n-alkane isomerization catalyzed by highly acidic ionic ... · n-alkane isomerization catalyzed by...
TRANSCRIPT
n-Alkane Isomerization Catalyzed by Highly Acidic
Ionic Liquids in Multiphase Reaction Systems
n-Alkanisomerisierung katalysiert durch hochacide ionische
Flüssigkeiten in mehrphasigen Reaktionssystemen
Der Technischen Fakultät der
Friedrich-Alexander-Universität Erlangen-Nürnberg
zur Erlangung des Grades
DOKTOR – INGENIEUR
vorgelegt von
Dipl.-Ing. Carolin Meyer
aus Bamberg
Erlangen, 2013
Als Dissertation genehmigt von der Technischen Fakultät der Universität Erlangen-Nürnberg
Tag der Einreichung: 27.08.2012
Tag der Promotion: 05.07.2013
Dekan: Prof. Dr.-Ing. habil. Marion Merklein
Berichterstatter: Prof. Dr. Peter Wasserscheid
Prof. Dr.-Ing. Andreas Jess
III
Vorwort
Die Ergebnisse der vorliegenden Arbeit entstanden von März 2009 bis April 2012 am
Lehrstuhl für Chemische Reaktionstechnik der Friedrich-Alexander-Universität
Erlangen-Nürnberg.
Mein ganz besonderer Dank gilt meinem Doktorvater Prof. Dr. Peter Wasserscheid für die
spannende Themenstellung, die Bereitstellung sehr guter Arbeitsbedingungen, die zahlreichen
Ratschläge und Anregungen, das stets offene Ohr, seinen unermüdlichen Enthusiasmus und
für sein großes, in mich gesetztes Vertrauen.
Herrn Prof. Dr.-Ing. Andreas Jess danke ich herzlich für die Übernahme des Zweitgutachtens.
Außerdem gilt mein Dank den weiteren Mitgliedern des Prüfungskollegiums, Herrn Prof. Dr.
Martin Hartmann, Herrn Prof. Dr. Jörg Libuda und Herrn Prof. Dr.-Ing. Malte Kaspereit.
Herrn Prof. Dr. Wilhelm Schwieger danke ich für wertvolle Diskussionen und Anregungen
rund um meine Arbeit.
Herrn Dr. Normen Szesni, Dr.-Ing. Jens Freiding und Dr. Rainer Albert Rakoczy vom
Industriepartner Süd-Chemie AG, die einen Teil dieser Arbeit begleitet haben, möchte ich für
die gute und kooperative Zusammenarbeit danken.
Für die Finanzierung meines Projektes im Rahmen des Exzellenzclusters „Engineering of
Advanced Materials (EAM)" der Deutschen Forschungsgemeinschaft (DFG) und für die
Finanzierung von Auslandskonferenzen, sowie weiterbildenden Kursen im Rahmen der im
Exzellenzcluster integrierten Graduate School „Advanced Materials and Processes“ möchte
ich danken.
Außerdem möchte ich an dieser Stelle allen meinen herzlichen Dank aussprechen, die zum
Gelingen dieser Arbeit beigetragen haben. Ein ganz besonderer Dank geht an meine
Studenten Carina Kohr, Veit Hager, Tobias Tengler, Thomas Jakuttis, Sven Jakubenko,
Denise Geburtig und Lisa Hopf. Es hat viel Spaß gemacht, mit Euch zu arbeiten. Durch Eure
Studien-, Bachelor- und Diplomarbeiten, oder auch im Rahmen eines HIWI-Jobs, habt ihr
einen großen Anteil zu dieser Doktorarbeit beigetragen.
Für die vielfältige Hilfe bei elektrischen oder computertechnischen Problemen möchte ich
mich bei Gerhard Dommer, Alexander Busch und Hendryk Partsch bedanken. Frau Menuet,
Frau Singer und Frau Bittan danke ich für die viele Arbeit im Hintergrund, die immer einen
IV
reibungslosen Ablauf ermöglicht hat. Für die große Unterstützung und den Einsatz bei den
immer wiederkehrenden Reaktorneuaufbauten und sonstigen mechanischen Problemen geht
ein ganz herzliches und großes Danke an Achim Mannke, Michael Schmacks und Julian Karl.
Einen besonderen Dank möchte ich auch den fleißigen Leuten vom Sorptionslabor, Stefan
Gütlein, Tilman Knorr und Alexandra Inayat, für die Vermessung meiner Proben und für die
Diskussionen zur Auswertung aussprechen, ebenso Peter Schulz für die Hilfe bei allen
analytischen Problemen und Fragestellungen. Auch Nicola Taccardi danke ich herzlich für die
ICP-AES Analytik, sowie für die Hilfe und Ratschläge rund um die Auswertung und
sonstigen Fragen zur Chemie.
Prof. Dr.-Ing. Andreas Jess, Stephan Aschauer, Lisa Schilder und Johannes Thiessen möchte
ich außerdem herzlich danken, dass ich beim CVT in Bayreuth Experimente und die
H2-Chemisorption durchführen durfte. Danke Aschi, Lisa und Johannes für Eure
Hilfsbereitschaft, Tipps und Anregungen.
Judith Scholz, Stefan Schlenk und René Wölfel danke ich dafür, dass sie sich die Zeit
genommen haben, meine Arbeit kritisch durchzulesen und für die daraus hervorgegangenen
wertvollen Anmerkungen.
Allen aktuellen und ehemaligen Kollegen und Freunden am CRT gebührt ein großer Dank für
die Unterstützung, die hilfreichen Diskussionen, die netten Unterhaltungen bei der einen oder
anderen Tasse Kaffee oder am Gang, sowie die schöne und vor allem unvergessliche Zeit am
Lehrstuhl. Ich freue mich schon jetzt auf unsere Pruggern-Alumni-Hütte. Bin ich froh, Judith,
dass wir uns am Lehrstuhl einen Abzug teilen mussten und uns dadurch kennengelernt haben.
Die gemeinsamen Konferenzfahrten waren einfach immer toll.
Bei meinen Freunden möchte ich mich sehr herzlich bedanken, dass sie immer für mich da
sind.
Dir, Stefan, danke ich von ganzem Herzen für deine Unterstützung, weit über die
Doktorarbeit hinaus.
Zum Schluss möchte ich mich herzlichst bei meiner Familie und besonders bei meinen Eltern
bedanken, die mir dies alles erst ermöglicht haben.
Publications V
Publications
Parts of this work have been previously published in the following publications or presented
at the conferences listed.
Patent
� C. Meyer, V. Hager, P. Wasserscheid, Bifunktionelles Katalysatorsystem für die
Hydroisomerisierung von Alkanen, DE10 2011 114 605.2 (priority 03/10/2011), status
pending.
Peer-reviewed articles
� C. Meyer, P. Wasserscheid, Chemical Communications, 46, 7625-7627, 2010.
� C. Meyer, V. Hager, W. Schwieger, P. Wasserscheid, Journal of Catalysis, 292,
157-165, 2012.
Conference contributions
� C. Meyer, V. Hager, M. Haumann, P. Wasserscheid, Acidic ionic liquids for n-alkane
isomerization in a liquid-liquid or slurry-phase reaction mode, presentation at
International DGMK Conference (Catalysis – Innovative Applications in
Petrochemistry and Refining), Dresden, 2011.
� C. Meyer, Solid catalysts with ionic liquid layer (SCILL) for n-alkane isomerization,
presentation at 3rd Symposium of the Cluster of Excellence, Engineering of Advanced
Materials (EAM), Oberhof, 2011.
� C. Meyer, P. Wasserscheid, Low temperature n-alkane isomerization catalyst:
superacidic supported ionic liquid in a slurry-phase reaction mode, poster at 8th
European Congress of Chemical Engineering (ECCE), Berlin, 2011.
� C. Meyer, C. Kohr, M. Haumann, P. Wasserscheid, Low temperature n-alkane
isomerization catalysts: Superacidic ionic liquids in a liquid-liquid biphasic or
VI Publications
supported ionic liquid phase (SILP) slurry reaction system, poster at 4th Congress On
Ionic Liquids (COIL-4), Washington D.C. / USA, 2011.
� C. Meyer, C. Kohr, M. Haumann, P. Wasserscheid, Superacidic ionic liquids - Low
temperature n-alkane isomerization catalysts in liquid-liquid biphasic and slurry-
phase reaction systems, 22nd North American Meeting (NAM) of the North American
Catalysis Society, Detroit / USA, 2011.
� C. Meyer, C. Kohr, M. Haumann, P. Wasserscheid, Superacidic ionic liquids - low
temperature n-alkane isomerization catalysts in liquid-liquid biphasic and slurry-
phase reaction systems, 44. Jahrestreffen Deutscher Katalytiker und Jahrestreffen
Reaktionstechnik, Weimar, 2011.
� C. Meyer, M. Haumann, P. Wasserscheid, Superacidic ionic liquids – low temperature
n-alkane isomerization catalysts, Green Solvents for Synthesis, Berchtesgaden, 2010.
� C. Meyer, M. Haumann, P. Wasserscheid, Superacidic ionic liquids – an effective
n-alkane isomerization catalyst, 3rd European Association for the Chemical and
Molecular Sciences (EuCheMS) Chemistry Congress, Nuremberg, 2010.
VII
Für Stefan und meine Familie
VIII
Table of contents IX
Table of contents
List of symbols and abbreviations XIV
1 Introduction 2
2 General part 6
2.1 Introduction to the alkane isomerization 6
2.1.1 Fundamentals 6
2.1.2 Relevance of alkane isomerization 8
2.1.3 Industrial processes 9
2.1.3.1 Processes using AlCl3 / HCl as monofunctional catalyst 10
2.1.3.2 Processes with bifunctional catalysts consisting of Pt supported on chlorinated alumina 11
2.1.3.3 Processes based on noble metal on amorphous silica-alumina as bifunctional catalyst 11
2.1.3.4 Processes using noble metal on zeolites 11
2.1.3.5 Processes using noble metal on metal oxides 12
2.2 Definition and classification of acids 13
2.3 Mechanism of the alkane isomerization 16
2.3.1 Mechanism of the acid catalyzed alkane isomerization 16
2.3.1.1 Initial step of the alkane isomerization reaction 16
2.3.1.2 Skeletal isomerization of carbenium ions 17
2.3.1.3 Chain propagation 18
2.3.1.4 Possible side reactions in acidic isomerization catalysis 19
2.3.1.4.1 Catalytic cracking on monofunctional acidic catalysts 19
2.3.1.4.2 Alkylation reaction 21
2.3.1.4.3 Coke formation on solid acidic catalysts 21
2.3.1.5 Cracking inhibitor hydrogen 22
X Table of contents
2.3.2 Alkane hydroisomerization using bifunctional or monofunctional metal catalysts 22
2.3.2.1 Initial step of the alkane isomerization reaction using bifunctional catalysts 22
2.3.2.2 Chain propagation 22
2.3.2.3 Additional possible side reactions using bifunctional and monofunctional metal catalysts 22
2.3.2.3.1 Hydrocracking 23
2.3.2.3.2 Hydrogenolysis 23
2.3.2.3.3 Coke formation on bifunctional catalysts 23
2.4 Introduction to acidic ionic liquids and their application in catalysis 23
2.4.1 Characteristic properties of acidic halometallate ionic liquids 24
2.4.2 Acidic halometallate ionic liquids – typical applications in catalysis 26
2.4.2.1 Friedel-Crafts reactions 27
2.4.2.2 Alkylation 30
2.4.2.3 Lewis acidic ionic liquid catalyzed oligomerization and polymerization 31
2.4.2.4 Acidic ionic liquid catalyzed cracking of hydrocarbons 33
2.5 Catalyst and immobilization concepts with ionic liquids 33
2.5.1 Multiphase reaction systems 34
2.5.2 Supported ionic liquid phase (SILP) catalyst 34
2.5.3 Solid catalyst with ionic liquid layer (SCILL) 37
2.5.4 Stabilization and immobilization of nanoparticles in ionic liquids 41
2.6 Catalyst systems for the alkane isomerization 42
2.6.1 Heterogeneous catalyzed alkane isomerization 42
2.6.2 Acidic ionic liquid catalysts for alkane isomerization 46
2.7 Objective of this work 47
3 Experimental 50
3.1 General working techniques 50
Table of contents XI
3.2 Chemicals 50
3.3 Catalyst preparation 52
3.4 Reaction 54
3.5 Analytics 58
3.5.1 Gas chromatography (GC) analysis 58
3.5.2 Gas chromatography - mass spectrometry (GC-MS) 58
3.5.3 Electrospray ionization – mass spectrometry (ESI-MS) 59
3.5.4 Inductively coupled plasma - atom emission spectroscopy (ICP-AES) 59
3.5.5 N2-adsorption 59
3.5.6 Nuclear magnetic resonance spectroscopy (NMR) 60
3.5.7 H2-chemisorption 60
3.5.8 X-rax diffraction (XRD) 60
3.6 Further calculations 61
3.6.1 Modified reaction time 61
3.6.2 Aspen Plus simulation 61
4 Results and discussion 64
4.1 Catalytic experiments with the monofunctional acidic ionic liquid catalysts in a liquid-liquid biphasic reaction mode 64
4.1.1 Reproducibility of the data 64
4.1.2 Catalyst screening experiments for system selection 67
4.1.2.1 Experiments with the system [BMIM]Cl / AlCl3 / H2SO4 67
4.1.2.2 Experiments with the system [BMIM]Cl / AlCl3 / CuCl2 70
4.1.2.3 Experiments with the system [BMIM]Cl / AlCl3 / 1-chlorooctane 72
4.1.3 Recyclability of the acidic ionic liquid systems [BMIM]Cl / AlCl 3 / H2SO4 and [BMIM]Cl / AlCl3 / 1-chlorooctane 73
4.1.4 Influence of hydrogen partial pressure 76
4.1.4.1 Influence of hydrogen partial pressure variation on n-octane isomerization at very mild reaction conditions 76
4.1.4.2 Influence of hydrogen partial pressure on n-octane isomerization at elevated reaction temperature 78
XII Table of contents
4.1.4.3 Influence of hydrogen on n-hexane isomerization at very mild reaction conditions 80
4.1.5 Variation of the reaction temperature in the hydroisomerization of n-hexane 82
4.1.6 Influence of the ionic liquid´s Lewis acidity on the hydroisomerization of n-octane 86
4.2 Monofunctional SILP catalysts in a slurry-phase reaction mode 89
4.3 Bifunctional SCILL catalyst for n-octane hydroisomerization in a slurry-phase reaction mode 91
4.3.1 Characterization of the SCILL material 92
4.3.2 Influence of hydrogen partial pressure on activity and selectivity of the bifunctional SCILL catalyst 97
4.3.3 Product distribution for the bifunctional SCILL catalyst and its comparison with results of the acidic ionic liquid in liquid-liquid biphasic catalysis 100
4.3.4 Influence of the reaction temperature on activity and selectivity 104
4.3.5 Influence of the ionic liquid´s acidity 105
4.3.6 Varying amounts of 1-chlorooctane and its influence on activity and selectivity 108
4.3.7 Catalyst stability 110
4.3.8 Catalyst recycling 113
4.4 In situ formed Pt nanoparticles in acidic ionic liquids for n-octane hydroisomerization 115
4.4.1 Activity and selectivity of the catalysts derived from PtCl2 or PtCl4 in the Lewis acidic ionic liquid [BMIM]Cl / AlCl3 116
4.4.2 Influence of the ionic liquid´s acidity in combination with the metal precursor PtCl2 118
4.5 Comparison of the bifunctional and monofunctional catalytic systems in the n-octane isomerization 120
4.5.1 Characterization of silica gel 100 and comparison with Pt / silica 120
4.5.2 Catalytic results of the four different catalytic systems 122
4.5.3 Reaction network 129
4.6 Final evaluation 131
Table of contents XIII
5 Summary / Abstract 134
6 Zusammenfassung / Kurzfassung 142
7 Appendix 150
7.1 Influence of hydrogen on n-hexane isomerization at very mild reaction conditions 150
7.2 Reproducibility of SCILL catalyzed experiments 152
7.3 Mass spectra of GC-MS analysis 156
7.4 Calculation of the RON 157
7.5 XRD diffractogram 158
8 References 160
XIV Symbols and abbreviations
List of symbols and abbreviations
Symbols
Latin letters
∆H�,�� standard reaction enthalpy at standard
pressure po = 1.01325 bar and temperature T J mol-1
∆H�,�� molar standard enthalpy of formation at standard pressure po = 1.01325 bar and temperature T J mol-1
a activity mol m-3
A area m2
B indicator base -
c concentration mol m-3
d diameter m
G free enthalpy J
G partial molar free enthalpy (= µi) J mol-1
H+ proton -
H0 Hammett acidity -
K equilibrium constant -
L length m
m mass kg
M molar mass kg mol-1
n molar amount mol
p pressure Pa
R organic rest -
R+ carbenium ion -
R= alkene -
RX organic halide -
Symbols and abbreviations XV
S selectivity %
t time s
T temperature K
tmod. modified reaction time min mol ionic liquid mol-1n-alkane
Tm melting point K
V volume m3
x mole fraction -
X conversion %
y variable -
Y yield %
z charge -
Greek letters
α ionic liquid pore filling degree -
δ deviation %
ε catalyst loading -
θ Bragg angle °
λ wavelength m
µi chemical potential J mol-1
υ stoichiometric coefficient -
Subscripts
0 value (concentration,...) at the beginning of the reaction (t = 0 s)
abs absolute
cat. catalyst
i, j component i,j
tot. total
XVI Symbols and abbreviations
Abbreviations
[BMIM] + 1-buty-3-methylimidazolium
[BMMIM] + 1-butyl-2,3-dimethylimidazolium
[Bpyr]+ 1-butylpyridinium
[DMIM] + 1-dodecyl-3-methylimidazolium
[EMIM] + 1-ethyl-3-methylimidazolium
[Epyr]+ 1-ethylpyridinium
[Et3NH]+ triethylammonium
[Hpyr]+ 1-hexylpyridinium
[Me3NH]+ trimethylammonium
[N(CN)2]- dicyanamide
[NTf2]- bis(trifluoromethylsulfonyl)imide
[OctSO4]- octylsulfate
[OMIM] + 1-octyl-3-methylimidazolium
[OTf] - trifluoromethanesulfonate
[TMSu]+ trimethylsulfonium
2,2-DMB 2,2-dimethylbutane
2,3-DMB 2,3-dimetylbutane
2-MP 2-methylpentane
3-MP 3-methylpentane
b cd-1 barrel per calender day
BET Brunauer-Emmett-Teller
BJH Barrett-Joyner-Halenda
Cl-octane 1-chlorooctane
COSMO-RS conductor-like screening model for real solvent
DCM dichloromethane
ESI electrospray ionization
FCC fluid catalytic cracking
Symbols and abbreviations XVII
FIA fluoride ion affinity
GC gas chromatography
HOMO highest occupied molecular orbital
ICP-AES inductively coupled plasma atom emission spectroscopy
IR infrared
LHSV liquid hourly space velocity
LUMO lowest unoccupied molecular orbital
MON motor octane number
MS mass spectrometry
MTBE methyl tert-butyl ether
NMR nuclear magnetic resonance
NRTL non-random two liquid
PCP protonated cyclopropane
RON research octane number
rpm revolutions per minute
SCILL solid catalyst with ionic liquid layer
SEM scanning electron microscope
SHOP Shell Higher Olefin Process
SILP supported ionic liquid phase
TIP Total Isomerization Process
UHV ultra-high vacuum
UOP Universal Oil Products
vol. % volume %
WHSV weight hourly space velocity
wt. % weight %
XAS X-ray absorption spectroscopy
XPS X-ray photoelectron spectroscopy
XRD X-ray diffraction
1
___________________________________________________________________________
1. Introduction
2 Introduction
Introduction 1
In 1876, Nikolaus August Otto developed a stationary, single-cylinder, four-stroke engine that
ran on coal gas [1]. This invention set a development in motion which substantially shaped
the industrial age. About ten years later, the Otto engine was firstly used to drive vehicles
(Daimler and Maybach automobiles, Benz Patent Motor Car). Attention focused on the
further improvement of engines by increasing the compression ratio (from circa 1918 on). The
increase in compression ratio was limited by abnormal combustion, so-called engine-knock,
which can damage the engine. Systematic research began to improve the knock resistance of
fuels. In 1922, the discovery of the antiknock effect of tetraethyllead by Midgley and Boyd
was a milestone in the development of gasoline. The further advances in the development of
engine technology and in the quality of gasoline were mutual dependent. Additionally, since
1970, efforts for improving the environmental compatibility of automotive use have become
increasingly important worldwide. Unleaded gasoline was developed which was a prerequisite
for the introduction of catalytic converters. In recent years, the reduction of sulfur content in
the gasoline occurred. This measure improves the efficiency and durability of exhaust gas
aftertreatment devices and reduces the emission of sulfur compounds. The limitation of
benzene, other aromatics and olefins in gasoline reduces the resulting post catalyst exhaust
emissions. A decrease of the Research Octane Number (RON) is the consequence due to
banned or limited concentration of fuel additives. Subsequently, the interest in large-scale
refinery production of highly branched alkanes has considerably increased since they combine
high RON with benign toxicology and ecotoxicology properties [2].
Major refinery processes to improve the RON include naphtha-isomerization, reforming,
addition of FCC-naphtha or alkylation. Naphtha-isomerization is a simple and very cost
effective technology for octane replacement. It is state-of-the-art in industry to isomerize
pentane / hexane feedstocks heterogeneously catalyzed by bifunctional catalysts, which
contain mostly Pt on various solid acids, like zeolites, chlorinated alumina or sulfated zirconia
[2, 3]. The metal sites provide an additional hydrogenation / dehydrogenation functionality in
the hydroisomerization whereby the skeletal rearrangement of the alkanes takes place on the
acid site of the catalyst. The mentioned catalysts are operated between 403 K and 573 K,
depending on the catalyst type applied. Isomerization of long-chain alkanes (C > 6) has not
been industrially applied up to now due to increased side reactions like cracking. Because of
limited gasoline additives, it will be necessary to find alternatives and thus, to isomerize even
Introduction 3
heptanes, octanes or longer chain alkanes. Recent research activities for the conversion of
alkanes with a chain length C > 6 focus on various heterogeneous catalytic systems [4-6]
operating at higher reaction temperatures.
Alkane isomerization is an equilibrium reaction and the formation of the desired high-RON
contributing highly branched isomers tends to occur at lower temperatures. Besides, coking
and along with it deactivation of the bifunctional heterogeneous catalysts is a typical and
well-known problem [7]. Therefore, recent research is based on the motivation to investigate
catalysts which are active and selective for the long-chain (C > 6) alkane isomerization at
mild reaction temperatures. Furthermore, they should avoid pore blocking with carbonaceous
and high boiling side products and subsequent loss of activity, selectivity and long-term
stability.
Highly acidic chloroaluminate ionic liquids gave proof of their ability to convert n-alkanes at
very mild reaction conditions (< 340 K) [8, 9]. Depending on the mole fraction of AlCl3,
chloroaluminate systems are Lewis basic, x(AlCl3) < 0.5, neutral, x(AlCl3) = 0.5, or Lewis
acidic, x(AlCl3) > 0.5 [10], whereupon the latter property is required for the alkane
isomerization. Brønsted superacidity of HCl in the acidic ionic liquid
1-ethyl-3-methylimidazolium chloride / aluminum chloride ([EMIM]Cl / AlCl3) was proved
and described by Smith [11]. Chloroaluminate ionic liquids are unavoidably contaminated by
small quantities of water which results in superacidic catalytic systems. In this work, different
highly acidic chloroaluminate based systems were investigated for their potential to isomerize
n-octane, a model compound of the long-chain alkanes (C > 6), in a liquid-liquid biphasic
reaction system. Attention focused on activity optimization and on possibilities to increase the
iso-alkane selectivity for the reactants n-hexane as well as n-octane.
In a second approach, a heterogeneous catalyst, Pt on silica (Pt / silica), was coated with a thin
film of chloroaluminate ionic liquid according to the principle of solid catalysts with ionic
liquid layer (SCILL). The resulting bifunctional SCILL catalyst – metal Pt and highly acidic
ionic liquid – was firstly applied in the alkane isomerization. The hydroisomerization of
n-octane was carried out under moderate temperature conditions (Treaction = 373 – 423 K) and
hydrogen pressures of up to 40 bar in a slurry-phase reaction mode (liquid organic phase,
solid SCILL catalyst). The beneficial interaction of hydrogen, Pt and acidic ionic liquid was
elucidated by comparing the new bifunctional SCILL catalyst in the presence of hydrogen
with the same catalyst in the absence of hydrogen and with the Pt-free, acidic ionic liquid on
silica, the so-called supported ionic liquid phase (SILP) catalyst, under hydrogen pressure.
4 Introduction
The combination of Pt and Lewis acidic ionic liquid can also be achieved by the addition of Pt
precursors to hallometallate ionic liquids and in situ reduction under hydrogen atmosphere of
the metal ions under hydrogen atmosphere. First experiments of this bifunctional catalyst
concept for the n-octane hydroisomerization at moderate reaction conditions were conducted
and a comparison was drawn between the two different bifunctional catalyst concepts and the
monofunctional acidic ionic liquid catalyst.
Catalyst immobilization strategies are indispensable for economically and ecologically
effective, industrially applied homogeneous catalysis. Different immobilization concepts for
catalysts using ionic liquids, namely multiphase reaction systems and SILP catalysts are
feasible. SCILL systems combine the immobilization of ionic liquid and the development of a
new catalyst concept: modification of the selectivity and / or activity by covering a
heterogeneous catalyst with an ionic liquid layer. Furthermore, ionic liquids can not only
stabilize but also immobilize metal nanoparticles. Throughout this work, the different
immobilization concepts are discussed and were investigated for their stability in terms of
leaching of catalyst components.
5
___________________________________________________________________________
2. General Part
6 General part
General part 2
2.1 Introduction to the alkane isomerization
Saturated hydrocarbons are the main components of crude oil [12] and play a key role for
energy supply and chemical feedstock demand. The straight-chain liquid hydrocarbons have a
very low octane number rendering them into undesirable gasoline components. To transform
these alkanes into chemicals suitable for gasoline, they have to undergo reactions like
isomerization or cracking. These reactions need acidic catalysts, noble metal catalysts or very
high reaction temperatures to activate the strong covalent C-H or C-C bonds. Isomerization is
of major economic importance as the branched C5 – C8 alkanes are the main components of
gasoline. The fundamentals of alkanes and alkane isomerization reaction as well as the
relevance of branched alkanes for high-quality fuel production are described in the following
chapter.
2.1.1 Fundamentals
The unreactive character of alkanes is explained by the unavailability of both, lone pairs and
empty orbitals. Carbon and hydrogen belong to the group of elements whose number of
valence electrons equals the number of valence orbitals available. The highest occupied
molecular orbitals (HOMOs) are deep lying σ bonding orbitals and the lowest unoccupied
molecular orbitals (LUMOs) are high lying σ* levels. Thus, the bonding between C and H is
very strong (dissociation energy: 377 – 461 kJ mol-1) [13]. The low polarity is another factor
contributing to the unreactivity of alkanes. Since branched alkanes are thermodynamically
more stable than linear alkanes, catalytic conversions to branched isomers are possible.
The criterion for the thermodynamical equilibrium under constant pressure and temperature in
one mixed phase is that the change of Gibbs free energy of every possible reaction in this
system is zero [14] (Equation 1).
∆ G = υi G� i=
i
υi µi=0
i
(1)
General part 7
The thermodynamical equilibrium composition of the five hexane isomers is depicted in
Figure 1 [15]. From a thermodynamic point of view, low reaction temperatures are favorable
to obtain highly branched and thus high-octane alkanes.
Figure 1: Temperature dependent equilibrium composition of hexane isomers [15].
The standard state reaction enthalpy (∆H�,���� ) can be calculated from the molar standard
state enthalpy of formation (∆H��,���� )i of every single reactant [16] according to Heß´s law
[14] (Equation 2).
∆ HR, 298 0 = ∑ υi (∆H� f,298
0)i i (2)
Alkane isomerization reactions are slightly exothermic. The reaction enthalpies ∆H�,���� for
all octane isomers, starting with reactant n-octane in each case, range from -16.98 kJ mol-1
to -1.96 kJ mol-1.
To summarize, high temperatures are necessary to activate and to convert the unreactive
alkanes, however, low temperatures favor the desired highly branched isomers.
The neccesity of alkane isomerization and the relevance of high-octane branched alkanes for
gasoline production are described in the following.
280 300 320 340 360 380 400 420 440 460 4800
10
20
30
40
50
60
70
Equ
ilibr
ium
com
posi
tion
/ mol
%
Temperature / K
n-hexane 2-methylpentane 3-methylpentane 2,2-dimethylbutane 2,3-dimethylbutane
8 General part
2.1.2 Relevance of alkane isomerization
Gasoline is a complex mixture of hydrocarbons which mainly consists of alkanes, with their
number of C atoms > 3, cycloalkanes, olefins and aromatics. The distilling range is between
300 K and 533 K at 1 bar [17]. The composition of gasoline can vary widely and depends on
the type of crude oil processed, the kind of refinery process and reaction conditions. Volume
fractions of the different hydrocarbons in gasoline are listed in Table 1 [18].
Table 1: Normal ranges for hydrocarbon types in gasoline [18].
Hydrocarbons Motor gasoline / vol. %
Alkanes 30 – 90
Cycloalkanes 1 – 35
Olefins 0 – 20
Aromatics 5 - 55
Gasoline is also blended with oxygenates, mainly ethers (e.g. methyl tert-butyl ether (MTBE))
and alcohols (e.g. ethanol). Additives, like anti-knock agents, anti-oxidants, corrosion
inhibitors, metal deactivators, combustion chamber scavengers, inlet system detergents,
carburettor anti-icing compounds and dyes, may be used to boost certain performance features
[19]. Hereby, the octane number indicates the anti-knock qualitiy of gasoline. It is defined as
the volume fraction (%) of iso-octane in a blend of n-heptane and iso-octane which produces
the same anti-knock intensity as the tested fuel. The higher the octane number, the more
compression the fuel can withstand before detonating. It is distinguished between the RON
and the motor octane number (MON). The same standard test engine is used for the RON and
MON determination however differences exist in the operating conditions. RON is measured
in an engine running at 600 rpm and a fuel / air mixture at 325 K, while the MON is
determined at 900 rpm and 422 K [20]. DIN EN 228 defines minimum values for both, the
RON and MON, which are shown for different types of gasoline in Table 2 [21].
General part 9
Table 2: Minimum values for the RON and MON of different types of gasoline according to DIN EN 228 [21].
Gasoline RON MON
Normal 91.0 82.5
Super 95.0 85.0
Super Plus 98.0 88.0
Knock results from premature self-ignition and combustion of certain molecules in the
fuel / air mixture before the ignition spark is reached during compression in the engine. The
expansive motion of the fuel / air mixture works against the compression stroke in the engine
and, hence, creates the knock. Thereby, the power of the engine is reduced and the engine
suffers damage. The anti-knock quality depends on the thermal stability of the molecules
which is, in turn, a function of its molecular structure [20]. The octane number of
hydrocarbons decreases with increasing carbon number and augments for a higher degree of
branching [22]. RONs for different linear and branched alkanes are listed in Table 24,
(appendix 7.4). Blending of the fuel with anti-knock agents like tetraethyl lead or aromatics
results in a higher RON. However, tetraethyllead has been replaced by MTBE with the
implementation of car exhaust catalysts. Application of MTBE is strictly reglemented in the
USA. Furthermore, environmental regulations limit the percentage of carcinogenic aromatics
and olefins in gasoline [2], [19]. One possibility to compensate the consequent loss of fuel´s
RON is the isomerization of linear alkanes to branched ones. Isomerization processes of
n-pentane, n-hexane and mixtures of both are already established in industry. Though, the
isomerization of C5 / C6 feedstock is not sufficient and the contribution of branched
long-chain alkanes with C > 6 to high-octane gasoline is necessary. Latest research activities
and catalyst developments of long-chain alkane isomerization reaction are presented in
chapter 2.6. The main focus of this chapter is on industrially applied isomerization processes
for the C5 / C6 feedstock.
2.1.3 Industrial processes
In 2009, the German production capacity of alkane isomerization, the skeletal isomerization
of saturated hydrocarbons, accounted for 94,226 b cd -1. The feedstock of this production
included C4, C5 and C5 / C6. For better evaluation, the charge capacitiy for crude distillation
10 General part
added up to 2,410,662 b cd -1 [23]. Butane isomerization process is not part of this work for
two reasons: firstly, isomerized butane is just an intermediate for the refinery alkylation,
which is the reaction of olefins with isobutane, to produce a high-octane gasoline, and
secondly, C4 and C > 4 alkanes are converted by different mechanisms.
Up to now, industrial processes for the isomerization of alkanes only exist for alkanes with
C ≤ 6 and can be classified according to the applied catalyst, into monofunctional and
bifunctional systems [24], [3]. The bifunctional systems consist of a noble metal for
hydrogenation / dehydrogenation on an acidic support, which catalyzes the skeletal
isomerization. In contrast, monofunctional catalysts exhibit only highly acidic active centers.
Detailed description of their catalytic mechanism follows in chapter 2.3. An overview of the
isomerization catalysts is given in Table 3.
Table 3: Classification of isomerization catalysts applied in industrial processes [24], [3].
Type Catalyst
Monofunctional AlCl 3 / HCl
Bifunctional Noble metal on chlorinated alumina
Bifunctional Noble metal on amorphous silica-alumina
Bifunctional Noble metal on an acidic form of zeolite
Bifunctional Noble metal on metal oxide
The processes of these categories are briefly presented in the following.
2.1.3.1 Processes using AlCl3 / HCl as monofunctional catalyst
The isomerization of Standard Oil Co. of Indiana is based on the monofunctional catalyst type
[24, 25]. The solid, supported AlCl3 catalyst is used in a fixed bed reactor. HCl acts as
promoter and is added with purified naphtha. These catalysts are sensitive to trace impurities
like water in the feed. Several reactors are necessary to replenish the deactivated catalyst
during operation to maintain constant catalytic activity and product yields.
Shell´s liquid-phase isomerization process applies the low-melting eutectic mixture of AlCl3
and SbCl3 which allows very mild isomerization temperatures (333 – 373 K) [26]. HCl is also
added to the naphtha stream in this process. The catalyst has to be separated in an extra
General part 11
column from the product. For this process fresh catalyst has to be added for activity
maintenance as well.
These Friedel-Crafts catalysts represent first-generation systems and suffer from corrosion
problems, high maintenance costs and low time-on-stream efficiency because of high catalyst
consumption. For this reason they have not been employed anymore [27]. Though, the low
reaction temperature leading to more favorable isomerization equilibrium is an advantageous
point of these processes.
2.1.3.2 Processes with bifunctional catalysts consisting of Pt supported on chlorinated
alumina
The Penex process of Universal Oil Products (UOP) is based on the bifunctional catalyst, Pt
supported on chlorinated alumina [28], [24], [3]. The reliability of this catalyst and process
has been commercially demonstrated since 1969 for C5 / C6 isomerization. The catalyst is
activated by chlorination with HCl. This is added as organic chloride which generates
hydrogen chloride in situ. Continuous addition of organic chloride activator maintains catalyst
activity. The desulfurized and dried feedstock C5, C6 or C5 / C6 passes the isomerization
reactors, commonly two in series, which are operated at 393 – 443 K, 20 – 70 bar and a
hydrogen to hydrocarbon ratio of 1:2. Due to the relatively low isomerization temperature,
highly branched products with a RON of about 84 – 85 can be obtained.
2.1.3.3 Processes based on noble metal on amorphous silica-alumina as bifunctional
catalyst
The Pentafining process of Atlantic Refining is an example including bifunctional catalysts
with amorphous silica-alumina and noble metal [29], [3]. A consequence of the relatively low
acid strength of the amorphous silica-alumina catalyst are high operating temperatures
(693 - 753 K), operating pressures between 20 - 50 bar and relatively low octane levels
because of the unfavorable thermodynamic equilibrium at these elevated temperatures. This
process has become obsolete and has been replaced because of the mentioned drawbacks.
2.1.3.4 Processes using noble metal on zeolites
Bifunctional catalysts based on acidic zeolites stand out by their robustness. In contrast to the
previously mentioned systems, they can withstand certain levels of water and sulphur [30],
[8]. These catalysts were firstly developed for the Shell Hysomer process [31], which is
12 General part
offered for licence by Shell and UOP and is operated with Pt on mordenite. Reaction
temperatures range between 513 – 553 K and isomerization is conducted in the presence of
hydrogen. The most advantageous features comprise long catalyst lifetimes, up to seven years
in the commercial Hysomer operation. n-Paraffins are not fully converted at these high
operating temperatures leading also to lower octane levels compared to the Penex process
(RON: 77 - 79 vs. 84 - 85). The normal paraffins are selectively adsorbed in the IsoSiv
process by molecular sieves. The unbranched alkanes are recycled to the isomerization reactor
after a desorption step. The IsoSiv process is basis of the joint Shell / UOP process, known as
TIP (Total Isomerization Process). Thus, the isomerization process can be combined
profitably with IsoSiv-iso / normal developed by UOP. The TIP process combines the Once-
Through Zeolitic Isomerization process, formely known as Hysomer process, with the IsoSiv
process. The product of the TIP process yields a RON of 87 – 90. This continuous gas-phase
process is operated between 518 – 643 K and at moderate pressures (14 – 34 bar) [3].
Hydrogen must be present at a sufficient partial pressure to prevent coking and deactivation of
the catalyst. The catalyst is regenerable using an oxidation procedure.
2.1.3.5 Processes using noble metal on metal oxides
Another possibility to isomerize C5 / C6 feedstock is the application of noble metal / metal
oxide catalysts as applied, for example, in the UOP Par-Isom process [3]. The first
commercial operation of this naphtha isomerization catalyst was in 1996 [32]. The catalysts
consist of sulfated zirconia, which can be classified as solid superacid. Therefore, this
catalytic system shows a higher activity compared to zeolitic based catalysts. The respective
continuous gas-phase process using a fixed bed reactor is operated between 453 – 513 K at
30 bar. Sulfated metal oxide catalysts are not as tolerant toward coordinating poisons in the
feedstock as zeolitic based ones. Nonetheless, these systems are not deactivated irreversibly
by water or oxygen and are fully regenerable under oxidative conditions. This Par-Isom
process increases the RON of light naphtha to 81.
In conclusion, isomerization catalysts and processes for C5 / C6 alkanes have been developed
and optimized over years, moreover already successfully operated. However, much research
effort has to be laid in the development of highly active, selective, as well as stable and robust
catalytic systems to isomerize long-chain alkanes. Thereby, difficulties arise from increased
cracking side reactions due to different cracking mechanisms for short (C5 / C6) and
long-chain alkanes (C > C6) (Scheme 9, chapter 2.3.1.4.1).
General part 13
2.2 Definition and classification of acids
All catalysts for skeletal isomerization consist of acids as only catalytically active material or
in combination with further active sites. Therefore, different types of acids and acidity scales
are presented in this chapter as they are inevitable for the understanding of alkane
isomerization reactions.
Arrhenius [33], Brønsted [34] and Lowry [35] defined acids as proton donors and bases as
proton acceptors, nowadays known as Brønsted acids and bases, respectively. Brønsted
acidity is expressed quantitatively in aqueous solution as the pH value, that is –lg a (H+, aq.).
Lewis extended and generalized the acid-base concept [36, 37]. He defined an acid as
substance that can accept electrons, like AlCl3, and a base as substance that can donate
electrons, for example NH3. Several procedures have been developed to evaluate the strength
of a Lewis acid [38, 39]. Since the work of Bartlett and coworkers [40], it is known that the
fluoride ion affinity (FIA) is a reliable measure of the Lewis acidity, combining the strength
of a Lewis acid A(g) with the energy that is realeased upon binding a fluoride ion F-(g)
(Scheme 1) [41, 42].
A(g) + F-(g) AF-
(g)
Scheme 1: Reaction of a Lewis acid A with a fluoride ion F- [43].
The FIA is defined as the negative of the reaction enthalpy ∆HR. Therefrom, the strength of a
Lewis acid corresponds to the absolute value of FIA.
