Gas phase catalysis by zeolites Michael Sto ¨ cker * SINTEF Materials and Chemistry, Department of Hydrocarbon Process Chemistry, P.O. Box 124 Blindern, N-0314 Oslo, Norway Received 25 April 2004; received in revised form 15 November 2004; accepted 20 November 2004 Available online 8 April 2005 Dedicated to Diplom-Chemiker Ulf Blindheim on the occasion of his 70th birthday Abstract This paper provides an overview about todayÕs use of zeolites and related microporous materials as catalysts within the fields of refining, petrochemistry and commodity chemicals. The content of this presentation is devoted to gas phase catalysis—with focus on acid catalysis, hydrocarbon conversion and formation, oil and natural gas upgrading as well as catalytic probe reactions for the characterisation of zeolites and related microporous materials. The review is primarily meant for beginners who intend to get acquainted with this field. However, for more detailed information the interested reader is invited to consult the dedicated papers cited throughout this overview. Ó 2005 Elsevier Inc. All rights reserved. Keywords: Zeolites; Microporous materials; Gas phase catalysis; Crude oil upgrading; Natural gas conversion 1. Introduction Catalysis by zeolites—with focus on hydrocarbon conversion and formation—covers nowadays a broad range of processes related to the upgrading of crude oil and natural gas. This includes, among others, fluid cata- lytic cracking (FCC), hydrocracking, dewaxing, aliphate alkylation, isomerisation, oligomerisation, transforma- tion of aromatics, transalkylation, hydrodecyclisation as well as the conversion of methanol to hydrocarbons. All these conversions are catalysed by zeolites or related microporous materials, based both on the acid pro- perties and shape-selective behaviour of this type of materials. The first part of this chapter deals with the under- standing of the chemistry of acid catalysis using zeolites or related microporous materials, including the for- mation of acid sites, carbocation chemistry and their reaction mechanisms, as well as the importance of the shape-selectivity of the microporous materials. The term ‘‘zeolite’’ is used for the microporous aluminosilicate systems, however, SAPO type catalysts belong to the family of microporous materials as well. The second part of this chapter covers the discussion of the present situation and the new developments re- lated to the above mentioned petroleum refining and natural gas conversion processes using microporous materials, with focus on the new requirements due to the introduction of new fuel specifications world-wide. Finally, the third part of this chapter is dedicated to the different probe reactions with respect to the charac- terisation of zeolites and related microporous materials. Since the micropores of zeolites and related compounds have diameters in the range of molecular dimensions, the shape-selective effect reveals unique possibilities in the catalytic conversion of, for example, hydrocarbons. However, proper utilisation of this behaviour requires that the conversion occurs at active sites connected to the internal pore structure and not at the external 1387-1811/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.micromeso.2005.01.039 * Tel.: +47 98 24 39 33; fax: +47 22 06 73 50. E-mail address: [email protected]www.elsevier.com/locate/micromeso Microporous and Mesoporous Materials 82 (2005) 257–292
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
www.elsevier.com/locate/micromeso
Microporous and Mesoporous Materials 82 (2005) 257–292
Gas phase catalysis by zeolites
Michael Stocker *
SINTEF Materials and Chemistry, Department of Hydrocarbon Process Chemistry, P.O. Box 124 Blindern, N-0314 Oslo, Norway
Received 25 April 2004; received in revised form 15 November 2004; accepted 20 November 2004
Available online 8 April 2005
Dedicated to Diplom-Chemiker Ulf Blindheim on the occasion of his 70th birthday
Abstract
This paper provides an overview about today�s use of zeolites and related microporous materials as catalysts within the fields ofrefining, petrochemistry and commodity chemicals. The content of this presentation is devoted to gas phase catalysis—with focus on
acid catalysis, hydrocarbon conversion and formation, oil and natural gas upgrading as well as catalytic probe reactions for the
characterisation of zeolites and related microporous materials. The review is primarily meant for beginners who intend to get
acquainted with this field. However, for more detailed information the interested reader is invited to consult the dedicated papers
cited throughout this overview.
� 2005 Elsevier Inc. All rights reserved.
Keywords: Zeolites; Microporous materials; Gas phase catalysis; Crude oil upgrading; Natural gas conversion
1. Introduction
Catalysis by zeolites—with focus on hydrocarbon
conversion and formation—covers nowadays a broad
range of processes related to the upgrading of crude oil
and natural gas. This includes, among others, fluid cata-
The classical cracking mechanism is based on the fact
that a carbenium ion abstracts a hydride from an alkane
forming another carbenium ion, which cracks by b-scis-sion (cleavage of the C–C bond located b to the trivalentpositively charged carbon atom), forming an alkene—
see also Fig. 4 [6].
The overall process is governed by the stability of the
carbenium ions in the different states of the reaction. In
addition, the reaction rate decreases in the sequence
for an alkane molecule [6]. Reproduced by permission of Elsevier,
Amsterdam.
260 M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292
bi-functional catalysts. However, classical cracking can
also take place on mono-functional acid catalysts, but
in this case the carbenium ions have to be formed in asterically demanding, bi-molecular hydride transfer [9].
Large pore zeolites like Y zeolite show usually a greater
tendency to crack the hydrocarbons according to the
classical cracking mechanism. However, the small and
medium pore zeolites like ZSM-5 favour the non-classi-
cal Haag-Dessau mechanism which allow mono-molecu-
lar reactions while restricting the bi-molecular (hydride
transfer) reactions due to steric limitations in the pores.The Haag-Dessau mechanism (see Fig. 6) is the key to
unravel the competing mechanisms of catalytic cracking,
including the classical cracking and oligomerisation
Fig. 5. b-Scission mechanism for secondary and tertiary alkylcarbenium
cracking. Understanding of non-classical cracking hashelped in the diagnosis of shape-selectivity and mass
transfer effects in zeolite-catalysed cracking [6].
From the work of Olah concerning the hydrocarbon
chemistry in superacids it was known that alkanes can
be protonated at low temperatures in liquid phase.
However, Haag and Dessau postulated their mechanism
in 1984 by demonstrating that even zeolites can proton-
ate alkanes to give carbonium ions—which are transi-tions states in cracking [10]. The carbonium ions
collapse to give the cracking products: to begin with alk-
anes (or hydrogen) and smaller carbenium ions, which
then release protons to form the final cracking product
ions [1,8]. Reproduced by permission of Wiley-VCH, Weinheim.
Fig. 7. Schematic reaction pathways for carbonium ion decay (protolytic cracking) of a protonated 3-methylpentane together with principle
transition states for dehydrogenation and cracking [1]. Reproduced by permission of Wiley-VCH, Weinheim.
M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292 261
consisting of alkenes [6]. Haag and Dessau�s suggestionis based on the carbonium ion decay of a protonated
3-methylpentane molecule, as shown in Fig. 7.Since the decay of the carbonium ion leads to break-
ing of the C–C or C–H bonds at the insertion point of
the proton, the reaction is called protolytic cracking as
well [1]. Furthermore, compared to the super acid chem-
istry in liquid phase, the formation of the carbonium
ions using zeolites is only significant at temperatures
higher than 450 �C, and they exist only in a transitionstate.In conclusion, Haag-Dessau cracking (also called
mono-molecular or protolytic cracking) dominates at
low conversions, high reaction temperatures, low reac-
tant pressures and with small and medium pore zeolites
having a low concentration of Brønsted acid sites. All
these conditions favour a low reactant concentration
in the pores and impede hydride transfer. The decay of
the carbonium ion into an alkane and a smaller carbe-nium ion is the main step in the Haag-Dessau cracking
mechanism [1].
Fig. 8. Simplified reaction network for the cracking of alkanes on zeolites
A simplified reaction network for the catalytic crack-
ing of alkanes using zeolites is shown in Fig. 8 [11].
At higher reactant partial pressure the classical crack-ing mechanism is gradually replaced by oligomerisation
cracking, where we observe substantial oligomerisation
preceding the cracking process. Experimental evidence
for such a route has been demonstrated by Werst et al.
[12] using labeling investigations, and revealing entire
scrambling of carbon-labeled olefinic cracking products.
The importance of this mechanism increases with higher
conversion and higher partial pressure as well as lowerreaction temperatures. However, the fundamental chem-
istry related to this cracking mechanism is basically the
same as observed for the classical cracking mechanism
[1].
2.4. Mechanistic pathway for the alkylation of isobutane
with n-butene on zeolites (C–C bond formation)
Zeolites and related microporous materials are also
used as catalysts for the formation of carbon–carbon
[11]. Reproduced by permission of Imperial College Press, London.
262 M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292
bonds, like the very demanding alkylation of isobutane
with n-butene. The reaction pathway of the isobutane/
n-butene alkylation is quite complicated since several
competing reactions can take place besides the regular
alkylation, like self-alkylation, destructive alkylation,
multiple alkylation as well as oligomerisation and crack-ing. Due to the lower reaction temperature for the ali-
phate alkylation (thermodynamically favoured), the
desorption step is often difficult in these reactions, since
the reaction product is often more strongly adsorbed
than the reactants. However, the obtained product mix-
ture is an excellent blending component for gasoline,
and the reaction is industrially carried out applying
either HF or sulfuric acid as acid catalysts. There is anintense search looking for attractive alternatives for
those acids, with large pore zeolites (among other solid
Fig. 9. Mechanism of aliphate alkylation [4]. Repro
catalysts) as promising candidates, however, so far these
systems suffer from an unsuitable catalyst lifetime
[1,4,13–18].
