The influence of carbon laydown on selectivity in the hydrogenation of pentenenitriles over supported-nickel catalysts

Article abstract of DOI:10.1016/j.apcata.2010.06.036

Pentenenitriles contain two-reducible functionalities: a carbon-carbon double bond and a nitrile group, either of which may undergo hydrogenation during reaction. In this work we show how the deposition of hydrocarbonaceous material on the catalyst surface during pentenenitrile hydrogenation over 16 wt.% Ni/Al₂O₃ and 10 wt.% Ni/SiO₂ catalysts has a significant impact on the observed catalytic activity and selectivity. The role of carbon laydown in controlling catalytic performance in this system has been evaluated through activity measurements and mechanistic studies employing a Tapered Element Oscillating Microbalance (TEOM) and a conventional flow-through reactor. TEOM data indicating the deposition of carbonaceous material during reaction are correlated with kinetic analysis which provides a description of catalyst deactivation in terms of the deactivation of groups of active sites. Specifically five distinct active sites are shown to exist on Ni/Al ₂O₃ including a hydrogenation site on the support, which is not present in the case of Ni/SiO₂. The nature and strength of these sites are discussed. Furthermore, deuteration studies provide mechanistic insights suggesting that the hydrogenation reaction proceeds via a cyclic intermediate. The reported data identify a correlation between mass laydown on specific active sites and deactivation, thereby demonstrating the influence of hydrocarbonaceous deposits on selectivity. Both the location and the nature of such deposits are crucial in determining its influence on reaction.

Full text of DOI:10.1016/j.apcata.2010.06.036

Applied Catalysis A: General 384 (2010) 192–200
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
The influence of carbon laydown on selectivity in the hydrogenation of
pentenenitriles over supported-nickel catalysts
James McGregora,∗, Arran S. Canningb, Scott Mitchellb, S. David Jacksonb, Lynn F. Gladdena
a University of Cambridge, Department of Chemical Engineering and Biotechnology, New Museums Site, Pembroke Street, Cambridge CB2 3RA, UK
b Centre for Catalysis Research, WestCHEM, Department of Chemistry, University of Glasgow, Glasgow G12 8QQ, Scotland, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 9 April 2010
Received in revised form 16 June 2010
Accepted 17 June 2010
Available online 25 June 2010
Pentenenitriles contain two-reducible functionalities: a carbon–carbon double bond and a nitrile group,
either of which may undergo hydrogenation during reaction. In this work we show how the deposition
of hydrocarbonaceous material on the catalyst surface during pentenenitrile hydrogenation over 16 wt.%
Ni/Al2O3 and 10 wt.% Ni/SiO2 catalysts has a significant impact on the observed catalytic activity and
selectivity. The role of carbon laydown in controlling catalytic performance in this system has been eval-
uated through activity measurements and mechanistic studies employing a Tapered Element Oscillating
Microbalance (TEOM) and a conventional flow-through reactor. TEOM data indicating the deposition of
carbonaceous material during reaction are correlated with kinetic analysis which provides a description
of catalyst deactivation in terms of the deactivation of groups of active sites. Specifically five distinct
active sites are shown to exist on Ni/Al2O3 including a hydrogenation site on the support, which is not
present in the case of Ni/SiO2. The nature and strength of these sites are discussed. Furthermore, deutera-
tion studies provide mechanistic insights suggesting that the hydrogenation reaction proceeds via a cyclic
intermediate. The reported data identify a correlation between mass laydown on specific active sites and
deactivation, thereby demonstrating the influence of hydrocarbonaceous deposits on selectivity. Both
the location and the nature of such deposits are crucial in determining its influence on reaction.
© 2010 Elsevier B.V. All rights reserved.
Keywords:
Catalytic hydrogenation
cis-2-Pentenenitrile
trans-3-Pentenenitrile
Kinetic analysis
TEOM
Alumina
Coke
Deactivation
1. Introduction
alyst less strongly than the nitrile moiety [14]. Despite this, under
hydrogenation conditions the carbon–carbon double bond is more
Selectivity is one of the key drivers in catalysis research: improv-
ing selectivity has direct economic and environmental benefits
associated with less waste and reduced material and energy inputs.
The single largest application of catalysis in chemical manufactur-
ing is catalytic hydrogenation, and hence the selective reduction of
one-functionality within a multi-functional unsaturated molecule
is a subject that receives much attention [1–4]. Hydrogenation reac-
tions over supported-metal catalysts are, in general, accompanied
by the deposition of carbonaceous material on the catalyst surface
[5]. This deposited coke can impart significant influence on cat-
alytic activity and selectivity most commonly through deactivation
of active sites [6–8], but also through action as a hydrogen-transfer
agent, or direct involvement in catalytically active or selective sites
[9–11]. In addition to simple chemical conversion data, kinetic anal-
ysis can greatly assist in elucidating the processes occurring on
the catalyst surface during reaction [12–14]. For instance, kinetic
studies on propenenitrile and cyclohexane hydrogenation have
shown that the alkene functionality adsorbs to a Raney nickel cat-
reactive than the carbon–nitrogen bond.
Following from the observation that nitrile functionalities are
more strongly adsorbed than olefinic functionalities, it might be
expected that in longer chained alkenenitrile molecules, where
the two moieties are a significant distance apart, high selectivi-
ties towards hydrogenation of the nitrile group could be obtained.
This result has indeed been observed experimentally [14–16].
In contrast, in shorter-chained molecules such as propenenitrile
and butenenitriles selectivity is directed towards reduction of the
carbon–carbon double bond [14,17]. Recent studies have focused
on the hydrogenation of five-carbon chain species: penteneni-
triles. Kukula and Koprivova investigated the hydrogenation of
cis-2-pentenenitrile (C2PN) and trans-3-pentenenitrile (T3PN) in
the liquid-phase over a variety of catalysts including Raney nickel
[15]. In the case of the conjugated isomer (C2PN) the primary prod-
uct was n-pentanenitrile (PN) while for the non-conjugated isomer,
trans-3-pentenenitrile (T3PN), the major primary product is trans-
3-pentenamine. Elsewhere, the formation of PN from both C2PN
and T3PN over Ni/Al2O3 has also been observed [18].
We have previously conducted independent investigations
into gas-phase pentenenitrile hydrogenation over an alumina-
supported-nickel catalyst employing a Tapered Element Oscillating
∗
Corresponding author. Tel.: +44 0 1223 330134; fax: +44 0 1223 334796.
E-mail address: jm405@cam.ac.uk (J. McGregor).
0926-860X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2010.06.036

J. McGregor et al. / Applied Catalysis A: General 384 (2010) 192–200
193
Table 1
Summary of our previous studies on the hydrogenation of pentenenitriles.