In 1927, the term “superacid” was introduced in the chemical literature by Hall and Conant
[44, 45]. In the 1960´s, Olah´s studies on carbocations, amongst others, focused on highly
acidic non-aqueous systems [46, 47]. Gillespie proposed an arbitrary and widely accepted
definition of superacids, defining them as any acid system that is stronger than 100 % sulfuric
acid, that is H0 ≤ -11.9 (Equation 3) [48-51]. Gillespie´s definition of superacids relates to
Brønsted acid systems. Hammett function and the H0 value were originally developed by
Hammett and Deyrup to measure the protonation of weakly basic indicators in acid solution
[52, 53]. It is derived from ionization equilibria of a particular class of indicators which
behave as uncharged bases in the Brønsted-Lowry sense (Scheme 2). H0 is defined by
Equation 3.
∆HR = -FIA
14 General part
Scheme 2: Ionization equilibrium of an indicator base.
H0 ≡ -logKBH+- logcBH+
cB (3)
with KBH+ = aB · aH+
aBH+
H0 is currently the most common parameter for quantifying the acidity of superacidic
solutions. Olah suggested to define Lewis superacids as those that are stronger than anhydrous
AlCl 3 in their reactivity [54]. Another proposal for the definition of the term Lewis superacid
can be found in literature: molecular Lewis acids which are stronger than monomeric SbF5 in
the gas-phase are Lewis superacids [43]. In contrast, the acidity of solids is still not well
known and difficult to measure. Numerous literature related to acidity characterization of
solids is available, for example reviews of Corma [55] or Farneth and Gorte [56]. Superacids
can be classified in solid and liquid superacids. The latter ones are grouped in detail according
to their number of components in the superacidic system, primary, binary as well as ternary
superacids. An overview of all different types of superacids is given in Table 4.
General part 15
Table 4: Types of superacids [54].
Type Examples
Primary superacids Brønsted superacids (HSO3Cl, HClO4)
Lewis superacids (SbF5, TaF5)
Binary superacids Binary Brønsted superacids (HF-HSO3F, HF-CF3SO3F)
Conjugated Brønsted-Lewis superacids (HBr / AlBr3, HCl / AlCl3, HF / SbF5; HSO3F / SbF5)
Ternary Superacids HSO3F-HF-SbF5
Solid Superacids Zeolitic acids
Polymeric resin sulfonic acids complexed with Lewis acids and perfluorinated polymer (e.g. Nafion-H)
Brønsted and Lewis acid-modified metal oxides and mixed oxides; as well as metal salts complexed with Lewis acids
Immobilized superacids and graphite-intercalated superacids
The first proposals for an unified acidity scale which would make it possible to compare
quantitatively acidity in spite of different media date back to the 1950s [57]. Due to
experimental complications, these proposals along with further approaches to thermodynamic
acidity values are not widely applied [58]. The group of Krossing proposed an unified
Brønsted acidity scale on the basis of the absolute chemical potential µabs of the proton in any
medium (gas, liquid, solid) according to Equation 4 [58, 59].
µabs�H+, medium�= µ�H+�-µ0 (H+, gas-phase, 298.15 K) (4)
with µabs�H+,gas� ≡ 0 kJ mol-1
16 General part
2.3 Mechanism of the alkane isomerization
The isomerization of n-alkanes proceeds by a mechanism with carbenium ions as key
intermediates. Carbenium ions are not only active species for the skeletal isomerization but
also for undesired side reactions which reduce the selectivity to the desired isomers. The
alkane isomerization mechanisms of acid and bifunctional catalysts are described in this
chapter, at first of the monofunctional homogeneous and then of the heterogeneous acid
catalysts. Bifunctional heterogeneous catalysts contain acid sites for skeletal isomerization
and metallic sites for hydrogenation / dehydrogenation reactions [60-65]. The isomerization
mechanism on the acidic sites of these bifunctional catalysts corresponds to the mechanism of
the monofunctional acidic catalysts. Thus, only mechanistic differences between the
monofunctional acid and bifunctional acid / metal catalysts are described in chapter 2.3.2.
2.3.1 Mechanism of the acid catalyzed alkane isomerization
The catalytic cycle of the acid catalyzed isomerization is described, involving chain initiation
to form the first active carbenium ion species, carbenium ion rearrangement and the chain
propagation. This chapter includes also possible side reactions and mechanistic considerations
of the cracking inhibitor hydrogen in the presence of carbenium ions.
2.3.1.1 Initial step of the alkane isomerization reaction
Olah et. al studied intensively superacids and hydrocarbon chemistry with these superacidic
catalysts, amongst others also the alkane isomerization [15, 54]. The initial step, namely the
formation of the carbenium ion, occurs in superacids by abstracting a hydride ion according to
Scheme 3.
Scheme 3: Formation of the carbenium ion in superacids [54].
Without an appropriate level of acidity, the isomerization reaction may occur at much lower
rate or even do not take place at all. Sulfuric acid (H0 = -11.9), the threshold to superacidity, is
not acidic enough for the skeletal isomerization of n-alkanes [15].
General part 17
Another possibility of carbenium ion generation involves the use of organic halides in the
presence of strong Lewis acids (Scheme 4) [66]. Lewis acids are capable of removing the
halide. This catalyst system is well known from the Friedel-Crafts chemistry [67].
Scheme 4: Generation of carbenium ion by halide removing from an organic halide RX using a Lewis acidic catalyst AlX3 [66].
In addition, acidic catalysts are capable of alkene protonation, which is shown in Scheme 5.
Scheme 5: Carbenium ion formation by the protonation of an alkene using acidic catalysts [66].
The generation of alkenes in turn depends on reaction intermediates and catalyst types. In
acidic, metal free catalysts, alkenes can originate from cracking side reactions [61] or can be
formed by deprotonation of carbenium ions. The carbenium ion is in equilibrium with the
corresponding alkene in every acidic media (Scheme 5) [66]. Another alkene origin is the
presence of olefin impurities in the feed.
Hydride abstraction by alkyl carbenium ions acting as strong Lewis acids is well known in
hydrocarbon chemistry as chain propagation step (Scheme 8) [54, 68]. Simultaneously, a new
carbenium ion is generated.
2.3.1.2 Skeletal isomerization of carbenium ions
It is convenient to distinguish between type A (non-branching) and type B (branching)
rearrangements of alkylcarbenium ions [69]. Type A rearrangements are generally much
faster than type B rearrangements [70]. It is widely accepted that type A rearrangements
proceed via classical alkyl and hydride shift (Scheme 6) and type B via non-classical
protonated cyclopropane (PCP) rearrangement (Scheme 7). Application of the classical
mechanism for type B rearrangement is implausible as it would result in primary carbenium
ions as intermediates which are known to have high energy contents. Only n-alkanes larger
than n-butane can be isomerized via the PCP intermediate. Evidence for this PCP mechanism
was provided by Brouwer [69] and Weitkamp [71, 72].
18 General part
The stability of carbenium ions depends on the number of organic rests on the positively
charged carbon. The stability of carbenium ions increases in the following order because of
the electron donating field effect (+I) of the alkyl groups: primary R+ < secondary R+ <
tertiary R+. Therefore, carbenium ions rearrange to the thermodynamically more stable
carbenium ion species.
Scheme 6: Type A (non-branching) skeletal rearrangement by hydride (a) or alkyl shift (b) (R1, R2, R3 = organic rest) [66].
Scheme 7: Type B (branching) skeletal rearrangement of a carbenium ion via the protonated cyclopropane (PCP) intermediate [54].
2.3.1.3 Chain propagation
The rearranged carbenium ions can abstract a hydride from an alkane to give the
corresponding iso-alkane and a new carbenium ion (Scheme 8) [54]. Therewith, the catalytic
cycle is closed.
General part 19
Scheme 8: Hydride-transfer between the carbenium ion and alkane [54].
This catalytic cycle is controlled by two kinetic and two thermodynamic parameters [54].
Skeletal isomerization of carbenium ions is kinetically controlled by the relative rates of
hydrogen shift, alkyl shift as well as PCP formation and it is thermodynamically controlled by
the relative stabilities of the carbenium ions. The chain propagation, however, is kinetically
controlled by the hydride transfer from excess of starting hydrocarbons and by the relative
thermodynamic stabilities of the various hydrocarbon isomers. The stability of intermediate
carbenium ions is different from that of hydrocarbons.
2.3.1.4 Possible side reactions in acidic isomerization catalysis
Carbenium ions are not only key intermediates for the isomerization but also for some side
reactions like cracking or alkylation. Thus, selectivity to the desired isomerized alkanes
decreases and the thermodynamical equilibrium within the isomers is never reached [4]. Many
parallel and consecutive reactions take place during acid catalyzed alkane isomerization. Even
with a pure model compound, like n-hexane or n-octane, a large number of products are
formed.
2.3.1.4.1 Catalytic cracking on monofunctional acidic catalysts
Two different mechanisms can be distinguished for the alkane cracking, the bimolecular
cracking via carbenium ions and the monomolecular cracking via carbonium ions. The
general concept of carbocations includes all cations of carbon-containing compounds, which
can be differentiated into two limiting cases: (i) trivalent (“classical”) carbenium ions and (ii)
five or higher coordinated (“nonclassical”) carbonium ions [54].
For bimolecular cracking the carbenium ion formation is followed by a bond scission [61, 72-
74]. The C-C-bond cleavage occurs in β-position to the positive charge of the carbenium ion,
whereby a new carbenium ion and an alkene is formed. Depending on the type of cracked
carbenium ion and the new formed carbenium ion (primary, secondary or tertiary), different
modes of β-scission can be distinguished (type A, type B1, type B2, type C and type D)
according to Scheme 9 [61]. Chain length and degree of branching of the carbenium ion
determine the mode of β-scission. Rate constants strongly decrease from type A to type D
20 General part
β-scissions [72, 75]. Theoretically, hexane isomers can only crack according to type D and C,
whereas type D is unlikely at mild temperatures (below 523 K) because of the high energy
content of the primary carbenium ion intermediate. On the contrary, all modes of β-scission
are open for octane isomers. Therefrom, lower selectivities for long-chain alkanes (C > 6) and
especially for multi-branched long-chain alkanes are the consequence.
Scheme 9: Modes of β-scission of alkylcarbenium ions [61]; parameter m is the minimum number of C-atoms of the carbenium ion to crack according to the corresponding mode of β-scission.
In the monomolecular cracking, called Haag and Dessau´s mechanism, a solid acidic catalyst
protonoates an alkane to give carbonium ion transition states that collapse to produce alkanes
(or hydrogen) and carbenium ions [76, 77]. The cracking products include hydrogen, methane
and ethane, in contrast to those of classical catalytic cracking. Reaction temperatures about
800 K are necessary for the Haag-Dessau-cracking mechanism. Protolytic cracking on solid
acids is only kinetically significant when alkene concentrations are low because alkenes are
much better proton acceptors than alkanes. Alkene protonation leads to classical catalytic
cracking.
Dec
reas
ing
rate
con
stan
ts o
f β-s
ciss
ions
General part 21
Scheme 10: Haag–Dessau cracking mechanism for an alkane molecule (RH) proceeding via a carbonium ion transition state [77].
2.3.1.4.2 Alkylation reaction
Another different type of side reaction involving carbenium ions as active species is the
alkylation reaction [78]. An alkene is alkylated by a carbenium ion, producing a new and
higher molecular carbenium ion. Again, the alkene can be formed by cracking via β-scission
or deprotonation of carbenium ions. The catalytic cycle is also maintained by hydride transfer
from an alkane to the carbenium ion.
Scheme 11: Alkylation of isobutane with 2-butene.
2.3.1.4.3 Coke formation on solid acidic catalysts
Formation and retention of heavy side products takes place on solid catalysts, in the pores, on
the outer surface or on both. The formation of these non-desorbed products is called coke and
represents the most frequent cause of catalyst deactivation in industrial processes. It is
differentiated between coke formation at low temperatures (< 473 K) and high temperatures
(> 623 K) [7]. Low temperature coke is non-aromatic. Carbonaceous deposits are directly
formed from unsatured compounds like alkenes or cyclic alkenes in
oligomerization / polymerization reaction but not from alkanes and cycloalkanes.
H+
RH2+
R2+
RH R1H
desorption
carbonium ion
alkene
22 General part
2.3.1.5 Cracking inhibitor hydrogen
Hydrogen can react with carbenium ions. This fact was shown for liquid superacidic catalysts
[79] and for solid metal free acidic catalysts [80-83]. Thus, hydrogen pressure can be
responsible for higher isomer selectivity circumventing consecutive cracking reactions of
branched alkanes. However, hydrogen can also have a negative influence on the rate of
isomerization and cracking because it is the reverse reaction of carbenium ion generation
(Scheme 3). A possible mechanism for the hydrogen – carbenium ion reaction was proposed
based on theoretical calculations: hydrogen and carbenium ions can interact via the formation
of the H2-carbenium ion-complex with subsequent evolution to the corresponding alkane and
proton [80].
2.3.2 Alkane hydroisomerization using bifunctional or monofunctional
metal catalysts
Bifunctional catalysts contain acid sites for skeletal isomerization and metallic sites for
hydrogenation / dehydrogenation reactions [60-65]. The main function of the metal is the
dehydrogenation of saturated reactant molecules to alkenes and the hydrogenation of olefinic
intermediates desorbed from acidic sites. The mechanisms on the acidic part of the catalysts
are widely accepted to be identical to those described for monofunctional catalysts.
2.3.2.1 Initial step of the alkane isomerization reaction using bifunctional catalysts
In the case of bifunctional catalysts, alkanes can be dehydrogenated on the metal catalyst [27,
84, 85]. The resulting alkenes desorb from the metal sites and diffuse to Brønsted acid sites
where they are protonated according to Scheme 5 to form the active carbenium ions [72].
2.3.2.2 Chain propagation
In the case of bifunctional catalysts, rearranged carbenium ions can desorb regenerating the
initial Brønsted acid site and and produce an iso-olefin. This iso-olefin is hydrogenated on the
metal part of the catalyst [4, 86, 87].
2.3.2.3 Additional possible side reactions using bifunctional and monofunctional metal
catalysts
Like for the monofunctional catalysts side reactions occur which lower the catalysts´s
selectivity as well as stability.
General part 23
2.3.2.3.1 Hydrocracking
Hydrocracking is mechanistically similar to catalytic cracking (chapter 2.3.1.4.1), especially
regarding the key step of C-C bond cleavage. However, different are the presence of hydrogen
and thus fully saturated products. Ideally, no catalyst deactivation occurs in hydrocracking as
carbonaceous deposits are hydrogenated. Cracking or hydrocracking reactions lead to
different distributions of the products [72].
2.3.2.3.2 Hydrogenolysis
Hydrogenolysis occurs on monofunctional metal catalysts. Hydrogenolysis is the C-C bond
splitting followed by hydrogenation of the fragments [88]. If methane and ethane is not part of
the cracked products, any hydrocracking mechanism on the noble metal (hydrogenolysis) can
be excluded as potential side reaction [89].
2.3.2.3.3 Coke formation on bifunctional catalysts
Low temperature coke formation on bifunctional metal / acid catalysts results from
conversion of reactive intermediates on the acid sites of the support or from products formed
on the metalic sites [7]. In comparison to monofunctional heterogeneous acidic catalysts,
oligomeric coke precursors can be hydrogenated on the metal part of the catalyst in the
presence of hydrogen [90] resulting in higher catalyst stability.
2.4 Introduction to acidic ionic liquids and their application in
catalysis
Ionic liquids consist completely of ions and have, by definition, a melting point of Tm < 373 K
[10, 91] as well as very low vapor pressures when used below their decomposition
temperatures [92]. The vapor pressure of ionic liquids was firstly measured by the group of
Rebelo [93] and Paulechka [94] by high vacuum effusion measurements in 2005. In 2006,
Earle showed that 1-ethyl-3-methylimidazolium bis(trifluoromethanesulfonyl)imide
([EMIM][NTf 2]) and 1-butyl-3-methylimidazolium bis(trifluoromethanesulfonyl)imide
([BMIM][NTf 2]) can be distilled without decomposition at pressures below 0.01 bar and
temperatures of about 573 K [95]. Changing physicochemical and catalytic properties by
simple anion or cation variation opens up a broad field of applications. Therefore, the name
“designer solvents” is often attributed to ionic liquids [96]. Catalysis in ionic liquids was
reviewed in detail by Welton [10, 97, 98], Seddon [99], Wasserscheid [91, 100], Olivier-
24 General part
Bourbigou [101, 102], Dupont [103] and Cole-Hamilton [104]. Literature review of this
chapter focuses on acidic halometallate ionic liquids, especially on chloro- and
bromoaluminates.
2.4.1 Characteristic properties of acidic halometallate ionic liquids
Chloroaluminate ionic liquids, low-temperature molten salts, were discovered in the 1940s
[105-108]. The application of mixtures of 1-ethylpyridinium bromide / aluminum chloride
([Epyr]Br / AlCl3) to the electrodeposition of aluminum was described. The system
[Epyr]Br / AlCl3 is only liquid at or below room temperature with compositions between 63
and 68 mol. % of AlCl3. Not until 1975, Osteryoung and Gilbert studied the chemical and
physical properties of ionic liquids made from 1-butylpyridinium chloride / aluminum
chloride ([Bpyr]Cl / AlCl3) [109]. This molten salt is liquid below 313 K in case of a AlCl3
content between x(AlCl3) = 0.43 – 0.66 [110, 111]. First applications of chloroaluminate ionic
liquids included mostly battery electrolytes [112-115]. The synthesis of the ionic liquid
[EMIM]Cl / AlCl 3 established a system with a liquid range much wider than that of Hurley
and Wier [107, 108] or Osteryoung [109]: it is a liquid at room temperature from
x(AlCl3) = 0.33 - 0.67 [112, 116-121]. Depending on the mole fraction of AlCl3,
chloroaluminate systems are Lewis basic at x(AlCl3) < 0.5, neutral at x(AlCl3) = 0.5 or Lewis
acidic at x(AlCl3) > 0.5 [10]. Well-known reactions occuring in chloroaluminate ionic liquids
and their corresponding equilibrium constants are listed in Table 5. Hence, predominant
species in basic melts are Cl- and [AlCl4]- and in acidic melts [Al2Cl7]
- and [AlCl4]-. Anionic
species that are present in the chloroaluminate ionic liquid as function of the mole fraction of
AlCl 3 are also shown in Figure 2. The maximum amount of AlCl 3 is limited by the liquid
range of the chloroaluminate ionic liquid to two mol AlCl 3 per mol [cation]Cl [121].
Table 5: Thermodynamic equilibrium parameter of reactions occuring in chloroaluminate ionic liquids [117].
Reaction log K (473 K)
I [AlCl 4]- Cl- + AlCl3 -16.67
II [Al 2Cl7]- [AlCl4]
- + AlCl3 -2.54
III [Al 3Cl10]- [Al 2Cl7]
- + AlCl3 -0.90
IV [Al 4Cl13]- [Al 3Cl10]
- + AlCl3 -0.34
General part 25
Figure 2: Anionic species at 473 K in the chloroaluminate ionic liquid [EMIM]Cl / AlCl 3 as function of the mole fraction x (AlCl3), calculated from thermodynamical model. X1 = Cl-,X4 = [AlCl 4]
-, X6 = Al2Cl6, X7 = [Al2Cl7]-, X10 = [Al3Cl10]
-, X13 = [Al4Cl13]-
[117].
The extremely hygroscopic organic chlorides and the anorganic AlCl3 used for the synthesis
of these chloroaluminate ionic liquids are unavoidably contaminated by small quantities of
water [11, 122, 123]. Accordingly, the resulting ionic liquid contains millimolar quantities of
both oxy / hydroxychloroaluminate species and molecular HCl when the melts are mixed with
AlCl 3 [123-126]. Only phosgene (COCl2) can remove all oxide contaminations from both,
basic and acidic, chloroaluminate ionic liquids [127]. Consequently, the chemistry and
properties of chloroaluminate ionic liquids with HCl was in the focus of many articles.
Brønsted superacidity of HCl in the acidic ionic liquid [EMIM]Cl / AlCl 3 [11] and
quantitative studies of the acidity of HCl in [EMIM]Cl / AlCl3 as a function of HCl pressure
and ionic liquid composition (51.0 – 66.4 mol. % AlCl3) were described by Smith [128].
Johnson constructed a phase diagram of the ternary system HCl / [EMIM]Cl / AlCl3 under
conditions of ambient temperature and pressure [129]. Until today, it is contrarily discussed, if
molecular HCl or other species are the active superacidic species in acidic chloroaluminate
ionic liquids [122, 129-133]. Since it has been shown that [AlCl4]- and [Al2Cl7]
- are only poor
hydrogen-bond acceptors [134], it is likely that the interaction with HCl is relatively weak.
Maybe it is a highly dynamic system with the HCl being solvated by the best hydrogen-bond
acceptor ions available in the ionic liquid [10]. Osteryoung showed that it is possible to
remove proton impurities from the melt by evacuation overnight at a pressure significantly
less than the HCl equilibrium pressure (1.3-9 – 1.3-10 bar) [123]. This ultra-high vaccum, in
turn, shifts the ionic liquid proton / oxide equilibrium toward the formation of HCl.
mol %
26 General part
Ethylaluminum dichloride is also capable to remove protons [124]. Though, it is difficult to
add exactly the required mass of the aluminiumalkyl to eliminate all protic impurities and to
avoid excess alkylaluminum chloride in the ionic liquid which changes the properties of the
ionic liquid [135, 136].
Acidic ionic liquids are not restricted to the chloroaluminate system [cation]Cl / AlCl3. More
acidic systems can be obtained by substituting the chloride ion with a bromide ion. The
Hammett acidity H0 of the bromide system is lower ([EMIM][Al2Br7] / HBr): H0 = -17)
compared to the chloroaluminate system ([EMIM][Al2Cl7] / HCl): H0 = -15) [137]. The first
room temperature bromide-based ionic liquid [EMIM]Br / AlBr3 was characterized by Hussey
[138]. Ma and Johnson prepared ionic liquids of the type trimethylsulfonium bromide
/ aluminum bromide / hydrogen bromide ([TMSu]Br / AlBr3 / HBr) and trimethylsulfonium
bromide / aluminum chloride / hydrogen bromide ([TMSu]Br / AlCl3 / HBr) [139]. They
examined to what extent a series of aromatic hydrocarbons is protonated from which the
liquid´s acidity was determined. Actual H0 values can be assigned by reference to HF. If a H0
value of -10.7 was taken for HF, the H0 values for the systems of AlCl3 / [TMSu]Br / HBr
(molar ratio AlCl3 / [TMSu]Br = 2/1) and AlBr3 / [TMSu]Br / HBr (molar ratio
AlBr3 / [TMSu]Br = 2/1) at 1 bar of HBr are about -14 and -16 to -17, respectively.
2.4.2 Acidic halometallate ionic liquids – typical applications in catalysis
A number of key industrial processes in oil refinery, petrochemistry and chemistry are
acid-catalyzed. The use of AlCl3 results in high levels of corrosive waste water which has to
be treated and in the impossibility of catalyst recycling. Lewis acids like AlCl3 or AlBr3 can
be directly transferred to the corresponding acidic chloroaluminate or bromoaluminate ionic
liquids which form polynuclear anionic species. Thereby, the Lewis acids are “immobilized”
in the ionic liquid phase. Chloroaluminate ionic liquids are successfully applied in catalysis
because of the possibility to tune their acidity by varying the ratio between [cation]Cl and
AlCl 3, their low freezing points and their wide liquid composition ranges. Important examples
of acidic halometallate ionic liquids acting as catalyst and solvent with the focus on
chloroaluminate ionic liquids are described in the following chapter. Literature review of the
acidic ionic liquid catalyzed alkane isomerization is presented separately in chapter 2.6.2.
General part 27
2.4.2.1 Friedel-Crafts reactions
Friedel-Crafts reactions are conversions which form one additional C-C-bond with an
aromatic compound catalyzed by strong Lewis acids. Important reactions are Friedel-Crafts
alkylation, Friedel-Crafts acylation and carbonylation.
Friedel-Crafts alkylation (Scheme 12) of benzene using halogenalkanes were among the
earliest reactions to be investigated in chloroaluminate ionic liquids using the acidic ionic
liquid [EMIM]Cl / AlCl 3 [140]. Primary and secondary halogenalkanes were applied yielding
polyalkylated products. An excess of aromatic substrate favors the formation of
monoalkylated products.
Scheme 12: Friedel-Crafts alkylation.
BP patented the synthesis of alkylated aromatics, the reaction of an olefin with benzene, in
acidic ionic liquids [141]. Advantages of this process compared to the reaction with AlCl3 in
organic solvents are simple product separation, catalyst recycling and higher selectivities to
alkylation products.
The alkylation of benzene with linear olefins (C10 – C14) is largely used industrially to
produce linear alkylbenzenes. They serve as precursors for alkylbenzenesulfonates, which are
employed as surfactants and detergents intermediates [137]. Akzo Nobel developed the ionic
liquid catalyst and solvent trimethylammonium chloride ([Me3NH]Cl) / AlCl3
(x(AlCl3) = 0.67) for the alkylation of benzene with 1-dodecene, a cheaper alternative to
imidazolium based ionic liquids [142]. This reaction proceeds in a biphasic mode implying
simple catalyst recovery and recycling. Moreover, this alkylation can also be performed with
chloroaluminates immobilized on different solid supports (silica, alumina or zirconia) [143].
Varying mole fractions of AlCl3 in the acidic ionic liquid leads to different ionic liquid
acidities. The effect of ionic liquid acidity on the yield and selectivity was tested in the
alkylation of diphenyl oxide with 1-dodecene using [BMIM]Cl / AlCl 3 [144]. Ionic liquids
with x(AlCl3) ≤ 0.55 were not catalytically active. A maximum yield (about 90 %) of
monoalkylated product was achieved with x(AlCl3) = 0.6 (Treaction = 353 K). Ranking of ionic
28 General part
liquids with mixed halogenaluminate anions according to their Lewis acidity gives the
following order: 1-dodecyl-methylimidazolium ([DMIM])[Al 2Cl6Br] <
1-octyl-methylimidazolium ([OMIM])[Al2Cl6Br] ≈ [BMIM][Al 2Cl7] < [BMIM][Al 2Cl6Br] <
[BMIM][Al 2Cl6I] [145]. These different ionic liquid acidities were compared using the results
of the alkylation of benzene with 1-dodecene. Lewis acidity alone could explain neither
conversion nor the selectivity of the reaction. Furthermore, ionic liquids with more Lewis
acidic anions are less hydrolytically stable.
Detailed kinetic investigations of the reaction of cumene with propene in [EMIM]Cl / AlCl3
(x(AlCl3) = 0.67) were conducted by Joni in a liquid-liquid biphasic reaction mode [146].
Various products (di-, tri- and tetraisopropylbenzene) result from a series of consecutive
alkylation reactions. It is necessary to take the solubility of these products into account to fit
the kinetic model to the data. A COSMO-RS model was used to predict the relative
solubilities of the product. Higher alkylated products are less soluble in the reactive ionic
liquid phase, leading to an improved selectivity for the monoalkylated product.
Polysubstitution is no issue in the Friedel-Crafts acylation (Scheme 13) in contrast to the
Friedel-Crafts alkylation. Friedel-Crafts acylation is selective for the monosubstituted ketonic
product which is deactivated for further electrophilic substitutions.
Scheme 13: Friedel-Crafts acylation.
Aromatic ketones are important fine chemicals and / or intermediates. A Lewis acid, e.g.
AlCl 3, is required to generate active acylium ions out of the acylation agent. Usually acid
chloride or acid anhydride, which attacks the aromatic compound, is used. The carbonyl
oxygen of the product forms an adduct with the highly oxophilic AlCl3. Thus, stoichiometric
amounts of AlCl3 related to the acylation agent are consumed and the product must be
liberated by hydrolysis.
First catalytic aromatic acylations in ionic liquids were performed in acidic chloroaluminates
[140]. Ionic liquids were extensively described later on as solvent to perform acylation
reaction [147]. Reactions of acetylchloride with carboxylic aromatic compounds in the acidic
ionic liquid [EMIM]Cl/AlCl 3 (x(AlCl3) = 0.67) were conducted and compared with the results
General part 29
of conventional molecular solvents [148]. The reaction still required a high amount of catalyst
as the Lewis acidic oxophilic ionic liquid forms also an adduct with the carbonyl oxygen of
the product [98, 149]. Hence, the focus of this research area lies on alternative systems which
are not based on chloroaluminates to overcome the hydrolysis step for product recovery [98].
Another important Friedel-Crafts reaction is the carbonylation of aromatic compounds
(Scheme 14). In 1985, Texaco patented the reaction of toluene with carbon monoxide to
p-tolualdehyde in a Lewis acidic ionic liquid [150]. Brausch used a new type of highly acidic
ionic liquid with the formula [cation][NTf2] / AlCl 3 for the carbonylation of toluene which
resulted in selectivities of about 85 % to p-tolualdehyde [151]. Product isolation was carried
out by a hydrolysis step. Though, the hydrophobic ionic liquid [cation][NTf2] could be
recycled after the hydrolysis step and reloaded with AlCl3. The group of White investigated
the toluene carbonlyation at room temperature in the acidic ionic liquids [EMIM]Cl / AlCl3 or
[BMIM]Cl / AlCl 3 combined with HCl [152]. High selectivities to the desired product
p-tolualdehyde could be achieved even at very high conversions. Yet, solid formation was
found at high conversion. The association of chloroaluminate anions with the tolualdehyde
was suggested as possible reason.
Scheme 14: Carbonlyation of toluene.
Carbonylation of benzene to benzaldehyde in different acidic ionic liquids of the type
[BMIM]Br or [BMIM]Cl with AlCl 3 resulted in high selectivities (96 %) toward
benzaldehyde at relatively high yields (91 %) [153]. After the reaction, benzaldehyde was
present in a complex with the acidic aluminate ionic liquid. Dichloromethane (DCM) and
ethyl ether were added to the reaction mixture for product extraction. 10 wt. % of the initial
AlCl 3 had to be added after each recycling to compensate the loss of AlCl3 by hydrolysis
during the separation process. A decreasing yield of benzaldehyde with each run down to
72 % in the fifth recycling run referred to the initial yield was observed.
30 General part
2.4.2.2 Alkylation
Alkylation (Scheme 15) or “refinery alkylation” is the reaction of isobutane with short-chain
olefins (C3 – C5) in the presence of highly acidic catalysts. Alkylates are particularly suited
gasoline-blending components due to their, nonaromatic, high-octane and paraffinic nature.
Catalysts based on acidic ionic liquids could be alternatives for processes based on sulfuric
and hydrofluoric acids which show drawbacks from corrosion, environmental and
toxicological point of view.
The group of Chauvin from the Institute Français du Pétrole (IFP) published the alkylation of
isobutane with ethene in the acidic ionic liquid [BMIM]Cl / AlCl 3 [154] and the alkylation of
isobutane with 2-butene [155]. Acidic chloroaluminate ionic liquids are highly-suited
catalysts for alkylation because of their tuneable Lewis acidity by AlCl3 mole fraction
variation [155]. Low temperature and fine-tuning of the acidity are required for the alkylation.
Cracking reactions are favored at high ionic liquid acidities and heavy byproduct formation,
which occurs preferably at lower ionic liquid acidities. Continuous butene alkylation was
performed by the group of Chauvin for 500 h without any loss of catalytic activity and
selectivity [156]. Thereby, sufficient level of mixing was required for high selectivities and a
good quality alkylate. Moreover, they found out that addition of CuCl to the acidic
chloroaluminate ionic liquid was beneficial for the reaction performance.
Scheme 15: Refinery alkylation.
A detailed study of the alkylation of isobutane with 2-butene in acidic ionic liquids using
[CnMIM]X / AlCl 3 (with n = 4, 6 or 8; X = Cl, Br, or I) was conducted by Yoo et al. [157].
[OMIM]Br / AlCl 3 showed highest catalytic activity. Catalytic results obtained with
[OMIM]Br / AlCl 3 were compared to those achieved with sulfuric acid catalyst. The authors
observed a more active ionic liquid catalyst which also exhibited slower deactivation due to
less heavy byproducts. However, lower concentrations of trimethylpentanes were produced
with the ionic liquid catalyst. Both results can be explained by the higher acidity of
[OMIM]Br / AlCl 3 compared to sulfuric acid. Moisture sensitivity of these ionic liquids led to
gradual deactivation of the ionic liquid catalyst and solvent.
General part 31
PetroChina has introduced a commercial alkylation process based on the composite catalyst
system chloroaluminate ionic liquid with CuCl. The addition of CuCl led to higher
selectivities to octanes and a higher trimethylpentanes / dimethylhexanes ratio [158, 159]. In
2006, this process replaced an existing 65,000 t a-1 sulfuric acid alkylation plant in China
[160].
The group of Jess investigated isobutene / 2-butene alkylation with promoted Lewis acidic
ionic liquid catalysts [BMIM]Cl / AlCl3 (x(AlCl3) = 0.64) [161]. The promoters tert-butyl
bromide and tert-butyl chloride were added to the Lewis acidic ionic liquid. Results were
compared to those obtained with neat Lewis acidic ionic liquid catalyst, sulfuric acid and to
catalytic runs in which water was added to the Lewis acidic ionic liquid. The reactions with
tert-butyl halides as promoter were accelerated by up to two orders of magnitude. Pure Lewis
acidic ionic liquid catalysts, [BMIM]Cl / AlCl3 with the tert-butyl halide promoter, were more
active than Brønsted acidic catalysts, water in [BMIM]Cl / AlCl 3 or sulfuric acid. Another
result of these experiments was the formation of a Brønsted acidic ionic liquid catalyst system
to some extent with increasing catalyst deactivation that is with increasing decomposition of
tert-butyl cation into H+ and isobutene. Catalyst deactivation yielded in inconstant product
selectivities.
2.4.2.3 Lewis acidic ionic liquid catalyzed oligomerization and polymerization
Acidic chloroaluminate ionic liquids are known to catalyze oligomerization / polymerization
(Scheme 16) of olefins via a cationic reaction mechanism even in the presence of the proton
scavenger ethylaluminum dichloride [162-164].
Scheme 16: Olefin oligomerization / polymerization.
Two mechanisms of the carbocation formation from olefins are possible in acidic
chloroaluminate ionic liquids. These are shown in Scheme 17.
32 General part
Scheme 17: Mechanisms of cation formation from olefins in acidic chloroaluminate ionic liquids.
In 1995, BP patented the polybutene synthesis out of a butene-rich C4 cut such as Raffinate I
or Raffinate II [165]. This technology offers many advantages compared to AlCl3 or BF3
catalysts: simple product separation due to biphasic reaction mode, increased rate of
polymerization, minimization of consecutive reactions of the olefinic end-group in the
polymer (e.g. isomerization) as the polymer product forms a separate phase, higher molecular
weight of the polymer even at higher reaction temperatures and incorporation of butene in the
polymer product. Thus, the direct use of Raffinate is possible without prior isolation of
isobutene.
In 2000, Symyx and Bayer patented the polymerization of isobutene into very high molecular
weight polymers (average molecular weight M > 100,000 g mol-1) which possess low oxygen
permeability and mechanical resilience and are therefore used in the automobile industry as
rubber products [166]. Chloroaluminate ionic liquid catalysts are advantageous because the
need of extremely low temperatures for the production of very high molecular weight
polymers with isobutene as monomer or with a variety of comonomers is not given. The
possibility of polyisobutene production was also shown by BASF in 2008 [167]. The use of
hydrocarbyl substituted pyrazolium cations in combination with heptachlorodialuminate
anions contributes to this innovation. The average molecular weight was more than four times
higher (M = 700,000 g mol-1) compared to the [EMIM]Cl / AlCl3 ionic liquid catalyst
(M = 150,000 g mol-1).
Synthetic lubricating oil with low pour point and high viscosity index is synthesized by
oligomerization of α-olefins like 1-decene or 1-dodecene. In 1997, BP showed that it is
possible to oligomerize a mixture of C6 – C12 α-olefins to high-viscosity polyalphaolefins
[168]. Chevron described a process for the production of very high-viscosity polyalphaolefins
using the acidic ionic liquid [Me3NH]Cl / AlCl 3 [169]. The reaction was carried out in the
absence of organic solvents, which have hitherto been used as a diluent for the feed that
General part 33
consisted of 1-decene or 1-dodecene. The continuous process for the production of
polyalphaolefins using the same catalysts was patented by Chevron in 2003 [170].
2.4.2.4 Acidic ionic liquid catalyzed cracking of hydrocarbons
The catalytic cracking of polyethylene has been used as a process for its recycling and
conversion into other valuable products. Acidic chloroaluminate ionic liquids
(x(AlCl3) = 0.67) cracked these polyethylenes (T = 393 – 473 K) in the presence of protons,
[EMIM][HCl 2] (1 mol. %) or sulfuric acid (2 mol. %) and yielded gaseous alkanes and cyclic
alkanes [171]. No significant amount of aromatic compounds or olefins was formed.
Cracking of alkanes (nonane, tetradecane, 2-methylpentane) (Scheme 18) was possible below
343 K with the acidic ionic liquid 1-hexylpyridinium chloride ([Hpyr]Cl) / AlCl3 [172]. The
addition of a proton scavenger like EtAlCl2 or CaH reduced the rate of cracking while the
addition of protons in the form of HCl increased the catalytic cracking activity.
Scheme 18: Alkane cracking.
2.5 Catalyst and immobilization concepts with ionic liquids
One problematic issue in homogeneous catalysis is the challenging product and catalyst
recovery. Thus, catalyst immobilization concepts are indispensable for economically and
ecologically effective, industrially applied homogeneous catalysis. Different immobilization
concepts for ionic liquids, namely multiphase reaction systems and SILP catalysts, are
described in the first and second part of this chapter. In a SILP system, an ionic liquid solvent
with a metal catalyst, or the catalytically active ionic liquid alone, is immobilized on a porous
support. The third part of this chapter deals with SCILL systems. The purpose of SCILL
systems is not only the immobilization of ionic liquid, but rather the creation of a new catalyst
concept: modification of the selectivity and / or activity by covering a heterogeneous catalyst
with an ionic liquid layer. Finally, nanoparticles, the “frontier between homogeneous and
heterogeneous catalysis” [173] can be stabilized and immobilized in ionic liquids. This special
multiphase reaction concept is described in chapter 2.5.4.
34 General part
2.5.1 Multiphase reaction systems
Mulitphase reaction systems consist of an immobilizing phase, which is catalytically active
itself or contains a catalyst and ligand, as well as a second phase, preferably solely the
product.
Generally, a suitable catalyst phase must exhibit the following properties:
� sufficient solubility for the reactants
� limited solubility for the product
� rapid mass transfer of the feedstock in the catalyst phase
� excellent solubility for the catalyst to ensure full catalyst immobilization and no
leaching
� no deactiviation of the immobilized catalyst by the catalyst solvent.
Problems of this catalyst concept may arise from mass transfer limitations between both liquid
phases. High stirring rates or external mixing devices can reduce this mass transfer influence.
Realistic possibilities using mulitphase operation of homogeneous catalysts are processes with
organic / organic, organic / aqueous or “fluorous” solvent pairs, supercritical fluids, systems
with soluble polymers and ionic liquids [174]. Prominent examples of successfully
industrially applied multiphase reaction systems are the Shell Higher Olefine Process (SHOP)
[174] and the Ruhrchemie / Rhône-Poulenc´s oxo process [175, 176].
Acidic ionic liquid catalyzed alkane isomerization is highly suited for a multiphase reaction
system in terms of simple product separation, catalyst recycling and low ionic liquid catalyst
leaching into the product phase as the unpolar alkanes exhibit a miscibility gap with
chloroaluminate ionic liquids. However, a sufficiently large phase boundary between alkanes
and ionic liquid catalyst phase has to be formed which is limited to a certain extent. Hence,
the immobilization of the ionic liquid catalyst via the SILP concept provides an improvement
and step further toward optimized immobilized homogeneous catalysts.
2.5.2 Supported ionic liquid phase (SILP) catalyst
A SILP catalyst comprises a thin film of ionic liquid, which is catalytically active itself or
dissolves a homogeneous catalyst, immobilized on a highly porous support material.
