in selective oxidation strongly depends on the
redox property of the vanadium species as well as on the acid
base character of the support and catalyst.36,37 Tagawa et al.
observed that the acid sites of the Hammett Ho values between
1.5 and −5.6 are active for ethylbenzene conversion, which
adsorb ethylbenzene reversibly. Oxidation of the same occurs
on the basic sites of pKa lying between 17.2 and 26.5.9 How-
ever, in Sm/V systems, no such sites were detected by the
Hammett indicator method, leading to the conclusion that the
influence of the acid base properties on the selective oxidation
activity is negligible in the present case. Hence, apart from the
acido basic properties, the redox property plays an important
role in selective oxidation over Sm/V catalyst systems.
molecule, thus improving the selectivity. Consequently, the lo-
cally limited amount of reactive lattice oxygen might be an ex-
planation for the selective oxidation activity of the Sm2O3/
V2O5 system.
As shown in Table 3, above 3% vanadia, the styrene selec-
tivity was almost constant for the supported system. This can
be explained based on the preferential formation of SmVO4 at
higher vanadia percentages. Up to 7% vanadia, surface species
may contribute to the styrene selectivity. The completion of
monolayer coverage occurs at about 7% vanadia. At higher va-
nadia concentration, vanadate has a prominent role, which ex-
hibits a constant selectivity towards styrene, which is in accor-
dance with Chang et al., who reported constant styrene selec-
tivity for tetraoxovanadates.10
The interaction of ethylbenzene with acidic and basic cen-
tres of the catalyst surface resulted in the production of ben-
zene and toluene. Wang et al. suggested that strong acidic sites
could abstract α-hydrogen of ethylbenzene, thus facilitating
the formation of benzene.46 The decrease in the benzene selec-
tivity with an increase in the vanadia content can be rational-
ized in terms of the reduction of strong acid sites, which has
also been confirmed through ammonia desorption studies.
Likewise, Krause proposed the role of strong basic sites for
toluene formation.47 Strong basic sites on a catalyst preferably
abstract beta hydrogen, leading to a cleavage of the C–C bond
of the side chain resulting in a high yield of toluene. Accord-
ing to the results obtained from electron-donor property stud-
ies, the number of strong basic sites is becoming reduced due
to vanadia impregnation. The expected decrease of the toluene
percentage is consistent with the results presented in Table 3.
3.2. Effect of the Reaction Temperature. V7 is taken
as a representative of a supported system to examine the effect
of various reaction parameters on the activity and selectivity of
the catalysts. ODH of ethylbenzene was performed at different
temperatures over V7 to understand the effect of the reaction
temperature on the system. The dependence of the catalytic
activity on the reaction temperature is given in Table 4. There
was an enhancement in the conversion of ethylbenzene as the
temperature was increased from 475 to 525 °C. However, no
remarkable change in the selectivity of styrene was observed
over the above temperature range. A similar observation was
also found with magnesium orthovanadate in the ODH of eth-
ylbenzene.10
According to Sachtler et al.,38,39 the selectivity in the oxida-
tion reaction is determined by two factors: 1) by the intrinsic
activity of lattice oxygen and 2) by their availability. The in-
trinsic activity depends on the nucleophilicity of the lattice ox-
ygen, which was proposed as a site capable of abstracting hy-
drogen atoms as hydrogen ions.40 This in turn depends on the
coordination environment of the vanadium ion in the supported
system. Grzybowska et al. have shown that the V5+ ion exists
in isolated VO4 tetrahedra and tetraoxovanadates.7 It is accept-
ed that vanadia in a tetrahedral coordination is most active in
oxidative dehydrogenation.4, 41–44 For the present system, both
