Flame-Made Catalytic Materials For Alkene PDF

Diss. ETH No.
23621
Flame-made Catalytic Materials for Alkene

Production
A thesis submitted to attain the degree of

DOCTOR OF SCIENCES of ETH ZURICH
(Dr. sc. ETH Zurich)
presented by
RAJESH KOIRALA
MSc, University of Cincinnati
Born on 17.05.1984
citizen of Nepal
accepted on the recommendation of

Prof. Dr. Sotiris E. Pratsinis, examiner
Prof. Dr. Alfons Baiker, co-examiner
2016
i
Acknowledgements
I would like to express my deepest gratitude to Prof. Dr. Sotiris E. Pratsinis for
providing the opportunity to perform Ph.D. in his lab and his continuous encouragement,
guidance, support and advice during this period. I am extremely grateful to Prof. Dr. Alfons
Baiker for his tireless scientific support and guidance, many inspiring and helpful discussions
and for agreeing to co-advise this thesis.
I am grateful to my past and present colleagues and all members of the Particle
Technology Laboratory (PTL) at ETH Zurich, where this work was carried out, for the
motivating and friendly working atmosphere. Special thanks go to Dr. Robert Bchel who
mentored me at the start of my Ph.D. and continuously help me both scientifically and
technically throughout my stay at PTL. Moreover, I am thankful for his help in translating the
abstract from english to german and Christoph for proof reading it. Many thanks go to Dr.
Karsten Wegner for several discussions related to the flame synthesis of nanoparticles. I wish
to acknowledge Rene Plss from the IPE workshop for solving technical challenges in the lab,
Agnes Gntner for her indispensable bureaucratic work and organizing various social events
and Dr. Frank Krumeich from the Electron Microscopy Center, EMEZ, of ETH Zurich for
providing microscopy images.
Many thanks go to the students who performed their projects and worked as an assistant
with me: Athanasia Tsoukalou, Edouard Gence, Sotiria Mostrou, Randika Sayakkarage,
Davide Stucchi and Simon Spiekermann. Many thanks go to all my amazing friends who
continuously surrounded and supported me during this journey and helped me to develop both
personally and scientifically.
Financial support by ETH Research Grant (ETH-39-12-2) is gratefully acknowledged.
Finally, I would like to express my deepest gratitude to my parents (Ram Babu and Bishnu
Kumari) and brother (Yogesh) for their love, support and faith in me. And last but not least I
would like to especially thank my wife Manasa for providing endless support, joy and
motivation in my life and numerous stimulating scientific discussions as a fellow researcher.
ii Acknowledgements
iii
Contents
Acknowledgements i
Contents iii
Summary vii
Zusammenfassung xi
1. Synthesis of Catalytic Materials in Flames: Opportunities and Challenges 1
1.1 Introduction 2
1.2 Experimental conditions of flame synthesis and their influence
on material properties 4
1.2.1 Particle formation mechanism 4
1.2.2 Reactor configurations 7
1.2.3 Flame parameters for catalyst synthesis 8
1.2.4 Temperature 13
1.3 Particle properties relevant for catalysis 14
1.3.1 Chemical composition 14
1.3.2 Multi-metal systems 16
1.3.3 Specific surface area 20
1.3.4 Particle size 22
1.3.5 Crystal structure and morphology 24
1.3.6 Spatial distribution of components 26
1.3.7 Coating of ceramic supports 30
1.3.8 Scale-up 32
1.4 Concluding Remarks 34
1.5 Ethene production from catatlytic conversion of natural gas 35
1.5.1 Overview 35
1.5.2 Oxidative Coupling of Methane (OCM) 36
1.5.3 Oxidative Dehydrogenation of Ethane (ODHE) 37
1.5.4 Summary and outlook 38
1.6 Acknowledgement 39
1.7 References 39
2. Oxidative Coupling of Methane on Flame-made Mn-Na2WO4/SiO2:
Influence of Catalyst Composition and Reaction Conditions 45
2.1 Introduction 47
2.2 Experimental 48
iv Contents
2.2.1 Catalyst preparation 48

2.2.2 Catalyst characterization 49
2.2.3 Catalyst tests 49
2.3 Results and Discussion 50
2.3.1 Structural properties of catalysts 50
2.3.2 Catalytic tests 57
2.3.2.1 Effect of Mn, Na2WO4 and SiO2 content in OCM 57
2.3.2.2 Influence of reaction conditions 62
2.4 Conclusions 70
2.5 Acknowledgements 70
2.6 References 71
3. Oxidative Dehydrogenation of Ethane with CO2 over Flame-made
Ga-loaded TiO2 75
3.1 Introduction 77
3.2 Experimental 79
3.2.2 Characterization 79
3.2.3 Catalytic studies 81
3.3.1 Textural and structural properties 82
3.3.2 Catalytic performance 91
3.3.3 Influence of reaction conditions on performance of
optimal catalyst 94
3.3.4 Carbon deposition and catalyst regeneration 100
3.4 Conclusions 104
3.6 References 105
4. Influence of Support in CO2-assisted Cobalt-Catalyzed Oxidative
Dehydrogenation of Ethane 109
4.1 Introduction 110
4.2 Experimental 111
4.2.3 Catalyt testing 113
v

4.3.1 Structural and chemical properties of catalysts 114
4.3.1.1 Crystallinity, surface area and morphology 114
4.3.1.2 NH3-TPD and DRIFTS 118
4.3.1.3 Raman 122
4.3.1.4 UV-Vis reflectance and reducibility 124
4.3.1.5 XPS 128
4.3.2 Catalytic performance 129
4.3.2.1 Identified products and relevant reactions observed 129
4.3.2.2 Influence of single and mixed oxides supports 130
4.3.2.3 Performance of SiO2-supported catalyst at
different reaction conditions 136
4.3.2.4 Stability of the catalyst 138
4.4 Conclusions 140
4.6 References 141
5. Intimate Cobalt-Silica Interaction: an Important Parameter for
Oxidative Dehydrogenation of Ethane 145
5.1 Introduction 146
5.2 Experimental 147
5.2.3 Catalyt testing 148
5.3.1 Structural and chemical properties of catalysts 149
5.3.2 Catalytic activity 156
5.4 Conclusions 159
5.6 References 160
6. Outlook and Research Recommendations 163
6.1 References 166
A. Supporting Information: Influence of Support in CO2-assisted
Cobalt-Catalyzed Oxidative Dehydrogenation of Ethane 167
A.1 Physical characterizations 167
vi Contents
A.2 Catalytic results 168

A.3 References 174
B. Supporting information: Intimate Cobalt-Silica Interaction: an
Important Parameter for Oxidative Dehydrogenation of Ethane 175
B.1 Catalyst characterization 175
B.2 Catalytic results 180
Curriculum Vitae 181
Publications and Presentations 183
vii
Summary
Recent discovery of vast reserves of natural gas has renewed interests from both industries
and academia to utilize this resource for the production of various chemicals. Such
technologies reduce reliance on a dwindling supply of crude oil and energy-intensive
processes applied for high value chemical production. Though there exist several research
works investigating the conversion of relatively inert methane and ethane (major components
of natural gas) to valuable chemicals such as ethene, successful commercialization has not yet
been realized. An important requirement is the development of highly active catalysts that can
accelerate these reactions. Generally, mixed metal oxides-based catalysts are used that are
typically synthesized using classical wet-methods such as wet-impregnation, sol-gel and co-
precipitation. Such synthesis procedures need multiple steps requiring, in some cases, many
days to obtain the final catalyst product. Additionally, precise control of the catalyst
composition and structure is usually difficult. Alternative to this, flame spray pyrolysis (FSP),
produces multi-component catalysts in single step with tailored physicochemical properties
and high reproducibility. Moreover, it is a proven scalable technique which is not the case for
most other synthesis techniques. Therefore, in this work FSP was utilized to develop novel
catalysts to catalyze oxidative coupling of methane (OCM) and oxidative dehydrogenation of
ethane (ODHE) reactions to produce ethene. Moreover, extensive characterization techniques
were applied to elucidate the catalysts structure-activity relationships which will in turn
enable the rational design of highly active catalysts.
In Chapter 1, an overview of a flame aerosol synthesis process with special focus on
the control of various physicochemical properties of the catalysts and their influence on the
catalytic activity is presented. Particle formation mechanisms, limitations and capabilities of
different burner configurations such as vapor-fed aerosol flame synthesis, flame-
assisted/flame spray pyrolysis in catalytic material synthesis are discussed. Particularly, ease
of controlling the particle properties which have major effect on the catalytic activity e.g.
particle size, crystal structure and surface area is highlighted. Potential and limitations of the
flame aerosol technique for the synthesis of wide range of materials, which finds applications
beyond catalysis, are also assessed. Finally, recent growth in the natural gas resources and
viii Summary
how its major components (e.g. methane and ethane) can be utilized in the efficient
production of ethene, through novel catalytic processes such as oxidative coupling of methane
(OCM) and oxidative dehydrogenation of ethane (ODHE), is briefly discussed.
In Chapter 2, the influence of Mn-Na2WO4/SiO2 composition on the OCM reaction
was investigated. All the as-prepared catalysts were found to be non-porous, XRD amorphous
and exhibiting relatively high specific surface area. All the components were found to be
essential to achieve high activity and through this, optimal composition was determined to be
1.9%Mn-3%Na2WO4/SiO2. Synergism between these elements was a crucial factor dictating
the catalytic reaction. Structural changes did not influence the activity, thus indicating the
structural insensitivity of the reaction. Increasing Na2WO4 loadings (> 3 wt%) exerted
negative effect on the activity, thus, highlighting the importance of surface species
composition and their distribution. In addition, higher activity of the catalyst obtained from
flame synthesis compared to that from wet-method at lower Na2WO4 loadings (~ 3 wt%)
further confirms the crucial role of the synthesis method in controlling the elemental
distribution.
In Chapter 3, a wide range of (0-17 wt%) gallium-loaded TiO2-supported catalysts
were produced using FSP and evaluated for ODHE where CO2 was employed as an oxidant.
Specific surface area of the catalysts remained virtually constant but gallium induced TiO2
rutilization with increasing gallium loadings. Gallium oxide was highly dispersed and in
amorphous form in all the catalysts. Increase in ethane conversion with increasing gallium
content indicated its active participation in the reaction. However, decrease in ethene
selectivity with increasing ethane conversion was observed. Highest yield was obtained over
10 wt% gallium-loaded TiO2. At gallium loadings > 10 wt% the activity decreased, primarily
due to rapid coke deposition. High CO2 conversion was observed and followed a closely
similar trend as that of ethane conversion, thus indicating its active participation during the
reaction. Coke removal and reverse water gas shift reaction are some of the speculated tasks
of CO2 during the reaction. Based on the observations, the underlying reaction pathways are
proposed. Amount of coke deposits seemed to be directly proportional to the gallium loadings
which indicated that in addition to the ODHE it also promotes coking during the reaction.
Catalyst that minimizes the coke deposition while exhibiting and maintaining high
activity was further sought by initiating the investigation over (4.5 wt%) cobalt-loaded
various single and mixed-metal oxide supports (Chapter 4). Catalytic activity was highly
influenced by the support. Significant differences in the reducibility and acidity of these
catalysts were evidenced by various characterization techniques. Similar catalytic activity was
ix
observed over SiO2-, ZrO2- and TiO2-ZrO2-supported catalysts which were higher than other
single and mixed metal oxides supported catalyst under the standard reaction conditions.
Further catalytic investigation at different contact times revealed that the performance of the
SiO2-supported catalyst is superior. Correlating the catalysts activity and physicochemical
properties, it appeared that the surface acidity plays major role in SiO 2-supported catalyst
whereas CoOx redox properties dictated the activities of ZrO2- and TiO2-ZrO2-supported
counterparts.
In Chapter 5, the influence of the cobalt loadings (0.1-4.5 wt%) on the activity of
SiO2-supported catalysts during ODHE reaction is presented. With increase in the cobalt
content, ethane conversion increased until 0.75 wt% loading and then decreased upon further
increase. The optimal activity was nearly 1.5 times higher than that obtained from the catalyst
containing 4.5 wt% cobalt. Cobalt-silicate structure where Co2+ is likely tetrahedrally
coordinated in SiO2 matrix was found to be crucial for achieving high activity. This structure
seemed to be fully developed at the cobalt loading of 0.75 wt% and further addition led to the
formation of smaller CoOx clusters, which lowered the catalytic activity. Based on the results,
it was speculated that the reaction proceeds via non-redox pathways and that the presence of
optimal tetrahedrally coordinated Co2+ in SiO2 matrix (i.e. cobalt silicate) is essential to
realize high activity.
This thesis showcases the capability of producing wide range of catalytic materials,
using flame spray pyrolysis technique, which can be utilized towards ethene production from
methane and ethane, the major components of natural gas. Moreover, through extensive
characterizations and parametric sensitivity testing, several descriptors have been identified
that will assist in the rational design of highly active catalysts.
x Summary
xi
Zusammenfassung
Die krzlich entdeckten riesigen Erdgasvorkommen sind von grossem Interesse fr Industrie
und Forschung, da es zur Produktion von verschiedenen Chemikalien genutzt werden kann.
Die Nutzung von Erdgas ermglicht die Abhngigkeit von Erdl reduzieren, was Sinnvoll ist
dadie Erdlfrdermengen rcklufig sind und die Gewinnung von wertvollen Chemikalien
aus Erld typischerweise sehr Energieintensiv sind. Obwohl die Umwandlung von relativ
inaktiven Methan (die Hauptkomponente von Erdgas) zu wertvollerem Ethan bereits in der
Forschung bekannt ist, werden diese Prozesse bis jetzt noch nicht erfolgreich umgesetzt. Eine
wichtige Voraussetzung ist die Entwicklung von sehr aktiven Katalysatoren, welche die
Umwandlungsreaktionen (d.h. die Umwandlung von Methan zu Ethan) beschleunigen. Bis
heute werden oft Metalloxid-basierende Katalysatoren typischerweise mit klassischen nassen
Synthesemethoden hergestellt wie zum Beispiel mit der Nass-Imprgnation, dem Sol-Gel-
Verfahren oder der Mitfllungsmethode. Die genannten Prozesse bentigen mehrere Schritte
und dauern oft mehrere Tage bis das Endprodukt hergestellt ist. Zudem ist die genaue
Katalysatorenzusammensetzung und -struktur oft schwierig zu kontrollieren. Im Gegensatz zu
diesen nassen Synthesemethoden, kann die Flammen Sprh Pyrolyse (FSP) solche
Katalysatoren in einem Schritt herstellen und dabei die Zusammensetzung sowie die
gewnschten physikalisch-chemisch Eigenschaften reproduzierbar einstellen. Zustzlich ist
die Produktion nachweislich skalierbar, was fr andere Syntheseverfahren oft nicht der Fall
ist. Aus diesen Grnden wurde in dieser Arbeit FSP benutzt um neue Katalysatoren zu
entwickeln fr die Oxidative-Koppelung-von-Methan (OCM) und fr die Oxidative-
Dehydrierung-von-Ethan (ODHE) zur Herstellung von Ethen. Die Struktur der Katalysatoren
wurde intensiv untersucht um ein Zusammenhang zwischen der Aktivitt und den
Katalysatoreigenschaften zu schaffen, um so hoch aktive Katalysator zu entwerfen.
Kapitel 1 zeigt eine bersicht der Herstellungsprozesse von in Flammen hergestellten
Partikel mit Fokus auf der Kontrolle von verschiedenen physikalisch-chemisch Eigenschaften
der Katalysatoren und deren Einfluss auf die katalytische Aktivitt. Die
Entstehungsmechanismen von Partikel wird in dieser Arbeit diskutiert und die Grenzen und
Kapazitten von verschiedenen Brennerauslegung zur Herstellung von Katalysatoren werden
vorgestellt wie zum Beispiel die Dampf-Flammen Synthese und die Flammen untersttzte
xii Zusammenfassung
FSP. Im Speziellen wird die einfache Kontrolle der Partikeleigenschaften hervorgehoben,

welche einen wichtigen Effekt auf die katalytische Aktivitt der Partikel haben wie z. B.
Partikelgrsse, -kristallstruktur und die -oberflche. Mgliche Anwendungen und die Grenzen
der Flammen Sprh Pyrolyse wurden untersucht, auch fr Anwendungen ausserhalb der
Katalyse. Schlussendlich wurde die Mglichkeit der Nutzung und der effizienten Produktion
von Ethen durch moderne katalytische Prozesse wie der OCM und der ODHE kurz
besprochen.
Im Kapitel 2 wird der Einfluss der Zusammensetzung von Mn-Na2WO4/SiO2
untersucht fr die OCM Reaktion. Alle unbehandelten Katalysatoren waren nicht pors,
amorph (laut XRD) und hatten eine hohe spezifische Oberflche. Die optimale
Zusammensetzung ist 1.9%Mn-3%Na2WO4/SiO2 wobei alle genannten Komponenten fr die
hohe Aktivitt erforderlich sind. Ihr Zusammenwirken bestimmt die katalytische Reaktion.
nderungen in der physikalische Struktur hatte keinen Einfluss auf die Aktivitt und somit ist
sie von sekundrer Wichtigkeit. Eine Erhhung der Na2WO4 Beladung (> 3 Gewichts%) hatte
einen negativen Effekt auf die Aktivitt, was die die Wichtigkeit der Oberflchenatome, deren
Zusammensetzung und Verteilung zum Vorscheinbringt. Zustzlich haben die in der Flamme
hergestellten Partikel eine hhere Aktivitt als Nasschemische, weil die Elemente viel besser
auf dem Trger verteilt sind. Dies gilt trotz des tieferen Na2WO4 Anteils (~ 3 Gewichts%) der
nasschemisch hergestellten Partikel.
Kapitel 3 prsentiert ein breites Spektrum von Gallium beladenen (0 17 Gewichts%)
TiO2 Katalysatoren. Sie wurden mit FSP hergestellt und als ODHE getestet mit CO2 als
Oxidationsmittel. Die spezifische Oberflche war nahezu konstant, Gallium induzierte
hauptschlich eine nderung in der Kristallstruktur von TiO2 von Anatas zu Rutil, wobei der
Rutilgehalt von der Ga Beladung abhing. Auf allen Katalysatoren war das Galliumoxid gut
verteilt und amorph. Die Ethanumsetzung stieg mit hherer Ga Beladung, ein Anzeichen fr
die aktive Teilnahme von Ga an der Reaktion. Die Ethen Selektivitt reduzierte sich mit
hheren Umsatzraten. Die hchste Ausbeute wurde mit 10 Gewichts% Ga auf TiO2 erreicht.
Bei hherer Ga Beladung >10 Gewichts% reduzierte sich die Aktivitt, weil sich Russ auf den
Partikeln absetzte. Ein hoher CO2 Umsatz wurde beobachtet und dieser Verlauf war
vergleichbar mit dem Ethanumsatz, was besttigt, dass CO2 aktiv an der Reaktion teil nimmt.
Mgliche Rollen von CO2 sind die Russentfernung und die Wassergas-Shift-Reaktion. Auf
den Beobachtungen basierend wurde ein mglicher Reaktionsmechanismen vorgeschlagen.
Da die Russablagerungen proportional zur Ga Beladung ist, knnte das Ga nicht nur die
ODHE sondern auch die Russbildung in der Reaktion frdern.
xiii
Verschiedene Trgermaterialien mit 4.5 Gewichts% Co wurden alsKatalysatoren mit

minimaler Russablagerung und gleichzeitig hhereren Umsatzraten untersucht (Kapitel 4).
Die Aktivitt war sehr vom Trger abhngig. Aussagekrftige Unterschiede in
Reduzierbarkeit und Aktivitt dieser Katalysatoren wurden durch unterschiedliche
Analyseverfahren erklrt. hnliche Umstze unter einer Standardkondition wurden bei den
Trgermaterialen SiO2, ZrO2 und TiO2-ZrO2 beobachtet. Ihr Umsatz war markant hher als
bei andern Materialien unabhngig ob Einkristalle oder Mischoxyden. Weitergehende
katalytische Untersuchen bei unterschiedlichen Kontaktzeiten zeigten, dass SiO2 als
Trgermaterial den Anderen berlegen ist. Die saure Oberflchenchemie ist im Fall von SiO2
die ausschlagende Eigenschaft, wobei die Redox Eigenschaft von CoOx die tragende Rolle bei
ZrO2 und TiO2-ZrO2 spielt.
In Kapitel 5 wird der Einfluss der Kobalt Beladung (0.1 -4.5 Gewichts%) auf SiO2 fr
die ODHE Reaktion vorgestellt. Mit hherem Kobaltgehalt erhhte sich die
Ethanumwandlung bis zu einer Beladung von 0.75 Gewichts%. Bei hherer Co Beladungen
sank die Ethanumwandlung wieder. Die optimale Aktivitt war beinahe 1.5 mal hher als von
quivalenten Katalysatoren mit 4.5 Gewichts% Co. Fr die Aktivitt war die tetraedische
Koordination von Co2+ in der SiO2 Matrix ausschlaggebend. Bei einer Beladung von 0.75
Gewichts% war diese Struktur vollstndig ausgebildet. Eine hhere Kobalt Beladung fhrte
zur Entstehung von kleinen CoOx Partikel, welche die katalytische Aktivitt reduierte.
Aufgrund dieser Beobachtung kann man Redox Reaktionen ausschliessen.
Diese Arbeit zeigt die grosse Kapazitt der Flammen Sprh Technologie fr die
Herstellung einer Vielzahl von Katalysatoren, die fr die Umwandlung von Ethen aus den
Hauptkomponenten von Erdgas, Methan und Ethan, genutzt werden knnen. Die genauen und
vielfltigen Charakterisierungen der analysierten Pulver sowie die genauen Untersuchungen
der Parameter von denen die Leistung der Katalysatoren abhngt, helfen bei der Auslegung
von aktiven Katalysatoren.
xiv Zusammenfassung
1
Chapter 1
1. Synthesis of Catalytic Materials in Flames:

Opportunities and Challenges
Abstract
The proven capacity of flame aerosol technology for rapid and scalable synthesis of
functional nanoparticles makes it ideal for the manufacture of an array of
heterogeneous catalysts. Capitalizing on the high temperature environment, rapid
cooling and intimate component mixing at either atomic or nano scale, novel catalysts
with unique physicochemical properties have been made using flame processes. This
tutorial review covers the main features of flame synthesis and illustrates how the
physical and chemical properties of as-synthesized solid catalytic materials can be
controlled by proper choice of the process parameters. Gas phase particle formation
mechanisms and the effect of synthesis conditions (reactor configuration, precursor
and dispersion gas flow rates, temperature and concentration fields) on the structural,
chemical and catalytic properties of as-prepared materials are discussed. Finally,
opportunities and challenges offered by flame synthesis of catalytic materials are
addressed.
Part of it is published in Chemical Society Reviews, 45 (2016), 3053-3068.

2 Chapter 1
1.1 Introduction
Flame aerosol technology for synthesis of nanoparticles, e.g. carbon black, dates back
to the beginning of human civilization as can be seen in the remnants of numerous art
works. Ancient Chinese were among the first to produce them in larger quantity.1
Industrial scale production of carbon black started with the discovery of its use in
reinforcing rubber, thus drastically increasing its demand especially by the motor
vehicle industry.1 So carbon black is the largest by volume and value flame-made
commodity. Nowadays, using this technology a few other commodities are produced
such as optical waveguides, pigmentary TiO2, fumed SiO2, Al2O3 and other oxides1
and even specialty chemicals (e.g. nanofluids). Due to its versatility and ease of
materials production, flame synthesis is displacing older technologies2, e.g. the
chloride versus the sulfate process for pigmentary TiO2. Industries like Evonik,
Cabot, Dupont etc. produce in large scale ceramic powders in flames indicating the
scale-up potential of this technology for a wide range of materials production.2
To start with, the widely-used photocatalytic material, nano-TiO2 P25 by
Evonik, is made by oxy-hydrogen flames.2 In addition to metal oxides, bi- or multi-
metal oxides as well as oxide-supported noble metals that are of prime interest in
catalysis, such as ZrO2, Fe2O3, ZnO, SnO2, TiO2/SiO2, SiO2/Al2O3, Pt/TiO2, LaBO3
(B = Co, Mn, Fe) and Pd/La2O3/Al2O3 have been made already using flame aerosol
technology.3 Flame methods offer a couple of attractive features that are not
achievable with classical wet-chemistry methods. A detailed overview of classical
catalyst preparation methods such as co-precipitation, impregnation, sol-gel and
hydrothermal syntheses is provided in an all-embracing review.4 As a representative
example, Figure 1.1 compares the main steps involved in a classical wet-chemistry
preparation (co-precipitation) and flame synthesis of a catalyst. A striking difference
between the classical preparation and flame aerosol synthesis is the number of steps
involved. Wet-chemistry methods generally consist of various time consuming steps,
while the flame methods facilitate rapid single step synthesis. In situ calcination
during the high temperature production alleviates the need of post thermal treatment.
Moreover, flame-made catalysts require no solvent-intensive washing that increases
the chance of altering the catalyst composition due to leaching of components.
Additionally, flame synthesis affords continuous production, while wet-chemistry
routes are usually batch processes. Rapid quenching of the flame-made particles gives
3
access to the formation of metastable phases and thus catalytic materials with
distinctly different properties compared to wet-chemistry made materials can be
produced. Flame aerosol technology is highly versatile as it can easily tune catalyst
characteristics such as specific surface area (SSA), particle size and crystallinity
through its process parameters.5 Moreover, spatial and also preferential deposition of
the active materials can be achieved by multiple flames in a single step.6 Some
restraints for the production of catalysts by flame methods arise from the fact that
suitable precursors are relative expensive and highly crystalline and porous materials
are difficult to produce. Moreover, not all precursors are easily mixed, explosive
precursors mixtures and conditions should be considered and in some cases products
of incomplete combustion could be present.
Applications of flame-made nanomaterials can be found in catalysis, sensors,
biomaterials, ceramics, composites and bioimaging.7 In catalysis, Strobel et al.
3
reviewed the use of flame-made catalysts for photocatalysis, epoxidation, selective
catalytic reduction of NOx and CH4 combustion. A follow-up review by Schimmoeller
et al.8 showed the fast growing production and utilization of flame-made catalysts
focusing on V2O5-based catalysts and their structural and chemical properties in
comparison to those prepared by classical wet methods. The growing interest in flame
aerosol synthesis of catalysts can be ascribed primarily to its flexibility, speed and
scalability.
The purpose of this tutorial review is to provide a holistic view on flame
methods and their potential for controlling physical and chemical characteristics (e.g.
surface area, particle size, crystal structure, chemical and phase composition etc.) of
catalytic materials. We will start with a brief description of particle formation
mechanisms in different flame reactors. Later on, the influence of synthesis conditions
on the physicochemical properties of as-prepared materials and their effect on the
catalyst performance is illustrated using various examples. At last, future
opportunities offered by flame techniques for the preparation of sophisticated
materials for catalysis will be discussed.
4 Chapter 1
Figure 1.1: Comparison of flame synthesis with classical wet-preparation method

(co-precipitation) of catalysts. Ready-to-use catalysts can be produced in a single step
using flame aerosol technology (red bracket), whereas wet-methods (broken blue
bracket) often require multiple steps, sometimes taking days, to produce them.
Depending on the application (reactor type) a possible forming step (e.g. extrusion,
granulation, pelletizing) at the end of both preparation routes may be necessary,
which has not been indicated in the schemes. Adapted from ref. 3 with permission
from Elsevier.
1.2 Experimental conditions of flame synthesis and their

influence on material properties
1.2.1 Particle formation mechanism
During flame aerosol synthesis, particle formation follows two main routes: droplet-
to-particle and gas-to-particle2 conversion resulting in particles by top-down or
bottom-up processes, respectively. Depending on the state of the metal precursor, we
distinguish vapor-fed and liquid-fed flame synthesis processes (Figure 1.2a).3 Vapor-
5
fed flame processes lead to particles solely by gas-to-particle conversion while liquid-
fed ones may involve particle formation by both routes.
In vapor-fed aerosol flame synthesis (VAFS), gaseous precursors (e.g. TiCl4,
SiH4 vapors) are fed to the flame resulting in product particles by nucleation, surface
growth and/or condensation that grow further by coagulation-agglomeration. Such
particles are aggregates (chemically-bonded primary particles) and agglomerates
(physically-bonded primary particles). Their state depends on material properties and
residence time in the flame.9 This is an industrially applied process for the synthesis
of photocatalytic TiO2 and catalyst supports (SiO2, Al2O3 etc.) as well as pigmentary
TiO2 and fumed silica, alumina etc. In multicomponent catalysts, metal oxides
(refractories) with high melting point condense out first while other
components/species (noble metals or soft oxides) precipitate or form on them
according to their melting temperature (e.g. V2O5 on TiO2 for NOx reduction)10 or
surface reactions. Supported catalysts such as Cu/ZnO/Al2O3 for methanol synthesis,
Pt/TiO2 for SO2 oxidation and TiO2/SiO2 for epoxidation of cyclohexenol have been
already produced via this technique.3
6 Chapter 1
(a) (b)
Figure 1.2: Schematic of (a) possible particle formation pathways during the aerosol synthesis and (b) FSP setup for nanoparticles synthesis
and a picture of the flame during nanoparticle production. Large or hollow particles are formed during FASP synthesis, whereas only partly in
FSP. Mainly, smaller particles are formed from FSP and VAFS where precursor droplets are transformed to vapor which after combustion,
nucleation, surface growth and condensation produce nanoparticles. Metal oxide nucleation and solid solution formation is governed by their
melting temperatures. Adapted with permission from ref. 3.
7
In liquid-fed flames we distinguish if the supporting fuel is mixed with the

metal precursor (flame spray pyrolysis, FSP)11 or it is provided separately (flame-
assisted spray pyrolysis, FASP)12. In both, the metal precursors are sprayed into fine
droplets that should evaporate to precursor vapor for nanoparticle synthesis by gas-to-
particle conversion. In FSP13, the precursor droplets contain the fuel that evaporates
and burns to drive particle formation. This phenomenon greatly facilitates the one-
step synthesis of noble metal clusters on catalytic supports, e.g. Pt/Al 2O3 for
enantioselective hydrogenation reaction.14 Due to its compact and easy operation, the
FSP technique for synthesis of catalysts3 has been popular and gaining recently lots of
interest from industry15. In FASP, the fuel combustion is decoupled from the
precursor droplets. Typically H2 or hydrocarbon gases are burned and the precursor
droplets are sprayed into their flame. In both FSP and FASP, droplets could be
converted to either hollow or large micron- or nano-particles and/or solid solutions of
varying morphologies and sizes. Hollow particles are formed due to precipitation of
precursor around the droplet before the evaporation of the solvent. Larger particles are
the result of drying, collapse and densification of droplets due to insufficient enthalpy
for the conversion of liquid solvent to vapor before precursor surface precipitation.
Ideally, precursor vapor is generated during FASP and FSP, which after chemical
reaction, give particles of nanometer size as in VAFS. Quite a few catalysts have been
produced by FASP (e.g. SrMnO3, SrTiO3 and MoO3/TiO2) and FSP (e.g. LaMnO3,
LaCoO3, Au/TiO2 and Ag/ZnO) and have been evaluated already for catalytic
reactions such as photocatalysis, CH4 combustion and CO oxidation.3
1.2.2 Reactor configurations
VAFS reactors require vapor precursor that is combusted or oxidized on its own or
assisted by a hydrocarbon or H2/O2 flame to produce nanoparticles. High cost and
difficulty in finding volatile precursors make VAFS challenging for catalyst synthesis
that requires several components. Moreover, in synthesis of multicomponent particles
by VAFS, it is difficult to achieve homogeneous distribution of metal oxides due to
differences in their precursor volatility.
The FSP and FASP reactor configurations overcome the necessity of volatile
metal precursors as they utilize precursors that dissolve typically in combustible
(FSP) and noncombustible (FASP) solvents. As a result, virtually all elements in the
8 Chapter 1
periodic table can be used by liquid-fed flame synthesis processes. Both

configurations are similar, differences exist mainly in their operating principle. The
FASP utilizes low enthalpy content or non-combustible aqueous solutions of typically
nitrate and acetate precursors that are rather cheap. Therefore they require an external
source of energy in the form of hydrocarbon or H2/O2 flames to drive droplet/gas-to-
particle conversion. The FASP12 produces nanometer to submicron sized particles,
depending on the applied combustion energy.
On the other hand, FSP, which was first introduced by Sokolowski et al.11 for
synthesis of Al2O3 is quite similar in concept to carbon black synthesis by the furnace
process7 and leads readily to nano-sized particles by gas-to-particle conversion.
During FSP13, the precursor solution is fed at the center of the reactor (Figure 1.2b) ,
e.g. with a syringe pump, dispersed by high velocity gas (e.g. O2) creating a fine spray
of droplets that is ignited and stabilized by a pre-mixed flame. After combustion,
particles are formed as in VAFS. Worth noting is that, more than 50% of energy is
contributed by the liquid precursor solution during the combustion process in FSP.3
Each reactor configuration has its own merit but FSP is generally preferred due to its
capacity to produce nano-sized homogeneous particles and its compact operation
provided that appropriate precursor-solvent mixtures are identified for a given catalyst
composition.
1.2.3 Flame parameters for catalyst synthesis
Precursor concentration, fuel, mixing, oxidant, entrainment, precursor/dispersion flow

rate ratio (P/D) and precursor solution composition are some of the process
parameters that affect product properties and in particular primary particle and
crystallite sizes that frequently affect catalytic performance. In VAFS, precursor
concentration in the carrier gas (e.g. O2, Ar) is primarily controlled by evaporator,
whereas carrier gas flow rate can also be used in regulating the rate at which precursor
vapor is fed to the flame. Moreover, fuel (H2 or hydrocarbon) and oxidant flow rates
and their mixing influence the product particle properties.2
9
80 100
(a) dp
dAnatase
dRutile 80
Average diameter, nm
60
Anatase, wt.%
Anatase (wt.%)
60
40
40
20
20
0 0
0 5 10 15 20 25 30
TTIP flow rate, g/h
Flame: A (b)
Primary particle diameter, nm
80
Air CH4
Ar/TiCl4
60
Flame: B
40 CH4 Air
Ar/TiCl4
20
200 300 400

Flow rate of methane, mL/min
Figure 1.3: (a) Primary particle diameter (circles, left axis), anatase (triangles, left
axis), and rutile (diamonds, left axis), phase composition (squares, right axis) of TiO2
powders made in flames with 2 L/min O2 flow rate and the quenching nozzle placed at
5 cm above the diffusion burner (VAFS) for TTIP flow rates of 1.6 to 26 g/h 16 and (b)
Primary particle diameter of TiO2 particles made in diffusion burner (VAFS) with two
configurations: flame A (red symbols) and B (black symbols) as a function of CH4
flow rate at constant 3.8 L/min air flow rate.5 Adapted from refs. 16 and 5 with
permission from John Wiley & Sons and Elsevier.
Figure 1.3a shows the effect of titanium isopropoxide (TTIP) flow rates (1.6-26
g/h) on the average particle diameters (dp, dAnatase and dRutile) and anatase content of
10 Chapter 1
TiO2 at constant CH4, carrier-gas (Ar) and O2 flow rates.16 The particle and crystallite
sizes increased with increasing TTIP flow rate until about 16 g/h and remained nearly
constant at higher rates. Increasing TTIP flow rate increases both concentration and
flame temperature that accelerate particle growth by coagulation and sintering.2
However, growth in primary particle size (dp) slows down at higher TTIP flow rates,
here (> 16 g/h) as particle growth by sintering is inversely proportional to particle size
requiring much longer residence times at high temperature with increasing particle
size.9 Anatase TiO2 content decreased significantly (85-18 wt%) with increasing TTIP
flow rates (1.6-26 g/h) due to increasing O2 deficiency that favors rutile formation16,
highlighting its importance in the flame aerosol synthesis.
Figure 1.4b shows the influence of hydrocarbon (e.g. CH4) flow rate and VAFS
burner configurations on the primary particle diameter of TiO2 at a TiCl4 flow rate of
1.6x10-4 mol/min.5 The dp increased with increasing CH4 flow rate for both gas
mixing patterns (Flame A and B). However, much bigger and less aggregated
particles were formed in classic diffusion flames (A) than in inverse (or double)
diffusion flames (B). In flame A high concentration of newly formed particles
experience high temperature from CH4 combustion that promotes coalescence (or
sintering) before they get cooled or diluted by air, whereas in flame B the sequence is
reversed. Sintering of the particles can be further suppressed by increasing the air
flow rate, which dilutes and cools the flame.5 Controlling the oxidant (air or O2) flow
rate, which in turn affects the particle residence time, TiO2 catalysts containing 10
wt.% V2O5 with a wide range of SSA were prepared.10 Their specific surface area
could be increased from 23 to 120 m2/g by increasing the O2 flow rate from 2 to 10
L/min in a diffusion flame and a homogeneous distribution or coating of vanadia on
TiO2 could be achieved.10 This change in SSA significantly influenced the selective
catalytic reduction of NO that will be further discussed in section 3.3.
High temperature residence time, a dominant characteristic of flame synthesis
that determines the extent of particle growth (by coalescence and sintering), can be
controlled also by applying an electric field in the flame.2 The average primary
particle diameter of TiO2 decreased with increasing the field strength using needle or
plate (corona discharge) electrodes. The field generated by the electrodes across the
flame reduces the particle residence time in the high temperature region. Moreover, it
also charges the newly formed particles, which creates electrostatic repulsion and
11
dispersion. Both phenomena favor resistance towards particle growth resulting in

smaller primary particles.
Figure 1.4: Conceptual schematic of effect of precursor/dispersion (P/D) ratio on

specific surface area (SSA) of material. SSA decreases with increasing the P/D ratio.
Furthermore, particle size can also be controlled by addition and changing the length
of the tube and/or air entrainment by lifting the tube (see scheme in the middle).
Adapted from ref. 41 with permission from Elsevier.
Concepts similar to VAFS are utilized in both FSP and FASP where particle
properties are controlled easily by P and D, which are equivalent to changing
simultaneously both precursor/fuel concentrations and oxidant ratio in VAFS. The
degree of dispersion, high temperature particle residence time and extent of
combustion are closely related to the P/D ratio. Its effect is quite similar for most
materials during their FSP synthesis, especially when particle formation takes place
solely by gas-to-particle conversion as in VAFS.13 High surface area materials can be
made by decreasing the P/D ratio which results in shorter visible flames thus lowering
the particle residence time in the high temperature zone and slowing down particle
12 Chapter 1
growth (Figure 1.4). The opposite effect occurs at higher P/D ratios. This has been
well demonstrated in the synthesis of various catalytic materials.3,8 However, this
trend does not always hold13 especially when there are products of incomplete
combustion.17 This can be mitigated sometimes by utilizing sufficient combustion
energy e.g. by using high enthalpy solvents.13
Additionally, flame-made particle characteristics can be tuned at constant P/D
ratio by enclosing the FSP with a tube, which blocks air entrainment to the flame
spray jet, thus preventing the cooling of the flame (Figure 1.4, inset).18 Moreover,
particle size can also be controlled by varying the length of such tube and essentially
controlling such high temperature residence time.
(a) (b) (c)
(d) (e) (f)
Figure 1.5: Electron micrographs of the structure of various flame-made catalytic

materials. Nanocrystals of (a) fresh and (b) calcined (900 C for 2 h in air)
Ce0.5Zr0.5O2. Lattice fringes indicate well developed single crystals in most cases and
calcination affords crystalline particles of similar size.19 (c) Pd supported SiO2 where
Pd was well dispersed with a size ranging from 1-2 nm.20 (d) Al2O3 supported Pt-Pd
catalyst where Pt-Pd coexisted as an alloy as confirmed by EDXS and EXAFS
analysis.21 (e) Pt-FeOx/CeO2 catalyst with Pt residing in FeOx crystal hinting to
strong interaction between them.22 (f) Flame-made ZnO nanorods with length and
diameter as high as 100 nm and 50 nm, respectively.23 Adapted from refs. 19, 20, 21,
22 and 23 with permission from Royal Society of Chemistry, Elsevier and Springer.
13
The FSP or FASP solvent composition can critically affect the product catalyst
characteristics. In FSP synthesis of CeO2-ZrO2, precursor solutions derived from the
mixture of lauric-acetic acid gave highly crystalline Ce0.5Zr0.5O2 (Figure 1.5a). In
contrast, a product containing ceria rich and zirconia-like phases was obtained when
mixing iso-octane, acetic acid and 2-butanol.19 In the latter precursor solution spray,
rapid release of all liquid solvent from droplets takes place (due to low boiling point,
around 100 C) leaving lumps of precursor salts. Due to difference in the
decomposition rate of the two salts inhomogeneous distribution of these materials
occurs resulting in products of different compositions. The highly crystalline
Ce0.5Zr0.5O2 still remained open with only few contacts between particles, thus
improving sintering-resistance as demonstrated by the good thermal stability of
Ce0.5Zr0.5O2 even when calcined at 900 C for 2 h in air (Figure 1.5b)19 indicating its
potential application in high temperature catalytic reactions, e.g. the partial oxidation
of CH4.
1.2.4 Temperature
Temperature strongly affects the characteristics of flame-made catalysts. Flame

aerosol synthesis, involves combustion of precursors to produce materials, and as a
consequence strong temperature profiles are established in the reactor, e.g. in the SiO 2
synthesis flame.13 These profiles highly depend on the applied process parameters
such as precursor composition, solvent and oxidant gas (O2/air) and have a significant
effect on the final structural and chemical properties of the materials.
Generally, the average flame temperature is around 2000 C just above the
nozzle.13 Usually it increases slightly above the burner as the precursor/solvent are
consumed and then decreases with further distance away from the burner as it mixes
with entrained oxidant or sheath gas. These temperature profiles can vary with solvent
type (e.g. ethanol or xylene). The highest measured temperature for pure ethanol and
xylene flames were ~2700 C and ~3400 C, respectively.24 This temperature
difference can be related to the lower flame temperature resulting from stoichiometric
combustion of O2 with the former solvent compared to that from the latter.
Temperature profiles obtained from a simulation model showed excellent agreement
with experimental values especially for larger heights, which opens up the possibility
14 Chapter 1
of determining flame temperature computationally for a wide range of materials

synthesis.24
1.3 Particle properties relevant for catalysis

1.3.1 Chemical composition
The chemical composition is a key parameter in the design of catalysts. Flame

methods are versatile as regards tailoring of elemental composition. Limitations exist
only concerning solubility of precursors in a single miscible solvent solution and
potential toxicity of vapors, e.g. nickel tetracarbonyl formation in the flame when
employing nickel precursors. In such cases special safety measures have to be taken,
e.g. particle production inside the glove box.25 Purity of the catalyst composition is
essential as even trace amounts (in ppm range) of impurities influence catalytic
performance. Wet-synthesis processes generally require multiple steps thereby
increasing the chance of losing active elements and also possible contamination from
residues or production equipment. In contrast, flame aerosol synthesis can produce
materials with well specified composition and high purity.2,7
Transition metal impurities (Co, Cr, Mn and Fe) were added deliberately in the
ppm range in the flame synthesis of TiO2/SiO2 epoxidation catalysts lowering their
selectivity to the epoxide compared to that of pure catalysts (Figure 1.6a).26 The
drastic reduction of olefin selectivity with increasing Cr addition (> 3 ppm) was due
to leaching of Cr into the reaction solution, leading to homogeneous catalysis,
catalyzing the substrate conversion to the corresponding ketone. This example shows
that doping in the ppm range is easily achievable with flame synthesis and highlights
the importance of catalyst purity.
15
Brnsted/Lewis (B/L) acid ratio, -

100
(a) (b)
Olefin selectivity, % 0.6
80
0.4
60
Co 0.2
40
Cr
Mn
Fe
20 0.0
10 100 1000 0 20 40 60 80 100
Content of transition metal in TiO2-SiO2, ppm SiO2 content in Pt/Al2O3, wt.%
200 500
(c) (d)
Enantioselectivity (ee), %
90
CH3CHO in the effluent, ppm
400
150
CO2 production, ppm

300
80
100
200
70
50
100
circle: Methyl benzoylformate hydrogenation
square: Ketopantolactone hydrogenation
60 0 0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 10 20 30
Brnsted/Lewis (B/L) ratio, - Relative contents of terminal TiOH group, %
Figure 1.6: (a) The selectivity of epoxide formation related to alkene consumption for
transition metal-doped TiO2-SiO2. Doping with Cr and Co reduce the efficiency of the
alkene usage, while incorporation of Mn and Fe does not lead to significant loss of
alkene reactant at dopant levels of up to 2000 ppm. Cr exhibits the most pronounced
effect, as it leaches into the reaction mixture, converting most of the substrate to the
corresponding ketone instead of epoxide formation.26 (b) Effect of SiO2 on
Brnsted/Lewis (B/L) ratio of Pt/Al2O3 catalysts and (c) its influence on
enantioselectivity, Pt/Al (filled symbols) and Pt/Al-Si (open symbols).27 (d) Variation
of acetaldehyde concentration in effluent after UV photodegradation and CO2
produced under steady-state conditions with the relative content of terminal TiOH
groups estimated by 1H MAS NMR studies (initial acetaldehyde concentration =
18310 ppm).28 Adapted from refs. 26, 27 and 28 with permission from Royal Society
of Chemistry and Elsevier.
16 Chapter 1
The acidity/basicity of catalysts can be tuned by their chemical composition.

