Hierarchical zeolites for catalytic hydrocarbon conversion
Citation for published version (APA):Tempelman, C. H. L. (2015). Hierarchical zeolites for catalytic hydrocarbon conversion. Technische UniversiteitEindhoven.
Document status and date:Published: 01/01/2015
Document Version:Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)
Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can beimportant differences between the submitted version and the official published version of record. Peopleinterested in the research are advised to contact the author for the final version of the publication, or visit theDOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and pagenumbers.Link to publication
General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, pleasefollow below link for the End User Agreement:www.tue.nl/taverne
Take down policyIf you believe that this document breaches copyright please contact us at:[email protected] details and we will investigate your claim.
Download date: 30. Jan. 2022
Hierarchical Zeolites for Catalytic
Hydrocarbon Conversion
PROEFSCHRIFT
ter verkrijging van de graad van doctor aan de
Technische Universiteit Eindhoven, op gezag van de
rector magnificus, prof.dr.ir. F.P.T. Baaijens, voor een
commissie aangewezen door het College voor
Promoties in het openbaar te verdedigen
op woensdag 28 oktober 2015 om 16.00 uur
door
Christiaan Herman Lucien Tempelman
geboren te Zaltbommel
Dit proefschrift is goedgekeurd door de promotor:
prof.dr.ir. E.J.M. Hensen
Copyright © 2015, Christiaan Tempelman
Hierarchical zeolites for catalytic hydrocarbon conversion
A catalogue record is available from the Eindhoven University of Technology Library
ISBN: 978-90-386-3938-3
The work described in this thesis has been carried out at the Schuit Institute of Catalalysis
within the Laboratory of Inorganic Chemistry and Catalysis of Eindhoven University of
Technology in The Netherlands. This work received financial support by the European
Community through the NEXT-GTL project (NMP3-LA-2009-229183).
Cover design: Ilse Weisfelt
Printed at Gildeprint
Contents
Chapter 1 General introduction…………….……………………………...
1
Chapter 2 Desilication and silylation of Mo/HZSM-5 for methane
dehydroaromatization…………………………………………..
17
Chapter 3 Activation of Mo/HZSM-5 for methane aromatization………...
51
Chapter 4 On the deactivation of Mo/HZSM-5 in the methane
dehydroaromatization reaction…………………………………
73
Chapter 5 One-step synthesis of nano-crystalline MCM-22………………
95
Chapter 6 Effect of zeolite crystalline domain size on the methane
aromatization performance of Mo/HZSM-5……………...…….
123
Chapter 7 Texture, acidity and fluid catalytic cracking performance of
hierarchical faujasite zeolite prepared by an amphiphilic
organosilane…………………………………………………….
137
Summary ………………………………………………………………….. 165
List of publications ………………………………………………………………….. 171
Acknowledgements …………………………………………………………………. 173
Curriculum Vitea ………………………………………………………………….. 177
Chapter 1
1
General introduction
1. lntroduction
1.1 Catalysis
Catalysis already played an important role in the lives of early humans. One can for
instance consider the fermentation of sugars into ethanol. The term catalysis was introduced
in 1835 by the Swedish scientist Jöns Jacob Berzelius [1]. It is a combination of the Greek
words κατά (kata, “down”) and λύω (luō, “loose”). Berzelius was the first to investigate and
report catalytic reactions in an organized manner and he can be considered the founder of
modern-day catalysis. The translation of academic catalysis research into industrial
applications started at the end of the nineteenth century. Rapid economic growth and the
increasing world population demanded the large scale production of base chemicals. The use
of catalytic processes made this economically and practically possible. The main enabling
aspect of catalysis was the increase of the rate of chemical reactions under practical
conditions. Other driving forces boosting development of catalytic processes were the
increasing need for mobility, increasing living standards and, less positively, both world wars.
During these periods large scale processes have been developed: the production of ammonia
for fertilizers (Haber-Bosch process), nitric acid production for explosives (Ostwald process)
and fuel production and processing (Fischer-Tropsch process, fluid catalytic cracking (FCC),
etc). Nowadays, catalysts are the key assets of the chemical industry, enabling approximately
85% of all industrial chemical processes [2]. From the 1970s onwards, emission control
became an important aspect in catalysis research. A very significant example in this field is
the three-way monolith catalytic converter of automotive exhaust gasses to reduce CO, NO
and hydrocarbon emissions [2-4]. Without catalysis, the current living standards of our
society would not have been achieved.
To explain the principle of catalytic action, let us consider the reaction between
molecules A and B (Fig. 1.1) to form molecule AB. In a typical chemical reaction, one has to
Chapter 1
2
overcome an energy barrier that often requires a substantial amount of energy. Very often, the
high temperatures required to overcome this uncatalyzed barrier will lead to undesired side
reactions. A successful catalyst will lower the energy barrier for the overall reaction by
providing an alternative reaction path in which a catalyst is involved. In the first step of a
catalytic reaction (adsorption), the reactants form bonds with the catalyst surface, lowering
the free energy. Subsequently, the adsorbed reactants may form bonds with each other to form
the desired product. To close the catalytic cycle, the products desorb from the catalyst surface.
Fig. 1.1. Schematic representation of a catalytic reaction between molecule A and B to form
molecule AB [2].
Catalysts come in great diversity in respect to composition, form, size and shape. One of the
largest adopters of catalysis is Nature. Almost all biological processes occur with the help of
enzymes. Examples are the build-up of proteins and DNA. Also the breakdown of ethanol in
the body by alcohol dehydrogenase and the conversion of CO2 and H2O to sugars by
chlorophyll in plants are important biochemical processes involving enzymes as catalysts. The
term heterogeneous catalysis describes a catalytic reaction in which the reactants are in a
different phase than the catalyst. Usually, reactants are in the gas or liquid phase and the
catalyst is a solid. In a homogeneous catalytic process, reactants and catalyst are in the same
phase. Typically, a heterogeneous catalyst has a high active surface area, provided by
stabilized of the active phase in the form of small particles on a relatively inert material that
Chapter 1
3
acts as a support. The reactivity of the active phase resides in the coordinative unsaturation of
the surface atoms.
An important class of heterogeneous catalysts are zeolites. Zeolites are crystalline
aluminosilicates characterized by a well-defined microporous structure; the pores have
dimensions close to those of molecules. In addition, zeolites have a tunable chemical
composition by which virtue the acidity can be controlled to some degree. The term zeolite
was first mentioned by the Swedish mineralist Axel Fredrik Cronstedt. It is a combination of
the Greek words ζέω (zéō), meaning "to boil" and λίθος (líthos), meaning "stone" [5]. Barrer
succeeded to synthesize zeolites under hydrothermal conditions, and with this a range of new
structures that did not occur in nature became available [6]. Currently, 218 zeolite frameworks
are known in the IZA database [7]. However, only a small number of these are actually used
in industrial catalytic processes. The largest application of zeolites is in laundry detergents,
which comprises an annual volume of 1.44 million metric tons [2]. As catalysts, zeolites
mainly find application as solid acids. Their high Brønsted acidity arises when a framework
oxygen atom is neighbored by Si and Al cations. The bridging oxygen atom has a -1 charge.
Compensating the negative charge with a proton results in a highly ionic bond, providing the
material its strong Brønsted acidity [8]. In addition to strong acidity, two other properties are
important to explain the widespread application of zeolites [9]. The first one is the shape
selectivity exterted by the presence of the acid sites in micropores [10, 11], with dimensions
in the range of molecules such as hydrocarbons. The second one is their outstanding thermal
and hydrothermal stability [12] that makes them applicable under harsh industrial conditions.
For instance, ZSM-5 with the MFI topology is widely used in the chemical industry. It is used
as an additive in the FCC process next to faujasite zeolite, and also in the isomerization of n-
alkanes into branched isomers to increase the octane number of gasoline. The topology of
ZSM-5 is such that a threedimensional pore system is obtained that is comprised of straight
(along the b-axis, Fig. 1.2a) and sinusoidal (along the a-axis, Fig. 1.2b) channels that
intersecting at the so-named channel intersections. The pores consist of 10-membered rings,
so that the pore diameter is 5.5 Å.
Other important zeolite types are MCM-22 (MWW topology) and zeolite Y (FAU
topology). The first one is an important catalyst for the alkylation of benzene with olefins [13],
while the second is the dominant acid component in FCC cracking catalysts [14]. The
micropore system of MWW consists of two separate two-dimensional channel systems. One
system consists of straight 10-membered rings (Fig. 1.2c, indicated in blue) [15, 16]. The
Chapter 1
4
second pore system is created when two cups located at the surface of adjacent MWW layers
are connected to form a super cage. These large ellipsoidal cages are typically 7.1 Å in
diameter and 18.2 Å in height [16]. The large cages are connected with each other through 10-
membered ring windows (Fig. 1.2c, indicated in red). The structure of the FAU family to
which zeolite Y belongs is built from sodalite cages connected to each other through a double
4-ring (Fig. 1.2d). In this configuration a supercage is formed with a diameter of 11.6 Å
[17,18]. The supercages are accessible through 12-membered windows (indicated in red)
having a diameter of 7.4 Å [17,18].
Fig. 1.2. Zeolite structures of a) ZSM-5 (MFI) viewed along the c-axis, b) ZSM-5 (MFI)
viewed along the a-axis, c) MCM-22 (MWW) and d) zeolite Y (FAU).
Most of the applications of zeolites in catalysis aim to valorize crude oil into fuels and
chemicals. Today, the production and use of fossil resources is debated because of climate
change issues. Burning fossil fuels releases vast amounts of CO2 into the atmosphere, causing
global warming [19]. Energy security is another driver to explore the use of alternative
feedstock. The largest oil reserves are located in politically unstable regions making the crude
oil supply chain uncertain. A final concern relates to the end of cheap oil; at the rate of current
consumption there is enough conventional oil for about half a century [20]. All these factors
form drivers for the search for alternative energy and chemical resources. Although during the
last decade new technologies have been discovered and are being developed, the road from
discovery to industrial application may take decades [21]. It is most desirable to convert our
current society that relies on the use of fossil resources to a carbon-neutral society based on
the use of renewable energy from the sun. To overcome long lead-times for novel renewable
energy technologies, there is a great need for transition technologies that are preferably based
on less-polluting feedstock. The option that is currently gaining much in attention is the use of
natural gas for the production of fuels and chemical [22]. Methane, the main component of
natural gas, is the cleanest of all fossil resources and it is widely available.
Chapter 1
5
1.2 Methane as feedstock
Historically, natural gas has been mainly used for heating and electrical power
generation. It can be efficiently distributed using pipelines to the end consumers. Transport
via pipelines over longer distances is also feasible, although the capital investments are
typically higher than the investments for compressed or liquefied natural gas transport.
Especially, LNG technology is rapidly becoming more important for natural gas transport
over long distances. As an alternative, the direct conversion of methane into valuable products
has gained significant interest. Methane has been used for a long time as the primary source of
hydrogen for such processes as methanol and ammonia synthesis. The use of natural gas for
the production of chemicals has certain advantages as compared to petroleum oil. The
composition of natural gas is more uniform and methane possesses the highest hydrogen to
carbon (H/C) ratio of all fossil resources. The total estimated natural gas reserves exceed that
of petroleum with a factor 10, especially when the shale gas deposits are also taken into
account [23, 24]. The extraction of natural gas from impermeable shale rock formation
recently took off with the development of horizontal drilling and hydraulic fracturing
techniques [24, 25]. This development has already dramatically changed energy scenarios and
energy markets. Despite environmental concerns about these new gas exploration techniques,
shale gas production is already fully implemented in the U.S. Other countries such as China,
Australia, South Africa and the United Kingdom are exploring the possibility of shale gas
production. As a result of the cheap shale gas production in the U.S., natural gas prices have
plummeted, making it a competitive carbon feedstock next to petroleum [26].
The catalytic conversion of cheap methane into valuable chemicals is one of the
interesting routes to decrease our dependency on petroleum. At present, Gas-to-liquid (GTL)
processes are already contributing to the conversion of methane via synthesis gas (a mixture
of CO and H2) into fuels and chemicals (Fig. 1.3). Despite the scale of these activities of Shell
and Sasol in Qatar, Nigeria and South Africa, the contribution of liquid fuels from natural gas
to satisfy our hunger for liquid fuels for mobility is still very small. The recent paper by Wood
et al. [27] presents a complete overview of both established GTL processes and those under
development. All currently commercially operated processes are based on the indirect
conversion of methane. First, methane is transformed into synthesis gas by partial oxidation,
steam reforming or autothermal reforming. The resulting syngas can then be converted into
methanol, which in turn can be used for the production of dimethyl ether (DME), light olefins
Chapter 1
6
or gasoline. Currently, China is rapidly expanding its methanol-to-olefins (MTO) capacity. A
significant fraction of the syngas for methanol production is derived from coal, although it is
projected that a number of new natural gas based plants will be built in the U.S. to produce
methanol for MTO plants at China’s east coast. Alternatively, syngas can also be directly
converted into hydrocarbons using the Fischer-Tropsch Synthesis (FTS) process [28-31]. In
FTS, a range of hydrocarbons is produced; the wax fraction is converted in a downstream
hydrocracking step to naphtha, diesel and lubricants. Due to the large investments required for
Fischer-Tropsch type GTL processes, operating this technology is only economically viable
close to very large natural gas fields. Investments in liquefaction of natural gas are also very
high so that such terminals can only be built close to large fields to be economically viable.
Monetizing on smaller fields can be done by compressed natural gas technology. However,
the majority of natural gas fields are too small to be exploited with current technology.
Fig. 1.3. Main routes from natural gas to value-added products: MDA (methane dehydro-
aromatization), POX (partial oxidation), OCM (oxidative coupling of methane), MTG
(methanol to gasoline), MTO (methanol to olefins), and FTS (Fischer–Tropsch synthesis).
The main problem with current GTL processes is the methane conversion step that
requires large capital investments (e.g., the oxygen plant for auto thermal reforming). The
economics of processes that directly convert methane into liquids would be much better,
because it would avoid the syngas generation step. An inherent problem with such direct
methane conversion routes to, for instance, methanol lies in the high reaction temperatures
needed to active the strong C-H bonds in methane. The formed products are highly reactive at
Chapter 1
7
such temperatures, often leading to undesired by-products and catalyst deactivation.
Activation of methane can be done under oxidative or non-oxidative conditions [28-31]. In
the oxidative routes, oxygen is the preferred oxidant. An example is the oxidative coupling of
methane (OCM) to ethane and ethylene. The process has not been applied commercially,
because of low product yields and the difficulty in heat management of the process. Another
oxidative route is the partial oxidation of methane to produce methanol or formaldehyde; such
approaches suffer from the higher reactivity of the product than the reactant so that reasonable
selectivities can be achieved at modest conversion.
Without an oxygen source, methane can be converted into aromatics and hydrogen. The
main process under investigation is the dehydroaromatization of methane, which was first
reported by Wang et al. in 1993 [32]. If it could be done in an economic manner, methane
dehydroaromatization (MDA) could convert methane into liquid benzene and hydrogen [33].
However, the unfavorable thermodynamic equilibrium of the methane benzene + hydrogen
reaction demands high reaction temperatures (>650°C) to attain reasonable methane
conversion [34, 35]. As a consequence of extensive coke deposition, catalyst stability is a
major problem. Usually, the process is operated at atmospheric pressure in a fixed-bed reactor
using Mo-modified zeolites. The most often used zeolite is Mo/HZSM-5 [36].
1.3 General aspects of the MDA catalyst
Since the early work by Wang et al. [32] many studies and reviews [37-42] have been
published on molybdenum modified zeolites as promising MDA catalysts. Large efforts have
been made to improve benzene selectivity and catalyst lifetime. Mo/zeolite catalysts are prone
to deactivation due to the formation of carbon species at the zeolite surface Also, substantial
research efforts has been dedicated to elucidate the reaction mechanism. The most accepted
reaction mechanism follows a cascade of reactions [41]. Initially, methane reacts with the Mo-
oxide to form the MoCx species. These species activate methane and catalyze the formation of
the C-C bonds, mainly resulting in ethylene. In the last step, the olefins aromatize at the
Brønsted acid sites to form benzene and heavier aromatics.
As mentioned, the first step in the MDA reaction is the activation of methane by breaking
one of the stable C-H bonds. Breaking the C-H bond requires high temperatures (>650°C) and,
at such conditions, the resulting CH3 specie is susceptible to further C-H bond breaking
leading to formation of coke. Theoretical calculations suggested that transition metal oxides
Chapter 1
8
are able to activate a single C-H bond without breaking the remaining C-H bonds [43, 44].
Several potential metals have been proposed including Zn, W, Re, Cu, Ni, Fe, Mn, Cr, V, Ga
and Pt [38, 39, 42]. Extensive screening revealed molybdenum to be the preferred metal-oxide
component for the activation and oligomerization of methane (see Table 1.1). In order to
obtain the active phase, the Mo-oxide is carburized upon reaction with methane to form
molybdenum carbide (MoCx) which is believed to be the active species. The formation
mechanism and properties of these MoCx species will be discussed in section 1.5.
Introducing the metal oxide onto the zeolite support has certain advantages. It improves
particle dispersion and, therefore, the catalytic activity. Furthermore, the Brønsted acidic
properties of the zeolite also contribute to the spreading of Mo over the surface and diffusion
into the micropores [45, 46]. The support also plays an important role in the stabilization of
the MoCx species and it enhances the proximity to the acid sites needed for the aromatization
step.
The aromatization step comprises transformation of the olefins over the Brønsted acid
sites into aromatic compounds, mainly benzene. Efforts have been made to correlate the acid
properties of zeolite to the performance in the MDA reaction. The influence of the Si/Al ratio
and Al distribution has been investigated [47-51]. The catalyst based on ZSM-5 as zeolite
support is the benchmark. The high selectivity towards benzene shown for Mo/ZSM-5 [37, 41,
42] originates from the micropore channel dimensions (5.4-5.6 Å) close to the kinetic
diameter of benzene. In this way, the micropores sterically hinder formation of larger
(poly)aromatic molecules. In addition to ZSM-5, other zeolite structures have been
investigated [38-40]. In particular, MCM-22 (MWW topology) showed promising catalytic
performance. activity of Mo/MCM-22 was higher and it was less prone to deactivation than
ZSM-5. The improved catalytic properties of the Mo/MCM-22 catalyst were attributed to the
more suitable topology. The micropores of MCM-22 are able to accumulate more carbon
compared to ZSM-5 before losing shape selectivity [52]. However, due to patent issues,
material costs and industrial relevance, the Mo/ZSM-5 catalyst is still considered to be the
benchmark for MDA and most of the research is focused on it.
Chapter 1
9
Table 1.1. Comparison MDA activities of different metals supported on HZSM-5 zeolite.
Adapted from [39].
Active
metals
Reaction conditions CH4
conversion
(%)
Selectivity (%)
T (ºC) Flow (ml·gcat-
1·h-1) Benzene Naphthalene
Mo 730 1500 16.7 60.4 8.1
Zn 700 1500 1.0 69.9 n.d.2
W 800 1500 13.3 52.0 n.d.2
Re 750 1440 9.3 52.0 0
Co-Ga 700 1500 12.8 66.5 7.2
Fe 750 8001 4.1 73.4 16.1
Mn 800 1600 6.9 75.6 11.9
V 750 8001 3.2 32.6 6.3
Cr 750 8001 1.1 72.0 3.7
1GHSV / h
-1;
b Not determined.
1.4 Catalyst preparation
The final performance of the catalyst in MDA strongly depends on the Mo introduction
procedure [9, 43]. A wide range of methods have been explored in the past, predominantly
related to the preparation of Mo/ZSM-5. These reports concluded that catalysts prepared by
incipient wetness impregnation procedure or solid state exchange showed the best catalytic
performance in MDA [39]. In the incipient wetness procedure, the zeolite is impregnated with
an aqueous ammonium heptamolybdate (AHM) solution. After impregnation, bulky
molybdate anions [Mo7O246-
] are deposited at the external surface of the zeolite. To
decompose the AHM precursor into smaller metal-oxide species and improve the dispersion
of the molybdate species, a calcination step is performed. Xu et al. studied the influence of
calcination after impregnation in more detail using FT-IR spectroscopy and differential
thermal analysis [53]. They found that AHM starts to decompose into MoO3 crystallites at
514 K. However these crystallites are large and remain at the external surface of the zeolite.
Increasing the temperature further to approximately 773 K, leads to degradation of the metal-
oxide into smaller fragments. These smaller fragments spread over the external surface of the
zeolite and are small enough to diffuse into the micropores. The group of Iglesia [54-56]
found the mechanism of metal oxide spreading and micropore diffusion in the solid state
Chapter 1
10
exchange procedure of H-ZSM-5 and MoO3 to be comparable with that of the incipient
wetness procedure. It was also found that the Mo species interact with the Brønsted acid sites
of the zeolite forming strong Mo-O-Al bonds. The exchange of Mo with the acidic BAS
protons leads to a decrease in acidity. Therefore, introducing large amounts of Mo can lead to
severe reduction of BAS concentration and, possibly, decrease of the catalytic activity.
Scheme 1 presents the most accepted mechanism on the formation Mo-oxo species exchanged
to the BAS. In the first step a small MoO3 fragment reacts with the BAS proton and forms a
MoO2-(OH)+ specie connected to the BAS. In the next step either monomeric (MoO2
2-) or
dimeric (Mo2O52+
) species are formed, depending on the BAS concentration. High BAS
concentration leads to the formation of predominantly monomeric Mo-oxo species (scheme 1,
equation 2). At lower BAS concentrations the dimeric Mo-oxo species and are formed when
two MoO2-(OH)+ species are in proximity (Scheme 1.1, equation 1). Extreme calcination
temperatures can lead to BAS hydrolyzation forming framework defects. The framework
defects can react with the Mo-oxo species leading to Al extraction and the formation of
Al2(MoO4)3 species.
Scheme 1.1. Interactions between MoOx species and Brønsted acid sites of HZSM-5. Adapted
from [39].
1.5 The active Mo phase in MDA
It is generally accepted that the MoCx particles are responsible for the activation and
oligomerization of methane in the MDA reaction. To obtain the active MoCx phase the MoOx
>773K MoO
3 +
O
H
Al Si
MoO3 +O2
>773K O
MoO2OH
Al Si
+O
MoO2OH
Al Si
O2
>773K + H2O (g)
Si
O
Mo
Al
O
O
Al
O
Mo
Si
O
O
O
O
H
Al Si
MoO3 +O2
>773K O
MoO2OH
Al Si
+O
MoO2OH
Al Si
O2
>773K + H2O (g)
Si
O
Mo
Al
O
O
Al
O
Mo
Si
O
O
O
O
H
Al Si
MoO3 +O2
>773K O
MoO2OH
Al Si
+O
MoO2OH
Al Si
O2
>773K + H2O (g)
Si
O
Mo
Al
O
O
Al
O
Mo
Si
O
O
O
+ O
H
Al Si
MoO3 +O2
>773K O
MoO2OH
Al Si
+O
MoO2OH
Al Si
O2
>773K + H2O (g)
Si
O
Mo
Al
O
O
Al
O
Mo
Si
O
O
O
+ H2O
O2 O
2
>773K
O
H
Al Si
MoO3 +O2
>773K O
MoO2OH
Al Si
+O
MoO2OH
Al Si
O2
>773K + H2O (g)
Si
O
Mo
Al
O
O
Al
O
Mo
Si
O
O
O
+ O
H
Al Si
MoO3 +O2
>773K O
MoO2OH
Al Si
+O
MoO2OH
Al Si
O2
>773K + H2O (g)
Si
O
Mo
Al
O
O
Al
O
Mo
Si
O
O
O
O2
>773K O
MoO2OH
Al Si
O
H
Al Si
+O2
773-793K+ H2O (g)O
Mo
O
O O
SiAl Al Si
+ H2O
O
H
Al Si
MoO3 +O2
>773K O
MoO2OH
Al Si
+O
MoO2OH
Al Si
O2
>773K + H2O (g)
Si
O
Mo
Al
O
O
Al
O
Mo
Si
O
O
O
+ O
H
Al Si
MoO3 +O2
>773K O
MoO2OH
Al Si
+O
MoO2OH
Al Si
O2
>773K + H2O (g)
Si
O
Mo
Al
O
O
Al
O
Mo
Si
O
O
O
O2
>773K O
H
Al Si
+O
H
Al Si
O2
>773K
O
SiAl+ H2O (g)
Al Si
O2
MoO3
“Extraframework“ Al2O3 or Al2(MoO4)3
(Reversible, no loss of crystalinity)(Reversible, no loss of crystalinity)
++ O
H
Al Si
+O
H
Al Si
O2
>773K
O
SiAl+ H2O (g)
Al Si
O2
MoO3
“Extraframework“ Al2O3 or Al2(MoO4)3
(Reversible, no loss of crystalinity)(Reversible, no loss of crystalinity)
+O
2
>773K + H
2O “Extraframework”
Al2O
3 or Al
2(MoO
4)
3
(1)
(2)
(3)
Chapter 1
11
species have to be carburized by reaction with methane. The carburization leads to an
induction period towards the formation of hydrocarbon products during the early stages of
reaction. The exact place and the nature of these carbide species, however, are still not fully
understood. Studies by the groups of Solymosi and Lunsford [57, 58] proposed that highly
dispersed Mo2C species at the external surface of the zeolite are responsible for methane
activation. Later, EXAFS studies by the group of Iglesia suggested that the active MoCx
particles are formed upon carburization of dimeric Mo-oxo species located inside the micro
pores [54, 56, 59] (Scheme 1.2). In their proposed mechanism, a part of the BAS are
regenerated upon the formation of these MoCx clusters. In this way, the BAS are in close
proximity to the MoCx phase. It possibly explains the better performance of the bi-functional
catalyst compared to catalytic systems in which the metal carbide and acidic component are
separated.
Scheme 1.2. Reaction of exchanged of MoOx/H-ZSM5 with CH4. Adapted from [43].
Monitoring the size of the MoCx clusters with time on stream, the particles were observed
to increase to approximately 0.6 nm in size (ca. 10 Mo atoms) [43], close to the diameter of
the ZSM-5 pores. A small fraction of these MoCx clusters were seen to agglomerate into
larger particles and exceed the zeolite pore dimensions forcing them to migrate to channel
intersections, crystalline defects and the external surface. The resistance to agglomeration of
the main fraction of MoCx particles was attributed to several factors including the low vapor
pressure, high melting point and the strong interaction of MoCx species with the zeolite
framework [60]. They showed that highly dispersed Mo-oxo species that strongly interact
with the zeolite surface were difficult to completely reduce to Mo2C. Instead, the carburized
metal-oxide phase was only partially reduced into MoCxOy particles [61, 62].
1.6 Catalyst deactivation
Poor catalyst stability is the major hurdle to be overcome in the further development of
the MDA reaction into an industrial process. Catalyst deactivation is attributed to the high
reaction temperatures required for C-H bond activation, which favors formation of coke; it
+ CH4 O
O
O
O
Si Si
Al
SiAl
Si
SiSi
+ COx + H2OO
H
Al Si
+
Si
O
Mo
Al
O
O
Al
O
Mo
Si
O
O
O
Chapter 1
12
results in the deposition of carbonaceous products and blockage of the micropores [38].
Another problem is sublimation of some Mo-oxide phases [63]. Some researchers also
attribute deactivation to the transformation of the η-Mo3C2 phase towards less active α-MoC1-
x and β-Mo2C phases upon interaction with carbonaceous deposits [37]. Coke formation is
widely regarded as the main contributor to catalyst deactivation [11, 64-66]. Brønsted acid
sites (BAS) located at the external surface are thought to cause extensive carbon deposition
[67], with the lack of shape selectivity of the micropores explaining the different products
mixture formed from benzene and toluene formed in the micropores. The polynuclear
aromatics tend to cover the external surface and block access to the micropores.
Several papers have characterized the carbonaceous deposits by temperature programmed
oxidation and thermogravic analysis methods to the spent samples. Such characterization
provides quantitative and qualitative information on the carbonaceous deposits composition
[65-69]. The different types of coke are typically distinguished by their combustion
temperature. Usually, two types of carbon species can be identified. An amorphous type of
carbon formed in the proximity of Mo can be oxidized at lower temperatures, the alternative
interpretation being that it pertains to oligomeric species in the micropores. A polyaromatic
type of carbon can only be oxidized at higher temperatures.
1.7 Catalyst regeneration
Since it appears that coke formation is an intrinsic property of the MDA reaction over
Mo-modified zeolites, it is likely that a viable industrial process would need to involve
periodic regeneration to remove the carbonaceous deposits. Several regeneration methods
have been extensively explored [45], [70, 71]. It has turned out to be a challenge to control the
high exothermicity of the coke oxidation, which will lead to catalyst damage [70].
Sublimation of Mo-oxides formed upon regeneration also negatively affects performance of
the regenerated catalyst. Typically, oxidation of the coke in air (above 723 K to remove hard
coke) leads to decreased activity after consecutive regeneration cycles. In order to prevent
catalyst damage, regeneration should preferably be carried out at lower temperatures. A
promising example is found in the study by Ma et al. [70], who showed the possibility to
remove carbon deposits using a mixture of 2 % NO in air). In such mixture, carbon could be
fully removed at 623 K and eight regeneration cycles could be carried out without activity
loss. The regenerability depends also on the degree of coking. Ismagilov et al. [72] showed
that it is possible to regain full catalytic activity when regeneration was applied after a period
Chapter 1
13
of 6 h on stream for five consecutive regeneration cycles. However, longer periods on stream
(15 h and 20 h) led to irreversible damage of the catalyst and decrease in activity after each
regeneration cycle. This most likely relates to the nature and amount of the different coke
species.
1.8 Scope of the thesis
The valorization of methane into aromatics has already been investigated in academic and
industrial research laboratories. Poor catalyst stability hinders commercial application of this
process. This thesis investigates the deactivation of Mo-containing zeolite catalysts in the
MDA reaction. Methods to improve catalyst performance, stability and selectivity are also
investigated. Chapter 2 discusses ways to improve its performance in MDA. One strategy to
improve catalytic performance is the introduction of mesopores in the acidic zeolite support to
decrease the size of the micropore domains; in this way diffusion can be enhanced and the
negative effects of carbon deposition can be alleviated. The alternative strategy involves the
chemical modification of the external zeolite surface with silica to decrease its activity
towards formation of coke. The catalyst performance is compared to the benchmark
Mo/ZSM-5. To this end, the Mo introduction procedure into zeolite ZSM-5 has been
optimized by screening several adaptable parameters. Chapter 2 showed the beneficial effect
of smaller micropore domains in the zeolite crystal on the catalytic stability in MDA. Chapter
3 covers the influence of the gas atmosphere during the pre-heating step prior to the MDA
reaction. Therefore, the benchmark Mo/HZSM-5 and its silylated counterpart were pre-treated
in inert, oxidizing and carburizing conditions. It was found that pre-carburization of the
catalyst led to the best catalytic performance. The chemical properties of the catalyst after
precarburization was further characterized and compared to that of the fresh catalyst. In
Chapter 4, detailed analysis of activated and spent Mo/HZSM-5 samples during the MDA
reaction is carried out to identify the main reasons for the deactivation. Although formation of
carbon is argued to be the main reason for the poor catalyst stability, relatively little is known
about the nature and location of the carbon deposits that deactivate the catalyst. Chapter 5
focuses on MCM-22 as an alternative zeolite component in the MDA catalyst. In this chapter,
a method is proposed to obtain nano-crystalline MCM-22 in a one-pot synthesis approach by
partial delamination of the zeolite crystallites by adding an organo-silane molecule to the
synthesis gel. In this approach the thickness of the MCM-22 crystals can be decreased in a
simpler manner than the procedure that leads to delaminated MCM-22 (ITQ-2). The prepared
Chapter 1
14
material is extensively characterized and compared to conventional MCM-22 and ITQ-2 in
the MDA reaction, and also in liquid phase benzene alkylation with propylene. In Chapter 6
the influence of the micropore domain of the ZSM-5 zeolite support is studied in more detail.
Therefore MFI supports with varying micropore domain size have been prepared ranging
from 10 µm to several nanometers. After introducing molybdenum the catalytic performance
was tested in MDA. Evaluation of the catalytic performance revealed improved stability and
selectivity upon hierarchical structuring of the ZSM-5 support. In Chapter 7 the potential of
mesopore-containing faujasite in FCC is investigated. Using an organosilane molecule,
mesopores are introduced during Y zeolite synthesis. The various preparation methods were
scaled up for the catalytic evaluation in FCC. A mesoporous Y zeolite is compared with bulk
zeolite Y in a procedure that involves binding of the zeolite in kaolin, accelerated steam
deactivation and fluid catalytic cracking of vacuum gas oil. The residual acidity of the zeolites
and composite catalysts is determined and the influence of interconnected mesopores on the
performance in catalytic FCC is discussed. The main findings and results of the thesis are
summarized in the Summary.
1.9 References
[1] B. Lindström, L.J. Petterson, Cattech 7 (2003) 130-138.
[2] I. Chorkendorff and J.W. Niemantsverdriet, Concepts of Modern Catalysis, Wiley, 2007.
[3] R.A. van Santen, P.W.N.M. van Leeuwen, J.A. Moulijn, B.A. Averill, Catalysis: An Integrated
Apporach, Elsevier, 1999.
[4] "The Catalytic Converter". The Petersen Automotive Troubleshooting & Repair Manual. New
York, N Grosset & Dunlap.
[5] A.F. Crondstedt, Kongl Vetenskaps Academiens Handlinger Stockholm, 17 (1756) 120.
[6] R.M. Barrer, J. Chem. Soc. (1948) 127-132.
[7] IZA Database of Zeolite Structures:
http://izasc.biw.kuleuven.be/fmi/xsl/IZA-SC/ft.xsl
[8] R.A. van Santen, M. Neurock, Molecular Heterogeneous Catalysis, Wiley, 2009.
[9] C. Martínez, A. Corma, Coordination Chem. Rev. 255 (2011) 1558-1580.
[10] P.B. Weisz, V.J. Frilette, J. Phys. Chem. 64 (1960) 289-384.
[11] J.T.F. Degnan, J. Catal., 216 (2003) 32-46.
[12] A. Corma, Chem. Rev. 97 (1997) 1737-2462.
[13] A. Corma, V. Martínez-Soria, E. Schnoeveld, J. Catal. 192 (2000) 163-173.
[14] A. Haas, D.A. Harding, J.R.D. Nee, Microporous Mesoporous Mater. 28 (1999) 325-333.
[15] M.E. Leonowicz, J.A. Lawton, S.L. Lawton, M.K. Rubin, Science 264 (1994) 1910-1913.
[16] S.L. Lawton, M.E. Leonowicz, R.D. Partridge, P. Chu, M.K. Rubin, Microporous Mesoporous
Mater. 23 (1998) 109-117.
[17] G. Bergerhoff, W.H. Baur, W. Nowacki, N. Jb, Miner. Mh. (1958) 193-200.
[18] J. Dhainaut, T.J.Daou, A.Chappaz, N.Bats, B.Harbuzaru, G.Lapisardi, H.Chaumeil, A. Defoin,
L.Rouleau, J.Patarin, Microporous Mesoporous Mater. 174 (2013) 117-125.
[19] "Intergovernemental Panel on Climate Change (IPCC). Climate change 2013: the physical
science basis." (2013)
Chapter 1
15
[20] British Petroleum Company. BP Statistical Review of World Energy. London: British Petroleum
Co, 2013. [21] G.J. Kramer, M. Haigh, Nature 426 (2009) 568-569.
[22] M.P. Hekkert, F.H.J.F Hendriks, A.P.C. Faaij, M.L. Neelis, Energy Policy 33(2005) 579-594.
[23] J.N. Armor, J. Energy Chem. 22 (2013) 21-26.
[24] Q. Wang, X. Chen, A.N. Jha, H. Rogers, Renew. Sus. Energy Rev. 30 (2014) 1-28.
[25] Annual Energy Outlook 2013 (AEO2013), in: U.S. Energy Information Administration (EIA),
U.S. Energy Information Administration; Office of Integrated and International Energy
Analysis, U.S. Department of Energy, Washington DC, 2013.
[26] F.Asche, A. Oglend, P. Osmundsen, Energy Policy 47 (2012) 117-124.
[27] D.A. Wood, C. Nwaoha, B.F. Towler, J. Nat. Gas Sci. Eng 9 (2012) 196-208.
[28] J.H. Lunsford, Catal. Today 63 (2000) 165-174.
[29] E.F. Sousa-Aguiar, L.G. Appel, C. Mota, Catal. Today 101 (2005) 3-7.
[30] A. Martínez, G. Prieto, A. García-Trenco, E. Peris, Zeolites and Catalysis: Synthesis, Reactions
and Application, Wiley, 2010, pp. 649-685.
[31] M. Gharibi, F.T. Zangeneh, F. Yaripour, S. Sahebdelfar, Appl. Catal. A 443-444 (2012) 8-26.
[32] L. Wang, L. Tao, M. Xie, G. Xu, J. Huang, Y. Xu, Catal. Lett. 21 (1993) 35-41.
[33] M.P. Hekkert, F.H.J.F Hendriks, A.P.C. Faaij, M.L. Neelis, Energy Policy 33 (2005) 579-594.
[34] F. Larachi, H. Oudghiri-Hassani, M.C. Iliuta, B.P.A. Grandjean, P.H. McBreen, Catal. Lett. 84
(2002) 183-192.
[35] C. Angelici, B.M. Weckhuysen, P.C.A. Bruijnincx, Chem. Sus. Chem. 6 (2013) 1595-1614.
[36] J. Zhu, X. Meng, F. Xiao, Front. Chem. Sci. Eng. 7 (2013) 233-248.
[37] Z.R. Ismagilov, E.V. Matus, L.T. Tsikoza, Energy Environ. Science 1 (2008) 526-541.
[38] S. Majhi, P. Mohanty, H. Wang, K.K. Pant, J. Energy Chem. 22 (2013) 543-554.
[39] S. Ma, X. Guo, L. Zhao, S. Scott, X. Bao, J. Energy Chem. 22 (2013) 1-20.
[40] Y. Xu, L. Lin, Appl. Catal. A 188 (1999) 53-67.
[41] Z.R. Ismagilov, E.V. Matus, M.A. Kerzhentsev, L.T. Tsikoza, I.Z. Ismagilov, K.D. Dosumov,
A.G. Mustafin, Petroleum Chem. 51 (2011) 174-186.
[42] W. Wei, J. Nat. Gas Chem. 9 (2000) 76-88.
[43] G. Fratesi, S. De Gironcoli, J. Chem. Phys. 125 (2006).
[44] A. Kokalj, N. Bonini, S. De Gironcoli, C. Sbraccia, G. Fratesi, S. Baroni, J. Am. Chem. Soc.
128 (2006) 12448-12454.
[45] D. Ma, Y. Shu, X. Han, X. Liu, Y. Xu, X. Bao, J. Phys. Chem. B 105 (2001) 1786-1793.
[46] D. Ma, X. Han, D. Zhou, Z. Yan, R. Fu, Y. Xu, X. Bao, H. Hu, S.C.F. Au-Yeung, Chem. Eur. J.
8 (2002) 4557-4561.
[47] X. Bao, W. Zhang, D. Ma, X. Han, X. Liu, X. Guo, X. Wang, J. Catal. 188 (1999) 393-402.
[48] A. Martínez, E. Peris, A. Vidal-Moya, in: Zeolites and Related Materials: Trends, Targets and
Challenges, Elsevier B.V., Amsterdam, 2008, pp. 1075-1080.
[49] C. Xu, H. Liu, M. Jia, J. Guan, S. Wu, T. Wu, Q. Kan, Appl. Surf. Sci. 257 (2011) 2448-2454.
[50] J. Dědeček, Z. Sobalík, B. Wichterlová, Catal. Rev. Sci. Eng. 54 (2012) 135-223.
[51] C.H. Bartholomew, R.J. Farrauto, Fundamentals of Industrial Catalytic Processes, 2nd ed.,
Wiley, 2006.
[52] Y. Shu, D. Ma, L. Xu, Y. Xu, X. Bao, Catal. Lett. 70 (2000) 67-73.
[53] Y. Xu, Y. Shu, S. Liu, J. Huang, X. Guo, Catal. Lett. 35 (1995) 233-243.
[54] W. Li, G.D. Meitzner, R.W. Borry Iii, E. Iglesia, J. Catal. 191 (2000) 373-383.
[55] Y.H. Kim, R.W. Borry Iii, E. Iglesia, Microporous Mesoporous Mater. 35-36 (2000) 495-509.
[56] R.W. Borry iii, Y.H. Kim, A. Huffsmith, J.A. Reimer, E. Iglesia, J. Phys. Chem. B 103 (1999)
5787-5796.
[57] F. Solymosi, J. Cserényi, A. Szöke, T. Bánsági, A. Oszkó, J. Catal. 165 (1997) 150-161.
[58] F. Solymosi, L. Bugyi, A. Oszkó, Catal. Lett. 57 (1999) 103-107.
[59] W. Ding, S. Li, G.D. Meitzner, E. Iglesia, J. Phys. Chem. B 105 (2001) 506-513.
[60] H.S. Lacheen, E. Iglesia, Phys. Chem. Chem. Phys. 7 (2005) 538-547.
[61] H. Liu, W. Shen, X. Bao, Y. Xu, J. Mol. Catal. A 244 (2006) 229-236.
[62] H. Liu, X. Bao, Y. Xu, J. Catal. 239 (2006) 441-450.
Chapter 1
16
[63] D. Ren, X. Wang, G. Li, X. Cheng, H. Long, L. Chen, J. Nat. Gas Chem. 19 (2010) 646-652.
[64] M. Nagai, Nishibayashi, S. T. Omi, Appl. Catal. A 253 (2003) 101-112.
[65] B. Weckhuysen, M. Rosynek, J. Lunsford, Catal. Lett. 52 (1998) 31-36.
[66] D. Ma, D. Wang, L. Su, Y. Shu, Y. Xu, X. Bao, J. Catal. 208 (2002) 260-269.
[67] C. Descorme, P. Gelin, C. Lecuyer, A. Primet, Appl. Catal. B 13 (1997) 185-195.
[68] H. Liu, L. Su, H. Wang, W. Shen, X. Bao, Y. Xu, Appl. Catal. A 236 (2002) 263-280.
[69] K. Honda, X. Chen, Z.-G. Zhang, Catal. Commun 5 (2004) 557-561.
[70] H. Ma, R. Kojima, R. Ohnishi, M. Ichikawa, Appl. Catal. A 275 (2004) 183-187.
[71] R.W. Borry III, E.C. Lu, Y.H. Kim, E. Iglesia, Stud. Surf. Sci. Catal. 119 (1998) 403-410.
[72] Ismagilov Z.R., Tsykoza L.T., Matus E.V., Litvak G.S., Ismagilov I.Z., Sukhova O.B. Eurasian
Chem. Tech. J. 7 (2005) 115-121.
Chapter 2
17
Desilication and silylation of Mo/HZSM-5 for methane
dehydroaromatization
Summary
The influence of mesoporosity and silylation on the physico-chemical and catalytic properties
of Mo/HZSM-5 in methane dehydroaromatization was investigated. The zeolites were
characterized by XRD, 27
Al and 95
Mo NMR, UV–Vis, UV Raman and pyridine IR
spectroscopy and TEM. Base-desilicated mesoporous and bulk HZSM-5 zeolites with
comparable Brønsted acidity were employed as acidic supports. Mo loading was optimized to
minimize loss of acidity. Surface silylation of Mo/HZSM-5 resulted in improved Mo-oxide
dispersion. More intensive silylation led to decreased Mo-oxide dispersion because of
increased hydrophobicity. High methane conversion rates were associated with small Mo-
oxide precursor particles. Silylation of the external surface of Mo/HZSM-5 led to higher
methane conversion and less coke formation. On contrary, silylation of HZSM-5 prior to Mo
introduction had a negative effect on the performance. Post-synthesis silylation of Mo/HZSM-
5 affected the Mo-oxide phase. The amount of hard coke decreased with increasing silylation
degree due to deactivation of acid sites at the external surface. It also decreased naphthalene
yield. Methane conversion and aromatics selectivity were lower for mesoporous Mo/HZSM-5
compared with bulk Mo/HZSM-5. Although the initial Mo-oxide dispersion was higher, the
different nature of the mesopore surface resulted in rapid formation of large Mo-carbide
particles with higher coke selectivity. Silylation slightly improved activity and selectivity to
benzene.
This chapter is published in Microporous Mesoporous Mater. 203 (2015) 259-273.
Catalyst performance
Chapter 2
18
2.1 Introduction
Increasing crude oil prices related to the depletion of petroleum reserves are a strong
driver for the search of alternative feedstocks to produce fuels and chemicals. Natural gas is a
viable feedstock to facilitate the transition to a carbon-neutral energy supply. It is abundant
and the cleanest of all fossil resources. Currently, the price of shale gas is very low in certain
regions. The proved global natural gas reserves are estimated at ca. 180 trillion cubic metres
according BP’s statistical review of world energy [1]. A large share of these reserves is
located in remote areas and it is often produced as “associated” gas during the production of
crude oil [2]. The remote location makes exploitation of this associated gas economically
unfeasible [2] and [3]. Technologies to employ natural gas reserves are already available, e.g.,
by transport in liquefied form and by conversion into transportation fuels via the syngas
platform (a mixture of CO and H2). There is, however, also great interest in direct conversion
of methane into liquid chemicals. Direct methane oxidation to methanol remains a great
unsolved challenge. A more promising approach involves non-oxidative methane
dehydroaromatization (MDA), first described in 1993 by Wang et al. [4]. In this process,
methane is converted into aromatics, predominantly benzene, and hydrogen.