species having a tetrahedral geometry, this condition is satis-
fied, indicating tetrahedral vanadia as the active species present
in the supported system. Oxygen attached to a V5+ in tetrahe-
dral coordination is found to be highly nucleophilic.7
Another factor on which the selectivity depends is the avail-
ability of lattice oxygen at the reaction site, which is expressed
as the average number of oxygen molecules that react with
each hydrocarbon molecule (also termed as average oxygen
stoichiometry, AOS).41 The number of vanadium(ꢁ) oxide
units that can effectively interact with the adsorbed hydrocar-
bon unit determines the AOS in the case of a particular sup-
ported system. If only a limited number of lattice oxygens are
available, the oxidation stops at a particular level, resulting in
partial oxidation products. V2O5 has a layer structure, and all
of the layers take part in the oxidation reaction because the dif-
fusion of oxygen from other layers of the lattice is much faster
in V2O5,45 leading to complete oxidation. In the case of a sup-
ported vanadia system, there is no large reservoir of bulk oxy-
gen available to abstract so many oxygens. In the present case,
VO4 units are isolated (sufficiently apart), which supplies only
a limited number of oxygen atoms to react with a hydrocarbon
3.3. Effect of Flow Rate. In another series of experi-
ments, the effect of the contact time on the catalytic activity
and the selectivity of products was studied. The activity and
selectivity variations as a function of the feed rate over V7 are
shown in Fig. 7. The conversion decreased along with an in-
Table 4. Influence of Reaction Temperature on Conversion and Selectivity
Temperature/°C
Conversion/%
Selectivity/%
Toluene Benzene
Styrene
84.6
85.0
C-oxides
9.7
10.0
475
500
525
13.1
18.0
24.0
2.7
2.0
1.2
3.0
3.0
3.5
83.3
12.0
Reaction conditions: catalyst, V7:2 g; flow rate: 6 mL/h; time on stream: 1 h; air flow
rate: 20 mL/min.

S. Sugunan et al.
Bull. Chem. Soc. Jpn., 75, No. 3 (2002) 469
3.4. Deactivation Study. Time-on-stream studies were
conducted on V7 to establish the stability of the supported cat-
alysts. The results are shown in Fig. 8. The conversion was
more or less constant, although a slight deactivation was ob-
served after 6 h. The selectivity also remained unaltered dur-
ing a prolonged treatment with reactants. The catalyst turned
dark grey after continuous use, probably due to carbon depos-
its and a reduction of the catalyst. The original color could be
regained by treating the catalyst in flowing air at 400 °C for 2
h.
3.5. Mechanism. In the oxidative dehydrogenation of
ethylbenzene, one of the most probable suggested mecha-
nisms, consists of an abstraction of “H” from the ethylbenzene
by the lattice oxygen to form styrene, and a reoxidation of the
catalyst by gas-phase oxygen.48 Similar conclusions regarding
the mechanism have been reached by Tagawa et al.9 To obtain
insight into the reaction pathway over Sm/V systems, the reac-
tion was carried out in the absence of air (Table 5). The fact
that the catalytic activity and percentage conversion were sig-
nificantly reduced under non-oxidative conditions pointed out
to the role of gaseous oxygen in the reaction. The reason for
the reduced conversion may be a reduction of the catalyst. The
formation of styrene reduces the catalyst system, and the pos-
sibility of reoxidation is limited under non-oxidative condi-
tions. The presence of air increases the conversion, which in-
dicates that the presence of gaseous oxygen favors the redox
operation of the catalyst by supplying oxygen needed for hy-
drogen abstraction by reoxidising the catalyst. Hence, the
aforesaid mechanism (Mars and van Krevelen) may be operat-
ing here, which can be represented as
Fig. 7. Effect of flow rate on the conversion of ethylbenzene
and selectivity of styrene.
EB + [O] → ST + [ ],
[
] + 1/2 O2 → [O],
where [O] is the lattice oxygen and [ ] is the lattice vacan-
cy.
Fig. 8. Deactivation studies conducted over V7.