The mechanism of the aliphate alkylation can be de-
scribed as follows: The reaction is initiated by the addi-
tion of a proton to the n-butene, forming the secondarybutyl-(2) cation, which abstracts a hydride ion from
isobutane forming tertiary butyl cations. These tertiary
butyl cations interact with n-butene forming isooctyl
cations (preferentially the high octane number repre-
senting trimethylpentanes). The isooctyl cations capture
hydride ions from isobutane forming isooctanes and
tertiary butyl cations, which then continue the reaction
cycle (see Fig. 9).The lifetime of the large pore zeolites (FAU, BEA
and EMT) is determined by the relative rates of hydride
duced by permission of Elsevier, Amsterdam.
M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292 263
transfer and n-butene addition. The latter is usually con-
trolled by using back-mixed reactors operating at
high conversions and low alkene concentrations. The
hydride transfer rate is depending on the stability of
the carbenium ion and the space limitations given by
the microporous framework of the zeolite. The slowerthe hydride transfer the more multiple alkylation takes
place, forming C12 and C16 units, which block the active
sites as well as the micropores and deactivate the cata-
lyst [1].
Typical side reactions are classical cracking and oli-
gomerisations over weak Brønsted acid sites, producing
larger olefins, which participate in alkylation and lead to
the formation of larger alkylate molecules, which againcontribute to increased deactivation of the catalyst [1].
The coke is formed through a combination of hydride
transfers, inter- and intra-molecular alkylation reactions
and oligomerisations leading to heavy, unsaturated cyc-
lic and acyclic compounds [4].
2.5. Shape selectivity of acid zeolites and related
microporous materials
In 1960, Weisz and Frilette [19] introduced the
expression ‘‘Shape-Selective Catalysis’’ by demonstrat-
ing that Ca A zeolite dehydrated 1-butanol at 260 �Cbut not isobutanol. This observation showed that the
conversion took place inside the microporous structure
of the Ca A zeolite (0.5 nm pore diameter), not available
for the branched isobutanol due to its large kineticdiameter. This size exclusion model has since then been
used to remove linear hydrocarbons from mixtures con-
taining both branched and linear hydrocarbons.
The mechanisms of molecular shape-selective cataly-
sis can be described and summarised as follows:
Reactant selectivity describes the phenomenon of
microporous catalysts acting as molecular sieves and
excluding bulky molecules from entering the intra-crys-talline void-structure while allowing smaller molecules
to enter. The critical exclusion limit can be varied over
a wide range of different zeolites and related micropo-
rous solids [4].
Product selectivity refers to discrete diffusivities of the
reaction products formed with respect to the micropo-
rous pore architecture and crystal size of the catalyst
particles. Sterically less hindered product moleculesmay easily leave the microporous framework, whereas
bulky product molecules may stay much longer in the
cavities of the zeolites. The term molecular traffic control
has been coined by Derouane and Gabelica in 1980
describing qualitatively the transport of molecules
with different shape and/or size in the microporous
framework of zeolites with two discrete sets of pores
[4,20].Restricted transition state-type selectivity occurs when
the spatial configuration around a transition state or a
reaction intermediate located in the intra-crystalline
volume is such that only certain configurations are pos-
sible. This means the formation of reaction intermedi-
ates and/or transition states is sterically limited due to
the shape and size of the microporous lattice allowing
the access of the species formed to interact with the ac-tive sites. This type of selectivity was first proposed by
Csicsery [21] and is usually connected to the suppres-
sion of undesired side reactions like coke formation.
Whereas the product selectivity depends on the crystal
size of the catalyst the restricted transition state-type
selectivity is not depending on the relative rates of dif-
fusion and reaction, hence both selectivities can easily
be distinguished by changing the crystal size of thecatalyst [22].
The different types of shape-selectivities of zeolites
and related microporous materials are summarised in
Fig. 10.
Zones et al. [23] introduced the term inverse shape-
selectivity for those cases where the restricted transition
state-type selectivity arises from a positive discrimina-
tion of specific transition states. An example is the skel-etal multiple branching of n-hexane in large pore zeolites
or related microporous solids, where the highest selectiv-
ity for multiple branched isohexanes was registered for
microporous solids with well defined pore diameters
and with an optimum interaction with the desirable
isomers [4]. See also Fig. 11.
The cage or window effect, observed in connection
with the hydrocracking of long n-alkanes, represents acertain case of molecular shape-selectivity in zeolites.
This effect is in operation when the diffusivities and/or
reactivities do not change monotonically within a
homologous series of compounds, due to the fact that
certain cracking products, which fit the cage dimensions
and are trapped in the cage of the zeolites, were not ob-
served as products [4]. Gorring reported for the first
time this effect in connection with the hydrocrackingof hexadecane using erionite [24]. Only small amounts
of C7–C9 alkanes were observed (although representing
the central cracking products of the probe molecule
and detected in large amounts when cracking hexade-
cane without using a shape-selective catalyst), since they
fit excellently within the cage of erionite [4].
Selective reactions at the pore mouth of zeolites have
been observed by Martens et al., for example the long-chain n-alkane isomerisation over Pt/H ZSM-22 (TON
structure) revealing large amounts of mono-branched
isomers although these isomers cannot desorb from
the narrow channels of this mono-dimensional zeolite
[25–28]. This observation has been termed pore mouth
catalysis since the product pattern is explained by
involving only the acid sites at the entrance of the small
pore zeolites such as Theta-1, ZSM-22 and erionite. Inaddition, the second branching of n-alkanes over Pt/H
ZSM-22 was shown to occur at approximately the
Fig. 11. Inverse shape selectivity observed for the skeletal multiple branching of n-hexane in large pore zeolites. AFI microporous solids represent the
optimum pore diameter [4]. Reproduced by permission of Elsevier, Amsterdam.
Fig. 10. Schematic representation of the three types of shape-selectivity. Concerning the restricted transition state-type selectivity: the lower
transition state molecule is easier to accommodate in the cavities than the upper one [1]. Reproduced by permission of Wiley-VCH, Weinheim.
264 M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292
distance between the pore openings at the surface of thezeolite crystallites [22]. This phenomenon has been
termed key-lock catalysis [26].
A schematic presentation of the latter type shape-selective effects in zeolites and related microporous sol-
ids is given in Fig. 12.
Fig. 12. Shape-selective environments in different zeolite structure types: (a) large molecules have access to interrupted cavities and channel
intersections for pore mouth catalysis; (b) molecules are plugged into the pore aperture; (c) molecules are converted in multiple pore mouths
according to key-lock catalysis; (d) molecules are converted in the intra-crystalline shape-selective environment [4]. Reproduced by permission of
Elsevier, Amsterdam.
M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292 265
3. Hydrocarbon conversion and formation, crude oil and
natural gas upgrading
3.1. World market of zeolites and related microporous
materials with focus on their application as catalysts
Zeolites and related microporous solids are impor-
tant components of adsorbents and catalysts used inconnection with the upgrading of crude oil and natu-
ral gas, among others due to the fact that they are
easily separated from the educts and products and
by representing a clean technology compared to their
predecessors.
The world market of zeolites and related micropo-
rous solids is still in a period of strong development.
Currently about 1.6 millions of tons are used per year,of which about 1.3 millions of tons refer to synthetic
zeolites and about 0.3 millions of tons to natural zeo-
lites, the latter mainly applied as adsorbent and ion ex-
changer [29,30].
Concerning the application of synthetic zeolites and
related microporous materials, the focus in terms of
amounts is definitely on detergent builders (1.05 millions
of tons per year), followed by catalysis (0.15 millions oftons per year) and finally adsorption (0.1 millions of
tons per year). A-type zeolites are by far the most com-
TON (Theta-1, ZSM-22) Long-chain paraffin isomerisation
AEL (SAPO-11) Long-chain paraffin isomerisation
266 M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292
depending on which country or part of the world they
are serving. Consequently, the process facilities at differ-
ent refineries can vary, however many processes can be
implemented. The market demand is mainly related to
gasoline, diesel, kerosene and fuel oils, which have to
meet specifications which are requested by national gov-
ernments or the European Commission [34]. As an
example, the most advanced category of the car indus-try�s World-Wide Fuel Charter (WWFC) concerninggasoline specifications is summarised in Table 2 [35].
The quality of gasoline is usually defined by the mo-
tor octane number (MON, built up with isoparaffins and
ethers and reduced by the presence of alkenes) and the
research octane number (RON, obtained through the
presence of aromatics and ethers). Diesel is mainly char-
acterised by the cetane number, which is linked to thepresence of alkanes. The cetane number is decreased
by the presence of higher aromatics [34].
