See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/269811476

Kinetic modelling of n -butane dehydrogenation over

CrO<SUB>x</SUB>VO<SUB>x</SUB>/MCM-41 catalyst in
a fixed bed reactor

Article  in  Progress in Reaction Kinetics and Mechanism · December 2014

DOI: 10.3184/146867814X14119972226885

CITATIONS READS

0 439

4 authors:

Babajide Patrick Ajayi Basim Abussaud

University of Louisville King Fahd University of Petroleum and Minerals
12 PUBLICATIONS   65 CITATIONS    30 PUBLICATIONS   208 CITATIONS   

SEE PROFILE SEE PROFILE

Rabindran Jermy Sulaiman al-khattaf

King Fahd University of Petroleum and Minerals King Fahd University of Petroleum and Minerals
48 PUBLICATIONS   825 CITATIONS    153 PUBLICATIONS   1,837 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Butane dehydrogenation View project

Oxidative dehydrogenation of lower alkanes View project

All content following this page was uploaded by Babajide Patrick Ajayi on 01 September 2015.

The user has requested enhancement of the downloaded file.

 Progress in Reaction Kinetics and Mechanism, 2014, 39(4), 341 – 353
doi:10.3184/146867814X14119972226885
RESEARCH PAPER
Kinetic modelling of n‑butane dehydrogenation
over CrOxVOx/MCM‑41 catalyst in a
fixed bed reactor
Babajide Patrick Ajayia*, Basim Abussaudb, Rabindran Jermyb
and Sulaiman Al Khattaf b
a
Chemical Engineering Department, University of Louisville, Louisville KY 40292, USA
b
Center for Refining & Petrochemicals, King Fahd University of Petroleum & Minerals,
Dhahran 31261, Saudi Arabia

*E-mail: bpajay01@louisville.edu

ABSTRACT
The kinetics of n‑butane dehydrogenation over CrOxVOx/MCM‑41 catalyst was
studied. The models were developed based on the catalyst tests carried out in
a packed bed reactor at reaction temperatures varied from 525 to 575 °C under
atmospheric pressure. The dehydrogenation of n‑butane mainly gives butene
isomers (over 90%), 1,3-butadiene and cracked products consisting of methane,
ethane, ethene and propene. Based on the experimental observations, power
law type models were formulated and parameters were estimated by fitting the
experimental data implemented in MATLAB. The activation energy for the
formation of butenes (96.2 kJ mol – 1) was found to be considerably less than
the activation energy for the formation of the undesirable cracked products
(130.4 kJ mol – 1).

KEYWORDS: n‑butane, dehydrogenation, kinetic modelling, cracking,

MCM-41, chromium, vanadium, power law, activation energy

1. INTRODUCTION
The importance of catalytic dehydrogenation of paraffin hydrocarbons to olefins
has been growing steadily in recent years. Light olefins, such as butenes, are
important raw materials for the synthesis of polymers, gasoline additives and
various other petrochemical products. The cracking processes, especially fluid
catalytic cracking and steam cracking cannot match the rising demand for the
341

