Selective oxidation and oxidative dehydrogenation of hydrocarbons on bismuth vanadium molybdenum oxide

Article abstract of DOI:10.1016/j.jcat.2015.02.015

A systematic investigation of the oxidative dehydrogenation of propane to propene and 1- and 2-butene to 1,3-butadiene, and the selective oxidation of isobutene to methacrolein was carried out over Bi_1-x/3V_1-xMo_xO₄ (x = 0-1) with the aim of defining the effects of catalyst and reactant composition on the reaction kinetics. This work has revealed that the reaction kinetics can differ significantly depending on the state of catalyst oxidation, which in turn depends on the catalyst composition and the reaction conditions. Under conditions where the catalyst is fully oxidized, the kinetics for the oxidation of propene to acrolein and isobutene to methacrolein, and the oxidative dehydrogenation of propane to propene, 1-butene and trans-2-butene to butadiene are very similar - first order in the partial pressure of the alkane or alkene and zero order in the partial pressure of oxygen. These observations, together with XANES and UV-Vis data, suggest that all these reactions proceed via a Mars van Krevelen mechanism involving oxygen atoms in the catalysts and that the rate-limiting step involves cleavage of the weakest C-H bond in the reactant. Consistent with these findings, the apparent activation energy and pre-exponential factor for both oxidative dehydrogenation and selective oxidation correlate with the dissociation energy of the weakest C-H bond in the reactant. As the reaction temperature is lowered, catalyst reoxidation can become rate-limiting, the transition to this regime depending on ease of catalyst reduction and effectiveness of the reacting hydrocarbons as a reducing agent. A third regime is observed for isobutene oxidation at lower temperatures, in which the catalyst is more severely reduced and oxidation now proceeds via reaction of molecular oxygen, rather than catalyst lattice oxygen, with the reactant.

Full text of DOI:10.1016/j.jcat.2015.02.015

Journal of Catalysis 325 (2015) 87–100
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Selective oxidation and oxidative dehydrogenation of hydrocarbons on
bismuth vanadium molybdenum oxide
⇑
Zheng Zhai, Xuan Wang, Rachel Licht, Alexis T. Bell
Department of Chemical and Biomolecular Engineering, The University of California Berkeley, Berkeley, CA 94720-1462, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A systematic investigation of the oxidative dehydrogenation of propane to propene and 1- and 2-butene
to 1,3-butadiene, and the selective oxidation of isobutene to methacrolein was carried out over
Bi1ꢀx/3V1ꢀxMo O (x = 0–1) with the aim of defining the effects of catalyst and reactant composition on
x 4
Received 6 January 2015
Revised 16 February 2015
Accepted 22 February 2015
the reaction kinetics. This work has revealed that the reaction kinetics can differ significantly depending
on the state of catalyst oxidation, which in turn depends on the catalyst composition and the reaction
conditions. Under conditions where the catalyst is fully oxidized, the kinetics for the oxidation of propene
to acrolein and isobutene to methacrolein, and the oxidative dehydrogenation of propane to propene, 1-
butene and trans-2-butene to butadiene are very similar—first order in the partial pressure of the alkane
or alkene and zero order in the partial pressure of oxygen. These observations, together with XANES and
UV–Vis data, suggest that all these reactions proceed via a Mars van Krevelen mechanism involving oxy-
gen atoms in the catalysts and that the rate-limiting step involves cleavage of the weakest CAH bond in
the reactant. Consistent with these findings, the apparent activation energy and pre-exponential factor
for both oxidative dehydrogenation and selective oxidation correlate with the dissociation energy of
the weakest CAH bond in the reactant. As the reaction temperature is lowered, catalyst reoxidation
can become rate-limiting, the transition to this regime depending on ease of catalyst reduction and
effectiveness of the reacting hydrocarbons as a reducing agent. A third regime is observed for isobutene
oxidation at lower temperatures, in which the catalyst is more severely reduced and oxidation now
proceeds via reaction of molecular oxygen, rather than catalyst lattice oxygen, with the reactant.
Ó 2015 Elsevier Inc. All rights reserved.
Keywords:
Bismuth vanadium molybdenum oxide
Selective oxidation
Oxidative dehydrogenation
Butene
1
. Introduction
has been shown to involve the abstraction of an H atom from the
methyl group of propene to form an allyl intermediate, which is
The oxidation of propene to acrolein and isobutene to
methacrolein and the oxidative dehydrogenation of n-butene to
,3-butadiene are used to produce commodity chemicals and
then stabilized on the catalyst surface as a vinyl alkoxide.
Acrolein is then produced by the abstraction of a second H atom
from the alkoxide species [24,25]. Much less is known, though,
about the reactions of other hydrocarbons over bismuth molyb-
date-based catalysts. Studies of the oxidative dehydrogenation of
n-butenes to butadiene over bismuth molybdate have been carried
out, and mechanism for this reaction is thought to resemble that
for the oxidation of propene [26–30]. On the other hand, several
different mechanisms have been proposed for the oxidation of iso-
1
monomers for a variety of polymers. Most of the catalysts used
to promote these reactions are based on bismuth molybdate to
which other metals are added to enhance activity and product
selectivity [1–22]. For this reason, there is considerable interest
in understanding the influence of catalyst composition on catalyst
activity and selectivity reaction process, and the effects of reactant
composition on rate of product formation and the distribution of
products formed. Surprisingly, though, there have been relatively
few in depth investigations conducted on this subject.
Of the several systems of interest, the one that has been inves-
tigated most extensively is the oxidation of propene to acrolein
over bismuth molybdate. This reaction has been shown to proceed
via a Mars van Krevelen mechanism [23]. The rate-limiting step
butene to methacrolein, including
a Langmuir–Hinshelwood
mechanism and a redox model [31–33].
It is notable that previous studies of reaction mechanism and
kinetics have tended to be reactant specific, and very few have
involved a systematic investigation of the effects of catalyst and
reactant composition. A notable exception has been the case of
propene oxidation to acrolein over Bi1ꢀx/3
34] prepared with a scheelite structure. Ueda et al. [12] and
Sleight et al. [35,36] have reported that Bi1ꢀx/3 is more
active for the oxidation of propene than either bismuth molybdate
x 4
V1ꢀxMo O (x = 0–1)
[
x 4
V1ꢀxMo O
⇑
Corresponding author.
E-mail address: bell@cchem.berkeley.edu (A.T. Bell).
http://dx.doi.org/10.1016/j.jcat.2015.02.015
021-9517/Ó 2015 Elsevier Inc. All rights reserved.
0