Reactant
C2PN
Catalyst
Ni/Al2O3
Technique
TEOM
Parameters investigated
Key conclusions
Ref.
Mass laydown during reaction
and variation in catalytic
Selectivity shift from
PA to PN with
[18]
activity with time-on-stream
increasing carbon
laydown
C2PN
C2PN
C2PN
Ni/Al2O3
Ni/Al2O3
␪-Al2O3
Flow reactor
Flow reactor
Flow reactor
Catalytic performance with
time-on-stream
Selective poisoning study using
CO and NH3
Catalytic performance with
time-on-stream
Selectivity shift from
PA to PN
[7]
[7]
[7]
At least 3 distinct
active sites exist
Catalytic activity of
support; formation of
ammonia
T3PN followed by C2PN
Ni/Al2O3
TEOM
Mass laydown during reaction
and variation in catalytic
activity with time-on-stream
T3PN deactivates
catalyst towards PN
formation
[18]
Microbalance (TEOM) [18] and an atmospheric flow-through reac-
tor [7]. Although separate, these two investigations reported very
similar results. For example, a significant variation in selectiv-
ity with time-on-stream during C2PN hydrogenation over 16 wt.%
Ni/Al2O3 was observed in both cases: initially the dominant prod-
uct is n-pentylamine (PA); as time-on-stream increases selectivity
shifts towards the product of selective hydrogenation of the
carbon–carbon double bond, PN. As the reaction proceeds further
the catalyst loses activity with PN production decreasing. Stud-
ies on the pure support, ␪-Al2O3, have shown that even in the
absence of Ni initial activity is observed towards PA formation. In
addition to catalytic selectivity and activity data, measurements of
mass change due to adsorption and carbon laydown over Ni/Al2O3
were also recorded in situ. The measured mass laydown profile
was correlated with changes in selectivity over time [18]. Specifi-
cally, during C2PN hydrogenation an initial rapid period of carbon
laydown was associated with a decrease in the production of the
total hydrogenation product, PA, while a second, slower, period of
carbon deposition was associated with deactivation towards the
selective hydrogenation product, PN. Our preliminary hypothesis
associated these two periods of mass laydown with two groupings
of active site: high energy sites which undergo rapid deactiva-
tion and lower energy sites which deactivate at a slower rate. The
adsorption of T3PN in H2 was also shown to deactivate certain
surface sites [18]. Exposing the catalyst for a short time (1 h) to
the non-conjugated isomer subsequently prevented the selective
hydrogenation of C2PN to PN. PA formation however was not deac-
tivated, the total hydrogenation product being the major product
over the cumulative 7 h timescale of reaction. Increasing produc-
tion of PA was observed from 3 h onwards, corresponding to a
decrease in the mass of the retained carbonaceous overlayer.
A more detailed description of the catalytically active sites
was provided through selective poisoning studies [7]. In summary,
adsorption of CO on metal sites prevented the formation of PN,
while adsorption of NH3 (which will adsorb on acid sites) poisons
the isomerisation reaction. As a result of these poisoning studies
a model describing three-distinct active sites was developed [7].
This however is insufficient to explain subsequent observations
reported in the present work.
experiments have been conducted to elucidate the mechanism of
reaction.
2. Experimental
2.1. Materials and catalyst characterisation
Al2O3- and SiO2-supported-nickel catalysts, in addition to the
pure support materials have been investigated. Ni/Al2O3 (HTC
400, Johnson Matthey) is characterised by a BET surface area of
106 m2 g−1, a pore volume of 0.39 cm3 g−1 and a metal loading of
16 wt.%). The support is ␪-Al2O3 (Johnson Matthey, BET surface area
101 m2 g−1, pore volume 0.42 cm3 g−1). Prior to use both the cat-
alyst and support were activated in situ by reduction in flowing
H2 at ≥523 K for 2 h. Previous TPR studies have demonstrated that
this temperature is sufficient to reduce the catalyst [19]. Ni/SiO2 is
characterised by a BET surface area of 132 m2 g−1, a pore volume
of 0.64 cm3 g−1 and a metal loading of 10 wt.%. The SiO2 support
is commercially available Aerolyst 3039 (Degussa) with a surface
area of 250 m2 g−1 and a pore volume of 0.9 cm3 g−1
.
2.2. Reaction studies
Catalytic reaction studies have been carried out in two differ-
ent reactors: an atmospheric flow-through reactor and a TEOM. In
both cases the as-received catalyst was ground and sieved prior
to use. For the atmospheric flow-through reactor a particle size
of 200–400 ␮m and a total of mass of 0.25 g was used, while the
corresponding values for the TEOM were 75–90 ␮m and 35 mg.
Both experimental set-ups have previously been described in detail
[7,18], while the operating principles and applications of the TEOM
have been reported by Zhu et al. [20] and Chen et al. [21]. Areas in
which the TEOM has previously been employed include adsorption,
diffusion, redox kinetics, reaction and deactivation due to carbon
deposition [21–25]. In the present study the TEOM was operated at
1.1 bar throughout catalyst pre-treatment and reaction. Corrections
for pressure differences between the pre-treatment and reaction
gas streams have been corrected for by conducting identical exper-
iments over a low-surface area ␣-Al2O3 (Fluka, BET surface area
0.10 m2 g−1) upon which only negligible adsorption occurs [18].
A key advantage of the TEOM is the ability to monitor mass lay-
down with high resolution (␮g) as a function of time-on-stream
(with time resolution ∼0.1 s), while simultaneously monitoring the
composition of the outlet stream, in this case using gas chromatog-
raphy (GC) (Hewlett-Packard 6890, FID detector). By reverting from
the reactant stream to an inert flow (He) the mass retained on
the catalyst surface, in the form of hydrocarbonaceous deposits or
coke, can be determined. In order to monitor the build-up of such
deposits with time-on-stream during TEOM studies the reactant
A summary of our previous studies indicating the reaction sys-
tems investigated and the principle conclusions drawn is shown in
Table 1. The aim of the present study is to extend these earlier stud-
ies and to investigate the parameters which influence selectivity in
the hydrogenation of C2PN and T3PN. In particular, understanding
the extent and role of carbon laydown and the nature of the cat-
alytic active site(s) are key parameters in this regard. The role of the
support in determining selectivity and deactivation has also been
investigated. Furthermore, in order to fully understand the origin of
observed selectivity and the origin and rate of catalyst deactivation,

194
J. McGregor et al. / Applied Catalysis A: General 384 (2010) 192–200
flow (50 cm3 min−1, H2:pentenenitrile ratio 12.5:1) was, after 1 h,
replaced by a flow of He (50 cm3 min−1) for 1 h, before reverting to
reactant flow. Seven such cycles were conducted, providing seven
1 h periods of pentenenitrile/H2 flow through the catalyst, each sep-
arated by 1 h of He flow. In both reactors the catalyst was reduced
by hydrogen in situ prior to use (Section 2.1) before being cooled to
a reaction temperature of 373 K. Pentenenitriles were introduced to
the catalyst via a vapouriser containing either C2PN (98%, Aldrich)
or T3PN (95%, Aldrich). In the atmospheric flow-through reactor
the reaction was monitored via GC (Varian 3300, TCD) for a period
of 3 h on-stream. The reactant flow rate was 50 cm3 min−1, with a
H2:pentenenitrile ratio 3.6:1. Samples were then cooled to room
temperature prior to temperature programmed oxidation (TPO)
studies.
