Fuel 128 (2014) 120–140

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Propylene production via propane oxidative dehydrogenation over

VOx/c-Al2O3 catalyst
Sameer A. Al-Ghamdi a,b, Hugo I. de Lasa a,⇑
a
Chemical Reactor Engineering Centre, University of Western Ontario, London, Ontario, Canada
b
Research & Development Center, Saudi Aramco Oil Company, Dhahran, Saudi Arabia

h i g h l i g h t s g r a p h i c a l a b s t r a c t

 We study propane oxidative

dehydrogenation in a fluidized-bed Propane CH
reactor under O2-free conditions. H3C CH2
Propylene
 We use fludizable VOx/c-Al2O3 CH2 100

catalyst prepared with close control CH3 CH3 5% VOx /Al2O3

7% VOx /Al2O3

of Brønsted acidity. 80 10% VOx /Al2O3

 We employ VOx supported on c-Al2O3

Selectivity (%)
60
catalysts with 5–10 wt.% vanadium
loading. Isolated VO4 surface species Polymeric VOx surface species C3 H6
40
(monovanadate) (polyvanadate)
 We employ NH3-TPD and pyridine
O O O O
FTIR to assess the effect of vanadium
O O ODH
V V V V V 20 COx

loading on acidity. O O O O O O O O O O O

 Reactivity tests gave 11.7–15.1% γ-Al2O3 Support 0

10 11 12 13 14 15 16 17 18
propane conversion and 67.6–85.9% Increasing vanadium loading (5-10 wt.%) C3H8 Conversion (%)

propylene selectivity.

a r t i c l e i n f o a b s t r a c t

Article history: This study contributes with a new vanadium oxide catalyst supported on c-Al2O3 especially designed for
Received 18 November 2013 propane ODH under oxygen-free conditions. This catalyst is prepared with different vanadium loadings
Received in revised form 12 February 2014 (5–10 wt.%) via wet impregnation using ammonium metavanadate as a precursor. The prepared catalyst
Accepted 15 February 2014
is characterized using BET surface area, H2-TPR, NH3-TPD, O2 chemisorption, Laser Raman Spectroscopy,
Available online 11 March 2014
pyridine FT-IR and XRD. The characterization of the supported VOx/c-Al2O3 catalysts by Raman spectros-
copy reveals that monomeric VOx species are dominant at low vanadium loadings. Polymeric VOx species
Keywords:
increase, however, at higher loadings until surface monolayer coverage is reached. Pyridine FTIR and
Propane ODH
Oxidative dehydrogenation
NH3-TPD reveal that this VOx/c-Al2O3 catalyst shows progressive acidity reduction with vanadium addi-
Vanadium oxide tion while compared to bare alumina. This controlled acidity reduction is achieved with a higher Brønsted
Lattice oxygen acidity. The performance of the prepared catalysts is established having in mind possible industrial pro-
CREC Riser Simulator pane ODH applications. This is accomplished using a CREC Fluidized Bed Riser Simulator at 475–550 °C
and at atmospheric pressure. Successive propane injections for ODH experiments (without catalyst
regeneration) over a partially reduced catalyst show that the prepared catalysts display promising pro-
pane conversions (11.73–15.11%) and propylene selectivities (67.65–85.89%) at 475–550 °C. Propylene
selectivity increases while that of COx decreases as the degree of catalyst reduction augments with
consecutive propane injections. Under such oxygen-free conditions oxygen from the catalyst lattice is
consumed via propane ODH. This demonstrates that a controlled degree of catalyst reduction is very
desirable for high propylene selectivity in ODH.
Ó 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +1 519 661 2144; fax: +1 519 850 2931.
E-mail address: hdelasa@eng.uwo.ca (H.I. de Lasa).

http://dx.doi.org/10.1016/j.fuel.2014.02.033
0016-2361/Ó 2014 Elsevier Ltd. All rights reserved.
 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140
Nomenclature
ri rate of reaction i (mol/g s)
VOx vanadium oxide surface species
XC3H8 propane conversion (%)
Si selectivity of component i (%)
Yi Yield of product i
SBET Brunauer–Emmet–Teller specific surface area (cm2/g)
Dp particle diameter (nm)
Vatom number of vanadium atoms
Nopropane initial moles of propane (mols)
Wcat. weight of catalyst (g)
i representing all chemical species including coke
Greek symbols
b rate of temperature increase (°C/min)
1. Introduction
The conversion of light alkanes to their corresponding olefins
has been the subject of intensive research over the last two dec-
ades [1–3]. This has been driven by a worldwide steady increase
on the demand for light olefins [4–20]. Currently, light olefins are
produced via steam cracking of naphtha or catalytic cracking/dehy-
drogenation. However, these processes suffer from several limita-
tions related to their high energy requirements, coke formation,
selectivity control and thermodynamic constraints [21–27]. Oxida-
tive dehydrogenation (ODH) has been investigated as an alterna-
tive technology for olefins production. ODH does not suffer in
principle from the drawbacks of traditional methods. Oxidative
dehydrogenation of propane is of particular importance with pro-
pane being a main component of natural gas. This makes propane
a preferable raw material, to be a substitute of naphtha in the
manufacturing of propylene [28].
Compared to the current olefin production technologies, pro-
pane oxidative dehydrogenation has the advantages of being a
low temperature process and having an enhanced catalyst lifetime
due to the presence of oxygen that limits coke formation. More-
over, as an exothermic process, propane ODH can overcome the
thermodynamic limitations of non-oxidative dehydrogenation
through the formation of water product. However, the low selec-
tivity toward the desired propylene product is still a main issue
in propane ODH process. This has been in fact, a major ODH draw-
back limiting ODH opportunities for reactor scale up. Propylene
higher reactivity than propane reactivity in ODH favors a compet-
itive propylene combustion with the undesired formation of addi-
tional amounts of carbon oxides [2,29–33].
Regarding ODH, new reactor concepts might also help to make
oxidative dehydrogenation a feasible alternative for olefin produc-
tion. Fluidized bed reactors offer some outstanding advantages in
comparison to conventional reactor systems, namely: controlled
isothermal conditions, uniform residence time distributions and
the absence of mass transfer limitations. Thus, one can circumvent,
with fluidized beds, the potential issues with hot spots in fixed bed
catalytic reactors that may interfere with reactor performance and
damage the catalyst. Moreover, fluidized-bed reactors with peri-
odic catalyst re-oxidation offer the means that enable the transport
of reduced catalyst species from the oxidative dehydrogenation
unit (zone) to the re-oxidation unit (zone). Therefore, the operation
of twin reactors, one for the ODH and one for catalyst regeneration,
appears to be an attractive alternative for the implementation of
this technology at the industrial scale.
121
Abbreviations
ODH oxidative dehydrogenation
MW molecular weight
CREC Chemical Reactor Engineering Center
XRD X-Ray Diffraction
TPR Temperature-Programed Reduction
TPO temperature-programed oxidation
TPD Temperature-Programed Desorption
LRS Laser Raman Spectroscopy
XPS X-ray photoelectron spectroscopy
FID flame ionization detector
TCD thermal conductivity detector
AOS Average Oxidation State
While there is significant research on ODH catalysts [34–38],
there is limited information that deals with light paraffins ODH
in dense fluidized beds, risers or downers reactors. These studies
are however confined to butane ODH [39–42].
Propane ODH offers new opportunities for innovative fluidized
bed technology using riser and downer reactors. In this regard, eth-
ane ODH over VOx/c-Al2O3 has been successfully produced in the
CREC Riser Simulator in the absence of molecular gas oxygen.
Promising ethane conversions and ethylene selectivities have al-
ready been demonstrated by CREC researchers [43,44]. Thus, it is
in the interest of the present study to extend our research and
explore propane ODH in the CREC Riser Simulator over a
VOx/c-Al2O3 catalyst in an oxygen-free atmosphere.
For propane conversion to propylene via ODH, supported vana-
dium oxide catalysts have been reported as the most active and
selective catalysts [1,32,45–55]. This is due to the vanadium ability
to provide lattice oxygen for hydrogen removal from alkanes
[2,56]. Several studies in the literature have shown that supported
vanadium catalysts are more active in ODH reactions with surface
VOx coverage in the sub-monolayer region [32,48,57–60]. How-
ever, due to the higher reactivity of propylene while compared to
propane, high propylene selectivities are only attainable at low
propane conversions. Thus, a catalyst for ODH should be designed
in such a way that it can selectively convert propane to propylene
without further oxidation of propylene to carbon oxides.
The catalytic behavior of supported vanadia catalysts in ODH
reactions is generally controlled by three factors: (i) the structure
of the VOx surface species; (ii) the redox properties of the VOx sur-
face species; and (iii) the acid–base character of the catalysts and
support material. These three factors can be strongly influenced
by the type of support and the vanadium loading [32,57,61–63].
The structure of alumina-supported vanadia has already been con-
sidered in several studies [1,8,32,48,57,64–68]. At low vanadium
loading, isolated VOx surface species (monovanadates) are exclu-
sively present. As VOx surface density increases with loading, sur-
face structures evolve from isolated monovanadates to polymeric
polyvanadates until monolayer coverage is reached. A crystalline
V2O5 phase was reported at high vanadium loading in addition to
amorphous vanadia phases. In this respect, variations of lattice
oxygen binding strength in different VOx surface species formed
on the surface of metal oxide supports have been suggested. This
appears to be a main factor affecting the catalytic performance of
alumina-supported vanadia catalysts in ODH reactions [66,69]. In
fact, it is generally accepted for alkane ODH that isolated tetrahe-
dral VOx species (which are obtained at low V-loadings) are more
122 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140
selective but less active than polymeric VOx species [70]. On the
other hand, there is a relationship between catalyst reducibility
and VOx surface structures on a given metal oxide support. In this
respect, the reducibility of the surface VOx species increases with
surface VOx coverage. Thus, the following trend for the reducibility
of the different supported vanadia species can be considered: poly-
meric surface VOx > isolated surface VOx > crystalline V2O5 nano-
particles [49,57,71–73]. Moreover, the extent of the reducibility
of supported vanadium oxide catalysts is affected by the type of
support used. This variation of vanadium oxide reducibility on dif-
ferent metal oxides supports could be related to the reducibility of
the different VAO-Support bonds existing on different support
types [50,74,1,57].
Several studies in the technical literature have shown that the
acid–base support character when using vanadium oxide catalysts
for ODH plays an important role in determining olefin selectivity
[57,62,70,73,75–78]. This was explained due to the fact that the
acid–base character of the support can control the vanadium cata-
lyst reactivity/selectivity. This occurs as a result that the acid–base
support properties influence both the reactants and product
adsorption/desorption. Thus, acidic catalysts should favor basic
reactant adsorption and acidic product desorption, thereby pro-
tecting these chemical species from further oxidation to carbon
oxides [79].
For instance, higher selectivities to ethylene have been obtained
over acid catalysts like the VOx/c-Al2O3 catalysts. These findings
were justified as a result of the fact that higher support acidity de-
creases the interaction between the ethylene product and the cat-
alyst [59]. In line with this, the significance of Lewis acidity for
ethane ODH was shown in our previous studies [43,44]. In these
studies, ethane ODH was reported using a VOx catalyst supported
on a chi-type alumina (c-Al2O3) support with controlled Lewis
acidity. For propane ODH reactions, low acidity with dominant
Brønsted acid sites is desirable to facilitate rapid propylene desorp-
tion, thus, avoiding its further oxidation to COx. However, the acid-
ity has to be kept at levels that can favor propane adsorption,
yielding, in this manner, acceptable overall catalytic activity [73].
Thus, there is a delicate balance in ODH between the support
acid–base character and the support acid–base strength.
Considering the different acidic properties of alkane feeds as
well as of olefins products and consistent with our efforts to estab-
lish a catalyst for ODH, a new VOx/c-Al2O3 fludizable catalyst for
propane ODH was prepared in the present study. This catalyst
was envisioned to have a controlled acid–base character on a gam-
ma-alumina (c-Al2O3) support. In line with this, the selected
c-Al2O3 type of support was thought to have a tailored low acidic
character, resulting in VOx/c-Al2O3 catalysts with high Brönsted
acid site density. Moreover, the new c-Al2O3 support was envi-
sioned as having a low degree of dehydroxylation during the ther-
mal pre-treatment phases, with most of hydroxyl groups
remaining on the support surface.
Given the above described facts, the goal for this study is to re-
port the performance of a new VOx/c-Al2O3 fludizable catalyst for
propane ODH in the absence of gas phase oxygen. This
VOx/c-Al2O3 catalyst is prepared with carefully chosen VOx load-
ings. This is done to ensure that surface sub-monolayer VOx cover-
age is achieved, with a close control of acidity and Brønsted sites.
2. Materials and methods
2.1. Catalyst preparation
The VOx/c-Al2O3 catalyst samples used in this investigation
were prepared by the wet saturation–impregnation of a commer-
cial porous c-Al2O3 support (SASOL, CataloxÒ SCCa 5/200), with
an aqueous solution of NH4VO3 (Sigma Aldrich, 99%) and oxalic
acid (Sigma Aldrich, 99%) in a 1:2 weight ratio at a pH of 2. This
solution was prepared by mixing ammonium metavanadate
(NH4VO3) with oxalic acid, heating and stirring the mixture until
clear before impregnating the support. Vanadium concentrations
in the solution were varied in order to ensure the desired VOx con-
centration in the catalyst. The impregnation was performed at
70 °C with continuous stirring. The solvent was then evaporated
under vacuum and the resulting cake was dried in air at 120 °C
for 16 h. Finally, the particles were calcined under a gas chroma-
tography quality air stream at 600 °C for 4 h. Once the catalyst
samples prepared, the bulk amount of vanadium in the catalyst
samples were determined using Inductively Coupled Plasma (ICP).
2.2. Catalyst characterization
2.2.1. BET surface area
The specific surface area and pore size distribution of the bare
c-Al2O3 support and the VOx/c-Al2O3 catalysts were determined
in an ASAP 2010 Analyser manufactured by Micromeritics using
Nitrogen adsorption–desorption at 77 K. Before each measure-
ment, a 0.2–0.30 g catalyst sample was degassed at 250 °C for 2 h
in order to ensure a clean and dry surface. The adsorption iso-
therms were measured in 106 to 1 pressure ranges and the surface
area and pore diameters were calculated using the method of
Brunauer, Emmett and Teller (BET).
2.2.2. X-ray Diffraction (XRD)
Powder X-ray Diffraction (XRD) measurements were considered
to identify the crystallographic structure of the prepared
VOx/c-Al2O3 catalyst. The XRD diffraction patterns of the catalyst
samples and the support material were obtained using an Ultima
IV instrument from Rigaku Instruments with a monochromatic
Cu Ka radiation source (wavelength = 0.15406 nm, 40 kV, 40 mA)
and a normal scan rate of 2° scan per minute within the 2h range
from 10° to 90° with a step size of 0.02.
2.2.3. Temperature Programed Reduction (TPR)
Temperature Programed Reduction (TPR) experiments on the
VOx/c-Al2O3 catalyst were used to assess the reducibility of the
catalyst samples. They were conducted using a Micromeritics
AutoChem II 2920 Analyzer. For each experiment, an amount of
100–150 mg of catalyst sample was first pretreated for 2 h at
250 °C in a 5% O2/He gas mixture at a flow rate of 50 ml/min. This
was followed by cooling down the sample to ambient temperature
using the same flow rate. Then, the pretreated sample was flushed
with Argon (99.99%) at a 50 ml/min flow rate for 30 min. Hereafter,
the argon flow was replaced by a flow of 10% H2/Ar gas mixture at a
flow of 50 ml/min while heating the catalyst to 900 °C at a heating
rate of 10 °C/min. Hydrogen consumption was monitored at the
exit gas stream by using a thermal conductivity detector (TCD).
The amount of hydrogen uptake by the sample can be calculated
through numerical integration of the area of the TPR profile.
2.2.4. NH3 Temperature Programed Desorption (TPD)
Temperature programed desorption of ammonia (NH3-TPD)
experiments was used to determine the acid property of the cata-
lysts. The AutoChem II Analyzer from Micrometrics was used for
NH3-TPD experiments. For each experiment, an amount of 100–
150 mg of catalyst sample was first pretreated for 1 h at 550 °C
in a 5% O2/He gas mixture at a flow rate of 50 ml/min. Then, the
samples were cooled to 100 °C and brought to saturation using a
4.55% NH3/He gas mixture at a flow rate of 50 ml/min for 1 h. After
that, the ammonia flow was switched off, and replaced by a flow of
a pure inert gas (He) at a rate of 50 ml/min for 1 h at the same tem-
perature to remove physically adsorbed ammonia. The tempera-
ture was then raised up to 600 °C at a rate of 10 °C/min. As the
 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140
temperature was increased, the ammonia desorbed as it gained en-
ough energy to overcome the activation energy barrier. The NH3
concentration of the effluent gas from the sample was monitored
by a thermal conductivity detector (TCD).
2.2.5. Pyridine FTIR
The nature of catalyst surface acid sites was determined using
Fourier Transform Infrared Spectroscopy (FTIR) and by using pyri-
dine as probe molecule. Before the analysis, the catalyst samples
were dried for 2 h under N2 flow at 550 °C and then cooled to
100 °C. The samples were kept at 100 °C and saturated with pyri-
dine using a N2 stream containing pyridine. Adsorption of pyridine
was allowed for 1 h. Then, the samples were flushed with pure N2
at 100 °C for 90 min, in order to remove weakly adsorbed pyridine.
Following this, samples were placed in a diffuse reflectance infra-
red spectroscopy (DRIFTS) cell. FTIR spectra were recorded at room
temperature using a Bruker IFS55 FTIR spectrometer operating at a
4 cm1 resolution and 100 scans.
2.2.6. O2 pulse chemisorption
Oxygen pulse chemisorption was used to determine the degree
of dispersion of the prepared VOx/c-Al2O3 catalysts. Experiments
were performed using the AutoChem II 2920 Analyzer from
Micromeritics. For each experiment, an amount of 100–150 mg of
fresh catalyst was first pretreated for 1 h at 550 °C in a 5% O2/He
gas mixture at a flow of 50 ml/min. This was followed by cooling
of the sample to ambient temperature under the same flow rate.
Then, the pretreated sample was flushed with Argon (99.99%) at
a flow rate of 50 ml/min for 30 min to remove any adsorbed oxy-
gen. Hereafter, the samples were pre-reduced at 370 °C for 2 h in
a flow of 10% H2/Ar gas mixture at a flow rate of 50 ml/min. Sam-
ples were then purged for 30 h at the same temperature using pure
He gas (99.99%) at a 30 ml/min flow rate to remove physically ad-
sorbed hydrogen. Subsequently, pulses of a known volume of oxy-
gen (1 ml loop volume, 5% O2/He gas mixture) were injected into
the carrier gas (He) until the saturation of the sample was attained.
A thermal conductivity detector (TCD) was employed to quantify
oxygen at the exit gas stream. Oxygen chemisorption uptake was
determined as the difference of two successive adsorption iso-
therms measured at 370 °C. The dispersion was then expressed
as the ratio between the oxygen uptake and the vanadium atoms
available at the catalyst surface (O/Vsurf), assuming an O to V
chemisorption stoichiometry of 1:1.
2.2.7. Laser Raman Spectroscopy (LRS)
Laser Raman Spectroscopy (LRS) was employed to gain funda-
mental information about molecular structures of different vana-
dium oxide surface species (VOx) on a c-Al2O3 support at
different V loadings. Laser Raman spectra were obtained using
dehydrated catalyst samples in a Renishaw Model 2000 Raman
Spectrometer equipped with a thermoelectrically cooled CCD
detector (73 °C). The samples were excited with a 633 nm Ar
ion laser line. A laser power of approximately 2 mW irradiating
the sample was used. Most of the samples were analyzed at full
power; only the un-supported V2O5 sample was analyzed at 10%
power as the signal was too strong. The sample dehydration was
carried out at 450 °C under dry air for 2 h and then cooled to room
temperature. Following this, Raman spectra were obtained at room
temperature at a spectral resolution of 1 cm1. Sample exposure
times were typically 10 and 30 s, with this being a function on
the vanadium loading.
2.3. Propane ODH catalytic test
The catalytic reaction runs of propane oxidative dehydrogena-
tion over supported vanadium oxide catalyst samples were estab-
123
lished using the Chemical Reactor Engineering Center (CREC) Riser
Simulator. The CREC Riser Simulator is a bench scale mini-fluidized
bed reactor invented by de Lasa (1992) [80]. This mini-fluidized
reactor (capacity of 50 cm3) operates under batch conditions and
is designed for catalyst evaluation and kinetic studies under fluid-
ized bed (riser/downer) reactor conditions. A detailed description
of the CREC Riser Simulator, the operational procedure and system
set-up have already been described in a recent publication by
Al-Ghamdi et al. [44]. A schematic diagram of the CREC Riser
Simulator’s components and the fluidization patterns inside the
Riser Simulator catalysts are provided in Fig. 1a.
Thermal and catalytic experiments of propane ODH were car-
ried out at four temperature levels (475, 500, 525 and 550 °C),
one catalyst loading (0.76 g) and four contact times (5, 10, 15
and 20 s). All thermal and catalytic runs were repeated 3 times
to ensure reproducibility of the experimental results. The carbon
mass balance closures, which considered CO, CO2, CH4, ethylene,
ethane, propylene, propane and carbon deposited over the catalyst
were in the ±2% range. Total carbon formed on the spent catalyst
was measured by a TOC Analyzer with a spent catalyst sample that
had been used in the most severe conditions in this study (i.e. after
10 consecutive propane injections, 20 s contact time and 550 °C).
Total carbon formed was found to be in the range of 1–1.5 wt.%.
The effluent from the reactor is analyzed online by a gas chro-
matograph (Shimadzu GC-2010) equipped with three packed col-
umns: HayeSepD 100/120 mesh (30 ft  1/800 O.D. S.S., Supelco)
column, Carboxen-1000 60/80 mesh (15 ft  1/800 O.D. S.S., Supe-
lco) column and Carboxen-1004 80/100 mesh (6.5 ft  1/1600 O.D.
S.S., Supelco) column. The Carboxen-1000 and Carboxen-1004 col-
umns were used for separating H2, O2, N2, Ar, CO, and CO2, and they
were connected in series to a thermal conductivity detector (TCD).
On the other hand, all hydrocarbons (methane, ethane, ethylene,
propane and propylene) were separated in the HayeSepD column
and detected by a flame ionization detector (FID).
The performance of the catalyst in the oxidative dehydrogena-
tion of propane was examined on the basis of propane conversion
and selectivities. Propane conversion, selectivity to a carbon con-
taining product and propylene yield were defined as follows:
P
i mi n i
Propane Conversion; X C3 H8 ð%Þ ¼ P  100 ð1Þ
3npropane þ i mi ni
mi ni
Selectivity to a product; Si ð%Þ ¼ P  100 ð2Þ
i mi ni
XC3 H8 ð%Þ  SC3 H6 ð%Þ
Propylene Yield ðYÞ ¼ ð3Þ
100
where ni is the moles of gaseous carbon containing product i, mi is
the number of carbon atoms in gaseous carbon containing product
i and npropane is the moles of unconverted propane in the product
stream.
Fluidizability is an important catalyst property for its use in flu-
idized-bed reactors. Thus, to confirm its adequacy for fluidized bed
conditions, the fluidizability of the prepared catalysts was assessed
based on their particle size distribution and apparent density. The
particles size distribution (PSD) of the prepared VOx/c-Al2O3 cata-
lysts was measured using a Mastersizer 2000 from Malvern Instru-
ments. The PSD of the various prepared VOx/c-Al2O3 catalysts
displayed a narrow particle size distribution around 90 lm. Almost
50% of the amount of catalyst particles was between 40 and 80 lm
in size while the other 50% of catalyst particles were in the range
between 80 and 110 lm.
The apparent particle density (AD) of the catalyst samples was
assessed using a method established at the CREC [81]. Apparent
particle density values for the various VOx/c-Al2O3 catalysts used
are reported in Table 1a.
124 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140