Outstanding advantages of the SILP system compared to a biphasic system are the very thin
General part 35
film of ionic liquid and the high specific surface area. Thus, mass transfer limitation or even
influence is excluded. The lack of mass transfer limitation leads to an extremely efficient use
of the ionic liquid catalyst volume and, therefore, to significant reduction of ionic liquid and
catalyst amount. Besides, the SILP concept combines both, the advantages of homogeneous
catalysis, high activity and selectivity, as well as of heterogeneous catalysis, simple product
separation, catalyst recycling and handling. Typically, the ionic liquid is held on the porous
support by physisorption. Capillary forces and electrostatic interactions (e.g. between anion
and cation of the ionic liquid and surface –OH groups of the support) immobilize the ionic
liquid on the porous support. SILP catalysts are highly suited for continuous gas-phase
processes in a fixed-bed reactor as only extremely slow evaporation and / or thermal
decomposition takes place. Slurry-phase reaction mode is only applicable if the cross
solubility of ionic liquid and reactant / product / solvent is very low like in the case of acidic
chloroaluminate ionic liquid catalyzed alkane isomerization.
The SILP system was introduced by Mehnert for the hydroformylation and hydrogenation
reactions [177-179]. During the last years, the promising SILP catalyst concept has attracted
significant research interest. SILP catalysts were successfully applied e.g. in the extensively
studied hydroformylation [180-185], hydrogenation [186, 187], enantioselective
hydrogenation [188, 189], water gas shift reaction [190, 191], alkene metathesis [192, 193],
hydroamination [194] and carbonylation of methanol [195]. Additional information to SILP
catalysts is given by reviews from Riisager [196, 197], Gu [198] or de Vos [199], amongst
others.
The ionic liquid film distribution in SILP catalysts was studied with solid-state NMR [200].
Thereby, no homogeneous ionic liquid film but island formation was proposed for ionic liquid
loadings < 10 vol. % (referred to the total pore volume of the porous support). Lemus
characterized SILP materials that were prepared with different supports [201]. Pore size
distributions with varying amounts of [OMIM][PF6] on activated carbon cloth were measured
and analyzed with a combination of adsorption-desorption isotherms of N2 at 77 K and
mercury porosimetry. Lemus suggested hierarchically pore filling, firstly the micropores,
followed by the mesopores and finally the macropores. Lercher wrote about complete
coverage of the porous silica surface with ionic liquid analyzed with IR absorption
spectroscopy [202]. Jess investigated [BMIM] octylsulfate ([OctSO4]) loadings on porous
solid catalyst Ni / SiO2 [203]. N2-adsorption measurements indicated that the surface area
decreases with increasing ionic liquid pore filling degree. The micro- and mesopores were
36 General part
partially or completely filled for a pore filling degree of 20 %. They concluded and calculated
that a monolayer with a thickness of 0.5 nm was formed at an ionic liquid pore filling degree
of around 10 %. The micro- and mesopores were successively blocked with increasing ionic
liquid loading.
In 2000, the first example of immobilization of acidic chloroaluminate ionic liquids on porous
supports for the alkylation of aromatics (benzene, toluene, naphthalene and phenol) with
dodecene has been reported by Hölderich [204]. The acidic ionic liquid [BMIM]Cl / AlCl3
was added to differently dried and calcined supports (SiO2, Al2O3, TiO2 or ZrO2) and the
excess of ionic liquid was extracted with DCM in a Soxhlet extraction system. Formation of
HCl was observed during immobilization. Only the silica-based supports were active in the
alkylation reaction. In 2002, the group of Hölderich published different immobilization
methods of ionic liquids on porous supports [205]. Thereby, acid catalysts are synthesized
which contain ionic complexes where either the anion or the cation is bonded covalently to
silica.
Joni developed a well-defined and optimized pretreatment of the silica support material prior
to impregnation with acidic chloroaluminate ionic liquid [206]. A solution of
n([EMIM]Cl) / n(AlCl 3) = 1/2 and DCM was contacted with the calcined silica. After this
chemical pretreatment step, surface –OH groups of the silica support were completely
removed as the anion was covalently bound to the basic surface sites of the silica support. The
pretreated supports themselves showed no catalytic activity in the diisopropylbenzene
isomerization. In fact, it is a suitable support for the immobilization of a defined amount of
acidic chloroaluminate ionic liquid to get a highly acidic SILP catalyst. Pictures of this acidic
SILP catalyst and the Lewis acidic ionic liquid n([EMIM]Cl) / n(AlCl 3) = 1/2 are shown in
Figure 3.
General part 37
Figure 3: Lewis acidic ionic liquid n([EMIM]Cl) / n (AlCl 3) = 1/2 (left), acidic SILP catalyst (n([EMIM]Cl) / n(AlCl 3) = 1/2 immobilized on pretreated silica according to Joni [206]) (right).
In contrast to the work reported by Hölderich, the ionic liquid film remains free flowing on
the support´s surface while being fixed to the support by capillary forces and physisorption.
These SILP catalysts were successfully tested in the slurry-phase Friedel-Crafts alkylation of
cumene and in the continuous gas-phase isopropylation of cumene and toluene. The catalysts
operated with constant high selectivity and unchanged catalytic activity for 210 h time-on-
stream [206, 207].
2.5.3 Solid catalyst with ionic liquid layer (SCILL)
The concept of SCILL systems, i.e. the coating of the internal surface of a heterogeneous
catalyst with a thin film of ionic liquid, whereupon the catalyst´s activity and / or selectivity is
modified, is explained in this chapter. A schematic picture of the SCILL concept is shown in
Figure 4. The activity and selectivity of the solid catalyst can be changed by the ionic liquid
coating in different ways:
� The ionic liquid can change the effective concentrations of reactant(s) and
intermediate(s), compared to the uncoated solid catalyst, provided that the solubility of
liquid or gaseous reactants in ionic liquids differ from those in the fluid phase [208].
� The ionic liquid can interact with the catalyst particles and modify their adsorption and
reaction properties [209, 210].
� The ionic liquid can inertize the support by covering active surface sites and suppress
unwanted side reactions [211].
38 General part
Figure 4: SCILL system – a schematic overview from the SCILL-catalyst powder up to the catalytic active sites.
A comparison of SILP and SCILL catalysts is drawn in Table 6 in order to summarize
important features of the respective catalyst concepts and to highlight differences and
similarities.
Jess introduced the SCILL concept using a model sequential hydrogenation reaction, the
hydrogenation of cyclooctadiene to cyclooctene and cyclooctane [208, 212]. A commercial
Ni / SiO2 catalyst was coated with the ionic liquid [BMIM][OctSO4]. The effect of ionic
liquid coating on the selectivity for the intermediate cyclooctene was very pronounced in a
positive way. One explanation for the high yield (70 %) of the intermediate cyclooctene by
applying the SCILL catalyst compared to the uncoated Ni / SiO2 catalyst (40 %) is the lower
solubility of the intermediate product which decreases the reaction rate for the consecutive
hydrogenation. The influence of ionic liquid layer on effective concentrations of
cyclooctadiene and cyclooctene alone could not explain the total change in selectivity. In fact,
the ionic liquid acted as cocatalyst and inhibited the (re)adsorption of cyclooctene on Ni sites.
However, the ionic liquid coating led also to a decrease of specific surface area of the porous
catalyst and to lower concentrations of cyclooctadiene in the ionic liquid than in the organic
phase. This resulted in an activity decline. Although, the SCILL catalyst was used in liquid
phase catalysis, no leaching of the ionic liquid into the organic phase was detectable. The Jess
group extended their SCILL research to further selective liquid and gas-phase hydrogenations
General part 39
[203]. For those systems in which the feedstock is less soluble in the ionic liquid compared to
the organic phase, the reaction rate was lowered (cyclooctadiene, octine, naphthaline). For a
higher solubility, this was reverse, like for cinnamaldehyde. The same correlation is valid for
the selectivity of intermediates. For intermediates with a lower solubility in the ionic liquid
than the feed (octene compared to octine, hydrocinnamaldehyde to cinnamaldehyde and
tetraline compared to naphthaline) the maximum yield of the intermediate increased compared
to the uncoated catalyst. The yield for the intermediate cinnamyl alcohol, which has a higher
solubility than the feedstock cinnamaldehyde, decreased when the SCILL catalyst was
applied.
Table 6: Comparison of SILP and SCILL catalysts.
SILP SCILL
Catalyst nature (Homogeneous metal catalyst dissolved in) ionic liquid immobilized on porous support
Heterogeneous catalyst coated with ionic liquid
Purpose Immobilization of homogeneous catalyst
Enhancement of activity and / or improvement of selectivity
Reactant/product solubility
Optimal: high for reactants, low for products
Optimal: high for reactants, low for products
Ionic liquid loading Low for high specific surface area but enough for homogeneous dissolution of metal (complex)
Low for high specific surface area but enough to cover active centers
Ionic liquid – support interaction
Good wettability required Good wettability required
Metal and metal complex solubility
As high as possible As low as possible
Ionic liquid – metal interaction
Desired to stabilize catalyst in solution
Desired to enhance activity and / or selectivity
Maximum reaction temperature
Limited by stability of metal (complexes) and ligand
Limited by ionic liquid stability
The group of Claus investigated SCILL catalysts in the regioselective hydrogenation of citral
[211, 213]. Supported Pd catalysts were coated with [N(CN)2]-, [NTf2]
-, and [PF6]--based
40 General part
ionic liquids. Solid Pd catalyst´s modification by ionic liquid coating led to very high
selectivities to citronellal because consecutive and side reactions were inhibited. The best
results were achieved with [N(CN)2]--based ionic liquids (90 % selectivity to citronellal at
80 % conversion). Except for [BMIM][PF6], the citral conversions were in a similar range
compared to the ionic liquid free catalyst. The lower hydrogen solubility in ionic liquids
compared to organic solvents [214] did not influence the conversion indicating no mass
transfer limitation.
Further studies were conducted to investigate the interaction of ionic liquid and active Pd
which is basis for the understanding of promoting effects in the citral hydrogenation [209].
The coating of Pd / SiO2 with ionic liquids lowers the hydrogen uptake and heats of
adsorption remarkably. The reduction of H2 uptake was much higher for [N(CN)2]--based
ionic liquids compared to [NTf2]--based ionic liquids. X-ray photoelectron spectroscopy
(XPS) measurements revealed that ionic liquids deposited on Pd / SiO2 modify the chemical
state of surface Pd which is partially transformed to Pd2+ and undergoes complexation with
the [N(CN)2]-. The ionic liquid anion acts as ligand to the metal. This finding was also
underlined by IR spectroscopy. Beside lower hydrogen solubility in ionic liquids [214], the
coordination between ionic liquid and Pd may also contribute to a lower accessibility of
hydrogen toward the Pd surface sites. X-ray absorption spectroscopy (XAS) characterizations
excluded an influence of the ionic liquid on the metal dispersion which may affect hydrogen
uptake and adsorption strength. Selective citral hydrogenation with SCILL catalysts based on
Pd and the ionic liquid [BMIM][N(CN)2] was also applied continuously in a trickle-bed
reactor [215].
The SCILL catalyst system Ru / Al2O3 / [cation][NTf2] also influenced the conversion and
selectivity of the citral hydrogenation [216]. The conversions decreased for all SCILL
catalysts in comparison to the Ru / Al2O3 system. Lower hydrogen solubility in ionic liquids
compared to organic solvents accounts reasonably for the decreasing conversion. [NTf2]--
based ionic liquids inhibit primarly the C=C hydrogenation which leads to different product
selectivities.
Silica supported Pt catalysts with a thin film of 1-butyl-2,3-dimethylimidazolium
trifluoromethanesulfonate ([BMMIM][OTf]) were investigated by the Lercher group with
focus on the interactions between ionic liquid, support and Pt particles [202]. Interaction of
silica and ionic liquid occurs via hydrogen bonds. Pt clusters modify the electron density of
the ionic liquid which changes the polarity of the ionic liquid within certain levels. The ionic
General part 41
liquid protects the Pt cluster from oxidation as the Pt particles in ionic liquid coated systems
are in zero oxidation state. SCILL catalyzed ethene hydrogenation showed that the interaction
of ionic liquid and Pt surface is weaker compared to the interaction of reactant molecules and
Pt surface as the ionic liquid coated and ionic liquid free Pt / SiO2 catalysts had similar
catalytic activities in the experiments.
The groups of Steinrück, Libuda and Wasserscheid studied the interaction of the ionic liquid
[BMIM][NTf 2] with Pt and Pd nanoparticles supported on thin and ordered alumina films in
UHV conditions [210, 217]. The interaction of [BMIM][NTf 2] with the nanoparticles is strong
enough to partially replace strongly chemisorbed adsorbates like CO. The ionic liquid showed
a ligand effect that is specific for the material and selective to particular sites.
2.5.4 Stabilization and immobilization of nanoparticles in ionic liquids
Ionic liquids are also valuable media for catalysis with metal nanoparticles. In most cases, the
reactions are multiphase systems. The nanoparticles dispersed in the ionic liquid form the
dense phase and the substrate and product remain in the upper phase. Hence, the ionic liquid
phase is easily recovered by simple decantation. Metal nanoparticles with 1 – 10 nm in
diameter have unique properties that result from their large surface-to-volume ratio and
quantum size effects [218]. Nanoparticles are only kinetically stable because the formation of
bulk metal is thermodynamically favored. Thus, nanoparticles must be stabilized in order to
prevent agglomeration by electrostatic and / or steric protection. Ionic liquids have been
shown to be a promising medium for the synthesis and stabilization of nanoparticles [219,
220]. Transition-metal nanoparticles are probably stabilized by protective layers of
supramolecular structures ([(DAI)x(X)x-n)]n+ [(DAI) x-n(X)x)]
n-)n (in which DAI is the
1,3-dialkylimidazolium cation and X the anion) through the loosely bound anionic moieties
and / or N-heterocyclic carbenes together with an oxide layer if present on the metal surface.
However, the surface-bound protective species are easily displaced by other substances in the
media. This fact explains on the one hand to some extend their catalytic activity, on the other
hand their relatively low stability against aggregation / agglomeration. More stable catalytic
systems in ionic liquids can be obtained by the addition of ligand or polymeric stabilizer.
Transition-metal nanoparticles dispersed in ionic liquids are active catalysts for various
reactions such as the hydrogenation of alkenes, arenes and ketones [218]. Further examples of
ionic liquid-based catalysis with metal nanoparticles are described in the review by Gu [221]
and Dupont [222].
42 General part
First examples can be found in literature that describes the immobilization of an ionic liquid
in which metal nanoparticles are immobilized on solid supports. This is in close analogy to
the SILP concept where homogeneously dissolved metal complexes are immobilized instead
of metal nanoparticles. Hagiwara et al. describe immobilized the precursor Pd(OAc)2 non-
covalently as a supported ionic liquid catalyst in a nanosilica dendrimer with the aid of ionic
liquid and tested this catalysts in the Suzuki-Miyaura reaction [223]. High loaded
imidazolium salts framework on the surface of silica and inside its pores acted as support for
Pd immobilization. The multi-layered, covalently supported ionic liquid phase was the
support for Pd nanoparticles and tested by Gruttadauria et al. in the Suzuki reaction in
aqueous medium [224].
2.6 Catalyst systems for the alkane isomerization
Active and selective C5 / C6 alkane isomerization catalysts have been successfully applied in
industry. It is necessary and desirable to extend the alkane isomerization to heavier alkanes
(larger than C6) provided that high isomerization yields can be achieved. The isomerization of
C7+ alkanes is very different from the isomerization of light paraffins, like n-C4 – n-C6,
because cracking side reactions become progressively competitive as the carbon chain length
increases. Besides, it would be favorable to decrease the reaction temperature of the alkane
isomerization from a thermodynamic point of view.
Alkane isomerization has been investigated by numerous groups over the last decades. Firstly,
heterogeneously catalyzed hydroisomerization / hydrocracking reactions are presented in this
chapter. The selected examples are grouped according to different catalysts types applied
which are all based on noble metals. Chosen literature deals with the model feedstock
n-heptane or n-octane to allow direct comparison with the ionic liquid based catalysts for the
n-octane isomerization of this work. The second part of this chapter comprises the literature
review of ionic liquid catalyzed alkane isomerization / cracking. Up to now, only a few
groups have investigated the isomerization / cracking with ionic liquid catalysts.
2.6.1 Heterogeneous catalyzed alkane isomerization
Weitkamp et al. elucidated in model compound studies the possibility to isomerize the C7 cut
obtained from atmospheric distillation of crude oil as an alternative to its aromatization [225].
Thereby, the zeolites 0.27 wt. % Pd / LaNaY-72, 0.27 wt. % Pd / HMCM-41 and 0.27 wt. %
Pd / HBeta were used for the hydroisomerization of n-heptane (phydrogen = 10 bar,
General part 43
pn-heptane = 0.1 bar, W / Fn-heptane = 400 g h mol-1 with W := dry mass of catalyst). Isomerization
of n-heptane was found to be possible up to medium conversions without considerable
hydrocracking. However, the RON of the resulting isomerization products was much too low
(37 - 43) for potential application as gasoline.
Hydroisomerization of n-octane was kinetically studied by Corma et al. using different
zeolitic structures with large pore sizes (12 MR) [4]. 0.5 % Pt was incorporated by incipient
wetness impregnation with an aqueous solution of H2PtCl6. It is evident that at temperatures at
which zeolites show catalytic activity, the cracking of mono- and specially dibranched
isomers is faster than the isomerization of mono- to dibranched products. Within the
experimental conditions of this study (Treaction = 493 – 723 K, ptotal = 20 bar,
hydrogen / n-octane = 10, WHSV = 5.13 h-1 referred to n-octane), the thermodynamic
equilibrium of monobranched products was obtained but this was not the case for dibranched
products due to fast cracking. The rate of cracking of dibranched products equaled almost the
isomerization of mono- to dibranched products. Consequently, high levels of di- or
tribranched C8 alkanes can only be reached by introducing a distillation and recycling unit or
by using membrane reactors.
The group of Ramos and Valverde studied the effect of metal loading in the
hydroisomerization of n-octane over beta agglomerated zeolites (B) containing Pt and / or Pd
and binder (Bent) [226]. Experiments were conducted in a tubular micro-reactor with the
following reaction conditions: Treaction = 523 - 683 K, ptotal = 10 bar,
WHSV = 10 g n-octane h-1 g-1zeolite, molar ratio hydrogen / n-octane = 14. It was observed that
the octane isomer yield rose with increasing metal / acid balance up to a certain ratio.
n-Octane conversion increased with a higher amount of Pt or Pd metal content for the same
reaction temperature and was lower for bimetallic catalysts in comparison to the
monometallic loaded zeolites. Highest achieved octane isomer selectivities at 603 K and 70 %
n-octane conversion were 82.6 % (58.2 % monobranched, 24.4 % multibranched isomers)
with 0.75 wt. % Pd / BBent catalyst, 94.8 % (70.5 % monobranched, 24.3 % multibranched
isomers) with 0.75 wt. % Pt / BBent catalyst and 91.9 % (62.5 % monobranched, 29.4 %
multibranched isomers) with 0.25 wt. % Pd / 0.75 wt. % Pt / BBent catalyst.
Liquid phase n-octane hydroisomerization in a stirred semi-batch microautoclave was
investigated over beta, USY and mordenite zeolites loaded with 1 wt. % Pt [227]. Best results
regarding multibranched isomer selectivity (Smultibranched = 36.6 %, Smonobranched = 55.7 %) were
44 General part
achieved with PtMOR (Treaction = 543 K, p= 90 bar, conversion = 77.4 %, reaction
time = 10.5 h, 7 g zeolite / mol n-octane).
The hydroisomerization of n-octane with the hierarchical (meso- and microporous) zeolite
Pt / ZSM-22 yielded 70 % monobranched isomers, 5 % dibranched and circa 5 % cracking
products at 80 % conversion (Treaction = 573 K, p = 50 bar, phydrogen = 5 bar,
W / F = 14 - 90-kg s mol-1) [228]. The hierarchical ZSM-22 zeolite exhibited an enhanced
sorption capacity for n-octane in contrast to it’s purely microporous equivalent. With insertion
of mesoporosity in addition to microporosity, higher activity and heptane isomer yields were
achieved with the hierarchical ZSM-22 catalyst compared to the microporous zeolite. This
was also attributed to an increased number of pore mouths and the associated decreased
diffusion length for reactants and products.
Bifunctional Pt / WOx-ZrO2 (Pt / WZr, 12.7 wt. % W) and Pt / Beta (Si / Al = 12) catalysts
were studied by Arribas and coworkers for the simultaneous n-alkane hydroisomerization and
aromatic hydrogenation using a n-heptane / benzene feed mixture (25 wt. % benzene) at
30 bar, WHSV = 3.1 h-1 and molar ratio of hydrogen / hydrocarbon = 10 [5]. The
Pt / WOx-ZrO2 (0.6 % Pt) yielded 51.7 % iso-heptane at 493 K. The obtained heptane isomers
contained 30.8 % of di- and tri-branched and 69.2 % of monobranched products. The
experiments with Pt / Beta (1 % Pt) resulted in 49.9 % iso-heptane yield at 533 K with
selectivities of 27.5 % for the multi-branched heptanes and 72.5 % for the monobranched
heptanes.
Multicomponent nanocomposite catalysts of the type Pt / WO3 / M-ZrOx were also tested in
the n-heptane hydroisomerization (398 – 473 K) [229]. 30 % n-heptane was converted with an
isomerization selectivity of greater than 99 % using a Pt / WO3 / Al-ZrOx catalyst (p = 7 bar,
hydrogen / n-heptane = 1, volume hour space velocity = 1 h-1) at 398 K under trickle bed
conditions.
Sulfated zirconia was tested in the n-octane hydroisomerization by Busto and coworkers
[230]. This study aimed for a maximum liquid product yield (C5 – C8) which was obtained at
18.5 % conversion for the Pt / SZ catalyst (Treaction = 573 K, p = 15 bar, WHSV = 4 h-1, molar
ratio hydrogen / n-octane = 6, time-on-stream = 30 min). Selectivities to mono- and
multibranched alkane isomers within the liquid products were not reported in this publication.
General part 45
Pt / WOx-ZrO2 catalysts were optimized for the isomerization / cracking of the model
compound n-octane (Treaction = 573 K, 10 bar, WHSV = 4, heptane / n-octane = 6) [231]. The
results revealed a strong influence of the calcination temperature on the final metal / acid
balance of the catalysts. The highest liquid and isoparaffin yields were obtained with samples
calcined at 973 K. Products included C1 – C8 alkanes and aromatics. All catalysts showed a
RON gain of 55 when the isomerized product is compared to the feedstock n-octane.
Silica supported tungsten zirconia catalysts were tested in the simultaneous
hydroisomerization and hydrocracking of n-octane (Treaction = 573 K, p = 1 bar,
WHSV = 1 h-1, molar ratio hydrogen / n-octane = 6 ) for the production of C4 – C8 products
with high octane numbers [90]. The catalysts applied in this study deactivated rapidly in the
absence of Pt which is required to hydrogenate coke precursors. Pt(0.5 %)W7.5Z1.0Si
performed best in n-octane isomerization with the following selectivities within the
iso-octanes: Smonobranched = 69.5 %, Sdibranched = 23.3 %, Stribranched = 7.2 %.
In a quite different approach, Ohno et al. studied the catalytic properties of H2-reduced
HxMoO3 in alkane isomerization in comparison to those of H2-reduced MoO3 and 0.01 wt. %
Pt / MoO3 [6]. The latter displayed the highest activity in n-heptane hydroisomerization
(96.7 % isomerization selectivity at 57.2 % conversion; 76.7 % monobranched, 19.1 %
dibranched and 4.2 % tribranched heptanes) at 423 K (atmospheric pressure, molar ratio
hydrogen / n-heptane = 10, flow rate of C7 = 0.02 mol h-1).
n-Octane hydroisomerization was also carried out using the metal-acid bifunctional catalyst
MoO2-x(OH)y [232]. This system achieved 73.5 % isomerization selectivity at 86.3 %
conversion, with 50.8 % monobranched and 22.7 % multibranched iso-octanes
(Treaction = 623 K, phydrogen = 5 bar, LHSV = 0.8 h-1, hydrogen / n-octane = 30.2). The product
mixture included alkanes ranging from C1 – C8 and small amounts of aromatics.
Another catalyst system for alkane isomerization to mention is the heteropolyacid H3PW12O40
supported on ZrO2, SiO2 or carbon [233]. These catalyst systems were investigated in the
hydrocracking and hydroisomerization of n-octane to gain fuel-grade gasoline (C4 – C7)
(Treaction = 573 K, p = 10 bar, WHSV = 1 h-1, hydrogen / n-octane = 6). The systems were
positively influenced by the addition of Pt with regard to their activity and stability.
A detailed review of the selective isomerization of high n-alkanes (C7+) was published which
covers further examples of all these different isomerization catalyst types [234].
46 General part
2.6.2 Acidic ionic liquid catalysts for alkane isomerization
The interest in running alkane isomerization reaction at very low temperatures has triggered
attempts in the past to apply highly acidic chloroaluminate ionic liquids as catalysts for this
reaction. Because of the limited number of publications in this field, the literature review was
not restricted to heptane or octane feedstock.
Superacidic ionic liquids of the type [Me3NH]Cl / AlCl 3 / HCl were tested by Berenblyum et
al. [8]. They studied saturated C7-mixtures that included already branched C7 alkanes at
298 K and 323 K in a batch autoclave. Higher HCl pressure and AlCl3 mole fraction of the
ionic liquid resulted in an increase of the multi-branched heptane concentration. Another
focus of their work was the investigation of the ionic liquid deactivation during reaction.
Multi-branched heptane formation decreased linearly if recycling runs were carried out
without HCl saturation before the next run. The HCl saturation of the feed as well as
simultaneous HCl saturation of feed and ionic liquid led to an increase in catalytic activity.
Moreover, the authors proposed the formation of so-called acid soluble oil in the course of
isomerization which interacted with the acidic ionic liquid and accounted also for catalyst
deactivation. No results of isomerization and cracking selectivities were displayed in this
paper.
Zhang and coworkers investigated the isomerization of n-pentane in the system
triethylammonium chloride ([Et3NH]Cl) / AlCl3 [9]. Here, some superacidity originated
obviously from traces of water in the system. The experiments were conducted in a batch
autoclave at p = 7 bar, Treaction = 303 K, 318 K or 333 K, t = 0.5 – 8 h, varying mole fractions
of AlCl3 in the ionic liquid and different catalyst to oil volume ratios. The conversion of
n-pentane increased with the enhancement of reaction temperature, reaction time and catalyst
to oil ratio, while the yield of isomerization oil and the selectivity for iso-paraffins decreased.
44.6 % n-pentane was converted with a selectivity for iso-parraffins of 90.5 % if the
determined optimal reaction conditions were applied (Treaction = 303 K, t = 3 h, volume ratio
catalyst / oil = 1/1).
A similar system using different acidic chloroaluminate ionic liquids with a nitrogen-
containing cation and no added Brønsted acid has been claimed earlier by Haldor Topsøe in a
patent for alkane isomerization under very mild conditions [235]. More patents by Haldor
Topsøe followed which based all on acidic chloroaluminate ionic liquids with a N-containing
cation and are described in the following.
General part 47
The hydrocarbon isomerization with composite catalysts comprising an ionic liquid, e.g.
[Me3NH]Cl / AlCl 3, and a metal salt additive, for example MoCl5, FeCl3 or CuCl2, was
patented [236]. The feed consisted of saturated C7 hydrocarbons, linear and branched alkanes,
which were converted at mild reaction conditions.
Furthermore, the influence of a cyclic hydrocarbon additive, like methylcyclohexane, on
hydrocarbon isomerization catalyzed by [Me3NH]Cl / AlCl 3 was protected in a patent [237].
A saturated C7 hydrocarbon feed, which included already branched alkanes, was converted
under mild reaction conditions. Higher selectivities for multi-branched C7 products and yields
of multi-branched products were achieved as a consequence of the added cyclic hydrocarbon
additive.
Moreover, systems of the general type [cation]Cl / AlCl3 / H2SO4 have been also disclosed
[238]. The systems have been claimed to isomerize a distillation cut of paraffinic
hydrocarbons, including branched starting material, at temperatures as mild as 298 K. The
patent gives no information on the role of the added sulfuric acid, on kinetics and on the
obtained product distribution. With respect to the latter point, the ratio of feedstock
isomerization versus feedstock cracking is of utmost importance for the practical relevance of
the catalyst system and its evaluation against state of the art heterogeneous isomerization
catalysts.
In 2011, BASF included in their patent the isomerization of linear, branched and / or cyclic
hydrocarbons in the presence of a chloroaluminate ionic liquid of the general type [organic
cation][Al2Cl7] and an olefin [239]. The combination of highly acidic ionic liquid and olefin,
whose amount is < 5 wt.-% referred to the saturated hydrocarbon feedstock, accelerated the
isomerization.
All these examples of acidic ionic liquid catalyzed alkane isomerization show the potential of
low temperature alkane conversion. Thus, examples and results of the literature review
presented in chapter 2.6.2 were the starting point of this work.
2.7 Objective of this work
Basic goal of this thesis was the investigation and optimization of highly acidic ionic liquid
based catalysts for n-alkane isomerization at low and moderate reaction temperatures. Highly
active and selective ionic liquid based catalysts should be developed for the unbranched
model compounds n-hexane and n-octane. The focus was on the long-chain alkane n-octane,
48 General part
which exhibits more reaction pathways of side reactions compared to the reactant n-hexane.
Due to the pronounced cracking ability, alkane isomerization feedstock with C > 6 has not
been industrially applied up to now. However, even the utilization of the longer-chain alkanes
and their contribution to high-octane gasoline production will be necessary. Recent
investigations conerning the isomerization of long-chain alkanes (C > 6) include mostly only
heterogeneous reaction systems which are active at higher temperature. From a
thermodynamic point of view, lower reaction temperatures are favorable for the formation of
highly branched isomers. Highly acidic ionic liquids offer the opportunitiy to convert alkanes
at very mild reaction conditions. Therefore, it is interesting, promising but also challenging to
investigate their ability to act as alkane isomerization catalyst.
The investigated catalytic systems can be basically divided into monofunctional and
bifunctional ionic liquid based catalysts. Main focus was on the optimization of the catalysts´s
activity and selectivity. Moreover, elucidation of the different mechanisms using either
monofunctional or bifunctional catalysts was part of this work.
49
___________________________________________________________________________
3. Experimental
50 Experimental
Experimental 3
General working techniques, used chemicals, catalyst preparations, reactor set-up and the
experimental procedure of the isomerization experiments are described in detail in this
chapter.
3.1 General working techniques
All experiments, moisture sensitive chemicals and catalysts were operated and handled under
inert conditions. Moisture sensitive chemicals were stored in a Plexiglas® Glovebox
(GS GLOVEBOX Systemtechnik GmbH) using argon as inert gas. Catalyst preparation and
filling of the glass liner of the applied batch reactor was also carried out in the glovebox. In
addition, Schlenk technique was applied for SILP and SCILL preparation. If necessary,
chemicals were dried overnight under high vacuum (p < 0.1 mbar). Glass liner and reactor pot
were stored in the oven at T = 343 K overnight prior to use to avoid moisture contamination.
3.2 Chemicals
The chemicals used for preparation of the catalysts are listed in Table 7. Platinum on silica
(Pt / silica) was dried at room temperature under high vacuum (p < 0.1 mbar) overnight and
stored under argon. Silica 100 was calcined under N2 prior to use according to the following
temperature program and stored under argon: 4 K min-1 � 423 K, 2 h isothermal at 423 K,
4 K min-1 � 873 K, 12 h isothermal at 873 K.
Table 8 includes all chemicals applied in the experiments as reactant, cocatalyst, promoter or
inert gas.
Table 9 shows all chemicals that were necessary for analysis of the products, catalysts and
organic phases.
Experimental 51
Table 7: Chemicals for catalyst preparation.
Chemical Short Purity Origin
1-Butyl-3-methylimidazolium tetrachloroaluminate
[BMIM][AlCl 4] 95 % Sigma Aldrich
Aluminum chloride, anhydrous, sublimed AlCl3 98 % Merck
1-Butyl-3-methylimidazolium bromide [BMIM]Br 97 % Fluka
Aluminum bromide, anhydrous AlBr3 98 % Merck
Platinum (1 wt. %) on silica Pt / silica Sigma Aldrich
Silica 100, 0.2 – 0.5 mm Silica Merck
Platinum(II) chloride PtCl2 98 % Sigma Aldrich
Platinum(IV) chloride PtCl4 99.9 % Alfa Aesar
Dichloromethane DCM 99.9 % Merck
Argon Ar 4.6 Linde
Nitrogen N2 5.0 Linde
Table 8: Reactants, cocatalysts, promoters and inert gases.
Chemical Short Purity Origin
n-Hexane n-C6 98.9 % AnalaR NORMAPUR
n-Octane n-C8 99 % Merck
Hydrogen H2 5.0 Linde
Sulfuric acid H2SO4 96 % Merck
Copper(II) chloride CuCl2 98 % Merck
1-Chlorooctane 1-Cl-C8 98 % Merck
Argon Ar 4.6 Linde
Helium He 4.6 Linde
52 Experimental
Table 9: Chemicals for analysis.
Chemical Short Purity Origin
Cyclohexane (water < 20 ppm, solvent purification system)
- 99.9 % BASF
Dichloromethane (water < 20 ppm, solvent purification system)
DCM 99.9 % Akzo Nobel
Propane C3 2.5 Linde
Propene C3 2.5 Linde
iso-Butane iso-C4 2.5 Linde
n-Butane n-C4 2.5 Linde
Isoparaffins Mix, analytical standard (Alphagaz PIANO Calibration Standard)
- Sigma Aldrich
Olefins Mix, analytical standard (Alphagaz PIANO Calibration Standard)
- Sigma Aldrich
Platinum plasma standard solution, (1000 µg ml-1)
- VWR Prolabo
Aluminum plasma standard solution, Specpure ® (1000 µg ml-1)
- Alfa Aesar
Silicon plasma standard solution, Specpure ® (1000 µg ml-1)
- Alfa Aesar
Copper plasma standard solution, Specpure ® (1000 µg ml-1)
- Alfa Aesar
Argon Ar 4.6 Rießner Gase
Hydrogen H2 3.0 Rießner Gase
3.3 Catalyst preparation
Preparation of Lewis acidic ionic liquids
The Lewis acidic chloroaluminate ionic liquids [BMIM]Cl / AlCl 3 were prepared by mixing
[BMIM][AlCl 4] with a defined molar amount of AlCl3. The mixture was stirred until
complete dissolution of the AlCl3 and a clear liquid was obtained. This acidic chloroaluminate
Experimental 53
ionic liquid was used directly as monofunctional catalyst in liquid-liquid biphasic alkane
isomerization experiments and for the preparation of the SILP and SCILL catalysts. The
bromoaluminate ionic liquid was prepared by very slow addition of a defined molar amount
of AlBr3 to [BMIM]Br until a clear ionic liquid was obtained.
Preparation of composite ionic liquid catalysts
Lewis acidic ionic liquid was mixed with CuCl2 and stirred at 363 K for 1.5 h. The obtained
slurry-phase ionic liquid system was used as catalysts in biphasic experiments. H2SO4 or the
promoter 1-chlorooctane was injected directly with a syringe into the catalytic phase in the
reactor prior to reaction start.
For the experiments with metal salts, the Lewis acidic ionic liquid was mixed with PtCl2 or
PtCl4 and stirred for 1 h at room temperature in the glovebox.
Preparation of the SILP and SCILL catalysts
Calcined silica was used as support for the preparation of the SILP catalyst. Pt / silica was
used for the preparation of the SCILL catalyst. Pt / silica was ground, sieved (particle size
125 – 400 µm) and dried at room temperature under high vacuum (p < 0.1 mbar) overnight
prior to catalyst preparation.
As interactions between the acidic ionic liquid and the basic surface sites of the silica support
would result in a loss of acidity, the silica and Pt / silica materials, respectively, were
pretreated according to a method developed by Joni [206]. In this pretreatment step, silica or
Pt / silica, respectively, were impregnated with the acidic ionic liquid to eliminate any
hydroxyl group or any basic site on the support surface. A defined amount of acidic ionic
liquid was diluted in DCM and the silica or Pt / silica, respectively, was added. After stirring
this suspension for at least 1 h at room temperature, DCM was removed by evaporation under
vacuum. In the next step, the impregnated support was washed with DCM to remove the
excess of acidic ionic liquid that was not covalently bound to the basic surface sites on the
silica support. The pretreated silica or pretreated Pt / silica was used for the preparation of the
SILP and SCILL catalyst, respectively, by impregnating it with a defined amount of acidic
ionic liquid. The applied amount of ionic liquid was chosen to obtain a catalyst material with
defined ionic liquid loading ε (Equation 5) and ionic liquid pore filling degree α (Equation 6).
According to Equation 5, the ionic liquid loading ε of the SILP and SCILL catalyst is defined
as the ratio of the used mass of acidic ionic liquid mionic liquid and the total mass mcat.,tot. of the
54 Experimental
SILP and SCILL catalyst, respectively. The α-value calculation is based on the volume of
Lewis acidic ionic liquid and the pore volume of the pretreated Pt / silica or silica as
calculated from the nitrogen adsorption isotherm at p / p0 nearby 1.0. The obtained catalyst
was stored inside the glovebox until usage for catalytic isomerization experiments. Figure 5
depicts the purchased Pt / silica and the SCILL catalyst before and after reaction.
ε = mionic liquid
mcat.,tot. (5)
α = V ionic liquid
Vpore (6)
Figure 5: Pt / silica as purchased (left), SCILL catalyst before reaction (Pt / silica, n([BMIM]Cl) / n(AlCl 3) = 1/2, ε = 0.5, α = 0.85) (middle), same SCILL catalyst after reaction (right).
3.4 Reaction
Reactor set-up
All experiments were carried out in a 300 ml Parr batch autoclave. The flowsheet of the
reactor is shown in Figure 6. The reactor was equipped with an internal cooling coil
connected to a cryostat and a heating jacket to ensure defined isothermal reaction conditions.
The reactor consisted mainly of Hastelloy C due to the corrosive character of halometallate
ionic liquids. Moreover, all experiments were conducted with a glass liner. Another advantage
of the glass liner is the filling of reactant and catalyst under inert conditions in the glovebox
and simple loading and unloading of the catalyst into and out of the reactor. Further
Experimental 55
specifications of the reactor are listed in Table 10. To guarantee hydrogen uptake in the
hydroisomerization experiments and sufficient mixing between the organic and ionic liquid
phases, a four blade gas entrainment stirrer was applied. A photo of the reactor, glass liner and
gas entrainment stirrer is displayed in Figure 7.
Figure 6: Flowsheet of the batch autoclave.
Filling of the reactor with hydrogen or inert gas was possible via valve V2, V3 or V4. Two
valves, V1a and V1b, were built in the sampling line. The organic product was cooled down
under reaction pressure in the pipe section between these two valves V1a and V1b before
releasing the organic. The sampling line was equipped with a frit which was applied for
experiments with SILP or SCILL catalyst. Reactor temperature, reactor pressure as well as
revolutions per minute were recorded for the whole reaction time.
56 Experimental
Table 10: Specification of the applied Parr batch-autoclave.
Component / operating conditions Specification
Reactor pot, stirrer, sampling line, cooling coil, thermocouple
Hastelloy C
Reactor periphery Stainless steel 1.4571
Sealing Viton
Bursting disk Inconel Au
Glass liner Quartz glass
Reactor pot volume 300 ml
Maximum pressure 200 bar
Temperature 263 - 573 K
Figure 7: Photo of the isomerization reactor set-up (left), applied glass liner and stirrer (right).