Generally, this is achieved by adding components that possess acidic or basic sites, or
by combining oxides which upon interaction form active sites with enhanced acidity,
e.g. by combining SiO2 and Al2O3. This has been exemplarily shown for Pt catalysts
supported on mixed SiO2-Al2O3 having acidic sites and for Pt/Cs2O possessing basic
sites.27 Figure 1.6b shows that by increasing the SiO2 content in Pt/Al2O3, the ratio of
Brnsted/Lewis acid sites (B/L) can be decreased. Increasing the B/L ratio directly
affected both the methyl benzoylformate and ketopantolactone hydrogenation reaction
rates and all SiO2-Al2O3 supported Pt catalysts showed higher activity than those on
pure Al2O3 (Figure 1.6c). Moreover, an optimal B/L acid ratio (around 0.25) was
achieved by adding 30 wt% SiO2 that exhibited the highest enantioselectivity for both
reactions.27 The change in acidity of the support also influences the electronic
properties of the loaded noble metals such as Pt27 and Pd20. These changes in the
properties of noble metals can be a decisive factor determining their catalytic
behavior. Similarly, silica addition to ZrO2 introduced Brnsted acid sites that
enhanced the mandelate yield from about 8% (for pure ZrO2) up to 52-67% in the
mixed oxides in the hydrogenation of phenylglyoxal to ethyl mandelate.29
Control over hydroxyl (OH) groups on FSP-made TiO228 and Pt/TiO230 by
doping Cu and fluorine (F) has been recently reported. The performance over F-doped
TiO2 was superior to that of Cu-doped TiO2 in the photodegradation of acetaldehyde,
as evidenced by the CO2 production (Figure 1.6d).28 This difference in the activity
could be directly related to their relative content of terminal TiOH, where former had
the lowest (~ 4%) and latter had the highest (~ 30%) values. During the photocatalytic
hydrogen production, an improvement of the performance was observed (from 19.1 to
22 mmol/h.gcat) over Pt/TiO2 after addition of 5 wt% F, which was attributed to an
increase in the population of surface OH groups.30 At higher loadings, the H2
production rate decreased due to the introduction of structural defects hindering the
interface electron transfer.
1.3.2 Multi-metal systems
Multicomponent materials such as mixed oxides have been widely investigated in

various catalytic reactions because they offer fascinating possibilities for tailoring
17
catalyst properties. FSP has been successively used for producing various high surface
area, nonporous, mixed metal oxide nanopowders of different composition consisting
of CeO2-Al2O3, CeO2-ZrO2, CeO2-ZrO2-Al2O3, CoO-Al2O3, NiO-Al2O3 and ZrO2-
Al2O3.31 High-throughput screening as well as continuous flow catalytic studies were
applied to assess the catalytic performance of these mixed oxides for NOx reduction
and propane/propene oxidation. A set of Ce1-xZrxO2 and Al-Ce1-xZrxO2 nanopowders
were shown to exhibit good catalytic activity for both NOx reduction and
propane/propene oxidation. Interestingly, all of these Pt-free catalysts showed
activities comparable to traditional Pt/Al2O3 catalysts. In a later study, the same group
describes the synthesis of CexZr1-xO2 and (Ce0.7Zr0.3O2)x(Al2O3)1-x core-shell
nanopowders.32 These studies, among others, demonstrate that complex metal mixed
oxides can be produced with good control of stoichiometry and phase purity using
FSP. Beside their use as active component, the application of such mixed metal
oxides as supports is also attractive due to their high thermal stability, e.g.
Ce0.5Zr0.5O2 (Figure 1.5a,b)19.
Flames are suitable for the rapid and single step synthesis of perovskites. These
catalysts have shown high performance in CH4 combustion (over LaCoO3)33, one
among several other catalytic reactions, where high thermal stability and oxygen
carrying capacity are important. Addition of an element into the perovskite structure
can create partial metal ion substitution, thereby improving the catalytic activity, as
e.g. in Ag-doping of LaMnO3 (Figure 1.7a)34.
Flameless combustion of CH4 was improved distinctly with increasing Ag substitution
of La in LaMnO3 made by both flame and sol-gel method, with the former showing
higher catalytic activity. The high activity of the flame-made catalysts was attributed
to their high SSA, which facilitates the rapid oxygen transfer from bulk to surface and
vice versa, resulting in high oxygen availability and faster regenerability. Moreover,
temperature programmed reduction (TPR) showed that partial substitution of La by
Ag decreased the onset of the first reduction peak and enlarged the Mn reduction
range also contributing to the higher catalytic activity.
18 Chapter 1
100 SG-LaMnO3 (a)

SG-La0.95Ag0.05MnO3
SG-La0.9Ag0.1MnO3
80
CH4 conversion, %
FP-LaMnO3
FP-La0.95Ag0.05MnO3
FP-La0.9Ag0.1MnO3
60
40
20
0
200 300 400 500 600
Temperature, C
(b)
Rh/MgAl2O4
80
NOx removal, %
Pt/MgAl2O4
60
40
Pd/MgAl2O4
20 Reductants:
CO, closed symbols
C3H6, open symbols
20 40 60 80 100 120
Reduction time, s
Figure 1.7: (a) Catalytic activity of fresh perovskite catalysts produced using flame
pyrolysis and sol-gel.34 (b) NOx storage-reduction (NSR) during lean-rich cycling at
350 C over 1 wt% M/MgAl2O4 (M = Pt, Pd, or Rh) with different duration of rich
period using CO and C3H6 as reductant. The NSR cycles were carried out with 360 s
lean and 10, 20, 30, 60 and 120 s rich periods and plotted after 10 steady cycles.35
Adapted from refs. 34 and 35 with permission from Royal Society of Chemistry and
Elsevier.
19
Spinels (e.g. MgAl2O4) have been widely applied as catalyst supports, apart
from their usage in other fields such as sensors, due to their high thermal stability.
Generally, active materials such as noble metals (Pt, Pd, Rh)35 and/or transition metals
(Mn, Fe, Co)36 are loaded on spinels depending on catalytic application. Noble metal
loading did not influence the spinel structure of MgAl2O4 and high dispersion was
achieved for Rh compared to Pt and Pd. Therefore, the high NOx removal activity of
Rh/MgAl2O4 storage-reduction catalysts can be attributed to the high Rh dispersion
and the activity could be further improved by changing the reductant from CO to
C2H6, during regeneration (Figure 1.7b). The wet synthesis of these noble metal-
loaded MgAl2O4 catalysts generally requires two separate steps: synthesis of
MgAl2O4 and later impregnation of noble metal(s), both requiring post thermal
pretreatments, costing time and leading to lower SSA of the catalyst.
These examples show the unique opportunity of flame aerosol technology to
produce materials with high SSA in a rapid single step, while maintaining other
important properties. Additionally, high SSA (174 to 212 m2/g) ternary spinels (e.g.
MgAl2-xMxO4, where M = Mn, Fe, Co) can also be formed by flame aerosol methods
exhibiting high resistance towards sintering during catalytic CH4 combustion.36 Flame
aerosol synthesis of these types of ternary spinel catalysts is possible due to the
intimate mixing in the flame of the constituents (elements) on atomic scale. In
contrast, if required, the Mn, Fe and Co loaded MgAl2O4 catalysts can be synthesized
utilizing two nozzles, where the dopant resides on the well-structured spinel structure,
as will be discussed in section 3.6. Flame synthesis of different combination of
WO3/CeOx-TiO2 improved the surface Ce3+ concentration, which together with
homogeneous coverage of particles (e.g. Ce-Ti) by amorphous WO3 resulted in high
NH3-SCR activity comparable to that of the widely used wet-made V2O5-WO3/TiO2
catalysts. However, after hydrothermal aging the flame-made catalysts exhibited
much higher catalytic activity compared to the V-based analogous catalyst and Ce3+
remained the dominating surface species indicating its crucial role in the reaction. 37
This is a unique characteristic of flame aerosol technology facilitating the single step
production of complex multicomponent catalyst materials with desired properties
such as high oxygen carrying capacity and thermal stability.
20 Chapter 1
1.3.3 Specific surface area
Many catalytic reactions are influenced by SSA. Active metals in their bulk form
possess fairly low surface areas and thus relatively little active sites per unit mass.
Furthermore, catalytic active metals are prone to sintering at temperatures higher than
their specific Tamman temperature. Active sites must be accessible by the reactants
and therefore those located at the surface are most important because transfer to them
is not limited by intraparticle diffusion. Therefore, the active components are
supported on thermally stable materials such as refractory oxides (e.g. SiO2, CeO2,
Al2O3, ZrO2, TiO2) with high SSA, resulting in high surface to volume ratio of the
active component (high dispersion e.g. Pd/SiO2 Figure 1.5c). Flame aerosol synthesis
can be considered ideal for the synthesis of high SSA materials due to fast quenching
of the product, which suppresses particle growth. Flame-made V2O5/TiO2, a widely
investigated catalyst for the selective catalytic reduction of NO by NH3, exhibits high
NO conversion (> 99%) at significantly lower temperature (200 C). Applying VAFS,
the SSA of V2O5/TiO2 could be increased from 23 to 120 m2/g by increasing the
oxidant flow rate (Figure 1.8a)10, as described in section 2.3. The NO conversion over
the high surface area catalysts was more than 95%, already at a low reaction
temperature of about 200 C, highlighting the importance of the catalysts SSA.
Furthermore, the increase in SSA reduced drastically the selectivity to undesired N 2O,
an effective greenhouse gas, in the whole temperature range, as illustrated in Figure
1.8a. It should be noted that such performance was achieved by scaling up the VAFS
of V2O5/TiO2 from 4 to 200 g/h as will be discussed in section 1.3.8.
Maintaining constant high catalytic activity is sometimes difficult as the
catalysts tend to deactivate with reaction time. Sintering, coking and poisoning are
some of the causes for catalyst deactivation.38 Especially, with supported metal
catalysts, sintering of the active component at high reaction temperatures has a major
influence on catalyst activity due to increase of the crystallite size and thus lowering
of the number of accessible active sites on the surface. However, in some cases
sintering of a catalyst component does not necessarily lead to decreased performance.
For example, Mn-Na2WO4/SiO2 catalysts showed significant structural change, i.e.
formation and growth of the cristobalite SiO2 phase during the oxidative coupling of
methane, but this had virtually no effect on the catalytic performance39 indicating that
the reaction is insensitive to this structural change.38
21
100 100
(a)
80 80
Selectivity to N2O, %
NO conversion, %
Catalysts SSA, m2/g
10V2 23
60 10V3 49 60
10V10 120
40 40
20 Filled: conversion 20
Open: selectivity
0 0
100 150 200 250 300 350 400
1.0 Temperature, C
closed: FSP
(b)
open: wet
0.8 fresh
2 h aged
CO conversion, -
6 h aged
0.6
0.4
0.2
0.0
150 200 250 300
Temperature, C
Figure 1.8: (a) Effect of SSA of V2O5/TiO2 on NO conversion (closed symbols) and
N2O selectivity (open symbols) in selective reduction of NO with NH3. Slight
improvement on NO conversion was obtained at lower reaction temperatures (< 200
C). Major influence of SSA was observed for N2O selectivity, the higher SSA the
lower was the selectivity (figure inset: xVy, where x: vanadia content and y: O2 flow
rate).10 (b) Comparison of CO conversion over fresh (square) and hydrothermally
aged (circle and down triangle) FSP (closed) and incipient wetness impregnation
(IWI) (open) Mn/Al2O3 catalysts.40 Adapted from refs. 10 and 40 with permission from
Elsevier.
22 Chapter 1
For reactions, however, that are sensitive to structural changes, it is important to

maintain the optimal particle size and crystallite structure, by making the catalyst
thermally stable. Recent work by Tepluchin et al.40 shows that even after
hydrothermal aging at 700 C for 12 h, only a minimal decrease in SSA of FSP-made
catalysts (20%Mn or 20%Fe/Al2O3) occurred (161 to 141 and 150 to 110 m2/g,
respectively) compared to that of a corresponding catalysts produced by impregnation
(139 to 95 and 136 to 72 m2/g, respectively). Manganese proved to be better than Fe
for CO conversion when loaded on Al2O3 produced by both flame and impregnation
methods. However, CO conversion over FSP-made Mn/Al2O3 was superior to wet-
made (Figure 1.8b). Though catalytic activity decreased on both catalysts after
hydrothermal aging, the flame-made one still showed better performance, owing to its
thermal stability. Additionally, the activity of the flame-made Mn/Al2O3 catalysts
could be partially regenerated after SO2 poisoning unlike the wet-made ones. This
behavior was attributed to the resistance of the flame-made catalyst against sintering
and poisoning, and to the homogeneous distribution of Mn species achieved.
1.3.4 Particle size
The influence of particle size and shape on catalyst performance is a well investigated
topic. Decrease in particle size can increase the number of active sites improving the
catalytic performance. Furthermore in the particle size range up to about 10 nm the
statistics of atoms with a particular coordination changes significantly, which can lead
to a drastic change in the catalysts performance if the target reaction is structure-
sensitive. However, controlling the size and homogenous distribution of supported
nanoparticles is often difficult since the particles tend to grow by sintering, especially
during prolonged high temperature calcination as used in classical wet preparation
methods. Noble metals are widely used in catalysis due to their excellent catalytic
properties for many reactions. However, the resources of these metals are limited as
they are expensive. A viable solution towards reducing their usage would be to
develop non-noble metal containing catalysts, a big challenge, or minimizing their
loading while maintaining catalytic activity. Flame aerosol synthesis, where
exceptional control over particle size can be achieved, is a valuable tool for targeting
these challenges, as explained in 1.2.3.
23
High catalytic activity can often be realized by decreasing the particle size of
the active component, enhancing the interfacial contact between active (metal)
particles and support. These interactions can induce electronic and geometric effects
in the active particles, which are favorable for the reaction. Figure 1.9 shows the
influence of Pt particle size on sucrose mineralization where ~1.6 nm sized Pt showed
optimal performance regardless of reaction conditions.41 This improved activity was
attributed to the creation of additional electronic surface states and more reactive
sites. Interestingly, the electronic effect induced in Pt/TiO2 containing the smallest Pt
particles (~1.4 nm) was not favorable, since the high photocurrent density of the Pt
deposits increased the electron-hole recombination. Furthermore, the deposit size was
too small to establish sufficient electrical contact for efficient interfacial charge
transfer between the photocatalyst and sucrose, and thus this catalyst showed
relatively low activity. With increasing Pt particle size (>1.6 nm), the photocurrent
started to accumulate, making formation of organic-metal deposit bonds difficult,
which also resulted in low activity. In some cases, slightly larger but well-dispersed,
active materials are necessary to achieve high catalytic activity, as e.g. for the
enantioselective hydrogenation of -ketoesters over Rh-loaded Al2O3.42
160
50% overall oxidation rate, g C/min
140
Pure TiO2
120
100
FSP
80 FSP O2 enriched
Degussa P25
Degussa P25 O2
60
0 1 2 3
Active Pt particle diameter, nm
Figure 1.9: Effect of active Pt particle diameter in flame-made Pt/TiO2 on the half-life
rates of 2000 g of carbon (as sucrose) and their comparison with pure FSP-TiO2
and Degussa P25. Pt particles of size ~1.6 nm showed optimal performance.41
24 Chapter 1
1.3.5 Crystal structure and morphology
Correlating catalytic activity with crystal structure and morphology of

technical catalysts is often challenging due to the difficultly in establishing a
conclusive relationship. Existing differences in the physicochemical properties such
as crystal composition, surface area and particle size of the analyzed materials render
a proper comparison difficult. This problem can be minimized by preparing materials
where all properties except the selected target property are similar. An example
towards this goal is the preparation of TiO2 with similar surface areas but different
phase composition (amount of rutile) for the photocatalytic decomposition of phenol
and salicylic acid43 and the production of H2 from methanol.44 The comparison of the
catalytic performance of flame-made TiO2 (F1-VAFS), commercial Degussa P25 and
UV-Titan, which all had similar surface areas (70-78 m2/g), showed that the presence
of both anatase and rutile is necessary to achieve high activity (Figure 1.10a).43 The
photocatalytic activity of Aldrich rutile was better than that of UV-Titan though their
SSAs were vastly different (2 vs. 70 m2/g). This can be traced back to the presence of
a small amount of anatase (3%) in the Aldrich rutile. TiO2 particles of 20-40 nm size,
which contained < 10 wt% rutile, were the most photoactive in destroying phenol.
Moreover, it was inferred that both anatase and rutile are necessary to achieve some
activity, and a synergistic effect between them results in an optimal performing
photocatalyst. Similar correlations were also observed in the photocatalytic
decomposition of salicylic acid.
For the photocatalytic H2-production from methanol also a synergistic effect
between anatase and rutile was observed (Figure 1.10b).44 The optimal anatase
content was 39 mol% affording a production of about 4500 mol of H2 in 8 h. Lower
activity of the catalysts containing both higher (95 mol%) and lower (4 mol%) anatase
content further substantiated the necessity of optimal composition. Interestingly,
mechanical mixing of TiO2, mimicking optimal anatase content (39 mol%), afforded
the lowest activity, which can be attributed to the absence of a beneficial synergistic
effect, as it was observed with the flame-made catalysts. In addition to the control
over the content of these conventional TiO2 phases, metastable phases such as
crystalline Ti3O5 and Ti4O7 can also be produced by FSP.45 These titanium suboxides
were discovered in flame-made Ag/TiO2 that showed very high visible-light activity
25
for the photodegradation of Cr6+ and methylene blue (down to 15 min half-life)
compared to that of conventional P25 (Figure 1.10c).
1.0 5000
Anatase composition (mol%) (b)
(a) 4%
Phenol concentration, C/Co
0.8 4000 39
95
Amount of H2, mol

39% (physically mixed)
TiO2 Rutile wt%
0.6 UV-Titan 100 3000
Aldrich rutile 97
Degussa P25 25
0.4 F1-VAFS <0.1 2000
F2-VAFS 7
0.2 1000
0.0 0
0 15 30 45 60 75 90 0 1 2 3 4 5 6 7 8 9
Illumination time, min Time, h
Methylene Blue concentration, C/C0
30
1.0 (c) Bi/Mo = 2:1 -Bi2MoO6 (d)
Bi/Mo = 1:1 -Bi2Mo2O9
Degussa P25 (TiO2) 25 Bi/Mo = 2:3 -Bi2Mo3O12
Propylene conversion, %
0.8 20Ag/TiO2 (P/D = 8/5)

20Ag/TiO2 (P/D = 3/5) 20
0.6
15
0.4
10
0.2
5
0.0 0
0 40 80 120 160 200 0.00 0.05 0.10 0.15 0.20 0.25 0.30
Irradiation time, min w/F, g.s.mL-1
Figure 1.10: (a) Photocatalytic decomposition of phenol with flame synthesized (F1
and F2-VAFS refers to TiO2 from two different flame synthesis conditions) and
commercial TiO2. Initial phenol concentration: 1mM, catalyst concentration: 0.5 g/L,
pH: 3.5, O2 flow rate: 1 L/min, 450 W UV-lamp.43 b) Photocatalytic H2 evolution over
flame-made TiO2 nanoparticles containing different amount of anatase (4, 39, and 95
mol%) and over TiO2 nanoparticles prepared by physical mixing method resulting in
39% anatase.44 (c) Photocatalytic reduction of methylene blue by Degussa P25 and
Ag/TiO2 prepared by FSP at P/D = 8/5 and 3/5 under visible-light irradiation (
>400 nm).45 (d) Propylene conversion as a function of contact time (w/F) over flame-
made Bi/Mo catalysts.46 Adapted from refs. 43, 44, 45 and 46 with permission from
Taylor & Francis, American Chemical Society, Elsevier and Royal Society of
Chemistry.
26 Chapter 1
The possibility of the formation of these types of metastable phases and crystal
structures by flame aerosol synthesis provides an interesting tool for tailoring
materials suitable for catalytic application. Another example is the preparation of low
temperature BaCO3 that was produced by applying two-nozzle synthesis. This
material exhibits high NOx storage capacity.6 Moreover, successful production of Pt-
Ba/Al2O3 catalysts virtually free of the high temperature BaCO3, in a wide range of
Ba loadings (4.5-33 wt%), led to improved NOx storage capacity of these catalysts. In
contrast, wet-made catalyst of similar composition contained high temperature BaCO3
that led to lower NOx storage capacity compared to flame-made ones.47 Other studies
indicate that higher dispersion of active species can be achieved by flame aerosol
synthesis compared to classical wet-chemistry methods, resulting in higher catalytic
activity, e.g. V2O5/TiO2 for methanol oxidation.48
Metastable phases such as -Bi2Mo2O9 with high SSA (19 m2/g) can also be
synthesized in a single step using FSP.46 Generally, this phase is obtained by a high
temperature (> 560 C) calcination process that results in low SSA. Propylene
conversion over this metastable phase was higher than that of other phases (-
Bi2Mo3O12 and -Bi2MoO6) and increased with increasing contact time (Figure
1.10d). Acrolein selectivity of all these catalysts was comparable (all around 70%) at
their highest conversion proving the superior catalytic performance of -Bi2Mo2O9.
The catalytic activity of both - and -phases was lower than that of the -phase
though their SSAs were similar (-phase, 18 m2/g) or even higher (-phase, 45 m2/g)
than that of the -phase (19 m2/g), illustrating the role of crystal structure in the
reaction. These metastable or mixed-oxide phases are formed in flame aerosol
synthesis due to rapid quenching of the particles that immediately leads to freezing of
the structure.
1.3.6 Spatial distribution of components
The nature and location of active catalyst components either on the surface or in the
subsurface of the support can greatly influence the catalytic activity. In flame
synthesis of Pt-Pd/Al2O3 catalysts, an alloy of Pt-Pd was formed, confirmed by EDXS
(Figure 1.5d) and EXAFS analyses.21 Catalytic activity of the combustion of CH4 was
high over the catalysts containing Pt-Pd alloy as Pt influenced the redox property of
Pd favoring its presence in reduced form. The H2-TPR results showed that Pt addition
27
broadens the Pd reduction signal and shifts it towards lower temperatures, indicating
easier reduction. A small increase of the Pt content of the catalyst improved its
resistance towards sintering thus enhancing the CH4 combustion activity.
In flame-made Co/ZrO2, Co was internally distributed and stabilized within the
ZrO2 matrix, unlike in conventionally prepared catalysts.49 This facilitated formation
of very fine Co0 clusters after the reduction and their amount was doubled after
addition of a small quantity (0.4 wt%) of noble metal (Ag, Ru, Pt, Rh, Pd) as
promoter, enhancing the reducibility of the Co-species. The high temperature CoOx
reduction peak shifted from about 550 C to around 320 C except in the case of Ag
promotion. Rh addition to Co/ZrO2 ameliorated the hydrogenation of CO and resulted
in higher methane selectivity with increasing hydrogen partial pressure compared to
the unpromoted catalyst.
As the number of components in the catalyst increases, the complexity in
understanding the underlying reasons for the catalyst activity also increases due to
their interdependent interactions. Therefore, control over the location/interaction of
these components can provide a better understanding of the function of catalysts and
eventually aid in the design of optimal performing catalysts. In this case, two-nozzle
synthesis can be applied for developing sophisticated materials where the intermixing
and interaction of the different components in the flame can be better controlled.6
Formation of mixed oxides is favored in single nozzle systems due to the

homogeneous distribution of the precursor components in the feed, which results in
intimate mixing in the flame, while this intermixing can be minimized with two
nozzles systems.6 Spraying Pt/Ba and Al2O3 precursor separately utilizing two
nozzles, crystalline BaCO3 was obtained, whereas it was amorphous in powders
prepared by single-nozzle synthesis. Moreover, BaCO3 crystallinity could be
improved by increasing the internozzle distance, which delays the intermixing of Ba
and Al, and consequently the interaction occurs at lower temperature. Higher NO x
uptake by Pt/Ba/Al2O3 catalyst prepared by two nozzles (3.6% vs. 0.9 wt%)
demonstrated the beneficial use of two-nozzle synthesis in this case.
Figure 1.11a shows schematically the two-nozzle set-up as it was used for the
spatial control of the different components of potassium promoted Pt/Al2O3 and Pd/
Al2O3 NOx storage-reduction catalysts. The precursor solution of the desired
components could be sprayed separately, e.g. by spraying solution mixture containing
28 Chapter 1
precursors of potassium (K) and Pt or Pd with one nozzle and precursor solution
containing only Al with the other.50 As a result, catalysts with spatially controlled
deposition were obtained where Pt or Pd are interacting preferentially with K 2CO3, or
catalysts where the noble metals were deposited on Al2O3 instead on K2CO3.
(a)
(b)
Figure 1.11: (a) Schematic of two-nozzle flame synthesis configurations leading to

preferential deposition of Pt/Pd on K2CO3 or Al2O3. (b) Performance of catalysts
prepared by preferential deposition in NO storage-reduction at 250 C. NOx
conversion as a function of lean-rich cycling of catalysts. Each cycle consisted of a
lean (oxidizing) period (3 min in 667 ppm NO and 3.3% O2 in He) and a rich
(reducing) period (1 min in 667 ppm NO and 1333 ppm C3H6 in He). The highest NOx
conversion was observed for catalyst with Pt on K2CO3 and Pd deposited on Al2O3.50
29
Using this set-up, during synthesis, particle size, composition and location of
the individual components can be controlled to some extent. The particles produced
from each nozzle condense first before mixing to the final product, promoting mixing
at the nanoscale rather than on atomic scale6 ensuring their predetermined interactions
and location. By variation of the intersection distance between the two nozzles,
catalyst can be optimized based on the feedback of performance tests. Utilizing this
technique, it was shown that Pd deposition on Al2O3 and Pt on K2CO3 exhibits best
performance for NOx storage-reduction, especially at the beginning of the cycling
between fuel-lean and fuel-rich periods (Figure 1.11b).50
Novel highly active Pt/FeOx-CeO2 catalysts for the preferential oxidation of
carbon monoxide (PROX) were developed by adjusting the intersection distance
between the two nozzles that controlled the interactions between Pt/FeOx and CeO2.22
Strong interaction between Pt and FeOx was evident by electron microscopy (Pt
residing in FeOx crystals Figure 1.5e). By adjusting the intersection distance, it was
possible to tune the morphology and the reducibility of the catalysts. Intimate
interactions between Pt/FeOx and CeO2 achieved at the greatest flame distance led to
reducibility of the material at the lowest temperature (-6 C). This tailor-made catalyst
showed CO conversion > 99.5% below 90 C, which is 30 C lower than that of
mechanically mixed catalyst of same composition.
Co-Mo/Al2O3 catalysts for hydrodesulfurization (HDS) were prepared
employing single-nozzle and two-nozzle FSP.51 The best performing catalysts were
those prepared with the two-nozzle FSP, which showed an improve of the relative
activity, compared to a commercial reference catalyst, of 91% while the
corresponding single-nozzle made catalyst showed a relative activity of 75%. The
better performance of the catalyst prepared by two-nozzle FSP was attributed to better
promotion of the active molybdenum sulfide phase, due to suppression of the
formation of the undesired phase CoAl2O4, which makes Co unavailable for
promotion.
The above examples illustrate that prior knowledge of the required physical and
chemical properties of the catalyst is essential in choosing the right synthesis method.
Another important advantage of the two-nozzle synthesis is that it overcomes
the problem of precursor immiscibility, because the solvent can be adopted to the
solubility of the precursor. The extension of the two-nozzle system to systems
containing further nozzles is feasible. However, to benefit from such an extension, a
30 Chapter 1
better understanding of the particle formation mechanism in multicomponent system

is crucial.
1.3.7 Coating of ceramic supports
Structured ceramic supports have found wide application, particularly in

environmental catalysis. Generally, the catalytic powders are wash-coated on
cordierite honeycomb and thermally treated at high temperatures before using them in
a catalytic process, e.g. in exhaust gas treatments. This requires various steps, and a
lengthy process, which leads to significant fabrication costs. Therefore, development
of a technique that can directly coat the catalytic materials onto the ceramic support,
which also avoids post-thermal treatment, will be beneficial for many catalyst
industries. The flame aerosol technique can be utilized for developing such catalyst-
coated ceramic supports, as has been demonstrated by Schimmoeller et al.52 The
ceramic foam was mounted into a double-walled water cooled tube and positioned
right above the flame producing V2O5/TiO2 (Figure 1.12a), which was tested in the
partial oxidation of o-xylene to phthalic anhydride. The coating amount was easily
controlled by varying the pressure drop and deposition time. Homogeneous coating
was achieved (Figure 1.12b) and the catalyst powder coating showed a homogeneous
particle size distribution (Figure 1.12c).
31
1.0 (d)
o-xylene conversion, Xo, -

0.8
(a)
0.6
0.4
foam-particles (93 m2/g)

0.2 foam-particles (53 m2/g)
FSP-made pellets
wet-made pellets
0.0
1E-3 0.01 0.1
(b)
Modified time, mod, gcat s cm-3
0.8
(e)
Selectivity, SPA,o, -
0.6
0.4
(c) 0.2
0.0
0.2 0.4 0.6 0.8 1.0
o-xylene conversion, Xo, -
Figure 1.12: (a) Schematic of the FSP coating of V2O5/TiO2 catalyst onto ceramic foam, (b) SEM image of catalyst coated ceramic foam and (c)
TEM image of V2O5/TiO2 nanoparticles. (d, e) Catalytic performance of various coated ceramic foams and pellets in o-xylene oxidation to
phthalic anhydride. (d) o-xylene conversion with modified time and (e) o-xylene conversion versus phthalic anhydride selectivity.52 Adapted from
ref. 52 with permission from Elsevier.
32 Chapter 1
The coated foam with high SSA (93 m2/g) exhibited significantly higher activity compared to
that of the coated foam with low SSA (53 m2/g) and pelletized (both FSP- and wet-made)
catalysts (Figure 1.12d). Moreover, coated forms showed higher phthalic anhydride selectivity
than that of others at 100% o-xylene conversion (Figure 1.12e). The high catalytic activity of
the coated ceramics was attributed to the high SSA and porosity (~98%) that enhanced the
intraparticle mass transfer, which was suppressed in pelletized catalysts of similar
composition due to their longer diffusion path and smaller pore sizes.
Direct deposition of Mn3O4 onto catalytic laboratory-scaled cordierite diesel
particulate filter was tested for soot oxidation using FSP.53 The oxidation of tight contact soot
occurred in the temperature range of 180-350 C, which indicates its potential for
instantaneous removal of soot under these conditions. As a result, particulate filter could be
continuously regenerated under realistic diesel exhaust conditions. The investigations
discussed above are promising and this novel direct catalyst coating technology using flame
methods is likely to be attractive for catalyst industries.
1.3.8 Scale-up
An important feature of any synthesis process is its scalability while maintaining the
physicochemical properties of the best catalyst emerging from laboratory tests. While many
synthesis methods show significant hurdles for scale-up, flame aerosol synthesis is already a
proven scale-up technology, as e.g. demonstrated by the industrial scale production of
photocatalytic TiO2.1 Physicochemical properties of the material are strongly influenced by
process parameters as described in section 2.3. Therefore, it requires a proper understanding
of the process from both experimental observations and theoretical simulations.
Understanding of the whole synthesis process is a prerequisite for successful scale-up.
Recently, Grhn et al.24 demonstrated that ZrO2 nanoparticles production can be scaled-up
from ~100 to 500 g/h by maintaining important product properties such as crystal size and
type. This was achieved by keeping the high temperature residence time constant during the
synthesis i.e. maintaining the FSP P/D ratio constant (Figure 1.13a). Variation of the P/D ratio
had virtually no effect on the mass fraction of tetragonal ZrO2 (92-96 wt%) except that the
crystal size increased from about 10 to 22 nm.
33
Oxygen flow rate, L/min

20 40 60 80
25 100
(a)
20 80
Tetragonal fraction, %
Crystal size, nm
15 60
10
40
5
20
0
10 20 30 40 50 60 70
Precursor flow rate, mL/min
(b)
100
80
NO removal, %
18 wt% VOx/TiO2,
flame-made, 100 g/h
60
20 wt% VOx/TiO2,
40 impregnated
20
160 200 240 280 320 360

Reactor temperature, C
Figure 1.13: (a) Product ZrO2 tetragonal phase fraction (right axis, solid symbols) and
crystal size (left axis, open symbols) as a function of precursor flow at constant 80 L/min of
O2 (triangles) and constant precursor to dispersion O2 (P/D) flow ratio (circles). Product
crystallinity is maintained during constant P/D ratio scaling.24 (b) Catalytic activity
comparison of flame- and wet-made V2O5/TiO2 catalysts for the NO removal by selective
catalytic reduction. Flame-made catalyst production was scaled up to 100 g/h and its
catalytic performance was superior to that of the catalyst prepared using wet-impregnation
method.54 Adapted from refs. 24 and 54 with permission from American Chemical Society and
John Wiley & Sons.
34 Chapter 1
The possibility of large scale production of materials such as SiO2 (> 1000 g/h), ZrO2
(up to 600 g/h) and Y2O3/ZrO2 (> 300 g/h) utilizing FSP has already been demonstrated by
various authors as has been summarized in a recent study.24 In addition, ZnO nanorods can
also be produced using FSP in high production rates (> 3 kg/h, Figure 1.5f).23 It seems
justified to state that multicomponent catalysts production can also be scaled-up maintaining
the physicochemical properties and thus in several cases the flame aerosol technique can
potentially replace the time consuming wet-synthesis methods. Successful pilot scale
production of binary catalyst (V2O5/TiO2) has already been demonstrated using a diffusion
flame reactor (VAFS) with a high production rate of 200 g/h.54 The catalyst showed better NO
removal activity (160-280 C) compared to a corresponding one with similar composition and
SSA produced by classical wet-chemistry method (Figure 1.13b).
1.4 Concluding Remarks
The flame aerosol technique is an attractive route for nanoparticle synthesis, which already is
employed in many powder synthesis industries and is currently being explored by many
academic and industrial laboratories for catalysis. However, there are still challenges in this
technique, one of them being precursor costs. Ideal precursors (e.g. alkoxides, organometallic
compounds) are relatively expensive, while cheaper and therefore preferred nitrate precursors
usually give inhomogeneous particles that are not desirable. However, utilizing proper
chemistry, homogeneous catalytic particles could be produced55 from these cheap precursors
opening up new avenues for utilizing a wide range of inexpensive precursors for catalyst
production.
Standard methodology for the inflight characterization of the aggregates and
agglomerates is necessary as it affects the assessment of the particle growth processes, which
in-turn affects their performance. Modification of nanoparticle surface with functional
materials (organic groups) during their production is another aspect that can be explored.
Development and successful implementation of such process can cut the cost of post
functionalization. Scale-up of material production is a challenge for many synthesis
techniques due to limited understanding of the dynamics involved in the production of
homogeneous multicomponent materials. In contrast, scalability of flame technology is
already proven in the production of nanomaterials.2,24 Forced nozzle quenching can be utilized
to reduce the particle size (e.g. TiO2).16 This possibility could be further explored in the
synthesis of catalysts. Utilizing the flame aerosol technique, synthesis of non-oxygen
35
containing metal fluorides, sulfides, nitrides, carbides and phosphides could be feasible
provided that O2 is controlled in the synthesis environment.
Flame aerosol synthesis produces non-porous nanoparticles and also less crystalline
products, due to low particle residence time in the high temperature zone of the flame.
Therefore, this technique may not be suitable for catalytic reactions where these properties of
the nanoparticles are crucial. However, by controlling the air entrainment and the length of
the tube enclosing the flame18, crystallinity of the particles can be improved by manipulating
their residence time at high temperatures5. Today, only a few investigations52,53 show the
possibility of developing direct catalyst coating technique utilizing FSP. However, these
examples show the feasibility and potential of this technique.
The future of this synthesis technique for the production of various sophisticated
materials with application not only in catalysis seems promising. This is also reflected by a
significant increase of the research activity in this field witnessed in the past few years.
However, more work will be required to fully understand the chemical and physical processes
occurring in the flame and to fully exploit the potential of this technique for material
synthesis. Research towards this aim will need a concerted effort of material scientists,
chemists, physicists and engineers.
1.5. Ethene production from catalytic conversion of natural gas
1.5.1 Overview
Ethene is an important feedstock in industry. It is used in the production of various high value
chemicals such as polyethylene, polychloroethene and ethylene oxide which are in turn
crucial in the manufacture of everyday materials (e.g. plastic bags, paint thinner etc.).56
Currently, ethene is exclusively produced by cracking, which is energetically-intensive,
suffers from severe coking and emits significant amounts of CO2 making it highly
undesirable.57 Since, the forecasts show an increase in the global ethene demand58, it is
imperative to develop alternative, more efficient processes for its production. Therefore, the
use of methane and ethane, the main components of the natural gas (NG), has been proposed
to produce ethene.59
Currently, NG has been mainly used for combustion-related processes such as
residential and industrial heating and in electricity production. With the recent growth in NG
reserves contributed by the rapid rise in shale gas extraction in the US, attention has now been
36 Chapter 1
turned towards upgrading of natural gas to high value chemicals.58 The development of highly
efficient catalytic processes is the key in realizing the solution for both the growing NG
reserves and increasing global ethene demand. Oxidative coupling of methane and oxidative
dehydrogenation of ethane are the two reactions which are of prime focus in finding a
common solution to these aforementioned problems.
1.5.2 Oxidative Coupling of Methane (OCM)
Industrial interest in the production of ethane and ethene by utilizing cheaply available
methane started after the early report on OCM from Keller and Bashin in 1980s.60 During the
OCM reaction, methane and oxygen are co-fed over the catalyst at relatively high
temperatures (> 700 C). The OCM reaction products are ethane, ethene, CO2, CO and H2O
(eqn. 1.1) and their distribution determines the efficiency and nature of the catalyst.
Particularly due to high reaction temperatures coupled with the reactive nature of ethane and
ethene with O2, achieving high C2 (ethane + ethene) selectivity has been a major challenge.
Therefore, developing catalysts that will minimize the undesired side reactions and thus
exhibit high selectivity while maintaining high and stable conversion are of significant
interests.61
CH4 + O2 C2 H6 + C2 H4 + CO2 + CO + H2 O (1.1)
Mainly due to high economic potential offered by this reaction, a large number of
catalysts of different types and compositions were produced and assessed.62 However, most of
them exhibited low activity and those which showed higher activity (e.g. Li/MgO) were
unstable, owing mainly to the structural-chemical changes (e.g. lithium sublimation in
Li/MgO) induced by high reaction temperatures.63 Owing to such difficulties in obtaining
high and stable C2-yield, the industries gradually lost interests in this reaction. However, with
the recent boom in NG reserves, attention towards OCM reaction has been revived.
Development and improvement of novel approaches such as synthesis methods (e.g. flame
spray pyrolysis39), reactor design (e.g. membrane reactors64) and rigorous computational
techniques (e.g. CFD65) are applied to gain fundamental understanding of the system and
through which it is aimed to optimize reaction conditions and catalysts to achieve higher C2-
yield.
Both methane conversion and C2-selectivity highly depend on the amount/nature of
oxidant (e.g. O2 and CO2) and also the residence time in the catalyst bed.66,67 Since the
products (ethene and ethane) are more reactive than the reactant (methane), improving both is
37
challenging as one counteracts the other. Zavyalova et al.68 recently demonstrated the
existence of this limitation by plotting selectivities and conversions obtained from several
high performing catalysts. Majority of the catalyst exhibited C2-yield between 25-30% and
only few showed values > 30%. Therefore, novel approaches are being pursued by several
researchers to break off this limitation.69 Partially or fully replacing O2 in the feed by CO2,
which is a milder oxidant, is one of the strategies that has been explored to improve the
product selectivity.67,70 Among several catalysts, Mn-Na2WO4/SiO2 catalyst has shown some
promising results in-terms of both activity and stability.68 However, conflicting speculations
on the roles of elements and active sites (e.g. Na-O-Mn and W-O-Si, )71, which highly
depends on the applied synthesis technique72, exist, demanding further in-depth analysis for
their rational design.
1.5.3 Oxidative Dehydrogenation of Ethane (ODHE)
Due to increasing global ethene demand and existing problems associated with current
production process, significant attention has been dedicated to ODHE reaction.59,73 The
reaction can be performed at relatively low temperatures (< 500 C) by co-feeding ethane and
oxygen over the catalyst (eqn. 1.2).74 However, due to the reactive nature of O2, achieving
high ethene selectivity has been a major challenge.75 Alternatively, CO2 and N2O can be used
as milder oxidants. However, due to the limited availability, use of the latter oxidant in large
scale is not viable making CO2 an obvious choice. CO2 is a greenhouse gas, whose emissions
are rising thus demanding urgent attention towards green and sustainable remediation.
Therefore, its usage towards making high value products (eqn. 1.3) would be beneficial from
both environmental and economic point of views.76
C2 H6 + O2 C2 H4 + H2 O (1.2)
C2 H6 + CO2 C2 H4 + CO + H2 O (1.3)
Though use of CO2 offers some benefits, slightly higher temperatures (> 600 C) are
necessary to perform the reaction due to its inert nature. As a result, active single metal oxide
catalysts suffer from sintering and coking during the reaction. Therefore, they are stabilized
on supports that possess high thermal stability and sometimes also partly contributes towards
minimizing the coke formation and deposition.75 The role of CO2 in the reaction has been
reported not only to promote redox behavior of the active metal oxides (e.g. Cr2O3 and Co3O4)
38 Chapter 1
but also to remove the deposited coke and consume H2 from ethane dehydrogenation via
reverse-boudouard and reverse-water gas shift reactions, respectively.77,78
Among several metal oxides that were assessed for ODHE reaction, gallium and
chromium oxides exhibited higher activity.79 The catalytic activity of gallium oxide supported
on titania was higher compared to other supported catalysts.80 However, ethane conversion
gradually decreased with reaction time owing to the formation and deposition of coke on the
active sites of the catalyst. Though surface acidity was considered as a crucial property
required for obtaining high activity but it also promoted coke formation and deposition.
Therefore, balance between the activity and surface acidic sites is essential in realizing high
performance. On the other hand, chromium oxide-based catalysts have been found to exhibit
higher and better activity compared to gallium oxide-based catalysts.81 Among several single
metal oxide-supported catalysts, the performance of Cr-supported on SiO2 was the highest
(ethene yield of 52%)81. Several zeolites as support materials (e.g TS-1 and ZSM-5) have also
been reported to improve the activity through acidic sites modification, however, the observed
performance was similar to that of the SiO2-supported catalysts.82,83 Though catalysts based
on both gallium and chromium oxides have been reported to show some promising results,
former metal oxide being relatively expensive and latter being toxic present challenges for
their future application. As an alternative, cobalt oxide-based catalysts have been proposed
which show activity similar to that reported over aforementioned metal oxide based catalysts.
In particular, cobalt supported on BaCO384, MCM-4185 and Mn-Na2WO4/SiO286 are some of
the reported active ODHE catalysts with reported ethene yield of 40-55%. However, there still
remains a lot to be investigated towards developing novel catalysts that exhibit high and
stable catalytic activity.
1.5.4 Summary and outlook
In recent years, ethene demand has increased and further growth has been predicted. And
current ethene production technology suffers from several problems causing both financial
and environmental damages. Therefore, with the increasing NG reserves, it appears viable to
produce ethene using components of NG (methane and ethane). However, such selective
conversions are only possible through the development of effective novel catalysts, a topic of
investigation for quite some time. Though considerable progress has been made, primarily
due to novel synthesis methods and development of sophisticated characterization techniques,
coking and stability still remain as major challenges. Therefore, further investigation is
needed including, but not limited to, synthesis methods, novel metal oxides and in situ
39
characterization to gain fundamental understanding at atomic level to tailor the catalyst

properties. In both the OCM and ODHE reactions, low selectivity towards the desired product
(ethene) is the key issue arising from its highly reactive nature (compared to that of the
reactant). Therefore, process modifications (e.g. reactor design) focusing on reducing their
residence time and using membrane reactors could also be explored.
1.6. Acknowledgement
Financial support by ETH Zurich Research Grant (ETH-39-12-2) is kindly acknowledged.
1.7. References
(1) Ulrich, G. D. Chem. Eng. News 1984, 62, 22-29.