PRK1300246_FINAL.indd 341 25/11/2014 11:04:45

 342 B.P. Ajayi, B. Abussaud, R. Jermy and S. Al Khattaf

product, hence the need for catalytic dehydrogenation processes. Over the past
several decades, the role of different metal oxides on different supports like
ZrO2, SiO2, and zeolites has been explored for dehydrogenation [1,2]. Among
the oxide catalysts, vanadium and chromium-oxide supported catalysts have
been most promising for dehydrogenation due to their high catalytic activity.
Hakuli et al. [3] reported that silica- and alumina-supported chromium
oxides were effective for the production of lower alkenes such as propene
and isobutene. Botavina et al. [4] stated that for oxidative dehydrogenation
of propane and isobutane, the highest activity, selectivity and stability were
achieved over 5.0 wt% CrOx/silica catalysts prepared by the wet impregnation
method. Volpe et al. [5] investigated n‑butane dehydrogenation using VOx
supported on USY, NaY, γ‑Al2O3 and α‑Al2O3. The high catalytic activity of
VOx/USY and butane selectivity was related to the VOx monolayer and its mild
acidity. Highly dispersed VOx/SBA 15 mesoporous silica has been found to
effect a highly efficient oxidative dehydrogenation of propane to light olefins
(propene + ethylene). The olefin selectivity was observed up to 80% at a high
propane conversion of 41.7%. The enhanced propene selectivities at higher
propane conversions have been attributed to the unique large pore diameters
and the low surface acidity of the SBA‑15 materials [6]. Chromium oxide
catalysts supported on high surface area, pore-expanded MCM‑41 silica was
investigated for ethane dehydrogenation; SBA‑15 and MCM‑41 are silica-
based molecular sieves with a wide application in selective oxidation reactions.
The study showed that the ethane conversion strongly depends on the chromia
loading and the catalyst preparation method [7]. Similarly, VOx species
supported on hexagonal mesoporous silica (HMS) were found to be active in
ODH (oxidative dehydrogenation) of n‑butane, where the isolated monomeric
VOx species were determined to play most active catalytic role [8].
For a new catalytic system such as CrOxVOx/MCM-41, a kinetic model
elucidating the rate mechanistic steps is very essential to the design of a reactor
and to predict reactor performance confidently. The literature is replete with
kinetic studies on monometal oxide (CrOx or VOx) systems, in the present
study, however we aim to develop a model that will predict the rate law and
mechanistic steps by utilising a CrOxVOx bimetallic system using the power
law model. The mechanistic models considered were established based on a
set of extensive experimental data generated using a fixed bed reactor under
atmospheric pressure conditions. The kinetic parameters were estimated by
least square fittings of the model’s ordinary differential equations (ODE) using
experimental data implemented in MATLAB.
www.prkm.co.uk

PRK1300246_FINAL.indd 342 25/11/2014 11:04:45

 Kinetics of n‑butane dehydrogenation over CrOxVOx /MCM‑41 343
2. EXPERIMENTAL

2.1 Catalyst preparation

The method for the preparation of highly ordered MCM‑41 (abbreviated as
M‑41) material and chromium/vanadium impregnated M‑41 catalysts has been
detailed in our earlier work [9].

2.2 Catalysts test

The catalytic runs were carried out in a fully integrated, fixed bed continuous
flow reactor system (BELCAT). The system consisted of one tubular quartz
reactor attached with stainless steel one-zone furnace assembly through the
reactor furnace wall thermo well. The catalyst sample (0.12 g) was packed in the
reactor and activated with flowing N2 at 823 K for 2 h, after which the contact
time was varied from 1.5 to 24 gcat.min/gbutane. The reaction temperatures were
525, 550 and 575 °C respectively. The gas product samples were analysed by
on‑line gas chromatography (Agilent, 7890N) using a J&W Scientific HP 5
capillary column (length: 30 m, ID: 0.32 mm, and film thickness 0.25 μm) and
FID detector. The gas products, including H2, CO and CO2, were analysed with
a TCD and a GS GASPRO capillary column. The products were identified by
comparison with authentic samples.
The term gas hourly space velocity (GHSV) is the inverse of contact time (τ)
and is defined as the mass flow rate of n‑butane in g/min over the weight of
the catalyst. The contact times were varied by changing the partial pressure of
n‑butane in the mixed gas containing n‑butane and nitrogen while keeping the
catalyst weight fixed.

GC-analysed data were further processed to evaluate the conversion and

selectivity using the following equations:

www.prkm.co.uk

PRK1300246_FINAL.indd 343 25/11/2014 11:04:46

 344 B.P. Ajayi, B. Abussaud, R. Jermy and S. Al Khattaf

3. RESULTS AND DISCUSSION

3.1 Physico-chemical properties of catalyst
The X‑ray diffraction patterns for CrOxVOx/M‑41 sample have been extensively
discussed elsewhere [9]. The catalyst showed four well-resolved diffraction
peaks corresponding to hexagonal symmetry indicating a well-ordered
mesoporous support with no detectable crystalline peaks at higher 2 θ range.
The textural characteristics of bimetal impregnated M‑41 samples, which have
also been well detailed in our previous report [9] are summarised in Table 1.
The surface area (1079 m2 g – 1), and pore volume (0.66 cm3 g – 1) calculated from
nitrogen adsorption isotherm were very high, typical of mesoporous materials.