Experimental 57
Isomerization experiments
Pressure test with ptest = 1.3 · preaction was carried out before every catalytic run. The desired
mass of catalyst ([BMIM]X / AlX3, [BMIM]X / AlX 3 with CuCl2, PtCl2 or PtCl4, SILP or
SCILL catalyst) and reactant (n-octane or n-hexane) were filled into the glass liner within the
glovebox. The defined amount of H2SO4 or 1-chloroctane, if applied, was injected into the
catalytic phase before reaction. The filled glass liner was placed inside the reactor under argon
counter flow and the reactor was flushed for three times with argon after closing. The
experiments for the catalyst screening were run under argon (pargon = 3 bar). For all other
experiments, the reactor was charged with helium or hydrogen pressure according to the
desired reaction conditions (potal = phydrogen + phelium) prior to heating. Upon reaching the
reaction temperature, the gas entrainment stirrer was set to 1000 rpm (slurry-phase
experiments with SILP or SCILL catalyst) or 1200 rpm (liquid-liquid biphasic experiments),
respectively. This moment was defined as reaction start (treaction = 0 h). Samples of the organic
phase were taken during the experiment after defined time periodes. Thereby, the stirrer was
turned off to enable phase separation. The organic product was filled in the sampling line
under reaction pressure and valve V1b and V1a were closed. The sampling line was cooled
down to 268 K to avoid evaporation of light boiling components while sampling. The organic
product was directly released in excess solvent cyclohexane (two samples per sampling time)
or DCM (one sample per sampling time), which could be handled very well because of the
needle valve V1b. The resulting mixture was either immediately analyzed via offline gas
chromatography (GC) or stored in a fridge (T = 259 K). For the catalyst screening
experiments with the systems [BMIM]Cl / AlCl3 / H2SO4 and [BMIM]Cl / AlCl3 / CuCl2
pressure was released after reaction and the reactor was opened. After 30 min at ambient
condition, the mass of liquid organic product (LOP) was determined by weighting. LOP and
gas-phase were analyzed via GC and only the LOP was used for further calculations. Results
of the analyzed reaction mixture in the presence of an alkyl halide promoter were also
confirmed with the analyzed data obtained with an external standard (n-heptane). The
application of an internal standard was not applicable for all experiments because one
coercively necessary property of a standard - no conversion if it is contacted with the catalyst
- was not granted [240-242].
58 Experimental
3.5 Analytics
3.5.1 Gas chromatography (GC) analysis
All organic product samples were analyzed using a Shimadzu 2010 Plus GC equipped with a
CP Sil PONA CS capillary column (50 m x 0.21 mm, average thickness: 0.5 µm) and a flame
ionization detector. Conversion of n-octane (Xn-octane) and selectivity for component i (Si)
were calculated according to Equation 7, 8 and 9, with Ai is the peak area of component i.
xi = Ai
∑ Aj�j=1
(7)
Xn-octane = 1 - xn-octane
xn-octane,0 (8)
Si = xi - xi,0
xn-octane,0 - xn-octane (9)
For the catalyst screening experiments with the systems [BMIM]Cl / AlCl3 / H2SO4 and
[BMIM]Cl / AlCl 3 / CuCl2, Xn-octane, Si and the yield of LOP (YLOP) were calculated according
to Equation 10, 11 and 12.
Xn-octane = 1 - nn-octane
nn-octane,0 (10)
YLOP = nLOP · � υn-octane�nn-octane,0 | υLOP|
(11)
Si = Yi(LOP)Xn-octane
(12)
3.5.2 Gas chromatography - mass spectrometry (GC-MS)
To check the intactness of the ionic liquid layer of the SCILL catalyst after reaction, the ionic
liquid layer was removed in a washing procedure with DCM after the experiment. Circa 50 ml
distilled water and 50 ml cyclohexane were added to the removed ionic liquid, stirred for
Experimental 59
16 h, followed by 5 h without stirring for phase separation. The organic phase was analyzed
using a Varian 450-GC equipped with a Varian VF-5ms column (30 m x 0.25 mm, average
thickness: 0.25 µm) and Varian 220-MS (ion trap mass spectrometer) with electron ionization.
3.5.3 Electrospray ionization – mass spectrometry (ESI-MS)
ESI-MS measurements of the organic product mixture were performed on an esquire 6000
(Bruker Daltonics Inc.). The organic phase was dissolved in methanol prior to analysis. Scans
ranging from 50 m/z to 1500 m/z were achieved with N2 dry gas, 220 V at capillary exit and
1562 µs accumulation time.
3.5.4 Inductively coupled plasma - atom emission spectroscopy (ICP-AES)
The Pt and Al content in the organic product mixture was analyzed by ICP-AES to determine
the degree of metal and ionic liquid catalyst leaching into the organic product. The acidic
ionic liquid was analyzed for Pt to exclude dissolution of the Pt from the silica support into
the acidic ionic liquid. For this analysis, the ionic liquid layer was washed off the SCILL
catalyst (five times) after use in catalysis using DCM. The amount of Si, Pt and Al of the
Pt / silica was analyzed as purchased, after pretreatment, after ionic liquid impregnation, after
reaction and after removing of the ionic liquid with DCM. The analyses were carried out
using a SPECTRO CIROS CCD (SPECTRO Analytical Instruments GmbH). The latter was
calibrated prior to measurement using plasma standard solutions of Si, Pt and Al.
The solubility of CuCl2 in the ionic liquid n([BMIM]Cl) / n(AlCl3) = 1/2 was analyzed by
ICP-AES. CuCl2 and the ionic liquid with n(AlCl3) / n(CuCl2) = 18 were stirred for 24 h.
Afterwards, the solid and liquid phases were separated by a centrifuge (VWR Galaxy
MiniStar). The liquid phase was analyzed for the Cu content with the same device mentioned
above which was calibrated before analysis using plasma standard solution of Cu.
3.5.5 N2-adsorption
Specific surface area and pore volume of porous catalyst samples were determined by
physisorption of N2 using a QUADRASORB SI (Quantachrome). Before N2-adsorption was
carried out at 77 K, the samples (Pt / silica and Pt / silica pretreated) were heated under
vacuum (423 K, 4 h). The SCILL systems with α = 0.85 were filled into the measuring device
under inert conditions in the glovebox and heated prior to N2-adsorption at 353 K for 12 h.
The specific surface area was calculated according to the Brunauer-Emmett-Teller (BET) or
60 Experimental
Barrett-Joyner-Halenda (BJH) method and the pore volume with the help of the isotherm at
p p0-1
nearby 1.
3.5.6 Nuclear magnetic resonance spectroscopy (NMR)
NMR was used to analyze the intactness of the cation of the ionic liquid after reaction. The
ionic liquid layer was washed off the SCILL catalyst (5 times) after use in catalysis using
DCM that was separated afterwards under UHV conditions for 18 h. Addition of CDCl3
resulted in the formation of two phases. The lower, CDCl3-rich phase was analyzed by 1H-NMR using a JEOL ECX-400 spectrometer (400 MHz) (JEOL Ltd.).
3.5.7 H2-chemisorption
Pulse H2 chemisorption was performed on a Chembet 3000 (Quantachrome) at T = 273 K to
determine the hydrogen uptake and dispersion of Pt on the porous silica. Two independent
investigations of the Pt / silica catalyst (dp = 125 - 400 µm) were carried out, after drying with
argon (T = 423 K, 10 K min-1, 1 h), and after reduction in hydrogen (10 vol. % H2 in argon,
T = 473 K, t = 0.5 h; desorption under argon flow at T = 473 K for 0.75 h). For analysis,
pulses of hydrogen in argon (10 vol. %) (Vpulse = 200 µl, ∆ tpulse = 3.5 min) were injected.
3.5.8 X-ray diffraction (XRD)
For the analysis of the in situ formed Pt-particles of the experiments with the catalytic system
PtCl2 / [BMIM]Cl / AlCl 3 and PtCl4 / [BMIM]Cl / AlCl 3, XRD was conducted. The
diffractograms were generated on an X´Pert Pro Diffractometer (Philips) under 40 kV and
40 mA with a CuKα-source (wavelength: 0.15418 nm). The samples were analyzed with a
step size of 0.02 °, 5 s analysis time each and in the range of 2 – 80 °. For the XRD analysis,
the AlCl3 was hydrolized with ethanol and sodium hydrogen carbonate after reaction. The
black particles were separated from the hydrolized ionic liquid by a centrifuge (VWR Galaxy
MiniStar), washed with ethanol and dried under vacuum. The sample was mixed with
charcoal to gain the minimum volume necessary to fill the sample holder.
The linear dimension of the particle L was calculated according to the Scherrer-formula [243]
(Equation 13) in which λ is the wavelength of the incident x-rays, θ is the Bragg angle, B the
half-value breadth of the diffracted beam and K is a numerical constant. A cubic three-
dimensional crystal is best described by K = 0.94 [243].
Experimental 61
L= K· λ
B ·cosθ (13)
3.6 Further calculations
3.6.1 Modified reaction time
A modified reaction time tmod. which referrs to the active component present in all
experiments, the ionic liquid, was used for evaluation of the experiments (Equation 14)
tmod. =treaction·n ionic liquid
n n-alkane (14)
The modified reaction time tmod. includes the reaction time treaction, the molar amount of
catalytically active Lewis acidic ionic liquid nionic liquid (molar amount of ionic liquid used in
liquid-liquid biphasic experiments or used for the impregnation of the SILP and SCILL
catalyst) and the molar amount of n-octane (nn-octane) and n-hexane (nn-hexane), respectively.
3.6.2 Aspen Plus simulation
Aspen Plus® V 7.2 by AspenTech was used to calculate the solubility of all product alkanes
in the solvent cyclohexane (20 mol. % product, 80 mol. % cyclohexane) used for GC analysis
to guarantee the feasibility of the sampling method. Moreover, calculations were also
conducted without the solvent cyclohexane as n-octane and further alkanes, which are liquid
at ambient conditions, were available in excess in the organic. These can also dissolve
products like isobutane whereby the vapor pressure is lowered. Due to practical reasons, only
n-alkanes with 3 ≤ C ≤10 were used for the calculation because numerous isomers exist for
each alkane (e.g. C6: 5 isomers, C7: 9 isomers, C8: 18 isomers) and the boiling points of the
respective iso-alkanes range between the used n-alkanes. The molar fractions of the alkanes
applied for the calculation correspond to the composition of the product at maximum
modified reaction and conversion gained in this thesis and thus describes the worst case. Flash
calculation was conducted and the molar fractions of the gas-phase in dependency on the flash
temperature are shown in Figure 8.
62 Experimental
Figure 8: Molar fraction in the gas-phase in dependency on the flash temperature (pflash = 1 bar) for the pure product and product in cyclohexane (20 mol. % product, 80 mol. % cyclohexane); NRTL gE-model was used for the flash calculation.
Boiling of the pure products starts at T = 296 K (p = 1 bar) and at T = 334 K (p = 1 bar) if the
products are dissolved in excess cyclohexane. Therefore, the sampling method described in
chapter 3.4 was applicable for the conducted alkane conversion experiments.
270 280 290 300 310 320 330 340 350 360 370 380
0
10
20
30
40
50
60
70
80
90
100
product in cyclohexane
pure product
mol
ar fr
actio
n in
the
gas
phas
e / m
ol %
temperature / K
63
___________________________________________________________________________
4. Results and discussion
64 Results and discussion
Results and discussion 4
The results of the n-alkane (n-hexane or n-octane) isomerization investigated in this thesis are
basically divided according to the two different catalyst types applied, into results of the
monofunctional and bifunctional acidic ionic liquid based catalysts. The experiments with the
monofunctional acidic ionic liquid were conducted in a liquid-liquid biphasic reaction mode
and moreover in a slurry-phase reaction mode if the acidic ionic liquid was immobilized on a
porous support (SILP catalyst). Further, two bifunctional catalyst types were tested, namely
the SCILL system (Pt / silica pretreated, coated with n([BMIM]Cl) / n(AlCl 3) = 1/2) and in
situ formed Pt nanoparticles in an acidic ionic liquid.
4.1 Catalytic experiments with the monofunctional acidic ionic
liquid catalysts in a liquid-liquid biphasic reaction mode
First experiments focus on a screening that was necessary to find active ionic liquid
isomerization catalysts. Besides, the selectivity for branched alkanes and especially for
iso-octanes is part of the discussion to prove the suitability of acidic ionic liquids for alkane
isomerization, especially for long-chain alkanes (C > 6). Different parameters were changed
(hydrogen partial pressure, reaction temperature and ionic liquid acidity) to check if the
selectivity for the iso-hexanes and iso-octanes, respectively, could be improved.
4.1.1 Reproducibility of the data
At the beginning, reproducibility experiments were conducted to ensure repeatable batch runs
as well as ionic liquid synthesis. Experiments presented in this chapter were performed with
the reactant n-hexane under helium atmosphere. Results of the influence of modified reaction
time on n-hexane conversion are depicted in Figure 9. The selectivity-conversion curve is
shown in Figure 10. Moreover, the deviation δ(X) of the conversions was calculated
according to Equation 15. The values of δ(X) are listed in Table 11.
Results and discussion 65
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n (H2SO4) / n (AlCl3) = 0.18, Treaction = 303 K, phelium = 15 bar.
Figure 9: Reproducibility of the experimental data in the n-hexane isomerization (influence of modified reaction time on n-hexane conversion).
deviation δ(y) = exp. data y (experiment 1)- exp. data y (experiment 2)
exp. data y (experiment 1) (15)
Table 11: Deviation δ(X) of the n-hexane conversion for experiment 1 and 2 (Figure 9).
tmod. / min molionic liquid mol-1n-hexane δ(X) / %
6.7 8.6
13.4 -4.4
20.1 -2.8
26.8 -2.7
Deviation δ(S) (Equation 15) was also calculated for the iso-hexane selectivities. As the
selectivity strongly dependens on n-hexane conversion, selectivity values of experiment 2
were calculated on the basis of an exponential fit of the experimental data. The same
0 5 10 15 20 25 300
10
20
30
40
50
60
X n
-hex
ane /
%
modified reaction time / min mol ionic liquid
mol-1 n-hexane
experiment 1 experiment 2
66 Results and discussion
conversions achieved in experiment 1 were used for the calculations of δ(S). Table 12
contains the as-obtained deviations δ(S).
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n(H2SO4) / n(AlCl3) = 0.18, Treaction = 303 K, phelium = 15 bar.
Fit (exponential), y = y0 + A·exp (R0·x); experiment 1: y0 = 22.68, A = 80.14, R0 = -0.097; R2 = 0.9982; experiment 2: y0 = 22.89, A = 86.35, R0 = -0.110; R2 = 0.9986.
Figure 10: Reproducibility of the experimental data in the n-hexane isomerization (influence of n-hexane conversion on selectivity for iso-hexanes).
Table 12: Deviation δ(S) of the selectivity for iso-hexanes for experiment 1 and 2 (Figure 10); experiment 1: Siso-hexanes (experimental data), Xn-hexane (experimental data); experiment 2: Siso-hexanes (fit 2), Xn-hexane = Xn-hexane (experiment 1).
Xn-hexane / % δ(S) / %
6.5 0.43
23.3 4.17
32.1 5.53
38.4 0.16
0 10 20 30 40 50 600
10
20
30
40
50
60
70
80
90
100
S is
o-he
xane
s / %
X n-hexane
/ %
experiment 1 experiment 2 fit for experiment 1 fit for experiment 2
Results and discussion 67
In conclusion, reproducibility of the batch runs and ionic liquid synthesis was sufficiently
given for the present purpose. The average deviation of the conversion δ(X) accounts for
4.6 % and of the selectivity δ(S) for 2.6 %.
4.1.2 Catalyst screening experiments for system selection
4.1.2.1 Experiments with the system [BMIM]Cl / AlCl 3 / H2SO4
Literature review (chapter 2.6.2) gave proof that chloroaluminate ionic liquids are possible
alkane isomerization catalysts. Moreover, higher proton concentrations in the highly acidic
ionic liquid increase the catalytic acitivty. One possibility to increase the proton concentration
in chloroaluminate ionic liquids is the addition of H2SO4. The effect of H2SO4 addition to
n([BMIM]Cl) / n(AlCl 3) = 1/2 was systematically studied by varying the molar ratio
H2SO4 / AlCl3 from 0 (pure chloroaluminate ionic liquid) to 0.75. The results of n-octane
conversion, the yield of liquid organic product (LOP), that is all liquid products without
n-octane, selectivity for iso-octanes and selectivity for branched alkanes within the LOP in
dependency on the molar ratio H2SO4 / AlCl3 are depicted in Figure 11.
It is evident that addition of the Brønsted acid H2SO4 to the Lewis acidic chloroaluminate
system increased n-octane conversion up to the H2SO4 / AlCl3 ratio of 0.18. At higher H2SO4
concentration, the catalytic activity decreased again. This indicates a maximum of the acidity
in this specific composition correlating to a maximum n-octane conversion of 77.2 % with a
yield of branched alkanes (C4 - C10) in the LOP of 61.7 %. The yield of branched C6 - C8
alkanes added up to 32.4 %.
These findings can be understood in the light of the known mechanism of alkane
isomerization (see chapter 2.3.1). The latter is considered to proceed as a chain reaction
including an initiation step that requires a superacidic proton to form the first carbenium ion
via direct hydride abstraction (Scheme 3). The highest catalytic activity found for the molar
ratio range 0.14 < H2SO4 / AlCl3 < 0.21 can be explained by the reactions and equilibria
present in the ionic liquid system to form highly acidic, “naked” protons (Scheme 19).
68 Results and discussion
Conditions: tmod. = 62 min molionic liquid mol-1n-octane, n([BMIM]Cl) / n(AlCl 3) = 1/2 , Treaction = 303 K, pargon = 3 bar,
Figure 11: Influence of the molar ratio H2SO4 / AlCl 3 in the acidic ionic liquid system [BMIM]Cl / AlCl 3 / H2SO4 on n-octane conversion, yield of liquid organic product (LOP), selectivity for iso-octanes and selectivity for iso-alkanes within the LOP; the rest to obtain 100 % yield reflects gaseous C5 and C4.
It is obvious that the highest acidity is obtained if each HCl formed by the complexation of Al
with the sulfate ion is reacting with one of the Lewis acidic ions [Al2Cl7]- (or [Al3Cl10]
-) to
form the “naked” protons. In this case, the highest concentration of “naked” protons is
generated in the least coordinating environment. From the stoichiometry of Scheme 19 it can
be assumed that the H2SO4 / [Al2Cl7]--ratio resulting in the highest ionic liquid acidity should
thus be in the ratio of 3/8. Given the fact that the formation of each [Al2Cl7]- ion requires two
molar equivalents of AlCl3, of which the first is only neutralizing the chloride salt to the
neutral [AlCl4]-, a H2SO4 / AlCl3 ratio of 3/16 or 0.1875 in the ionic liquid synthesis would be
expected to provide the highest acidity. This predicted ideal composition corresponds well to
the H2SO4 / AlCl3-ratio found to be optimum in the catalytic experiments. Exceeding this
optimum molar ratio, more [Al2Cl7]- and [Al3Cl10]
- anions are consumed by reacting to
Al 2(SO4)3. In this case, the molar amount of formed HCl exceeds the molar amount of
remaining acidic anions and therefore HCl gas leaves the system and the amount of “naked”
proton formed drops. Thus, the overall system acidity drops again at higher H2SO4 contents.
In all these experiments, the formation of solid Al2(SO4)3 in the reaction system was observed
and confirmed by isolation and analysis of the precipitate using BaCl2. However, isolation and
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.80
10
20
30
40
50
60
70
80
90
100
Xn-
octa
ne, Y
LOP, S
iso-
alka
nes /
%
n(H2SO
4) / n(AlCl
3)
Xn-octane
YLOP
Siso-octanes
Siso-alkanes (within the LOP)
Results and discussion 69
removal of Al2(SO4)3 was not necessary to apply the highly acidic ionic liquid system in
catalysis. The most basic component of the system, the sulfate ion, is removed from the
system by the precipitation of Al2(SO4)3.
Scheme 19: In situ HCl generation by H2SO4 addition to a Lewis acidic ionic liquid [cation]Cl / AlCl 3 with x(AlCl 3) > 0.5.
The in situ formation of HCl in the Lewis acidic ionic liquid n([BMIM]Cl ) / n(AlCl3) = 1/2
generates a superacidic catalyst according to Smith [11, 128]. Though, the
[BMIM]Cl / AlCl 3 / H2SO4 system is much easier to handle and safer to work with compared
to the [cation]Cl / AlCl3 / HCl systems [8]. The handling of toxic and highly corrosive HCl
gas is avoided. Proof of superacidity of the applied reaction system was not part of this work.
Daniel Schmitt investigated the acidity of different highly acidic ionic liquids, among them
chloroaluminate ionic liquids, with the Gutmann acceptor number based on the 31P-NMR
chemical shift using the indicator triethylphosphine oxide [244].
The [BMIM]Cl / AlCl 3 / H2SO4 system was already catalytically active at a temperature as
low as 303 K. In a next series of experiments, it was tested whether the isomerization rate and
selectivity could be further improved by temperature variation (T reaction = 283 – 323 K). For
this, the most active [BMIM]Cl / AlCl3 / H2SO4 system (n(H2SO4) / n(AlCl3) = 0.18) was
used. As can be seen from Figure 12, it was found that low temperatures are favorable to
maximize the selectivities for branched alkanes within the LOP (99.1 % at 283 K versus
93.4 % at 323 K) and selectivity for iso-octanes (16.7 % at 283 K versus 7.4 % at 323 K),
while temperature increase led to higher overall n-octane conversion (53.5 % at 283 K versus
91.1 % at 323 K).
Generally, selectivity for the isomerized alkane has always to be discussed in the light of the
alkane conversion. According to Weitkamp [61], the rate of β-scission of carbenium ions is
drastically increased if the number of C atoms in the respective alkane is ≥ 8. Once the
carbenium ions are generated, consecutive cracking reactions of mono- and dibranched octane
isomers are faster than hydride transfer from other alkanes in the reaction mixture to generate
the iso-octanes. The decrease of selectivity for branched octanes with increasing n-octane
70 Results and discussion
conversion is in agreement with this mechanistic interpretation. Therefore, S-X-plots are
coercively necessary if selectivities are compared.
Conditions: tmod. = 62 min molionic liquid mol-1n-octane, n(H2SO4) / n(AlCl3) = 0.18, pargon = 3 bar, n([BMIM]Cl) / n(AlCl 3) = 1/2 .
Figure 12: Influence of reaction temperature on n-octane conversion, yield of LOP, selectivity for iso-octanes and selectivity for iso-alkanes within the LOP.
Compared to classical bifunctional heterogeneous isomerization catalysts (see chapter 2.6.1),
catalytic activity of the ionic liquid systems was proven at much lower reaction temperatures.
The yield of LOP showed a maximum at 303 K. Cracking reactions got dominant for higher
temperatures, which was also reflected in the decrease of selectivity for iso-octanes. The
selectivity for branched octane isomers was low (below 20 % for all experiments). However,
products consisted solely of alkanes in the range of C4 – C10, whereupon most alkanes were
branched. Cracked branched alkane products which are liquid at ambient conditions can also
contribute to gasoline providing that the RON is high enough.
4.1.2.2 Experiments with the system [BMIM]Cl / AlCl 3 / CuCl2
Another possibility of in situ HCl generation and with it the creation of a superacidic ionic
liquid catalyst is the addition of CuCl2 to chloroaluminate ionic liquids.
The low temperature conversion of n-pentane [245] and the alkylation of benzene with
iso-pentane [246, 247] in the presence of AlCl3 and CuCl2 was described in literature. Olah
280 290 300 310 320 3300
10
20
30
40
50
60
70
80
90
100
X
n-oc
tane, Y
LOP, S
iso-
alka
nes
/ %
temperature / K
Xn-octane
YLOP
Siso-octanes
Siso-alkanes (within the LOP)
Results and discussion 71
suggested the protonolysis of iso-pentane, with protic impurities acting as coacids for AlCl3,
forming the corresponding tert-alkyl cation. CuCl2 oxidizes the hydrogen formed in the
protonolysis, thus making the reaction thermodynamically more feasible. This redox reaction
is shown in general in Scheme 20.
Scheme 20: Alkane activation in the presence of AlCl3 and CuCl2 [246].
This reaction mechanism can be transferred to chloroaluminate ionic liquids with
x(AlCl3) > 0.5. Despite handling all chemicals, catalysts and the reaction mixture under
argon, protic impurities in the acidic chloroaluminate ionic liquids are unavoidable, but, at the
same time, were even necessary in this case to start the redox reaction.
The catalytic system [BMIM]Cl / AlCl3 / CuCl2 was tested in the n-octane conversion and
compared to the experiment with solely Lewis acidic ionic liquid
n([BMIM]Cl) / n(AlCl 3) = 1/2 (Table 13).
Table 13: Influence of CuCl2 addition to the Lewis acidic ionic liquid [BMIM]Cl / AlCl 3 on n-octane conversion, selectivity for iso-octanes and yield of LOP.
[BMIM]Cl / AlCl 3 [BMIM]Cl / AlCl 3 / CuCl2
Xn-octane / % 30.6 42.2
Siso-octanes / % 7.8 16.8
YLOP / % 12.4 26.3
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n(CuCl2) / n(AlCl3) = 0.06, Treaction = 303 K, tmod. = 62 min molionic liquid mol-1n-octane, pargon = 3 bar.
CuCl2 increased the n-octane conversion, selectivity for iso-octanes and yield of LOP which
can be explained by the increased carbocation concentration in the chloroaluminate system
according to Scheme 20. Higher catalytic activity, yield of LOP and selectivity for branched
alkanes as consequence of increasing ionic liquid acidity was also observed for the
[BMIM]Cl / AlCl 3 / H2SO4 system (0 < H2SO4 / AlCl3 < 0.21) (Figure 11). Screening and
optimization of the [BMIM]Cl / AlCl3 / CuCl2 system was not extended as CuCl2 was only
RH + H+ + 2 CuCl2 R+ + 2 CuCl + 2 HClAlCl3
RH + HCl + AlCl3R+[AlCl4]- + H2
72 Results and discussion
partially soluble in n([BMIM]Cl) / n(AlCl3) = 1/2 which was confirmed by ICP-AES
measurements and resulted in a slurry-phase reaction mixture. The soluble amount of CuCl2
in n([BMIM]Cl) / n(AlCl 3) = 1/2 was restricted to n(CuCl2) / n(AlCl3) = 6.6·10-3.
4.1.2.3 Experiments with the system [BMIM]Cl / AlCl 3 / 1-chlorooctane
The third approach of an acidic ionic liquid based isomerization catalyst comprises the
addition of an alkyl halide to the Lewis acidic ionic liquid and utilizes the promoting effect of
this alkyl halides in acidic chloroaluminate ionic liquids [161]. The effect of added promoter
1-chlorooctane to the Lewis acidic [BMIM]Cl / AlCl3 was tested systematically for varying
molar ratios of 1-chlorooctane / AlCl3 ranging from 0 (pure chloroaluminate ionic liquid) to
0.6. As the octylcarbenium is coercively an intermediate in the n-octane isomerization,
1-chlorooctane was chosen as promoter. It is very likely that the generated primary octyl
carbenium ion is immediately rearranged to more stable carbocations because the primary one
is very unstable in a thermodynamic point of view. Figure 13 illustrates the promoting effect
of 1-chlorooctane in the n-octane conversion.
n-Octane conversion increased with promoter addition up to
n(1-chlorooctane) / n(AlCl 3) = 0.5. The curve of n-octane conversions can be explained by the
higher concentration of carbenium ions in the system with an increasing concentration of
1-chlorooctane. However, the slope of this graph lowers down with higher promoter
concentrations. This can be rationalized by the loss of acidity in the catalytic system
according to Scheme 21 as the concentration of Lewis neutral [AlCl4]- anions in the ionic
liquid system rises if more 1-chlorooctane is added.
Scheme 21: [AlCl4]- formation by alkyl halide addition to a Lewis acidic chloroaluminate ionic
liquid.
If the amount of 1-chlorooctane moles equaled half of the moles of AlCl3 used for
n([BMIM]Cl) / n(AlCl 3) = 1/2 synthesis, maximum conversion was reached because all acidic
anions are converted to neutral [AlCl]4- under this prerequisite. Thus, the conversion for
n(1-chlorooctane) / n(AlCl3) = 0.5 and 0.6 were found to be almost identical (46.1 % at the
molar ratio of 0.5 versus 46.4 % at the molar ratio of 0.6) (Figure 13).
Results and discussion 73
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 303 K, tmod. = 27 min molionic liquid mol-1n-octane, phelium = 15 bar.
Figure 13: Influence of the molar ratio 1-chlorooctane / AlCl3 on the n-octane conversion, selectivity for iso-octanes and selectivity for iso-alkanes.
The selectivity for iso-octanes decreased with higher amount of 1-chlorooctane (29.5 % at the
molar ratio of 0 down to 17.8 % at the molar ratio of 0.6) which can be explained by the
dependency of the selectivity on the conversion.
For the whole series of experiments all products were alkanes ranging from C4 to C10 with
very high selectivities for branched alkanes (> 97 %).
4.1.3 Recyclability of the acidic ionic liquid systems
[BMIM]Cl / AlCl 3 / H2SO4 and [BMIM]Cl / AlCl 3 / 1-chlorooctane
Three different acidic ionic liquid systems were tested for their ability to convert n-octane
(see chapter 4.1.2). The recyclability should be tested for the two most promising ionic liquid
systems [BMIM]Cl / AlCl3 / H2SO4 and [BMIM]Cl / AlCl3 / 1-chlorooctane. The results are
listed in Table 14 and Table 15. For the recycling experiments, most of the organic phase was
drawn off with a syringe after settling time at ambient reaction condition and replaced with
n-octane. A small amount of organic left in the glass liner acted as protective layer against
moisture for the ionic liquid layer during loading of n-octane.
0.0 0.1 0.2 0.3 0.4 0.5 0.60
10
20
30
40
50
60
70
80
90
100
Xn-
octa
ne, S
iso-
alka
nes /
%
n(1-chlorooctane) / n(AlCl3)
Xn-octane
Siso-octanes
Siso-alkanes
74 Results and discussion
Both systems showed catalytic activity after recycling. The [BMIM]Cl / AlCl 3 / H2SO4
catalyst was tested in four experiments (Table 14) with different reaction times and masses of
ionic liquid as well as n-octane.
Table 14: Four different recycling experiments with the catalytic system [BMIM]Cl / AlCl 3 / H2SO4.
1st run 2nd run decrease Xn-octane / %
Experiment 1: Xn-octane / % 43.6 34.0 22.0
Experiment 2: Xn-octane / % 44.9 29.5 34.3
Experiment 3: Xn-octane / % 10.6 5.0 52.8
Experiment 4: Xn-octane / % 9.0 4.1 54.4
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2 , n(H2SO4) / n(AlCl3) = 0.18, Treaction = 303 K, pargon = 3 bar;
Experiment 1: mn-octane = 30 g, mionic liquid = 20 g, treaction = 2.5 h (per run), tmod. = 25.9 min molionic liquid mol-1n-octane; Experiment 2: mn-octane = 40 g, mionic liquid = 20 g, treaction = 4.0 h (per run), tmod. = 31.1 min molionic liquid mol-1n-octane; Experiment 3: mn-octane = 92 g, mionic liquid = 20 g, treaction = 2.0 h (per run), tmod. = 6.8 min molionic liquid mol-1n-octane; Experiment 4: mn-octane = 92 g, mionic liquid = 20 g, treaction = 2.0 h (per run), tmod. = 6.8 min molionic liquid mol-1n-octane.
The loss of catalytic activity could be assigned to loss of HCl which is in equilibrium between
the three phases, the ionic liquid, organic and gas-phase. Because of reactor opening, pressure
release and removal of the organic, some of the HCl might have been lost which led to lower
acidity in the catalytic phase. The addition of HCl to the reaction mixture could propably
maintain constant catalytic activity in the highly acidic chloroaluminate system. Recycling
without significant loss of activity applying highly acidic molten salt catalysts under HCl
atmosphere has been successfully demonstrated in our work group [248]. However, a
different reactor set-up, which consists of glass, would be necessary to conduct recyling
experiments under HCl atmosphere and was not available for this work. Moreover, short air
contact while n-octane loading was unavoidable which resulted in irreversible deactivation of
the oxophilic chloroaluminate anions. The significant deviation in conversion decrease for the
different experiments could be referred to unequal reactant and catalyst masses and therefore
varying modified reaction times. Experiment 3 and 4 were conducted under the same reaction
conditions. Thus, their deviations are congruent. Higher modified reaction times (experiment
Results and discussion 75
1 and 2) in constrast to those in experiment 3 and 4 led to better catalytic activity in the
second recycling run.
Another two different experiments were run with the [BMIM]Cl / AlCl 3 / 1-chlorooctane
system (Table 15). In experiment 1, the separated mass of organic was replaced with n-octane
for the second and third run. In contrast, n-octane and 1-chlorooctane were added for the
second run as well as third run and replaced the drawn off organic phase in experiment 2. The
same molar ratio of 1-chlorooctane / n-octane was adjusted for the first, second and third run
in experiment 2.
Table 15: Two different recycling experiments with the catalytic system [BMIM]Cl / AlCl 3 / 1-chlorooctane.
1st run 2nd run 3rd run decrease Xn-octane / % (1�2; 2�3)
Experiment 1: Xn-octane / % 15.5 5.7 2.6 63.2; 54.4
Experiment 2: Xn-octane / % 12.3 9.1 6.6 26.0; 27.5
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2 , n(1-chlorooctane) / n(AlCl3) = 0.36, Treaction = 303 K, pargon = 3 bar;
Experiment 1: mn-octane = 92 g, mionic liquid = 20 g, treaction = 2 h (per run), tmod. = 6.8 min molionic liquid mol-1n-octane; Experiment 2: mn-octane = 92 g, mionic liquid = 20 g, treaction = 1 h (per run), tmod. = 3.4 min molionic liquid mol-1n-octane, n(1-chlorooctane) / n(n-octane) = 0.04.
A significant decrease of n-octane conversion in the second and third run of experiment 1
could be explained by lower carbenium ion concentration compared to the first run. A similar
behavior was observed by the group of Jess in continuous alkylation of isobutane with
2-butene catalyzed with [BMIM]Cl / AlCl3 (x(AlCl3) = 0.64) and one-time addition of a
tert-alkyl halide promoter [161]. They proposed that isobutene was generated out of the
tert-butyl cation. Simultaneously, a proton was released which is highly Brønsted acidic since
the corresponding anion [AlCl4]- is a very weak base in the ionic liquid catalyst. The proton
could protonate an alkene or react with [AlCl4]- according to Scheme 22. The hypothetical
H[AlCl 4] acid [133] would decompose instantaneously into HCl and AlCl3.
76 Results and discussion
Scheme 22: Decomposition of hypothetical H[AlCl4] [161].
In turn, like in the case of the [BMIM]Cl / AlCl 3 / H2SO4 system, it is propable that some
amount of the HCl was lost because of reactor opening and organic withdrawal. This could
explain the conversion decrease in experiment 1. It is imaginable that the generated alkene
underwent an alkylation reaction with an alkyl cation yielding in higher molecular weight
alkanes [47] in the reaction which was conducted in a liquid-liquid biphasic reaction mode in
the batch autoclave.
In experiment 2, 1-chlorooctane and n-octane were added to the reaction mixture before the
second and third run providing the possibility that new carbenium ions were formed. Thus,
the decrease of conversion amounted only to 26.0 % and 27.5 %, respectively, in contrast to
63.2 % and 54.4 % in experiment 1. However, deactivation still took place because it might
be probable that the ionic liquid acidity decreased with every run as the concentration of
[AlCl 4]- increased. Furthermore, short air contact while n-octane loading was unavoidable.
4.1.4 Influence of hydrogen partial pressure
The ability of hydrogen to act as cracking inhibitor was proposed by Olah for AlCl3 based
catalysts [15] (see also chapter 2.3.1.5). Therefore, the effect of hydrogen on the catalyst´s
activity and selectivity was tested, firstly for the n-octane conversion.
4.1.4.1 Influence of hydrogen partial pressure variation on n-octane isomerization at
very mild reaction conditions
The catalytic system [BMIM]Cl / AlCl3 / H2SO4 was applied for the next series of
experiments. Figure 14 and Figure 15 contain the influence of hydrogen partial pressure
variation (0, 15 and 40 bar) on n-hexane conversion and iso-hexanes selectivity at 303 K.
The n-octane conversions of the three experiments were in a similar range (Xn-octane = 36.2 %
at phydrogen = 0 bar, Xn-octane = 38.0 % at phydrogen = 15 bar and Xn-octane = 38.7 % at
phydrogen = 40 bar) at maximum applied modified reaction time. The GC analyses of the
products of the conducted n-octane experiments showed alkanes in the range of 4 ≤ C ≤ 10
whereupon the selectivity for branched alkanes was always > 90 %. The conversions differed
more at lower reaction times. No definite explanation can be given for this behavior.
Results and discussion 77
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n(H2SO4) / n(AlCl3) = 0.18, Treaction = 303 K, ptotal = 15 bar for phydrogen = 0 or 15 bar, ptotal = 40 bar for phydrogen = 40 bar, ptotal = phelium + phydrogen.
Figure 14: Influence of modified reaction time and hydrogen partial pressure on the n-octane conversion.
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n(H2SO4) / n(AlCl3) = 0.18, Treaction = 303 K, ptotal = 15 bar for phydrogen = 0 or 15 bar, ptotal = 40 bar for phydrogen = 40 bar, ptotal = phelium + phydrogen.
Figure 15: Influence of n-octane conversion and hydrogen partial pressure on the selectivity for iso-octanes.
0 5 10 15 20 25 300
10
20
30
40
50
60
Xn-
octa
ne /
%
modified reaction time / min molionic liquid
mol-1 n-octane
phydrogen
= 0 bar (n-octane)
phydrogen
= 15 bar (n-octane)
phydrogen
= 40 bar (n-octane)
0 10 20 30 40 50 600
10
20
30
40
50
60
70
80
90
100
S iso-
octa
nes /
%
Xn-octane
/ %
phydrogen
= 0 bar (n-octane)
phydrogen
= 15 bar (n-octane)
phydrogen
= 40 bar (n-octane)
78 Results and discussion
The selectivity values of all three isomerization experiments exhibited nearly identical
dependency on n-octane conversion. Thus, it could be concluded that the reactant hydrogen
could not improve the selectivity for iso-octanes.
4.1.4.2 Influence of hydrogen partial pressure on n-octane isomerization at elevated
reaction temperature
The next series of catalytic experiments tested the influence of hydrogen pressure on the
n-octane isomerization at higher reaction temperature (Treaction = 393 K). Moreover,
1-chlorooctane was added to the Lewis acidic ionic liquid catalyst [BMIM]Cl / AlCl3.
Figure 16 indicates the results of varying hydrogen partial pressure (phydrogen = 0, 15 and
40 bar) on n-octane conversion at 393 K for two different amounts of the promoter
1-chlorooctane (n(Cl-octane) / n(AlCl3)) = 0.036 and 0.36). From the results it is again
evident that the presence of hydrogen and the level of hydrogen pressure did not result in
significantly higher catalytic activity. The higher conversions obtained for the higher
concentration of the 1-chlorooctane promoter can be rationalized by the higher concentration
of carbenium ions in the system which was already shown and discussed in chapter 4.1.2.3.
Selectivity for iso-octanes was found to be a function of n-octane conversion but independent
of the hydrogen partial pressure for this monofunctional catalyst (Figure 17). Thus, also at a
higher temperature level hydrogen did not act as cracking inhibitor in the case of the
monofunctional Lewis acidic ionic liquid catalyst.
Results and discussion 79
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 393 K, ptotal = 40 bar.
Figure 16: Influence of hydrogen partial pressure on n-octane conversion for the monofunctional Lewis acidic ionic liquid catalyst using different concentrations of the promoter 1-chlorooctane.
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 393 K, ptotal = 40 bar.
Figure 17: Influence of hydrogen partial pressure and n-octane conversion on the selectivity for iso-octanes for the monofunctional Lewis acidic ionic liquid catalyst using different concentrations of the promoter 1-chlorooctane.