(2) Pratsinis, S. E. Prog. Energy Combust. Sci. 1998, 24, 197-219.
(3) Strobel, R.; Baiker, A.; Pratsinis, S. Adv. Powder Technol. 2006, 17, 457-480.
(4) Schwarz, J. A.; Contescu, C.; Contescu, A. Chem. Rev. 1995, 95, 477-510.
(5) Pratsinis, S. E.; Zhu, W.; Vemury, S. Powder Technol. 1996, 86, 87-93.
(6) Strobel, R.; Mdler, L.; Piacentini, M.; Maciejewski, M.; Baiker, A.; Pratsinis, S. E.
Chem. Mater. 2006, 18, 2532-2537.
(7) Strobel, R.; Pratsinis, S. E. J. Mater. Chem. 2007, 17, 4743-4756.
(8) Schimmoeller, B.; Pratsinis, S. E.; Baiker, A. ChemCatChem 2011, 3, 1234-1256.
(9) Tsantilis, S.; Pratsinis, S. E. Langmuir 2004, 20, 5933-5939.
(10) Stark, W. J.; Wegner, K.; Pratsinis, S. E.; Baiker, A. J. Catal. 2001, 197, 182-191.
(11) Sokolowski, M.; Sokolowska, A.; Michalski, A.; Gokieli, B. J. Aerosol Sci. 1977, 8,
219-230.
(12) Marshall, B. S.; Telford, I.; Wood, R. Analyst 1971, 96, 569-578.
(13) Mdler, L.; Kammler, H. K.; Mueller, R.; Pratsinis, S. E. J. Aerosol Sci. 2002, 33,
369-389.
(14) Strobel, R.; Stark, W. J.; Mdler, L.; Pratsinis, S. E.; Baiker, A. J. Catal. 2003, 213,
296-304.
40 Chapter 1
(15) Thibaut, B. Platinum Met. Rev. 2011, 55, 149-151.

(16) Wegner, K.; Pratsinis, S. E. AlChE J. 2003, 49, 1667-1675.
(17) Demokritou, P.; Bchel, R.; Molina, R. M.; Deloid, G. M.; Brain, J. D.; Pratsinis, S. E.
Inhal. Toxicol. 2010, 22, 107-116.
(18) Waser, O.; Groehn, A. J.; Eggersdorfer, M. L.; Pratsinis, S. E. Aerosol Sci. Technol.
2014, 48, 1195-1206.
(19) Stark, W. J.; Madler, L.; Maciejewski, M.; Pratsinis, S. E.; Baiker, A. Chem. Commun.
2003, 588-589.
(20) Huang, J.; Jiang, Y.; van Vegten, N.; Hunger, M.; Baiker, A. J. Catal. 2011, 281, 352-
360.
(21) Strobel, R.; Grunwaldt, J.-D.; Camenzind, A.; Pratsinis, S.; Baiker, A. Catal. Lett.
2005, 104, 9-16.
(22) Dreyer, J. A. H.; Grossmann, H. K.; Chen, J.; Grieb, T.; Gong, B. B.; Sit, P. H. L.;
Mdler, L.; Teoh, W. Y. J. Catal. 2015, 329, 248-261.
(23) Hembram, K.; Sivaprakasam, D.; Rao, T. N.; Wegner, K. J. Nanopart. Res. 2013, 15,
1-11.
(24) Grhn, A. J.; Pratsinis, S. E.; Snchez-Ferrer, A.; Mezzenga, R.; Wegner, K. Ind. Eng.
Chem. Res. 2014, 53, 10734-10742.
(25) Athanassiou, E. K.; Krumeich, F.; Grass, R. N.; Stark, W. J. Phys. Rev. Lett. 2008,
101, 166804.
(26) Stark, W. J.; Strobel, R.; Gunther, D.; Pratsinis, S. E.; Baiker, A. J. Mater. Chem.
2002, 12, 3620-3625.
(27) Hoxha, F.; Schimmoeller, B.; Cakl, Z.; Urakawa, A.; Mallat, T.; Pratsinis, S. E.;
Baiker, A. J. Catal. 2010, 271, 115-124.
(28) Jiang, Y.; Scott, J.; Amal, R. Appl. Catal. B: Environ. 2012, 126, 290-297.
(29) Wang, Z.; Jiang, Y.; Hunger, M.; Baiker, A.; Huang, J. ChemCatChem 2014, 6, 2970-
2975.
(30) Chiarello, G. L.; Dozzi, M. V.; Scavini, M.; Grunwaldt, J.-D.; Selli, E. Appl. Catal. B:
Environ. 2014, 160161, 144-151.
(31) Weidenhof, B.; Reiser, M.; Stwe, K.; Maier, W. F.; Kim, M.; Azurdia, J.; Gulari, E.;
Seker, E.; Barks, A.; Laine, R. M. J. Am. Chem. Soc. 2009, 131, 9207-9219.
(32) Kim, M.; Laine, R. M. J. Am. Chem. Soc. 2009, 131, 9220-9229.
(33) Chiarello, G. L.; Rossetti, I.; Forni, L. J. Catal. 2005, 236, 251-261.
41
(34) Buchneva, O.; Rossetti, I.; Oliva, C.; Scavini, M.; Cappelli, S.; Sironi, B.; Allieta, M.;
Kryukov, A.; Forni, L. J. Mater. Chem. 2010, 20, 10021-10031.
(35) Roy, S.; van Vegten, N.; Maeda, N.; Baiker, A. Appl. Catal. B: Environ. 2012, 119
120, 279-286.
(36) van Vegten, N.; Baidya, T.; Krumeich, F.; Kleist, W.; Baiker, A. Appl. Catal. B:
Environ. 2010, 97, 398-406.
(37) Michalow-Mauke, K. A.; Lu, Y.; Kowalski, K.; Graule, T.; Nachtegaal, M.; Krcher,
O.; Ferri, D. ACS Catal. 2015, 5, 5657-5672.
(38) Bartholomew, C. H. Appl. Catal. A: Gen. 2001, 212, 17-60.
(39) Koirala, R.; Bchel, R.; Pratsinis, S. E.; Baiker, A. Appl. Catal. A: Gen. 2014, 484,
97-107.
(40) Tepluchin, M.; Kureti, S.; Casapu, M.; Ogel, E.; Mangold, S.; Grunwaldt, J. D. Catal.
Today 2015, 258, 498-506.
(41) Teoh, W. Y.; Mdler, L.; Beydoun, D.; Pratsinis, S. E.; Amal, R. Chem. Eng. Sci.
2005, 60, 5852-5861.
(42) Hoxha, F.; van Vegten, N.; Urakawa, A.; Krumeich, F.; Mallat, T.; Baiker, A. J.
Catal. 2009, 261, 224-231.
(43) Fotou, G. P.; Pratsinis, S. E. Chem. Eng. Commun. 1996, 151, 251-269.
(44) Kho, Y. K.; Iwase, A.; Teoh, W. Y.; Madler, L.; Kudo, A.; Amal, R. J. Phys. Chem. C
2010, 114, 2821-2829.
(45) Fujiwara, K.; Deligiannakis, Y.; Skoutelis, C. G.; Pratsinis, S. E. Appl. Catal. B:
Environ. 2014, 154155, 9-15.
(46) Schuh, K.; Kleist, W.; Hoj, M.; Trouillet, V.; Jensen, A. D.; Grunwaldt, J. D. Chem.
Commun. 2014, 50, 15404-15406.
(47) Piacentini, M.; Strobel, R.; Maciejewski, M.; Pratsinis, S. E.; Baiker, A. J. Catal.
2006, 243, 43-56.
(48) Kumar, V.; Lee, N.; Almquist, C. B. Appl. Catal. B: Environ. 2006, 69, 101-114.
(49) Teoh, W. Y.; Doronkin, D. E.; Beh, G. K.; Dreyer, J. A. H.; Grunwaldt, J.-D. J. Catal.
2015, 326, 182-193.
(50) Bchel, R.; Pratsinis, S. E.; Baiker, A. Appl. Catal. B: Environ. 2011, 101, 682-689.
(51) Hj, M.; Pham, D.; Brorson, M.; Mdler, L.; Jensen, A.; Grunwaldt, J.-D. Catal. Lett.
2013, 143, 386-394.
(52) Schimmoeller, B.; Schulz, H.; Pratsinis, S. E.; Bareiss, A.; Reitzmann, A.; Kraushaar-
Czarnetzki, B. J. Catal. 2006, 243, 82-92.
42 Chapter 1
(53) Wagloehner, S.; Nitzer-Noski, M.; Kureti, S. Chem. Eng. J. 2015, 259, 492-504.
(54) Stark, W. J.; Baiker, A.; Pratsinis, S. E. Part. Part. Syst. Charact. 2002, 19, 306-311.
(55) Biemelt, T.; Wegner, K.; Teichert, J.; Kaskel, S. Chem. Commun. 2015, 51, 5872-
5875.
(56) Zimmermann, H.; Walzl, R. Ullmann's Encyclopedia of Industrial Chemistry Wiley-
VCH: Weinheim, Germany, 2000.
(57) Ren, T.; Patel, M.; Blok, K. Energy 2006, 31, 425-451.
(58) Sattler, J. J. H. B.; Ruiz-Martinez, J.; Santillan-Jimenez, E.; Weckhuysen, B. M.
Chem. Rev. 2014, 114, 10613-10653.
(59) Cavani, F.; Trifir, F. Catal. Today 1995, 24, 307-313.
(60) Keller, G. E.; Bhasin, M. M. J. Catal. 1982, 73, 9-19.
(61) Lunsford, J. H. Catal. Today 2000, 63, 165-174.
(62) Hutchings, G. J.; Scurrell, M. S.; Woodhouse, J. R. Chem. Soc. Rev. 1989, 18, 251-
283.
(63) Arndt, S.; Laugel, G.; Levchenko, S.; Horn, R.; Baerns, M.; Scheffler, M.; Schlgl, R.;
Schomcker, R. Cataly. Rev. 2011, 53, 424-514.
(64) Godini, H. R.; Trivedi, H.; de Villasante, A. G.; Grke, O.; Jao, S.; Simon, U.;
Berthold, A.; Witt, W.; Wozny, G. Chem. Eng. Res. Des. 2013, 91, 2671-2681.
(65) Salehi, M.-S.; Askarishahi, M.; Godini, H. R.; Grke, O.; Wozny, G. Ind. Eng. Chem.
Res. 2016,
(66) Pak, S.; Qiu, P.; Lunsford, J. H. J. Catal. 1998, 179, 222-230.
(67) Wang, D.; Xu, M.; Shi, C.; Lunsford, J. H. Catal. Lett. 1993, 18, 323-328.
(68) Zavyalova, U.; Holena, M.; Schlgl, R.; Baerns, M. ChemCatChem 2011, 3, 1935-
1947.
(69) Kondratenko, E. V.; Schluter, M.; Baerns, M.; Linke, D.; Holena, M. Catal. Sci.
Technol. 2015, 5, 1668-1677.
(70) He, Y.; Yang, B.; Cheng, G. Catal. Today 2004, 98, 595-600.
(71) Arndt, S.; Otremba, T.; Simon, U.; Yildiz, M.; Schubert, H.; Schomcker, R. Appl.
Catal. A: Gen. 2012, 425426, 53-61.
(72) Wang, J.; Chou, L.; Zhang, B.; Song, H.; Zhao, J.; Yang, J.; Li, S. J. Mol. Catal. A:
Chem. 2006, 245, 272-277.
(73) Cavani, F.; Ballarini, N.; Cericola, A. Catal. Today 2007, 127, 113-131.
(74) Grtner, C. A.; van Veen, A. C.; Lercher, J. A. ChemCatChem 2013, 5, 3196-3217.
43
(75) Bhasin, M. M.; McCain, J. H.; Vora, B. V.; Imai, T.; Pujad, P. R. Appl. Catal. A:
Gen. 2001, 221, 397-419.
(76) Ansari, M. B.; Park, S.-E. Energy Environ. Sci. 2012, 5, 9419-9437.
(77) Nakagawa, K.; Kajita, C.; Okumura, K.; Ikenaga, N.-o.; Nishitani-Gamo, M.; Ando,
T.; Kobayashi, T.; Suzuki, T. J. Catal. 2001, 203, 87-93.
(78) Koirala, R.; Buechel, R.; Krumeich, F.; Pratsinis, S. E.; Baiker, A. ACS Catal. 2015, 5,
690-702.
(79) Nakagawa, K.; Okamura, M.; Ikenaga, N.; Suzuki, T.; Kobayashi, T. Chem. Commun.
1998, 1025-1026.
(80) Nakagawa, K.; Kajita, C.; Ide, Y.; Okamura, M.; Kato, S.; Kasuya, H.; Ikenaga, N.;
Kobayashi, T.; Suzuki, T. Catal. Lett. 2000, 64, 215-221.
(81) Wang, S.; Murata, K.; Hayakawa, T.; Hamakawa, S.; Suzuki, K. Appl. Catal. A: Gen.
2000, 196, 1-8.
(82) Zhao, X.; Wang, X. Catal. Commun. 2006, 7, 633-638.
(83) Mimura, N.; Okamoto, M.; Yamashita, H.; Oyama, S. T.; Murata, K. J. Phys. Chem. B
2006, 110, 21764-21770.
(84) Zhang, X.; Ye, Q.; Xu, B.; He, D. Catal. Lett. 2007, 117, 140-145.
(85) Corts Corbern, V.; Jia, M. J.; El-Haskouri, J.; Valenzuela, R. X.; Beltrn-Porter, D.;
Amors, P. Catal. Today 2004, 9192, 127-130.
(86) Zhu, J.; Qin, S.; Ren, S.; Peng, X.; Tong, D.; Hu, C. Catal. Today 2009, 148, 310-315.
44 Chapter 1
45
Chapter 2
2. Oxidative Coupling of Methane on Flame-made

Mn-Na2WO4/SiO2:Influence of Catalyst
Composition and Reaction Conditions
Abstract
Mn-Na2WO4/SiO2 catalysts of different composition were made in a single step by
flame spray pyrolysis, a process that can be scaled up to a production rate in the range of kg/h.
The multicomponent catalysts containing 0 - 5 wt% Mn and 0 - 6 wt% Na2WO4 were
characterized by nitrogen adsorption, XRD, TEM, STEM, EDXS and TPR, and tested in the
oxidative coupling of methane (OCM) in a continuous flow microreactor at different reaction
conditions (temperature, CH4/O2 feed ratio, space time). As-prepared catalysts showed much
higher specific surface area (SSA) and a more homogenous spatial distribution of the
constituents than corresponding catalysts prepared by wet-impregnation. Upon exposure to
reaction conditions at 810 C the amorphous SiO2 component gradually transformed to
crystalline cristobalite, as shown by in situ XRD. However, the cristobalite formation, which
was accompanied by a decrease of the SSA had only a marginal influence on catalyst
performance. Conversion and selectivity to C2 compounds (ethene, ethane) increased with Mn
content up to 1.9 wt% and remained virtually constant up to 5 wt%, affording a constant C2-
yield of 17%. Increasing the Na2WO4 content was most effective up to 1 wt%, while at
contents higher than 3 wt% the conversion declined. Only catalysts containing all metal
constituents (Mn, W, Na, Si) afforded good catalytic behavior. Flame-derived catalysts with a
composition 1.9%Mn-3%Na2WO4/SiO2 exhibited the highest C2-yield of 18.5 % at a
selectivity of 70%. Both, the flame-made catalyst and a corresponding wet-impregnated
reference catalyst showed stable conversion, but different C2-selectivity behavior with time-
on-stream, decreasing with the flame-derived catalyst and increasing with the wet-
impregnated counterpart. The initial C2-selectivity as well as that after 6 h on stream when a
steady-state was achieved was higher for the flame-made catalyst and this at considerably
46 Chapter 2
higher conversion, indicating the superior catalytic behavior of this catalyst. The better
performance of the flame-derived catalyst is traced back to its more homogeneous spatial
distribution of the constituents, which favors their beneficial mutual interaction.
Part of it is published in Journal of Applied Catalysis A: General, 484 (2014), 97-107.

47
2.1 Introduction
Ethylene can be synthesized by oxidative coupling of methane (OCM) in a single pass

reaction using natural gas, in contrast to current production by cracking naphtha and gas-oil
requiring multiple steps.1 After its discovery the OCM reaction has been in the focus of
several industrial and academic research laboratories.2 A variety of catalysts has been
investigated 3 after the initial report on OCM by Keller and Bhasin4. The reaction, however,
remains a challenge due to the still limited C2+-yield achieved and the lack of a complete
understanding of the relation between catalyst properties and reaction conditions and their
influence on reaction performance. Catalyst properties such as basicity5, specific surface area
6
as well as reaction conditions7 play an important role for achieving optimal C2+-yield.
Though some success has been reported, methyl radical generation8, non-selective oxidation9,
high reaction temperatures and catalyst stability10 are still some of the key obstacles in
achieving optimal catalytic and process systems. Multicomponent catalysts have shown
typically better performance compared to pure metal oxide based catalysts11 according to
recent statistical analysis of OCM catalysts3. Among them, Mn-Na2WO4/SiO2 is of significant
industrial interest due to its superior performance and stability.12,13
The Mn-Na2WO4/SiO2 catalysts have shown 20-37% CH4 conversion and 65-80% C2+-
selectivity in single pass reaction.13,14 To increase the product yield, addition of CeO215 or
replacing of W by V, Mo, Nb or Cr16 have been investigated. The C2+-yield increased from
13.8 to 21.1% with addition of CeO2 but it decreased upon replacing W. Furthermore, Mn-
Na2WO4 supported on MgO showed similar catalytic activity to that on SiO2 with conversion
and selectivity around 20% and 80%, respectively.13 During extended operation, however,
Mn-Na2WO4 supported on SiO2 was more stable than on MgO due to the catalysts ability to
retain the Mn ions in the near surface region.10
Wang et al.13 proposed that Mn is the active metal and the combination of Na-O-Mn
species forms an active center, whereas Na improves selectivity and W stabilizes the catalyst.
Wu et al.17 and Ji et al.18 proposed that the distorted tetrahedral structure of WO4 is
responsible for the activity where Na-O-W and Na-O-Mn could be the active center. Jiang et
14
al. demonstrated the formation of a structure containing a W=O and three W-O-Si surface
bonds from reconstruction of surface WO4 tetrahedral units. This was suggested to play a key
role in the reaction. Palermo et al.19 set forth that the SiO2 phase transition to cristobalite is
essential to selectivity and that Na plays a dual role in structural modification and as chemical
promoter.
48 Chapter 2
Flame spray pyrolysis (FSP)20 is a powerful tool for the synthesis of nanoparticles with
favorable properties for catalytic applications21. Flame-made catalysts exhibit high dispersion
of active metals, high thermal stability as well as specific surface area22. In addition, solid
solutions and mixed oxide phases can be easily obtained that have crucial catalytic or thermal
stability in various reactions.23 So the catalytic performance of flame-made catalysts is
comparable to the best conventionally prepared catalysts.24
Here, we have investigated the influence of catalyst composition on the structural and
catalytic properties of flame-made Mn-Na2WO4/SiO2 catalysts used for the OCM reaction.
The catalyst with optimal composition was studied under various reaction conditions to
elucidate the role of temperature, molar feed ratio, space time and the ratio of catalyst surface
area to volume in the fixed bed reactor. Finally, the best performing flame-derived catalyst is
compared to a corresponding catalyst prepared by classic wet-impregnation, which shows its
superior catalytic behavior.
2.2 Experimental
2.2.1 Catalyst preparation
Catalysts were made by FSP.25 The spray flame was ignited and supported using a pilot flame
of premixed CH4/O2 (>99.9%, Pangas) with flow rates of 1 and 2 L/min, respectively. By
means of a syringe pump (Lambda, VIT-FIT) 5 mL/min of precursor solution was injected
and dispersed into fine droplets by co-flowing 5 L/min of O2 (>99.9%, Pangas) maintaining a
pressure drop of 2 bar at the nozzle tip. The gas flow rates were controlled with mass flow
controllers (Bronkhorst). The metal concentration in all precursor solutions was 1 M.
Precursor solutions were prepared by dissolving in xylene (Sigma Aldrich) appropriate
amounts of manganese 2-ethylhexanoate (~6% Mn, Strem Chemicals), sodium 2-
ethylhexanoate (Strem Chemicals), tungsten (VI) ethoxide 5%w/v in ethanol (99.8% metal
basis, Alfa Aesar) and hexamethyldisiloxane (Sigma Aldrich). The fine powder products from
combustion of these precursors were collected 52 cm above the burner on water-cooled glass
fiber filters (Whatman GF 6, 25.7 cm in diameter) using a vacuum pump (Busch, Seco SV
1040C). The collected powders were used for the catalytic test without any further thermal
treatment as this synthesis process involves a severe temperature impact which makes
additional calcination obsolete.
As a reference, 1.9%Mn/SiO2 catalysts containing 3 and 5 wt% of Na2WO4 were also

prepared by a wet-impregnation according to the procedure described in ref.13. In brief, the
49
catalyst was prepared by mixing the slurry of silica gel (Davisil 636, 35-60 mesh particle size,
Sigma Aldrich) with appropriate concentration of Mn(NO3)3.4H2O (Sigma Aldrich) and
Na2WO4 (Sigma Aldrich) at 85 C. The solution was then heated to 130 C for 4-8 h which
was later collected (at room temperature) and calcined for 8 h at 800 C.
2.2.2 Catalyst characterization
The specific surface area (SSA) of the nanoparticles was measured by N2 adsorption at 77 K
(Tristar 3000, Micromeritics) applying the Brunauer-Emmett-Teller (BET) method. Prior to
measurement, the samples were preheated at 150 C for 1 h under nitrogen flow. X-ray
diffraction (XRD) measurements of powders were performed on a Bruker D8 Advance
diffractometer (40 kV, 40 mA, CuK) over the 2 range of 10-70. In situ high temperature
XRD was also performed in the same instrument using a high temperature cell-chamber
(Anton Paar, HTK 1200N) and heating the sample till 810 C at 10 C/min. Temperature-
programmed reduction (TPR) was carried out on a Micrometrics Autochem II 2920 with 40
mg of sample under flowing 5% H2 in Ar (Pan Gas, >99.999%, 30 mL/min) between 50 C
and 870 C with a heating rate of 10 C/min. High resolution transmission electron
microscopy (HR-TEM) images were obtained using a Tecnai F30 ST (FEI, FEG, operated at
300 kV) microscope. Scanning transmission electron microscopy (STEM) and energy-
dispersive X-ray spectroscopy (EDXS) studies were performed using Hitachi HD-2700CS
(aberration-corrected dedicated STEM, cold FEG, 200 kV). An EDX spectrometer (Gemini
system, EDAX) mounted on electron column right above the sample allowed record the
spectra of the selected spots.
2.2.3 Catalyst tests
The catalysts were tested in a quartz glass fixed-bed reactor of 4 mm inner diameter with co-
fed methane (CH4, 99.5% PanGas) and oxygen (O2, 99.95% PanGas) at total flow rate of 90
mL/min. The CH4 and O2 gas flow rates were regulated by mass flow controllers (Brooks)
and set to a ratio of 5, if not otherwise. The fixed bed consisted of 208 mg of catalyst (80-140
m mesh) and 292 mg of silicon carbide (SiC, 160-190 m mesh) and was fixed by quartz
wool. The ratio of catalyst weight to the reactant gas molar flow rate (W/F) was 3.1 kg.s/mol
which was kept constant in all tests unless otherwise mentioned. The furnace temperature was
increased to 810 C at a rate of 10 C/min. Both inlet and outlet reaction temperatures of the
catalyst bed was measured by inserting a thermocouple axially which showed a maximum
50 Chapter 2
temperature difference between outlet and inlet of 13 C at the highest conversion. Since the
catalyst bed was diluted with SiC to decrease the temperature gradient, the radial temperature
gradients are expected to be small in our 4 mm diameter reactor. The reactor inner diameter
was reduced to 2 mm after the catalyst bed to reduce the dead volume. After reaching the
reaction temperature, CH4 and O2 gas was switched on and the catalyst performance
monitored for 2 h, if not stated otherwise. Negligible conversion (1-2%) was observed with a
reactor filled with only quartz wool and SiC. The outlet gas was analyzed using on-line gas
chromatography (HP-PLOT Q column 30 m long, 0.32 mm in diameter and 0.2 m thick
fitted with TCD and FID, Agilent Technologies). In addition, the exhaust gas was monitored
continuously using a non-dispersive infrared analyzer for CO and CO2 (ABB Uras 26) and a
quadrupole mass spectrometer (Pfeiffer Vacuum, Thermostar GSD 300). The carbon mass
balance was calculated using the amount of methane (C1) fed and all the C1 and C2 products
(i.e. CO, CO2, C2H4, C2H6) plus the unreacted methane. The balance was always better than
90%.
Methane conversion, C2-selectivity and C2-yield were calculated as follows:
moles of CH4 consumed

CH4 conversion (%) = 100 (2.1)
moles of CH4 in the feed
(n moles of C2 products)
Selectivity towards C2 (%) = 100 (2.2)
moles of CH4 consumed
CH4 conversion (%) C2 selectivity(%)

Yield(%) = (2.3)
100
The effect of a change of the catalyst surface (S) to void volume of the catalytic fixed bed (0.5
cm3, bed length 4 cm) containing SiC as inert diluent was investigated by progressively
replacing SiC by catalyst while keeping the bed length and interparticle void volume constant.
2.3 Results and Discussion
2.3.1 Structural properties of catalysts
Particle formation in flame depends among other parameters on the difference in volatility of
the materials that result in segregated particle components due to their sequential
formation.26,27 Since SiO2 is less volatile than the other metal oxides, during flame synthesis,
51
it condenses earlier followed by the sequential condensation of other components. Therefore,

metal dopants, Mn, Na and W are expected to be preferentially deposited on the SiO2 surface.
These flame-made Mn-Na2WO4/SiO2 catalysts at lower loadings (1 wt%) of both Mn and
Na2WO4 showed a surface area >300 m2/g which decreased upon increasing their loadings.
The reduction was mainly observed during the addition of Na2WO4 in 1.9%Mn/SiO2 resulting
SSA of 210 m2/g at 6 wt% loading. Substantial reduction in SSA was observed when these
catalysts were exposed to catalytic tests at high temperature. All standard catalytic tests
involved heating of the catalyst bed to 810 C in helium and then switch of the gas flow to
CH4/O2 feed and reaction for 2 h. The SSA of 1.9%Mn/SiO2 decreased from 409 to 239 m2/g
and that for 1.9%Mn-4%W/SiO2 from 415 to 189 m2/g after the standard test. Sodium
containing catalysts were particularly prone to sintering. With these catalysts the SSA
decreased to less than 5 m2/g after standard test.
300
specific surface area, m2/g
275 as-prepared
250
15
10
5 spent
0
0 1 2 3 4 5
Mn in 3%Na2WO4/SiO2, %
Figure 2.1: BET surface areas of flame-made Mn-doped 3%Na2WO4/SiO2 catalysts.

Comparison between as-prepared and spent catalysts (after catalytic test at 810 C for 2 h).
The SSA of both fresh and spent (0-5)%Mn-3%Na2WO4/SiO2 catalysts are shown in Figure
2.1. For the as-prepared catalysts, higher Mn loading led to lower SSA, which is attributed to
increased sintering of silica in the presence of Mn.28 The pore volume of these flame-made
Mn-Na2WO4/SiO2 powders was around 0.12 cm3/g and its reduction upon exposure to
52 Chapter 2
reaction conditions was inevitable due to severe sintering. The SSA of both wet-prepared
reference catalysts, 1.9%Mn-3%Na2WO4/SiO2 and 1.9%Mn-5%Na2WO4/SiO2, was around 2
m2/g which is comparable to that reported by Wang et al..13 Moreover, the wet-impregnated
catalysts had a very low pore volume of around 0.006 cm3/g, which is much lower than that of
the fresh silica gel support (0.75 cm3/g). These results suggest severe sintering and pore
mouth plugging, potentially by melt of Na2WO4, of these catalysts during the phase
transformations induced by the calcination step.29
Figure 2.2: TEM images of as-prepared and spent flame-made (a,b) and wet-prepared (c,d)
1.9%Mn-3%Na2WO4/SiO2 catalysts.
Figure 2.2 shows TEM images of both as-prepared and spent flame-made and wet-
impregnated 1.9%Mn-3%Na2WO4/SiO2 catalysts. During reaction, the particle size increased
from about 7.8 nm to 5.2 m for 1.9%Mn-3%Na2WO4/SiO2. The as-prepared flame-made
particles (Fig. 2a) were amorphous, whereas larger crystalline particles were observed (Figure
2.2b) after reaction. In the spent catalyst, a change in surface concentration of active
components can be expected due to their potential coverage by a homogeneous layer of melt.
During the exposure to catalytic tests probably a melt composed of Na 2WO4 was formed,
53
since, this component has a melting temperature of about 700 C.30 Similar morphological
changes were also observed for the wet-prepared catalysts (Figure 2.c,d). Larger particles (in
m range) with some size variation were observed in fresh samples. These particles were
irregular with rough surface, however, after catalytic tests large particles (> 0.2 m) with
smooth surface were discernible indicating that further sintering of particles has occurred
during reaction.
In addition to the textural differences of the flame- and wet prepared catalysts
described above, STEM and EDXS analyses of both flame and wet-made 1.9%Mn-
3%Na2WO4/SiO2 catalysts revealed significant structural differences (Figure 2.3). These
investigations showed that the flame-made catalyst is made up of nm-sized particles, which
are homogeneously distributed on the SiO2 support. This well-distributed nanoparticles
contained W and/or Mn. In wet-made catalysts, larger crystals (ca. 50 nm) were observed with
more frequent MnOx crystals. EDXS showed the presence of Mn, Na, W, and Si, except in
few cases where Na was not detected probably due to its low contrast (Z = 11), compared to
the other elements. The signals due to Si and W at about 1.8 keV overlap and this in
combination with the low W concentration made quantification difficult.31 Narrow focused
EDXS measurements showed the presence of W in the flame-made powder, whereas W was
scarcely detected in wet-made ones, indicating a more homogeneous spatial distribution of
these elements in the flame-derived catalyst. The more homogeneous distribution of the
constituents of the flame-derived catalyst was further corroborated by analysis of larger
selected areas. In wet-made catalyst, the constituents were mainly agglomerated due to its low
surface area. Overall, it can be inferred that the flame-made catalysts showed a significantly
better homogeneous spatial distribution of the elements compared to the wet-made
counterpart.
54 Chapter 2
a b
c d
e f
Figure 2.3: STEM and EDXS analyses of as-prepared flame (a,b) and wet-made (c-f)
1.9%Mn-3%Na2WO4/SiO2 catalysts.
All as-prepared flame-made powders were XRD amorphous. Figure 2.4a shows XRD
patterns of 1.9%Mn-(0-6)%Na2WO4/SiO2 catalysts after catalytic tests. Cristobalite and
quartz SiO2 phases were present in all catalysts, only those without Na remained completely
amorphous. Low intensity reflections at 16, 27.5 and 31.5 corresponding to Na2WO4 were
detected similar to the observation made by Simon et al..29 The typical MnWO4 reflection was
only detected for the 1.9%Mn-4%W/SiO2 (without Na). However, the possible presence of
55
amorphous or very small crystals of MnWO4 cannot be ruled out. In contrast to the
investigation by Simon et al.29, tridymite SiO2 was absent, instead quartz SiO2 was detected,
similar to the observation made by Dedov et al..25 In the 1.9%Mn-4%W/SiO2 catalyst, the -
cristobalite SiO2 signals were very weak compared to catalysts containing Na (Figure 2.a),
indicating that Na promotes the cristobalite formation. This transformation occurred already
at 810 C which is much lower than 1500 C32, where amorphous silica is transformed to
cristobalite. This observation corroborates the behavior reported by Palermo et al..19
a) cristobalite SiO2 b) cristobalite SiO2

# quartz SiO2 # quartz SiO2
Na2WO4 Na2WO4
MnWO4
#

intensity, a.u.
intensity, a.u.

#

1.9%Mn-6%Na2WO4/SiO2 #
# 5%Mn-3%Na2WO4/SiO2
1.9%Mn-5%Na2WO4/SiO2
4%Mn-3%Na2WO4/SiO2
1.9%Mn-3%Na2WO4/SiO2 3%Mn-3%Na2WO4/SiO2
1%Mn-3%Na2WO4/SiO2
1.9%Mn-4%W/SiO2
1.9%Mn/SiO2 3%Na2WO4/SiO2
10 20 30 40 50 60 70 10 20 30 40 50 60 70
2 degree
2 degree
Figure 2.4: XRD patterns of various spent flame-made Mn-Na2WO4/SiO2 catalysts.
Figure 2.4b shows XRD patterns of spent (0-5)%Mn-3%Na2WO4/SiO2 catalysts. Major

reflections belonging to cristobalite and quartz are discernible, similar to those in Figure 2.4a.
With increasing Mn content, the quartz reflections became stronger. Among the various
phases detected (cristobalite, quartz, Na2WO4 and Mn2O3), the Mn2O3 phase of wet-prepared
catalyst disappeared after the reaction. The disappearance of this phase could be linked to its
56 Chapter 2
possible active participation in the reaction. Lower number of reflections belonging to quartz
were found in spent wet-prepared catalysts containing higher Na2WO4 content, indicating that
Na2WO4 inhibits quartz formation. No other bimetallic compounds were detected except
Na2WO4. Apart from stabilizing W by forming Na2WO4, Na seemed to also interact with
amorphous SiO2 transforming it primarily to crystalline cristobalite (Figure 2.4a,b). This
suggests that Na plays a dual role as structural and chemical promoter.19
1.9%Mn(X)%Na2WO4/SiO2
X = 5, FSP
X = 3, FSP
X = 5, FSP-calcined
X = 3, FSP-calcined
TCD signal, a.u.
X = 5, Wet
X = 3, Wet
Mn2O3, FSP WO3, FSP
200 300 400 500 600 700 800

temperature,C
Figure 2.5: Temperature programmed reduction (TPR) of as-prepared and calcined FSP-
made and wet-prepared reference catalysts.
Figure 2.5 compares the TPR profiles of flame- and wet-prepared catalysts. In addition,
TPR profiles of flame-made pure Mn2O3 and WO3 powders are also presented as a reference.
As-prepared FSP powders showed a main reduction peak at about 840 C which is more than
100 C higher than that of the calcined flame-made or the wet-prepared catalysts that
exhibited broader peaks at lower temperatures (696-760 C), indicating easier reducible metal
57
oxides with bulk properties.33 For flame-made catalysts, a slight shift of the reduction peak
from 836 to 844 C was observed at the highest Na2WO4 content (X = 5 wt%). The wet-
prepared 1.9%Mn-3%Na2WO4/SiO2 catalyst showed a major broad peak around 765 C with
a small shoulder at about 660 C. The broad major peak split into two, positioned at about
730 C and 810 C, when the Na2WO4 content was increased to 5 wt%.
The high temperature reduction peaks (> 550 C) of all the powders are attributed to the
reduction of W-species. The reduction of bulk WO3 is reported to occur in three steps: WO3
WO2.9 WO2 W, however, the number of reduction peaks was found to be highly
dependent on the sample size.34 These reduction peaks are reported to be highly influenced by
the type of support and the size of the WOx domains impairing any direct comparison.33 The
low temperature peaks (ranging from 400-600 C) can be assigned to the reduction of Mn-
species which is also reported to occur in three steps: MnO2 Mn2O3 Mn3O4
MnO.35,36
Both flame-made fresh and calcined samples were investigated in air using TGA-MS.
Only a small weight loss of ~4% in fresh and <1% in calcined samples was measured mainly
due to evolving water and negligible CO2. Thus, reduction peaks due to reduction of
carbonaceous residues can be ruled out in the TPR profiles shown in Figure 2.5.
2.3.2 Catalytic tests
2.3.2.1 Effect of Mn, Na2WO4 and SiO2 content in OCM
Figure 2.6 shows, CH4 conversion, C2-selectivity and C2-yield, and CO/CO2 ratio in the
reactor effluent as a function of Na2WO4 content. Pure 1.9%Mn/SiO2 (without Na2WO4)
showed a very low C2-yield of 2.2% at 15% conversion. Addition of Na2WO4 improved the
performance: for a loading of 1 wt% Na2WO4, the C2-yield was 13.5% and increased to 16%
for 3 wt% Na2WO4. Higher than 3 wt% Na2WO4 did not further improve the catalytic
performance. Also Jiang et al.14 with wet-prepared 1.9%Mn/SiO2 catalysts found an optimal
loading at 4 wt% Na2WO4. For 1.9%Mn/SiO2 with more than 4 wt% Na2WO4 a decrease in
conversion and yield was observed that could have originated from the coverage of active
surface sites by Na2WO4 as its melting point (~700 C) is below the reaction temperature
810 C.
58 Chapter 2
50 100
conversion or yield, %
40 80
C2-selectivity
C2-selectivity, %
30 60
conversion
20 40
C2-yield
10 20
0 0
2.4
1.5
2.0
C2H4/C2H6, -
CO/CO2, -
1.0
1.6
1.2
0.5
0.8
0.0 0.4
0 1 2 3 4 5 6
Na2WO4 content in 1.9%Mn/SiO2, wt %