Table 1 Physico-chemical properties of Cr- and V‑supported M‑41 catalyst

Metal content (wt. %) dspacing Surface area Pore volume Pore diameter
Catalyst
Cr V (Å) (m 2 g–1) (cm3 g–1) (nm)
CrOxVOx/M-41 1.2 2.8 36.78 1079 0.66 2.5

3.2 Model development

It is instructive to develop a model with a reasonably low number of reaction
steps and kinetic parameters that will be just adequate to enable us to describe
the essential features of the n‑butane dehydrogenation and cracking reactions.
Rather than develop a detailed model incorporating all GC‑detected chemical
species and possible reaction steps, we simplified the model by lumping all
cracked products consisting of methane, ethane, ethene, propane, propene as
one entity, and the butene isomers; namely 1‑butene, cis-2-butene, trans-2-
butene were regarded as a single specie called C4 olefin. Also, some reaction
steps with no significant contribution to the overall rate were ignored. Examples
of such cases are Eqns (2) and (3) which is the single and double step biolefin
route respectively. The selectivity to butadiene in the entire sample size was
less than 5%, so the reaction steps leading to butadiene in the system were
neglected. See Table 2 for product distribution.

Dehydrogenation steps:
C4H10 C4H8 + H2 (1)
C4H8 C4H6 + H2 (2)
C4H10 C4H6 + 2H2 (3)

www.prkm.co.uk

PRK1300246_FINAL.indd 344 25/11/2014 11:04:46

 Kinetics of n‑butane dehydrogenation over CrOxVOx /MCM‑41 345
Cracking steps:
C4H10 C2H4 + C2H6 (4)
C4H10 CH4 + C3H6 (5)
C4H8 2C2H4 (6)

Therefore, Scheme 1 was proposed after making the simplifications explained

in the foregoing paragraph:
Further modification was made to Scheme  1 after the model failed to work.
As can be seen by Eqn (4), the stoichiometric proportion of ethene to ethane is
1 : 1 which is expected to increase if the secondary cracking reactions [Eqn (6)]
were to be factored in. However, the product distribution table suggested a
decrease in the ratio, hence secondary monomolecular cracking of olefin was
neglected in the new scheme:

Table 2 Product distribution of n‑butane dehydrogenation over CrOxVOx/M‑41

Contact Yield (wt%)
Time Conv.
(gcatmin/
gbutane) CH4 C2H6 C2H4 C3H8 C3H6 trans 2 nC4H8 iC4H8 cis 2 C4H6 (%)
butene butene
525 °C
1.5 0.02 0.02 0.02 0.06 0.04 0.57 0.46 0.01 0.43 0.04 1.75
3 0.03 0.03 0.03 0.07 0.07 0.89 0.7 0.02 0.67 0.07 2.65
8 0.06 0.05 0.05 0.07 0.14 1.43 1.11 0.09 1.09 0.18 4.36
12 0.1 0.09 0.09 0.12 0.23 1.78 1.4 0.14 1.36 0.27 5.68
16 0.24 0.23 0.15 0.14 0.5 1.77 1.84 0.45 1.55 0.18 7.06
19 0.29 0.28 0.17 0.2 0.57 1.98 2.11 0.58 1.75 0.2 8.18
24 0.35 0.35 0.19 0.25 0.64 2.35 2.72 0.67 2.18 0.25 9.99
550 °C
1.5 0.03 0.02 0.05 0.06 0.09 1.24 1.03 0.05 0.94 0.39 2.95
3 0.06 0.05 0.06 0.07 0.15 1.16 0.96 0.05 0.89 0.14 3.97
8 0.13 0.1 0.13 0.08 0.33 1.4 1.14 0.12 1.07 0.27 4.87
12 0.19 0.17 0.18 0.11 0.48 2.1 1.72 0.21 1.62 0.41 7.38
16 0.58 0.55 0.41 0.26 1.27 2.01 2.23 0.58 1.79 0.33 10
19 0.64 0.6 0.43 0.32 1.33 2.25 2.62 0.81 2.06 0.37 11.52
24 0.83 0.82 0.47 0.46 1.56 2.84 3.61 1.02 2.72 0.46 14.91
575 °C
1.5 0.09 0.08 0.1 0.07 0.24 1.09 0.95 0.05 0.84 0.15 3.74
3 0.14 0.12 0.15 0.09 0.37 1.38 1.2 0.09 1.06 0.22 4.9
8 0.36 0.3 0.37 0.16 0.86 3.14 2.2 0.47 2.22 0.59 10.86
12 0.39 0.29 0.43 0.14 1 1.99 1.72 0.21 1.54 0.65 16.59
16 1.35 1.33 0.84 0.71 2.48 3.39 4.64 1.7 3.35 0.68 20.88
19 1.67 1.51 1.16 0.63 3.29 2.75 3.48 1.09 2.59 0.7 23.12
24 1.72 1.69 0.99 0.9 2.92 3.79 5.07 2.15 3.7 0.77 27.24

www.prkm.co.uk

PRK1300246_FINAL.indd 345 25/11/2014 11:04:46

 346 B.P. Ajayi, B. Abussaud, R. Jermy and S. Al Khattaf

k1
C4H10
k–1 C4 Olefin + H2

k2 k3

Cracked Products
Scheme 1 Catalytic reaction scheme of n‑butane dehydrogenation.

k1
C4H10
C4 Olefin + H2

k–1

k2

Cracked Products
Scheme 2 Modified catalytic reaction scheme of n‑butane dehydrogenation.