0 5 10 150
10
20
30
40
50
60
70
80
Xn-
octa
ne /
%
modified reaction time / min molionic liquid
mol-1 n-octane
phydrogen
= 0 bar phydrogen
= 15 bar phydrogen
= 40 bar
n(Cl-octane) / n(AlCl3) = 0.036
phydrogen
= 0 bar phydrogen
= 15 bar phydrogen
= 40 bar
n(Cl-octane) / n(AlCl3) = 0.36)
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
S iso-
octa
nes /
%
Xn-octane
/ %
phydrogen
= 0 bar phydrogen
= 15 bar phydrogen
= 40 bar
n(Cl-octane) / n(AlCl3) = 0.036
phydrogen
= 0 bar phydrogen
= 15 bar phydrogen
= 40 bar
n(Cl-octane / n(AlCl3) = 0.36
80 Results and discussion
4.1.4.3 Influence of hydrogen on n-hexane isomerization at very mild reaction
conditions
Finally, the effect of hydrogen on n-hexane, which is not as pronounced to cracking side
reactions as n-octane, was investigated using the [BMIM]Cl / AlCl 3 / H2SO4 system at 303 K.
Reactant n-hexane was converted under different hydrogen partial pressures (0, 5, 10 and
15 bar). The results of the hydrogen influence on conversion and selectivity for iso-hexanes
are presented in Figure 18 and Figure 19. The conversions at the maximum modified reaction
time applied differed significantly and rose (Xn-hexane = 3.0, 6.7, 36.2 and 38.3 %) with
decreasing hydrogen partial pressure (phydrogen = 15, 10, 5 and 0 bar) (Figure 18).
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n(H2SO4) / n(AlCl3) = 0.18, Treaction = 303 K, ptotal = 15 bar, ptotal = phelium + phydrogen.
Figure 18: Influence of hydrogen partial pressure and modified reaction time on n-hexane conversion.
These results can be explained with the help of theoretical considerations of chapter 2.3.1.5
(cracking inhibitor hydrogen). Hydrogen can react with carbenium ions releasing an alkane
and a proton. In the case of the experiments with high hydrogen pressure (phydrogen = 15 or
10 bar), it might be possible that carbenium ions were generated but their skeletal
rearrangement was rate determing and slow compared to their reaction with hydrogen.
Cracking reactions of linear carbenium ions are thermodynamically not favoured because they
belong to type D mode of β-scission according to Weitkamp (Scheme 9). Furthermore, the
0 5 10 15 20 25 300
10
20
30
40
50
60
Xn-
hexa
ne /
%
modified reaction time / min molionic liquid
mol-1n-hexane
phydrogen
= 0 bar
phydrogen
= 5 bar
phydrogen
= 10 bar
phydrogen
= 15 bar
Results and discussion 81
pathways for cracking side reactions are limited for hexane. Only monobranched hexane
isomers are potential reactants for cracking. An isomerization step to form a methylpentane is
necessary at the beginning (Scheme 23). The consecutive character of the isomerization and
cracking reaction can be seen for the experiment with phydrogen = 5 bar. The rate of conversion
was low at the beginning. However, with rising monobranched hexane isomer concentration
cracking drastically increased leading to higher conversions.
Scheme 23: Possible isomerization reactions starting with n-hexane.
This explanation was underlined by the significant selectivity drop (Figure 19,
phydrogen = 5 bar: Siso-hexanes = 65.0 % at Xn-hexane = 9.6 % versus Siso-hexanes = 32.2 % at
Xn-hexane = 30.6 %). The curve of selectivity for iso-hexanes in dependency on the conversion
was slightly shifted to higher values for phydrogen = 5 bar compared to the catalytic run with
phydrogen = 0 bar. For the other two experiments (phydrogen = 10 and 15 bar), high selectivities
(82 – 90 %) could be attributed to overall low conversions (2.7 – 6.7 %). At phydrogen = 15 bar,
almost no n-hexane was converted at all (Xmax = 3.0 %). In general, all products of these
experiments with reactant n-hexane ranged between C4 – C9.
In conclusion, hydrogen influenced the rate of reaction negatively. Moreover,
hydroisomerization of n-hexane might only lead to slightly increased selectivities for
iso-hexanes compared to experiments without hydrogen. The small differences between
phydrogen = 0 bar and phydrogen = 5 bar could also be explained by the deviation δ(S) of the
experimental data. However, selectivity increase was basically just a function of reduced
n-hexane conversion in the presence of hydrogen.
82 Results and discussion
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n(H2SO4) / n(AlCl3) = 0.18, Treaction = 303 K, ptotal = 15 bar, ptotal = phelium + phydrogen.
Figure 19: Influence of hydrogen partial pressure and n-hexane conversion on the selectivity for iso-hexanes.
4.1.5 Variation of the reaction temperature in the hydroisomerization of
n-hexane
Temperature variation (Treaction = 293, 303 and 313 K) under hydrogen pressure
(phydrogen = 10 bar) was conducted for the short chain reactant n-hexane to test whether the
isomerization selectivity could be improved. Lower temperature favors higher branched
alkanes thermodynamically (Figure 1) and might not favor endothermic cracking side
reactions.
Figure 20 shows a significant temperature influence on the conversions. Considerable
differences in n-octane conversion occured (Xn-octane = 3.72 % at 293 K, Xn-octane = 6.7 % at
303 K and Xn-octane = 50.3 % at 313 K) at the same modified reaction time
(tmod. = 26.8 min molionic liquid mol-1n-alkane). The isomerization rate increased with temperature
according to Arrhenius law (see Figure 22 and Figure 23). Moreover, it is likely that only
branched hexanes undergo cracking reactions according to the mechanistic proposal. Cracking
reactions, which contributed mostly to the n-hexane conversion as the skeletal isomerization
step was obviously very slowly, occured predominantly at 313 K.
0 10 20 30 40 50 600
10
20
30
40
50
60
70
80
90
100
S iso-
hexa
nes /
%
Xn-hexane
/ %
phydrogen
= 0 bar
phydrogen
= 5 bar
phydrogen
= 10 bar
phydrogen
= 15 bar
Results and discussion 83
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n(H2SO4) / n(AlCl3) = 0.18, phydrogen = 10 bar, ptotal = 15 bar, ptotal = phelium + phydrogen.
Figure 20: Influence of reaction temperature and modified reaction time on n-hexane conversion.
This explanation was underlined by the selectivities for iso-hexanes (Figure 21). Selectivity
seemed to be just dependent on the hexane conversion and was not shifted by reaction
temperature change. However, the conversions differed extremely and therefore it was not
possible to compare selectivities at similar n-hexane conversions. The comparison of
experiments with varying reaction temperatures conducted under hydrogen (phydrogen = 10 bar),
shown in this chapter, and under helium atmosphere is illustrated in Figure 59 and Figure 60
(appendix section 7.1) to underline repeatedly the effect of hydrogen on conversion and
selectivity.
0 5 10 15 20 25 300
10
20
30
40
50
60
Xn-
hexa
ne /
%
modified reaction time / min molionic liquid
mol-1 n-hexane
Treaction
= 293 K
Treaction
= 303 K
Treaction
= 313 K
84 Results and discussion
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n(H2SO4) / n(AlCl3) = 0.18, phydrogen = 10 bar, ptotal = 15 bar, ptotal = phelium + phydrogen.
Figure 21: Influence of reaction temperature and n-hexane conversion on the selectivity for iso-hexanes.
The consecutive reaction character of the skeletal isomerization of n-hexane (Scheme 24) can
be followed by the course of selectivities over modified reaction time for the three
investigated reaction temperatures (Figure 22 and Figure 23). The selectivities of the isomers
2-methylpentane (2-MP), 3-methylpentane (3-MP), 2,2-dimethylbutane (2,2-DMB) and
2,3-dimethylbutane (2,3-DMB) add up to 100 %. Hexane was excluded for this illustration
because the reactant was available in excess. The concentration of 2,2-DMB, the last
compound in the consecutive reaction, increased significantly with longer reaction times for
Treaction = 313 K compared to 303 K and 293 K and the 2,3-DMB-fraction was lowered at
313 K as it reacts to 2,2-DMB. In conclusion, the higher the reaction temperature, the faster
was the consecutive reaction. However, the fraction of hexane isomers at 313 K equaled not
the thermodynamic equilibrium (Figure 1) because the reaction time at this catalyst / reactant
ratio was too short and furthermore cracking reactions occured.
0 10 20 30 40 50 600
10
20
30
40
50
60
70
80
90
100
S iso-
hexa
nes /
%
Xn-hexane
/ %
Treaction
= 293 K
Treaction
= 303 K
Treaction
= 313 K
Results and discussion 85
Scheme 24: Reaction network of the n-hexane isomerization [54] including 2-methylpentane (2-MP), 3-methylpentane (3-MP), 2,2-dimethylbutane (2,2-DMB) and 2,3-dimethylbutane (2,3-DMB).
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n(H2SO4) / n(AlCl3) = 0.18, phydrogen = 10 bar, ptotal = 15 bar, ptotal = phelium + phydrogen; the selectivities for 2-MP, 3-MP, 2,2-DMB and 2,3-DMB sum up to 100 %.
Figure 22: Influence of reaction temperature and modified reaction time on the selectivity for 2-methylpentane (2-MP) and 3-methylpentane (3-MP).
n-hexane
2-MP 3-MP
2,3-DMB 2,2-DMB
+H+
-H2
+H2
+H2
+H2
+H2
-H+
-H+
-H+
-H+
0 5 10 15 20 25 300
10
20
30
40
50
60
70
80
90
100
S hexa
ne-is
omer
/ %
modified reaction time / min molionic liquid
mol-1n-hexane
2 MP 3 MP (Treaction
= 293 K)
2 MP 3 MP (Treaction
= 303 K)
2 MP 3 MP (Treaction
= 313 K)
86 Results and discussion
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n(H2SO4) / n(AlCl3) = 0.18, phydrogen = 10 bar, ptotal = 15 bar, ptotal = phelium + phydrogen; the selectivities for 2-MP, 3-MP, 2,2-DMB and 2,3-DMB sum up to 100 %.
Figure 23: Influence of reaction temperature and modified reaction time on the selectivity for 2,2-dimethylbutane (2,2-DMB) and 2,3-dimethylbutane (2,3-DMB).
4.1.6 Influence of the ionic liquid´s Lewis acidity on the
hydroisomerization of n-octane
Results presented in chapter 4.1.4 revealed that hydrogen was ineffective to change the
catalyst´s selectivity for iso-octanes. Hence, another approach for the selectivity improvement
was the use of an ionic liquid with lower acidity. The molar ratio of the used ionic liquid was
lowered from n([BMIM]Cl) / n(AlCl3) = 1/2.0 to the molar ratio of 1/1.7. Lewis acidity and
the anionic species present in the chloroaluminate ionic liquid are a function of the mole
fraction of AlCl3 (Figure 2). Reduced Lewis acidity should decrease the overall acidity of the
catalytic system [BMIM]Cl / AlCl3 / H2SO4 and its ability to crack branched octanes to a
great extent even at low reaction temperatures (Treaction = 303 K). The experiments were
additionally conducted under hydrogen atmosphere (phydrogen = 15 bar).
0 5 10 15 20 25 300
5
10
15
20
25
30
35
40
S hexa
ne-is
omer
/ %
modified reaction time / min molionic liquid
mol-1 n-hexane
2,2 DMB 2,3 DMB (Treaction
= 293 K)
2,2 DMB 2,3 DMB (Treaction
= 303 K)
2,2 DMB 2,3 DMB (Treaction
= 313 K)
Results and discussion 87
Conditions: n(H2SO4) / n(AlCl3) = 0.18, Treaction = 303 K, ptotal = 15 bar, phydrogen = 15 bar.
Figure 24: Influence of the ionic liquid´s Lewis acidity and modified reaction time on n-octane conversion.
n-Octane hydroisomerization was influenced by the acidity change (Figure 24). The values of
conversions were almost parallel shifted to smaller ones for all modified reaction times if the
less Lewis acidic chloroaluminate ionic liquid was applied. The conversion decrease was not
unexpected because the Lewis acidity was lower for the molar ratio of
[BMIM]Cl / AlCl 3 = 1.7.
Though, the selectivity for iso-octanes remained unchanged (Figure 25). The selectivity was
just dependent on the converted moles of n-octane and was almost identical for both Lewis
acidic ionic liquids at similar conversions (e.g. Siso-octanes = 22.9 % at Xn-octane = 19.2 % for
n([BMIM]Cl) / n(AlCl 3) = 1/2.0; Siso-octanes = 22.8 % at Xn-octane = 18.9 % for
n([BMIM]Cl) / n(AlCl 3) = 1/1.7). In summary, the highly acidic protons available in the ionic
liquid n([BMIM]Cl) / n(AlCl 3) = 1.7 with H2SO4 could also crack the generated iso-octanes
and this, in the same manner, like the more Lewis acidic ionic liquid system. The product
spectrum was also equal for both catalysts (4 ≤ C ≤10) with a selectivity for
iso-alkanes > 90 %.
0 5 10 15 20 25 30 350
10
20
30
40
50
60
X n
-oct
ane /
%
modified reaction time / min mol ionic liquid
mol-1 n-octane
n([BMIM]Cl) / n(AlCl3) = 1/2.0
n([BMIM]Cl) / n(AlCl3) = 1/1.7
88 Results and discussion
Conditions: n(H2SO4) / n(AlCl3) = 0.18, Treaction = 303 K, ptotal = 15 bar, phydrogen = 15 bar.
Figure 25: Influence of ionic liquid acidity and n-octane conversion on the selectivity for iso-octanes.
In conclusion, chapter 4.1 focused on monofunctional acidic ionic liquid based catalysts for
the n-hexane and n-octane isomerization in a liquid-liquid biphasic reaction mode. Three
different catalytic systems, [BMIM]Cl / AlCl3 / H2SO4, [BMIM]Cl / AlCl 3 / CuCl2 and
[BMIM]Cl / AlCl 3 / 1-chlorooctane were screened for their ability to convert n-octane. All
catalytic systems showed high activity even at the applied mild reaction conditions.
[BMIM]Cl / AlCl 3 / H2SO4 and [BMIM]Cl / AlCl3 / 1-chlorooctane were optimized regarding
their molar composition to get the highest possible concentrations of active carbenium ions
and along with it highest conversion. Another main focus of this chapter was on the
selectivity for the respective iso-alkanes as cracking is a predominant side reaction in the
isomerization. The selectivities for iso-octanes and iso-hexanes, respectively, were not very
high because cracking of branched alkanes occurred. However, the products consisted almost
completely of branched alkanes in the range of 4 ≤ C ≤ 10 and are thus also desirable for
gasoline production provided that the RON of each component is sufficient. The addition of
hydrogen to the reaction mixture might give small selectivity improvements for iso-hexanes.
However, the presence of hydrogen was effectless for iso-octanes selectivity improvement.
Furthermore, the selectivity was dependent on n-alkane conversion for all tested catalysts.
This could be logically explained by mechanistic considerations regarding possible reaction
pathways of carbenium ions, the active species of the alkane isomerization. Additional
0 10 20 30 40 50 600
5
10
15
20
25
30
35
40
S iso-
octa
nes /
%
X n-octane
/ %
n([BMIM]Cl) / n(AlCl3) = 1/2.0
n([BMIM]Cl) / n(AlCl3) = 1/1.7
Results and discussion 89
approaches, lower reaction temperature in the hydroisomerization of n-hexane or reduced
Lewis acidity of the chloroaluminate ionic liquid in the hydroisomerization of n-octane
resulted not in increased selectivities for the respective iso-alkanes.
All following experiments were only conducted with reactant n-octane. This alkane possesses
increased attractiveness for further investigations. The long-chain alkane has not been applied
industrially in the isomerization up to now because of the pronounced cracking side reactions.
Therefore, alternative catalytic systems in contrast to the classical bifunctional heterogeneous
catalysts and processes need to be investigated to elucidate their potentials in the alkane
isomerization in general and especially of long-chain alkanes (C > 6).
4.2 Monofunctional SILP catalysts in a slurry-phase reaction
mode
SILP catalysts offer a possibility to minimize the drawbacks of a biphasic ionic
liquid / organic liquid reaction – difficult product separation and catalyst recycling as well as
a small phase boundary between ionic liquid and reactant - while maintaining the attractive
features of an ionic liquid. The obtained catalytic material appears macroscopically as a solid,
thus, it can be handled like a heterogeneous catalyst, whereas microscopically the
homogeneous character of the catalyst is preserved.
Another different aspect to mention is that an extremely low reaction temperature in the
alkane isomerization would be only justified if the selectivity for iso-alkanes was
outstandingly high. Otherwise, the low reaction temperature has negative effect on the
reaction rate according to Arrhenius law. Moreover, the isomerization of alkanes is part of oil
refineries that work at a temperature level higher than room temperature depending on each
processing step. For example, catalyst poisons like sulfur compounds are removed by
hydrogenation in the hydrotreating step which is operated at temperatures between 573 and
673 K and pressures ranging from 25 to 60 bar [12]. The hydrotreating is a necessary step
before refining that also includes isomerization. A time and cost intensive procedure would be
necessary to cool down the reaction mixture to a temperature as low as 303 K. Therefore,
experiments with the SILP catalyst and all following experimental runs were conducted at
higher reaction temperature.
To elucidate the effectiveness of ionic liquid immobilization experiments with the
monofunctional solid supported system (SILP catalyst) and monofunctional unsupported
90 Results and discussion
acidic ionic liquid catalyst were carried out and compared. All experiments were conducted
without addition of an alkyl halide. Besides, for both catalysts the effect of hydrogen versus
helium pressure was investigated. The same mass of Lewis acidic ionic liquid
[BMIM]Cl / AlCl 3 was used for the SILP preparation and for the ionic liquid catalyst.
Different conclusions can be drawn from the catalytic results (Figure 26 and Figure 27).
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 373 K; SILP: mSILP = 23 g, ε = 0.4, α = 0.85.
Figure 26: Comparison of n-octane conversion over modified reaction time for the SILP and acidic ionic liquid catalyst under hydrogen (phydrogen = 40 bar) and helium atmosphere (phelium = 40 bar), respectively.
Firstly, the immobilization of the ionic liquid on pretreated silica led to higher n-octane
conversions compared to the ionic liquid in a biphasic reaction mode, under helium as well as
hydrogen atmosphere. Secondly, the experiments conducted under hydrogen resulted in lower
conversions, especially at very low amounts of converted n-octane (Xn-octane < 10 %), that is
for the whole modified reaction time applying the ionic liquid catalyst and until
8.9 min molionic liquid mol-1n-octane for the SILP system. The concentration of carbenium ions
was very low for the used catalytic systems as additives like H2SO4 or 1-chlorooctane were
not present in the reaction mixture. Superacidic protons originated only from traces of water
in the ionic liquid. In addition, the isomerization and cracking reactions have consecutive
character. Some of the carbenium ions might be captured by hydrogen according to chapter
2.3.1.5. With rising monobranched octane isomer concentration, cracking drastically
increased which was obviously faster than the skeletal isomerization. In general, the
0 5 10 150
10
20
30
40
50
60
70
80
Xn-
octa
ne /
%
modified reaction time / min molionic liquid
mol-1n-octane
SILP (phelium
= 40 bar)
SILP (phydrogen
= 40 bar)
ionic liquid (phelium
= 40 bar)
ionic liquid (phydrogen
= 40 bar)
Results and discussion 91
selectivity-values decreased with higher conversions according to the mechanism of the
octane isomerization. All four experiments showed the same trend for selectivities and their
decrease. The high selectivities in the experiments can be attributed to the extremely low
n-octane conversions.
It can be noted that the pretreated silica system itself showed no catalytic activity in the
isomerization of n-octane (Treaction = 373 K, phelium = 40 bar, treaction = 4 h).
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 373 K; SILP: mSILP = 23 g, ε = 0.4, α = 0.85.
Figure 27: Comparison of selectivity for iso-octanes over n-octane conversion for the SILP and acidic ionic liquid catalyst under hydrogen (phydrogen = 40 bar) and helium atmosphere (phelium = 40 bar), respectively.
In summary, the SILP system offers an improved recycling and immobilization concept for
the catalytically active acidic ionic liquid. The ionic liquid n([BMIM]Cl) / n(AlCl3) = 1/2 was
successfully immobilized on pretreated silica support and showed a higher catalytic activity in
the n-octane conversion compared to the uncoated acidic ionic liquid.
4.3 Bifunctional SCILL catalyst for n-octane hydroisomerization
in a slurry-phase reaction mode
Monofunctional acidic ionic liquids are active and attractive isomerization catalysts because
they are able to convert even n-alkanes at mild reaction conditions. However, the catalyst´s
selectivity could not be improved for the monofunctional ionic liquid based catalysts by
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
S iso-
octa
nes /
%
Xn-octane
/ %
SILP (phelium
= 40 bar)
SILP (phydrogen
= 40 bar)
ionic liquid (phelium
= 40 bar)
ionic liquid (phydrogen
= 40 bar)
92 Results and discussion
variation of hydrogen partial pressure, reaction temperature or ionic liquid acidity. For further
catalytic improvements the noble metal Pt, which can hydrogenate alkenes and dehydrogenate
alkanes, was added to the ionic liquid catalyst in form of Pt / silica to synthesize systems with
bifunctional properties. A SCILL system was synthesized by coating the pretreated Pt / silica
with Lewis acidic chloroaluminate ionic liquid. Acidity was provided by the ionic liquid and
Pt offered the hydrogenation function which is important in the hydroisomerization of
alkanes.
It could be proved that the monofunctional heterogeneous Pt / silica catalyst, one component
of the SCILL system, can not convert n-octane. As expected, the untreated and uncoated
Pt / silica catalyst (ε = 0, α = 0) showed no catalytic acitivity for n-octane conversion under
hydrogen atmosphere (Treaction = 393 K, phydrogen = 40 bar, mPt / silica = 10.4 g,
n(Cl-octane) / n(AlCl3) = 0.36, mn-octane = 55 g, treaction = 4 h). Due to the lack of any
sufficiently acidic site in this system, no carbenium ions are generated and no alkane
isomerization was observed.
The influence of hydrogen partial pressure, reaction temperature, ionic liquid Lewis acidity as
well as 1-chlorooctane addition on the catalytic activity and selectivity were investigated for
the SCILL system and are presented. Complete product distribution of the SCILL catalyzed
experiment and its comparison with ionic liquid catalyzed runs is also part of this chapter.
Further, the stability of the SCILL system and its recyclability were part of the studies.
Firstly, the characterization of the SCILL material is described. Reproducibility of the data
with SCILL catalysts was proven and the results are shown in appendix 7.2.
4.3.1 Characterization of the SCILL material
The hydrogen uptake and dispersion of Pt on the porous silica were determined by
H2-chemisorption on the dried Pt / silica and its reduced counterpart (Table 16). The
dispersion of the Pt was found to be in a similar range for both, the dried and reduced
Pt / silica. It could be concluded from this result that the commercial Pt / silica was already in
its reduced form. Therefore, the material was applied without further reduction step prior to
SCILL preparation.
Results and discussion 93
Table 16: Hydrogen uptakes of Pt / silica catalysts at T = 273 K (for 0.57 wt. % Pt).
Material Pretreatment H2 uptake / µl g-1Pt Dispersion / %
Pt / silica Dried 125 37
Pt / silica Reduced 105 32
The ICP-AES analysis (Pt, Si) of the Pt / silica resulted in a mean Pt value of 0.566 wt. %
which was used for all further calculations (Table 17). The Pt / Si ratio calculation is
important in the context of investigations to the catalyst stability (see chapter 4.3.7).
Table 17: ICP-AES (Pt, Si) results of the heterogeneous Pt / silica catalyst.
Pt / Silica (sample) Pt / wt. % Pt / Si ratio / mol mol-1
1 0.521 1.71·10-3
2 0.592 1.66·10-3
3 0.586 1.90·10-3
Mean value 0.566 1.76·10-3
As expected, both, the specific surface area ABET and the pore volume Vpore decreased if the
Pt / silica was pretreated according to the method described by Joni et al. [206] (see Table 18,
Figure 28, Figure 29).
Table 18: Specific surface area ABET and pore volume VPore for the Pt / silica and pretreated Pt / silica.
Material ABET / m2 g-1 Vpore / ml g-1
Pt / silica 236 1.086
Pretreated Pt / silica 203 0.878
The isotherms of these materials (Figure 28) can be assigned to type IV isotherms according
to IUPAC [249]. The amount of micropores is negligible. Only mesopores are contributing to
94 Results and discussion
the pore volume in the range of 6 – 15 nm. The results of the normalized cumulative surface
areas Apore (dp) / Apore,total of the two materials are shown in Figure 30. The curve of the
pretreated silica was shifted to smaller pore diameters. The decrease of the pore diameter was
about 1.8 nm which equals the layer covalently bond to the surface hydroxyl groups of the
silica. A narrow distribution of the pores can be concluded from the normalized cumulative
surface area (Figure 30), the cumulative pore volume (Figure 29) and the adsorption
isotherms (Figure 28). The applied porous support was suitable for the immobilization of
ionic liquid because micropores would be filled anyway by the ionic liquid (maximum size
[Al 2Cl7]- = ca. 0.69 nm).
N2-adsorption measurements were also performed for the SCILL system (α = 0.85, ε = 0.5).
The measured data are depicted in Figure 31. Of course, the pore volume and specific surface
area were reduced for the SCILL system (ABET = 30 m2 g-1, Vpore = 0.135 ml g-1) compared to
the values of Pt / silica and pretreated Pt / silica (Table 18). However, the results of the
normalized cumulative surface areas Apore (dp) / Apore,total (Figure 32) revealed that the curve of
the SCILL material was not shifted to smaller pore diameters compared to the pretreated
Pt / silica in spite of the ionic liquid loading. Moreover, the hysteresis of all three materials
was in the same range of partial pressures. These results indicate a constant pore diameter for
the materials, even at the high pore filling degree (α = 0.85). Therefore, pore closing and
inclusion of gas by the ionic liquid during heating procedure before N2-adsorption was likely,
especially at the high ionic liquid loading (α = 0.85). Measured data of the SCILL system
corresponded probably not to the real system.
Results and discussion 95
Figure 28: N2-adsorption isotherms of the Pt / silica and the pretreated Pt / silica.
Figure 29: Cumulative pore volume calculated with BJH method (desorption) of the Pt / silica and the pretreated Pt / silica.
0.0 0.2 0.4 0.6 0.8 1.00
100
200
300
400
500
600
700
800
Vol
ume
N2 /
ml S
TP g
-1
Relative pressure p p-10 / -
Pt / silica Pt / silica pretreated
2 4 6 8 10 12 14 16 18 20 22 24 26
0.00.10.20.30.40.50.60.70.80.91.01.11.2
Cum
ulat
ive
pore
vol
ume
/ m
l g-1
Pore diameter / nm
Pt / silica Pt / silica pretreated
96 Results and discussion
Figure 30: Cumulative normalized surface area Apore (dp) / Apore,total calculated with BJH method (desorption) of the Pt / silica and the pretreated Pt / silica.
Figure 31: N2-adsorption isotherms of the Pt / silica, pretreated Pt / silica and SCILL system (α = 0.85).
4 6 8 10 12 14 16 18 20 22 24 26
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Cum
ulat
ive
surf
ace
area
Apo
re(d
p) /A
pore
,tota
l / -
Pore diameter / nm
Pt / silica Pt / silica pretreated
0.0 0.2 0.4 0.6 0.8 1.00
100
200
300
400
500
600
700
800
Vol
ume
N2 /
ml S
TP g
-1
Relative pressure p p-10 / -
Pt / silica Pt / silica pretreated SCILL ( αααα = 0.85)
Results and discussion 97
Figure 32: Cumulative normalized surface area Apore (dp) / Apore,total calculated with BJH method (desorption) of the Pt / silica, pretreated Pt / silica and SCILL system (α = 0.85).
4.3.2 Influence of hydrogen partial pressure on activity and selectivity of
the bifunctional SCILL catalyst
The influence of hydrogen partial pressure (phydrogen = 0, 15 and 40 bar) on activity and
selectivity of the bifunctional SCILL catalyst was investigated in the hydroisomerization of
n-octane (T = 393 K, n(Cl-octane) / n(AlCl3) = 0.036). The results for the n-octane conversion
are depicted in Figure 33. They revealed a dependency of the catalytic activity on the
hydrogen partial pressure. Higher hydrogen partial pressures resulted in a significant increase
in n-octane conversion and iso-octane selectivity. At a hydrogen partial pressure of 40 bar the
n-octane conversion was 73.7 % after maximum applied modified reaction time, while in the
absence of hydrogen the conversion was less than 20 % under otherwise identical conditions.
Comparing these results with the outcome of the reactions with the same acidic ionic liquid in
liquid-liquid biphasic mode (see chapter 4.1.4, Figure 16 and Figure 17,
n(Cl-octane) / n(AlCl3) = 0.036), it is evident that the bifunctional SCILL system provided a
much more active and selective catalyst system although the same mass of ionic liquid was
used for both experiments. A rise in total pressure from 15 to 40 bar helium in the absence of
hydrogen showed no significant effect on n-octane conversion.
4 6 8 10 12 14 16 18 20 22 24 26
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Cum
ulat
ive
surf
ace
area
Apo
re(d
p) /A
pore
,tota
l / -
Pore diameter / nm
Pt / silica Pt / silica pretreated SCILL ( αααα = 0.85)
98 Results and discussion
In spite of equal masses of the acidic ionic liquid [BMIM]Cl / AlCl 3, the experiment with the
SCILL catalyst under helium atmosphere (see Figure 33) resulted in higher conversion
(Xn-octane = 16. 6 % at 10.3 min molionic liquid mol-1n-octane) in comparison to the ionic liquid
catalyst without hydrogen (Xn-octane = 6.9 % at 13.4 min molionic liquid mol-1n-octane) (see
Figure 16, n(Cl-octane) / n(AlCl3) = 0.036). In both experiments, only the acidic ionic liquid
was the catalytically active compound because the reactions were run under helium.
Therefore, higher activity can be referred to the immobilization of ionic liquid on porous
silica in case of the SCILL systems. The immobilization of ionic liquid leads to a larger phase
boundary between the ionic liquid and alkane phase. This effect was also proven in
experiments with SILP and uncoated ionic liquid catalyst without any promoter (chapter 4.2,
Figure 26).
Conditions: mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 393 K, n(Cl-octane) / n(AlCl3) = 0.036.
Figure 33: Influence of the hydrogen partial pressure and modified reaction time on n-octane conversion.
Apart from the enhanced activity, the almost parallel shift of the iso-octanes selectivity in the
Siso-octane / Xn-octane-plot with rising hydrogen partial pressures is obvious (Figure 34). At the
same level of n-octane conversion, a hydrogen partial pressure of 15 bar increased the
selectivity for iso-octanes by an absolute value of 10 % compared to the reaction in the
absence of hydrogen. Comparing the reaction at a hydrogen partial pressure of 40 bar with the
hydrogen-free system, this difference increased to an absolute value of up to 15 %. Like in the
0 5 10 150
10
20
30
40
50
60
70
80
90
100
Xn-
octa
ne /
%
modified reaction time / min molionic liquid
mol-1n-octane
phydrogen
= 0 bar, ptotal
= 15 bar
phydrogen
= 0 bar, ptotal
= 40 bar
phydrogen
= 15 bar
phydrogen
= 40 bar
Results and discussion 99
case of n-octane conversion, the effect of total pressure on the selectivity was checked by
pressurizing the system to the same values with helium (15 and 40 bar) but in this case the
selectivity effect was negligible.
Conditions: mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 393 K, n(Cl-octane) / n(AlCl3) = 0.036.
Figure 34: Influence of the hydrogen partial pressure and n-octane conversion on the selectivity for iso-octanes.
The selectivities of the SCILL experiment with n(Cl-octane) / n(AlCl3) = 0.036 (Figure 34)
can only be compared with those of the ionic liquid catalyst using
n(Cl-octane) / n(AlCl3) = 0.36 (chapter 4.1.4, Figure 17) because similar conversion values
like in the SCILL runs have to be reached to make the selectivities compareable. In contrast to
the enhancing effect of hydrogen on the selectivities for iso-octanes applying the bifunctional
SCILL system, the increase of hydrogen partial pressure was effectless for the
monofunctional acidic ionic liquid (Figure 17).
Closing, due to the lack of ethane in the cracked products of the SCILL catalyzed experiment,
any hydrocracking mechanism on the noble metal (hydrogenolysis) could be excluded as
potential side reaction [89]. Further, it can be stated that the applied molar ratio of
hydrogen / n-octane was about 0.8 and thereby low compared to recent catalytic
investigations concerning heterogeneously catalyzed hydroisomerization of n-octane where
the molar ratios ranged from 6 – 30 (see the literature review given in chapter 2.6.1).
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
Sis
o-oc
tane
s / %
phydrogen
= 0 bar, ptotal
= 15 bar
phydrogen
= 0 bar, ptotal
= 40 bar
phydrogen
= 15 bar
phydrogen
= 40 bar
Xn-octane
/ %
100 Results and discussion
4.3.3 Product distribution for the bifunctional SCILL cat alyst and its
comparison with results of the acidic ionic liquid in liquid-liquid
biphasic catalysis
Up to now, conversion and selectivity for the respective iso-alkanes were key aspects of the
discussion. In this part, the whole product distribution gained with a SCILL catalyst under
hydrogen atmosphere is presented and compared with the products obtained using the acidic
ionic liquid in biphasic catalysis. Closing, the RONs of the products for the SCILL and ionic
liquid catalyzed experiments were calculated and are presented in this chapter.
Figure 35 represents the composition of the octane isomers as a function of n-octane
conversion for the slurry-phase experiment using the bifunctional SCILL catalyst and the
biphasic experiment using the monofunctional Lewis acidic ionic liquid (see also Figure 34
and Figure 17, phydrogen = 40 bar). 1-Chlorooctane was added in both experiments, with the
molar ratio n(Cl-octane) / n(AlCl3) = 0.036 for the SCILL experiment and a ten-time higher
ratio for the liquid-liquid biphasic runs to reach similar conversions because otherwise the
comparison of product selectivities would be meaningless.
Both catalyst systems yielded only monobranched and dibranched octane isomers and this in
very similar distributions. The absence of even higher branched octane isomers is due to their
high cracking rate. When comparing the composition of the alkane products with a carbon
chain length other than 8, differences between the bifunctional SCILL catalyst and the liquid-
liquid catalysis with the acidic ionic liquid can be observed.
In Figure 36, the selectivities for products of different chain lengths are given as selectivity
for the respective carbon fraction. For the SCILL catalyst, the fraction of propane and butanes
within the by-products increased while the fraction of pentanes, hexanes and heptanes
decreased in contrast to the monofunctional Lewis acidic ionic liquid catalyst. The fraction of
the by-products with a chain length higher than 8 was unaffected. The differences in the
distributions of the by-products can be explained by the hydrogenation activity of the
Pt-containing catalyst. While the shorter chain by-products (C3 – C4) are mainly formed by
β-scission, the formation of by-products with a higher chain length requires a combination of
cracking and alkylation reactions. The alkylation reactions are dependent on the existence of
olefines which are formed by cracking reactions via β-scission or by deprotonation of
carbenium ions. In the presence of Pt and hydrogen, a significant part of the olefins is
hydrogenated. With decreasing concentration of olefins the alkylation rate decreases, leading
Results and discussion 101
to a lower fraction of C5 – C7 products. For both, the monofunctional and the bifunctional
catalyst, the formed by-products consisted mainly of branched alkanes (around 90 %
iso-alkanes and 10 % n-alkanes).
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 393 K, phydrogen = 40 bar; SCILL catalyst: mSCILL = 20 g, ε = 0.5, α = 0.85, n(Cl-octane) / n(AlCl3) = 0.036; ionic liquid catalyst: mionic liquid = 10 g, n(Cl-octane) / n(AlCl3) = 0.36.
Figure 35: Composition of the formed branched octane isomers as a function of n-octane conversion for the monofunctional ionic liquid catalyst and for the bifunctional SCILL system.
0 10 20 30 40 50 60 70 800
10
20
30
40
50
60
70
80
90
100
SCILL catalyst ionic liquid catalyst monobranched C8 monobranched C8
dibranched C8 dibranched C8
S iso-
octa
nes /
%
Xn-octane
/ %
102 Results and discussion
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 393 K, phydrogen = 40 bar; SCILL catalyst: mSCILL = 20 g, ε = 0.5, α = 0.85, n(Cl-octane) / n(AlCl3) = 0.036; ionic liquid catalyst: mionic liquid = 10 g, n(Cl-octane) / n(AlCl3) = 0.36.
Figure 36: Composition of by-products (selectivity for the carbon fraction) as a function of n-octane conversion for the monofunctional ionic liquid and the bifunctional SCILL catalyst; the fractions contain n- and iso-alkanes; the sum of all fractions adds up to 100 %.
The RONs were calculated according to Equation 16 for the two experiments already
discussed in this chapter, where υi is the volume fraction of molecule i in the sample.
RON = υ ii
RON i (16)
The used RONs for the respective alkanes are listed in appendix 7.4 (Table 24). The densities
necessary for the υi calculations are taken from literature [250] [251].
Two different cases were calculated:
a) Product mixture without C3, C4 and n-octane
b) Product mixture without C3, C4 and n-alkanes.
Butanes are the lowest boiling paraffins practicable to incorporate in gasoline. However, in
practice only n-butane is of value because iso-butane is too volatile and required in large
quantities for alkylation reactions [252]. Branched and linear alkanes can be seperated easily
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
50
SCILLcatalyst C3
C4
C5
C6
C7
≥≥≥≥ C9
ionic liquidcatalyst C3
C4
C5
C6
C7
≥≥≥≥ C9
S by-p
rodu
cts /
%
Xn-octane
/ %
Results and discussion 103
by molecular sieves which only allow the straight-chain paraffins to pass through. n-Octane
was excluded from the RON calculation as the reactant was available in excess after reaction
and thus would falsify the RON.
Figure 37 shows the calculated RONs obtained in SCILL and ionic liquid catalyzed
experiment.
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 393 K, phydrogen = 40 bar; SCILL catalyst: mSCILL = 20 g, ε = 0.5, α = 0.85, n(Cl-octane) / n(AlCl3) = 0.036; ionic liquid catalyst: mionic liquid = 10 g, n(Cl-octane) / n(AlCl3) = 0.36.
Figure 37: Calculated RON of the product mixture of the SCILL and ionic liquid catalyzed experiments over the modified time.
The RON approached for the SCILL and ionic liquid catalyst beginning at
tmod. = 6.1 min molionic liquid mol-1n-octane. Up to that, the ionic liquid catalyst resulted in slightly
higher RON values compared to the experiment with the bifunctional catalyst. The selectivity
for iso-octanes was higher for the SCILL catalyst. However, these iso-octanes consisted
solely of mono- and dibranched octanes which have a lower RON compared to cracking
products like iso-pentane or branched hexanes because the RON decreases for higher chain
lengths and increases with the degree of branching. Thus, the RON of both experiments
approached with decreasing iso-octanes fraction in the SCILL experiment and higher n-octane
conversion. For both, the SCILL and ionic liquid catalyst, the calculated RONs were higher
for the case if all n-alkanes were excluded. This was not suprising but led only to small
differences as most products were branched.
0 5 10 150
10
20
30
40
50
60
70
80
90
100
SCILL: without C3, C
4 and n-octane
SCILL: without C3, C
4 and n-alkane
ionic liquid: without C3, C
4 and n-octane
ionic liquid: without C3, C
4 and n-alkane
RO
N /
-
modified reaction time / min mol ionic liquid
mol-1 n-octane
104 Results and discussion
Though, the calculated RON gained with a SCILL or ionic liquid catalyst is still below the
minimum values for gasoline (see Table 2). Nonetheless, gasoline is a mixture of different
hydrocarbon types beside alkanes (Table 1). Alternative routes for the synthesis of high
RON-exhibiting components are for example the alkylation of isobutane with 2-butene or the
dimerization of isobutene with following hydrogenation.
4.3.4 Influence of the reaction temperature on activity and selectivity
To investigate the influence of the reaction temperature on the activity and the selectivity of
the SCILL catalyst, hydroisomerization experiments were carried out between 373 K and
423 K. Figure 38 shows the conversion of n-octane as a function of the modified reaction
time. As expected from Arrhenius law, the reaction rate of the SCILL-catalyzed reaction and
thus the conversion of n-octane over time increased with higher reaction temperature. The
selectivity for iso-octanes was found to be nearly identical for all three temperatures at
comparable n-octane conversions (see Figure 39). This indicated that, at least within the
considered temperature range, the selectivity for iso-octanes was independent of the applied
reaction temperature.