The oxygen-supplying entity in the case of the ODH reac-
tion is a matter of conflict, even now. Several oxygen species
were suggested as the “H” abstracters, including terminal
VwO, bridging oxygen of M–O–V bond, and Brønsted acid
sites.7 However, studies conducted on alumina exclude VwO
from the list, since such an active site is not necessary to create
an active site.7 This is supported by the fact that the activity
and selectivity do not vary much with the percentage of vana-
dia, even though the VwO species at higher and lower vanadia-
loaded systems differ widely. According to reports, it is the
highly nucleophilic bridging oxygen that furnishes active and
selective sites in the case of tetraoxovanadate systems, in gen-
eral. Hence, the possibility of bridging oxygen acting as the
crease in the feed rate i.e., at a low contact time. The selectivi-
ty towards styrene was increased along with an increase in the
feed rate. A small feed rate would increase the residence time
of the reactant on the catalyst surface. A high conversion of
ethylbenzene at higher contact times (small feed rates) can be
attributed to the higher probability of the reactive adsorption.
The possibility of product readsorption and the production of
more oxygenated products are probable at low flow rates,
which leads to less selectivity, as is evident from the figure.
This is of prime importance while considering the highly mo-
bile lattice oxygen of samaria, which replenishes the con-
sumed oxygen at the reaction site, leading to a high percentage
of carbon oxides.
Table 5. Dehydrogenation of Ethylbenzene over Sm/V System
Catalyst
Conversion/%
Selectivity/%
Styrene
82.9
90.8
90.0
87.9
Toluene
Benzene
2.8
1.4
1.5
1.8
C-oxides
4.6
V3
V7
V11
V15
6.4
6.8
7.2
7.5
9.8
4.0
4.0
4.0
3.7
3.4
3.2
Reaction conditions: reaction temperature: 475 °C; feed rate: 6 mL/h; catalyst, V7: 2 g; time on stream: 1 h.

470 Bull. Chem. Soc. Jpn., 75, No. 3 (2002)
Oxidative Dehydrogenation of Ethylbenzene
active species that abstract the hydrogen from ethylbenzene is
tentatively suggested for the present system. This explains the
negligible activity variation with the increase of vanadia com-
position in the supported system. The formation of amorphous
vanadium(ꢁ) oxide entities might not increase with the per-
centage of vanadia. However, the probability of Brønsted sites
as the active sites cannot be ruled out here, since it was ob-
served from ammonia desorption studies that vanadia addition
enhances the Brønsted acidity from TPD studies using ammo-
nia.
In the ODH of ethylbenzene, the activation of a C–H bond
should be considered first, which is often regarded as the rate-
determining step.9 In the activation of hydrocarbons contain-
ing π electrons, the abstraction of H atoms on an oxide centre
is facilitated by the donation of π electrons to a Lewis site,
which leads to a weakening of the C–H bond (allylic type oxi-
dation). The oxygen in the catalyst abstracts the H in ethylben-
zene. The hydrocarbon activation centre is thus considered to
be a couple of Mn+–O, which is regarded as an acid-base pair.
Reduced vanadium ions chemisorb gaseous oxygen reversibly,
and are converted to lattice O2−. The dissociation of molecular
oxygen with electron transfer according to overall reaction is
shown below.49,50
The authors wish to acknowledge their sincere gratitude to
CSIR, New Delhi for the award of SRF to N. K. Renuka and
AICTE (New Delhi) for the financial assistance.
References
1
(1988).
2
C. J. Ottamari and A. Anderson, Catal. Today, 3, 211
T. J. Dines, C. H. Rochester, and A. M. Wars, J. Chem.
Soc., Faraday Trans., 87, 1473 (1991).
J. M. Lopez-Nieto, G. Kremenik, and J. G. Fiero, Appl.
Catal., 61, 235 (1990).
3
4
A. Corma, J. M. Lopez-Neito, N. Paredes, M. Parez, Y.
Shen, H. Cao, and S. L. Suib, Stud. Surf. Sci.Catal., 72, 213
(1992).
5
Y. Nagakawa, T. Ono, H. Miyata, F. J. Hatayama, and Y.
Kubokawa, J. Chem. Soc., Faraday Trans., 1, 79, 2929 (1983).