The experimental results were modelled using the catalyst deactivation function
model based on the reaction conversion catalyst decay function as developed
by Al Khattaf and de Lasa [10]. In this model, deactivation is dependent on the
conversion of n‑butane which in turn is linked to the coke formed on the surface
of the active sites of the catalysts. This approach appears to be more accurate
than the time-on-stream approach widely used in the literature to account for
catalyst deactivation. Based on the fact that the temperature difference across
the catalyst bed is negligible, the fixed bed was considered to be an isothermal
plug flow reactor. The total reaction rate for each chemical specie i expressed
by ri was evaluated by summing up all the reaction rates at every step j in which
specie i was involved. For n number of reaction steps, the rate is given by:

(7)

From the reaction network shown in Scheme 2 and on the assumption that
butadiene formation is neglected, the following set of mole balances can be
written:
www.prkm.co.uk

PRK1300246_FINAL.indd 346 25/11/2014 11:04:46

 Kinetics of n‑butane dehydrogenation over CrOxVOx /MCM‑41 347
Rate of disappearance of n‑butane, rB

* (8)

Rate of appearance of C4 olefin, rB*

* (9)
* *

Rate of appearance of cracked products, rCP

(10)

Rate of appearance of hydrogen, rH

*
(11)

where Ci is the molar concentration of each of the chemical species, i being B

(n‑butane), B* (C4 olefin), CP (Cracked Products) and H (Hydrogen), and the
contact time, τ is given by:
(12)

from which W is the catalyst weight (0.12 g), F is the total flow rate of the mixed
gas into the reactor comprising n‑butane and nitrogen. W was fixed at 0.12 g
while F was varied keeping the partial pressure of n‑butane constant in each
case. The term φ is the catalyst decay function, which accounts for the loss
of catalytic activity as a result of catalytic deactivation due to coking. For the
reaction conversion model, the catalyst decay function is given by:

(13)

where λ is the catalyst decay constant and XB is the conversion of n‑butane

given by:
(14)

The temperature-dependent equilibrium constant (Kc) for the reversible reaction

was found to be [11]:
www.prkm.co.uk

PRK1300246_FINAL.indd 347 25/11/2014 11:04:47

 348 B.P. Ajayi, B. Abussaud, R. Jermy and S. Al Khattaf

(15)

The activation energy of reaction step j, Ej, is related to the temperature

dependent rate constant kj by the Arrhenius equation:

(16)

koj is the pre exponential factor. Elsewhere [12], it has been suggested that
reparameterisation of the Arrhenius equation will help reduce parameter
interaction during modelling. As a result Eqn (16) was reparameterised by
centring the values around koj which is the value of the rate constant for reaction
step j at the average temperature T0:

(17)

The experimental runs were carried out at 525, 550, 575 °C, and therefore
550 °C was selected as the average temperature, T0.

3.3 Model assumptions:

Below are the premises on which the model was developed.
(1) Mass transfer limitation is negligible.
(2) Catalyst deactivation is a function of n‑butane conversion.
(3) A single deactivation constant is adopted for all reaction steps.
(4) Effectiveness factor is unity.
(5) Reactor operates isothermally.
(6) Butadiene yield is negligible.
(7) Secondary cracking of olefin is neglected.
(8) Model assumes only catalytic conversion.

3.4 Determination of model parameters and model discrimination

The kinetic parameters in Eqns (8) – (11) were estimated using a non-linear
regression analysis. To estimate the parameters and evaluate the models,
individual specific rate equations were combined with mole balance
equations. The ordinary differential equations obtained were then solved
numerically together with a least-squares fitting of the experimental n‑butane
dehydrogenation data. The MATLAB ODE 45 subroutine (Runge Kutta Gill
method) was applied in solving the ODE while the LSQCURVEFIT subroutine
www.prkm.co.uk