The preferred catalyst for MDA is the bifunctional Mo/HZSM-5 zeolite. At the start of
the reaction, the Mo-oxide (MoOx) phase reacts with CH4 to form molybdenum carbides
(MoCx) [5]. The MoCx phase is believed to activate methane and provide sites for C-C
coupling into ethylene [6]. The zeolitic Brønsted acid sites convert ethylene into benzene and
other aromatic molecules. A major drawback of this reaction is the low stability of
Mo/HZSM-5. The high temperatures employed, typically 973 K or higher required to activate
the strong C-H bonds of methane, result in extensive formation of carbonaceous deposits that
block the micropores and deactivate the catalyst [7-11]. The adverse effects of coking can be
lowered by introducing mesopores that reduce diffusion pathways and improve the efficiency
factor of the zeolite crystals [12, 13]. Previous studies have shown the beneficial effect on the
MDA reaction of (i) mesopore introduction by silicon extraction from HZSM-5 [13] and (ii)
carbon black templating in HZSM-5[14] and HMCM-22 [15].
Another way to improve the stability is to reduce the rate of carbon formation. The
Brønsted acid sites (BAS) located at the external surface have been implicated in the rapid
formation of large carbonaceous deposits [17], mostly polynuclear aromatic molecules too
large desorb from the zeolite surface [14]. These carbon deposits eventually block the
Chapter 2
19
micropore entrances and deactivate the catalyst [18-20]. Deactivation of the external BAS in
the parent catalyst might, therefore, be a strategy to suppress formation of this type of
coke [17-19].
In this study, we investigated the influence of mesopore introduction and external surface
BAS deactivation on the catalytic MDA performance of Mo/HZSM-5. We employed
desilication for generation of mesopores because of its low cost and facile optimization. We
first optimized the Mo introduction method for a bulk microporous zeolite, before applying
the optimized approach to the desilicated zeolite. The catalytic performance of the resulting
mesoporous Mo/HZSM-5 zeolite is compared with that of conventional Mo/HZSM-5 zeolite
with similar acidity. The second approach is to deactivate the BAS located at the external
surface by silylation with the aim to reduce coke formation. In this study, particular attention
is given to silylation before and after Mo introduction. The catalysts were extensively
characterized by FTIR, 27
Al MAS NMR, UV-Raman and UV–Vis spectroscopy. Also, spent
catalysts were characterized for changes in their textural properties. The influence of
desilication and silylation on the methane dehydroaromatization performance is discussed.
2.2 Experimental methods
2.2.1. Synthesis of materials
The parent zeolite (HZSM-5) was obtained in the ammonium form from Albemarle
Catalysts. Zeolites with different Si/Al ratios (19.4, 28.2 and 43.6 as determined by ICP
analysis) were used to optimize the desilication procedure. The zeolites are denoted as ZSM-
5(20), ZSM-5(30) and ZSM-5(40), respectively. For the introduction of mesopores in HZSM-
5, we optimized the base-catalyzed desilication procedure described by Groen et al. [20] by
varying the NaOH concentration and the temperature and time of base leaching (Table 2.1).
Typically, 1.66 g of dried ZSM-5 was suspended in a 50 ml NaOH solution. After desilication
the hot liquor was filtered and washed with copious amounts of deionized water. The
desilicated samples were exchanged three times with a 1 M NH4NO3 solution for 4 h at 353 K
followed by drying overnight at 383 K. To convert the ammonium form to the proton form,
the dried zeolite was calcined at 523 K for 4 h. The zeolites are denoted as HZSM-5 and
HZSM-5(meso) for the parent and desilicated samples, respectively.
Molybdenum oxide was introduced onto the zeolites by physical mixing with MoO3 or
incipient wetness impregnation with an aqueous solution of ammonium heptamolybdate
Chapter 2
20
tetrahydrate (AHM, Merck). For physical mixing, MoO3 and the zeolite were thoroughly
grinded in a mortar. For the incipient wetness impregnation procedure, the dried zeolite was
impregnated with a solution of appropriate concentration AHM. After impregnation the
samples were dried for 1 h. The targeted Mo content was 4 wt%. The Mo-loaded zeolites
were calcined in artificial air at varying temperatures and dwell times. The heating rate was
1.5 K/min. Mo-modified zeolites are denoted as Mo/HZSM-5 and Mo/HZSM-5(meso).
For silylation, the method described by Zheng et al. was used [21]. Typically, 2 g of
zeolite was dried overnight at 373 K and then dispersed in 50 ml n-hexane. To this
suspension, 0.3 ml tetraethylorthosilicate (TEOS, Merck) was added and stirred for 1 h under
reflux. The amount of TEOS corresponded to 0.4 wt% based on the amount of zeolite in the
suspension. Thereafter, the catalyst was filtered and dried overnight at 373 K. The samples
underwent a two-step calcination process in artificial air. The first step consisted of heating
the sample at a rate of a 2 K/min to 393 K followed by an isothermal period of 2 h. In the
second step, the temperature was further increased to 773 K at a rate of 0.2 K/min, followed
by an isothermal period for 4 h. Single and triple silylated treated Mo/HZSM-5 are
abbreviated as Mo/HZSM-5(Mo,Si1) and Mo/HZSM-5(Mo,Si3), respectively. Mo/HZSM-5
silylated before Mo introduction is referred to as Mo/HZSM-5(Si1,Mo). The same naming
method is used for the mesoporous zeolites, e.g., Mo/HZSM-5(meso,Mo,Si3).
2.2.2. Characterization
The Mo and Al contents of the samples were determined by inductively coupled plasma
optical emission spectroscopy (ICP-OES) on a Spectro CIROS CCD spectrometer equipped
with a free-running 27.12 MHz generator at 1400 W. Prior to analysis, samples were digested
in a mixture of HF/HNO3/H2O (1:1:1).
XRD patterns were recorded on a Bruker D4 Endeavor powder diffraction system using
Cu Kα radiation with a scanning speed 0.01° sec−1
in the 2θ range 5–60°. XRD crystallinities
were determined using the Bruker TOPAS 3.0 software.
Infrared spectra for the determination of IR crystallinity were recorded on a Nicolet Avatar
360 spectrometer with a KBr pellet (1 mg of zeolite in 100 mg of KBr). Zeolite crystallinity
was estimated from the ratio of the intensities of bands at 450 cm−1
and 550 cm−1
[22].
Argon sorption isotherms were measured at 87 K on a Micromeritics ASAP2020 system
in static measurement mode. The samples were outgassed at 623 K for 8 h prior to the
Chapter 2
21
sorption measurements. The Brunauer–Emmett–Teller (BET) equation was used to calculate
the specific surface area (SBET) from the adsorption data obtained (p/p0 = 0.05–0.25). The
mesopore volume (Vmeso) and mesopore size distribution were calculated using the Barrett–
Joyner–Halenda (BJH) method on the adsorption branch of the isotherm. The micropore area
(Smicro) and micropore volume (Vmicro) were calculated from the t-plot curve with the thickness
range being 3.5 and 5.4 Å [23].
Infrared spectra were recorded in the 4000–400 cm−1
range using a Bruker Vertex 70v
apparatus. Samples were pressed into a self-supporting wafer with a density of about
10 mg/cm2. To remove adsorbed water the sample was evacuated for 2 h at 773 K. After
evacuation the sample was cooled to 323 K followed by recording of the background
spectrum. The total concentration of the Brønsted acid sites was determined by measuring IR
spectra of adsorbed pyridine. Pyridine adsorption was carried out on the dehydrated zeolite
wafer at 423 K. After saturation was reached, the sample was evacuated at 423 K for 2 h and a
spectrum was recorded. The amount of Brønsted acid sites on the external surface was
determined by similar procedures using 2,4,6-collidine as the base. The spectra were
deconvoluted by standard procedures and for quantification the extinction coefficients
reported by Datka et al. [24] and Nesterenko et al. [25] were used for pyridine and 2,4,6-
collidine, respectively.
The surface composition of the samples was analyzed by X-ray Photoelectron Spectroscopy
(XPS) using a Thermo Scientific K-Alpha equipped with a monochromatic small-spot X-ray
source and a 180° double focusing hemispherical analyzer with a 128-channel detector.
UV–Vis spectra were recorded on a Shimadzu UV-2401 PC spectrometer in diffuse-
reflectance mode with a 60 mm integrating sphere. BaSO4 was used as the reference.
UV Raman spectra were recorded with a Jobin–Yvon T64000 triple stage spectrograph
with spectral resolution of 2 cm−1
. The laser line at 244 nm of a Lexel 95-SHG laser was used
as exciting source with an output of 20 mW. The power of the laser on the sample was about
2 mW. The excitation laser line at 325 nm was produced by a Kimmon He–Cd laser. The
power of the laser on the sample was 4 mW.
Magic angle spinning (MAS) 27
Al single pulse NMR spectra were recorded on a Bruker
Avance DMX-500 NMR spectrometer equipped with a 2.5 mm MAS probe head operating at
an 27
Al NMR resonance frequency of 130.3 MHz. The 27
Al chemical shift is referred to a
saturated Al(NO3)3 solution. In a typical experiment 10 mg of well-hydrated sample was
packed in a 2.5 mm zirconia rotor. The MAS sample rotation speed was 20 kHz. Single-pulse
Chapter 2
22
excitation was used with a 18° pulse of 1 μs and an interscan delay of 1 s. The relaxation time
was 1 s and the pulse length was 1 s.
95Mo NMR measurements were carried out on a Agilent 850 NMR spectrometer at
a 95
Mo NMR frequency of 55 MHz. MAS NMR spectra of 95
Mo-enriched Mo/HZSM-5
zeolites were recorded using a 4-mm MAS probehead and a sample rotation rate of 8 or
16 kHz. Reference spectra of MoO3 and Al2(MoO4)3 were recorded with a MAS rate of
40 kHz by use of a 1.6 mm MAS probehead. To obtain the strongest possible 95
Mo NMR
signal, spinning sideband double-frequency sweep (SSDFS) signal enhancement was used
prior to the observation pulse. To remove probe ringing effects from the spectra, the free
induction decays were left-shifted 10 points (20 μs at a spectral width of 500 kHz) prior to
applying the Fourier transformation.
Transmission electron micrographs were obtained with a FEI Tecnai 20 transmission
electron microscope (TEM) at an electron acceleration voltage of 200 kV. Typically, a small
amount of sample was suspended in ethanol, sonicated and dispersed over a Cu grid with a
holey carbon film. Scanning electron microscopy (SEM) was performed using a Philips
environmental FEIXL-30 ESEM FEG in high-vacuum mode at low voltage.
Weight-loss curves of spent catalysts after 12 h on stream in methane aromatization were
measured by thermogravimetric analysis (TGA) using a Mettler Toledo TGA/DSC 1
apparatus. Samples were heated in uncovered alumina crucibles at a rate of 5 K/min to
1023 K in a He/O2 mixture containing 33.3 vol% O2.
2.2.3. Catalytic activity measurements
An amount of 0.5 g of catalyst was introduced in a tubular quartz reactor with a length of
490 mm and an internal diameter of 4.0 mm. The catalyst was supported on a quartz wool
plug in the isothermal zone of the oven. All gases were fed using thermal mass controllers.
The temperature was increased at a rate of 5 K/min to 973 K in a He gas flow of 25 ml/min.
The reaction was started by switching the reactor feed to a N2/CH4(5 vol% N2, internal
standard) mixture at a WHSV of 1710 ml CH4/gcat h. Products were analyzed by an online
Interscience CompactGC gas chromatograph equipped with three analysis channels for
separate analysis of light gases (Molsieve 5A, TCD), light hydrocarbons (Al2O3/KCl, TCD)
and aromatics (Rtx-1, FID).
Chapter 2
23
2.3. Results and discussion
2.3.1. Preparation of hierarchical HZSM-5
To introduce mesoporosity in HZSM-5, the desilication procedure of Groen et al. was
employed [20]. We optimized this procedure in order to obtain a hierarchical HZSM-5 zeolite
with a Si/Al ratio similar to that of a conventional bulk HZSM-5 (Si/Al ≈ 20) for proper
comparison of catalytic performance in the MDA reaction. We also aimed to retain the high
crystallinity and Brønsted acidity of the parent sample as much as possible. HZSM-5 zeolites
with varying Si/Al ratios were treated with NaOH solutions at different conditions. Table 2.1
summarizes the most important physico-chemical properties as a function of the treatment
procedure. All parent zeolites already contain a small amount of mesopores. These mesopores
are predominantly related to interparticle voids. NaOH treatment increased the mesopore
volume. It is also seen that the mesopore volume increased with decreasing Al content of the
starting zeolite. For instance, treatment of HZSM-5(20) with 0.6 M NaOH for 0.5 h at 358 K
did not generate additional mesoporosity as compared with the parent zeolite, whereas
substantial mesoporosity was introduced in HZSM-5(40), concomitant with a strong decrease
of the XRD crystallinity. These trends are consistent with those reported by Groen et al. [20].
From Table 1, we selected treatment of treatment of HZSM-5(30) with a 0.2 M NaOH
solution for 0.5 h at 338 K as a suitable method to prepare the desired hierarchical starting
zeolite. This HZSM-5(meso) zeolite has a relatively high mesopore volume of 0.27 cm3/g and
its crystallinity loss due to the base treatment is minor. The physico-chemical properties of the
material prepared following the optimized desilication procedure (ZSM-5(meso)) were
characterized in more detail and are listed in Table 2.2.
Chapter 2
24
Table 2.1. Physico-chemical properties of parent and desilicated ZSM-5 zeolites.
Sample Si/Alparent CNaOH
(M)
Temperature
(K)
time
(h)
CRXRD1
(%)
Vmeso
(cm3/g)
Si/Alfinal
ZSM-5 21.3 - - - 100 0.03 21.3
ZSM-5(meso) 21.3 0.2 338 0.5 106 0.03 -
21.3 0.2 338 2.0 105 0.04 -
21.3 0.2 358 0.5 106 0.04 -
21.3 0.6 338 0.5 90 0.04 -
21.3 0.6 338 2.0 50 0.43 -
21.3 0.6 358 0.5 45 - -
ZSM-5 28.2 - - - 100 0.08 28.2
ZSM-5(meso) 28.2 0.2 338 0.5 89 0.27 19.5
28.2 0.2 338 2.0 83 0.36 17.2
28.2 0.2 358 0.5 83 - -
28.2 0.6 338 0.5 87 0.17 12.6
28.2 0.6 338 2.0 70 0.24 -
ZSM-5 43.6 - - - 100 0.11 43.6
ZSM-5(meso) 43.6 0.2 338 0.5 97 0.24 35.5
43.6 0.2 338 2.0 78 0.49 -
43.6 0.2 358 0.5 88 0.38 29.7
43.6 0.2 358 2.0 83 - -
43.6 0.6 338 0.5 22 0.9 9.4 1 Crystallinity.
Table 2.2. Physico-chemical properties of ZSM-5(meso) prepared under optimized
desilication conditions and the untreated ZSM-5 counterparts.
Sample Si/Alfinal
Vmicro
(cm3/g)
Vmeso
(cm3/g)
Smicro
(m2/g)
Smeso
(m2/g)
CRXRD1
(%)
Particle
size2
(μm)
ZSM-5 21.3 0.13 0.03 141 15 100 0.5
ZSM-5(meso) 19.5 0.14 0.31 147 114 89 0.8
ZSM-5 28.2 0.13 0.06 195 74 100 1.5 1 Crystallinity.
2 Particle size determined by SEM analysis.
2.3.2. Optimization Mo loading procedure
Mo was loaded onto the parent HZSM-5(20) zeolite by incipient wetness impregnation of
an AHM solution and by physical mixing with MoO3. The aim was to obtain a highly
dispersed Mo-oxide precursor phase and, at the same time, limit the extraction of framework
Al (FAl). The calcination temperature, calcination time and the Mo precursor (AHM vs.
MoO3) were varied. The Al distribution in the Mo-containing zeolites was investigated
by 27
Al MAS NMR spectroscopy (Fig. 2.1). The spectra contain four bands at chemical shifts
Chapter 2
25
of 55 ppm, 14 ppm, 0 ppm and −11 ppm. The feature at 55 ppm is due to FAl. The peak
around 0 ppm originates from extraframework Al (EFAl) atoms in octahedral coordination.
The signals at 14 and −11 ppm are due to Al2(MoO4)3, which are also part of the EFAl
phase [27-30]. The relative contributions of the various Al species obtained by deconvolution
of the NMR spectra are presented in Table 2.3. In HZSM-5, Al is predominantly present as
FAl. After impregnation with AHM and drying, a small amount of Al2(MoO4)3 was observed.
This phase was not observed upon physical mixing of MoO3 with the zeolite. All of these
samples contained no or at most a very small amount of Al2(MoO4)3, as long as the
calcination temperature was lower than 823 K. Above this temperature, the amount of
Al2(MoO4)3 increased significantly, and at the same time, the amount of EFAl species
characterized by the 0 ppm feature increased. The NMR spectra also show that the amount of
FAl species decreased, indicating that dealumination of the framework occurred [26].
Fig. 2.1 27
Al MAS NMR spectra of (left) AHM impregnated HZSM-5 with a) HZSM-5, b)
Mo/HZSM-5 (0 h, 298 K), c) Mo/HZSM-5 (5 h, 773 K), d) Mo/HZSM-5 (8 h, 773 K), e)
Mo/HZSM-5 (5 h, 823 K), f) Mo/HZSM-5 (5 h, 873 K) and g) Mo/HZSM-5 (5 h, 973 K). 27
Al MAS NMR spectra of (right) of MoO3 physical mixed HZSM-5 with a) HZSM-5,
Mo/HZSM-5 (0 h, 298 K), c) Mo/HZSM-5 (5 h, 773 K), d) Mo/HZSM-5 (5 h, 823 K), e)
Mo/HZSM-5 (5 h, 873 K) and f) Mo/HZSM-5 (5 h, 973 K). The information in between the
parentheses represents calcination time and calcination temperature.
Table 2.3 also summarizes the crystallinities of the Mo-containing zeolites as determined
by XRD and IR. The parent HZSM-5(20) zeolite is highly crystalline. Impregnation with an
100 50 0 -50
Inte
nsit
y (
a.u
.)
Chemical shift (ppm)
100 50 0 -50
Chemical shift (ppm)
Chapter 2
26
AHM solution and drying decreased the crystallinity. Physical mixing of HZSM-5 with
MoO3 did not affect the crystallinity when the calcination temperature was below 873 K.
Above this temperature, the crystallinity decreased. Together with the NMR data, we
conclude that extraction of Al by Mo-oxide resulting in Al2(MoO4)3 is the likely reason for
partial structural collapse of the zeolites.
Table 2.3. Physico-chemical properties of the Mo/ZSM-5 zeolite catalysts.
Sample Mo
introduction1
Tcalc
(K)
tcalc
(h)
CRXRD
(%)
CIR
(%)
AlIV
(%)
AlVI
(%)
AlAl2Mo2O4
(%)
ZSM-5 - - - 100 100 94 6 0
- 973 5 97 102 - - -
Mo/ZSM-5 IMP - - 80 97 82 16 3
IMP 773 2 81 101 - - -
IMP 773 5 81 93 70 21 9
IMP 773 8 81 93 73 20 7
IMP 823 5 78 110 82 14 4
IMP 873 2 49 74 - - -
IMP 873 5 46 60 53 29 18
IMP 873 8 47 55 - - -
IMP 973 2 23 55 - - -
IMP 973 5 33 53 43 23 34
IMP 973 8 28 49 - - -
Mo/ZSM-5 PM - - 102 102 95 5 0
PM 773 2 86 98 - - -
PM 773 5 77 96 72 18 10
PM 773 8 67 95 - - -
PM 823 5 82 98 79 20 1
PM 873 2 52 77 - - -
PM 873 5 32 51 49 29 22
PM 873 8 24 39 - - -
PM 973 2 28 49 - - -
PM 973 5 20 37 40 23 37
PM 973 8 22 40 - - -
ZSM-5(Meso) - - - 89 - 78 22 -
Mo/ZSM-5(Meso) IMP 823 5 89 - 66 25 9 1IMP = impregnation with an AHM solution, PM = physical mixing of MoO3 with zeolite.
Transmission electron micrographs of the impregnated sample calcined at 973 K points to
partial destruction of the zeolite (Fig. 2.2c) and formation of mesopores. Some of the cavities
created in the zeolite crystal are large enough to be seen in SEM images (Fig. 2.2f). Prolonged
calcination did not further change the morphology. The TEM images also point to better
dispersion of the Mo-oxide phase in the impregnated samples compared with the physically
Chapter 2
27
mixed samples (Fig. 2.2b). TEM images of the physically mixed sample show oval-shaped
particles with a typical size of ∼10 nm on the external zeolite surface.
Fig. 2.2. Transmission electron microscopy micrographs of Mo/HZSM-5 prepared by MoO3
(a) Mo/HZSM-5 (5 h, 773 K) and AHM impregnation (b) Mo/HZSM-5 (5 h, 773 K), (c)
Mo/HZSM-5 (5 h, 873 K), (d) Mo/HZSM-5 (5 h, 973K). Scanning electron micrographs of
(e) Mo/HZSM-5 (5 hours, 773K) and (f) Mo/HZSM-5 (5 h, 973K). The information in
between the parentheses represent calcination time and calcination temperature.
From all these observations, we selected AHM impregnation followed by calcination at
823 K for 5 h as the preferred method for the preparation of the Mo/HZSM-5 and applied it to
HZSM-5(meso). The XRD crystallinity of the parent HZSM-5(meso) was not affected by
introduction of Mo. However, the NMR spectrum of Mo/HZSM-5(meso) shown in Fig.
1 reveals that greater amounts of EFAl including Al2(MoO4)3formed upon calcination at
823 K compared with Mo/ZSM-5. Deconvolution of the relevant spectrum confirms this
(Table 2.3).
We also investigated the state of Mo as a function of the Mo precursor for HZSM-5 and
HZSM-5(meso) by 95
Mo MAS NMR spectroscopy. Fig. 2.3 shows NMR spectra of
Mo/HZSM-5 and Mo/HZSM-5(meso) with Mo introduced in two different ways. Spectra of
bulk MoO3 and Al2(MoO4)3 are included for comparison. The NMR detectable isotope 95
Mo
has a quadrupolar spin 5/2 and a low gyromagnetic ratio. To obtain the highest possible
signal, 95
Mo-enriched samples were prepared using 95
Mo-labeled MoO3 as the primary Mo
source. In addition, the 95
Mo NMR was run at ultrahigh magnetic field (20 T) and a specific
Chapter 2
28
pulse sequence (SSDFS) was used to enhance the NMR signal. All co-calcined mixed
Mo/HZSM-5 samples (Fig. 2.3b and c) show the same lineshape as bulk MoO3 (Fig. 2.3a).
This lineshape results from the second-order quadrupolar line-broadening, which cannot be
completely averaged by magic-angle spinning. The quadrupolar MAS NMR lineshape
observed is consistent with the one observed by Hu et al. [30], apart from a minor signal at
−73 ppm in the spectra of bulk MoO3 and co-calcined microporous Mo/HZSM5 (Figs. 2.3a
and b). This minor signal may reflect a Mo impurity in MoO3 with higher oxygen
coordination symmetry, because its chemical shift equals the isotropic shift (−73 ppm) of the
main signal component without the quadrupolar effect. The minor signal is not visible for co-
calcined mesoporous Mo/HZSM-5 (Fig. 3c), but this sample shows broad weak spectral
features in the range where one expects Al2(MoO4)3 features (Fig. 2.3d, [32]), which points to
formation of Al-O-Mo species at the internal surface of the mesopores. The spectra of the
Mo-impregnated zeolites are entirely different. For the impregnated microporous Mo/HZSM-
5 (Fig. 2.3e) a signal is observed around 0 ppm, which is practically the position of
(NH4)2MoO4 dissolved in water. It thus appears that impregnation followed by calcination
yields different MoOx species than the bulk-like MoO3 species resulting from physical mixing
and co-calcination. The signal is relatively narrow, suggesting fairly monodispersed Mo
species with highly symmetric oxygen coordination. In contrast, the 95
Mo MAS NMR signal
of the impregnated mesoporous Mo/HZSM-5 zeolites is extremely broad. This is indicative of
a broad variation of MoOx species with low oxygen coordination symmetry attached to the
internal surface of the mesopores.
Chapter 2
29
Fig 2.3. (thick lines) 95
Mo MAS NMR spectra of (a) bulk MoO3, (b,c) 5 wt.% 95
MoO3
physically mixed and co-calcined with (b) HZSM-5 and (c) HZSM-5(meso), (d) bulk
Al2(MoO4)3, (e,f) (NH4)2MoO4 -impregnated and calcined (e) microporous and (f) HZSM-
5(meso). The thin-line spectrum inserted below (a) reflects the simulated lineshape for a 95
Mo
spin with quadrupole-tensor asymmetry parameter h = 0.3. Spectra (a) and (b) contain an
additional signal at ~70 ppm, and spectrum (c) indicates the possible presence of weak signals
in the range of aluminium molybate (d). The shift range in (f) is larger than in (a-e). Sample
rotation rate: (a and d) 40 kHz, (b, c, e, f) 8 kHz; spinning sidebands are marked with a star.
2.3.3. Silylation
In order to lower the acidity of the external zeolite surface, its hydroxyl groups were
silylated. We employed the method described by Zheng et al. using TEOS [4]. The effect of
silylation was first investigated by inspection of the hydroxyl stretching regions of the IR
spectra (Fig. 2.4). Bands around 3610 cm−1
, 3665 cm−1
and 3740–3745 cm−1
are attributed to
bridging hydroxyl groups (denoted as BAS, Brønsted acid sites, further on), extraframework
hydroxyl groups and silanol groups, respectively. Comparison of the spectra of HZSM-5 and
HZSM-5(meso) shows that the mesoporous zeolite contains much more silanol groups. Also,
a more intense background extending to lower wavenumbers due to hydrogen-bonded silanol
groups is evident for this sample. On top of that, HZSM-5(meso) also contains a small
100 0 -100 -200 -300 -400 -500
95Mo NMR shift (ppm)
1000 500 0 -500 -1000
Chapter 2
30
amount of EFAl-related OH species, which are not present in HZSM-5. The introduction of
Mo by AHM impregnation and calcination led to relatively small changes in the IR spectra. A
small loss of BAS is noted. Silylation affected the IR spectra more substantially in the sense
that the silanol density strongly decreased. The decrease was more pronounced for
Mo/HZSM-5. These changes in the silanol band intensity suggest that the introduced Mo
reacted with the silanol groups. The reaction of Mo-oxides with silanol groups on silica upon
calcination has been described before [32]. It leads to dispersion of Mo-oxides into smaller
clusters such as tetrahedral di-oxo, pentahedral mono-oxo as well as polymeric species [33,
34]. It also causes a further decrease of the concentration of BAS. Repetitive silylation and
calcination had a different effect and resulted in a higher concentration of BAS, possibly due
to reinsertion of earlier formed EFAl into the zeolite framework under the influence of the
additional silicon species. When the parent zeolites were first silylated and then co-calcined
with physically admixed MoO3, the spectra were very similar to those of the two starting
materials.
Fig. 2.4. Infrared spectra in the hydroxyl region of (left) bulk HZSM-5 zeolite and (right)
mesoporous HZSM-5 zeolite with (a) the parent zeolite, (b) Mo/HZSM-5, (c) Mo/HZSM-
5(Mo,Si1), (d) Mo/HZSM-5(Mo,Si3) and (e) Mo/HZSM-5(Si1,Mo).
3800 3600 3400 3200
Ab
so
rban
ce (
a.u
.)
Wavenumber (cm-1)
3800 3600 3400 3200
Wavenumber cm-1
Chapter 2
31
2.3.4. Catalyst characterization
The resulting two sets of catalysts were characterized in more detail so as to obtain an
overview of the changes brought about by Mo introduction and silylation. The BAS and
Lewis acid sites (LAS) contents were determined by IR spectroscopy of adsorbed pyridine.
The presence of external acid sites was probed by 2,4,6-trimethylpyridine. In the latter case,
the absorption bands related to 2,4,6-trimethylpyridine coordination to LAS (1633 cm−1
) and
BAS (1638 cm−1
) overlap and attempts to deconvolute these spectra did not result in
meaningful results. Therefore, we present only the values for the total acid site densities on
the external surface. The results are collected in Table 2.4. The total acidity of HZSM-5 is
consistent with the total Al content. The mesoporous zeolite have slightly less BAS than
HZSM-5. The bulk BAS concentration is lowered, which can be attributed to the introduction
of mesopores. Kox et al. [35] have shown that mesopore introduction by silicon extraction
results in siliceous mesopore walls. The amount of external acid sites in HZSM-5(meso) is
only slightly higher than in HZSM-5, despite the much higher combined mesopore and
external surface area.
Chapter 2
32
Table 2.4. Textural and physical properties of the prepared silylated catalysts.
Sample Vmic
(cm3/g)
Vmeso
(cm3/g)
Smic
(m2/g)
Smeso
(m2/g)
Cryst.XRD1
(%)
Si/Al
(XPS)
Si/Mo
(XPS)
Al
(wt%)
Mo
(wt%)
NBAS
(μmol/g)
NLAS
(μmol/g)
Nacid,ext.
(μmol/g)
HZSM-5 0.13 0.03 141 15 100 22 ∞ 2.2 ∞ 679 214 61
Mo/HZSM-5 0.07 0.07 113 49 86 49 3 2.1 4.12 533 272 51
Mo/HZSM-5(Mo,Si1) 0.04 0.02 123 15 90 58 6 2.0 3.53 314 127 21
Mo/HZSM-5(Mo,Si3) 0.09 0.02 149 14 95 30 15 1.7 3.22 659 117 34
Mo/HZSM-5(Si1,Mo) 0.11 0.02 255 18 96 31 112 1.9 1.96 639 118 25
HZSM-5(Si1) 0.13 0.02 247 16 - - - - - 697 339 11
HZSM-5(Meso) 0.14 0.31 147 114 89 15 ∞ 2.2 ∞ 544 182 66
Mo/HZSM-5(Meso) 0.07 0.23 69 105 89 20 55 2.0 3.97 471 211 71
Mo/HZSM-5(Meso,Mo,Si1) 0.13 0.34 90 106 90 30 39 2.0 3.57 356 239 69
Mo/HZSM-5(Meso,Mo,Si3) 0.13 0.21 132 79 94 23 26 1.5 3.06 375 144 43
Mo/HZSM-5(Meso,Si1,Mo) 0.11 0.19 120 74 92 16 ∞ 1.7 4.35 554 164 66 1 Parent HZSM-5(30) was used as reference zeolite to determine the XRD crystallinity of HZSM-5(meso) and the zeolite materials derived from
HZSM-5(meso).
Chapter 2
33
Modification of the parent HZSM-5 and HZSM-5(meso) with Mo decreased the BAS
content due to the exchange of protons with Mo and/or the extraction of framework Al [36].
At the same time, the LAS concentration increased. The Lewis acidity originates from EFAl,
partly in the form of Al2(MoO4)3, but likely also from Mo-oxo species [37]. Single and triple
silylation of Mo/HZSM-5 zeolites strongly affected the acidity. Silylation resulted in strong
decrease of BAS and LAS concentrations for Mo/HZSM-5. For Mo/HZSM-5(meso), the
same decrease in the BAS concentration is observed, yet the LAS concentration slightly
increased. This is due to the higher dispersion of Mo-oxo species when deposited on HZSM-
5(meso). Triple silylation recovers a significant part of the BAS in Mo/HZSM-5(Mo,Si3).
Such behavior was not observed for Mo/HZSM-5(meso,Mo,Si3). The results for the samples
that were silylated prior to introduction of Mo are different. Silylation did not lower the
internal acidity, and in this case the external acidity was substantially lowered. The same
effect is seen when silylation was done on HZSM-5 (HZSM-5(Si1)). It suggests that the less
pronounced changes in the external acidity content upon silylation for the Mo-containing
zeolites is caused by the Lewis acidity of the dispersed Mo-oxide phase.
The textural properties determined by Ar physisorption are also reported in Table 2.4.
Introduction of Mo led to smaller micropore volumes of HZSM-5 and HZSM-5(meso). The
micropore volume of Mo/HZSM-5 decreased further upon silylation. Silylation of the parent
HZSM-5 itself did not decrease the micropore volume in line with literature data [21]. Thus,
the presence of the Mo-oxide phase in Mo/HZSM-5 affects the silylation process. Together
with the changes in the concentration of BAS upon silylation for HZSM-5 and Mo/HZSM-5,
we infer that silylation of the internal zeolite surface is more substantial for the Mo-containing
sample. This possibly relates to the lower silanol density at the external surface after Mo
introduction. The presence of the additional mesoporosity induced by the Mo deposition and
calcination processes (Table 2.4) may also facilitate the diffusion of the silylating agent inside
the zeolite crystals. Repetitive silylation resulted in a higher micropore volume and a decrease
of the mesopore volume. A minor decrease in micropore volume was observed for Mo/ZSM-
5(Si1,Mo). All of these changes are trendwise similar but less pronounced for the mesoporous
HZSM-5 zeolite.
The wide-angle XRD patterns (Fig. 2.5) recorded for most of the zeolites were similar to
the pattern of HZSM-5(20). The patterns of Mo/HZSM-5(Si1,Mo) and Mo/HZSM-
5(meso,Si1,Mo) contain additional reflections at 12.8°, 25.7° and 38.9°, which belong to bulk
MoO3. This result points to the low dispersion of the Mo-oxide phase after physical mixing
Chapter 2
34
with MoO3. In all cases, the relatively small changes in the intensities of the XRD reflections
belonging to MFI show that crystallinity was only affected slightly upon Mo introduction
(Table 2.4).
Fig. 2.5. Wide angle XRD patterns of (left) bulk HZSM-5 zeolite and (right) mesoporous
HZSM-5 zeolite with (a) the parent zeolite, (b) Mo/HZSM-5, (c) Mo/HZSM-5(Mo,Si1), (d)
Mo/HZSM-5(Mo,Si3) and (e) Mo/HZSM-5(Si1,Mo).
Fig. 2.6 shows the 27
Al MAS NMR spectra of these zeolite catalysts. The samples
derived from the microporous ZSM-5 zeolite contained mainly FAl species with a small
fraction present as EFAl. None of these samples contained Al2(MoO4)3. The FAl peak is seen
to decrease upon introduction of Mo-oxide and silylation. Concomitantly, the tetrahedral Al
feature becomes broader. Tessonier et al. ascribed this behavior to the changing chemical
environment of the Al species upon exchange of Mo-oxo species with the BAS [26]. The
NMR spectrum of Mo/ZSM-5(Si1,Mo) is similar to that of the parent HZSM-5, indicating
that Mo did not affect the BAS as is clearly visible for the other samples. Al2(MoO4)3 was
only present in the Mo/HZSM-5(meso) samples. In all Mo-containing samples prepared by
impregnation, the FAl signal is much weaker and broadened upon Mo introduction and
further silylation. All this indicates that a portion of the BAS was replaced by Mo-oxo
species, resulting in perturbation of the symmetry around the tetrahedral Al species, making
them partially NMR-invisible.
10 20 30 40 50 60
cp
s (
a.u
.)
Angle (2)
10 20 30 40 50 60
cp
s (
a.u
.)
Angle (2)
Chapter 2
35
Fig. 2.6. 27
Al-MAS-NMR of (left) bulk HZSM-5 zeolite and (right) mesoporous HZSM-5
zeolite with (a) the parent zeolite, (b) Mo/HZSM-5, (c) Mo/HZSM-5(Mo,Si1), (d)
Mo/HZSM-5(Mo,Si3) and (e) Mo/HZSM-5(Si1,Mo).
XPS was used to study the external region of the zeolite crystals. Table 2.4 lists the
atomic Si/Al and Si/Mo ratios. The bulk Mo and Al contents are also tabulated here. It was
found that the Si/Al ratio in the surface region increased upon introduction of Mo. This
suggests that the Al present in the surface region preferentially interacts with the Mo-oxo
phase. For both series, triple silylation resulted in a decrease of the Si/Al ratio. These changes
were less pronounced for the sample that were first silylated. The changes in the Si/Mo ratios
upon silylation point to loss of Mo-oxide dispersion. The sample with the lowest Mo-oxide
dispersion has the highest XPS Si/Mo ratio. Note that the changes in Al and Mo content show
the expected trend following deposition of Si by the silylation steps. The Mo content of
Mo/ZSM-5(Si1,Mo) is lower than expected.
The nature of the Mo-oxide phase was characterized by UV–Vis and UV Raman
spectroscopy. The UV–Vis spectra shown in Fig. 2.7 contain bands characteristic for different
forms of Mo-oxide [38]. The absorption bands between 200 and 400 nm are assigned to
ligand to metal CT transitions (O2−
→ Mo6+
). CT bands in the range 210–250 nm and around
280 nm have typically been assigned to isolated monomeric Mo-oxo species [39, 40]. These
bands are related to tetrahedral Mo species, while those at higher wavelengths (300–330 nm)
are due to isolated Mo-oxo centers in octahedral geometry [41]. Jezlorowski and Knözinger
100 50 0 -50
Chemical shift ppm
100 50 0 -50
Inte
nsit
y (
a.u
.)
Chemical shift (ppm)
Chapter 2
36
contended that the band around 280 nm is more likely due to Mo-O-Mo containing
structures [42]. The tail into the 400–450 nm region is indicative for the presence of larger
molybdate clusters or MoO3crystallites [43]. Fig. 2.7 shows that single silylation of
Mo/HZSM-5 resulted in an overall increase of the intensity of the UV–Vis spectrum. This
sample contains more tetrahedral Mo-oxo species. Triple silylation resulted in Mo/ZSM-
5(Mo,Si3), whose spectrum contains less intense bands of highly dispersed Mo species and
more intense bands in the 400–450 nm range. This points to agglomeration of the Mo-oxide
phase, consistent with the XPS Si/Mo trend. The spectrum of Mo/ZSM-5(Si1,Mo), obtained
by silylation prior to Mo introduction, looks very different. The comparatively low intensities
in the molybdate region below 330 nm and the intense feature around 400–450 nm show that
Mo is mainly present as MoO3. This is due to the hydrophobic nature of the external surface
of the silylated HZSM-5(Si1) zeolite. The spectral trends for the mesoporous zeolites were
very similar and are therefore not discussed in detail.
Fig. 2.7. UV–Vis of (left) bulk HZSM-5 zeolite and (right) mesoporous HZSM-5 zeolite with
Mo/HZSM-5 (solid line), Mo/HZSM-5(Mo,Si1) (dashed line), Mo/HZSM-5(Mo,Si3) (dotted
line) and Mo/HZSM-5(Si1,Mo) (dashed-dotted line).
The zeolites were also investigated by UV Raman spectroscopy. The spectra obtained by
excitation with 244 and 325 nm lasers are shown in Fig. 2.8 and Fig. 2.9, respectively. Due to
the resonance Raman effect, the use of specific laser excitation lines enhances the spectral
features of highly dispersed, isolated tetrahedral molybdates (244 nm) and polymeric
200 300 400 500 600 700 800
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Ab
so
rban
ce
Wavenumber cm-1
200 300 400 500 600 700 800
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Wavenumber cm-1
Chapter 2
37
octahedral molybdates (325 nm). It should also be noted that the use of the 244 nm laser will
probe more of the surface region of the zeolite crystals than the use of the 325 nm laser,
whose light will penetrate deeper into the crystals [44]. For comparison, the Raman spectra of
bulk MoO3and Al2(MoO4)3 were also recorded.
Raman spectra for the microporous zeolites recorded upon excitation with the 244 nm
laser are shown in Fig. 2.8. The spectrum of HZSM-5 shows typical bands of MFI
zeolite [45]. Modification of HZSM-5 with Mo (Mo/HZSM-5) and silylation (Mo/HZSM-
5(Si1,Mo) and Mo/HZSM-5(Si3,Mo)) resulted in substantial spectral changes. The peaks
characteristic for HZSM-5 are not clearly observed anymore. Instead, new molybdenyl
stretching mode bands (960–970 cm−1
and 980–995 cm−1
[46-48]) are present due to small
Mo-oxo species [49-52]. The decreased intensity of the HZSM-5 bands and appearance of
Mo-oxo bands points to the coverage of the zeolite external surface with a Mo-oxide phase. It
is worthwhile to discuss the differences in the spectra following various treatments. The
spectrum of Mo/HZSM-5 contains a pronounced feature at 970 cm−1
due to the Mo=O stretch
of dimeric Mo structures inside the micropores [53] with a shoulder at 953 cm−1
(highly
dispersed octahedral molybdenum surface species [54]) on top of the broad 960–
970 cm−1
band. The peak at 970 cm−1
is not visible for Mo/HZSM-5(Mo,Si1) and Mo/HZSM-
5(Mo,Si3). The shoulder at 953 cm−1
has lower intensity for Mo/HZSM-5(Mo,Si1) and is
absent for Mo/HZSM-5(Mo,Si3). Bands due to aluminum molybdates are not observed. The
small feature at 360 cm−1
is a Mo-O bending mode [39, 55]. The subtle changes upon
silylation point to Mo-oxide agglomeration. The spectrum of Mo/HZSM-5(Si1,Mo) is very
different and resembles the spectrum of HZSM-5. This is caused by the low dispersion of the
MoO3 phase, so that Raman spectroscopy mainly probes the surface of the zeolite crystals.
The spectra for HZSM-5(meso) and thereof derived Mo-containing samples are collected
in Fig. 2.8. Different from the spectra for bulk HZSM-5, these contain a shoulder at
1025 cm−1
indicative of the presence of Al2(MoO4)3 [56], and a shoulder at 953 cm−1
of highly
dispersed octahedral molybdenum species. Upon single silylation of Mo/HZSM-5(meso,Mo)
a new feature at 849 cm−1
appeared due to β-MoO3 [57]. This is an intermediate phase during
transformation of amorphous Mo-oxides towards thermodynamically more stable α-
MoO3[57]. After further silylation the β-MoO3 peak was not observed anymore.
Chapter 2
38
Fig. 2.8. UV-Raman spectra recorded with a 244 nm laser of (left) bulk HZSM-5 and (right)
mesoporous HZSM-5 with (a) HZSM-5 (b) Mo/HZSM-5, (c) Mo/HZSM-5(Mo,Si1), (d)
Mo/HZSM-5(Mo,Si3), (e) Mo/HZSM-5(Si1,Mo), (f) bulk MoO3 and (g) bulk Al2(MoO4)3.
The spectra recorded with a 325 nm excitation laser are shown in Fig. 2.9. Due to
fluorescence interference, no clear features could be observed for Mo/HZSM-5(Si1,Mo). The
spectrum of Mo/HZSM-5 strongly resembles the spectrum of amorphous Mo-oxide
phases [57]. It is characterized by weak and broad bands at ∼860 cm−1
and ∼960 cm−1
. After
single and triple silylation of Mo/HZSM-5 peaks became visible at 280, 336 and 995 cm−1
.
They point to the formation of a microcrystalline α-MoO3 phase upon silylation [53, 57]. The
band at 820 cm−1
can be ascribed to α-MoO3 in close contact with the support surface[53, 57].
The spectrum still contains weak bands around 860 and 960 cm−1
due to amorphous Mo-
oxides. The spectra for the HZSM-5(meso)-derived samples are given in Fig. 2.9. The
findings are very similar with those for the microporous zeolite, albeit that the agglomeration
of dispersed Mo-oxo species into α-MoO3 was less pronounced for the mesoporous
Mo/HZSM-5 set. The latter is consistent with the observation of the intermediate β-
MoO3 phase in Fig. 2.8.
200 400 600 800 1000
Inte
nsit
y (
a.u
.)
Wavenumber (cm-1)
200 400 600 800 1000
Wavenumber (cm-1)
Chapter 2
39
Fig. 2.9. UV-Raman spectra recorded with a 325 nm laser of (left) bulk HZSM-5 and (right)
mesoporous HZSM-5 with (a) HZSM-5 (b) Mo/HZSM-5, (c) Mo/HZSM-5(Mo,Si1), (d)
Mo/HZSM-5(Mo,Si3), (e) Mo/HZSM-5(Si1,Mo), (f) bulk MoO3 and (g) bulk Al2(MoO4)3.
Careful analysis of UV–Vis and UV Raman spectra of Mo/HZSM-5 and Mo/HZSM-
5(meso) highlights specific changes of the Mo-oxide phase during silylation. The main effect
is transformation of small Mo-oxo species into larger crystallites upon extensive silylation.
The silylating agent competes with the Mo species for interaction with the zeolite surface. The
data show that the Mo-oxide phase is initially better dispersed in Mo/HZSM-5(meso)
compared with Mo/HZSM-5. This relates to the higher external (mesopore) surface area. The
higher dispersion also retards formation of α-MoO3 in comparison to the microporous zeolite
samples. As a consequence, the samples contain some β-MoO3. Silylation prior to Mo
introduction resulted in poor dispersion of the Mo-oxide phase.
The morphology of the Mo phase in the microporous and mesoporous sample series was
analyzed by TEM. The micrograph of Mo/HZSM-5 (Fig. 2.10a) revealed the presence of dark
oval-shaped spots at the crystal surface due to the Mo-oxide phase. After single silylation
treatment of Mo/HZSM-5 (Mo/HZSM-5(Mo,Si1)), it is difficult to observe these particles
(Fig. 2.10b). It points to the redispersion Mo-oxide into small particles. Inspection of the
micrographs after triple silylation (Mo/HZSM-5(Mo,Si3)) shows that larger particles were
formed (Fig. 2.10c). These particles are similar in size as those in Mo/HZSM-5. The
agglomeration of the initially small MoO3 particles is probably related to the weak interaction
200 400 600 800 1000
In
ten
sit
y (
a.u
.)