Figure 2.6: Influence of Na2WO4 content in the performance of flame-made 1.9%Mn-
Na2WO4/SiO2 catalyst for oxidative coupling of methane (OCM).
The TEM image of the spent catalyst (Figure 2.2b) supports this explanation.
Moreover, complete coverage of active metal oxides, e.g. MnOx13 can be expected with
increasing Na2WO4 loading. Improvement of catalytic activity was reported upon loading
with small amounts of Na2WO4 i.e. 2.2 wt% on 2%Mn/SiO2 due to prevention of the
59
formation of large agglomerate clusters.37 Na was found to be essential for the activity since
Na-free 1.9%Mn-4%W/SiO2 showed a conversion and selectivity of 22% and 45% (yield
~10%), respectively, which is significantly less than that of the catalyst containing 4 wt%
Na2WO4 (yield 17%) (Figure 2.6). Similar behavior was reported by Ji et al.18 and Palermo et
al.19, with low C2-yield of 0.5% and 6%, respectively. Na promotes the formation of distorted
WO4 tetrahedrons which could be the active center for the reaction.14,17,38 In the absence of
Na, formation of octahedral WO6 takes place which exhibits a lower catalytic activity and
selectivity likely due to presence of non-selective pathways.18,19 In all the flame-made
catalysts, the tetrahedral phase (Na2WO4) was detected only after reaction indicating its
presence in as-prepared catalysts as amorphous phase or very small crystals which can go
undetected by XRD. This tetrahedral phase was found to promote selective reaction pathways,
as reflected by the strong increase in selectivity upon addition of this component (Figure 2.6).
The selectivity remained virtually constant, though conversion decreased at higher loading of
Na2WO4 (>3%). No significant trend in the CO/CO2 molar ratio was observed when
increasing the Na2WO4 loading in 1.9%Mn/SiO2 while the C2H4/C2H6 ratio showed a
maximum at around 3 wt% (Figure 2.6, bottom).
The effect of Mn loading on the catalytic performance of the optimal 3%Na2WO4/SiO2
is shown in Figure 2.7. In the absence of Mn, <1% yield at ~5% conversion was observed.
Addition of Mn improved both conversion and selectivity with a maximum yield of 18.5% at
1.9 wt% Mn loading. Further increase in Mn content showed a slight decrease in conversion
and selectivity with a C2-yield of about 17%. These results indicate that the presence of Mn is
crucial for high conversion and selectivity.39 This point has been addressed by Wang et al. 13
who suggested that Na-O-Mn species could be the active site for the reaction. However, there
are also reports on undiminished catalytic activity in the absence of Mn, that is for
Na2WO4/SiO2.19 Our results show that high catalytic activity can be obtained only when Mn
and Na2WO4 were supported on SiO2. The decrease in C2-yield with increasing Mn loading
has been attributed to an accumulation of MnOx on the catalyst surface.18,40 Therefore,
decrease in C2-yield is expected at higher Mn loadings (>5 wt%) most likely due to formation
of MnMn6SiO12, which has been reported to favor non-selective reaction pathways.18
Nevertheless, MnOx (most likely Mn2O3) favored the selective reaction pathways evident by
increase in selectivity which remained virtually constant until the Mn loading of 5%. The
CO/CO2 molar ratios decreased with increasing Mn loadings in the 3%Na2WO4/SiO2 catalyst
indicating that the formation of CO2 is favored by Mn (Figure 2.7, bottom).
60 Chapter 2
conversion or yield, % 50 100
40 80
C2-selectivity
C2-selectivity, %
30 60
conversion
20 40
C2-yield
10 20
0 0
2.4
1.2
2.0
1.0 1.6
C2H4/C2H6, -
CO/CO2, -
1.2
0.8
0.8
0.6
0.4
0.4 0.0
0 1 2 3 4 5
Mn content in 3%Na2WO4/SiO2, wt %
Figure 2.7: Influence of Mn content in the performance of flame-made Mn-3%Na2WO4/SiO2

catalysts for oxidative coupling of methane (OCM).
The C2H4/C2H6 ratio depended on the Na2WO4 and Mn content in the catalysts. The
ratio was about 0.7 in the absence of Na2WO4 in 1.9%Mn/SiO2, which increased and showed
maxima (~1.9) at 3 wt% loading (Figure 2.6, bottom). A similar trend was observed with
increasing Mn content from 0 to 1.9% in 3%Na2WO4/SiO2, with a maximum ratio of 1.9
(Figure 2.7, bottom). The ratio decreased to about 1.5 with 3%Mn loading and remained
61
virtually constant on further addition of Mn (>3%). Based on these results, it can be inferred
that in the absence or at lower concentration of both Na2WO4 and Mn, formation of ethane
increases.
100
CH4 conversion
conversion or selectivity or yield, %
C2-selectivity
C2-yield
80
Wet method
FSP FSP-calcined
800 C
2h
8h
60
40
20
0
3 wt. % 5 wt. % 3 wt. % 5 wt. % 3 wt. %
Na2WO4 content in 1.9%Mn/SiO2, wt.%
Figure 2.8: Comparison of performance of fresh and calcined flame-made (FSP) and wet-
prepared 1.9%Mn-3%Na2WO4/SiO2 and 1.9%Mn-5%Na2WO4/SiO2 catalysts.
In Figure 2.8, the catalytic performance of the best FSP-made catalysts is compared to
wet-prepared catalysts of similar composition. The wet-prepared 1.9%Mn/SiO2 catalyst
containing 5%Na2WO4 showed comparable performance to that of the corresponding flame-
made 1.9%Mn-3%Na2WO4/SiO2 with a yield of 17.8%. XRD patterns of spent wet-prepared
catalysts showed the disappearance of the Mn2O3 phase supporting the argument on the
crucial role of Mn for activity. The difference in the catalytic performance of the flame-made
62 Chapter 2
and wet-prepared catalysts with similar composition (Figure 2.8) is not so prominent in view
of the many influencing factors such as distribution of active metals arising from different
synthesis techniques41,42, and precursors types.43
Overall, it can be inferred that all the elements (Mn, Na, W and Si) are necessary to
get the optimal performance. Their individual role during the reaction as metal oxide or mixed
oxides seemed to be similar in both the flame and wet-made catalysts. The more
homogeneous spatial distribution and thus more intimate mixing of these constituents in the
flame-made catalysts seem to be beneficial for their catalytic behavior. Owing to the
complexity of their interactions and melting of Na2WO4 during the reaction, the possible
reaction pathways can only be speculated based on literature data.44 A redox reaction
mechanism is expected involving Mn and W-based solid phases, where O2 activation occurs
on Mn (Mn3+/Mn2+), which initiates the formation of CH3 radical (W6+/W4+) on the surface.
Ethane is produced by coupling of two CH3 radicals which is assumed to produce via
dehydrogenation route to give C2H4.
2.3.2.2 Influence of reaction conditions
In order to elucidate the influence of the heat treatment during calcination, the flame-made
powders were also tested after calcination at 800 C for 2 and 8 h, respectively. All the
calcined flame-made catalysts showed stable conversion and selectivity (in the 2 h
experiments) regardless of the pretreatment duration. However, the catalytic performance was
inferior to that of corresponding fresh catalysts, most likely due to the structural changes
(partial crystallization) induced during calcination. For the fresh 1.9%Mn-3%Na2WO4/SiO2,
the C2-yield decreased from 18.5% to 16% after 2 h and to 13% after 8 h of calcination
(Figure 2.8). This shows that calcination of the flame-made catalysts is not beneficial.
Calcination not only results in a loss of SSA but also changes in surface composition due to
phase changes.45 XRD results of calcined samples show two distinct phases, cristobalite SiO2
and Na2WO4, and a very low intensity peak corresponding to quartz SiO2 at about 2 = 20.8
irrespective of the duration of calcination. XRD measurements of these pretreated catalysts
after the reaction showed patterns similar to those of spent 1.9%Mn-3%Na2WO4/SiO2 (Figure
2.4). The increase in intensity of the XRD peak indicative of quartz SiO 2 in these pretreated
catalysts only after reaction indicates that this phase formation is favored mainly under the
reaction conditions.
In the XRD patterns of spent catalysts (Figure 2.4a,b), a crystal phase change is
discernible with increasing Mn or Na2WO4 content. To clarify whether this change affected
63
the catalytic performance, in situ XRD measurements of 1.9%Mn-3%Na2WO4/SiO2 were

carried out at 810 C for 2 h. Figure 2.9 shows the gradual formation of -cristobalite (2 =
22) resulting in a crystal size of about 64 nm after 2 h. Interestingly, the cristobalite
formation did not influence much the catalytic performance (yield ~18.5%), which remained
virtually constant during this transformation. Therefore, it appears that the catalytic
performance is if at all only marginally affected by this phase transformation which is
accompanied by a massive loss in SSA from 271 to 2 m2/g. No reflection indicative of
Na2WO4 and manganese oxides were discernible during in situ measurement at 810 C. This
indicates that the Mn2O3 phase might exist in X-ray amorphous form, whereas Na2WO4 has
molten (~700 C melting point).
cristobalite SiO2

# quartz SiO2
intensity, a.u.

18.5
120
100
18.8
80
in
60
m
e,
tim
40
18
C2-yield, % 20
19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0

2, degree
Figure 2.9: In situ high temperature XRD at 810 C of flame-made 1.9%Mn-3%Na2WO4/SiO2
and corresponding C2-yields.
64 Chapter 2
conversion or yield, % 50 100
40 80
C2-selectivity, %
C2-selectivity
30 conversion 60
20 40
C2-yield
10 20
0 0
3
CO/CO2
0
760 780 800 820 840 860 880 900
reaction temperature, C
Figure 2.10: Temperature dependence of catalytic performance of 1.9%Mn-3%Na2WO4/SiO2.
Conditions: total gas flow rate, 90 mL/min; CH4/O2 ratio, 5; W/F, 3.1 kg.s/mol.
Figure 2.10 shows the effect of temperature on the catalytic performance of flame-
made 1.9%Mn-3%Na2WO4/SiO4. CH4 conversion at 770 C is 16% and increases with
increasing temperature until 810 C to 27%. This indicates that the catalyst was not able to
significantly reduce the activation energy of methane as high thermal energy was required to
break the strong C-H bond of methane (bond dissociation energy of 439 kJ/mole).46
Maximum conversion of ~27% was obtained at 810 C, with a minimal decrease at higher
temperatures in agreement with the behavior reported for sol-gel-made catalysts.41 This could
be attributed to the alterations in the physicochemical properties of the catalyst induced by
significant reduction in SSA at high reaction temperatures ( 830 C). Though high
65
temperature is detrimental as it promotes sintering resulting in change of catalyst surface

properties, it is essential for the activation of CH4 to dissociate the C-H bond. Molar ratios of
CO/CO2 increased at higher temperature probably due to the contribution of the reverse water
gas shift reaction (CO2 + H2 CO + H2O)47 as observed in other Mn-Na2WO4/SiO2 OCM
studies.15
50 100
40 80
C2-selectivity, %
C2-selectivity
30
60
conversion
20
40
C2-yield
10
20
0
0.9
CO/CO2
0.8
0.7
0.6
0.5
2 3 4 5 6
CH4/O2, -
Figure 2.11: Dependence of catalytic performance of 1.9%Mn-3%Na2WO4/SiO2 on CH4/O2
feed ratio. Conditions: temperature, 810 C; total gas flow rate, 90 mL/min; CH4/O2 ratio, 5;
W/F, 3.1 kg.s/mol.
In Figure 2.11, the effect of the molar reactant gas ratio, CH4/O2, on the catalytic
performance was investigated at 810 C. Lower conversion was observed when increasing the
CH4/O2 ratio from 2 to 6, accompanied by increasing C2-selectivity. At low CH4/O2 ratios,
total oxidation of CH4 is favored and also C2-products (C2H4 and C2H6) may be further
66 Chapter 2
oxidized, resulting in decreased selectivity. The highest yield of 22.5% was achieved at
CH4/O2 = 2, mainly due to high CH4 conversion.7 Similar observation was reported by Ghose
et al.48 who achieved a yield of ~25% with the same feed ratio which also decreased above at
750 C. This steady decrease of both conversion and yield at higher CH 4/O2 ratios can be
attributed to an oxygen deficient reaction environment.49 Takanabe and Iglesia7 reported that
the presence of H2O favors the OH-mediated pathways that ease the H-abstraction from
hydrocarbons. However, this pathway becomes complicated as prevalent H2O concentration
varies with conversion along the reactor and so does the effect of O2 pressure, complicating
the system. The CO/CO2 molar ratio increased up to a CH4/O2 ratio of 4 and then slightly
decreased at higher values.
W/F, kg.s/mol
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
50 100
40 C2-selectivity 80
C2-selectivity, %
30 60
conversion
20 40
10 C2-yield 20
0 0
3
CO/CO2
0
0 50 100 150 200 250 300
2 3
catalyst surface area/bed void volume, m /cm
Figure 2.12: Influence of ratio of catalyst surface area to void volume of fixed bed (S/V) on
the performance of 1.9%Mn-3%Na2WO4/SiO2. Conditions: temperature, 810 C; total gas
flow rate, 90 mL/min; CH4/O2 ratio, 5; W/F, 3.1 kg.s/mol.
67
In order to investigate the influence of catalyst bed dilution with inert material, a series
of experiments have been performed with different amount of SiC (diluent). In these
experiments the volume of the fixed bed was kept constant. Catalyst and diluent were well-
mixed to reduce the possibility of by passing the catalyst surface by the reactant gas. 50
Dilution has two effects, possible local hot spots can be suppressed and the ratio of catalyst
surface to void volume of the bed (S/V) is changed.51 The latter may alter the contributions of
heterogeneous (occurring at the catalyst surface) and possible homogeneous gas phase
reaction (occurring in the void space of the catalyst bed).
Figure 2.12 shows that the S/V ratio has a significant influence on the catalyst activity,
optimal performance (23.5% conversion at 75% C2-selectivity) was achieved at S/V ratios of
about 150 m2/cm3. At S/V >150 m2/cm3, a slight decrease in selectivity was observed, which
may be attributed to further oxidation of C2-compounds. Very low conversion of 2% with the
catalyst bed (length 4 cm) containing only inert SiC indicates a contribution of homogeneous
reaction pathways at 810 C and a CH4/O2 ratio of 5. Contribution from homogeneous
reaction in OCM has been previously addressed by Hatano et al.52 and Kalenik and Wolf53. In
the latter study the authors showed that methane conversion from homogeneous pathways
decreased when the free space in the reactor was reduced by filling it with inert quartz chips.
However, the results were reported to highly depend on residence time, gas composition,
temperature and even the geometry and contamination of the reactors. Nevertheless, the
increase in CH4 conversion shown in Figure 2.12 with increasing catalyst amount (S/V ratio
or W/F) can be mainly attributed to heterogeneous reaction. The effect was found to be more
significant at lower catalyst loadings (<150 m2/cm3 of S/V).
Formation of hot spots in the catalyst bed has been reported by Pak et al.10 for Mn-
Na2WO4/SiO2 catalysts during OCM at 800 C, inducing adverse effects on the catalytic
activity. Lee et al.54 also reported hot spot formation, increasing the reactor temperature as
high as 150 C, in mid- and bench-scale reactors due to poor heat transfer. Possible hot spot
formation can also not be completely ruled out in our micro-scale reactor, though it is less
likely.
The results shown in Figure 2.12 can also be interpreted as the influence of the contact
time (W/F) on the catalytic performance with optimal activity at 1.9 kg.s/mol. The decrease in
selectivity with increasing contact time could be attributed to further oxidation of C2 products.
CH4 conversion increased with increasing contact time and remained constant at higher W/F
values (1.9-3.6 kg.s/mol), however, this behavior has been reported to be strongly influenced
by the reaction conditions.7 The CO/CO2 ratio in the reactor effluent significantly decreased at
68 Chapter 2
low S/V (W/F) reaching a constant level of ca. 0.5 at an S/V of about 35 m 2/cm3 or at W/F =
0.4 kg.s/mol. This suggests that the formation of CO2 is potentially favored by the
heterogeneous pathways.7
80
flame-made
wet prepared
conversion or selectivity, %
60
C2-selectivity
40
conversion
20
0
0 100 200 300 400 500 600
time, min
Figure 2.13: Methane conversion (filled black symbols) and selectivity (open red symbols) by
flame- (square) and wet-made (triangle) 1.9%Mn-3%Na2WO4/SiO2 catalysts as a function of
time-on-stream. Conditions: temperature, 810 C; total gas flow rate, 90 mL/min; CH4/O2
ratio, 5; W/F, 3.1 kg.s/mol.
Finally, time-on-stream tests of both flame- and wet-made 1.9%Mn-3%Na2WO4/SiO2

were performed. Conversion and selectivity as a function of time-on-stream are presented in
Figure 2.13. A steady small increase in CH4 conversion was observed for the flame-made
catalyst, whereas conversion remained virtually unchanged for the wet-prepared catalyst. CH4
conversion after 10 h on stream was 29% and 19.0% for flame- and wet-made catalysts,
respectively. In the initial non-steady state period, the C2-selectivity of both catalysts showed
opposite trends, i.e. increasing in wet-made and decreasing in flame-made catalysts. However,
69
after about 6 h, when steady-state was reached the selectivity of the flame-made catalyst was
60% while that of the wet-impregnated reference catalyst was only 49% and this at
significantly lower conversion. The C2-yield achieved with the flame-made catalyst (17%)
was almost twice that of the wet-made counterpart (9 %). BET measurement of these spent
catalysts showed an SSA of about 2 and 1.7 m2/g for wet-and flame-made catalysts,
respectively. These results are similar to their SSA after 2 h of reaction indicating that an
increase in duration of the catalyst test had virtually no effect on the surface area anymore.
XRD analyses of the spent catalysts were performed to investigate the change in the bulk
properties of the catalysts after the stability tests. It indicated that the major structural change
occurs in the SiO2 phase. Transformation of cristobalite to quartz SiO2 was observed with
both catalysts. MnWO4 was identified at 2 of 18.5, 23.6, 24.2, 29.9 and 30.4, whereas
no traces of Mn2O3 were discernible, indicating its active participation in the reaction in wet-
prepared catalysts, as mentioned earlier. Moreover, very low intensity reflections indicative of
MnWO4 were also observed at 24.2 and 30.2 in spent flame-made catalysts, though no
phases of Mn were identified in XRD patterns of both fresh and calcined catalysts. Na2WO4
reflections were present in both samples indicating that their crystallinity is retained even after
reaction.
The observed dissimilar changes in selectivity in the time-on-stream tests for flame-
and wet-made catalysts are attributed to the different textual and structural properties of these
catalysts generated by the different preparation methods and the fact that the wet-made
catalyst had to be subjected to calcination prior to use, whereas the flame-made catalysts
could be used as-prepared. These structural differences seem to be partly eliminated if the two
differently prepared catalysts are exposed to longer time-on-stream. It has been shown that the
near surface composition of fresh catalysts is depended on the synthesis method and changes
can be expected during high temperature reaction.41 Changes could be induced by Na2WO4 as
its melting temperature is lower than the reaction temperature, while the phase transformation
of SiO2 had virtually no influence, as previously reported.41 Another reason for the change of
catalyst performance with time-on-stream could be related to the formation of MnWO4, which
contains distorted WO42- tetrahedrons, that are thought to be essential for catalytic
performance.29 Gradual formation of the MnWO4 phase in spent flame-made catalysts was
observed by XRD.
70 Chapter 2
2.4 Conclusions
Multicomponent Mn-Na2WO4/SiO2 catalysts with various compositions were prepared by

flame synthesis and their performance in the oxidative coupling of methane was investigated
in a flow micro-reactor at different conditions (temperature, CH4/O2 feed ratio, space time,
ratio of catalyst surface to void volume of fixed bed). Among these catalysts with varying Mn
(0-5wt%) and Na2WO4 (0-6 wt%) compositions, 1.9%Mn-3%Na2WO4/SiO2 was the most
efficient under the standard reaction conditions used (810 C, feed ratio CH4/O2 = 5, W/F =
3.1 kg.s/mol) affording 18.5% yield at 70% C2-selectivity. The presence of all components
(Mn, W, Na, Si) and their intimate mixing was crucial for optimal performance of the flame-
derived catalysts.
Under these conditions gradual phase transformation and crystallization of the amorphous
SiO2 to cristobalite occurred as evidenced by in situ high temperature XRD. This
transformation was accompanied by a dramatic loss of the specific surface area but had only a
marginal effect on catalyst performance. Catalysts containing Na were more prone to
sintering. Calcination of the flame-made catalyst negatively affected their performance.
Catalytic tests with different ratio of catalyst surface to void volume of the fixed bed
(S/V) revealed that the contribution of homogeneous reaction pathways was minor at the
conditions used.
Catalysts with similar composition prepared by wet-impregnation showed inferior

catalytic performance at optimal Na2WO4 content of 3wt%. In time-on-stream tests the flame-
made catalysts showed remarkably higher C2-selectivity and higher methane conversion
compared to a corresponding wet-impregnated reference catalyst. This behavior is traced back
to the more homogenous spatial distribution of the constituents in the flame-made catalysts,
which seems to be crucial for optimal catalytic performance.
2.5 Acknowledgements
We thank Dr. Frank Krumeich (ETH) for TEM/STEM/EDXS investigations and the Electron
Microscopy Center of ETH Zurich (EMEZ) for providing the necessary infrastructure. Thanks
are also due to Oliver Waser for helping with in situ high temperature XRD. Financial support
by ETH Research Grant ETH-39-12-2 and the European Research Council under the
European Unions Seventh Framework Program (FP7/2007-2013, ERC grant agreement no.
247283) is kindly acknowledged.
71
2.6 References
(2) Lunsford, J. H. Catal. Today 2000, 63, 165-174.
(3) Zavyalova, U.; Holena, M.; Schlgl, R.; Baerns, M. ChemCatChem 2011, 3, 1935-
1947.
(4) Keller, G. E.; Bhasin, M. M. J. Catal. 1982, 73, 9-19.
(5) Rane, V. H.; Chaudhari, S. T.; Choudhary, V. R. J. Nat. Gas Chem. 2008, 17, 313-
320.
(6) Kuo, Y.; Behrendt, F.; Lerch, M. Z. Phys. Chem. 2007, 221, 1017-1037.
(7) Takanabe, K.; Iglesia, E. J. Phys. Chem. C 2009, 113, 10131-10145.
(8) Pak, S.; Qiu, P.; Lunsford, J. H. J. Catal. 1998, 179, 222-230.
(9) Roos, J. A.; Korf, S. J.; Veehof, R. H. J.; van Ommen, J. G.; Ross, J. R. H. Appl.
Catal. 1989, 52, 147-156.
(10) Pak, S.; Lunsford, J. H. Appl. Catal. A-Gen. 1998, 168, 131-137.
(11) Dubois, J.-L.; Rebours, B.; Cameron, C. J. Appl. Catal. 1990, 67, 73-79.
(12) Fang, X.; Li, S.; Lin, J.; Gu, J.; Yang, D. J. Mol. Catal. (China) 1992, 6, 427.
(13) Wang, D. J.; Rosynek, M. P.; Lunsford, J. H. J. Catal. 1995, 155, 390-402.
(14) Jiang, Z. C.; Yu, C. J.; Fang, X. P.; Li, S. B.; Wang, H. L. J. Phys. Chem. 1993, 97,
12870-12875.
(15) Shahri, S. M. K.; Pour, A. N. J. Nat. Gas Chem. 2010, 19, 47-53.
(16) Mahmoodi, S.; Ehsani, M. R.; Ghoreishi, S. M. J. Ind. Eng. Chem. 2010, 16, 923-928.
(17) Wu, J.; Li, S. J. Phys. Chem. 1995, 99, 4566-4568.
(18) Ji, S.-F.; Xiao, T.-C.; Li, S.-B.; Xu, C.-Z.; Hou, R.-L.; Coleman, K. S.; Green, M. L.
H. Appl. Catal. A-Gen. 2002, 225, 271-284.
(19) Palermo, A.; Holgado Vazquez, J. P.; Lee, A. F.; Tikhov, M. S.; Lambert, R. M. J.
Catal. 1998, 177, 259-266.
369-389.
(21) Strobel, R.; Baiker, A.; Pratsinis, S. E. Adv. Powder Technol. 2006, 17, 457-480.
(22) Schimmoeller, B.; Schulz, H.; Ritter, A.; Reitzmann, A.; Kraushaar-Czarnetzki, B.;
Baiker, A.; Pratsinis, S. E. J. Catal. 2008, 256, 74-83.
(23) van Vegten, N.; Baidya, T.; Krumeich, F.; Kleist, W.; Baiker, A. Appl. Catal. B-
Environ. 2010, 97, 398-406.
72 Chapter 2
(24) Stark, W. J.; Wegner, K.; Pratsinis, S. E.; Baiker, A. J. Catal. 2001, 197, 182-191.
(25) Dedov, A. G.; Nipan, G. D.; Loktev, A. S.; Tyunyaev, A. A.; Ketsko, V. A.;
Parkhomenko, K. V.; Moiseev, I. I. Appl. Catal. A-Gen. 2011, 406, 1-12.
(26) Hung, C.-H.; Miquel, P. F.; Katz, J. L. J. Mater. Res. 1992, 7, 1870-1875.
(28) Stark, W. J.; Strobel, R.; Gunther, D.; Pratsinis, S. E.; Baiker, A. J. Mater. Chem.
2002, 12, 3620-3625.
(29) Simon, U.; Grke, O.; Berthold, A.; Arndt, S.; Schomcker, R.; Schubert, H. Chem.
Eng. J. 2011, 168, 1352-1359.
(30) Pistorius, C. W. F. T. J. Chem. Phys. 1966, 44, 4532-4537.
(31) Fiori, C. E.; Newbury, D. E. SEM Inc. AMF O'Hare, IL 1978, I, p. 401.
(32) Kracek, F. C. J. Phys. Chem. 1929, 34, 1583-1598.
(33) Barton, D. G.; Soled, S. L.; Meitzner, G. D.; Fuentes, G. A.; Iglesia, E. J. Catal. 1999,
181, 57-72.
(34) Vermaire, D. C.; van Berge, P. C. J. Catal. 1989, 116, 309-317.
(35) Malekzadeh, A.; Khodadadi, A.; Dalai, A. K.; Abedini, M. J. Nat. Gas Chem. 2007,
16, 121-129.
(36) Kapteijn, F.; Singoredjo, L.; Andreini, A.; Moulijn, J. A. Appl. Catal. B-Environ.
1994, 3, 173-189.
(37) Jao, S.; Sadjadi, S.; Godini, H. R.; Simon, U.; Arndt, S.; Grke, O.; Berthold, A.;
Arellano-Garcia, H.; Schubert, H.; Schomcker, R.; Wozny, G. J. Nat. Gas Chem.
2012, 21, 534-543.
(38) Ji, S.; Xiao, T.; Li, S.; Chou, L.; Zhang, B.; Xu, C.; Hou, R.; York, A. P. E.; Green, M.
L. H. J. Catal. 2003, 220, 47-56.
(39) Rodemerck, U.; Ignaszewski, P.; Lucas, M.; Claus, P. Chem. Eng. Technol. 2000, 23,
413-416.
(40) Malekzadeh, A.; Abedini, M.; Khodadadi, A. A.; Amini, M.; Mishra, H. K.; Dalai, A.
K. Catal. Lett. 2002, 84, 45-51.
(41) Wang, J.; Chou, L.; Zhang, B.; Song, H.; Zhao, J.; Yang, J.; Li, S. J. Mol. Catal. A:
Chem. 2006, 245, 272-277.
(42) Khan, A. Z.; Ruckenstein, E. Appl. Catal. A-Gen. 1992, 90, 199-207.
(43) Yu, Z. Q.; Yang, X. M.; Lunsford, J. H.; Rosynek, M. P. J. Catal. 1995, 154, 163-173.
(44) Arndt, S.; Otremba, T.; Simon, U.; Yildiz, M.; Schubert, H.; Schomcker, R. Appl.
Catal. A-Gen. 2012, 425426, 53-61.
73
(45) Korf, S. J.; Van Ommen, J. G.; Ross, J. R. H. Stud. Surf. Sci. Catal. 1991, 67, 117-
126.
(46) Blanksby, S. J.; Ellison, G. B. Acc. Chem. Res. 2003, 36, 255-263.
(47) Nikoo, M. K.; Amin, N. A. S. Fuel Process. Technol. 2011, 92, 678-691.
(48) Ghose, R.; Hwang, H. T.; Varma, A. Appl. Catal. A-Gen. 2013, 452, 147-154.
(49) Talebizadeh, A.; Mortazavi, Y.; Khodadadi, A. A. Fuel Process. Technol. 2009, 90,
1319-1325.
(50) Berger, R. J.; Prez-Ramr ez, J.; Kapteijn, F.; Moulijn, J. A. Chem. Eng. Sci. 2002, 57,
4921-4932.
(51) Taniewski, M.; Lachowicz, A.; Skutil, K.; Czechowicz, D. Chem. Eng. Sci. 1996, 51,
4271-4278.
(52) Hatano, M.; Hinson, P. G.; Shay Vines, K.; Lunsford, J. H. J. Catal. 1990, 124, 557-
561.
(53) Kalenik, Z.; Wolf, E. E. J. Catal. 1990, 124, 566-569.
(54) Lee, J. Y.; Jeon, W.; Choi, J.-W.; Suh, Y.-W.; Ha, J.-M.; Suh, D. J.; Park, Y.-K. Fuel
2013, 106, 851-857.
74 Chapter 2
75
Chapter 3
3. Oxidative Dehydrogenation of Ethane with CO2

over Flame-made Ga-loaded TiO2
Abstract
The influence of the Ga content (0-17 wt%) on the structural properties and the
catalytic behavior of flame-synthesized TiO2-supported gallium oxide in the oxidative
dehydrogenation of ethane (ODHE) has been investigated in a continuous tubular
microreactor using CO2 as a mild oxidant. The gallium oxide-titania powders consisted of
non-porous spherical particles of about 10 nm average diameter, as indicated by HRTEM and
X-ray line broadening, and had a specific surface area of about 120 m2/g. XRD showed no
reflections corresponding to Ga2O3 in the as-prepared samples indicating high dispersion of
the Ga constituent. At higher Ga-loading (> 14 wt%) stronger acidic sites became prominent
as indicated by NH3-TPD and DRIFTS. The ethene yield increased with Ga-loading up to
about 10 wt%. The molar CO2/C2H6 ratio in the feed, reaction temperature and space velocity
were decisive parameters for achieving maximum ethene yield. The ethene yield achieved
was 22% using a CO2/C2H6 ratio of 2.5, 700 C and a space velocity of 6,000 L/kgcat.h,
corresponding to 57% ethane conversion at 38% selectivity to ethene. After reaction (1 h on
stream) all catalysts showed significant coking and a loss of surface area depending on the
Ga-content. Catalysts with higher Ga-content were more resistant toward sintering but
showed more severe coking due to the presence of stronger acidic sites. Raman spectroscopy
revealed that all spent catalysts were covered with both D- and G-type carbon. Both carbon
deposition as well as reduction of catalyst surface area lead to a significant decrease in
activity with time-on-stream, while the selectivity to ethene increased up to above 70%. The
catalysts could efficiently be regenerated by exposing them for 10 min to air at the reaction
76 Chapter 3
temperature of 700 C, providing a base for the development of a dynamic process consisting
of alternative ODHE/regeneration cycles.
Part of it is published in ACS Catalysis, 5 (2015), 690-702.

77
3.1 Introduction
One of the most important feedstocks in chemical industry is ethene, a building block for
synthesis of commercially important products, such as polyethylene, ethylene oxide, styrene,
acetaldehyde, vinyl acetate, ethylene di(chloride/bromide), ethanol, ethylbenzene and many
more.1-3 The demand for these materials is growing due to the rising global market.3
Currently, ethene is produced mostly by steam cracking of naphtha and ethane. These raw
materials constitute approximately 70% of the production cost resulting in million tons of CO2
emission worldwide.4 The cracking process requires relatively high temperatures (>1000 C)
and severely suffers from coking that forces a complete shutdown of the operation for
periodic cleaning.4 As a consequence, there is a compelling need for the development of an
alternative, cost-effective process. Thermal dehydrogenation of ethane can also effectively
produce ethene at high temperatures (50% conversion, 720 C), however, it suffers from
uncontrollable side reactions as well as rapid coking.5
0
C2 H6 C2 H4 + H2 H298 = +136.5 kJ/mol (1)
Therefore, the use of catalysts to minimize the aforementioned problems while

achieving high conversion and selectivity is of great technological and economical interest.
As an alternative, oxidative dehydrogenation of ethane (ODHE) has been proposed over
thermal dehydrogenation. The current state of ODHE has been covered in a recent review,
where common principles and mechanistic aspects are discussed. 2
0
C2 H6 + 0.5O2 C2 H4 + H2 O H298 = 105.3 kJ/mol (2)
This reaction is highly favored even at lower temperatures (400-600 C). However, the
process is exothermic requiring efficient removal of the excess heat. Moreover, though
oxygen effectively oxidizes the deposited carbon from the catalyst surface, it also decreases
ethene selectivity by deep oxidation. Therefore, the use of a milder oxidant is preferred to
overcome the low selectivity. Recently, CO2 has been applied as a soft oxidant for oxidative
dehydrogenation of various alkanes (ethane, propane) and ethylbenzene and methane coupling
for ethene production.6 Such a strategy to utilize CO2 for commercial purposes is attractive
and promising, as it can be used for the synthesis of valuable products rather than releasing it
into the atmosphere.
Nakagawa et al.7 studied a series of metal oxides and found that Ga2O3 is an efficient
catalyst for ODHE with a yield (18.6%) doubled in the presence of CO2 than in its absence
(9%). Moreover, the ethene yield increased from about 3 to 25% with increasing Ga2O3
78 Chapter 3
surface area (from about 1 to 50 m2/g).8 The catalytic performance of Ga2O3 supported on
TiO2 was shown also to be superior to that on ZrO2, ZnO, Al2O3 and SiO2 supports for ODHE
with CO2. However, the ethene yield decreased remarkably when the Ga2O3/TiO2 catalyst was
subjected to time-on-stream test reaching a value of ~2% within 3 h due to carbon deposition.
Surface properties, such as acidity and basicity of Ga2O3 (amphoteric) are believed to play a
key role for the catalyst performance.8-10 And the slightly acidic CO2, upon coming in contact
with the Ga2O3 surface may alter its acidic/basic properties. To reduce coke deposition, tuning
acidity of Ga2O3 by addition of basic metal oxides (e.g. potassium oxide) decreased the alkane
conversion.11 Shen et al.12 reported that higher Si/Al ratio HZSM-5 supported Ga2O3 catalysts
are more resistant towards deactivation. They achieved an ethane conversion of 15% at 94%
ethene selectivity after 70 h reaction. The enhancement of the selectivity was attributed to a
reduction of the catalysts acidity suppressing undesired side reactions. The amount of
deposited coke on Ga2O3/HZSM-5(97) and Ga2O3 catalysts was virtually the same.
However, the latter catalyst which had a ten times lower surface area deactivated faster,
resulting in blocked active sites. Gallium oxide based catalysts have already shown potential
in various dehydrogenation reactions of alkanes, however, it is crippled by coking.13,14 Thus,
further research is essential to make such catalysts suitable for an efficient and stable alkane
dehydrogenation process.
Flame spray pyrolysis (FSP)15 has proven to be a versatile material synthesis technique
with great control over particle morphology16 and various other physical properties favorable
for catalytic applications17. Moreover, it is a single step process, with fast and high production
rates unlike most conventional catalyst preparation routes. Flame-made particles experience
rapid cooling due to the short flame reaction zone18, which prevents further sintering. In this
process novel composition of mixed oxide phases can form19 along with well-defined
structures20. The potential of this method for the synthesis of catalytic materials has been
proven in tailoring textural properties as well as dispersion and spatial distribution of active
components.21
With this in mind we synthesized a series of flame-made Ga2O3/TiO2 catalysts with
different Ga-loading and investigated their structural properties and catalytic potential for
ODHE using CO2 as mild oxidant. In the main focus were the influence of the Ga-loading and
reaction conditions (CO2/C2H6 feed ratio, temperature, space velocity) on the performance,
deactivation and regeneration behavior of these catalysts.
79
3.2 Experimental
Rapid and scalable single nozzle FSP15 was used for catalyst synthesis22. The primary flame
was created using premixed CH4 and O2 (>99.9%, Pangas) with flow rates of 1 and 2 L/min,
respectively. Gallium acetylacetonate (99.99%, Sigma Aldrich) and titanium isopropoxide
(TTIP) (97%, Sigma Aldrich) were dissolved in xylene to form precursor solutions of total
metal concentration of 0.45 M, which were injected at 5 mL/min by a syringe pump (Lambda,
VIT-FIT) through the FSP nozzle and dispersed into fine droplets by co-flowing 5 L/min of
O2 (>99.9%, Pangas) and maintaining a pressure drop of 2 bar at the nozzle tip. 15 Gas flow
rates were regulated by mass flow controllers (Bronkhorst). Resulting powders were collected
on water-cooled glass fiber filters (Whatman GF 6, 25.7 cm in diameter) 54 cm above the
burner with the aid of a vacuum pump (Busch, Seco SV 1040C). In addition to Ga/TiO2
catalysts, pure Ga2O3 and TiO2 were also synthesized as reference materials using FSP.
Commercial Ga2O3 (Alfa Aesar, 99.99% metal basis) was used as received as a reference.
3.2.2 Characterization
Nitrogen adsorption-desorption isotherms were measured using a Micromeritics Tristar 3000

instrument at 77 K. Samples were pretreated at 150 C for 1 h under continuous flow of
nitrogen to remove volatile impurities. The Brunauer-Emmett-Teller (BET) method was
utilized to determine the specific surface area (SSA) of the nanoparticles. Average particle
diameters were estimated from the SSAs by dp = 6/avg*SSA. The average density was
calculated using weighted densities of Ga2O3 (5.88 g/cm3), anatase (3.84 g/cm3) and rutile
TiO2 (4.26 g/cm3), respectively.
X-ray diffraction (XRD) was performed on a Bruker D8 Advance diffractometer
(40 kV, 40 mA, CuK) over the 2 range of 10-70. Average crystal sizes were estimated
using the TOPAS Rietveld method measuring the peak widths and the fraction of rutile in
TiO2 was calculated using the formula given by Spur and Myer23.
0.8IA 1
wt% rutile = (1 + ) (3.3)
IR
where: IA and IR are the intensities of anatase and rutile obtained from XRD measurements at
2 = 25 and 27, respectively.
80 Chapter 3
Acidic sites on the catalyst surfaces were probed by NH3-temperature programmed

desorption (TPD) using a Micromeritics Autochem II 2920 instrument. The samples (ca.
100 mg) were placed in a U-shaped glass tube, heated to 400 C (10 C/min) in a 5%O2/He
flow (20 mL/min) (Pangas) and held at this temperature for 30 min. Then the gas flow was
switched to He (Pangas, 99.999%) before cooling down to 50 C. After this pretreatment,
10 mol.% NH3/Ar (Pangas) was passed over the samples with a flow rate of 20 mL/min for 90
min, then switched back to He flow and ramped to 100 C (10 C/min) and held at this
temperature for 20 min. Finally, NH3-TPD was performed by heating the sample to 840 C
with a heating rate of 10 C/min and at a He flow rate of 20 mL/min. The evolved gases were
monitored by a mass spectrometer (Pfeiffer Vacuum, Thermostar, m/z = 15). CO2-TPD
measurements were performed following the aforementioned process, but using CO2 (m/z =
44) instead of NH3 to probe the basic sites of the samples.
Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) of all fresh
powders and commercial Ga2O3 was performed on a Vertex 70v spectrometer (Bruker
Optics). Spectra were obtained by averaging 100 scans at 4 cm-1 resolution. The powder
samples were placed in an in situ DRIFTS cell which was heated to 400 C in 5%O2/He
(Pangas) and held at this temperature for 30 min before cooling down to 40 C under He flow.
For NH3 adsorption the gas was switched to 10 mol.% NH3 in Ar, which was passed over the
sample for 15 min. Afterwards the sample was flushed with He for 15 min before measuring
the spectra.
Raman spectroscopy (Renishaw InVia Raman microscope) analysis of both the fresh
and spent powders were performed with a 514 nm laser at a range of 100-800 cm-1 and 100-
1800 cm-1, respectively. The laser power was set to 10% of 24 mW with exposure time of 10
s. Thermal analyses of spent catalysts were carried out using a thermogravimetric analyzer
(TGA/SDTA851e, LF/1100 C, Mettler Toledo AG) and a mixture of 40 mL/min O2 and 10
mL/min Ar as reactant gas, which was passed over the samples to oxidize the deposited
carbon. The evolving gases were monitored by an attached well-calibrated mass spectrometer
(Quadstar 422, Balzers) to identify the released volatiles from the sample. Same amounts of
sample (~ 28 mg) were taken for all the analyses.
High-resolution transmission electron microscopy (HRTEM) was performed on a

Tecnai F30 microscope (FEI; FEG cathode, at 300 kV, point resolution approx. 2 ).
Scanning transmission electron microscopy (STEM) and energy-dispersive X-ray
spectroscopy (EDXS) studies were done using an aberration-corrected, dedicated STEM
(Hitachi HD-2700CS) apparatus equipped with a probe corrector (CEOS). STEM imaging
81
was performed at an acceleration potential of 200 kV (electron gun: cold-field emitter)

detecting incoherently scattered electrons with a high-angle annular dark-field detector
(HAADF-STEM). An EDX spectrometer (Gemini system, EDAX) mounted in the electron
column above the sample facilitated recording spectra of selected spots and elemental
distribution maps (recording times ca. 30 min).
3.2.3 Catalytic studies
Reactions were carried out at an atmospheric pressure in a tubular quartz fixed-bed

microreactor of 4 mm inner diameter, which was contracted to 2 mm after the catalyst bed.
The catalyst bed consisted of a mixture of 150 mg of catalyst (80-140 m mesh) and 50 mg of
inert SiC (160-190 m mesh), which was fixed at the center of the reactor with quartz wool.
The catalyst bed length was ca. 2 cm. If not otherwise stated, the reactant gases, ethane
(99.5% PanGas) and carbon dioxide (99.95% PanGas), were passed at a total flow rate of 15
mL/min through the catalyst bed. Flow rates were regulated by calibrated mass flow
controllers (Brooks). If not otherwise stated, the molar feed gas ratio CO2/C2H6 was 2.5.
The reactor was placed in a temperature-controlled electrically-heated furnace and
heated to the reaction temperature (700 C at 10 C/min in standard experiments), which was
monitored with two thermocouples placed inside the reactor touching both inlet and outlet of
the catalyst bed. Generally, in catalytic runs the temperature difference between inlet and
outlet was < 4 oC. Reactor effluent gas lines were heated to 120 C with heating tapes to
prevent possible condensation. On-line gas chromatography (HP-PLOT Q column 30 m long,
0.32 mm in diameter and 0.2 m film thickness fitted with TCD and FID, Agilent
Technologies) was utilized to analyze gases in the product stream. A quadrupole mass
spectrometer (MS; Pfeiffer Vacuum, Thermostar GSD 300) and a non-dispersive infrared
analyzer for measuring CO and CO2 (ABB Uras 26) were also used to complement the GC
analysis. Cold water traps were used to condense water formed during reaction. All the
catalytic data are reported after 1 h reaction (time-on-stream) unless otherwise noted. C2H6
conversion, C2H4 selectivity and C2H4 yield were calculated as:
moles of C2 H6 consumed
C2 H6 conversion (%) = 100 (3.4)
moles of C2 H6 in the feed
moles of C2 H4 produced
C2 H4 selectivity (%) = 100 (3.5)
82 Chapter 3
C2 H6 conversion (%) C2 H4 selectivity(%)

C2 H4 yield (%) = (3.6)
100
H2 concentration was not measured during the reaction. Catalyst regenerability

experiments were performed using two different oxidants, CO2 and synthetic air (O2 20%
balance N2), keeping the aforementioned reaction conditions constant. Synthetic air was
utilized, after 90 min reaction for 10 min, to remove coke from the catalysts surface.
3.3.1 Textural and structural properties
Figure 3.1a shows the XRD patterns of as-prepared flame-made Ga2O3, TiO2 and various
gallium loaded TiO2 catalysts. Pure flame-made Ga2O3 showed low intensity reflections at
about 2 = 32, 35 and 46 in addition to some broader humps prominent at 2 = 64,
indicating that a significant fraction of this material was amorphous. TiO2 mainly consisting
of anatase (>80%). However, upon introducing gallium, the intensity of the anatase
reflections decreased while the rutile ones increased, indicating that Ga promoted rutile
formation. In similar catalysts prepared by laser-induced pyrolysis, the anatase to rutile
transformation was observed only upon calcination above 900 C.24 This promotion of the
anatase-rutile transformation has been attributed to the similar ionic radii of Ga3+ and Ti4+ i.e.
0.62 and 0.61 , respectively, favoring substitution25 of Ti4+ by Ga3+ in the TiO2 lattice
similar to Al3+ in titania26. Vemury et al.27 and Teoh et al.28 also had reported the replacement
of titanium with slightly larger metal ions i.e. Sn4+ = 0. 69 and Fe3+ = 0.64 , respectively,
during flame synthesis of doped TiO2. No reflections indicative of Ga2O3 or mixed oxides
were observed in any of the Ga-loaded TiO2 catalysts, indicating that the Ga component was
highly dispersed. However, when comparing the pattern of flame-made pure Ga2O3 to that of
the Ga/TiO2 powders, in the latter a small, broad hump was discernible at about 2 = 64
whose intensity increased with increasing gallium loadings. This could be attributed to
amorphous Ga2O3 at these loadings.
After ODHE, the rutile content of the TiO2 was >90 wt% (Figure 3.1b). Highest rutile
content was observed for pure TiO2 (~100%), while for the Ga-loaded samples it was in the
range of 90-98 wt%. The addition of Ga increased the catalysts resistance to sintering. This is
indicated by the decreasing rutile crystal size from 65 to 20 nm with increasing gallium
83
content from 0 to 17 wt%. The rutilization occurring during ODHE can be attributed to the
favorable reducing conditions at high temperature (700 C).29
A : Anatase
a R : Rutile
R
R Ga2O3
A R
A
17%Ga/TiO2
14%
intensity, a.u.
10%
7%
5%
3%
1%Ga/TiO2
TiO2
FSP-Ga2O3