PRK1300246_FINAL.indd 348 25/11/2014 11:04:47

 Kinetics of n‑butane dehydrogenation over CrOxVOx /MCM‑41 349
(Modified Marquad method) was used for parameter estimation. To obtain
more reliable model and model parameters, the experiments were conducted at
three different reaction temperatures (525 – 575 °C) and seven different contact
times (1.5 – 24  gcat.min.gbutane). Because there are sufficient data for iteration,
the degrees of freedom are more than is needed, hence we have very accurate
model results.
The model evaluation criteria are that all the rate parameters should be consistent
with physical principles which are the specific reaction and adsorption rates,
the activation energies for reactions and the heat of adsorption being positive.
The optimisation criteria used was based on a minimum sum of squares criteria
defined by:

(18)

Model discrimination was based on:

(9) Correlation coefficient closer to one.
(10) Lower sum of square residuals.
(11) Lower confidence interval for model parameters.
The values of the model parameters with their corresponding 95% confidence
intervals are shown by Table 3 while the cross correlation matrix is as depicted
by Table 4. As can be seen from Table 3, the estimated apparent activation

Table 3 Estimated values of the kinetic parameters at 95% confidence intervals

Reaction conversion model
Parameters R2
E(kJ mol – 1) Pre-exponential factor
k1 96.2 ± 12.2 5.9 ± 0.5 × 10 – 3
0.9994
k2 130.4 ± 16.9 5.0 ± 0.5 × 10 – 3
λ 0.97 ± 0.71

Table 4 Correlation matrix for n‑butane dehydrogenation over CrOxVOx/M‑41 catalyst

k1 E1 k2 E2 λ
k1 1.00 0.29 0.60 0.41 0.88
E1 0.29 1.00 0.50 0.01 0.53
k2 0.60 0.50 1.00 0.07 0.77
E2 0.41 0.01 0.07 1.00 0.38
λ 0.88 0.53 0.77 0.38 1.00

www.prkm.co.uk

PRK1300246_FINAL.indd 349 25/11/2014 11:04:47

 350 B.P. Ajayi, B. Abussaud, R. Jermy and S. Al Khattaf
Table 5 Apparent activation energy of n‑butane dehydrogenation and cracking reactions
reported in selected works
Apparent activation energy (kJ mol – 1)
Author Catalyst
Dehydrogenation Cracking
Lecher et al.14 H ZSM 5 105 140
Gounder et al.15 H100Na0MOR T 177 191
Gounder et al.15 H100Na0MOR Z 194 208
Gounder et al.15 H45Na55MOR Z 186 205
Macgregor et al.13 VOx/Al2O3 132  – 
Airaksinen et al.11 CrOx/Al2O3 137  – 
This study CrOxVOx/M-41 96 130

energies of dehydrogenation and cracking are 90.2 and 130.4 kJ mol – 1

respectively. The apparent activation energy of n‑butane dehydrogenation was
found to be lower than those found in the literature [11,13 – 15]. A similar trend
exists for the apparent activation energy of cracking (see Table 5). It was also
observed that the apparent activation energy for the production of C4 olefin is
much smaller (96.2 kJ mol – 1) than the cracking reactions producing methane,
ethane, ethene and propene (130.4 kJ mol–1). These results were expected since
the CrOxVOx/M‑41 catalyst is more effective for the dehydrogenation to give
C4 olefin than the cracking reaction which requires a high activation energy
to produce the undesired products [9]. In Table 4, the correlation between
E1 – k1 and E1 – λ for all estimated parameters for dehydrogenation reaction
and the correlation between E2 – k2 and E2 – λ for all estimated parameters
for cracking reaction have low-to-medium parameter interaction, however;
there are exceptions. The parameter interactions between k1 – λ and k2 – λ for
dehydrogenation and cracking reactions have been found to be significant.
The yield of butene as seen from Figure 1 was found to increase with both
contact time and temperature, which is also true for n‑butane conversion.
The maximum yields of 17, 13 and 9% were recorded at 575, 550 and 525 °C
respectively. The numerical results are compared with experimental data as
depicted by Figure 2. As shown the model prediction compares favourably with
the experimental data. This is a verification that the proposed kinetic model
is a good fit for the obtained experimental results. To further give credence to
the determined model, a parity plot (Figure 3), showing the overall agreement
of the experimental data and the model predictions, was performed. Also, the
value of the regression coefficient (R2 = 0.9994) suggests a strong correlation
between the experimental results and the model.
www.prkm.co.uk

PRK1300246_FINAL.indd 350 25/11/2014 11:04:47

 Kinetics of n‑butane dehydrogenation over CrOxVOx /MCM‑41 351

°C
°C
°C
°C Pred
°C Pred
°C Pred

Figure 1 C4 Olefin yields against n‑butane conversion at various temperatures for

CrOxVOx/M-41.