3. Results
In order to more fully understand the factors that influence
selectivity and deactivation in the hydrogenation of C2PN, and of
multi-functional molecules in general, experiments were designed
as follows. Firstly, in order to further determine the number and
nature of the catalytically active sites present on Ni/Al2O3, the
hydrogenation of C2PN over the pure ␪-Al2O3 support is inves-
tigated. The roles played by the metal and support sites are
additionally probed by changing the support material from alumina
to silica, hence C2PN hydrogenation studies have been conducted
over Ni/SiO2 and SiO2. Carbon laydown on catalytic sites is corre-
lated with changes in selectivity and catalyst deactivation, thereby
indicating a dynamic catalytic system under reaction conditions.
The hydrogenation of T3PN and the sequential hydrogenation of
T3PN and C2PN provide insights into the nature of the deactivating
carbonaceous deposits and their interaction with catalytic sites as
well as illustrating differences between the behaviour of the con-
jugated and non-conjugated pentenenitrile isomers. The location
and character of carbonaceous material deposited during reaction
is further characterised through TPO measurements.
In addition to probing the nature of the catalytically active
sites, studies were undertaken in order to determine the reaction
mechanism. These experiments employed D2 in place of H2. By con-
sidering the results of these deuteration studies with regard to the
hydrogenation experiments described above it is possible to pro-
posea detailedreactionmechanism. Togetherthese studiesidentify
the influence of carbon laydown on different active sites and hence
on catalytic selectivity and activity.
Fig. 2. Relative quantity of species detected by GC (FID) during hydrogenation of
C2PN over Al2O3 at 373 K, 1.1 bar, 50 cm3 min−1 flow rate and H2:pentenenitrile
ratio 12.5:1. Species detected are C2PN ( ), cis-2-pentene ( ), T3PN ( ) and trans-
2-pentene ( ). Curves are fitted only as a guide to the reader.
Ni/SiO2 and SiO2 has been investigated. Additionally the hydro-
genation of T3PN over Ni/Al2O3 and ␪-Al2O3 and the sequential
reaction of T3PN followed by C2PN have been carried out in order
to investigate the influence of deposited carbonaceous material on
specific active sites and therefore on catalytic performance. Finally
TPO studies provide an indication as to the number and strength of
the adsorption sites present.
3.1.1. C2PN + H2 over Ni/Al2O3 and ꢀ-Al2O3
Hydrogenation studies, complementary to those carried out
over Ni/Al2O3, have been conducted over the pure support in order
to deconvolve the roles of the Ni and Al2O3. TEOM data from the
reaction of C2PN with H2 over ␪-Al2O3 shown in Fig. 1a reveal that
the levels of mass laydown observed are significantly lower than
those observed over Ni/Al2O3. During the first 1 h pulse the level
of adsorbed material reaches a maximum of 45.2 ␮g/mgcat over
Ni/Al2O3 but only 18.3 ␮g/mgcat over the support. Note that this
does not necessarily imply that the additional carbonaceous mate-
rial formed over Ni/Al2O3 resides solely on Ni sites as spill-over
of this material onto the support may occur. The reaction pro-
file, for C2PN hydrogenation over ␪-Al2O3 as a function of time,
is shown in Fig. 2. This profile demonstrates surprising activity of
the ␪-Al2O3 support. Initially C2PN and cis-2-pentene are detected
in approximately equal amounts. At increasing times-on-stream
activity towards formation of the olefin begins to decrease. After
four 1 h pulses, cis-2-pentene represents only ∼20% of the outlet
3.1. Nature of the catalytically active sites
In order to investigate the nature of the catalytically active
sites present on Ni/Al2O3 the hydrogenation of C2PN over ␪-Al2O3,
Fig. 1. Quantity of mass adsorbed on catalyst during reaction at 373 K, 1.1 bar, 50 cm3 min−1 flow rate and H2:pentenenitrile ratio 12.5:1. Mass measurement is taken at the
end of each 1 h pulse of reactant + H2. (a) Data for C2PN hydrogenation over Ni/Al2O3 ) and Al2O3 ). (b) Data for T3PN hydrogenation over Ni/Al2O3 ) and Al2O3 ).
(
(
(
(

J. McGregor et al. / Applied Catalysis A: General 384 (2010) 192–200
195
Fig. 3. Reaction profile for C2PN hydrogenation over Ni/SiO2 in the atmospheric
flow-through reactor at 373 K, 50 cm3 min−1 flow rate and H2:pentenenitrile ratio
3.6:1, showing conversion ( ) and fractional composition of the gas-phase products,
Fig. 4. Reaction profile T3PN hydrogenation (373 K. 1.1 bar, 50 cm3 min−1 flow rate
and H2:pentenenitrile ratio 12.5:1) over Ni/Al2O3 in the TEOM. Species detected
are: PA ( ), T3PN ( ) and species assigned to cis-3-pentenenitrile ( ) and trans-
3-pentenamine ( ).
namely PN ( ), PA ( ), T2PN (+) and T3PN ( ).
stream. The formation of cis-2-pentene suggests that a hydrogenol-
ysis or similar reaction is occurring leading to the scission and
hydrogenation of C2PN producing the alkene and ammonia. Indeed,
ammonia was detected in the outlet gas stream from the TEOM dur-
ing hydrogenation studies both over the pure support and over the
catalyst in the present study. Ammonia has also previously been
detected during the reaction of C2PN and H2 in the atmospheric
flow-through reactor, although no cis-2-pentene was observed in
that work [7].
towards both the total hydrogenation and isomerisation products
decreases rapidly, with >90% of the detected exit gas comprising
the initial reactant. The remaining activity is towards the formation
of small amounts of a species believed to be trans-3-pentenamine.
The selective hydrogenation of the nitrile functionality can be ratio-
nalised by considering the high mass laydown that occurs in this
reaction, which may sterically hinder the formation of an appropri-
ate intermediate to facilitate C C hydrogenation. As a result, only
the nitrile functionality, adsorption of which is favoured kinetically,
undergoes reaction. Atmospheric flow-through reactor studies pro-
vide complementary data to TEOM measurements, confirming that
PA is the major product produced in the early stages of reaction and
that no PN, the product of selective C C hydrogenation, is observed.