8
8
Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
(
x = 1) or bismuth vanadate (x = 0). Our previous work proposed a
generalized model for the kinetics of propene oxidation over
, and explained the exact roles of Bi, Mo, and V
in affecting the activity and selectivity of Bi1ꢀx/3 [37–39].
Laboratory (ANL) on beam line 10BM. Measurements were per-
formed as described previously [37]. Data were acquired at the Bi
L -edge, and at the Mo and V K-edges before and after exposure
3
Bi1ꢀx/3
V
1ꢀxMo
x
O
4
V1ꢀxMo O
x 4
to reactant at 713 K. Additional data were acquired in situ at the
Mo K-edge under steady-state reaction conditions. These experi-
ments were carried out in a controlled-atmosphere cell that could
be heated up to 743 K in the presence of flowing gas [40].
Diffuse reflectance UV–VIS-NIR spectra were acquired using a
Fischer Scientific EVO 300 spectrometer equipped with a Praying
Mantis reflectance chamber and an in-situ high-pressure cell
(Harrick Scientific, Inc.). Spectra were referenced to the diffuse
reflectance spectrum of a Teflon reference.
By contrast, very little is known about the oxidation and oxidative
dehydrogenation of butene isomers over vanadium-substituted
bismuth molybdate. Therefore, many questions are still open. For
example, will both reactions follow a mechanism similar to that
propene oxidation for all catalyst compositions, and how does
the composition and structure of the reactant affect reactant
reactivity? Another question not fully understood is whether the
reaction kinetics are the same independent of reaction conditions
and if different for different reaction conditions, what are the
mechanistic implications?
The work reported here was undertaken in order to determine
catalyst and reactant composition, as well as reaction conditions
affect the kinetics for the oxidative dehydrogenation of propane
and 1- and 2-butene, and the selective oxidation of propene
x 4
and isobutene over Bi1ꢀx/3V1ꢀxMo O . The oxidation state of
the catalyst reduction was probed by in situ XANES and UV–Vis
2.3. Catalyst activity and selectivity
Measurements of reaction rates and product distributions were
performed using a packed-bed quartz tube reactor (10 mm in
diameter) loaded with 50–200 mg of catalyst. Prior to reaction,
the catalyst was preheated to the reaction temperature in air. All
experiments were carried out at atmospheric pressure with 3.3–
spectroscopy. The results of this work demonstrate that
Bi1ꢀx/3V1ꢀxMo O can operate in one of three regimes depending on
x 4
1
6.7% propene (99.9%, Praxair), 1-butene (99%, Praxair), trans-2-
butene (99%, Praxair), isobutene (99%, Praxair), propane (99%,
Praxair), and 3.3–16.7% oxygen (supplied from 20% oxygen in
helium, Praxair), balanced as needed with additional helium
the reactant and catalyst composition, the reaction temperature,
and the partial pressure of the reactants. Under conditions where
the catalyst is maintained in its fully oxidized state, all reactions
follow a Mars van Krevelen mechanism. Under these conditions,
the reaction kinetics are first order in the partial pressure of the
reactant and zero order in oxygen, and both the apparent activa-
tion energy and the apparent pre-exponential factor increase with
the strength of the weakest CAH bond of the reactant involved in
the rate-limiting step. When the rate of catalyst reoxidation cannot
keep up with the rate of catalyst reduction, the reaction becomes
zero order in reactant and fractional order in oxygen partial pres-
sure. In this case, the apparent activation energy and pre-exponen-
tial factor become independent of the reactant composition and
reflective of the activation energy for catalyst reoxidation. When
the catalyst is more severely reduced, reaction kinetics become
inverse order in reactant and first order in oxygen partial pressure.
Under these conditions, the apparent activation energy is very high
and the reaction is thought to proceed via the reaction of adsorbed
(
5
99.995%, Praxair). Data were collected at steady state between
73 K and 713 K. Products were analyzed using a gas chro-
matograph (Agilent 6890A) equipped with a 30 m HP-PLOT Q col-
umn and flame ionization detector (FID), for analyzing
a
hydrocarbons. An Alltech Hayesep DB packed column connected
to a thermal conductivity detector (TCD) was used to analyze for
oxygen, and carbon mono- and di-oxides. Reactant conversion
was calculated on the basis of products formed. Product selectivity
was defined as the moles of reactant converted to the product over
the sum of the moles of reactant converted to all products, based
on a carbon balance. All selectivities reported in this study are
intrinsic selectivities, extrapolated to a conversion of <1%.
3. Results
molecular O
2
and the reactant.
3.1. Kinetics
3
.1.1. Product distribution
The main product of 1-butene and trans-2-butene oxidation
2
. Methods
x 4
over Bi1ꢀx/3V1ꢀxMo O is 1,3-butadiene, and the principle by-prod-
ucts are trans-, cis-2-butene, and 1-butene. The influence of vana-
2.1. Catalyst preparation
dium content on catalyst activity for 1,3-butadiene formation from
Catalysts were prepared by the complexation procedure [37].
The metal precursors, ammonium molybdate tetrahydrate
99.98%, Sigma–Aldrich), bismuth (III) nitrate pentahydrate
99.98%, Sigma–Aldrich), and ammonium metavanadate (99%,
1
-butene and trans-2-butene is presented in Fig. 1. In both cases,
the activity passes through a maximum at x = 0.45 in a manner
similar to that observed for propene oxidation to acrolein produc-
tion over the same catalysts [37].
The selectivities to products formed from 1-butene and
trans-2-butene are presented in Fig. 2. For both 1-butene and
trans-2-butene, the product selectivities are similar and in each
case the main product is 1,3-butadiene. However, the selectivity
to 1,3-butadiene is higher starting from 1-butene than from
trans-2-butene. For both isomers of butene, the selectivity to
(
(
Sigma–Aldrich),
1 ꢀx/3):(1 ꢀx):x (x = 0–1.0), were mixed with citric acid (1:1 M
ratio with metal precursors) to produce materials with the stoi-
chiometry Bi1ꢀx/3 . Metal precursors with citric acid were
dissolved separately and then mixed together slowly. 2 M HNO
at
the
atomic
ratios
of
Bi:V:Mo =
(
x 4
V1ꢀxMo O
3
was used in place of water to dissolve bismuth nitrate to prevent
precipitation of bismuth hydroxides. The resulting solution was
dried at 60 °C for about 24 h in air to form a gel, which was then
dried at 120 °C and calcined in flowing air at 600 °C for 6 h.
Powdered catalysts were obtained.
1
,3-butadiene passes through a shallow minimum at x = 0.45 with
increasing value of x. By contrast, the selectivity to isomers of the
reactant passes through a maximum at the same value of x.
The main product of isobutene oxidation is methacrolein, and
the principle by-products are CO, CO2, and ethene. As shown in
Fig. 3, the activity for methacrolein formation at 703 K increases
2
.2. Catalyst characterization
with the value of x, and then reaches
(x = 0.45). Product selectivities are also pre-
sented in Fig. 3. The selectivity to methacrolein is 55% for x = 0,
a maximum for
X-ray absorption spectroscopy (XAS) measurements were per-
formed at the Advanced Photon Source at Argonne National
0.55 4
Bi0.85Mo0.45V O

Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
89
6
5
4
3
2
1
0
3.1.2. Partial pressure dependences
The rate of 1,3-butadiene formation from both 1-butene and
trans-2-butene can be represented by
1
-butene
trans-2-butene
m
1;2-C
n
O
rate1;3-butadiene ¼kapp
P
P
ð1Þ
4
H
8
2
where m and n are the apparent reaction orders of the partial pres-
sure of butenes and oxygen. Values of m and n are given in Tables 1
and 2 at temperatures of 673 K and 693 K (the high temperature
range) and 633 K (the low temperature range) for values of x = 0,
0
.45, and 1.0.
For BiVO
4
2 3
and Bi Mo O12, the data show that for both the high-
and low-temperature regimes, the rate of 1-butene (and trans-2-
butene) oxidation to 1,3-butadiene is nearly first order with
respect to the partial pressure of the alkene and nearly zero order
0
.0
0.2
0.4
0.6
0.8
1.0
x in Bi1-x/3Mo V1-xO
x
4
2
with respect to the partial pressures of O , regardless of the tem-
perature. The observed reaction orders are identical to those
reported by Keizer et al. [41] for the oxidative dehydrogenation
Fig. 1. Variation in the rate of the oxidative dehydrogenation of 1-butene and trans-
2
x 4
-butene to 1,3-butadiene on Bi1ꢀx/3V1ꢀxMo O with x at 673 K. Total flow rate:
of 1-butene over Bi
2
Mo
3
O12 at temperatures between 541 K and
6
0 mL/min. Left: 1-butene, P1ꢀC4H8 = 0.067 atm, PO2 = 0.167 atm. Right: trans-2-
6
88 K.
butene, Ptrans-2-C4H8 = 0.067 atm, PO2 = 0.167 atm.
For Bi0.85Mo0.45V O , the reaction orders were observed to
0.55 4
vary with temperature for both reactants. The alkene partial pres-
sure dependence increased from nearly zero order at low tempera-
ture (633 K) to close to first order at high temperature (673 and
rises slightly to 60% for x = 0.45, and then decreases to 48% for
x = 1.0. CO
tivity to CO
2
and CO (CO
x
) are the primary by-products, and selec-
6
93 K), whereas the oxygen partial pressure dependence decreased
x
follows a trend with catalyst composition that is oppo-
from a positive fractional order at lower temperatures to zero
order at higher temperatures.
The rate of isobutene oxidation to methacrolein can be repre-
sented by
site to that for methacrolein. The selectivity to methacrolein
observed for isobutene oxidation is about 10% lower than that to
acrolein for propene reported for identical reaction conditions
and catalysts [37]. It is notable, though, that the rate of methacro-
lein formation at a lower temperature (673 K) and a higher partial
pressure of isobutene (0.067 atm) is nearly independent of the
m
iso-C
n
O
ratemethacrolein ¼kapp
P
P
ð2Þ
H
8
2
value of x (Fig. 4). The product selectivities for BiVO
than those observed at 703 K and a propene partial pressure of
.0167 atm (Fig. 3). However, for Bi MoO12 and Bi0.85
, the selectivity to methacrolein (40%) is about 10% lower and
the selectivity to CO is 10% higher than those observed at 703 K
4
are identical
4
Table 3 lists the apparent reaction orders for methacrolein pro-
duction. At both 673 K and 703 K, the isobutene partial pressure
0
2
V0.55Mo0.45
O
4
dependence over BiVO
pressure dependence is nearly zero order. For Bi
, the reactant partial pressure dependences at
03 K are again first order in isobutene and zero order in oxygen.
4
is close to unity and the oxygen partial
2
Mo 12 and
3
O
2
and a propene partial pressure of 0.167 atm.
Fig. 5 illustrates the effects of catalyst composition on the activ-
ity and product selectivities for the oxidative dehydrogenation of
0.55 4
Bi0.85Mo0.45V O
7
However, at 673 K, the dependence on isobutene partial pressure
becomes negative first order and the oxygen partial pressure
dependence becomes nearly first order for both Bi Mo O12 and
2 3
propane over Bi1ꢀx/3
propene, and the activity at 753 K is shown in Fig. 5. The main
by-products are acrolein and CO . The product selectivities are also
shown in Fig. 5. The selectivities to propene and CO pass through
x 4
V1ꢀxMo O . The main product in this case is
2
0.55 4
Bi0.85Mo0.45V O .
2
Table 4 shows the reaction orders for propane dehydrogenation
to propene at 773 K. As expected, the rate of propene formation is
first order in propane and zero order in oxygen for all values of x.
shallow minima near x = 0.45, whereas the selectivity to acrolein
passes through a weak maximum at the same value of x.
1
00
100
1
,3-butadiene
1,3-butadiene
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
0
90
80
70
60
50
40
30
20
10
0
trans-2-butene
cis-2-butene
1-butene
cis-2-butene
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
x in Bi1-x/3Mo V1-xO4
x in Bi1-x/3Mo V1-xO4
x
x
x 4
Fig. 2. Dependence of the selectivity of Bi1ꢀx/3V1ꢀxMo O for the oxidative dehydrogenation of 1-butene and trans-2-butene at 673 K. Left: 1-butene, P1-C4H8 = 0.067 atm,
P
O2 = 0.167 atm. Right: trans-2-butene, Ptrans-2-C4H8 = 0.067 atm, PO2 = 0.167 atm.