Fig. 1a. Overview of the CREC riser simulator reactor body. Adapted from Al-Ghamdi et al. [44].

Table 1a vanadium loadings. The surface area of the fresh (activated) bare
Apparent particle densities of VOx/Al2O3 catalysts. c-Al2O3 support was found to be 203.52 m2/g. After calcination at
Sample Apparent particle density (g/cm3)
600 °C during 4 h, the metal-free calcined c-Al2O3 support showed
a slight decrease in the specific surface area to 190.84 m2/g. This
c-Al2O3 1.21
mild reduction in surface area was attributed to the collapse/block-
5% VOx/c-Al2O3 1.53
7% VOx/c-Al2O3 1.70 age of some of the support pores during calcination. Furthermore,
10% VOx/c-Al2O3 1.81 after impregnation with the vanadium metal and calcination at
600 °C for 4 h, the specific surface area of the V-loaded catalyst
samples displayed a further gradual decrease with vanadium load-
ing. This VOx/c-Al2O3 surface area reduction with the vanadium
addition can be ascribed to the blocking of some of the alumina
pores by VOx species.
Moreover, Fig. 2 shows the pore size distributions for the bare
c-Al2O3 support and the VOx/c-Al2O3 catalyst samples with differ-
ent vanadium loadings, as determined from the desorption branch
of the N2 isotherms. The pore size distribution displays a 7.2 nm
average for the samples. A very small number of pore counts have
pore size values below 2 nm and almost none above 50 nm. This
indicates that very few of the pores lie outside of the mesoporous
range of 2–50 nm.

3.1.2. VOx surface species and monolayer coverage

Fig. 1b. Geldart diagram for classification of particles fluidization [82].
Using the average particle size, the particle apparent density
obtained above, the fluidization regime of the VOx/c-Al2O3 parti-
cles were determined, as shown in Fig. 1b using Geldart’s powder
classification chart [82]. These characteristics were further con-
firmed experimentally using a Plexiglas model of the CREC Riser
Simulator, specially manufactured for flow visualization [81].
Finally, the chemical species transport processes from the bulk
of the fluid to the particle external surface and inside the particles
are neglected [81].
3. Results and discussion
3.1. Catalyst characterization
3.1.1. Surface area
Table 1b reports the BET surface areas of the bare c-A12O3
supports and the VOx/c-Al2O3 catalyst samples with different
Table 1 reports the nominal vanadium oxide surface density on
a c-Al2O3 support. Vanadium surface density is defined as the
number of vanadium atoms present per unit specific surface area
of the catalyst and is expressed in terms of Vanadium atoms per
nm2. In order to find the VOx surface density for the various cata-
lyst samples, the vanadium weight fraction and BET surface area
of each sample are required as shown in the following equation:
Surface density ðVatoms =nm2 Þ ¼
wt:% Vanadium loading  6:023  1023
ð4Þ
Molecular Weight of Vanadium  100  Surface Area
Fig. 3a and b describes the catalyst specific surface areas and the
calculated VOx surface densities for the VOx/c-Al2O3 catalysts as a
function of vanadium loading. It can be seen that the presence of
VOx species only had a modest effect on the specific surface area
of the support as shown in Fig. 3a. As a result and as shown in
Fig. 3b, VOx surface densities increased almost linearly with
increasing vanadium content.
Regarding vanadium oxide coverage, the maximum amount of
amorphous two-dimensional vanadium oxide surface species
 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140 125

Table 1b
Characteristics of alumina support and supported vanadia catalyst.

Sample V contenta (wt.%) SBET (m2/g) Vpb (cm3/g) Dpc (nm) Surface density (V/nm2) % Monolayerd
c-Al2O3 (activated) 0 203.5 0.54 7.13 – –
c-Al2O3 (calcined) 0 190.8 0.52 7.27 – –
5% V/c-Al2O3 5 188.6 0.45 6.74 3.1 31.3
7% V/c-Al2O3 7 179.2 0.46 7.46 4.6 46.2
10% V/c-Al2O3 10 153.3 0.39 7.17 7.7 77.1
a
V content, in wt.% was determined by ICP.
b
Vp is a single-point pore volume calculated from the adsorption isotherm at P/Po = 0.99.
c
The average pore sizes were calculated by applying the BJH method on the desorption branches of the isotherms.
d
Calculated assuming surface area of 0.1 nm2 per VO34 species.

monolayer coverage. It is at these conditions at which crystalline

3.0
Al2O3 V2O5 surface species start forming. For such a surface density
5% VO x/Al2O3 range, VOx surface species exist simultaneously with different frac-
2.5 7% VO x/Al2O3
tions as isolated vanadate species (monovanadate) and polymeric
10% VO x/Al2O3
vanadate species (polyvanadates). However, the 5% VOx/c-Al2O3
2.0 sample shows a surface density value of 3.1 which is near the the-
dV/dlog (D)

oretical monolayer coverage of monovanadate surface species.