10 20 30 40 50 60 70
A : Anatase
b R
R : Rutile
R
R R
A R
17%Ga/TiO2
intensity, a.u.
14%
10%
7%
5%
3%
1%Ga/TiO2
TiO2
10 20 30 40 50 60 70
2,degree
Figure 3.1: XRD patterns of flame-made (a) as-prepared pure Ga2O3, TiO2 and various
gallium-loaded TiO2, and (b) Spent gallium-loaded TiO2 after 1 h on stream at CO2/C2H6
ratio, 2.5; space velocity, 6,000 L/kgcat.h; temperature, 700 C.
84 Chapter 3
Complete rutilization was reported for wet-made powders containing gallium and titanium
after calcination at > 950 C.30 The absence of rutilization in as-prepared wet-made catalysts
is attributed to the different interaction of the gallium and titanium species in liquid and gas
phase and also to the exposure to different temperatures during their preparation.30,31
Figure 3.2a shows nitrogen adsorption-desorption isotherms of representative samples.
The isotherms of all powders showed the same characteristic shape, indicating that particles
were non-porous.22 The hysteresis at high P/Po is due to the interstitial void volumes of the
powders. TEM images of all samples indicated nearly spherical particles as exemplified by
the representative examples of pure TiO2 and 10 wt% Ga-loaded TiO2 (Figure 3.2b,c).
a b
adsorbed volume, a.u.
17%Ga/TiO2 c
C
10%Ga/TiO2
FSP-made TiO2
0.0 0.2 0.4 0.6 0.8 1.0

relative pressure, P/Po
Figure 3.2: Nitrogen adsorption-desorption isotherms (a) of FSP-made TiO2, 10%Ga/TiO2

and 17%Ga/TiO2 and TEM images of as-prepared (b) TiO2, and (c) 10%Ga/TiO2.
85
a b
c d
Figure 3.3: (a) HAADF-STEM image and combined map (b) of the 10%Ga/TiO2 sample. The
elemental maps of Ga (c) and Ti (d) reveal a homogeneous distribution of both elements. At
every measured spot, the EDX spectrum (not shown here) shows the peaks corresponding to
Ga and Ti simultaneously.
Figure 3.3 shows the HAADF-STEM image and the elemental maps of gallium and
titanium of as-prepared 10%Ga/TiO2 catalysts. The serially measured elemental maps reveal
the presence of well-distributed gallium (red) over titanium (green). Moreover, the EDX
spectrum (not shown) on every measured spot contained the peaks of gallium and titanium
simultaneously, providing another indication that probably part of Ga3+ ions substitute some
Ti4+ ions in the TiO2 lattice similar to Al3+.26
86 Chapter 3
a 100%Ga2O3 b 17%Ga/TiO2
MS signal intensity, a.u.
17%Ga
14%Ga
10%Ga
7%Ga

5%Ga
3%Ga
1%Ga
0%Ga
200 400 600 800 10%Ga/TiO2

temperature, C
2.5 2.5
CO2 desorbed (mol/m2)

NH3 desorbed (mol/m2)
c
2.0 2.0
1.5 1.5
1.0 1.0
0.5 0.5
0.0 0.0
0 20 40 60 80 100
200 400 600 800
Ga-loading of TiO2, wt% temperature, C
3150
3195
d
Kubelka-Munk, a.u.
3299
3033
3396
3673
1267
1426
1609
1496
1154
1675
1228
17%Ga/TiO2
10%Ga/TiO2
TiO2
4000 3500 3000 2500 2000 1500 1000

wavenumber, cm-1
Figure 3.4: Characterization of acidic and basic sites of Ga-loaded TiO2. (a) NH3-TPD
profiles, (b) NH3-TPD profiles of 10% and 17% Ga/TiO2 catalysts deconvoluted with
Gaussian function, (c) Desorbed amount of NH3 and CO2 during TPD analysis and (d)
DRIFT spectra of FSP-made pure TiO2, 10% and 17% Ga-loaded TiO2 after NH3 adsorption.
87
In order to gain information about the surface acidic and basic sites we applied NH3-
and CO2-TPD measurements. In their pure form both oxide constituents, TiO2 and Ga2O3 are
amphoteric, possessing acidic and basic sites.32 The NH3-TPD profiles (Figure 3.4a) show that
the pure TiO2 as well as the Ga-loaded TiO2 possess acidic sites of various strengths.
Strongest acidic sites (high temperature peak marked with dotted lines in the NH3-TPD) are
most prominent on the higher Ga-loaded catalysts (14%, 17% and the pure Ga2O3).
Deconvolution of the TPD profiles of 10% and 17% Ga-loaded TiO2 catalysts showed
multiple peaks with obvious increase of high temperature peaks, indicating promotion of
medium to strong acidic sites with increasing Ga-loading (Figure 3.4b). These strong acidic
sites are assumed to be mainly responsible for the enhanced coke deposition at higher gallium
loadings.33,34 The NH3 uptake of the pure flame-made TiO2 was 2.3 mol/m2, which is within
the range of previously reported values 3.235 and 1.536 mol/m2 and close to the one reported
by Gervasini and Auroux (2.5 mol/m2)37, on differently synthesized and pretreated TiO2.
Although at higher Ga-loadings new strong acid sites appeared, a decrease in the population
of acidic sites was observed upon addition of gallium in agreement with the observation of
Petre et al.10 in their pyridine adsorption experiments. This decrease is prominent for gallium
loadings up to about 8%; further increase of Ga-loading did not have a significant effect on
the total population of acidic sites, probably due to the appearance of strong acid sites as
evidence in the NH3-TPD profiles.
Figure 3.4c indicates that the population of basic sites changed only relatively little
with the gallium loading. The CO2 uptake by TiO2 (0.23 mol/m2) and Ga2O3 (0.28 mol/m2)
are close to those reported by Gervasini and Auroux.37 In Figure 3.4d, DRIFT spectra of
flame-made pure TiO2 and various Ga-loaded TiO2 catalysts are presented. Formation of new
bands and also band widening with increasing gallium content indicates alteration of the
distribution of the different acidic sites (Brnsted and Lewis sites) when TiO2 is loaded with
gallium. The multiple bands at 3396, 3299, 3195 and 3150 cm-1 can be associated to
stretching vibrations (both asymmetric and symmetric NH) of NH3 on the TiO2 surface.38
Moreover, the spectra around 367339 and 303340 cm-1 indicates the existence of residual OH
groups on the pre-treated Ga/TiO2 surface. The bands at 1496 and 1426 cm-1 are assigned to
Brnsted acid sites41, whereas bands at 1675, 1609, 1228 and 1154 cm-1 arise from
coordinatively adsorbed NH3 on Lewis acid sites.38 Upon Ga-loading of the TiO2, a new
distinct band appeared around 1267 cm-1, which is attributed to NH3 bonding on Lewis acid
sites of the Ga2O3 or mixed oxides (Ga/TiO2).42 This band became more prominent with
increasing Ga-content. From the DRIFT spectra and NH3-TPD measurements we infer that
88 Chapter 3
with increasing Ga-loading the concentration of coordinatively unsaturated Ga3+ on the

catalyst surface increases,43,44 as indicated by the formation of new bands and the creation of
strong acid sites.
120 a
as-prepared
BET surface area, m2/g
100
80
60
40
spent (700 C)
20
0
0 3 6 9 12 15 18
100
14
danatase b
drutile
average diameter, nm
dBET 80
12 Rutile wt.%
rutile wt., %
60
10
40
8
6 20
4 0
0 3 6 9 12 15 18
Ga content in TiO2, wt%
Figure 3.5: (a) Comparison of specific surface areas of as-prepared and spent (after 1 h on
stream at 700 oC) Ga/TiO2 catalysts, and (b) Average particle diameter and rutile content
(wt%) of TiO2 in various as-prepared Ga/TiO2 catalysts.
89
Figure 3.5a shows the SSA of all as-prepared and corresponding spent catalysts. All
as-prepared Ga-loaded TiO2 had an SSA around 120 m2/g. The rapid particle production in
the flame (low residence time, high temperature, fast quenching) results in higher surface area
because further particle growth is prevented.45 However, a dramatic loss of surface area was
observed for spent catalysts (after 1h on stream at 700 C). The presence of Ga, however,
mitigated that loss of SSA. As an example, the SSAs of spent TiO2 and 17%Ga/TiO2 were 1
and 40 m2/g, respectively.
In Figure 3.5b the average particle diameters estimated from BET and X-ray
diffraction of as-prepared catalysts are plotted as a function of the Ga content. The average
particle sizes estimated using these methods show similar trend: they decline with increasing
gallium content of the mixed oxide. The average crystal size of both anatase and rutile phases
continuously decreased with increasing Ga loading, however, they attained similar values
(~7 nm) at a loading of 17 wt%. The rutile content in as-prepared catalysts was about 20 wt%
which increased with Ga loading reaching, for example, about 76% at 17 wt% Ga.
Raman spectroscopy of as-prepared TiO2 showed typical bands belonging to anatase at
about 144, 400, 515 and 636 cm-1, whereas no bands due to rutile were identified (Figure
3.6a).46 However, with increasing Ga loading, some weak broad signals evolved at about 446
and 609 cm-1, indicating the presence of rutile. No bands of Ga2O3 were observed in the Ga-
loaded TiO2 powders, in agreement with XRD (Figure 3.1a). In Figure 3.6b Raman spectra of
all spent catalysts (after 1 h on stream) are shown. Both anatase and rutile bands were
observed along with bands belonging to diamond (D) and graphitic (G) carbon at 1356 and
1605 cm-1, respectively.47
Significant changes in the intensities of Raman bands due to TiO2 were observed in
spent catalysts (after 1 h on stream) depending on the gallium content (Figure 3.6b). Raman
band intensities of anatase at 144, 514 and 642 cm-1 decreased, while those of rutile increased
with increasing gallium content up to 5 wt%. At higher loadings, the intensity of the
corresponding bands decrease compared to carbon peaks but their ratio (rutile/anatase)
remains the same.
At 7 wt% Ga, bands at 514 and 642 cm-1 completely disappeared, while bands of rutile
at 445 and 615 cm-1 were still present though at low intensities. No Raman bands indicative of
Ga2O3 or mixed oxides were observed in spent catalysts, in agreement with XRD (Figure
3.1b), confirming the high dispersion of Ga even at high loadings. Figure 3.6b also depicts the
Raman bands of D- and G- carbon. While on pure TiO2 predominantly graphitic carbon (not
90 Chapter 3
shown) was deposited after 1 h on stream, on Ga-loaded TiO2 the distribution between D and
G carbon was more equalized with an estimated ratio around 1.6.
A: Anatase
A R: Rutile
A R A R A
17%Ga/TiO2
14%
intensity, a.u.
10%
7%
5%
3%
1%Ga/TiO2
TiO2
200 400 600 800
(a)
A: Anatase
R: Rutile D G
D: Diamond
G: Graphitic
A R R
intensity, a.u.
17%Ga/TiO2
14%
10%
7%
5% A A
3%
1%Ga/TiO2
200 400 600 800 1000 1200 1400 1600 1800

-1
Raman shift, cm
(b)
Figure 3.6: Raman spectra of (a) as-prepared and (b) spent (after 1 h on stream) Ga/TiO2
catalysts.
91
3.3.2 Catalytic performance
Figure 3.7a shows the catalytic performance of the various Ga/TiO2 catalysts with different
Ga-loadings after 1 h on stream. Pure flame-made TiO2 showed about the same ethane
conversion (24%) as observed with the empty reactor tube (23%), indicating that TiO2 was
virtually inactive after 1 h on stream. Ethane and CO2 conversions as well as ethene yield
showed a maximum at a gallium loading of ca. 10 wt%. The maximum ethene yield was 22%
corresponding to 57% ethane conversion at 38% selectivity to ethene. The only significant by-
product detected beside coke formation was methane (only traces of C3 were observed),
similar to the observation of Nakagawa et al..8 The increase in ethane conversion with
increasing Ga loading indicates that Ga2O3 is crucial for catalytic activity. However, the
undesired formation of methane and carbon deposits also indicates its potential role in alkane
cracking. Carbon deposition on the catalyst surface increased with increasing Ga-loading as
indicated by the thermoanalytical results shown in Figure 3.7b. At high Ga-loading the
catalyst surface was fully covered by carbon deposits and the catalyst activity and selectivity
became similar to that of pure TiO2.
Carbon dioxide conversion showed a behavior similar to that of ethane conversion
when the Ga-loading was increased. This indicates active participation of CO2 during the
reaction partly by removal of deposited carbon from the catalyst surface via reverse
Boudouard reaction.14 However, at higher Ga loadings (>10 wt%), the rate of carbon
deposition apparently surpasses its removal by oxidation with CO2 leading to rapid and
complete coverage and thus blocking of the active surface sites. These changes in CO2
conversion also lead to variation in the amount of CO in each experiment. A strong decrease
in ethane conversion from 57% to 24% was observed in the absence of CO2, that means when
CO2 was replace by He in the reactor feed over the 10%Ga/TiO2 catalysts, indicating the
beneficial role of CO2 in this reaction system.
Figure 3.7b shows how carbon deposition increased with the Ga-loading. Interestingly,
the catalyst with optimal Ga loading (10 wt%Ga/TiO2) showed lower rate of deactivation than
that observed by Nakagawa et al.9, maintaining 22% ethene yield compared to 18% after 1 h
time-on-stream. Nakagawa et al.8 showed that an increase in surface area of Ga2O3 increases
the ethene yield. Therefore, the observed high initial catalytic activity can be assigned to the
high initial surface area of the flame-made catalysts (~120 m2/g) and enhanced sintering
resistance in the presence of Ga.
92 Chapter 3
conversion, selectivity, yield, %

80
60 C2H4-selectivity
C2H6-conversion
40
20
C2H4-yield
CO2-conversion
0
0 3 6 9 12 15 18
5 b
total wt. loss
4
wt. loss, %
3
wt. loss due to carbon
100
2
99
TGA wt. loss, %
98
1 97
H2O
96 CO2
95
0 0 200 400 600 800
temperature, C
0 3 6 9 12 15 18
Ga loading of TiO2, wt%
Figure 3.7: (a) Effect of Ga-loadings on the oxidative dehydrogenation of ethane.

(Conditions: total gas flow, 15 mL/min; CO2/C2H6 ratio, 2.5; space velocity, 6,000 L/kgcat.h;
temperature, 700 C.), and (b) TGA analysis of mass loss due to water desorption and
burning of carbon deposit of spent Ga-loaded TiO2 catalysts as a function of Ga-loading,
inset: mass loss and MS signal of 10%Ga/TiO2.
93
TiO2
1%Ga/TiO2
80
3%Ga/TiO2
5%Ga/TiO2
7%Ga/TiO2
60 10%Ga/TiO2
selectivity, %
14%Ga/TiO2
17%Ga/TiO2
40
20
0
20 30 40 50 60 70 80 90
conversion, %
Figure 3.8: Selectivity vs conversion of various Ga-loaded TiO2 catalysts with progressing
reaction time. (Conditions: total gas flow, 15 mL/min; CO2/C2H6 ratio, 2.5; space velocity,
6,000 L/kgcat.h; temperature, 700 C; data taken, 20-60 min of the reaction, first point after
20 min of reaction and rest after 10 min from high to low conversion)
However, Ga2O3 seems to be very active for ethane ODH as well as ethene cracking.
Figure 3.8 shows the interdependence of selectivity and conversion with progressive time-on
stream. After 20 min of reaction, almost all Ga/TiO2 catalysts showed similar conversion
(83%) and selectivity (10%) values. However, conversion decreased due to active site
blocking by coke with concomitant increase in selectivity. These curves corroborate that the
10%Ga/TiO2 showed the best performance among the tested catalysts.
In Figure 3.9, both observed (solid lines) and expected (broken lines) reaction
pathways during ODHE are depicted. Reaction with only ethane mixed with helium showed
formation of ethene, CH4 and coke indicating the existence of pathways (a, c and d). The
weight loss obtained from TGA of the spent catalysts in the absence of CO2 was found to be
higher (3.3%) than in its presence (2.8%). This indicates that more coke was formed in the
absence of CO2 proving that CO2 mitigates coke deposition and likely oxidizes the deposited
coke during the ODHE reaction (pathway e). TGA measurement showed oxidation of coke by
94 Chapter 3
CO2 (total weight loss about 2.7%) was efficient in the temperature window of 800-1000 C
similar to the behavior reported by Nakagawa et al..9 The weight loss due to coke removal up
to 700 C was only 0.24%, while the rest (2.46%) was only removed at higher temperatures.
Large amount of carbon deposits (O2/Ar oxidation of coke) was also obtained from reaction
with mixture of ethene and CO2 (3.6 %), corroborating the existence of pathway (b). To
confirm the significance of the reverse water-gas shift (RWGS) reaction, ethane was replaced
by H2, keeping other reaction conditions similar. Under these conditions, CO2 conversion was
about 14% with CO as a major detectable product (pathway f) and the presence of traces of
CH4 indicated the significance of methanation. H2O produced from RWGS reaction is also
expected to follow steam reforming of ethane (g). The significance of these reactions
probably strongly depends on catalyst composition and reaction conditions. Especially
reactions (b-d) seem to be enhanced at higher Ga loadings (>10 wt%) leading to enhanced
catalyst deactivation.
Figure 3.9: Scheme of Observed (solid lines) and expected (broken lines) reaction pathways
during ODHE. (a) dehydrogenation of ethane; (b-d) coking reactions; (e) reverse Boudouard
reaction; (f) reverse water-gas shift reaction; (g) steam reforming of ethane. (CO2 (red) is
expected to consume H2 (green) from ethane dehydrogenation).
3.3.3 Influence of reaction conditions on performance of optimal catalyst
Figure 3.10 shows the effect of reaction temperature on the behavior of the best performing
10%Ga/TiO2 catalyst. Conversion increased almost linearly with increasing reaction
temperature, similar to Nakagawa et al.8, while CO2 conversion showed a maximum around
95
700 C. Ethene selectivity was higher at lower reaction temperature (600 C), reaching a flat
minimum between 650-700 C and then increased to 68% at 800 C. The increase in ethane
conversion with increasing temperature can be attributed to gradual evolution of thermal
dehydrogenation of ethane48, potentially enhanced by CO249. The lowering of CO2 conversion
at higher reaction temperatures (>700 C) could be attributed to the rapid deactivation of the
catalyst due to the inability of CO2 to oxidize and gasify the deposited carbon at sufficient
rate. High temperature accelerates formation and deposition of carbon on the catalyst
surface.11 The high ethane conversion at high temperatures (>700 C) is mainly attributed to
thermal dehydrogenation.
80 C2H6-conversion
60
C2H4-selectivity
40
C2H4-yield
20 CO2-conversion
0
600 650 700 750 800
temperature, C
Figure 3.10: Influence of temperature on the catalytic performance of 10%Ga/TiO2.
Conditions: total gas flow, 15 mL/min; CO2/C2H6 ratio, 2.5; space velocity, 6,000 L/kgcat.h.
High selectivity at high temperatures (800 C) arises from the unavailability of active
surface sites for further oxidation of ethene due to enhanced coking. Though a high reaction
temperature seems favorable due to high ethane conversion and ethene selectivity, it causes
96 Chapter 3
massive carbon deposition on the catalyst surface and also on the reactor wall requiring
extensive cleaning, which is expected to increase costs of production.4
Figure 3.11 shows the effect of the molar ratio of carbon dioxide to ethane in the feed
on the catalytic performance of the 10%Ga/TiO2 catalyst. The measurements were carried out
keeping the space velocity constant (6,000 L/kgcat.h) at 700 C. In absence of CO2, both the
yield and conversion were 12% and 5.5%, respectively, proving active participation of CO2 in
the reaction. Conversion increased with increasing CO2/C2H6 ratio, whereas the ethene
selectivity decreased strongly. This reduction in ethene selectivity is attributed to deep
oxidation of ethene and/or cracking.50 At the conditions given, a molar CO2/C2H6 ratio of 2.5
seemed to be optimal resulting in a maximal ethene yield of 22%.
80
C2H6-conversion
60
C2H4-selectivity
40
CO2-conversion
20
C2H4-yield
0
0 2 4 6 8
feed CO2/C2H6, -
Figure 3.11: Influence of CO2/C2H6 feed ratio on the catalytic performance of 10%Ga/TiO2.
Conditions: total gas flow, 15 mL/min; CO2/C2H6 ratio, 2.5; space velocity, 6,000 L/kgcat.h;
temperature 700 C.
Figure 3.12a shows the influence of the space velocity on the performance of the
10%Ga-loaded TiO2 catalyst. Ethane and CO2 conversions showed similar dependence on the
space velocity. The ethene yield slightly decreased with higher space velocity as a result of
97
the counteracting effects of decreasing ethane and CO2 conversion and increasing ethene
selectivity. Increasing the space velocity resulted in higher ethene selectivity reaching 53% at
a space velocity of 12,000 L/kgcat.h. The selectivity increase at higher space velocity is
ascribed to lower hydrocracking and deep oxidation of ethene. High CO2 conversion at lower
space velocities indicates an enhanced removal of coke, consequently, exposing the active
sites of the catalyst responsible for high C2H6 conversions. This role of CO2 as an oxidant in
the removal of coke as CO has been well documented.51,9 Figure 3.12b shows decreasing
selectivity with increasing conversion indicating existence of consecutive reaction of desired
product (ethene). The catalytic performance of these samples seemed to be similar indicating
differences in selectivity might be only due to change in conversion, due to coke deposition.
a b 1%Ga/TiO2
10%Ga/TiO2
80 80
17%Ga/TiO2
C2H6-conversion
selectivity, %
60 60
C2H4-selectivity
40 40
CO2-conversion
20 20
C2H4-yield
0 0
4000 8000 12000 15 30 45 60 75
space velocity, L/kgcat.h conversion, %
Figure 3.12: (a) Influence of space velocity on catalytic performance of 10%Ga/TiO2 and (b)
selectivity versus conversion of fresh Ga/TiO2 catalysts containing 1, 10 and 17% of Ga
(space velocity 2,000-12,000 L/kgcat.h) both at 700 C with CO2/C2H6 ratio of 2.5.
98 Chapter 3
a R
A : Anatase
R : Rutile
Ga2O3
intensity, a.u.
R R
900 C, 1 h A

800 C, 1 h
700 C, 1 h
Fresh
10 20 30 40 50 60 70
2,degree
100
b 10%Ga/TiO2
10%Ga/TiO2, 900 C
80
selectivity, %
60
40
20
0
10 20 30 40 50 60 70
conversion, %
Figure 3.13: (a) XRD patterns of 10%Ga/TiO2 calcined at various temperatures. (b)
Selectivity versus conversion of fresh (black squares) and calcined (red diamonds) (900 C
for 1 h) 10%Ga/TiO2 catalysts. Conditions: total gas flow, 15 mL/min; CO2/C2H6 ratio, 2.5;
space velocity, 2,000-12,000 L/kgcat.h; temperature 700 C.
99
In addition to the study of the influence of the reaction conditions, we also

investigated the effect of thermal pretreatment of the optimal 10%Ga/TiO2 catalyst on its
performance. For this purpose as-prepared catalysts were calcined at different temperatures
(700 C, 800 C and 900 C), with a ramp rate of 10 C/min and held at the target temperature
for 1 h. The intensity of the main rutile reflection (2 = 27) increased with increasing
calcination temperature while that of anatase (2 = 25) gradually disappeared (Figure 3.13a).
This indicates that Ga could not suppress rutile formation at higher calcination temperatures
(700-900 C), in contrast to the observation of Depero et al.24 and Benjaram et al.31 with
catalysts prepared by laser-induced pyrolysis and wet-chemistry, respectively, where >90%
anatase was present up to a calcination temperature of 800 C. Moreover, defects in crystal
structure of TiO2 during flame synthesis (Figure 3.1a), which were absent in powders made
by laser-induced pyrolysis and wet-chemistry, might have favored the anatase to rutile
transformation at lower temperatures. New reflections corresponding to Ga2O3 were observed
at 2 = 30, 32, 33 and 35 only after calcining the powder at 900 C implying that high
temperatures are required to transform amorphous Ga2O3 to a crystalline phase. Similar
observation was reported by Depero et al.24 where crystalline Ga2O3 was discernible only
after calcining Ga-loaded TiO2 at 1000 C.
Catalytic results indicated a gradual reduction in ethene yield with increasing
calcination temperature until 800 C (not shown). However, the sample calcined at 900 C
afforded an ethene yield (20%) similar to that of the as-prepared 10%Ga/TiO2 (22%) at a
space velocity of 6,000 L/kgcat.h. But, the ethane conversion decreased from 57 to 38%. The
selectivity versus conversion plot (Figure 3.13b) shows that at isoconversion there is no
significant difference in the performance of these catalysts. Therefore, the differences in
selectivity may be only due to different conversion. However, at the same space velocity (e.g.
6,000 L/kgcat.h), the low ethane conversion of the calcined samples can be attributed to this
thermal pretreatment which induced changes in the physicochemical properties of the catalyst
as its surface area decreased from 120 to 24 m2/g and crystalline Ga2O3 was formed (Figure
3.13a) after calcination at 900 C. These results indicate that high dispersion of Ga2O3 is
essential for achieving optimal catalyst activity. Such dispersion is facilitated during co-
synthesis of Ga/Ti oxides by flame aerosol technology.44,16
100 Chapter 3
3.3.4 Carbon deposition and catalyst regeneration
Severe coking on the catalyst surface is the major reason for its deactivation requiring
further investigation. Raman spectroscopy (Figure 3.6b) and thermal analyses (Figure 3.7b)
were applied to identify and quantify the deposited carbon on spent catalysts. Additionally,
insight on the time dependent carbon deposition was also gained. The intensity of the D-band
of carbon deposits on pure TiO2 (not shown) was smaller compared to that of the G-band,
however, with higher Ga loading the intensities of both bands became similar (Figure 3.6b).
The D-bands were broader than the G-bands in all spent catalysts and no major effect of Ga
content on the type of deposited carbon was observed. The amount of carbon belonging to D-
and G-bands was estimated from the ratio of the areas under the 1356 and 1605 cm-1 peaks,
respectively. The fraction of D-type carbon was always higher on Ga-loaded catalysts and
showed a ratio of D/G = ~1.66 that remained nearly constant up to 10 wt% Ga and slightly
decreased upon further Ga addition. Estimation of the particle size of the deposited carbon
using the intensities of D- and G-bands52,47 indicated particles in the range of 4.5-5.4 nm.
Figure 3.7b shows the amount of deposited carbon as a function of gallium content of
the Ga-loaded TiO2 derived from TGA analysis. Mass spectroscopy (MS) was used to analyze
the gaseous species evolving during the thermal analysis. The MS signal along with mass loss
as a function of temperature was plotted to identify the species responsible for the mass
change (inset, Figure 3.7b). All spent Ga-containing catalysts showed a two-step mass loss
(H2O loss between 80-200 C and carbon loss between 300-520 C). No noticeable weight
loss was observed with spent TiO2, indicating no significant amount of carbon (Figure 3.7b)
on its surface.
Carbon deposited primarily on the Ga2O3 surface or at the Ga2O3-TiO2 interface as
increasing weight loss was observed with increasing Ga content. The mass loss due to carbon
increased from 1.04 to 3.5% in spent catalysts containing 1 and 17 wt% of gallium,
respectively. A nearly linear relationship between deposited carbon and nominal gallium
content was observed (Figure 3.7b). Moreover, a small upshift in carbon oxidation
temperature (~50 C) with increasing Ga-loading was observed. The upshift of the onset of
the oxidation could be attributed to the higher carbon amount on the catalyst surface. 53 This
increasing amount of carbon on the catalyst surface with increasing Ga-loading is responsible
for the observed stronger loss of activity of higher Ga-loaded catalysts.
101
5 100
a
amount of carbon deposits, wt%
4 80
C2H6, conversion, %
3 60
2 40
1 20
0 0
20 40 60 80 100 120
time, min
b
G
D
Reaction time
intensity, a.u.
120 min
60 min
40 min
20 min
1200 1400 1600 1800

-1
Raman shift, cm
Figure 3.14: (a) Influence of reaction time (time-on-stream) on the amount of carbon
deposition on 10%Ga/TiO2, (b) Corresponding Raman spectra. Conditions: total gas flow, 15
mL/min; CO2/C2H6 ratio, 2.5; space velocity, 6,000 L/kgcat.h; temperature 700 C.
102 Chapter 3
Figure 3.14a shows the amount of carbon deposition and ethane conversion as a
function of time-on-stream for the 10%Ga/TiO2. Carbon deposition on the catalyst increased
almost linearly reaching a saturation level after 60 min on stream, while ethane conversion
decreased steadily till 120 min on stream. Note that the ethane conversion of 25% after 120
min is similar to the one obtained if the reaction is performed in the empty reactor tube
(23.1%) suggesting that the catalyst was virtually inactive after this time. Raman spectra of
the catalysts after different time-on-stream (Figure 3.14b) showed that at all times both types
of carbon, D and G, are present on the catalyst. The size of carbon particles ranged from 4.2
to 5.1 nm based on calculations utilizing the intensities of the specific D- and G-bands.52,47
However, as mentioned before, it appears that the change in catalyst activity was rather
dependent on the amount of deposited carbon than on its particle size. Overall, it is deduced
that carbon deposition is the major cause for the loss in catalyst performance.
The regeneration of the flame-made 10%Ga/TiO2 catalyst was accomplished by
oxidizing (gasifying) the carbon deposits using synthetic air after each cycle (90 min time-on-
stream). The 10%Ga/TiO2 catalyst exhibited good regenerability as it nearly regained its
initial conversion (83.7%) even at such short regeneration time (10 min) (Figure 3.15a). The
small loss in activity after the first regeneration might be due to some early loss in surface
area or loss of some temporary active sites.14 Gradual reduction in ethane conversion with
time-on-stream is associated with continuing deposition and accumulation of carbon on the
active sites of the catalyst. This increase in coking results from the insufficient ability of CO2
to remove coke from the catalyst surface.14 Similar behavior was observed with the
commercial Ga2O3, however, the deactivation rate was much faster exhibiting lower ethane
conversion compared to 10%Ga/TiO2. Commercial Ga2O3 catalysts showed an ethene
selectivity >90% whereas >60% was achieved with 10%Ga/TiO2 at the end of the first cycle.
After regeneration, the selectivity of the commercial Ga2O3 decreased (87%), whereas that of
10%Ga/TiO2 increased to 70%, which remained almost constant even after third regeneration.
The high activity of 10%Ga/TiO2 is attributed to the high dispersion of Ga on the TiO2. The
increase in selectivity with time indicates that the density of active sites decreases due to coke
deposition thereby slowing down the desired and particularly the undesired reactions. For
both catalysts the major cause for the continuous reduction in ethane conversion with time is
coke deposition on the catalyst surface.
103
100
80
C2H6 conversion, %
60
40
20
0
100
80
C2H4 selectivity, %
60
40
20
commer. Ga2O3
Oxidant CO2 10 wt%Ga/TiO2
Oxidant air
0
0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 400
time on stream, min time on stream, min
(a) (b)
Figure 3.15: Catalyst regeneration tests of 10%Ga/TiO2 and commercial Ga2O3 with air for
10 min after 90 min reaction with different oxidants in the feed (a) CO2/C2H6, (b) air/C2H6.
Conditions: total gas flow, 15 mL/min; (CO2 or Air)/C2H6 ratio, 2.5; space velocity, 6,000
L/kgcat.h; temperature 700 C (period between red dotted lines represents regeneration in
air).
Further investigations on the regenerability of the catalysts were performed by

replacing the oxidant CO2 in the feed with air, keeping other aforementioned reaction
parameters constant. High rate of decoking is expected due to the use of air, resulting in high
activity. Both commercial Ga2O3 and the 10%Ga/TiO2 catalysts showed very high initial
conversion (around 90%), however, the activity also declined with time-on-stream (Figure
3.15b), but less rapidly. Upon regeneration, both catalysts almost regained their initial
conversion values. Though both of them showed higher activity with air as oxidant, the
104 Chapter 3
selectivity to ethene was much lower (~40%), even after 90 min time-on-stream compared to
that of the reaction performed using CO2 as oxidant (around 80%). In addition to ethene,
methane, CO2 and CO were also detected in the product stream. The high ethane conversion
with low ethene selectivity can be attributed to their total oxidation by air. Similar to previous
regenerability test with CO2, the gradual decline in catalysts activity is attributed to coking. In
both cases, ethene selectivity improved with time-on-stream (increasing coking), indicating a
decrease of active sites responsible for both the desired and undesired reactions. Raman
spectra of these spent catalysts were comparable to those recorded for experiments with CO2
as oxidant, showing similar ratio of D- and G-bands. Overall, though coking reduced the
performance of both the commercial Ga2O3 and 10%Ga/TiO2, the use of the latter as catalyst
seems beneficial due to its higher performance even with relatively low loading of gallium.
3.4 Conclusions
Flame synthesis of Ga-loaded (0-17 wt%) TiO2 results in spherical virtually non-porous
particles of about 10 nm diameter in which the Ga constituent is well dispersed. Ga-loadings
up to about 8 wt% lead to a strong decrease of the population of acidic sites while further
increase of the Ga-content has little effect on the acidic site population but alters the
distribution of their strength. Strongest acidic sites are most prominent on the highly Ga-
loaded catalysts. The population of basic sites changes relatively little with Ga-loading. These
materials show remarkable catalytic activity in the oxidative dehydrogenation of ethane
(OHDE) with CO2 as oxidant. While flame-made pure titania is nearly inactive, the catalytic
performance of Ga-loaded TiO2 increases strongly with increasing Ga-content up to about 10
wt% affording an ethene yield of 22%. Higher Ga-loading proved to be unfavorable due to
enhanced cracking and carbon deposition. Besides Ga-loading, the feed ratio CO2/C2H6,
temperature and space velocity in the tubular fixed-bed microreactor were found crucial for
achieving optimal ethene yield. Ethane conversion and ethene selectivity showed a clear
interdependence, higher ethane conversion resulting in lower ethene selectivity. Raman
spectra revealed the presence of both D- and G- type carbon in all the spent catalysts. This
carbon deposition leads to enhanced ethene selectivity due to decreased ethane conversion.
The deactivated catalysts could be easily regenerated by air in a short time (10 min) at the
reaction temperature of 700 C. If air was used instead of CO2 as oxidant in the reaction,
carbon deposition was lower resulting in less prominent catalyst deactivation. However, with
air as oxidant, the selectivity to ethene was much lower due to enhanced further oxidation of
ethene affording significantly lower ethene yield compared to the reaction with CO2. The ease
105
and efficiency of the catalyst regeneration may provide a base for developing a dynamic
process consisting of alternative reaction/regeneration cycles.
This work is financially supported by ETH Research Grant ETH-39-12-2 and the European
Research Council under the European Unions Seventh Framework Program (FP7/2007-2013,
ERC grant agreement no. 247283) is kindly acknowledged.
3.6 References

(5) Kung, H. H. Adv. Catal. 1994, 40, 1-38.
(6) Ansari, M. B.; Park, S.-E. Energ. Environ. Sci. 2012, 5, 9419-9437.
1998, 1025-1026.
(9) Nakagawa, K.; Kajita, C.; Okumura, K.; Ikenaga, N.; Nishitani-Gamo, M.; Ando, T.;
Kobayashi, T.; Suzuki, T. J. Catal. 2001, 203, 87-93.
(10) Petre, A. L.; Auroux, A.; Glin, P.; Caldararu, M.; Ionescu, N. I. Thermochim. Acta
2001, 379, 177-185.
(11) Michorczyk, P.; Ogonowski, J. Appl. Catal. A: Gen. 2003, 251, 425-433.
(12) Shen, Z.; Liu, J.; Xu, H.; Yue, Y.; Hua, W.; Shen, W. Appl. Catal. A: Gen. 2009, 356,
148-153.
(13) Xu, B.; Zheng, B.; Hua, W.; Yue, Y.; Gao, Z. J. Catal. 2006, 239, 470-477.
(14) Zheng, B.; Hua, W.; Yue, Y.; Gao, Z. J. Catal. 2005, 232, 143-151.
106 Chapter 3
369-389.
(17) Strobel, R.; Baiker, A.; Pratsinis, S. E. Adv. Powder Technol. 2006, 17, 457-480.
(18) Rudin, T.; Pratsinis, S. E. Ind. Eng. Chem. Res. 2012, 51, 7891-7900.
(19) Ernst, F. O.; Kammler, H. K.; Roessler, A.; Pratsinis, S. E.; Stark, W. J.; Ufheil, J.;
Novk, P. Mater. Chem. Phys. 2007, 101, 372-378.
(20) van Vegten, N.; Baidya, T.; Krumeich, F.; Kleist, W.; Baiker, A. Appl. Catal. B:
Environ. 2010, 97, 398-406.
(22) Strobel, R.; Stark, W. J.; Mdler, L.; Pratsinis, S. E.; Baiker, A. J. Catal. 2003, 213,
296-304.
(23) Spurr, R. A.; Myers, H. Anal. Chem. 1957, 29, 760-762.
(24) Depero, L. E.; Marino, A.; Allieri, B.; Bontempi, E.; Sangaletti, L.; Casale, C.; Notaro,
M. J. Mater. Res. 2000, 15, 2080-2086.
(25) Shannon, R. Acta Crystallogr. A 1976, 32, 751-767.
(26) Akhtar, K. M.; Pratsinis, S. E.; Mastrangelo, S. V. R. J. Mater. Res. 1994, 9, 1241-
1249.
(27) Vemury, S.; Pratsinis, S. E. J. Am. Ceram. Soc. 1995, 78, 2984-2992.
(28) Teoh, W. Y.; Amal, R.; Mdler, L.; Pratsinis, S. E. Catal. Today 2007, 120, 203-213.
(29) Wegner, K.; Pratsinis, S. E. AlChE J. 2003, 49, 1667-1675.
(30) Mohammadi, M. R.; Fray, D. J. Acta Mater. 2007, 55, 4455-4466.
(31) Reddy, B. M.; Ganesh, I.; Reddy, E. P.; Fernndez, A.; Smirniotis, P. G. J. Phys.
Chem. B 2001, 105, 6227-6235.
(32) Auroux, A.; Gervasini, A. J. Phys. Chem. 1990, 94, 6371-6379.
(33) Bartholomew, C. H. Appl. Catal. A: Gen. 2001, 212, 17-60.
(34) Campelo, J. M.; Lafont, F.; Marinas, J. M.; Ojeda, M. Appl. Catal. A: Gen. 2000, 192,
85-96.
(35) Liu, Z. F.; Tabora, J.; Davis, R. J. J. Catal. 1994, 149, 117-126.
(36) Shibata, K.; Kiyoura, T.; Kitagawa, J.; Sumiyoshi, T.; Tanabe, K. Bull. Chem. Soc.
Jpn. 1973, 46, 2985-2988.
(37) Gervasini, A.; Auroux, A. J. Therm. Anal. 1991, 37, 1737-1744.
(38) Tsyganenko, A. A.; Pozdnyakov, D. V.; Filimonov, V. N. J. Mol. Struct. 1975, 29,
299-318.
107
(39) Primet, M.; Pichat, P.; Mathieu, M. V. The Journal of Physical Chemistry 1971, 75,
1216-1220.
(40) Risti, M.; Popovi, S.; Musi, S. Mater. Lett. 2005, 59, 1227-1233.
(41) Yamazoe, S.; Okumura, T.; Hitomi, Y.; Shishido, T.; Tanaka, T. J. Phys. Chem. C
2007, 111, 11077-11085.
(42) Busca, G. Phys. Chem. Chem. Phys. 1999, 1, 723-736.
(43) Chen, M.; Xu, J.; Su, F. Z.; Liu, Y. M.; Cao, Y.; He, H. Y.; Fan, K. N. J. Catal. 2008,
256, 293-300.
(44) Pushkar, Y. N.; Sinitsky, A.; Parenago, O. O.; Kharlanov, A. N.; Lunina, E. V. Appl.
Surf. Sci. 2000, 167, 69-78.
(45) Mdler, L.; Stark, W. J.; Pratsinis, S. E. J. Mater. Res. 2002, 17, 1356-1362.
(46) Porto, S. P. S.; Fleury, P. A.; Damen, T. C. Phys. Rev. 1967, 154, 522-526.
(47) Knight, D. S.; White, W. B. J. Mater. Res. 1989, 4, 385-393.
(48) Burch, R.; Crabb, E. M. Appl. Catal. A: Gen. 1993, 97, 49-65.
(49) Choudhary, V. R.; Mondal, K. C.; Mulla, S. A. R. J. Chem. Sci. 2006, 118, 261-267.
(50) Baidya, T.; Vegten, N.; Baiker, A. Top. Catal. 2011, 54, 881-887.
(51) Krylov, O. V.; Mamedov, A. K.; Mirzabekova, S. R. Ind. Eng. Chem. Res. 1995, 34,
474-482.
(52) Jawhari, T.; Roid, A.; Casado, J. Carbon 1995, 33, 1561-1565.
(53) Monti, D. A. M.; Baiker, A. J. Catal. 1983, 83, 323-335.
108 Chapter 3
109
Chapter 4
4. Influence of Support in CO2-assisted Cobalt-

Catalyzed Oxidative Dehydrogenation of Ethane
Abstract
Catalysts containing 4.5 wt% cobalt supported on single oxides of SiO2, Al2O3, TiO2 and
ZrO2 or on mixed ones of SiO2-Al2O3, SiO2-TiO2, SiO2-ZrO2 and TiO2-ZrO2 were produced
in a single step by flame spray pyrolysis. These nonporous catalysts were characterized by
nitrogen adsorption, XRD, HR-TEM, EDXS, NH3-TPD, H2-TPR, TGA, DRIFTS, XPS and
Raman and UV-Vis spectroscopy and tested for the oxidative dehydrogenation of ethane
(ODHE) in a continuous fixed-bed microreactor using CO2 as oxidant. Depending on the
supporting oxides the reducibility of the cobalt species varied in a broad range, indicating
vastly different support interactions. For catalysts supported on Al2O3 and SiO2-Al2O3 no
significant H2 consumption was observed up to 840 C, while ZrO2-supported CoOx was
already reduced at T< 500 C. The silica-supported catalysts showed a broad reduction peak
at ~780 C attributed to well dispersed CoOx embedded in the SiO2 matrix forming cobalt
silicate, consistent with HR-TEM, XPS and UV-Vis analyses. Among all catalysts, silica-
supported cobalt showed the highest ethene yield of 34% under standard reaction conditions
(700 C with CO2/C2H6 = 2.5). Admixing Al2O3, TiO2 and ZrO2 to SiO2 did not result in
improved catalytic performance. Reducibility of the CoOx species as well as the strength and
surface density of acidic sites were identified as crucial catalyst properties for optimal ethene
yield. Control of the surface acidity seems particularly important to avoid undesired side
reactions (cracking, coking) impairing activity and selectivity. Parametric sensitivity studies,
including reaction temperature, CO2/C2H6 feed ratio and space velocity, revealed the
important role of CO2. In the absence of CO2, only low C2H6 conversion was observed and
change of the oxidant to O2 resulted in poorer catalytic performance.
Part of it is published in Applied Catalysis A:General, 527 (2016), 96-108.