9
0
Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
80
70
methacrolein
CO
CO2
60
50
40
30
20
10
0
Ethene
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
x in Bi1-x/3Mo V1-xO4
x in Bi1-x/3Mo V1-xO4
x
x
Fig. 3. Variation in the rate and selectivity of iso-butene oxidation to methacrolein on Bi1ꢀx/3
V
x
1ꢀxMo O
4
with x. For rate measurement total flow rate: 60 mL/min.
Piso-C4H8 = 0.0167 atm, PO2 = 0.167 atm, 703 K.
0
0
0
0
0
0
0
.6
.5
.4
.3
.2
.1
.0
80
methacrolein
CO
7
6
5
0
0
0
CO2
Ethene
40
30
20
10
0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
x in Bi1-x/3Mo V1-xO4
x in Bi1-x/3Mo V1-xO4
x
x
Fig. 4. Variation in the rate and selectivity of iso-butene oxidation to methacrolein on Bi1ꢀx/3
V
x
1ꢀxMo O
4
with x. For rate measurement total flow rate: 60 mL/min.
P
iso-C4H8 = 0.067 atm, PO2 = 0.167 atm, 673 K.
1
00
3
3
2
2
1
.5
.0
.5
.0
.5
9
8
7
6
0
0
0
0
propene
CO2
acrolein
50
4
3
2
1
0
0
0
0
0
0
.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
x in Bi1-x/3Mo V1-xO4
x in Bi1-x/3Mo V1-xO4
x
x
x 4
Fig. 5. Variation in the rate and selectivity of the oxidative dehydrogenation of propane to propene on Bi1ꢀx/3V1ꢀxMo O with x. For rate measurement total flow rate:
6
0 mL/min. PC3H8 = 0.067 atm, PO2 = 0.167 atm, 753 K.
Table 1
Reaction orders at different temperatures for the oxidation of 1-butene to produce 1,3-butadiene on Bi1ꢀx/3
V
1ꢀxMo
x
O
4. The partial pressures of 1-butene and oxygen were fixed at
693 K
0.067 and 0.167 atm when varying the other one.
Composition (x)
633 K
-Butene
673 K
1
Oxygen
1-Butene
Oxygen
1-Butene
Oxygen
0
0
1
0.9 ± 0.0
0.1 ± 0.0
0.8 ± 0.0
0.0 ± 0.0
0.3 ± 0.0
ꢀ0.1 ± 0.1
1.0 ± 0.0
0.9 ± 0.1
1.0 ± 0.0
ꢀ0.1 ± 0.1
0.0 ± 0.1
ꢀ0.1 ± 0.1
0.9 ± 0.0
1.0 ± 0.0
0.9 ± 0.0
0.0 ± 0.0
0.0 ± 0.1
0.0 ± 0.0
.45

Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
91
Table 2
Reaction orders at different temperatures for trans-2-butene and oxygen to produce 1,3-butadiene on Bi1ꢀx/3
fixed at 0.067 and 0.167 atm when varying the other one.
V
1ꢀxMo
x
O
4. The partial pressures of trans-2-butene and oxygen were
693 K
Composition (x)
633 K
673 K
Trans-2-butene
Oxygen
Trans-2-butene
Oxygen
Trans-2-butene
Oxygen
0
0
1
0.8 ± 0.1
0.0 ± 0.1
0.9 ± 0.0
0.2 ± 0.1
0.3 ± 0.0
0.0 ± 0.1
0.8 ± 0.1
0.9 ± 0.0
1.0 ± 0.1
0.2 ± 0.1
0.1 ± 0.0
0.0 ± 0.0
0.8 ± 0.1
0.9 ± 0.1
1.0 ± 0.0
0.3 ± 0.0
0.1 ± 0.1
0.2 ± 0.0
.45
Table 3
Reaction orders at different temperatures for the oxidation of isobutene to produce methacrolein on Bi1ꢀx/3
V
x
1ꢀxMo O4. The partial pressures of isobutene and oxygen were fixed at
0.067 and 0.167 atm when varying the other one at 673 K (left). The partial pressures of isobutene and oxygen were fixed at 0.0167 atm and 0.167 atm when varying the other
one at 703 K (right).
Composition (x)
673 K
703 K
Isobutene (0.05–0.15 atm)
Oxygen (0.067–0.167 atm)
Isobutene (<0.0167 atm)
Oxygen (0.067–0.167 atm)
0
0
1
0.9 ± 0.1
ꢀ0.8 ± 0.1
ꢀ0.9 ± 0.0
0.0 ± 0.1
0.8 ± 0.1
1.1 ± 0.1
1.0 ± 0.1
0.9 ± 0.2
0.9 ± 0.1
0.0 ± 0.1
0.0 ± 0.0
0.0 ± 0.0
.45
Table 4
Reaction orders at different temperatures for the oxidative dehydrogenation of
propane to produce propene on Bi1ꢀx/3 1ꢀxMo 4. The partial pressures of propane
and oxygen were fixed at 0.067 and 0.167 atm when varying the other one at 773 K.
2-butene oxidation. The activation energy for 1-butene oxidation
to 1,3-butadiene obtained on bismuth molybdate (x = 1.0) is simi-
lar to that reported by Keizer et al. [41], 11 kcal/mol.
V
x
O
While the apparent activation energies for 1-butene and trans-
Composition (x)
773 K
2-butene oxidative dehydrogenation over Bi
the same for temperatures below 663 K as they are for higher tem-
peratures, for Bi0.85Mo0.45 the apparent activation energies
2 3 4
Mo O12 and BiVO are
Propane
Oxygen
V O
0.55 4
0
0
1
1.0 ± 0.1
1.1 ± 0.2
1.0 ± 0.1
0.0 ± 0.0
0.0 ± 0.1
0.0 ± 0.0
.45
below 663 K are higher than those measured at high temperatures,
24.3 kcal/mol in the case of 1-butene and 24.7 kcal/mol in the case
of trans-2-butene. A similar pattern has been reported by Linn and
Sleight [16] for 1-butene oxidation on bismuth molybdate,
14 kcal/mol for temperatures above 663 K, and 38 kcal/mol for
temperatures below 663 K.
3.1.3. Temperature dependence
Arrhenius plots for the oxidative dehydrogenation of 1-butene
and trans-2-butene to 1,3-butadiene, the oxidation of isobutene
to methacrolein, and the oxidative dehydrogenation of propane
to propene over Bi1ꢀx/3V1ꢀxMo O are shown in Figs. 6–8 for tem-
x 4
The apparent activation energy for the oxidation of isobutene to
methacrolein exhibits a complex dependence upon the catalyst
composition, the temperature regime, and the partial pressure of
perature in the range of 613–773 K. Values of the apparent activa-
tion energies data are listed in Table 5–8.
For temperatures above 663 K, the apparent activation energies
for 1-butene and trans-2-butene oxidative dehydrogenation to
isobutene. For BiVO
14.7 kcal/mol, independent of temperature or isobutene partial
pressure in the feed (0.067 atm or 0.0167 atm). For Bi Mo
the apparent activation energy is 21.6 kcal/mol when the isobu-
tene partial pressure is 0.0167 atm and increases to 44.0 kcal/mol
when the isobutene partial pressure is 0.067 atm, and for both
4
, the apparent activation energy is the same,
2
3 12
O ,
1
,3-butadiene increase slightly with x, from 6.5 to 9.2 kcal/mol
for 1-butene oxidation, and from 14.0 to 17.4 kcal/mol for trans-
-
-
-
7.5
8.0
8.5
-7.0
-7.5
-8.0
-8.5
-9.0
-9.5
x=0.45
x=1.0
x=0.45
-9.0
-
9.5 x=1.0
-
-
-
10.0
10.5
11.0
x=0
-
-
-
10.0
10.5
11.0
x=0
-11.5
12.0
-12.5
-
0
.00135
0.00140
0.00145
0.00150
0.00155
0.00160
0.00135 0.00140 0.00145 0.00150 0.00155 0.00160 0.00165
-1
-1
1
/T (K )
1/T (K )
x 4
Fig. 6. Arrhenius plot of the oxidative dehydrogenation of 1-butene and trans-2-butene oxidation to 1,3-butadiene over Bi1ꢀx/3V1ꢀxMo O catalysts at 613–733 K. Left: 1-
butene, P1ꢀC4H8 = 0.067 atm, PO2 = 0.167 atm. Right: trans-2-butene, Ptrans-2ꢀC4H8 = 0.067 atm, PO2 = 0.167 atm. Total flow rate: 60 mL/min.