World-wide are about 40 million barrels (one barrel
corresponds to 159 l) crude oil refined very day in the
refineries. The basic processes for refining crude oil are
still the same but the tendency is in the direction of more
complex process technology. To begin with crude oil is
divided into various fractions by atmospheric distilla-
Table 2
Most advanced gasoline specifications proposed in the World-Wide
Fuel Charter (WWFC) [35]
Parameter
Low octane gasoline [(MON + RON)/2] 86.8
High octane gasoline [(MON + RON)/2] 93.0
Oxygen (wt.% max.) 2.7
Benzene (vol.%) 1.0
Aromatics (vol.%) 35.0
Olefins (vol.%) 10.0
Sulfur (ppm) 5–10
tion. During this procedure the main fractions of oil
products are obtained, which cover
Liquefied petroleum gas (LPG)
C1–C4 cut
Naphtha
C5 to about 180 �C Middle distillates 130–300 �C Diesel/gas oil 150–370 �C Lube base oils/atm. residue higher than 370 �C
The residue of the atmospheric distillation can be
used as feed for a vacuum distillation, leading to thefractions termed vacuum gas oil (VGO, 370–540 �C)and vacuum residue (higher than 540 �C).All these fractions are the primary oil products,
however, many of them have to be upgraded before
they can meet the requested specifications and be used
as commercial products [36]. In the case of large
amounts of gasoline to be produced, a FCC unit will
be installed (feed stocks: VGO, residues). Furthermore,polymerization of the light alkenes obtained or their
alkylation with isobutane is implemented in the refin-
ery. Isomerisation of C5/C6 n-paraffins may be installed,
if there is a need for an increase in the octane number
(getting branched paraffins). If there is a stronger need
for diesel oil a hydrocracker unit may be added, and if
the residue is too viscous to be handled, a visbreaker
unit may be requested. The refineries apply a lot of cat-alytic units in order to upgrade, to convert or to purify
their product streams, only visbreaking and coking are
thermal processes. Only a few catalytic units use liquid
catalysts, like the isobutane/n-butene alkylation (HF or
H2SO4), the conversion of mercaptans into disulfides
(Co-phthalocyanines) or alkene dimerisation (Ziegler–
Natta type catalysts). All other catalysts are solids,
and the processes applying zeolites or related micropo-rous materials will be presented and discussed in the
following sub-chapters [34]. However, a presentation
Fig. 13. Thermodynamic equilibrium for hexane isomerisation [36].
Reproduced by permission of Elsevier, Amsterdam.
M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292 267
of a modern refinery is not complete without taking
into account the need of the petrochemical industry
which requests synthesis gas, olefins and aromatics
(benzene, toluene and xylenes—BTX) as basic feed
stocks. Olefins and aromatics are usually manufactured
from naphtha via steam cracking [36], however, theincreasing demand of propene can only be satisfied by
using other technologies like deep catalytic cracking
(DCC) and dehydrogenation of propane in addition
to the steam cracker production. Furthermore, a new
route has been introduced by Weitkamp et al. convert-
ing surplus aromatics from pyrolysis gasoline into high
value steam cracker feed (ethane, propane and n-bu-
tane) through a hydrodecyclisation step using a bi-func-tional Pd/H-ZSM-5 catalyst [37]. This technology
combines a considerable decrease of the requested ben-
zene gasoline content with an improved utilisation of
surplus aromatics by increasing the yields and selectiv-
ities of ethene and propene in the steam cracking pro-
cess and represents a real break-through with respect
to high value basic products within refinery and
petrochemistry.The stagnation of crude oil reserves and the increase
of their prices have recently driven the attention towards
the production of fuels and chemicals from natural gas
via synthesis gas (syngas). This route is also known as
the ‘‘gas to liquids (GTL)’’-technology. Fuels produc-
tion from syngas (in former times obtained from coal)
has been reported by Fischer and Tropsch in 1923 for
the first time [38], using an alkali-promoted iron cata-lyst. Fuels manufactured via the Fischer–Tropsch route
reveal an excellent quality since they consist mainly of
linear paraffins and a-olefins and do not contain sulfurand aromatics. A Co-containing catalyst is applied for
the production of heavy paraffins via the Fischer–Trop-
sch route starting with natural gas, a technology devel-
oped by Shell and named the ‘‘Shell Middle Distillate
Synthesis (SMDS)’’ route [39,40]. Finally, diesel fuel(or gasoline) is produced by hydrocracking of the more
or less sulfur and nitrogen-free wax obtained through
the SMDS process using noble metal containing zeolites.
The more restricted fuel specifications currently intro-
duced in order to reduce the environmental impact of
hazardous emissions represent a driving force with
respect to an increased use of fuels prepared via the
Fischer–Tropsch route as a blending component of thegasoline and diesel pools in the future [39]. Besides
the SMDS technology an alternative has been presented
by SASOL/Chevron termed as the ‘‘Slurry-Phase-Distil-
late’’ process, again based on the Fischer–Tropsch route
producing wax (using a Co-containing catalyst) fol-
lowed by a hydrocracking step in order to get diesel or
gasoline [41]. Finally, besides the dehydrogenation of
propane and the DCC process other alternatives haverecently been introduced with respect to meet the
increasing demand of propene, like the ‘‘Methanol to
Propene (MTP)’’ process of Lurgi applying H-ZSM-5
based catalysts [42].
The following sub-chapters will focus on the current
technology and future developments related to different
processes dealing with upgrading of crude oil and natu-
ral gas and applying zeolites or related microporoussolids as catalysts.
3.3. Isomerisation of n-paraffins
3.3.1. C5/C6 isomerisation (light straight
run—LSR—naphtha)
Environmental restrictions have caused the phase-out
of lead additives and the elimination (or lowering) ofbenzene from gasoline, with the consequence of an in-
creased demand of isoparaffins in order to improve the
octane number of gasoline. Therefore, isomerisation of
light straight run naphtha, containing C5/C6 n-paraffins,
has advanced to be an important process in the oil
refinery. Skeletal isomerisation of n-paraffins is an
acid-catalysed and equilibrium limited reaction, which
is thermodynamically favoured at lower temperatures,see Fig. 13 [39].
Industrially, the C5/C6-isomerisation is performed
using a bi-functional catalyst (noble metal together with
an acidic carrier) and in the presence of hydrogen. The
main advantage of applying a bi-functional catalyst is
that stable operations are possible under a sufficiently
high hydrogen pressure. The mechanism of this reaction
is well accepted, and can be summarised as follows (seealso Fig. 14):
1. The n-paraffins are initially dehydrogenated on the
noble metal sites to give the corresponding n-alkenes.
2. The n-alkenes are then protonated at the acid sites,
resulting in carbenium ions.
Fig. 14. Mechanism of the n-paraffin isomerisation [36]. Reproduced by permission of Elsevier, Amsterdam.
268 M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292
3. The carbenium ions undergo a skeletal isomerisation
to form more stable branched carbenium ions, fol-
lowing the protonated cyclopropane (PCP) interme-
diate mechanism suggested by Brouwer [43].
4. Finally, the branched carbenium ions are hydroge-
nated at the noble metal sites and desorb as isoparaf-
fins [39].
Fig. 14 suggests that isomerisation and cracking may
occur as parallel reactions in the presence of hydrogen,
in addition to sequential cracking of pre-isomerised
compounds and post-isomerised cracking products.
Since cracking and isomerisation are both catalysed by
similar acid catalysts, it is not surprising that cracking
takes place besides isomerisation. Skeletal n-paraffin
isomerisation requires at least a chain of four carbon
atoms, whereas cracking requests a minimum of seven
atoms in the carbon chain. That means that pentanes
and hexanes easily can be isomerised but not easily
cracked. For paraffins higher than hexanes, cracking is
usually a competing reaction to skeletal isomerisation,
resulting in a lower selectivity for the isomerisation reac-
tion, in spite of the fact that isomerisation is carried out
at lower temperatures than cracking [44]. Weitkampdemonstrated the different results on isomerisation and
cracking of C6–C10 n-paraffins using a bi-functional Pt/
Ca zeolite Y. See Fig. 15 [45].
Commercial isomerisation catalysts contain both a
noble metal (Pt) based hydrogenation-dehydrogenation
function and an acid function. The acid function is pro-
vided by either a halogenated (Cl, F) alumina carrier, a
sulfated zirconia substrate or by a zeolite (usually mord-
Fig. 15. Comparison of the hydroisomerisation and hydrocracking of
C6–C10 n-paraffins over Pt/Ca zeolite Y [45]. Reproduced by permis-
sion of Elsevier, Amsterdam.
M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292 269
enite (MOR)). A zeolite omega (MAZ structure type)
based catalyst has been demonstrated to be superior to
the mordenite based system (higher activity and better
selectivity), however, no commercial use has been re-
ported so far. The higher yield achieved by using a Pt/
H-MAZ catalyst has been reported to be due to both
the unique structural properties of MAZ and the higher
acid strength of its Brønsted sites [46]. Besides mordenite(MOR) and zeolite omega (MAZ), Pt/Beta zeolites
(BEA) have been investigated as isomerisation catalysts
as well, however, the Pt/H-MOR is the only system
which has been commercialised so far [39].