Conditions: n(Cl octane) / n(AlCl3) = 0.036, mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, phydrogen = 40 bar.
Figure 38: Influence of reaction temperature and modified reaction time on n-octane conversion.
0 5 10 150
10
20
30
40
50
60
70
80
90
100
Treaction
= 373 K
Treaction
= 393 K
Treaction
= 423 K
X n
-oct
ane /
%
modified reaction time / min mol ionic liquid
mol-1n-octane
Results and discussion 105
Conditions: n(Cl-octane) / n(AlCl3) = 0.036, mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, phydrogen = 40 bar.
Figure 39: Influence of reaction temperature and n-octane conversion on selectivity for iso-octanes.
4.3.5 Influence of the ionic liquid´s acidity
Varying acidity of the chloroaluminate ionic liquid [BMIM]Cl / AlCl 3 was already part of the
studies with the ionic liquid catalyst in a liquid-liquid biphasic reaction mode (see chapter
4.1.6). Higher Lewis acidity of the ionic liquid (n([BMIM]Cl) / n(AlCl 3) = 1/2) caused higher
n-octane conversion but almost unchanged selectivity for the branched octanes if the ionic
liquid catalyst was applied. Different results in comparison to the monofunctional ionic liquid
catalyst were expected for the SCILL system because of the bifunctional catalyst character
and the possible interaction between the Lewis acidic ionic liquid
(n([BMIM]Cl) / n(AlCl 3) > 1/1) and Pt.
Therefore, varying Lewis acidity was also tested for the SCILL catalyst under hydrogen
atmosphere (phydrogen = 40 bar). The same mass of ionic liquid was used for both SCILL
preparations. One has to take into consideration when evaluating the obtained data that
n([BMIM]Cl) / n(AlCl 3) = 1/2.0 exhibits a higher densitiy [253] and thus the SCILL system
exhibits a lower pore filling degree α compared to the SCILL with
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
Treaction
= 373 K
Treaction
= 393 K
Treaction
= 423 KS is
o-oc
tane
s / %
Xn-octane
/ %
106 Results and discussion
n([BMIM]Cl) / n(AlCl 3) = 1/1.7. The Lewis acidity depends on the molar ratio
[cation]Cl / AlCl3 (Figure 2).
Figure 40 revealed a significant change in conversion (about 20 %). The effect on the
selectivity was not as pronounced as for the conversion. However, a slight shift to higher
iso-octanes selectivities could be observed (Figure 41) when the more Lewis acidic ionic
liquid was loaded on the pretreated Pt / silica.
The results can be explained by the possible hydrogen activation in the presence of both, Pt
and the Lewis acidic ionic liquid. The presence of hydrogen causes a more active catalytic
system (Figure 33). It is propable that the hydrogen molecule could be better activated if Pt
was surrounded by a more Lewis acidic environment. Further explanation to the hydrogen
acitivation using a SCILL catalyst follows in chapter 4.5.
The molar ratio n([EMIM]Cl) / n(AlCl3) = 1/2 corresponds to the maximum solubility of
AlCl 3 in the chloroaluminate mixture to get an ionic liquid that is liquid at room temperature
[117]. Thus, further experiments with a more acidic chloroaluminate ionic liquid in the SCILL
system were not conductable.
Results and discussion 107
Conditions: mSCILL = 20 g, ε = 0.5, α = 0.85 for n([BMIM]Cl) / n(AlCl3) = 1/2.0, α = 0.86 for n([BMIM]Cl) / n(AlCl 3) = 1/1.7, n(Cl-octane) / n(AlCl3) = 0.036, phydrogen = 40 bar, Treaction = 393 K.
Figure 40: Influence of the ionic liquid´s Lewis acidity and modified reaction time on SCILL catalyzed n-octane conversion.
Conditions: mSCILL = 20 g, ε = 0.5, α = 0.85 (n([BMIM]Cl) / n(AlCl3) = 1/2.0), α = 0.86 (n([BMIM]Cl) / n(AlCl 3) = 1/1.7), n(Cl-octane) / n(AlCl3) = 0.036, phydrogen = 40 bar, Treaction = 393 K.
Figure 41: Influence of the ionic liquid´s Lewis acidity and n-octane conversion on the selectivity for iso-octanes in SCILL catalyzed experiments.
0 5 10 150
10
20
30
40
50
60
70
80
90
100
Xn-
octa
ne /
%
modified reaction time / min molionic liquid
mol -1n-octane
n([BMIM]Cl) / n(AlCl3) = 1/2.0
n([BMIM]Cl) / n(AlCl3) = 1/1.7
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
Sis
o-oc
tane
s / %
n([BMIM]Cl) / n(AlCl3) = 1/2.0
n([BMIM]Cl) / n(AlCl3) = 1/1.7
Xn-octane
/ %
108 Results and discussion
4.3.6 Varying amounts of 1-chlorooctane and its influence on activity and
selectivity
The molar ratio 1-chlorooctane / AlCl3 showed a significant influence on the n-octane
conversion for the catalytic system with [BMIM]Cl / AlCl3 and 1-chlorooctane (Figure 13) in
a liquid-liquid biphasic reaction mode. Therefore, the influence of four different molar ratios
1-chlorooctane / AlCl3 (0.0, 0.017, 0.036, 0.36) was also tested for the SCILL catalyzed
hydroisomerization of n-octane (phydrogen = 15 bar). The effect of alkyl halide addition on the
n-octane conversion as a function of modified reaction times (Figure 42) was by far not as
pronounced as in the liquid-liquid biphasic reaction system. The curves of the four
experiments differed but showed similar conversion for longer reaction time, even the
experiment without any alkyl halide addition (Xn-octane for n(Cl-octane) / n(AlCl3); 45.5 % for
0.0; 55.7 % for 0.017; 51.3 % for 0.036; 46.0 % for 0.36). No clear trend was obvious that
higher concentrations of 1-chlorooctane led to increased n-octane conversion. The iso-octane
selectivity exhibited also no recognizable changes and trends if the alkyl halide was added to
the reaction mixture (Figure 43).
The unequal influence of an alkyl halide in the presence of the ionic liquid catalyst in contrast
to the influence of 1-chlorooctane for the SCILL system gave already a hint that the reactions
were catalyzed with different mechanisms. Moreover, it is supposable that the effect of
carbenium ion generation by the addition of an alkyl halide to a Lewis acidic ionic liquid was
superposed by a mechanism that produces also carbenium ions in a larger extent when the
SCILL system was applied. Therefore, addition of an alkyl halide is needless if the alkane
conversion is SCILL catalyzed. The addition of any alkyl halide was omitted for further
experiments.
Results and discussion 109
Conditions: mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 393 K, phydrogen = 15 bar.
Figure 42: Influence of the molar ratio 1-chlorooctane / AlCl3 (n(Cl-octane) / n(AlCl3)) and modified reaction time on n-octane conversion.
Conditions: mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 393 K, phydrogen = 15 bar.
Figure 43: Influence of the molar ratio 1-chlorooctane / AlCl3 (n(Cl-octane) / n(AlCl3)) and n-octane conversion on selectivity for iso-octanes.
0 5 10 150
10
20
30
40
50
60
70
80
90
100
Xn-
octa
ne /
%
modified reaction time / min molionic liquid
mol-1n-octane
n(Cl-octane) / n(AlCl3) = 0
n(Cl-octane) / n(AlCl3) = 0.017
n(Cl-octane) / n(AlCl3) = 0.036
n(Cl-octane) / n(AlCl3) = 0.36
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
Sis
o-oc
tane
s / %
n(Cl-octane) / n(AlCl3) = 0
n(Cl-octane) / n(AlCl3) = 0.017
n(Cl-octane) / n(AlCl3) = 0.036
n(Cl-octane) / n(AlCl3) = 0.36
Xn-octane
/ %
110 Results and discussion
4.3.7 Catalyst stability
An important aspect for further development of these bifunctional SCILL systems towards
technical application is the stability of the catalyst. ICP-AES analyses of the product proved
that no Pt and Al leached into the organic phase (detection limits are given in Table 19). As
no Al leaching was observed, it can be concluded that no Lewis acidic ionic liquid was lost to
the product phase during reaction. This indicates that the very unpolar feedstock / product
phase was not able to dissolve the ionic liquid off the support. Moreover, the results revealed
that the ionic liquid physical adsorption on the support was strong enough to avoid removal of
ionic liquid droplets from the support.
Table 19: ICP-AES (Pt, Al) results of the organic and removed ionic liquid phase.
Element Analyzed phase Leaching / wt. % Detection limit / wt. %
(referred to total Pt, Al) (referred to total Pt, Al)
Pt Organic - 0.03
Al Organic - 0.03
Pt Ionic liquid 0.21 0.03
In addition, the Pt leaching into the immobilized acidic ionic liquid was analyzed. The
question whether the Pt stayed on the support (the bifunctional SCILL concept) or dissolved
in the acidic ionic liquid during reaction (leading to a bifunctional SILP concept) is of
conceptual importance. In the SCILL case, the support material is of direct relevance for the
nature of the catalytic Pt centre, while in the SILP case only the embedding ionic liquid
defines the reactivity of the dissolved Pt-complexes. The Pt leaching into the ionic liquid layer
was checked in two independent ways: a) the acidic ionic liquid was removed after reaction
from the support by a washing procedure with DCM and the washing solution was checked
for its Pt-content by ICP-AES; b) the Pt / Si ratio of the applied Pt / silica catalyst was
determined by ICP-AES prior to the ionic liquid coating as well as after catalytic reaction and
removal of the ionic liquid. The results are given in Table 20. The deviation of the Pt / Si ratio
of SCILL catalyst and Pt / silica (entry 1, 2 and entry 1, 3, respectively) was higher than the
standard deviation of the Pt / silica mean value calculation (Table 17 and Table 20). This fact
indicates a small amount of Pt leaching from the silica into the Lewis acidic ionic liquid
Results and discussion 111
which was proved by the ICP-AES results of the washed-off acidic ionic liquid. The Pt
leaching (0.21 wt. % of the total Pt) could take place during pretreatment or during catalysis.
However, by far the largest part of the Pt remained on the support making the whole catalyst a
true bifunctional SCILL system.
Table 20: ICP-AES (Pt, Si, Al) results of the applied Pt / silica material, SCILL catalyst before reaction (ε = 0.5, α = 0.85), SCILL catalyst after reaction and removal of acidic ionic liquid.
Entry Sample Pt / Si - ratio / mol mol-1
1 Pt / silica (mean value)
Standard deviation (for mean value calculation)
1.76·10-3
1.03·10-4
2 SCILL catalyst, before reaction 1.36·10-3
3 SCILL catalyst, ionic liquid washed off 1.52·10-3
To exclude the theoretical possibility of chlorine insertion from the chloroaluminate ionic
liquid into the organic products, the organic phase (experiments without 1-chlorooctane
addition) was analyzed by ESI-MS. The chlorine concentration was found to be below the
detection limit as no characteristic isotope peaks of chlorine compounds were found in the
mass spectrum (appendix chapter 7.3).
Typical heterogeneous, state-of-the-art isomerization / cracking catalysts operate at much
higher temperatures than the SCILL systems applied in this work. The deposition of heavy
products on the catalytic surface and coking is a major deactivation path in these high
temperature systems. To demonstrate the advantage of the SCILL systems operating at mild
temperatures in this respect, the ionic liquid was tested after reaction for any accumulation of
high boiling, unsaturated or cyclic byproducts. The analysis was carried out for a SCILL
system operated under hydrogen pressure (phydrogen = 40 bar) and for a SCILL system which
was applied under helium atmosphere (phelium = 40 bar) in the n-octane conversion. The
removed ionic liquid from the SCILL catalyst (washing procedure with DCM) was contacted
with an excess water to hydrolyze the ionic liquid anion. The resulting aqueous solution was
extracted with cyclohexane and the organic extraction phases were investigated with GC-MS
to analyze all organic products accumulated in the ionic liquid phase during reaction. For the
SCILL catalyst that was applied under hydrogen atmosphere, the highest m/z ratio found in
112 Results and discussion
the mass spectrum in this analysis was 164 which might originate from cyclic unsaturated
hydrocarbons. However, this mass and the respective size of the molecule were still small
enough to expect this species to be in extraction-equilibrium with the organic product phase.
Thus, accumulation of carbon-rich species in the SCILL system could be excluded under the
conditions applied in this study. The GC-MS analysis of the ionic liquid phase which was
washed off from the SCILL catalysts operated under helium resulted in a highest m/z ratio in
the mass spectrum of 282. Some cyclic and / or unsaturated compounds might be
hydrogenated if the SCILL system was operated in the presence of hydrogen, thus a lower
maximum m/z ratio was found in the mass spectrum for the hydroisomerization experiment.
Catalyst stability including coke formation under extended operation or in a continuous
reaction mode has not been investigated up to now.
The highly acidic catalysts applied in this work are able to convert n-alkanes. Thus, it is
imaginable that the butyl side chain of the [BMIM]-cation in the ionic liquid was transformed
during reaction. The ionic liquid layer was washed off the SCILL catalyst (5 times) after use
in catalysis using DCM that was separated afterwards under UHV conditions. Complete
dissolution of the ionic liquid phase after reaction in a solvent was not possible. All oxygen
containing solvents would react with the oxophilic chloroaluminate ionic liquid
[BMIM]Cl / AlCl 3. CDCl3 resulted in the formation of two phases. The lower, CDCl3-rich
phase was analyzed by 1H-NMR. The upper phase showed solid formation when it was
contacted with DMSO. The interesting cutout of the 1H spectrogram is shown in Figure 44.
The integrals of the two peaks belonging to the methyl group (4) and the methylene group (3)
exhibited the ratio 3 : 2 which proved that the butyl side chain of the [BMIM]+ cation was still
intact after reaction. Peaks of the methylene groups (3), (5), (6) as well as (7) were visible in
the spectrum. However, quantitative analysis of these peaks was not possible because product
alkanes were still dissolved in the ionic liquid phase and interfered with these peaks.
Results and discussion 113
Figure 44: Cutout of the 1H-NMR spectrum of the ionic liquid phase, which is soluble in CDCl3, from the SCILL system after use in catalysis.
4.3.8 Catalyst recycling
The recycling of the catalyst is an issue of practical relevance. The results discussed in
chapter 4.3.7 already demonstrated the stability of SCILL catalysts as no Al and Pt could be
detected in the organic product phase after reaction. Consequently, the recyclability of the
SCILL catalysts under hydrogen atmosphere was tested at 373 K in two recycling runs
without the addition of any alkyl halide (see Figure 45 and Figure 46). For the recycling
experiments, most of the organic phase was drawn off with a syringe from the SCILL catalyst
after settling time at ambient reaction condition and replaced with new n-octane. A small
amount of organic left in the glass liner acted as moisture protective layer for the SCILL
catalyst during loading of n-octane. It was unavoidable that small amounts of the solid SCILL
catalyst stuck to the stirrer and cooling coil and were therefore shortly exposed to air while
the organic phase was changed in the glass liner.
4 3 2 1 0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
abun
danc
e / -
1H-NMR / parts per million
(3)
Area:
2.02
(4)
Area:
3.00
Inte
nsity
/ -
114 Results and discussion
Conditions: mSCILL = 12 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 373 K, phydrogen = 40 bar.
Figure 45: Comparison of n-octane conversion over modified reaction time for three catalytic runs with the same SCILL catalyst.
Conversion of n-octane decreased slightly with every catalyst recycling from 9.6 % (recycling
run 0) to 8.4 % (recycling run 1) and finally 6.6 % (recycling run 2) after
tmod. = 1.64 min molionic liquid mol-1n-octane. In contrast, the curve of iso-octanes selectivity over
n-octane conversion was almost identical for all three catalytic runs (Figure 46). These results
showed the recyclability of the SCILL catalyst. Losses in catalytic activity can be ascribed to
the deactivated SCILL catalyst which was in contact with air.
In contrast to the recycling experiments with the monofunctional ionic liquid catalyst (see
Table 14 and Table 15), the conversion decrease turned out to be lower for the SCILL catalyst
(run 0 � run 1: ∆Xn-octane = 12.5 %; run 1 � run 2: ∆Xn-octane = 21.4 %) whereupon the
deactivated SCILL catalyst which stuck at the stirrer also accounted for the conversion
decrease. This might be explained by a different reaction mechanism for the SCILL catalyst.
No or only small amounts of HCl, which originate only from traces of water in the
chloroaluminate ionic liquid, might exist in the SCILL experiment which could be lost via the
gas-phase during loading with new n-octane.
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.00
2
4
6
8
10
12
14
X n
-oct
ane /
%
modified reaction time / min mol ionic liquid
mol-1n-octane
recycling run 0 recycling run 1 recycling run 2
Results and discussion 115
Conditions: mSCILL = 12 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 373 K, phydrogen = 40 bar.
Figure 46: Comparison of selectivity for iso-octanes as function of n-octane conversion for three catalytic runs with the same SCILL catalyst.
4.4 In situ formed Pt nanoparticles in acidic ionic liquids for
n-octane hydroisomerization
The bifunctional SCILL system and its promosing performance in the hydroisomerization of
n-octane were investigated in detail. The combination of metal Pt and Lewis acidic ionic
liquid can also be achieved by the addition of metal precursors to the ionic liquid and in situ
reduction of the metal ions under hydrogen atmosphere. The extension of the bifunctional
catalyst concept offers the oppurtunity to understand the interactions of ionic liquid and metal
nanoparticles and to investigate differences in the catalytic performance if the metal is added
to the ionic liquid which is afterwards immobilized on a porous support (analogy to SILP
concept) or if the metal is immobilized on a porous support and coated with ionic liquid
(SCILL concept). Besides, systematic variation of the Pt concentration would be possible in a
simple way if metal precursors are applied. In further consequence, the immobilized Lewis
acidic ionic liquid and metal nanoparticles could be applied in a slurry-phase reaction mode.
Thereby, simple product separation and catalyst recycling would be guaranteed.
0 2 4 6 8 10 12 140
5
10
15
20
25
30
35
40
45
S iso-
octa
nes /
%
X n-octane
/ %
recycling run 0 recycling run 1 recycling run 2
116 Results and discussion
This thesis includes the first steps of these ideas mentioned above. In the beginning, two
different metal precursors, PtCl2 and PtCl4, were tested in combination with the ionic liquid
n([BMIM]Cl) / n(AlCl 3) = 1/2. Moreover, the influence of ionic liquid acidity was elucidated
by the use of bromoaluminate systems in addition to the chloroaluminate ionic liquids. The
catalytic reactions were multiphase systems in which the metal particles dispersed in the ionic
liquid formed the denser phase and the organic remained in the upper phase.
4.4.1 Activity and selectivity of the catalysts derived from PtCl2 or PtCl4
in the Lewis acidic ionic liquid [BMIM]Cl / AlCl 3
The experiments with n([BMIM]Cl) / n(AlCl3) = 1/2 and PtCl2 or PtCl4 were conducted at
373 K and without alkyl halide addition. The influence of hydrogen was elucidated for the
PtCl2 system. Mainly two conclusions can be obtained from the n-octane conversions
(Figure 47).
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 373 K, ptotal = 40 bar, nplatinum = 0.3 mmol.
Figure 47: Influence of the precursor type (PtCl2 or PtCl4), hydrogen and modified reaction time on n-octane conversion.
Firstly, very similar results were achieved for PtCl2 or PtCl4 in the Lewis acidic ionic liquid.
Secondly, hydrogen addition to the PtCl2 system led to a conversion increase
(∆Xn-octane = 6.6 % at tmod. = 11.5 min molionic liquid mol-1n-octane). Similar conclusions can be
derived from the iso-octane selectivity curve (Figure 48). The PtCl2 and PtCl4 containing
0 5 10 150
5
10
15
20
25
30
Xn-
octa
ne /
%
modified reaction time / min molionic liquid
mol-1n-octane
ionic liquid and PtCl2 (p
hydrogen = 40 bar)
ionic liquid and PtCl2 (p
helium = 40 bar)
ionic liquid and PtCl4 (p
hydrogen = 40 bar)
Results and discussion 117
catalysts showed the same behavior in the experimental run under hydrogen. Direct
comparison of the helium and hydrogen run was not possible because the n-octane
conversions were not in the same range. Though, a steeper selectivity decrease with
increasing conversion seems to exist for the helium experiment compared to the catalytic
systems under hydrogen atmosphere.
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 373 K, ptotal = 40 bar, nplatinum = 0.3 mmol.
Figure 48: Influence of the precursor type (PtCl2 or PtCl4), hydrogen and n-octane conversion on the selectivity for iso-octanes.
Pt0 particles were formed in situ, but of course only under the reductive hydrogen atmosphere.
Black particles were found in the ionic liquid phase after catalysis for both experiments with
hydrogen. This was not the case if helium was applied. XRD analysis revealed that no PtCl2
or PtCl4 was left after catalysis and Pt0 was formed (XRD diffractogram (reflexes): Figure 49;
original XRD diffractogram: Figure 68, appendix 7.5). The remaining peaks in the
diffractogram of the sample corresponded very well to NaCl which might originate from the
hydrolysis of the ionic liquid´s anion with ethanol and sodium hydrogen carbonate after
reaction. The dimension L of the Pt particles, which was calculated according to the Scherrer-
formula (Equation 13), accounts for 41 nm (at 2θ = 39.94 °). Pretty large Pt particles were
formed in the applied in situ process.
0 5 10 15 20 25 300
5
10
15
20
25
30
35
40
45
S iso-
octa
nes /
%
Xn-octane
/ %
ionic liquid and PtCl2 (p
hydrogen = 40 bar)
ionic liquid and PtCl2 (p
helium = 40 bar)
ionic liquid and PtCl4 (p
hydrogen = 40 bar)
118 Results and discussion
Additionally, it can be noted that PtCl2 showed no catalytic activity in the n-octane
hydroisomerization without the acidic ionic liquid (Treaction = 373 K, phydrogen = 40 bar,
nplatinum = 0.3 mmol (PtCl2), mn-octane = 55 g, treaction = 4 h).
Figure 49: X-ray diffractogram (reflexes) of the analyzed sample and reference diffractograms of Pt and NaCl; sample: particles separated from the ionic liquid after catalysis (n([BMIM]Cl) / n(AlCl 3) = 1/2, nplatinum = 0.3 mmol (PtCl4), Treaction = 373 K, ptotal = 40 bar).
4.4.2 Influence of the ionic liquid´s acidity in combination with the metal
precursor PtCl2
It is known from literature that the Hammett acidity of bromoaluminate systems is even
higher than the acidity of chloroaluminate systems ([EMIM][Al 2Br7] / HBr: H0 = -17,
[EMIM][Al 2Cl7] / HCl: H0 = -15) [137]. Therefore, one experiment was run with the
bromoaluminate ionic liquid n([BMIM]Br) / n(AlBr3) = 1/2 and PtCl2 under hydrogen
atmosphere (phydrogen = 40 bar). Equal molar amounts of the acidic ionic liquids were applied
Position [°2Theta] (Kupfer (Cu))
20 30 40 50 60 70
Sample
Pt
NaCl
Inte
nsity
/ -
Inte
nsity
/ -
Inte
nsity
/ -
2 θ / °
Results and discussion 119
for the chloro- and bromoaluminate system. The higher acidity in the bromoaluminate system
resulted in a significant increase of converted n-octane molecules compared to the
chloroaluminate ionic liquid (∆Xn-octane = 16.1 %, tmod. = 11.5 min molionic liquid mol-1n-octane)
(Figure 50). The only selectivity data at similar conversion (Xn-octane = 10.4 % for
[BMIM]Cl / AlCl 3 and 11.1 % for [BMIM]Br / AlBr3) differed (∆Siso-octanes = 18.2 %).
Though, the extremely high acidity of the [BMIM]Br / AlBr3-system resulted in low
iso-octane selectivities, between 20.2 and 18.3 % (Figure 51). However, the selectivity
decrease was lower compared to the chloroaluminate system. Therefore, it is worth to
investigate the bromoaluminate based catalysts in further experiments, especially to get
selectivity values at higher n-octane conversions.
Conditions: n([BMIM]X) / n(AlX3) = 1/2, Treaction = 373 K, phydrogen = 40 bar, nplatinum = 0.3 mmol.
Figure 50: Influence of the ionic liquid and modified reaction time on n-octane conversion.
It is only reasonable to benefit from the higher acidity of the ionic liquid and thus from the
higher octane conversion if the selectivity for iso-octanes can be improved. This was not the
case for the bromoaluminate ionic liquid based catalyst. Immobilization of the ionic liqud on a
porous support and the resulting thin film of ionic liquid might change the diffusion influence
and the availability of hydrogen at the active metal Pt in contrast to experiments in a
liquid-liquid biphasic experiment and with it the selectivity.
0 5 10 150
10
20
30
40
50
X n
-oct
ane /
%
modified reaction time / min mol ionic liquid
mol-1 n-octane
[BMIM]Cl / AlCl3 and PtCl
2
[BMIM]Br / AlBr3and PtCl
2
120 Results and discussion
Conditions: n([BMIM]X) / n(AlX3) = 1/2, Treaction = 373 K, phydrogen = 40 bar, nplatinum = 0.3 mmol.
Figure 51: Influence of the ionic liquid and n-octane conversion on the selectivity for iso-octanes.
4.5 Comparison of the bifunctional and monofunctional catalytic
systems in the n-octane isomerization
Further results of the two bifunctional systems - SCILL and in situ formed Pt nanoparticles in
acidic ionic liquid - and of the two monofunctional catalysts - SILP catalyst and uncoated
acidic ionic liquid - are presented and compared in this chapter. Further, the differences in the
catalytic results and reaction pathways are elucidated. First part of this chapter presents the
silica and Pt / silica characterization.
4.5.1 Characterization of silica gel 100 and comparison with Pt / silica
For both, the SCILL and SILP system, silica was taken as porous support and pretreated
according to Joni et al. [206]. However, these silicas were not exactly identical materials
because the Pt / silica catalyst was purchased and not synthesized starting with porous silica.
To demonstrate the similarity of both silica supports and to guarantee comparable catalytic
results, N2-adsorption measurements of the silica and pretreated silica were conducted. The
N2-adsorption isotherms of the calcined silica, pretreated silica, as well as of the commercial
0 10 20 30 40 500
5
10
15
20
25
30
35
40
45
S iso-
octa
nes /
%
X n-octane
/ %
[BMIM]Cl / AlCl3 and PtCl
2
[BMIM]Br / AlBr3 and PtCl
2
Results and discussion 121
Pt / silica and pretreated Pt / silica are depicted in Figure 52. Pore volume Vpore and specific
surface area ABET of all four materials are listed in Table 21.
Figure 52: N2-adsorption isotherms of the commercial Pt / silica, pretreated Pt / silica, calcined silica and pretrated silica.
Table 21: Specific surface area ABET and pore volume Vpore for Pt / silica and silica 100.
Material ABET / m2 g-1 Vpore / ml g-1
Pt / silica (commercial) 236 1.086
Pt / silica (pretreated) 203 0.878
Silica 100 (calcined) 333 0.982
Silica 100 (pretreated) 274 0.744
Total pore volume Vpore of the Pt / silica was higher than that of silica 100. 81 % of the initial
pore volume remained after the pretreatment step for Pt / silica and 76 % for the silica 100.
The specific surface area resulted in higher values for both, the calcined silica 100 and
pretreated silica 100 compared to the Pt containing support. This could be attributed to a
higher fraction of micropores. The mean pore diameter of silica 100 is 10.5 nm and 14.9 nm
for Pt / silica. But, micropores (< 2 nm) are filled by the ionic liquid (maximum size [Al2Cl7]-
0.0 0.2 0.4 0.6 0.8 1.00
100
200
300
400
500
600
700
800
V
olum
e N
2 / m
l ST
P g-1
Relative pressure p p -1 0 / -
Pt / silica Pt / silica pretreated silica calcined silica pretreated
122 Results and discussion = ca. 0.69 nm) anyway, particularly if the silica is pretreated and the ionic liquid exceeds one
monolayer. The SILP and SCILL catalysts under investigation were loaded with the same
mass of Lewis acidic ionic liquid after the support pretreament and exhibited the same pore
filling degree α. Thus, the ionic liquid loading ε differed. The catalytic results of the SCILL,
Pt nanoparticles in acidic ionic liquid, SILP and unsupported acidic ionic liquid catalysts are
described and discussed in the following section.
4.5.2 Catalytic results of the four different catalytic systems
This chapter comprises results of four different catalytic systems without 1-chlorooctane or
H2SO4 addition at 373 K as the mechanism of the bifunctional sytems can be best understood
in direct and systematic comparison with the results of SILP and unsupported acidic ionic
liquids catalyzed n-octane isomerzation. Equal masses of n([BMIM]Cl) / n(AlCl3) = 1/2 were
used for the catalyst syntheses. Further, the molar amounts of Pt in the SCILL sytem and the
Pt added in form of PtCl2 were the same. For all catalysts, the effect of hydrogen versus
helium pressure was investigated.
To further elucidate the performance of the bifunctional SCILL system under hydrogen
pressure and to underline the necessity of all three components of the catalytic system – Pt,
Lewis acidic chloroaluminate ionic liquid and hydrogen - experiments with a Pt-free,
monofunctional solid supported system (SILP catalyst) and monofunctional unsupported
acidic ionic liquid catalyst were carried out. The catalytic results showed very similar curves
for the SILP catalyst under hydrogen atmosphere and for the SCILL catalyst under helium
(Figure 53 and Figure 54). This behavior suggests that the bifunctional SCILL catalyst - with
Pt as differentiating factor - acted like the monofunctional SILP catalyst in the absence of
hydrogen. Thus, the presence of Pt only became relevant if the reaction was carried out under
hydrogen pressure. Only the presence of hydrogen, Pt on support and acidic ionic liquid led to
a significant increase in catalytic activity.
Results and discussion 123
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 373 K; SILP: mSILP = 23 g, ε = 0.4, α = 0.85; SCILL: mSCILL = 20 g, ε = 0.5, α = 0.85.
Figure 53: Comparison of n-octane conversion over modified reaction time for the SILP catalyst under hydrogen atmosphere (phydrogen = 40 bar) versus the SCILL catalysts under helium (phelium = 40 bar) and hydrogen atmosphere (phydrogen = 40 bar).
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 373 K; SILP: mSILP = 23 g, ε = 0.4, α = 0.85; SCILL: mSCILL = 20 g, ε = 0.5, α = 0.85.
Figure 54: Comparison of selectivity for iso-octanes as function of n-octane conversion for the SILP catalyst under hydrogen atmosphere (phydrogen = 40 bar) versus the SCILL catalysts under helium (phelium = 40 bar) and hydrogen atmosphere (phydrogen = 40 bar).
0 5 10 150
10
20
30
40
50
60
70
80
Xn-
octa
ne /
%
modified reaction time / min molionic liquid
mol-1n-octane
SILP (phydrogen
= 40 bar)
SCILL (phelium
= 40 bar)
SCILL (phydrogen
= 40 bar)
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
S iso-
octa
nes /
%
Xn-octane
/ %
SILP (phydrogen
= 40 bar)
SCILL (phelium
= 40 bar)
SCILL (phydrogen
= 40 bar)
124 Results and discussion
The fact that the SCILL system showed significantly higher n-octane conversion and
selectivity for iso-octanes in the presence of hydrogen in comparison to the SCILL catalyst
under helium atmosphere further confirmed the importance of hydrogen activation by the
catalytic Pt-centre for the observed reactivity. According to the mechanism of isomerization,
firstly the n-octane molecule is converted to an unbranched octanyl cation by hydride
abstraction through a highly acidic proton or by another carbenium ion. The linear octanyl
cation undergoes isomerization to form a branched octanyl cation. The latter can react via
hydride transfer with any alkane of the reaction mixture to yield the corresponding branched
octane isomer and a new carbenium ion. However, besides the desired hydride transfer, the
branched octanyl cation can also undergo undesired side reactions. These include cracking
reactions via β-scission or alkylation reactions of carbenium ions with certain amounts of
olefins that form e.g. by deprotonation of carbenium ions (Scheme 25) [66] or cracking [61].
Both reactions decrease the selectivity for branched iso-octanes significantly.
Scheme 25: Equilibrium reaction of a carbenium ion and an olefin in acidic media [66].
In the presence of hydrogen, Lewis acidic ionic liquid and active Pt centres, most of the
olefins formed by deprotonation are likely to be hydrogenated prior to the undesired
alkylation reaction. This hydrogenation activity explains the higher selectivity for iso-octanes.
The role of hydrogenation reactions in the bifunctional SCILL system can also be recognized
by monitoring the total reactor pressure during the different isomerization experiments
(Figure 55). The hydrogen pressure of the experiment conducted with the SCILL catalyst
showed a notable continuous decrease, while the pressure in the two comparison runs –
monofunctional SILP under hydrogen and bifunctional SCILL under helium – remained
almost constant (the little jags in the graphs are due to reactor sampling). Only in the presence
of Pt and hydrogen, the hydrogenation of olefins that are formed via β-scission of carbenium
ions or by carbenium ion deprotonation took place. Only the hydrogenation of alkenes
originiating from β-scission results in steady hydrogen consumption throughout the reaction
time. No net hydrogen consumption should be observed if olefins are hydrogenated that are
formed by deprotonation of a carbenium ion, as the released acidic proton will typically react
with an alkane to yield a new carbenium ion and hydrogen molecule (Scheme 26).
Results and discussion 125
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 373 K; SILP: mSILP = 23 g, ε = 0.4, α = 0.85; SCILL: mSCILL = 20 g, ε = 0.5, α = 0.85.
Figure 55: Reactor pressure over modified reaction time for SILP catalyst under hydrogen atmosphere (phydrogen = 40 bar) versus SCILL catalysts under helium (phelium = 40 bar) and hydrogen atmosphere (phydrogen = 40 bar).
Scheme 26: Proposed reaction mechanism of the SCILL catalyzed n-alkane isomerization (RH = alkane, H+ = proton, R+ = carbenium ion, R= = alkene).
The hydrogenation activity gave a good explanation for the improved selectivity of the
bifunctional SCILL system. Different possibilities for the higher activity could be taken into
account. Firstly, the fact that the catalytic activity correlated also strongly to the level of
hydrogen partial pressure (Figure 33) suggests that the concentration of acidic protons and
0 5 10 1515
20
25
30
35
40
45
50
55
60
p reac
tor /
bar
modified reaction time / min molionic liquid
mol-1n-octane
SILP (hydrogen) SCILL (helium) SCILL (hydrogen)
126 Results and discussion
thus the acidity of the catalytic system increases if hydrogen dissolved in the Lewis acidic
ionic liquid is in contact with the active Pt centre. It is known that hydrogen can homolytically
split on Pt [254]. The catalytic result is in-line with the interpretation that in the strongly
Lewis acidic environment of the chloroaluminate melt, electron transfer may lead to the
formation of Pt-hydride and additional protons in the ionic liquid. This interpretation is
supported by the fact that 1-chlorooctane added to the bifunctional SCILL system caused no
additional enhancement of the catalytic activity (Figure 42). Apparently, the amount of
carbenium ions formed from hydrogen at the catalytic Pt-centre is dominating the catalytic
activity of the system. Previous publication by Lee and Song showed also the synergetic
effect of Pd / C with the Lewis acidic ionic liquid n([BMIM]Cl) / n(AlCl 3) = 1/2 and the
activation of hydrogen in the hydrogenation of arenes [255]. Pd / C, Pd nanoparticles and
solely the Lewis acidic ionic liquid n([BMIM]Cl) / n(AlCl3) = 1/2 were ineffective catalysts
for the hydrogenation of benzene and other hydrocarbon-based aromatics. They proposed an
ionic hydrogenation mechanism in the presence of the Lewis acidic ionic liquid and Pd / C. A
further possibility for the activity increase is described in the following. If a carbenium ion
undergoes a hydride transfer reaction with an alkene, an alkane and an unsaturated carbenium
ion is generated, which can be considered as protonated diene. This unsaturated compound is
much more basic and binds a proton much more strongly than monoalkenes. It is well-known
from traditional heterogeneous bifunctional isomerization catalysts that the alkene
concentration can be kept low by saturation with hydrogen avoiding the formation of di- or
polyunsaturated compounds [24]. At least, it can be concluded that the obtained catalytic
results indicate a gradual increase of the system´s acidity with increasing hydrogen pressure.
The next part of this chapter comprises the results of the two bifunctional catalysts and the
uncoated ionic liquid.
Figure 56 depicts the n-octane conversion of three different catalytic systems: the bifunctional
SCILL system, the bifunctional catalyst consisting of ionic liquid
n([BMIM]Cl) / n(AlCl 3) = 1/2 and PtCl2 and finally the pure chloroaluminate ionic liquid.
The results and observations are discussed step by step. Firstly, the activity of the bifunctional
catalyst ionic liquid / PtCl2 exhibited higher catalytic activity if the catalytic run was
conducted under hydrogen. Same reasons as already discussed for the bifunctional SCILL
system may account for this behavior. Secondly, the catalyst ionic liquid / PtCl2 showed
slightly decreased activity compared to the ionic liquid alone if the experiments were
conducted under helium. This fact can be explained with an acidity decrease in the
Results and discussion 127
chloroaluminate ionic liquid when PtCl2 was dispersed. The molar ratio AlCl3 / Cl- amounted
to 70.9 (the Cl--ions originate from PtCl2 dissolved in the chloroaluminate ionic liquid), that is
1.4·10-2 mol of AlCl3 are converted to the Lewis neutral anions [AlCl4]-. Thus, the increase in
octane conversion for the PtCl2 / ionic liquid catalyst under hydrogen can only be evaluated in
direct comparison with the PtCl2 / ionic liquid system under helium atmosphere and not with
the PtCl2-free catalyst. Thirdly, n-octane conversions were slightly higher, if the pure acidic
ionic liquid was used under helium. Very low conversions and thus concentration of
carbenium ions may explain this fact. Possibly, hydrogen supresses the branching step and
therefore possible cracking side reactions of branched octane isomers.
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 373 K, ptotal = 40 bar; SCILL: mSCILL = 20 g, ε = 0.5, α = 0.85; ionic liquid: mionic liquid = 10 g; ionic liquid and PtCl2: mionic liquid = 10 g, nplatinum = 0.3 mmol.
Figure 56: Influence of the catalyst system, hydrogen and modified reaction time on n-octane conversion.
When a comparison was drawn between the two bifunctional systems – SCILL and
PtCl2 / ionic liquid – operated under hydrogen, it is evident that the n-octane conversion
varied extremely (∆Xn-octane = 32.6 % at tmod. = 11.5 min molionic liquid mol-1n-octane). Different
reasons can account for this result. Firstly, in the SCILL system, the Pt was already in its
reduced form whereby the Pt nanoparticles in the ionic liquid were reduced in situ in the
catalytic system under reaction conditions. Secondly, the Pt particles in the ionic liquid might
tend to agglomerate. Ionic liquids have a beneficial effect on nanoparticle stabilization
0 5 10 150
10
20
30
40
50
60
70
80
Xn-
octa
ne /
%
modified reaction time / min molionic liquid
mol-1n-octane
SCILL phelium
= 40 bar phydrogen
= 40 bar
ionic liquid and PtCl2
phelium
= 40 bar phydrogen
= 40 bar
ionic liquid phelium
= 40 bar phydrogen
= 40 bar
128 Results and discussion
however, the addition of further stabilizer might be necessary. High dispersion of Pt during
catalysis is more supposable for the Pt in the SCILL system. Further, it is a difference if the Pt
particles are only embedded in the ionic liquid or if there is an additional Pt – support
interaction as it is the case for the SCILL catalysts. Finally, the ionic liquid was immobilized
on the porous Pt / silica in the SCILL system which led to a large phase boundary between the
ionic liquid and organic phase and a thin film of ionic liquid that shortens the diffusion
pathway of hydrogen to the Pt species. The PtCl2 was dispersed in the ionic liquid and the
phase boundary between ionic liquid and organic, which has to be generated by intensive
stirring, is lower in contrast to the SCILL system.
