Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The initial steps

Article abstract of DOI:10.1016/j.cattod.2015.07.006

In this work, a 2 wt.% Ni/Mg(Al)O catalyst was subjected to kinetic studies for the ethanol steam reforming (ESR) reaction at 500 °C, with space-time ranging from 0.03 to 0.50 mg min/ml and with P_C2H5OH:P_H2O:P_Inert = 0.0163:0.0500:0.93 (atm) as standard conditions. The results indicate that dehydrogenation and dehydration of ethanol were the predominant reactions. CH₃CHO and C₂H₄ were primary products and formed in parallel on different active sites. H₂O and C₂H₅OH competed for the same active sites on the catalyst surface and the reaction order with respect to H₂O was negative. The apparent activation energy for ethanol conversion was 110 kJ/mol. Furthermore, temperature-programmed desorption experiments confirmed the competing adsorption of C₂H₅OH and H₂O. Temperature-programmed deuteration of used catalyst showed that the catalyst contained C₂H₄, CH_x, acetate and carbonate species during ESR reaction.

Full text of DOI:10.1016/j.cattod.2015.07.006

G Model
CATTOD-9668; No. of Pages11
ARTICLE IN PRESS
Catalysis Today xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Kinetic and process study of ethanol steam reforming over
Ni/Mg(Al)O catalysts: The initial steps
Guangming Zenga,b, Yongdan Lia, Unni Olsbyeb,∗
a State Key Laboratory for Chemical Engineering (Tianjin University), School of Chemical Engineering and Technology, Tianjin University, 300072 Tianjin,
China
b inGAP Centre of Research-Based Innovation, Department of Chemistry, University of Oslo, 0315 Oslo, Norway
a r t i c l e i n f o
a b s t r a c t
Article history:
In this work, a 2 wt.% Ni/Mg(Al)O catalyst was subjected to kinetic studies for the ethanol steam
reforming (ESR) reaction at 500 ◦C, with space-time ranging from 0.03 to 0.50 mg min/ml and with
PC2H5OH:PH2O:PInert = 0.0163:0.0500:0.93 (atm) as standard conditions. The results indicate that dehydro-
genation and dehydration of ethanol were the predominant reactions. CH3CHO and C2H4 were primary
products and formed in parallel on different active sites. H2O and C2H5OH competed for the same active
sites on the catalyst surface and the reaction order with respect to H2O was negative. The apparent activa-
tion energy for ethanol conversion was 110 kJ/mol. Furthermore, temperature-programmed desorption
experiments confirmed the competing adsorption of C2H5OH and H2O. Temperature-programmed
deuteration of used catalyst showed that the catalyst contained C2H4, CHx, acetate and carbonate species
during ESR reaction.
Received 6 January 2015
Received in revised form 23 June 2015
Accepted 6 July 2015
Available online xxx
Keywords:
Ethanol steam reforming
Kinetic
Carbonaceous residue
Acid-base pair site
Ni-based catalyst
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
explored the deactivation and reaction mechanism of ESR over
Co/CeO2 and Pt/CeZrO2 catalysts. They proposed that the adsorbed
Syngas (CO and H2) is a key intermediate in the chemical indus-
try. A wide range of valuable fuels and chemicals are produced
via syngas, such as long-chain hydrocarbons via the F-T synthesis,
methanol by the methanol process and H2 by the water-gas-shift
(WGS) reaction and preferential oxidation of CO. Steam reform-
ing (SR) of bio-ethanol is currently being considered as one of the
most sustainable and economical alternative to fossil fuel-based
processes for syngas and H2 production. The potential industrial
application of ethanol steam reforming (ESR) would most prob-
ably build on the already commercialized natural gas and light
hydrocarbons reforming system in which Ni is the preferred active
metal since the 1960s. However, compared with CH4, C2H5OH is a
very reactive molecule whose decomposition over a catalyst sur-
face (or in gas phase) can occur at relatively low temperatures [1,2].
Furthermore, dehydration and C–C bond scission in the C2H5OH
molecule poses additional challenges for the catalyst during ESR,
which is not encountered with CH4.
ethoxy species may follow one of the two distinct pathways:
(1) decompose to CO, CH4 and H2 or (2) further dehydrogenate
to CH3CHO and acetyl species. The dehydrogenated species may
undergo oxidation to acetate species, which is facilitated by the
presence of either hydroxyl groups or oxygen on the catalyst
surface. The acetate species can either decompose to form CH4
and CO or further oxidize to produce carbonate species, which
is also favored by the presence of O species. However, once the
metal–support interface is lost due to blockage by CHx species (coke
precursors), the demethanation of acetate becomes hindered, lead-
ing to the deactivation of the catalyst. Song et al. [6–8] conducted
similar experiments over Co catalyst supported on ZrO2 and CeO2
and proposed mechanistic steps starting with either molecular or
dissociative adsorption of C2H5OH on the Co sites. The adsorbed
C2H5OH molecules can decompose to form single C-species or
dehydrate to C2H4. The dissociatively adsorbed C2H5OH (ethoxy
species) can move to the Co-support interface and be oxidized by an
additional H abstraction forming CH3CHO. The CH3CHO can either
desorb or decompose to CH4 and CO, or further oxidize to form
acetic acid and surface acetate species. Once the acetate species are
formed, multiple routes are possible: (1) formation of CH3COCH3
through aldol condensation, followed by a dehydrogenation and
decarboxylation; (2) C–C bond cleavage leading to the formation
of single carbon species; (3) dissociation to form surface CH3 and
Over the past decade, temperature-programmed desorption
(TPD) and infrared spectroscopic investigations have yielded signif-
icant insight in the reaction mechanism of ESR. de Lima et al. [3–5]
∗
Corresponding author. Tel.: +47 22855456.
E-mail address: unni.olsbye@kjemi.uio.no (U. Olsbye).
http://dx.doi.org/10.1016/j.cattod.2015.07.006
0920-5861/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: G. Zeng, et al., Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The
initial steps, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.006

G Model
CATTOD-9668; No. of Pages11
ARTICLE IN PRESS
2
G. Zeng et al. / Catalysis Today xxx (2015) xxx–xxx
carbon–oxygen species. The formed carbon–oxygen species may
desorb or further oxidize to give carbonate species. On the other
hand, the CH3 fragment may undergo oxidation to form formate
and carbonate. The catalyst surface is then regenerated through
CO2 desorption following carbonate decomposition and ready for
the next catalysis cycle regardless of the route followed.
Although the understanding of ESR reaction has advanced
significantly, the exact surface mechanism and especially the rate-
determining step (RDS), is still under considerable debate [9–25].
The available kinetic studies suggest various reaction mechanisms
involving one or more RDSs: (1) molecular adsorption of C2H5OH,
(2) dissociative adsorption of C2H5OH, (3) dissociation of adsorbed
C2H5OH, (4) dehydrogenation of ethoxy species, (5) C–C bond scis-
sion, (6) surface reaction between two adsorbed species. It thus
appears that the RDS will critically depend on the exact operation
conditions and the catalyst applied.