Wavenumber (cm-1)
200 400 600 800 1000
Wavenumber (cm-1)
Chapter 2
40
between the Mo phase and the zeolite surface after desilication. The micrograph of
Mo/HZSM-5(Si1,Mo) contains a large particle at the zeolite surface. Formation of relatively
large MoO3 crystallites is consistent with the XRD patterns. The TEM micrographs recorded
for the mesoporous sample series show the presence of large, agglomerated MoO3 particles.
The trends upon silylation are similar to those observed for the microporous sample. For
instance, single silylation of mesoporous Mo/HZSM-5 led to redispersion of the oval-shaped
Mo-oxide into smaller particles (inset in Fig. 2.10f). The Mo dispersion in Mo/HZSM-
5(Si1,Mo) was low (Fig. 10h), which agrees with the XRD findings.
Fig. 2.10. Transmission electron microscopy micrographs of bulk HZSM-5 zeolite of (a)
Mo/HZSM-5, (b) Mo/HZSM-5(Mo,Si1), (c) Mo/HZSM-5(Mo,Si3) and (d) Mo/HZSM-5. The
TEM micrographs of mesoporous HZSM-5 zeolite of (a) Mo/HZSM-5, (b) Mo/HZSM-
5(Mo,Si1), (c) Mo/HZSM-5(Mo,Si3) and (d) Mo/HZSM-5.
2.3.5. Catalytic activity measurements
The effect of the silylation treatment on the catalytic performance in the MDA was then
studied. The results for the catalysts derived from bulk HZSM-5 are shown in Fig. 2.11. For
all samples, the methane conversion rate and benzene selectivity decreased with time on
stream (Fig. 2.11a). The methane conversion rates were highest for Mo/HZSM-5(Mo,Si1) and
Mo/HZSM-5(Mo,Si3). The deactivation rate was similar for Mo/HZSM-5 and Mo/HZSM-
5(Mo,Si1). Compared with these two samples, triple silylated Mo/HZSM-5(Mo,Si3)
deactivated at a slightly lower rate. The highest benzene selectivities were observed for
Chapter 2
41
Mo/HZSM-5(Mo,Si1) and Mo/HZSM-5(Mo,Si3) with values of ∼80 and ∼70 wt%,
respectively. The total aromatics selectivity including toluene, xylenes and naphthalene for
these two materials were 85 wt% and 81 wt%, respectively. These selectivities were
substantially higher than the selectivities for Mo/ZSM-5 (benzene and total aromatics
selectivities of 50 wt% and 60 wt%, respectively). On the contrary, the methane conversion
rate and benzene/aromatics selectivity for Mo/HZSM-5(Si1,Mo) were lower compared to
Mo/HZSM-5. The selectivity trends displayed in Fig. 2.11b and c further show that the
benzene selectivity decreased with time on stream concomitant with an increase of the
ethylene selectivity. This points to deactivation of BAS. Notably, for Mo/HZSM-5 and
Mo/HZSM-5(Si1,Mo) the aromatics product distribution shifted from benzene towards
naphthalene after prolonged reaction. In comparison, the changes in aromatics product
distribution with time on stream for Mo/HZSM-5(Mo,Si1) and Mo/HZSM-5(Mo,Si3) were
much smaller and less naphthalene was formed.
Chapter 2
42
Fig. 2.11. Methane aromatization reaction rate (a) and selectivities to benzene (open symbols) and naphthalene (closed symbols) (b) and
ethylene (open symbols) and coke (closed symbols) (c) over bulk HZSM-5 with (■) Mo/HZSM-5, (●) Mo/HZSM-5(Mo,Si1), (▲)
Mo/HZSM-5(Mo,Si3) and ( ) Mo/HZSM-5(Si1,Mo).
0 2 4 6 8 10
0
2
4
6
8
10
12
14
CH
4 r
eacti
on
rate
(m
mo
l/h
.gc
at)
Time on stream (h)
0 2 4 6 8 10
0
20
40
60
80
100
Sele
cti
vit
y (
wt%
)
Time on stream (h)
0 2 4 6 8 10
0
20
40
60
80
100
Time on stream (h)
0
10
20
30
40
Chapter 2
43
Fig. 2.12. Methane aromatization reaction rate (a) and selectivities to benzene (open symbols) and naphthalene (closed symbols) (b) and
ethylene (open symbols) and coke (closed symbols) (c) over mesoporous HZSM-5 with (■) Mo/HZSM-5, (●) Mo/HZSM-5(Mo,Si1), (▲)
Mo/HZSM-5(Mo,Si3) and ( ) Mo/HZSM-5(Si1,Mo).
0 2 4 6 8 10
0
20
40
60
80
100
Time on stream (h)
0
5
10
15
20
25
30
0 2 4 6 8 10
0
20
40
60
80
100
Sele
cti
vit
y (
wt%
)
Time on stream (h)
0 2 4 6 8 10
0
2
4
6
8
10
12
14
CH
4 r
eacti
on
rate
(m
mo
l/h
.gc
at)
Time on stream (h)
Chapter 2
44
The reactivity trends for the catalysts based on HZSM-5(meso) were similar, albeit that
the overall methane conversion rates of these catalysts were lower and the deactivation was
also more severe (Fig. 2.12). Similar to the bulk Mo/HZSM-5 catalysts silylation of
Mo/HZSM-5(meso) improved the benzene selectivity. However, the magnitude of this effect
was smaller for the mesoporous than for the microporous zeolites. The mesoporous sample
silylated prior to Mo introduction also exhibited higher benzene selectivity than the
Mo/HZSM-5(meso) reference. All of the Mo/HZSM-5(meso) catalysts had lower methane
conversion rates and produced less benzene and more coke than their microporous analogs.
The nature of the carbonaceous deposits formed during reaction was characterized by
oxidation by TGA. Spent samples were retrieved after 12 h on stream in the MDA reaction
(Fig. 2.13). The TG curves contain two main features and a small negative feature around
∼700 K, which is attributed to the adsorption of oxygen on molybdenum carbide species and
their subsequent oxidation [39, 58, 59]. The main oxidation feature around ∼750 K belongs to
aliphatic hydrocarbon species, likely formed in the proximity of Mo (soft coke) [39, 58, 59].
The other feature, which occurs around 840 K, represents hard coke and is likely the result of
polynuclear aromatic species, whose formation is catalyzed by BAS [39, 58, 59]. The results
of deconvolution of these TG curves are given in Table 2.5. Consistent with the selectivity
differences observed in Fig. 2.11 and Fig. 2.12, spent Mo/HZSM-5(meso) zeolites contained
more coke than spent Mo/HZSM-5 zeolites. For Mo/HZSM-5, the amount of coke decreased
with increasing silylation degree. Most notably, it resulted in a strong decrease of the hard
coke, while the amount of soft coke varied much less. The textural properties of these spent
zeolites (Table 2.5) show that, except for Mo/HZSM-5(Mo,Si3), all of the spent Mo/HZSM-5
catalysts did not contain accessible micropores anymore. Spent Mo/HZSM-5(Mo,Si3)
contained much less coke, presumably because of deactivation of the external BAS.
Consistent with this, some of the micropores in the spent sample are still accessible. The spent
Mo/HZSM-5(meso) catalysts contained more coke than their microporous counterparts. Spent
Mo/HZSM-5(meso,Mo,Si1) and Mo/HZSM-5(meso,Mo,Si3) contained also substantially
more soft coke. The lower amount of hard coke, likely due to the weak acidity of the
mesopore walls, is also evident from the less significant decrease of micropore accessibility
for spent Mo/HZSM-5(meso) samples.
Chapter 2
45
Fig. 2.13. TGA weightloss curves of (left) bulk HZSM-5 zeolite and (right) mesoporous
HZSM-5 of (a) Mo/HZSM-5, (b) Mo/HZSM-5(Mo,Si1), (c) Mo/HZSM-5(Mo,Si3) and (d)
Mo/HZSM-5(Si1,Mo).
Table 2.5. Textural properties and coke content (C) of spent Mo/ZSM-5 catalysts after 12 h
on stream in the MDA reaction.
Sample Vmicro
(cm3/g)
Vmeso
(cm3/g)
Smicro
(m2/g)
Smeso
(m2/g)
Ctotal
(g/gcat)
Csoft
(g/gcat)
Chard
(g/gcat)
Mo/HZSM-5 0 0.02 0 8 0.13 0.022 0.108
Mo/HZSM-5(Mo,Si1) 0 0.02 0 8 0.10 0.024 0.076
Mo/HZSM-5(Mo,Si3) 0.11 0.02 149 15 0.03 0.030 0
Mo/HZSM-5(Si1,Mo) 0 0.02 0 9 0.10 0.031 0.079
Mo/HZSM-5(meso) 0.05 0.16 21 91 0.19 0.095 0.095
Mo/HZSM-5(meso,Mo,Si1) 0.06 0.11 35 65 0.21 0.187 0.023
Mo/HZSM-5(meso,Mo,Si3) 0.0 0.31 0 179 0.20 0.152 0.048
Mo/HZSM-5(meso,Si1,Mo) 0.05 0.10 45 51 0.16 0.088 0.072
2.3.6. General discussion
One of the goals of the present study was to determine the influence of mesoporosity in
ZSM-5 on the performance of Mo/HZSM-5 catalysts in the dehydroaromatization of methane.
Base leaching was employed to obtain a hierarchical HZSM-5 zeolite with a Si/Al ratio of 20,
comparable in acidity with the acidity of a conventional HZSM-5 zeolite. These two zeolites
formed the basis for the synthesis, characterization and catalytic activity measurements of
Mo/HZSM-5. Silylation of the samples before and after Mo loading was employed as a
method to deactivate the external surface.
600 700 800 900 1000
Weig
htl
oss (
mg
/K)
Temperature (K)
600 700 800 900 1000
Temperature (K)
Chapter 2
46
During Mo modification of HZSM-5 (AHM impregnation vs. mixing with MoO3) the
final physico-chemical properties were strongly influenced by the calcination step. When the
samples were calcined above 823 K in artificial air, it was always seen that extensive
Al2(MoO4)3 formation took place, concomitant with partial destruction of the zeolite
framework. Calcination at lower temperatures prevented dislodging of Al from the framework
in Mo/HZSM-5, when the starting zeolite had only micropores. The use of incipient wetness
impregnation with AHM and calcination at 823 K for 5 h resulted in Mo/HZSM-5 without
Al2(MoO4)3formation and minor loss of Brønsted acidity. The small loss of acidity is likely
due to partial ion-exchange with mobile Mo-oxo complexes [60]. Mo-oxide was mainly
present at the external surface as an amorphous oxide phase in isolated or slightly aggregated
forms. Preparation of Mo/HZSM-5 from mesoporous HZSM-5 led to formation of a greater
amount of Al2(MoO4)3. This is attributed to a higher fraction of molybdates entering the
micropores during impregnation or subsequent as a consequence of the increased
mesopore/external surface area to micropore surface area ratio of the desilicated zeolite.
Comparatively, Mo/HZSM-5 prepared by physical mixing of MoO3 with HZSM-5 always led
to low dispersion of the MoO3phase and there was little influence on the acidity as long as the
calcination temperature did not exceed 823 K.
Upon silylation significant changes were seen in the textural properties, the acidity and
the nature of the Mo-oxide phase. Consecutive silylation steps led to further changes. When
the silylation was done in the conventional manner, that is before Mo loading, it was found
that AHM wetness impregnation was not effective due to the considerable hydrophobicity of
the zeolite support. Therefore, we employed physical mixing with MoO3. Expectedly, the
result was very similar with results for the sample prepared by physical mixing of the starting
zeolite with MoO3. That is, the Mo-oxide dispersion was low and the Mo loading did not
affect the acidity.
A single silylation step of Mo/HZSM-5 resulted in improved dispersion of the Mo-oxide
phase. The increased Mo spreading resulted in a higher fraction of isolated Mo species, a
decrease of the concentration of internal and external BAS as a result of their reaction with
Mo-oxo species and minor decrease of the microporosity. For the mesoporous Mo/HZSM-5
zeolite similar but less pronounced changes were seen. When the silylation procedure was
repeated another two times on the same sample (3 times in total), the samples exhibited very
different properties. More complete silylation resulted in a substantial decrease of the Mo-
oxide dispersion as evidenced by UV–Vis and UV Raman spectroscopy. In addition, it was
Chapter 2
47
seen that, compared with single-silylated Mo/HZSM-5, the Brønsted acidity increased. 27
Al
NMR spectra show an increase of framework Al species. It is speculated that more intensive
silylation led to migration of Mo species from the micropores to the external surface and also
to partial removal of charge-compensating extraframework Al species. All this is consistent
with the increase of the micropore volume. Again, these differences were trendwise similar
for the mesoporous HZSM-5 zeolite, but less pronounced.
Next, we discuss the influence of these changes on the MDA activity. Comparing the
bulk HZSM-5 derived catalysts, it is seen that the poor Mo dispersion in Mo/HZSM-
5(Si1,Mo) is the cause of the relatively low methane conversion rates. The coke selectivity of
this silylated catalyst was also higher than of the other ones. High coke selectivity has been
related to large MoCx particles by Solymosi et al. [61], but can also be interpreted in this
particular case in terms of high Brønsted acidity on the external surface because of the low
Mo dispersion. The improved Mo dispersion in Mo/HZSM-5(Mo,Si1) and Mo/HZSM-
5(Mo,Si3) resulted in higher initial methane conversion rates. These silylated catalysts also
produced much less coke. With increasing silylation degree the amount of hard coke
decreased and it was nearly absent in spent Mo/HZSM-5(Mo,Si3). An important finding is
that the silylated samples displayed much lower rate of naphthalene formation than the non-
silylated ones. We explain this by deactivation of the external BAS, which can catalyze
formation of polynuclear aromatic hydrocarbons.
It was found that a higher proportion of the Mo species interacted with the acid sites in
the micropores of the mesoporous Mo/HZSM-5 zeolites. Although the initial dispersion of the
Mo-oxide phase in mesoporous Mo/HZSM-5 was high, pretreatment in He followed by
carburization resulted in sintering of the Mo-carbide particles, likely because of the siliceous
nature of the mesopore surface. The low dispersion of the Mo-carbide phase resulted in lower
methane conversion rates [62, 63]. These catalysts also produced more coke deposits than the
microporous ones. Analysis of the spent catalysts shows that these deposits were mainly of
the soft coke type. It appears that the coke was predominantly formed on the larger Mo-
carbide particles. The lower hard coke content is consistent with the lower acidity of the
mesopore/external surface.
Chapter 2
48
2.4. Conclusions
The influence of mesopores generated by zeolite desilication and silylation on the
catalytic performance of Mo/HZSM-5 for the dehydroaromatization of methane was
investigated. The base desilication procedure for HZSM-5 was optimized to have a mesopore-
containing ZSM-5 starting material comparable in acidic properties to a bulk HZSM-5 zeolite
(Si/Al = 20). The physico-chemical properties of Mo-modified HZSM-5 (AHM impregnation
or physical mixing with MoO3) were strongly influenced by the calcination step. The
optimum calcination temperature for high Mo dispersion without extensive
Al2(MoO4)3 formation and loss of Brønsted acidity was around 823 K. Mo/HZSM-5 prepared
by physical mixing of MoO3 had low MoO3 dispersion and acidity was not affected upon
calcination at 823 K. Silylation also led to significant changes. A single silylation step of
Mo/HZSM-5 resulted in improved Mo-oxide dispersion, a small loss of the internal and
external acidity and loss of external silanol groups. Repeated silylation and calcination led to
decreased Mo-oxide dispersion because of the high hydrophobicity. Brønsted acidity was
increased upon repetitive silylation which was attributed to removal of extraframework
aluminum exchanged to BAS. For methane dehydroaromatization high dispersion of the Mo-
oxide precursor is beneficial for high methane conversion rate and low rate of coke formation.
High Mo-oxide dispersion of single-silylated Mo/HZSM-5 led to higher methane conversion
rates and lower coke selectivity. Increasing the silylation degree resulted in less hard coke
formation, because the acid sites at the external surface become deactivated. The naphthalene
yield is lowest for the repeatedly silylated catalysts. Introducing mesoporosity in bulk HZSM-
5 does not improve the catalytic performance in methane dehydroaromatization. Both activity
and aromatics selectivity are lower than for bulk Mo/HZSM-5. Also, the effect of silylation is
less favorable for mesoporous Mo/HZSM-5. Characterization of fresh mesoporous
Mo/HZSM-5 showed an improved Mo-oxide spreading. However, when activating the
mesoporous catalyst in He followed by carburization larger Mo-carbide particles are formed,
which results in higher propensity to coke formation. Although mesoporosity does not
improve the catalytic performance of Mo/HZSM-5 for methane dehydroaromatization,
silylation of the Mo/HZSM-5 catalyst improves activity and selectivity to the desired benzene
product. Silylation prior to Mo introduction into bulk ZSM-5 was detrimental to the catalytic
performance.
Chapter 2
49
2.5 References
[1] British Petroleum Company. BP Statistical Review of World Energy. London: British Petroleum
Co, 2013, pp. 20.
[2] M. Golombok, W. Teunissen, Ind. Eng. Chem. Res. 42 (2003) 5003–5006.
[3] S. Thomas, R.A. Dawe, Energy 28 (2003) 1461–1477.
[4] L. Wang, L. Tao, M. Xie, G. Xu, Catal. Lett. 21 (1993) 35–41.
[5] F. Solymosi, A. Erdohelyi, A. Szöke, Catal. Lett. 32 (1995) 43–53.
[6] F. Solymosi, J. Cserényi, A. Szöke, T. Bánsági, A. Oszkó, J. Catal. 165 (1997) 150–161.
[7] L. Chen, L. Lin, Z.S. Xu, X.S. Li, T. Zhang, J. Catal. 157 (1995) 190–200.
[8] S. Liu, L. Wang, R. Ohnishi, M. Ichikawa, Kinet. Catal. 41 (2000) 132–144.
[9] B.M. Weckhuysen, M.P. Rosynek, J.H. Lunford, Catal. Lett. 52 (1998) 31–36.
[10] D. Ma, D. Wang, L. Su, Y. Shu, Y. Xu, X. Bao, J. Catal. 208 (2002) 260–269.
[11] C. Descorme, P. Gelin, C. Lecuyer, A. Primet, Appl. Catal. B 13 (1997) 185–195.
[12] M. Wojdyr, J. Appl. Crystallogr. 43 (2010) 1126–1128.
[13] L. Su, L. Liu, J. Zhuang, H. Wang, Y. Li, W. Shen, Y. Xu, X. Bao, Catal. Lett. 91 (2003) 155–
167.
[14] A. Martinez, E. Peris, M. Derewinski, A. Burkat-Dulak, Catal. Today 169 (2011) 75–84.
[15] N.B. Chu, J.Q. Wang, Y. Zhang, J.H. Yang, J.M. Lu, D.H. Yin, Chem. Mater. 22 (2010) 2757–
2763.
[16] T. Behrsing, H. Jaeger, J.V. Sanders, Appl. Catal. 54 (1989) 289–302.
[17] W.P. Ding, G.D. Meitzner, E. Iglesia, J. Catal. 206 (2002) 14–22.
[18] H.M. Liu, Y. Li, W.J. Shen, X.H. Bao, Y.D. Xu, Catal. Today 93–95 (2004) 65–73.
[19] S. Kikuchi, R. Kojima, H.T. Ma, J. Bai, M. Ichikawa, J. Catal. 242 (2006) 349–356.
[20] J.C. Groen, W. Zhu, S. Brouwer, S.J. Huynink, F. Kapteijn, J.A. Moulijn, J. Pérez-Ramírez, J.
Am. Chem. Soc. 129 (2007) 355–360.
[21] S. Zheng, H.R. Heydenrych, A. Jentys, J.A. Lercher, J. Phys. Chem. B 106 (2002) 9552–9558.
[22] J.A. Jacobs, E.G. Derouane, J. Weitkamp, Chem. Commun. 194 (1981) 591–593.
[23] B.C. Lippens, J.H. de Boer, J. Catal. 4 (1965) 319–323.
[24] J. Datka, A.M. Tutek, J.H. Jehng, I.E. Wachs, J. Catal. 135 (1992) 186–199.
[25] N.S. Nesterenko, F. Thibault-Starzyk, V. Montouillout, V.V. Yuschenko, C. Fernandez, J.-P.
Gilson, F. Fajula, I.I. Ivanova, Kinet. Catal. 47 (2006) 40–48.
[26] J.-P. Tessonnier, B. Louis, S. Rigolet, M.J. Ledoux, C. Pham-Huu, Appl. Catal. A 336 (2008)
79–88.
[27] W. Zhang, D. Ma, X. Han, X. Liu, X. Bao, X. Guo, X. Wang, J. Catal. 188 (1999) 393–402.
[28] W. Liu, Y. Xu, S.-T. Wong, L. Wang, J. Qiu, N. Yang, J. Mol. Catal. A Chem. 120 (1997) 257–
265.
[29] J.-Z. Zhang, M.A. Long, R.F. Howe, Catal. Today 44 (1998) 293–300.
[30] J.Z. Hu, J.H. Kwak, Y. Wang, C.H.F. Peden, H. Zheng, D. Ma, X. Bao, J. Phys. Chem. C 113
(2009) 2936–942.
[31] J.-B. d’Espinose de Lacaillerie, Z. Gan, Appl. Magn. Reson. 32 (2007) 499–511.
[32] K. Amakawa, Active site for propene metathesis in silica-supported molybdenum oxide
catalysts, (Ph.D. thesis), Berlin, 2013.
[33] J. Leyrer, R. Margraf, E. Taglauer, H. Knözinger, Surf. Sci. 201 (1988) 603–623.
[34] J. Leyrer, D. Mey, H. Knözinger, J. Catal. 124 (1990) 349–356.
[35] M.H.F. Kox, E. Stavitski, J.C. Groen, J. Pérez-Ramírez, F. Kapteijn, B.M. Weckhuysen, Chem.
Eur. J. 14 (2008) 1718–1725.
[36] R.W. Borry III, Y.H. Kim, A. Huffschmith, J.A. Reimer, E. Iglesia, J. Phys. Chem. B 103
(1999) 5787–5796.
[37] I.E. Wachs, Catal. Today 27 (1996) 437–455.
[38] N.W. Ngobeni, A.F. Carley, M.S. Scurell, C.P. Nicolaides, J. Mol. Catal. A Chem. 305 (2009)
40–46.
Chapter 2
50
[39] C.C. Williams, J.G. Ekerdt, J.M. Jehng, F.D. Hardcastle, A.M. Turek, I.E. Wachs, J. Phys.
Chem. 95 (1991) 8781–8791.
[40] E.L. Lee, I.E. Wachs, J. Phys. Chem. C 111 (2007) 14410–14425.
[41] M. Fournier, C. Louis, M. Che, P. Chaquin, D. Masure, J. Catal. 119 (1989) 400–414.
[42] H. Jezlorowski, H. Knözinger, J. Phys. Chem. 83 (1979) 1166–1173.
[43] G. Xiong, Z. Feng, J. Li, Q. Yang, P. Ying, Q. Xin, J. Phys. Chem. B 104 (2000) 3581–3588.
[44] F. Fan, K. Sun, Z. Feng, H. Xia, B. Han, Y. Lian, P. Ying, C. Li, Chem. Eur. J. 15 (2009) 3268–
3276.
[45] P.K. Dutta, M. Puri, J. Phys. Chem. 91 (1987) 4329–4333.
[46] Y. Lou, H. Wang, Q. Zhang, Y. Wang, J. Catal. 247 (2007) 245–255.
[47] H. Hu, I.E. Wachs, S.R. Bare, J. Phys. Chem. 99 (1995) 10897–10910.
[48] E.L. Lee, I.E. Wachs, J. Phys. Chem. C 112 (2008) 20418–20428.
[49] G. Xiong, Z. Feng, J. Li, Q. Yang, P. Ying, Q. Xin, C. Li, J. Phys. B 104 (2000) 3581–3588.
[50] G. Mestl, T.K.K. Srinivasan, Catal. Rev. Sci. Eng. 40 (1998) 451–570.
[51] H. Jezlorowski, H. Knözinger, J. Phys. Chem. 9 (1979) 1166–1173.
[52] L. Mosqueira, G.A. Fuentes, Mol. Phys. 100 (2002) 3055–3057.
[53] A.M. Rzehvskii, P. Choi, F.H. Ribeiro, R.J. Gulotty Jr., M.M. Olken, Catal. Lett. 73 (2001)
187–191.
[54] J.A. Mendoza-Nieto, O. Vera-Vallejo, L. Escobar-Alarcon, D. Solis-Casados, T. Klimova, Fuel
110 (2013) 268–277.
[55] M. de Boer, A.J. van Dillen, D.C. Koningsberger, J.W. Geus, M.A. Vuurman, I.E. Wachs,
Catal. Lett. 11 (1991) 227–239.
[56] G. Mestl, T.K.K. Srinivasan, Catal. Rev. Sci. Eng. 40 (1998) 451–570.
[57] E. Haro-Poniatowski, C. Julien, B. Pecquenard, J. Livage, M.A. Camacho-López, J. Mater. Res.
13 (1998) 1033–1037.
[58] B.S. Liu, L. Jiang, H. Sun, C.T. Au, Appl. Surf. Sci. 253 (2007) 5092–5100.
[59] Z.R. Ismagilov, E.V. Matus, L.T. Tsikoza, Energy Environ. Sci. 1 (2008) 526–541.
[60] W. Li, G.D. Meitzner, R.W. Borry III, E. Iglesia, J. Catal. 191 (2000) 373–383.
[61] F. Solymosi, A. Szöke, J. Cserényi, Catal. Lett. 39 (1996) 157–161.
[62] D.J. Wang, J.H. Lunsford, M.P. Rosynek, Top. Catal. 3 (1996) 289–293.
[63] D.J. Wang, J.H. Lunsford, M.P. Rosynek, J. Catal. 169 (1997) 347–358.
Chapter 3
51
Activation of Mo/HZSM-5 for methane aromatization
Summary
The effect of pretreatment of Mo/HZSM-5 at 973 K in inert (He), oxidizing (artificial air) and
carburizing (a CH4/He mixture) atmosphere on its performance in the non-oxidative methane
dehydroaromatization (MDA) was investigated. The influence of post-synthesis silylation to
deactivate external acid sites was also studied. Precarburization resulted in increased
aromatics selectivity and improved catalyst stability. The benzene selectivity was highest for
the silylated Mo/HZSM-5 catalyst (benzene + naphthalene selectivity after 1 h on stream was
close to 100%). Deactivation of the precarburized zeolites was less pronounced than of the
zeolites heated in air or He. During heating in air or He, a larger fraction of the Mo-oxo
species diffused into the micropores than during heating in methane. Carburization of the
Mo-oxide species in the micropores during the MDA reaction resulted in Mo-carbide
particles that contribute to pore blocking, making the Brønsted acid sites inaccessible. The
formation of Mo-carbides during heating in methane resulted in a less mobile Mo phase. It is
argued that the presence of Mo-carbide particles in the micropores contributes to rapid
catalyst deactivation in addition to the formation of hard coke on the external surface.
This work is published in Chin. J. Catal. 36 (2015) 829-837.
Chapter 3
52
3.1 Introduction
Because petroleum oil reserves are dwindling, the identification of alternative feedstocks
for fuels and chemical production is necessary. Natural gas is increasingly considered as a
feedstock for energy, fuels, and chemicals, because it is abundant and the cleanest of all fossil
resources. In the 2013 BP Statistical Review of World Energy, it was reported that proven
natural gas reserves amount to approximately 180 trillion m3 [1]. Large amounts of these
natural gas reserves are located in remote areas, which makes their valorization costly. In
addition to direct liquefaction, there are several options for the large‐scale conversion of
natural gas to liquefied fuels via the syngas platform (a mixture of H2 and CO). These include
methanol synthesis, dimethyl ether synthesis, and Fischer‐Tropsch synthesis [2–4]. Facilities
that produce syngas are capital intensive; only large‐scale syngas production is cost effective
[5]. A long‐term goal of the chemical industry is therefore to develop alternative routes for
directly converting methane to easily transportable liquid intermediates that can serve as
platforms for fuels and chemicals. The direct oxidation of methane to methanol is a significant
scientific challenge. Non‐oxidative methane dehydroaromatization (MDA), which converts
methane to aromatics, mainly benzene, and hydrogen, is more promising than oxidative
approaches.
MDA was first described by Wang et al. [6], and has since been widely investigated by
industry and academia. Catalyst screening has shown that Mo/HZSM‐5 is the preferred
catalyst for MDA [7–9]. The molybdenum oxide phase is carburized by CH4 to a
molybdenum carbide (MoCx) phase. Although the exact nature of this carbide phase has not
been determined, its function is to convert methane to ethylene [10,11], which is then
oligomerized and cyclized to aromatic compounds at the Brønsted acid sites (BAS) of the
zeolite. The harsh process conditions, with typical temperatures of 973 K and above,
necessary for the activation of methane [12] result in poor catalyst stability; industrial
applications of this reaction are therefore challenging. The poor catalyst stability is mainly
caused by the formation of carbonaceous deposits, which block the zeolite micropores [13].
The BAS located at the external surface are considered to be involved in the formation of
large amounts of carbonaceous deposits [14]. These BAS catalyze the formation of polycyclic
aromatic hydrocarbons [15], which are deposited on the external surface and eventually block
the entrances to the micropores [16–24]. The results of our comprehensive study of catalyst
deactivation in MDA suggested that the formation of a polycyclic carbon layer around the
Chapter 3
53
zeolite crystal with progressive time on stream was the main reason [25]. It has also been
reported that deactivation of the external surface BAS suppresses the formation of coke
during MDA [21–24]. The pretreatment procedure used to activate the Mo/HZSM‐5 catalyst
strongly influences the catalytic performance.
The pretreatment gas directly affects the state of the Mo phase and indirectly affects the
state of the zeolite at the start of the reaction [26–28]. Although a significant number of
studies include data on catalyst activation [7, 26–34], comprehensive studies are lacking;
comparisons among available studies is greatly hampered by the often very different reaction
conditions and differences among catalyst samples. Two studies [28, 33] have directly
compared the influence of the pretreatment procedure, but they were conducted at lower
weight hourly space velocity (WHSV) values than those typically used in MDA. In the
present study, we systematically compared the influence of the gas (inert He, oxidizing
artificial air, and reducing/carburizing CH4) used for catalyst activation on the catalytic MDA
performance. As we showed previously, catalyst silylation significantly affects the MDA
performance [24], therefore we determined the influence of catalyst pretreatment on fresh and
silylated samples. We showed that precarburization resulted in improved catalytic
performance (improved catalyst stability and increased benzene selectivity) for silylated and
non‐silylated Mo/ZSM‐5. We determined the detailed physicochemical properties of the
fresh, carburized, and spent catalysts.
3.2 Experimental
3.2.1 Synthesis
The parent NH4ZSM‐5 was obtained from Akzo‐Nobel (now Albemarle Catalysts). The
parent zeolite had a Si/Al ratio of 19.4, determined by inductively coupled plasma optical
emission spectroscopy (ICP‐OES). Mo was introduced by incipient wetness impregnation
with an aqueous solution of ammonium heptamolybdate tetrahydrate (Merck) of appropriate
concentration. Prior to impregnation, the zeolite was dried overnightat 383 K. After
impregnation, the samples were dried for 1 h. The target Mo content was 4 wt%. The Mo‐
containing zeolites were calcined in artificial air at 823 K for 5 h after heating at a rate of 1.5
K/min. The parent ZSM‐5 zeolite is denoted by HZSM‐5. The zeolite after Mo introduction is
denoted byMo/HZSM‐5.
A method based on that reported by Zheng et al. [35] was used to silylate the external
surfaces of the zeolites. Typically, Mo/HZSM‐5 (2 g) was dried overnight at 373 K and
Chapter 3
54
dispersed in n‐hexane (50 mL). Tetraethyl orthosilicate (TEOS, 0.3 mL; Merck) was added
and the suspension was stirred for 1 h under reflux. The amount of TEOS corresponded to 0.4
wt% of the zeolite. After silylation, the zeolite was filtered, dried overnight at 373 K, and
calcined in artificial air. The catalyst was first heated at a rate of 2 K/min to 393 K, followed
by an isothermal period of 2 h. The temperature was then further increased to 773 K at a rate
of 0.2 K/min. The catalyst was kept at this temperature for 4 h. The silylated Mo/HZSM‐5
zeolite is denoted by Mo/HZSM‐5(Si).
Prior to the MDA reaction, the catalyst was heated to 973 K at a rate of 10 K/min in He,
artificial air, or an 80%/20% (v/v) CH4/He flow. In addition to activity testing, another set of
samples was prepared by rapidly cooling the catalyst after pretreatment in a He flow. The
catalyst was transported under exclusion of air into a N2‐filled glove‐box. The activation
procedure is indicated by adding the gas type as a suffix to the sample name.
3.2.2 Characterization
The Mo and Al contents of the samples were determined by ICP‐OES, using a Spectro
CIROS CCD spectrometer equipped with a free‐running 27.12 MHz generator running at
1400 W. Prior to analysis, the samples were dissolved in a mixture of HF/HNO3/H2O (1:1:1).
Ar sorption isotherms were measured at 87 K, using a Micromeritics ASAP2020 system
in static measurement mode. The samples were outgassed at 623 K for 8 h prior to the
sorption measurements. BET equation was used to calculate the specific surface area (SBET)
from the adsorption data obtained (p/p0 = 0.05–0.25). The mesopore volume (Vmeso) and the
mesopore size distribution were determined, using the BJH method, from the adsorption
branch of the isotherm. The micropore area (Smicro) and the micropore volume (Vmicro) were
determined using the t‐plot method, at a thickness range between 3.5 and 5.4 Å [36].
Infrared (IR) spectra were recorded in the 4000–400 cm−1
range using a Bruker Vertex
70V spectrometer. Samples were pressed into a self‐supporting wafer of typical density 8
mg/cm2. Adsorbed water was removed by evacuating the sample at 773 K for 2 h. After
evacuation, the sample was cooled to 323 K, and then the background spectrum was recorded.
The total concentration of BAS was determined by IR spectroscopy of adsorbed pyridine. The
dehydrated zeolite wafer was cooled to 323 K and exposed to pyridine until the bands related
to pyridine were saturated. The sample was then evacuated at 423 K for 2 h and the spectrum
Chapter 3
55
was recorded. The spectra were deconvoluted by standard procedures, and the extinction
coefficient values reported by Datka et al. [37] were used for quantification.
Ultraviolet (UV) Raman spectra were recorded using a Jobin‐Yvon T64000 triple‐stage
instrument with a spectral resolution of 2 cm−1
. The excitation laser line at 325 nm was
produced by a Kimmon He‐Cd laser. The power of the laser on the sample was 4 mW.
Magic angle spinning (MAS) 27
Al single‐pulse nuclear magnetic resonance (NMR)
spectroscopy was performed using a Bruker Avance DMX‐500 NMR spectrometer equipped
with a 2.5 mm MAS probe head operated at an 27
Al NMR resonance frequency of 130.3 MHz.
The 27
Al chemical shift was referenced to saturated Al(NO3)3 solution. In a typical
experiment, a well‐hydrated sample (10 mg) was packed in a 2.5‐mm zirconia rotor. The
MAS sample rotation speed was 20 kHz. Single‐pulse excitation was used, with a 18° pulse of
1 μs and an interscan delay of 1 s. The relaxation time was 1 s and the pulse length was 1 s.
Transmission electron microscopy (TEM; FEI Tecnai 20) was performed at an electron‐
accelerating voltage of 200 kV. Typically, a small amount of sample was suspended in
ethanol, sonicated, and dispersed over a Cu grid with a holey carbon film.
Weight‐loss curves were determined for the spent catalysts after 12 h on stream in MDA.
These measurements were carried out using a Mettler Toledo TGA/DSC 1 instrument.
Samples were heated to 1023 K in uncovered alumina crucibles at a rate of 5 K/min in a 33/67
(v/v) O2/He flow. Catalytic H/D exchange between C6H6 and C6D6 was performed using a
ten‐channel parallel microflow reactor. Typically, the zeolite material (50 mg) was loaded
into a quartz tubular reactor with an internal diameter of 4.0 mm. The zeolites were first
pelletized and sieved to 125–250 μm mesh. The ten quartz tubes were then placed in the
paallel setup. Before reaction, the samples were heated to 723 K at a rate of 5 K/min,
followed by an isothermal period of 6 h, to remove water. The reactor was then cooled to 303
K. The reaction was started by switching the reactor feed to a gas flow containing a 90/10
(v/v) C6H6/C6D6 mixture. The effluent products were analyzed using mass spectrometry. The
rate of H/D exchange between benzene and perdeuterobenzene was determined as a function
of temperature.
Chapter 3
56
3.2.3 Catalytic activity measurements
The zeolite catalyst (0.5 g) was placed in a tubular quartz reactor of length 490 mm and
internal diameter 4.0 mm. The catalyst was supported on quartz wool in the isothermal zone
of the oven. All gases were fed using thermal mass flow controllers. The temperature was
increased at a rate of 5 K/min to 973 K in a 25 mL/min flow of He, CH4/He, or artificial air.
The reaction was started by switching the reactor feed to a 5/95 (v/v) N2/CH4 mixture (N2
was used as the internal standard) at a WHSV of of 1.22 h-1
(1710 ml CH4/gcat.h at STP
conditions). The products were analyzed using an online gas chromatograph (Interscience
Compact GC) equipped with three channels for separate analyses of light gases [Molsieve 5A,
thermal conductivity detector (TCD)], light hydrocarbons [Al2O3/KCl, flame ionization
detector (FID)], and aromatics (Rtx‐1, TCD). N2 was used as an internal standard to close the
carbon balance and account for the amount of carbon deposits formed during the reaction. The
carbon formation rate rcoke was determined based on the components ethylene, ethane,
propylene, propane, benzene, toluene, and naphthalene (eq. 3.2). The methane conversion
(XCH4) and benzene selectivity (SC6H6) were calculated using eqs. (3.1) and (3.3).
blanksaromatics
blanksslightgasse
TCD
CH
TCD
CH
TCD
N
TCD
N
TCD
product
TCD
product
In
Naromatic
A
A
f
A
f
A
r
4
4
2
2
2
(3.1)
in
iiCHCoke MWrrr4
(3.2)
100% 6666
66
in
products
HCHC
HC
MWr
MWrwtS (3.3)
Chapter 3
57
Symbols
4CHr = reaction rate of CH4
ir = formation rate of compound i
In
i = molar flow rate of compound i into the reactor Out
i = molar flow rate of compound i out of the reactor TCD
if = TCD sensitivity factor of compound i FID
if = FID sensitivity factor of compound i
iA = measured peak area of compound i in the chromatogram
MWi = molecular weight compound i
3.3 Results and discussion
3.3.1 Catalytic activity measurements
Figures 3.1 and 3.2 show the time‐on‐stream reaction data for Mo/HZSM‐5 and
Mo/HZSM‐5(Si) pretreated with three different gases (CH4, He, and O2) at 973 K. In all
cases, the methane conversion rate decreased with time on stream. The rates of deactivation of
the two zeolite samples activated in CH4 were lower than those of the samples heated in air
and He. The reaction data show that initially few hydrocarbons were present in the effluent
stream; this observation is in line with previous reports [29, 38–40]. Methane carburizes the
molybdenum oxide phase during the initial stages of the reaction. As low hydrocarbon
selectivity is also seen for the precarburized sample, we can conclude that carburization of the
molybdenum oxide precursor during catalyst activation in methane is not complete. After this
initial phase, the hydrocarbon selectivity increased. The main hydrocarbon product was
benzene, and naphthalene was the main aromatic side product. The highest benzene
selectivity was observed after 1 h on stream; the highest benzene selectivity was 65 wt% for
the Mo/HZSM‐5 samples and 80 wt% for the Mo/HZSM‐5(Si) samples. In line with
previously reported data [21–24], the maximum aromatic selectivities (benzene + naphthalene
selectivity) were higher for the silylated catalysts. It can also be seen that the combined
selectivities for benzene and naphthalene were close to 100% for Mo/HZSM‐5(Si,CH4/He).
After 1 h on stream, the benzene selectivities of the catalysts pretreated in air and He
decreased rapidly. In comparison, the decrease in the benzene selectivity was less pronounced
for the precarburized Mo/HZSM‐5 and Mo/HZSM‐5(Si) catalysts. For the precarburized
catalysts, the benzene selectivities started to decrease more pronouncedly only after 4 h on
stream. The total amounts of aromatics produced were the highest for the precarburized
Chapter 3
58
catalysts. In all cases, the decrease in aromatic selectivity was accompanied by a substantial
increase in ethylene selectivity, and a smaller increase in naphthalene selectivity. This is
probably the result of deactivation of the acid sites in the micropores; this leads to higher
selectivity for the intermediate product ethylene, which is formed on the molybdenum carbide
phase from methane. The high naphthalene selectivity is related to the conversion of ethylene
on BAS on the external surface.
Chapter 3
59
Fig. 3.1. MDA reaction data: (a) methane reaction rate, (b) benzene (closed symbols) and naphthalene (open symbols) selectivities and (c)
ethylene selectivity for (■) Mo/HZSM-5(air), (●) Mo/HZSM-5(He) and (▲) Mo/HZSM-5(CH4/He).
0 2 4 6 8 10
0
2
4
6
8
10
12
14
CH
4 r
eacti
on
rate
(m
mo
l/h
.gcat)
Time on stream (h)
0 2 4 6 8 10
0
10
20
30
40
50
60
70
80
90
100
Sele
cti
vit
y (
wt%
)
Time on stream (h)
0 2 4 6 8 10
0
10
20
30
40
50
60
Eth
yle
ne s
ele
cti
vit
y (
wt%
)
Time on stream (h)
Chapter 3
60
Fig. 3.2. MDA reaction data: (a) methane reaction rate, (b) benzene (closed symbols) and naphthalene (open symbols) selectivities and (c)
ethylene selectivity for (■) Mo/HZSM-5(Si,air), (●) Mo/HZSM-5(Si,He) and (▲) Mo/HZSM-5(Si,CH4/He).
0 2 4 6 8 10
0
2
4
6
8
10
12
14
CH
4 r
eacti
on
rate
(m
mo
l/h
.gcat)
Time on stream (h)
0 2 4 6 8 10
0
10
20
30
40
50
60
70
80
90
100
Sele
cti
vit
y (
wt%
)
Time on stream (h)
0 2 4 6 8 10
0
10
20
30
40
50
60
Eth
yle
ne s
ele
ctv
ity (
wt%
)
Time on stream (h)
Chapter 3
61
3.3.2 Catalyst characterization
The textural properties of the fresh, pretreated, and spent catalysts were determined using
Ar physisorption measurements. The results are shown in Table 3.1. The introduction of Mo
into the parent zeolite led to a decrease in the micropore volume (Vmicro). This is probably
because of diffusion of part of the molybdenum oxide phase into the HZSM‐5 micropores
during calcination. The increase in the mesopore volume (Vmeso) is presumably the result of
the damage caused by Al extraction from the zeolite framework [24,41]. The micropore
volume of the silylated Mo/HZSM‐5(Si) catalyst was smaller than that of the parent
Mo/HZSM‐5 zeolite. It should be stressed that silylation was performed after the Mo‐loading
step, which is more beneficial for the MDA reaction than the usual approach in which HZSM‐
5 is silylated before Mo introduction [24]. We previously showed that silylation and
calcination of Mo/HZMS‐5 result in higher molybdenum oxide dispersion over the external
surface and an increase in the fraction of Mo entering the micropores [24]. Carburization in
methane led to small decreases in the pore volumes of Mo/HZSM‐5 and Mo/HZSM‐5(Si).
The carburization step led to the formation of different types of carbonaceous species in the
catalysts. These species were investigated using thermogravimetric analysis (TGA). The TG
curves of the carburized and spent Mo/HZSM‐5 and Mo/HZSM‐5(Si) catalysts are shown in
Fig. 3.3. The curves display three main oxidation peaks. The main peak at 700 K is
characteristic of oxidation of molybdenum carbide (CMoC) [42]. The second peak, at a higher
temperature (~750 K), is related to coke species formed in the proximity of molybdenum
carbide [42]. This is usually amorphous polyolefinic coke and is referred to as soft coke
(Csoft). The small peak at 840 K is related to polycyclic aromatics, referred to here as hard
coke (Chard) [19,40]. The quantitative data obtained by deconvolution of these traces are
collected in Table 1. The TG curves for the precarburized Mo/HZSM‐5 and Mo/HZSM‐5(Si)
catalysts contain all three types of coke. Carburization of Mo/HZSM‐5(Si) led to a larger
amount of carbon species associated with the Mo phase and soft coke compared with
carburization of Mo/HZSM‐5. This difference is in line with higher molybdenum oxide
dispersion in the silylated catalyst. The decrease in the micropore volume was slightly higher
for the silylated catalyst.
Chapter 3
62
Table 3.1. Textural properties of the fresh and spent Mo/HZSM-5 and Mo/HZSM-5(Si) and the chemical nature of the carbon deposits.