TEOM data for the reaction of T3PN and H2 with ␪-Al2O3 are
shown in Fig. 1b. As with C2PN, a significantly lower mass uptake
occurs than is observed over Ni/Al2O3. The decrease in mass lay-
down is greater than twofold, the retained overlayer totalling
31.3 ␮g/mgcat after seven 1 h pulses as compared to 75.4 ␮g/mgcat
over Ni/Al2O3. Unlike C2PN however, T3PN is largely unreactive
over the support. Greater than 95% of the detected material from the
second pulse onwards comprises T3PN. During the first pulse small
amounts of the material assigned above as trans-3-pentenamine
are observed, reaching a maximum of 25% of the exit gas compo-
sition after 5 min. After 30 min however this value has fallen to
<5%.
3.1.2. C2PN + H2 over Ni/SiO2 and SiO2
Considering the hydrogenation of C2PN over Ni/Al2O3, further
insights may be gained into the roles played by the metal and by
the ␪-Al2O3, by varying the nature of the support. To this end, we
describe studies of C2PN hydrogenation over Ni/SiO2, and over the
SiO2 support, in an atmospheric flow-through reactor. In contrast
to ␪-Al2O3, SiO2 is apparently inert and shows no evidence of PA
formation. The main products formed, with a combined selectivity
of ∼90%, are isomerisation products, namely trans-3-pentenenitrile
and trans-2-pentenenitrile. In contrast, Ni/SiO2, as with Ni/Al2O3,
does produce PA in the initial stages of reaction. The overall reac-
tion profile for C2PN hydrogenation over Ni/SiO2 is shown in Fig. 3.
PA is produced initially, however the catalyst then rapidly deac-
tivates towards this reaction with PN and isomerisation products
dominating from 40 min onwards. This is a very similar reaction
profile to that observed when Ni/Al2O3 is employed as the catalyst,
as will be discussed in Section 4.3.
3.1.4. Sequential hydrogenation of T3PN followed by C2PN
3.1.3. T3PN + H2 over Ni/Al2O3 and ꢀ-Al2O3
The role of the carbonaceous overlayer deposited by T3PN, as
described in Section 3.1.3, on the hydrogenation of C2PN can be
probed through sequential hydrogenation studies. This in turn pro-
vides insights into the interaction of carbonaceous material with
catalytically active sites, which plays a key role in directing catalytic
selectivity. To this end, studies of the sequential hydrogenation
of T3PN followed by C2PN have been carried in our atmospheric
flow-through reactor over Ni/Al2O3 and over ␪-Al2O3. After 30 min
of T3PN hydrogenation, the flow was switched to helium to flush
the reactor and then, when no material was detected in the efflu-
ent stream, the flow was switched to C2PN in 5% H2/N2. Over
Ni/Al2O3, as with the catalyst which was not first exposed to T3PN,
PA was the initial product of C2PN hydrogenation. In contrast to
the untreated catalyst however, instead of decaying rapidly PA was
a major product for over 1 h. Considering ␪-Al2O3, C2PN hydro-
TEOM and atmospheric flow-through reactor experiments were
carried out for T3PN in the same manner as for C2PN in order
to investigate the influence of adsorbate structure on reactivity,
carbon laydown and deactivation. TEOM studies demonstrate a
broadly similar pattern of mass laydown over Ni/Al2O3 as for C2PN,
however as shown in Fig. 1b, significantly greater mass laydown is
obtained with T3PN. After seven 1 h pulses T3PN hydrogenation
results in a retained carbonaceous overlayer of 75.4 ␮g/mgcat as
opposed to 51.1 ␮g/mgcat for C2PN hydrogenation. The reaction
profile, shown in Fig. 4, is quite different for the trans isomer than
was observed for the conjugated compound. Initially, as with C2PN,
the non-selective hydrogenation product, i.e. PA, is the main species
observed, alongside smaller amounts of the products of isomeri-
sation reactions. After ∼20 min time-on-stream however, activity

196
J. McGregor et al. / Applied Catalysis A: General 384 (2010) 192–200
Fig. 5. Hydrogenation of C2PN over Al2O3 first exposed to T3PN in the atmospheric
flow-through reactor at 373 K, 50 cm3 min−1 flow rate and H2:pentenenitrile ratio
3.6:1 as a function of time. Conversion ( ) is shown on the left-hand axis and the
Fig. 6. Mass spectrometry data from TPO experiments performed on Ni/Al2O3 after
use for C2PN hydrogenation: (a) sample from initial, rapid, deactivation period; (b)
sample from second, slower, deactivation period.
fractional composition of the gas-phase products, namely PA ( ) and C2PN (
shown on the right-hand axis.
)
genation reactions, bar the substitution of D2 for H2. Reactant
C2PN was not deuterated. The species evolved from the reactor
in these experiments comprised unreacted cis-2-pentenenitrile,
n-pentylamine, PN, T3PN and trans-2-pentenenitrile in the same
proportions as for the hydrogenation experiments. The unreacted
C2PN evolved was not deuterated in any position: that is, no
H/D exchange took place. In contrast all reaction products, of
both deuteration and isomerisation processes, were observed to
be deuterated in all positions. These results indicate all hydrogen
atoms undergo H/D exchange, thereby implying that all five-carbon
atoms must interact with the catalyst surface during the hydro-
genation reaction thereby providing direct insight in the catalytic
reaction mechanism, as discussed in Section 4.1.
genation over the T3PN treated alumina resulted in a maximum
conversion of ∼25%, achieved in the early stages of reaction (Fig. 5).
Carbonaceous deposits however, represented the major product
formed under these conditions with ∼68% of the reactant con-
sumed being converted to such deposits based on a simple mass
balance. The remaining conversion was to PA. No isomerisation
was detected. After ∼25 min production of PA ceased, with a cor-
responding decrease in conversion observed. The conversion did
not drop to zero for another 2 h, indicating that carbon laydown
continued during this time.