The halogenated alumina and sulfated zirconia based
isomerisation catalysts are more strongly acidic than the
zeolite based catalysts, which means that they can
isomerise the LSR naphtha at temperatures below150 �C. Consequently, this favours the formation ofthe desired isoparaffins due to the thermodynamic con-
ditions. On the other hand, the tolerance level of the
non-zeolitic catalysts against water and sulfur is not
high, and this leads to a fast deactivation of the haloge-
nated alumina and sulfated zirconia based isomerisation
catalysts and to the request of a severe pre-treatment of
the naphtha feed. Finally, a continuous stream of halo-gen has to be added in order to keep the catalyst active,
leading to corrosion problems for the reactor system
[39]. The zeolite based isomerisation catalysts are also
lacking sulfur tolerance, however, to a much lesser ex-
tent [46].
The Pt/H-MOR catalysts are less acidic than the
Pt/halogen–alumina catalysts, and, consequently, they
have to be applied at higher reaction temperatures
(about 250 �C), which limits the formation of isoparaf-fins due to the thermodynamic conditions. However,
they are more robust than the non-zeolitic catalystsand can withstand low levels of impurities such as sulfur
and water in the feed [36,39].
The most known example of a zeolitic isomerisation
catalyst is the Pt/H-MOR catalyst, first developed for
the Shell Hysomer process. The process operates at
27–30 bar hydrogen pressure and a reaction temperature
of 250 �C. At this temperature, not all normal paraffinscan be converted to branched paraffins (see also Fig. 13).Therefore, it seems to be attractive to combine the
Hysomer isomerisation process with the ISOSIV iso/
normal paraffin separation process (using a Ca A zeolite
as selective adsorbent for the n-paraffins), commercia-
lised by Union Carbide (now UOP). The combined tech-
nology is known as ‘‘total isomerisation process (TIP)’’
and commercialised by UOP (see Fig. 16). The octane
gain in the TIP process is reported to be in the rangeof about 10 octane numbers [46]. UOP�s zeolitic isomeri-sation catalyst (Pt loaded mordenite) is commercialised
under the trade name HS-10.
The stability of the Hysomer catalyst is also demon-
strated by its long lifetime: catalyst charges have been
used for up to seven years in commercial operation, how-
ever, a catalyst deactivated by operational mishaps can in
most cases be regenerated by a simple coke burn-off [44].An important parameter controlling the isomerisa-
tion activity and selectivity is the framework Si/Al ratio
of the zeolite. A maximum of activity for n-pentane
isomerisation was observed for a Si/Al ratio of about
10, for which all framework aluminium atoms are iso-
lated, and therefore, supporting the strongest frame-
work Brønsted acid site possible within the mordenite
structure [47–49]. Increasing the lattice Si/Al ratio bydealumination is beneficial as well with respect to cata-
lyst deactivation by reducing the coking rate [50,51].
Dealumination by acid leaching decreases the number
of Brønsted acid sites and creates mesoporosity inside
the zeolite crystallites, which again leads to a decrease
of the diffusion limitations. The mesoporosity causes
shorter residence times and facilitates easier desorption
of the products, avoiding secondary reactions of theintermediates formed and improving the overall selectiv-
ity [46]. Besides the final lattice Si/Al ratio, the method
of dealumination applied has an influence on the cata-
lyst activity: steam dealumination leaves the lattice
aluminium removed in extra-framework positions
(EFAl), whereas acid leaching results in almost EFAl-
free samples. In addition, acid leaching forms alumin-
ium gradients along the zeolite crystallites, whereassteam treatment produces a more uniform aluminium
distribution. The most active Pt/H-MOR has been
Fig. 16. Paraffin Total Isomerisation Process (TIP) [36]. Reproduced by permission of Elsevier, Amsterdam.
270 M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292
obtained by dealuminating the zeolite through acid
leaching followed by mild steaming, leading to small
and controlled amounts of EFAl which create a syner-
getic effect on the Brønsted acid sites associated toframework aluminium (FAl), improving their overall
acid strength [39].
CEPSA and Sud-Chemie AG have commercialised
their isomerisation catalyst based on a strongly acidic
Pt/H-mordenite catalyst under the trade name HYSO-
PAR [46].
3.3.2. Isomerisation of long-chain n-paraffins
There is a strong interest to extend the isomerisation
reaction to n-paraffins containing carbon chains longer
than C6, mainly with respect to produce higher octane
multi-branched isomers. However, as mentioned earlier,
one has to take into account that the cracking tendency
of branched paraffins increases with the length of the
hydrocarbon chain and with the degree of branching.
As an example, Pt/H-MOR produces low yields of iso-mers in connection with n-heptane isomerisation due
to extensive cracking of the isoheptanes formed [11,39].
Furthermore, isomerisation of long-chain n-paraffins
has been used to improve the pour point, viscosity,
cloud point and freeze point of middle distillates and
lube oils. In this respect very good catalytic perfor-
mances have been observed for the n-heptane isomerisa-
tion using Pt supported on nano-crystalline Beta zeolite[52]. This experimental result is explained by a combina-
tion of Brønsted acid sites of lower acid strength than in
mordenite and a faster diffusion of the branched isomers
through the small crystallites (10–20 nm) of the nano-
crystalline Beta zeolite, leading to a decrease in the
cracking rate [11].
In addition, besides high isomerisation selectivities
combined with low cracking rates, the degree of branch-
ing should be minimized in order to keep a high quality
of the paraffinic product. In this respect, Pt/SAPO-11 (amedium pore sized microporous solid) has been demon-
strated to display high isomerisation selectivity and low
yields to multi-branched species in the n-octane isomeri-
sation [53]. The suppression of the formation of multi-
branched isomers on Pt/SAPO-11 catalysts has been
interpreted in terms of a transition state shape-selective
effect induced by the uni-dimensional pore structure of
SAPO-11 [39].However, in the domain of middle paraffin isomerisa-
tion (C7–C9 carbon chain length) there is still a need for
catalyst improvement in order to improve a thorough
isomerisation selectivity (two branches or more) while
minimizing the cracking rate [30].
Isomerisation of higher alkanes in the wax range is
going to play a much more important role in the future
since moderately branched alkanes formed by isomer-ised/cracked wax represent excellent components for
lube oils. The isomerisation and cracking of synthetic
wax produced through the Fischer–Tropsch route repre-
sent an excellent alternative for the manufacture of high-
quality fuels which are currently derived from petroleum
[44].
3.3.3. Isomerisation of n-butane
Isomerisation of n-butane can take place following
two different mechanisms, either via a direct isomerisa-
tion involving the formation of a highly unstable
primary carbenium ion (mono-molecular) or via a
dimerisation-cracking mechanism (bi-molecular). The
isomerisation of n-butane is even more thermodynami-
M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292 271
cally limited than the C5/C6 n-paraffin isomerisation.
When running the reaction at low temperatures, very
strong acid sites are requested. In fact, commercial n-
butane isomerisation processes make use of HCl/AlCl3(Phillips [54], Shell [55]) or Pt supported on chlorided
alumina as in the Butamer technology by UOP [56].Alternative solid acid catalysts—more beneficial with re-
spect to the environment—have been studied, among
others sulfated zirconia, heteropolyacids (in particular
12-tungstophosphoric acid) and zeolites. Sulfated zirco-
nia, however, deactivates very fast. Pt/H-mordenite is
active for n-butane isomerisation, however, this catalyst
requires higher reaction temperatures than sulfated zir-
conia or heteropolyacids in order to achieve reasonableconversions. In addition, the zeolite based catalyst forms
larger amounts of different by-products than the non-
zeolitic catalysts [11,39].
The present n-paraffin isomerisation capacity world
wide is roughly split equally between n-butane isomeri-
sation, LSR naphtha isomerisation using Pt/H-MOR
and LSR naphtha isomerisation over Pt/Cl/alumina
[44,57].
3.4. Skeletal isomerisation of light n-alkenes
C4 and C5 alkene skeletal isomerisation has been re-
garded as a suitable alternative for increased production
of isobutene and isopentenes. These isoalkenes are
mainly obtained from FCC units or steam crackers
and are used for the production of methyl tert-butyl
Fig. 17. Reaction mechanism for the acid catalysed skeletal isomerisation o
London.
and tert-amyl methyl ethers (MTBE and TAME), which
represent excellent fuel oxygenates with good octane
blending properties, in spite of the fact that MTBE
has been questioned as fuel additive for environmental
reasons [58]. In addition, isobutene is also an important
reactant for the petrochemical industry [39].C4 and C5 alkene skeletal isomerisation is an acid
catalysed reaction requesting strong acid sites. The reac-
tion mechanism can be described by the initial double
bond cis–trans isomerisation taking place at the acid
sites before skeletal isomerisation. The protonation of
the double bond leads to the formation of a secondary
carbenium ion, which then rearranges into a protonated
cyclopropane (PCP) structure, and finally ending upwith an isoalkene via the formation of an unstable pri-
mary carbenium ion in the case of n-butene, as shown
in Fig. 17 [11].
For thermodynamic reasons, low isomerisation tem-
peratures should be applied, since the equilibrium con-
centration of branched olefins decreases with increasing
temperature. However, at low temperatures the selectiv-
ity to isoalkenes decreases due to the competing olefinoligomerisation reactions. The extent of these side reac-
tions can be lowered by applying higher reaction temper-
atures and low alkene partial pressure, however, at
higher temperatures other non-desired reactions take
place, like cracking, hydrogen transfer and coking, lead-
ing to the catalyst deactivation [59].