The present work focused on ESR over hydrotalcite-derived (∼2
wt.%) Ni/Mg(Al)O, which has previously been tested as propane dry
reforming catalyst [26]. Kinetic investigations were carried out by
varying the space-time, H2O partial pressure (PH2O) and C2H5OH
partial pressure (PC2H5OH) to elucidate the possible RDS and gain
insight into the reaction mechanism of the ESR reaction. More-
over, temperature-programmed (TP) experiments were performed
to identify reaction intermediates and carbonaceous residue on the
catalyst surface after reaction.
an offline GC–MS (Hewlett Packard) equipped with a GS–GASPRO
column (60 m, 0.32 mm) and a HP-5973 mass selective detector.
2.3.1. Catalytic activity
The catalytic performance was evaluated at 500 ◦C and atmo-
spheric pressure. Typically, 5 mg catalyst (0.149–0.177 mm) was
mixed with 1.0 g of quartz (0.149–0.177 mm) and then loaded into
the reactor. Prior to reaction, the catalyst was reduced in a flow of
10% H2/Ar (50 ml/min) by heating from room temperature to 600 ◦C
at 10 ◦C/min and holding at the temperature for 10 min. The reac-
tor was then purged with 50 ml/min Ar for 20 min at 600 ◦C before
cooling down to reaction temperature and maintaining there for
another 30 min. The reactant mixture generated from two com-
bined saturator-cooling systems was further diluted with inert Ar
and N2 (internal standard) before entering the reactor. For pre-
oxidation reaction, the same procedure was applied except feeding
a flow of 6% O2/Ar (50 ml/min) during the pretreatment step.
For kinetic study, preliminary experiments were performed to
surpass the inter- and intra-particle diffusion resistance as sug-
gested by Idem and Bakhshi [28] (see Supplementary information,
Section S3). Kinetic data were collected by varying one parameter
at a time, keeping the other operating conditions constant. Specif-
ically, a constant reaction temperature of 500 ◦C, space-time in the
range of 0.03–0.50 mg min/ml, PH2O between 0 and 0.1551 atm,
PC2H5OH between 1.07 and 0.0448 atm were employed. The total
flow rate was kept 100 ml/min for all runs. During kinetic studies,
each data point was obtained in a separate experiment, starting
with a fresh catalyst. Unless otherwise specified, all kinetic data
were collected after 10 h on-stream by which quasi-steady state
conversion was approached.
2. Experimental
2.1. Catalyst preparation
The ethanol conversion and product yields were calculated from
the effluent concentrations via
The 2 wt.% Ni/Mg(Al)O catalyst was prepared by synthesis of the
corresponding hydrotalcite-like material, followed by calcination
and reduction in one step in a flow of 10 vol.% H2/N2, as reported
in detail in ref. [26].
ꢀ
ꢁ
out
ethanol
n
Xethanol
=
1 −
× 100%
out
n
C−containing species
noi ut
2.2. Characterization
Yield of i =
out
n
C-containing species
X-ray diffraction (XRD) patterns were acquired at room tem-
perature in a Siemens Bruker D5000 instrument in transmission
Debye–Scherrer geometry using Cu K␣1 radiation.
noi ut
Selectivity of i =
× 100%
out
The N2 physisorption isotherm was measured at −196 ◦C using a
BELSORP-mini II instrument. The sample was outgassed in vacuum
at 300 ◦C for 3 h prior to measurements.
n
C-containing products
noHut
2
Yield of H2
=
n
out
C-containing species
2.3. Catalytic test apparatus
noHut
2
All catalytic tests and TP experiments were conducted in the
apparatus schematically shown in Fig. 1. The system consists of a
feed section, a reaction section and an analysis section. The feed
section contains an array of mass flow controllers and two com-
bined saturator-cooling systems [27] for generating steam- and
ethanol-saturated gas, respectively. Ar was led through a boiling
saturator containing C2H5OH or H2O, and the resulting mixture
was cooled to the desired temperature by an attached column. The
alumina reactor consisted of an outer tube (8 mm i.d. and 12 mm
o.d.) and a smaller, closed tube (alumina, 8 mm o.d.) inserted from
the bottom of the outer tube. Alumina wool was placed on top
of the insert tube to hold the catalyst in place. A third alumina
tube (one end closed, 3 mm o.d. and 1 mm i.d.) contained a ther-
mocouple was inserted from the top of the outer tube, so that
the thermocouple could measure the temperature in the catalyst
bed. The analysis section consisted of a mass spectrometer (Pfeif-
fer Omnistar) (MS), an on-line Micro-GC (Agilent 3000) equipped
with two columns (molsieve and U-Plot) and a TC detector, and
Selectivity of H2
=
× 100%
out
n
H-containing products
Xethanol × Fint
ethanol
rethanol
=
mcat
where i: C-containing products in the effluent gas, i.e. CO, CO2,
CH4, C2H4, C2H6, CH3CHO, C2H5OH, C3H8 and C3H6; F: gas flow rate
(mol/h); r: conversion rate (mol gcat−1 h−1). The C and H numbers
in each effluent species were taken into account for the calculation.
In all experiments the carbon balance was within a range of 5%.
In addition, thermodynamic equilibrium calculations were carried
out on HSC 4.0 [29] at the same conditions as reaction.
2.3.2. TP experiments
In TP reduction (TPR) experiment, the sample (200 mg) was in-
situ pretreated at 300 ◦C for 1 h under a flow of Ar (50 ml/min) in
order to remove traces of adsorbates. Reduction profiles were then
Please cite this article in press as: G. Zeng, et al., Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The
initial steps, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.006

G Model
CATTOD-9668; No. of Pages11
ARTICLE IN PRESS
G. Zeng et al. / Catalysis Today xxx (2015) xxx–xxx
3
Fig. 1. Schematic diagram of experimental apparatus for ESR.
average pore diameter of the passivated catalyst were 140 m2/g
and 9.51 nm, respectively.
recorded by passing a stream of 10% H2/Ar (50 ml/min) through the
sample, while heating at a rate of 10 ◦C/min from ambient temper-
ature to 800 ◦C. The on-line MS was used to trace the variation of
H2 signal.
3.2. Preliminary investigation
In TPD experiments, the sample was prereduced in a flow of 10%
H2/Ar (50 ml/min) by heating from room temperature to 600 ◦C at
10 ◦C/min and holding at 600 ◦C for 10 min. After reduction, the sys-
tem was purged with 50 ml/min Ar for 30 min at 600 ◦C and then
cooled down to 50 ◦C. The adsorption of C2H5OH was performed
at 50 ◦C by introducing a flow of 2.1% C2H5OH/Ar (50 ml/min) for
30 min. In the case of C2H5OH + H2O, the system was purged with
Ar for 30 min after the adsorption of C2H5OH, and then exposed to
2.7% H2O/Ar (50 ml/min) for another 30 min. After adsorption, the
system was purged with Ar for another 30 min, and then subjected
to a linear temperature increase at a rate of 10 ◦C/min under flow-
ing Ar (50 ml/min). The effluent gas was analyzed by monitoring
characteristic ion fragments [8,30] by the on-line MS.