Sample Activation TOS1
(h)
Vmicro
(cm3/g)
Vmeso
(cm3/g)
Smicro
(m2/g)
Smeso
(m2/g)
SBET
(m2/g)
Ctotal
(wt%)
CMoC
(wt%)
Csoft
(wt%)
Chard
(wt%)
HZSM-5 - 0 0.12 0.02 246 10 277 - - - -
Mo/HZSM-5 - 0 0.10 0.06 182 44 272 - - - -
CH4/He 0 0.09 0.03 171 22 228 3.3 0.9 0.9 1.5
CH4/He 12 0 0.02 3 12 12 13.5 0 2.7 10.8
Air 12 0.02 0.02 32 14 59 11.7 0 3.4 8.3
He 12 0.02 0.02 27 14 61 11.1 0 3.6 7.5
Mo/HZSM-5(Si)
- 0 0.07 0.01 129 6 143 - - - -
CH4/He 0 0.05 0.02 97 12 130 4.5 1.2 1.8 1.5
CH4/He 12 0.03 0.02 64 13 90 9.5 0.5 2.4 6.6
Air 12 0.02 0.02 46 14 72 7.7 0 3.7 4.0
He 12 0.04 0.03 76 15 109 9.7 0 4.1 5.6 1 Time On Stream (TOS) in MDA.
Chapter 3
63
Fig. 3.3. TGA weight-loss curves after 12 hours of MDA reaction of non-silylated (left) and
silylated Mo/HZSM-5 (right) pretreated in (a) He, (b) air or (c) a CH4/He mixture. Spectrum
(d) corresponds to the catalyst pretreated in CH4/He which was subsequently recovered from
the reactor after activation (5x magnified).
During MDA, the micropore volumes of the zeolite catalysts decreased substantially. The
decreases in the micropore volumes of the non‐silylated catalysts were greater than those of
the silylated ones. This difference is consistent with the smaller amount of carbonaceous
species formed in the silylated samples during the reaction. TGA of the spent samples shows
that more hard coke was formed on the non‐silylated catalysts. This hard coke is usually
associated with the formation of polycyclic aromatics on the external surface of the zeolite
[14, 15]. Silylation effectively reduces the external acidity [14, 15] and, accordingly, the
formation of hard coke [24]. The lower residual micropore volume of the spent non‐silylated
samples can therefore be correlated with high coverage of the external surface by polycyclic
aromatics, which block the micropore entrances. Despite the lower rate of deactivation in
MDA, the TG data show that the precarburized spent samples always contained more coke,
especially in the form of hard coke, than did the samples activated in He or air. The carbon
speciations for the spent samples pretreated in He and air were similar.
27Al MAS NMR spectroscopy was used to determine the Al speciation in the samples.
The NMR spectra are shown in Fig. 3.4. The dominant feature, at 55 ppm, is related to
framework Al (FAl). The small peak at 0 ppm is attributed to extraframework Al (EFAl). The
600 700 800 900 1000
Weig
ht-
loss (
mg
/K)
Temperature (K)
600 700 800 900 1000
Temperature (K)
Chapter 3
64
feature at 14 ppm also arises from EFAl, in the form of Al2(MoO4)3. The increased EFAl
concentration and the formation of aluminum molybdates observed for the Mo‐modified
ZSM‐5 zeolites suggest dealumination of the framework. The FAl peaks for the Mo‐modified
zeolites are broader than that for the parent HZSM‐5 zeolite. Tessonier et al. [42] attributed
this broadening to a change in the symmetry around the Al tetrahedra following exchange of
protons with molybdenum oxide species. The resonance area of the FAl peak was lower for
the Mo‐modified samples, probably because of asymmetric Al coordination when Mo is
present at the exchange sites. On carburization, the intensity of the FAl peak decreased, which
indicates further dealumination. This is probably related to the high temperatures of the
carburization step. The decrease in the intensity of the FAl peak was stronger when
Mo/HZSM‐5 or Mo/HZSM‐5(Si) was treated in He or air. This suggests a higher degree of
proton exchange with molybdenum oxide species at the BAS than in activation by
precarburization. A portion of the EFAl is present in the form of Al2(MoO4)3, indicated by the
small peak at 14 ppm. The NMR data suggest that more FAl species are retained after
precarburization than after pretreatment in He or air.
Fig. 3.4. 27
Al MAS NMR spectra of non-silylated (left) and silylated Mo/HZSM-5 (right)
showing (a) parent ZSM-5 (b) fresh Mo/HZSM-5 and Mo/HZSM-5 pretreated in (c) He, (d)
air and (e) CH4/He mixture.
Before discussing the acidities of the most active precarburized samples in detail, we
describe the Fourier‐transform (FT)IR spectra of the activated samples. Because carbon
100 50 0 -50
Chemical shift (ppm)
100 50 0 -50
Chemical shift (ppm)
Chapter 3
65
formation resulted in blackening of the sample, FTIR experiments were only performed on
the zeolites after activation in air or He at 973 K. The hydroxyl stretching region of the
spectra contain bands at 3745, 3665, and 3610 cm−1
; these are attributed to silanols,
extraframework hydroxyls, and bridging hydroxyl, respectively. As expected, modification of
the parent zeolite with Mo led to a decrease in the intensity of the bridging hydroxyl band.
This indicates a lower BAS density as a result of proton exchange with molybdenum oxide
species. The silylated Mo/HZSM‐5 contained fewer acid sites. Pretreatment of the parent and
silylated Mo/HZSM‐5 in He or air led to the nearly complete disappearance of all hydroxyl
features, including the BAS. The acid site contents of these zeolites were also investigated
using pyridine adsorption measurements (Table 3.2). Modification of the parent zeolite with
Mo followed by silylation led to a decrease in the BAS content. After activation in He or air,
the BAS content decreased further. The BAS densities for these two activated zeolites are
higher than those suggested by theintensity of the bridging hydroxyl band (Fig. 3.5). This
discrepancy can be explained by the close proximity of molybdenum oxide species to the
bridging hydroxyl groups, which leads to perturbation of these hydroxyls.
Fig. 3.5. FTIR spectra of non-silylated (left) and silylated Mo/HZSM-5 (right) showing (a)
Mo/HZSM-5 pretreated in He, (b) Mo/HZSM-5 pretreated in air, (c) fresh Mo/HZSM-5 and
(d) parent ZSM-5.
3800 3600 3400 3200
Ab
so
rban
ce (
a.u
.)
Wavenumber (cm-1)
3800 3600 3400 3200
Wavenumber (cm-1)
Chapter 3
66
Table 3.2. Acidity of fresh and air- and He-activated Mo/HZSM-5 and Mo/HZSM-5(Si) as
measured by pyridine IR measurements. The air and He pretreated catalysts were activated at
973 K.
Sample Pretreatment
gas used
BAS
(mmol/gcat)
LAS
(mmol/gcat)
HZSM-5 - 0.679 0.284
Mo/HZSM-5 -
He
Air
0.615
0.229
0.361
0.136
0.055
0.197
Mo/HZSM-5(Si) -
He
Air
0.530
0.278
0.239
0.060
0.148
0.145
The samples obtained by precarburization were investigated in more detail, because they
performed substantially better than did the catalysts activated in air or He. FTIR spectroscopy
of adsorbed pyridine was unsuccessful, because of the presence of carbonaceous deposits,
therefore the Brӧnsted acidity was investigated by H/D exchange between benzene and
deuterated benzene. Haw and coworkers characterized the acid sites in faujasite zeolites using
1H NMR spectroscopy to track their H/D exchange with perdeuterobenzene [43]. Poduval et
al. [44] used the IR spectra of H/D exchanged faujasite zeolites to determine the strengths and
numbers of strong acid sites. Here, we determined the rate of H/D exchange between benzene
and d6‐benzene in a fixed bed reactor. A benzene to d6‐benzene ratio of 10:1 was used in the
feed, and the rate of the reaction was determined by measuring the conversion of benzene to
d1‐benzene. The obtained reaction rates are listed in Table 3.3. The exchange rates of the Mo‐
modified zeolites were slightly lower than that of the parent HZSM‐5 zeolite. This is because
of exchange of some of the protons at exchange positions with molybdenum oxide species.
Carburization further decreased the reaction rate. The decrease in the acidity was the strongest
for the silylated sample. After activation, only 28% of the original acidity of the HZSM‐5 was
retained.
The external surfaces of the zeolite crystals were examined using X‐ray photoelectron
spectroscopy (XPS). The surface Si/Al and Si/Mo ratios for the various catalysts are listed in
Table 3.3. The Si/Al ratios of the Mo‐modified zeolites before and after precarburization were
similar to the values for the parent HZSM‐5. The Si/Mo ratios increased on precarburization
of Mo/HZSM‐5 and Mo/HZSM‐5(Si). This increase suggests agglomeration of the Mo phase
during formation of the carbide phase. The agglomeration extents were similar for the two
Chapter 3
67
samples. Elemental analysis also shows that there was only a small loss of Mo during the
activation step in methane.
Table 3.3. Characterization of the elemental composition and benzene H/D exchange reaction
rates of Mo/HZSM-5 and Mo/HZSM-5(Si) before and after precarburization.
Sample Pretreatment
gas used
Al1
(wt%)
Mo1
(wt%)
Si/Al2
(XPS)
Si/Mo2
(XPS)
rC6H63
(mmol/min)
Relative
H/D
exchange
rate4
HZSM-5 - 2.2 ∞ 22 ∞ 1.65 1.00
Mo/HZSM-5 - 1.9 3.4 27 3.5 1.30 0.79
CH4/He 1.9 3.2 24 8.1 0.79 0.48
Mo/HZSM-5(Si) - 1.9 3.4 28 3.2 1.35 0.82
CH4/He 1.8 3.2 28 8.1 0.47 0.28 1Determined by ICP-OES;
2Determined by XPS experiments;
3Determined by catalytic H/D exchange
reaction; 4Remaining C6H6 reaction rate relative to that of parent HZSM-5.
The nature of the Mo phase was further investigated using UV Raman spectroscopy.
The sample was excited with a 325 nm laser. The spectra are shown in Fig. 3.6. The spectrum
of Mo/HZSM‐5 has a broad absorption band extending over the 600–1000 cm−1
region, with
more clearly defined bands at 860 and 960 cm−1
. This spectrum resembles that of amorphous
molybdenum oxide [45]. The weak band at 820 cm−1
is attributed to microcrystalline α‐MoO3
embedded in an amorphous molybdenum oxide matrix. The 380 cm−1
band characteristic of
HZSM‐5 is not observed for Mo/HZSM‐5, because the molybdenum oxide phase covers a
substantial part of the zeolite surface. The spectrum of Mo/HZSM‐5(Si) is different from that
of Mo/HZSM‐5. The former spectrum contains bands at 280, 336, 820, and 995 cm−1
; these
show the presence of a much greater proportion of microcrystalline α‐MoO3. This spectrum
also contains some features attributable to amorphous molybdenum oxide. Bands from
molybdenum oxides are not observed after carburization of Mo/HZSM‐5, indicating that the
particle surfaces were completely converted to molybdenum carbide. The weak peaks at 336
and 995 cm−1
indicate that a small amount of α‐MoO3 is retained in Mo/HZSM‐5(Si,CH4/He).
The less extensive carburization in the silylated sample may be the result of higher dispersion
of the initial molybdenum oxide phase [24].
Chapter 3
68
Fig. 3.6. UV-Raman spectra of (a) parent HZSM-5, (b) Mo/HZSM-5, (c) Mo/HZSM-
5(CH4/He), (d) Mo/HZSM-5(Si), (e) Mo/HZSM-5(Si,CH4/He), (f) bulk MoO3 and (g) bulk
Al2(MoO4)3.
TEM images of Mo/HZSM‐5 (Fig. 3.7a) show that the molybdenum oxide phase is
present as small particles with typical diameters of 1 nm. After precarburization, these
particles become more clearly visible, because they are larger (Fig. 3.7b). This indicates that
sintering occurred during the conversion of the molybdenum oxide particles to molybdenum
carbide particles. This is in line with the XPS results. The images suggest that the particle size
distribution also broadened during carburization. The TEM images confirm that the
molybdenum oxide particles (Fig. 3.7c) are significantly smaller in silylated Mo/HZSM‐5
than in non‐silylated Mo/HZSM‐5, in line with the discussion above. Similarly, carburization
of Mo/HZSM‐5(Si) led to sintering of these small particles to large molybdenum carbide
particles (Fig. 3.7d).
Extensive characterization of the activated and spent catalysts indicated that pretreatment
in air or He led to a larger fraction of molybdenum oxide species diffusing into the
micropores than did activation in methane. This is supported by the 27
Al NMR and FTIR
spectra. The mobility of molybdenum oxide species at high temperature, probably in the
partially reduced form (MoO2) [46], explains the decreased acidity. Although inspection of
the hydroxyl region of the IR spectra indicates that after heating in air or He there are few
acid sites left, the pyridine IR results show that the BAS density for these two samples is
200 400 600 800 1000 1200
Inte
nsit
y (
a.u
.)
Wavenumber (cm-1)
Chapter 3
69
similar to that of the sample activated in methane. Accordingly, we suggest that the higher
concentration of molybdenum oxide species in the micropores after He or air pretreatment
results in faster deactivation, because the molybdenum carbide particles formed during MDA
block the micropores. When the catalyst is activated in methane, less of the molybdenum
oxide phase ends up in the micropores, because of the conversion of molybdenum oxides to
molybdenum carbides. It has been reported that molybdenum carbides are less mobile than
molybdenum oxides [47]. All these data provide a satisfactory explanation for the lower
deactivation rate and improved stability in benzene selectivity of the precarburized catalysts.
The conclusion that fewer BAS are accessible in the micropores is also supported by the
higher ethylene selectivities of the He‐ and air‐pretreated catalysts when the MDA activity
decreases. The carbon speciation determined by TGA suggests that the formation in the
micropores of soft coke from molybdenum carbides may explain the rapid deactivation of the
He‐ or air‐pretreated samples. Our explanation for the improved catalyst stability after
pretreatment in methane is different from the previous claim that molybdenum carbides
obtained by precarburization are more stable under MDA conditions than those formed during
the MDA reaction [32, 48, 49].
Fig. 3.7. TEM micrographs of (a) Mo/HZSM-5, (b) Mo/HZSM-5(CH4/He), (c) Mo/HZSM-
5(Si) and (d) Mo/HZSM-5(Si,CH4/He).
Chapter 3
70
3.4 Conclusions
The effects of pretreatment of (silylated) Mo/HZSM‐5 in various gas atmospheres (artificial
air, He, or a CH4/He mixture) at 973 K on their catalytic performances in MDA were
investigated. Precarburization in methane gave catalysts with the highest aromatic selectivities
and the lowest rates of catalyst deactivation. The benzene selectivity was the highest for the
silylated Mo/HZSM‐5 catalyst. Deactivation of the precarburized catalysts was less
pronounced in the MDA reaction than for catalysts heated in air or He. This is because a
greater amount of Mo diffuses into the zeolite micropores in the form of mobile molybdenum
oxide species during heating in air or He than in heating in methane. Carburization of the
molybdenum oxide particles present in the micropores resulted in molybdenum carbide
particles, which contributed to pore blocking, making the BAS inaccessible. The deactivation
can also be partly attributed to the formation of soft coke in the micropores, and is probably
associated with the presence of molybdenum carbides. The more rapid formation of
molybdenum carbides during heating in methane decreased the amount of mobile
molybdenum oxide species and their diffusion into the micropores.
3.5 References
[1] BP Statistical Review of World Energy (2013) bp.com/statisticalreview.
[2] A. Holmen, Catal. Today 142 (2009) 2-8.
[3] E.F. Sousa-Aguiar, L.G. Appel, C. Mota, Catal. Today 101 (2005) 3-7.
[4] C. Perego, R. Bortolo, R. Zennaro. Catal. Today 142 (2009) 9-16.
[5] J.R. Rostrup-Nielsen, Catal. Today 71 (2002) 243-247.
[6] W. Zhang, D. Ma, X. Han, X. Liu, X. Bao, X. Gao, X. Wang, J. Catal. 188 (1999) 393-402.
[7] L. Wang, L. Tao, M. Xie, G. Xu, J. Huang, Y. Xu, Catal. Lett. 21 (1993) 35-41.
[8] Y. Xu, S. Liu, X. Guo, L. Wang, M. Xie, Catal Lett. 30 (1995) 135-149.
[9] W. Liu, Y. Xu, S.T. Wong, L. Wang, J. Qiu, N. Yang, J. Mol. Catal. A: Chem. 120 (1997) 257-
265.
[10] S. Majhi, P. Mohanty, H. Wang, K.K. Pant, J. Ener. Chem. 22 (2013) 543-554.
[11] D.M. Ren, X.S. Wang, G. Li, X.J. Cheng, H.Y. Long, J. Nat. Gas Chem. 19 (2010) 646-652.
[12] S. Ma, X. Guo, L. Zhao, S. Scott, X. Bao, J. Ener. Chem. 22 (2013) 1-20.
[13] M.C. Iliuta, I. Iliuta, B.P.A. Grandjean, F. Larachi, Ind. Eng. Chem. Res. 42 (2003) 3203-3209.
[14] T. Behrsing, H. Jaeger, J.V. Sanders, Appl. Catal. 54 (1989) 289-302.
[15] L. Su, L. Liu, J. Zhuang, H. Wang, Y. Li, W. Shen, Y. Xu, X. Bao, Catal. Lett. 91 (2003) 155-
167.
[16] L. Chen, L. Lin, Z.S. Xu, X.S. Li, T. Zhang, J. Catal. 157 (1995) 190-200.
[17] S. Liu, L. Wang, R. Ohnishi, M. Ichikawa, Kinet. Catal. 41 (2000) 132-144.
[18] B.M. Weckhuysen, M.P. Rosynek, J.H. Lunford, Catal. Lett. 52 (1998) 31-36.
[19] D. Ma, D. Wang, L. Su, Y. Shu, Y. Xu, X. Bao, J. Catal. 208 (2002) 260-269.
[20] C. Descorme, P. Gelin, C. Lecuyer, A. Primet, Appl. Catal. B 13 (1997) 185-195.
[21] W.P. Ding, G.D. Meitzner, E. Iglesia, J. Catal. 206 (2002) 14-22.
[22] H.M. Liu, Y. Li, W.J. Shen, X.H. Bao, Y.D. Xu, Catal. Today 93 (2004) 65-73.
Chapter 3
71
[23] S. Kikuchi, R. Kojima, H.T. Ma, J. Bai, M. Ichikawa, J. Catal. 242 (2006) 349-356.
[24] C.H.L. Tempelman, V.O. de Rodrigues, E.R.H. van Eck, P.C.M.M. Magusin, E.J.M. Hensen,
Microporous Mesoporous Mater. 203 (2015) 259-273.
[25] Z.R. Ismagilov, E.V. Matus, L.T. Tsikoza, Energ. Environ. Sci. 1 (2008) 526-541.
[26] H.S. Lacheen, E. Iglesia, Phys. Chem. Chem. Phys. 7 (2005) 538-547.
[27] J. Shu, A. Adnot, B.P.A. Grandjean, Ind. Eng. Chem. Res. 38 (1999) 3860-3867.
[28] S. Zheng, H.R. Heydenrych, A. Jentys, J.A. Lercher, J. Phys. Chem. B 106 (2002) 9552-9558.
[29] B.C. Lippens, J.H. de Boer, J. Catal. 4 (1965) 319-323.
[30] J. Datka, A.M. Tutek, J.H. Jehng, I.E. Wachs, J. Catal. 135 (1992) 186-199.
[31] Y. Shu, D. Ma, L. Xu, Y. Xu, X. Bao, Catal. Lett. 70 (2000) 67-73.
[32] D. Ma, Y. Shu, X. Bao, Y. Xu, J. Catal. 189 (2000) 314-325.
[33] D. Ma, Y. Shu, M. Cheng, Y. Xu, X. Bao, J. Catal. 194 (2000) 105-114.
[34] F. Solymosi, J. Cserényi, A. Szöke, T. Bánsági, A. Oszkó, J. Catal. 165 (1997) 150-161.
[35] B.S. Liu, L. Jiang, H. Sun, C.T. Au, Appl. Surf. Sci. 253 (2007) 5092-5100.
[36] J-P. Tessonnier, B. Louis, S. Rigolet, M.J. Ledoux, C. Pham-Huu, Appl. Catal. A 336 (2008)
79-88.
[37] L.W. Beck, T. Xu, J.B. Nicholas, J.F. Haw, J. Am. Chem. Soc. 117 (1995) 11594-11595.
[38] E.J.M. Hensen, D.G. Poduval, D.A.J.M. Ligthart, J.A.R. van Veen, M.S. Rigutto, J. Phys.
Chem. C 114 (2010) 8363-8374.
[39] E. Haro-Poniatowski, C. Julien, B. Pecquenard, J. Livage, M.A. Camacho-López, J. Mater. Res.
13 (1998) 1033-1037.
[40] R.W. Borry III, Y.H. Kim, A. Huffsmith, J.A. Reimer, E. Iglesia, J. Phys. Chem. B 103 (1999)
5787-5796.
[41] H.S. Lacheen, E. Iglesia, J. Catal. 230 (2005) 173-185.
[42] L. Wang, Y. Xu, W. Wong, W. Cui, X. Guo, Appl. Catal. A 152 (1997) 173-182.
[43] H. Liu, X. Bao, Y. Xu, J. Catal. 239 (2006) 441-450.
[44] C. Bouchy, I. Schmidt, J.R. Anderson, C.J.H. Jacobsen, E.G. Derouane, S.B. Derouane-Abd
Hamid, J. Mol. Catal. A 163 (2000) 283-296.
Chapter 3
72
Chapter 4
73
On the deactivation of Mo/HZSM-5 in the methane
dehydroaromatization reaction
Summary
The deactivation of Mo/HZSM-5 during the non-oxidative methane aromatization (MDA)
reaction that yields benzene and hydrogen was investigated. Catalysts were recovered from
the reactor after pre-activation and after increasing time on stream in methane. The physico-
chemical properties of the spent catalysts were characterized in detail by Ar physisorption, 27
Al MAS NMR and X-ray photoelectron spectroscopy. The nature of the carbon deposits was
determined by UV Raman spectroscopy and TGA, and the size and location of the Mo-carbide
particles by TEM and STEM-HAADF. The results show that the main cause for catalyst
deactivation is the formation of a carbonaceous layer at the external zeolite surface. This
layer is made up from polyaromatic hydrocarbons and decreases the accessibility of the
Brønsted acid sites in the micropores. At the same time, the decreased interaction of the Mo-
carbide particles with the external zeolite surface results in their sintering. The lower Mo-
carbide dispersion decreases methane conversion rates. The decreased accessibility of the
Brønsted acid sites shifts the selectivity from benzene to unsaturated intermediates formed on
the Mo-carbide particles. Silylation of the external surface mainly results in lower rate of
coke formation at the external surface, slowing down catalyst deactivation.
This chapter is published in Appl. Catal. B: Environ. 176 (2015) 731-739.
Chapter 4
74
4.1 Introduction
Dwindling fossil resources and concerns about the effect of their combustion on our
climate add urgency to the replacement of these non-renewable resources by renewable ones.
For example, to surpass petroleum as the main resource for aromatics scientists explore the
use of biomass as alternative feedstock [1]. A general problem is that the conversion
technologies for renewable resources such as biomass and solar energy are often at an early
stage of development. Usually, the development from the discovery of a novel energy
technology to large-scale commercial implementation takes decades [2]. Accordingly,
transition technologies based on relatively clean feedstock such as natural gas are increasingly
considered to fill the gap between conventional oil and coal based and future sustainable
processes. In energy scenarios, the abundance of natural gas is evident with estimated reserves
of 180 trillion ton cubic meters [3]. New technologies to extract natural gas from shale rock
are already leading to drastic changes in the energy and chemicals industry at the global scale
[1, 4]. On the other hand, a significant fraction of the natural gas reserves is located in
remotely located fields. Because of the small size of many of these fields, investments in
capital intensive transportation via pipelines or liquefaction are often not justified. For similar
reasons, it is common practice to flare associated gas from oil production in order to meet
safety and environmental legislations. Chemical conversion towards high-value fuels and
chemicals would be an alternative to make recovery of these natural gas resources
economically more attractive.
Currently, catalytic steam and autothermal reforming and gasification are common
technologies to convert natural gas into synthesis gas; synthesis gas serves as the platform for
the manufacture of a wide range of chemicals. The CAPEX and OPEX for the syngas
production step are very high, so that it is only profitable to construct large plants.
Accordingly, it remains a strong desire of the chemical industry to develop a simple process to
upgrade natural gas to liquids. One such process may be the direct aromatization of methane
to benzene under non-oxidative conditions (methane dehydroaromatization, MDA). Benzene
is an attractive intermediate, because it can be more easily transported than natural gas. In
addition, the increasing use of ethane from wet shale gas instead of naphtha to produce
ethylene is putting pressure on the aromatics supply. Direct methane to aromatics conversion
would be very desirable in this context [1].
Methane dehydroaromatization was first discussed by Wang et al. in 1993 [5].
Chapter 4
75
Mo/HZSM-5, the most common catalytic material for this reaction, is a bifunctional catalyst.
The molybdenum carbide phase is formed during carburization of the initial molybdenum
oxide phase. It converts methane into ethylene, while Brønsted acid sites in the shape-
selective micropores of HZSM-5 zeolite convert ethylene to benzene and other aromatics. The
process is typically operated at temperatures higher than 873 K because of the low reactivity
of methane. Thermodynamic equilibrium limits the reaction. In principle, higher reaction
temperatures than the most frequently reported one (973 K) result in higher benzene yield, but
also lead to rapid carbon laydown on the catalyst. The formation of coke, which causes
relatively fast deactivation of the catalyst, is the main challenge in the development of a
commercial process for methane dehydroaromatization [6].
Several studies have attempted to elucidate the reasons for the rapid deactivation of
Mo/HZSM-5 catalysts during the MDA reaction [7-17]. Extensive formation of polyaromatics
hydrocarbon carbon deposits was identified as the main reason for catalyst deactivation. The
formation of this type of carbon is assumed to take place at the Brønsted acid sites (BAS)
located on the external surface of the zeolite crystals [13, 14]. Some authors suggested that
such carbon species may eventually block the micropore apertures [7, 8]; this would explain
the increased formation of ethylene at the expense of benzene [9]. Deactivation of the external
surface BAS by silylation has been shown to decrease to some extent the formation rate of
such unwanted carbon species, but catalyst deactivation cannot be completely prevented in
this way [15-17]. Therefore, the coverage of the BAS inside the micropores by carbonaceous
species is also considered to contribute to catalyst deactivation [10]. We have recently
reported that silylation after Mo introduction yields better MDA catalysts than Mo
introduction after zeolite silylation [17].
Improvements of the MDA process may involve improved catalyst technology but also
reactor engineering approaches to cope with rapid catalyst deactivation. Such developments
will strongly hinge on better insight in the deactivation of the catalyst. In the present study, we
have investigated in detail the different stages in the life of a Mo/HZSM-5 catalyst during
methane dehydroaromatization. The precursor activated and deactivated catalysts after
different times on stream were extensively characterized for their physical and chemical
properties. The obtained results are captured in a model that describes deactivation during the
MDA reaction.
Chapter 4
76
4.2 Experimental
4.2.1 Catalyst synthesis
The proton form of ZSM-5 (Alsi-Penta) was obtained from Süd-Chemie (now Clarian).
The starting material had a Si/Al ratio of 15 as determined from ICP-AES elemental analysis.
For Mo loading, the zeolite was impregnated (incipient wetness impregnation) with an
aqueous solution of ammonium heptamolybdate tetrahydrate (AHM, Merck). The target Mo
content was 6 wt%. After impregnation, the material was dried for 1 h at room temperature
followed by calcination in artificial air at 823 K for 5 h after heating to the final temperature
at a rate of 1.5 K/min. Molybdenum modified zeolites are denoted by Mo/HZSM-5.
A portion of the molybdenum modified zeolite was silylated following a method adapted
from Ding et al. [14]. Typically, 2 g of zeolite was dried overnight at 373 K and then
dispersed in 50 ml n-hexane. To the suspension, an amount of 0.3 ml tetraethyl orthosilicate
(TEOS, Merck) was added and stirred for 1 h under reflux. The amount of TEOS added
corresponds to 0.4 wt% based on the amount of zeolite in the suspension. The treated zeolite
was then filtered off and dried overnight at 373 K. The resulting zeolite was calcined by a
two-step procedure in artificial air. The first step consisted of heating the sample at a rate of a
2 K/min to 393 K followed by an isothermal period of 2 h. In the second step the temperature
was increased to 773 K at a rate of 0.2 K/min followed by an isothermal period for 4 h. The
silylated Mo/HZSM-5 is denoted as Mo/HZSM-5(Si).
4.2.2 Characterization
The Mo and Al content of the zeolites was determined by inductively coupled plasma optical
emission spectroscopy (ICP-OES, Spectro CIROS CCD spectrometer). Prior to ICP
measurements, the zeolite samples were dissolved in a mixture of HF/HNO3/H2O (1:1:1).
UV Raman spectra were recorded with a Jobin–Yvon T64000 triple stage spectrograph
with a spectral resolution of 2 cm−1
. The 244 nm line at of a Lexel 95-SHG laser was used as
the excitation source. The power of the laser on the sample was about 2 mW.
Magic angle spinning (MAS) 27
Al single pulse NMR spectra were recorded on a Bruker
Avance DMX-500 NMR spectrometer (11.7 T; the Al resonance frequency at this field is
130.3 MHz). A 2.5 mm MAS probe head was used. The 27
Al chemical shift was referenced to
a saturated Al(NO3)3 solution. In a typical experiment, 10 mg of well-hydrated sample was
packed in a 2.5 mm zirconia rotor. The MAS sample rotation speed was 20 kHz. 27
Al NMR
Chapter 4
77
spectra were recorded with a single pulse sequence with 18 o pulse duration of 1 μs and a
interscan delay of 1 s.
Argon sorption isotherms were measured at 87 K on a Micromeritics ASAP2020 system
in static measurement mode. The samples were outgassed at 623 K for 8 h prior to the
sorption measurements. The Brunauer–Emmett–Teller (BET) equation was used to calculate
the specific surface area (SBET) in the pressure range p/p0 = 0.05–0.25. The mesopore volume
(Vmeso) and mesopore size distribution were calculated using the Barrett–Joyner–Halenda
(BJH) method on the adsorption branch of the isotherm. The micropore area (Smicro) and
micropore volume (Vmicro) were calculated from the t-plot curve using the thickness range
between 3.5 and 5.4 Å [19].
Transmission electron micrographs were obtained with a FEI Tecnai 20 instrument at an
electron acceleration voltage of 200 kV. Typically, a small amount of sample was suspended
in ethanol, sonicated and dispersed over a Cu grid with a holey carbon film.
Weight-loss curves were obtained from thermogravimetric analysis (TGA) using a
Mettler Toledo TGA/DSC 1 apparatus. Samples were heated in uncovered alumina crucibles
at a rate of 5 K/min to 1023 K in a 2/1 (v/v) He/O2 flow.
Isotopic H/D exchange between C6H6 and C6D6 was carried out in a 10-channel parallel
microflow reactor setup. Typically, 50 mg of zeolite was loaded in each quartz tubular reactor
with an internal diameter of 4.0 mm. Zeolites were first pelletized and then crushed and
sieved in a 125-250 μm mesh fraction. The ten quartz tubes with the catalyst contained
between quartz wool plugs were then placed in the parallel reactor setup. Before reaction
samples were dehydrated in a He flow to 723 K at a rate of 5 K/min followed by an
isothermal period of 6 h. After cooling to 303 K, the reaction was started by switching the
reactor feed to the gas flow containing a 90/10 (v/v) C6H6/C6D6 mixture. The effluent
products were analyzed by mass spectrometry. The rate of H/D exchange was determined as a
function of temperature.
4.2.3 Catalytic activity measurements
An amount of 0.5 g of catalyst was introduced in a tubular quartz reactor with a length of
490 mm and an internal diameter of 4.0 mm. The catalyst was supported on quartz wool in the
isothermal zone of the oven. All gases were fed using thermal mass controllers. The catalyst
was activated by increasing the temperature at a rate of 10 K/min to 973 K in a (80/20) (v/v)
Chapter 4
78
CH4/He gas flow of 25 ml/min. The reaction was started by switching the reactor feed to a
N2/CH4 mixture (5 vol% N2 in CH4) at a WHSV of 1.22 h-1
(1710 ml CH4/gcat.h at STP
conditions). Products were analyzed by an online gas chromatograph (Interscience
CompactGC) equipped with three analysis channels for analysis of light gasses (Molsieve 5A,
TCD), aromatics (Al2O3/KCl, TCD) and light hydrocarbons (Rtx-1, FID).
Nitrogen was used as the internal standard in order to determine the methane conversion
(XCH4) and the methane conversion rate (rCH4). The weight-based selectivities and reaction
rates of the products (rproduct) were determined using response factors for the various
compounds (main products are benzene, toluene, naphthalene, ethylene, ethane, propylene,
propane). The coke selectivity (Scoke) and the rate of coke formation (rcoke) were determined
by mass balance considerations.
To follow the deactivation process, samples were recovered after different reaction times
by rapidly cooling the reactor under flowing He. The Mo/HZSM-5 samples were then
recovered from the reactor in air. This procedure was carried out after activation of the
catalyst in CH4/He and activation in CH4/He followed by reaction for 5 min, 2 h, 5 h and 10 h.
The corresponding samples are denoted by Mo/HZSM-5(act), Mo/HZSM-5(0.08h),
Mo/HZSM-5(2h), Mo/HZSM-5(5h) and Mo/HZSM-5(10h).
4.3 Results and discussion
4.3.1 Catalytic activity measurements
Catalytic performance data for Mo/HZSM-5 and Mo/HZSM-5(Si) catalysts in the MDA
reaction are shown in Fig. 4.1. The trends in catalytic activity and selectivity with time on
stream were similar for both catalysts. In line with earlier reports [20], the methane
conversion rate gradually decreased. After 10 h on stream, nearly all catalytic activity was lost.
During reaction, the benzene formation rate strongly decreased. At the same time, the
ethylene selectivity increased. These changes point to rapid deactivation of the BAS that
convert ethylene to aromatics. It is also seen that, at the initial stages of the reaction, the
benzene selectivity was low and the coke selectivity was as high as 50 wt%. In this induction
period, part of the methane in the feed is used to convert the Mo-oxide phase into Mo-carbide
[20]. This implies that the Mo-oxide precursor was not totally converted to Mo-carbides
during heating in CH4/He. Concomitant with Mo-carbide formation, the coke selectivity
decreased and the benzene selectivity increased. The highest benzene selectivity was observed
Chapter 4
79
after a reaction time of 1-2 h. The benzene selectivity was higher for Mo/HZSM-5(Si) (75
wt%) than for Mo/HZSM-5 (64 wt%). After 2 h, the benzene selectivity decreased, first
relatively slowly, and then more rapidly after 5 h. Especially during the rapid decrease of
methane conversion, the ethylene and coke selectivity strongly increased. The coke selectivity
for Mo/HZSM-5(Si) was lower compared with its non-silylated counterpart. This difference is
in agreement with the coke content of the spent catalysts recovered after 10 h on stream
(Table 4.1). The silylated sample contained less hard coke as determined by TGA. This hard
coke is made up from polyaromatic hydrocarbons (PAHs).
Table 4.1. Carbon analysis of the spent Mo/HZSM-5 and Mo/HZSM-5(Si) catalysts
determined by TGA.
Sample Ctotal
(wt%)
CMoC
(wt%)
Csoft
(wt%)
Chard
(wt%)
HZSM-5 - - - -
Mo/HZSM-5 - - - -
Mo/HZSM-5(act) 3.7 0.9 2.8 0
Mo/HZSM-5(0.08h) 3.7 0.7 3.0 0
Mo/HZSM-5(2h) 7.8 0 4.0 3.8
Mo/HZSM-5(5h) 11.3 0 4.1 7.2
Mo/HZSM-5(10h) 14.2 0 3.8 10.4
Mo/HZSM-5(Si) - - - -
Mo/HZSM-5(Si,act) 4.7 4.1 0.6 0
Mo/HZSM-5(Si,2h) 9.3 0 3.3 6.0
Mo/HZSM-5(Si,10h) 12.3 0 2.5 9.8
Chapter 4
80
Fig. 4.1. Activity data for Mo/HZSM-5 (squares) and Mo/HZSM-5(Si) (circles) for methane dehydroaromatization showing (a) the CH4
reaction rate, (b) benzene selectivity and (c) coke (closed symbols) and olefin selectivity (open symbols) as a function of the reaction time.
0 2 4 6 8 10
0
2
4
6
8
10
12
14
16
18
20
CH
4 r
eacti
on
rate
(m
mo
l/h
.gc
at)
Time on stream (h)
0 2 4 6 8 10
0
20
40
60
80
100
Ben
zen
e s
ele
cti
vit
y (
wt%
)
Time on stream (h)
0 2 4 6 8 10
0
20
40
60
80
100
Sele
cti
vit
y (
wt%
)
Time on stream (h)
Chapter 4
81
4.3.2 Catalyst characterization
The XRD patterns of the fresh catalysts show that the introduction of Mo did not strongly
affect the crystallinity of the zeolites (Table 4.2). There are no indications for the presence of
large MoO3 crystallites in the diffractograms. The textural properties of the zeolite catalysts
are listed in Table 4.2. The decrease in the micropore volume upon introduction of Mo
indicates that a fraction the Mo phase is located in the zeolite micropores [20]. Activation in a
mixture of CH4/He at 973 K led to a decrease of the micropore volume, which may be due to
formation of Mo-carbide species and the deposition of carbonaceous species close to these
Mo-carbides [10]. The catalysts recovered directly after activation and after 5 min and 2 h on
stream all had comparable micropore volume (Vmicro ≈ 0.07 cm3/g). This suggests that the rate
of carbon deposition during the first 2 h of the reaction was relatively low. Samples recovered
after 5 h and 10 h had much lower micropore volumes of 0.04 cm3/g and 0 cm
3/g,
respectively. The decrease shows that substantial amounts of coke formed inside the
micropores or at the external surface, blocking the micropore entrances during the second
stage of the reaction. We verified for spent Mo/HZSM-5 that the complete loss in micropore
volume was not due to the amorphization of the zeolite; after a reaction time of 10 h, the
zeolite crystallinity remained at 64%.
Table 4.2. Textural properties of fresh and spent Mo/HZSM-5 and Mo/HZSM-5(Si).
Sample Vmicro
(cm3/g)
Vmeso
(cm3/g)
Smicro
(m2/g)
Smeso
(m2/g)
SBET
(m2/g)
XRD crystallinity
(%)
HZSM-5 0.13 0.02 246 10 251 100
Mo/HZSM-5 0.10 0.03 204 14 244 93
Mo/HZSM-5(act) 0.07 0.04 130 18 164 -
Mo/HZSM-5(0.08h) 0.07 0.03 135 19 179 -
Mo/HZSM-5(2h) 0.07 0.04 140 22 196 -
Mo/HZSM-5(5h) 0.04 0.023 87 12 115 -
Mo/HZSM-5(10h) 0 0.02 3 7 8 -
Mo/HZSM-5(Si) 0.06 0.02 131 11 162 90
Mo/HZSM-5(Si,Act) 0.05 0.03 105 15 137 -
Mo/HZSM-5(Si,2h) 0.03 0.03 6 18 104 -
Mo/HZSM-5(Si,10h) 0.00 0.01 5 5 4 -
The Al speciation in the fresh and spent zeolites was characterized by 27
Al MAS NMR
spectroscopy. The weight-normalized 27
Al MAS NMR spectrum of HZSM-5 (Fig. 4.2)
contains a dominant feature at δ = 55 ppm (δ, chemical shift) due to framework Al (FAl)
Chapter 4
82
atoms and a smaller feature at δ = 0 ppm due to extraframework Al (EFAl) species.
Modification of HZSM-5 with Mo (Mo/HZSM-5) led to a decrease and broadening of the FAl
signal caused by the proximity of cationic Mo-oxo complexes that replace the protons [21].
Two weak features at δ = -11 ppm and δ = 14 ppm are due to EFAl species in the form of
Al2(MoO4). The NMR spectra of the spent samples show a decreasing FAl content for
increasing MDA reaction time. Although the weak shoulder at δ = 30 ppm due to distorted
four-coordinated or five-coordinated Al species [22] increased slightly, the decrease in FAl
content is not paralleled by an increase in EFAl content. The increasing amount of NMR-
invisible Al may be related to framework damage occurring during the MDA reaction. It is
however at odds with the remaining crystallinity of the spent Mo/HZSM-5(10h) sample.
Another explanation is that rehydration of the spent samples before the 27
Al NMR
measurements may be incomplete because of the decreased micropore accessibility. The
resulting asymmetric Al coordination environments can in principle also explain the
decreased NMR-visibility of the Al atoms.
Fig. 4.2. 27
Al MAS NMR spectra of the parent, activated and spent Mo/HZSM-5 catalysts.
Conventional characterization of the surface acidity by IR spectroscopy of the activated and
spent samples was not possible, because the samples were black due to carbon laydown. We
used the low-temperature isotopic exchange reaction between perdeuterobenzene and benzene
to probe the Brønsted acidity of the activated and spent Mo/HZSM-5 catalysts. The isotopic
exchange reaction between C6H6 and C6D6 can be catalyzed by BAS at relatively low
150 100 50 0 -50
(ppm)
Chapter 4
83
temperatures [23]. This reaction has been earlier successfully employed in order to determine
the concentration of BAS of zeolites, clays and amorphous silica-alumina [24-26]. We used
the rate of C6H5D formation at a temperature of 313 K as a measure for the number of BAS.
This isotopomer is the main reaction product of isotopic exchange reaction of C6H6, because
C6H6 was present in tenfold excess to C6D6 in the reactant feed mixture. The absolute reaction
rates and the relative reaction rates compared with the parent HZSM-5 zeolite are given in
Table 4.3. The reaction rate of Mo/HZSM-5 was only slightly lower than that of the parent
HZSM-5 zeolite, consistent with the relatively small decrease in the bridging hydroxyl
density upon Mo introduction as probed by IR spectroscopy [17, 27, 28]. Activation in
methane resulted in a significant decrease of the Brønsted acidity. Although it is usually
assumed that some BAS are regenerated by the carburization of cationic Mo-oxo complexes,
we found that the overall acidity decreased further during the carburization step. Together
with the textural data, we conclude that blockage of the micropores was the main cause of the
decreased acidity. The H/D exchange data show that the accessible acidity was lower after
prolonged reaction. After 5 min on stream, already half of the acid sites of the fresh
Mo/HZSM-5 were not involved in the H/D exchange reaction anymore. The acidity gradually
decreased for prolonged reaction times and, after 10 h on stream, the accessible acidity was
nearly completely lost. The acidity decrease strongly correlates with the loss in micropore
volume as determined by Ar physisorption.
Table 4.3. Physico-chemical properties of fresh and spent Mo/HZSM-5 and Mo/HZSM-5(Si).
Sample Mo/Al1 IFAl
2
(%)
rH/D3
(mmol/min.gcat)
normalized rH/D 4
HZSM-5 - 100 1.27 1
Mo/HZSM-5 0.39 53 1.18 0.92
Mo/HZSM-5(act) 0.39 37 0.79 0.62
Mo/HZSM-5(0.08h) 0.39 40 0.61 0.46
Mo/HZSM-5(2h) 0.39 34 0.36 0.28
Mo/HZSM-5(5h) 0.39 23 0.14 0.11
Mo/HZSM-5(10h) 0.39 11 0 0
Mo/HZSM-5(Si) - 42 - -
Mo/HZSM-5(Si,act) - 23 - -
Mo/HZSM-5(Si,2h) - 14 - -
Mo/HZSM-5(Si,10h) - 15 - - 1 Atomic Mo/Al ratio as determined by ICP-OES analysis;
2 FAl concentration references to HZSM-5
as determined by 27
Al MAS NMR spectroscopy; 3 Isotopic exchange rate.; Relative isotopic exchange
rate normalized to HZSM-5.
Chapter 4
84
The Mo speciation in the external surface region of the zeolite crystals was investigated
by XPS. Table 4.4 collects the XPS results including reference binding energies for various
Mo species taken from literature [29]. It is difficult to discern between metallic Mo and highly
dispersed Mo2C (Mo2Csmall), because these species have nearly similar binding energies [30].
Wang et al. mentioned that carburization of MoO3 to Mo2C particles is thermodynamically
favored over full reduction of Mo species to metallic Mo under MDA conditions [30]. The
Mo speciation is given in Table 4.4. The parent Mo/HZSM-5 zeolite mainly comprised MoO3.
Upon activation in methane, MoO3 was almost completely converted into highly dispersed
Mo2C. Small amounts of MoO2 and large Mo2C particle (Mo2Clarge) were also observed. After
5 min of reaction (Mo/HZSM-5(0.08)), the oxidation degree of the Mo phase was higher than
directly after carburization. A likely explanation is that the replacement of the diluted
methane feed used for pre-carburization by the pure methane reactant feed led to a much
higher carburization rate of the remaining oxides and formation of water. The formation of
water may result in re-oxidation of some of the Mo-carbides at the surface. After prolonged
reaction, the amount of low-dispersed Mo2C particles has increased at the expense of highly
dispersed Mo2C particles, pointing to slow sintering of the initially highly dispersed Mo2C
particles during the MDA reaction. The XPS data show that, at the same time, the Si/C and
Al/C ratios in the surface region decreased (Table 4.4). We interpret this in terms of the
formation of a carbonaceous layer around the zeolite that separates the Mo-phase from the
zeolite surface.