In former times different solid acid catalysts have been
applied as light alkene isomerisation catalysts, like metal
f n-butenes [11]. Reproduced by permission of Imperial College Press,
272 M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292
halides and supported phosphoric acid. Later, environ-
mentally more friendly zeolite catalysts have been intro-
duced successfully, with medium-pore size zeolites as the
most promising systems since the competing (bi-molecu-
lar) oligomerisations can be lowered within the limited
space available in the smaller channels of these micropo-rous solids. Among the medium pore sized zeolites those
having a one-dimensional channel structure (like ZSM-
22 and Theta-1, both having the TON structure, as well
as ZSM-23) turned out to be the best catalysts. They gave
the highest yield of isobutene by working at relatively
low temperatures (below 400 �C) and with a low olefinpartial pressure. The selectivity to isobutene has been
shown to improve be decreasing the density of theBrønsted acid sites (increasing Si/Al ratio) and by
decreasing their acid strength by replacing lattice alumin-
ium by other trivalent cations, like gallium and iron. This
behaviour is explained by suppression of bi-molecular
reactions leading to side-products as the acid sites be-
come more and more isolated, and can be regarded as
indirect evidence that isobutene is mainly formed via a
mono-molecular mechanism, as shown in Fig. 17 [11,39].Finally, a ferrierite based catalyst has been intro-
duced by Shell, giving high yields of isobutene at
350 �C and long catalyst lifetimes [60]. This behaviourhas been attributed to the particular structure of ferrie-
rite possessing intersecting 10- and 8-membered ring
channels, which induce the selective formation of tri-
methyl pentene dimers and their cracking into C4 frag-
ments, including isobutene [61].In conclusion, quite active and selective alkene isom-
erisation catalysts can be prepared from microporous
solids by combining the proper pore architecture with
the presence of isolated and/or mild Brønsted acid sites
[39].
3.5. Aliphate alkylation
Aliphate alkylation in connection with oil refinery re-
fers to a technology dealing with the partial conversion
Fig. 18. Conversion of a liquid isobutane/1-butene mixture on a CeY zeolite.
ratio = 11, pressure: 31 bar) [63]. Reproduced by permission of Wiley-VCH,
of the C4 cut (isobutane and n-butenes) into the so-
called alkylate. This alkylate represents a very high qual-
ity and valuable component for the refinery gasoline
pool due to the high octane numbers of the formed iso-
octanes, preferentially represented by the trimethyl-
pentanes. Current alkylation technology applies eitherhydrofluoric (UOP and Phillips) or sulfuric acid (Stracto
and Kellogg). These traditional processes suffer from
several safety risks and drawbacks, like the high toxicity,
volatility and corrosiveness of hydrofluoric acid and the
high catalyst consumption of sulfuric acid, which re-
quires a regeneration plant close to the alkylation facil-
ity. Therefore, replacement of the existing alkylation
processes by new technology based on non-toxic, non-corrosive and environmentally friendly solid acid cata-
lysts is one of the most important research challenges
in the field of heterogeneous catalysis. Significant re-
search during the last two decades resulted in a variety
of different solid acids capable to produce an alkylate
with the same quality features as the conventionally
manufactured product. However, none of the alternative
solid catalysts have been commercially applied so far,mainly due to the short catalyst lifetime caused by the
decline of their hydride transfer activity [14,39].
The mechanism of aliphate alkylation has been de-
scribed in Section 2.4 for the case of isobutane/n-butene
alkylation. Due to typical side reactions like oligomeri-
sation and classical cracking the question arises as to
how long the formed product can be considered as an
alkylate. Arbitrarily, Weitkamp et al. placed this limitat a content of alkanes in the C8 product fraction of
90 mol.% [62]. In the example shown in Fig. 18, the alkyl-
ation stage then ends after about half an hour. This illus-
trates that the time-on-stream (TOS) behaviour of the
solid catalysts is still unsatisfactory—due to the decline
of the hydride transfer activity. In addition, the ratio be-
tween the trimethylpentanes (TMP) and the dimethylhex-
anes (DMH) formed can be taken as a measure of thealkylation/oligomerisation ratio for a certain solid acid
catalyst [11]. However, future research within this sub-
Composition of the C8 fraction (temperature 80 �C, isobutane/1-buteneWeinheim.
M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292 273
field of acid catalysis should concentrate on the reasons
for this loss of hydride transfer activity and measures to
extend the solid catalyst lifetime considerably [14].
In the following, the most recent advances related to
the use of solid acid catalysts for the aliphate alkylation
will be summarised, with focus on the zeolites. An excel-lent review covering this sub-field of heterogeneous
catalysis has recently been written by Weitkamp and
Traa, including the early studies with respect to the
search for solid alkylation catalysts [14].
Large-pore zeolites, like FAU, EMT and BEA, have
been in the focus concerning their application as ali-
phate alkylation catalysts. Extensive studies of the
behaviour of rare earth exchanged faujasites (RE-Y)were performed by Weitkamp, where he combined a
sampling system downstream of the reactor with high-
resolution GC in order to obtain detailed results on
the olefin conversion and product yields as a function
of time on stream [63,64]. With Ce–Y as catalyst he ob-
tained 100% butene conversion and a high quality alkyl-
ate during the first half hour on stream, after that the
activity decreased rapidly and the selectivity changedfrom the alkylate to the oligomerisate (C8 and C12 ole-
fins) as the zeolite became deactivated [62].
Corma et al. found a maximum initial conversion of
2-butene for USY zeolites with different unit cell sizes
(which means different framework compositions) cover-
ing a0 values between 2.435 and 2.450 nm [15]. The
TMP/DMH ratio continuously increased with the unit
cell parameter and the hydride transfer activity washigher for the mildly dealuminated samples [39]. In addi-
tion, the extra-framework aluminium (EFAl) formed
during the steam dealumination influenced the catalytic
performance of USY zeolites in the aliphate alkylation
as well [65].
Cardona et al. investigated the aliphate alkylation
using an USY zeolite at 50 �C, concluding with theobservation that their reaction pathway is in line withthe mechanism discussed in Section 2.4 [66]. Gardos
et al. applied rare earth exchanged Y zeolites in the ali-
phate alkylation using a batch-type autoclave and reac-
tion temperatures between 50 and 100 �C, arriving at theconclusion that the composition of the alkylate is
entirely controlled by the kinetics [67,68].
EMT, the hexagonal faujasite has been investigated
as aliphate alkylation catalyst by Stocker et al., usinga stirred-tank reactor and a reaction temperature of
80 �C [69–71]. La-H-EMT with a La3+ exchange degreeof 40% was observed to reveal the best alkylation perfor-
mance, followed by H-EMT and H-FAU and Ce Y zeo-
lite. The better alkylation performance of EMT was
ascribed to a higher strength of the Brønsted acid sites
and slightly larger cages in EMT as compared to the
FAU zeolite [39].Zeolite Beta (BEA), another large-pore zeolite with a
three-dimensional pore structure, has been used as cata-
lyst for the aliphate alkylation [72–76]. As for Y-type
zeolites, zeolite Beta deactivated rapidly as well [72].
Corma et al. demonstrated in their more in-depth stud-
ies that the performance of Beta as aliphate alkylation
catalyst depends on the synthesis recipe, the crystallite
size, the chemical composition, the post-preparationtreatments, the nature and amount of the extra-frame-
work Al and the density and strength of the Brønsted
acid sites [73,74]. As an example, it was concluded from
these studies that H-Beta zeolite prepared from tetraeth-
ylorthosilicate (TEOS) was more active than that syn-
thesised from amorphous silica, and that samples with
crystal sizes of 0.35 lm were more active than those of0.1 lm [39,73,74]. Finally, Kiricsi et al. [75] and Flegoet al. [76] studied La H-Beta zeolites and the nature of
the carbonaceous deposits formed after adsorption and
conversion of isobutene/1-butene mixtures.
Twelve-membered ring zeolites other than FAU,
EMT and BEA have only found limited attention as ali-
phate alkylation catalysts, as for example, ZSM-4 (zeo-
lite Omega), ZSM-20 (inter-growth between FAU and
EMT), ZSM-3, ZSM-18 and mordenite [14].Medium pore zeolites, like ZSM-5 and ZSM-11, have
also been investigated as aliphate alkylation catalysts,
however, these zeolites were found to be active for the
alkylation only at temperatures higher than 100 �C,which is not of interest from a thermodynamic point
of view [77,78]. In addition, these medium pore zeolites
produced less trimethylpentanes, indicating serious pore
restrictions for the formation of the desired alkylate [39].MCM-22 has been investigated, however, its behaviour
as aliphate alkylation catalyst was found to be in be-
tween those of 10– and 12-MR systems [72,79].
Non-zeolitic systems, like sulfated metal oxides, het-
eropoly acids, supported Lewis and Brønsted acids
and resins have been studied as aliphate alkylation cat-
alysts as well, however, a presentation of these systems
would be outside the scope of this overview. Finally, thischapter should not be finished before the current process
developments at the pilot plant stage are summarised
concerning the use of alternative solid aliphate alkyl-
ation catalysts, however, as far as this is known, no zeo-
lite based systems are among these catalysts [14]:
1. Trifluoromethanesulfonic acid on a porous carrier
(Haldor Topsøe A/S).2. Antimonypentafluoride on acid-washed silica (Chem-
ical Research & Licensing Co., Chevron Corp.).