TP decomposition, hydrogenation (TPH) and deuteration (D2) of
used catalysts were carried out using the following procedure: for
TP decomposition, the used catalyst was cooled down to less than
200 ◦C under Ar flow (50 ml/min), and then maintained at that tem-
perature for 30 min, followed by ramping the temperature to 800 ◦C
at 10 ◦C/min. For TPH, the spent catalyst was cooled down to 50 ◦C,
and then fed a flow of 10% H2/Ar (50 ml/min) for 30 min, followed
by ramping the temperature from 50 to 800 ◦C at 10 ◦C/min. In the
case of TP-D2, the same procedure as TPH was applied except feed-
ing a flow of 10% D2/Ar (50 ml/min) instead. The effluent gas was
analyzed on-line by the MS.
3.2.1. Blank tests
In order to address the contribution of gas-phase ESR, blank tests
were carried out in the test reactor loaded with quartz at different
temperatures with a total flow rate of 100 ml/min (PC2H5OH = 0.0163
atm). The results are shown in Fig. S3. Thermal decomposition
led to minor ethanol conversion (∼2.5%) at 500 ◦C, but the rate
increased with increasing temperature and reached 88% conver-
sion at 800 ◦C. The main reactions occurred during gas-phase ESR
were the dehydrogenation and dehydration of C2H5OH, as well as
the decomposition of CH3CHO.
C2H5OH ↔ CH3CHO+H2
C2H5OH ↔ C2H4+H2O
CH3CHO ↔ CH4+CO
Based on the results of the blank tests, 500 ◦C was selected as an
appropriate temperature for the kinetic study.
(1)
(2)
(3)
3.2.2. Elimination of internal and external diffusional resistance
Experiments were carried out with various catalyst particle sizes
(0.125–0.250 mm) and total flow rates (65–200 ml/min) at 500 ◦C
to select appropriate variables for accurate kinetic analysis. Test
results are reported in Fig. S5. A particle size of 0.149–0.177 mm
and a total flow rate of 100 ml/min were found appropriate for the
study.
3. Results and discussion
3.1. Catalyst characterization
3.2.3. Thermodynamic analysis
The XRD patterns of the precursor and the passivated catalyst
(Fig. S1) showed that both the HT-like precursor and the final
Mg(Al)O mixed oxide were well crystalline materials. Further-
more, a tiny diffraction peak of Ni was observed. The TPR profile
of the passivated catalyst (Fig. S2) displayed that the reduction
temperature of NiO maximized at 350 ◦C. The BET surface area and
Thermodynamic calculations showed that C2H5OH will be com-
pletely converted to H2 and C1 products (CO2, CO, CH4 and C) at
equilibrium (Fig. S6). If no C–C bond scission occurred, the C2H5OH
would also be converted completely, but leaving CH3CHO, C2H4, H2
and H2O as the only products, as shown in Fig. S7.
Please cite this article in press as: G. Zeng, et al., Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The
initial steps, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.006

G Model
CATTOD-9668; No. of Pages11
ARTICLE IN PRESS
4
G. Zeng et al. / Catalysis Today xxx (2015) xxx–xxx
Fig. 2. The ethanol conversion as
a function of time-on-stream over 2
wt.% Ni/Mg(Al)O catalyst. (Total flow rate = 100 ml/min, PC2H5OH = 0.0163 atm,
PH2O = 0.0500 atm, space-time = 0.05 mg min/ml, 500 ◦C.)
3.3. Catalyst performance with time-on-stream
The time-on-stream activity of the Ni/Mg(Al)O catalyst at 500 ◦C
and space-time of 0.05 mg min/ml is shown in Fig. 2. Ethanol con-
version decreased significantly during the first hour on-stream, but
became relatively stable (quasi-steady state) after 3 h on-stream.
The main products, i.e., H2, CH3CHO and C2H4, followed the same
trend during the time region tested, indicating that the catalyst
was mainly active for the dehydrogenation and dehydration of
ethanol. The selectivity of CH3CHO, as shown in Fig. S8, slowly
decreased with time-on-stream, whereas that of C2H4 correspond-
ingly increased.
Three plausible hypotheses exist for the initial deactivation,
i.e., Ni metal oxidation, coke deposition and surface restructur-
ing/sintering. While the last hypothesis was beyond the scope of
this study, the first two were experimentally tested. First, a pos-
sible, initial oxidation state change was elucidated by pretreating
the catalyst under three different conditions: pre-reduction, pre-
oxidation and calcination only. As shown in Fig. 3, the same initial
conversion, deactivation behavior as well as the same steady state
ethanol conversion and product yields were observed for all tests,
suggesting that a possible oxidation change (metallic oxidation)
have been completed before the first analysis (around 3 min on-
stream).
Since the Ni may possibly be oxidized, H2 co-feed experiments
were attempted to possibly enhance Ni reduction and thereby
promote syngas formation. The resulting conversion versus time-
on-stream and main product yields are shown in Fig. 4a and b,
respectively. As can be seen, similar trends but lower reactivities
versus time-on-stream were observed for the tests with the addi-
tion of H2 (Fig. 4a). Moreover, the steady state ethanol conversion
decreased from 24% to 19% upon H2 addition, irrespective of the H2
amount fed. CH3CHO and C2H4 were still the dominating products,
whereas the H2 addition influenced the CH3CHO yield more than
the C2H4 yield (Fig. 4b).
The hypothesis of deposition of carbonaceous compounds (or
‘coke’) as the reason for the initial deactivation was elucidated by
monitoring the mass balance. The carbon mass balances for three
individual tests performed under the same conditions (standard
conditions) over the same catalyst are shown in Fig. 5a. It is clear
that the C-balances within the time period shown were very similar
Fig. 3. The effect of catalyst pretreatment on: (a) time-on-stream ethanol conver-
sion;(b) quasi-steady state activities. (Total flow rate = 100 ml/min, PC2H5OH = 0.0163
atm, PH2O = 0.0500 atm, space-time = 0.05 mg min/ml, 500 ◦C). (Reduced-1 and -2
represent the reproducible tests over the same catalyst under the same conditions.)
except for the first analysis point, suggesting that some C species
may accumulate on the catalyst surface during the initial few min-
utes. More information about the initial period was gained from
tracing the real-time concentration of different components by MS
(Fig. 5b). After switching to reactant feed, H2O (m/e 17) was identi-
fied as the first molecule to leave the reactor, followed by H2 (m/e 2),
and then the C-containing species. The intensities of C2H5OH (m/e
31) and other C-containing species (e.g. CH3CHO (m/e 29)) were far
from steady state when the first GC analysis was measured (5 min
on-stream). Therefore, the first GC analysis presented a poor C-
balance and consequently showed a much higher conversion than
the second one (18 min on-stream), which in turn suggests that the
initial activity of the catalyst is not appropriate as basis for kinetic
studies. In addition, a remarkably high concentration of H2 was
observed within 2–4 min on-stream, likely being related to the high
activity of the fresh catalyst, and possibly indicating the deposition
of carbonaceous compounds in addition to C2 formation. In order
to limit the effect of the initial rapid deactivation and the state dif-
ference of the catalyst in separate experiments, it was decided to
use the steady state activity of the catalyst, obtained after 10 h on-
stream, as basis for the kinetic study in this work, which is in line
with some previous literature [10,14,22].