Chapter 4
85
Table 4.4. Mo and C speciation in surface region of spent Mo/HZSM-5 and Mo/HZSM-5(Si)
catalysts as determined by XPS.1
Sample MoO3
(%)
MoO2
(%)
Mo2Csmall
(%)
Mo2Clarge
(%)
Atomic ratios
Si/Al Al/C Si/C
Mo/HZSM-5 100 0 0 0 18.5 0.43 1.64
Mo/HZSM-5(act) 1 15 65 18 19.3 0.23 0.92
Mo/HZSM-5(0.08h) 16 32 27 25 19.4 0.13 0.53
Mo/HZSM-5(2h) 11 30 23 37 20.8 0.08 0.35
Mo/HZSM-5(5h) 10 30 5 55 19.3 0.05 0.19
Mo/HZSM-5(10h) 7 27 11 55 22.3 0.02 0.09
Mo/HZSM-5(Si) 100 0 0 0
Mo/HZSM-5(Si,act) 33 39 4 23 22.2 0.12 0.52
Mo/HZSM-5(Si,2h) 15 35 19 31 22.1 0.09 0.43
Mo/HZSM-5(Si,10h) 1 29 11 60 22.0 0.004 0.14 1 Binding energy of Mo species from Ref. 29: MoO3: 232.7 eV; MoO2: 229.8 eV; Mo2Csmall: 227.6 eV;
Mo2Clarge: 228.0 eV.
The nature of the Mo phase and carbonaceous deposits in the fresh, activated and spent
catalysts was investigated by UV Raman spectroscopy (Fig. 4.3). The Raman signal upon 244
nm excitation mostly derives from the species present at the external surface of the zeolite
crystals [12]. The Raman spectrum of HZSM-5 shows bands that are typical for MFI zeolite
[18]. The most prominent band at 380 cm-1
can be assigned to the double-five-ring vibration
of the MFI framework. In Mo/HZSM-5, the intensity of this band is very low, which is caused
by the presence of Mo-oxo species at the external zeolite surface [31]. Compared with
HZSM-5, the spectrum of Mo/HZSM-5 contains additional bands at 280 cm-1
, 336 cm-1
, 821
cm-1
and 995 cm-1
, which can be attributed to α-MoO3 [32-34,8]. Upon activation in methane,
the α-MoO3 signals disappeared as a result of carburization of the Mo-oxides [35]. New broad
bands appear at ~1400 cm-1
and ~1600 cm-1
. A closer look at this region (right panel of Fig. 3)
shows that the band at 1600 cm-1
shifted towards lower wavenumbers for prolonged reaction
times. Li et al. have assigned Raman bands in this region to various types of carbon [37].
PAHs and graphitic carbon give rise to bands at 1595 cm-1
and 1585 cm-1
, respectively [37].
Accordingly, we attribute the spectral changes to the formation of increasing amounts of
PAHs and graphitic carbon during the MDA reaction. The shoulders visible at 1610 cm-1
and
1560 cm-1
relate to adsorbed naphthalene molecules and to conjugated olefinic species,
respectively [37]. The broad band at 1385 cm-1
is typical for coke formed from olefinic
precursors. The band originally positioned at 1385 cm-1
, which shifted towards 1365 cm-1
after prolonged reaction; is characteristic for graphitic carbon [37]. In summary, the Raman
Chapter 4
86
spectra point to the formation of significant amounts of PAHs and graphitic carbon in the
spent samples.
Fig. 4.3. UV Raman spectra (λexcitation = 244 nm) of the parent, activated and spent Mo/HZSM-
5 catalysts.
TGA was employed to characterize the carbonaceous deposits on the spent catalysts (Fig.
5.4). Three types of carbon were distinguished as a function of the calcination temperature
during TGA in artificial air. These include relatively light carbonaceous species associated
with Mo-carbides (CMoC, ~693 K), soft coke (Csoft, ~753 K) and hard coke (Chard, ~813 K).
Soft coke is thought to be amorphous in nature and likely formed in the proximity of Mo-
carbide particles [38]. Hard coke is mainly comprised of PAHs formed by reactions of olefins
on BAS located at the external surface of the zeolite. Table 4.1 lists the results of
deconvolution of the TGA curves in Fig. 4.4. The catalysts recovered after the activation step
and after 5 min reaction (Mo/HZSM-5(0.08)) contained about 25% CMoC and 75% Csoft and
very little hard coke. The total amount of carbon was similar for these samples. Longer
reaction times led to a significant increase of the total carbon content, almost exclusively in
the form of hard coke.
1200 1400 1600
Inte
nsit
y (
a.u
.)
Wavenumber (cm-1)
400 800 1200 1600
Inte
nsit
y (
a.u
.)
Wavenumber (cm-1)
Chapter 4
87
Fig. 4.4. TGA weight-loss curves of spent Molybdenum modified ZSM-5.
We used transmission electron microscopy to study the dispersion of the Mo-oxide/carbide
phase in more detail. Bright-field TEM images are shown in Fig. 4.5; Fig. 4.6 reports
HAADF-STEM images. For fresh Mo/HZSM-5, a few large MoO3 particles with sizes up to
100 nm are visible at the external zeolite surface. As the XRD patterns did not contain
evidence for such large MoO3 particles, we conclude that the amount of such large particles is
relatively small. Activation in methane converted the Mo-oxide particles to small Mo2C
particles (Fig. 4.5b). Only a relatively small amount of Mo2C particles larger than 10 nm are
visible in the EM images. Close inspection of the images reveals that an amorphous carbon
layer has formed at the external surface of the zeolite crystals. Some of the Mo2C particles are
separated from the zeolite crystal by this carbonaceous layer. After activation, also some
carbon nanotubes are visible in line with earlier findings [39]. The TEM images of the
catalyst recovered after 2 h of reaction (Fig. 4.5c) showed a much lower density of small-
sized Mo2C particles. Clearly, the Mo-carbide particles sinter during the MDA reaction. After
2 h, the amorphous carbon layer covers a significant fraction of the external zeolite surface.
Compared with Mo/HZSM-5(act), more Mo2C particles were seen to be separated from the
zeolite surface by this carbonaceous layer. After 10 h on stream, the carbonaceous layer was
much thicker; it is difficult to discern the external surface of the zeolite (Fig. 4.5d). In this
sample, the Mo2C phase was present as relatively large particles. The images taken in STEM-
HAADF mode (Fig. 4.6) serve to illustrate the gradual transformation of the Mo-carbide
600 800 1000
Weig
ht
loss r
ate
(m
g/K
)
Temperature (K)
Chapter 4
88
phase from highly dispersed particles in the activated catalyst towards large agglomerated
particles after 10 h reaction. It supports the conclusion that the dispersion of the Mo-oxide
phase is initially very high. Upon carburization, they slowly sinter during the MDA reaction;
this process is accelerated by the detachment of these particles from the external zeolite
surface support, which is caused by the carbonaceous layer.
Fig. 4.5. Transmission electron micrographs of (a) Mo/HZSM-5, (b) Mo/HZSM-5 (act), (c)
Mo/HZSM-5 (2h) and (d) Mo/HZSM-5 (10h).
20 nm
100 nm 20 nm
Small Mo2C
Large Mo2C
Carbon nanotube
Carbon deposits + Mo2C
Carbon deposits
Large Mo2C
20 nm
Small Mo2C
Large Mo2C
Carbon deposits
b
c d
200 nm
a
Chapter 4
89
Fig. 4.6. STEM-HAADF micrographs of (a) Mo/HZSM-5, (b) Mo/HZSM-5 (act), (c)
Mo/HZSM-5 (2h) and (d) Mo/HZSM-5 (10h).
We also characterized some of the spent Mo/HZSM-5(Si) catalysts. The results were
qualitatively similar to those obtained for the spent Mo/HZSM-5 samples. A gradual decrease
in the micropore volume (Table 4.2) was observed with reaction time. The micropore volume
of the sample retrieved after 10 h reaction was also negligible for the silylated zeolite catalyst.
The FAl content as determined by NMR spectroscopy also decreased with the progressing
reaction. For silylated Mo/HZSM-5, XPS data point out the formation of a carbonaceous layer
around the zeolite crystals. TGA confirms that PAHs are the main compounds in the
carbonaceous layer formed around the zeolite crystals. Although the general trends are
similar, some subtle differences can be noted between Mo/HZSM-5 and Mo/HZSM-5(Si).
The micropore volumes for the spent Mo/HZSM-5(Si) catalysts were lower than those of the
Mo/HZSM-5 analogues. This is due to the improved spreading of the Mo phase upon
silylation [17]. Activated Mo/HZSM-5(Si) contained more MoO3 and MoO2 in comparison
with Mo/HZSM-5(act). This difference suggests slower carburization of the Mo-oxide
precursor in silylated Mo/HZSM-5, possibly due to the increased interaction of the Mo-oxide
0.5 μm
50 nm
0.2 μm0.2 μm
50 nm
a b
c d
Chapter 4
90
precursor with the zeolite surface. TGA of the carbonaceous deposits revealed a relative large
amount of soft coke after activation of Mo/HZSM-5(Si)). Combined with the XPS and
textural analysis data, these findings support the conclusion that the Mo dispersion upon the
silylation treatment of Mo/HZSM-5 was improved. The overall carbon content of the spent
Mo/HZSM-5(Si) after 10 h was lower compared to that of the non-silylated analogue. The
initial total carbon content of Mo/HZSM-5(Si) was higher than that of Mo/HZSM-5. We
believe that the amount of carbon formed due to undesired side-reactions in Mo/HZSM-5(Si,2
h) is overestimated by the TGA analysis. This is caused by the encapsulation of MoCx with a
layer of carbonaceous deposits. As these deposits have to be oxidized (at higher temperature)
before the MoCx particles can be oxidized, the TGA curves cannot be used to determine the
content of the Mo-carbide particles. In this way, the amount of coke is overestimated and the
amount of MoCx is underestimated. Subtracting the initial CMoC content from the Ctotal in the
spent catalysts recovered after 2 h in the MDA reaction indicates that less carbon deposits
were formed on Mo/HZSM-5(Si,2 h) (9.3–4.1 = 5.2wt%) compared with Mo/HZSM-5(2h)
(7.8–0.9 = 6.9wt%).
4.3.3 Deactivation mechanism
This study investigated the deactivation of the Mo/HZSM-5 catalyst in the bifunctional
MDA reaction. It is generally assumed that the Mo-carbide phase converts methane into
ethylene and hydrogen. The olefins are then reacted on the BAS to aromatic compounds. The
main focus of the present investigation was on the changes of the catalyst upon activation and
during reaction that lead to catalyst deactivation. An important finding of the present study is
that a thick carbonaceous layer is formed at the external zeolite surface. This layer blocks the
access of the olefinic intermediates to the acid sites. In addition, it leads to the detachment of
the Mo-carbide particles from the zeolite surface, thereby accelerating their sintering. While
the lower accessibility of the acid sites mainly affects the product distribution, the sintering of
the Mo-carbide phase leads to decreased methane conversion rates. During catalyst
preparation, Mo modification of the parent zeolite (Fig. 4.7a) leads to the migration of a small
fraction of mobile Mo-oxo species into the micropore space during the calcination step (Fig.
4.7b). Most of the Mo-oxo species remain at the external surface, predominantly in the form
of highly dispersed particles, but also as some larger MoO3 crystallites. Upon activation in
methane, the Mo-oxide carburizes (Fig. 4.7c). The resulting Mo-carbide (MoCx) particles are
Chapter 4
91
mainly present in highly dispersed form at the external surface and inside the micropores.
During the carburization process, the micropore volume became lower, which may be
attributed to the growth of the Mo-carbide species in the micropores as well as the formation
of amorphous polyolefinic (soft coke) associated with Mo-carbides. At the external surface, a
small amount of PAHs (hard coke) forms. The BAS located at the external surface are most
likely involved in the formation of these hard coke deposits. We speculate that also silanol
groups may be implicated in oligomerization reactions of unsaturated intermediates at the
very high temperatures used for the MDA reaction.
During the first 2 h of reaction, the micropore volume did not substantially change. After
2 h, part of the highly dispersed Mo2C particles at the external surface agglomerated into
larger particles as evidenced by TEM measurements (Fig. 4.7d). Agglomeration of the MoCx
phase may be attributed to the decreased interaction of the particles with the external zeolite
surface. This is caused by the formation of a carbon layer at the external surface that separates
the MoCx particles from the zeolite external surface as seen in the TEM images. It has been
reported that large MoCx particles are undesired, because they exhibit high selectivity towards
coke and, accordingly, accelerate deactivation [39]. Based on our results, we argue that the
loss in MoCx dispersion is the main reason for the decreased methane conversion rate
observed in the catalytic performance data.
After 2 h, the PAHs layer rapidly grows over the external zeolite surface (Figs. 4.7e and
4.7f) and, in this way, lowers the micropore volume. Consequently, the amount of BAS
accessible to olefinic intermediates is decreased. As the formation of hard coke correlates well
with the loss in micropore volume, we attribute the decrease in aromatics selectivity with time
on stream to the decreased accessibility of the shape-selective BAS located in the micropores.
Agglomeration of MoCx particles and formation of soft coke inside the micropores may also
contribute to the decreasing pore volume. After 10 h on stream, the thickness of the
polyaromatic carbon layer has grown to 20 nm (Fig. 4.7f). It leads to the nearly complete
inaccessibility of the micropores.
Chapter 4
92
Fig. 4.7. Schematic representation of the catalysts physical state after various times on stream. The figures represent (a) parent ZSM-5, (b)
Mo/HZSM-5, (c) Mo/HZSM-5 (act), (d) Mo/HZSM-5 (2h), (e) Mo/HZSM-5 (5h) and (f) Mo/HZSM-5 (10h).
a b c
d e f
zeolite
MoO3
MoCx
hard coke
soft coke
Chapter 4
93
4.4 Conclusions
The deactivation of Mo/HZSM-5 catalysis in the MDA reaction was investigated. The
parent zeolite, the Mo-modified zeolite catalyst and activated and spent catalysts were
characterized in detail for their textural properties, the dispersion and location of the Mo-
oxide/carbide phase and the nature of the carbonaceous deposits. After carburization of the
Mo-oxide phase, optimum performance in terms of benzene yield is reached. The formation
of a carbonaceous layer consisting of PAHs at the external surface decreases the accessibility
of the BAS in the zeolite micropores. Simultaneously, this carbonaceous layer lowers the
interaction of the Mo-carbide particles with the external zeolite surface. As a consequence, the
Mo-carbide particles sinter, which explains the decreased methane conversion rate. The lower
Brønsted acidity shifts the selectivity from benzene to unsaturated olefinic intermediates. The
hard coke layer at the external surface is due to acid sites that can be partially deactivated by
silylation.
4.5 Reference
[1] P.C.A. Bruijnincx, B.M. Weckhuysen, Angew. Chem. Int. Ed. 52 (2013) 11980-11987.
[2] G.J. Kramer, M. Haigh, Nature 462 (2009) 568-569.
[3] British Petroleum Company. BP Statistical Review of World Energy. London: British Petroleum
Co, (2013) pp. 20.
[4] H. Rogers, Oxford Rev. Econ. Pol. 27 (2011) 117-143.
[5] L. Wang, L. Tao, M. Xie, G. Xu, Catal. Lett. 21 (1993) 35-41.
[6] B.S. Liu, J.W.H. Leung, L. Li, C.T. Au, A.S-C. Cheung, Chem. Phys. Lett. 430 (2006) 210-214.
[7] J. Bai S. Liu, S. Xie, L. Xu, L. Lin, Catal. Lett. 90 (2003) 123-130.
[8] J. Bai, S. Lie, S. Xie, L. Xu, L. Lin, React. Kinet. Catal. Lett. 82 (2004) 279-286.
[9] S. Li, d. MA, Q. Kan, P. Wu, Y. Peng, C. Zhang, M. Li, Y. Fu, J. Shen, T. Wu, X. Bao, React.
Kinet. Catal. Lett. 70 (2000) 349-356.
[10] D. Ma, Y. Shu, M. Cheng, Y. Xu, X. Bao, J. Catal. 194 (2000) 105-114.
[11] E.V. Matus, I.Z. Ismagilov, O.B. Sukhova, V.I. Zaikovskii, L.T. Tsikoza, Z.R. Ismagilov, Ind.
Eng. Chem. Res. 46 (2007) 4063-4074.
[12] F. Fan, K. Sun, Z. Feng, H. Xia, B. Han, Y. Lian, P. Ying, C. Li, Chem. Eur. J. 15 (2009) 3268-
3276.
[13] T. Behrsing, H. Jaeger and J. V. Sanders, Appl. Catal. 54 (1989) 289–302.
[14] W.P. Ding, G.D. Meitzner, E. Iglesia, J. Catal. 206 (2002) 14-22.
[15] H.M. Liu, Y. Li, W.J. Shen, X.H. Bao, Y.D. Xu, Catal. Today 93–95 (2004) 65-73.
[16] S. Kikuchi, R Kojima, H.T. Ma, J. Bai, M. Ichikawa, J. Catal. 242 (2006) 349-356.
[17] C.H.L. Tempelman, V.O. de Rodrigues, E.H.R. van Eck, P.C.M.M. Magusin, E.J.M. Hensen,
Microporous Mesoporous Mater. 203 (2015) 259-273.
[18] J. Chen, Z. Feng, P. Ying, C. Li, J. Phys. Chem. B 108 (2004) 12669-12676.
[19] B.C. Lippens, J.H. de Boer, J. Catal. 4 (1965) 319–323.
[20] Y. Shu, D. Ma, L. Xu, Y. Xu, X. Bao, Catal. Lett. 70 (2000) 67-73.
[21] J-P Tessonnier, B. Louis, S. Rigolet, M. J. Ledoux, C. Pham-Huu, Appl. Catal. A 336 (2008)
79–88.
[22] R.K. Sato, P.F. McMillan, P. Dennison, R. Dupree, J. Phys. Chem. 95 (1991) 4483-4489.
Chapter 4
94
[23] R.A. van Santen, G.J. Kramer, Chem. Rev. 95 (1995) 637-660.
[24] L.W. Beck, T. Xu, J.B. Nicholas, J.F. Haw, J. Am. Chem. Soc. 117 (1995) 11594-11595.
[25] E.J.M. Hensen, D.G. Poduval, V. Degirmenci, D.A.J.M. Ligthart, W. Chen, F. Maugé, M.S.
Rigutto, J.A.R. van Veen, J. Phys. Chem. C 114 (2010) 8363-8374.
[26] E.J.M. Hensen, D.G. Poduval, D.A.J.M. Ligthart, J.A.R. van Veen, M.S. Rigutto, J. Phys.
Chem. C 114 (2010) 8363-8374.
[27] R.W. Borry III, Y.H. Kim, A. Huffsmith, J.A. Reimer, E. Iglesia, J. Phys. Chem. B 103 (1999)
5787-5796.
[28] J-P. Tessonnier, B. Louis, S. Walspruger, J. Sommer, M-J. Ledoux, C. Pham-Huu, J. Phys.
Chem. B 110 (2006) 10390-10395.
[29] F. Solymosi, J. Cserényi, A. Szöke, T. Bánsági, A. Oszkó, J. Catal. 165 (1997) 150-161.
[30] D. Wang, J.H. Lunsford, M.P. Rosynek, Top. Catal. 3 (1996) 289-297.
[31] C. Li, J. Catal. 216 (2003) 203-212.
[32] G. Mestl, T.K.K. Srinivasan, Catal. Rev.-Sci. Eng. 40 (1998) 451-570.
[33] W.Li, G.D. Meitzner, R.W. Borry III, E. Iglesia, J. Catal. 191 (2000) 373-383.
[34] A.M. Rzehvskii, P. Choi, F.H. Ribeiro, R.J. Gulotty Jr. M.M. Olken, Catal. Lett. 73 (2001) 187-
191.
[35] D. Ma, Y. Shu, X. Han, X. Liu, Y. Xu, X. Bao, J. Phys. Chem. B 105 (2001) 1786-1793.
[36] C. Li, P.C. Stair, Catal. Today 33 (1997) 353-360.
[37] B.S. Liu, L. Jiang, H. Sun, C.T. Au, Appl. Surf. Sci. 253 (2007) 5092-5100. [38] S.Qu, B. Yang, Catal. Today 98 (2004) 639-645.
[39] F. Solymosi, A. Szöke, J. Cserényi, Catal. Lett. 39 (1996) 157-161.
Chapter 5
95
One-step synthesis of nano-crystalline MCM-22
Summary
Nano-crystalline MCM-22 zeolite was synthesized in a one-pot procedure by the use of an
organosilane (dimethyl-octadecyl-(3-trimethoxysilylpropyl)-ammonium chloride, TPOAC) in
the zeolite synthesis gel. This crystal growth inhibition procedure introduced mesopores in
the MCM-22 crystallites. The lower mechanical stability of the nano-crystalline MCM-22
zeolite compared with bulk MCM-22 can be countered to some extent by pillaring. The
increased external surface of the microporous zeolite domains resulted in increased
accessibility of the Brønsted acid sites, as followed from the better performance in liquid-
phase benzene alkylation with propylene as compared with bulk MCM-22. The increased
accessibility of the internal acid sites in Mo-loaded hierarchical MCM-22 was also evident
from the improved benzene selectivity during methane aromatization. Silylation of
hierarchical Mo/MCM-22 was detrimental for the catalytic performance in MDA. The nano-
crystalline MCM-22 has physico-chemical and catalytic properties intermediate between
those of MCM-22 and ITQ-2 with the benefit over ITQ-2 that it can be synthesized in a single
step.
This chapter has been accepted for publication in Microporous Mesoporous Mater. (2015)
Chapter 5
96
5.1 Introduction
Since the first report in the early 1990s [1], MCM-22 zeolite has attracted considerable
attention of the catalysis community. MCM-22 is a medium-pore zeolite, prepared at typical
Si/Al ratios between 10 and 20. The MWW pore topology endows promising catalytic
properties to MCM-22 zeolites. For instance, MCM-22 exhibits shape selectivity properties in
the trans-alkylation of toluene to p-xylene [2]. Its potential for the hydroisomerization of n-
alkanes into branched alkanes [3, 4] and the alkylation of benzene to cumene and
ethylbenzene [5, 6, 7] has also been demonstrated. MCM-22 has also been explored as an
acidic support in methane dehydroaromatization [8, 9].
As-synthesized MCM-22 consists of MWW layers that are kept together by hydrogen
bonds between the silanol groups that terminate the layers. Calcination of this precursor leads
to condensation of these silanol groups, resulting in a rigid crystal. The micropore system of
MCM-22 is made up from two separate two-dimensional channel systems [1]. The micropore
channels in the [001] direction consist of straight 10-membered rings with a diameter of
typically 5.6 Å [1, 10]. The second pore system is created when two cups located at the
surface of adjacent MWW layers are connected to form a supercage. These large ellipsoidal
cages are typically 7.1 Å in diameter and 18.2 Å in height [10]. The large cages are connected
with each other through 10-membered ring windows.
MCM-22 is employed at the industrial scale for the liquid-phase alkylation of benzene to
cumene and ethylbenzene [11-13]. The application of a solid catalyst such as MCM-22 for
benzene alkylation is important, because it can replace environmentally stressing AlCl3 [14].
Sastre et al. have shown that benzene does not enter the micropores of MCM-22 under liquid-
phase conditions, which has been related to the slightly twisted micropore entrances [15].
Accordingly, it has been assumed that alkylation takes mainly place over Brønsted acid sites
(BAS) located at the external surface of MCM-22. In agreement with this, increasing the
external surface by delamination improves the catalytic performance in alkylation reactions
[5, 16]. Delaminated MCM-22 (ITQ-2) consists of single MWW layers and contains a high
concentration of surface BAS, which are accessible to relatively bulky molecules [5]. The
preparation of ITQ-2 involves a number of delicate steps, including swelling of as-synthesized
MCM-22 by a surfactant, delamination by ultra-sonication and hydrolysis [17].
One of the other possible uses of MCM-22 zeolite is as the acidic component in
bifunctional catalysts for non-oxidative methane dehydroaromatization (MDA). This reaction
Chapter 5
97
converts methane into benzene and hydrogen. The most studied catalyst for this reaction is
Mo/ZSM-5 [18], although also other zeolite types - mainly medium pore structures - have
been tested. Several studies report that MCM-22 zeolite performs better than ZSM-5 [19-21].
MDA catalysts are comprised of an acidic zeolite and Mo-oxide. The Mo-oxide precursor
supported on the zeolite is rapidly converted into Mo-carbide under reaction conditions. These
Mo-carbide particles convert methane into a C2-intermediate, presumably ethylene, which is
further converted over BAS into benzene, toluene and naphthalene. The main challenge to
further develop the MDA technology is to overcome the rapid deactivation of the catalyst by
coke formation [22-26]. Decreasing the crystal size has been shown to lower the deactivation
rate for Mo/ZSM-5 and Mo/MCM-22 catalysts [27-29].
A large number of methods to synthesize hierarchically structured zeolites have been
described along the last decade. A number of reviews categorize and discuss these approaches
[30-32]. Broadly speaking, two strategies can be followed, namely bottom-up and top-down
approaches. In top-down approaches, mesopores are introduced after the zeolite has been
synthesized, and it is usually achieved by extracting either Al or Si atoms from the zeolite
framework. Common methods for demetallation include steaming [33-40], acid-leaching [30-
46] and base treatment [47, 48]. In bottom-up approaches, the mesopores are introduced in the
framework during the synthesis. This approach usually involves the use of a space-filling
molecule (mesoporogens). For example, mesoporous voids can be introduced by the simple
addition of carbon beads to the synthesis gel of MFI [49-51]. Choi et al. were the first to
report the introduction of intra-crystalline mesoporosity using an amphiphilic organosilane
surfactant molecule (dimethyl-octadecyl-(3-trimethoxysilylpropyl)-ammonium chloride,
TPOAC) [41]. Recently, Carvalho et al. have prepared nano-crystalline ZSM-12 by this
approach [42]. In contrast to MFI for which several direct routes for hierarchical structuring
have been explored, only one route has been reported for the preparation of hierarchical
MCM-22. This route involves the addition of carbon black pearls to the synthesis gel [29, 52].
In this study, we report about a one-pot synthesis approach of nano-crystalline MCM-22
by use of TPOAC as a crystal growth inhibitor. The addition of TPOAC impedes the
crystallite growth, in a similar manner as has been shown for nano-crystalline ZSM-12 [42].
The mechanical stability of this nano-crystalline MCM-22 zeolite can be improved by
pillaring. The resulting zeolite material shows improved performance in the liquid-phase
alkylation of benzene and the aromatization of methane compared to catalysts based on bulk
MCM-22.
Chapter 5
98
5.2 Experimental
5.2.1 Synthesis
A literature procedure was employed for the preparation of bulk MCM-22 [53]. This
recipe was modified to generate mesoporosity by adding an organosilane. In a typical
synthesis, an amount of 5.85 g silica gel (Sigma Aldrich) was mixed with 2.97 g of
hexamethylene imine (HMI) in a round-bottom flask. In a second round-bottom flask, 0.385 g
sodium hydroxide and 0.48 g sodium aluminate were dissolved in 30 ml water. The latter
solution was added to the first one and the mixture was stirred overnight at room temperature.
Then, dimethyl-octadecyl-(3-trimethoxysilylpropyl)-ammonium chloride (60 wt% in
methanol, ABCR) was added and the stirring was continued for 4 h. The resulting gel was
transferred to a Teflon-lined stainless-steel autoclave and kept at 423 K for 7 days under
rotation (60 rpm). The Si/Al ratio of the final synthesis gel was 16.3. The HMI/TPOAC ratio
in the synthesis gel was varied (HMI/TPOAC = 6, 12 and 120). The resulting solids are
denoted as MCM-22(x) with x being the HMI/TPOAC ratio. MCM-22 prepared without
TPOAC in an otherwise similar synthesis gel served as the reference. The organics were
removed from the as-synthesized materials using calcination in artificial air (20/80 (v/v)
O2/He) at 623 K for 6 h using a heating rate of 1 K/min.
Silica pillars were introduced in MCM-22 by impregnating 1 g of the non-calcined
MCM-22(12) with 4 g of tetrapropylammonium hydroxide solution (40 wt% TPOAH). The
suspension was stirred for 16 h at room temperature. To 1 g of the resulting zeolite/TPOAH
mixture, 5 g of tetraethylorthosilicate (TEOS) was added under vigorous stirring in Ar
atmosphere. This mixture was stirred at 351 K for 25 h. This procedure was followed by
addition of HCl until the pH was 1. The resulting mixture was stirred for 6 h at 313 K. The
solid was recovered by filtration and washed with copious amounts of water. The material was
dried overnight at 373 K and finally calcined. Calcination was done by heating in N2
atmosphere to 723 K followed by exposure to artificial air (20/80 (v/v) O2/He) at 823 K for 12
h. The pillared sample was denoted as MCM-22(12)-pillared.
Another reference sample was ITQ-2. This delaminated zeolite was prepared following
the procedure outlined by Corma and co-workers [54]. As-synthesized MCM-22 was used as
the starting material for the delamination procedure. To 1 g of the MCM-22 precursor 3.9 ml
of CTAB solution (29 wt%, CTAB) was added. After dispersion 1.2 g TPAOH (40 wt% in
water) was added. The final mixture was heated for 16 h at 353 K. After this, the mixture was
Chapter 5
99
ultrasonicated for 1 h. Finally, the pH was adjusted to 2 and stirred for 1 h and subsequently
filtered. To remove the organic constituents the solid was calcined in artificial air at 823 K for
6 h.
To obtain the proton form, the zeolites were ion-exchanged in a 1 M NH4NO3 solution
for 4 h followed by filtration. This procedure was repeated two times. After drying overnight
at 383 K, the solids were calcined at 723 K for 4 h in an artificial air flow.
For methane dehydroaromatization, molybdenum was loaded onto the MCM-22 zeolites. For
this purpose, the dried zeolite was impregnated with a solution of appropriate concentration
ammonium heptamolybdate tetrahydrate (AHM, Merck). After impregnation the samples
were dried for 1 h. The targeted Mo content was 4 wt%. The Mo-loaded zeolites were
calcined in artificial air after heating to 823 K at a rate of 1.5 K/min. The final temperature
and the dwell time at this temperature were varied. Zeolites modified with molybdenum are
denoted by the prefix “Mo/”.
A portion of the Mo-modified zeolites was silylated by a procedure described in the
literature [55]. For this purpose, 2 g of zeolite was dried at 373 K overnight and then
dispersed in 50 ml n-hexane. To this suspension, 0.3 ml tetraethylorthosilicate (TEOS, Merck)
was added under stirring and, subsequently, refluxed for 1 h. The amount of TEOS
corresponded to 0.4 wt% based on the amount of zeolite in the suspension. Thereafter, the
catalyst was filtered and dried at 373 K overnight. Finally, the zeolite was calcined in two
steps. The first step consisted of heating the sample at a rate of 2 K/min to 393 K followed by
an isothermal period of 2 h. The second step further increased the temperature to 773 K at a
rate of 0.2 K/min followed by an isothermal period for 4 h. The silylation treatment was
denoted by adding “Si” to the catalyst notation, e.g., Mo/MCM-22(12, Si).
5.2.2 Characterization
XRD patterns were recorded on a Bruker D4 Endeavor machine using Cu Kα radiation.
Diffraction patterns were measured in the 5° ≤ 2Ө ≤ 60° range using a step size of 0.1°.
Ar sorption isotherms were measured at 87 K on a Micromeritics ASAP2020 system in static
measurement mode. The samples were outgassed at 623 K for 8 h prior to the sorption
measurements. The Brunauer–Emmett–Teller (BET) equation was used to determine the
surface area (SBET) from the adsorption data in the p/p0 = 0.05–0.25 range. The mesopore
volume (Vmeso) and mesopore size distribution were determined by applying the Barrett–
Chapter 5
100
Joyner–Halenda (BJH) method to the adsorption branch of the isotherm. The micropore area
(Smicro) and micropore volume (Vmicro) were calculated by the NLDFT method (Ar at 87 K on
oxides as the model, assuming cylindrical pores, without regularization).
Infrared spectra were recorded in the 4000-400 cm-1
range using a Bruker Vertex 70v
apparatus. Samples were pressed into a self-supporting wafer with a density of about 10
mg/cm2. To remove adsorbed water the sample was evacuated for 2 h at 773 K. After
evacuation the sample was cooled to 323 K followed by recording of the background
spectrum. The BAS concentration was determined by measuring IR spectra of adsorbed
pyridine. Pyridine adsorption was carried out on the dehydrated zeolite wafer at 423 K. After
saturation was reached following exposure to gaseous pyridine, the sample was evacuated at
523 K for 2 h and a spectrum was recorded. The evacuation step was repeated at 623 K and
673 K. The acidity of the samples was determined by deconvolution of the weight normalized
spectra according standard procedures and the obtained values are expressed in arbitrary units.
Magic-angle spinning (MAS) nuclear magnetic resonance (NMR) measurements were
performed on a 11.7 T Bruker DMX500 NMR spectrometer operating at a frequency of 500
MHz for 1H measurements. The
27Al MAS NMR spectra were measured using a Bruker 2.5-
mm MAS probehead spinning at 20 kHz. The 1H and
29Si MAS NMR measurements were
carried out using a 4-mm MAS probehead at sample rotation rates of 12.5 kHz for 1H and 10
KHz for 29
Si NMR measurements, respectively.
Quantitative 1H NMR spectra were recorded with a Hahn-echo p1-τ1-p2-τ2-aq pulse
sequence with a 90o pulse with p1 = 5 μs and a 180º pulse with p2 = 10 μs. The interscan
delay was set at 120 s. Quantitative 29
Si NMR spectra were recorded using a high power
proton decoupling direct excitation (DE) pulse sequence with a 45 o pulse duration of 2.5 μs
and an interscan delay of 160 s. 1H-
29Si cross-polarization (CP) spectra were obtained using
an interscan delay of 3 s and a contact time of 3 ms. 27
Al NMR spectra were recorded with a
single pulse sequence with a 18o pulse duration of 1 μs and an interscan delay of 1 s.
1H and
29Si NMR shifts were referred to tetramethylsilane (TMS), while saturated Al(NO3)3 solution
was used for 27
Al NMR shift calibration. For 1H MAS NMR measurements, the zeolites were
first dehydrated at a temperature of 723 K at vacuum lower than 10-5
mbar for 6 h. The
dehydrated zeolites were placed into the 4 mm MAS NMR zirconia rotor under inert
conditions. The deconvolution of the NMR spectra was done using DMfit2011.
Transmission electron micrographs were obtained with a FEI Tecnai 20 transmission
electron microscope (TEM) at an electron acceleration voltage of 200 kV. Typically, a small
Chapter 5
101
amount of sample was suspended in ethanol, sonicated and dispersed over a Cu grid with a
holey carbon film. Scanning electron microscopy (SEM) was performed using a Philips
environmental FEIXL-30 ESEM FEG in high-vacuum mode at low voltage.
5.2.3 Catalytic activity measurements
The catalytic activity in the alkylation of benzene with propylene was evaluated in a
high-pressure stainless-steel reactor at 398 K and 3.5 MPa. For these measurements, the
zeolites were pelletized, crushed and sieved in a 0.25–0.42 mm fraction, and diluted with
silicon carbide to a total bed volume of 3.6 ml. The molar benzene/propylene ratio in the feed
was 3.5. The weight-hourly space velocity (WHSV) was 25 h-1
to the olefin in the feed.
Samples were analyzed as a function of time on stream by online gas chromatography (5%
phenyl–95% dimethylpolysiloxane column, length 30 m, internal diameter of 0.25 mm, 1 μm
thick stationary phase film).
The catalytic performance of Mo-loaded zeolites in methane dehydroaromatization was
measured in a fixed-bed reactor at 973 K, 0.1 MPa and a contact time of 16 g cat·h/mol CH4
(GHSV = 1500 h-1
). The catalyst weight was 0.5 g, and was diluted with silicon carbide to
achieve a bed volume of 2.8 cm3. We verified that the diluent was not catalytically active.
Prior to reaction, the catalysts were heated from room temperature to 973 K at a heating rate
of 10 K/min in a methane/nitrogen mixture (80 vol% methane). After reaching the reaction
temperature, the reactor was purged with N2 for 0.5 h. Finally, the feed was switched to
methane. The reactor outlet was analyzed by a two-channel online gas chromatograph. N2
used as internal standard, H2, CO, CO2 and CH4 were separated over HayeSep N (0.5 m),
HayeSep Q (1.5 m) and 13X molecular sieve (1.2 m length) columns (TCD). In the second
channel, hydrocarbons were separated over a pre-column (CP-Wax capillary column, 5.0 m
length and 0.32 mm inner diameter). The light hydrocarbon products were further separated in
a CP-Porabond Q (25 m length and 0.32 mm inner diameter) and detected by a FID. The
aromatics were detected by FID after separation over a CP-Wax column (1.0 m length and
0.32 mm inner diameter).
Chapter 5
102
5.3 Results and discussion
5.3.1 Catalyst preparation
Acid catalyst preparation and physico-chemical properties
The XRD patterns of the as-prepared MCM-22 zeolites are shown in Fig. 5.1. MCM-22,
MCM-22(12) and MCM-22(120) have the MWW structure [53], but the diffraction peaks are
less intense and broader for the materials prepared at lower HMI/TPOAC ratio. The material
prepared at a HMI/TPOAC ratio of 6 was almost completely X-ray amorphous. The ITQ-2
reference material was obtained by exposing as-synthesized MCM-22 (MCM-22(p)) to a
delamination step followed by calcination. The XRD pattern of ITQ-2 matches the one given
in the literature [56]. Table 1 shows that the Si/Al ratios of the MCM-22 materials prepared
with and without TPOAC are similar. The textural properties of the crystalline materials were
determined by Ar physisorption and the results are listed in Table 5.1. The micropore volume
of MCM-22 was 0.15 cm3/g, in good agreement with literature values [57]. The use of
TPOAC led to a small decrease in the micropore volume. The mesopore volume increased
with decreasing HMI/TPOAC ratio. The zeolite prepared at a HMI/TPOAC ratio of 12 had
the largest mesopore volume (Vmeso = 0.3 cm3/g). The higher surface area of ITQ-2 compared
with bulk MCM-22 (414 m2/g vs. 117 m
2/g) indicates that MCM-22(p) was delaminated to a
significant extent. Still, the surface area of our ITQ-2 is lower compared with some of the
values reported before for this material starting from laminar precursors prepared with higher
Si/Al ratios [58]. The lower surface area of our ITQ-2 material is likely due to the difficulty of
delaminating MCM-22(p) with relatively high Al content (Si/Al ratio ~15) [58].
Chapter 5
103
Fig. 5.1. Wide angle XRD reflection patters of calcined (a) MCM-22, (b) MCM-22(120) (c)
MCM-22(12), (d) MCM-22(6), (d) MCM-22(12)-pillared and ITQ-2.
10 20 30 40 50 60
Inte
nsit
y (
a.u
)
Angle (°)
Chapter 5
104
Table 5.1. Physico-chemical properties and elemental composition of the materials before and after applying external pressure.
Sample HMI/TPOAC Si/Al1
Si/Al2 Pressure
3
(106
N/m2)
Vmicro
(cm3/g)
Vmeso
(cm3/g)
Smicro
(m2/g)
Smeso
(m2/g)
SBET
(m2/g)
ΔVmeso
(%)
MCM-22 ∞ 16.3 19.2 0 0.15 0.08 332 54 449 -
∞ 16.3 19.2 16 0.10 0.09 191 54 308 +12
MCM-22(120) 120 16.3 - 0 0.14 0.18 263 95 370 -
MCM-22(12) 12 16.3 20.6 0 0.13 0.31 184 122 359 -
12 16.3 20.6 16 0.04 0.18 94 94 219 -40
MCM-22(12)-pillared 12 16.3 21.5 0 0.05 0.32 126 182 406 -
12 16.3 21.5 16 0.05 0.26 118 145 347 -16
ITQ-2 ∞ 16.3 23.7 0 0.14 0.26 294 160 708 - 1 Ratio of Si/Al in the synthesis gel;
2 Determined from ICP-AES elemental analysis;
3 Pressure was applied for 60 s.
Chapter 5
105
In general, the mechanical stability of hierarchical zeolites is lower than that of bulk zeolites
and exerting mechanical forces usually reduces the beneficial mesoporosity. Pillaring is an
established procedure to increase the mechanical stability of such hierarchical zeolites [59]. In
the present work, we applied this procedure to MCM-22(12) because of its favorable textural
properties. Comparison of the XRD patterns before and after pillaring (Fig. 5.1) shows that
the crystallinity was not affected. Ar physisorption data point to the substantial decrease of the
micropore volume due to pillaring, suggesting that part of the micropores were blocked due to
deposition of TEOS in the micropores. The textural properties were also determined after
pressing bulk MCM-22, MCM-22(12) and pillared MCM-22 (Fig. 5.2). Comparison of the Ar
sorption isotherms before and after pressing the samples shows that the decrease of the
mesopore volume upon mechanical stress was lower for MCM-22(12)-pillared than for
MCM-22(12). This is particularly evident from the decrease of the hysteresis loop. On
contrary, the isotherm of bulk MCM-22 is hardly affected by the pressing procedure. All this
is also evident from the textural data derived from the Ar physisorption isotherms listed in
Table 5.1. It is interesting to note that the mechanical stability test led to a decrease of the
micropore volume of MCM-22(12) and MCM-22, but not for the pillared zeolites.
Fig. 5.2. Ar physisorption isotherms of (square) MCM-22, (circle) MCM-22(12), (triangle)
MCM-22(12)-pillared and (diamond) ITQ-2. Isotherms with open symbols correspond to
samples measured after exposing to external mechanical force. The isotherms are presented in
a stacked fashion for clarity. The Y offsets are progressively increased with 100 cm3/g for
each subsequent sample.
0.0 0.2 0.4 0.6 0.8 1.0
Vo
lum
e a
dso
rbed
(cm
3/g
.ST
P)
Relative pressure
100 cm3/g
Chapter 5
106
The morphology of these zeolites was investigated by electron microscopy. Representative
SEM images are shown in Fig. 5.3. The primary crystals of MCM-22 show the well-known
platelet morphology of MWW zeolite arranged in larger spherical secondary particles. The
crystal size of these platelets is several nanometers. From SEM, the morphology of MCM-
22(12) and MCM-22(12)-pillared appears to be more open. The morphology of ITQ-2 is also
platelet-like, arranged in a more aligned fashion, likely due to the single layer morphology at
the nano-scale.
Fig. 5.3. SEM micrographs of (a) MCM-22, (b) MCM-22(12), (c) MCM-22(12)-pillared and
(d) ITQ-2.
Fig. 5.4 depicts representative TEM images of the same materials. The thickness of the
MWW crystallites for MCM-22(12) (Fig. 5.4b), which are in the range of 30-40 nm, are seen
to be similar to MCM-22 (Fig. 5.4a). More detailed inspection of the TEM images of MCM-
22 shows that the zeolite crystals are aligned forming a supra-crystallite structure of
approximately 100 nm. The crystallites of MCM-22(12) seem to be organized in a more
random fashion. The addition of TPOAC possibly causes the zeolite crystals to become
separated, which is likely due to the hydrophobic nature of TPOAC grafted at the zeolite
crystal surface. The random organization of the agglomerated crystallites results in interstitial
voids that are in part responsible for the mesoporosity of the material. Furthermore, some
crystallites show a decreased crystal size (inset Fig. 5.4b) pointing to the inhibiting effect of
a b
c
d
Chapter 5
107
TPOAC on crystal growth. The TEM images show that some of the MWW layers of MCM-
22(12) are separated due to the presence of TPOAC. In this way, additional interstitial voids
were created, which contribute to the mesopore volume. The morphology of MCM-22(12)-
pillared (Fig. 5.4c) is similar to that of MCM-22(12). The TEM image for ITQ-2 (Fig. 5.4d)
confirms that this material contains single MWW layers.
Fig. 5.4. TEM micrographs of (a) MCM-22, (b) MCM-22(12), (c) MCM-22(12)-pillared and
(d) ITQ-2.
29Si MAS-NMR spectroscopy was employed to distinguish between the terminal silanol
and framework silicate species. The amount of terminal silanol groups can be related to the
external surface area. First, 1H→
29Si cross-polarization (CP) measurements were conducted to
prove the presence of Q3 (Si(OSi)3OH and Si(1Al)) and Q2 sites (Si(OSi)2(OH)2), next to the
predominant Q4 sites (SiO4). The Q2 sites are characterized by a peak at -93 ppm, while Q4
species are identified by overlapping peaks in the range of -104 to -120 ppm [60]. Q3 species
including Si(OSi)3OH and Si(1Al) are identified by peaks at -98 ppm and -101 ppm,
respectively [60]. To quantify the various silicate species, the spectra were measured in high-
power decoupling (hp-dec) mode (Fig. 5). The results of deconvolution of these spectra are
given in Table 2. Hierarchical MCM-22 contains more Q3 species than bulk MCM-22 zeolite.
50 nm
20 nm
0.2 µm
50 nm
a b
c d
Chapter 5
108
As the Al bulk content is nearly the same for all MCM-22 zeolites, the higher Q3
concentration can be attributed to the higher silanol content and the larger external surface
area. The larger external surface area is due to the inhibited crystal growth upon addition of
TPOAC. It is consistent with the increasing fraction of Q2+Q3 species.
Chapter 5
109
Fig. 5.5. 29
Si MAS NMR spectra measured in hp-dec mode of (a) MCM-22, (b) MCM-22(12), (c) MCM-22(12)-pillared and (d) ITQ-2.
-80 -100 -120 -140
(ppm)
-80 -100 -120 -140
Inte
nsit
y (
a.u
.)