1.5
1.0
0.5
0.0
1 10
Pore Diameter (nm)
Fig. 2. Pore size distribution for the c-Al2O3 support and the VOx/c-Al2O3 catalyst
samples determined from nitrogen desorption isotherms.
(VOx) in contact with the oxide support as a monomolecular layer
is called monolayer coverage. The horizontal lines in Fig. 3b show
the VOx surface densities corresponding to theoretical monolayer
coverage of monovanadates (2.9 Vatoms/nm2) or polyvanadates
(8.8 Vatoms/nm2). These values of theoretical monolayer coverage
were found based on structural calculations given by Centi [83].
The values of surface density were found to be 3.1, 4.6 and
7.7 Vatoms/nm2 for the 5% VOx/c-Al2O3, 7% VOx/c-Al2O3 and 10%
VOx/c-Al2O3 catalyst samples, respectively. These values are be-
tween the 2.9 and 8.8 Vatoms/nm2 values for theoretical monolayer
coverage of monovanadate and polyvanadates, respectively. This
allows us to argue that the VOx/c-Al2O3 catalysts of the present
study are likely to be in the monolayer coverage configuration
which is below the upper limit for theoretical polymeric vanadia
3.1.3. H2-TPR and degree of reduction
TPR experiments were used to investigate the reducibility of
vanadium oxide surface species deposited on the c-Al2O3 support
H2-TPR profiles of bulk V2O5 and VOx/c-Al2O3 samples with differ-
ent vanadium loadings are shown in Fig. 4. The reduction profile
for bulk V2O5 exhibits four major peaks at 650, 690, 740 and
850 °C. Similar observations on the reduction of bulk V2O5 were re-
ported by Bosc et al. [84] and Koranne et al. [85]. These authors ob-
served similar multiple reduction peaks when bulk V2O5 was
treated in a 5% H2 in argon up to 1000 °C. The presence of multiple
peaks is attributed to the reduction of V2O5 to V2O3 through the
intermediate formation of oxide species with different oxidation
states according to the following sequence:
V2 O5 ! V6 O13 ! V2 O4 ! V2 O3
On the other hand, H2-TPR profiles for the VOx/c-Al2O3 catalyst
samples show only a single symmetric reduction peak extending
from 310 °C to 550 °C with a Tmax between 432 and 452 °C. How-
ever, the 10% VOx/c-Al2O3 sample shows a second weak peak at
about 575 °C. The first low temperature reduction peak observed
in all VOx/c-Al2O3 prepared samples can be assigned to the reduc-
tion of amorphous monomeric and polymeric VOx surface species.
The second highest temperature peak displayed for the 10%
VOx/c-Al2O3 sample can only be assigned to the reduction of bulk
V2O5-like surface species. It can also be noticed in Fig. 4 that the
first reduction peak shifts to a lower temperature with increasing

300
(a) 10 (b)
Polyvanadate
VOx Surface density (Vatoms/nm2)

250
8
Surface Area (m2/g)

200

6
150

4
Monovanadate
100

2
50

0
0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
V content (wt.%) V content (wt.%)

Fig. 3. BET surface areas (a) and VOx surface densities (b) for supported vanadium oxide catalysts.
126 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140

0.14 reported in Table 2. In a similar manner, the degree of reduction

was developed assuming that V2O5 was the sole initial species
0.12 present on the supported catalyst. With these data, the Average
Oxidation States (AOS) of reduced vanadium oxides were
Bulk V2O5
0.10 determined.
TCD signal (a.u.)

On this basis, the oxidation states of vanadium oxides found

0.08 were in the V2O3.02–V2O3.37 range. One should notice that the most
10% VO x/γ-Al2O3
likely vanadium oxide structure around this value is V2O3. This
0.06 vanadium species has been reported by several authors as a very
7% VO x/γ-Al2O3 stable structure among various vanadium oxides [86,87]. This sug-
0.04 gests that the vanadium phase after complete reduction of V2O5 in
5% VO x/γ-Al2O3 the TPR experiments is the V2O3 or mixed phases of V2O3, with
0.02 V3O3.4 and V3O3.7.

0.00
50 150 250 350 450 550 650
Temperature (οC)
Fig. 4. H2-TPR profile of bulk V2O5 and calcined VOx/c-Al2O3 catalyst samples:
temperature up to 900 °C (operating conditions: heating rate: 10 °C/min; reducing
agent: 10 mole% H2; balance Ar at 50 cm3/min; oxidizing agent: 5 mole% O2;
balance He at 50 cm3/min).
vanadium loading. This suggests that a sub-monolayer coverage
with easily reducible species is formed at higher loading. Thus, it
is speculated that at higher vanadium loadings, there is a higher
abundance of polymeric VOx species which is considered more
reducible than the monomeric VOx that exists at lower vanadium
loadings. This was further confirmed by determining the degree
of reduction (% reduction) of each catalyst sample.
The degree of reduction of a supported vanadium oxide catalyst
can be defined as the ratio of the exposed reducible vanadium
oxide to the actual vanadium oxide amount on the catalyst. This
degree of reduction can be calculated for each sample by knowing:
(a) the amount of reducible metal in the catalyst sample, which can
be related to the amount of hydrogen consumed during the TPR
experiments, (b) the actual vanadium amount on the catalyst, (c)
the initial reducible V2O5 species present on the supported catalyst
(Reaction 1).
Table 2 reports the hydrogen consumption, the maximum
reduction temperature and the reduction degree for each catalyst.
It can be observed in Table 2 that the degree of catalyst reduction
increases with vanadium loading. These results support the
assumption that there is increased catalyst reducibility under
hydrogen atmosphere for catalyst samples with higher vanadium
loading.
V2 O5 þ 2H2 ! V2 O3 þ 2H2 O
Furthermore, in an attempt to determine the nature of vana-
dium oxide surface species, the average oxidation state of vana-
dium was calculated using the amount of hydrogen consumed
via TPR. This allowed us to qualitatively establish the nature of
existing phases by comparing their average oxidation states. These
average oxidation states were compared with those expected as
750 850 3.1.4. NH3-TPD
In addition to the redox properties, surface acidity is a very
important aspect of supported vanadia catalysts. Surface acidity
can determine their catalytic activity; and it is governed by the
nature of the support used and the surface structure of surface van-
adia species (VOx). Thus, a correlation of surface acidity with the
structure of supported vanadia and hence with their catalytic
activity is very important.
TPD of ammonia was used in this study to compare the total
acidity and acidity strength of the bare c-Al2O3 support and the
VOx/c-Al2O3 catalyst samples with different vanadium loadings.
Fig. 5 shows the relationship between the thermal conductivity
of the desorbed species (NH3 pre-adsorbed at 100 °C) with a linear
increase in temperature (15 °C/min).
Table 3 reports ammonia uptake by various samples and their
desorption temperatures (Tdes.). The total acidity of each catalyst
sample was established on the basis of integration of desorbed
NH3 as given by the TCD (temperature conductivity detector) pro-
files. One should note that the NH3-TPD profile for bare c-Al2O
showed a broad initial desorption peak at 170 °C followed by a
high temperature desorption peak at 350 °C. These two peaks can
be attributed to NH3 desorption from weak and strong Lewis acid
sites, respectively. It can also be seen in Fig. 5 that the addition
of vanadium on c-Al2O3 resulted in acidity changes: from
c-Al2O3 related Lewis sites to vanadium oxide promoted Brönsted
sites. As a result, this leads to a decreased acid strength, yielding a
lower density of weaker acid sites. Moreover, for the vanadium-
containing samples, it was observed that raising the vanadium
loading resulted in increased acidity of catalyst samples. This
was confirmed by the increased NH3 uptake at higher vanadium
species loading.
The acidity of supported-vanadia catalysts has been the subject
ð5Þ
of several studies [88–93]. Unsupported bulk V2O5 powder was
found by Busca et al. [91] to show both surface Brönsted and Lewis
acid sites. Moreover, it was also observed by Datka et al. [94] and
Miyata et al. [95] that the formation of VOx surface species on oxide
supports such as Al2O3 and TiO2 is accompanied by a decrease in
the density of surface Lewis acid sites and an increase in the
density of surface Brönsted acid sites.

Table 2
TPR data comparing hydrogen consumption using VOx/c-Al2O3 catalysts with varying amounts of vanadium (wt.%).

Sample Tmax (°C) H2 uptake (cm3 STP/g) % Reductiona AOSb VOx surface species
T1 T2
5% V/c-Al2O3 451.8 – 16.42 81.28 3.37 V2O3.37
7% V/c-Al2O3 445.7 – 22.74 83.06 3.34 V2O3.34
10% V/c-Al2O3 432.6 575 37.03 99.18 3.02 V2O3.02
Average oxidation state of V V2 O5 ! V6 O13 ! VO2 ! V4 O7 ! V2 O3
þ5 þ4:33 þ4 þ3:5 þ3

a
Based on V2O5 + 2H2 ? V2O3 + 2H2O.
b
Average Oxidation State (AOS).
 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140 127

0.09 1490 Py-Bronsted

Py-Lewis
10%VO x/Al2O3 1540 1610 1640
1448
0.08 Old 10% V/c-Al 2O3

0.07
7%VO x/Al2O3
0.06 10% V/γ-Al2O3

Absorbance
TCD (a.u.)

0.05
5%VO x/Al2O3 Fresh γ-Al2O3

0.04
7% V/γ-Al2O3

0.03

0.02 5% V/γ-Al2O3

0.01
γ-Al2O3
0.00
100 150 200 250 300 350 400 450 500 550 600 1350 1400 1450 1500 1550 1600 1650 1700
Temperature (οC) Wavenumber (cm-1)

Fig. 5. NH3-temperature programed desorption profiles as measured with a TCD for Fig. 6. FTIR spectra of pyridine adsorbed on c-Al2O3 and VOx/c-Al2O3 samples.
fresh c-Al2O3 and various VOx/c-Al2O3 catalyst samples. (Sample weight: 0.1 g;
heating rate: 15 °C/min, He flow rate: 30 ml/min, NH3 pre-adsorbed at 100 °C.)

on surface Brønsted acid sites showing that the VOx/c-Al2O3 sam-

Table 3
Temperature programed desorption of NH3 for VOx/c-Al2O3 catalyst samples and bare
c-Al2O3 support.
Catalyst sample Tdes. (°C)
c-Al2O3 200, 350
5% V/c-Al2O3 194
7% V/c-Al2O3 180
10% V/c-Al2O3 196
Zou et al. [88] studied the surface acidity of SiO2, c-Al2O3 and
TiO2 supported vanadia catalysts. This was done using both micro-
calorimetry and infrared spectroscopy as well as using ammonia as
a probe molecule. The acidity in terms of nature, number and
strength was correlated to surface structures of vanadia species
on the catalyst surface. It was found that surface acidity of sup-
ported vanadia catalysts is strongly affected by the vanadia species
dispersion, surface structure, nature of the support and vanadium
loading. In addition, microcalorimetric studies of NH3 adsorption
showed a gradual decrease in the heat of ammonia adsorption
and ammonia coverage with the addition of VOx on the c-Al2O3
support. The heat of ammonia adsorption appears to decrease pro-
gressively with the increase of the vanadium loading up to the
point when the monolayer coverage is complete. This means that
VOx species may cover some of the strong Lewis acid sites of the
c-Al2O3 support. Thus, as vanadium loading increases, this leads
to a higher abundance of Brönsted sites.
3.1.5. Pyridine FTIR
To gain additional information about the acidity of the prepared
VOx/c-Al2O3 catalysts samples, Fourier Transform Infrared Spec-
troscopy (FTIR) using pyridine as a probe molecule was used. By
using this characterization technique, one is able to assess the cat-
alyst’s acidity based on Brønsted or Lewis sites concentrations.
Fig. 6 shows the FTIR spectra of the c-Al2O3 support and all the
VOx/c-Al2O3 samples used in the present study. It can be noted in
Fig. 6 that the FTIR spectra of adsorbed pyridine on the bare
c-Al2O3 support do not exhibit any bands other than that at
1448 cm1 which arise from the adsorption of pyridine on surface
Lewis acid sites. This shows that the c-Al2O3 support does not con-
tain any Brønsted acid sites strong enough to react with pyridine.
On the other hand, in addition to the band at 1448 cm1, the FTIR
spectra of adsorbed pyridine on the VOx/c-Al2O3 samples showed a
1540 cm1 band. This band arises from the adsorption of pyridine
ples possess both Lewis and Brønsted acid properties. It can also
be observed that the relative intensity of the bands at 1540 cm1
augments with increasing vanadium loading. This is in agreement
NH3 uptake (cm3 STP/g) with the results obtained by NH3-TPD where the total acidity of the
VOx/c-Al2O3 samples increases with vanadium loading.
8.26
10.94
Based on these findings, it can be concluded that the total acid-
11.50 ity and the number of surface Brønsted acid sites augment with
13.02 increasing surface VOx coverage. These results are consistent with
findings of other studies [32,57,73]. In addition, the multi-bands in
the spectral region between 1580 and 1660 cm1 are generally as-
signed to hydrogen-bonded pyridine while the bands at 1490 cm1
relate to both Lewis and Brønsted acid species [96,97].
Furthermore, the 1540 cm1 and 1448 cm1 peaks were used to
quantify the Brønsted and Lewis acid sites concentrations, respec-
tively. Table 4 reports the Brønsted/Lewis acid sites ratios for all
VOx/c-Al2O3 samples. It can be noticed that Brønsted acid sites
are dominant over Lewis acid sites. This was, in fact, true for all
samples, with the Brønsted/Lewis ratio increasing steadily with
vanadium loading. Such an increase in Brønsted acidic sites of
VOx-based catalysts with vanadium loading has been reported in
several studies. This is an important factor to improve VOx catalytic
activity for ODH of light alkanes [1,32,49,57,71,73].
Fig. 6 and Table 4, both report data for the 10%VOx/c-Al2O3 cat-
alyst sample prepared in our previous study for ethane ODH [44]. It
can be noticed that at the same V-loading (10%), the less-acidic
c-Al2O3 support resulted in a higher Brønsted/Lewis ratio and a
larger number of Brønsted acid sites. This acidity difference is par-
ticularly noticeable if one compares the properties of the catalyst
of this study with the ones of other catalysts developed by our
research group for ethane ODH [44].
3.1.6. Metal dispersion
In supported metal oxide catalysts, metal dispersion plays an
important role in determining their catalytic activity. There are
several factors which can control the dispersion of active metal
on the support materials such method of preparation, type of
Table 4
Acid properties of VOx/c-Al2O3 catalyst samples using pyridine FTIR.
Catalyst sample Brønsted/Lewis sites ratio
5% V/c-Al2O3 1.8
7% V/c-Al2O3 4.2
10% V/c-Al2O3 5.1
)old) 10% V/c-Al2O3 3.2
128 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140

support and metal precursor used. The most commonly used 20000
180
method for the determination of metal oxide dispersion of a cata-
lyst is the selective chemisorption of probe gases like oxygen on a 325 993
700
pre-reduced catalyst sample. 15000 448 520 567
235 345 944 1033
O2 chemisorption experiments were conducted by dosing small (e)
amounts (17.2 lmol) of O2 over a pre-reduced catalyst sample.