Please cite this article in press as: G. Zeng, et al., Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The
initial steps, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.006

G Model
CATTOD-9668; No. of Pages11
ARTICLE IN PRESS
G. Zeng et al. / Catalysis Today xxx (2015) xxx–xxx
5
Fig. 4. The effect of H2 co-feeding: (a) time-on-stream ethanol conversion;(b)
quasi-steady state activities. (Total flow rate = = 100 ml/min, PC2H5OH = 0.0163 atm,
Fig. 5. The initial C-balances of several reproducible time-on-stream tests (a) and
the MS traced real-time composition of the effluent gas during the initial 20 min
(b). (Total flow rate = 100 ml/min, PC2H5OH = 0.0163 atm, PH2O = 0.0500 atm, space-
time = 0.05 mg min/ml, 500 ◦C.)
PH2O = 0.0500 atm, space-time = 0.10 mg min/ml, 500 ◦C.) (The two legend of
and
in a represent the reproducible tests performed at the same conditions.)
catalyst surface, and that the sites responsible for CH3CHO for-
mation deactivates more rapidly than those responsible for C2H4
formation.
3.4. Kinetic investigation
3.4.1. Effect of space-time
Fig. 8 shows the individual product yields versus ethanol con-
version for the space-times investigated. The plots of CH3CHO and
C2H4 increased monotonically from origo with ethanol conver-
sion, suggesting that the two products are primary and formed in
parallel. The H2 yield curve also crossed origo and increased mono-
tonically with ethanol conversion until 60% conversion, and then
shifted to a higher slope. This curve shape suggests that H2 is both
a primary and a secondary product. Significant amounts of CH4, CO
and CO2 were only observed at conversions exceeding 60% (CO2)
and 80% (CH4 and CO), indicating that these products are secondary
products of reaction. The parallel formation of CH4 and CO further
suggests that they are formed by CH3CHO splitting.
Ethanol conversion and product yields versus space-time after
10 h on-stream at 500 ◦C are shown in Fig. 6a and b, respectively.
Ethanol conversion increased and approached an asymptotic value
for space-time ≤0.3 mg min/ml, but further increased for the high-
est space-time studied (0.5 mg min/ml). Inspection of the product
yield curves showed that the inversion of the ethanol conversion
curve around 0.3 mg min/ml was due to a stabilization, followed
by an increase, in the net CH3CHO product yield with increasing
space-time. Furthermore, significant formation of CH4, CO and CO2
was observed at the highest space-time. Repeated measurements
showed that the observed effect of space-time was reproducible.
Fig. 7 shows CH3CHO and C2H4 selectivities versus ethanol con-
version, obtained at different space-times at 500 ◦C. As can be
seen, selectivities to CH3CHO and C2H4 were varied contrarily,
while their values globally decrease and increase with increas-
ing space-time, respectively. For a given conversion, e.g. 35%,
CH3CHO selectivity was systematically higher for a fresh catalyst
(space-time 0.10 mg min/ml) than for a deactivated one (space-
time 0.20 mg min/ml). This observation suggests that C2H4 and
CH3CHO are (at least partially) formed on different sites on the
3.4.2. Effect of H2O partial pressure
The effect of PH2O on ethanol conversion rate is shown in Fig. 9.
At a constant PC2H5OH and space-time, an increase in PH2O signifi-
cantly decreased the ethanol conversion rate, suggesting a negative
reaction order with respect to H2O. Increasing the constant PC2H5OH
from 0.0163 to 0.0408 atm moderately increased the conversion
rate. However, the negative order in PH2O on ethanol conversion
Please cite this article in press as: G. Zeng, et al., Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The
initial steps, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.006

G Model
CATTOD-9668; No. of Pages11
ARTICLE IN PRESS
6
G. Zeng et al. / Catalysis Today xxx (2015) xxx–xxx
was maintained. Together, these findings indicate that the H2O
and C2H5OH compete for the same active sites on the catalyst sur-
face. Supporting evidence for this conclusion was found by TPD
experiments below.
The effect of PH2O on the individual CH3CHO and C2H4 prod-
uct yields is shown in Fig. 10, together with lines indicating the
expected product yields at thermodynamic equilibrium without
C–C bond scission. The influence of H2O addition is stronger on C2H4
formation than on CH3CHO formation, again pointing to different
sites being active for the formation of these two products.
3.4.3. Effect of C2H5OH partial pressure
The ethanol conversion rate as a function of PC2H5OH is plot-
ted in Fig. 11. Apparently, the curves have the form of a Langmuir
isotherm, suggesting that the active sites reached saturation in
C2H5OH at high PC2H5OH. An increase in the constant PH2O led to a
general decrease in conversion rate, but without significantly alter-
ing the reaction order in C2H5OH, again pointing to a competitive
adsorption between the two reactants. Plotting the ethanol conver-
sion rate as a function of H2O/C2H5OH ratio shows the link between
the two variables (Fig. 12). The correlation shown suggests that the
effects of PC2H5OH and PH2O could essentially be attributed to that of
H2O/C2H5OH ratio. The product yields of CH3CHO and C2H4 versus
PC2H5OH are illustrated in Fig. 13. Similar to that of PH2O, increasing
PC2H5OH led to decreased CH3CHO and C2H4 yields, but favored the
selectivity of C2H4 (Fig. S11).
3.4.4. Apparent activation energy
In order to evaluate the apparent activation energy (EA)
of ethanol reforming, experiments were performed at con-
stant PH2O (0.0500 atm), PC2H5OH (0.0163 atm) and space-time
(0.05 mg min/ml), while the temperatures were varied from 489
to 511 ◦C. The Arrhenius plot of ln(rethanol) versus 1/T is shown in
Fig. 14.The apparent activation energy of ESR was calculated to
be 110.34 kJ/mol. In comparison with the EA reported elsewhere
[9–20,22–24], it is apparent that the EA seems to depend on the
extent of reaction and the exact reaction conditions, which prob-
ably would also alter the rate determining step in the reaction
sequence.
Fig. 6. The effect of space-time on quasi-steady state ethanol conversion and prod-
uct distribution. (Total flow rate = 100 ml/min, PC2H5OH = 0.0163 atm, PH2O = 0.0500
atm, 500 ◦C.)
3.5. Temperature-programmed experiments
3.5.1. TPD of C2H5OH and C2H5OH + H2O
Fig. 15 presents the TPD profiles of adsorbed C2H5OH and
C2H5OH + H2O over the 2 wt.% Ni/Mg(Al)O catalyst. As revealed
by Fig. 15a, the main desorbed species were C2H5OH (m/e 31),
H2O (m/e18), CH3CHO (m/e 29), C2H4 (m/e 28, 27, 26), H2, CO (m/e
28), CO2 (m/e 44). Specially, large amount of H2O was detected in
the temperature range of 50–550 ◦C (maxima at around 250 ◦C),
being related to the high extent of ethanol dehydration and the
subsequent adsorption-desorption on the catalyst. At low temper-
atures (<300 ◦C), the catalyst exhibited wide and intense C2H5OH,
CH3CHO and C2H4 desorption peaks (maxima at around 150 ◦C),
being assigned to the ethanol physically adsorbed and its dehydro-
genation and dehydration products. Some authors [4,5,31,32] also
reported broad desorption peaks of C2H5OH, CH3CHO and C2H4
centered at around 200 ◦C over supported noble metal catalysts.