3. Proprietary catalysts (UOP and Catalytica Inc, Neste
Oy, Conoco Inc.).
3.6. Catalytic reforming
Besides catalytic cracking, catalytic reforming is one
of the most important processes within a modern
Fig. 19. Schematic diagram of a typical FCC unit [39]. Reproduced by
permission of Wiley-VCH, Weinheim.
274 M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292
refinery, where low-octane naphtha components (nor-
mal- and cycloalkanes) are converted into high-octane
isoalkanes and aromatics suitable for gasoline produc-
tion. The catalytic reforming process consists of a num-
ber of hydrogenation and dehydrogenation reactions, in
addition to isomerisation, cyclisation and cracking withrespect to the formation of isoalkanes and aromatics.
The process is run at 425–525 �C with hydrogen pres-sures in the range of 0.5–3.0 MPa [39].
Conventional reforming catalysts are based on Pt/Cl
and Pt–Sn supported on alumina, however, the focus
here will be concentrated on zeolite based reforming cat-
alysts. The non-acidic Linde Type L zeolite (LTL, with
potassium as counter ions) as support for Pt and Bahas been investigated by several groups [80–82]. The
presence of highly dispersed Pt clusters inside the zeolite
channels and the shape-selective effects imposed by the
mono-directional framework structure of the KL-zeolite
is responsible for the very good aromatisation perfor-
mance of this zeolite, which has a pore diameter of
0.71 nm. One of the main drawbacks of this zeolite is
the high sensitivity towards sulfur poisoning, responsi-ble for the fast catalyst deactivation. The preparation
method and especially the Pt incorporation have a
strong influence of the catalyst performance with respect
to activity and stability [39].
Other zeolite based reforming catalysts have been
investigated as well, like Pt/ZSM-12, Pt/Beta and sul-
(see also Fig. 23) [39,96,103].Another solution with respect to improvement of the
reactant accessibility and decrease of secondary cracking
reactions has relied on the preparation of small Y zeolite
crystallites (<0.5 lm). Furthermore, the preparation ofmesoporous Al-MCM-41 materials and the application
of their Ni/Mo derivatives with respect to hydrocracking
of vacuum gas oil revealed a good selectivity for middle
distillates and demonstrated a higher activity comparedwith Ni/Mo supported on amorphous silica–alumina
[104,105].
Besides Y zeolite, other large pore zeolites or related
microporous and mesoporous materials, such as Omega,
L zeolite, Beta, VPI-5, mordenite, UTD-1, MCM-48
and SBA-15, have been investigated as components for
hydrocracking catalysts, however, these systems have
not yet led to commercial application [39,96,106].Recent developments within hydrocracking catalysis
have been concentrated on the improvements in the
amorphous silica–alumina and Y zeolite base materials
and the control of catalyst pore architecture. Although
zeolite based hydrocracking catalysts offer the best pros-
pects for heavy feeds conversion, their activity is sup-
pressed due to the restricted access of the heavier
lumination [103]. Reproduced by permission of Imperial College Press,
M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292 279
molecules to the zeolite pores. Therefore, the main
focus concerning further research will be concentrated
on finding a zeolitic catalyst component dedicated to
the production of middle distillates, associated with the
activity, stability and proper pore architecture of the
final hydrocracking catalyst.
3.9. Catalytic dewaxing
Heavier oils, like gas oils, diesel oils or lube oils, con-
tain often wax (long-chain normal and/or slightly
branched C18 and higher paraffins), which crystallise at
low temperatures (below 100 �C). This affects their vis-cosity deleteriously, as quantified by pour point determi-nation [107]. These compounds are therefore often
removed, either by physical processes such as extraction
in solvent dewaxing, or nowadays by catalytic dewaxing
using shape selective catalysts. During catalytic dewax-
ing (carried out at about 400 �C), normal and slightlybranched long-chain paraffins are removed by selective
cracking to lighter products including gas. The basic
process resembles selectoforming, where C5–C9 normalparaffins are cracked to LPG. The selective removal of
the long-chain normal paraffins in gas oil dewaxing
using ZSM-5 as catalyst is illustrated in Fig. 24, where
the sharp peaks of the normal paraffins in the gas chro-
matograph of the feed are absent in the product after
catalytic dewaxing [108].
Fig. 24. Catalytic dewaxing of gas oil using shape selective ZSM-5 as catalys
chromatograph after catalytic dewaxing, RT means retention time). Reprod
Catalytic dewaxing has been connected to the med-
ium pore zeolites and related microporous solids, includ-
ing ZSM-5, ZSM-11, ZSM-23 and SAPO-11 [109]. BP
introduced the first generation dewaxing catalysts,
which was based on mordenite in order to remove nor-
mal alkanes from lube oils. However, Mobil�s discoveryof ZSM-5 forced the development of the next generation
of dewaxing catalysts, applicable to both gas oils
(MDDW process) and lube oils (MLDW process). The
new generation of ExxonMobil catalysts (MSDW: Mo-
bil Selective DeWaxing) selectively converts linear long-
chain paraffins to their corresponding isoparaffins
instead of cracking to lower molecules. Studies in this
direction involved zeolite Beta and MCM-22 as catalystcomponents. Chevron commercialised their Isodewax-
ing process, which is based on isomerisation rather than
cracking as well, using SAPO-11 as dewaxing catalyst
[116], whereas the alkylation of naphthalene with meth-
anol can either yield preferentially 1-methylnaphthalene
(ZSM-12, 12-MR system, kinetic control) or 2-methyl-
naphthalene (dealuminated mordenite or ZSM-5) [111].
3.10.2. Isomerisation and transalkylation
of alkylaromatics
The transfer of alkyl groups between aromatic mole-
cules, also termed transalkylation, and the intra-mole-
cular isomerisation are commercially applied in large
scale. Both reactions are acid catalysed processes, and
most of the catalysts used are solid systems, with zeolites
as the most prominent representatives [117].The pyrolysis gasoline from naphtha crackers and the
naphtha reformate consist mainly of xylenes and ethyl-
benzene with respect to their C8 aromatics fractions.
They are isolated from these streams by distillation
and solvent extraction. Xylenes are used on a large scale
industrially, and they are precursors for a number of
important petrochemicals. Especially para-xylene has
an enormous market potential (terephthalic acid pro-duction for polyester formation), with an annual in-
crease of about 7% [93,118].
When C8 aromatics fractions are processed with re-
spect to xylene isomerisation, the remaining ethylben-
zene must be converted to xylenes. Catalysts used for
the xylene and ethylbenzene isomerisation contain al-
ways platinum, in former times mainly on chlorinated
aluminas or steamed silica–aluminas (Octafining processdeveloped by Atlantic Refining Corp.), nowadays pref-
erentially on mordenite. The process is conducted in a
M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292 281
fixed-bed reactor at temperatures between 370 and
430 �C and a pressure range of 7–15 bar. New genera-tions of zeolite based catalysts have been introduced
by UOP (Isomar process) and IFP (Oparis), claiming
even higher yields of para-xylene. Metal modified
H-ZSM-5 catalysts are most effective when the xyleneisomerisation is accompanied by ethylbenzene dealkyla-
tion to benzene and ethene under similar process condi-
tions (ExxonMobil) [118].
Disproportionation is a special case of transalkyla-
tion and is in place when alkyl groups are transferred
between identical molecules. Several processes are
commercialised dealing with toluene disproportionation
using mordenite or ZSM-5 as catalysts and arriving atthermodynamic mixtures of xylenes (ExxonMobil,
UOP and TotalFinaElf). Selective toluene dispropor-
tionation processes are on the market as well, using
ZSM-5 catalysts and arriving at para-xylene rich prod-
uct mixtures with selectivities of more than 80% (Eni-
Chem, ExxonMobil and UOP) [117,118].
The disproportionation of ethylbenzene to benzene
and diethylbenzenes has been studied extensively[119,120], but has not been commercialised until re-
cently. With large pore zeolites (Y zeolite), the reaction
occurs via a hydride transfer chain reaction through
diphenylethanes as intermediates. Medium pore zeolites
(ZSM-5) cannot accommodate this bulky intermediate,
and ethylbenzene disproportionation proceeds via a
dealkylation–realkylation path. This technology is com-
mercialised using a modified ZSM-5 catalyst [117]. TheCatalysis Commission of the International Zeolite
Association (IZA) has recommended the disproportion-
ation of ethylbenzene using LaNaY zeolite as a stan-
dard reaction for acidity characterisation of acid
zeolites [121].
3.10.3. Conversion of surplus aromatics
In connection with the introduction of the EuropeanAuto Oil Programme, the aromatics content of gasoline
has to be reduced currently from 43 to 35 vol.% in 2005.