110 Chapter 4
4.1 Introduction
Increasing demand for ethene (C2H4) derivatives such as polyethene, ethene oxide,
ethylbenzene and many more has made ethene one of the main chemical feedstocks in the
chemical industry.1,2 Conventionally, ethene is made by naphtha and ethane (C2H6) cracking,
both energy intensive processes that promote high rate of coke deposition, a continuous
operation hurdle, apart from high CO2 emission.3,4 Recently, ethane cracking has significantly
picked up the pace, due to current growth in discoveries of natural gas, especially shale gas in
the US, making it a cheap feed for ethene production.5 The aforementioned inherent problems
of these processes can be tackled by developing a relatively low temperature catalytic process.
Oxidative dehydrogenation of ethane (ODHE) takes place at lower temperatures where
oxidants such as oxygen6, carbon dioxide7 and nitrous oxide8 are utilized. However, O2 results
in an exothermic reaction requiring efficient heat removal and it also easily oxidizes ethane,
whereas utilizing N2O in large scale, is rather challenging due to its limited availability.
Therefore, CO2, a greenhouse gas, comes as an obvious choice due to its high availability and
relatively low cost.9
Studies on ODHE employing CO2 have been reported over Cr2O3-10,11 and Ga2O3-
based12-14 catalysts, where the former showed superior activity. However, the use of Cr2O3 is
problematic due to toxicity issues while Ga2O3 is relatively expensive and shows severe
sintering and coking. The latter metal oxide has been reported to exhibit higher activity
compared to that of the former.15 But when supported on generic supports (e.g. Al2O3, SiO2
etc),7,16 the observed activity was reversed, i.e. the Cr-based catalyst was superior. One of the
issues that crippled the performance of Ga3O3-based catalyst is high rate of coke deposition
resulting in gradual and early deactivation16, requiring frequent regeneration.12 By tuning the
acidity and basicity of the catalyst, to some extent, the coke deposition has been minimized14,
however, the necessity of catalyst regeneration seemed to be inevitable. The reported initial
ethane conversion and ethene selectivity over Ga2O3-based catalysts are in the range of 20-
30% and 80-90%, respectively, which gradually decrease, primarily ethane conversion, with
time.16,17 In contrast, relatively high and stable ethane conversion of around 50% and ethene
selectivity of > 90% were obtained over SiO2-supported Cr2O3 compared to that of other
supported catalysts (e.g. Al2O3, TiO2, and ZrO2).7 The observed high activity was attributed to
the redox property and acidity/basicity of the Cr2O3/SiO2. Some improvement on the
Cr2O3/SiO2 activity has been achieved by addition of acidic and basic oxide promoters.18
111
Cobalt oxide based catalysts are attractive for oxidative dehydrogenation of ethane19-
21
, propane22,23 and isobutane24 using O2 and also for dry reforming of methane25,26. Li et al.27
used 3%Co-MCM-41 with CO2 as oxidant and reported ethane conversion of about 40% with
a selectivity of 99%. Zhang et al.28 found 7% Co loading on BaCO3 to be optimal with a very
stable (> 5 h) ethane conversion of 48% at a C2H4 selectivity of 92%. The high activity and
the stable performance was attributed to the redox property of CoOx as well as synergism
between BaCO3 and BaCoO3.
The support characteristics have major influence on the performance in ODHE, as has
been shown for Cr2O3-7 and Ga2O3-based16 catalysts. To our knowledge, there exists no
systematic study of the catalytic performance of single and mixed oxide supported cobalt
catalysts on CO2-assisted ODHE. With this in mind, we investigated the influence of single
and mixed-oxide supports on the performance of Co-based ODHE catalysts. A scalable
single-step flame spray pyrolysis (FSP) method was used for their preparation.29 This
synthesis technique is flexible, easily operable and offers high possibility of metastable phase
formation due to which rapid growth in its utilization for catalyst synthesis is observed in
recent years.29,30
In this comparative study we report the physicochemical properties of flame-made
single and mixed-oxide supported Co catalysts and their performance in ODHE using CO2 as
a mild oxidant. The physicochemical properties of the catalysts were analyzed by nitrogen
adsorption, XRD, HR-TEM, EDX, NH3-TPD, TPR, DRIFTS, XPS and Raman- and UV-Vis
spectroscopy and crucial properties for optimal catalytic performance are identified.
4.2 Experimental

All catalysts were made by FSP.12 Briefly, precursor solutions were prepared by
mixing appropriate amounts of metal-precursors in toluene (Sigma Aldrich), except for
catalysts containing (a) Al2O3, where diethylene glycol monobutyl ether (Sigma-Aldrich) and
acetic anhydride (Sigma-Aldrich) (vol. ratio of 2:1) and (b) ZrO2, where a mixture of toluene,
2-ethylhexanoate (Sigma-Aldrich, purity > 99%) and acetonitrile (Sigma-Aldrich, purity >
99.5%) at vol. ratio of 3:1:1 were used. All the metal precursors, cobalt (III) acetylacetonate
(98%), hexamethyldisiloxane ( 98%), titanium iso-propoxide (> 97% purity) and aluminum
tri-sec-butoxide (97% purity) from Sigma-Aldrich and zirconium 2-ethylhexanoate (~6% Zr)
from Strem chemicals, were used as received. The total molar concentration of metal (i.e. Co
112 Chapter 4
+ metal in support(s)) was set to 0.45 M except for Zr-containing catalysts where the molarity
was set to 0.31 M due to solubility limitation. When preparing mixed-oxide supports, both
oxides were of equivalent weight (i.e. 1:1 ratio of 95.5 wt%). The cobalt content was always
4.5 wt%. During FSP, the precursor solution was pumped at 5 mL/min into the flame reactor
nozzle using a syringe pump (Lambda, VIT-FIT). The feed was dispersed into fine droplets
by co-flowed O2 of 5 L/min (> 99.9% PanGas) and immediately ignited by pilot flame fed by
CH4/O2 of 1/2 L/min (> 99.9% PanGas). The pressure drop at the nozzle-tip was maintained
at 2 bar. Product powders were collected on glass fiber filter (ALBET GF 6, 25.7 cm in
diameter) with the aid of a vacuum pump (Busch, Seco SV 1040C). All gas flow rates were
regulated by mass flow controllers (Bronkhorst).
X-ray diffractograms were recorded on a Bruker D8 Advance diffractometer (40 kV,

20 mA, Cu K) over the 2 range of 10-70 with a step size of 0.019 and scan speed of 0.1
sec/step. Nitrogen adsorption-desorption isotherms were measured on a Micromeritics Tristar
instrument and the specific surface area (SSA) of both fresh and spent catalyst powders was
determined by applying the Brunauer-Emmett-Teller (BET) method at 77 K. All samples
were kept at 150 C in N2-atmosphere for 1 h prior to measurements.
Micromeritics Autochem 2920 was used for H2-temperature programmed reduction
(H2-TPR) and NH3-temperature programmed desorption (NH3-TPD). For all measurements, a
sample mass of ca. 100 mg, a temperature ramp rate of 10 C/min and total gas flow rate of
20 mL/min was applied. The NH3-TPD was performed by heating the sample from room
temperature to 400 C in 5% O2 in He and maintaining it at this temperature for 2 h before
cooling down to 50 C in helium. Then, the gas flow was switched to 10% NH3 in Ar, which
was passed over the sample for 90 min. Temperature was increased to 100 C and the gas
flow switched back to helium to remove physisorbed NH3 from the surface of the sample.
Finally, the temperature was ramped to 840 C and the desorbed NH3 was monitored by a
mass spectrometer (m/z = 15). For H2-TPR measurements, the samples were heated from
room temperature to 200 C in helium and held at this temperature for 2 h before cooling
down to 50 C. Then the gas flow was switched to 5%H2/Ar before ramping to 840 C. The
H2-consumption was monitored by a mass spectrometer (m/z = 2).
Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) of all fresh
catalysts was performed on a Vertex 70v spectrometer (Bruker Optics). The powder samples
were placed in an in-situ DRIFTS cell that was heated to 400 C in 5%O2 in He at 20 mL/min
113
and held at this temperature for 30 min. After cooling down to 40 C in helium at 20 mL/min,
background (sample + KBr) was taken and then the gas was switched to 10% NH3/Ar, which
was passed over the sample for 15 min. Afterward the weakly adsorbed NH3 was flushed by
helium for 15 min before measuring the spectra.
UV-Vis diffuse reflectance spectra (Cary 500 UV-Vis-NIR Spectrophotometer) were
recorded from 300 to 800 nm for the fresh supported Co-catalysts and bare single metal
oxides. The samples were diluted with BaSO4 keeping constant the mass ratio of the sample
and BaSO4 and spectra were measured against a BaSO4 background. Raman (Renishaw InVia
Raman microscope) spectra of both fresh and spent catalysts were obtained using a 785 nm
laser with 5% of the laser power (500 mW) and an exposure time of 10 sec. Accumulation of
spectra was set to 20 for better resolution.
XPS measurements were performed using a VG ESCALAB 220iXL spectrometer
(Thermo Fisher Scientific) with monochromatic Al K source (spot size: 500 m, power: 150
W). The powder samples were placed on an adhesive copper tape which was mounted on a
stainless steel holder that was later transferred to the instrument chamber and kept under
vacuum. In built electron and ion neutralizers were used to compensate the charge generated
from the samples. The spectra were recorded in constant analyzer energy of 30 eV. The XPS
system was calibrated using the Ag 3d5/2, the Cu 2p3/2 and the Au 4f7/2 peaks.
High resolution transmission electron microscopy (HR-TEM) was performed by a
Tecnai F30 ST (FEI, FEG, operated at 300 kV) microscope. Electron-dispersive X-ray
spectroscopy (EDXS) measurements were taken by an EDX spectrometer (Gemini system,
EDAX), mounted on the electron column right above the sample that allowed recording the
spectra of selected spots. Amount of coke in the spent catalyst was quantified using a
thermogravimetric analyzer (TGA/SDTA851e, LF/1100 C, Mettler Toledo AG). A mixture
of 40 mL/min CO2 and 10 mL/min Ar as reactant gas was passed over the samples to oxidize
the deposited carbon. Approximately 20 mg of each sample was taken for the analysis.
4.2.3 Catalyst testing
The catalysts were tested at atmospheric pressure in a tubular continuous flow reactor
with 4 mm inner diameter that was contracted to 2 mm after the catalyst bed. Prior to use, the
catalyst powders were pelletized, crushed and sieved, affording a 80-140 m size fraction.
Mixtures of 150 mg of catalyst (80-140 m) and 50 mg SiC (160-190 m) were used to
establish a catalyst bed, which was fixed at the center using quartz wool.31 The tubular reactor
114 Chapter 4
was placed in a temperature-controlled electrically heated furnace which was ramped

(10 C/min) to the desired temperature (typically 700 C).
If not otherwise stated, the standard reaction conditions were: 700 C, reactant gas
ratio CO2/C2H6 = 2.5 and a total flow rate of reactant gases (CO2, 99.95 % PanGas and C2H6,
99.5 % PanGas) of 15 mL/min. During reaction, the temperature of the catalyst bed was
monitored using two thermocouples at its inlet and outlet. Generally, their temperature
difference was < 4 C during all testing. The effluent gases were analyzed, after passing
through cold traps (to condense H2O), using online gas chromatography (HP-PLOT Q column
30 m long, 0.32 mm in diameter and 0.2 m film thickness fitted with TCD and FID, Agilent
Technologies). All the catalytic test results are presented after 2 h reaction, if not otherwise
noted. The C2H6 conversion, C2H4 selectivity, C2H4 yield and CO2 conversion are calculated
as follows:
C2 H6 conversion (%) = 100 (4.1)
C2 H4 selectivity (%) = 100 (4.2)

C2 H4 yield (%) = (4.3)
100
moles of CO2 consumed

CO2 conversion (%) = 100 (4.4)
moles of CO2 in the feed
4.3.1 Structural and chemical properties of catalysts
4.3.1.1 Crystallinity, surface area and morphology
The flame-made catalytic powders were nonporous as evidenced by their nitrogen adsorption-
desorption isotherms (Fig. A.1, Appendix A).12 Figure 4.1a shows the XRD patterns of fresh
powders containing 4.5 wt% Co supported on SiO2, Al2O3, TiO2, ZrO2 and corresponding
mixed oxide supports.
115
:srilankite :srilankite
a b :tetragonal ZrO2
:tetragonal ZrO2
:anatase TiO2 :monoclinic ZrO2
:rutile TiO :anatase TiO2
2
:-Al2O3 :rutile TiO2
:-Al2O3

TiO2-ZrO2 TiO2-ZrO2
Intensity, a.u.

SiO2-ZrO2 SiO2-ZrO2
SiO2-TiO2
SiO2-TiO2

SiO2-Al2O3
SiO2-Al2O3

ZrO2

ZrO2

TiO2
TiO2

Al2O3
Al2O3
SiO2
SiO2
10 20 30 40 50 60 70 10 20 30 40 50 60 70
2, degree 2, degree
Figure 4.1: XRD patterns of supported cobalt catalysts (a) fresh and (b) spent (4.5 wt% Co
loadings in all catalysts. Mixed oxide supports contain an equal mass ratio of both oxides.)
The SiO2- and SiO2-Al2O3-supported catalysts were XRD-amorphous while -Al2O3

patterns were observed with pure Al2O3 support. As XRD reflections of -Al2O3 and CoAl2O4
are similar, small amounts of CoAl2O4 could be present.32 In TiO2-supported catalyst, both
anatase and rutile TiO2 reflections were observed with the latter being dominant, indicating
that cobalt promotes rutilization. This can be attributed to the substitution of Co3+ (0.55 )
into the Ti4+ (0.61 ) lattice, due to the difference in their ionic radii 33 similar34 to Al3+
substitution of Ti4+. The intensity of anatase and rutile reflections is lower in the SiO2-TiO2-
supported catalyst than for the TiO2-supported catalyst, because SiO2 suppresses TiO2
crystallization and/or there is a significant amorphous fraction. In the ZrO2-supported catalyst
essentially all reflections can be assigned to tetragonal zirconia (t-ZrO2). Slightly broader
reflections implying smaller crystallite size were observed in SiO2-ZrO2-supported catalysts, a
similar action to Si-doping of TiO2 and SnO2 where amorphous interstitial Si leads to smaller
crystals and blocks crystal growth during their co-oxidation. In the TiO2-ZrO2-supported
catalyst, only reflections corresponding to the mixed oxide phase, (ZrTiO4) srilankite35, were
116 Chapter 4
observed. No reflections belonging to the crystalline TiO2 and ZrO2 were observed but their
presence in amorphous form and/or as very small crystals cannot be ruled out.
fresh
spent
400
SiO2 Al2O3 TiO2 ZrO2
300 SiO2- SiO2- SiO2- TiO2-

BET SSA, m2/g
Al2O3 TiO2 ZrO2 ZrO2
200
100
0
2 O
O
3
O
O
2
2
O
3
rO
2
iO
2
rO
2
Ti
Zr
Si
Al
-T
l 2
-Z
-Z
-A
O
2
O
2
O
2
O
2
Si
Ti
Si
Si
Figure 4.2: Comparison of BET SSAs of fresh (black) and spent (red) 4.5%Co-loaded
catalysts. Inset shows photographs of the powders of fresh single and mixed oxides loaded
with 4.5% Co.
After ODHE at standard conditions, only the XRD patterns of TiO2- and ZrO2-
supported catalysts were changed (Fig. 4.1b). In such spent TiO2-supported catalyst the
anatase was transformed to rutile TiO2. Its reflection was narrower than in the fresh sample,
indicating crystal growth. In the spent ZrO2-supported catalyst, the majority of t-ZrO2 was
transformed to larger monoclinic (m) crystals. The high resistance towards sintering of the
Al2O3-, SiO2- and their corresponding mixed oxide-supported catalysts can be attributed to the
intrinsic thermal stability of these oxides or large amorphous fraction. No reflections of CoOx
117
and its mixed oxide were detected in fresh and spent catalysts, indicating their high dispersion
or rather amorphous state.
Figure 4.2 shows the SSA of fresh and spent catalysts. The SSA of all catalysts
decreased after standard ODHE test. The decrease was most significant for TiO2- and ZrO2-
supported catalysts, whereas both SiO2- and Al2O3-supported catalysts showed remarkable
resistance towards sintering, in agreement with XRD (Fig. 4.1). The SSA of fresh TiO2-ZrO2-
supported catalyst was similar to that of the corresponding fresh single oxide supported
catalysts (~ 110 m2/g). However, the decrease in SSA was smaller in the case of the mixed
metal oxides supported catalyst (48 m2/g) due to the formation of the thermally stable
srilankite phase36 (Fig. 4.1). In SiO2-containing catalysts (e.g. SiO2-TiO2), the decrease in
SSA after ODHE was minimal due to resistance of SiO2 towards sintering.
Figure 4.3: HR-TEM images of (a) fresh and (b) spent 4.5%Co/SiO2 catalysts with
corresponding local EDXS measurements. Arrows point towards spots of EDX measurements.
The freshly prepared catalyst powders differed significantly in their color as indicated
by the inset of Fig. 4.2. The catalysts color also confirms the formation of cobalt containing
mixed oxides during synthesis, e.g. blue in SiO2- (cobalt silicate) and Al2O3- (cobalt
118 Chapter 4
aluminate) supported catalysts.37 The black color of the ZrO2-supported catalyst could be an
indication of CoOx species formation. In Figure 4.3 the HR-TEM images of fresh and spent
SiO2-supported catalysts are shown exemplarily. In the fresh catalyst, cobalt-containing
particles were around 2 nm and increased to about 10 nm after reaction. The bright spots in
the images are due to cobalt species embedded in the silica matrix (e.g. cobalt silicate) in the
fresh catalysts, whereas the observed bright spots on the corresponding spent catalysts are
from Co in SiO2 and/or CoOx as indicated by EDX measurements.
4.3.1.2 NH3-TPD and DRIFTS
NH3-TPD measurements on fresh single and mixed oxide supported catalysts were performed
to examine acidic sites on their surfaces (Fig. 4.4a). A very broad NH3-desorption signal was
observed with the SiO2-supported catalyst compared to that of Al2O3- and TiO2-supported
ones. The ZrO2-supported catalyst showed the narrowest low temperature desorption peak
(around 200 C) among all catalysts. The SiO2-Al2O3-supported catalyst possessed both weak
and medium acidic sites (similar to that observed for the Al2O3-supported catalyst) beside
strong acidic sites, indicated by the peak at ~500 C. For the TiO2-ZrO2-supported catalyst,
the NH3 desorption profile is broader than that of the ZrO2-supported catalyst and similar to
that of the TiO2-supported catalyst.
Figure 4.4b shows the deconvoluted profiles of some selected catalysts. In Al2O3- and
SiO2-Al2O3-supported catalysts, the low temperature (100-250 C) signal was composed of
physisorbed NH3 and weakly adsorbed NH3 at Brnsted acid sites, respectively.38 The
deconvoluted peak around 290 C can be attributed to strong Lewis acid sites, whereas higher
desorption temperatures can be attributed to strong Brnsted acid sites.38 Among single metal
oxide supported catalysts, the SiO2-supported one showed the highest total NH3 uptake (210
mol NH3/gcat) followed by those with TiO2, Al2O3 and ZrO2 supports (Table 4.1). Mixed
oxide supports showed relatively higher NH3 uptake than single ones. However, when the
amount of NH3 desorbed is referred to the catalysts SSA, different tendencies are observed,
particularly for single metal supported catalysts (Table A.2). Silica-supported catalyst
exhibited the lowest surface density of acidic sites (0.50 mol NH3/m2cat), whereas for the
TiO2-supported one highest value was measured (1.8 mol NH3/m2cat). High surface densities
of acidic sites were also found for TiO2-ZrO2 and SiO2-TiO2-supported catalysts.
119
Figure 4.4: (a) NH3-TPD analysis of differently supported 4.5%Co-catalysts and (b)
corresponding deconvoluted plots of selected patterns.
Figure 4.5 shows the DRIFT spectra of NH3 adsorbed on a) single and b) mixed oxide
supported catalysts. All bands at 3600-3750 cm-1 can be attributed to NH3 reaction with the
residual OH groups of these flame-made oxides.39 Multiple bands in the range of 3100 to
3500 cm-1 observed in SiO2- and Al2O3-supported catalysts arise from the NH3 stretching
modes on the support.40 Moreover, both catalysts showed a band at 1619 cm-1 due to
asymmetric NH3 deformation in addition to 1282 (on Co/SiO2) and 1242 cm-1 (on Co/Al2O3)
due to symmetric NH3 deformation. Both catalysts contained Brnsted acid sites as evidenced
by a band around 1450 cm-1. The multiple bands in the range of 3100 to 3500 cm-1 can be
attributed to the stretching vibration of NH3 on the TiO2 surface.40 The bands at 1414 and
1540 cm-1 are indicative of Brnsted acid sites and the existence of NH2-bending vibration,
respectively41, and 1594, 1158 and 1112 cm-1 of Lewis-acid centers.40 The latter doublet
120 Chapter 4
bands are due to inversion splitting.42 The bands at 1615 and 1223 cm-1 are due to N-H
deformation and the 3308, 3203, 3168 cm-1 signals related to N-H stretching are assigned to
the Lewis-acidic sites on the ZrO2-supported catalyst.40 Additionally, a band at 1460 cm-1 is
assigned to Brnsted acid sites.40
Table 4.1: Amount of NH3-desorbed and H2-uptake of single and mixed oxide supported
4.5%Co catalysts.
4.5%Co- mol NH3/gcat mol H2/gcat
loaded
100-350 C 350-650 C total < 500 C > 500 C
SiO2 190 20 210 - 403
Al2O3 184 - 184 - -
TiO2 179 16 195 - 757
ZrO2 89 - 89 912 113
SiO2-Al2O3 150 65 224* - -
SiO2-TiO2 287 2 289 - 173
SiO2-ZrO2 179 - 179 254
TiO2-ZrO2 172 34 206 146 568

*Additional 9 molNH3/gcat originate from strong acid sites (> 650 C, Fig. 4.4b)
121
3308
a
3168
3203
3400
3677
1223
1615
1460
ZrO2 x1/5
Kubelka-Munk, a.u.
1158
1594
1112
1414
1540
TiO2
3747
1619
1242
1449
3707
Al2O3
1282
SiO2
x1/10
4000 3500 3000 2000 1500 1000

1190
3245
b
3157
3308
3672
3383
1603
3747
1464
Kubelka-Munk, a.u.
TiO2-ZrO2
1619
3707
1278
SiO2-ZrO2
x1/3
SiO2-TiO2
1289
SiO2-Al2O3
4000 3500 3000 2000 1500 1000

Wavenumber, cm-1
Figure 4.5: DRIFTS spectra of NH3 adsorbed on (a) single and (b) mixed metal oxides
supported 4.5%Co catalysts. Conditions see experimental part.
122 Chapter 4
The spectra in the NH3-stretching range of 3100-3400 cm-1 for mixed oxide supported
catalysts remained similar to the single oxide ones (Fig. 4.5a). A broad hump ranging from
2700 to 3100 cm-1 in catalysts containing SiO2 can be attributed to perturbed OH vibrations43
in Si-OHNH3. The band at 1464 cm-1 is relatively intense and broader than that of single
oxides, indicating an increased density of Brnsted acid sites. Moreover, a slight shift in the
position of the bands belonging to the Lewis acid sites of Al2O3-, TiO2- and ZrO2-supported
catalysts was observed upon introduction of the second metal oxide, an indication of an
increase in catalyst acidity.44 These observations are in agreement with the NH3-TPD results
(Table 4.1).
4.3.1.3 Raman
Figure 4.6a shows Raman spectra of fresh catalysts. No significant signals attributable to Si-O
stretches or Al2O3 were observed in the SiO2 and SiO2-Al2O3-supported catalysts, probably
due to the low Raman scattering of these oxides.36 A very small signal at 693 cm-1 was
observed with the Al2O3-supported catalyst but could not be assigned unambiguously due to
the similarity in band position of cobalt oxide and CoAl2O4.45 However, the color of the
Al2O3-supported catalyst was blue (Fig. 4.2 inset) indicating the formation of Co-mixed
oxides32, e.g. CoAl2O4, consistent with the observation made by Hj et al.32 and Minnermann
et al.46 for flame-made Co-Mo/Al2O3 and Co/Al2O3 catalysts, respectively.
For TiO2-supported catalyst Raman bands at 439 and 610 cm-1 (due to rutile TiO2)47
and 146 cm-1 (due to anatase TiO2)48, were detected in agreement with XRD (Fig. 4.1a). The
band around 250 cm-1 is the result of the second-order Raman scattering of the rutile
structure.49 For the ZrO2-supported catalyst50, only the bands at 146, 266, 317, 461 and 646
cm-1 corresponding to t-ZrO2 were observed. The band at 693 cm-1 reveals the presence of
Co3O4.45 In the SiO2-TiO2-supported catalyst, the band at 447 cm-1 belonging to rutile TiO2
(rest anatase) was observed further confirming its presence in the catalyst. The absence of
Raman bands of the SiO2-ZrO2-supported catalyst even though the crystal phase of ZrO2 was
distinctly detected by XRD, indicates that SiO2 impaired the measurement. The observed
Raman band positions of the TiO2-ZrO2-supported catalyst are different than those of the
TiO2- and ZrO2-supported catalysts, further confirming the formation of a mixed oxide
phase51, namely, ZrTiO4 (srilankite)52, as was indicated by XRD (Fig. 4.1a). The bands at 286,
326, 414 and 634 cm-1 can be assigned to ZrTiO4.52 The broad band around 146 cm-1 could be
an overlap of anatase TiO2, t-ZrO2 and ZrTiO4 (~160 cm-1) due their close proximity.
123
146 a
286 414
326
634
TiO2-ZrO2
SiO2-ZrO2
396 447
Intensity, a.u.
517 640
SiO2-TiO2
SiO2-Al2O3
266
317 461 646 693 ZrO
2
439
610
250
TiO2
Al2O3
SiO2
200 400 600 800

144 b
TiO2-ZrO2
SiO2-ZrO2
398 443 516 608 638 SiO -TiO

Intensity, a.u.
2 2
177 474 SiO2-Al2O3

188 332 380 614
300 500 549 634
221 266 685
ZrO2
439
609
236266
335
143 693
TiO2
Al2O3
SiO2
200 400 600 800

Raman shift, cm-1
Figure 4.6: Raman Spectra of (a) fresh and (b) spent supported 4.5%Co-loaded catalysts.
Raman spectroscopy of most spent catalysts showed features similar to the fresh ones
with only a slight change mainly due to fluorescence (Fig. 4.6b). Significant changes were
observed only for spent TiO2- and ZrO2-supported catalysts. New bands at 236, 266, 335 and
693 cm-1 were detected matching the vibrational modes of CoTiO3 (693 cm-1 could be also
from Co3O4) in the TiO2-supported catalysts.53,54 The rest of the bands belong to rutile TiO2.
124 Chapter 4
For the spent ZrO2-supported catalyst, all bands can be assigned to m-ZrO2 except the bands
at 266 and 685 cm-1, which belong to t-ZrO250 and Co3O445, respectively. Raman bands of the
spent TiO2-ZrO2-supported catalyst show low intensity but broad features around 200, 450
and 600 cm-1 belonging to ZrTiO4 and indicating that the crystal structure is retained during
ODHE in agreement with XRD (Fig. 4.1b). The suppression of the Raman spectra intensity
could be due to the interference created by deposited coke on the catalyst surface. In all spent
catalysts, Raman bands around 1300 and 1600 cm-1 were observed (not shown), belonging to
disordered (D) and graphitic (G) carbon, an indication of coke deposition during the
reaction.55
4.3.1.4 UV-Vis reflectance and reducibility
Figure 4.7 shows UV-Vis spectra of a) pure single metal oxides, b) Co-containing single
metal oxides and c) Co-containing mixed metal oxides. Fairly flat (400-800 nm) spectra were
obtained with pure single metal oxide supports (Fig. 4.7a). Therefore, the observed changes in
the spectra of Fig. 4.7bc can be attributed to the CoOx and/or its interaction with the support.
Flame-made pure CoOx exhibited bands around 400 and 710 nm (Fig. 4.7a), close to the
reported values of around 430 and 720 nm for Co3O4.21 In all catalysts containing TiO2, red-
shifting of its band at 300-400 nm is observed that can be attributed to TiO2 rutilization by
cobalt. A broad feature is observed in the range of 575-735 nm, which can be assigned to
CoTiO3.56 A similar feature (500-700 nm) was also observed with the TiO2-ZrO2-supported
catalyst indicating the formation of mixed oxide in accordance with XRD (Fig. 4.1). Two
broad signals around 380 and 690 nm were observed with the ZrO2-supported catalyst that
can be attributed to CoOx. These signals were absent in other catalysts. However, since
relatively high H2-uptake was observed over TiO2- and TiO2-ZrO2-supported catalysts (Table
4.1), the features of CoOx could have been overlapped/suppressed by bands from other
structures. More specifically, the first signal from CoOx (around 380 nm) is clearly masked by
the TiO2 absorption spectrum, whereas the absence of the second signal (around 690 nm) is
likely due to different interaction of CoOx with TiO2 and TiO2-ZrO2 compared to that with
ZrO2, which will be discussed later. Three distinct bands were observed around 520, 590 and
635 nm and 540, 580 and 630 nm with SiO2- and Al2O3-supported catalysts, respectively.
125
pure CoOx
pure ZrO2
pure TiO2
pure Al2O3
pure SiO2
300 400 500 600 700 800
ZrO2
F(R), a.u.
TiO2
Al2O3
SiO2
300 400 500 600 700 800
TiO2-ZrO2
SiO2-Al2O3
SiO2-ZrO2
SiO2-TiO2
300 400 500 600 700 800

Wavelength, nm
Figure 4.7: UV-Vis reflectance spectra of (a) pure single metal oxides, (b) Co-containing
single metal oxide and (c) Co-containing mixed metal oxides. In all the measurements, the
powder sample was mixed with BaSO4 keeping the mass ratio approx. 1/10.
126 Chapter 4
The structural feature of SiO2-ZrO2-supported catalyst was similar to that of the SiO2-
supported one. The SiO2-TiO2-supported catalyst showed spectral features (3 bands) similar
to that of SiO2- supported one. Such spectral features of the single metal oxide supported
catalysts (Al2O3 and SiO2) are attributed to the three spin allowed d-d transitions of
tetrahedrally coordinated Co2+ in Co-aluminate57 and Co-silicate species58. The absence of
octahedral Co3+ species indicates that flame synthesis favors formation of tetrahedral species
incorporated into the Al2O332- and SiO259-matrix.
H2-TPR profiles of all as-prepared catalysts are presented in Fig. 4.8 along with
deconvolution of the SiO2-, ZrO2- and TiO2-ZrO2-supported ones. Significant differences in
the reduction profiles indicate that the oxidation state and the interaction of the cobalt species
strongly depend on support composition. For Co3O4 generally, a two-step reduction is
observed: Co3O4 CoO and CoO Co.60 This was also confirmed by TPR over
mechanically mixed commercial Co3O4 (Sigma Aldrich) with flame-made SiO2 (90 wt%) (not
shown). The powder prepared by mechanical mixing showed a broad reduction peak ranging
from 200 to 450 C. Generally, relatively strong interaction between the CoOx species and the
support was observed with all catalysts, indicated by a shift of the reduction peaks to higher
temperatures (> 450 C).60
The silica-supported catalyst could only be reduced above 750 C, indicating strong
interaction between cobalt oxide and silica, most likely due to formation of cobalt silicate.61 A
distinct shoulder was observed (~820 C), indicating the existence of different types of
interactions between CoOx and SiO2. No reduction peaks were observed in both the Al2O3-
and SiO2-Al2O3-supported catalysts, which could be attributed to the possible formation of
CoAl2O4 that can only be reduced above 900 C.32,62 The high temperature reduction peak
(~600 C) of the TiO2-supported catalyst can be attributed to the strong binding between
CoOx and rutile TiO2.63
127
TiO2-ZrO2
H2-MS signal intensity, a.u.

SiO2-ZrO2
SiO2-TiO2
SiO2-Al2O3
ZrO2
TiO2
Al2O3
SiO2
200 400 600 800

Temperature, C
Figure 4.8: H2-TPR of differently supported 4.5%Co-loaded catalysts and deconvolution of

selected profiles. Conditions see experimental part.
For ZrO2-supported catalysts the low temperature reduction peak (around 200 C) can
be assigned to nano-sized easily reducible surface Co3O4 species.64 The high temperature
reduction peak was shifted to higher temperature (around 500 C) compared to that of pure
commercial Co3O4 (around 400 C), which can be ascribed to the reduction of strongly bound
CoOx on the ZrO2 surface.64 The presence of a shoulder between low (~210 C) and high
(~280 C) temperature reduction peaks can be attributed to the stepwise reduction of a
fraction of Co3O4 to Co0. Silica containing TiO2 and ZrO2 mixed oxide supported catalysts
showed high temperature reduction peaks of low intensity (~ 810 C) assignable to the
reduction of cobalt silicate. With the TiO2-ZrO2-supported catalyst, a high temperature
128 Chapter 4
reduction peak (~ 500 C) was observed, also indicating strong interaction between CoO x and
the support. This reduction peak of the TiO2-ZrO2-supported catalyst is positioned at higher
temperature compared to that of ZrO2- and at slightly lower position than that of TiO2-
supported catalysts. The broad feature observed for the TiO2-ZrO2-supported catalyst implies
that virtually all CoOx species may have been embedded in the support matrix making them
difficult to reduce at lower temperatures.63,64
Hydrogen uptakes of the various catalysts are summarized in Table 4.1. The ZrO2-
supported catalyst exhibited the highest total H2-uptake (~1025 mol/gcat), followed by TiO2-,
TiO2-ZrO2-, SiO2-, SiO2-ZrO2- and SiO2-TiO2-supported catalysts. We can distinguish
between H2-uptake due to CoOx species reducible at lower temperature (< 500 C), as it
occurred with the ZrO2-supported catalyst and higher temperature (> 500 C), as observed
with the SiO2-supported one. Al2O3- and SiO2-Al2O3-supported catalysts showed no
significant H2-uptakes, indicating that cobalt species were not reducible up to 840 C. Lower
H2-uptake by TiO2- and TiO2-ZrO2-supported catalysts compared to the ZrO2-supported one
indicates that CoOx-admixing to TiO2 results in less reducible species (stronger support
interaction), e.g. CoTiO3.54 Correlating this with the UV-Vis results (Fig. 4.7) of the former
two catalysts (TiO2- and TiO2-ZrO2), the absence of the second signal of CoOx (around 690
nm) can be attributed to this Co-Ti interaction.
4.3.1.5 XPS
Figure 4.9 shows the Co 2p spectra obtained from XPS analysis over selected supported
catalysts. The intensity of signals of the ZrO2-supported catalyst is higher compared to those
from TiO2-ZrO2- and SiO2-supported catalysts, which could be related to the presence and/or
amount of reducible CoOx species.65 The Co 2p3/2 and Co 2p1/2 signals appeared at binding
energies (BEs) of 779.4 eV and 794.9 eV, respectively, which also had corresponding satellite
peaks appearing at 784.5 eV and 801.0 eV, over ZrO2 supported catalysts (Table A.1). Similar
values were also obtained over TiO2-ZrO2-supported catalyst. These BE values are close to
those reported for Co3O4 and CoO confirming the presence of reducible cobalt oxides, which
is in good agreement with reducibility results (Fig. 4.8).65,66 However, a small difference in
the BEs between these aforementioned supported catalysts exists (around 0.4 eV) pointing to
the possible presence of a slightly higher amount of Co3O4 in the catalyst surface exhibiting
the low values (i.e. over ZrO2). Over the SiO2-supported catalyst, the observed signals are
vastly different from those of the other two supported catalysts, implying the absence of
129
reducible CoOx species and the formation of Co-Si mixed oxide (e.g. cobalt silicate)67, in
agreement with our UV-Vis and H2-TPR results.
Co 2p3/2
Co 2p1/2
satellite satellite
TiO2-ZrO2
Counts/s, a.u.
ZrO2
SiO2
810 800 790 780 770 760