The fact that there is only trace amount of CH4 and/or CO (covered
by abundant C2H4) at the low temperature region suggests that the
extent of CH3CHO decomposition is low under the conditions used.
Some literature [3–5,8] reported the formation of H2, CH4 and CO in
the low-temperature region during TPD of C2H5OH over Pt/CeZrO2
and Co/CeO2 catalysts. They suggested that C2H5OH adsorbs on the
catalyst surface as ethoxy species that can be decomposed to H2,
CH4 and CO over metal particles. However, in this work, a signifi-
cant peak for H2 desorption (one of the three) was only detected at
Fig. 7. The C2 selectivities as a function of ethanol conversion over 2 wt.% Ni/Mg(Al)O
catalyst. (Total flow rate = 100 ml/min, PC2H5OH = 0.0163 atm, PH2O = 0.0500 atm,
500 ◦C.)
Please cite this article in press as: G. Zeng, et al., Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The
initial steps, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.006

G Model
CATTOD-9668; No. of Pages11
ARTICLE IN PRESS
G. Zeng et al. / Catalysis Today xxx (2015) xxx–xxx
7
Fig. 8. The product distribution as a function of ethanol conversion over 2 wt.% Ni/Mg(Al)O catalyst. (Total flow rate = 100 ml/min, PC2H5OH = 0.0163 atm, PH2O = 0.0500 atm,
500 ◦C.)
temperatures above 200 ◦C, indicating that generated H2remained
on the catalyst surface.
species. Although it is common to observe CH4 within the temper-
ature range tested, there is only trace CH4 signal in our experiment.
This may be due to the ultimately decomposition of CHx fragment
with the aid of highly dispersed Ni particles on the catalyst surface,
in line with the proposal of Marton et al. [35]. In addition, there
is also a small peak, corresponding to C2H4, that appears between
286 and 400 ◦C, coinciding with the second significant peak of H2
desorption. This strongly suggests that C2H4 and H2 are formed via
a common intermediate.
To examine the role of H2O and its interaction with C2H5OH as
well as with catalyst, a TPD of C2H5OH + H2O experiment was per-
formed by a sequential adsorption step where C2H5OH adsorption
was followed by H2O adsorption. The procedure for C2H5OH + H2O
adsorption is presented in Fig. 15b. As can be seen, the C2H5OH
adsorbed was surprisingly wiped out by H2O in the H2O adsorption
step. Generally, there could be two reasons for this observation: (1)
At higher temperatures, significant peaks for H2 production
(two of the three, centered at 356 ◦C and 493 ◦C, respectively), a
broad peak of CO (> 419 ◦C), a much smaller peak for CO2 as well
as a trace peak of CH4 were detected. Several previous studies
[4,5,8,33,34] detected the formation of CO2, CO and CH4 at relatively
high temperatures (>250 ◦C) during TPD of C2H5OH or CH3CHO
over Pt/CeZrO2, Co/CeO2, Ir/CeO2, and PtNi/␥-Al2O3 catalysts, and
assigned their formation to the decomposition and oxidation of
carbon species (i.e., CH3CHO and acetate species) formed previ-
ously. In the current study, it is highly possible that some of the
ethoxy species are rapidly converted into acetate and carbonate
species on the catalyst surface upon heating. The decomposition of
acetate species is responsible for the formation of CO and CH4 at
temperatures above 400 ◦C, whereas the formation of small amount
of CO2 is probably mainly due to the decomposition of carbonate
Fig. 10. The effect of PH2O on CH3CHO and C2H4 product yields. (Lines are equi-
librium composition calculated basing on dehydrogenation and dehydration of
ethanol) (Total flow rate = 100 ml/min, space-time = 0.05 mg min/ml, 500 ◦C.)
Fig. 9. The effect of PH2O on ethanol conversion rate. (Total flow rate = 100 ml/min,
space-time = 0.05 mg min/ml, 500 ◦C.)
Please cite this article in press as: G. Zeng, et al., Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The
initial steps, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.006

G Model
CATTOD-9668; No. of Pages11
ARTICLE IN PRESS
8
G. Zeng et al. / Catalysis Today xxx (2015) xxx–xxx
Fig. 11. The effect of PC2H5OH on ethanol conversion rate. (Total flow
rate = 100 ml/min, space-time = 0.05 mg min/ml, 500 ◦C.)
Fig. 14. Arrhenius plot of ln(rethanol) versus 1/T for ESR. (Total flow rate = 100 ml/min,
PC2H5OH = 0.0163 atm, PH2O = 0.0500 atm, space-time = 0.05 mg min/ml, 489–511 ◦C.)
reconstruction of the hydrotalcite phase by the addition of steam;
(2) competing adsorption of H2O on the ethanol adsorption site. The
first possibility was ruled out by characterizing the catalyst by XRD
after subjecting it to steam at 50 ◦C followed by cooling to room
temperature. No phase change was observed. It was therefore con-
cluded that the replacement of C2H5OH by H2O in this work was
mainly due to the competitive adsorption on the same active site,
confirming the hypothesis in Section 3.4.2.
When subjected to temperature program (Fig. 15c), an intense
desorption peak of H2O and much weaker desorption peaks (max-
ima at around 315 and 470 ◦C) of H2 were detected, while only a
trace peak of CO2 (m/e 44) (maxima at around 485 ◦C) was observed
as C-containing species in the temperature range investigated.
The fact that only trace amounts of C-containing species such as
CH3CHO, C2H4, CO and CO2 were observed during TPD, in line with
the small amount of retained C2H5OH after H2O adsorption. The
effective substitution of C2H5OH by H2O made it possible to esti-
mate the amount of adsorption sites for C2H5OH and H2O on the
catalyst surface. Table 1 presents the calculation parameters and
results by assuming that one molecule covered one adsorption site.
As listed, the amount of C2H5OH adsorption site was approximately
1.55 mmol/g while the amount of sites for H2O was almost three
times higher, around 4.44 mmol/g.
Fig. 12. The effect of H2O/C2H5OH molar ratio on ethanol conversion rate. (Total
flow rate = 100 ml/min, space-time = 0.05 mg min/ml, 500 ◦C.)
3.5.2. Carbonaceous residue on the spent catalyst
The main reactions that may contribute to coke formation dur-
ing ESR are the following: (1) ethanol dehydration to C2H4, followed
by polymerization to coke; (2) aldol-condensation of CH3CHO to
CH3COCH3, followed by further condensation and dehydration to
mesityl oxide/heavy hydrocarbon [5,8]; (3) the Boudouard reaction
and (4) the decomposition of light hydrocarbons such as CH4 and
C2H4 to carbon and H2.