A simpler conclusion can be drawn from the selectivity curve which is shown in Figure 57 for
all six experiments in dependency on the n-octane conversion. The three experiments
conducted with different catalytic systems resulted in a steep selectivity decrease if helium
was added. In contrary, if a Pt component and hydrogen was applied, the selectivity decline
was decreased. Nevertheless, highest concentrations for iso-octanes were the benefit if the
SCILL system was applied.
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, Treaction = 373 K, ptotal = 40 bar; SCILL: mSCILL = 20 g, ε = 0.5, α = 0.85; ionic liquid: mionic liquid = 10 g; ionic liquid and PtCl2: mionic liquid = 10 g, nplatinum = 0.3 mmol.
Figure 57: Influence of the catalyst system, hydrogen and n-octane conversion on the selectivity for iso-octanes.
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
S iso-
octa
nes /
%
Xn-octane
/ %
SCILL phelium
= 40 bar phydrogen
= 40 bar
ionic liquid and PtCl2
phelium
= 40 bar phydrogen
= 40 bar
ionic liquid phelium
= 40 bar phydrogen
= 40 bar
Results and discussion 129
To conclude, a positive effect on n-octane conversion and on the selectivity for iso-octanes
was observed when the bifunctional catalyst, PtCl2 or PtCl4 in combination with a highly
Lewis acidic ionic liquid, and hydrogen was applied. An additional reaction pathway, the
hydrogenation of formed iso-alkenes, is also likely for these bifunctional systems. Moreover,
a higher concentration of protons in the acidic ionic liquid, like in the SCILL case, might
occur which could explain the enhanced activity compared to the helium experiments.
Further investigations are necessary to overcome the drawbacks like missing immobilization
of the ionic liquid phase with dispersed metal particles. Besides, it was more efficient to
prepare Pt0 nanoparticles before the reaction to avoid a possible induction period until the
active metal catalyst is formed. Furthermore, the interaction of Pt particles, ionic liquid and
solid support has to be understood in detail.
Closing, no evidence of the Pt-hydride species could be provided which would have exceeded
the scope of this work. But, the experimental results which were explained in detail in this
chapter implicate the possibility of Pt-hydride species formation and the activation of
hydrogen in the presence of the bifunctional catalyst.
4.5.3 Reaction network
A possible reaction network for the monofunctional and bifunctional catalyzed n-alkane
isomerization is depicted in Figure 58. The dashed lines correspond to the additional reaction
pathways for the bifunctional catalyst, which contains Pt, under hydrogen atmosphere. The
dehydrogenation of alkanes was excluded as this reaction is impropable at the relatively mild
reaction temperatures applied for the experiments. Alkylation and cracking are considered as
possible side reactions. Oligomerization of alkenes was not taken into account because the
concentration of alkenes is low and the reaction was carried out in a stirred autoclave which
implies fast backmixing of the produced alkenes.
In the following, the reaction mechanism is described. Firstly, the n-octane molecule is
converted to an unbranched octanyl cation by hydride abstraction through a highly acidic
proton or by another carbenium ion. Once the carbenium ion has been produced it can
undergo an alkylation reaction with an alkene, generated by cracking reactions or out of a
carbenium ion by deprotonation.
130 Results and discussion
Figure 58: Possible reaction network of the n-alkane isomerization and side reactions (alkylation, cracking) at relatively mild reaction conditions (RH = alkane, R+ = carbenium ion, H+ = proton, R= = alkene, alkylate+ = reaction of an alkene with a carbenium ion to yield a new carbenium ion with higher chain length).
The carbenium can also be deprotonated to yield an alkene. In turn, the alkene can further
react with a carbenium ion to yield an alkylated carbenium ion. Only in the presence of Pt and
hydrogen, alkenes can be hydrogenated. Desired is the skeletal rearrangement of the
carbenium ions to give (highly) branched carbocations. Besides, these branched carbocations
are thermodynamically favored compared to the linear ones. The same reaction pathways like
described above are possible if the iso-carbenium ion is the starting point. In addition,
isomerized carbenium ions can undergo cracking reactions which leads to an alkene and a
carbenium ion. Cracking of linear alkanes is extremely unlikely (see chapter 2.3.1.4.1). Even
in the absence of Pt, alkenes can further be converted. They can react with carbenium ions.
Moreover, alkenes are by far more reactive than alkanes. The hydrogenation step on the Pt
catalyst opens another pathway for the reaction with an alkene but not a new one. However,
the hydrogenation of isomerized carbeniums ions on Pt can explain the higher iso-octanes
selectivity because further reactions of the iso-carbenium ions are minimized and the
hydrogenation of the iso-alkene might be very fast in presence of the bifunctional catalyst.
n-RH n-R iso-R iso-RH
n-R=iso-R=
alkylate
alkylate
alkylate cracked products (alkene + R )
alkylate
H2
H2
PtPt
H ; -H2
H2; -H
RH; -R
R ; -RH
H ; -H2
H2; -H
RH; -R
R ; -RH
H- H
R -R -RR
R=- R= R=- R=
H- H
Results and discussion 131
The increased activity in case of bifunctional catalyzed hydroisomerization of n-octane could
origin from a higher proton concentration present in the reaction mixture (see chapter 4.5.2)
which could generate a higher concentration of carbenium ions out of the feedstock n-alkane.
4.6 Final evaluation
Aim of this thesis was the investigation and optimization of highly acidic ionic liquid based
catalysts for n-alkane isomerization at moderate reaction temperatures. All applied catalysts in
this work, the acidic ionic liquid, SCILL system and in situ formed Pt nanoparticles in a
Lewis acidic ionic liquid, were active in the n-octane conversion below 393 K. Classical
heterogeneous catalysts can only be operated at higher reaction temperatures compared to the
ionic liquid based systems which has a negative impact on the thermodynamic equilbrium
within the isomers as high fractions of branched isomers should be produced. However, the
advantage of a low reaction temperature regarding the isomers equilibrium becomes only
important if cracking reactions can be minimized. Otherwise, the equilibrium between the
isomers is never reached and especially the multi-branched high RON contributing
components are lost. The selectivities for iso-octanes of the ionic liquid based catalysts used
in this work were dependent on the n-octane conversion. Mostly, they were lower as the
selectivies for the respective iso-alkanes of heterogeneous catalyzed heptane or octane
isomerizations, e.g. of zeolitic or sulfated zirconia based bifunctional catalysts (see chapter
2.6.1). Though, the selectivites for dibranched isomers within the iso-octanes, which exhibit
higher RONs compared to the monobranched isomers, were similar for the bifunctional
heterogeneous and ionic liquid based catalyts (see chapter 4.3.3). The fraction of desirable
tribranched octane isomers was insignificant for both catalyst types as these highly branched
alkanes are very pronounced for cracking side reactions.
In addition, the selectivity for iso-alkanes of the bifunctional SCILL system could be further
improved if a continuous reaction mode and shorter reactions times would be applied
compared to the batch runs in a stirred tank reactor. It has to be elucidated in further studies if
highly acidic ionic liquid based systems are potential catalysts for the isomerization of alkanes
and with it for the fuel-upgrade in a refinery, also regarding their robustness as halometallate
ionic liquids are irreversibly destroyed by water. and thus make high demands on the
pretreatment of the feedstock (feed dryer) and on the material of the reactor as corrosion
occurs.
132 Results and discussion
Independent of the selectivity, activity and stability of ionic liquid based catalysts in the
alkane isomerization, the beneficial interaction of supported Pt, acidic ionic liquid and
hydrogen was studied in various experiments and elucidated in detail for the bifunctional
SCILL systems. These catalysts offer the potential to be investigated in the field of further
acidic catalyzed reactions or for the hydrogenation of unsaturated compounds.
133
___________________________________________________________________________
5. Summary / Abstract
134 Summary / Abstract
Summary / Abstract 5
The aim of this thesis comprised the development of highly acidic ionic liquid based catalysts
for the n-alkane isomerization. Focus was on the reactant n-octane, a model compound for
long-chain alkanes (C > 6), which are more pronounced to cracking side reactions compared
to the state-of-the art feedstocks pentane / hexane and therefore has not been industrially
applied up to now for the up-grading of fuels and production of high RON gasoline. However,
this will be necessary in near future due to stricter environmental regulations.
The investigated catalysts can be basically categorized in two different catalytic systems, the
monofunctional and bifunctional catalysts. In the monofunctional category, uncoated acidic
ionic liquids and SILP catalysts were investigated. SCILL systems, which are synthesized by
coating pretreated Pt / silica with a thin film of the acidic ionic liquid [BMIM]Cl / AlCl3, as
well as Pt nanoparticles based systems, which were formed in situ under hydrogen by the
addition of Pt-precursors to a Lewis acidic hallometallate ionic liquid, can be attributed to the
bifunctional catalysts. Table 22 provides an overview of the investigated reaction systems.
The obtained results for the two different catalyst categories are presented in the following
and Table 23 summarizes the influence of important reaction parameters and the results of
catalyst recycling experiments.
Monofunctional acidic ionic liquid based catalysts
Three different catalytic systems, [BMIM]Cl / AlCl3 / H2SO4,
[BMIM]Cl / AlCl 3 / 1-chlorooctane as well as [BMIM]Cl / AlCl3 / CuCl2 were screened for
system selection in the n-octane isomerization at Treaction = 303 K. The first two systems were
further optimized for highest carbenium ion generation and therewith highest n-octane
conversion. Optimum molar ratio was found to be n([BMIM]Cl) / n(AlCl 3) = 1/2 with
n(H2SO4) / n(AlCl3) = 0.18 and n(1-chlorooctane) / n(AlCl3) = 0.5, respectively. Moreover,
these two optimized systems were tested for their recyclability. Both catalytic sytems could be
recycled, though with reduced catalytic activity in the second or third run. n-Octane
conversion decrease could be referred to the loss of HCl during reactor loading for the
recyling run or to decreased Lewis acidity of the ionic liquid in the case of repeated
1-chlorooctane addition in the recycling experiment.
Summary / Abstract 135
Table 22: Overview of the reaction systems studied in this thesis.
The cracking inhibitor hydrogen (phydrogen = 0 – 40 bar) was totally effectless in the n-octane
isomerization at mild reaction temperatures (catalyst: [BMIM]Cl / AlCl 3 / H2SO4,
Treaction = 303 K) and also at elevated temperatures (catalyst:
[BMIM]Cl / AlCl 3 / 1-chlorooctane, Treaction = 393 K).
The influence of hydrogen (phydrogen = 0 – 15 bar) was also tested for the reactant n-hexane
(catalyst: [BMIM]Cl / AlCl3 / H2SO4, Treaction = 303 K), which is not as pronounced to
cracking side reactions as n-octane. Hydrogen influenced the reaction rate negatively. The
selectivities might only be slightly shifted to higher values with increasing hydrogen partial
Monofunctional catalysts Bifunctional catalysts
� Acidic ionic liquid n([BMIM]Cl) / n(AlCl 3) = 1/1.7;1/2.0 n(H2SO4) / n(AlCl3) = 0 – 0.75 n(1-chlorooctane) / n(AlCl3) = 0 – 0.6 n(CuCl2) / n(AlCl3)= 0 – 0.06 Treaction = 283 – 393 K ptotal = 3 – 40 bar phydrogen = 0 - 40 bar n-hexane and n-octane liquid-liquid biphasic
� SCILL
n([BMIM]Cl) / n(AlCl 3) = 1/1.7;1/2.0 coated on pretreated Pt / silica α = 0.85 ε = 0.5 n(1-chlorooctane) / n(AlCl3) = 0 – 0.36 Treaction = 373 – 423 K ptotal = 15 - 40 bar phydrogen = 0 - 40 bar n-octane slurry-phase
� SILP n([BMIM]Cl) / n(AlCl 3) = 1/2.0 coated on pretreated silica α = 0.85 ε = 0.4 Treaction = 373 K ptotal = 40 bar phydrogen = 0 - 40 bar n-octane slurry-phase
� Pt nanoparticles in an acidic
ionic liquid n([BMIM]Cl) / n(AlCl 3) = 1/2.0; n([BMIM]Br) / n(AlBr 3) = 1/2.0 PtCl2 or PtCl4
n(Pt) = 0.3 mmol Treaction = 373 K ptotal = 0 - 40 bar phydrogen = 0 - 40 bar n-octane multiphase
silica support
ionic liquid film
silica support
Pt
ionic liquid film
136 Summary / Abstract
pressure. However, basically, the selectivity decrease was just a function of reduced n-hexane
conversion as the selectivty for iso-alkanes is generally strongly dependent on the n-alkane
conversion.
Reaction temperature variation (Treaction = 293 – 313 K) in the n-hexane isomerization
(catalyst: [BMIM]Cl / AlCl3 / H2SO4) showed an influence on the catalytic activity but the
selectivity seemed to be just dependent on the hexane conversion and was not shifted by
reaction temperature change.
Another measure regarding the selectivity improvement, decreased Lewis acidity of the ionic
liquid (n([BMIM]Cl) / n(AlCl 3) = 1/1.7 compared to n([BMIM]Cl) / n(AlCl3) = 1/2.0),
resulted not in higher iso-octanes selectivities at 303 K, merely the n-octane conversion was
parallelly shifted to lower values.
The acidic ionic liquid based catalysts were applied in a liquid-liquid biphasic reaction mode.
Alkanes show a miscibility gap with the polar chloroaluminate ionic liquid. Therefore,
product separation was possible by decantation. An improved recycling and immobilization
concept of the catalytically active acidic ionic liquid offers the SILP system. The ionic liquid
n([BMIM]Cl) / n(AlCl 3) = 1/2.0 was successfully immobilized on pretreated silica support
and applied in a slurry-phase reaction mode (liquid organic phase, solid SILP catalyst). The
tested SILP catalyst showed higher catalytic activity in the n-octane conversion compared to
the classical biphasic system (Treaction = 373 K, phelium = 40 bar) which can be explained by a
larger phase boundary between the acidic ionic liquid and organic phase in the SILP catalyst.
Bifunctional ionic liquid based catalytic systems
The coating of pretreated Pt / silica with a thin film of chloroaluminate ionic liquid, the
synthesis of a SCILL system, resulted in a modified and enhanced activity and selectivity
compared to the monofunctional acidic ionic liquid based catalyst and the heterogeneous
Pt / silica catalyst. The SCILL material was characterized by H2-chemisorption, N2-adsorption
and ICP-AES before its application in catalysis. The SCILL systems revealed a strong
dependency of the catalytic activity on the hydrogen partial pressure. Higher hydrogen partial
pressure resulted in a significant increase in n-octane conversion as well as iso-octanes
selectivity. At a hydrogen partial pressure of 40 bar the n-octane conversion was 73.7 % after
maximum applied modified reaction time, while in the absence of hydrogen the conversion
Summary / Abstract 137
was less than 20 % under otherwise identical conditions (Treaction = 393 K, ptotal = 40 bar).
Comparing the selectivity for iso-octanes at a hydrogen partial pressure of 40 bar with the
hydrogen-free system this difference increased to an absolute value of up to 15 %.
Temperature variation (Treaction = 373 – 423 K) was a further part of the investigative
experiments. According to Arrhenius law, the reaction rate of the SCILL-catalyzed reaction
and the n-octane conversion increased with higher reaction temperature. The selectivity for
iso-octanes was independent of the applied reaction temperature within the considered
temperature range. A less Lewis acidic ionic liquid (n([BMIM]Cl) / n(AlCl 3) = 1/1.7
compared to (n([BMIM]Cl) / n(AlCl3) = 1/2.0) on the pretreated Pt / silica led to a lower
amount of converted n-octanes, like in the case of the monofunctional ionic liquid based
catalysts, but also to slightly reduced selectivities for iso-octanes. In contrast to the catalytic
system [BMIM]Cl / AlCl3 / 1-chlorooctane, no clear trend was obvious that higher
concentrations of the promoter 1-chlorooctane resulted in increased n-octane conversions for
the SCILL system. The iso-octane selectivities exhibited also no recognizable trend if the
alkyl halide was added.
Different analytical techniques were applied to prove the stability of the SCILL catalysts
under reaction conditions. ICP-AES revealed that no Pt or Al leached into the organic phase
within the detection range and only minor amounts of Pt (0.21 wt. % Pt referred to the total
Pt) were found in the washed-off acidic ionic liquid. Further, chlorine insertion from the
chloroaluminate ionic liquid into the organic products could be excluded by ESI-MS. Coking
and accumulation of high-boiling compounds is a major issue for heterogeneous catalyzed
alkane isomerization reactions. Though, highest m/z ratio found in the spectrum of the extract
of the washed-off ionic liquid and hydrolized ionic liquid´s anion applying GC-MS analysis
was 164 if the SCILL catalyzed experiment was conducted under hydrogen atmophsere
(phydrogen = 40 bar) which might origin from cyclic unsaturated hydrocarbons. Thus,
accumulation of carbon-rich species and coke formation in the SCILL system could be
excluded under the conditions applied in this work. Moreover, it could be shown by 1H-NMR
analysis that the butyl side chain of the [BMIM]-cation was still intact after catalysis. The
stability of the SCILL systems was also reflected in the recycling experiments. Almost
constant iso-octanes selectivities and slightly decreasing n-octane conversions with every
recycling run are the results of three catalytic runs in a row with the same SCILL system.
Further, a comparison between the whole product spectra, which consisted of 4 ≤ C ≤ 10, and
RON calculations of a SCILL and monofunctional acidic ionic liquid based catalyzed
n-octane isomerization was conducted. Both catalyst systems yielded only monobranched and
138 Summary / Abstract
dibranched octane isomers and this in very similar distributions. Differences between the
bifunctional SCILL system and monofunctional ionic liquid existed if the composition of the
alkane products with a carbon chain length other than 8 was compared. The differences in the
distributions of the by-products can be explained by the hydrogenation activity of the
Pt-containing catalyst leading to a lower fraction of C5 – C7 products.
A positive effect on n-octane conversion and on the selectivity for iso-octanes was also
observed when the bifunctional catalyst, PtCl2 or PtCl4 in combination with a highly Lewis
acidic ionic liquid, and hydrogen was applied in a multiphase reaction system at 373 K. XRD
analysis proved the in situ formation of Pt nanoparticles in the chloroaluminate ionic liquid
under hydrogen atmosphere.
Comparative experiments for all four catalytic systems (uncoated acidic ionic liquid, SILP,
SCILL and Pt nanoparticles) in the n-octane isomerization under hydrogen or helium
atmosphere without any alkyl halide addition were carried out at 373 K. Firstly, results of
bifunctional SCILL and monofunctional SILP catalysts were compared and discussed.
Bifunctional and monofunctional systems revealed different reaction pathways. The
bifunctional SCILL system acted like the monofunctional SILP catalyst in the absence of
hydrogen. The hydrogenation activity of Pt gave a good explanation for the improved
selectivity of the bifunctional SCILL system under hydrogen. The fact that the catalytic
activity correlated also strongly to the level of hydrogen partial pressure suggests that the
concentration of acidic protons and thus the acidity of the catalytic system might increase if
the Lewis acidic ionic liquid and the active Pt centers are applied in combination with
hydrogen. The Pt nanoparticles in Lewis acidic ionic liquid exhibited enhanced activity and
selectivity in n-octane hydroisomerization in contrast to the monofunctional uncoated acidic
ionic liquid. However, the in situ formed Pt nanoparticles and with it bifunctional catalysts
were less active and showed slightly decreased selectivity if this catalytic system was
compared with the SCILL system.
The interaction of acidic ionic liquid, Pt and hydrogen in the SCILL and Pt nanoparticles
catalyzed n-alkane isomerization opens an additional reaction pathway. A possible reaction
network for the bifunctional and monofunctional ionic liquid based catalysts was proposed.
Summary / Abstract 139
Table 23: Summarized results of the different catalytic systems for the n-octane isomerization studied in this thesis.
Monofunctional catalysts Bifunctional catalysts
Acidic ionic liquids
SILP SCILL Pt nanoparticles in an acidic ionic liquid
Influence of hydrogen
None None Yes Yes
Influence of 1-chlorooctane
Yes / None (no clear trend)
/
Influence of the ionic liquid´s acidity
Yes / Yes Yes
Recyclability Possible (but with high activity decrease)
/ Possible (but with activity decrease)
/
The bifunctional SCILL system composed of Pt / silica coated with a thin layer of acidic
chloroaluminate ionic liquid was the first time applied for the alkane isomerization. The
beneficial interaction of supported Pt, acidic ionic liquid and hydrogen is attractive for further
research and development in the field of alkane isomerization as well as beyond this reaction.
Bifunctional catalysts based on a supported metal, like Pt or Pd, coated with an acidic ionic
liquid offer the opportunity to be investigated for further acid catalyzed reactions or for the
hydrogenation of unsaturated compounds.
140
141
___________________________________________________________________________
6. Zusammenfassung / Kurzfassung
142 Zusammenfassung / Kurzfassung
Zusammenfassung / Kurzfassung 6
Ziel dieser Arbeit war die Entwicklung von Katalysatorsystemen für die
n-Alkanisomerisierung, die auf hochaciden ionischen Flüssigkeiten basieren. Der Fokus lag
dabei auf dem Edukt n-Octan, eine Modellkomponente für längerkettige Kohlenwasserstoffe
(C > 6), welche im Vergleich zu den bisher industriell eingesetzten Ausgangskomponenten
Pentan / Hexan anfälliger für Crackreaktionen ist. Daher wurde Octan bis jetzt noch nicht als
Ausgangsstoff in den Raffinerien verwendet, um Treibstoff mit einer hohen RON zu erhalten.
Auf Grund strengerer umwelttechnischer Regulierungen wird in Zukunft jedoch eine
Ausweitung der Eduktalkane auf längerkettige Kohlenwasserstoffe, wie beispielsweise
n-Octan, notwendig sein.
Die untersuchten Katalysatoren können grundlegend in zwei Kategorien unterteilt werden,
nämlich in monofunktionelle und bifunktionelle Systeme. Zu den Monofunktionellen zählen
die aciden ionischen Flüssigkeiten und SILP Katalysatoren. Die SCILL Systeme, bei welchen
vorbehandeltes Pt / Silica mit einer aciden ionischen Flüssigkeit des Typs [BMIM]Cl / AlCl3
modifiziert wird, und ebenso die Katalysatoren, welche auf in situ gebildeten Pt
Nanopartikeln basieren, werden dem bifunktionellen Katalysatortyp zugeordnet. Tabelle 1
gibt einen Überblick über die in der Arbeit untersuchten Katalysatoren. Die
Versuchsergebnisse der zwei verschiedenen Katalysatorkategorien werden im Folgenden
beschrieben und Tabelle 2 fasst weiterhin den Einfluss wichtiger Reaktionsparameter, sowie
das Ergebnis der Katalysatorrecycling-Versuche zusammen.
Monofunktionelle Katalysatoren, die auf aciden ionischen Flüssigkeiten basieren
Drei verschiedene Katalysatorsysteme, [BMIM]Cl / AlCl3 / H2SO4,
[BMIM]Cl / AlCl 3 / 1-Chloroctan, [BMIM]Cl / AlCl3 / CuCl2 wurden in der
n-Octanisomerisierung bei 303 K getestet, um eine Systemauswahl zu treffen. Die beiden
erstgenannten Katalysatoren wurden weiterhin dahingehend optimiert, eine maximale
Konzentration an Carbeniumionen und damit an n-Octan Umsatz zu erhalten. Außerdem
wurden diese beiden Katalysatorsysteme auf ihre Recyclierbarkeit untersucht. Die
katalytische Aktivität verringerte sich mit dem zweiten bzw. dritten Recycling. Die Abnahme
an n-Octan Umsatz kann auf den Verlust von HCl während der Reaktorbeladung mit neuem
Zusammenfassung / Kurzfassung 143
Octan zurückgeführt werden bzw. auf eine verringerte Lewis Acidität der ionischen
Flüssigkeit, wenn jedem Recycling-Versuch nicht nur n-Octan, sondern auch 1-Chloroctan
zugegeben wurde.
Tabelle 1: Übersicht der in dieser Arbeit untersuchten Katalysatorsysteme.
Beim Edukt n-Octan zeigte der Crack-Inhibitor Wasserstoff (pWasserstoff = 0 – 40 bar) sowohl
bei geringen Reaktionstemperaturen (Katalysatorsystem: [BMIM]Cl / AlCl 3 / H2SO4,
TReaktion = 303 K) also auch bei höheren (Katalysatorsystem:
[BMIM]Cl / AlCl 3 / 1-Chlorooctan, TReaktion = 303 K) überhaupt keine Wirkung.
Monofunktionelle Katalysatoren Bifunktionelle Katalysatoren
� Acide ionische Flüssigkeiten n([BMIM]Cl) / n(AlCl 3) = 1/1,7;1/2,0 n(H2SO4) / n(AlCl3) = 0,0 – 0,75 n(1-Chloroctan) / n(AlCl3) = 0,0 – 0,6 n(CuCl2) / n(AlCl3)= 0 – 0,06 TReaktion = 283 – 393 K pgesamt = 3 – 40 bar pWasserstoff = 0 - 40 bar n-Hexan und n-Octan flüssig-fllüssig zweiphasig
� SCILL
vorbehandeltes Silika beschichtet mit n([BMIM]Cl) / n(AlCl 3) = 1/1,7;1/2,0 α = 0,85 ε = 0,5 n(1-Chloroctan) / n(AlCl3) = 0 – 0,36 TReaktion = 373 – 423 K pgesamt = 15 - 40 bar pWasserstoff = 0 - 40 bar n-Octan Slurry-System
� SILP vorbehandeltes Silika beschichtet mit n([BMIM]Cl) / n(AlCl 3) = 1/2,0 α = 0,85 ε = 0,4 TReaktion = 373 K pgesamt = 40 bar pWasserstoff = 0 - 40 bar n-Octan Slurry-System
� Pt Nanopartikel in acider
ionischer Flüssigkeit n([BMIM]Cl) / n(AlCl 3) = 1/2,0; n([BMIM]Br) / n(AlBr 3) = 1/2,0 PtCl2 oder PtCl4
n(Pt) = 0,3 mmol TReaktion = 373 K pgesamt = 0 - 40 bar pWasserstoff = 0 - 40 bar n-Octan mehrphasig
Silika Träger
ionischer Flüsisgkeitsfilm
Silika Träger
Pt
ionischer Flüssigkeitsfilm
144 Zusammenfassung / Kurzfassung
Der Einfluss von Wasserstoff (pWasserstoff = 0 – 15 bar) auf die katalytische Aktivität und
Selektivität wurde auch für das kurzkettige n-Hexan, welches weniger anfällig für
Crackreaktionen ist, getestet (Katalysatorsystem: [BMIM]Cl / AlCl 3 / H2SO4,
TReaktion = 303 K). Die Zugabe von Wasserstoff beeinträchtigte die Reaktionsgeschwindigkeit
negativ. Im Vergleich zu dem Versuch unter Helium wurde lediglich eine sehr geringfügige
Verschiebung der Selektivitäten hin zu höheren Werten festgestellt. Jedoch können die
erzielten besseren Selektivitäten im Prinzip nur auf die niedrigeren n-Hexan Umsätze
zurückgeführt werden, denn iso-Alkan Selektivitäten sind bei der Alkanisomerisierung stets
umsatzabhängig.
Eine Variation der Reaktionstemperatur (TReaktion = 293 – 313 K) bei der
n-Hexanisomerisierung unter Verwendung des Katalysatorsystems
[BMIM]Cl / AlCl 3 / H2SO4 zeigte nur Einfluss auf die Katalysatoraktivität. Die Selektivität zu
den iso-Hexanen blieb im untersuchten Temperaturbereich unverändert.
Eine weitere Maßnahme zur Selektivitätsverbesserung für das Edukt n-Octan, die
Verringerung der Lewis-Acidität der ionischen Flüssigkeit (n([BMIM]Cl) / n(AlCl 3) = 1/1,7
im Vergleich zu n([BMIM]Cl) / n(AlCl3) = 1/2,0), resultierte ebenfalls nicht in höheren iso-
Octan Selektivitäten bei 303 K, lediglich der Umsatz an n-Octan verringerte sich.
Die Katalysatoren, welche auf aciden ionischen Flüssigkeiten basieren, wurden in flüssig-
flüssig zweiphasiger Reaktionsführung verwendet. Das Reaktionssystem war hierfür sehr gut
geeignet, da Alkane mit den polaren Chloroaluminatschmelzen eine Mischungslücke bilden.
Daher konnten die Produktabtrennung und das Katalysatorrecycling einfach durch
Dekantieren erfolgen. Eine optimierte Möglichkeit zum Katalysatorrecycling und zum
Immobilisieren der katalytisch aktiven ionischen Flüssigkeit stellt das SILP Konzept dar. Es
wurde gezeigt, dass die ionische Flüssigkeit n([BMIM]Cl) / n(AlCl 3) = 1/2,0 erfolgreich auf
einem vorbehandelten Silika Träger immobilisiert und in einem Slurry-System verwendet
werden konnte (flüssige Organik, fester SILP Katalysator). Der SILP Katalysator zeigte im
Vergleich zur ungeträgerten aciden ionischen Flüssigkeit eine höhere katalytische Aktivität
(TReaktion = 373 K, pHelium = 40 bar). Dies kann auf eine größere Phasengrenzfläche zwischen
der geträgerten aciden ionischen Flüssigkeit und der organischen Phase im SILP Katalysator
zurückgeführt werden.
Zusammenfassung / Kurzfassung 145
Bifunktionelle Katalysatorsysteme, die auf ionischen Flüssigkeiten basieren
Die Verwendung eines SCILL Systems, welches durch das Beschichten von vorbehandeltem
Pt / Silika mit einer dünnen Schicht an acider Chloroaluminatschmelze hergestellt wird, führte
zu modifizierter Aktivität und Selektivität im Vergleich zu monofunktionellen Katalysatoren,
die auf aciden ionischen Flüssigkeiten basieren oder im Vergleich zu dem heterogenen
Katalysator Pt / Silika. Zunächst wurde das SCILL System mittels H2-Chemisorption,
N2-Adsorption und ICP-AES-AES charakterisiert, bevor es in der Katalyse eingesetzt wurde.
Die katalytische Aktivität der SCILL Katalysatoren war stark vom Wasserstoffpartialdruck
abhängig. Höhere Wasserstoffpartialdrücke resultierten in einer deutlichen Steigerung des
n-Octan Umsatzes und der iso-Oktan Selektivität. Bei einem Wasserstoffpartialdruck von
40 bar wurden nach maximaler modifizierter Reaktionszeit 73.7 % n-Octan umgesetzt,
während es unter Helium nur 20 % unter ansonsten identischen Versuchsbedingungen waren
(TReaktion = 393 K, pgesamt = 40 bar). Vergleicht man die Selektivität zu den iso-Octanen bei
einem Wasserstoffpartialdruck von 40 bar mit den Ergebnissen des Wasserstoff-freien
Reaktionssystems, ergibt sich eine Selektivitätssteigerung von 15 %. Weitere Experimente
beinhalteten eine Temperaturvariation (TReaktion = 373 – 423 K). Entsprechend dem Gesetz
von Arrhenius erhöhte sich die Reaktionsgeschwindigkeit und somit auch der n-Octan Umsatz
mit steigender Temperatur. Innerhalb des untersuchten Temperaturbereichs war die iso-Octan
Selektivität unabhängig von der Reaktionstemperatur. Eine Beschichtung des vorbehandelten
Pt / Silikas mit einer weniger Lewis aciden ionischen Flüssigkeit
(n([BMIM]Cl) / n(AlCl 3) = 1/1,7 im Vergleich zu n([BMIM]Cl) / n(AlCl3) = 1/2,0) führte,
wie dies auch schon für die acide ionische Flüssigkeit beobachtet wurde, zu geringerem
Umsatz an n-Octan. Jedoch verschlechterte sich bei dem SCILL System im Vergleich zum
monofunktionellen Katalysator bei dieser Aciditätsvariation auch leicht die Selektivität. Im
Vergleich zu dem Katalysator [BMIM]Cl / AlCl3 / 1-Chloroctan konnte bei höheren
Konzentrationen des Promoters 1-Chloroctan in Kombination mit dem SCILL System kein
Trend hin zu höheren n-Octan Umsätzen ausgemacht werden, dies war ebenso für die
iso-Octan Selektivität der Fall.
Verschiedene Analytikmethoden wurden angewandt, um die Stabilität des SCILL Systems
unter Reaktionsbedingungen nachzuweisen. Mittels ICP-AES konnte gezeigt werden, dass
kein Pt und Al in die Organik ausgetragen wurde und nur geringe Mengen an Pt (0,21 Gew. %
Pt der Gesamtmenge an Pt) wurden in der aciden ionischen Flüssigkeit gefunden, die zuvor
vom SCILL-Katalysator abgewaschen wurde. Weiterhin konnte der Einbau von Chlorid aus
146 Zusammenfassung / Kurzfassung
der ionischen Flüssigkeit in die Produkte durch die Anwendung von ESI-MS ausgeschlossen
werden. Verkokung und die Anreicherung hochsiedender Komponenten ist eine Hauptursache
für die Deaktivierung von heterogenen bifunktionellen Katalysatoren für die
Alkanumsetzung. In den Spektren der GC-MS Analytik des Extraktes der abgewaschenen
ionischen Flüssigkeit und des hydrolysierten Anions der ionischen Flüssigkeit, welche bei
einem SCILL Experiment unter Wasserstoff (pWasserstoff = 40 bar) eingesetzt wurde, belief sich
der maximale m/z-Wert auf 164, der auf cyclische ungesättigte Kohlenwasserstoffe
zurückgeführt werden könnte. Daher konnten unter den angewandten Bedingungen die
Anreicherung kohlenstoffreicher Verbindungen und Verkokung des SCILL Systems
ausgeschlossen werden. Weiterhin konnte durch 1H-NMR Messungen gezeigt werden, dass
die Butyl-Seitenkette des [BMIM]-Kations auch nach dem Versuch noch vorhanden war und
nicht in der Reaktion umgesetzt wurde. Die Stabilität der SCILL Systeme spiegelte sich auch
in den Recyclingversuchen wieder. Annähernd konstante iso-Octan Selektivitäten und sich
leicht verringernde n-Octan Umsätze wurden für die drei Recyclingexperimente mit
demselben Katalysator erreicht. Abschließend wurden ein Vergleich der Produktspektren,
welche beide aus Alkanen mit 4 ≤ C ≤10 bestanden, und eine Berechnung der RON der
Produkte für zwei Experimente durchgeführt, die mit einem SCILL System bzw. acider
ionischer Flüssigkeit katalysiert wurden. Beide Katalysatortypen produzierten einfach und
zweifach verzweigte Octanisomere, wobei die Selektivitäten hierzu sehr ähnlich waren. Beim
Vergleich der Produkte, deren Kohlenstoffanzahl von 8 verschieden war, ergaben sich
Unterschiede. Diese Abweichungen in der Nebenproduktverteilung können durch die
Hydrierfähigkeit des Pt enthaltenden Katalysators erklärt werden, wodurch eine geringere
Konzentration an Alkanen mit einer Kettenlänge von C5 – C7.
Auch mit den bifunktionellen Katalysatoren, PtCl2 oder PtCl4 in Kombination mit einer stark
Lewis aciden ionischen Flüssigkeit, konnte ein positiver Effekt von Wasserstoff auf den
n-Octan Umsatz und die Selektivität zu iso-Oktanen bei 373 K festgestellt werden. Mit XRD
konnte die in situ Bildung von Pt Nanopartikeln in der Chloroaluminatschmelze unter
Wasserstoffatmosphäre nachgewiesen werden.
Vergleichende Experimente wurden für alle vier katalytischen Systeme (ungeträgerte acide
ionische Flüssigkeit, SILP, SCILL und Pt Nanopartikel) in der n-Octanisomerisierung,
sowohl unter Helium als auch unter Wasserstoffatmosphere, ohne Zusatz von Promotoren bei
373 K durchgeführt. Zuerst wurden die Ergebnisse der SCILL und SILP Katalysatoren
verglichen und diskutiert. Daraus lässt sich ableiten, dass unterschiedliche Reaktionspfade für
Zusammenfassung / Kurzfassung 147
die mono- und bifunktionellen Katalysatoren vorliegen. In der Abwesenheit von Wasserstoff
verhält sich das SCILL System wie ein monofunktioneller Katalysator. Erhöhte iso-Oktan
Selektivitäten unter Anwendung des SCILL Katalysators unter Wasserstoff lassen sich gut
durch die Hydrierfähigkeit des Pt begründen. Die Tatsache, dass die katalytische Aktivität
eine starke Abhängigkeit vom Wasserstoffpartialdruck zeigt, impliziert, dass die
Konzentration acider Protonen und damit die Acidität des katalytischen Systems zunehmen
könnte, wenn die Lewis acide ionische Flüssigkeit und die aktiven Pt Zentren in Kombination
mit Wasserstoff bei der Reaktion eingesetzt werden. Im Vergleich zu der ungeträgerten
monofunktionellen ionischen Flüssigkeit zeigten die Pt Nanopartikel in acider ionischer
Flüssigkeit eine verbesserte Aktivität und Selektivität. Jedoch waren die Pt
Nanopartikelsysteme immer noch weniger aktiv und geringfügig unselektiver als die
bifunktionellen SCILL Systeme in der Hydroisomerisierung von n-Octan.
Tabelle 2: Übersicht der Ergebnisse für die in dieser Arbeit untersuchten Katalysatorsysteme für die n-Octanisomerisierung.
Monofunktionelle Katalysatoren Bifunktionelle Katalysatoren
Acide ionische Flüssigkeiten
SILP SCILL Pt Nanopartikel in acider ionischer Flüssigkeit
Einfluss von Wasserstoff
Nein Nein Ja Ja
Einfluss von 1-Chloroctan
Ja / Nein (kein Trend zu erkennen)
/
Einfluss der Acidität der ionischen Flüssigkeit
Ja / Ja Ja
Recyclierbarkeit Möglich (aber hohe Verringerung der Aktivität)
/ Möglich (aber Verringerung der Aktivität)
/
Das Zusammenspiel von acider ionischer Flüssigkeit, Pt und Wasserstoff eröffnet zusätzliche
Reaktionspfade. Es wurde ein mögliches Reaktionsnetzwerk für die Alkanisomerisierung,
148 Zusammenfassung / Kurzfassung
katalysiert mit bifunktionellen oder monofunktionellen auf aciden ionischen Flüssigkeiten
basierenden Katalysatorsystemen, aufgestellt.
Das bifunktionelle Katalysatorsystem, bestehend aus Pt / Silica, welches mit einem dünnen
Film einer aciden Chloroaluminatschmelze beschichtet ist, wurde zum ersten Mal bei der
Alkanisomerisierung eingesetzt. Die Interaktion von geträgertem Pt, acider ionischer
Flüssigkeit und Wasserstoff erweist sich als attraktiv für weitere Forschung und Entwicklung,
sowohl auf dem Gebiet der Alkanisomerisierung, als auch darüber hinaus. Bifunktionelle
Katalysatoren, welche auf geträgerten Metallen, wie Pt oder Pd, basieren, die mit einer aciden
ionischen Flüssigkeit beschichtet sind, bieten die interessante Möglichkeit, dass deren Einsatz
auch für andere säurekatalysierte Reaktionen und Hydrierungen ungesättigter Verbindungen
erforscht wird.
149
___________________________________________________________________________
7. Appendix
150 Appendix
Appendix 7
7.1 Influence of hydrogen on n-hexane isomerization at very mild
reaction conditions
Conditions: n([BMIM]Cl) / n(AlCl3) = 1/2, n(H2SO4) / n(AlCl3) = 0.18, ptotal = 15 bar, ptotal = phelium + phydrogen.
Figure 59: Influence of modified reaction time and hydrogen partial pressure on n-hexane conversion.
0 5 10 15 20 25 300
10
20
30
40
50
60
70
X n
-hex
ane /
%
modified reaction time / min mol ionic liquid
mol-1 n-hexane
Treaction
= 293 K Treaction
= 303 K Treaction
= 313 K (phydrogen
= 10 bar)
Treaction
= 293 K Treaction
= 303 K Treaction
= 313 K (phydrogen
= 0 bar)
Appendix 151
Conditions: n([BMIM]Cl / n(AlCl3) = 1/2, n(H2SO4) / n(AlCl3) = 0.18, ptotal = 15 bar, ptotal = phelium + phydrogen.