Intensity (a.u.)
Oxygen doses were repeated on the sample every 60 s until two (d)
10000
consecutive peaks with constant area were recorded in the TCD.
Oxygen uptake after each pulse was calculated from the area dif- (c)

ference between two consecutive peaks. The pulses were termi-

5000 (b)
nated when the areas of two consecutive peaks were the same,
indicating that no more O2 uptake was taking place. O2 uptake
curves obtained at 370 °C, on various VOx/c-Al2O3 catalyst samples, (a)

as a function of O2 pulses, are shown in Fig. 7. 0

Table 5 reports the cumulative oxygen uptake values obtained 200 300 400 500 600 700 800 900 1000 1100

at 370 °C on various VOx/c-Al2O3 catalyst samples as a function Raman shift (cm-1)

of V loading. It is shown that oxygen uptake increases with vana-
Fig. 8. Dehydrated Raman spectra for (a) calcined c-Al2O3, (b) 5% VOx/c-Al2O3, (c)
dium content up to 10 wt.% VOx loading. Therefore, using a stoichi-
7% VOx/c-Al2O3, (d) 10% VOx/c-Al2O3 and (e) bulk V2O5.
ometry of O/V = 1/1, the metal dispersion, defined as the ratio of
oxygen uptake to V content, was estimated for each sample. This
was achieved by utilizing the total number of V atoms present in 2 h. Raman spectra results are shown for bulk V2O5, activated
the sample as well as the number of chemisorbed oxygen atoms. c-Al2O3 support and various VOx/c-Al2O3 catalysts with different
Metal dispersion is found to decrease with the increase in vana- V loadings. Table 6 also provides the positions of the bands associ-
dium loading. Moreover, the metal dispersions as reported in Ta- ated with various VOx species. The narrow 1030–1035 cm1 band
ble 5 also give an active equivalent particle diameter of vanadia is assigned to the stretching mode of the V@O bond in isolated
particles smaller than 4 nm. This vanadia particle diameter was monovanadate surface species. On the other hand, the broad bands
found to be consistent with the X-ray diffraction patterns where in the 700–945 cm1 region are ascribed to the stretching modes of
no diffraction lines above the background were observed. This V@O (945–1030 cm1) and VAOAV (670–945 cm1) in polyvana-
was true for all catalyst samples prepared. date surface species [62,72,98]. The remaining bands, appearing
at 180, 235, 325, 345, 448, 520, 567 and 993 cm1 are assigned
to bulk V2O5 crystals [59,99].
3.1.7. Laser Raman Spectroscopy
Fig. 8 also shows that the c-Al2O3 support does not exhibit any
Fig. 8 reports the Raman spectra obtained at ambient tempera-
Raman bands in the 100–1100 cm1 region due to the ionic charac-
ture after sample thermal treatment at 450 °C under air flow for
ter of the AlAO bonds [32]. Moreover, it can be seen that the struc-
tures of the VOx surface species as detected by Raman
16 spectroscopy are dependent on vanadium loading and hence VOx
5% VO x/g-Al2O3 surface density. Thus, as the vanadium loading increases from
14 7% VO x/g-Al2O3 5 wt.% to 10 wt.%, the proportion of V2O5 or polyvanadate to
Cumulative O2 Uptake (cm3 STP/g)

10% VO x/g-Al2O3 monovanadate species increases. For example, in the catalyst sam-
12
ple with 5% V loading, the only band detected is that for mono-
vanadate species at 1033 cm1. The absence of polyvanadate
10
bands in Raman spectra suggests that VOx species exist as isolated
8 monovanadate structures. On the other hand, Raman bands attrib-
uted to polyvanadate species are detected in the 750–1000 cm1
6 range for the catalyst sample with 7% V loading and higher which
indicates the coexistence of both polyvanadate and monovanadate
4
species on c-Al2O3. At higher loadings of 10% V, Raman bands cor-
responding to crystalline V2O5 are also detected.
2
In summary, the Raman spectra results presented in Fig. 8 and
0 Table 6 suggest that isolated tetrahedral monovanadate structures
1 3 5 7 9 11 13 15 17 19 21 23 25 anchored on the support with VAOAA1 bonds are prevalent at low
O2 Pulse No. vanadium loading. As V loading increases, two-dimensional poly-
vanadates form, leading to the formation of VAOAV bonds which
Fig. 7. Cumulative oxygen uptake versus number of oxygen pulses on VOx/c-Al2O3 connect to neighboring VOx species. At higher V loadings, three-
catalyst samples (Tads. = Tred. = 370 °C). dimensional structures form via the reaction of VOx species
exceeding monolayer coverage with polyvanadates structures,
which ultimately crystallize into bulk V2O5 form.
Table 5
Oxygen uptake, metal dispersion and active particle diameter on various VOx/c-Al2O3
catalysts.
3.1.8. X-ray diffraction
Sample O2 uptake (cm3/g) % Dispersiona (O/V) Active particle Fig. 9 describes the X-ray diffraction patterns of bulk V2O5, the
diameter (nm)
c-Al2O3 support and various VOx/c-Al2O3 catalysts of various VOx
5% V/c-Al2O3 6.96 63.26 1.97 loadings calcined at 600 °C. No diffraction lines corresponding
7% V/c-Al2O3 9.98 43.25 1.99
to the vanadium oxide compounds were detected in the
10% V/c-Al2O3 13.16 59.81 2.09
VOx/c-Al2O3 samples except in those for the bare c-Al2O3 support
a
Dispersion = fraction of vanadium atoms at the surface, assuming Oads/Vsurf = 1. at 2h positions of 48° and 67°. This may be an indication to either
 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140 129

Table 6
Raman band assignments for vanadium oxide species (Sameer: table was expanded to see all numbers clearly).

Sample Monovanadate (V@O) Polyvanadates (V@O, VAOAV and VAOAAl) V2O5 Support
Bulk V2O5 None None 180, 235, 325, 345, 448, 520, 567 and 993 –
c-Al2O3 (activated) – – – None
5% V/c-Al2O3 1033 None None –
7% V/c-Al2O3 1033 650–945 None –
10% V/c-Al2O3 1033 650–945 180, 235, 325, 448, 520, 567 and 993 –

that VOx species exist as a dispersed amorphous phase on the alu-

300
mina surface or that V2O5 if present, exist as small XRD undetect-
able crystalline nanoparticles (<4 nm). Based on Laser Raman
250
Spectroscopy (LRS) and H2-TPR results reported previously in Sec-

+
tion 3.1.7 of this manuscript, the samples with 5% and 7% V loading
showed only an amorphous VOx phase. However, the sample with 200

+
Intensity (a.u.)
10% V loading showed some crystalline V2O5 nanoparticles coexis- Spent 10%VO x/ γ−Al2O3
ting with the amorphous VOx phase. All these phases were in the 150
sub-monolayer coverage as indicated by the O2-chemisorption dis-
persion results. These findings are in agreement with the published
100 Spent 7%VO x/ γ−Al2O3
data with similar VOx loadings on c-Al2O3, where VOx primarily
present as vanadate or polyvanadate which is known to be X-ray
amorphous. Some crystalline V2O5 nanoparticles could exist at 50 Spent 5%VO x/ γ−Al2O3
lower V loadings which might form due to catalyst preparation
procedures [32,65,85]. Moreover, no evidence of AlV3O9 was de- 0
tected in any of the samples, suggesting that the reaction between 10 20 30 40 50 60 70 80 90
VOx and c-Al2O3 is negligible during the treatment at 600 °C. Intensity (a.u.)
Furthermore, in order to establish the structural stability of the
Fig. 10. XRD patterns of Spent VOx/c-Al2O3 catalysts with different vanadium
VOx/c-Al2O3 catalysts, XRD were considered for spent catalysts
loadings after 10 consecutive propane injections. (Operation conditions: T = 550 °C,
after 10 consecutive propane injections for propane ODH reactions contact time = 20 s, C3H8 injected = 10 ml, catalyst loaded = 0.76 g.).
at the most severe reaction conditions considered in this study (i.e.
20 s contact time and 550 °C). Fig. 10 reports the XRD patterns of Catalytic propane ODH runs were studied under an oxygen-free
the spent VOx/c-Al2O3 catalysts. One can notice the absence of atmosphere employing lattice oxygen of vanadium oxide in the
any diffraction patterns due to structural changes other than that VOx/c-Al2O3 catalysts. To achieve this, two sets of experiments
for the bare c-Al2O3 support at 2h positions of 48° and 67°. This were considered: single-injection and multi-injections experi-
can be viewed as an indication of high catalyst stability under ments. Single-injection experiments are used to study the interac-
the selected reaction conditions. tion of propane with the fully oxidized catalyst in a reaction/
regeneration cycles. In contrast, multi-injection experiments are
used to change the catalyst state from completely oxidized to a
3.2. Propane ODH in the CREC Riser Simulator partially reduced one and to study the influence of the reduction
degree of the catalyst on its performance in propane ODH in con-
Propane ODH experiments were developed in a CREC Riser secutive catalyst reduction cycles. In each set of experiments, the
Simulator operated as a batch reactor mode under fluidized bed only identifiable carbon-containing products other than propylene
conditions. The fluidizability of the various VOx/c-Al2O3 catalyst were CO, CO2, CH4, C2H6 and C2H4.
samples used in this study was ensured in the riser basket by con-
ducting the experiments at high impeller speeds of 5500 rpm. 3.2.1. Thermal runs
Blank thermal runs (without catalyst) were performed using the
empty reactor prior to the catalytic studies to determine the possi-
350 ble contribution of homogenous gas-phase reactions. This allowed
one to clearly distinguish between the catalytic oxidative dehydro-
*

*
*

300 genation effects versus the thermal conversion effect. The thermal
*
*
*

Bulk V 2 O 5 runs were tested using the same reactant (propane) at a tempera-
*
*
*

*
*

250
ture range of 475–550 °C and 20 s contact time which were consid-
Intensity (a.u.)

200 10% VO x / γ− Al 2 O 3 ered as the most severe conditions in the experimental plan. Fig. 11
shows propane conversion in the reactor without any catalyst at
150 the temperature range used in the catalytic runs. It can be seen
7% VO x / γ− Al 2 O 3
in Fig. 11 that propane conversion due to thermal effect displayed
100 low values between 0.21% and 3.12%. Based on these results, ther-
5% VO x / γ− Al 2 O 3
mal cracking was neglected under the conditions studied, and the
+

50
+

γ− Al 2 O 3
conversion observed during the catalytic runs truly represents the
0 catalytic activity of VOx/c-Al2O3 catalysts.
10 20 30 40 50 60 70 80 90
2θ
3.2.2. Single-injection propane ODH experiments
Fig. 9. XRD patterns for Al2O3 support and VOx/Al2O3 catalysts with different Single-injection ODH experiments with cycles of reaction
vanadium loadings (*: V2O5, +: Al2O3). followed by regeneration were developed. Thus, in these
130
3.5
3.0
Propane Conversion (%)
2.5
2.0
1.5
1.0
0.5
0.0
450
Fig. 11. Propane conversion as a function of temperature for the blank runs.
(Operating conditions: reaction time = 20 s, C3H8 injected = 10 ml). Error bars
correspond to standard deviation of three repeats.
experiments, the catalyst was repeatedly reduced by reaction with
propane and then re-oxidized with air at various reaction temper-
atures and contact times. Between the reduction and oxidation, the
catalyst was flushed with pure Argon flow for 15 min. For each
reaction cycle, the conversion and selectivity for the main products
were obtained. At the beginning of each experiment, the fresh cat-
alyst was first heated to the reaction temperature under inert
(Argon) gas flow and then pretreated with air flow for 20 min to
ensure that the catalyst was in a fully oxidized state.
Fig. 12 reports propane ODH over repeated reaction–regenera-
tion cycles at 550 °C using the various VOx/c-Al2O3 catalysts of
the present study. Results show that propane conversions and
product selectivity remain stable during the different runs with
the following trends being observed:
(a) Propane conversion stays at the 60%, 73.2% and 80.1% levels
over the 5%VOx/c-Al2O3, 7%VOx/c-Al2O3 and 10%VOx/c-Al2O3
catalysts samples, respectively.
(b) Propylene selectivities are in the 2.1%, 0.93% and 0.57% levels
over the 5%VOx/c-Al2O3, 7%VOx/c-Al2O3 and 10%VOx/c-Al2O3
catalysts samples, respectively.
(c) COx remaining stable at 88%, 90% and 93% levels over the
5%VOx/c-Al2O3, 7%VOx/c-Al2O3 and 10%VOx/c-Al2O3 cata-
lysts samples, respectively.
As gas phase oxygen was not present in the propane injection, it
can be concluded that all of the reaction products were the result
90
80
C3H8 Conversion (%)
70
60
50
40
30
0 1
Fig. 12. (a) Propane conversion and (b) propylene and COx selectivities during consecutive ODH cycles over the various VOx/c-Al2O3 catalysts. (Operating conditions: catalyst
regenerated after every injection, T = 550 °C, C3H8 injected = 10 ml, reaction time = 20 s, catalyst loaded = 0.76 g). Error bars correspond to standard deviation of three repeats.
S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140
of the interaction with lattice oxygen of the VOx surface species.
Furthermore, low propylene and high COx selectivities at any given
propane conversion were obtained during single injection ODH
experiments as shown in Fig. 12. These low propylene and high
COx selectivities can be related to both the amount and type of sur-
face oxygen species available on the fresh/regenerated catalyst sur-
face. An excessive amount of non-stoichiometric, available lattice
oxygen as well as the existence of non-selective oxygen species
on the catalyst surface is believed to favor the conversion of pro-
pane and the produced propylene into combustion products.
Fig. 12 shows the effect of an excess amount (non-stoichiome-
tric) of lattice oxygen where propane conversion augments with
VOx loading on the catalyst. This is accompanied with a parallel in-
crease in COx selectivity and a decrease in propylene selectivity.
This finding leads us to conclude that an optimum amount of
475 500 525 550 575
VOx loading is needed to have the stoichiometric amount of lattice
Temperature (oC) oxygen necessary to selectively convert propane to propylene dur-
ing ODH. On the other hand, there exist non-selective oxygen spe-
cies on the catalyst surfaces that could be either loosely bound
lattice oxygen from the surface and bulk of the catalyst or weakly
adsorbed oxygen species produced from gas-phase O2 during the
pre-treatment/regeneration steps. This loosely bound lattice oxy-
gen is considered more reactive, and thus, likely to cause car-
bon–carbon bond breakage and promote total oxidation. As a
result, both types of oxygen species may participate in the total
oxidation of propane and contribute to low propylene selectivity.
Fig. 13 describes the selectivity of ODH products over various
catalysts as a function of propane conversion at different reaction
temperatures. One can notice that there is a consistent decrease
in propylene selectivity with a corresponding increase in COx selec-
tivity at high conversions. This shows that propylene is the primary
reaction product of the propane ODH while COx (CO and CO2) are
the secondary products of the consecutive oxidation of propane
and propylene. Based on these results, a classical ‘‘triangular’’
reaction network with ‘‘direct alkane combustion’’ together with
alkene formation and later ‘‘alkene combustion’’ could be envi-
sioned for propane ODH. Moreover, non-zero selectivities of COx
and C3H6 can be expected upon extrapolation to the zero degree
of propane conversion. The non-zero selectivities are an indication
that all reaction products are formed from propane via a competi-
tive (parallel) reaction network.
The interaction of propane feed with the oxygen species on the
catalyst during ODH reactions in the absence of gas phase oxygen
have been the subject of several studies in the literature [28,100–
104]. It has often been suggested for oxidative dehydrogenation
reactions that the selectivity is dependent on the binding strength
of the oxygen. Moreover, it has also been proposed that there may
be different types of oxygen species on the catalyst surface,
5 100
4 90
COx Selectivity (%)
(a) (b) COx
C3H6 Selectivity (%)
3 80
C3H6
2 70
5% VOx /Al 2O3
7% VOx /Al 2O3 1 60
10% VOx /Al 2O3
0 50
2 3 4 5 6 7 8 9 10 11 0 1 2 3 4 5 6 7 8 9 10 11
Cycle # Cycle #
 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140 131

20 100 10 100
(a) (b)
90

Propylene Selectivity (%)

8
Propylene Selectivity (%)

COx Selectivity (%)

15

COx Selectivity (%)

80 95
6
10 70

4
60 90
5
2
50

0 40 0 85
0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80
Propane Conversion (%) Propane Conversion (%)

6 100

5
(c)
Propylene Selectivity (%)

COx Selectivity (%)

4 475 oC
500 oC
525 oC
3 95 550 oC

0 90
0 10 20 30 40 50 60 70 80
Propane Conversion (%)