3.1.5. TPO studies
In order to more clearly understand carbon deposition and to
assist in identifying the number and nature of catalytically active
sites TPO studies on catalysts after reaction have been under-
taken. Kinetic analysis of the rate of catalyst deactivation in C2PN
hydrogenation has previously identified two deactivation stages,
and these are discussed in detail in Section 4.3. Briefly, an ini-
tial, rapid, deactivation period is observed before isomer formation
occurs. This is followed by a second, slower, period of deactivation
where isomerisation of C2PN occurs. TPO analysis was therefore
conducted on two catalyst samples, one from each stage of deacti-
vation. Both of these samples were removed from the atmospheric
flow-through reactor. The TPO experiments were run between 298
and 873 K and are shown in Fig. 6. The catalyst sample from the
rapid deactivation phase showed CO2 desorption peaks at 519,
541, 557, 571 585 and 591 K, while the catalyst sample from the
slower deactivation phase displayed desorption peaks at 531, 549,
570, 586, 591, 602 and 618 K. In general the TPO profile for the
catalyst sample from the second, slower, deactivation period has
larger peaks. Furthermore, more peaks at higher temperatures are
observed for this sample than for that characterising the first, more
rapid, deactivation phase.
4. Discussion
Experimental studies have been carried in order to probe the
number and nature of the catalytically active sites, including the
influence of carbon deposition upon those sites, and to elucidate
the reaction mechanism. In order to understand the requirements
for an active site it is first necessary to determine the reactions that
occur on them. To this end the elucidated reaction mechanism is
reported in Section 4.1, while the active sites present are described
in Section 4.2. Finally, the role of carbonaceous material deposited
on the catalyst surface during reaction is discussed in Section 4.3.
4.1. Hydrogenation reaction mechanism
Deuteration studies reported herein have provided evidence in
support of a reaction mechanism for the hydrogenation of C2PN
over Ni/Al2O3. The reaction of C2PN with D2 over Ni/Al2O3 yielded
reaction products deuterated in all positions, with unreacted C2PN
being fully protiated. The observed production of fully deuterated
products would not be possible if the molecule reacted in a flat,
linear geometry as the C–H bonds in the methyl group would
not interact with the surface and hence no deuteration of these
would occur. This therefore implies that the reaction proceeds via
a cyclic intermediate such as that shown in Fig. 7. This mecha-
nism is consistent with previous hypotheses based on the reaction
of H2 (as opposed to D2) with C2PN over Ni/Al2O3 [7] and of H2
with butenenitrile over Raney-Ni [14,27]. In the proposed mecha-
nism illustrated in Fig. 7, pentenenitrile co-ordinates to the surface
3.2. Probing the catalytic mechanism
The use of D2 in place of H2 is a well established technique
in probing hydrogenation reaction mechanisms [26]. Therefore, in
addition to the hydrogenation of pentenenitriles, the reaction of
deuterium with C2PN over Ni/Al2O3 has also been investigated.
These studies were conducted in the atmospheric flow-through
reactor described earlier under identical conditions as per hydro-

J. McGregor et al. / Applied Catalysis A: General 384 (2010) 192–200
197
proposed previously [7,18]. On the basis of selective poisoning of
metal sites with carbon monoxide and acid sites with ammonia, a
three-site model was suggested to describe the surface sites active
in the hydrogenation of C2PN. Those sites comprised: a nickel site
capable of forming PN; a lower energy isomerisation site on the
␪-Al2O3 support and finally, a high energy support site which catal-
yses the total hydrogenation reaction, producing PA. The presence
of a support site with such high hydrogenation activity is surprising.
However, this same ␪-Al2O3 support has previously demonstrated
activity for the hydrogenation of trans-2-pentene to n-pentane [29].
In that work 13C-DEPT NMR spectroscopy provided a direct, in situ
probe of the activity of ␪-Al2O3. While this three-site hypothesis
provided an adequate description of results previously described,
it is unable to account for the observations reported in the present
study. Specifically, TPO studies identify either six or seven dis-
tinct carbon species depending on the stage at which the reaction
is stopped. While there evidently need not be a 1:1 relationship
between TPO peaks and active sites this does perhaps hint at the
need to expand upon the previous three-site model. Additionally,
the previous three-site model is unable to explain either the for-
mation of PA over Ni/SiO2 or the poisoning of activity towards
PN production through pre-treatment with T3PN. As such, a five-
site model is proposed with regard to hydrogenation of C2PN over
Ni/Al2O3.
Fig. 7. Proposed interaction of C2PN with the catalyst surface leading to formation
of a ␲-allyl species which allows hydrogen exchange and is an intermediate in the
hydrogenation process.
The five sites determined are: (i) moderate energy Ni site for PN
formation; (ii) moderate energy Al2O3 site for C2PN isomerisation;
iii) high energy Al2O3 site for PA formation; (iv) high energy Ni site
for PA formation; and (v) high activity Ni site for PN formation. The
five sites have been determined as follows:
through the nitrile functionality, as is favoured kinetically. In order
to selectively hydrogenate the carbon–carbon double bond, it is
necessary for the reaction to proceed via a cyclic intermediate.
Hydrogenation then takes place by a concerted hydrogen transfer.
That the alkene moiety does not adsorb directly to an active nickel
centre is supported by the observation that no skeletal isomerisa-
tion of the carbon backbone is observed. Following rearrangement
to a cyclic geometry the adsorbed nitrile forms a ␲-allyl species
which allows for the interaction of carbons 1–4 with the catalyst
surface. This in turn is in equilibrium with a second ␲-allyl species
which permits interaction of the final carbon atom, C5. The final
step is necessary in order to produce reaction products deuterated
in all positions, as is observed experimentally. Catalytically active
sites are also present on the ␪-Al2O3 (Section 4.2) and contribute
to the activity the Ni/Al2O3 catalyst. The products from these sites
must also be fully deuterated. A potential structure of the active alu-
mina site and a hydrogenation mechanism are discussed in Section
4.2.
Hydrogenation of non-conjugated alkenenitriles, namely 3-
butenenitrile and T3PN has also been proposed to proceed via
a cyclic intermediate [7,14]. This is consistent with the results
obtained from the atmospheric flow-through reactor in the present
work, however the lack of the selective hydrogenation product,
PN, in TEOM experiments is surprising. This however could be
assigned to a combination of the lower reactivity of T3PN and
variations in experimental conditions such as the higher gas flow
rate-to-catalyst mass ratio employed in TEOM studies. As a result
the residence time may be too low for the less reactive T3PN to
hydrogenate on the low energy sites responsible for C C hydro-
genation, as discussed in Section 4.2. Lower activity of T3PN cf.
C2PN has also been observed over ruthenium-complex catalysts
ꢀ Site (i) is identified through the selective poisoning of metal sites
by CO which has previously been shown [7] to prevent formation
of PN formation thereby confirming that this reaction takes place
on metal sites. Through reaction studies these sites are shown
to deactivate slowly thereby identifying site (i) as a moderate
energy site.
ꢀ Site (ii) is identified through NH3 adsorption studies which
result in the poisoning of acid sites on the support surface
and the cessation of the isomerisation reaction [7]. This iso-
merisation site is thereby confirmed as an acid site. Elsewhere,
studies of the liquid-phase hydrogenation of pentenenitriles
over Raney-Ni did not observe the formation of isomerisation
products, lending weight to the concept that this reaction takes
place only over support sites [15].