Figure 60: Influence of n-hexane conversion and hydrogen partial pressure on the selectivity for iso-hexanes.
0 10 20 30 40 50 60 700
10
20
30
40
50
60
70
80
90
100
S iso-
hexa
nes /
%
X n-hexane
/ %
Treaction
= 293 K Treaction
= 303 K Treaction
= 313 K (p hydrogen
= 10 bar)
Treaction
= 293 K Treaction
= 303 K Treaction
= 313 K (p hydrogen
= 0 bar)
152 Appendix
7.2 Reproducibility of SCILL catalyzed experiments
Conditions: mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 373 K, phydrogen = 40 bar, n(Cl-octane) / n(AlCl3) = 0.036.
Figure 61: n-Octane conversion over modified reaction time.
0 5 10 150
10
20
30
40
50
60
70
80
90
100X
n-oc
tane /
%
modified reaction time / min molionic liquid
mol-1n-octane
experiment 1 (Treaction
= 373 K)
experiment 2 (Treaction
= 373 K)
Appendix 153
Conditions: mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 373 K, phydrogen = 40 bar, n(Cl-octane) / n(AlCl3) = 0.036.
Figure 62: Selectivity for iso-octanes over n-octane conversion.
Conditions: mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 393 K, phydrogen = 40 bar, n(Cl-octane) / n(AlCl3) = 0.036.
Figure 63: n-Octane conversion over modified reaction time.
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
experiment 1 (Treaction
= 373 K)
experiment 2 (Treaction
= 373 K)S is
o-oc
tane
s / %
Xn-octane
/ %
0 5 10 150
10
20
30
40
50
60
70
80
90
100 experiment 1 (T
reaction = 393 K)
experiment 2 (Treaction
= 393 K)
Xn-
octa
ne /
%
modified reaction time / min molionic liquid
mol-1n-octane
154 Appendix
Conditions: mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 393 K, phydrogen = 40 bar, n(Cl-octane) / n(AlCl3) = 0.036.
Figure 64: Selectivity for iso-octanes over n-octane conversion.
Conditions: mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 393 K, phelium = 15 bar (experiment 1), phelium = 40 bar (experiment 2), n(Cl-octane) / n(AlCl3) = 0.036.
Figure 65: n-Octane conversion over modified reaction time.
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
experiment 1 (Treaction
= 393 K)
experiment 2 (Treaction
= 393 K)S is
o-oc
tane
s / %
Xn-octane
/ %
0 5 10 150
10
20
30
40
50
60
70
80
90
100
experiment 1 (Treaction
= 393 K)
experiment 2 (Treaction
= 393 K)
Xn-
octa
ne /
%
modified reaction time / min molionic liquid
mol-1n-octane
Appendix 155
Conditions: mSCILL = 20 g, n([BMIM]Cl) / n(AlCl3) = 1/2, ε = 0.5, α = 0.85, Treaction = 393 K, phelium = 15 bar (experiment 1), phelium = 40 bar (experiment 2), n(Cl-octane) / n(AlCl3) = 0.036.
Figure 66: Selectivity for iso-octanes over n-octane conversion.
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
45
experiment 1 (Treaction
= 393 K)
experiment 2 (Treaction
= 393 K)S is
o-oc
tane
s / %
Xn-octane
/ %
156 Appendix
7.3 Mass spectra of GC-MS analysis
Figure 67: Mass spectra (positive and negative mode) of the organic product phase.
Appendix 157
7.4 Calculation of the RON
Table 24: RON of different alkanes according literature [20, 22].
Alkane RON Alkane RON
n-Butane 94 Monomethylheptanes 52
iso-Butane 102 Dimethylheptanes 94
n-Pentane 62 2,2,3-Trimethylbutane 113
iso-Pentane 92 n-Octane -15
n-Hexane 25 Monomethyloctanes 25
Monomethylhexanes 76 Dimethyloctanes 69
2,2-Dimethylbutane 92 Trimethyloctanes 105
2,3-Dimethylbutane 106 iso-Nonanes 91
n-Heptane 0 iso-Decanes 81
158 Appendix
7.5 XRD diffractogram
Figure 68: X-ray diffractogram of the particles separated from the ionic liquid after catalysis (n([BMIM]Cl) / n(AlCl 3) = 1/2, Treaction = 373 K, nplatinum = 0.3 mmol (PtCl4), phydrogen = 40 bar).
Position [°2Theta] (Kupfer (Cu))
10 20 30 40 50 60 70
Impulse
0
500
1000
lang_2-80°_5s_0,02 step_x1_CAB198
2 θ / °
Inte
nsity
/ -
159
___________________________________________________________________________
8. References
160 References
References 8
1. W. Dabelstein, A. Reglitzky, A. Schütze, K. Reders, Automotive Fuels, in Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA, 2007.
2. H. Weyda, E. Kohler, Catalysis Today, 81, 51-55, 2003.
3. N. A. Cusher, Handbook of Petroleum Refining Processes, 3rd ed, Isomerization, ed. Meyer, R. A., Boston, Mass. , McGraw-Hill, 9.3-9.5, 2004.
4. A. Chica, A. Corma, Chemie Ingenieur Technik-CIT, 79(6), 857-870, 2007.
5. M. A. Arribas, F. Marquez, A. Martinez, Journal of Catalysis, 190, 309-319, 2000.
6. T. Ohno, Z. Li, N. Sakai, H. Sakagami, N. Takahashi, T. Matsuda, Applied Catalysis, A: General, 389, 52-59, 2010.
7. M. Guisnet, P. Magnoux, Applied Catalysis, A: General, 212, 83-96, 2001.
8. A. S. Berenblyum, E. A. Katsman, Y. Z. Karasev, Applied Catalysis, A: General, 315, 128-134, 2006.
9. R. Zhang, X. Meng, Z. Liu, J. Meng, C. Xu, Industrial & Engineering Chemistry Research, 47(21), 8205-8210, 2008.
10. T. Welton, Chemical Reviews, 99, 2071-2083, 1999.
11. G. P. Smith, A. S. Dworkin, R. M. Pagni, S. P. Zingg, Journal of the American Chemical Society, 111, 525-530, 1989.
12. M. Baerns, A. Behr, A. Brehm, J. Gmehling, H. Hofmann, U. Onken, A. Renken, Technische Chemie, Weinheim, Wiley-VCH, 491-506, 2006.
13. R. H. Crabtree, Chemical Reviews, 85(4), 245-269, 1985.
14. G. Emig, E. Klemm, Technische Chemie, 5. ed, Heidelberg, Springer Verlag, 107-139, 2005.
15. G. A. Olah, Á. Molnár, Hydrocarbon Chemistry, 2. ed, Hoboken, New Jersey, Wiley, 160-214, 2003.
16. B. E. Poling, J. M. Prausnitz, O. C. John Paul, The Properties of Gases and Liquids, 5. ed, Boston, McGraw-Hill A29-A30, 2001.
17. Europäische Gemeinschaft, Verordnung (EG) Nr. 1907/2006 des europäischen Parlaments und des Rates vom 18. Dezember, 2006.
18. Concawe, in Gasoline, Product Dossier No. 92/103, Brüssel, 1992.
19. S. Hirzel, Benzin (Ottokrafstoffe): BUA-Stoffbericht 200, 1997.
References 161
20. P. Ghosh, K. J. Hickey, S. B. Jaffe, Industrial & Engineering Chemistry Research, 45(1), 337-345, 2006.
21. DIN EN 228, Kraftstoffe für Kraftfahrzeuge - Unverbleite Ottokraftstoffe - Anforderungen und Prüfverfahren, 2008.
22. W. G. Lovell, Journal of Industrial and Engineering Chemistry, 40, 2388-2438, 1948.
23. L. Koottungal, Oil and Gas Journal, 2009.
24. S. T. Sie, Handbook of Heterogeneous Catalysis, ed. Ertl, G., Knözinger, H., Schüth, F.,Weitkamp, J., Vol. 7, Weinheim, Wiley-VCH, 2809-2830, 1997.
25. H. G. Krane, E. W. Kane, Petroleum Refiner, 36, 177-182, 1957.
26. W. M. J. Ruedisulj, H. D. Evans, E. B. Fountain, Hydrocarbon Processing and Petroleum Refiner, 42, 125-130, 1963.
27. Y. Ono, Catalysis Today, 81(1), 3-16, 2003.
28. D. H. Belden, Petroleum Refiner, 35, 149-152, 1956.
29. J. J. Hur, J. K. Deichler, G. R. Worrell, Oil and Gas Journal, 54, 103-105,107, 1956.
30. I. E. Maxwell, J. E. Naber, Catalysis Letters, 12(1), 105-115, 1992.
31. H. Kouwenhoven, W. C. V. Z. Langhout, Chemical Engineering Progress, 67(4), 65, 1971.
32. T. Kimura, Catalysis Today, 81(1), 57-63, 2003.
33. S. Arrhenius, Zeitschrift fuer Physikalische Chemie, 1, 631, 1887.
34. J. N. Bronsted, Recueil des Travaux Chimiques des Pays-Bas et de la Belgique, 42, 718-728, 1923.
35. T. M. Lowry, Journal of the Society of Chemical Industry, 42(3), 43-47, 1923.
36. G. N. Lewis, Transactions of the Faraday Society, 19, 452-458, 1923.
37. G. N. Lewis, Journal of the Franklin Institute, 226, 293-313, 1938.
38. A. Corma, H. Garcia, Chemical Reviews, 102, 3837-3892, 2002.
39. A. Corma, H. Garcia, Chemical Reviews, 103, 4307-4365, 2003.
40. T. E. Mallouk, G. L. Rosenthal, G. Mueller, R. Brusasco, N. Bartlett, Inorganic Chemistry, 23, 3167-3173, 1984.
41. K. O. Christe, D. A. Dixon, D. Mclemore, W. W. Wilson, J. A. Sheehy, J. A. Boatz, Journal of Fluorine Chemistry, 101, 151-153, 2000.
42. I. Krossing, I. Raabe, Chemistry - A European Journal, 10, 5017-5030, 2004.
162 References
43. L. O. Mueller, D. Himmel, J. Stauffer, G. Steinfeld, J. Slattery, G. Sabtiso-Quinones, V. Brecht, I. Krossing, Angewandte Chemie, International Edition, 47, 7659-7663, 2008.
44. N. F. Hall, J. B. Conant, Journal of the American Chemical Society, 49, 3047-3061, 1927.
45. J. B. Conant, N. F. Hall, Journal of the American Chemical Society, 49, 3062-3070, 1927.
46. G. A. Olah, G. K. S. Prakash, J. Sommer, Science, 206(4414), 13-20, 1979.
47. G. A. Olah, Angewandte Chemie International Edition in English, 34(1314), 1393-1405, 1995.
48. R. J. Gillespie, T. E. Peel, Advances in Physical Organic Chemistry, 9, 1-24, 1971.
49. R. J. Gillespie, T. E. Peel, E. A. Robinson, Journal of the American Oil Chemists' Society, 93, 5083-5087, 1971.
50. R. J. Gillespie, T. E. Peel, Journal of the American Oil Chemists' Society, 95, 5173-5178, 1973.
51. R. J. Gillespie, Canadian Chemical Education, 4, 9-10, 1969.
52. L. P. Hammett, A. J. Deyrup, Journal of the American Chemical Society, 54, 2721-2739, 1932.
53. M. A. Paul, F. A. Long, Chemical Reviews, 57, 1-45, 1957.
54. G. A. Olah, G. K. S. Prakash, Á. Molnár, J. Sommer, Superacid Chemistry, 2nd ed, Hoboken, New Jersey, Wiley, 1-34; 83-87; 501-539, 2009.
55. A. Corma, Chemical Reviews, 95, 559-614, 1995.
56. W. E. Farneth, R. J. Gorte, Chemical Reviews, 95, 615-635, 1995.
57. M. Mandel, Nature, 176(4486), 792-793, 1955.
58. D. Himmel, S. K. Goll, I. Leito, I. Krossing, Angewandte Chemie, International Edition, 49, 6885-6888, 2010.
59. D. Himmel, S. K. Goll, I. Leito, I. Krossing, Chemistry - A European Journal, 17, 5808-5826, 2011.
60. J. Weitkamp, W. Gerhardt, P. A. Jacobs, Acta Physica et Chemica, 31, 261-270, 1985.
61. J. Weitkamp, P. A. Jacobs, J. A. Martens, Applied Catalysis, 8(1), 123-141, 1983.
62. M. Steijns, G. Froment, P. Jacobs, J. Uytterhaeven, J. Weitkamp, Industrial & Engineering Chemistry Product Research and Development, 20(4), 654-660, 1981.
63. J. A. Martens, P. A. Jacobs, J. Weitkamp, Applied Catalysis, 20(1-2), 239-281, 1986.
References 163
64. G. Kinger, H. Vinek, Applied Catalysis, A: General, 218(1-2), 139-150, 2001.
65. M. Guisnet, C. Thomazeau, J. L. Lemberton, S. Mignard, Journal of Catalysis, 151(1), 102-110, 1995.
66. G. A. Olah, Friedel-Crafts and Related Reactions, Isomerization of Saturated Hydrocarbons, ed. Pines, H.,Hoffman, N. E., Vol. 2, New York, Interscience Publishers, 1211-1217, 1964.
67. G. A. Olah, Friedel-Crafts and Related Reactions, Alkylation of Aromatics with Haloalkanes, ed. Drahowzal, F. A., Vol. 1, New York, Interscience Publishers, 417-475, 1964.
68. J. Sommer, R. Jost, M. Hachoumy, Catalysis Today, 38(3), 309-319, 1997.
69. D. M. Brouwer, J. M. Oelderik, Recueil des Travaux Chimiques des Pays-Bas, 87, 721-736, 1968.
70. J. Weitkamp, H. Farag, Acta Physica et Chemica, 24, 327-333, 1978.
71. J. Weitkamp, Industrial & Engineering Chemistry Product Research and Development, 21, 550-558, 1982.
72. J. Weitkamp, ChemCatChem, 4, 292-306, 2012.
73. A. Corma, A. V. Orchillés, Microporous and Mesoporous Materials, 35-36, 21-30, 2000.
74. A. Corma, J. Planelles, F. Tomás, Journal of Catalysis, 94(2), 445-454, 1985.
75. A. Corma, A. Martinez, S. Pergher, S. Peratello, C. Perego, G. Bellusi, Applied Catalysis, A: General, 152(1), 107-125, 1997.
76. S. Kotrel, H. Knözinger, B. C. Gates, Microporous and Mesoporous Materials, 35-36, 11-20, 2000.
77. W. O. Haag, R. M. Dessau, R. M. Lago, Studies in Surface Science and Catalysis, 60, 255-265, 1991.
78. A. Corma, A. Martínez, Catalysis Reviews. Science and Engineering, 35(4), 483-570, 1993.
79. A. F. Bickel, C. J. Gaasbeek, H. Hogeveen, J. M. Oelderik, J. C. Platteeuw, Chemical Communications, 634-635, 1967.
80. A. Corma, J. Sánchez, F. Tomás, Journal of Molecular Catalysis, 19(1), 9-15, 1983.
81. K. Ebitani, J. Tsuji, H. Hattori, H. Kita, Journal of Catalysis, 138(2), 750-753, 1992.
82. J. Meusinger, A. Corma, Journal of Catalysis, 152(1), 189-197, 1995.
83. J. Meusinger, A. Corma, Journal of Catalysis, 159(2), 353-360, 1996.
84. T. F. Degnan, C. R. Kennedy, AIChE journal, 39(4), 607-614, 1993.
164 References
85. A. Holló, J. Hancsók, D. Kalló, Applied Catalysis, A: General, 229(1-2), 93-102, 2002.
86. P. B. Weisz, E. W. Swegler, Science, 126, 31-32, 1957.
87. H. L. Coonradt, W. E. Garwood, Industrial & Engineering Chemistry Process Design and Development, 3(1), 38-45, 1964.
88. J. H. Sinfelt, Catalysis Reviews, 9, 145-166, 1974.
89. J. R. Anderson, Advances in Catalysis, 23, 1-90, 1973.
90. M. Busto, M. E. Lovato, C. R. Vera, K. Shimizu, J. M. Grau, Applied Catalysis, A: General, 355(1-2), 123-131, 2009.
91. P. Wasserscheid, W. Keim, Angewandte Chemie, International Edition, 39(21), 3772-3789, 2000.
92. O. Aschenbrenner, S. Supasitmongkol, M. Taylor, P. Styring, Green Chemistry, 11, 1217-1221, 2009.
93. L. P. N. Rebelo, J. N. C. Lopes, J. M. S. S. Esperanca, E. Filipe, Journal of Physical Chemistry B, 109, 6040-6043, 2005.
94. Y. U. Paulechka, D. H. Zaitsau, G. J. Kabo, A. A. Strechan, Thermochimica Acta, 439, 158-160, 2005.
95. M. J. Earle, J. M. S. S. Esperanca, M. A. Gilea, L. J. N. Canongia, L. P. N. Rebelo, J. W. Magee, K. R. Seddon, J. A. Widegren, Nature, 439, 831-834, 2006.
96. M. Freemantle, Chemical & Engineering News, 76(13), 32-37, 1998.
97. T. Welton, Coordination Chemistry Reviews, 248(21-24), 2459-2477, 2004.
98. J. P. Hallett, T. Welton, Chemical Reviews, 111(5), 3508-3576, 2011.
99. J. Holbrey, K. Seddon, Clean Technologies and Environmental Policy, 1(4), 223-236, 1999.
100. S. Werner, M. Haumann, P. Wasserscheid, Annual Review of Chemical and Biomolecular Engineering, 1, 203-230, 2010.
101. H. Olivier-Bourbigou, L. Magna, Journal of Molecular Catalysis A: Chemical, 182-183, 419-437, 2002.
102. H. Olivier-Bourbigou, L. Magna, D. Morvan, Applied Catalysis, A: General, 373(1-2), 1-56, 2010.
103. J. Dupont, R. F. de Souza, P. A. Z. Suarez, Chemical Reviews, 102, 3667-3691, 2002.
104. D. J. Cole-Hamilton, Science, 299(5613), 1702, 2003.
105. T. P. Wier, Jr., F. H. Hurley, US2446349, 1948.
References 165
106. F. H. Hurley, US2446331, 1948.
107. F. H. Hurley, T. P. Wier, Jr., Journal of The Electrochemical Society, 98, 207-212, 1951.
108. F. H. Hurley, T. P. Wier, Jr., Journal of The Electrochemical Society, 98, 203-206, 1951.
109. R. J. Gale, B. Gilbert, R. A. Osteryoung, Inorganic Chemistry, 17, 2728-2729, 1978.
110. S. Takahashi, N. Koura, S. Kohara, M. L. Saboungi, L. A. Curtiss, Plasmas Ions, 2, 91-105, 1999.
111. J. Robinson, R. A. Osteryoung, Journal of the American Chemical Society, 101, 323-327, 1979.
112. J. S. Wilkes, J. A. Levisky, R. A. Wilson, C. L. Hussey, Inorganic Chemistry, 21(3), 1263-1264, 1982.
113. Z. J. Karpinski, R. A. Osteryoung, Inorganic Chemistry, 23(10), 1491-1494, 1984.
114. R. T. Carlin, P. C. Truelove, R. A. Osteryoung, Electrochimica Acta, 37(14), 2615-2628, 1992.
115. J. C. Nardi, C. L. Hussey, L. A. King, US4122245A, 1978.
116. A. M. Elias, J. S. Wilkes, Journal of Chemical & Engineering Data, 39, 79-82, 1994.
117. H. A. Oye, M. Jagtoyen, T. Oksefjell, J. S. Wilkes, Materials Science Forum, 73-75, 183-189, 1991.
118. J. S. Wilkes, J. A. Levisky, J. L. Pflug, C. L. Hussey, T. B. Scheffler, Analytical Chemistry, 54, 2378-2379, 1982.
119. J. S. Wilkes, J. S. Frye, G. F. Reynolds, Inorganic Chemistry, 22, 3870-3872, 1983.
120. A. A. Fannin, Jr., L. A. King, J. A. Levisky, J. S. Wilkes, Journal of Physical Chemistry, 88, 2609-2614, 1984.
121. A. A. Fannin, Jr., D. A. Floreani, L. A. King, J. S. Landers, B. J. Piersma, D. J. Stech, R. L. Vaughn, J. S. Wilkes, J. L. Williams, Journal of Physical Chemistry, 88, 2614-2621, 1984.
122. P. C. Trulove, R. A. Osteryoung, Inorganic Chemistry, 31(19), 3980-3985, 1992.
123. M. a. M. Noel, P. C. Trulove, R. A. Osteryoung, Analytical Chemistry, 63(24), 2892-2896, 1991.
124. T. A. Zawodzinski Jr, R. T. Carlin, R. A. Osteryoung, Analytical Chemistry, 59(21), 2639-2640, 1987.
125. T. A. Zawodzinski, Jr., R. A. Osteryoung, Inorganic Chemistry, 29, 2842-2847, 1990.
166 References
126. A. a. K. Abdul-Sada, A. M. Greenway, K. R. Seddon, T. Welton, Organic Mass Spectrometry, 28(7), 759-765, 1993.
127. K. Ala'a, A. G. Avent, M. J. Parkington, T. A. Ryan, K. R. Seddon, T. Welton, Journal of the Chemical Society, Dalton Transactions, (22), 3283-3286, 1993.
128. G. P. Smith, A. S. Dworkin, R. M. Pagni, S. P. Zingg, Journal of the American Chemical Society, 111, 5075-5077, 1989.
129. J. L. E. Campbell, K. E. Johnson, Inorganic Chemistry, 32, 3809-3815, 1993.
130. J. L. E. Campbell, K. E. Johnson, Journal of the American Chemical Society, 117(29), 7791-7800, 1995.
131. E. J. Angueira, M. G. White, AIChE Journal, 51(10), 2778-2785, 2005.
132. A. H. Otto, T. Steiger, S. Schrader, Chemical Communications, 1998(3), 391-392, 1998.
133. C. Sikorska, S. Freza, P. Skurski, Journal of Physical Chemistry A, 114, 2235-2239, 2010.
134. A. Elaiwi, P. B. Hitchcock, K. R. Seddon, N. Srinivasan, Y.-M. Tan, T. Welton, J. A. Zora, Journal of the Chemical Society, Dalton Transactions, 3467-3472, 1995.
135. R. T. Carlin, R. A. Osteryoung, J. S. Wilkes, J. Rovang, Inorganic Chemistry, 29, 3003-3009, 1990.
136. B. Gilbert, Y. Chauvin, H. Olivier, M.-V. T. F. Di, Journal of the Chemical Society, Dalton Transactions, 3867-3871, 1995.
137. B. Cornils, W. A. Herrmann, I. T. Horváth, W. Leitner, S. Mecking, H. Olivier-Bourbigou, D. Vogt, Multiphase Homogeneous Catalysis, Vol. 2, Weinheim, Wiley-VCH, 511-513, 2005.
138. J. R. Sanders, E. H. Ward, C. L. Hussey, Journal of the Electrochemical Society, 133, 325-330, 1986.
139. M. Ma, K. E. Johnson, Journal of the American Chemical Society, 117(5), 1508-1513, 1995.
140. J. A. Boon, J. A. Levisky, J. L. Pflug, J. S. Wilkes, The Journal of Organic Chemistry, 51(4), 480-483, 1986.
141. A. a. K. Abdul-Sada, M. P. Atkins, B. Ellis, P. K. G. Hodgson, M. L. M. Morgan, K. R. Seddon, WO9521806A1, 1995.
142. F. G. Sherif, L.-J. Shyu, C. C. Greco, A. G. Talma, C. P. M. Lacroix, WO9803454A1, 1998.
143. F. G. Sherif, L.-J. Shyu, WO9903163A1, 1999.
144. L. Y. Piao, X. Fu, Y. L. Yang, G. H. Tao, Y. Kou, Catalysis Today, 93-95, 301-305, 2004.
References 167
145. H. Xin, Q. Wu, M. Han, D. Wang, Y. Jin, Applied Catalysis, A: General, 292, 354-361, 2005.
146. J. Joni, D. Schmitt, P. S. Schulz, T. J. Lotz, P. Wasserscheid, Journal of Catalysis, 258, 401-409, 2008.
147. Z. C. Liu, X. H. Meng, R. Zhang, C. M. Xu, Petroleum Science and Technology, 27, 226-237, 2009.
148. M. J. Earle, K. R. Seddon, C. J. Adams, G. Roberts, Chemical Communications, 2097-2098, 1998.
149. J. K. D. Surette, L. Green, R. D. Singer, Chemical Communications, 2753-2754, 1996.
150. J. F. Knifton, US4554383A, 1985.
151. N. Brausch, A. Metlen, P. Wasserscheid, Chemical Communications, 2004(13), 1552-1553, 2004.
152. E. J. Angueira, M. G. White, Journal of Molecular Catalysis A: Chemical, 227, 51-58, 2005.
153. W. J. Zhao, X. Z. Jiang, Catalysis Letters, 107, 123-125, 2006.
154. H. Olivier, Y. Chauvin, A. Hirschauer, Preprints - American Chemical Society, Division of Petroleum Chemistry, 37, 780-785, 1992.
155. Y. Chauvin, A. Hirschauer, H. Olivier, Journal of Molecular Catalysis, 92(2), 155-165, 1994.
156. Y. Chauvin, A. Hirschauer, H. Olivier, US5750455, 1996.
157. K. Yoo, V. V. Namboodiri, R. S. Varma, P. G. Smirniotis, Journal of Catalysis, 222(2), 511-519, 2004.
158. C. P. Huang, Z. C. Liu, C. M. Xu, B. H. Chen, Y. F. Liu, Applied Catalysis A, General, 277(1-2), 41-43, 2004.
159. Y. Liu, R. Hu, C. Xu, H. Su, Applied Catalysis, A: General, 346, 189-193, 2008.
160. Z. Liu, R. U. I. Zhang, C. Xu, R. Xia, Oil & Gas Journal, 104(40), 52-56, 2006.
161. S. Aschauer, L. Schilder, W. Korth, S. Fritschi, A. Jess, Catalysis Letters, 141(10), 1405-1419, 2011.
162. Y. Chauvin, B. Gilbert, I. Guibard, Journal of the Chemical Society, Chemical Communications, 1715-1716, 1990.
163. M. Goledzinowski, V. I. Birss, J. Galuszka, Industrial & Engineering Chemistry Research, 32, 1795-1797, 1993.
164. O. Stenzel, R. Brull, U. M. Wahner, R. D. Sanderson, H. G. Raubenheimer, Journal of Molecular Catalysis A: Chemical, 192, 217-222, 2003.
168 References
165. A. a. K. Abdul-Sada, P. W. Ambler, P. K. G. Hodgson, K. R. Seddon, N. J. Stewart, WO9521871A1, 1995.
166. V. Murphy, WO2000032658A1, 2000.
167. K. Wissel, K. Halbritter, K. Massonne, V. Stegmann, DE102007040919A1, 2008.
168. M. P. Atkins, M. R. Smith, B. Ellis, EP791643A1, 1997.
169. K. D. Hope, M. S. Driver, T. V. Harris, US6395948B1, 2002.
170. K. D. Hope, D. W. Twomey, D. A. Stern, J. B. Collins, WO2003089390A2, 2003.
171. C. J. Adams, M. J. Earle, K. R. Seddon, Green Chemistry, 2, 21-24, 2000.
172. L. Xiao, K. E. Johnson, R. G. Treble, Journal of Molecular Catalysis A: Chemical, 214, 121-127, 2004.
173. D. Astruc, F. Lu, J. R. Aranzaes, Angewandte Chemie International Edition, 44, 7852-7872, 2005.
174. B. Cornils, W. A. Herrmann, I. T. Horváth, W. Leitner, S. Mecking, H. Olivier-Bourbigou, D. Vogt, Multiphase Homogeneous Catalysis, Vol. 1, Weinheim, Wiley-VCH, 3-23, 2005.
175. C. W. Kohlpaintner, R. W. Fischer, B. Cornils, Applied Catalysis, A: General, 221, 219-225, 2001.
176. E. Wiebus, R. Cornils, Chemie Ingenieur Technik, 66, 916-923, 1994.
177. C. P. Mehnert, R. A. Cook, N. C. Dispenziere, M. Afeworki, Journal of American Chemical Society, 124, 12932-12933, 2002.
178. C. P. Mehnert, E. J. Mozeleski, R. A. Cook, Chemical Communications, 3010-3011, 2002.
179. C. P. Mehnert, Chemistry – A European Journal, 11(1), 50-56, 2005.
180. A. Riisager, R. Fehrmann, S. Flicker, H. R. Van, M. Haumann, P. Wasserscheid, Angewandte Chemie, International Edition, 44, 815-819, 2005.
181. A. Riisager, P. Wasserscheid, H. R. Van, R. Fehrmann, Journal of Catalysis, 219, 452-455, 2003.
182. M. Haumann, M. Jakuttis, S. Werner, P. Wasserscheid, Journal of Catalysis, 263, 321-327, 2009.
183. M. Haumann, M. Jakuttis, R. Franke, A. Schoenweiz, P. Wasserscheid, ChemCatChem, 3, 1822-1827, 2011.
184. U. Hintermair, Z.-X. Gong, A. Serbanovic, M. J. Muldoon, C. C. Santini, D. J. Cole-Hamilton, Dalton Transactions, 39, 8501-8510, 2010.
185. M. Haumann, A. Riisager, Chemical Reviews, 108, 1474-1497, 2008.
References 169
186. J.-P. T. Mikkola, P. P. Virtanen, K. Kordas, H. Karhu, T. O. Salmi, Applied Catalysis, A: General, 328, 68-76, 2007.
187. P. Virtanen, H. Karhu, K. Kordas, J.-P. Mikkola, Chemical Engineering Science, 62, 3660-3671, 2007.
188. E. Oechsner, M. J. Schneider, C. Meyer, M. Haumann, P. Wasserscheid, Applied Catalysis. A, General, 399(1-2), 35-41, 2011.
189. U. Hintermair, T. Hoefener, T. Pullmann, G. Francio, W. Leitner, ChemCatChem, 2, 150-154, 2010.
190. S. Werner, N. Szesni, A. Bittermann, M. J. Schneider, P. Haerter, M. Haumann, P. Wasserscheid, Applied Catalysis, A: General, 377, 70-75, 2010.
191. S. Werner, N. Szesni, M. Kaiser, R. W. Fischer, M. Haumann, P. Wasserscheid, ChemCatChem, 2, 1399-1402, 2010.
192. R. Duque, E. Oechsner, H. Clavier, F. Caijo, S. P. Nolan, M. Mauduit, D. J. Cole-Hamilton, Green Chemistry, 13, 1187-1195, 2011.
193. J. Scholz, S. Loekman, N. Szesni, W. Hieringer, A. Goerling, M. Haumann, P. Wasserscheid, Advanced Synthesis & Catalysis, 353, 2701-2707, 2011.
194. S. Breitenlechner, M. Fleck, T. E. Muller, A. Suppan, Journal of Molecular Catalysis A: Chemical, 214, 175-179, 2004.
195. A. Riisager, B. Jorgensen, P. Wasserscheid, R. Fehrmann, Chemical Communications, 994-996, 2006.
196. A. Riisager, R. Fehrmann, M. Haumann, P. Wasserscheid, Topics in Catalysis, 40(1), 91-102, 2006.
197. A. Riisager, R. Fehrmann, M. Haumann, P. Wasserscheid, European Journal of Inorganic Chemistry, 2006(4), 695-706, 2006.
198. Y. Gu, G. Li, Advanced Synthesis & Catalysis, 351(6), 817-847, 2009.
199. C. Van Doorslaer, J. Wahlen, P. Mertens, K. Binnemans, D. De Vos, Dalton Transactions, 39(36), 8377-8390, 2010.
200. M. Haumann, M. Jakuttis, H. Maier, H. Breitzke, G. Buntkowsky, A. Schönweiz, 43. Jahrestreffen Deutscher Katalytiker. Weimar, 2010.
201. J. Lemus, J. Palomar, M. Gilarranz, J. Rodriguez, Adsorption, 17(3), 561-571, 2011.
202. R. Knapp, A. Jentys, J. A. Lercher, Green Chemistry, 11(5), 656-661, 2009.
203. A. Jess, C. Kern, W. Korth, OIL GAS European Magazine, 38-45, 2012.
204. C. Decastro, E. Sauvage, M. H. Valkenberg, W. F. Hölderich, Journal of Catalysis, 196(1), 86-94, 2000.
205. M. H. Valkenberg, C. C. De, W. F. Hölderich, Green Chemistry, 4, 88-93, 2002.
170 References
206. J. Joni, M. Haumann, P. Wasserscheid, Advanced Synthesis & Catalysis, 351, 423-431, 2009.
207. J. Joni, M. Haumann, P. Wasserscheid, Applied Catalysis, A: General, 372(1), 8-15, 2010.
208. U. Kernchen, B. Etzold, W. Korth, A. Jess, Chemical Engineering & Technology, 30(8), 985-994, 2007.
209. J. R. Arras, E. Paki, C. Roth, J. R. Radnik, M. Lucas, P. Claus, The Journal of Physical Chemistry C, 114(23), 10520-10526, 2010.
210. H. P. Steinrück, J. Libuda, P. Wasserscheid, T. Cremer, C. Kolbeck, M. Laurin, F. Maier, M. Sobota, P. S. Schulz, M. Stark, Advanced Materials, 23(22-23), 2571-2587, 2011.
211. J. Arras, M. Steffan, Y. Shayeghi, D. Ruppert, P. Claus, Green Chemistry, 11(5), 716-723, 2009.
212. A. Jess, W. Korth, B. Etzold, DE102006019460A1, 2007.
213. J. Arras, M. Steffan, Y. Shayeghi, P. Claus, Chemical Communications, 4058-4060, 2008.
214. P. J. Dyson, G. Laurenczy, C. A. Ohlin, J. Vallance, T. Welton, Chemical Communications, 2418-2419, 2003.
215. N. Wörz, J. Arras, P. Claus, Applied Catalysis, A: General, 391(1-2), 319-324, 2011.
216. J. Arras, D. Ruppert, P. Claus, Applied Catalysis A: General, 371(1-2), 73-77, 2009.
217. M. Sobota, M. Happel, M. Amende, N. Paape, P. Wasserscheid, M. Laurin, J. Libuda, Advanced Materials, 23, 2617-2621, 2011.
218. P. Migowski, J. Dupont, Chemistry - A European Journal, 13, 32-39, 2007.
219. D. Astruc, Nanoparticles and Catalysis, Transition-metal Nanoparticle Catalysis in Imidazolium Ionic Liquids, ed. Dupont, J.,Silva, D. D. O., Weinheim, WILEY-VCH, 195-218, 2008.
220. M. Antonietti, D. Kuang, B. Smarsly, Y. Zhou, Angewandte Chemie International Edition, 43, 4988-4992, 2004.
221. Y. Gu, G. Li, Advanced Synthesis & Catalysis, 351(6), 817-847, 2009.
222. J. D. Scholten, B. C. Leal, J. Dupont, ACS Catalysis, 2, 184-200, 2012.
223. H. Hagiwara, H. Sasaki, N. Tsubokawa, T. Hoshi, T. Suzuki, T. Tsuda, S. Kuwabata, Synlett, 1990-1996, 2010.
224. M. Gruttadauria, L. F. Liotta, A. M. P. Salvo, F. Giacalone, P. V. La, C. Aprile, R. Noto, Advanced Synthesis & Catalysis, 353, 2119-2130, 2011.
References 171
225. S. Ernst, C. Bischof, M. Hartmann, J. Weitkamp, Proceedings of the World Petroleum Congress, 16th, 27-37, 2000.
226. A. De Lucas, P. Sánchez, F. Dorado, M. J. Ramos, J. L. Valverde, Applied Catalysis, A: General, 294(2), 215-225, 2005.
227. A. Fúnez, J. W. Thybaut, G. B. Marin, P. Sánchez, A. De Lucas, J. L. Valverde, Applied Catalysis, A: General, 349(1-2), 29-39, 2008.
228. D. Verboekend, K. Thomas, M. Milina, S. Mitchell, J. Perez-Ramirez, J.-P. Gilson, Catalysis Science & Technology, 1, 1331-1335, 2011.
229. J. Xu, J. Y. Ying, Angewandte Chemie, International Edition, 45, 6700-6704, 2006.
230. M. Busto, K. Shimizu, C. R. Vera, J. M. Grau, C. L. Pieck, M. A. D'amato, M. T. Causa, M. Tovar, Applied Catalysis, A: General, 348, 173-182, 2008.
231. J. M. Grau, C. R. Vera, V. M. Benitez, J. C. Yori, Energy & Fuels, 22(3), 1680-1686, 2008.
232. H. Al-Kandari, F. Al-Kharafi, A. Katrib, Applied Catalysis, A: General, 383(1-2), 141-148, 2010.
233. J. C. Yori, J. M. Grau, V. M. Benítez, J. Sepúlveda, Applied Catalysis, A: General, 286(1), 71-78, 2005.
234. V. M. Akhmedov, S. H. Al Khowaiter, Catalysis Reviews, 49(1), 33-139, 2007.
235. V. T. Vladimirovna, K. L. Modestovich, K. Vladislav, Z. Y. Egorovich, J. Houzvicka, J. Zavilla, EP1310472A1, 2003.
236. K. Herbst, J. Houzvicka, J. B. Tofte, J. Zavilla, EP Patent 1,346,768, 2003.
237. J. Houzvicka, J. Zavilla, K. Herbst, US20040059173A1, 2004.
238. J. Zavilla, K. Herbst, J. Houzvicka, S. I. Hommeltoft, EP1402950A1, 2004.
239. S. Tschirschwitz, S. Deuerlein, J. Buerkle, M. Schmitt, S. Oehlenschlaeger, K. Wissel-Stoll, V. Wloka, US20110137098A1, 2011.
240. G. A. Olah, J. M. Bollinger, C. A. Cupas, J. Lukas, Journal of American Chemical Society, 89, 2692-2694, 1967.
241. V. A. Ksenofontov, T. V. Vasina, Y. E. Zubarev, L. M. Kustov, Reaction Kinetics and Catalysis Letters, 80, 329-335, 2003.
242. J. Abbot, Journal of Catalysis, 123, 383-395, 1990.
243. A. L. Patterson, Physical Review, 56, 978-982, 1939.
244. D. Schmitt, Unpublished PhD thesis, Lehrstuhl für Chemische Reaktionstechnik, Friedrich-Alexander Universität Erlangen-Nürnberg, 2012.
245. Y. Ono, T. Tanabe, N. Kitajima, Journal of Catalysis, 56(1), 47-51, 1979.
172 References
246. G. A. Olah, P. Schilling, J. S. Staral, Y. Halpern, J. A. Olah, Journal of the American Chemical Society, 97(23), 6807-6810, 1975.
247. L. Schmerling, J. A. Vesely, The Journal of Organic Chemistry, 38(2), 312-315, 1973.
248. J. Messner, Unpublished PhD thesis, Lehrstuhl für Chemische Reaktionstechnik, Friedrich-Alexander Universität Erlangen-Nürnberg, 2012.
249. K. S. W. Sing, D. H. Everett, R. a. W. Haul, L. Moscou, R. A. Pierotti, J. Rouquerol, T. Siemieniewska, Pure and Applied Chemistry, 57, 603-619, 1985.
250. K. Griesbaum, A. Behr, D. Biedenkapp, H.-W. Voges, D. Garbe, C. Paetz, G. Collin, D. Mayer, H. Höke, Hydrocarbons, in Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA, 2000.
251. Chemical Rubber Company, CRC Handbook of Chemistry and Physics, Boca Raton, Fla., CRC Press, Taylor & Francis Group, 2007.
252. E. G. Hancock, Critical Reports on Applied Chemistry, 10, 20-56, 1985.
253. M. Eichmann, PhD thesis, Institut für Technische Chemie und Petrolchemie, Rheinisch-Westfälische Technische Hochschule Aachen, 1999.
254. K. Christmann, G. Ertl, T. Pignet, Surface Science, 54(2), 365-392, 1976.
255. R. R. Deshmukh, J. W. Lee, U. S. Shin, J. Y. Lee, C. E. Song, Angewandte Chemie, International Edition, 47, 8615-8617, 2008.