(ppm)
-80 -100 -120 -140
(ppm)
-80 -100 -120 -140
(ppm)
Chapter 5
110
Acidity characterization
The Al coordination in the MCM-22 zeolites was determined by 27
Al MAS NMR
spectroscopy. The spectra are shown in Fig. 5.6. The two peaks at 55 ppm (FAl1) and 48 ppm
(FAl2) in the spectrum of MCM-22 indicate the presence of two types of framework Al
species in line with the literature [60, 61]. The FAl1 species are located inside the micropores,
while the FAl2 species reside at the external surface and/or the large cavities [61]. Another
less pronounced feature at 0 ppm is due to extraframework aluminum (EFAl). The Al
speciation was quantified by deconvolution of the 27
Al MAS NMR spectra (Table 5.2). The
fraction of EFAl in MCM-22(12) is 30%, which is significantly higher than the EFAl content
(21%) for MCM-22. After pillaring, MCM-22(12)-pillared contains a similar amount of
EFAl. The addition of TPOAC led to a decrease in the FAl2 content for MCM-22(12) and
MCM-22(12)-pillared, which we attribute to the change of the amount of FAl species at the
external surface. The removal of surface Al is also evident when the as-synthesized MCM-22
was delaminated to obtain ITQ-2. This phenomenon has already been reported for the
synthesis of ITQ-2 [54, 58].
Fig. 5.6. 27
Al MAS NMR spectra of (a) MCM-22, (b) MCM-22(12), (c) MCM-22(12)-
pillared and (d) ITQ-2.
The 1H MAS NMR spectra help to characterize the various hydroxyl groups (Fig. 5.7). The
bands at 1.9 ppm and 2.6 ppm are assigned to terminal hydroxyls associated with silanol
100 50 0 -50
Inte
nsit
y (
a.u
.)
(ppm)
Chapter 5
111
groups and EFAL species, respectively [62]. The protons of the Brønsted acid sites (BAS)
give rise to the 4.0 ppm peak in the 1H NMR spectrum. Internal silanols are characterized by
the peak at 6.1 ppm. The content of the various hydroxyl species as determined from these
NMR spectra are listed in Table 5.2. Comparing the quantitative data reveals the BAS content
in MCM-22(12) to be lower than in MCM-22. After pillaring of MCM-22(12) the BAS
concentration increased, indicated by the increase of the corresponding NMR signal. Further
inspection of the NMR spectra suggests a decrease in EFAl content after pillarization of
MCM-22(12), which is confirmed by the quantitative data. These observations are
inconsistent with the findings from the 27
Al MAS NMR experiments. Seemingly, a fraction of
the EFAl is invisible to 27
Al NMR [59, 63-66]. The larger BAS concentration of MCM-
22(12)-pillared compared with MCM-22(12) is most likely due to leaching of the EFAl
present upon acid treatment. In this way, part of the EFAl that was compensating the negative
framework charge is removed, thus increasing the amount of BAS. The removal of FAl
during delamination explains the lower BAS content of ITQ-2 as compared with the parent
MCM-22. These findings are in agreement with the trends observed by 27
Al MAS NMR
spectroscopy and literature [58].
The acidic properties of the zeolites were characterized in more detail using FTIR
spectroscopy. The spectra in Fig. 5.8 contain two main absorption bands at 3612 cm-1
(BAS)
and 3749 cm-1
(silanol groups). MCM-22(12) and MCM-22(12)-pillared contain more silanol
groups than MCM-22, while the reverse holds for the BAS content. The silanol and BAS
contents of ITQ-2 were comparable with MCM-22(12). The acidity of MCM-22(12)-pillared
was larger than that of ITQ-2 and MCM-22(12), albeit lower than that of MCM-22. These
findings are consistent with the results from 1H MAS NMR results. Acid site quantification
by FTIR spectroscopy after exposure of pyridine to the parent zeolites (Table 5.3) trends with
the observations from the FTIR spectra and the 1H MAS NMR results.
Chapter 5
112
Fig. 5.7. 1H MAS NMR spectra of (a) MCM-22, (b) MCM-22(12), (c) MCM-22(12)-pillared and (d) ITQ-2.
Table 5.2. Relative contributions of Si, Al and H species in the MWW zeolites as determined by deconvolution of 29
Si, 27
Al and 1H MAS
NMR spectra.
Nucleus 29
Si NMR 27
Al NMR 1H NMR
Sample Q4
(%)
Q3
(%)
Q2
(%)
FAl1
(%)
FAl2
(%)
EFAl
(%)
SiOHext
(a.u)
SiOHint
(a.u)
Si(OH)Al
(a.u)
EFAlOH
(a.u)
MCM-22 92 7 1 67 12 21 25 22 44 7
MCM-22(12) 84 15 1 62 8 30 49 13 26 17
MCM-22(12)-pillared 82 16 2 61 7 32 63 29 39 10
ITQ-2 79 19 2 73 6 21 52 19 36 14
0 2 4 6 8 10
Inte
nsit
y (
a.u
.)
(ppm)
0 2 4 6 8 10
(ppm)
0 2 4 6 8 10
(ppm)
0 2 4 6 8 10
(ppm)
Chapter 5
113
Fig. 5.8. FTIR spectra of (a) MCM-22, (b) MCM-22(12), (c) MCM-22(12)-pillared and (d)
ITQ-2 (samples dehydrated in vacuo at 723 K).
Catalytic activity measurements: benzene alkylation with propene
Fig. 5.9 depicts the time on stream behavior (Fig. 5.9a) and the selectivity (Fig. 5.9b) of
the various MCM-22 zeolites and ITQ-2 in the liquid-phase alkylation of benzene with
propylene. All catalysts initially show high activity, but deactivate with time on stream. ITQ-2
showed the highest stability in this reaction. The lower rate of deactivation of ITQ-2 as
compared with MCM-22 is in line with previous reports [5]. The nano-crystalline MCM-
22(12) exhibited better stability than the other MCM-22 materials, although the rate of
deactivation is higher as compared with ITQ-2. The higher catalytic stability of ITQ-2
compared with bulk MCM-22 has been explained by the larger amount of BAS accessible to
benzene at the external zeolite surface [5, 67, 68]. The deactivation of MCM-22 is usually
attributed to carbonaceous deposits that cover these BAS at the external surface [69-71]. We
attribute the lower rate of deactivation of MCM-22(12) compared with MCM-22 to the
decreased crystallite size; the decreased crystallite size implies a higher external surface area
and improved accessibility of benzene to the BAS at the external surface. Moreover, the
decrease of the diffusion path length will allow desorption of undesired products (oligomers,
multiple alkylated aromatics) before they are converted into bulkier coke species. The
stability towards deactivation of pillared MCM-22(12)-pillared was the lowest among the
MCM-22 zeolites. This may be correlated with the higher acidity of this sample. Such
3800 3600 3400 3200
Ab
so
rban
ce (
a.u
.)
Wavenumber (cm-1)
Chapter 5
114
deactivation has also been reported by the group of Corma [69], who showed that the
selective deactivation of external BAS with 2,6-di-tert-butyl pyridine (DTPBPy) led to rapid
decrease of the alkylation activity with time on stream. The explanation of low external BAS
content of MCM-22(12)-pillared is supported by the lower selectivity to diisopropylbenzene
(Fig. 5.9b) compared with MCM-22, MCM-22(12) and ITQ-2. The selectivity to multi-
alkylated benzenes products is higher for MWW zeolites with a higher external surface area
[67, 71]. The lower external acidity of MCM-22(12)-pillared due to silica deposition explains
the lower selectivity to diisopropylbenzene.
Fig. 5.9. Catalytic performance in benzene alkylation (T = 398 K; 3.5 MPa; 1/3.5 (mol/mol)
Benzene/propylene): (a) propylene conversion and (b) cumene (closed symbols) and
diisopropylbenzene (open symbols) selectivities of Beta (■), MCM-22 (●), MCM-22(12) (▲)
and MCM-22(12)-pillared (♦) and ITQ-2 (▼).
5.3.2 Preparation, characterization and catalytic testing of bifunctional Mo/zeolites.
The various MCM-22 zeolites were also used as acid supports for the preparation of Mo-
modified zeolite catalysts for the MDA reaction. The Mo was introduced by incipient wetness
impregnation followed by calcination. The targeted Mo content was 4 wt%.
Characterization of bifunctional Mo/zeolites catalysts
The acidity of the Mo-modified zeolites was determined by pyridine adsorbed FTIR (see
Table 5.3). Modification of the parent materials with Mo led to a decrease in acidity for all
0 60 120 180 240 300
0
20
40
60
80
100S
ele
cti
vit
y (
wt%
)
Time on stream (min)
0 60 120 180 240 300
0
10
20
30
40
50
60
70
80
90
100
Pro
pyle
ne c
on
vers
ion
(%
)
Time on stream (min)
Chapter 5
115
samples. The decrease in acidity is due to exchange of the protons associated to the BAS with
Mo-oxo species. Silylation led to further decrease of the acidity in agreement with an earlier
report for Mo/ZSM-5 [71]. The acidity decrease in this case can be attributed to the improved
spreading of Mo over the zeolite and to the deactivation of the external acid sites by the
silylation treatment.
Catalytic activity measurements: methane dehydroaromatization
The time on stream behavior of the Mo/zeolites is shown in Fig. 5.10. While the methane
conversion decreased with time on stream (Fig. 5.10a), the benzene selectivity exhibits an
optimum around 4 h. The initially low benzene selectivity relates to the conversion of Mo-
oxide in the precursor material into Mo-carbide. The highest benzene selectivity (55 wt%)
was observed for Mo/MCM-22(12) after 4 h on stream. The benzene selectivities for
Mo/MCM-22 and Mo/MCM-22(12)-pillared were lower (~40 wt%). Mo/ITQ-2 showed the
lowest benzene selectivity (~30 wt%). The benzene selectivity inversely correlates with the
coke selectivity. Several reports have discussed the formation of coke at the external surface
BAS [72]. Deactivation of these external surface BAS by the addition of a small amount of
silica has been shown to improve benzene selectivity [71]. Therefore, we evaluated to what
extent the performance of these materials can be improved by silylation with TEOS. The
methane conversion as a function of time on stream did not change upon silylation. Silylation
also did not affect the selectivities (Figs. 5.11b and 5.11c) for Mo/MCM-22(12), Mo/MCM-
22(12)-pillared and Mo/ITQ-2 upon silylation. However, the benzene selectivity of
Mo/MCM-22 strongly improved by silylation. We surmise that the less pronounced effect of
silylation on nano-crystalline MCM-22 and ITQ-2 is due to their much higher silanol content.
We also evaluated the usefulness of an earlier developed reaction/regeneration cycle
procedure for the MDA reaction [78]. A typical reaction/regeneration cycle consisted of
reaction for 1.5 h at 973 K, followed by an oxidation step in artificial air at 773 K. Cooling
was done in inert atmosphere but the reaction temperature was recovered by heating in the
feed mixture. The results are shown in Fig. 5.12 (each data point represents catalytic
performance 70 min after the regeneration cycle). These data point out the increased stability
and benzene selectivity of silylated Mo/MCM-22, consistent with the data in Fig. 5.10 and
5.11. Both silylated and non-silylated Mo/MCM-22 displayed increased benzene selectivity
with each consecutive regeneration cycle. The methane conversion increased during the first 3
reaction/regeneration cycles. Thereafter, a small decrease in the methane conversion can be
Chapter 5
116
noted. Acidity characterization (Table 5.3) on the calcined Mo/MCM-22 and Mo/MCM-
22(Si) catalysts after exposure of 12 consecutive reaction cycles reveals a substantial loss in
acidity (~-60%) compared to the fresh catalyst. The reason for this is not clear, but possibly
the loss of the integrity of the zeolite structure, ineffective removal of refractory coke or
agglomeration of the Mo-phase during the oxidative regeneration step may provide an
explanation [79].
Chapter 5
117
Fig. 5.10. Catalytic performance in methane dehydroaromatization (T = 973 K; 0.1 MPa; 5/95 (v/v) N2/CH4): (a) methane conversion, (b)
benzene (closed symbols) and naphthalene (open symbols) selectivities and (c) coke (closed symbols) and olefin (open symbols) selectivities
of MCM-22 (■), MCM-22(12) (●), MCM-22(12)-pillared (▲) and ITQ-2 (♦).
0 2 4 6 8
0
5
10
15
20
25
CH
4 c
on
vers
ion
(%
)
Time on stream (h)
0 2 4 6 8
0
20
40
60
80
100
Sele
cti
vit
y (
wt%
)Time on stream (h)
0 2 4 6 8
0
20
40
60
80
100
Sele
cti
vit
y (
wt%
)
Time on stream (h)
Chapter 5
118
Fig. 5.11. Catalytic performance in methane dehydroaromatization (T = 973 K; 0.1 MPa; 5/95 (v/v) N2/CH4): (a) methane conversion, (b)
benzene (closed symbols) and naphthalene (open symbols) selectivities and (c) coke (closed symbols) and olefin (open symbols) selectivities
of silylated MCM-22 (■), MCM-22(12) (●), MCM-22(12)-pillared (▲) and ITQ-2 (♦).
0 2 4 6 8
0
5
10
15
20
25
CH
4 c
on
vers
ion
(%
)
Time on stream (h)
0 2 4 6 8
0
20
40
60
80
100
Sele
cti
vit
y (
wt%
)Time on stream (h)
0 2 4 6 8
0
20
40
60
80
100
Sele
cti
vit
y (
wt%
)
Time on stream (h)
Chapter 5
119
Fig. 5.12. Catalytic performance in methane dehydroaromatization with intermediate regeneration ((T = 973 K; 0.1 MPa; 5/95 (v/v) N2/CH4);
each point represents a time on stream of 70 min): (a) methane conversion, (b) benzene (square) and naphthalene (circle) selectivities and (c)
coke (square) and olefin (circle) selectivities of non-silylated (closed symbols) and silylated (open symbols) MCM-22.
0 2 4 6 8 10 12
0
2
4
6
8
10
12
Meth
an
e c
on
vers
ion
(%
)
Regeneration cycle
0 2 4 6 8 10 12
0
20
40
60
80
100
Sele
cti
vit
y (
wt%
)Regeneration cycle
0 2 4 6 8 10 12
0
20
40
60
80
100
Sele
cti
vit
y (
wt%
)
Regeneration cycle
Chapter 5
120
5.4 Conclusions
A one-pot synthesis procedure for the preparation of nano-crystalline MCM-22 was
developed. It involves the addition of an amphiphilic organosilane to the synthesis gel. The
total Brønsted acidity of this nano-crystalline MCM-22 zeolite is lower than that of bulk
MCM-22. Nevertheless, nano-crystalline MCM-22 shows higher catalytic performance in the
liquid-phase alkylation of benzene with propylene due to the increased accessibility of the
Brønsted acid sites. The low mechanical stability of the hierarchical material was improved
by pillaring as followed from textural analysis before and after exerting mechanical forces by
pelletizing the powdered zeolite. The addition of silica during the pillaring with TEOS led to a
substantial decrease of the acidity and the catalytic alkylation performance. The shorter
diffusion pathways through the MCM-22 with reduced crystal size also led to an improved
benzene selectivity in the methane aromatization reaction. External surface modification of
the hierarchical Mo/MCM-22 catalyst following a silylation treatment was detrimental to the
catalytic performance of this catalyst in MDA.
5.5 References
[1] M.E. Leonowicz, J.A. Lawton, S.L. Lawton, M.K. Rubin, Science 264 (1994) 1910-1913.
[2] W.O. Haag, R.M. Lago, P.B. Weisz, Faraday Discuss. 72 (1981) 317-330.
[3] A. Corma, C. Corell, F. Llopis, A. Martinez, J. Pérez-Pariente, Appl. Catal., A 115 (1994) 121-
134.
[4] R. Ravishankar, S. Sivasanker, Appl. Catal., A 142 (1996) 47-59.
[5] P.J. Van den Brink, A. Corma, E.J. Creyghton, V. Fornes, V. Martines-Soria, WO patent
01/21562, 2001.
[6] C. Flego, G. Pazzuconi, E. Bencini C. Perego, Stud. Surf. Sci. Catal. 126 (1999) 461-464.
[7] C. Perego, S. Amarilli, R. Millini, G. Bellussi, G. Girotti, G. Terzoni, Microporous Mesoporous
Mater. 6 (1996) 395-404.
[8] Y. Shu, D. Ma, L. Xu, Y. Xu, X. Bao, Catal. Lett. 70 (2000) 67-73.
[9] N. Chu, J. Yang, J. Wang, S. Yu, J. Lu, Y. Zhang, D. Yin, Catal. Comm. 11 (2010) 513-517.
[10] S.L. Lawton, M.E. Leonowicz, R.D. Partridge, P. Chu, M.K. Rubin, Microporous Mesoporous.
Mater. 23 (1998) 109-117.
[11] C.T. Kresge, Q.N. Le, W.J. Roth, R.T. Thompson, U.S. Patent 5,259,565, 1993.
[12] P. Chu, M.E. Landis, Q.N. Le, U.S. Patent 5,334,795, 1994.
[13] D.L. Holtermann, R. Innes, U.S. Patent 5,149,894, 1992.
[14] M.F. Bentham, G.J. Gajda, R.H. Jensen, H.A. Zinnen, Erdöl Erdgas Kohle 113 (1997) 84-88.
[15] G. Sastre, C. Richard, A. Catlow, A Corma, J. Phys. Chem. B 103 (1999) 5187-5196.
[16] A. Corma, V. Fornes, S.B. Pergher, Th.L.M. Maesen, J.G. Buglass, Nature 396 (1998) 353-356.
[17] A. Corma, V. Fornés, J.M. Guil, S. Pergher, Th. L. M. Maesen, J.G Buglass, Microporous
Mesoporous Mater. 38 (2000) 301-309.
[18] L. Wang, L. Tao, M. Xie, G. Xu, Catal. Lett. 21 (1993) 35-41.
[19] Y. Shu, D. Ma, L. Xu, Y. Xu, X. Bao, Catal. Lett. 70 (2000) 67-73.
[20] V.T.T. Ha, L.V. Tiep, P. Meriaudeau, C. Naccache, J. Mol. Catal., A 181 (2002) 283-290.
Chapter 5
121
[21] M.T. Portilla, F.J. Llopis, C. Martínez, S. Valencia, A. Corma, Appl. Catal., A 393 (2011)
257-268.
[22] L. Chen, L. Lin, Z.S. Xu, X.S. Li, T. Zhang, J. Catal. 157 (1995) 190–200.
[23] S. Liu, L. Wang, R. Ohnishi, M. Ichikawa, Kinet. Catal. 41 (2000) 132–144.
[24] B.M. Weckhuysen, M.P. Rosynek, J.H. Lunford, Catal. Lett. 52 (1998) 31–36.
[25] D. Ma, D. Wang, L. Su, Y. Shu, Y. Xu, X. Bao, J. Catal. 208 (2002) 260–269.
[26] C. Descorme, P. Gelin, C. Lecuyer, A. Primet, Appl. Catal., B 13 (1997) 185–195.
[27] L. Su, L. Liu, J. Zhuang, H. Wang, Y. Li, W. Shen, Y. Xu, X. Bao, Catal. Lett. 91 (2003) 155-
166.
[28] A. Martinez, E. Peris, M. Derewinski, A. Burkat-Dulak, Catal. Today 169 (2011) 75-84.
[29] N.B. Chu, J.Q. Wang, Y. Zhang, J.H. Yang, J.M. Lu, D.H. Yin, Chem. Mater. 22 (2010) 2757-
2763.
[30] S. van Donk, A. H. Janssen, J. H. Bitter and K. P. de Jong, Catal. Rev. Sci. Eng., 45 (2003)
297–319.
[31] J. Pérez-Ramírez, C.H. Christensen, K. Egeblad, C.H. Christensen, J.C. Groen, Chem. Soc. Rev.
37 (2008) 2530-2542.
[32] D.P. Serrano, J.M. Escola, P. Pizarro, Chem. Soc. Rev. 42 (2013) 4004-4035.
[33] S. Morin, P. Ayrault, N.S. Gnep, M. Guisnet, Appl. Catal., A 166 (1998) 281-292.
[34] J. Lynch, F. Raatz, C. Delande, Stud. Surf. Sci. Catal. 39 (1987) 547-557.
[35] J. Lynch, F. Raatz, P. Dufresne, Zeolites 7 (1987) 333-340.
[36] V. Patzelova, N.I. Jaeger, Zeolites 7 (1987) 240-242.
[37] A. Zukal, V. Patzelova, U. Lohse, Zeolites 6 (1986) 133-136.
[38] Y. Sasaki, T. Suszuki, Y. Takamura, A. Saji, H. Saka, J. Catal. 178 (1998) 94-100.
[39] C. Choi-Feng, J.B. Hall, B.J. Huggins, R.A. Beyerlein, J. Catal. 140 (1993) 395-405.
[40] R.A. Beyerlein, C. Choi-Feng, J.B. Hall, B.J. Huggins, G.J. Ray, Top. Catal. 4 (1997) 27-42.
[41] M. Choi, H. S. Cho, R. Srivastava, C. Venkatesan, D.-H. Choi, R. Ryoo, Nat. Mater., 5 (2006)
718–723.
[42] K.T.G. Carvalho, E.A. Urquieta-Gonzalez., Catal. Today 243 (2015) 92-102.
[43] G.R. Meima, Cattech 2 (1998) 5-12.
[44] P. Hudec, J. Novansky, S. Silhar, T.N. Trung, M. Zubek, J. Madar, J. Ads. Sci. Tech. 3 (1986)
159-166.
[45] B.L. Meyers, T.H. Fleisch, G.J. Ray, J.T. Miller, J.B. Hall, J. Catal. 110 (1988) 82-95.
[46] M. Tromp, J.A. van Bokhoven, M.T. Garriga Oostenbrink, J.H. Bitter, K.P. de Jong, D.C.
Koningsberger, J. Catal. 190 (2000) 209-214.
[47] J.C. Groen, W. Zhu, S. Brouwer, S.J. Huynink, F. Kapteijn, J.A. Moulijn, J. Pérez-Ramírez, J.
Am. Chem. Soc. 129 (2007) 355–360.
[48] J. C. Groen, L. A. A. Peffer, J. A. Moulijn and J. Pérez-Ramírez, J. Mater. Chem. 16 (2006)
2121-2131.
[49] C. J. H. Jacobsen, C. Madsen, J. Houzvicka, I. Schmidt and A. Carlsson, J. Am. Chem. Soc. 122
(2000) 7116–7117.
[50] K. Zhu, K. Egeblad and C. H. Christensen, Eur. J. Inorg. Chem. 25 (2007) 3955–3960
[51] A.H. Janssen, I. Schmidt, C.J.H. Jacobsen, A.J. Koster, K.P. De Jong, Microporous Mesoporous
Mater. 65 (2003) 59-75.
[52] J.Yang, J. Chu, J. Wang, D. Yin, J. Lu, Y. Zhang, Chin. J. Catal. 35 (2014) 49-57.
[53] I. Güray, J. Warzywoda, N. Baç, A. Sacco, Microporous Mesoporous Mater. 31 (1999) 241–
251.
[54] A. Corma, U. Diaz, V. Fornés, J.M. Guil, J. Martínez-Triguero, E.J. Creyghton, J. Catal. 191
(2000) 218-224.
[55] S. Inagaki, K. Kamino, E. Kikuchi, M. Matsukata, App. Catal., A 318 (2007) 22-27.
[56] M. Cheng, D. Tan, X. Bao, Chem. Comm. (2000) 1713-1714.
[57] A. Corma, C. Corell, J. Peréz-Pariente, J.M. Guil, R. Guil-Lopéz, S. Nicolopoulos, J. Gonzalez
Calbet, M. Vallet-Regi, Zeolites 16 (1996) 7-14.
[58] A. Martínez, E. Peris, G. Sastre, Catal. Today 107-108 (2005) 676-684.
Chapter 5
122
[59] Y.J. He, G.S. Nivarthy, F. Eder, K. Seshan, J.A. Lercher, Microporous Mesoporous Mater. 25
(1998) 207-224.
[60] W. Kolodziejski, C. Zocovich-Wilson, C. Corell, J. Pérez-Pariente, A. Corma, J. Phys. Chem.
99 (1995) 7002-7008.
[61] P. Mériaudeau, A. Tuel, T.T.H. Vu, Catal. Lett. 61 (1999) 89-92.
[62] M. Hunger, S. Ernst, S. Steuernagel, J. Weitkamp, Microporous Mater. 6 (1996) 349-353.
[63] E. Brunner, H. Ernst, D. Freude, T. Fröhlich, M. Hunger, H. Pfeifer, J. Catal. 127 (1991) 34-41.
[64] S.M. Campbell, D.M. Bibby, J.M. Coddington, R.F. Howe, R.H. Meinhold, J. Catal. 161 (1996)
338-349.
[65] A. Maijanen, E.G. Derouane, J.B. Nagy, Appl. Surf. Sci. 75 (1994) 204-212.
[66] C.A. Fyfe, J.L. Bretherton, L.Y. Lam, J. Am. Chem. Soc. 123 (2001) 5285-5291.
[67] P. J. Van den Brink, A. Corma, E. J. Creyghton, V. Fornes, V. Martines, Soria, WO 01/21562,
2001.
[68] A. Corma, V. Martínez-Soria, E. Schnoeveld, J. Catal. 192 (2000) 163–173.
[69] J.C. Cheng, T. F. Degnan, J.S. Beck, Y.Y. Huang, M. Kalyanaraman, J.A. Kowalasky, C.A.
Loehr and D.N. Mazzone, Stud. Surf. Sci. Catal. 121 (1999) 53-60.
[70] C. Perego, P. Ingallina, Green Chem. 6 (2004) 274-2794.
[71] C.H.L. Tempelman, V.O. de Rodrigues, E.R.H van Eck, P.C.M.M. Magusin, E.J.M. Hensen,
Microporous Mesoporous Mater. 259 (2015) 259-273.
[72] X. Dong, Y. Song, W. Lin, Catal. Lett. 8 (2007) 539-542.
[73] M.T. Portilla, F.J. Llopis, C. Martinez, Catal. Sci. Technol. (2015), DOI: 10.1039/C5CY00356C
[74] H. Ma, R. Kojima, R. Ohnishi, M. Ichikawa, Appl. Catal., A 275 (2004) 183–187.
Chapter 6
123
Effect of zeolite crystalline domain size on the methane
aromatization performance of Mo/HZSM-5
Summary
The influence of the micropore domain size on the catalytic performance of Mo/HZSM-5
catalysts in the methane dehydroaromatization reaction is evaluated. For this purpose,
zeolites with varying micropore domain sizes were prepared using different hierarchical
structuring approaches. These include extraction of silica and the use of mesoporogens. The
particle size of bulk ZSM-5 zeolite was also compared by using bulk ZSM-5 zeolites with
average crystal sizes of 10 µm and 1 µm. In this way, the micropore domain size of the zeolite
supports was varied between several nm’s and 10 µm. The structure and texture of the
materials was characterized by XRD, Ar physisorption and electron microscopy. After Mo
introduction, the catalytic performance of these Mo/ZSM-5 zeolites was determined in the
dehydroaromatization of methane was at 973 K. Catalysts derived from bulk ZSM-5 crystals
are prone to rapid deactivation due to blocking of the micropores with carbonaceous
deposits. Larger crystals deactivate more rapidly. Hierarchical structuring was beneficial in
decreasing the rate of catalyst deactivation and improved the benzene formation rate.
10 μm 1 nm 1 μm 100 nm 10 nm
20 nm 30 nm 50 nm
9.66 μm
2 μm
ZSM-5 crystalline micropore domain size
Chapter 6
124
6.1 Introduction
The use of methane, the main constituent of natural gas, as a feedstock for the production
of fuels and chemicals is considered an interesting alternative to petroleum as carbon source.
A promising route is the valorization of methane in the methane dehydroaromatization
(MDA) reaction. In this process methane is converted to benzene and hydrogen. The most
studied catalyst for this reaction is the bifunctional Mo/ZSM-5 catalyst. Under reaction
conditions, the Mo-oxide supported on the zeolite is first rapidly converted into Mo-carbide
species [1]. The Mo-carbide species facilitate oligomerization of methane into olefins [2]. The
produced olefins, presumably ethylene, are then aromatized to benzene over the Brønsted acid
sites (BAS) in the zeolite [3]. The main challenge for industrial application of the MDA
process is rapid catalyst deactivation. As discussed in Chapter 4, the formation of a carbon
layer at the external surface is considered the main reason for catalyst deactivation. The
formation of a carbon layer leads to micropore blockage thereby making the BAS inside the
micropores inaccessible.
In Chapter 2, it was reported that the rate of deactivation in MDA was lower for
hierarchical Mo/HZSM-5 with mesopores than for bulk Mo/HZSM-5. It is speculated that the
presence of mesopores and the corresponding decrease of the micropore domain size of the
MFI zeolite slows down the negative effect of micropore blocking with carbonaceous
deposits. Carbon deposition is one of the major causes of catalyst deactivation in the MDA
reaction. The benefit of decreasing the micropore domain size of zeolites for the catalytic
performance has been demonstrated before. Examples include the prolonged life of
hierarchical Fe/ZSM-5 for the benzene-to-phenol [4] and SSZ-13 for the methanol-to-olefins
[5] reactions. Similar to the MDA reaction, catalyst deactivation in these two reactions can be
largely attributed to coke deposition in the micropore space. The improved lifetime is due to
the more efficient use of the zeolite crystal.
In recent years, the development of novel methods for the preparation of hierarchical
structured zeolites has gained momentum, resulting in a variety of synthesis procedures to
decrease the micropore domain size. Several papers have reviewed the plethora of preparation
methods now available to synthesize such hierarchial zeolites [6-8]. Two synthesis strategies
can be distinguished, namely bottom-up and the top-down approaches. Top-down approaches
use post-synthesis treatment of the zeolite to extract either Al or Si atoms from the zeolite
framework. Often used methods for metal extraction are steaming [9-16], acid-leaching [14-
Chapter 6
125
20] and base treatment [21, 22]. Bottom-up approaches make use of void-filling mesopore
templates or mesoporogen. The early bottom-up routes [23-25] involved addition of space-
filling hard templates such as carbon beads to the synthesis gel of MFI, creating mesoporous
voids. A problem of this approach which is shared with some of the top-down approaches is
that the micropore domains remain relatively large. Recently, Ryoo et al. [26] showed the
possibility to prepare MFI with highly interconnected micropores and mesopores by using an
organosilane surfactant molecule. An alternative approach [27] to prepare MFI with highly
interconnected mesopores is to create a matrix of agglomerated protozeolite crystals. In this
way, the interstitial voids formed between the agglomerated nano-crystallites form the
mesopores. Ryoo et al. also reported the synthesis of MFI nanosheets [28]. The zeolite in this
case crystallizes in the form of sheets with crystal dimension of only a few nm in the direction
of the straight channels.
In this chapter we investigate the role of the micropore domain size of the zeolite support
on the performance of the Mo/HZSM-5 catalyst in the MDA reaction. To this end, we
prepared various mesoporous ZSM-5 zeolites. The size of the microporous domains vary
from several nm to 10 µm. After modifying the zeolites with Mo, the catalysts were evaluated
for their performance in the MDA reaction. The differences in performance are discussed in
terms of morphology and texture.
6.2 Experimental methods
6.2.1. Synthesis of zeolites
Reference ZSM-5
NH4ZSM-5 with Si/Al = 19.4 was obtained from AkzoNobel Catalysts. The parent zeolite
was converted to the proton form by calcination at 723 K using a heating rate of 1 K/min.
This sample is denoted as HZSM-5-ref.
Large ZSM-5 crystals
A basic solution was prepared by dissolving 11.268 g NaOH and 0.876 g KOH in 585.9 ml
water. Then, 266.1 ml Ludox AS was added dropwise under vigorous stirring followed by
further stirring for 1 h. To this suspension, first 100.16 g TPABr was added followed by
addition of 134 ml of an aqueous solution containing 5.13 g NaAlO2. Finally, a solution of
26.9 g ammonium carbonate dissolved in 134 ml water was added to the synthesis gel. The
Chapter 6
126
suspension was aged overnight, transferred to a Teflon-lined autoclave and heated for 10 days
at 443 K under stirring. The organic constituents were removed by calcination. For this, the
material was heated in artificial air at a rate of 1 K/min to 823 K followed by an isothermal
period of 6 h. The final calcined zeolite is denoted as HZSM-5-large.
Desilicated ZSM-5
For desilication, an optimized method described earlier in Chapter 2 was followed. In a
Teflon beaker, 2 g NH4ZSM-5 (AkzoNobel) with a Si/Al of 30 was suspended in 50 ml of an
aqueous NaOH solution (0.2 M). The suspension was stirred vigorously at 338 K for 0.5 h.
The hot suspension was filtered off and washed with copious amounts of water. The
desilicated ZSM-5 zeolite is denoted as HZSM-5-des.
Organosilane templated mesoporous ZSM-5
Tetrapropyloctadecyl ammonium (trimethoxy) silane (TPOAC, 60 wt% in MeOH, ABCR)
was used as the mesoporogen to synthesize hierarchical ZSM-5. A solution (A) was prepared
by dissolving 0.21 g sodium aluminate (NaAlO2), 4.2 g tetrapropyl ammonium bromide
(TPABr) and 1.2 g NaOH in 202.5 ml water. A second solution (B) was prepared by mixing
12.855 g TEOS with 2.07 g TPOAC solution in methanol. Mixture B was added dropwise to
solution A. After 2 h of aging, the resulting gel was transferred into a 125 ml Teflon-lined
autoclave. The autoclave was heated for 5 days at 423 K under rotation. Then, the suspension
was filtered and the recovered solid was washed with copious amounts of water. The sample
was finally calcined at 823 K (HZSM-5-org).
Small ZSM-5 crystals
A solution A was prepared by adding 13.3 ml TEOS to 3.9 ml tetrapropyl ammonium
hydroxide (TPAOH, 40 wt% in water). A second solution B was prepared by dissolving 0.245
g NaAlO2 in 23.49 ml water. After dissolving, solution B was added to solution A. The
resulting mixture was refluxed for 20 h at 363 K under stirring. To this mixture 2.07 g of
phenylammonium trimethoxy silane (PHAPTMS, ABCR) was added and was kept refluxing
for another 6 h at 363 K. After this period, the resulting gel was transferred to a Teflon lined
autoclave and heated at 443 K for 5 days. The solid was recovered by filtration followed by
washing with copious amounts of water and then calcined at 823 K at a rate of 1 K/min in
artificial air for 6 h (HZSM-5-nano).
Chapter 6
127
Nanosheet ZSM-5
A template solution A was prepared by dissolving 1.4 g of C22-6-3 template (bromide form) in
7.8 ml of water at 333 K. To solution A, 0.21 g of NaOH was added followed by stirring for
another 4 h at 323 K. A second solution B was prepared by mixing 0.12 g Al(OH)3 and 5.54
ml TEOS for 1 h in 10 ml water. After cooling solution A to room temperature, solution B
was added dropwise under vigorous stirring. The resulting suspension was stirred for 1 h in an
open vessel at room temperature. The final gel was transferred to a Teflon-lined stainless steel
autoclave. The autoclave was heated under rotation at 443 K for 5 days. The product was
recovered by filtration. The residue was first washed with copious amounts of demineralized
water followed by washing with ethanol. The product was dried overnight at 383 K. The
organics were removed by calcination in air at 823 K (HZSM-5-sheet).
Mo loading
Prior to Mo loading, the calcined zeolites were exchanged three times with an aqueous 1 M
NH4NO3 solution at 353 K for 4 h. The sample was then dried overnight at 383 and calcined
at 723 K for 4 h in artificial air. Mo was loaded onto the zeolites by impregnation with an
aqueous solution of ammonium heptamolybdate tetrahydrate (AHM, Merck). The targeted
Mo content was 4 wt%. After drying for 1 h at ambient, the Mo-modified zeolites were
calcined in artificial air at 823 K for 5 h. The heating rate was 1.5 K/min. The Mo-containing
zeolites are denoted by the prefix “Mo/”.
6.2.2. Characterization
The Mo and Al content of the Mo-modified zeolites was determined by inductively coupled
plasma optical emission spectroscopy (ICP-OES) using a Spectro CIROS CCD spectrometer
equipped with a free-running 27.12 MHz generator at 1400 W. Prior to analysis, samples were
digested in a mixture of HF/HNO3/H2O (1:1:1).
XRD patterns were recorded on a Bruker D4 Endeavor powder diffraction system using
Cu Kα radiation with a scanning speed 0.01 ° sec−1
in the 2θ range 5–60 °. The zeolite
crystallinity was determined using the Bruker TOPAS 3.0 program.
Ar sorption isotherms were measured at 87 K on a Micromeritics ASAP2020 system in
static mode. The samples were outgassed at 623 K for 8 h prior to the sorption measurements;
samples were transferred to the measuring port using TranSeals. The Brunauer–Emmett–
Teller (BET) equation was used to calculate the specific surface area (SBET) from the
Chapter 6
128
adsorption data in the p/p0 range of 0.05–0.25. The mesopore volume (Vmeso) and mesopore
size distribution were calculated using the Barrett–Joyner–Halenda (BJH) method applied to
the adsorption branch of the isotherm. The micropore area (Smicro) and micropore volume
(Vmicro) were calculated using the t-plot method using a thickness range of 3.5-5.4 Å [30].
Transmission electron micrographs were obtained with a FEI Tecnai 20 transmission
electron microscope (TEM) at an electron acceleration voltage of 200 kV. Typically, a small
amount of sample was suspended in ethanol, sonicated and dispersed over a Cu grid with a
holey carbon film. Scanning electron microscopy (SEM) was performed using a Philips
environmental FEIXL-30 ESEM FEG in high-vacuum mode at low voltage.
6.2.3. Catalytic activity measurements
An amount of 0.5 g of catalyst was introduced in a tubular quartz reactor with a length of
490 mm and an internal diameter of 4.0 mm. The length of the catalyst bed is 50 mm. The
catalyst was supported on a quartz wool plug in the isothermal zone of the oven. All gases
were fed using thermal mass controllers. Prior to reaction, the temperature was increased at a
rate of 5 K/min to 973 K in a He gas flow of 25 ml/min. The reaction was started by
switching the reactor feed to a mixture of CH4/N2 (5 vol% N2, used as internal standard)
mixture at a WHSV of 1.22 h-1
. This corresponds to a space velocity of 1710 ml CH4/gcat h.
Products were analyzed by an online Interscience CompactGC gas chromatograph equipped
with three analysis channels for separate analysis of light gases (Molsieve 5A, TCD), light
hydrocarbons (Al2O3/KCl, FID) and aromatics (Rtx-1, TCD).
6.3. Results and discussion
The XRD patterns of the starting zeolites are collected in Fig. 6.1. All samples have the
zeolite MFI structure and they do not contain indications for the presence of impurity phases.
The absence of a broad reflection around 23 º shows that the samples contain little amorphous
silica. The diffraction lines of HZSM-5-des, HZSM-5-org HZSM-5-nano and HZSM-5-sheet
are broader than those of the bulk HZSM-5 zeolites. This broadening is due to the smaller size
of the coherent scattering domains [31, 32].
Chapter 6
129
Fig. 6.1. XRD reflection patterns of (a) HZSM-5-ref, (b) HZSM-5-des, (c) HZSM-5-large, (d)
HZSM-5-org, (e) HZSM-5-nano and (f) HZSM-5-sheet.
The textural properties of the zeolites were measured by Ar physisorption. The corresponding
data determined by analysis of the adsorption branches of the isotherms are listed in Table 6.1.
The micropore volume of some of the nanocrystalline zeolites is slightly lower as compared
with those of the bulk zeolites. The micropore volume of HZSM-5-sheet is substantially lower.
This is in part due to the problem of using the t-plot method to analyze the micropore volume
for these sheet-like zeolites [34]. As expected, the bulk HZSM-5-large and HZSM-5-ref
zeolites do not contain mesopores. HZSM-5-des obtained by base leaching contains a
considerable amount of mesopores. HZSM-5-nano and HZSM-5-org also contain mesopores,
but in smaller amount than HZSM-5-des. The large mesopore volume in HZSM-5-des is due
to the presence of relatively large mesopores formed due to ill-controlled extraction of Si
from the zeolite framework [33]. Compared with the other zeolites, the nanosheet ZSM-5
sample has the largest mesopore volume. The mesopores in HZSM-5-sheet are mainly located
between the zeolite sheets [32].
10 20 30 40 50 60
Inte
nsit
y (
a.u
)
Angle (°)
Chapter 6
130
Table 6.1. Physico-chemical properties of the prepared ZSM-5 zeolite supports.
Sample Al1
(Wt%) Mo
1
(wt%)
Vmicro (cm
3/g)
Vmeso (cm
3/g)
Smicro (m
2/g)
Smeso (m
2/g)
SBET (m
2/g)
NBAS2
(μmol/g)
HZSM-5-large 1.0 3.04 0.15 0 265 2 445 554
HZSM-5-ref 1.9 4.12 0.14 0 257 16 432 679
HZSM-5-des 1.9 3.97 0.14 0.28 236 97 457 544
HZSM-5-nano 1.3 3.5 0.12 0.1 292 81 495 -
HZSM-5-org 3.2 4.5 0.11 0.12 129 67 333 1292
HZSM-5-sheet - - 0.02 0.76 89 390 570 280[29]
1 Elemental bulk content as determined by ICP-OES analysis.
2 Determined by infrared spectroscopy after desorption of pyridine at 423 K.
The crystal size and morphology of the zeolites were analyzed by TEM. It can be appreciated
that the set of zeolites covers a range of crystalline domain sizes from several nanometers to
tens of microns. The zeolite particles in HZSM-5-large (Fig. 6.2a) and HZSM-5-ref (Fig.
6.2b) have typical sizes of 10 μm and 1 μm, respectively. The crystals of HZSM-5-des (Fig.
6.2c), HZSM-5-org (Fig. 6.2d) and HZSM-5-nano (Fig. 6.2e) are substantially smaller. The
average micropore domain size of HZSM-5-des is ~100 nm. HZSM-5-org and HZSM-5-nano
comprise crystalline particles of about 10-30 nm. From the EM images the thickness of the
HZSM-5-sheet zeolite is estimated several nanometers.
The concentration of BAS of the zeolite samples was determined by IR spectroscopy of
adsorbed pyridine and the results are listed in Table 6.1. The desilicated ZSM-5 (HZSM-5-
des) material contains less BAS than HZSM-5-ref. Based on the data in Chapter 2, we can
ascribe this to the formation of some EFAl species. Large crystal ZSM-5 (HZSM-5-large)
zeolite contains a similar amount of BAS as HZSM-5-des; the lower acidity compared with
HZSM-5-ref is due to the lower Al content (Table 6.1). The acidity of HZSM-5-sheet sample
was substantially lower than of the other zeolites. The lower acidity of the nanosheet samples
has been discussed before in literature [32]. The acidity of HZSM-5-org is the highest which
is due to the high Al content of the material.
Chapter 6
131
Fig. 6.2. Electron Microscopy micrographs of (a) HZSM-5-large, (b) HZSM-5-ref, (c)
HZSM-5-des, (d) HZSM-5-org, (e) HZSM-5-nano and (f) HZSM-5-sheet.
6.3.2 Catalytic activity measurements
Prior to catalytic testing in MDA, the zeolites were loaded with approximately 4 wt% of Mo.
The bulk content was determined by ICP-OES analysis (Table 6.1). Evaluation of the catalytic
performance in the MDA reaction shows that Mo/HZSM-5-org has the highest methane
conversion rate (Fig. 6.3). The methane conversion rates for Mo/HZSM-5-large, Mo/HZSM-
5-nano and Mo/HZSM-5-des were slightly lower than that of the reference Mo/HZSM-5-ref
catalyst. The activity of Mo/HZSM-5-sheet was very low. These activity differences seem to
vary with the acidity; a larger concentration of BAS results in a higher activity, which is in
keeping with earlier reports [36, 37]. The relation between acidity and methane reactivity,
however, is not well understood. Tessonier et al. suggested that increased acidity improves the
Mo dispersion to explain the increased methane activity [38]. The benzene selectivities are
presented in Fig. 6.3b. The lower initial benzene selectivity of the Mo/HZMS-5-des,
Mo/HZSM-5-nano and Mo/HZSM-5-sheet samples compared with the reference trends with
the lower acidity of these samples. The aromatization step is considered to take place at the
BAS located inside the micropores. In line with this, the initial benzene selectivity of the large
50 nm
30 nm 20 nm 20 nm
a b c
d e f
9.66 μm
Chapter 6
132
bulk Mo/HZSM-5 (Mo/HZSM-5-large) is also lower compared with that of Mo/HZSM-5-ref.
The lower initial benzene selectivity for the large ZSM-5 crystals may also be due to the
longer diffusion lengths that will result in more secondary coking reactions. The high initial
benzene selectivity of Mo/HZSM-5-org is attributed to high acidity. The improved stability in
benzene formation of the small micropore domain sized catalysts (Mo/HZSM-5-des,
Mo/HZSM-5-nano, Mo/HZSM-5-org and Mo/HZSM-5-sheet) can be ascribed to the
improved accessibility of the benzene selective BAS located in the zeolites micropore system.
In this way, the detrimental effect of carbon laydown at the catalysts external surface with
progressive time on stream is suppressed. In Chapter 4 we attributed deactivation in benzene
formation to the growth of a carbon layer at the external surface of the zeolite surface. Such
carbon layer blocks the micropores and, thus, the benzene selective BAS. Furthermore,
decreasing the micropore domain size improves the diffusion of products out of the zeolite
crystals and thereby limiting secondary reactions leading to the formation of carbon. In
Chapter 3 it was suggested that the formation of carbon inside the micropores contributes to
some extent to catalyst deactivation. However, the effect of such intracrystalline carbon
formation on catalyst deactivation was argued to be smalled compared with the effect of
carbon formation at the external surface.