Fig. 13. C3H6 selectivity as a function of C3H8 conversion at various temperatures for single-injection propane ODH over (a) 5% VOx/c-Al2O3 (b) 7% VOx/c-Al2O3 and (c) 10%
VOx/c-Al2O3 catalyst samples. (Operating conditions: C3H8 injected = 10 ml, catalyst loading = 0.76 g). Error bars correspond to standard deviation of three repeats.

including some that are reactive enough to cause the direct and to- In view of the above, it can be concluded that fully oxidized
tal oxidation of propane. Haber and Witko [100] have proposed (fresh) VOx/c-Al2O3 catalysts are active but not selective for pro-
two type of active oxygen surface species: electrophilic oxygen pane ODH reactions. This could be attributed to the availability
forms (e.g. O2, O 
2 , O ) which have been claimed to perform total of an excess amount (non-stoichiometric) of lattice oxygen as well
oxidation, and nucleophilic lattice oxygen ions O2 which have as the existence of non-selective oxygen species on the catalyst
been claimed to be the more selective toward propylene. More- surfaces. Both types of oxygen species participate in the total oxi-
over, Che et al. and Libre et al. [105–107] have found that oxygen dation of propane and propylene product. Thus, an optimized and
species with different bond strength and electronic character could controlled character (nucleophilic or electrophilic) and an appro-
exist on the catalyst surface. As a result, different types of oxygen priate number of oxygen species on the catalyst surface are impor-
species could form on the catalyst surface resulting in different tant factors to be considered for more selective catalysts.
selectivities toward the ODH products. To elucidate this matter, a second set of experiments was con-
In addition to the nucleophilic and electrophilic natures of dif- sidered in the context of the present study. These experiments
ferent surface oxygen species, Kung and Andersen [108] studied were performed with partially reduced catalysts using a series of
the oxidation of butane over two V/c-Al2O3 catalysts with loadings consecutive propane injections. These experiments were devel-
of 2.9 and 8.2 V/nm2 using both pulses and steady butane feeds. oped for propane ODH reaction with no catalyst regeneration in
These authors found that the number of the oxygen atoms on the between injections. In this manner, the effect of the degree of
catalyst surface and the place where these atoms are incorporated reduction (i.e. the number of available active oxygen species) on
in the reactant molecule are two important factors for determining the activity and selectivity of the VOx/c-Al2O3 catalysts in ODH
the selectivity of these catalysts for butane ODH. Thus, it is impor- reaction was examined very effectively. More details about consec-
tant to control the number of oxygen atoms that are incorporated utive-injections propane ODH experiments are given in the follow-
into the reactant molecule. This number is normally controlled via ing section of the present study.
the residence time of the molecule on the surface, the number of
oxygen atoms available at the active sites and the reactivity of oxy- 3.2.3. Consecutive-injections propane ODH experiments
gen on the surface (i.e. its electrophilicity or nucleophilicity). A As stated above, as an alternative, catalyst samples were subject
similar conclusion on the effect of the number of active oxygen to successive propane injections in the CREC Riser Simulator. In
species was made by Dinse et al. [109] in his study to investigate these ODH experiments, the VOx/c-Al2O3 catalysts are progres-
the oxidative dehydrogenation (ODH) of ethane on alumina-sup- sively reduced via the consecutive propane injections without cat-
ported vanadia. In this study, transient-response experiments were alyst regeneration between the injections. With these data and for
carried out with both a fully oxidized sample of 10 wt.% VOx/c- each of the injections, the instantaneous conversion and selectivi-
Al2O3 and a sample that had been partially reduced in H2 to remove ties for the main products were calculated.
half of the reducible oxygen from the vanadia. The rate of ODH was Every experimental set consisted of 10 successive propane
found to depend on the amount of reactive oxygen available on the injections, contact times between 5 and 20 s and temperatures
catalyst surface. between 475 and 550 °C. In order to minimize the effect of the
132 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140

catalyst to feed ratio, all the experiments were conducted using the decreases as the degree of reduction of the catalyst augments with
same amount of catalyst (0.76 g) and injecting the same amount of the number of consecutive propane injections. This significant
propane (10 ml). The degree of reduction of the catalyst and its change in selectivity indicates that a certain degree of catalyst
selectivities for ODH products was determined by analyzing the reduction is needed in order to obtain good propylene selectivity.
various products resulting from each propane injection. For example, high propylene selectivities of 92.1%, 84.9% and
Fig. 14 reports propane conversion and product selectivities as 77.6% were obtained using the 5%VOx/c-Al2O3, 7%VOx/c-Al2O3
functions of the number of propane injections (pulses) over differ- and 10%VOx/c-Al2O3 catalysts, at conditions where these catalysts
ent VOx/c-Al2O3 catalysts at 550 °C. It can be observed in Fig. 14a were reduced up to 83.2%, 67.4% and 55.06%, respectively.
that the conversion of propane is highest in the first injection for Thus, the above reported results support the view that propyl-
all catalysts and then decreases progressively with increasing ene selectivity in propane ODH reaction over VOx-based catalysts
number of propane injections. However, a small but noticeable is strongly influenced by the binding energy of lattice oxygen.
activity level was maintained even after 10 pulses of propane over When the catalyst is in a more oxidized state, lattice oxygen is
the different VOx/c-Al2O3 catalysts. This could be ascribed to the loosely bound and it is more likely to promote deep oxidation of
availability of an extended supply of lattice oxygen as will be dis- hydrocarbons to carbon oxides. Moreover, there is also the possi-
cussed in more details later in this section. As no molecular oxygen bility that in the first injection, adsorbed oxygen from the regener-
is fed with the injection, the catalyst is believed to supply oxygen ation step (which has a mainly non-selective contribution) could
for the ODH reaction. Following the first propane injection, pro- remain on the VOx/c-Al2O3 catalyst surface. In this case, a selective
pane conversion declined drastically from 58.8%, 76.5% and catalyst surface would be obtained only after the adsorbed oxygen
81.98% to 12.5%, 14.4% and 16.4% over the 5%VOx/c-Al2O3, had been consumed via the first propane injection. This is quite
7%VOx/c-Al2O3 and 10%VOx/c-Al2O3 catalysts samples, respec- apparent for the subsequent propane injections with the propylene
tively. There was a further gradual decrease to 11.5%, 12.4% and selectivity augmenting with the number of propane injections. It is
13.6% with the subsequent propane injections. Thus, it is plausible under these conditions that combustion products (COx) decrease.
that the decrease in propane conversion over VOx/Al2O3 catalysts is Thus, these findings show that in propane ODH reaction network
mainly the result of progressive consumption of the reactive lattice with the catalyst not being regenerated in between propane injec-
oxygen species. tions (oxygen-free environment), propane conversion could be
Fig. 14b also shows the relative variation of propylene and COx achieved with high propylene selectively. This high selectivity for
selectivities. It can be seen that for all catalysts the first injection propylene is the result of the lattice oxygen being the main driver
has very low propylene selectivity and very high COx selectivity. of the ODH reaction. One can also notice that following the first
However, the selectivity for propylene increases and that for COx ethane injection (first pulse), the catalyst tested displays a stable

100 100

90 (a) (b)
80 80
70
C3H8 Conversion (%)

Selectivity (%)

60 60 C3H6
50

40

15

20 COx

10 0
0 1 2 3 4 5 6 7 8 9 10 11 0 1 2 3 4 5 6 7 8 9 10 11
C3H8 Injection # C3H8 Injection #

0.9

0.8
(c)
% Reduction of Catalyst

0.7

0.6

0.5 5% VO x /Al2 O3
7% VO x /Al2 O3
0.4 10% VO x /Al2 O3

0.3

0.2

0.1

0.0
0 1 2 3 4 5 6 7 8 9 10 11
C3H8 Injection #

Fig. 14. ODH propane conversion with successive propane injections (a) propylene and COx selectivity and (b) catalyst degree of reduction as a function of the successive
propane injections ODH over various VOx/c-Al2O3 catalyst samples. (Operating conditions: T = 550 °C, reaction time = 20 s, C3H8 injected = 10 ml, catalyst loaded = 0.76 g).
Error bars correspond to standard deviation of three repeats.
 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140
performance in terms of propane conversion and propylene selec-
tivity even after ten successive reduction runs.
The effect of the catalyst oxidation state on oxidative dehydro-
genation selectivity over vanadium based catalysts has been the
subject of several studies in the literature [43,44,49,104,110–
112]. In these studies, the ODH reaction was carried out in the ab-
sence of gas-phase O2 and the catalyst was reduced via successive
pulses of paraffins (e.g. ethane, propane and butane). It was shown
that olefin selectivities of ODH reactions over various vanadium
based catalysts augments when increasing the reduction degree
of the catalyst. Yoon et al. [113], in a study considering propane
oxidative dehydrogenation over a MgAMoAO catalyst attributed
this to the progressive catalyst reduction. In this respect, the bind-
ing strength of the remaining lattice oxygen augments and as a re-
sult the catalyst becomes more selective. On the other hand, López-
Nieto et al. [49] and Balcaen et al. [104] found in their studies on
propane and butane ODH over VOx/c-Al2O3 catalysts that the
reducibility of the V-atoms can influence the selectivity to ODH
products. This effect was assigned to the selective redox rate of
VOx surface species. In this respect, a higher reaction temperature
was found to accelerate the redox exchanges on the catalyst sur-
face and hence increases the selectivity. Thus, the rate at which
the redox processes take place on the catalyst surface was found
to be another relevant factor in determining the catalytic perfor-
mance of V-based oxidative dehydrogenation catalysts. In another
recent study from our research group, promising results were re-
ported by Al-Ghamdi et al. [43,44] on ethane ODH reaction over
VOx/c-Al2O3 catalysts in CREC Riser Simulator in the absence of
molecular oxygen. It was shown that in the absence of oxygen, eth-
ylene could be produced via ODH with 57.684.5% ethylene selec-
tivity at ethane conversions between 6.5% and 27.6% and at
temperatures between 550 and 600 °C. It was concluded that the
absence of gas-phase oxygen is critical for the selective conversion
of ethane into ethylene with a major role of lattice oxygen in sus-
taining the ODH over several reaction cycles.
The degree of catalyst reduction reported in Fig. 14c is defined
as the ratio of the number of oxygen molecules consumed from
the catalyst during the ODH reaction to the total amount of avail-
able exchangeable lattice O atoms. The later was determined by
first reducing the VOx/c-Al2O3 catalyst sample in a flow of H2. This
was then followed by successive pulses of molecular O2 (O2 titra-
tion) of the reduced sample until no further O2 uptake was moni-
tored. Thus, the oxygen consumed during the catalyst re-oxidation
is believed to re-oxidize the surface and replenish the lattice oxy-
gen. On the other hand, the number of oxygen molecules con-
sumed from the catalyst was calculated after each propane
injection by analyzing the amounts of oxygen-containing products
(i.e. CO, CO2 and H2O). As the ODH reaction was conducted in the
absence of a gas-phase oxygen, removable lattice oxygen was as-
sumed to be the only source of O-atoms in the ODH products.
It can be also observed in Fig. 14c that different VOx/c-Al2O3
catalysts show different degrees of reduction. The catalyst sample
with the highest V loading (10% V) showed a degree of reduction of
55.1% after 10 propane injections while that with the lowest V
loading (5% V) showed degree of reduction of 80.4% by the 10th
injection. This was attributed to the differences in VOx content of
the VOx/c-Al2O3 catalysts. In fact higher levels of lattice oxygen
supply can extend the ODH reaction significantly.
Besides the VOx loading, the extent of the lattice oxygen supply
can be affected by the reaction temperature. Fig. 15 reports the ex-
tent of reduction for the 5% VOx/c-Al2O3 catalyst sample as a func-
tion of reaction temperature and propane injections. It can be
observed that an extended supply of lattice oxygen could be main-
tained by the catalysts at a lower temperature. Thus, as tempera-
ture increases, the degree of catalyst reduction per injection
increases and hence more lattice oxygen is consumed. This results
133
1.0
5% VO x/Al2O3 Catalyst
0.9
0.8
Reduction degree of Catalyst
0.7
0.6
0.5
0.4
0.3
475 oC
0.2 500 oC
525 oC
0.1
550 oC
0.0
0 1 2 3 4 5 6 7 8 9 10 11
C3H8 Injection #
Fig. 15. Variation of catalyst degree of reduction with reaction temperature as a
function of successive-injection propane ODH over a 5% VOx/c-Al2O3 catalyst.
(Operating conditions: reaction time = 20 s, C3H8 injected = 10 ml, catalyst loa-
ded = 0.76 g). Error bars correspond to standard deviation of three repeats.
as a consequence in a smaller lattice oxygen supply. In a study by
Balcaen et al. [114,115], the activation of propane both in the
absence and presence of gas-phase O2 or CO2 over a CuOACeO2/
c-Al2O3 catalyst was investigated. These researchers used a TAP
reactor in the 350–600 °C range. It was found by performing isoto-
pic multi-pulse experiments with C18O2 over the 16O2-pretreated
catalyst at 600 °C that at higher temperatures there is an increased
mobility of lattice oxygen.
To further investigate the extent of lattice oxygen supply by the
VOx/c-Al2O3 catalysts of the present study, a set of experiments
were carried out where the catalyst sample with the lowest V load-
ing (5%) was considered for successive propane injections. This was
done at the most severe experimental conditions (i.e. 20 s contact
time and 550 °C reaction temperature). This experiment was con-
tinued until all the lattice oxygen supply was depleted, as shown
by the full reduction (degree of reduction of 1) reported in
Fig. 16(a).
Fig. 16(a) reports the propane conversion for 23 consecutive
injections. One can notice a progressive decrease in propane con-
version with consecutive injections. In addition, one can observe
that propylene selectivity, as illustrated in Fig. 16(b), showed an
initial increase from a very low value of 2% to around 91%, remain-
ing stable up to the 14th injection. Following this, propylene selec-
tivity showed a gradual decrease to 67.4%. Moreover, this decrease
in propylene selectivity after the 14th injection is accompanied by
a gradual decrease in both COx selectivity and propylene yield as
shown in Fig. 16(c). This is an indicator of a decreased supply of lat-
tice oxygen for the ODH reaction. Hence, lower amount of products
are formed. By the 23rd injection, the catalysts showed a 100% de-
gree of reduction with no further COx formation (0% COx selectiv-
ity). It is interesting to see that the values of propane conversion
and propylene selectivity, at this point, are similar to those ob-
tained in thermal runs. At this stage, this is also a strong indicator
of the absence of any lattice oxygen available for the ODH reaction.
Based on these findings, it can be concluded that the lattice oxy-
gen reservoir on the VOx/c-Al2O3 catalysts makes their use possible
in continuous twin redox fluidized bed reactors. In this type of unit,
only a small fraction of the catalyst or about 1/23 would be contin-
uously transferred from the reduction ODH unit (where lattice oxy-
gen is consumed in the ODH reaction) to the oxidizing unit (where
lattice oxygen is replenished). Once the catalyst is oxygen replen-
ished, it should return to the ODH unit.
Table 7 reports propane conversion and product selectivities for
successive-injections propane ODH experiments with pure C3H8
134 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140