ꢀ Site (iii) is confirmed through the observation of the formation of
PA over the pure alumina support, in the absence of nickel. Nei-
ther adsorption of CO nor NH3 disrupts the activity of this site,
hence it can be concluded that it is neither metallic nor acidic
in nature. A tentative reaction mechanism and corresponding
reaction site are shown in Fig. 8. This mechanism once again
invokes the cyclic intermediates discussed in Section 4.1; in
this case adsorption takes place at a bridging hydroxyl group
on the support surface. Further work is required to determine
whether this mechanism is correct and the nature of the inter-
mediate steps. Any mechanism must also take into account that
all the deuteration products over Ni/Al2O3 are deuterated in all
positions.
in water [28]. One reason for the differing activity towards C
C
hydrogenation between the two isomers may be the influence of
conjugation. Conjugation to nitrile groups, such as in the case of
cis-2-pentenenitrile is known to activate olefinic groups resulting
in increased hydrogenation rates [15,27,28].
ꢀ Hydrogenation experiments over Ni/SiO2 and SiO2 have eluci-
dated the presence of site (iv). While ␪-Al2O3 clearly catalyses
the total hydrogenation of pentenenitriles, as reported in Sec-
tion 3.1, SiO2 does not. Despite this, PA is produced in the early
stages of C2PN hydrogenation over Ni/SiO2, before rapid deacti-
vation. This indicates that Ni is also capable of catalysing the total
hydrogenation reaction, identifying site (iv). Similarly, Kukula
and Koprivova [15] have previously observed the formation of
PA over Raney-Ni, starting from C2PN. That this is a high energy
4.2. Nature and number of catalytically active sites
That a number of distinct active sites, of differing strength and
consequently differing activity, are present on Ni/Al2O3 has been

198
J. McGregor et al. / Applied Catalysis A: General 384 (2010) 192–200
Fig. 8. Proposed structure of active site for the hydrogenation of C2PN to PA over Al2O3. The reaction mechanism is also shown.
site is confirmed by its rapid deactivation: if deactivation of this
site did not occur it is likely that PA would be formed from
the hydrogenation of PN, as has been reported previously for
the liquid-phase reaction of PN employing 1-propanol as the
H-transfer agent [30].
where CA0 represents the initial reactant concentration, CA is the
concentration of the reactant with time, k is the reaction rate,
tw represents weight time (gcat−1 hlfluid), kd is the rate of deacti-
vation and t is time. From a plot of ln[(CA0/CA) − 1] vs time the
rate of deactivation can be evaluated. Fig. 9b shows the profile
for the deactivation relationship. From this model it is clear that
the point where there is a dramatic change in the deactivation
rate coincides with the formation of trans-2-pentenenitrile and
T3PN isomers. This two-stage deactivation model is in excellent
agreement with the TEOM mass change analysis. The initial rapid
(0.0516 min−1) deactivation stage corresponds to an initial rapid
build-up of carbonaceous material. During this time the catalyst
deactivates towards PA formation providing a strong indication
that the sites responsible for this reaction (sites (iii) and (iv) dis-
cussed in Section 4.2) are deactivated directly through carbon
laydown. The second, slower (0.0005 min−1), deactivation phase
corresponds to the slower period of carbon accumulation on the
surface and is associated with the deactivation of the catalyst
toward PN production, described in Section 4.3. Again, this deac-
tivation is at least in part a result of carbon laydown. This second
period corresponds with the observation of isomerisation products.
In addition to quantifying the rate of mass laydown and its cor-
relation with the deactivation of groups of active sites, there is
evidence as to the nature of the carbonaceous material deposited.
Ammonia has been clearly demonstrated to poison acid sites on the
support thereby inhibiting the isomerisation reaction. It is likely
that PA, produced in the early stages of reaction, behaves in a sim-
ilar way. The surface affinity of the amine will be comparable to
that of ammonia: this strong surface affinity arises from the inter-
action of the lone pair present in the sp3 hybridised orbital with
the catalyst [32]. This is consistent with the observation that iso-
merisation products are not formed simultaneously with PA. The
switch in selectivity of the alumina support from PA production to
isomerisation can therefore be explained in terms of the desorp-
tion of PA retained on active acid sites, making them available for
reaction. A similar phenomenon has previously been observed in
TEOM studies where a decrease in the mass of a retained hydro-
carbonaceous overlayer was correlated with product desorption,
as observed by GC [18]. Further to the poisoning role likely played
by PA, ammonia is also produced directly during reaction. Studies
on the pure ␪-Al2O3 support material demonstrated the formation
of ammonia and cis-2-pentene from C2PN, indicating that the sup-
port is active for this reaction in addition to the total hydrogenation
of C2PN. While cis-2-pentene is not observed in the reaction prod-
ucts over Ni/Al2O3 it is likely that Al2O3 continues to catalyse this
reaction, forming ammonia. The hydrocarbon fragment, which des-
orbs as the alkene in the absence of Ni, may instead be retained as
a tightly bound hydrocarbon fragment on the Ni surface, further
ꢀ The final active site, site (v), is a high activity site capable of the
hydrogenation of cis-2-pentenenitrile to PN. It is these higher
activity sites which are deactivated through exposure of the cat-
alyst to T3PN in hydrogen prior to reaction of cis-2-pentnenitrile
(Section 3.2). The lower energy active sites, (i), remain unaf-
fected.
A more detailed description of the kinetics of deactivation of
these sites, and hence their assignment as “high” or “moderate”
energy is provided in Section 4.3.
4.3. Deactivation and influence of carbon laydown
In light of the identification of five distinct active sites on
Ni/Al2O3 (Section 4.2) it is important to consider that each of these
will likely show differing deactivation kinetics based on the adsorp-
tion and reaction processes occurring upon on them. That adsorbed
material is retained on the catalyst surface in the form of carbona-
ceous material is confirmed through TPO studies of Ni/Al2O3 after
reaction. The extent of this carbon laydown during C2PN hydro-
genation has been quantified through TEOM analysis indicating
that 51.1 ␮g/mgcat of material is retained on the catalyst surface
after seven 1 h reaction periods (Fig. 1a). Additionally, the rela-
tively high level of mass laydown observed on the pure support
(13.6 ␮g/mgcat at the end of the reaction) indicates that much of
the retained carbon on Ni/Al2O3 may be present on alumina sites.