This number is also expressed by the most advanced gas-
oline specifications proposed in the World-Wide Fuel
Charter (WWFC, see Table 2) [35]. As a consequence,
there will be a surplus of aromatics. The main source
of aromatics is the so-called pyrolysis gasoline, a by-
product rich in aromatics from the manufacture of eth-ene and propene by steam-cracking of straight run
naphtha (light hydrocarbons). Due to the predicted
world-wide increasing demand of ethene and propene,
the surplus of pyrolysis gasoline will increase further
[122]. A novel catalytic process for hydrogenative ring
opening of aromatics has been introduced, which allows
the conversion of pyrolysis gasoline from naphtha
steam-crackers into a high quality synthetic steamcracker feed composed of C2–C4 n-alkanes [37,123].
There are two process variants, the direct conversion
of aromatics with hydrogen on bi-functional zeolites
and the two-stage process comprising a conventional
ring hydrogenation to cycloalkanes followed by ring
opening of these compounds using acidic zeolites. A
large part of this research has been carried out in the
laboratories of J. Weitkamp at the University of Stutt-gart (Germany) [37,123–127].
The main advantage of the direct route is the fact that
one single reactor is sufficient to perform the desired cat-
alytic conversion, which is carried out using shape selec-
tive bi-functional zeolites, like Pd/H-ZSM-5 (Si/Al ratio
of about 20). For example, ring hydrogenation (6 MPa
hydrogen pressure) of toluene to methylcyclohexane
and skeletal isomerisation of the cycloalkane into ethyl-cyclopentane and dimethylcyclopentanes occurs at tem-
peratures up to about 250 �C, however, ring openingwith respect to the desired formation of C2þ n-alkanes
(73% of the total synthetic steam cracker feed) requires
temperature of up to 400 �C [124]. C2þ n-alkanes refersin this context to ethane, propane and n-butane.
Advantages of the two stage route are an easier re-
moval of the heat generated in the two exothermic stepsand the possibility to optimise the reaction conditions of
ring hydrogenation (performed by conventional technol-
ogy using metal catalysts) and ring opening (carried out
by novel shape selective zeolites in their acid form as cat-
alysts). Applying H-ZSM-5 with a Si/Al ratio of 20,
yields of C2þ n-alkanes, comparable to those obtained
during the direct conversion of toluene on Pd/H-ZSM-
5, are achieved when processing methylcyclohexane at400 �C as second step of the two stage route. ZSM-5and ZSM-11 turned out to be the best suited catalysts
for this technology, whereas large pore zeolites (like
Beta or Y zeolite) deactivated rapidly. Zeolites with
too narrow pores, like ZSM-35 or ZSM-22, did not
show high degrees of ring opening capacities, probably
due to the hindered diffusion of the hydrocarbons in-
volved [124].
3.11. Aromatisation of light alkanes
Aromatics (BTX) are precursor compounds for the
petrochemical industry of large importance. Since the
liquified petroleum gas (LPG) fractions in the refinery
are supposed to increase due to more severe operations
of the FCC units, there is a need to convert the lowvalue LPG into aromatics.
The aromatisation of light alkanes is easier to per-
form as the size of the alkane increases, taking into
the thermodynamic relations. For example, propane
and higher alkanes can be aromatised at temperatures
lower than 500 �C, whereas temperatures of up to575 �C are requested for the aromatisation of ethane.In addition, the proportion of benzene in the BTX frac-tion tends to decrease with increasing size of the alkane
[11,39,128].
282 M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292
The mechanism of the light alkane aromatisation in-
volves the initial dehydrogenation of paraffins to form
lower olefins and hydrogen, followed by olefin oligomer-
isation and cyclisation reactions to produce aromatics,
in which hydrogen transfer plays an important role
[36], see also Fig. 25.Several zeolite based industrial processes have been
introduced with respect to the aromatisation of light
alkanes. The most known process is BP/UOP�s CY-CLAR process using C3/C4 refinery gases as feedstock
and Ga/H-ZSM-5 as catalyst (1–5 wt.% Ga, Si/Al ratio
between 15 and 30). However, other technologies are
known, including Mobil�s M2 Forming process (usingH-ZSM-5), Chevron�s AROMAX process applyingC6–C8 alkanes as feed and Pt/L zeolite as catalyst as well
as Selectoforming over erionite [128]. An advantage for
the CYCLAR process is the fact that significant
amounts of hydrogen are formed, which is needed in
the refinery and which makes this process economically
very attractive [11].
The preferred catalyst for alkane aromatisation is Ga/
H-ZSM-5. The incorporation of Ga can be performedeither by direct synthesis, impregnation or ion exchange.
The success of the modified H-ZSM-5 catalyst is con-
nected to the limited coke formation by using this
zeolite. Ga modification improves this behaviour even
more [128]. Besides gallium, other metals such as Zn,
Pt, Ni and Ag have been used in combination with
ZSM-5 for the light alkane aromatisation, however,
the Ga- and Zn/H-ZSM-5 modified catalysts are the pre-ferred systems. Finally, Pt/KL zeolite has been applied
in the selective formation of benzene from hexane.
There is a strong interest in the aromatisation of eth-
ane since this compound is an important component of
Fig. 25. Simplified reaction mechanism for the aromatisation of propane
Imperial College Press, London.
both refinery and natural gases. However, from a ther-
modynamic point of view, it is very difficult to convert
ethane into aromatic hydrocarbons, since a highly
unstable primary carbenium ion will be formed in the
acid catalysed oligomerisation step, whereas for higher
alkanes a more stable secondary carbenium ion isformed. Anyway, several investigations have been re-
ported using Pt, Pd, Ga or Zn modified H-ZSM-5 zeo-
lites and temperatures of about 575 �C [39,128–130].
3.12. Natural gas upgrading/gas conversion technologies
The evolution of the known crude oil and natural gas
reserves world-wide indicates a dramatic increase in thelatter compared to a levelling off concerning the crude
oil. This trend is expected to continue, which will—in
addition to the price development with respect to the
crude oil based upgrading—most likely generate a grad-
ual shift towards the application of natural gas as a feed-
stock for the production of fuels and petrochemicals.
This situation has forced an enhanced global interest
in processes, which can convert natural gas into liquidsand higher added value products—without going via
methanol as intermediate. This route is known as the
‘‘gas to liquids (GTL)’’-technology, based on the
Fischer–Tropsch route. The interest to manufacture
fuels and petrochemicals from natural gas is driven by
the desire to apply this technology directly, for example
at remote natural gas field sites, in order to minimize
transportation costs and gas burning at the recoverysites [36].
The following sub-chapters will deal with the
‘‘GTL’’-technology, based on the Fischer–Tropsch
route as well as the methanol to hydrocarbon conver-
over modified H-ZSM-5 catalysts [11]. Reproduced by permission of
M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292 283
sions, where zeolites and related microporous materials
have been demonstrated to be superior catalysts.
3.12.1. ‘‘Gas to Liquids (GTL)’’/conversion of synthesis
gas to fuel
Fuels production directly from syngas (in formertimes obtained from coal) has been reported by Fischer
and Tropsch in 1923 for the first time [38], using an al-
kali-promoted iron catalyst. Fuels manufactured via the
Fischer–Tropsch route reveal an excellent quality since
they consist mainly of linear paraffins and a-olefinsand do not contain sulfur and aromatics. A Co-contain-
ing catalyst is applied for the production of heavy paraf-
fins via the Fischer–Tropsch route starting with naturalgas, a technology developed by Shell and named the
[39,40]. In addition, diesel fuel (or gasoline) is produced
by hydrocracking of the more or less sulfur and nitro-
gen-free wax obtained through the SMDS process using
noble metal containing zeolites. The more restricted fuel
specifications currently introduced in order to reduce the
environmental impact of hazardous emissions representa driving force with respect to an increased use of fuels
prepared via the Fischer–Tropsch route as a blending
component of the gasoline and diesel pools in the future
[39].
Besides the SMDS technology an alternative has been
presented by SASOL/Chevron termed as the ‘‘Slurry-
Phase-Distillate’’ process, again based on the Fischer–
Tropsch route producing wax (using a Co-containingcatalyst) followed by a hydrocracking step in order to
get diesel or gasoline [41].
The methanol to gasoline (MTG) plant in New
Zealand has been combined with a methane steam
reforming unit for production of synthesis gas and a
methanol plant to produce gasoline from natural gas.
The process economics can be improved considerably
by a clever combination and close integration of the dif-ferent steps. In the TIGAS process developed by Haldor
Topsøe AS for the manufacture of gasoline in a pilot
plant scale, the methanol synthesis and the MTG reac-
tions are integrated—without the separation of metha-
nol as an intermediate product. A multi-functional
catalyst has been developed, however, these process
technologies do not usually apply catalysts based on
zeolites or related microporous materials [36]. Finally,ExxonMobil has introduced the so-called ‘‘Advanced
Gas Conversion for the 21st Century’’ (AGC-21) tech-
nology, again based on the Fischer–Tropsch route [131].
3.12.2. Methanol to hydrocarbons
Besides the direct route from synthesis gas to hydro-
carbons (GTL-technology, based on the Fischer–Trop-
sch route), a strong focus has been concentrated onthe indirect route via the production of methanol from
synthesis gas (mixture of carbon monoxide and hydro-
gen), which is made by steam reforming of natural gas
or gasification of coal, and the consecutive formation
of hydrocarbons. Of coarse, these technologies are alter-
natives to the chemical conversion of methane, either via
direct coupling, which is thermodynamically not favour-
able, or via oxidative coupling, a route not successfullyso far from an industrial point of view.