1.0 13.4%, and 15.1% at 550 °C, for the same catalysts samples,
0.9 respectively.
(a) On the other hand, in contrast to the ODH experiments over
0.8
Reduction degree of Catalyst

fully oxidized catalysts (i.e. single-injection ODH experiments),

0.7
propylene selectivity showed a gradual increase with increasing
0.6 reaction temperature, as shown in Fig. 17(b). For example, the cat-
0.5 alyst sample with the 5% vanadium loading gives the highest C3H6
selectivities between 84.4% and 85.9% over the temperature range
0.4
from 475 to 550 °C. For the sample with 7% vanadium loading, the
0.3 C3H6 selectivity increases from 71.2% at 475 °C to 75.3% at 550 °C.
0.2 The C3H6 selectivity of the 10% V catalyst sample is the lowest of
0.1 any of the catalysts examined in this work, with C3H6 selectivity
of 64.1% at 475 °C. Furthermore, it increases to 67.8% at 550 °C. This
0.0
increase in propylene selectivity with reaction temperature is
100
attributed to the variation of catalyst degree of reduction with both
90 (b) reaction temperatures and the number of propane injections.
80 As discussed earlier in this section, the catalyst sample is pro-
Conversion/Slectivity (%)

70 gressively reduced with the successive propane injections. It has

been also shown that the reduction degree of the catalyst is more
60
pronounced at higher temperatures due to higher mobility of the
C 3H 8 conversion
50 lattice oxygen. As a result of this, as the number of propane injec-
C 3H 6 selectivity
40 CO x selectivity tions is increased at higher reactions temperatures, a higher degree
30
of reduction of the catalyst is attained. At such higher degrees of
reduction of the catalysts, the selective pathway toward ODH is
20
preferred over that for combustion and COx formation due to lower
10 availability of lattice oxygen. This is apparent from the observed
0 catalyst performance reported in Fig. 17(c): The increase in propyl-
20
ene selectivity with reaction temperature is accompanied by a
18 decrease in the COx selectivity (CO and CO2).
(c)
16
3.2.3.2. Effect of reaction contact time. The effect of the reaction con-
14
Propylne Yield (%)

tact time was examined by conducting the experiments at four dif-

12
10
8
6
4 version at 5 s contact time and increase to 11.7%, 13.4%, and 15.1%
2
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
C3H8 Injection #
Fig. 16. Changes of degree of catalyst reduction (a) propane conversion, COx
selectivity and propylene selectivity and (b) propylene yield (c) with the number of
propane injections propane ODH over a 5% VOx/c-Al2O3 catalyst. Operating
conditions: T = 550 °C, reaction time = 20 s, C3H8 injected = 10 ml, catalyst
loaded = 0.76 g). Error bars correspond to standard deviation of three repeats.
over various VOx/c-Al2O3 catalysts at various contact times and dif-
ferent temperature levels. The only identifiable carbon-containing
products other than propylene were CO, CO2, CH4, C2H6 and C2H4.
3.2.3.1. Effect of reaction temperature. The effect of the reaction
temperature was examined by conducting successive-injection
propane ODH experiments over the various VOx/c-Al2O3 catalysts
at reaction temperatures between 475 and 550 °C. Fig. 17 reports
propane conversions and selectivities to propylene and COx during
the ODH reaction runs at various reaction temperatures.
It can be observed that catalyst samples with different vana-
dium loadings showed different reactivity during the successive-
injection propane ODH experiments, resulting in 5.4%, 6.4% and
7.2% propane conversion at 475 °C for the 5%V, 7%V and 10%V cat-
alysts samples, respectively. As shown in Fig. 17(a), as the reaction
temperature increases, propane conversion increases to 11.7%,
ferent contact times: 5, 10, 15 and 20 s. Fig. 18(a, c and e) reports
propane conversions and selectivities to propylene and COx during
the ODH reaction runs at various reaction times and 550 °C. It is
observed that propane conversion for all catalysts increases as
the reaction time increases giving 4.1%, 5.2% and 6.2% propane con-
at 20 s reaction time for the 5%V, 7%V and 10%V catalysts samples,
respectively.
One can notice that propylene selectivity augments slightly
with increasing reaction time as shown in Fig. 18(b, d and f). The
catalyst sample with the 5% vanadium loading shows C3H6 selec-
tivities between 70.9% and 76.4% at 5 s contact time and increase
to 84.4–85.9% at 20 s contact time. For the samples with 7% and
10% vanadium loading, the C3H6 selectivity increases from 60.7%
to 66.2% and 55.1% to 60.7% at 5 s contact time to 71.2–75.3%
and 64.13–67.8% at 20 s contact time, respectively. This increase
in propylene selectivity with reaction contact time can be assigned
to the variation of catalyst degree of reduction with contact time is
analogous to the effect of reaction temperature discussed earlier.
Longer contact times between the catalyst and reactant at any
given propane injection yield higher consumption of catalyst lat-
tice oxygen, resulting in higher degree of catalyst reduction for
the subsequent injections.
On the other hand, and as a complementary evidence of this
finding, one can observe that the product combustion selectivity
(COx), shown in Fig. 18(b, d and e), displays a gradual decrease with
increasing reaction time. This demonstrates that limited lattice
oxygen is available at higher extents of reduction and longer con-
tact times. Both conditions favor high propane conversions with
high propylene selectivities.
Thus, one can conclude that the performance of the various
VOx/c-A12O3 catalysts of the present study is affected considerably
by both ODH operation conditions and catalyst regeneration. It ap-
pears that consecutive reaction injections (pulses) are a preferred
 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140 135

Table 7
Conversion and product distribution results for Successive-Injections propane ODH over various VOx/c-Al2O3 catalysts. Reported values are the average of 10 consecutive C3H8
injections.

Catalyst T (°C) Time (s) XC3 H8 (%) Selectivity (%) C3H6 yield (%)
C3H6 COx CH4 C2H6 C2H4
5% VOx/c-Al2O3 475 5 2.35 70.89 86.49 2.76 0.24 0.70 1.63
10 3.29 76.86 89.88 2.97 0.27 0.67 2.50
15 4.19 80.35 91.53 2.97 0.29 0.67 3.35
20 5.38 84.42 94.45 2.94 0.31 0.61 4.53
500 5 3.08 73.37 86.26 4.73 0.60 1.30 2.18
10 4.59 77.53 89.96 4.20 0.69 1.06 3.46
15 5.77 81.89 91.89 4.98 0.94 1.12 4.71
20 7.32 84.70 95.22 4.15 0.89 0.96 6.21
525 5 3.83 75.44 88.01 6.14 1.51 2.26 2.85
10 6.32 79.92 91.09 5.67 2.10 2.03 5.03
15 7.99 83.28 92.90 5.58 2.60 1.99 6.67
20 8.88 85.79 95.52 4.84 2.30 1.75 7.64
550 5 4.08 76.38 89.66 6.30 1.54 3.78 3.13
10 6.08 81.24 91.86 5.49 2.20 2.76 4.98
15 9.12 84.52 93.88 5.05 3.30 2.44 7.73
20 11.73 85.94 96.90 5.41 3.62 1.55 10.11
7% VOx/c-Al2O3 475 5 3.24 60.73 35.81 2.57 0.25 0.64 1.62
10 4.18 67.20 29.30 2.61 0.30 0.59 2.59
15 5.14 67.68 28.76 2.64 0.32 0.61 3.29
20 6.42 71.20 25.19 2.73 0.34 0.55 4.36
500 5 3.74 64.75 29.00 4.45 0.65 1.16 2.14
10 5.60 68.94 24.78 4.29 0.89 1.11 3.53
15 6.89 71.58 22.00 4.37 1.01 1.03 4.69
20 8.11 73.09 21.03 4.02 0.99 0.87 5.82
525 5 4.60 65.29 24.36 6.31 1.76 2.28 2.63
10 6.55 70.60 19.46 5.79 2.20 1.94 4.34
15 9.77 71.98 19.11 5.00 2.31 1.61 6.78
20 11.88 73.79 17.51 4.78 2.40 1.51 8.45
550 5 5.06 66.20 20.01 7.12 2.43 4.24 3.18
10 8.41 71.49 15.64 6.60 2.98 3.30 5.90
15 11.21 74.46 11.67 6.73 3.92 3.22 8.36
20 13.36 75.34 10.43 7.21 4.44 2.58 10.06
10% VOx/c-Al2O3 475 5 3.73 55.12 41.52 2.48 0.27 0.61 1.48
10 4.67 61.99 34.59 2.49 0.33 0.59 2.57
15 5.90 61.67 34.82 2.60 0.34 0.57 3.32
20 7.16 64.13 32.45 2.53 0.36 0.53 4.24
500 5 4.22 57.24 36.74 4.27 0.71 1.04 2.01
10 6.27 63.43 30.05 4.44 1.03 1.06 3.42
15 7.68 64.64 29.27 4.13 1.03 0.92 4.52
20 9.14 65.73 28.66 3.72 1.07 0.82 5.70
525 5 5.26 58.79 30.29 6.60 2.06 2.26 2.47
10 7.24 64.75 25.05 6.04 2.28 1.88 4.10
15 10.79 65.78 25.89 4.85 2.07 1.41 6.56
20 13.41 67.29 23.58 5.21 2.48 1.43 8.32
550 5 6.16 60.64 23.79 7.83 3.18 4.56 3.49
10 10.50 66.18 18.79 7.51 3.68 3.84 6.64
15 12.93 66.81 16.13 8.53 4.75 3.78 8.53
20 15.05 67.77 15.01 8.72 4.99 3.51 10.16

mode of operation for ODH. As a result, regeneration should only
be allowed once this series of consecutive reaction cycles is com-
pleted. One can, for instance, envision that this is equivalent to
operating a twin circulating fluidized reactor process (reactor–
regenerator), where the following occurs: (a) most of the catalyst
stream leaving the reactor is returned to the ODH unit directly;
(b) only a small fraction of the catalyst stream is directed toward
the regenerator. This operational strategy can allow the function-
ing of industrial scale ODH processes with high propylene selectiv-
ities, under conditions close to the ones reported in Fig. 14(a and
b).
3.2.3.3. Effect of vanadium loading. VOx/c-Al2O3 catalysts for pro-
pane ODH have been reported in the technical literature as being
affected by both vanadium loading and hence vanadium surface
density [1,32,50,62,73,116–118]. At low vanadium loading (lower
VOx surface density), a highly dispersed amorphous vanadate
phase is formed. Its structure changes from isolated VO4 species
(monovanadates) to polymeric VO4 species (polyvanadates) at
medium loading. This occurs until the monolayer coverage is
achieved with 100% polymerized species at monolayer surface
coverage. Thus, the observed activities and selectivities toward dif-
ferent products obtained during propane ODH over the various
VOx/c-Al2O3 catalysts can be further related to their surface
density. This would allow the determination of the relative contri-
bution of isolated and polymeric surface VOx species for this cata-
lytic reaction. In order to do this, the initial propane ODH rate
(normalized per gram of catalyst) was compared.
In this respect, one can calculate the initial average reaction rate
(per gram of catalyst) of propane ODH using the following
equation:
 
Nopropane X propane mol
r0ODH ¼ ð6Þ
W cat t gs
where N opropane is the number of moles of propane injected in each
run into the CREC Riser Simulator, Xpropane is the conversion of pro-
pane, Wcat is the weight of catalyst used and t is the reaction time.
136 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140

16 90

14
(a) (b)
85
12
C3H8 Conversion (%)

C3H6 Selectivity (%)

80
10

8 75

6
70
4
65
2

0 60
460 470 480 490 500 510 520 530 540 550 560 470 480 490 500 510 520 530 540 550 560
Temperature ( C) ο Temperature (οC)

35

30 (c)
25
COx Selectivity (%)

5% VOx /Al2O3
7% VOx /Al2O3
20 10% VOx /Al2O3

15

10

0
470 480 490 500 510 520 530 540 550 560
Temperature (οC)

Fig. 17. Temperature effect on ODH propane conversion and product distribution: (a) C3H8 conversion, (b) C3H6 selectivity and (c) COx selectivity for successive-injections
propane ODH over various VOx/c-Al2O3 catalyst samples. (Operating conditions reaction time = 20 s, C3H8 injected = 10 ml, catalyst loaded = 0.76 g). Data points are the
averages values of 10 successive injections excluding the first injection on fully oxidized (fresh) catalyst). Error bars correspond to standard deviation of three repeats.

Fig. 19(a) reports initial propane ODH rates as a function of surface density while the initial rate of formation (per gram of cat-
vanadium surface density at different reaction temperatures. As alyst) of COx increased with VOx surface density. Thus, the propyl-
shown in Fig. 19a, for all temperatures, the ODH rate increases ene selectivity changes as shown in Fig. 20 can be traced to the
with increasing VOx surface density. This suggests that polyvana- formation of extra COx at higher VOx surface species concentration.
date surface species, which are more dominant at higher VOx sur- This increase in COx formation with VOx surface density can be
face densities, are more active than monovanadate species. This attributed to the excess surface oxygen species available at higher
type of dependence is expected as the catalysts used in this work loading. These findings provide further evidence of the role as-
are all in the sub-monolayer VOx surface coverage as shown in signed to the polyvanadates species as being more active but less
Fig. 4. It is expected that the initial propane ODH rate will increase selective than monovanadate. With the increased activity of poly-
with VOx surface density. This will occur until monolayer surface vanadate species, more oxygen species are available for the ODH
coverage (8.8 VOx/nm2) is reached. After this, a decrease in the reaction resulting in reduced propylene and increased COx selectiv-
initial propane ODH rate will start to develop with further in- ities as shown in Fig. 20.
creases in VOx surface density. This decrease in the propane ODH The reactivity variation of different types of VOx surface species
rate with VOx surface density above monolayer coverage has been (monovanadate versus Polyvanadates) could also be explained
attributed to the formation of crystalline V2O5 nanoparticles. These from their ease-of-reduction. Early in this section, H2-TPR was used
nanoparticles decrease the number of exposed catalytic active sites to determine the extent of reduction of each catalyst used in this
[48,57]. study. The obtained% reductions were 81.3%, 83.1% and 99.2% for
Furthermore, Fig. 20 describes the influence of VOx surface den- the 5% VOx/c-Al2O3, 7% VOx/c-Al2O3 and 10% VOx/c-Al2O3 catalysts,
sity on product selectivities during propane ODH over VOx/c-Al2O3 respectively. It appears that the V5+ species, and the (VAOAAl)
catalysts. It can be seen that at all reaction temperatures, the sam- bond in the monomeric VOx surface species are more difficult to
ple with the lowest vanadia surface density (3.1 VOx/nm2) and con- reduce than the (VAOAV) or (V@O) bonds in the polymeric VOx
taining predominately monovanadates surface species shows the surface species. This can also be justified considering that propane
highest propylene selectivities. However, increasing vanadia sur- ODH reaction initial reduction rates over the VOx/c-Al2O3 catalyst
face density yields a decrease in propylene and increase in COx involve formation of easy-to-reduce VOx surface species. Previous
selectivities products. studies [49,59,71] have suggested that the higher propane ODH
These results are in well agreement with the initial rates of for- rates observed on supported metal oxide catalysts depend on their
mation of propylene and COx (per gram of catalyst) reported in reducible nature. It was found that for a given metal oxide (e.g.
Fig. 19(b). It is shown in Fig. 19(b) that the initial rate of formation VOx, Mo W, or Nb), propane turnover rates increased as the reduc-
of propylene (per gram of catalyst) was almost independent of VOx tion rate of the oxide catalyst using H2 increased.
 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140 137

90 50

(a) o
475 C
500 oC
80
(b)
40
10 525 oC 70
o
550 C C3H6
C3H8 Conversion (%)

C3H6 Selectivity (%)

COx Selectivity (%)

60
30
50

40 COx
5 20
30

20 10

10

0 0 0
0 5 10 15 20 0 5 10 15 20
Reaction Time (s) Reaction Time (s)

15 80 60

(c) 70 (d) 50
60
C3H6
C3H8 Conversion (%)

C3H6 Selectivity (%)

COx Selectivity (%)

10 40
50

COx
40 30

30
5 20
20
10
10

0 0 0
0 5 10 15 20 0 5 10 15 20
Reaction Time (s) Reaction Time (s)

70
70

15 (e) (f) 60
60
C3H6
C3H8 Conversion (%)

50
C3H6 Selectivity (%)

COx Selectivity (%)

50

10 40
40 COx

30
30

5 20 20

10 10

0 0 0
0 5 10 15 20 0 5 10 15 20
Reaction Time (s) Reaction Time (s)

Fig. 18. Influence of contact time on propane conversion, propylene and COx selectivities over: (a and b) 5%V, (c and d) 7% V and (e and f) 10%V catalyst samples. (Operating
conditions: T = 550 °C, C3H8 injected = 10 ml, catalyst loaded = 0.76 g). Data points are the average values of 10 successive injections excluding the first injection on a fully
oxidized (fresh) catalyst). Error bars correspond to standard deviation of three repeats.