In addition to enhance the conversion of ethanol and the for-
mation of C1 products, TP oxidation of the catalysts after the
space-time investigations demonstrate that higher space-times
also favor the deposition of carbon (not shown), in good agreement
with the results of Co/CeO2 [36]. In this case, further TP experiments
were carried out over catalysts after running at high space-time
to investigate the nature of the residues that may contribute to
catalyst deactivation. Fig. 16a illustrates the TP decomposition of
the catalyst after running at a space-time of 0.20 mg min/ml. As
can be seen, the major species detected were C2H4 (m/e 26, 27,
28), H2 (m/e 2), CO (m/e 28), CO2 (m/e 44) and CH4 (m/e 15).
At temperatures < 420 ◦C, significant peaks corresponding to C2H4,
Fig. 13. The effect of PC2H5OH on CH3CHO and C2H4 product yields. (Lines are equi-
librium composition calculated basing on dehydrogenation and dehydration of
ethanol) (Total flow rate = 100 ml/min, space-time = 0.05 mg min/ml, 500 ◦C.)
Please cite this article in press as: G. Zeng, et al., Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The
initial steps, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.006

G Model
CATTOD-9668; No. of Pages11
ARTICLE IN PRESS
G. Zeng et al. / Catalysis Today xxx (2015) xxx–xxx
9
Fig. 16. TP characterization of spent 2 wt.% catalysts: (a) TP decomposition (space-
time = 0.20 mg min/ml); (b) TPH (space-time = 0.20 mg min/ml); (c) TP-D2 (space-
time = 0.50 mg min/ml).
Fig. 15. Gas-phase species monitored with MS during TPD of C2H5OH (a), adsorption
of H2O after C2H5OH adsorption (b) and TPD of C2H5OH + H2O (c).
Table 1
The calculated adsorption sites for C2H5OH and H2O on the 2 wt.% Ni/Mg(Al)O catalyst.
Reactant
Feed (ml)
Area inC2H5OH
adsorption step
Area in H2O
adsorption step
Area during
desorption step
Volume adsorbed
on catalyst (ml)
Catalyst (g)
Adsorption site
(mmol/g)
C2H5OH
H2O
30.89
36.68
16.11
–
1.05
32.73
–
5.70
1.89
5.45
0.0501
0.0501
1.55
4.44
Please cite this article in press as: G. Zeng, et al., Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The
initial steps, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.006

G Model
CATTOD-9668; No. of Pages11
ARTICLE IN PRESS
10
G. Zeng et al. / Catalysis Today xxx (2015) xxx–xxx
accompanied by a much smaller peak of H2, were observed between
290 and 415 ◦C. This is probably related to desorption and decom-
position of adsorbed or slightly polymerized C2H4 species. At higher
temperature, the simultaneous formation of CO2, CH4, CO and H2
(all centered at around 530 ◦C) were associated with the decompo-
sition of acetate species [37,38], which accumulated on the catalyst
surface during reaction.
Fig. 16b presents the TPH profiles of the used catalyst which
were also performed at a space-time of 0.20 mg min/ml. As shown,
the main evolution peaks observed were C2H4 (maxima at around
350 ◦C), CH4 (510 ◦C) and CO (550 ◦C), suggesting again that the
carbon species deposited during ESR are C2H4 intermediates and
acetate species. The fact that there was almost no CO2 detected
in this case suggests that the generation of CO2 is not favored in
the reducing atmosphere. Besides, no CH4 was observed in the low
temperature region (< 400 ◦C) [39], implying that almost no CHx is
generated or remained on the catalyst surface under the conditions
applied.
Fig. 16c shows the TP-D2 profiles of the used catalyst performed
at the highest space-time for more than 16 h. As illustrated, the
most abundant desorption peak evolving at temperatures <400 ◦C
was HD (m/e 3), indicating that huge amount of H species remained
on the catalyst surface after test. Meanwhile, the evolution peaks of
CH3D (m/e 17), CH2D2 (m/e 18) and CHD3 (m/e 19) (centered around
200 ◦C) were detected simultaneously, confirming the presence of
CHx group on the catalyst surface. Besides, a trace peak assigned to
CO2 was observed, and can be attributed to the decomposition of
carbonate species formed during reaction. At higher temperatures,
a pronounced peak of CH2D2 (m/e 18, centered at around 620 ◦C)
was accompanied by a much smaller peak of CHD3 (m/e 19) as well
as a trace peak of CH4 (m/e 16). Moreover, significant formation
of HD (m/e 3) and CO were observed in the closely followed tem-
perature region. Together, these observations at high temperature
range again confirm the deposition of acetate species during reac-
tion and its decomposition to form CHx, H2 and CO during heating.
Furthermore, the most abundant CH2D2 (m/e 18) detected among
the CHx derivatives is likely correlated with the stability of CH2
species among the CHx group formed right after the decomposition
of acetate species, further indicating that the C–C bond breaking
may occur through CH2–CO species [40].
which means the co-fed H2 may directly affect the adsorption
and conversion of C2H5OH, especially the dehydrogenation reac-
tion.
Even though 2 wt.% Ni may be too little to maintain suffi-
cient C–C bond rupture activity for syngas formation under the
conditions studied, this study provides interesting information
concerning the reaction scheme and the effect of Ni2+ addition to
Mg(Al)O. Di Cosimo et al. have previously studied the effect of Al
incorporation on MgO on the surface basicity and ethanol conver-
sion [43,44]. They reported that there are surface sites of low (OH−
group), medium (Mn+–O2−), and strong (O2− anions) basic centers
in the surface of Mg(Al)O mixed oxide with 5 > Mg/Al > 1. They fur-
ther reported that the dehydrogenation of C2H5OH to CH3CHO is
favored on the Mg-rich Mg(Al)O oxides because these samples con-
tain a large number of properly positioned Al3+ Lewis acid sites and
Mg2+–O2− basic pairs, which are required for H abstraction steps
leading to ethoxy formation. On the other hand, the dehydration
of C2H5OH also involves the initial surface ethoxy formation on a
weak Lewis acid-strong base pair site. Recently, Jensen et al. [45]
investigated the acid-base property of 2 wt.% Ni/Mg(Al)O catalyst
and concluded that the sample has an intermediate basic strength
of O2− sites and Mn+–O2− pairs and acidic Lewis sites in comparison
with Ni/CaO, Ni/MgO and Ni/Al2O3. Therefore, the ethanol dehy-
drogenation and dehydration could proceed simultaneously and
might also share common steps for H abstraction from C2H5OH (the
ethoxy species) since O2− sites and Mn+–O2− pairs and acidic Lewis
sites exist together over the Ni/Mg(Al)O catalyst. In the current
study, kinetic investigation suggests that CH3CHO and C2H4 are
primary products which are formed in parallel on different active
sites. H2, C2H5OH and H2O compete for the same active sites on
the catalyst surface. Moreover, TPD experiment suggests that C2H4
and H2 are formed via a common intermediate. On the basis of these
observations, it is reasonable that dehydrogenation and dehydra-
tion share the same elementary steps for ethoxy species formation
(C2H5OH ↔ C2H5OH* ↔ C2H5O* + H*).