To begin with, the methanol–hydrocarbons technol-
ogy was primarily regarded as a powerful method to
convert coal into high-octane gasoline. This concept
has been expanded since, not only with respect to the
formation of other fuels, but also to chemicals in gen-
eral. Of coarse, light olefins are important components
for the petrochemical industry, and the demand ofhigh-quality gasoline is increasing as well. In fact, with
this new technology, one can make almost anything
out of coal or natural gas that can be made out of crude
oil.
Methanol is converted into an equilibrium mixture
of methanol, dimethylether and water, which can be
processed catalytically to either gasoline (methanol to
gasoline, MTG) or olefins (methanol to olefins, MTO),depending on the catalyst and/or the process operation
conditions (see Fig. 26).
Although methanol itself is a potential motor fuel or
can be blended with gasoline, it would require large
investments to overcome the technical problems con-
nected with it. The commercial MTG reaction runs at
temperatures around 400 �C at a methanol pressure ofseveral bars and uses a ZSM-5 catalyst. These are theoptimal conditions for converting the olefins that form
within the catalyst into paraffins and aromatics. How-
ever, at one point in the MTG reaction, the product
mixture consists of about 40% light olefins. The impor-
tance of light olefins as intermediates in the conversion
of methanol to gasoline was recognised early. Conse-
quently, a number of attempts were made to selectively
form light olefins from methanol, not only on medium-pore zeolites but also on small-pore zeolites, SAPO type
molecular sieves and over large-pore zeolites (however,
to a much lesser extent). If one interrupted the reaction
at the point of about 40% light olefin formation, one
could harvest these C2–C4 olefins. By adjusting the reac-
tion conditions (such as, for example, raising the tem-
perature to 500 �C) as well as the catalyst applied, onecan increase dramatically the olefin yield. This discoveryled to the development of the MTO process, which gen-
erates mostly propene and butenes, with high-octane
gasoline as a by-product. However, the catalyst can be
modified in such a way that even more ethene is pro-
duced [132].
1973 marked the beginning of the energy crisis, and
new interest in synfuels and other chemicals favoured
the continuation of the methanol–hydrocarbon research[133,134]. Already the MTG and MTO processes repre-
sent a sort of chemical factory, to be brought on stream
Fig. 26. Methanol to hydrocarbons reaction path [132]. Reproduced by permission of Elsevier, Amsterdam.
284 M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292
as the technological and/or economic demands arise.
One can go a step further and convert the olefins to
an entire spectrum of products, through another ZSM-
5 based process: Mobil�s olefin–gasoline and distillateprocess (MOGD), originally developed as a refinery pro-
cess, which works well coupled with the MTO process.In the MOGD reaction, ZSM-5 oligomerises light ole-
fins, from either refinery streams or MTO, into higher-
molecular-weight olefins that fall into the gasoline,
distillate and lubricant range (see also Fig. 27) [135].
In 1979 the New Zealand government selected the
MTG process over the Fischer–Tropsch (SASOL) pro-
cess for converting natural gas from their extensive
Maui field to gasoline. At that time, Mobil�s fixed-bedMTG process was unproven commercially, whereas
the SASOL technology was already commercialised
[135]. The New Zealand plant started to produce about
600000 ton per year gasoline from April 1986, supplying
one-third of the nation�s gasoline demand [133]. Thegasoline production part of the factory was later closed
down, due to the price available for gasoline vs. the price
Fig. 27. Gasoline and distillate production via methanol and Mobil�s ZSM-5
of methanol, however, the methanol production part is
still in operation.
In the scale-up to commercial operation of the MTG
plant two factors were of certain interest, catalyst deac-
tivation and heat production, respectively (the MTG
reaction is highly exothermic). Despite the relativelyhigh stability and selectivity of the ZSM-5 catalyst
(due to the high silica content and the catalytic effective-
ness based on the unique pore structure), deactivation
during the MTG operation is pronounced. The catalyst
is progressively coked and must be regenerated by calci-
nation in air. The heat production in the MTG process
is high, leading to an adiabatic temperature rise of some
650 �C [36,136,137].The MTO technology seems now to be ready for
commercial use. The MTO process of Mobil has been
demonstrated in the same experimental 4000 ton per
year plant at Wesseling (Germany) used to prove the
fluid-bed MTG process, applying ZSM-5 as catalyst
[133]. The fluid-bed technology provided all advantages
in terms of increased product yield, better quality and a
technology [132]. Reproduced by permission of Elsevier, Amsterdam.
M. Stocker / Microporous and Mesoporous Materials 82 (2005) 257–292 285
very efficient heat recovery. Depending on the world
market situation (price, demand, etc.) the fluidised-bed
technology is available to produce liquid fuels via meth-
anol [138].
UOP has, in co-operation with Norsk Hydro, an-
nounced in 1996 their SAPO-34 based MTO processto be realised for construction of a 250000 ton per year
plant using a natural gas feedstock for production of
ethene. A 0.5 ton per year demonstration unit operated
by Norsk Hydro has verified the olefin yields and cata-
lyst performance. SAPO-34 is extremely selective to-
wards ethene and propene formation (about 80%) with
the flexibility of altering the ratio between the two ole-
fins by varying the reactor conditions. The MTO processcan be designed for an ethene to propene ratio between
0.75 and 1.5—at nearly complete methanol conversion.
The high selectivity to ethene gives SAPO-34 (which
has the chabazite (CHA) structure) a significant advan-
tage over other types of catalyst systems, like ZSM-5 or
SSZ-13 (synthetic aluminosilicate with CHA structure).
In addition, the SAPO-34 has a significantly better sta-
bility due to a lower rate of coke formation than theother catalytic systems with comparable acid site densi-
ties. The need to remove the high exothermic heat of the
MTO reaction as well as the need for frequent regener-
ation led to a fluidised-bed reactor and regenerator
design. UOP and Norsk Hydro have commercially
manufactured the MTO catalyst (MTO-100TM), based
on SAPO-34, which has shown the type of attrition
resistance and stability suitable to handle multipleregeneration steps and fluidised-bed conditions [139,
140].
The discovery of the MTG reaction happened by
accident. One group at Mobil was trying to convert
methanol to other oxygen-containing compounds over
a ZSM-5 catalyst. Instead, they received unwanted
hydrocarbons. Somewhat later, another Mobil group,
working independently, was trying to alkylate isobutanewith methanol over ZSM-5 and identified a mixture of
paraffins and aromatics boiling in the gasoline range—
all coming from methanol. Although the discovery of
MTG was accidental, it occurred due to a balanced ef-
fort in catalysis over many years. The MTO reaction
seems to benefit from this development, although inde-
pendent research has been performed since. The evolu-
tion of the methanol–hydrocarbons technology, fromits discovery until its realisation on a demonstration
and/or commercial scale, has been accompanied by
extensive research related to the basic question of the
mechanism of formation of the initial C–C bond. To ad-
dress this topic in detail would be out of scope for this
paper, however, the interested reader can follow this dis-
cussion in the reviews mentioned in the list of Refs.
[132–137].Finally, an alternative to produce propene from
methanol has been introduced by Lurgi, the so-called
‘‘Methanol-to-Propene (MTP)’’ process, applying an
H-ZSM-5 based catalyst from Sud-Chemie AG
[42,131].
4. Catalytic probe reactions with respect to the porearchitecture characterisation of zeolites and related
porous materials
4.1. General remarks
Different types of probe molecules have been used to
investigate various properties of zeolites and related
microporous and mesoporous materials, for example,the pore architecture with respect to internal channel
dimensions, possible network connectivities, contribu-
tions of the external surface to the overall behaviour
of the materials, accessibility of pockets and caves in
connection with the structure of those compounds. Fur-
thermore, probe molecules are used to evaluate the sur-
face hydrophobicity and hydrophilicity as well as to
determine the acidity and basicity of porous systems.These type of investigations are done either by phys-
ico-chemical studies of the interaction of probe mole-
cules with the surface of the porous material (for
example by adsorption of the probe molecule and mon-
itoring the interaction by applying spectroscopic, calori-
metric, volumetric, gravimetric or other methods) or by
using probe molecules as model substrates in catalytic
reactions. Since the pores of zeolites and related micro-porous materials have dimensions comparable to those
of actual probe molecules, the phenomenon of shape-
selectivity in catalytic reactions can be observed. How-
ever, in order to take advantage of this, the catalytic
reaction must take place at active sites on the internal
surface and not on the external one. Therefore, the
probe molecules applied are usually designed with re-
spect to evaluate the influence of both the internal andexternal surface activities on the properties under inves-
tigation, like for example, a mixture of linear and
branched/bulky molecules. Branched and bulky probe
molecules are usually incapable to penetrate the micro-
pores whereas linear molecules can.
Concerning the selection of suitable probe molecules
one should note that the crystallographic pore diameters
do not represent suitable threshold molecular diameters,since molecules about 20% larger than this diameter can
be accommodated, especially at elevated temperatures.
Furthermore, the type of activity under investigation