The results obtained in the present study with regard to the VOx species, in the terminal (V@O) bond in the isolated (mono-
reactivity-structure of the VOx/c-Al2O3 catalysts for propane ODH meric) VO4 surface species or in the bridging (VAO-support) bond
are in good agreement with several previously reported studies available at a given VOx surface density on a specific type of
for supported vanadium oxide catalysts in propane ODH support.
[48,119,120]. In all these studies, it was found that the specific pro- In summary, it is proven in the present study that in the
pane reaction rate (expressed in turnover frequency TOF) in ODH absence of gas phase molecular oxygen, propane could be con-
augments with increasing VOx surface density. This indicated that verted with high selectivity toward propylene via the contribution
the polyvanadate phase is more active than the monovanadate of lattice oxygen. Such high propylene selectivity is achieved via
phase. This variation in ODH catalytic activity with VOx surface the design of a catalyst with carefully selected properties such as
density was attributed to the different contributions of lattice oxy- acidity, reducibility and balance of active phases (i.e. polyvanadate
gen atoms in the bridging (VAOAV) bonds in the polymeric surface and monovanadate phases). With this being accomplished and
138 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140

4.5x10-3 (a) A new VOx/c-Al2O3 catalyst for propane ODH with carefully
475 oC
selected properties such as acidity, reducibility and balance
4.0x10-3 o
500 C of active phases (i.e. polyvanadate and monovanadate
o
525 C
phases) was developed. This catalyst was characterized
Reaction Rate (mol/gcat.s)

o
550 C
3.5x10-3
using BET surface area analysis and showed a moderate
decrease in the total surface area of the catalyst after vana-
3.0x10-3
dium loading on the calcined c-Al2O3 support.
(b) It was shown that the VOx surface structure, reducibility and
2.5x10-3
acidity of the developed catalyst depend on vanadium load-
2.0x10-3
ing. Monomeric VOx surface species were a dominant species
at low vanadium loadings while polymeric VOx species were
1.5x10-3 formed at higher loadings. In addition, it was noticed that
(a) the polymeric surface VOx species were more reducible
1.0x10-3 and displayed a higher abundance of Brønsted acidity than
2 3 4 5 6 7 8 9 the isolated VOx species.
VOx Surface Density (V/nm2) (c) It was proven that the prepared catalyst VOx/c-Al2O3
displayed good propane conversions (11.73–15.11%)
0.6 with promising propylene selectivity (67.65–85.89%) at
475–550 °C. This was demonstrated in a CREC fluidized-bed
0.5
Riser Simulator under consecutive propane injections. It
was found that recently regenerated VOx/c-Al2O3 catalysts
Formation Rate (mol/gcat.s)

were active, however, not selective for propane ODH conver-

0.4
sion. On the other hand, it was proven that the VOx/c-Al2O3
catalysts following the second and subsequent propane injec-
0.3
tions yielded steady high propylene selectivities in ODH. This
promising propylene selectivity was attained under an oxy-
0.2 C3H6 gen free atmosphere with a controlled degree of catalyst
reduction via the contribution of VOx/c-Al2O3 lattice oxygen.
COx
0.1

(b)
0.0
2 3
VOx Surface Density (V/nm2)
Fig. 19. Effect of VOx Surface density on: (a) propane ODH (per gram of catalyst)
and (b) propylene and COx formation rates (per g-catalyst) over VOx/c-Al2O3
catalysts. (Operating conditions: reaction time = 20 s, C3H8 injected = 10 ml, cata-
lyst loaded = 0.76 g). Error bars correspond to standard deviation of three repeats.
Acknowledgements
4 5 6 7 8
The authors wish to acknowledge the National Sciences and
Engineering Research Council of Canada (NSERC) for their financial
support of this research and the Saudi Arabian Oil Company (Saudi
ARAMCO), Saudi Arabia, for the scholarship awarded for the devel-
opment of this Ph.D. research at the University of Western Ontario.

References
90 100
[1] Blasco T, López-Nieto JM. Oxidative dehydrogenation of short chain alkanes
90
85 475 oC on supported vanadium oxide catalysts. Appl Catal A Gen 1997;157:117–42.
500 oC 80 [2] Mamedov EA, Corberan VC. Oxidative dehydrogenation of lower alkanes on
80 525 oC vanadium oxide-based catalysts. The present state of the art and outlooks.
550 oC
C3H6 Selectivity (%)

COx Selectivity (%)