9
2
Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
-
-
-
-
7.0
7.5
8.0
8.5
-
-
-
8.5
9.0
9.5
x=0.45
x=1.0
x=0.45
x=1.0
-9.0
-9.5
-
-
-
-
-
-
10.0
10.5
11.0
11.5
12.0
12.5
-
-
-
-
-
-
10.0 x=0
10.5
x=0
11.0
11.5
12.0
12.5
-13.0
0.00135
0.00135
0.00140
0.00145
0.00150
0.00155
0.00140
0.00145
0.00150
0.00155
-
1
1/T (K-1)
1
/T (K )
x 4
Fig. 7. Arrhenius plot of the isobutene oxidation to methacrolein over Bi1ꢀx/3V1ꢀxMo O catalysts at 613–733 K. Left: PisoꢀC4H8 = 0.0167 atm, PO2 = 0.167 atm. Right:
PisoꢀC4H8 = 0.067 atm, PO2 = 0.167 atm. Total flow rate: 60 mL/min.
-
-
-
-
-
-
-
12.0
12.5
13.0
13.5
14.0
14.5
15.0
Table 6
Apparent activation energies under different temperature ranges for oxidation
x=0.45
of
trans-2-butene
to
produce
1,3-butadiene
on
x 4.
Bi1ꢀx/3V1ꢀxMo O
P
trans-2-C4H8 = 0.067 atm, PO2 = 0.167 atm.
Composition (x)
E
app (kcal/mol)
x=0
663–713 K
613–663 K
x=1.0
0
14.0
14.4
17.4
14.0
24.7
17.4
0.45
1
Table 7
Apparent activation energies under different temperature ranges for the oxidation of
isobutene
to
methacrolein
on
Bi1ꢀx/3
V
1ꢀxMo
x
O
4.
P
iso-C4H8 = 0.0167 atm,
P
O2 = 0.167 atm.
Composition (x)
E
app (kcal/mol)
0
.00130
0.00135
0.00140
7
03–733 K
643–703 K
-1
1
/T (K )
0
14.7
17.0
21.6
14.7
44.8
21.6
0.45
Fig. 8. Arrhenius plot of the oxidative dehydrogenation of propane to propene over
catalysts at 703–773 K and PC3H8 = 0.067 atm, PO2 = 0.167 atm.
1
x 4
Bi1ꢀx/3V1ꢀxMo O
Total flow rate: 60 mL/min.
Table 8
Apparent activation energies for the oxidative dehydrogenation of propane to
propene on Bi1ꢀx/3 1ꢀxMo C3H8 = 0.167 atm, PO2 = 0.167 atm.
Table 5
Apparent activation energies under different temperature ranges for 1-butene and
oxygen to produce 1,3-butadiene on Bi1ꢀx/3
V
x 4.
O P
V
1ꢀxMo
x
O
4.
P
1-C4H8 = 0.067 atm,
P
O2 = 0.167 atm.
Composition (x)
Eapp (kcal/mol)
Composition (x)
E
app (kcal/mol)
63–713 K
7
03–773 K
6
613–663 K
0
24.4
25.5
31.5
0.45
0
6.5
7.6
9.2
6.5
24.3
9.2
1
0
.45
1
3
.2. Catalyst characterization
isobutene partial pressures the apparent activation energy is
independent of the temperature. In the case of
4, the apparent activation energy is found to
depend on both the isobutene partial pressure and the reaction
temperature. For an isobutene partial pressure of 0.0167 atm, the
apparent activation energy is 17 kcal/mol for temperatures above
XANES spectra were acquired in order to establish which ele-
ments are reduced at elevated temperature in the presence of dif-
ferent alkenes on Bi Mo
Measurements were taken before and after exposure to 1-butene,
trans-2-butene, isobutene, and propene at 713 K for 2 h. All the
spectra are shown in the Supplementary Information.
0.55
Bi0.85Mo0.45V O
2
3 12 0.55 4
O , BiVO4, and Bi0.85Mo0.45V O .
7
03 K and it is 44.0 kcal/mol for temperatures below 703 K.
However, when the isobutene partial pressure is raised to
.067 atm, the apparent activation energy becomes 44.0 kcal/mol
independent of the reaction temperature (Fig. 7).
The position of the Bi L
3
-edge did not change with time and
independent of the
0
remained identical to that for Bi
2 3
O

Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
93
composition of the reducing agent or the catalyst compositions.
After exposure to propene, 1-butene, trans-2-butene, and isobu-
tene, the Mo K-edge shifted to lower energy, and the height of
the pre-edge feature decreased. Similar trends were observed for
the V K-edge. After reduction, the edge shifts and the pre-edge fea-
ture height decreased indicating a reduction of V. For both the Mo
and V K-edges, a larger shift in the edge position and a smaller pre-
edge feature were observed after reduction with all butene isomers
than with propene, indicating that butenes are better reducing
agents than propene. It is also noted that the reported changes in
supported by the observation that the apparent activation energy
for the oxidative dehydrogenation of propane, 1-butene, and
trans-2-butene correlates with the strength of the weakest CAH
bond of the reactant. Getsoian et al. [38,39] have demonstrated
that of the three types of Mo@O bonds present on the surface of
2
Bi MoO12, and the one that weakly interacts with the lone pair
on Bi (e.g., BiꢁꢁꢁO@Mo) is the most active for activating the CAH
bond in the methyl group of propene. We, therefore, assume that
this will also be the most active form of oxygen in the present sys-
tem. The next step following CAH bond cleavage of the reactant
involves abstraction of a hydrogen atom from the methyl group
of the methyl-allyl radical (Step 3), which reduces the adjacent
the Mo and V K-edge spectra were larger for Bi0.85V0.55Mo0.45O
4
than for BiVO or Bi Mo
4
2
3 12
O .
5
+
Diffuse reflectance UV–VIS spectra were acquired in order to
further characterize the reducing power of different butenes.
When the catalyst is heated to 673 K and exposed to a reducing
agent, the absorbance of the baseline at a wave number below
the adsorption edge increased as a function of time, due to catalyst
reduction. The absorbance at 800 nm was plotted as a function of
time of exposure to reducing conditions, and the initial reduction
rate was calculated (Supplementary Information). The correlation
Mo@O to Mo and produces 1,3-butadiene. In the final two steps
of the cycle, water is formed and desorbed reversibly producing
two oxygen vacancies (Step 4) and the catalyst is then reoxidized
by oxygen.
A rate expression for the rate of 1-butene consumption shown
in Eq. (3) can be derived under the assumption that Step 2 is
rate-limiting and that the pseudo-steady-state assumption is valid
for intermediates [37]:
k
1
k
2
P
1
C4H8 ½Sꢃ
rate ¼
h
ꢀ
ꢁ
i
ð3Þ
k
1
P
C
H
k
1
k
2
P
C
H
k
k
2
P
C
H
k
1
k
2
k
ꢀ4
P
C
H
P
H
O
k
1
k
2
P
C
H
4 8
þkꢀ1
4
8
4
8
4
8
4
8
2
ðk
2
þkꢀ1Þ1 þ
þk3
Þþ
Þþk
þk5P
k
2
þkꢀ1
ðk
2
þkꢀ1
k
4
ðk
2
þkꢀ1
4
k
5
P
O
ðk
2
þkꢀ1
Þ
O
ðk
2
Þ
2
2
of the apparent rate coefficient kapp for different butene isomers
In Eq. (3), k
i
and kꢀi represent the rate coefficients for reaction i
with the initial rate of reduction measured by UV–VIS on
in the forward and reverse directions, respectively, K
i
represents
Bi0.85V0.55Mo0.45
O
4
is shown in Fig. 9. The observed linear correla-
the equilibrium constant for reaction i, P represents the partial
j
tion suggests that the oxidation of alkenes proceeds via a Mars van
Krevelen mechanism involving catalyst oxygen atoms. The more
readily the reactant can be reduced, the higher is the reaction rate.
The reducibility of isobutene and trans-2-butene is lower than that
of 1-butene, but greater than that of propene.
pressure of reactant or product j, and [S] is the number of active
sites per unit BET catalyst surface area.
When the rate of catalyst reoxidation is rapid relative to the rate
of catalyst reduction, catalyst surface is fully oxidized, and if it is
2
assumed that kꢀ1 ꢂk , then Eq. (3) simplifies to
1 2
k k
ðkꢀ1 þk Þ
½Sꢃ
4
. Discussion
rate ¼
PC4H8 ffiK k P
½Sꢃ¼kappPC4H8
ð4Þ
1
4 8
2 C H
2
4.1. Mechanism and kinetics of the oxidative dehydrogenation of
propane, 1-butene, and trans-2-butene, and oxidation of isobutene
Tables 1 and 2 show that the kinetics for the oxidative dehydro-
genation of propane to propene and of 1-butene and trans-2-
butene to butadiene are very similar—essentially first order in
the partial pressure of the alkane or alkene and zero order in the
1
1
1
40
20
00
partial pressure of oxygen—for BiVO
at temperatures above 663 K. Taken together,
the UV–Vis and XANES data presented in Fig. 9 and Figs. S1–S5
Supplementary Information) suggest that in all cases oxidative
4 2
and Bi MoO12, and for
Bi0.85V0.55Mo0.45O
4
8
6
4
0
0
0
(
dehydrogenation proceeds via a Mars van Krevelen mechanism
involving oxygen atoms of the catalysts. The observed kinetics
can be rationalized on the basis of the mechanism shown in
Scheme 1, which is illustrated for the oxidative dehydrogenation
1
-butene
Isobutene
of 1-butene on Bi
explain the kinetics of propene oxidation to acrolein [37].
Reaction begins with the physical adsorption of 1-butene (Step 1)
2
MoO12 and is similar to that developed to
20
0
Trans-2-butene
Propene
followed by hydrogen abstraction from the ACH
2
A group of 1-
-
2
0
2
4
6
8
10
12
14
butene by one of the Mo@O bonds of the catalyst and concurrent
6
+
5+
Delta Absorbance/Second
reduction of Mo to Mo (Step 2). As in the case for propene oxi-
dation, Step 2 is taken to be the rate-limiting step. Hydrogen
Fig. 9. Correlation of the apparent rate coefficient for the production of the
principle product, at 703 K for different butene isomers (PC4H8 = 0.0167,
O2 = 0.167 atm) and initial rate of reduction measured using UV–VIS at 703 K
(PC4H8 = 0.0167 atm) on Bi0.85
2
abstraction is taken to occur from the ACH A group of 1-butene
because the strength of this CAH bond is weaker than that of other
CAH bonds in the reactant. As discussed below, this assumption is
k
app
,
P
V
0.55Mo0.45O
4.