°C
°C
°C
°C Pred
°C Pred
°C Pred

Figure 2 Comparison of experimental and model-predicted conversions at different

temperatures.
www.prkm.co.uk

PRK1300246_FINAL.indd 351 25/11/2014 11:04:48

 352 B.P. Ajayi, B. Abussaud, R. Jermy and S. Al Khattaf

Figure 3 Parity plot between the experimental results and model predictions for
CrOxVOx/M-41.

4. CONCLUSIONS
Extensive experimental kinetic data were obtained for n‑butane dehydrogenation
over CrOxVOx /M‑41 catalyst in a fixed bed reactor. A power law model was
adopted for evaluating the kinetic parameters based on the reaction conversion
catalyst decay function. The experimental data showed a reasonable fit for the
model at 95% confidence interval with a correlation coefficient of R2 = 0.9994.
The apparent activation energies of dehydrogenation and cracking are 96.2 and
130.4 kJ mol – 1 respectively. From these results, the catalyst can be said to be a
reliable catalyst for dehydrogenation when compared with some catalysts from
the literature because of its lower apparent activation energy. Also, the catalyst
was observed to be selective towards dehydrogenation rather than cracking
based on its lower apparent activation energy for dehydrogenation over cracking.

5. REFERENCES
[1] Boot, L.A., van Dillen, A.J., Geus, J.W. and van Buren, F.R.(1996) J. Catal., 163,
186 – 194.
[2] Jiménez-López, A., Rodriguez-Castellón, E., Maireles-Torres, P., Diaz, L. and Mérida-
Robles, J. (2001) Appl. Catal. A, 218, 295 – 306.
www.prkm.co.uk

PRK1300246_FINAL.indd 352 25/11/2014 11:04:48

 Kinetics of n‑butane dehydrogenation over CrOxVOx /MCM‑41 353
[3] Hakuli, A., Harlin, M.E., Backman, L.B. and Krause, A.O.I. (1999) J. Catal., 184,
349 – 335.
[4] Botavina, M.A., Martra, G., Yu, A., Agafonov, N.A., Gaidai, N.V., Nekrasov, N.,
Trushin, D.V., Coluccia, S. and Lapidus, A.L. (2008) Appl. Catal. A, 347, 126 – 132.
[5] Volpe, M., Tonetto, G. and de Lasa, H. (2004) Appl. Catal. A, 272, 69 – 78.
[6] Liu, Y.M., Cao,Y., Zhu, K.K., Yan, S.R., Dai, W.L., He, H.Y. and Fan, K.N. (2002)
Chem. Commun., 2832 – 2833.
[7] Malleswara-Rao, T.V., Mamoune Zahidi El and Sayari, A., (2009) J. Mol. Catal. A, 301,
159 – 165.
[8] Setnicka, M., Bulanek, R., Capek, L. and Cicmanec, P. (2011) J. Mol. Catal. A, 344,
1 – 10.
[9] Ajayi, B.P., Rabindran Jermy, B., Ogunronbi, K.E., Abussaud, B.A. and Al Khattaf, S.
(2013) Catal. Today, 204, 189 – 196.
[10] Al Khattaf, S. and de Lasa, H. (2001) Ind. Eng. Chem. Res., 40, 5398 – 5404.
[11] Airaksinen, S.M.K., Harlin, M.E. and. Krause, A.O.I.,(2002) Ind. Eng. Chem. Res., 40,
5619 – 5626.
[12] Agarwal, A.K. and Brisk, M.L.(1985) Ind. Eng. Chem. Proc. Des. Dev., 24, 203 – 207.
[13] McGregor, J., Huanga, Z., Parrotta, E.P.J., Zeitlera, B.J.A., Nguyena, K.L., Rawsonc,
J.M., Carleye, A., Hansenf, T.W., Tessonnierf, J.P., Suf, D.S., Teschnerf, D., Vassf, E.M.,
Knop Gerickef, A., Schlöglf, R. and Gladdena, L.F. (2010) J. Catal., 269, 329 – 339.
[14] Lercher, J.A., van Santen, R.A. and Vinek, H. (1994) Catal. Lett., 269, 91 – 96.
[15] Gounder, R. and Iglesia, E. (2010) Angew. Chem. Int. Ed., 49, 808 – 811.

www.prkm.co.uk

PRK1300246_FINAL.indd
View publication stats 353 25/11/2014 11:04:48