70 Appl Catal A Gen 1995;127:1–40.

75 [3] Baerns M, Buyevskaya O, Mulla SAR. A comparative study on non-catalytic
60
and catalytic oxidative dehydrogenation of ethane to ethylene. Appl Catal A
C3H6
70 50 Gen 2002;226:73–8.
[4] Shen Z, Liu J, Xu H, Yue Y, Hua W, Shen W. Dehydrogenation of ethane to
40 ethylene over a highly efficient Ga2O3/HZSM-5 catalyst in the presence of
65
COx
30 CO2. Appl Catal A Gen 2009;356:148–53.
60 [5] Čapek L, Bulánek R, Adam J, Smoláková L, Sheng-Yang H, Čičmanec P.
20 Oxidative dehydrogenation of ethane over vanadium-based hexagonal
55 mesoporous silica catalysts. Catal Today 2009;141:282–7.
10
[6] Shi X, Ji S, Wang K. Oxidative dehydrogenation of ethane to ethylene with
50 0 carbon dioxide over Cr–Ce/SBA-15 catalysts. Catal Lett 2008;125:331–9.
2 3 4 5 6 7 8 [7] Čapek L, Adam J, Grygar T, Bulánek R, Vradman L, Košová-Kučerová G, et al.
Oxidative dehydrogenation of ethane over vanadium supported on
VOx Surface Density (V/nm2)
mesoporous materials of M41S family. Appl Catal A Gen 2008;342:99–106.
[8] Klose F, Wolff T, Lorenz H, Seidelmorgenstern A, Suchorski Y, Piorkowska M,
Fig. 20. Dependence of propylene and COx selectivities on VOx surface density for et al. Active species on c-alumina-supported vanadia catalysts: nature and
VOx/c-Al2O3 catalysts. reducibility. J Catal 2007;247:176–93.
[9] Danica B, Desislava A, Mirko P, Stefan H. An experimental study of the partial
based on the promising propane ODH results obtained in the pres- oxidation of ethane to ethylene in a shallow fluidized bed reactor. J Serbian
Chem Soc 2007;72:183–92.
ent study in the CREC Fluidized-Bed Riser Simulator, industrial
[10] Haddad N, Bordes Richard E, Hilaire L, Barama A. Oxidative dehydrogenation
scale propane ODH application and scale up can be envisioned. of ethane to ethene on alumina-supported molybdenum-based catalysts
modified by vanadium and phosphorus. Catal Today 2007;126:256–63.
[11] Karamullaoglu G, Dogu T. Oxidative dehydrogenation of ethane over
4. Conclusions chromium–vanadium mixed oxide and chromium oxide catalysts. Ind Eng
Chem Res 2007;46:7079–86.
[12] Tope B, Zhu Y, Lercher JA. Oxidative dehydrogenation of ethane over Dy2O3/
On the basis of the data and results obtained in the present MgO supported LiCl containing eutectic chloride catalysts. Catal Today
study, the following key conclusions can be reported: 2007;123:113–21.
 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140
[13] Cavani F, Ballarini N, Cericola A. Oxidative dehydrogenation of ethane and
propane: how far from commercial implementation? Catal Today
2007;127:113–31.
[14] Lemonidou AA, Heracleous E. Reaction pathways of ethane oxidative and
non-oxidative dehydrogenation on c-Al2O3 studied by temperature-
programmed reaction (TP-reaction). Catal Today 2006;112:23–7.
[15] Solsona B, Dejoz A, Garcia T, Concepcion P, López-Nieto JM, Vazquez M, et al.
Molybdenum–vanadium supported on mesoporous alumina catalysts for the
oxidative dehydrogenation of ethane. Catal Today 2006;117:228–33.
[16] Heracleous E, Lemonidou AA. Ni–Nb–O mixed oxides as highly active and
selective catalysts for ethene production via ethane oxidative
dehydrogenation. Part II: Mechanistic aspects and kinetic modeling. J Catal
2006;237:175–89.
[17] Heracleous E, Lemonidou AA. Ni–Nb–O mixed oxides as highly active and
selective catalysts for ethene production via ethane oxidative
dehydrogenation. Part I: Characterization and catalytic performance. J Catal
2006;237:162–74.
[18] Botella P, Dejoz A, López-Nieto JM, Concepcion P, Vazquez M. Selective
oxidative dehydrogenation of ethane over MoVSbO mixed oxide catalysts.
Appl Catal A Gen 2006;298:16–23.
[19] Nakagawa K, Miyake T, Konishi T, Suzuki T. Oxidative dehydrogenation of
ethane to ethylene over NiO loaded on high surface area MgO. J Mol Catal A
Chem 2006;260:144–51.
[20] López-Nieto JM, Botella P, García-González E, Dejoz A, Vazquez M, González-
Calbet J. Selective oxidative dehydrogenation of ethane on MoVTeNbO mixed
metal oxide catalysts. J Catal 2004;225:428–38.
[21] Anon R. New propylene production technologies hold great promise. Oil Gas J
2004;102:50–2.
[22] Bhasin M. Is true ethane oxydehydrogenation feasible? Top Catal
2003;23:145–9.
[23] Resasco D, Haller G. Catalytic dehydrogenation of lower alkanes. Catal R Soc
Chem 1994:11.
[24] Chan KYG, Inal F, Senkan S. Suppression of coke formation in the steam
cracking of alkanes: ethane and propane. Ind Eng Chem Res 1998;37:
901–7.
[25] Dharia D, Letzsch W, Kim H, McCue D, Chapin L. Increase light olefins
production. Hydrocarb Process 2004;83:61–6.
[26] Ren T, Patel M, Blok K. Olefins from conventional and heavy feedstocks:
energy use in steam cracking and alternative processes. Energy
2006;31:425–51.
[27] Farrauto RJ, Bartholomew CH. Fundamentals of industrial catalytic processes.
1st ed. London: Chapman & Hall; 1997.
[28] Kondratenko EV, Steinfeldt N, Baerns M. Transient and steady state
investigation of selective and non-selective reaction pathways in the
oxidative dehydrogenation of propane over supported vanadia catalysts.
Phys Chem Chem Phys 2006;8:1624–33.
[29] Chen K, Xie S, Bell AT, Iglesia E. Alkali effects on molybdenum oxide catalysts
for the oxidative dehydrogenation of propane. J Catal 2000;195:244–52.
[30] Kung H. Oxidative dehydrogenation of light (C2 to C4) alkanes. Adv Catal
1994;40:1–38.
[31] Chen K, Khodakov A, Yang J, Bell AT, Iglesia E. Isotopic tracer and kinetic
studies of oxidative dehydrogenation pathways on vanadium oxide catalysts.
J Catal 1999;333:325–33.
[32] Bell AT, Khodakov A, Olthof B, Iglesia E. Structure and catalytic properties of
supported vanadium oxides: support effects on oxidative dehydrogenation
reactions. J Catal 1999;181:205–16.
[33] Chen K, Xie S, Bell AT, Iglesia E. Structure and properties of oxidative
dehydrogenation catalysts based on MoO3/Al2O3. J Catal 2001;198:232–42.
[34] Fukudome K, Ikenaga N, Miyake T, Suzuki T. Oxidative dehydrogenation of
propane using lattice oxygen of vanadium oxides on silica. Catal Sci Technol
2011:987–98.
[35] Al-Zahrani SM, Abasaeed AE, Putra M. Kinetics of oxidehydrogenation of
propane over alumina-supported Sr–V–Mo catalysts. Catal Commun
2012;26:98–102.
[36] Lin X, Poeppelmeier KR, Weitz E. Oxidative dehydrogenation of ethane with
oxygen catalyzed by K–Y zeolite supported first-row transition metals. Appl
Catal A Gen 2012;381:114–20.
[37] Piumetti M, Bonell B, Massiani P, Millot Y, Dzwigaj S, Gaberova L, et al. Novel
vanadium-containing mesocellular foams (V-MCF) obtained by direct
synthesis. Microporous Mesoporous Mater 2011;142:45–54.
[38] Piumetti M, Bonell B, Massiani P, Dzwigaj S, Rossetti I, Casale S, et al. Effect of
vanadium dispersion and of support properties on the catalytic activity of V-
containing silicas. Catal Today 2012;179:140–8.
[39] Rubio O, Herguido J, Menendez M. Oxidative dehydrogenation of n-butane on
V/MgO catalysts-kinetic study in anaerobic conditions. Chem Eng Sci
2003;58:4619–27.
[40] Soler J, López-Nieto JM, Herguido J, Menendez M, Santamaria J. Oxidative
dehydrogenation of n-butane in a two-zone fluidized-bed reactor. Ind Eng
Chem Res 1999;38:90–7.
[41] Rubio O, Herguido J, Menendez M. Oxidative dehydrogenation of butane in an
interconnected fluidized-bed reactor. AIChE J 2004;50:1510–22.
[42] Lemonidou AA. Oxidative dehydrogenation of C4 hydrocarbons over VMgO
catalyst—kinetic investigations. Appl Catal A Gen 2001;216:277–84.
[43] Al-Ghamdi SA, Hossain MM, de Lasa HI. Kinetic modeling of ethane oxidative
dehydrogenation over VOx/Al2O3 catalyst in a fluidized-bed riser simulator.
Ind Eng Chem Res 2013;52:5235–44.
139
[44] Al-Ghamdi SA, Volpe M, Hossain MM, de Lasa HI. VOx/c-Al2O3 catalyst for
oxidative dehydrogenation of ethane to ethylene: desorption kinetics and
catalytic activity. Appl Catal A Gen 2013;450:120–30.
[45] Bañares MA, Khatib SJ. Structure-activity relationships in alumina-supported
molybdena–vanadia catalysts for propane oxidative dehydrogenation. Catal
Today 2004;96:251–7.
[46] Sugiyama S, Sugimoto N, Hirata Y, Nakagawa K, Sotowa K. Oxidative
dehydrogenation of propane on vanadate catalysts supported on various
metal hydroxyaptites. Phosphorus Res Bull 2008;22:13–6.
[47] Ballarini N, Cavani F, Cericola A, Cortelli C, Ferrari ML, Capannelli G, et al.
Supported vanadium oxide-based catalysts for the oxidehydrogenation of
propane under cyclic conditions. Catal Today 2004;92:99–104.
[48] Bell AT, Iglesia E, Argyle MD, Chen K. Effect of catalyst structure on oxidative
dehydrogenation of ethane and propane on alumina-supported vanadia. J
Catal 2002;208:139–49.
[49] López-Nieto JM, Soler J, Concepcion P, Herguido J, Menendez M, Santamaria J.
Oxidative dehydrogenation of alkanes over V-based catalysts: influence of
redox properties on catalytic performance. J Catal 1999;332:324–32.
[50] Lemonidou AA, Nalbandian L, Vasalos IA. Oxidative dehydrogenation of
propane over vanadium oxide based catalysts effect of support and alkali
promoter. Catal Today 2000;61:333–41.
[51] Grzybowska B, Klisinska A, Samson K, Gressel I. Effect of additives on
properties of V2O5/SiO2 and V2O5/MgO catalysts: I. Oxidative
dehydrogenation of propane and ethane. Appl Catal A Gen 2006;309:10–6.
[52] Solsona B, López-Nieto JM, Pena ML, Rey F, Vidal-Moya A, Blasco T. Vanadium
oxide supported on mesoporous MCM-41 as selective catalysts in the
oxidative dehydrogenation of alkanes. J Catal 2001;203:443–52.
[53] Dinse A, Frank B, Hess C, Habel D, Schomäcker R. Oxidative dehydrogenation
of propane over low-loaded vanadia catalysts: Impact of the support material
on kinetics and selectivity. J Mol Catal A Chem 2008;289:28–37.
[54] Taylor MN, Carley AF, Davies TE, Taylor SH. The oxidative dehydrogenation of
propane using vanadium oxide supported on nanocrystalline ceria. Top Catal
2009;52:1660–8.
[55] Routray K, Reddy KRS, Deo G. Oxidative dehydrogenation of propane on V2O5/
Al2O3 and V2O5/TiO2 catalysts: understanding the effect of support by
parameter estimation. Appl Catal A Gen 2004;265:103–13.
[56] Bell AT, Dinse A, Schomäcker R. The role of lattice oxygen in the oxidative
dehydrogenation of ethane on alumina-supported vanadium oxide. Phys
Chem Chem Phys 2009;11:6119–24.
[57] Martinez-Huerta MV, Gao X, Tian H, Wachs IE, Fierro JLG, Bañares MA.
Oxidative dehydrogenation of ethane to ethylene over alumina-supported
vanadium oxide catalysts: relationship between molecular structures and
chemical reactivity. Catal Today 2006;118:279–87.
[58] Argyle MD. Oxidative dehydrogenation of light alkanes on metal oxide
catalysts. University of California, Berkeley; 2003.
[59] Blasco T, Galli A, López-Nieto JM, Trifiro F. Oxidative dehydrogenation of
ethane and n-butane on VOx/Al2O3 catalysts. J Catal 1997;169:203–11.
[60] Fierro JLG, López-Nieto JM, Kremenic G. Selective oxidation of propene over
supported vanadium oxide catalysts. Appl Catal 1990;61:235–51.
[61] Bañares MA. Supported metal oxide and other catalysts for ethane
conversion: a review. Catal Today 1999;51:319–48.
[62] Wachs IE, Weckhuysen BM. Structure and reactivity of surface vanadium
oxide species on oxide supports. Appl Catal A Gen 1997;157:67–90.
[63] Bozon-Verduraz F, Enache DI, Bordes Richard E, Ensuque A. Vanadium oxide
catalysts supported on zirconia and titania I. Preparation and
characterization. Appl Catal A Gen 2004;278:93–102.
[64] Wu Z, Kim H, Stair PC, Rugmini S, Jackson SD. On the structure of vanadium
oxide supported on aluminas: UV and visible Raman spectroscopy, UV–
visible diffuse reflectance spectroscopy, and temperature-programmed
reduction studies. J Phys Chem B 2005:2793–800.
[65] Reddy EP, Varma RS. Preparation, characterization, and activity of Al2O3-
supported V2O5 catalysts. J Catal 2004;221:93–101.
[66] Sham E, Murgia V, Gottifredi JC, Torres EMF. Oxidative dehydrogenation of
propane and n butane over alumina supported vanadium catalysts. Lat Am
Appl Res 2004;34:75–82.
[67] Arena F, Frusteri F, Parmaliana A. Structure and dispersion of supported-
vanadia catalysts. Influence of the oxide carrier. Appl Catal A Gen
1999;176:189.
[68] Auroux A, Le Bars J, Forissier M, Vedrine JC. Active sites of V2O5-c-Al2O3
catalysts in the oxidative dehydrogenation of ethane. J Catal 1996;162:250–9.
[69] Yoshida S, Tanaka T, Nishima Y, Mizutani H. The local structures of vanadium
oxide on silica and c-alumina studied by X-ray absorption (XANES-EXAFS)
spectroscopy – the effect of hydration. 9th Int congr catal; 1988. p. 1473.
[70] Corma A, López-Nieto JM, Paredes N, Perez M, Shen Y, Cao H, et al. Oxidative
dehydrogenation of propane over supported-vanadium oxide catalysts. Stud
Surf Sci Catal 1992:213–20.
[71] Chen K, Bell AT, Iglesia E. The relationship between the electronic and redox
properties of dispersed metal oxides and their turnover rates in oxidative
dehydrogenation reactions. J Catal 2002;209:35–42.
[72] Roozeboom F, Mittelmeijer-Hazeleger MC, Moulijn JA, Medema J, Beer VHJ
De, Gellings PJ. Vanadium oxide monolayer catalysts. 3. A Raman
spectroscopic and temperature-programmed reduction study of monolayer
and crystal type vanadia on various supports. J Phys Chem 1980;84:2783–91.
[73] López-Nieto JM. The selective oxidative activation of light alkanes. From
supported vanadia to multicomponent bulk V-containing catalysts. Top Catal
2006;41:3–15.
140 S.A. Al-Ghamdi, H.I. de Lasa / Fuel 128 (2014) 120–140
[74] Wachs IE. Recent conceptual advances in the catalysis science of mixed metal
oxide catalytic materials. Catal Today 2005;100:79–94.
[75] Ishida S, Imamura S, Fujimura Y. Acid–base properties of supported vanadium
oxide catalyst. React Kinet Catal Lett 1991;43:453–9.
[76] Blasco T, López-Nieto JM, Dejoz A, Vazquez M. Influence of the acid–base
character of supported vanadium catalysts on their catalytic properties for
the oxidative dehydrogenation of n-butane. J Catal 1995;157:27–282.
[77] Shen J, Zou H, Li M. Surface acidity of supported vanadia catalysts. J Therm
Anal Calorim 2003;72:209–21.
[78] Deo G, Wachs IE. Reactivity of supported vanadium oxide catalysts: the
partial oxidation of methanol. J Catal 1994;146:323–34.
[79] Kung H, Michalakos PM, Kung MC, Jahan I. Selectivity patterns in alkane
oxidation over Mg3(VO4)2–MgO, Mg2V2O7, and (VO)2P2O7. J Catal
1993;140:226–42.
[80] de Lasa HI. Canadian patent 1,284,017 (1991). USA patent 5,102,628 (1992);
1992.
[81] Al-Ghamdi SA. Thesis: oxygen-free propane oxidative dehydrogenation over
vanadium oxide catalysts: reactivity and kinetic modelling. University of
Western Ontario; 2013.
[82] Geldart D. Types of gas fluidization. Powder Technol 1973;7:285–92.
[83] Centi G. Nature of active layer in vanadium oxide supported on titanium
oxide and control of its reactivity in the selective oxidation and
ammoxidation of alkylaromatics. Appl Catal A Gen 1996;147:267–98.
[84] Bosc H, Bert JK, Van Ommen JG, Gellings PJ. Factors influencing the
temperature-programmed reduction profiles of vanadium pentoxide. J
Chem Soc Faraday Trans 1984;80:2479–88.
[85] Koranne MM, Goodwin JG, Marcelin G. Characterization of silica- and
alumina-supported vanadia catalysts using temperature programmed
reduction. J Catal 1994;148:369–77.
[86] Muhler M, Besselmann S, Freitag C, Hinrichsen O. Temperature-programmed
reduction and oxidation experiments with V2O5/TiO2 catalyst. Phys Chem
Chem Phys 2001;3:4633–8.
[87] Van Vuuren CP, Stander F. The reduction of V2O5. Part II. Reduction by
hydrogen. Thermochim Acta 1990;165:85–91.
[88] Zou H, Li M, Shen J, Auroux A. Surface acidity of supported vanadium
catalysts. J Therm Anal Calorim 2003;72:209–21.
[89] Kantcheva MM, Hadjiivanov KI, Klissurski DG. An IR spectroscopy study of the
state and localization of vanadium-oxo species adsorbed on TiO2 (anatase). J
Catal 1992;134:299–310.
[90] Turek AM, Wachs IE. Acidic properties of alumina-supported metal
oxide catalysts: an infrared spectroscopy study. J Phys Chem 1992;96:
5000–7.
[91] Busca G, Ramis G, Lorenzelli V. FT-IR study of the surface properties of
polycrystalline vanadia. J Mol Catal 1989;50:231–40.
[92] Khader MM. Surface acidity of V2O5/Al2O3 catalysts: IR and TPD studies. J Mol
Catal A Chem 1995;104:87–94.
[93] Le Bars J, Vedrine JC, Auroux A. Role of surface acidity on vanadia/silica
catalysts used in the oxidative dehydrogenation of ethane. Appl Catal A Gen
1992;88:179–95.
[94] Datka J, Turek AM, Jheng M, Wachs IE. Acidic properties of supported niobium
oxide catalysts: an infrared spectroscopy investigation. J Catal 1992;135:
186–99.
[95] Miyata H, Fujii K, Ono T. Acidic properties of vanadium oxide on titania. J
Chem Soc Faraday Trans 1 1988;84:3121–8.
[96] Tonetto G, Atias J, de Lasa HI. FCC catalysts with different zeolite crystallite
sizes: acidity, structural properties and reactivity. Appl Catal A Gen
2004;270:9–25.
[97] Chester AW, Derouane EG. Zeolite characterization and catalysis. 5th ed. New
York: Springer; 2009.
[98] Wachs IE. Raman and IR studies of surface metal oxide species on oxide
supports: supported metal oxide catalysts. Catal Today 1996;27:437–55.
[99] Zhao B, Xu X, Ma H, Sun D, Gao J. Monolayer dispersion of oxides and salts on
surface of ZrO2 and its application in preparation of ZrO2-supported catalysts
with high surface areas. Catal Lett 1997;45:237–44.
[100] Haber J, Witko M. Quantum-chemical modelling of hydrocarbon oxidation on
vanadium-based catalysts. Catal Today 1995;23:311–6.
[101] Arnold EW, Sundaresan S. The role of lattice oxygen in the dynamic behaviour
of oxide catalysts. Chem Eng Commun 2007:37–41.
[102] Genser F, Pietrzyk S. Oxidative dehydrogenation of propane on V2O5/TiO2
catalysts under transient conditions. Chem Eng Sci 1999;54:4315–25.
[103] Crapanzano S, Babich I, Lefferts L. Selection of mixed conducting oxides for
oxidative dehydrogenation of propane with pulse experiments. Appl Catal A
Gen 2011;391:70–7.
[104] Balcaen V, Sack I, Olea M, Marin GB. Transient kinetic modeling of the
oxidative dehydrogenation of propane over a vanadia-based catalyst in the
absence of O2. Appl Catal A Gen 2009;371:31–42.
[105] Che M, Tench AJ. Characterization and reactivity of molecular oxygen species
on oxide surfaces. Adv Catal 1982:1–148.
[106] Che M, Tench AJ. Characterization and reactivity of mononuclear oxygen
species on oxide surfaces. Adv Catal 1982;31:77–133.
[107] Libre JM, Barbaux Y, Grzybowska B, Bonnelle JP. A surface potential study of
adsorbed oxygen species on a Bi2Mo3O12 catalyst. React Kinet Catal Lett
1982;20:249–54.
[108] Kung H, Andersen PJ. The effect of oxygen binding energy on the selective
oxidation of butane over V/c-Al2O3. Stud Surf Sci Catal 1993;75:205–17.
[109] Dinse A, Schomäcker R, Bell AT. The role of lattice oxygen in the oxidative
dehydrogenation of ethane on alumina-supported vanadium oxide. Phys
Chem Chem Phys 2009;11:6119–24.
[110] Owen OS, Kung MC, Kung H. The effect of oxide structure and cation
reduction potential of vanadates on the selective oxidative dehydrogenation
of butane and propane. Catal Lett 1992;12:45–50.
[111] Creaser D, Andersson B, Hudgins RR, Silverston PL. Transient kinetic analysis
of the oxidative dehydrogenation of propane. J Catal 1999;182:264–9.
[112] Creaser D, Andersson B, Hudgins RR, Silverston PL. Transient study of
oxidative dehydrogenation of propane. Appl Catal A Gen 1999;187:147–60.
[113] Yoon YS, Ueda W, Moro-oka Y. Oxidative dehydrogenation of propane over
magnesium molybdate catalysts. Catal Lett 1995;35:57–64.
[114] Balcaen V, Roelant R, Poelman H, Poelman D, Marin GB. TAP study on the
active oxygen species in the total oxidation of propane over a CuO–CeO2/
c-Al2O3 catalyst. Catal Today 2010;157:49–54.
[115] Balcaen V, Poelman H, Poelman D, Marin GB. Kinetic modeling of the total
oxidation of propane over Cu- and Ce-based catalysts. J Catal
2011;283:75–88.
[116] Mattos ARJ, da Silva San Gil RA, Rocco MLM, Eon JG. Zinc-modified, alumina-
supported vanadium oxides as catalysts for propane oxidative
dehydrogenation. J Mol Catal A Chem 2002;178:229–37.
[117] Heracleous E, Machli M, Lemonidou AA, Vasalos IA. Oxidative
dehydrogenation of ethane and propane over vanadia and molybdena
supported catalysts. J Mol Catal A Chem 2005;232:29–39.
[118] Schwarz O, Frank B, Schomäcker R. Characterisation and catalytic testing of
VOx/Al2O3 catalysts for microstructured reactors. Catal Commun
2008;9:229–33.
[119] Khodakov A, Olthof B, Bell AT, Iglesia E. Structure and catalytic properties of
supported vanadium oxides: support effects on oxidative dehydrogenation
reactions. J Catal 1999;181:205–16.
[120] Tian H, Ross EI, Wachs IE. Quantitative determination of the speciation of
surface vanadium oxides and their catalytic activity. J Phys Chem B
2006;110:9593–600.