Carbon laydown is in turn associated with a change in product
selectivity and deactivation of the catalyst. That different surface
sites exhibit different behaviour with respect to carbon laydown
is evidenced by the apparent 2-component behaviour observed in
TEOM data (Fig. 9a), where an initial, rapid, period of mass increase
is followed by a second, slower, period beginning after ∼2 h (in
total) of reaction. The extent of C2PN loss to the catalyst in the atmo-
spheric flow reactor is in agreement with the TEOM results: initially
there is significant laydown with the extent decreasing with time-
on-stream. The deactivation profile is here examined in more detail
using a deactivation model derived for the analysis of data for a
constant flow/mixed/batch reactor with independent deactivation
[31]:
ꢀꢁ
ꢂ
ꢃ
CA0
CA
ln
− 1 = ln(ktw) − kdt,

J. McGregor et al. / Applied Catalysis A: General 384 (2010) 192–200
199
Fig. 9. (a) TEOM data showing mass laydown (blue) and species detected by GC during TEOM studies of the hydrogenation of C2PN over Ni/Al2O3 at 373 K, 1.1 bar, 50 cm3 min−1
flow rate and H2:pentenenitrile ratio 12.5:1. Species detected are PA ( ), C2PN ( ) and PN ( ). Adapted from Fig. 6(a) and (b) [18]. (b) Deactivation kinetics of C2PN
hydrogenation over Ni/Al2O3 as derived from atmospheric flow-through reactor studies; 50 cm3 min−1 flow rate and H2:pentenenitrile ratio 3.6:1. The two deactivation
stages mirror the two mass laydown stages observed in (a). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of
the article.)
contributing to catalyst deactivation. This mechanism is indicated
schematically in Fig. 10. A closely related category of compounds,
alkylamines, are well known to behave in a similar manner, under-
going hydrogenolysis at the C–N bond in the presence of hydrogen
and an appropriate catalyst [32–35]. As with C2PN hydrogenation,
these reactions produce ammonia and either a gas-phase hydro-
carbon species, or a strongly adsorbed surface residue which can
poison catalytically active sites [32,33,36,37]. Nickel is known to
promote such reactions through the weakening of the C–N bond
upon adsorption [38,39].
Studies of the sequential reaction of T3PN followed by C2PN
also reveal a greater understanding of the nature of the carbona-
ceous residues implicated in catalyst deactivation. Data from both
TEOM and atmospheric flow-through reactor studies over Ni/Al2O3
show that the hydrocarbonaceous overlayer deposited as a result
of T3PN hydrogenation prevents, with a degree of selectivity, the
transformation of C2PN to PN. Correspondingly, no PN formation is
observed during T3PN hydrogenation. Specifically it can be deduced
that T3PN deactivates both the high energy Ni site (site (v), Section
4.2) and the Al2O3-based isomerisation site (site (ii), Section 4.2).
This latter point is also evidenced from studies of the sequential
reaction over the pure support where, in contrast to the untreated
support, no isomerisation products were observed. As such, the
principle reaction over both the pure support and Ni/Al2O3 is the
total hydrogenation of C2PN to PA. The greater mass of retained
carbon formed during T3PN hydrogenation compared to during
C2PN hydrogenation is reflected in the differing activities of the
isomer: mass laydown is greatest where the reactant (or a species
derived from it) is not hydrogenated to a less tightly adsorbed
species. The description of active surface sites provided herein can
also explain previous experimental observations. In particular, dur-
ing T3PN hydrogenation the formation of PA has been shown to
directly correlate with mass laydown on the catalyst surface [18].
An increase in the adsorbed mass was accompanied by a decrease
in PA formation, however the removal of adsorbed species through
flushing with He resulted in an increase in activity towards PA for-
mation. As a result of the studies reported in the present work this
can now be assigned to the removal of hydrocarbonaceous mate-
rial from site (v) described above. Concerning the hydrogenation
of C2PN over Ni/Al2O3 it is likely that the reactant molecule con-
tributes to deactivation, in particular of sites (ii) and (v) in a similar
manner to T3PN.
A key point to note is that the nature of the retained overlayer
differs depending on the pentenenitrile isomer from which it orig-
inates. As such, it necessary to think beyond generic concepts such
as “coke” and instead consider the structural nature and poten-
tial catalytic role of any carbonaceous residue. The importance
of this has been shown in studies of alkyne hydrogenation over
Pd catalysts where retained carbon, in the form of Pd–C surface
phase, forms the catalytically selective sites for alkene formation
[9]; in xylene isomerisation where “selectivation” through carbon
laydown is commonly employed [40,41]; and in the dehydrogena-
tion of n-butane over VOx/Al2O3 where retained carbon directly
provides the catalytically active centres [42].
Studies of the hydrogenation of C2PN over Ni/Al2O3, including
studies on related systems such as Ni/SiO2, have therefore allowed
us to develop a model of deactivation in this reaction, caused in
large part by carbon laydown, in terms of the deactivation of groups
of active sites on the catalyst surface.
5. Conclusions
Understanding the factors which influence selectivity in the
hydrogenation of multi-functional molecules is a key driver in
catalysis research. In this study we have correlated carbon lay-
down with catalyst deactivation and changes in selectivity in the
hydrogenation of cis-2-pentenenitrile over Ni/Al2O3. In particular,
five distinct active sites have been identified which are subject to
different deactivation processes and rates. Deactivation of these
sites therefore directly controls selectivity in this reaction. The
sites identified are: (i) moderate energy Ni site for PN formation;
(ii) moderate energy Al2O3 site for C2PN isomerisation; (iii) high
energy Al2O3 site for PA formation; (iv) high energy Ni site for PA
Fig. 10. Possible routes to the formation of ammonia from C2PN. The top route is
a result of hydrogenolysis and is the mechanism proposed to occur on Al2O3. The
bottom route involves the decomposition of PA and the formation of a retained
carbonaceous deposit and is the mechanism proposed to occur on Ni/Al2O3.

200
J. McGregor et al. / Applied Catalysis A: General 384 (2010) 192–200
formation; and (v) high activity Ni site for PN formation. Rapid mass
laydown in the early stages of reaction, quantified by TEOM mea-
surements, can be directly implicated in the deactivation of sites
responsible for PA formation, therefore changing reaction selectiv-
ity from PA to PN. The rates of deactivation have been quantified
through kinetic analysis. Of particular note is site (ii): an alumina
site capable of the total hydrogenation of C2PN. The identification
of such a site is surprising as alumina is not generally known as
a highly active hydrogenation catalyst. Finally, a reaction mech-
anism, involving a cyclic intermediate has been proposed. These
results indicate the key role carbon laydown plays in controlling
catalytic activity and selectivity in reactions of multi-functional
molecules, with the nature and the location of the deposited car-
bonaceous material directly influencing the formation of specific
reaction products.
[12] G. Neri, G. Rizzo, L. De Luca, F. Corigliano, I. Arrigo, A.S. Arico, A. Donato, Appl.