9
4
Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
O
H
O
O
H
H
H
C
H
H
H
C
Mo
Bi
O
C
C
H
C
H
C
H
O
HC
C
H
H
O
O
O
O
k
1
O
H
k
2
O
H
O
H
O
H
C
Bi
O
H
Mo
O
Mo
Bi
O
O
O
Bi
O
C
H
Mo
k
-1
OH C
O
Physical
Adsorption
Step 1
H abstraction
H
O
O
O
O
O
H
C
H
Step 1
O
H
O
Bi
O
O
Mo
O
Bi
O
Mo
O
O
O
O
Re-oxidation
Step 5
+
O
k5
2
k
3
nd
2
H
O
O
O
H
O
O
O
abstraction
Step 3
H
C
Bi
O
Bi
O
Mo
O
Mo
C
H
k
4
OH C
H
O
O
O
O
H
C
H
O
k-4
O
O
Bi
O
Bi
HO
Mo
Mo
O
-
2
H O;
O
O
-
1,3-butadiene
O
Step 4
2 3
Scheme 1. The mechanism of the oxidative dehydrogenation of 1-butene to 1,3-butadiene over Bi Mo O12 proposed on the basis of data reported in this study.
and the apparent rate constant kapp is now given by K
form of Eq. (4) is completely consistent with the kinetics observed
for the oxidative dehydrogenation of 1-butene and trans-2-butene
1
k
2
½Sꢃ. The
here for the oxidative dehydrogenation of 1-butene and trans-2-
butene to 1,3-butadiene, and the kinetics for the oxidative
dehydrogenation of propane to propene. For BiVO , the kinetics
4
on BiVO
Bi0.85V0.55Mo0.45O
4
4
and Bi
at temperatures above 663 K. It is also noted
that Eq. (4) also holds for the oxidative dehydrogenation of propane
at all temperatures for all three catalysts (BiVO
and Bi MoO12).
Tables 1 and 2 show that for Bi0.85
2
MoO12 at all temperatures and for
are first order in isobutene and zero order in oxygen independent
of the isobutene partial pressure (0.0167 atm or 0.067 atm) and for
Bi MoO12 and Bi0.85V0.55Mo0.45O only for isobutene partial pres-
2 4
sures below 0.0167 atm. We note that under these constraints
the apparent activation energies for isobutene oxidation are
4
, Bi0.85
V0.55Mo0.45O
4
,
2
V0.55Mo0.45
O
4
the orders in 1-
14.7 kcal/mol for BiVO
4
,
17.0 kcal/mol for Bi
2
MoO12
,
and
butene and trans-2-butene change significantly for temperatures
below 663 K. Thus, at 633 K, the order in alkene decreases to nearly
zero and the order in oxygen rises to 0.3. We also note that the
apparent activation energy increases for both isomers of butene.
Notably, though, changes in reaction order in alkene and oxygen,
and the apparent activation energy, do not change for the oxidative
21.6 kcal/mol for Bi0.85V0.55Mo0.45O .
4
The kinetics of isobutene oxidation have been discussed to only
a limited degree, with very different conclusions drawn by differ-
ent authors [31–33]. Mann and Ko [31] considered 13 different
mechanisms, and concluded rate of reaction on copper-promoted
bismuth molybdate is described best by a mechanism in which
the rate-controlling step is assumed to be the surface reaction
between charged absorbed 2-methylpropene and oxygen mole-
cules. By contrast, Benyahia and Mearns [32,33] found fractional
apparent orders with respect to isobutene and oxygen for isobu-
tene oxidation over multicomponent bismuth molybdate, which
they rationalized on the basis of a redox model. It should be noted
though none of these studies reported sufficient rate data and cata-
lyst characterization from which one could develop a compelling
case for a particular reaction mechanism.
The kinetics of isobutene oxidation for the conditions noted
above can be rationalized using the mechanism proposed in
Scheme 2, which is very similar to that reported in our earlier work
on the oxidation of propene to acrolein [37], as well as being simi-
lar to that shown in Scheme 1 for the oxidative dehydrogenation of
alkenes and alkanes. The primary difference is that the 2-methyl-
allyl radical formed in Step 2 is now trapped by reaction with a
molybdenyl oxygen atom. This step, in turn, leads to the formation
of methacrolein. By contrast, the 1-methyl-allyl radical formed
4
dehydrogenation of propane to propene over Bi0.85V0.55Mo0.45O .
We attribute the changes in the reaction orders and the apparent
activation energy to a transition of the catalyst from a state in
which the catalyst surface is fully saturated with oxygen to one
in which the catalyst is partially reduced. Under the latter circum-
stance, the rate of catalyst reoxidation becomes rate limiting. The
reason that the transition in kinetic regimes is not observed for
the oxidative dehydrogenation of propane is that the rate of this
reaction is two orders of magnitude lower than that for oxidative
dehydrogenation of either 1-butene or 2-butene. Reference to Eq.
(
2 5
3) shows that in the limit that the ratio of k /k becomes very
large, the last term in the denominator becomes dominant and
the rate expression reduces to
rate ¼k
5
P
O2
ð5Þ
The kinetics for isobutene oxidation to methacrolein for tem-
peratures above 663 K are very similar to those reported earlier
for the oxidation of propene to acrolein, to the kinetics reported

Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
95
O
O
O
O
Bi
O
Mo
H
C
H
H
H
H
H
C
C
C
H
k
1
H
C
O
k2
C
O H3
C
H
CH3
H
O
O
O
O
k
-1
O
O
O
Bi
Mo
O
O
O
Bi
O
Mo
Physical
Adsorption
Step 1
H abstraction
O insertion
Step 2
O
Mo
Bi
O
O
O
H
H
O
C
O
H
O
O H3C
Bi
C
O
O
Bi
C H
Mo
O
O
O
H
Mo
O
O
O
O
O
Re-oxidation
Step 5
k
5
+
O
2
O
O
O
k3
O
O
O
O
Bi
O
Mo
Step 3
H
Bi
O
H
Mo
C
H
O
k
4
O H3C
Bi
O
C
H
O
C
O
O
O
H
O
Bi
k
-4
Mo
Mo
O
-
H
2
O;
O
O
- Methacrolein
O
O
Step 4
Scheme 2. The mechanism of the oxidation isobutene to methacrolein over Bi
2
3
Mo O
12 under lower partial pressure of isobutene proposed on the basis of data reported in
this study.
from 1-butene or trans-2-butene is not stabilized and instead loses
an H atom to form 1,3-butadiene. Moro-oka has noted that oxygen
addition is only observed when the intermediate formed upon
cleavage of the weakest CAH bond of the reactant does not have
a weak CAH bond that can be easily cleaved [42]. In accord with
Moro-oka’s statement, we note that the CAH bond in the methyl
group of 1-methyl-allyl will cleave easily leading to the formation
of 1,3-butadiene. However, the CAH bond in the methyl group of
methyl-allyl (from 1-butene and trans-2-butene) will form 1,3-
butadiene directly after dehydrogenation.
As noted earlier, the kinetics of isobutene are significantly dif-
ferent for Bi
2
MoO12 and Bi0.85
V
0.55Mo0.45
O
4
for high isobutene par-
at
tial pressures, and in the case of Bi0.85
V
0.55Mo0.45O
4
temperatures below 703 K. Under these conditions, the isobutene
partial pressure becomes inverse first order and the oxygen partial
pressure becomes first order, and the apparent activation energy
increases to 44 kcal/mol for both catalysts. In situ UV–Vis spectra
(see Supporting Information) also show more significant reduction
2
-methyl-allyl is harder to cleave and consequently is not possible
to form a stable product. As a result, 2-methyl-allyl (from isobu-
tene) will undergo oxygen insertion to form methacrolein, but 1-
4
of Bi0.85V0.55Mo0.45O at 673 K particularly for an isobutene partial
O
H
O
O
H
O
O
Mo
O
H
C
H
C
Bi
O
Bi
O
Mo
C
H
C
H
O
OH C
HC
C
H
H
1
-butene:
O
O
O
H
O
C
H
H
H
H
H
O
O
O
Bi
O
O
Bi
O
Mo
Mo
O
O
O
O
O
O
O
O
O
O
H
Bi
O
Mo
H
Bi
O
Mo
H
H
OH CH
OH
C
Trans-2-butene:
C
H
C
O
O
O
O
C
H
C
H
O
O
H
C
H
C
H
O
Bi
HO
O
Bi
HO
Mo
Mo
H
O
O
O
O
Scheme 3. Differences in the mechanisms of isomers formation from 1-butene and trans-2-butene.

9
6
Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
H H C
H
C H
nd
O*
O*
3
H C
H
C H
2
H
3
Gas phase
H C
C
C
H
H
H
H C C C H
O
H
O
C
C
O
O
abstraction
oxygen O*
O
O
O
H
O
H
H
O
O
Mo
O
Mo
Mo
O
O*
C
O
H
H
H
H C C C H
O*
O*
H
O
C
O
Mo
H
O
C
O
CO
production
Re-oxidize
CO
2
production
H
H C
H
C
H
C H
O
O
O
O
Mo
H abstraction
H
C
H
C H
H C
O
H
O
O
O
Mo
Gas phase
oxygen O*
O*
H
C
H
C H
C
O
H
H
O
O
O
Mo
O*
C
O
H
C
H
C H
H
O
O
O
Mo
H
Scheme 4. Possible pathway for CO and CO
2
2 3
formation on Bi Mo O12 during isobutene oxidation.
from white for the fully oxidized catalyst to gray, and
changes from bright yellow to dark yellow.
The kinetics of isobutene oxidation observed on the reduced
forms of Bi MoO12 and Bi0.85 can be rationalized in
Table 9
Bi0.85V0.55Mo0.45O
4
Symmetry number (r) for different reactants.
Reactant
Propene
Symmetry number (r)
2
V0.55Mo0.45O
4
3
2
6
6
6
terms of the following reactions sequence:
1
-Butene
Trans-2-butene
Isobutene
Propane
K1
(1) Ãþiso-C
4
H
8
ꢂꢂ!ꢂꢂ piso-C
4
H
8
Ã
K2
(
(
(
2) ÃþO
3) iso-C
4) C
2
ꢂꢂ!ꢂꢂ O
2
Ã
k3
4
H
8
ÃþO
2
Ãꢂꢂ!C
4
H
6
O ÃþH OÃ
2
K4
4
H
6
O Ãꢂꢂ!ꢂꢂ C
4
H
6
O þÃ
pressure of 0.067 atm than is observed the same experiment is car-
ried out in 1-butene. Visual inspection of the catalyst after opera-
tion under steady-state reaction conditions also confirms that
isobutene reduces both Bi
greater extent than does 1-butene. After reaction in 0.067 atm of
isobutene and 0.167 atm of oxygen, the color of Bi MoO12 changes
K5
(5) H
2
O Ãꢂꢂ!ꢂꢂ H
2
O þÃ
⁄
2
MoO12 and Bi0.85V0.55Mo0.45
O
4
to a
where ‘‘ ’’ represents an active site. In contrast to the mechanism
presented in Scheme 2, it is assumed here that oxidation occurs
2
2
via reaction with adsorbed O . Implicit in the development of this

Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
97
pathway is that the reduction of the catalyst is much more rapid
than its reoxidation. Step 1 and Step 2 of the proposed sequence
represent the adsorption of isobutene and oxygen onto active sites.
Step 3 is the bimolecular reaction of the adsorbed molecules. Steps
Bi1
ꢀx/3 1ꢀx x 4
Mo O catalysts are other butene isomers, but the
V
by-products of isobutene are mainly CO, CO2, and ethene.
For the oxidative dehydrogenation of 1-butene and trans-2-
butene, once the 1-methyl-allyl intermediate is formed following
H abstraction, the intermediate can undergo further hydrogen
abstraction, yielding 1,3-butadiene, or reinsertion of the first
abstracted hydrogen at the end of the allylic moiety to yield other
isomers. One might expect this process to lead to the same product
distribution regardless of the starting alkene. However, our experi-
mental data show that the product distributions resulting from the
allylic intermediates produced from 1-butene and trans-2-butene
are different (Fig. 2). The selectivity to 1,3-butadiene is higher than
that for 1-butene compared to trans-2-butene, indicating oxidative
dehydrogenation is favored for 1-butene, but isomerization is
favored for trans-2-butene. Similar results have been observed by
Portela [8], and the reason can be explained as follows. For 1-
butene, once 1-methyl-allyl intermediate is formed, the insertion
must occur at the end of the intermediate. However, the position
of the first abstracted hydrogen is not at the end of the intermedi-
ate. For trans-2-butene, the position of the first abstracted hydro-
gen is at the end of the intermediate, so it becomes much easier
to form an isomerization product, cis-2-butene. This difference is
illustrated in Scheme 3. The proposed pathways also help to
explain the higher selectivity to cis-2-butene than 1-butene from
trans-2-butene.
4
and 5 are desorption of methacrolein and water. If Steps 1, 2, 4,
and 5 are assumed to be quasi-equilibrated and Step 3 is the irre-
versible, rate-determining step, then the rate of methacrolein for-
mation is given by the following relation:
k
3
K
1
K
2
P
iso-C4H8
P
O2 ½Sꢃ
P
C H O
ratemethacrolein ¼ꢀ
ꢁ
ð6Þ
2
P
H
O
4
6
2
5
1
þK
1
P
iso-C4H8 þK
2
P
O2
þ
þ
K4
K
Since the conversion of isobutene was always below 5%, we
assume that PC4H6
O O
and PH2 are negligible. If we assume that at
higher isobutene partial pressures, isobutene is strongly adsorbed
onto the catalyst surface, so that adsorbed isobutene is the most
abundant surface intermediate, then
k
3
K
2
P
O2 ½Sꢃ
ratemethacrolein
¼
ð7Þ
P
iso-C4H8
Eq. (7) is consistent with the observed kinetics shown in
Table 3, in which the rate of isobutene oxidation to methacrolein
is negative first order in isobutene and positive first order in oxy-
gen. This rate expression suggests that the high apparent activation
The by-product formation occurring during the oxidation of iso-
butene is different from that observed for 1- and trans-2-butene.
Isobutene does not undergo isomerization because the structure
of the 2-methyl-allyl intermediate does not lend itself to isomer-
energy measure for reduced Bi
attributable to an even higher activation barrier for Step 3.
To this point, we have only discussed the mechanism and kinet-
ics for formation of primary products. Therefore, we turn next to a
discussion of the mechanism by which by-products are formed.
The by-products of 1-butene and trans-2-butene over
2 4
MoO12 and Bi0.85V0.55Mo0.45O is
ization, and thus only CO and CO
2
are formed as by-products.
for CO formation is 0.4
The partial pressure dependence on O
2
2
3
2
2
2
1
1
2
8
4
0
6
2
8
4
propane
isobutene
propene
t-2-butene
1-butene
80
85
90
95
100
Weakest C-H Bond Dissociation Energy / (kcal/mol)
32
28
24
20
16
12
8
4
3
2
2
2
2
8
4
0
propane
isobutene
propene
t-2-butene
1-butene
16
12
8
propane
isobutene
propene
t-2-butene
1
-butene
4
82
84 86 88 90 92 94 96 98
82 84 86 88 90 92 94 96 98
Weakest C-H Bond Dissociation
Energy / (kcal/mol)
Weakest C-H Bond Dissociation
Energy / (kcal/mol)
Fig. 10. Correlation of the apparent activation energy for formation of the principle product with the dissociation energy of the weakest CAH bond involved in the rate-
determining step occurring on Bi Mo 12 (top), BiVO (bottom left), and Bi0.85 (bottom right). All bond dissociation energy data come from: Handbook of bond
dissociation energies in organic compounds. Yu-Ran Luo. CRC Press, 2003.
2
3
O
4
V0.55Mo0.45O
4