Binding energy, eV
Figure 4.9: XPS measurement over selected fresh 4.5%Co-loaded catalysts.
4.3.2 Catalytic Performance
4.3.2.1 Identified products and relevant reactions observed
During the ODHE reaction, besides the desired ethene also the formation of water, H2, CO,
coke and methane were detected indicating simultaneous occurrence of several side reactions.
Therefore, based on observations and with literature support, the following reactions are
proposed to be relevant. With all catalysts, ethene was the major product confirming ODHE
(C2H6 + CO2 C2H4 + CO + H2O) and/or DHE (C2H6 C2H4 + H2) as the governing
130 Chapter 4
reactions. In the former case, CO2 has been reported to oxidize the reduced active sites7,28
(redox pathways) whereas in the latter case, it pushes the equilibrium towards product side by
consuming H2 via reverse water gas shift (RWGS) reaction (CO2 + H2 CO + H2O).11 The
RWGS reaction has been investigated separately over some selected catalysts (Fig. A.4) and
will be discussed later. Production of CO from the reaction can be related to RWGS, dry
reforming of ethane (C2H6 + 2CO2 4CO + 3H2) and/or reverse Boudouard reaction (CO2 +
C 2CO) but to separate contributions of these feasible reactions is rather challenging. Coke
formation by cracking was observed with all catalysts but to different extent (Table A.2).
Methane production (Fig. A.3) can be related to CO and/or CO2 methanation or to ethane
hydrogenolysis (C2H6 + H2 2CH4). The above mentioned reactions have been observed in
several studies.11,12,68 The elucidation of the individual contributions of all side reactions is
very demanding and was beyond the scope of this work.
4.3.2.2 Influence of single and mixed metal oxides supports
In a first step the ODHE performance of all catalysts was compared at standard conditions
(700 C, feed ratio CO2/C2H6 = 2.5, total reactant gas (C2H6 and CO2) flow rate 15 mL/min,
W/F = 0.6 g.s/mL) after 2 h on stream (Fig. 4.10). SiO2- and ZrO2-supported cobalt catalysts
showed superior performance compared to those supported on other single metal oxide
supports. A similar ethene yield (around 23%) was observed for SiO2- and ZrO2-supported
catalysts, with an ethene selectivity of 81% and 62%, respectively. The activity of TiO2- and
Al2O3-supported catalysts was similar to that of the homogeneous gas phase reaction in the
absence of catalyst (ethene yield ~15%). Wang et al7 reported a similar behavior for Cr-doped
catalysts for ODHE where the performance of the SiO2-supported catalyst was superior
compared to other supported catalysts. The yield of SiO2-supported catalyst was better than
that of catalysts where SiO2 was mixed with ZrO2, TiO2 or Al2O3. The TiO2-ZrO2-supported
catalyst exhibited an ethene yield of ~24%, which is comparable to that of both SiO2- and
ZrO2-supported catalysts, though this mixed-oxide-supported catalyst showed lower ethene
selectivity than that of the optimal SiO2-supported catalyst. However, the lower selectivity is
compensated by an increase in ethane conversion resulting in comparable overall
performance. The improvement in ethane conversion can be attributed to its higher resistance
towards sintering (Fig. 4.2) and an increase in the surface acidity (Table A.2). Correlating
CO2 conversions (Table A.2) with the reaction results (Fig. 4.10), it can be inferred that
conversion of CO2 is highly affected by the support type, which will be discussed further in
detail.
131
100
C2H6 conversion
C2H4 selectivity
Conversion, selectivity, yield %
C2H4 yield
80
60
40
20
0
O
3
O
2
O
O
2
iO
rO
2
lO
2
3
rO
2
Zr
Ti
Si
Al
2
-T
-Z
2
-Z
-A
O
2
O
2
O
2
O
2
Si
Si
Ti
Si
Figure 4.10: Influence of support on ODHE reaction over 4.5%Co-loaded catalysts. Reaction
conditions: T = 700 C; CO2/C2H6 = 2.5; total flow rate = 15 mL/min; catalyst wt. = 150 mg;
2 h time on stream.
Reaction rates referred to unit surface area (mol C2H6/m2.s) of the catalysts were
evaluated to gain insight into the activity of the catalysts (Fig. A.2). Interestingly, the SiO2-
supported catalyst showed the lowest rate (~13 mol C2H6/m2.s), while the ZrO2- and TiO2-
ZrO2- supported ones showed the highest rates (~60 molC2H6/m2.s). Among all the catalysts,
Al2O3-, ZrO2- and TiO2-ZrO2-supported ones showed highest CH4 selectivity, while the
lowest was observed over the SiO2-supported one under the applied standard reaction
conditions (Table A.2). This indicates that cracking or methanation is promoted over these
three former catalysts. Highest amount of coke was deposited over the Al2O3-supported
catalyst (~4 wt%) followed by catalysts based on TiO2-ZrO2 and SiO2-Al2O3 SiO2 (1.9
wt%) supports, which can be associated with their acidity (Table 4.1). Efficient removal of
132 Chapter 4
coke and/or catalyst physicochemical changes (e.g. SSA and acidity) during the reaction
could be the reason for relatively low coke deposition over the other catalysts (Table A.2).
80
C2H4 selectivity, %
60
40
SiO2 SiO2-Al2O3
20 Al2O3 SiO2-TiO2
TiO2 SiO2-ZrO2
ZrO2 TiO2-ZrO2
0
15 20 25 30 35 40 45 50
C2H6 conversion, %
Figure 4.11: Selectivity versus conversion of differently supported 4.5%Co-loaded catalysts

at contact times of 0.1-1.2 g.s/mL. Reaction conditions: T = 700 C; CO2/C2H6 = 2.5; total
flow rate = 15 mL/min; catalyst wt. = 25-300 mg; 2 h time-on-stream.
Deeper insight into the catalyst performance is gained by considering the selectivity
versus conversion plots. For that purpose the catalysts were tested at different contact times
(0.1-1.2 g.s/mL), while keeping the other reaction conditions constant (Fig. 4.11). At
isoconversion values > 35%, the SiO2-supported catalyst showed the highest ethene
selectivity among all tested catalysts confirming that it is the best performing catalyst for
ODHE under these conditions. This catalyst showed ethane conversion and ethene selectivity
as high as 46% and 74%, respectively. Similar high ethane conversion (46%) was also
observed over ZrO2- and TiO2-ZrO2-supported catalysts but at the expense of ethene
selectivity. The undesired side reactions of ethene (e.g. formation of CH4 and coke) became
133
increasingly significant at higher conversion/contact time. The CH4 selectivity of SiO2-

supported catalyst was virtually constant at all contact times indicating the occurrence of
ethene conversion mainly to coke with increasing contact time (Fig. A.3). While over ZrO2-
and TiO2-ZrO2-supported catalysts conversion of ethene to CH4 seemed to be favored with
increasing contact time as evidenced by a small increment in its selectivity. The opposing
trend in ethane conversion and ethene selectivity over these catalysts indicates a trade-off
between them that poses a critical challenge for developing an efficient catalyst.
At isoconversion values < 35%, SiO2- and SiO2-ZrO2-supported catalysts exhibited
better performance than TiO2-, SiO2-Al2O3- and SiO2-TiO2-supported ones. Though similar
catalytic behavior was observed over SiO2- and SiO2-ZrO2-supported catalysts, improvement
in ethane conversion with minimal change in ethene selectivity could not be achieved over the
latter even when using the highest contact time (1.2 g.s/mL). Overall, the SiO2-supported
catalyst was the most efficient catalyst exhibiting an ethene selectivity of 74% at 46% ethane
conversion, resulting in an ethene yield of 34%.
During ODHE, CO2 was consumed over all catalysts confirming its active
participation in this reaction (Table A.2). Higher CO2 conversion was observed over ZrO2-
(17%) and TiO2-ZrO2- (22.3%) supported catalysts than over the others (3.4-9.5%). Oxidation
of the deposited coke from the catalyst surface, reaction with H2 favoring the forward ethane
dehydrogenation reaction12 and oxidation of reduced active species are some of the
pathways28 consuming CO2. The high CO2 conversion over ZrO2- and TiO2-ZrO2-supported
catalysts can be directly correlated to the high ethane conversion and low ethene selectivity
(Fig. 4.10), where CO2 assists in favoring the forward dehydrogenation reaction by
consuming H2 and also to its beneficial action in regenerating active sites blocked by
carbonaceous deposits.12,28 The reverse water gas shift (RWGS) reaction was performed
independently over SiO2-, ZrO2- and TiO2-ZrO2-supported catalysts to evaluate its role in the
reaction network (Fig. A.4). High RWGS activity was observed over the ZrO2- and TiO2-
ZrO2-supported catalysts evidenced by high CO2 conversion. Therefore, during the ODHE
reaction over these two catalysts, high ethane conversion can be directly related to the high
RWGS reaction rate, which shifts the equilibrium of the ODHE reaction towards the products.
Moreover, compared to the SiO2-supported catalyst, these two catalysts exhibited slightly
higher CH4 selectivity indicating high cracking or methanation rates. Since ethene is more
reactive than ethane, the decrease in its selectivity is most likely due to cracking.
Additionally, the low ethene selectivity can also arises from unselective reaction of ethane
and CO2 (dry reforming of ethane).69 The CO2 conversion increased with increasing contact
134 Chapter 4
time, mainly over SiO2-, ZrO2- and TiO2-ZrO2-supported catalysts, further corroborating the
CO2 participation in ODHE (Fig. A.5).
SiO2-, ZrO2-, and TiO2-ZrO2-supported catalysts (Fig. 4.10 and 4.11) may be taken to
elucidate the influence of the various characteristics (reducibility, Fig. 4.8, Table 4.1; surface
acidity, Fig. 4.4, Table 4.1; surface area, Fig. 4.2) of the catalysts on their activity and
selectivity. We speculate that depending on the reducibility of the cobalt species and the
surface acidity two different reaction pathways may prevail, a redox and a non-redox
mechanism. In the case of SiO2-supported catalyst, cobalt silicate was formed, in which the
cobalt species are fairly stable against reduction.61 It can be assumed that the tetrahedrally
coordinated Co2+ ions that are present in cobalt silicate58 are essential for the catalytic
performance. Similar observation was also reported for cobalt-exchanged hydroxyapatite19
and Co/MCM-4124 catalysts during oxygen-assisted ODHE and oxidative dehydrogenation of
isobutane, respectively. Since the bare flame-made SiO2 is virtually inactive, showing a
performance similar to that of an empty tube (~15% yield), it can be concluded that the high
activity of the SiO2-supported catalyst is exclusively due to cobalt that exists predominantly
as cobalt silicate likely containing tetrahedrally coordinated Co2+ ions (Fig. 4.7). In contrast,
octahedrally coordinated Co3+ ions (CoOx species) prevail in ZrO2- and TiO2-ZrO2-supported
catalysts, as corroborated by the low temperature reduction profiles (Fig. 4.8) and Co 2p3/2,1/2
spectra of XPS (Fig. 4.9).
As concerns the role of surface acidity, SiO2- and TiO2-ZrO2-supported catalysts
showed similar overall NH3 uptake, whereas that of the ZrO2-supported catalyst was much
lower (Table 4.1). The surface acidity was normalized by the catalysts SSA and correlated to
the activity of the catalyst to further understand the activity-acidity relationship (Fig. A.6).
Regardless of the difference in the surface density of acidic sites, all the catalysts showed
virtually similar ethane conversion except the SiO2-, ZrO2- and TiO2-ZrO2-supported ones.
Among these, the SiO2-supported catalyst possessed the lowest density of acid sites followed
by ZrO2- and TiO2-ZrO2-supported ones (Table A.2). An increase in ethane conversion was
observed with increasing surface density of acidic sites, whereas an opposite trend was
observed when relating them to the ethene selectivity. This indicates that these sites are
reactive towards both ethane and ethene and indicates that beside the state and reducibility of
the cobalt species, surface acidity plays a role in ODHE, particularly by promoting side
reactions. Nevertheless, regardless of these differences, the overall performance (ethene yield)
of these catalysts was virtually similar, lending further support that most likely different
reaction pathways occur with these catalysts.
135
As regards the role of the surface area, it seems to be of lesser importance than
reducibility and surface acidity. This emerges from Fig. 4.2 where mixed oxide supported
catalysts containing SiO2 had relatively high SSA (> 150 m2/g) but their activity was low
compared to that of ZrO2 and TiO2-ZrO2.
With the SiO2-supported catalyst tetrahedrally coordinated Co2+ ions in the SiO2
matrix and surface acidity seem crucial. Recently, Hu et al.70 showed that the selective
propane dehydrogenation proceeds via a non-redox mechanism over Co/SiO2 where Co2+ ions
are tetrahedrally coordinated in the SiO2 matrix. It was speculated that the heterolytic
cleavage of the ethane C-H bond is the key step in this reaction and the process is catalyzed
by the non-redox single site i.e. Co2+ that is tetrahedrally coordinated with the SiO2 structure.
This conclusion was derived based on the resistance of these ions to both oxidation and
reduction, in agreement with previous reports on Co exchanged zeolite ZSM-5.71 In addition,
the residual Brnsted acid sites also significantly contributed to the activity of the zeolite
catalyst.70,71 Thus, CO2 mainly reacts with coke and/or H2 over the SiO2-supported catalyst.
The reaction is expected to occur primarily as a consequence of the redox cycle of Co-O
species over the ZrO2- and TiO2-ZrO2-supported catalysts, wherein CO2 acts as an oxidant28,
in addition to its two previously mentioned contributions. An increase in ethane conversion
with a concomitant decrease in ethene selectivity was observed that could be attributed to the
higher acidity of the TiO2-ZrO2-supported catalyst.
Summarizing, evidence has been accumulated that both reducibility of the cobalt
species as well as the surface acidity are important descriptors for optimizing supported cobalt
catalysts for CO2-assisted ODHE. The superior performance of the SiO 2-supported cobalt
catalysts can be attributed to the role of the relatively stable tetrahedrally coordinated Co 2+
ions near the surface, whereas the redox behavior of easier reducible CoOx species seems to
be crucial for explaining the performance of ZrO2- and TiO2-ZrO2-supported catalysts.
The best performing catalysts were also evaluated for their potential for the ODHE
reaction using O2 as an oxidant (Table A.3). All the reaction conditions were kept similar as
described in the experimental section except the reaction temperature and the reactant gas
composition, which were changed to 500 C and C2H6:O2:He = 6.4:3.2:5.4 mL/min,
respectively. This temperature was chosen since it seemed to represent the mean value of the
generally reported reaction temperatures (300-700 C) and the reactant gas ratio of 2 is equal
to the stoichiometric ratio of the reaction (C2H6 + 0.5O2 C2H4 + H2O).1,6 Ethane
conversion and ethene selectivity were less than 10% and 40%, respectively, over SiO 2- and
ZrO2-supported catalysts. However, these values were higher over TiO2-ZrO2-supported
136 Chapter 4
catalysts (C2H4-yield of ~16%). Former two supported catalysts promoted the unselective
reactions, whereas the latter one favored the desired reaction. Performance of these catalysts
was generally lower compared to most other true ODHE catalysts.1,72,73 Nevertheless, the
activity of the mixed oxides supported catalyst seemed to be comparable to some of the
reported values exhibiting some potential as ODHE catalyst under true oxidative conditions
(with O2 or air).19,21 Overall, comparing the catalytic results from O2 and CO2 assisted ODHE,
it can be inferred that the use of the latter oxidant is beneficial.
4.3.2.3 Performance of SiO2-supported catalyst at different reaction conditions
For deeper insight into the behavior of the best performing SiO2-supported catalyst, the
influence of reaction temperature (650-725 C), feed gas composition (CO2/C2H6 = 0-4.5) and
gas hourly space velocity (3,000-36,000 L/kgcat.h) as well as the catalysts stability was
investigated. Figure 4.12a shows the influence of reaction temperature on the ODHE
performance of the SiO2-supported catalyst. At higher reaction temperatures, in addition to
the catalytic reaction, some contribution of homogeneous gas phase reactions seems to
become significant. An ethane conversion of 20% was achieved at 650 C, which increased
with increasing reaction temperature, attaining nearly 50% at 725 C. However, at T > 675
C, the ethane selectivity was adversely affected. At higher reaction temperatures, formation
of larger metallic cobalt ensembles can be expected25, which promotes the unselective and/or
secondary reactions resulting in lower ethene selectivity.69 Concomitant increase in CO2
conversion was observed with increasing ethane conversion indicating that it assists the
reaction by gasifying the deposited coke and also by reacting with H2, which is primarily
produced by DHE.12 During this CO2 assisted coke gasification, the catalysts surface is
partially cleaned exposing sites (e.g. cobalt metal ensembles) that transform ethene to coke,
thus lowering selectivity to ethene. In addition, CO2 reforming of ethane could also contribute
to high conversions (both ethane and CO2) at higher temperatures.69
137

a
80 C2H4 selectivity
60
40
C2H6 conversion
C2H4 yield
20
CO2 conversion
0
650 675 700 725
Reaction temperature, C
b
80
C2H4 selectivity
60
40 C2H6 conversion
C2H4 yield
20
CO2 conversion
He/C2H6 = 2.5
0
0 1 2 3 4 5
CO2/C2H6 gas ratio, -
Figure 4.12: Influence of (a) reaction temperature and (b) CO2/C2H6 ratio on the catalyst
activity of SiO2-supported Co catalyst.
Figure 4.12b shows the influence of the feed gas composition on the performance of
the SiO2-supported catalyst. The reaction was carried out at constant gas hourly space velocity
of 6,000 L/kgcat.h at 700 C. The ethane conversion dropped from ~30% to ~20% when CO2
was replaced by helium while maintaining the gas ratio of 2.5, in the reactant gas feed.
138 Chapter 4
However by lowering the reactant gas ratio (CO2/C2H6) to 1.5, ethane conversion increased to
~34%, confirming that CO2 plays an important role during reaction, e.g. by removing coke, as
discussed earlier.12,16 After a slight decrease, ethene selectivity increased and remained
virtually constant (around 80%), irrespective of a further increase in the CO2/C2H6 ratio. A
similar trend was found for ethane conversion, which showed a small initial decrease
followed by an increase to a stable value of ~35% resulting the highest ethene yield of ~28%.
Minimal decrease in CO2 conversion was observed with increasing CO2/C2H6 ratio. This
behavior is similar to that observed by Zu et al.69 for ODHE over Co-loaded Mn-
Na2WO4/SiO2 catalysts. Rather high CO2 conversions (22-45%) were achieved at lower
CO2/C2H6 ratio ( 3), however, at the expense of ethene selectivity reducing their yield to
25%. This was attributed to CO2 reforming of C2H6 due to formation and growth of metallic
cobalt domains.69 So high catalyst activity can be achieved by maintaining the oxidic state of
cobalt. Any further improvement in the catalytic activity seems to be mainly limited by the
contact time (Fig. 4.11).
The effect of gas hourly space velocity (GHSV) on the activity of SiO2-supported
catalyst at 700 C and a CO2/C2H6 ratio of 2.5 is shown in Fig. 4.13a. With increasing
GHSVs, ethane conversion decreased, whereas ethene selectivity increased to ~24% and
~86%, respectively, at 36,000 L/kgcat.h. This decrease in conversion can be ascribed to the
low residence time for reactant gases in the catalyst bed. Improvement in the selectivity can
be attributed to the reduction/elimination of secondary reaction pathways of ethene, e.g.
cracking.28,69 At higher GHSVs, ethene selectivity seemed to have reached the optimal state
though a decreasing trend in the ethane conversion was still perceivable indicating similar
product distribution. The ethene yield followed a trend similar to ethane conversion with
increasing GHSV. The CO2 conversion increased with decreasing GHSV corroborating its
active role in the ODHE reaction.
4.3.2.4 Stability of the catalyst
Sintering and coking are the two main reasons for deactivation of ODHE catalysts.10,12
Therefore, an investigation of the stability of the catalyst is necessary to evaluate its efficacy
and identify descriptors for tailoring the optimal catalyst. The time-on-stream test was
performed using the SiO2-supported catalyst for 5 h under standard reaction conditions. The
catalyst was stable showing nearly 30% ethane conversion and 80% ethene selectivity,
exhibiting virtually constant ethene yield, even after 5 h of reaction (Fig. 4.13b). The initial
unsteady period can be attributed to early changes in the physicochemical properties of the
139
catalyst, e.g. SSA decrease and coke deposition. Steady CO2 conversion, which followed the
trend similar to that of ethane conversion, indicates that CO2 was actively involved in the
reaction. XRD patterns of the spent catalyst were similar to that of the fresh one (Fig. 4.1a)
and the SSA was still relatively high (247 m2/g) after the stability test. Therefore, the stable
performance of the catalyst can be attributed to its resistance towards sintering and the ability
to retain most of cobalt-embedded in the SiO2 matrix as cobalt silicate.
a
Conversion, selectivity, yield%
80 C2H4 selectivity
60
40
C2H6 conversion
20 C2H4 yield
CO2 conversion
0
4000 8000 12000 36000 40000
GHSV, L.kg-1.h-1
b
80
C2H4 selectivity
60
40 C2H6 conversion
20 C2H4 yield
CO2 conversion
0
0 50 100 150 200 250 300
Time, min
Figure 4.13: Influence of (a) gas hourly space velocity and (b) time-on-stream on the catalyst
activity of SiO2-supported Co catalyst.
140 Chapter 4
4.4 Conclusions
Highly dispersed CoOx species on various single (SiO2, Al2O3, TiO2 and ZrO2) and mixed
metal oxides (SiO2-Al2O3, SiO2-TiO2, SiO2-ZrO2 and TiO2-ZrO2) were produced using flame
spray pyrolysis. XRD and Raman analyses showed that Co promotes rutilization in TiO2
containing catalysts. Depending on the support different types of interaction between CoOx
and support were revealed by H2-TPR, XPS and UV-Vis spectroscopy. Well reducible CoOx
species were formed in ZrO2-, TiO2- and TiO2-ZrO2-supported catalysts, whereas the SiO2-
containing catalysts showed very high reduction temperature of the cobalt constituent due to
formation of cobalt silicate containing tetrahedrally coordinated Co2+ species. NH3-TPD as
well as DRIFTS showed that SiO2- and TiO2-ZrO2-supported catalysts possessed moderately
strong acidic sites, whereas SiO2-TiO2- and SiO2-Al2O3-supported catalysts showed stronger
acidic sites.
Best catalytic performance in CO2-assisted ODHE was observed with the silica
supported catalyst, which outperformed all others in a wide range of contact times, exhibiting
74% ethene selectivity (others < 60%) at isoconversion of 46% leading to a yield of 34% at
the standard conditions used. The studies revealed that both the reducibility of the Co species
as well as the strength and surface density of acidic sites are critical catalyst properties for
achieving optimal performance in the oxidative dehydrogenation of ethane. However, these
properties seem to be interdependent to some degree, so that an absolute assignment of their
contribution cannot be given without ambiguity. Clear is that the surface acidity also plays a
significant role by promoting side reactions (e.g. reverse water gas shift and cracking).
Coke deposition was observed with all catalysts and Raman spectroscopy revealed it
to be mainly D- and G-type carbon. Parametric sensitivity studies of the best performing
SiO2-supported catalyst showed that high reaction temperature (> 675 C) is detrimental for
ethene selectivity. Low performance in the absence of CO2 confirmed that CO2 plays an
important role in ODHE increasing the conversion by 50%. Change of the oxidant from CO 2
to O2 led to significantly lower ethene yields, indicating that the use of CO2 is beneficial for
these catalysts. The SiO2-supported catalyst showed stable performance for 5 h time-on-
stream. Optimization of this catalyst by increasing the fraction of Co species, which are
embedded in the silica matrix and strongly interact with the silica support (cobalt silicate),
seems a promising strategy to further enhance the performance of this catalyst in ODHE with
CO2.
141
We thank Dr. Frank Krumeich (ETH) for STEM/EDXS investigations and the Electron
Microscopy Center of ETH Zurich (EMEZ) for providing the necessary infrastructure and Dr.
Mario El Kazzi for the XPS measurements. Financial support by ETH Research Grant ETH-
39-12-2 is kindly acknowledged and from the European Research Council under the European
Union's Seventh Framework Program (FP7/2007-2013)/ERC grant agreement (247283).
4.6 References
(2) Bhasin, M. M.; McCain, J. H.; Vora, B. V.; Imai, T.; Pujad, P. R. Appl. Catal. A:
Gen. 2001, 221, 397-419.
Chem. Rev. 2014, 114, 10613-10653.
(6) Cavani, F.; Ballarini, N.; Cericola, A. Catal. Today 2007, 127, 113-131.
2000, 196, 1-8.
(8) Held, A.; Kowalska, J.; Nowiska, K. Appl. Catal. B: Environ. 2006, 64, 201-208.
(10) Zhao, X.; Wang, X. Catal. Commun. 2006, 7, 633-638.
(11) Deng, S.; Li, H.; Li, S.; Zhang, Y. J. Mol. Catal. A: Chem. 2007, 268, 169-175.
(12) Koirala, R.; Buechel, R.; Krumeich, F.; Pratsinis, S. E.; Baiker, A. ACS Catal. 2015, 5,
690-702.
(14) Shen, Z.; Liu, J.; Xu, H.; Yue, Y.; Hua, W.; Shen, W. Appl. Catal. A: Gen. 2009, 356,
148-153.
1998, 1025-1026.
142 Chapter 4
(17) Cheng, Y.; Gong, H.; Miao, C.; Hua, W.; Yue, Y.; Gao, Z. Catal. Commun. 2015, 71,
42-45.
(18) Wang, S.; Murata, K.; Hayakawa, T.; Hamakawa, S.; Suzuki, K. Catal. Lett. 2001, 73,
107-111.
(19) Kabouss, K. E.; Kacimi, M.; Ziyad, M.; Ammar, S.; Bozon-Verduraz, F. J. Catal.
2004, 226, 16-24.
(20) Schuurman, Y.; Ducarme, V.; Chen, T.; Li, W.; Mirodatos, C.; Martin, G. A. Appl.
Catal. A: Gen. 1997, 163, 227-235.
(21) Brik, Y.; Kacimi, M.; Ziyad, M.; Bozon-Verduraz, F. J. Catal. 2001, 202, 118-128.
(22) Mitran, G.; Cacciaguerra, T.; Loridant, S.; Tichit, D.; Marcu, I.-C. Appl. Catal. A:
Gen. 2012, 417418, 153-162.
(23) Davies, T. E.; Garcia, T.; Solsona, B.; Taylor, S. H. Chem. Commun. 2006, 3417-
3419.
(24) Corts Corbern, V.; Jia, M. J.; El-Haskouri, J.; Valenzuela, R. X.; Beltrn-Porter, D.;
Amors, P. Catal. Today 2004, 9192, 127-130.
(25) Ruckenstein, E.; Wang, H. Y. Catal. Lett. 2001, 73, 99-105.
(26) Nakagawa, K.; Kikuchi, M.; Nishitani-Gamo, M.; Oda, H.; Gamo, H.; Ogawa, K.;
Ando, T. Energ. Fuel 2008, 22, 3566-3570.
(27) Li, Y.-N.; Guo, X.-H.; Zhou, G.-D.; He, X.; Bi, Y.-L.; Li, W.-X.; Cheng, T.-X.; Wu,
T.-H.; Zhen, K.-J. Pol. J. Chem. 2005, 79, 1357-1364.
(29) Koirala, R.; Pratsinis, S. E.; Baiker, A. Chem. Soc. Rev. 2016,
DOI:10.1039/c5cs00011d,
(31) Schimmoeller, B.; Jiang, Y.; Pratsinis, S. E.; Baiker, A. J. Catal. 2010, 274, 64-75.
(32) Hj, M.; Linde, K.; Hansen, T. K.; Brorson, M.; Jensen, A. D.; Grunwaldt, J.-D. Appl.
Catal. A: Gen. 2011, 397, 201-208.
(33) Shannon, R. Acta Crystallogr. A 1976, 32, 751-767.
(34) Akhtar, K. M.; Pratsinis, S. E.; Mastrangelo, S. V. R. J. Mater. Res. 1994, 9, 1241-
1249.
(35) Kong, L. B.; Ma, J.; Zhu, W.; Tan, O. K. J. Alloys Compd. 2002, 335, 290-296.
(36) Reddy, B. M.; Ganesh, I.; Chowdhury, B. Catal. Today 1999, 49, 115-121.
143
(37) Khodakov, A. Y.; Chu, W.; Fongarland, P. Chem. Rev. 2007, 107, 1692-1744.
(38) Martins, G. V. A.; Berlier, G.; Bisio, C.; Coluccia, S.; Pastore, H. O.; Marchese, L. J.
Phys. Chem. C 2008, 112, 7193-7200.
(39) Primet, M.; Pichat, P.; Mathieu, M. V. J. Phys. Chem. 1971, 75, 1216-1220.
(40) Tsyganenko, A. A.; Pozdnyakov, D. V.; Filimonov, V. N. J. Mol. Struct. 1975, 29,
299-318.
(41) Yamazoe, S.; Okumura, T.; Hitomi, Y.; Shishido, T.; Tanaka, T. J. Phys. Chem. C
2007, 111, 11077-11085.
(42) Busca, G.; Saussey, H.; Saur, O.; Lavalley, J. C.; Lorenzelli, V. Appl. Catal. 1985, 14,
245-260.
(43) Griffiths, D. W. L.; Hallam, H. E.; Thomas, W. J. T. Faraday Soc. 1968, 64, 3361-
3369.
(44) Spielbauer, D.; Mekhemer, G. A. H.; Zaki, M. I.; Knzinger, H. Catal. Lett. 1996, 40,
71-79.
(45) Jongsomjit, B.; Panpranot, J.; Goodwin Jr, J. G. J. Catal. 2001, 204, 98-109.
(46) Minnermann, M.; Grossmann, H. K.; Pokhrel, S.; Thiel, K.; Hagelin-Weaver, H.;
Bumer, M.; Mdler, L. Catal. Today 2013, 214, 90-99.
(47) Porto, S. P. S.; Fleury, P. A.; Damen, T. C. Phys. Rev. 1967, 154, 522-526.
(48) Bronkema, J. L.; Leo, D. C.; Bell, A. T. J. Phys. Chem. C 2007, 111, 14530-14540.
(49) Parker, J. C.; Siegel, R. W. J. Mater. Res. 1990, 5, 1246-1252.
(50) Shi, L.; Tin, K.-C.; Wong, N.-B. J. Mater. Sci. 1999, 34, 3367-3374.
(51) Manriquez, M. E.; Picquart, M.; Bokhimi, X.; Lpez, T.; Quintana, P.; Coronado, J.
M. J. Nanosci. Nanotechno. 2008, 8, 6623-6629.
(52) Gajovi, A.; Furi, K.; Musi, S.; Djerdj, I.; Tonejc, A.; Tonejc, A. M.; Su, D.;
Schlgl, R. J. Am. Ceram. Soc. 2006, 89, 2196-2205.
(53) Baraton, M. I.; Busca, G.; Prieto, M. C.; Ricchiardi, G.; Escribano, V. S. J. Solid State
Chem. 1994, 112, 9-14.
(54) Jongsomjit, B.; Sakdamnuson, C.; Goodwin, J., Jr.; Praserthdam, P. Catal. Lett. 2004,
94, 209-215.
(56) Iwasaki, M.; Hara, M.; Kawada, H.; Tada, H.; Ito, S. J. Colloid Interface Sci. 2000,
224, 202-204.
(57) Matralis, H.; Papadopoulou, C.; Lycourghiotis, A. Appl. Catal. A: Gen. 1994, 116,
221-236.
144 Chapter 4
(58) Okamoto, Y.; Nagata, K.; Adachi, T.; Imanaka, T.; Inamura, K.; Takyu, T. J. Phys.
Chem. 1991, 95, 310-319.
(59) Turr, N.; Acua, A.; Schimmller, B.; Mayr-Schmlzer, B.; Mania, P.; Hermans, I.
Top. Catal. 2011, 54, 737-745.
(60) Brown, R.; Cooper, M. E.; Whan, D. A. Appl. Catal. 1982, 3, 177-186.
(61) Ernst, B.; Libs, S.; Chaumette, P.; Kiennemann, A. Appl. Catal. A: Gen. 1999, 186,
145-168.
(62) Wang, W.-J.; Chen, Y.-W. Appl. Catal. 1991, 77, 223-233.
(63) Li, J.; Lu, G.; Wu, G.; Mao, D.; Guo, Y.; Wang, Y.; Guo, Y. Catal. Sci. Technol.
2014, 4, 1268-1275.
(64) Teoh, W. Y.; Setiawan, R.; Mdler, L.; Grunwaldt, J.-D.; Amal, R.; Pratsinis, S. E.
Chem. Mater. 2008, 20, 4069-4079.
(65) Jimnez, V. M.; Espins, J. P.; Gonzlez-Elipe, A. R. Surf. Interface Anal. 1998, 26,
62-71.
(66) Vo, M.; Borgmann, D.; Wedler, G. J. Catal. 2002, 212, 10-21.
(67) Coulter, K. E.; Sault, A. G. J. Catal. 1995, 154, 56-64.
(68) Myint, M.; Yan, B.; Wan, J.; Zhao, S.; Chen, J. G. J. Catal. 2016,
DOI:10.1016/j.jcat.2016.02.004,
(70) Hu, B.; Bean Getsoian, A.; Schweitzer, N. M.; Das, U.; Kim, H.; Niklas, J.;
Poluektov, O.; Curtiss, L. A.; Stair, P. C.; Miller, J. T.; Hock, A. S. J. Catal. 2015,
322, 24-37.
(71) Li, W.; Yu, S. Y.; Meitzner, G. D.; Iglesia, E. J. Phys. Chem. B 2001, 105, 1176-1184.
(72) Zhu, H.; Rosenfeld, D. C.; Anjum, D. H.; Sangaru, S. S.; Saih, Y.; Ould-Chikh, S.;
Basset, J.-M. J. Catal. 2015, 329, 291-306.
(73) Heracleous, E.; Lee, A. F.; Wilson, K.; Lemonidou, A. A. J. Catal. 2005, 231, 159-
171.
145
Chapter 5
5. Intimate Cobalt-Silica Interaction: an Important

Parameter for Oxidative Dehydrogenation of
Ethane
Abstract
The influence of cobalt loading on the catalytic performance of SiO2-supported catalysts was
investigated for the oxidative dehydrogenation of ethane (ODHE) with CO2 in a plug flow
reactor at 700 C. A maximum ethene yield of ~39% was obtained at 0.75 wt% cobalt loading
with a CO2 to C2H6 ratio of 2.5 and total flow rate of 15 mL/min. Both fresh and spent
catalysts were XRD amorphous indicating high cobalt dispersion and resistance towards
sintering. In contrast to higher cobalt loaded on silica catalysts, the catalyst with optimal
cobalt loading of 0.75 wt% showed no significant H2 consumption during TPR hinting to
strong binding between Co and Si in the form of cobalt silicate. UV-Vis and XPS indicated
that this mixed oxide where Co2+ is tetrahedrally coordinated in a SiO2 matrix was present
over the whole range of investigated cobalt loadings (0.1-4.5 wt%). This structure seems to be
fully developed at a cobalt loading of around 0.75 wt%. Raman spectra showed no significant
changes with increasing cobalt loading indicating that not all silanol groups were utilized for
anchoring the cobalt. A higher cobalt loadings (> 1.7 wt%), low concentrations of cobalt
oxide were detected. Coke deposition in the form of D- and G-type carbon increased with
higher cobalt loading as evidenced by TGA and Raman spectroscopy. The optimum catalyst
showed nearly stable activity with ethane conversion of ~37% and ethene selectivity ~84%
during 10 h of reaction.
Part of it is be submitted to ACS Catalysis, In preparation.

146 Chapter 5
5.1 Introduction
Recent surge in shale gas reserves consisting predominantly of methane and in lower amounts
of ethane and propane1 has attracted significant attention in both industry and academia. This
originates from the need of finding alternatives to crude oil, which reserves are finite, to
produce industrially valuable chemicals such as ethylene and propylene.2,3 In addition, current
industrial processes such as steam cracking and fluidized catalytic cracking that produce
olefins are energy-intensive and suffer from coke formation and deposition.4 Therefore, more
efficient processes that are driven by discoveries of novel catalysts are necessary. Oxidative
conversion processes are attractive primarily due to their low operating temperatures in
comparison to thermal processes.3 Achieving high product selectivity during oxidative
dehydrogenation of ethane (ODHE) has been an issue mainly due to the strong oxidizing
property of O2. Therefore, use of a milder oxidant such as CO2 has garnered intense interests.5
Moreover, global climate change offers additional impetus to find methods converting CO2 to
useful chemicals rather than their underground sequestration.
In ODHE reaction, sintering and coke deposition are the major problems that have
crippled the performance of some of the well investigated chromium6,7- and gallium8,9-based
catalysts. Among these, chromium-based catalysts were found to show superior performance
with ethane conversion and ethene selectivity values ranging between 30-60% and 60-90%,
respectively. Redox behavior of these catalysts was reported to be critical during the
reaction.10,11 In a different system i.e. Ca-doped ThO2, oxide ion vacancies were identified as
active sites for conversion of ethane.12 These vacancy sites were created by Ca and their
number could be manipulated by varying the Ca amount. The highest ethene selectivity of
97% and ethane conversion of 46% was obtained over Th0.75Ca0.25O2 at 725 C.
In our previous investigation13, where we investigated a series of single and mixed
oxides supported cobalt catalysts containing 4.5 wt% Co for ODHE with CO2, SiO2-supported
catalysts performed better affording an ethene yield of ~34%. The strong interaction between
cobalt and silica resulting in the formation of cobalt silicate, where likely tetrahedrally
coordinated Co2+ is anchored to SiO2 matrix, was suggested to be crucial for realizing good
catalytic performance of this catalyst. Generally, in wet-methods, synthesis of cobalt silicate
requires multiple steps, relatively high calcination temperatures ( 1000 C) and longer
duration.14 Interestingly, such a phase, which was very well dispersed, was produced in single
step during flame spray pyrolysis (FSP) synthesis.13 This synthesis method offers flexibility
and thus such structures can be further tuned and produced rapidly in large scale.15 Cobalt
147
silicate formation and their amount in the catalyst highly depend on the cobalt content.
Therefore, an in-depth investigation to develop further understanding of the interrelationship
between catalytic performance and cobalt silicate structure is needed.
In this work, we have produced the SiO2-supported catalysts with cobalt loadings
varying between 0.1 to 4.5 wt% and investigated their ODHE performance. The
physicochemical properties of the catalysts were analyzed using several characterization
techniques such as XRD, BET, H2-TPR, HR-TEM, XPS, Raman- and UV-Vis spectroscopy.
The crucial factors affecting the catalytic activity are identified and discussed with special
focus on cobalt silicate structure.
5.2 Experimental

Flame spray pyrolysis (FSP) technique was used for all the catalysts production. Precursor
solutions were prepared by mixing appropriate amounts of metal-precursors, cobalt (III)
acetylacetonate (98%) and hexamethyldisiloxane ( 98%), in toluene (Sigma Aldrich) setting
the total metal molar concentration to 0.45 M. During flame synthesis, a syringe pump
(Lambda, VIT-FIT) was used to pump the precursor solution at 5 mL/min into the flame
reactor nozzle, which was dispersed into fine droplets by co-flowed O2 of 5 L/min (> 99.9%
PanGas) while maintaining the pressure drop at 2 bar at the nozzle-tip. The resultant fine
spray was immediately ignited by a pilot flame (CH4/O2 of 1/2 L/min,> 99.9% PanGas). All
gas flow rates were regulated by mass flow controllers (Bronkhorst). A vacuum pump (Busch,
Seco SV 1040C) was utilized to collect the product powders from the combustion on glass
fiber filter (ALBET GF 6, 25.7 cm in diameter).

X-ray diffractograms were recorded over a 2 range of 10-70 with a step size of 0.019 and
0.1 sec/step on a Bruker D8 Advance diffractometer (40 kV, 20 mA, Cu K). A
Micromeritics Tristar instrument was used for nitrogen adsorption-desorption measurements
and the specific surface area (SSA) of both fresh and spent catalyst powders was determined
by applying the Brunauer-Emmett-Teller (BET) method at 77 K. Prior to the measurements,
all samples were kept at 150 C in N2-atmosphere for 1 h. H2-temperature programmed
reduction (TPR) was carried out using a Micromeritics Autochem 2920 instrument. For all
measurements, a sample mass of ca. 100 mg, a temperature ramp rate of 10 C/min and total
gas flow rate of 20 mL/min was applied. During H2-TPR measurements, the samples were
148 Chapter 5
heated from room temperature to 200 C in helium and held at this temperature for 2 h. After
cooling down to 50 C, the gas flow was switched to 5%H2/Ar then the temperature was
ramped to 840 C. The H2-consumption was monitored by a mass spectrometer (m/z = 2).
UV-Vis diffuse reflectance spectra (Cary 500 UV-Vis-NIR spectrophotometer) were
recorded from 300 to 800 nm for the fresh cobalt loaded catalysts. The samples were diluted
with BaSO4 keeping constant the mass ratio of the sample and BaSO4 and spectra were
measured against a BaSO4 background. Raman (Renishaw InVia Raman microscope) spectra
of fresh catalysts were obtained using a 785 nm laser with 50% of the laser power (500 mW).
For deposited coke analysis, a 514 nm laser with 1% of the laser power (24 mW) was used. A
thermogravimetric analyzer (TGA/SDTA851e, LF/1100 C, Mettler Toledo AG) was used to
quantify the amount of coke in the spent catalyst. In all measurements, approximately 20 mg
of powder was taken, which was heated from 30 to 950 C (10 C/min) while following the
reactant gas mixture (CO2, 40 mL/min and Ar, 10 mL/min) over it. High resolution
transmission electron microscopy (HR-TEM) was performed using a Tecnai F30 ST (FEI,
FEG, operated at 300 kV) microscope. Electron-dispersive X-ray spectroscopy (EDXS)
measurements were taken by an EDX spectrometer (Gemini system, EDAX), mounted on the
electron column right above the sample that allowed recording the spectra of selected spot.
X-ray photoelectron spectroscopy (XPS) measurements were performed using a VG
ESCALAB 220iXL spectrometer (Thermo Fisher Scientific) with focused monochromatic Al
K source (spot size: 500 m, power: 150 W). The pressure in the analysis chamber was
approximately 2 109 mbar. The spectrometer was calibrated on a clean silver surface by
measuring the Ag 3d5/2 peak at a binding energy (BE) of 368.25 eV with a full-width-at-half-
maximum (FWHM) of 0.78 eV. All the spectra were recorded using pass energy of 30 eV in
steps of 50 meV and a dwell time of 50 ms. The powder samples were placed on an adhesive
copper tape, which was mounted on a stainless steel holder that was later transferred to the
XPS spectrometer chamber and kept under vacuum. A low energy electron gun (Flood gun)
was used to compensate the charging effect generated from the nonconductive powder. The
calibration of the peaks binding energy position was applied by aligning the SiO2 component
in the Si2p core level peak to 104 eV.
5.2.3 Catalyst testing

A tubular continuous flow reactor with 4 mm inner diameter that was contracted to 2 mm
after the catalyst bed was utilized for catalyst testing under atmospheric pressure. The catalyst
powders were pressed into pellets, crushed and sieved and the desired size fraction (i.e. 80-
149
140 m) were collected. A catalyst bed was prepared from mixtures of 150 mg of catalyst
(80-140 m) and 50 mg SiC (160-190 m), which was fixed at the center using quartz wool.
The tubular reactor is placed in a temperature-controlled electrically heated furnace, which is
ramped to the desired reaction temperature (typically 700 C at 10 C/min) during the
catalytic reaction. The standard reaction conditions applied were: 700 C, reactant gas ratio
CO2/C2H6 = 2.5 and a total flow rate of reactant gases (CO2, 99.95 % PanGas and C2H6,
99.5 % PanGas) of 15 mL/min, which is fixed if not otherwise stated. The catalyst bed
temperature was monitored using two thermocouples at its inlet and outlet, which showed,
generally the difference of < 4 C. The effluent gases were analyzed using online gas
chromatography (HP-PLOT Q column 30 m long, 0.32 mm in diameter and 0.2 m film
thickness fitted with TCD and FID, Agilent Technologies). All the catalytic test results are
presented after 2 h reaction, if not otherwise noted. The C2H6 conversion, C2H4 selectivity,
C2H4 yield and CO2 conversion are calculated as follows:
C2 H6 conversion (%) = 100 (5.1)
C2 H4 selectivity (%) = 100 (5.2)

C2 H4 yield (%) = (5.3)
100
moles of CO2 consumed

CO2 conversion (%) = 100 (5.4)
moles of CO2 in the feed

5.3.1 Structural and chemical properties of catalysts
Figure 5.1a shows the selected XRD patterns of some SiO2-supported cobalt catalysts.
All the fresh catalysts showed XRD patterns similar to amorphous SiO2 regardless of their
cobalt content, which was present in low loadings and at high dispersion. Moreover, rapid
quenching of particles during the FSP synthesis prevented their further growth. Similar XRD
patterns were observed over the spent catalysts indicating that no significant phase change
occurred during the reaction (Fig. B.1, Appendix B). Specific surface areas (SSA) of all fresh
150 Chapter 5
as well as spent catalysts are presented in Figure 5.1b. SSAs of all the fresh catalysts were
around 300 m2/g. Slight increase in SSA with increasing cobalt loadings indicates that cobalt
acts as a surface area promoter. Decrease in the catalyst surface area after the reaction
indicates that sintering had occurred but does not result in any XRD-visible phase change.
600
a b
BET surface area, m2/g

fresh
450
Intensity, a.u.
%Co/SiO2
spent
4.5% 300
2.0%
0.75%
150
0%
0
10 20 30 40 50 60 70 0 1 2 3 4 5
2, degree Cobalt content in SiO2, wt%
Figure 5.1: (a) XRD patterns and (b) BET surface area of selected SiO2-supported Co
catalysts.
Figure 5.2 shows the UV-Vis spectra of various SiO2-supported cobalt catalysts. Bare
SiO2 showed no features in the measured range, whereas flame-made pure CoOx exhibited
bands around 400 and 710 nm (not shown) close to the reported values.16 These features
observed in Co/SiO2 catalysts must arise from either CoOx or are due to interaction between
Co and Si. Addition of only 0.1 wt% of cobalt gave rise to two distinct signals around 600 and
650 nm. Another band around 520 nm was observed when the cobalt content was increased to
0.3 wt%. This band became more pronounced with increasing cobalt loading and the crevice
between the bands around 600 and 650 nm gradually disappeared. These three bands
characterize the existence of tetrahedrally coordinated Co2+ in the SiO2 matrix.17 Since, the
features around 430 and 700 nm belonging to CoOx species16,18 were absent, it can be inferred
that the catalyst mostly contains Co-Si (cobalt silicate) species19. At higher cobalt loadings,
the spectra of 2 and 4.5%Co/SiO2 showed (slight) broad features at higher wavelength, which
may be attributed to the presence of small amounts of CoO.
151
wt%Co wt%Co
4.5
0.75
F (R), a.u.
2.0
0.5
1.7
0.3
0.1 1.3
bare SiO2 1.0
500 600 700 800 500 600 700 800
Wavelength, nm Wavelength, nm
Figure 5.2: UV-Vis measurements over low cobalt loaded SiO2 catalysts.
In an attempt to get further information from UV-Vis measurements, each spectrum

was treated separately and the signals were deconvoluted after subtracting the baseline. The
results are presented in Figure B.2. Three distinct signals (designated as I, II & III) were
observed upon deconvolution except in the case of the lowest cobalt-loaded catalyst (0.1
wt%). However, a very small peak is visible in the main spectrum. The percentage area of
each deconvoluted signal was calculated using the area contributed from the three signals to
the total area (Table B.1). Generally, an increasing trend in the percentage of peak area I and
III was observed while that of peak area II decreased with increasing cobalt loadings
(neglecting values from 0.1% loading). Similar observations were also reported by Ddeek
and Wichterlova20 for their Co2+ exchanged pentasil-containing zeolites. The assignment of
the deconvoluted spectra was based on their location in the zeolite with spectral contribution
from the lowest and the highest wavenumber increasing, whereas medium wavenumber
decreasing with increasing cobalt content.
152 Chapter 5
H2-MS signal intensity, a.u. 4.5%Co/SiO2
2%Co/SiO2
0.75%Co/SiO2
200 400 600 800