Fig. 6.3. Methane reaction rate (a) and benzene selectivity (b) of spent Mo/HZSM-5-ref
(square), Mo/HZSM-5-des (circle), Mo/HZSM-5-large (triangle), Mo/HZSM-5-org
(diamond), Mo/HZSM-5-nano (pentagonal) and Mo/HZSM-5-sheet (star).
0 2 4 6 8 10
0
2
4
6
8
10
12
14
Acti
vit
y
mm
ol/
gcat.h
Time on stream (h)
0 2 4 6 8 10
0
20
40
60
80
100
Ben
zen
e s
ele
cti
vit
y (
wt%
)
Time on stream (h)
Chapter 6
133
The nature of the carbonaceous deposits on the spent catalysts was characterized by TG
analysis. The weight-loss curves shown in Fig. 6.4 contain various features. The peak at ~
750 K relates to carbon formed in the proximity of molybdenum [39-41] and is referred to as
soft coke (Csoft). Carbon with a poly-aromatic nature [39-41] is most likely formed at the
external surface BAS and is indicated by a peak at ~ 850 K. This coke is typically assumed to
be of poly-aromatic nature and, accordingly, named hard coke (Chard). The weight-loss curves
were deconvoluted into contributions of the various C types and the results listed in Table 6.2.
These data indicate that the carbon content is lower for the bulk MFI materials (Mo/HZSM-5-
large and Mo/HZSM-5-ref) compared with the hierarchical structured materials. The larger
amount of Csoft in spent hierarchical structured catalysts Mo/HZSM-5-nano, Mo/HZSM-5-des
and Mo/HZSM-5-sheet suggests the presence of less well dispersed Mo-oxo phase. These
large metal-oxide particles lead to the formation of large Mo2C particles, which are known to
produce more soft coke nature. The substantial Csoft content in the spent Mo/HZSM-5-sheet
sample indicates that the Mo phase is poorly dispersed in the nano-layered sample. Analysis
of the carbon deposits in spent Mo/HZSM-5-org revealed a high concentration of Chard and
suggests a large amount of external surface BAS, which is in line with the large bulk BAS
concentration in Mo/HZSM-5-org (Table 6.2).
Fig. 6.4. TGA weight loss curves of spent (a) Mo/HZSM-5-ref, (b) Mo/HZSM-5-des, (c)
Mo/HZSM-5-large, (d) Mo/HZSM-5-org, (e) Mo/HZSM-5-nano and (f) Mo/HZSM-5-sheet.
600 700 800 900 1000
Weig
ht
loss (
mg
/K)
Temperature (K)
Chapter 6
134
Table 6.2. Coke content on the spent catalyst after 12 h on stream in the MDA reaction.
Sample Ctotal
(g/gcat) Csoft
(g/gcat)
Chard
(g/gcat)
Mo/HZSM-5-large 0.08 0.01 0.07
Mo/HZSM-5-ref 0.13 0.02 0.11
Mo/HZSM-5-des 0.19 0.10 0.10
Mo/HZSM-5-org 0.21 - 0.21
Mo/HZSM-5-nano 0.12 0.06 0.06
Mo/HZSM-5-sheet 0.19 0.14 0.05
6.4 Conclusions
In this chapter we investigated the effect of the micropore domain size of the ZSM-5 zeolite
component in the MDA catalyst on the final catalytic performance in MDA. To this end, a set
of ZSM-5 zeolites was prepared varying in micropore domain size. Characterization of the
prepared zeolite materials revealed the crystalline domain sizes to vary from 10 μm to several
nm’s. Evaluating the performance of the Mo modified zeolites in the MDA reaction revealed
a beneficial effect of a decreased crystalline domain size on the stability in benzene formation.
We argue that the better stability of the hierarchical catalysts in MDA is due to improved
accessibility of the BAS located inside the micropores. Hierarchical structuring decreases the
adverse effect of micropore blockage. The results from catalytic testing indicate a relation
between the Brønsted acidity of the zeolite support and the methane reaction rate.
6.5 References
[1] F. Solymosi, A. Erdohelyi, A. Szöke, Catal. Lett. 32 (1995) 43–53.
[2] F. Solymosi, J. Cserényi, A. Szöke, T. Bánsági, A. Oszkó, J. Catal. 165 (1997) 150–161.
[3] D. Ma, Y. Shu, M. Cheng, Y. Xu, X. Bao, J. Catal. 194 (2000) 105-114.
[4] A.J.J. Koekkoek, C.H.L. Tempelman, V. Degirmenci, M. Guo, Z. Feng, C. Li, E.J.M. Hensen,
Catal. Today, 168 (2011) 96-111. [5] L. Wu, E.J.M. Hensen, Catal. Today, 235 (2014) 160-168. [6] S. van Donk, A. H. Janssen, J. H. Bitter and K. P. de Jong, Catal. Rev. Sci. Eng., 45 (2003)
297–319. [7] Javier Pérez-Ramírez, C.H. Christensen, K. Egeblad, C.H. Christensen, J.C. Groen, Chem. Soc.
Rev. 37 (2008) 2530-2542. [8] D.P. Serrano, J.M. Escola, P. Pizarro, Chem. Soc. Rev. 42 (2013) 4004-4035. [9] S. Morin, P. Ayrault, N.S. Gnep, M. Guisnet, Appl. Catal. A Gen. 166 (1998) 281. [10] J. Lynch, F. Raatz, C. Delande, Stud. Surf. Sci. Catal. 39 (1987) 547. [11] J. Lynch, F. Raatz, P. Dufresne, Zeolites 7 (1987) 333. [12] V. Patzelova, N.I. Jaeger, Zeolites 7 (1987) 240. [13] A. Zukal, V. Patzelova, U. Lohse, Zeolites 6 (1986) 133. [14] Y. Sasaki, T. Suszuki, Y. Takamura, A. Saji, H. Saka, J. Catal. 178 (1998) 94.
Chapter 6
135
[15] C. Choi-Feng, J.B. Hall, B.J. Huggins, R.A. Beyerlein, J. Catal. 140 (1993) 395. [16] R.A. Beyerlein, C. Choi-Feng, J.B. Hall, B.J. Huggins, G.J. Ray, Top. Catal. 4 (1997) 27. [17] G.R. Meima, CATTECH 2 (1998), 5. [18] P. Hudec, J. Novansky, S. Silhar, T.N. Trung, M. Zubek, J. Madar, J. Ads. Sci. Tech. 3 (1986)
159. [19] B.L. Meyers, T.H. Fleisch, G.J. Ray, J.T. Miller, J.B. Hall, J. Catal. 110 (1988) 110, 82. [20] M. Tromp, J.A. van Bokhoven, M.T. Garriga Oostenbrink, J.H. Bitter, K.P. de Jong, D.C.
Koningsberger, J. Catal. 190 (2000) 209. [21] J.C. Groen, W. Zhu, S. Brouwer, S.J. Huynink, F. Kapteijn, J.A. Moulijn, J. Pérez-Ramírez, J.
Am. Chem. Soc. 129 (2007) 355–360. [22] J. C. Groen, L. A. A. Peffer, J. A. Moulijn and J. Pérez-Ramírez, J. Mater. Chem., 16 (2006)
2121-2131. [23] C. J. H. Jacobsen, C. Madsen, J. Houzvicka, I. Schmidt and A. Carlsson, J. Am. Chem. Soc., 122
(2000) 7116–7117. [24] K. Zhu, K. Egeblad and C. H. Christensen, Eur. J. Inorg. Chem., 2007, 3955–3960 [25] A..H. Janssen, I. Schmidt, C.J.H. Jacobsen, A.J. Koster, K.P. De Jong, Microporous
Mesoporous Mater., 65 (2003) 59-75. [26] M. Choi, H. S. Cho, R. Srivastava, C. Venkatesan, D.-H. Choi, R. Ryoo, Nat. Mater., 5 (2006)
718–723. [27] D.P. Serrano, J. Aguado, J.M. Escola, J.M. Rodríguez, A. Peral, Chem. Mater. 18 (2006) 2462–
2464. [28] K. Na, C. Jo, J. Kim, K. Cho, J. Jeng, Y. Seo, R.J. Messinger, B.F. Chmelka, R. Ryoo, Science,
333 (2011) 328–332. [29] S.Z. Chen, K. Huddersman, D. Keir, L.V.C. Rees 8 (1988) 106-109. [30] B.C. Lippens, J.H. de Boer, J. Catal. 4 (1965) 319–323. [31] K. Na, W. Park, Y. Seo, R. Ryoo, Chem. Mater. 23 (2011) 1273-1279.
[32] X.Zhu, L.Wu, P.C.M.M. Magusin, B. Mezari, E.J.M. Hensen, J. Catal., 327 (2015) 10-21.
[33] J.C. Groen, L.A.A. Peffer, J.A. Moulijn, J. Pérez-Ramírez, Chem. Eur. J., 11 (2005) 4983-4994.
[34] M.A. Camblor, A. Corma, S. Valencia, Microporous Mesoporous Mater., 25 (1998) 59-74.
[35] E. Verheyen, C. Jo, M. Kurttepeli, G. Vanbutsele, E. Gobechiya, T.I. Korányi., S. Bals, G. van
Tendeloo, R. Ryoo, C.E.A. Kirschhock, J.A. Martens, J. Catal. 300 (2013) 70-80.
[36] J. Shu, A. Adnot, B.P.A. Grandjean, Ind. Eng. Chem. Res. 38 (1999) 3860-3867.
[37] M. Marczewski, H. Marczewska, K. Mazowiecka, React. Kinet. Catal. Lett. 54 (1995) 81-86.
[38] J-P. Tessonier, B. Louis, S. Rigolet, M.J. Ledoux, C. Pham-Huu, App. Catal. A, 336 (2008) 79-
88.
[39] C.C. Williams, J.G. Ekerdt, J.M. Jehng, F.D. Hardcastle, A.M. Turek, I.E. Wachs, J. Phys.
Chem. 95 (1991) 8781–8791. [40] B.S. Liu, L. Jiang, H. Sun, C.T. Au, Appl. Surf. Sci. 253 (2007) 5092–5100. [41] Z.R. Ismagilov, E.V. Matus, L.T. Tsikoza, Energy Environ. Sci. 1 (2008) 526–541.
Chapter 6
136
Chapter 7
137
Texture, acidity and fluid catalytic cracking performance
of hierarchical faujasite zeolite prepared by an
amphiphilic organosilane
Summary
Mesoporous zeolite Y was synthesized by using an amphiphilic organosilane. The texture and
the acidity of the mesoporous zeolite samples were compared with a microporous faujasite
reference zeolite. The synthesis of the most suitable mesoporous zeolite Y was scaled up in
order to prepare composite catalysts that could be tested for fluid catalytic cracking.
Composite catalysts were prepared by spray-drying the zeolite with Kaolin as filler and an
alumina sol as binder. The acidic properties of these composite FCC catalysts prepared from
conventional and mesoporous faujasite zeolites were compared. While IR spectroscopy after
H/D exchange with deuterated benzene indicates that strong bridging hydroxyl groups are
present in the freshly prepared composite catalysts, these zeolitic Brønsted acid sites are not
observed anymore in the lab-deactivated composite catalysts. These samples contain a
significant number of weaker Brønsted acid sites. The strength of the acid sites in the
composite catalysts is comparable with the acidity of amorphous silica-alumina. The
composite catalysts show excellent catalytic performance in the fluid catalytic cracking of
vacuum gas oil. These data indicate that relatively weak acid sites are responsible for the
FCC activity. The well-embedded mesoporosity in the parent hierarchical zeolite crystals
results in increased diesel and decreased gasoline and coke yield.
This chapter is accepted for publication in Fuel Process. Techn. (2015) In press.
Chapter 7
138
7.1 Introduction
Fluidized catalytic cracking (FCC) is one of the key processes in the refinery industry,
converting heavy gas oils obtained from atmospheric and vacuum distillation into lighter
products [1-4]. The products from FCC units are valuable for the gasoline and diesel pool.
Currently, more than 400 FCC units are operated worldwide, converting approximately 10
million barrels heavy feedstock per day [1]. The success of this process is based on its
simplicity, relatively low construction and operation costs and the flexibility in processing
various types of heavy feedstock [4]. Zeolite Y has been the main acid component of FCC
catalysts. Composite FCC catalysts used in industry further comprise clays, binders and other
additives [5]. Zeolite Y is the preferred zeolite component because of its relatively large pore
openings (7.4 Å), strong Brønsted acidity and good (hydro)thermal stability [6, 7]. Another
important aspect is that it can be synthesized at low cost without organic structure-directing
agents.
Under FCC conditions, the pore openings of 7.4 Å impose diffusion limitations and limit
the conversion of the larger hydrocarbon molecules in the feed. One approach to overcome
this would be to prepare zeolites with larger pores, but such zeolites are usually not very
stable and their synthesis requires expensive organic templates [8-11]. Another approach to
ease diffusion of large molecules in zeolites is the fabrication of hierarchical zeolites that
contain mesopores well-interconnected with the micropore network in the zeolite [12-14].
One distinguishes bottom-up and top-down approaches. In top-down approaches, usually Si
[15-18] or Al [19-24] atoms are extracted from the zeolite framework. In bottom-up
approaches, the mesoporous zeolite is formed in one step, usually by adding to the synthesis
gel a second template as void filling spheres such as carbon black particles or in the form of
organic molecules that act as mesoporogen during the formation of the zeolite [24].
Hierarchization has most frequently been applied in the synthesis of ZSM-5 and BEA
zeolites [12, 14, 24, 25]. It is well-known that mesopores are created in zeolite Y crystals
during steam treatment as employed to convert low-acidic freshly prepared zeolite Y into
strongly acidic ultrastabilized Y zeolites, which are the main acid component in
hydrocracking operations [18]. Usually, the mesopores are not uniformly distributed over the
zeolite crystals and, sometimes, they are also not connected to the external surface [26]. De
Jong and co-workers investigated how steam treatment followed by acid and base leaching
steps improves the mesopore interconnectivity, which is useful to limit secondary cracking
Chapter 7
139
reactions [27]. The benefit of mesoporosity on the cracking performance of vacuum gas oil
and bulky model molecules has been well established [28-32].
Mesoporous Y zeolite can also be obtained using surfactant templates [33-36, 37-40].
Garcia-Martinez et al. reported about the scale-up of a surfactant-templated process to prepare
mesoporous Y zeolite; composite catalysts based on such hierarchical zeolite Y showed
improved yield of valuable gasoline and light cycle oil (LCO) products over bottoms and coke
in FCC catalyst evaluation [41]. Another versatile method to introduce mesopores in zeolites
involves the use of organosilanes that covalently bind to the growing zeolite surface [33-37].
This approach was first described by the Ryoo group in the preparation of mesoporous ZSM-5
[33]. The organosilane dimethyl-octadecyl-(3-trimethoxysilylpropyl)-ammonium chloride
(TPOAC) has also been used in the preparation of hierarchical zeolite Y [34-36]. For instance,
Fu et al. reported improved catalytic performance in the hydrodesulfurization of 4,6-
dimethyldibenzothiophene using a TPOAC-templated mesoporous Y zeolite support [34].
Another example is the benefit of mesopores in zeolite Y in the aldol condensation of n-
butanol with benzaldehyde [35]. Jin et al. showed that replacing a small portion of bulk
zeolite Y with hierarchical zeolite Y led to a shift in the FCC selectivity from coke to gasoline
and dry gas products [36].
In the present study, we investigated the feasibility of using TPOAC in the direct
synthesis of hierarchical zeolite Y for use in FCC composite catalyst. We first optimized the
synthesis of mesoporous Y zeolite at small (gram) scale. The most promising mesoporous
zeolite was scaled up. For comparison, a bulk zeolite Y was prepared and its Al content was
lowered by substitution of framework Al by Si. The acidic properties of the scaled-up
zeolites, the composite catalysts derived thereof and the lab-deactivated and regenerated FCC
catalysts were characterized in detail. The FCC performance of two lab-deactivated composite
catalysts was evaluated in an Advanced Catalytic Evaluation (ACE) unit.
7.2 Experimental
7.2.1 Zeolite synthesis
For the synthesis of reference bulk zeolite Y, a seed gel was prepared by dissolving 4.04 g
NaOH and 2.0 g NaAlO2 in 19.97 g water. Then, 22.80 g sodium silicate solution (Prolabo,
25.5 – 26.5% SiO2) was added dropwise under vigorous stirring. The resulting seed gel (gel
A) was aged overnight at room temperature. In a second round bottom flask, a feedstock gel
Chapter 7
140
(gel B) was prepared. After dissolving 0.04 g NaOH and 3.31 g NaAlO2 in 33.19 g water,
35.56 g sodium silicate (26 wt% in water, Prolabo) was added dropwise under vigorous
stirring. The Si/Al ratio of the feedstock gel was varied between 2.5 and 5.0 by adjusting the
amount of sodium aluminate. To prepare the final synthesis gel, an amount of 4.46 g of the
aged seed gel A was added to the feedstock gel B under vigorous stirring and was stirred for
another hour. The resulting gel was transferred into a 125 mL Teflon-lined stainless-steel
autoclave and heated in a static oven at 373 K for 24 h.
The gel for obtaining mesoporous zeolite Y was prepared in the same way as described
above. Prior to the hydrothermal step, dimethyl-octadecyl-(3-trimethoxysilylpropyl)-
ammonium chloride (TPOAC, ABCR, 60 wt% in methanol) was added dropwise to the
synthesis gel; this gel was further stirred for 4 h. The Si/TPOAC ratio was varied between 10
and 125. The gels were then hydrothermally treated at 373 K for 72 h. The solid materials
were recovered by filtration of the suspension, followed by washing with copious amounts of
water. To remove TPOAC, the solids were calcined in artificial air (20/80 (v%/v%) O2/He).
The materials are denoted by FAU(x, y) with x being the SiO2/TPOAC ratio (∞, 125, 45, 20,
10) and y the Si/Al ratio (2.5, 3.5, 5.0) in the starting gel.
A portion of the mesoporous zeolite Y prepared with an SiO2/TPOAC ratio of 45 and an
initial Si/Al ratio of 5.0 in the synthesis gel (Y(45,5.0)) was treated with ammonium
hexafluorosilicate (AHFS). The zeolite was first ion-exchanged four times with 1 M KNO3,
followed by four exchange cycles with 1 M NH4NO3 under reflux. After drying the zeolite
overnight, 10 g of zeolite was slurried in 100 ml of 3.4 M ammonium acetate at 348 K. An
amount of 135 ml of 0.1622 M AHFS was added dropwise over a period of 2 h. The final
slurry was stirred overnight at 348 K. The solid was recovered by filtration and washed with 1
L of hot (363 K) demineralized water. The washed solid was dried in a vacuum oven at 293 K.
The synthesis of several zeolite materials was scaled up by increasing the reactant amount
by a factor of 16. The hydrothermal synthesis was done in an 1.5 L Teflon-lined autoclave. At
this scale, a standard zeolite Y was synthesized at a Si/Al ratio of 2.5 and a mesoporous
zeolite Y at a Si/Al ratio of 5.0 in the presence of TPOAC (SiO2/TPOAC = 45). The Si/Al
ratio of the microporous zeolite Y was increased by AHFS treatment according to the
procedure outline above. The proton forms of these materials were obtained by suspending 1
g of calcined zeolite in 10 ml 1 M NH4NO3 for 4 h at 353 K. The ion-exchange was repeated
twice. The final step was calcination in artificial air at 723 K for 4 h. The calcined zeolites are
denoted as FAU(∞, 4.1)-large and FAU(45, 2.9)-large, reflecting the final Si/Al ratios as
Chapter 7
141
determined by ICP analysis.
Composite catalysts were prepared by spray-drying the zeolite with Kaolin as filler and
alumina sol as the binder. The resulting catalyst composite consisting of 35 wt % zeolite, 50
wt % Kaolin and 15 wt % alumina were subjected to steam-calcination to simulate catalyst
regeneration. The composite catalysts were steamed at 1023 K for 4 h using 100 % steam
followed by calcination at 873 K for 1 h. The deactivated catalyst was then sieved to obtain
particles in the range 38−212 μm and calcined at 873 K for 2 h.
7.2.2 Characterization
Elemental analysis was done by inductively coupled plasma optical emission
spectroscopy (ICP-OES) on a Spectro CIROS CCD spectrometer equipped with a free-
running 27.12 MHz generator at 1400 W. Zeolite samples were dissolved in a mixture of
HF/HNO3/H2O (1:1:1).
XRD patterns were recorded on a Bruker D4 Endeavor powder diffraction system using
Cu Kα radiation with a scanning speed 0.01 °·sec−1
in the 2θ range 5-60 °. The crystallinity
was determined according the standardized procedure ASTM D 3906-80. To this end, the
intensities of the 15.7 °, 18.8 °, 20.5 °, 23.8 °, 27.2 ° and 34.3 ° 2θ reflections (corresponding
to the [331], [511], [440], [533], [642] and [555] hkl planes) were taken after background
subtraction and related to the intensities of the highly crystalline sample FAU(∞, 2.5)
prepared in this work FAU(∞, 2.5).
Ar physisorption isotherms were measured at 87 K on a Micromeritics ASAP2020 system
in static measurement mode. The samples were outgassed at 623 K for 8 h prior to the
sorption measurements. The Brunauer–Emmett–Teller (BET) equation was used to calculate
the specific surface area (SBET) in the pressure range p/p0 = 0.05–0.25. The mesopore volume
(Vmeso) and mesopore size distribution were calculated using the Barrett–Joyner–Halenda
(BJH) method on the adsorption branch of the isotherm. The micropore area (Smicro) and
micropore volume (Vmicro) were calculated from the t-plot curve using the thickness range
between 3.5 and 5.4 Å [42].
Nuclear Magnetic Resonance (NMR) measurements were performed on a 11.7 Tesla
Bruker DMX500 NMR spectrometer operating at 500 MHz for 1H, 99 MHz for
29Si and 132
MHz for 27
Al. The 27
Al MAS NMR was done using a Bruker 2.5-mm MAS probehead
spinning at 20 kHz; 27
Al NMR spectra were recorded with a single pulse sequence with a 18°
pulse duration of 1 μs and a interscan delay of 1 s. A saturated Al(NO3)3 solution was used for
Chapter 7
142
27Al NMR shift calibration. The
1H and
29Si MAS NMR measurements were carried out using
a 4-mm MAS probehead with sample rotation rates of 12.5 kHz for 1H and 10 KHz for
29Si
NMR measurements. 1
H and 29
Si NMR shifts were referred to Tetramethylsilane (TMS).
Quantitative 29
Si NMR spectra were recorded using a high power proton decoupling direct
excitation (HP-DEC) pulse sequence with a 45 o pulse duration of 2.5 μs and an interscan
delay of 160 s. For 1H MAS NMR measurements, the zeolites were first dehydrated at a
temperature of 723 K at a pressure lower than 10−5
mbar for 6 h. The dehydrated zeolites were
placed into the 4-mm MAS NMR zirconia rotor under inert conditions and transferred to the
NMR probehead. Quantitative 1H NMR spectra were recorded with a Hahn-echo pulse
sequence p1-τ1-p2-τ2-aq with a 90o pulse p1 = 5 μs and a 180º p2 = 10 μs and an interscan
delay of 120 s.
Infrared spectra were recorded in the 4000-400 cm−1
range using a Bruker Vertex 70v
apparatus. Samples were pressed into a self-supporting wafer with a density of about 10
mg/cm2. To remove physisorbed water, the sample was evacuated for 2 h at 773 K at a
pressure lower than 2 x 10−6
mbar. After evacuation, the sample was cooled to 323 K; then, a
background spectrum was recorded. The total concentration of the Brønsted acid sites was
determined by monitoring the H/D exchange reaction with d6-benzene (C6D6, Sigma Aldrich)
following a literature procedure [43]. C6D6 was kept in a glass ampoule connected to an
evacuated gas supply system. C6D6 was dosed into the cell with a computer controlled
pneumatic valve, delivering a dose of 0.33 mmol C6D6. The sample was exposed for various
times to the probe, followed by evacuation for 1 h. The sequence was repeated to record the
spectra of partially exchanged samples with exposure times of 30 min, 30 min and 60 min at
303 K; 30 min at 323 K, 30 min at 373 K, and 30 min at 523 K. The total concentration of the
Brønsted acid sites was determined by IR spectroscopy of adsorbed pyridine. Pyridine
adsorption was carried out on the dehydrated zeolite wafer at 373 K. After saturation, the
sample was evacuated at 423 K for 1 h and a spectrum was recorded. This desorption step was
repeated at 573 K and 723 K. After each desorption step a spectrum was recorded at 423 K.
The spectra were deconvoluted and the acidity was quantified using the extinction coefficient
values reported by Datka [44].
Transmission electron micrographs were obtained with a FEI Tecnai 20 transmission
electron microscope (TEM) at an electron acceleration voltage of 200 kV. Typically, a small
amount of sample was suspended in ethanol, sonicated and dispersed over a Cu grid with a
holey carbon film. Scanning electron microscopy (SEM) was performed using a Philips
Chapter 7
143
environmental FEIXL-30 ESEM FEG in high-vacuum mode at low voltage.
7.2.3 Catalytic activity measurements
The Brønsted acidity of some of the materials was evaluated by catalytic activity
measurements in the bifunctional hydroconversion of n-heptane [43, 45, 46]. For this purpose,
the zeolites were loaded with 0.4 wt% Pd following incipient wetness impregnation with an
aqueous solution of Pd(NH3)4(NO3)2. The impregnated materials were calcined at 573 K and
sieved in a 250-500 μm mesh fraction. Prior to catalytic testing, the catalyst was reduced in a
H2 of 100 ml/min at 713 K and 30 bar. The reaction temperature was lowered with 0.2 K/min
from 713 K to 473 K. The product stream was analyzed by online gas chromatography.
The catalytic activity in fluid catalytc cracking was evaluated in an Advanced Catalytic
Evaluation unit (ACE, Kayser Technology, USA). The reaction temperature was 803 K. The
feedstock was a vacuum gas oil (VGO) obtained from a PetroChina refinery in Dalian. An
amount of 9 g composite catalyst was weighed into the reactor; the amount of VGO feed was
1.5 g. The contact time was 90 s. Gas products were analyzed using an online refinery gas
analyzer (M/s AC Analyticals). The boiling point distributions of the liquid products were
analyzed using a simulated distillation gas chromatograph (M/s AC Analyticals). Coke
deposited on the catalyst was burnt in a catalyst regeneration step and quantified using an
online CO2 analyzer.
7.3 Results and discussion
7.3.1 Optimization and scale-up of the synthesis procedure
We first varied the crystallization time for zeolite Y in the presence of TPOAC. The XRD
patterns of the obtained materials are shown in Fig. 7.1. Without TPOAC, highly crystalline
zeolite Y was obtained in 24 h. The optimal hydrothermal synthesis time in the presence of
TPOAC was 72 h. A shorter crystallization time resulted in a higher fraction of amorphous
silica; longer crystallization times led to the formation of zeolite P. The formation of zeolite P
as a competitive phase during zeolite Y synthesis has been reported before [47]. The need for
longer crystallization times in the presence of TPOAC in the synthesis gel is in line with
results of other studies [35, 48]. Fig. 7.2 shows the XRD patterns of the calcined zeolites
synthesized at different gel SiO2/TPOAC ratios. The XRD pattern of FAU(∞,2.5) is similar to
the one reported for crystalline zeolite Y [49, 50]. FAU(125,2.5) and FAU(45,2.5) also have
Chapter 7
144
the FAU structure, but their crystallinities are lower (Table 7.1). The materials prepared at
Si/TPOAC < 45 did not crystallize under the given conditions; the broad reflection observed
for these materials around 2θ = 23° shows that mainly amorphous silica was formed.
Fig. 7.1. XRD patterns of calcined zeolites: (a) FAU(∞,2.5), (b) FAU(125,2.5), (c)
FAU(45,2.5), (d) FAU(20,2.5) and (e) FAU(10,2.5).
Fig. 7.2. XRD patterns of calcined zeolites: (a) FAU(∞,2.5), (b) FAU(125,2.5), (c)
FAU(45,2.5), (d) FAU(20,2.5) and (e) FAU(10,2.5).
10 20 30 40 50 60
Inte
nsit
y (
a.u
.)
Angle (°)
10 20 30 40 50 60
Inte
nsit
y (
a.u
.)
Angle (°)
Chapter 7
145
The textural properties of the crystalline zeolites were investigated by Ar physisorption.
Fig. 7.3 shows the isotherms for FAU(∞, 2.5), FAU(45, 2.5) and FAU(125, 2.5). The
hysteresis loops in the relative pressure region of 0.4-0.8 evidence that FAU(45, 2.5) and
FAU(125, 2.5) contain mesopores. The textural data are collected in Table 7.1. FAU (∞, 2.5)
contains only micropores; the textural data agree with reported values for zeolite Y [51]. The
two crystalline mesoporous zeolites, FAU(45,2.5) and FAU(125,2.5), possessed a significant
amount of mesopores. The mesopore volume increased with increasing TPOAC content in the
synthesis gel. The lower micropore volume as compared with FAU (∞, 2.5) is most likely due
to the decreased crystallization degree, which was also apparent from the XRD patterns.
Chapter 7
146
Table 7.1. Textural properties of the zeolite Y materials.
Sample SiO2/TPOAC Si/Al1
Si/Al2
Vmicro
(cm3/g)
Vmeso
(cm3/g)
Smicro
(m2/g)
Smeso
(m2/g)
SBET
(m2/g)
CXRD3
(%)
FAU(∞, 2.5) ∞ 2.5 2.05 0.28 0.02 571 15 612 100
FAU(∞, 3.5) ∞ 3.5 2.51 0.23 0 404 4 463 98
FAU(125, 2.5) 125 2.5 2.14 0.22 0.11 447 94 639 84
FAU(45, 2.5) 45 2.5 2.58 0.16 0.15 325 143 566 79
FAU(45, 3.5) 45 3.5 3.04 0.26 0.11 512 72 619 67
FAU(45, 5.0) 45 5.0 3.50 0.16 0.10 207 50 420 78 1 Si/Al ratio in the synthesis gel;
2 Determined by ICP elemental analysis;
3 Relative XRD crystallinity determined according the ASTMD
3906-80 standard procedure.
Chapter 7
147
Fig. 7.3. Ar physisorption isotherms of calcined zeolites: (a) FAU(∞, 2.5), (b) FAU(125, 2.5)
and (c) FAU(45, 2.5). The isotherms are presented in a stacked fashion for clarity. The Y
offsets are progressively increased with 100 cm3/g for each subsequent sample.
FAU(∞, 2.5) and FAU(45, 2.5) were treated with AHFS with the aim to increase the
framework Si/Al ratio. FAU(45, 2.5) was chosen because of its relatively large mesopore
volume and its good crystallinity. The framework structure of microporous FAU(∞, 2.5) was
largely retained during the AHFS treatment (Fig. 7.4), whereas mesoporous FAU(45, 2.5)
zeolite was found to collapse when treated in the same manner.
0.0 0.2 0.4 0.6 0.8 1.0
Vo
lum
e a
dso
rbed
cm
3/g
.ST
P
Relative pressure
a
c
b
0.0 0.2 0.4 0.6 0.8 1.0
Vo
lum
e a
dso
rbed
cm
3/g
.ST
P
Relative pressure
Chapter 7
148
Fig. 7.4. (left) XRD patterns of calcined zeolites: (a) FAU(∞,2.5), (b) FAU(∞,3.5) (c)
FAU(∞,5.0) and (d) FAU(∞,2.5) after treatment with AHFS; (right) XRD patterns of calcined
zeolites: (a) FAU(45,2.5), (b) FAU(45,3.5), (c) FAU(45,5.0) and (d) FAU(45,5.0) after
treatement with AHFS.
It was also attempted to increase the framework Si/Al ratio by lowering the Al content of
the synthesis gel. For the conventional synthesis without TPOAC, crystalline zeolites were
obtained at Si/Al ratios of 2.5 and 3.5 (Fig. 7.4). The XRD pattern of FAU(∞, 5.0) shows that
crystallization was not possible at this Si/Al ratio, in line with literature [52]. On contrary,
when zeolite Y was synthesized at Si/Al ratios of 3.5 and 5.0 in the presence of TPOAC,
highly crystalline materials were obtained as follows from the XRD patterns in Fig. 7.4.
However, the bulk Al content of the resulting samples was higher than targeted. For FAU(45,
3.5) and FAU(45, 5.0), the Si/Al ratios were 3.04 and 3.50, respectively, as determined by
elemental analysis. Textural analysis of FAU(45, 5.0) shows that this material combines
micropores and mesopores. The TEM images of FAU(45, 5.0) also reveal that the mesopores
are well integrated into the primary zeolite particles (Fig. 6c).
As larger quantities of zeolites were required for the ACE test, the synthesis of FAU(45,
5.0) was scaled-up. The resulting zeolite is denoted as FAU(45, 5.0)-large. For comparison,
the synthesis of microporous FAU(∞, 2.5) was also carried out at the same scale; this zeolite
was then treated with AHFS to decrease the framework Al content. This reference zeolite is
denoted as FAU(∞, 4.1)-large. The XRD patterns of the zeolites in Fig. 7.5 (the two bottom
10 20 30 40 50 60
Inte
nsit
y (
a.u
.)
Angle (°)
10 20 30 40 50 60In
ten
sit
y (
a.u
.)
Angle (°)
Chapter 7
149
patterns) point out the higher crystallinity of calcined FAU(∞, 4.1)-large as compared with
FAU(45, 2.9)-large. This is in accordance with the data in Table 7.3. The crystallinity of
FAU(∞, 4.1)-large was slightly lower compared to that of the FAU(∞, 2.5) zeolite suggesting
scaling-up of the synthesis procedure to hamper the crystallization process. The bulk Si/Al
ratio of FAU(∞, 4.1)-large after the AHFS treatment was 4.1. The final Si/Al ratio of FAU(45,
2.9)-large was lower (Si/Al = 2.9) compared with FAU(∞, 4.1)-large. Ar physisorption data
confirm that FAU(45, 2.9)-large contained mesopores; these mesopores are also visible in the
TEM image shown in Fig. 7.6d. The diameter of the mesopores is approximately 5 nm. The
TEM images of the microporous zeolite do not contain evidence for such mesopores (Fig.
7.6b), in keeping with the low mesopore volume of this sample. FAU(∞, 4.1)-large (Fig. 7.6e)
is made up from cubic crystals. The morphology of FAU(45, 2.9)-large is different; the SEM
image in Fig. 7.6f shows that, in addition to cubic crystals, also thin sheet-like crystals are
present, which are intergrown with the larger zeolite crystals.
Table 7.2. Textural and acidic properties of the large-scale conventional and mesoporous Y
zeolites, FCC composite catalysts before and after lab-deactivation and after ACE and
regeneration by calcination.
Sample Vmicro
(cm3/g)
Vmeso
(cm3/g)
Smicro
(m2/g)
Smeso
(m2/g)
SBET
(m2/g)
CXRD1
(%)
FAU(∞, 4.1)-large 0.24 0.04 475 24 560 93
FCC(micro, fresh) 0.07 0.06 139 28 192 23
FCC(micro, steamed) 0.03 0.06 54 32 96 15
FCC(micro, regenerated) 0.02 0.08 47 43 103 15
FAU(45, 2.9)-large 0.14 0.11 286 41 383 69
FCC(meso, fresh) 0.03 0.04 55 32 127 13
FCC(meso, steamed) 0.006 0.04 14 22 37 5
FCC(meso, regenerated) 0.004 0.05 12 28 46 5 1 Relative XRD crystallinity determined according the ASTM D 3906-80 standard procedure.
Chapter 7
150
Fig. 7.5. XRD patterns of zeolites and composite catalysts: (left) (a) FAU(∞, 4.1), (b) FCC
(micro), (c) FCC(micro, steamed) and (d) FCC(micro, regenerated); (right) (a) FAU(45, 2.9),
(b) FCC(meso), (c) FCC(meso, steamed), (d) FCC(meso, regenerated).
Fig. 7.6. TEM micrographs of calcined zeolites: (a) FAU(∞, 2.5), (b) FAU(∞, 4.1)-large, (c)
FAU(45, 5.0), d) FAU(45, 5.0)-large, and SEM images of e) FAU(∞, 2.5)-large and f)
FAU(45, 5.0)-large.
10 20 30 40 50 60
Inte
nsit
y (
a.u
.)
Angle (°)
10 20 30 40 50 60
Inte
nsit
y (
a.u
.)
Angle(°)
10 20 30 40 50 60In
ten
sit
y (
a.u
.)
Angle (°)
20 nm
20 nm
20 nm
20 nm
b a c
d e f
Chapter 7
151
7.3.2 NMR spectroscopy
27Al MAS NMR spectroscopy was used to determine the Al coordination of the zeolites.
NMR spectra are shown in Fig. 7.7. The spectra are dominated by a feature at δ = 55 ppm (δ,
chemical shift), corresponding to Al in the zeolite framework (FAl). A second smaller feature
at δ = 0 ppm is due to extraframework Al (EFAl). Both zeolite samples contain most of the Al
in the framework, but small amounts of EFAl were also present. The presence of EFAl trends
with the lower crystallinity of these materials as determined from XRD experiments.
Fig. 7.7. 27
Al MAS NMR spectra of (a) FAU(∞, 4.1)-large and (b) FAU(45, 2.9)-large.
The framework Si/Al ratio (Si/AlFW) of the zeolites was determined by 29
Si MAS NMR
spectroscopy. The spectra shown in Fig. 7.8 contain features due to Si species in different
coordination environments (denoted as Si(nAl) with n indicating the number of Al atoms in
the next-nearest-neighbour (NNN) positions). Peaks at δ = -90 ppm, δ = -95 ppm and δ = -101
ppm are attributed to Si(3Al), Si(2Al) and Si(1Al) species [53, 54]. Si(0Al) species in the
zeolite framework give rise to the feature at δ = -107 ppm [53, 54]. The results of the
deconvolution of these spectra are given in Table 7.3. The estimated framework Si/Al ratios
are 4.8 and 3.5 for FAU(∞, 4.1)-large and FAU(45, 2.9)-large, respectively. Both samples
contain less Al in the framework than the bulk Al content and this is consistent with the
presence of EFAl.
Table 7.3. Physico-chemical properties of the zeolites prepared at larger scale.
Sample Si(3Al)1
(%)
Si(2Al)1
(%)
Si(1Al)1
(%)
Si(0Al)1
(%)
Si/AlFW1 Si/Albulk
2
FAU(∞, 4.1)-large 1.8 14.6 48.3 35.3 4.8 4.05
FAU(45, 2.9)-large 7.2 26.5 39.5 26.8 3.5 2.90 1Determined from
29Si MAS NMR spectra;
2Bulk Si/Al ratio determined by ICP-OES analysis.
Chapter 7
152
Fig. 7.8. 29
Si MAS NMR spectra of (a) FAU(∞, 4.1)-large and (b) FAU(45, 2,9)-large.
The hydroxyl groups in the zeolites were characterized by 1H MAS NMR spectroscopy
(Fig. 7.9). Peaks at δ = 4.6 ppm and δ = 4.0 ppm represent BAS located in the sodalite cages
and supercages, respectively. The peak at δ = 2.6 ppm is related to hydroxyl groups associated
with EFAl. Silanol groups are identified by the signal at δ = 1.9 ppm. The content of the
various hydroxyl groups was quantified by deconvolution of the 1H NMR spectra and the
results are listed in Table 7.4. FAU(∞, 4.1)-large contains more BAS than FAU(45, 2.9)-large.
For FAU(45, 2.9)-large, the amount of BAS in the sodalite cages is lower than the amount of
BAS in the supercages. FAU(45, 2.9)-large zeolite has a larger content of hydroxyls groups
related to EFAl and silanols. On contrary, the amount of BAS in the sodalite and supercages is
similar for FAU(∞, 4.1)-large.
Chapter 7
153
Fig. 7.9.
1H MAS NMR spectra of (a) FAU(∞, 4.1)-large and (b) FAU(45, 2.9)-large.
Table 7.4. Concentration of OH groups determined by deconvolution of 1H MAS NMR
spectra.
Sample FAU(∞, 4.1)-large FAU(45, 2.9)-large
Chemical shift
(ppm)
µmol/g Chemical shift
(ppm)
µmol/g
BAS
sodalite cages 4.6 519 4.6 83
Supercages 4.0 502 3.8 138
AlOH 2.6 276 1.8 190
SiOH 1.9 149 2.4 128
7.3.3 FTIR spectroscopy of H/D exchange with C6D6
To quantify the BAS, the selective H/D exchange of the bridging hydroxyl groups with
deuterated benzene was monitored by FTIR spectroscopy. The relevant FTIR spectra of
FAU(∞, 4.1)-large and FAU(45, 2.9)-large are shown in Figs. 7.10 and 7.11. The spectra of
the dehydrated zeolites contain features at 3550 cm−1
and 3631 cm−1
in the OH stretching
region due to bridging hydroxyl groups in the sodalite cages and the supercages, respectively
[44]. The spectra also contain features of silanol groups (3745 cm−1
) and aluminol groups
associated with EFAl species (3670 cm−1
) [43]. The intensity of the BAS bands is higher for
FAU(∞, 4.1)-large as compared with FAU(45, 2.9)-large. After exposure to C6D6, the
intensities of the 3550 cm−1
and 3631 cm−1
bands are lower. The new bands at 2630 cm−1
and
2680 cm−1
are due to deuteroxyl groups corresponding to BAS in sodalite cages and
supercages, respectively, after selective exchange of the acidic protons with deuterium [43].
Chapter 7
154
Inspection of the spectrum in the OD stretching region indicates that FAU(∞, 4.1)-large
contains less BAS in the sodalite cages than in the supercages, in line with the difference
noted by 1H NMR spectroscopy. The corresponding spectrum of FAU(∞, 4.1)-large points to
nearly equivalent amounts of BAS in sodalite cages and supercages for this sample. The H/D
exchange occurs at higher rate for FAU(∞, 4.1)-large as compared with FAU(45, 2.9)-large.
The difference relates to the higher framework Al content of the latter zeolite [43]. Not all of
the BAS in these zeolites could be exchanged following extensive H/D exchange at 523 K in
contrast to earlier results for zeolite Y [43]. The reason for this difference is not clear, but it
might mean that some parts of the crystals are not accessible to C6D6. The BAS content of the
two zeolites was estimated by deconvolution of the OD stretching region after H/D exchange
at 523 K according to established procedures [43]. The data are reported in Table 7.5. The
BAS concentrations are 1.51 mmol/g and 0.65 mmol/g for FAU(∞, 4.1)-large and FAU(45,
2.9)-large, respectively. The BAS concentrations are significantly lower compared with the
theoretical values based on the framework Si/Al ratios determined from the 29
Si MAS NMR
data. Decreased crystallinity due to the presence of TPOAC in the gel and, in case of the
microporous zeolite, the AHFS treatment may explain this. EFAl species may also partially
compensate the negative framework negative charge instead of protons. The acidity of
FAU(45, 2.9)-large is lower than that of FAU(∞, 4.1)-large; this difference is in keeping with
the 1H MAS NMR data.
Chapter 7
155
Table 7.5. Results of acidity characterization of zeolite component and FCC composite catalysts.
Sample NBAS1
(μmol/g)
T40%2
NBAS-4233
(μmol/g)
NBAS-5733
(μmol/g)
NBAS-7233
(μmol/g)
NLAS-4233
(μmol/g)
NLAS-5733
(μmol/g)
NLAS-7233
(μmol/g)
FAU(∞, 4.1)-large 1512 - - - - - - -
FCC(micro,fresh) 340 - - - - - - -
FCC(micro,steamed) 0 592 100 60 21 154 94 46
FCC(micro,regenerated) 0 - 74 49 17 48 26 8.7
FAU(45, 2.9)-large 648 - - - - - - -
FCC(meso,fresh) 286 - - - - - - -
FCC(meso,steamed) 0 609 30 15 3.5 118 64 29
FCC(meso,regenerated) 0 - 40 22 5.8 68 36 14 1 Concentration of BAS determined by H/D exchange FTIR at 523 K [31];
2 Temperature required to reach n-heptane conversion of 40%;
3 Concentration
of BAS and LAS determined after evacuation for 1 h at 423 K, 573 and 723 K.
Chapter 7
156
Fig. 7.10. FTIR spectra showing the (left) OH and (right) OD stretching regions of FAU(∞, 4.1)-large recorded after exposure to d6-benzene
at different times and temperatures.
3800 3700 3600 3500 3400 3300 3200
0.0
0.1
0.2
0.3
0.4
Ab
so
rban
ce
Wavenumber cm-1
2800 2700 2600 2500 2400
0.0
0.1
0.2
0.3
Wavenumber (cm-1)
Activated 303 K, 1800 sec 303 K, 1800 sec 303 K, 3600 sec 323 K, 1800 sec 373 K, 1800 sec 523 K, 1800 sec
303 K, 1800 sec 303 K, 1800 sec 303 K, 3600 sec 323 K, 1800 sec 373 K, 1800 sec 523 K, 1800 sec
Chapter 7
157
Fig. 7.11. FTIR spectra showing the (left) OH and (right) OD stretching regions of FAU(45, 2.9)-large recorded after exposure to d6-benzene
at different times and temperatures.