9
8
Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
on Bi
as for propene oxidation. This suggests that the oxygen atoms in
CO come from both gas phase oxygen and lattice oxygen; how-
ever, the oxygen atom in CO only comes from gas phase oxygen.
Possible pathways to these products are shown in Scheme 4. We
2
Mo
3
O
12 and is first order for CO formation, which is the same
rate-limiting step is assumed to involve cleavage of the weakest
CAH bond in the reactant. As noted earlier, under these circum-
stances, the rate of alkane of alkene consumption can be described
by Eq. (4). This expression can now be written to identify the sig-
nificance of the apparent pre-exponential factor and the apparent
activation barrier. Thus
2
still assume that CO and CO
methacrolein, in a manner similar to what we proposed earlier
for CO and CO formation from propene. However, since the selec-
2
originate from the precursor to
ꢃꢀ
ꢄ
E
RT
app
2
k
app ¼K
1
k
2
½Sꢃ¼
r
ꢃꢀ
A
app exp
½Sꢃ
tivity of methacrolein from isobutene is lower (ꢄ50%, Fig. 3) than
the selectivity of acrolein from propene (ꢄ70%) [37] on bismuth
ꢄ
D
Hads ꢀEint
¼
r
A
app exp
½Sꢃ
ð8Þ
2
molybdate, the selectivities to CO and CO are higher for isobutene
RT
than for propene. As is shown in Scheme 4, for isobutene, the reac-
tion sequence begins with the 2-methyl-allyl species formed in the
where
r
is the symmetry number (i.e., how many equivalent CAH
rate-limiting step. Reaction of adsorbed O
posed to form an intermediate leading to CO and CO
dization, a new intermediate is formed, which can also produce CO
and CO . This intermediate is exactly the same as the starting point
of the allyl species formed in the RDS for propene to acrolein [37].
As a result, one isobutene molecule can produce more CO and CO
than propene, leading to a higher selectivity of CO and CO . In this
mechanism, the oxygen in CO comes from O , whereas in CO one
and the other from lattice
2
with this species is pro-
bonds could participate in the rate-limiting step). The symmetry
number for different reactants is shown in Table 9. The apparent
activation energy is the sum of the change in enthalpy of the adsor-
bate upon adsorption of a hydrocarbon molecule onto the surface,
2
. After re-oxi-
2
DHads, and the intrinsic activation energy of the CAH bond breaking,
2
Eint. If we assume that the heat of adsorption is similar for all the
butene isomers, and the difference between propene and 1-butene
on bismuth molybdate is ꢄ3 kcal/mol [43], we can expect a similar
positive correlation between the intrinsic activation energy and the
weakest CAH bond dissociation energy. The stronger the CAH bond,
the more energy is needed to break it. We, therefore, conclude that
a plausible descriptor for relating reactant activity is the strength of
the weakest CAH bond.
2
2
2
of the oxygen atoms comes from O
oxygen.
2
4.2. Correlation of the apparent activation energies and pre-
exponential factors with the molecular structure of the reactants
Based on the preceding discussion, we plot the apparent activa-
tion energy versus the weakest CAH bond dissociation energy
measure for the oxidative dehydrogenation of propane, 1-butene,
and trans-2-butene, and for the oxidation of propene [37] and iso-
The discussion of the mechanism and kinetics for the oxidative
dehydrogenation of propane, 1-butene, and trans-2-butene and for
the oxidation of isobutene is very similar at temperatures above
6
73 K, for which the catalyst is saturated with oxygen and the
butene carried out over Bi
2 3
Mo O
12, Bi0.85V0.55Mo0.45O4, and on
1
0
8
6
4
2
0
2
4
propane
isobutene
propene
-
-
t-2-butene
1-butene
8
12
16
20
24
28
32
Apparent Activation Energy on Bi Mo O / (kcal/mol)
2
3
12
1
0
8
6
4
2
0
2
4
1
0
8
6
4
2
0
propane
propane
isobutene
propene
isobutene
propene
-
-
t-2-butene
-2
-4
t-2-butene
1-butene
1-butene
5
10
15
20
25
5
10
15
20
25
Apparent Activation Energy on BiVO / (kcal/mol)
Apparent Activation Energy on
4
Bi0.85Mo0.45V0.55O / (kcal/mol)
4
Fig. 11. Correlation of the natural log of apparent pre-exponential factor with the apparent activation energy for formation of the principle products occurring over Bi
top), BiVO (bottom left), and Bi0.85 (bottom right).
2 3 12
Mo O
(
4
V0.55Mo0.45O
4

Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
99
BiVO
4
. Fig. 10 shows that for reactions over a given metal oxide, the
kinetics are strongly dependent on the state of catalyst reduction,
which in turn depends on the catalyst and reactant composition,
and on the reaction temperature and reactant partial pressures.
For conditions where the catalyst is fully oxidized state, the
kinetics of all reactions are nearly identical. The rate of reaction
is linearly dependent on the alkane or alkene partial pressure
and independent of the oxygen partial pressure, and the apparent
activation energy ranges from 6.5 kcal/mol to 31.5 kcal/mol
depending on the reactant composition. The kinetics for oxidative
dehydrogenation and oxidation over fully oxidized catalysts can be
interpreted in terms of a Mars van Krevelen mechanism in which
the reactant adsorbs reversibly on the catalyst surface and then
undergoes cleavage of the weakest CAH bond in the rate-limiting
step. In the case of 1-butene and trans-2-butene, the allyl radical
formed contains a relatively weak CAH bond which can undergo
cleavage leading to 1,3-butadiene. By contrast, the allyl radical
formed from isobutene does not contain such a CAH bond and
instead is first stabilized by addition to an oxygen atom on the
apparent activation energy correlates with the strength of the
weakest CAH bond. We note that the observed correlation is con-
sistent with the theoretical interpretation of the intrinsic activa-
tion energy developed by Getsoian et al. for the oxidation of
propene to acrolein [44]. The authors of this study reported that
the intrinsic activation energy (Eint) determined from density func-
tional theory can be interpreted in terms of the following sum:
E
int ¼ECAH þE
G
þEOAH
ð9Þ
in which ECAH, EOAH, and E
sociation of a CAH bond, formation of an OAH bond, and the band
gap of the catalyst, respectively.
For a given reactant, the bond dissociation energy of the CAH
bond involved in the rate-limiting step is the same and indepen-
dent of catalyst composition. The DFT calculations suggest that
the OAH bond formation energy does not depend strongly on the
identity of the nearest neighbor to oxygen, leading to the conclu-
sion that the principle term in Eq. (9) depending on catalyst com-
G
refer to the energy required for dis-
catalyst surface. Subsequent cleavage of an H atom from the a-car-
bon atom of the vinyl alkoxide intermediate leads to the formation
of methacrolein. This pattern of reaction is similar to that observed
previously for the oxidation of propene to acrolein [37]. The appar-
ent activation energy and pre-exponential factors for both the
oxidative dehydrogenation of propane, 1-butene, and trans-2-
butene, and for the oxidation of propene and isobutene increase
with the strength of the weakest CAH bond of the reactant
involved in the rate-limiting step. A correlation of the apparent
pre-exponential factor with the apparent activation energy, and
hence the weakest CAH bond of the reactant is also observed. It
is proposed that this trend is attributable to a transition state
structure that is more strongly bound to the catalyst surface and
is more negative in entropy change the stronger is the CAH bond
that must be cleaved. For a given reactant, it is also observed that
the lower the apparent activation energy is the lower the band gap
of the catalyst, and hence, it eases of undergoing reduction.
A second regime is observed when the catalyst cannot be main-
tained in a fully oxidized state, and the rate of reaction becomes
rate-limited by the reoxidation of the catalysts. This is the case
for the oxidative dehydrogenation of 1- and trans-2-butene over
position is E
apparent activation energy should increase with this parameter.
A correlation of Eint with E for propene oxidation to acrolein was
reported by Getsoian et al. [44] A similar trend can be deduced
from Fig. 10, consistent with the decrease in the measured band
G
, and hence, that for a given value of ECAH, the
G
gap in the order Bi
2.32 eV) > BiVO (2.06 eV).
The next question is whether the apparent pre-exponential fac-
2 3 4
Mo O12 (2.70 eV) > Bi0.85V0.55Mo0.45O
(
4
tor, Aapp, correlates with the apparent activation energy. This ques-
tion is difficult to address unambiguously because it is not known
whether the number of active sites, [S], is constant with changes in
catalyst composition. For this reason, Fig. 11 shows plots of
ln(Aapp[S]) versus Eapp for three representative catalysts, since it
is reasonable to assume that the number of active sites for a given
catalyst composition is independent of the composition of the
reactant. It is evident from Fig. 11 that for each catalyst there is a
positive correlation between ln(Aapp[S]) and Eapp. It is notable that
the data points for isobutene oxidation lie farthest from the trend
line and always above this line. Although the reason for this devia-
tion is not known, we hypothesize that nature of the deviation sug-
gests that the isobutene may access a larger number of surface
oxygen atoms than that can any of the other reactants. This inter-
pretation is certainly qualitatively consistent with the observation
that isobutene is a more effective reducing agent than 1-butene or
trans-2-butene.
4
Bi0.85V0.55Mo0.45O at low-temperature regime. In this case, the
rate of oxidative dehydrogenation becomes zero order in alkene
partial pressure and fractional oxygen partial pressure, and the
apparent activation energy becomes independent of reactant
composition.
A third reaction regime is observed for isobutene oxidation
under conditions that result in a substantial degree of catalyst
reduction. For Bi Mo O12 and Bi0.85Mo0.45V O4, the activation
2 3 0.55
energy is 44.0 kcal/mol at all temperatures when the isobutene
partial pressure is raised to 0.067 atm. This same apparent activa-
The positive correlation of Aapp with Eapp indicates that the pro-
duct of the pre-exponential factor for k
adsorption for K in Eq. (8) increases with the activation energy
for this reaction. While it is possible that changes in the entropy
of adsorption also contribute to changes in the magnitude of Aapp
2
and entropy term of
1
,
0.55 4
tion energy is observed for Bi0.85Mo0.45V O at temperatures
below 703 K when the isobutene partial pressure is 0.0167 atm.
this seems unlikely, since the values for Eapp and Aapp are very simi-
lar for both propene oxidation and butene oxidative dehydrogena-
tion. We, therefore, suggest that the principle cause of the
increases in Aapp is attributable to increase in the activation
entropy for Step 2. This trend is what would be expected for a
looser transition state involving a higher activation barrier.
Acknowledgments
This work was funded by Director, Office of Science, Office of
Basic Energy Sciences of the U.S. Department of Energy under
Contract No. DE-AC02-05CH11231. Portions of this research were
carried out at the Argonne National Laboratory (ANL), managed
by the University of Chicago-Argonne, LLC, for the U.S.
Department of Energy, Office of Science.
5
. Conclusions
The oxidative dehydrogenation of 1-butene and trans-2-butene
to 1,3-butadiene, the oxidative dehydrogenation of propane to pro-
pene, and the oxidation of isobutene to methacrolein were investi-
gated over Bi1ꢀx/3
V
1ꢀxMo
x
O
4
for a range of temperatures and
Appendix A. Supplementary material
reactant partialpressures. These investigations were complemented
by characterization of the oxidation state of the catalyst by XANES
and UV–Vis spectroscopy. These studies show that the reaction
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2015.02.015.