H. Miyata, M. Kohno, T. Ohno, F. J. Hatayama, and Y.
Kubokawa, J. Chem. Soc., Faraday Trans., 1, 85, 3663 (1989).
6
7
(1997).
8
(1982).
9
B. Grzybowska-Swierkosz, Appl. Catal. A: Gen., 157, 409
T. Tagawa, T. Hattori, and Y. Murakami, J. Catal., 75, 56
T. Tagawa, T. Hattori, and Y. Murakami, J. Catal., 75, 66
(1982).
−
O
2 gas ꢀ O2 ads ꢀ O2 ads → 2O−
10 W. S. Chang, Y. Z. Chen, and B. L. Yang, Appl. Catal. A:
Gen., 124, 221 (1995).
11 Y. Murakami, K. Iwayama, H. Uchida, T. Hattori, and T.
Tagawa, J. Catal., 71, 257 (1981).
12 “Encyclopedia of Industrial Chemical Analysis,” ed by F.
D. Snell and L. S. Ettre, Interscience, NewYork (1973), Vol. 17, p.
475.
13 K. V. Narayana, A. Venugopal, K. S. Rama Rao, V. Venkat
Rao, S. Khaja Masthan, and P. Kanta Rao, Appl. Catal. A, 150,
269 (1977).
14 I. E. Wachs, Catal. Today, 27, 437 (1996).
15 L. D. Frederickson, Jr. and D. M. Hausen., Anal. Chem.,
35, 818 (1963).
16 N. Das, H. Eckert, H. Hu, I. E. Wachs, J. Walcer, and F.
Fehre, J. Phys. Chem., 97, 8240 (1993).
17 H. Eckert and I. W. Wachs, J. Phys. Chem., 93, 6796
(1989).
18 Xingtao Gao, J. L.G. Fierro, and I. E. Wachs, Langmuir,
15, 3169 (1999).
It is clearly apparent from Table 5 that the selectivity towards
styrene is increased under non-oxidative reaction conditions,
revealing that gaseous oxygen participates in the non-selective
oxidation of hydrocarbons. The decrease of the styrene selec-
tivity in the presence of air may not be due to an increase of the
ethylbenzene conversion, because the styrene selectivity does
not depend on the conversion. An alternative suggestion for
the high styrene selectivity under the non-oxidative condition
is the absence of gaseous oxygen. A reduction of the styrene
selectivity in the presence of gaseous oxygen is also reported
by Chang et al.10 Yang et al.51 and Shakhnovich et al.52 also
proposed a similar degradation of the dehydrogenation product
by gaseous oxygen.
Conclusion
Vanadate species both in amorphous and crystalline forms
were formed by supporting vanadia on samaria by wet impreg-
nation. Surface vanadia existed as tetrahedral units in the sup-
ported system. Lewis acid and base sites of the support were
reduced upon vanadia incorporation, concomitant with an en-
hancement of the Brønsted acid sites. Surface studies using
different experimental methods indicated the participation of
basic OH groups and Lewis-acid sites of the support in bond
formation with vanadia species. Vanadia created redox sites on
samaria. In the oxidation of ethylbenzene, the activity for par-
tial oxidation was enhanced through supporting vanadia on sa-
maria. Vanadium ion in tetrahedral coordination was found to
be selective towards styrene, and the active site was concluded
to be the bridging oxygen of the surface vanadium(ꢁ) oxide
species. The higher vanadia-loaded systems exhibited con-
stant styrene selectivity, where crystalline tetraoxovanadates
were predominantly formed.
19 J. A. Gadsden, “IR Spectra of Minerals and Related Com-
pounds,” (1975), p. 26.
20 H. P. Boehnm and H. Knozinger, “Catalysis,” ed by J. R.
AndersonandM. Moudart, Springer, Berlin (1983), Vol. 4, Chap. 2.
21 H. P. Oliveira, F. J. Anaissi, and H. E. Toma, Mat. Res.
Bull., 33, 1783 (1998).
22 A. Auroux and A. Gervasini, J. Phys. Chem., 94, 6371
(1990).