3800 3700 3600 3500 3400 3300 3200
0.00
0.05
0.10
0.15
0.20
Ab
so
rban
ce
Wavenumber (cm-1)
2800 2700 2600 2500 2400
0.00
0.05
0.10
Wavenumber (cm-1)
Activated 303 K, 1800 sec 303 K, 1800 sec 303 K, 3600 sec 323 K, 1800 sec 373 K, 1800 sec 523 K, 1800 sec
303 K, 1800 sec 303 K, 1800 sec 303 K, 3600 sec 323 K, 1800 sec 373 K, 1800 sec 523 K, 1800 sec
Chapter 7
158
7.3.4 Composite catalyst characterization
Composite catalysts were prepared by spray-drying the zeolite with Kaolin as filler and
alumina sol as the binder. The resulting catalysts were deactivated by steaming at 1023 K for
4 h using 100 % steam followed by calcination at 873 K for 1 h. After FCC performance
evaluation, the catalysts were regenerated in air at 823 K. The physical properties of the
freshly prepared, lab-deactivated and regenerated composite catalysts were determined by
XRD and Ar physisorption. The contribution of the zeolite component in the composite
catalysts is clearly visible in the XRD patterns (Fig. 7.5). The additional broad feature in these
patterns around 23° is due to amorphous silica originating from the kaolin component. The
textural data of the composite catalysts listed in Table 7.2 point out the lower micropore
volume as compared with the parent zeolites. The decrease in micropore volume of the
composite catalysts trends well with the XRD crystallinity. The decrease is only slightly
higher than the zeolite content; this suggests that the procedure to obtain the composite
catalysts has only slightly damaged the zeolite structure. The BAS content of the composite
FCC(meso, fresh) catalyst as measured by H/D exchange FTIR is nearly proportional to the
zeolite content of the composite catalyst. For FCC(micro, fresh), the BAS density is slightly
lower than the zeolite content.
As customary in FCC catalyst evaluation, the composite catalysts were subjected to a
steam-calcination treatment that simulates the deactivation in the FCC regenerator. Fig. 7.5
shows that this treatment led to a significant decrease of the zeolite crystallinity. The stronger
decrease of the crystallinity of mesoporous Y zeolite points out its lower hydrothermal
stability, which can be linked to the higher Al content. The lower stability is also evident from
the stronger decrease of the micropore volume during the steam treatment step. After FCC
evaluation and regeneration at 773 K, the XRD patterns were almost unchanged. This shows
that the accelerated deactivation treatment yields representative equilibrium catalysts. This
conclusion based on XRD is underpinned by the textural data. According to H/D exchange
FTIR, the deactivated and regenerated composite catalysts do not contain BAS. The low BAS
content of the steamed and regenerated catalysts is also indicated by the absence of features at
3550 cm-1
and 3631 cm-1
in their corresponding FTIR spectra after dehydration (Fig. 7.12).
We estimate that the detection limit of the H/D exchange FTIR method is 0.002 mmol/g.
Chapter 7
159
Fig. 7.12. FTIR spectra showing the OH stretching regions of zeolites and composite
catalysts: (left) (a) FAU(∞, 4.1)-large (b) FCC(micro), (c) FCC(micro,steamed), (d)
FCC(micro, regenerated); (right) (a) FAU(45, 2.9)-large (b) FCC(meso), (c) FCC(meso,
steamed) and (d) FCC(meso, regenerated).
The acidity of the steam-deactivated and regenerated composite catalysts was also probed
by FTIR spectroscopy of adsorbed pyridine. While H/D exchange FTIR mainly probes strong
zeolite acidity, pyridine probes a broader range of BAS [55]. BAS and LAS contents
determined by deconvolution of the FTIR spectra as function of the evacuation temperature
are collected in Table 7.5. Both zeolites contain only a small amount of BAS as represented
by the values after evacuation at 423 K. The Brønsted and Lewis acidities were higher in the
FCC(micro) composite catalysts as compared with the FCC(meso) ones. The regenerated
composite catalysts had nearly similar acidities as the parent samples before FCC catalyst
activation. The amount of pyridine after evacuation at 723 K represents strong BAS. The
amount of such strong BAS in the steam-deactivated and regenerated zeolites is very low; it is
much lower than the amount of strong BAS in amorphous silica-aluminas [55]. H/D exchange
FTIR method is able to probe this small amount of strong BAS in amorphous silica-aluminas.
Thus, we speculate that the low acidity in the composite catalysts is mainly due to an
amorphous silica-alumina phase, which is likely closely integrated in the zeolite material. The
number of strong zeolite BAS in the composite catalysts is too low to be titrated by H/D
exchange FTIR. Overall, the Brønsted acidity of FCC(micro) is higher than that of
FCC(meso). Also, the FCC(micro) composite zeolites contain more Lewis acid sites.
3800 3700 3600 3500 3400 3300 3200
Ab
so
rban
ce
Wavenumber (cm-1)
3800 3700 3600 3500 3400 3300 3200
Wavenumber (cm-1)
Chapter 7
160
We further determined the acidity of the composite catalysts by measuring their catalytic
performance in the bifunctional hydroisomerization of n-heptane. For this purpose, we loaded
the catalysts with 0.4 wt% Pd. At this Pd content, the isomerization of hydrocarbons is limited
by the Brønsted acidity of the catalyst [43]. The acidity of zeolites, clays and silica-alumina in
terms of hydroisomerization activity correlates with the concentration of strong BAS of
zeolitic strength titrated by the H/D exchange FTIR method [43]. The temperature required
for a conversion of 40% (T40) is used to measure the acidity. The T40 values for FCC(micro)
and FCC(meso) were 592 and 609 K. These values are typical for amorphous silica-alumina
samples, which contain only few zeolite-like BAS [43, 56]. Thus, we speculate that only very
few zeolite-type acid sites remain due to the extensive steam-deactivation procedure.
7.3.5 Catalytic activity measurements
The lab-deactivated composite catalysts were then evaluated for their FCC catalytic
performance in an ACE testing unit using a VGO feed. Relevant data about the VGO are
given in Table 7.6. It is mainly composed of saturated and aromatic compounds. The sulfur
content is relatively low for VGO. In addition, the VGO contains small amounts of Ni and V.
The ACE test was carried out at a catalyst-to-oil ratio of 6 and a contact time of 90 s. The
temperature was 803 K. The products were analyzed by established techniques and grouped
into dry gas, gasoline, light cycle oil (LCO), bottoms and coke product classes. Table 7.7
shows that both composite zeolite catalysts can achieve high conversion of the feed. The
FCC(micro)-based catalyst exhibited a slightly higher VGO conversion than the catalyst
based on the FCC(meso) zeolite, which is most likely due to the higher acidity of the former
composite catalyst [57, 58]. We cannot draw firm conclusions about the influence of the
mesopores on the feed conversion because of the acidity difference of the parent zeolite
components. FCC(meso) shows significantly higher LCO (diesel) yield at nearly similar
gasoline yield. The combined yield of less valuable coke and LPG products is lower
compared with the reference. Although the bottoms yield is higher than for the microporous
reference catalysts, this will not be a significant drawback in practice because this fraction can
be recycled to the riser. By comparison with yield-conversion data in fluid catalytic cracking
[59,60], the higher diesel and lower gasoline yield can be reasonably linked to the improved
diffusion due to presence of mesopores that limits secondary cracking reactions of LCO and
bottoms [32].
Chapter 7
161
Table 7.6. VGO composition and properties used as feedstock for FCC catalyst evaluation.
Density (293 K)
(g/cm3)
0.93
Viscosity (373 K)
(mPa•s) 13.84
Carbon residue (wt.%) 4.19
C (wt.%) 86.76
H (wt.%) 11.63
S (wt.%) 0.40
N (wt.%) 0.69
Saturates (wt.%)
Aromatics (wt.%)
Resins (wt.%)
Asphaltenes (wt.%)
64.68
28.44
6.68
0.21
Ni (ppm)
V (ppm)
4.9
3.7
Table 7.7. Product distribution after FCC catalyst evaluation of composite catalysts (VGO, T
= 803 K, catalyst-to-oil ratio = 1.7, contact time 90 s).
Composite
catalyst dry gas
(wt%) LPG
(wt%)
gasoline
(wt%) LCO
(wt%) coke
(wt%) bottoms
(wt%) conversion
(wt%) FCC(micro) 3.2 16.7 39.6 22.9 10.7 6.9 93
FCC(meso) 2.3 12.7 36.3 29.3 8.1 11.3 89
7.4 Conclusions
The synthesis of mesoporous faujasite zeolite using an amphiphilic organosilane was
optimized. Using TPOAC, longer crystallization times were needed as compared with
conventional zeolite Y synthesis. The TPOAC-modified syntheses gave mesoporous zeolites
with appreciable mesoporosity, yet decreased crystallinity. While it was possible to selectively
remove Al from the framework of well-crystallized microporous zeolite Y by treatment with
AHFS, such treatment led to the collapse of the mesoporous Y zeolites. Two zeolites prepared
at larger scale, namely microporous Y zeolite followed by AHFS treatment and mesoporous Y
zeolite were used to prepare composite FCC catalysts by spray-drying the zeolite with Kaolin
as filler and alumina sol as binder. Infrared spectroscopy after H/D exchange shows the
presence of strong bridging hydroxyl groups in the freshly prepared composite catalysts in
amounts that are in keeping with the zeolite content. After accelerated steaming, these strong
zeolitic Brønsted acid sites are not observed anymore. These samples contain a significant
Chapter 7
162
number of weaker Brønsted acid sites. The strength of the acid sites in the composite catalysts
is comparable with the acidity of amorphous silica-alumina. The composite catalysts show
excellent catalytic performance in the fluid catalytic cracking of a vacuum gas oil. The
catalytic data indicate that the relatively weak acid sites are responsible for the FCC activity.
The well-embedded mesoporosity in the parent zeolite crystals results in higher diesel and
lower gasoline yield.
7.5 References
[1] C.I.C. Pinheiro, J.L. Fernandes, L. Domingues, A.J.S. Chambel, I. Graça, N.M.C. Oliveira,
H.S. Cerqueira, F Ramôa Ribeiro, Ind. Eng. Chem. Res. 51 (2012) 1-29.
[2] W. Vermeiren, J.-P. Gilson, Top. Catal. 52 (2009) 1131-1161.
[3] H.S. Cerquira, G. Caeiro, l. Costa, F. Ramôa Ribeiro, J. Mol. Catal. A 292 (2008) 1-13.
[4] A.A. Avidan, Oil Gas J. 90 (1992) 59-67.
[5] C. Liu, Y. Deng, Y. Pan, S. Zheng, X. Gao, App. Catal. A 257 (2004) 145-150.
[6] G. Bergerhoff, W.H. Baur, W. Nowacki, N. Jb, Miner. Mh. (1958) 193-200.
[7] J. Dhainaut, T.J.Daou, A.Chappaz, N.Bats, B.Harbuzaru, G.Lapisardi, H.Chaumeil, A. Defoin,
L.Rouleau, J.Patarin, Microporous Mesoporous Mater. 174 (2013) 117-125.
[8] J. Sun, C. Bonneau, Á. Cantín, A. Corma, M. J. Díaz-Cabañas, M. Moliner, D. Zhang, M. Li, X.
Zou, Nature 458 (2009) 1154–1157 .
[9] A. Corma, M. J. Díaz-Cabañas, J. Jiang, M. Afeworki, D. L. Dorset, S. L. Soled, K. G.
Strohmaier, Proc. Natl. Acad. Sci. U. S. A. 107 (2010) 13997–14002.
[10] J. Jiang, J. L. Jorda, J. Yu, L. A. Baumes, E. Mugnaioli, M. J. Diaz-Cabanas, U. Kolb, A.
Corma, Science 333 (2011) 1131–1134.
[11] J. Jiang, Y. Yun, X. Zou, J.L. Jorda, A. Corma, Chem. Sci. 6 (2015) 480-485.
[12] S. Van Donk, A.H. Janssen, J.H. Bitter, and K.P. De Jong, Catal. Rev. Sci. Eng. 45 (2003) 297-
319.
[13] K. Li, J. Valla, J. Garcia-Martinez, ChemCatChem 6 (2014) 46-66.
[14] K. Möller, T. Bein, Chem. Soc. Rev. 42 (2013) 3689-3707.
[15] A. Čižmek, B. Subotic, I. Smit, A. Tonejc, A. Rosario, F. Crea, A. Nastro, Microporous Mater.
8 (1997) 159-169.
[16] J.C. Groen, J.C. Jansen, J.A. Moulijn, J. Pérez-Ramírez, J. Phys. Chem. B 108 (2004) 13062-
13065.
[17] J.C. Groen, L.A.A. Peffer, J.A. Moulijn, J. Pérez-Ramírez, Chem.-Eur. J. 11 (2005) 4983-4994.
[18] J.C. Groen, L.A.A. Peffer, J.A. Moulijn, J. Pérez-Ramírez, Colloids Surf. A 241 (2004) 53-58.
[19] R.M. Lago, W.O. Haag, R.J. Mikovsky, D.H. Olson, S.D. Hellring, K.D. Schmitt, G.T. Kerr,
Stud. Surf. Sci. Catal. 28 (1986) 677-684.
[20] G.J. Ray, B.L. Meyers, C.L. Marshall, Zeolites 7 (1987) 307-310.
[21] Q.L. Wang, G. Giannetto, M. Torrealba, G. Perot, C. Kappenstein, M. Guisnet, J. Catal. 130
(1991) 459-470.
[22] J.L. Molz, H. Heinichen, W.F. Hölderich, J. Mol. Catal. A. 136 (1998) 175-184.
[23] C.S. Triantafillidis, A.G. Vlessidis, L. Nalbandian N.P. Evmiridis, Microporous Mesoporous
Mater. 47 (2001) 369-388.
[24] Javier Pérez-Ramírez, C.H. Christensen, K. Egeblad, C.H. Christensen, J.C. Groen, Chem. Soc.
Rev. 37 (2008) 2530-2542.
[25] D.P. Serrano, J.M. Escola, P. Pizarro, Chem. Soc. Rev. 42 (2013) 4004-4035.
[26] J. Zečevic, C.J. Gommes, H. Friederich, P.E. de Jongh, K.P. de Jong, Angew. Chem. 51 (2012)
4213-4217.
Chapter 7
163
[27] K.P. de Jong, J. Zečevic, H. Friedrich, P.E. de Jongh, M. Bulut, S. van Donk, R. Kenmogne, A.
Finiels, V. Hulea, F. Fajula, Angew. Chem. 49 (2010) 10074-10078.
[28] A. Corma, Stud. Surf. Sci. Catal. 49 (1989) 49-67.
[29] A. Corma, Chem. Rev. 97 (1997) 2373-2420.
[30] K. Sato, Y. Nishimura, H. Shimada, Catal. Lett. 60 (1999) 83-87.
[31] Z. Qin, B. Shen, Z. Yu, F. Deng, L. Zhao, S. Zhou, D. Yuan, X. Gao, B. Wang , H. Zhao , H.
Liu, J. Catal. 298 (2013) 102-111.
[32] J.L. Agudelo , E.J.M. Hensen, S.A. Giraldo, L.J. Hoyos, Fuel Proc. Tech. 133 (2015) 89-96.
[33] M. Choi, H.S. Cho, R. Srivastava, C. Venkatesan, Dae-Heung Choi, R. Ryoo, Nature
Materials 5 (2006) 718 – 723.
[34] W. Fu, L. Zhang, T. Tang, Q. Ke, S. Wang, J. Hu, G. Fang, J. Li F.-S. Xiao, J. Am. Chem. Soc.
133 (2011) 15346–15349.
[35] B. Liu, F. Chen, L. Zheng, J. Ge, H. Xi and Y. Qian, RSC Adv. 3 (2013) 15075–15084.
[36] J. Jin, C. Peng, J. Wang, H Liu, X. Gao, H. Liu, C. Xu, Ind. Eng. Chem. Res. 53 (2014) 3406-
3411.
[37] K. Tang, X. Hong, Y. H. Zhao, Y. G. Wang, Petr. Sci. Tech. 28 (2011) 779-787.
[38] Y. Tao, H. Kanoh, K. Kaneko, J. Phys. Chem. B 107 (2003) 10974-10976
[39] Y. Tao, H. Kanoh, Y. Hanzawa, K. Kaneko, Colloids Surf. A: Physicochem. Eng. Aspects 241
(2004) 75–80.
[40] F.N. Gu, F. Wei, J.Y. Yang, N. Lin, W.G. Lin, Y. Wang, J.H. Zhu, J. Am. Chem. Soc. 22
(2010) 2442-2450.
[41] J. García-Martínez, K. Li, G. Krishnaiah, Chem. Comm. 48 (2012) 11841-11843.
[42] B.C. Lippens, J.H. de Boer, J. Catal. 4 (1965) 319–323.
[43] E.J.M. Hensen, D.G. Poduval, V. Degirmenci, D.A.J.M. Ligthart, W. Chen, F. Maugé, M.S.
Rigutto, J.A.R. van Veen, J. Phys. Chem. C 114 (2010) 8363-8374.
[44] J. Datka, A.M. Tutek, J.H. Jehng, I.E. Wachs, J. Catal. 135 (1992) 186-199.
[45] G.E. Gianetto, G.R. Perot, M.R. Guisnet, Ind. Eng. Chem. Prod. Res. Dev. 25 (1986) 481-490.
[46] J.W. Thybaut, C.S. Laxmi Narasimhan, J.F. Denayer, G.V. Baron, P.A. Jacobs, J.A. Martens,
G.B. Marin, Ind. Eng. Chem. Res. 44 (2005) 5159-5169.
[47] H. Faghihian, N. Godazandeha, J. Porous Mater. 16 (2009) 331-335.
[48] A. Inayat, I. Knoke, E. Spiecker, W. Schwieger, Angew. Chem. 51 (2012) 1962-1965.
[49] H. Yin, T. Zhou, Y. Liu, Y. Chai and C. Liu, J. Porous Mater. 19 (2011) 277–281.
[50] M.M.J. Treacy, J.B. Higgins, Collection of Simulated XRD Powder Patterns for Zeolites,
Elsevier, 2001.
[51] E.A. Pidko, S.M. T. Almutairi, B. Mezari, P.C.M.M. Magusin, E.J.M. Hensen, ACS Catal. 3
(2013) 1504-1517.
[52] S.M.T. Almutairi, B. Mezari, G.A. Filonenko, P.C.M.M. Magusin, M.S. Rigutto, E.A.
Pidko, E.J.M. Hensen, ChemCatChem 5 (2013) 452-466.
[53] J. Sanz, V. Fornés, A. Corma, J. Chem. Soc. Faraday Trans. 84 (1988) 3113-3119.
[54] W. Lutz, H. Toufar, D. Heidemann, N. Salman, C.H. Rüscher, T.M. Gesing, J.-Chr. Buhl, R.
Bertram, Microporous Mesoporous Mater. 104 (2007) 171-178.
[55] E.J.M. Hensen, D. Poduval, V. Degirmenci, D.A.J.M. Ligthart, W. Chen, F. Maugé, M.S.
Rigutto and J.A.R. van Veen, J. Phys. Chem. C 116 (2012) 21416-21429.
[56] D.G. Poduval, J.A. van Veen, M.S. Rigutto, E.J.M. Hensen, Chem. Comm. 46 (2010) 3466-
3468.
[57] R.J. Gorte, Catal. Lett. 62 (1999) 1-13.
[58] A.I. Biaglow, C. Gittleman, R.J. Gorte, R.J. Madon, J. Catal. 120 (1991) 88-93.
[59] J. Scherzer, Stud. Surf. Sci. Catal. 76 (1993) 145-182.
[60] J. García-Martínez, M. Johnson, J.Valla, K. Li, J.Y. Ying, Catal. Sci. Technol. 2 (2012)987-994.
Chapter 7
164
Summary
165
Hierarchical zeolites for catalytic hydrocarbon conversion
Currently, our society is largely fueled by crude oil, coal and gas. The depletion of these fossil
reserves, geo-political issues and environmental concerns are main drivers to look for other
carbon feedstocks to secure the fuels and chemicals demand in the future. The growing
concern about climate change also drives the search for renewable resources. As the lead time
for the development of a novel technology from invention to industrial use at a level that it
impacts the economy takes several decades, it is important to improve the efficiency of the
current industrial processes that convert fossil feedstock into fuels and chemicals. In addition,
it is necessary to use as clean as possible fossil feedstock. Natural gas with methane as the
main component has the highest H/C ratio of all fossil fuels and is accordingly considered as
the most important feedstock for the midterm when novel low-carbon technologies are being
developed. A further advantage of natural gas is that large reserves are available. However,
many of these reserves are stranded so that it is difficult to monetize them. Transport via
pipelines or in liquefied form either by liquefaction or after conversion into hydrocarbons via
the Fischer-Tropsch process are very costly. Accordingly, several approaches are under
investigation to convert methane directly into liquids. With methane oxidation typically
yielding only very low methanol yields, the option of aromatization of methane into benzene
and hydrogen has drawn most attention from industry and academia. In methane
dehydroaromatization (MDA), methane is directly converted into high value and easy to
handle liquid aromatic compounds, mainly benzene. Poor stability of the MDA catalyst has
hampered implementation of this reaction in the industry.
In Chapter 2, the influence of mesopores created by base desilication and silylation on the
catalytic performance of Mo/HZSM-5 in MDA was investigated. The desilication procedure
of HZSM-5 was optimized to have a mesoporous material with acidic sites similar to those in
bulk HZSM-5. The calcination procedure used during Mo introduction was shown to have a
strong influence on the physical properties of the final Mo/HZSM-5 material. It was shown
that impregnation of the zeolite with an aqueous ammonium heptamolybdate (AHM) solution
followed by calcination at 823 K for 5 h was optimal. The material obtained in this way
Summary
166
contained highly dispersed Mo-oxide phase and had a high zeolite crystallinity with only
small amounts of extraframework aluminum. The Mo/HZSM-5 material prepared upon
physical mixing with MoO3 had a low Mo-oxide dispersion after calcination at 823 K.
Evaluation of the catalytic performance of bulk and mesoporous Mo/HZSM-5 in MDA
showed that the mesoporous zeolite catalyst exhibited better stability. This improvement is
attributed to the smaller micropore domains of the zeolite support decreasing the negative
effect of micropore blockage by carbon deposition. Characterization of fresh mesoporous
Mo/HZSM-5 showed an improved Mo-oxide spreading over the surface. The methane
conversion rate and aromatics selectivity of mesoporous Mo/HZSM-5 was lower compared to
bulk Mo/HZSM-5. These difference are due to the formation of larger Mo-carbide (MoCx)
particles in mesoporous Mo/HZSM-5 when pretreating the catalysts in He followed by
exposure to methane-rich reaction conditions. The large MoCx particles formed in this way
display higher propensity to coke formation. In an alternative approach, a post-synthesis
silylation treatment was employed. Silylation selectively deactivates the Brønsted acid sites
(BAS) located at the external surface. Such BAS are suspected to form polyaromatic carbon
species during the MDA reaction leading to micropore blockage and, with this, catalyst
deactivation. The silylation of Mo/HZSM-5 led to an increase of the benzene selectivity and
methane conversion rate for both bulk and mesoporous Mo/HZSM-5 compared to their non-
silylated analogues. Post-mortem analysis of used catalysts showed smaller amounts of
aromatic hard coke in the silylated catalysts, which is in keeping with the lower rate of carbon
formation. The higher methane conversion rate after catalyst silylation was attributed to the
improved spreading of the Mo-oxide phase. It was found that silylation of the zeolite before
modification by Mo was less preferred as compared with silylation of the Mo/HZSM-5 zeolite.
The hydrophobicity of the zeolite after silylation led to poor Mo-oxide dispersion upon
modification of HZSM-5 with Mo, resulting in low methane conversion rate and benzene
selectivity in MDA.
In Chapter 3, the influence of the gas used during the pre-treatment of the catalyst on the
MDA reaction was investigated. Silylated and non-silylated Mo/HZSM-5 were exposed to
inert (He), oxidizing (artificial air) and carburizing (CH4 diluted in He) gases.
Characterization of the catalysts after such pretreatment steps showed that a larger fraction of
the Mo species was present in the micropores after pretreatment in He or air than after
precarburization in methane. The lower diffusivity of Mo species into the micropore channels
upon precarburization in methane was attributed to faster formation of immobile Mo carbides.
Summary
167
Catalytic evaluation revealed the highest catalytic stability and benzene selectivity for the
precarburized catalysts. The lower stability and benzene selectivity of air and He pretreated
catalysts was attributed to the greater amount of Mo-oxide species in the micropores for these
catalysts after pretreatment. The carburization of Mo-oxide particles inside the micropores
results in MoCx particles, which partially block the zeolite pores. A larger part of the
micropores is blocked after carburization for He and air pretreated catalysts, making a larger
part of the BAS inaccessible, than for precarburized catalysts.
Chapter 4 presents a comprehensive study to understand catalyst deactivation in the
MDA reaction. Catalysts were recovered after precarburization and after various times on
stream. These recovered catalysts were thoroughly characterized. With progressing reaction, a
poly-aromatic layer grows over the external surface of the catalyst. The formation of this layer
is detrimental for the catalyst performance in two ways. Firstly, the carbon layer blocks the
micropores, decreasing the accessibility to the BAS located inside the micropores. As a result,
the benzene formation decreases. Secondly, the carbon layer forms a barrier between the
initially highly dispersed MoCx particles and the zeolite surface. Due to lower interactions of
the Mo phase with the zeolite support, the MoCx particles agglomerate into larger particles,
resulting in a decreased methane conversion rate. Besides, the large MoCx particles are also
more susceptible to coke formation. The formation of the polyaromatic layer is likely due to
acid sites located at the external surface. As discussed in Chapter 2, these surface BAS can be
partially deactivated by a silylation treatment, which surpresses to some degree the growth of
the aromatic layer at the external surface.
In Chapter 5, a novel method for the preparation of nano-crystalline MCM-22 is
presented. To decrease the MCM-22 crystallite size an organosilane molecule (octadecyl-(3-
trimethoxysilylpropyl)-ammonium chloride, TPOAC) was added to the synthesis gel. When
grafted to the zeolite surface, the hydrophobic tail limits the growth of the MCM-22 crystal
and thus the average zeolite crystal size. The catalytic performance of the nano-crystalline
MCM-22 was evaluated in the MDA reaction as well as in liquid phase benzene alkylation.
Although the nano-crystalline MCM-22 had a lower Brønsted acidity than bulk MCM-22,
nano-crystalline MCM-2 showed improved performance in the MDA reaction compared with
conventional MCM-22. This is attributed to the smaller micropore domain size. In liquid
phase benzene alkylation, nano-crystalline MCM-22 showed catalytic performance
intermediate between bulk MCM-22 and ITQ-2. The improved performance of nano-
crystalline MCM-22 compared with bulk MCM-22 was attributed to the increased
Summary
168
accessibility of the zeolite BAS to benzene.
In Chapter 6, a brief study of the influence of the micropore domain size of the ZSM-5
zeolite support on the performance in the MDA reaction is presented. To this end, a series of
zeolites was prepared varying in micropore domain size. After Mo modification the various
catalysts were screened in the MDA reaction. The catalytic evaluation showed improved
benzene selectivities and catalytic stabilities for the hierarchical structured zeolites containing
smaller micropore domain sizes compared to their bulk analogues. We conclude that the
better stability of the hierarchical catalysts in the MDA reaction relates to the improved
accessibility of the BAS located inside the micropores. Furthermore, hierarchical structuring
decreases the adverse effect of micropore blockage.
Chapter 7 presents a study on the use of an organosilane molecule (octadecyl-(3-
trimethoxysilylpropyl)-ammonium chloride, TPOAC) as mesoporogen for the preparation of
hierarchical zeolite Y as potential acid component in FCC catalysts. The use of TPOAC in the
synthesis gel led to the formation of interconnected mesopores in the final zeolite crystals.
After optimization at the labscale, the synthesis of mesoporous and microporous zeolite Y
was scaled up with the purpose to prepare a bindered FCC catalyst, which could be evaluated
for its FCC performance in an Advanced Catalytic Evaluation (ACE) unit. A part of the
framework aluminum in the microporous material was selectively exchanged with silicon
following treatment with AHFS. Such treatment with the mesoporous material was not
possible, because it led to amorphization of the zeolite. The zeolites were used to prepare FCC
catalysts by spray-drying with Kaolin as filler and alumina sol as binder. Acidity
characterization by H/D exchange showed the presence of strong BAS in the freshly bindered
zeolites. After accelerated deactivation of the FCC catalysts by steam calcination, no strong
BAS were observed anymore. The lab-deactivated catalyst contained a substantial amount of
weaker BAS. The acid strength of these acid sites are comparable to those in amorphous silica
aluminas. Evaluation of these lab-deactivated FCC catalysts in an ACE unit using a vacuum
gas oil fluid catalytic showed high conversion rates and product distributions typical for
zeolite Y based cracking catalysts. These findings suggest that the residual weak acid sites are
responsible to catalyze the FCC reactions. The well-embedded mesoporosity resulted in
higher diesel and lower gasoline yield.
One of the major issues in the catalytic conversion of hydrocarbons is the coke formation
resulting in catalyst deactivation in processes such as dehydroaromatization of methane and
FCC of heavy oils. In general, catalyst stability largely determines the economic viability of a
Summary
169
process. Poor catalyst stability due to coke formation in the MDA reaction is an illustrative
example, where catalyst deactivation poses such a challenge that large scale implementation
in the chemical industry is hampered. To tackle deactivation problems, combining catalyst
development with advanced reaction engineering solutions appear important. To identify
better MDA catalyst formulations and process conditions, it is essential to understand better
the reaction mechanism, the exact state of the Mo component [1], the the role of the zeolite
support, the need for acid sites in the catalytic reaction and/or in stabilizing certain Mo states.
Two approaches must be considered. The design of better models with more controlled and
preferably homogeneous speciation of the Mo component to be able to better follow the
evolution of the Mo species and the acid sites during the MDA reaction. The other one is to
use advanced techniques to better investigate the catalysts during activation and operation.
For instance, high resolution electron microscopy and electron tomography can help to
establish the size and location of the Mo phase in these catalysts. A re-investigation of the Mo
phase by X-ray absorption spectroscopy starting from well-defined models will also be
helpful to resolve the location and structure of the catalytic sites that activate methane. Whilst
deactivation of the Mo/HZSM-5 seems to be a fact, it is worthwhile to investigate in more
detail the possibility to regenerate the catalysts by oxidation or reduction reactions. Oxidation
helps to remove the coke built up on the catalyst but has the drawback that oxidation of the
Mo-carbide leads to sublimation of the resulting Mo-oxide. Hydrogenation helps to reduce the
coke content also, but only as long as no large condensed aromatic products are formed.
Combinations of these approaches can also be considered.
Possible opportunities for catalyst improvement evolve from the rapid development in the
field of zeolite science. New topologies are reported frequently and routes for the hierarchical
structuring of zeolites to decrease micropore domain size are intensively investigated.
Tailoring the properties of the zeolite support could lower the rate of formation of
hydrocarbon species that deactivate the catalyst. The potential of new zeolite topologies and
procedures to introduce mesopores are not limited to the MDA reaction. Although processes
like FCC of heavy oil fractions are well established, there is still room for improvements. For
instance, Rive technology has shown the promise of hierarchically structured fuajasite zeolite
at the commercial scale [2]. The introduction of well interconnected mesopores in faujasite
improves the accessibility of the acid sites to bulky molecules present in heavy feedstocks
such that the conversion can be improved. Although many of the procedures used to generate
mesoporosity in zeolite require organic molecules as mesoporogens, our increasing
understanding about zeolite growth [3] may help to control the formation of advanced porous
solids with controlled porosity at all length scales and ideally placed active sites.
Summary
170
References
[1] J. Gao, Y. Zheng, J.M. Jehng, Y. Tang, I.E. Wachs, S.G. Podkolzin, Science 348 (6235) 686-690.
[2] K. Li, J. Valla, J. Garcia-Martinez, ChemCatChem. 6 (2014) 46-66.
[3] A.I. Lupulescu, J.D. Rimer, Science 344 (2014) 729-732.
List of publications
171
Journal publications
C.H.L. Tempelman, V.O. de Rodrigues, E.H.R. van Eck, P.C.M.M. Magusin, E.J.M
Hensen, Desilication and silylation of Mo/HZSM-5 for methane dehydroaromatization,
Microporous Mesoporous Mater. 203 (2015) 259-273.
2
C.H.L. Tempelman, X. Zhu, X., E.J.M. Hensen, Activation of Mo/HZSM-5 for methane
aromatization. Chin. J. Catal., 36 (2015) 829-837.
C.H.L. Tempelman, E.J.M. Hensen, On the deactivation of Mo/HZSM-5 in methane
dehydroaromatization. Applied Catal. B, 176 (2015) 731-739.
4 C.H.L. Tempelman, X. Zhu, K. Gudun, B. Mezari, B. Shen, E.J.M. Hensen, Texture,
acidity and fluid catalytic cracking performance of hierarchical faujasite zeolite prepared
by an amphiphilic organosilane. Fuel Process. Technol. (2015), in press.
e
r
5
C.H.L. Tempelman, M. T. Portilla, M.E. Martínez-Amero, B. Mezari, N.G.R. de
Caluwé, C. Martínez, E.J.M. Hensen, One-step synthesis of nano-crystalline MCM-22,
Microporous Mesoporous Mater. (2015), in press.
The author also contributed to the following publication outside scope of this thesis:
e
r
5
A.J.J. Koekkoek, C.H.L. Tempelman, V. Degirmenci, M. Guo, Z. Feng, C. Li, E.J.M.
Hensen, Hierarchical zeolites prepared by organosilane templating: a study of the
synthesis mechanism and catalytic activity, Catal. Today 168 (2011) 96-111.
Contribution to book publication
t
e
r
5
M.T. Portilla, C.H.L. Tempelman, C. Martinez, E.J.M. Hensen, New trends in catalyst
design for methane dehydroaromatization. In G. Centi (Ed.), Recent Advances in Gas to
Liquid Technologies (2015).
Conferences
C.H.L. Tempelman, V.O. de Rodrigrues, E.J.M. Hensen, Optimizing zeolites for non-
oxidative dehydrogenation of methane to benzene: Effect of mesopore introduction and
silylation, Europacat XI. 1 – 6 September 2013, Lyon, France. [Oral]
2
C.H.L. Tempelman, N.G.R. de Caluwé, B. Mezari, E.J.M. Hensen, One-pot synthesis of
mesoporous MCM-22 for benzene alkylation, 17th International Zeolite Conference (IZC
17). 7 – 12 July 2013, Moscow, Russia. [Poster]
List of publications
172
C.H.L. Tempelman, E.J.M. Hensen, An exploratory study on the deactivation of methane
aromatization catalysts: Opportunities for improved catalyst performance, 2nd
International Conference on Materials for Energy (EnMat II). 12 – 16 May 2013,
Karlsruhe, Germany. [Oral]
4 C.H.L Tempelman, V.O. de Rodrigues, E.J.M. Hensen, Optimizing zeolites for non-
oxidative dehydrogenation of methane to benzene: Effect of mesopore introduction and
silylation, 14th Netherlands' Catalysis and Chemistry Conference (NCCC XIV). 11 – 13
March 2013, Noordwijkerhout, The Netherlands. [Oral]
e
r
5
C.H.L. Tempelman, V.O. de Rodrigues, E.J.M. Hensen, Developing hierarchical zeolites
for MDA catalysis, Summer School Energy and Materials from the Sun. 20 – 23 June
2011, Rolduc Abbey, The Netherlands. [Poster]
V.O. de Rodrigues, C.H.L. Tempelman, P.C.M.M. Magusin, E.J.M. Hensen, Methane
dehydroaromatization: the influence of hierarchical structuring, 12th Netherlands'
Catalysis and Chemistry Conference (NCCC XII). 28 February – 2 March 2011,
Noordwijkerhout, The Netherlands. [Oral]
Acknowledgements
173
The long road has come to an end. The ride was sometimes rough and bumpy, but it made me
grow as a scientist and as a person. It was a privilege to work with people from so many
different backgrounds and to make a lot of new friends. Needless to say, without the help of
many people this thesis would never be possible. Probably I will forget some people in these
acknowledgements and therefore I apologize on beforehand.
First of all I would like to thank Emiel. You introduced me in the world of zeolite
catalysis when I was starting as a VKO student at the university. You showed me the
complexity of heterogeneous catalysis and thaught me to look critical at thing. Thank you for
all the advice and support during the project, but also the freedom to explore the broad world
of zeolite catalysis. It gave me the chance to evolve not only as a chemical engineer and
scientist, but also personally. Especially the opportunity to work in a European project with so
many partners and the possibility do a part of the research in Valencia were eye-opening
experiences. I am also grateful to you for giving me the chance to work in the CatchBio
project.
Also, I would like to thank the commitee members prof. Rutger van Santen, prof. Freek
Kapteijn, dr. Pieter Magusin, prof. Bert Weckhuysen and prof. Volker Hessel who took the
time to read and evaluate the thesis and the suggestions they gave me to further improve this
thesis.
Thank you Cristina, Teresa and Belene for fascilitating the benzene alkylation
experiments in Valencia and the scientific discussions during the project. Your hospitality
made the stay in Spain a great experience, which I will never forget!
My master graduation project with you Arjan was a perfect introduction to the IMC
group. You taught me the first steps as a scientist who came in handy the rest of the project.
You are a great teacher, a good scientist and a nice collegue to work with.
After working the first year of PhD with Victor I was somewhat spoiled. You generated a
great amount of data which largely contributed to the scientific output of the project. So
special thanks to you for kick-starting the project, it helped me during the rest of my PhD.
During the project I worked with several students who did a great part of the work and
initiated new ideas. Good luck Arno, Robin Broos and Robin Willems finishing your PhD
projects and Kristina and Niek with your careers.
Acknowledgements
174
Of course very important, and I guess this is true for everybody in the IMC group, is the
help I got from Emma. Your knowledge about the university rules has prevented major
disasters. I really admire your to-the-point acting and your willingness to help people. Several
times I promised you to devote a special chapter to you in the thesis, but unfortunately I had
to keep it limited to these few lines due to budget problems ;). I really enjoyed working with
you Brahim. Your knowledge on zeolites was not only limited to your expertise on NMR. I
like your persistence to understand “weird” findings and your willingness to help other
people. Discussing the results during the project and putting scientific weight to them resulted
in new insides. Johan and Tiny, also known as the “Buurman and Buurman” of IMC, you guys
are an essential part of the IMC group. The way you solve technical issues is similar to that of
a formula 1 pitstop crew; fast and good. I learned a lot of new skills, which are also useful at
home. Adelheid, I am thankful for the help during the ICP sessions, but foremost I remember
the Inorganic Chemistry lab practice course we supervised. Besides the new chemistry tricks I
learned, we talked a lot about things outside chemistry. What I really appreciated is that I
could always talk to you about personal matters which were bothering me.
Alessandro, I enjoyed our separate coffee breaks we had all over the campus. During our
private coffee sessions we had lively discussions about everything. The obviousness of you to
help people without anything in return is heart striking. I often had to laugh about your typical
“Alessandroiaanse” way of explaining thing; wordy but clear. William thanks for the fruitful
collaboration on the Sn-Beta project. But even more for teaching me politics, how to grow
peppers, the country of Austria and its inhabitants, how to speak proper German and more of
these wisdoms from Lage Zwaluwe. And of course the walk to the Jan Linders during the
lunch breaks. You are a nice person to work with and I admire your patience with people,
even when they are shouting at you. Nikolay and georgy, you showed me that Russians are
pleasant people to hang out with and that you can even be friends with them. Nikolay, it is a
shame that we never succeeded in winning the NIOK catalysis football cup with the Bulls of
Hensen. A likely reason is that we always peaked too late, the third half. Good luck with the
MDA project! Georgy, thanks for showing me the nightlife in Eindhoven and surroundings
and providing bed and breakfast when needed. Oh yeah, and the infamous stay in Lyon where
we cannot talk about. I hope you have a great time in Japan. Lara, I enjoyed the conversations
we had over a beer or a coffee. I hope you will finish your thesis soon! Evgeny P., you
encouraged me to get out of my comfort zone which boosted my confidence in new social
environments. I will remember traveling to the Next-GTL meetings in the so-called “flying
Acknowledgements
175
death-traps”. I had a lot of fun during these meetings. It was an absolute pleasure to work with
you during the CatchBio project. I like your non-conventional way of looking at things, it is
very refreshing. Volkan, I learned a lot from you during the graduation project and the first
years of PhD. Although we only met a few times after you left IMC, these moments were
legendary. During the last year I missed complaining to you, Aysegul, in our beloved misery
corner. I am looking forward to the next time we meet. Thanks to the other IMC group
members with who I shared good times during and after work: Andrey, Anton, Aleksander,
Arno, Abdul, Burcu, Chao Chao, Douglas, Evgeny U., Esther, Giulia, Jan-Philip, Jan, Juan,
Freke, Giulia, Gabriela, Guanna, Ivo, James, Kaituo, Lennart, Lingqian, Lei Lei, Long Chen,
Long Fei, Lu, Maarten, Robert, Robin, Sami, Tamas, Tobias, Wilburt, Xaoming, Xiaochun,
Xuefang, Yibin ….
Besides my friends of IMC I have to acknowledge some of my friends from home, which
are named here in no particular order. Thanks to you Kees; you brought me relieve in the
darkest times with your sense of humor. And you Marica for being even more dark than
Kees... ough! Together with little Boris you formed a family to go to when I felt absolutely
miserable. I also have to thank you Levijn. You are my friend since… highschool? You are
always there, whenever I am angry, anxious or sad, but above all you bring joy! Veerle,
besides being a good friend I learned from you that doing things without too much thinking
can be quite useful to get further in life. I am confident you will be Holland’s most famous
tattoo artist. Geert thank you for being... you! You and Daniëlle give sanity to this world with
your upright no-nonsense views. Therefore this need-to-keep-it-short-acknowledgement. Esgo
and Roland “Roel Driesvink” Gieles, the jamming sessions at the Tammer we had are
unforgetable and were a great relieve during times of stress. Roel, I am still waiting for your
book of memoires when you were sailing to the old Indies! Cor, you are a true friend without
agenda and politics. Naturally, I have to mention Jos, the most charismatic guy from Leiden.
The nice thing about you is that once you start talking you never stop. The sound waves you
produce in this way can be very relaxing and are often saturated with humor. And don’t forget
your voice imitation of a screaming pig, it is legendary. Nooks, Robin, Magriet, Githa and Eté
you always provide me a warm welcome to your home. Thank you for caring about me
whenever I needed it. Kwiest, Henriette, Nout, Sharon, Mik and Nomi; thanks for all the
Sunday mornings we were running with or without hangover and the holidays in Ireland.
Daniel, looking at you how you cope with your setbacks in live reminds me to put things in
perspective when things are tough. I respect you for your perservance and positivity. I am sure
Acknowledgements
176
you will conquer the world with your plates and build your imperium as a designer. Peer,
Neel, Lara and Freek, thank you for all the meals you cooked for me. It prevented me from
starving. And of course the rest of gang: Clemens, Dennis, Helma, Ilse, Krista, Marcel,
Marlieke, Remo, Silvie, Stefan, Thijs and Vince. You guys are the best and are special to me!
Judith, Cecile, Mathias, Jos, Marthe, Hugo and Merel, for almost 6 years I enjoyed being at
your home. It felt like I was a part of the family. Unfortunately, things went different but I will
always cherish that period.
Last but not least, my family. Thank you grandparents, aunt Luus, uncle Dik, aunt Marion
and uncle Gerard for always being supportive. It didn’t matter if it was school, sports or
music. A special thanks to uncle Herman who always believed in me. You took and still take
us everywhere we want and I relish the discussions we have during the long hikes we make.
Mirjam and Ivo, the fact that I can always come to your place to eat, drink beers, sport, work
in the garden, watch Dotje and Jaggers fight, complain, watch TV series but above all enjoy
your company is an enrichment to my life. Linda, thank you for entering my life during the
last bits of writing. The process became way lighter and made me start to enjoy things again.
And then finally my loving mom and dad! Although you pushed me the most of all people
you also always believed in my abilities. You taught me to respect myself and others and to do
what you think is right. Those lessons turned out to be very useful in life.
Curriculum vitea
177
Christiaan Tempelman was born on 02 September 1984 in Zaltbommel. After finishing HAVO in
2002 at Scholengemeenschap Cambium in Zaltbommel, The Netherlands, he studied Chemistry at the
Fontys Hogescholen in Eindhoven. After receiving his bachelor degree, he continued his master
studies Chemical Engineering at the Eindhoven University of Technology. He graduated in 2010
within the Inorganic Materials Chemistry (IMC) group on “Synthesis mechanism of mesoporous
ZSM-5” supervised by prof. dr. ir. E.J.M. Hensen. He continued to work in the same group by starting
a PhD project in 2010. His PhD study concerned the conversion of hydrocarbons over zeolite catalysts
with the main focus on the methane dehydro-aromatization reaction to directly convert low-value
methane to high-value benzene. The project was part of the Next-GTL consortium project funded by
the European Union. The results are presented in this dissertation and have been published as 5
research papers in peer reviewed international scientific journals. The findings in this work have been
presented at several national and international conferences. In 2013, he worked 6 weeks at the Instituto
Tecnología Químíca (ITQ) in Valencia as a visiting researcher. From March 2014 he started to work
as a researcher within the IMC group on the catalytic conversion of glucose based biomass towards
renewable fuels and chemicals. As of September 2015 he is employed at DAF Trucks N.V where he
works as development engineer on heavy-duty diesel after treatment systems.