Although dehydrogenation and dehydration could proceed
simultaneously, much higher selectivities to CH3CHO were
achieved at low space-time region, as shown in Fig. 7. Interest-
ingly, this result is quite different from that obtained for Mg(Al)O
without incorporation of Ni2+ (Fig. S9). As shown, C2H4 selectiv-
ity is always higher than that of CH3CHO. Increasing space-time
increased the selectivity to C2H4, while its value varied lightly with
ethanol conversion at a fixed space-time. A comparison between
the selectivity versus ethanol conversion curves obtained over
Mg(Al)O (Fig. S9) and Ni/Mg(Al)O (Fig. 7), respectively, at differ-
ent space-times reveal two additional differences. First, the C2H4
and CH3CHO selectivities varied with deactivation over Ni/Mg(Al)O
but not over Mg(Al)O. This result suggests that the selective deac-
tivation of the Ni/Mg(Al)O sample is related to catalytic sites which
are created by Ni addition to Mg(Al)O, and that those sites favor
CH3CHO formation. On the other hand, Mg(Al)O deactivated very
little during 1000 min on-stream. Second, Mg(Al)O was selective
towards C2H4, and yielded C2H4:CH3CHO ratios in the range 1.5–3
(Fig. S9), while Ni/Mg(Al)O was selective towards CH3CHO, and
yielded C2H4:CH3CHO ratios in the range 0.33–1 (Fig. 7) under
the same conditions. The opposite selectivity trend observed for
the two materials strongly suggest that sites involving Ni domi-
nate product formation over the Ni/Mg(Al)O material. This effect
is likely related to either a) the abstraction of Ni from the original
hydrotalcite-like precursor during calcination creates defects on
the Mg(Al,Ni)O support which facilitate the adsorption of ethanol
and the activation of H2O to OH− or O2-, or b) partial substitution of
Mg2+ (72 pm) by smaller Ni2+ (69 pm) modifies the abundance of
4. Overall discussion
The Ni/Mg(Al)O catalyst used in this study has previously been
successfully tested as catalyst for the dry reforming of propane
to syngas (CO and H2) at 600 ◦C [26]. A major observation from
the test data presented in this study is that the same Ni/Mg(Al)O
catalyst was barely active for syngas formation from ethanol and
water at 500 ◦C. It mainly produced C2H4, CH3CHO, H2 (and H2O)
under all tested conditions, while CO and CO2 were detected in
the effluent only at the highest space-time employed (Fig. 6 and
8). The reason could be that metallic Ni was rapidly covered
and deactivated by coke, as suggested by the poor initial carbon
mass balance (Fig. 5), or that Ni was mainly present in an oxi-
dized state under the conditions employed, or both. Instead of
improving the catalyst performance, H2 addition (Fig. 4) signif-
icantly reduced the ethanol conversion in terms of suppressing
C2H4 and, even more strongly, CH3CHO generation. Laosiripojana
et al. [41,42] studied the effect of co-fed H2 on ESR and CH4 dry
reforming reactions. They attributed the negative effect of H2 on
CH4 conversion to inhibition of the hydrocarbon - lattice oxygen
interaction, and to possible coverage of active Ni sites by hydrogen.
In our case, the negative impacts of H2 co-feed on ethanol conver-
sion and product distribution are probably due to the competition
for the same active sites with C2H5OH and H2O. Another possi-
ble explanation is the consumption of surface OH-/O2− species,
M
n+–O2− pair sites that promote dehydrogenation. However, this
advantage can be diluted by increasing space-time, since much
more acid-base pair sites are available for ethanol conversion,
Please cite this article in press as: G. Zeng, et al., Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The
initial steps, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.006

G Model
CATTOD-9668; No. of Pages11
ARTICLE IN PRESS
G. Zeng et al. / Catalysis Today xxx (2015) xxx–xxx
11
indicating that ethanol dehydration might be a preferential reac-
tion in comparison with dehydrogenation under some specific
conditions, further suggesting that dehydration may be unpre-
ventable during ESR reaction but can be suppressed efficiently by
the presence of active metal.
Changjiang Scholars and Innovative Research Teams in Universi-
ties under file number IRT 0641. G.M. Zeng thanks the Chinese
Scholarship Council for a fellowship to work in the University of
Oslo.
Among the several available kinetic investigations, varied
kinetic regimes with the variations of PC2H5OH and PH2O were also
observed by Simson et al. [20]. It was found that the reaction order
for C2H5OH was 1.2 for PC2H5OH of 0.03–0.045 atm, while little
dependence on C2H5OH was observed at lower concentrations.
On the other hand, changes in PH2O (<0.14 atm) did not affect
ethanol conversion, while a negative dependence was found for
higher concentrations. Örücü et al. [15] studied the kinetics on
the basis of initial reaction rate over Pt-Ni/Al2O3 at 400 ◦C and
proposed that the rate of ESR is directly proportional to PC2H5OH
and is mildly inhibited by PH2O. Recently, a sole dependence of ESR
reaction on C2H5OH was found by Ciambelli et al. [22] over Pt/CeO2
catalyst at 300 ◦C with various PH2O (0.005–0.030 atm) and PC2H5OH
(0.001–0.010 atm). Although some part of information/conclusions
proposed by literature seem in favor of our work, the different
operating conditions and various catalyst systems preclude direct
comparison. However, the extensive kinetic data obtained in this
work demonstrates that the competitive interaction of C2H5OH
and H2O with the support is critical for ESR reaction over 2 wt.%
Ni/ mg(Al)O. Furthermore, kinetic regime may change depending
Appendix A. Supplementary data
Supplementary material related to this article can be found, in
the online version, at doi:10.1016/j.cattod.2015.07.006.
References
[1] J. Sun, X.P. Qiu, F. Wu, W.T. Zhu, Int. J. Hydrogen Energy 30 (2005) 437.
[2] Y.D. Li, D.X. Li, G.W. Wang, Catal. Today 162 (2011) 1.
[3] S.M. de Lima, A.M. da Silva, L.O.O. da Costa, U.M. Graham, G. Jacobs, B.H. Davis,
L.V. Mattos, F.B. Noronha, J. Catal. 268 (2009) 268.
[4] S.M. de Lima, I.O. da Cruz, G. Jacobs, B.H. Davis, L.V. Mattos, F.B. Noronha, J.
Catal. 257 (2008) 95.
[5] S.M. de Lima, A.M. Silva, U.M. Graham, G. Jacobs, B.H. Davis, L.V. Mattos, F.B.
Noronha, Appl. Catal. A 352 (2009) 95.
[6] H. Song, U.S. Ozkan, J. Catal. 261 (2009) 66.
[7] H. Song, L.Z. Zhang, U.S. Ozkan, Ind. Eng. Chem. Res. 49 (2010) 8984.
[8] H. Song, X.G. Bao, C.M. Hadad, U.S. Ozkan, Catal. Lett. 141 (2011) 43.
[9] D.A. Morgenstern, J.P. Fornango, Energy Fuels 19 (2005) 1708.
[10] P.D. Vaidya, A.E. Rodrigues, Ind. Eng. Chem. Res. 45 (2006) 6614.