23 F. Hatayama, T. Ohno, T. Maruoka, T. Ono, and H. Miyata,
J. Chem. Soc., Faraday Trans., 87, 2629 (1991).
24 M. M. Kantcheva, K. I. Hadijiivanov, and D. G. Klissurski,
J. Catal., 134, 299 (1992).
25 J. R. Sohn, S. G. Cho, Y. I. Pae, and S. Hayashi, J. Catal.,
159, 170 (1996).
26 A. M. Turek, I. E. wachs, and E. De Canio, J. Phys. Chem.,
96, 5000 (1992).
27 K. Esumi and K. Meguro, J. Colloid Interface Sci., 66 (1),

S. Sugunan et al.
Bull. Chem. Soc. Jpn., 75, No. 3 (2002) 471
192 (1978).
(1968), p. 454.
28 S. Sugunan and G. D. Rani, J. Mat. Sci. Lett., 10, 887
(1991).
40 A. Bielanski and J. Haber, “Oxygen in Catalysis,” Marcel
Dekker, New York, NY (1991).
29 R. P. Porter and W. K. Hall, J. Catal., 5, 366 (1966).
30 J. Le Bars, J. C. Vedrine, A. Auroux, S. Trautman, and M.
Baerns, Appl. Catal. A: Gen., 119, 341 (1994).
31 K. Esumi and K. Meguro, J. Adhesion Sci. Technol., 4, 393
(1990).
32 K. Meguro and K. Esumi, J. Colloid Interface Sci., 59 (1),
93 (1977).
33 B. D. Flockhart, I. R. Leith, and R. C. Pink, Trans. Faraday
Soc., 65, 542 (1969).
41 H. H. Kung and M. C. Kung, Appl. Catal. A: Gen., 157,
105 (1997).
42 J. G. Eon, R. Olier, and J. C. Volta, J. Catal., 145, 318
(1994).
43 T. Blasco, A. Galli, J. M. Lopez-Nieto, and F. Trifiro, J.
Catal., 169, 203 (1997).
44 D. Patel, P. J. Andersen, and H. H. Kung, J. Catal., 125,
132 (1990).
45 C. Doornkamp, M. Clement, X. Gao, G. Deo, I. E. Wachs,
and V. Ponec, J. Catal., 185, 415 (1999).
46 I. Wang, W. S. Chang, R. J. Sheiu, J. C. Wu, and C. S.
Chang, J. Catal., 83, 438 (1983).
34 Mamoru Ai, Bull. Chem. Soc. Jpn., 50, 2579 (1997).
35 C. P. Bezouhanava and M. A. Al-Zihari, Catal. Lett., 11,
245 (1991).
36 T. Blasco and J. M. Lopez-Nieto, Appl. Catal. A: Gen.,
157, 117 (1997).
37 E. A. Mamedov and V. Cortes-Corberan, Appl. Catal. A:
Gen., 127, 1 (1995).
38 W. M. H. Sachtler and N. H. De Boer, “Proceedings 3rd In-
ternational Congress on Catalysis, Amsterdam, 1964,” Wiley,
New York (1965), p. 240.
39 W. M. H. Sachtler, G. J. H. Dorgelo, J. Fahrenfort, and R.
J. H. Voorhoev, “Proceedings 4th International Congress on Catal-
ysis, Moscow, 1968,” ed by B. A. Kazanski, Adler, New York
47 A. Krause, Sci. Pharm., 38, 266 (1970).
48 J. Hanuza and B. Jezowska-Trzebiatowska, J. Mol. Catal.,
4, 271 (1978).
49 J. H. Lunsford, Catal. Rev., 8, 135 (1973).
50 A. Beilanski and J. Haber, Catal. Rev., 19, 1 (1979).
51 B. L. Yang and H. H. Kung, J. Catal., 77, 410 (1982).
52 G. V. Shaknovich, I. P. Belomestnykh, N. V. Nekrasov, M.
M. Kostyukovsky, and S. L. Kiperman, Appl. Catal., 12, 23
(1984).