Catal. A: Gen. 350 (2008) 169–177.
[13] M.M. Telkar, C.V. Rode, V.H. Rane, R. Jaganathan, R.V. Chaudhari, Appl. Catal. A:
Gen. 216 (2001) 13–22.
[14] J.L. Dallons, G. Jannes, B. Delmon, Catal. Today 5 (1989) 257–264.
[15] P. Kukula, K. Koprivova, J. Catal. 234 (2005) 161–171.
[16] G.D. Yadav, M.R. Kharkara, Appl. Catal. A: Gen. 126 (1995) 115–123.
[17] T. Tihanyi, K. Varga, I. Hannus, I. Kiricsi, P. Fejes, React. Kinet. Catal. Lett. 18
(1981) 449–454.
[18] J. McGregor, L.F. Gladden, Appl. Catal. A: Gen. 345 (2008) 51–57.
[19] B.W. Hoffer, A. Dick van Langeveld, J.-P. Janssens, R.L.C. Bonné, C.M. Lok, J.A.
Moulijn, J. Catal. 192 (2000) 432–440.
[20] W. Zhu, J.M. van de Graaf, L.J.P. van den Broeke, F. Kapteijn, J.A. Moulijn, Ind.
Eng. Chem. Res. 37 (1998) 1934–1942.
[21] D. Chen, E. Bjorgum, K.O. Christensen, A. Holmen, R. Lodeng, Advances in Catal-
ysis, vol. 51, Elsevier Academic Press Inc., San Diego, 2007, pp. 351–382.
[22] J. McGregor, Z. Huang, G. Shiko, L.F. Gladden, R.S. Stein, M.J. Duer, Z. Wu, P.C.
Stair, S. Rugmini, S.D. Jackson, Catal. Today 142 (2008) 143–151.
[23] J. Perez-Ramirez, A. Gallardo-Llamas, C. Daniel, C. Mirodatos, Chem. Eng. Sci. 59
(2004) 5535–5543.
[24] S. van Donk, A. Broersma, O.L.J. Gijzeman, J.A. van Bokhoven, J.H. Bitter, K.P. de
Jong, J. Catal. 204 (2001) 272–280.
[25] L.M. Petkovic, G. Larsen, J. Catal. 191 (2000) 1–11.
Acknowledgements
[26] G.V. Smith, F. Notheisz, Heterogeneous Catalysis in Organic Chemistry, Aca-
demic Press, New York, 1999.
This work was supported by the ATHENA project (EPSRC
GR/R47523/01) funded by EPSRC and Johnson Matthey.
[27] J.L. Dallons, G. Jannes, B. Delmon, Acta Chim. Hung. 119 (1985) 223–232.
[28] D.C. Mudalige, G.L. Rempel, J. Mol. Catal. A: Chem. 116 (1997) 309–316.
[29] M.D. Mantle, P. Steiner, L.F. Gladden, Catal. Today 114 (2006) 412–417.
[30] R.C. Mebane, D.R. Jensen, K.R. Rickerd, B.H. Gross, Synth. Commun. 33 (2003)
3373–3379.
References
[1] H.-U. Blaser, C. Malan, B. Pugin, F. Spindler, H. Steiner, M. Studer, Adv. Synth.
Catal. 345 (2003) 103–151.
[31] R.J. Farranto, C.H. Bartholemew, Fundamentals of Industrial Catalytic Processes,
Chapman and Hall, London, 1997, pp. 306–311.
[2] A.J. Plomp, D.M.P. van Asten, A.M.J. van der Eerden, P. Maki-Arvela, D.Y. Murzin,
K.P. de Jong, J.H. Bitter, J. Catal. 263 (2009) 146–154.
[3] J. Springerova, P. Kacer, L. Cerveny, Res. Chem. Intermed. 31 (2005) 785–
795.
[32] J.R. Anderson, N.J. Clark, J. Catal. 6 (1966) 20–25.
[33] C. Kemball, R.L. Moss, Proc. R. Soc. Lond. A 238 (1956) 107–116.
[34] G. Meitzner, W.J. Mykytka, J.H. Sinfelt, Catal. Lett. 32 (1995) 335–344.
[35] G. Meitzner, W.J. Mykytka, J.H. Sinfelt, Catal. Lett. 37 (1996) 137–141.
[36] P.H. Emmett, R.W. Harkness, J. Am. Chem. Soc. 54 (1932) 538–548.
[37] C. Kemball, F.J. Wolf, Trans. Faraday Soc. 51 (1955) 1111–1119.
[38] I. Chorkendorff, J.N. Russell, J.T. Yates, J. Chem. Phys. 86 (1987) 4692–4700.
[39] S.X. Huang, D.A. Fischer, J.L. Gland, J. Phys. Chem. 100 (1996) 13629–13635.
[40] J.S. Beck, R.A. Crane, M.F. Mathias, J.A. Kowalski, D.N. Lissy, D.L. Stern, WO Patent
99 52842 (1999) to Mobil Oil Corporation.
[41] F. Bauer, W.-H. Chen, Q. Zhao, A. Freyer, S.-B. Liu, Micropor. Mesopor. Mater. 47
(2001) 67–77.
[42] J. McGregor, Z. Huang, E.P.J. Parrott, J.A. Zeitler, K.L. Nguyen, J.M. Rawson, A.
Carley, T.W. Hansen, J.-P. Tessonnier, D.S. Su, D. Teschner, E.M. Vass, A. Knop-
Gericke, R. Schloegl, L.F. Gladden, J. Catal. 269 (2010) 329–339.
[4] P. Gallezot, D. Richard, Catal. Rev. Sci. Eng. 40 (1998) 81–126.
[5] A.S. Alammar, G. Webb, J. Chem. Soc., Faraday Trans. 1 74 (1978) 657–664.
[6] S.D. Jackson, A. Monaghan, Catal. Today 128 (2007) 47–51.
[7] A.S. Canning, S.D. Jackson, S. Mitchell, Catal. Today 114 (2006) 372–376.
[8] A. Sarkany, React. Kinet. Catal. Lett. 96 (2009) 43–54.
[9] D. Teschner, E. Vass, M. Havecker, S. Zafeiratos, P. Schnorch, H. Sauer, A. Knop-
Gericke, R. Schloegl, M. Chamam, A. Wootsch, A.S. Canning, J.J. Gamman, S.D.
Jackson, J. McGregor, L.F. Gladden, J. Catal. 242 (2006) 26–37.
[10] S.J. Thomson, G. Webb, Chem. Commun. (1976) 526–527.
[11] C.A. Hamilton, S.D. Jackson, G.J. Kelly, R. Spence, D. de Bruin, Appl. Catal. A: Gen.
237 (2002) 201–209.