[11] A. Akande, A. Aboudheir, R. Idem, A. Dalai, Int. J. Hydrogen Energy 31 (2006)
1707.
[12] D.R. Sahoo, S. Vajpai, S. Patel, K.K. Pant, Chem. Eng. J. 125 (2007) 139.
[13] E. Akpan, A. Akande, A. Aboudheir, H. Ibrahim, R. Idem, Chem. Eng. Sci. 62
(2007) 3112.
on the PC2H5OH and PH2O
.
[14] P.V. Mathure, S. Ganguly, A.V. Patwardhan, R.K. Saha, Ind. Eng. Chem. Res. 46
(2007) 8471.
5. Conclusions
[15] E. Örücü, F. Gökaliler, A.E. Aksoylu, Z.I. Önsan, Catal. Lett. 120 (2008) 198.
[16] V. Mas, M.L. Dieuzeide, M. Jobbágy, G. Baronetti, N. Amadeo, M. Laborde,
Catal. Today 133–135 (2008) 319.
In this work, kinetic investigation and TP characterization were
carried out over 2 wt.% Ni/Mg(Al)O catalyst to gain insight into
the reaction mechanism and deactivation. The kinetic investiga-
tion demonstrated that C–C bond scission is strongly restricted
over the Ni/Mg(Al)O catalyst at 500 ◦C and space-times of
0.03–0.5 mg ml/min, being attributing to the deposition of carbona-
ceous species and/or the oxidation of Ni active metal. CH3CHO
and C2H4 are primary products and formed in parallel on different
active sites. Dehydrogenation and dehydration are the predomi-
nant reactions and may share the same elementary steps for the
ethoxy species. H2O and C2H5OH compete for the same active sites
on the catalyst surface and the reaction order with respect to H2O
is negative. The apparent activation energy for ethanol conversion
is 110 kJ/mol.
TPD of C2H5OH indicate that the conversion of C2H5OH proceeds
through dehydration and dehydrogenation, followed by further
oxidation and decomposition to form C1 products and H2. TPD of
C2H5OH + H2O confirmed the competing adsorption of C2H5OH and
H2O on the catalyst surface, although the adsorption of H2O was
three times higher than that of C2H5OH. The TP decomposition and
TPH of the used catalysts suggested the presence of C2H4 inter-
mediates and acetate species during ethanol conversion and their
decomposition to produce C2H4, H2, CH4, CO and CO2. TP-D2 of
the used catalyst running at the highest space-time confirmed that
the residues contain H species, CHx group, carbonate and acetate
species, and further indicates that C–C bond breaking may occur
through CH2–CO species.
[17] V. Mas, G. Baronetti, N. Amadeo, M. Laborde, Chem. Eng. J. 138 (2008) 602.
[18] V. Mas, M.L. Bergamini, G. Baronetti, N. Amadeo, M. Laborde, Top. Catal. 51
(2008) 39.
[19] I. Llera, V. Mas, M.L. Bergamini, M. Laborde, N. Amadeo, Chem. Eng. Sci. 71
(2012) 356.
[20] A. Simson, E. Waterman, R. Farrauto, M. Castaldi, Appl. Catal. B 89 (2009) 58.
[21] O. Görke, P. Pfeifer, K. Schubert, Appl. Catal. A 360 (2009) 232.
[22] P. Ciambelli, V. Palma, A. Ruggiero, Appl. Catal. B 96 (2010) 190.
[23] C. Graschinsky, M. Laborde, N. Amadeo, A.L. Valant, N. Bion, F. Epron, D.
Duprez, Ind. Eng. Chem. Res. 49 (2010) 12383.
[24] N.R. Peela, D. Kunzru, Ind. Eng. Chem. Res. 50 (2011) 12881.
[25] V. Palma, F. Castaldo, P. Ciambelli, G. Iaquaniello, Appl. Catal. B 145 (2014)
73–84.
[26] A. Olafsen, Å. Slagtern, I.M. Dahl, U. Olsbye, Y. Schuurman, C. Mirodatos, J.
Catal. 229 (2005) 163.
[27] G. Ertl, H. Knözinger, J. Weitkamp, Handbook of Heterogeneous Catalysis, vol.
III, Wiley-VCH, Weinheim, 1997.
[28] R.O. Idem, N.N. Bakhshi, Chem. Eng. Sci. 51 (1996) 3697.
[29] A. Roine A, Outokumpu HSC Chemistry 4.0, Poris, Finland.
[30] NIST Chemistry WebBook, http://webbook.nist.gov
[31] A. Yee, S.J. Morrison, H. Idriss, J. Catal. 186 (1999) 279.
[32] A. Erdohelyi, J. Raskó, T. Kecskés, M. Tóth, M. Dömök, K. Baán, Catal. Today 116
(2006) 367.
[33] W.J. Cai, F.G. Wang, E.S. Zhan, A.C. Van Veen, C. Mirodatos, W.J. Shen, J. Catal.
257 (2008) 96.
[34] M.C. Sanchez-Sanchez, R.M.N. Yerga, D.I. Kondarides, X.E. Verykios, J.L.G.
Fierro, J. Phys. Chem. A 114 (2010) 3873.
[35] E. Marton, J.M. Vohs, ACS Catal. 1 (2011) 1414.
[36] A.M. da Silva, L.O.O. da Costa, K.R. Souza, L.V. Mattos, F.B. Noronha, Catal.
Commun. 11 (2010) 736.
[37] L.V. Mattos, G. Jacobs, B.H. Davis, F.B. Noronha, Chem. Rev. 112 (2012) 4094.
[38] S.M. de Lima, A.M. da Silva, G. Jacobs, B.H. Davis, L.V. Mattos, F.B. Noronha,
Appl. Catal. B 96 (2010) 387.
[39] A. Olafsen, C. Daniel, Y. Schuurman, L.B. Råberg, U. Olsbye, C. Mirodatos, Catal.
Today 115 (2006) 179.
[40] H.F. Wang, Z.P. Liu, J. Am. Chem. Soc. 130 (2008) 1099.
[41] N. Laosiripojana, S. Assabumrungrat, Appl. Catal. B 60 (2005) 107.
[42] N. Laosiripojana, S. Assabumrungrat, S. Charojrochkul, Appl. Catal. A 327
(2007) 180.
Acknowledgments
[43] J.I. Di Cosimo, V.K. Díez, M. Xu, E. Iglesia, C.R. Apesteguía, J. Catal. 178 (1998)
499.
[44] J.I. Di Cosimo, C.R. Apesteguía, M.J.L. Ginés, E. Iglesia, J. Catal. 190 (2000) 261.
[45] M.B. Jensen, S. Morandi, F. Prinetto, A.O. Sjåstad, U. Olsbye, G. Ghiotti, Catal.
Today 197 (2012) 38.
Y.D. Li thanks NSF of China for the financial support under con-
tract numbers 21076150 and 21120102039.Y.D. Li’s work has been
also supported by the Program of Introducing Talents to the Uni-
versity Disciplines under file number B06006, and the Program for
Please cite this article in press as: G. Zeng, et al., Kinetic and process study of ethanol steam reforming over Ni/Mg(Al)O catalysts: The
initial steps, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.006