1
00
Z. Zhai et al. / Journal of Catalysis 325 (2015) 87–100
[
[
[
[
[
[
[
[
[
[
21] E.V. Hoefs, J.R. Monnier, G.W. Keulks, J. Catal. 57 (1979) 331–337.
References
22] M.W.J. Wolfs, P.H.A. Bastist, J. Catal. 32 (1974) 25–36.
23] P. Mars, D.W. van Krevelen, Chem. Eng. Sci. 3 (1954) 41–59.
24] R.K. Grasselli, Top. Catal. 21 (2001) 79–88.
[
[
[
[
1] J.D. Idol, U.S. Patent 2, 904, 580, 1959.
2] G.W. Hearne, M.L. Adams, U.S. Patent 2, 451, 485 1948.
3] P.H.A. Batist, J.F.H. Bouwens, G.C.A. Schuit, J. Catal. 25 (1972) 1–11.
4] J.M. Lopez Nieto, E.A. Burga, G. Kremenic, Ind. Eng. Chem. Res. 29 (1990) 337–
25] R.K. Grasselli, J. Chem. Educ. 63 (1986) 216–221.
26] C.R. Adams, H.H. Voge, C.Z. Zorgan, W.E. Armstrong, J. Catal. 3 (1964) 379–386.
27] P.H.A. Batist, C.J. Kapteijns, B.C. Lippens, G.C.A. Schuit, J. Catal. 7 (1967) 33–49.
28] C.R. Adams, T.T.J. Jennings, J. Catal. 2 (1963) 63–68.
29] A.C.A.M. Bleijenberg, B.C. Lippens, G.C.A. Schuit, J. Catal. 4 (1965) 581–585.
30] P.H.A. Batist, A.H.W.M. Der Kinderen, Y. Leeuwenburgh, F.A.M.G. Mets, G.C.A.
Schuit, J. Catal. 12 (1968) 45–60.
3
42.
[
[
5] S. Breiter, H.-G. Lintz, Chem. Eng. Sci. 50 (1995) 785–791.
6] H. Lee, J.C. Jung, H. Kim, Y.-M. Chung, T.J. Kim, S.J. Lee, S.-H. Oh, Y.S. Kim, I.K.
Song, Catal. Commun. 9 (2008) 1137–1172.
[
[
[
7] D.A.G. van Oeffelen, J.H.C. van Hooff, G.C.A. Schuit, J. Catal. 95 (1985) 84–100.
8] M. Farinha Portela, Top. Catal. 15 (2001) 241–243.
9] H.H. Voge, C.D. Wagner, D.P. Setvenson, J. Catal. 2 (1963) 58–62.
[
[
[
[
31] R.S. Mann, D.W. Ko, J. Catal. 30 (1973) 276–282.
32] F. Benyahia, A.M. Mearns, Appl. Catal. 66 (1990) 383–393.
33] F. Benyahia, A.M. Mearns, Appl. Catal. 70 (1991) 149–159.
34] M. Cesari, G. Perego, A. Zazzetta, G. Manara, G. Notari, J. Inorg. Nucl. Chem. 33
[
[
[
10] W.M.H. Sachtler, Recl. Trav. Chim. 82 (1963) 243–245.
11] L.D. Krenzke, G.W. Keulks, J. Catal. 61 (1980) 316–325.
12] W. Ueda, K. Asakawa, C.L. Chen, Y. Moro-Oka, T. Ikawa, J. Catal. 101 (1986)
(
1971) 3595–3597.
[
[
[
[
[
[
[
[
[
[
35] A.W. Sleight, K. Aykan, D.B. Rogers, J. Solid State Chem. 13 (1975) 231–236.
36] A.W. Sleight, W.J. Linn, Ann. N. Y. Acad. Sci. 272 (1976) 22–44.
37] Z. Zhai, A. Getsoian, A.T. Bell, J. Catal. 308 (2013) 25–36.
38] A. Getsoian, V. Shapovalov, A.T. Bell, J. Phys. Chem. C 117 (2013) 7123–7137.
39] A. Getsoian, A.T. Bell, J. Phys. Chem. C 117 (2013) 25562–25578.
40] R.E. Jentoft, S.E. Deutsch, B.C. Gates, Rev. Sci. Instrum. 67 (1996) 2111–2112.
41] K. Keizer, PH.A. Batist, G.C. Schuit, J. Catal. 15 (1969) 256–266.
42] Y. Moro-oka, Catal. Today 45 (1998) 3–12.
3
60–368.
[
[
[
[
[
[
13] G.W. Keulks, J. Catal. 19 (1970) 232–235.
14] B. Grzybowska, J. Haber, J. Janas, J. Catal. 49 (1977) 150–163.
15] J. Haber, W. Turek, J. Catal. 190 (2000) 320–326.
16] W.J. Linn, A.W. Sleight, J. Catal. 41 (1976) 134–139.
17] I. Matsuura, G.C.A. Schuit, J. Catal. 20 (1971) 19–39.
18] M. Oliveira, M.F. Portela, M.J. Pires, F.R. Riebeiro, Can. J. Chem. Eng. 61 (1983)
8
7–92.
43] M. Krivanek, P. Jiri, J. Strnad, J. Catal. 23 (1971) 259–269.
44] A. Getsoian, Z. Zhai, A.T. Bell, J. Am. Chem. Soc. 136 (2014) 13684–13697.
[
[
19] J.C. Jung, H. Kim, A.S. Choi, Y.-M. Chung, T.J. Kim, S.J. Lee, S.-H. Oh, I.K. Song, J.
Mol. Catal. A 259 (2006) 166–170.
20] B. Grzybowska, J. Haber, J. Komorek, J. Catal. 25 (1972) 25–32.