Temperature, C
Figure 5.3: H2-TPR over selected cobalt-loaded SiO2 catalysts
Figure 5.3 shows the reduction behavior of selected cobalt-loaded SiO2 catalysts.
Virtually no reduction peak was observed for the 0.75%Co/SiO2 catalyst whereas broad
reduction profiles were observed over 2 and 4.5%Co/SiO2 catalysts both around 780 C.
Moreover, the catalyst containing 4.5 wt% cobalt exhibited a shoulder around 810 C,
indicating the heterogeneity in the interactions between cobalt and silica. The high
temperature reduction indicates that cobalt is strongly bound to the SiO2 support likely at
lower oxidation state.19 A small shift of the reduction peak towards low temperature when
cobalt loading was increased to 4.5 wt% indicates the weakening of the cobalt support
interaction. This shift is particularly remarkable because higher loading of the reducible
species normally leads to a shift to higher temperature.21 Strong bond between the cobalt
species and SiO2 or cobalt species buried in the bulk of SiO2 and therefore inaccessible for the
reactants can be the reason for the absence of the reduction peak over the 0.75%Co/SiO2. The
latter reason can be ruled out based on the spectroscopic and catalytic results. It can be
speculated that the interaction might be even stronger at cobalt loading lower than 0.75 wt%
and therefore the CoOx species cannot be reduced under these conditions. In microscopy
images (Fig. B.3), bright spots consisting of CoOx species and/or cobalt silicate were present
in 4.5%Co/SiO2 catalysts, whereas in the 0.75wt%Co-catalyst no such spots were observed.
Since both UV-Vis and H2-TPR measurements showed the presence of cobalt silicate in
0.75%Co/SiO2, it can be inferred that they are small and highly dispersed and are below the
detection limit of the electron microscopy instrument.
153
Co 2p3/2
Co 2p1/2
satellite satellite
Counts/s, a.u.
wt%Co
4.5%
2.0%
0.75%
810 800 790 780 770 760

Binding energy, eV
Figure 5.4: XPS spectra of selected cobalt-loaded SiO2 catalysts.
In Figure 5.4, XPS spectra of some cobalt-loaded SiO2 catalysts are presented. The
Co 2p spectrum shows the characteristic doublet peaks at 781.8 and 797.7 eV corresponding
to 2p3/2 and 2p1/2, respectively, arising from spin-orbit coupling.22 Each of them also had its
shake-up satellite peak at binding energies of about 787 and 802.8 eV, respectively. The
difference in the binding energies of Co 2p3/2 and Co 2p1/2 main peaks is about 15.8 eV.
Whereas a difference of about 5.2 eV is obtained from the main and its shake-up satellite
peaks. Generally, a difference of about 15 eV in the former case and 9.5 eV (CoO) in the
latter case has been reported.23,24 These differences caused by the shift of the peak positions
can be related to the formation of mixed Co-Si oxide (cobalt silicate). The binding energy of
the Co 2p3/2 peak for cobalt silicate has been reported at about 781.3 eV, which is close to the
observed value (781.8 eV).25 Slight broadening of the Co 2p3/2 peak seemed to occur with
increase in cobalt content, an indication of possible CoO formation. Moreover, the presence
of satellite peaks indicates that the Co(II), likely in tetrahedral state (e.g. Co2SiO4).26,27 These
154 Chapter 5
results further confirm the presence of cobalt silicate, where Co2+ ions are likely tetrahedrally
coordinated with SiO2, on the surface of catalysts and suggest the presence of CoO (in low
concentration) at high cobalt loadings which is in agreement with UV-Vis results (Fig. 5.2)
where slight broadening of the band was observed with increasing cobalt loadings ( 2 wt%).
493
603
976
wt%Co
Raman Intensity, a.u.
679
795
10.0%
4.5%
2.0%
1.3%
0.75%
0.3%
0%
200 400 600 800 1000 1200

Wavenumber, cm-1
Figure 5.5: Selected Raman spectra of different cobalt-loaded SiO2 catalysts (785 nm, laser
power: 50%, 10%Co/SiO2 spectra for comparison).
Figure 5.5 shows selected Raman spectra of various cobalt-loaded SiO2 catalysts. High
laser power (50% of 24 mW) was used to detect the siloxane rings, silanols and the possible
change that may have occurred due to cobalt addition. The two bands at 493 and 603 cm-1 can
be assigned to the 4- and 3-membered siloxane rings, respectively.28 The bands at 795 and
976 cm-1 are primarily related to a Si-(OH) stretching vibration or silanol groups.29 No
155
observable changes in these bands were detected upon increasing the cobalt loadings until
about 1.7 wt% (Fig. B.5). A small yet distinct hump is formed at around 679 cm-1 when the
cobalt loading was increased to 2 wt%, which became prominent at the highest loading
(10%Co/SiO2 showed for comparison). This new band at 679 cm-1 matches with one of the
bands from Co3O4.30 UV-Vis showed the three characteristic humps that are attributed to Co2+
tetrahedrally coordinated to the SiO2 matrix forming cobalt silicate mixed oxide, which was
further confirmed by XPS. Therefore, at lower loadings it can be expected that virtually all
cobalt is used for the formation of this mixed oxide.
Recently, Hu et al.31 reported that the band belonging to the trisiloxane rings (~603
cm-1) disappears when cobalt was grafted/anchored to these rings suggesting them as the main
anchoring sites. Only half of the calculated amount of Co (4.3 wt%) was found to be anchored
to the silica (300 m2/g) though sufficient amount of cobalt (5 wt%) was introduced during the
synthesis. It was postulated that not all the anchoring sites have geometries conducive to
participate in anchoring and that the cobalt grafting is self-limiting owing to the hindrance or
repulsion created by the positive charged Co(NH3)63+ ions in the solution. Though the latter
postulation doesnt apply in our case since we have gas-phase synthesis, the former one can
be valid. However, during rapid gas-phase synthesis, the interaction between the ions (Co and
SiOH) is highly affected by the residence time and the environment in the vapor phase. All
the produced catalysts were light blue (bluish) in color, an indication of cobalt silicate
formation, generally obtained during high calcination temperatures.32 Presence of the band
belonging to the siloxane rings (603 cm-1) indicates their partial utilization for anchoring and
the formation and growth of CoOx (679 cm-1) at Co loading 2wt% indicates the existence of
kinetically driven limitation on anchoring.
Raman spectra of spent catalysts were obtained to examine the degree and type of
coke deposition (Fig. B.4). All the catalysts showed two major peaks at around 1350 and 1650
cm-1 indicating the formation of disordered (D)- and graphitic (G)-like carbon on the surface
of the catalysts.33 Over lower cobalt loaded catalysts, these carbon structures were not well
developed. Prominent and high intensity peaks belonging to these carbon structures started to
form with increasing amount of cobalt in the catalysts. Overall, these results confirm the
deposition of coke during the reaction and identify their nature or type. Coke deposition was
also confirmed by TGA analysis of spent catalysts. An increasing weight loss due to coke was
evident with increasing cobalt loading (Table B.1). This indicates that coke deposits mainly
on CoOx or Co-Si mixed species and that these structures also promote cracking in addition to
the ODHE reaction.
156 Chapter 5
5.3.2 Catalytic activity
Figure 5.6 shows the effect of low cobalt loadings on ODHE reaction over SiO2
support. Ethane conversion increased with increasing cobalt loading until 0.75% reaching
about 46% and then decreased with further cobalt addition. However, only marginal increase
in the ethene selectivity was observed with a value of about 85% at the highest ethane
conversion (46%).
60
C2H4 selectivity 80
45
Conversion, yield, %
60
Selectivity,%
C2H6 conversion
30
40
C2H4 yield
15
20
CO2 conversion
0 0
0 1 2 3 4 5
Co content in SiO2, wt%
Figure 5.6: ODHE reaction over various SiO2-supported with different cobalt loadings. (Rxn
conditions: T = 700 C; CO2/C2H6 = 2.5; total flow rate = 15 mL/min; catalyst wt. = 150 mg;
2 h rxn time)
Low CO2 conversion of ~4% was obtained over 0.1%Co/SiO2 catalyst, which
increased with increasing the cobalt loading reaching a value of around 8% at 0.5 wt% and
then remained virtually constant upon further increase (Fig. B.6). Different roles of CO2 in
reverse water gas shift (RWGS) reaction, reverse boudouard (removing coke) and as a
promoter of redox capacity of the active materials have been suggested.34-36 Since the H2-TPR
results revealed that these catalysts are difficult to reduce under the applied reaction
conditions (700 C), other pathways seem plausible during the reaction. Hu et al.31 has
157
reported similar non-reducible Co/SiO2 catalyst for propane dehydrogenation reaction where
they proposed the occurrence of non-redox pathway during the reaction, where heterolytic
cleavage of the C-H bond is speculated to be the key step, which is catalyzed by tetrahedral
Co2+ present in SiO2 matrix. UV-Vis results show the presence of such a structure (cobalt
silicate) in all the Co/SiO2 catalysts, however, the structure seemed to be fully developed only
at cobalt loadings higher than 0.5 wt%. Therefore, an initial increase in the catalytic activity
(up to 0.75 wt% loading) can be directly related to the formation of the cobalt silicate
structure. The decreasing activity of the catalyst with increasing cobalt loading can be
attributed to the formation of cobalt oxide species that are not bound to the SiO2 matrix. These
species are prone to sintering resulting in the formation of less active CoOx clusters that are
detrimental for the activity.37,38 Moreover, such structures seemed to promote coke deposition
as observed from TGA analysis (Table B.1). However, no influence on the methane
selectivity (~ 1% selectivity) was observed with varying cobalt content. Based on the
observed results, it can be inferred that the ODHE occurs likely via dehydrogenation in which
CO2 assists/promotes primarily by consuming H2 and partly by removing coke from the
surface.
Similar catalytic behavior of cobalt at lower loadings during O2-assisted oxidative
dehydrogenation of ethane has been previously reported by Elkabouss et al.18,39 The cobalt
exchanged hydroxyapatite catalyst showed the highest ethene yield of ~17% (ethene
selectivity of ~57%) at cobalt loading of about 0.9 wt%. The reported low selectivity can be
attributed to the use of oxygen, which is a strong oxidant that oxidizes the desired products.
The specific dehydrogenation activity was found to highly depend on the presence of isolated
Co2+ sites and was highest when these sites were prevalent. At higher loadings (e.g. > 1.7
wt%), Co2+-O-Co3+ ensembles are created resulting in lower activity due to the formation of
Co3O4. However, this oxide was absent on SiO2-supported catalysts, which is likely due to
their high dispersion and rapid quenching. Another possibility is that their sizes were below
the detection limit of the instrument. Nevertheless, UV-Vis, XPS and H2-TPR results showed
the presence of isolated Co2+, which seemed to dominate at cobalt loadings lower than 1 wt%.
Correlating the catalytic activity and the UV-Vis spectra, it can be inferred that the formation
and optimal amount of tetrahedrally coordinated Co2+ in the SiO2 matrix is essential for high
ODHE activity.
Figure 5.7 shows the selectivity versus conversion results of 0.3, 0.75 and 4.5 wt%
cobalt containing catalysts. Both the 0.3 and 0.75% cobalt-loaded SiO2 catalysts showed
158 Chapter 5
virtually similar ethene selectivity (around 84%) regardless of the changes in the ethane
conversion. In contrast, over 4.5%Co/SiO2, a continuous decrease in ethene selectivity was
observed at conversions > 25%. Such behavior of the cobalt-loaded catalysts (< 1 wt%) hints
to the absence of secondary reactions that consume ethene, whereas the existence of such
consecutive reactions is clearly evident at higher loadings. These observations can be directly
related to the presence of a small concentration of CoOx species, which are likely to form
metallic clusters that are known to promote the undesired reactions at higher loadings.
Presence of such oxides at higher cobalt loadings (likely > 0.75 wt% cobalt) has been
suggested by XPS measurement.
90
C2H4 selectivity, %
80
70
60
0.3%Co/SiO2
0.75%Co/SiO2
4.5%Co/SiO2
50
20 30 40 50
C2H6 conversion, %
Figure 5.7: Selectivity versus conversion of various Co-loaded SiO2 catalysts at contact times
of 0.1-1.2 g.s/mL. Reaction conditions: T = 700 C; CO2/C2H6 = 2.5; total flow rate = 15
mL/min; catalyst wt. = 25-300 mg; 2 h time-on-stream.
159
100
Conversion, selectivity, % C2H4 selectivity
80
60
C2H6 conversion
40
C2H4 yield
20
CO2 conversion
0
0 100 200 300 400 500 600
Time, min
Figure 5.8: Stability test over 0.75%Co-loaded SiO2 catalysts. (Rxn conditions: T = 700 C;
CO2/C2H6 = 2.5; total flow rate = 15 mL/min; catalyst wt. = 150 mg).
Figure 5.8 shows a time-on-stream test of the 0.75%Co/SiO2 catalyst under the
standard reaction conditions. The catalyst showed fairly stable ethene selectivity of about 84%
though ethane conversion decreased from ~46% to 37% during 10 h on stream. CO2
conversion of around 7% was obtained during this extended reaction period. In addition to
ethene, methane and coke were also produced during the reaction. Since the methane
selectivity is about 1%, coke can be considered as the second major product of the reaction
with selectivity value of around 15%. The initial unsteady state activity manifested in low
ethene selectivity and slightly high ethane conversion can be attributed to the early
physicochemical changes of the catalyst that occurred at the initial stage of the reaction.
Decrease in ethane conversion with time indicates the occurrence of catalyst deactivation,
which could arise from structural changes (sintering) and to some extent coke deposition.
Moreover, during extended operation some cobalt species transformed to metallic cobalt
ensembles which reduce the catalytic activity and possibly increase the coke deposition.
However, these changes were not significant enough to be detected by XRD measurements.
160 Chapter 5
Nevertheless, from the observed performance it can be inferred that most of cobalt stays intact
in the SiO2 matrix and such structure (cobalt silicate) is essential in realizing good ODHE
performance.
5.4 Conclusions
Flame spray pyrolysis prepared SiO2-supported catalysts with highly dispersed cobalt
at loadings from 0.1 to 4.5wt% were structurally characterized and tested in the oxidative
dehydrogenation of ethane with CO2. UV-Vis, XPS and H2-TPR revealed the formation of the
cobalt silicate structure in all the catalysts. At higher cobalt loadings (> 2%), UV-Vis and
XPS suggest the formation of small amounts of CoOx. In the cobalt silicate structure, Co2+
ions are likely tetrahedrally coordinated and play a crucial role for the CO2-assisted oxidative
dehydrogenation of ethane. The optimal amount of tetrahedrally coordinated Co2+ in SiO2
seemed to be achieved at 0.75 wt% cobalt loading, which exhibited the highest ethene yield of
~39% (~ 85% ethene selectivity). This catalyst showed fairly stable activity during 10 h time-
on-stream. Coke deposition in the form of D- and G-type carbon was observed over all
catalysts and the amount increased with increasing cobalt loading as revealed by Raman
spectroscopy and thermogravimetry.
We thank Dr. Frank Krumeich (ETH) for STEM/EDXS investigations and the Electron
Microscopy Center of ETH Zurich (EMEZ) for providing the necessary infrastructure and Dr.
Mario El Kazzi for XPS measurements. Financial support by ETH Research Grant ETH-39-
12-2 is kindly acknowledged.
5.6 References
(1) Jacoby, H. D.; O'Sullivan, F. M.; Paltsev, S. Econ. Energy Environ. Policy 2012, 1,
37-51.
Chem. Rev. 2014, 114, 10613-10653.
161

(6) Mimura, N.; Takahara, I.; Inaba, M.; Okamoto, M.; Murata, K. Catal. Commun. 2002,
3, 257-262.
(7) Shi, X.; Ji, S.; Wang, K. Catal. Lett. 2008, 125, 331-339.
1998, 1025-1026.
2000, 196, 1-8.
(11) Asghari, E.; Haghighi, M.; Rahmani, F. J. Mol. Catal. A: Chem. 2016, 418419, 115-
124.
(12) Baidya, T.; Vegten, N.; Baiker, A. Top. Catal. 2011, 54, 881-887.
(13) Koirala, R.; Buechel, R.; Pratsinis, S. E.; Baiker, A. In preparation 2016,
(14) Sazonov, A.; Meven, M.; Hutanu, V.; Kaiser, V.; Heger, G.; Trots, D.; Merz, M. Acta
Crystallographica Section B 2008, 64, 661-668.
(15) Koirala, R.; Pratsinis, S. E.; Baiker, A. Chem. Soc. Rev. 2016,
DOI:10.1039/c5cs00011d,
(16) Brik, Y.; Kacimi, M.; Ziyad, M.; Bozon-Verduraz, F. J. Catal. 2001, 202, 118-128.
(17) Okamoto, Y.; Nagata, K.; Adachi, T.; Imanaka, T.; Inamura, K.; Takyu, T. J. Phys.
Chem. 1991, 95, 310-319.
(18) El Kabouss, K.; Kacimi, M.; Ziyad, M.; Ammar, S.; Bozon-Verduraz, F. J. Catal.
2004, 226, 16-24.
(19) Ernst, B.; Libs, S.; Chaumette, P.; Kiennemann, A. Appl. Catal. A: Gen. 1999, 186,
145-168.
(20) Ddeek, J.; Wichterlov, B. J. Phys. Chem. B 1999, 103, 1462-1476.
(21) Monti, D. A. M.; Baiker, A. J. Catal. 1983, 83, 323-335.
(22) Valero-Romero, M. J.; Sartipi, S.; Sun, X.; Rodriguez-Mirasol, J.; Cordero, T.;
Kapteijn, F.; Gascon, J. Catal. Sci. Technol. 2016, 6, 2633-2646.
(23) Chuang, T. J.; Brundle, C. R.; Rice, D. W. Surf. Sci. 1976, 59, 413-429.
(24) Mayer, B.; Uhlenbrock, S.; Neumann, M. J. Electron. Spectrosc. Relat. Phenom. 1996,
81, 63-67.
162 Chapter 5
(25) Ming, H.; Baker, B. G. Appl. Catal. A: Gen. 1995, 123, 23-36.
(26) Du, V. A.; Gross, S.; Schubert, U. Chem. Commun. 2010, 46, 8549-8551.
(27) Coulter, K. E.; Sault, A. G. J. Catal. 1995, 154, 56-64.
(28) Humbert, B.; Burneau, A.; Gallas, J. P.; Lavalley, J. C. J. Non-Cryst. Solids 1992, 143,
75-83.
(29) Astorino, E.; Peri, J. B.; Willey, R. J.; Busca, G. J. Catal. 1995, 157, 482-500.
(30) Jongsomjit, B.; Panpranot, J.; Goodwin Jr, J. G. J. Catal. 2001, 204, 98-109.
(31) Hu, B.; Bean Getsoian, A.; Schweitzer, N. M.; Das, U.; Kim, H.; Niklas, J.;
Poluektov, O.; Curtiss, L. A.; Stair, P. C.; Miller, J. T.; Hock, A. S. J. Catal. 2015,
322, 24-37.
(32) Khodakov, A. Y.; Chu, W.; Fongarland, P. Chem. Rev. 2007, 107, 1692-1744.
(34) Deng, S.; Li, H.; Li, S.; Zhang, Y. J. Mol. Catal. A: Chem. 2007, 268, 169-175.
(35) Myint, M.; Yan, B.; Wan, J.; Zhao, S.; Chen, J. G. J. Catal. 2016,
DOI:10.1016/j.jcat.2016.02.004,
(38) Ruckenstein, E.; Wang, H. Y. Catal. Lett. 2001, 73, 99-105.
(39) El Kabouss, K.; Kacimi, M.; Ziyad, M.; Ammar, S.; Ensuque, A.; Piquemal, J.-Y.;
Bozon-Verduraz, F. J. Mater. Chem. 2006, 16, 2453-2463.
163
Chapter 6
6. Outlook and Research Recommendations
In this thesis various catalysts were produced using flame spray pyrolysis (FSP) technique and
tested for their potential in converting cheaply available natural gas into high value chemicals.
FSP technique offers precise control over particle properties such as size, surface area,
crystallinity and scalability which in turn improves the catalytic activity. 1 Through this
technique, catalysts can be produced rapidly in single step. These features are generally
difficult to achieve in classic wet-chemistry methods. In addition to these features, the
flexibility and the scalability of the FSP technique have marked an increase in its application
for the synthesis of several catalyst materials.2
Natural gas is an abundant source of energy that can potentially replace crude oil as an
alternative feedstock for the production of valuable chemicals (e.g. ethene and propene). With
its ever-increasing production stemming from the discovery of new reserves and technological
advancements that increase the extraction efficiency, significant attention has been directed
towards catalytic upgrading of this cheaply available resource.3 Oxidative conversion of the
components of the natural gas (e.g. methane and ethane) to ethene is a promising approach
that can overcome the inefficiencies associated with the traditional cracking process (e.g.
requires high temperatures and frequent regeneration due to coking)4 used for its production.5
This thesis aims to move toward the successful realization of such a process through the
development of highly efficient novel materials that can catalyze the oxidative conversion of
methane and ethane.
Although Mn-Na2WO4/SiO2 catalyst has been investigated for a long time as a
potential catalyst for oxidative coupling of methane (OCM), conclusive evidence regarding
the active center for the reaction is still missing. This in turn restricts the rational design of
better catalysts. It is well known that the synthesis method significantly influences the catalyst
property and thus its activity. In this thesis, a novel flame-synthesis method, FSP, was applied
164 Chapter 6
to systematically investigate the effect of elemental composition and physical-chemical

characteristics on the activity of Mn-Na2WO4/SiO2. All the elements were found to be
essential to achieve better catalytic activity suggesting the existence of a multi-component
synergistic effect during the OCM reaction. The surface species related to Mn, W and Na,
seemed to drive the reaction. A two nozzle6 FSP synthesis technique can be utilized to
decouple the elemental interactions and study their role by assessing their performance.
However, major breakthroughs can be achieved only if real-time changes in surface/structure
of the catalyst and the product distributions are monitored using advanced characterization
techniques that are typically available at a synchrotron.7 By employing synchrotron vacuum
ultra-violet (VUV) photoionization combined with molecular-beam mass spectroscopy over
Li/MgO, the intermediate species (methyl radicals) that are likely to dictate the catalytic
activity were identified, thus, paving way for their application in the rational design of highly
active catalysts.
Catalytic production of ethene using ethane from natural gas is another topic of
immense industrial interest. Though oxidative conversion using O2 occurs at slightly lower
temperatures, due to the highly reactive nature of the oxidant (resulting in low selectivity), use
of a milder oxidant, e.g. CO2, has been suggested. Moreover, usage of this abundant
greenhouse gas to produce high value chemicals is always an ecological and economically
lucrative approach. Gallium-based catalysts (Ga/TiO2) have shown relatively high activity for
oxidative dehydrogenation of ethane (ODHE), however, continuous deactivation is observed
primarily due to coke deposition. Moreover, the ethane conversion and ethene selectivity
showed opposite trends during the reaction, indicating that the active sites react with both.
Therefore, it is essential to decrease/control these sites to minimize the undesired secondary
reactions that consumes ethene. Both phenomena are inter-related and highly depended on the
acid-base characteristics of the catalyst. Though some structural changes were observed after
the reaction, the ability of the catalyst to regain its activity after subsequent regeneration
cycles shows the key role of surface properties in the reaction. Therefore, future focus should
be directed towards controlling the acidity and basicity of the catalysts using FSP, as shown
by Hoxha et al.8 On the other hand, promoters can also be introduced to reduce the coke
deposition either by poisoning some reactive sites or facilitating the oxidation of coke from
the surface. Additionally, process modifications, such as introducing water vapor and/or low
partial pressure of oxygen during the reaction can alleviate coke deposition.
In this thesis, cobalt oxide-based catalysts were also investigated as alternatives to
gallium oxide-based ones which critically suffered from coke deposition during ODHE
165
reaction. Investigation of a wide range of cobalt-loaded single and mixed metal oxides
supported catalysts showed that SiO2, ZrO2 and TiO2-ZrO2 supports are favorable for the
reaction. Further catalytic investigation performed at different contact times revealed SiO 2 as
the best support. Cobalt-silicate phase in Co/SiO2 is speculated to play a crucial role in
achieving high activity during the reaction. Some coke deposition was also observed on the
catalysts surface after the reaction. Therefore, it would be interesting to introduce promoters
such as ceria, lanthanum oxide and manganese oxide that are well known coke suppressants.
In addition, most of these promoters are also active ODHE catalysts so additional benefit can
be harnessed from their addition.
In the later part of the study, influence of low cobalt loadings on SiO2 over ODHE
reaction was investigated. Ethane conversion increased with increasing the cobalt loadings
until 0.75 wt% and then decreased upon further increase. Highly dispersed cobalt species
particularly in the form of cobalt-silicate and/or those that are strongly embedded into SiO2
matrix were found to entail high activity. However, in situ/operando techniques such as X-ray
absorption spectroscopy that investigate the structural-chemical properties under catalytic
conditions can provide more fundamental insight into plausible descriptors of catalytic
activity. Moreover, since the anchored cobalt species in SiO2 matrix appear to be the active
center, flame synthesis conditions can be modified to obtain high concentration of such
centers. In comparison with Ga/TiO2 (chapter 3), cobalt-based catalysts showed better
performance for the ODHE reaction, particularly when supported on SiO2 (chapter 4&5).
Moreover, the latter catalysts exhibited lower coke deposition than Ga/TiO2, which was the
major cause for catalyst deactivation. Therefore, future efforts should be dedicated towards
the improvement of the ODHE performance of cobalt-based catalysts by gaining in-depth
understanding of structural-chemical-activity relationship through several investigations as
mentioned earlier.
It is truly remarkable how the development of highly efficient catalysts has facilitated
the industrial transition towards a more sustainable sector. However, more research efforts are
needed to provide crucial fundamental understanding of the catalytic processes at the atomic
level so as to engineer sophisticated high performance catalysts. In this regard, flame aerosol
process, particularly FSP, has garnered major attention from both academia and industries for
the synthesis of nano-materials for several applications including catalysis, due to its several
merits over wet-methods.9,10 Application of flame-made catalysts for ethene production
seemed promising, however, further optimization on both material properties and process
conditions are necessary to obtain the activity that is of industrial interest. Therefore,
166 Chapter 6
experimental research together with dedicated computational efforts studying the dynamics of
particle formation during the synthesis and in situ characterizations (e.g. XAS) of catalytic
materials can open gates towards novel and functional catalysts production.
6.1 References:
(1) Strobel, R.; Baiker, A.; Pratsinis, S. Adv. Powder Technol. 2006, 17, 457-480.
Chemical Reviews 2014, 114, 10613-10653.
(6) Bchel, R.; Pratsinis, S. E.; Baiker, A. Appl. Catal. B: Environ. 2011, 101, 682-689.
(7) Luo, L.; Tang, X.; Wang, W.; Wang, Y.; Sun, S.; Qi, F.; Huang, W. Scientific Reports
2013, 3, 1625.
(8) Hoxha, F.; Schimmoeller, B.; Cakl, Z.; Urakawa, A.; Mallat, T.; Pratsinis, S. E.;
Baiker, A. J. Catal. 2010, 271, 115-124.
(10) Thibaut, B. Platinum Met. Rev. 2011, 55, 149-151.
167
Appendix A
Supporting Information: Influence of Support in

CO2-assisted Cobalt-Catalyzed Oxidative
Dehydrogenation of Ethane
A.1 Physical characterization
Figure A.1 shows the isotherms of the selected cobalt-loaded catalysts. The isotherms show
that the catalysts were non-porous. The small hysteresis at high P/P0 is attributed to interstitial
voids.
SiO2
Al2O3
Adsorbed volume, a.u.
TiO2
ZrO2
SiO2-Al2O3
0.0 0.2 0.4 0.6 0.8 1.0

Pi/P0,-
Figure A.1: BET isotherms of selected supported 4.5%Co containing catalysts.
168 Appendix A
Table A.1: Binding energies of Co 2p3/2 and Co 2p1/2 for the selected catalysts.
4.5%Co-loaded Co2p3/2 Co2p1/2
SiO2 781.7 - 797.1 -
ZrO2 779.4 784.5 794.9 801.0
TiO2-ZrO2 779.7 784.9 795.4 801.6
Table A.1 shows the binding energies (BE) of optimal performing 4.5 wt% cobalt loaded
catalysts. Both ZrO2- and TiO2-ZrO2-supported cobalt catalysts showed low BE values
compared to SiO2-supported one at Co 2p peak positions. This indicates the presence of CoOx
species, likely Co3+, over the former catalysts. This slightly high BE of SiO2-supported
catalyst indicates the presence of Co (II), which could be from cobalt silicate or small amount
of CoO due to their close proximity.
A.2 Catalytic results

Rate of reaction, molC2H6/m2.s
60
45
30
15
0
O
2
O
2
lO
3
iO
rO
2
rO
2
lO
3
O
2
Zr
Si
-T
2
Ti
-Z
-Z
2
-A
A
O
2
O
2
O
2
O
2
Si
Si
Ti
Si
Figure A.2: Rate of C2H6 consumption over several supported catalysts.

169
Figure A.2 shows the rate of reaction of several supported catalysts. Rate of C 2H6 reaction
referred to unit specific surface area of catalysts under standard reaction conditions. The
values were calculated according to the following formula:
2 6
( )
2 .
molar flowrate26
=

The results show that ZrO2- and TiO2-ZrO2-supported catalysts exhibited the highest rate of
reaction. Note that the SiO2-supported catalyst exhibited the lowest reaction rate per unit
specific surface area (mol C2H6/m2.s).
5 5
SiO2 ZrO2
CH4 selectivity, %
CH4 selectivity, %
4 4
3 3
2 2
1 1
0 0
0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5
Contact time (W/F), g.s/mL Contact time (W/F), g.s/mL
5
TiO2-ZrO2
CH4 selectivity, %
0
0.0 0.5 1.0 1.5
Contact time (W/F), g.s/mL
Figure A.3: Methane selectivity of selected supported Co catalysts at different contact times.
170 Appendix A
Figure A.3 shows the methane selectivity of selected supported catalysts. Ethene and methane
selectivity calculation were based on the amount of the product formed and the ethane amount
consumed (eqn. 4.2 in chapter 4).
Methane selectivity was calculated according to:
mole of 4 formed
4 (%) = 100
2 6
The methane selectivity was generally low in all the supported catalysts (below 2%) except
Al2O3-, ZrO2- and TiO2-ZrO2-supported catalysts. Methane selectivity increased from ~1% to
~4% with increasing contact times. Among the best performing catalysts (Fig. A.3), the ZrO 2-
supported catalyst showed similar trend to Al2O3-supported catalyst, whereas the TiO2-ZrO2-
supported catalyst exhibited higher values, which only slightly increased with increasing
contact time.
30
RWGS reaction: CO2/H2 ratio 2.5
25
CO2 conversion, %
20 TiO2-ZrO2
15
ZrO2
10 SiO2
5
20 40 60 80 100 120
time, min
Figure A.4: Reverse water gas shift (RWGS) reaction over selected supported catalysts.
Reaction conditions: T = 700 C, catalyst = 150 mg, SiC = 50 mg, CO2/H2 = 2.5 and total
flow 15 mL/min, 2 h of reaction.
171
Figure A.4 shows the reverse water gas shift (RWGS) activity of selected best performing
catalysts (SiO2-, ZrO2- and TiO2-ZrO2-supported cobalt catalysts). The reactions were
performed under similar conditions used during the standard ODHE reaction. All the catalysts
showed RWGS activity, the highest one was observed over TiO2-ZrO2-supported catalysts
followed by ZrO2- and SiO2-supported ones.
30
contact time (W/F), g.s/mL
0.1
25 0.3
0.6
CO2 conversion, %
0.8
20 1.2
15
10
0
lO
3
O
2
rO
iO
lO
2
3
O
2
rO
2
O
2
Si
Zr
Ti
-T
2
2
-Z
-Z
-A
A
O
2
O
2
O
2
O
2
Si
Si
Ti
Si
Figure A.5: CO2 conversion of several supported catalysts at different contact times.
Figure A.5 shows the CO2 conversion over various supported cobalt catalysts at different
contact times (0.1-1.2 g.s/mL). In general, an increase in CO2 conversion with increasing
contact time was observed indicating the active role of CO2 in the ODHE reaction. The
observed high CO2 conversion over the ZrO2- and TiO2-ZrO2-supported catalysts can be
related to their high RWGS rate during the reaction (Fig. A.4).
172 Appendix A
45
TiO2-ZrO2
Ethane conversion, % 40
ZrO2
35
SiO2
30
25 TiO2
Al2O3
20
SiO2-ZrO2 SiO2-TiO2
SiO2-Al2O3
15
0.5 1.0 1.5
NH3 uptake, mol NH3/m2cat
SiO2-TiO2
SiO2-ZrO2
90
Ethene selectivity, %
SiO2
TiO2
Al2O3
75
SiO2-Al2O3
ZrO2
60
TiO2-ZrO2
0.5 1.0 1.5
NH3 uptake, mol NH3/m2cat

Figure A.6: Surface density of acidic sites versus activity and ethene selectivity of catalysts.
Figure A.6 shows the dependence of ethane conversion and ethene selectivity on the density
of acidic sites. The results show no clear trend between the activity and acidity. The TiO2- and
TiO2-ZrO2-supported catalyst had similar NH3 uptake per unit specific surface area but the
latter one showed higher ethane conversion. Moreover, though ZrO2- and SiO2-supported
catalyst had lower NH3 uptake compared with many others, the ethane conversion was higher.
Similarly, no clear trend was also observed for ethene selectivity. However, considering only
173
SiO2-, ZrO2- and TiO2-ZrO2-supported catalysts, it seems that with increasing density of acid
surface sites the ethane conversion increases while the ethene selectivity decreases. The
decrease in selectivity can be attributed to the higher reactivity of ethene compared to ethane.
In addition, acid density of the latter materials (ZrO2 and TiO2-ZrO2) seemed to promote also
the RWGS reaction (Fig. A.4). The TiO2-ZrO2-supported catalysts exhibited the highest
RWGS activity followed by ZrO2- and SiO2-supported ones.
Table A.2: Amount of NH3-desorbed per unit area, CH4 selectivity, CO2 conversion and coke
amount of various single and mixed oxide supported 4.5%Co catalysts.
4.5%Co- mol NH3/m2cat CH4 CO2 Coke,
loaded selectivity, % conversion, wt%
%
100-350 C 350-650 C total
SiO2 0.4 0.1 0.5 0.7 7.1 1.8
Al2O3 1.1 - 1.1 2.5 9.5 4.0
TiO2 1.6 0.2 1.8 0.8 8.1 0.1
ZrO2 0.9 - 0.9 3.7 17.0 0.6
SiO2-Al2O3 0.8 0.3 1.2* 0.8 7.2 1.9
SiO2-TiO2 1.6 ~0.0 1.6 1.4 4.9 0.6
SiO2-ZrO2 0.9 - 0.9 0.9 3.4 0.3
TiO2-ZrO2 1.4 0.3 1.7 2.6 22.3 2.8
*Additional 0.1 mol NH3/m2cat originate from strong acid sites (> 650 C, Fig. 4.4b)
Note: Both ethene and methane selectivity were calculated by using the amount of ethane consumed at the
denominator. Since the consumption of ethane is not limited to only ethene and methane (also converted to
coke), the resulted selectivity is lower. However, when only the products that were measured is used for the
calculation1,2 (equation below), ethene selectivity > 95% (balanced methane) is obtained.
n moles of C2 H4
C2 H4 selectivity (%) = 100
n moles of CH4 + n moles C2 H4
174 Appendix A
n is number of carbons.
Table A.2 shows the normalized NH3-desorbed per unit area, methane selectivity, CO2
conversion and deposited amount of coke of several supported catalysts. The results show that
the SiO2-supported catalyst exhibits the lowest density of acidic surface sites, whereas TiO2
shows the highest. Methane selectivity was highest over Al2O3-, ZrO2- and TiO2-ZrO2-
supported catalysts, whereas highest coke deposition was observed over Al2O3- and TiO2-
ZrO2-supported catalysts.
Table A.3 Oxidative dehydrogenation of ethane over selected supported cobalt catalysts using
oxygen as an oxidant.
4.5%Co-loaded C2H6 conversion,% C2H4 selectivity, COx selectivity,
%* %*
SiO2 6.2 27.5 72.5
ZrO2 7.2 34.8 65.2
TiO2-ZrO2 17.5 90.4 9.6
Note: Presented data are after 2 h of reaction. Reaction conditions: T = 500 C, catalyst = 150 mg, SiC = 50 mg,
total flow 15 mL/min, C2H6 = 6.4 mL/min, O2 = 3.2 mL/min and balanced helium.
*some trace amount of CH4 was also observed in the reaction.
Table A.3 shows the oxidative dehydrogenation of ethane over selected supported catalysts in
the presence of oxygen. Both SiO2- and ZrO2-supported cobalt catalysts showed low activity
compared to the TiO2-ZrO2-supported one. The latter catalysts performance was obtained
under non-optimized conditions. Therefore, further investigation can be performed to improve
the activity.
A.3 References
(1) Shi, X.; Ji, S.; Wang, K. Catal. Lett., 2008, 125, 331-339.
(2) Zhang, X.;Ye, Q.; Xu, B.; He, D. Catal. Lett., 2007, 117, 140-145.
175
Appendix B
Supporting Information: Intimate Cobalt-Silica

Interaction: An Important Parameter for Oxidative
Dehydrogenation of Ethane
B.1 Catalyst characterization
Figure B.1 shows the XRD patterns of selected spent catalysts. Absence of any reflections
belonging to cobalt oxide or its mixed oxides (cobalt silicate) in spent catalysts indicates that
the catalyst showed resistance towards sintering and/or were below the detection limit of the
instrument.
Spent Catalysts
Intensity, a.u.
wt%Co/SiO2
4.5%
2.0%
0.75%
0%
10 20 30 40 50 60 70
2, degree
Figure B.1: XRD patterns of some selected spent cobalt-loaded SiO2 catalysts.
176 Appendix B
Figure B.2: Deconvoluted UV-Vis spectra of different cobalt-loaded SiO2 catalysts.
In Figure B.2, deconvoluted UV-Vis spectra of several Co/SiO2 catalysts are presented. Three
deconvoluted signals (indicated as I, II & III) were observed for all catalyst, except for
0.1%Co/SiO2, where the fitting was difficult. These bands are characteristic for tetrahedrally
coordinated Co2+ ions anchored to the SiO2 matrix. The peak areas were calculated to estimate
their overall contributions, which are tabulated below (Table B.1). Both areas of peak I and III
increased with increasing cobalt content whereas peak II area decreased. Since, the
performance of 0.75%Co/SiO2 was superior compared to that of others, it can be inferred that
at this cobalt loading the optimal contribution from these peaks exist that is crucial for
realizing the high activity.
177
Table B.1: Percentages of peak areas (determined from peak deconvolution) of UV-Vis
spectra and the TGA analysis
wt% of cobalt Peak Area I, % Peak Area II, % Peak Area III, % Coke, wt%
SiO2
0.1 - 62.3 37.7 0.32
0.3 12.7 58.1 29.2 0.62
0.5 14.4 57.5 28.1 0.96
0.75 15.7 55.8 28.5 0.98
1.0 15.6 49.7 34.6 1.4
1.3 21.2 46.8 32.0 1.5
1.7 19.8 44.0 36.2 1.5
2.0 25.3 38.0 36.7 1.5
4.5 33.1 25.9 41.0 1.8
Figure B.3: HR-TEM images of 0.75 and 4.5% Co-loaded SiO2 catalysts.
Figure B.3 shows the HR-TEM images of 0.75 and 4.5% Co-loaded SiO2 catalysts. No bright
spots in the microscopy image of 0.75%Co/SiO2 catalyst was observed indicating very high
dispersion of CoOx species and/or Co-Si mixed oxide, whereas such spots were observed over
the catalysts containing 4.5 wt% cobalt.
178 Appendix B
G D
D G
4.5%
0.75%
Intensity, a.u.
0.5% 2.0%
1.7%
0.3%
1.3%
0.1%
1.0%
1000 1200 1400 1600 1800 2000 1000 1200 1400 1600 1800 2000
Raman Shift, cm-1 Raman Shift, cm-1

Figure B.4: Raman spectra of several spent Co-SiO2 catalysts.
Figure B.4 shows the Raman spectra of spent catalysts where two bands belonging to
disordered (D)-and graphitic (G)-type carbon was detected. The spectra were more developed
at higher cobalt loadings likely due to formation of some metallic cobalt ensembles that
promoted further coking. Spectra obtained from low cobalt-loaded catalysts (0.1-0.75 wt%)
were of very low intensity with lots of background, which was subtracted before presenting in
a figure. These results indicate that at lower cobalt loadings virtually all the cobalt is buried or
anchored to SiO2 minimizing coke deposition and as the loading is increased cobalt is
susceptible to migrate to the surface promoting coking. This can also be directly related to the
TGA results (Table B.1) which showed an increase in the amount of coke deposits with
increasing cobalt amount. Low coke deposition (0.32 wt%) was obtained at the lowest cobalt
loading (0.1 wt%), which increased with higher Co-loading. Coke amounts were virtually
same over the catalysts containing 1 to 2 wt% cobalt indicating that loadings variation in this
range had little effect on the coke deposition.
179
493
603 %Co/SiO2
679
795 976
10.0%
Raman Intensity, a.u.
4.5%
2.0%
1.7%
1.3%
1.0%
0.75%
0.5%
0.3%
0.1%
0%
200 400 600 800 1000 1200

Wavenumber, cm-1
Figure B.5: Raman spectra of all fresh Co/SiO2 catalysts.
In Figure B.5, Raman spectra of various cobalt-loaded SiO2 catalysts are presented. The bands
at 795 and 976 cm-1 are primarily related to a Si-(OH) stretching vibration or silanol groups.
The two bands at 493 and 603 cm-1 can be assigned to the 4- and 3-membered siloxane rings,
respectively. The band at 679 cm-1 is formed at higher cobalt loadings (> 2 %), which
intensity became more pronounced at the cobalt loading of 10 wt%. This band matches with
the bands of Co3O4 indicating the formation of CoOx species at higher cobalt loadings. Such
species may have already formed even at cobalt loadings of around 1 wt%, which may have
gone undetected due to instrument limitation. This assumption is based on the Raman (Fig.
B.4) and TGA results (Table B.1), where much developed carbon bands and the rather rapid
growth in the coke deposits were observed.
180 Appendix B
B.2 Catalytic results
12
CO2 conversion, %
0
0.0 0.5 1.0 1.5 2.0 4.5
Co content in SiO2, wt%
Figure B.6: CO2 conversion over several cobalt-loaded SiO2 catalysts.
Figure B.6 shows the CO2 conversion measured over several cobalt-loaded SiO2 catalysts.
Low conversion around 4% was obtained at the lowest cobalt loading (0.1%), which increased
with higher cobalt loading up to 0.5 wt% and then virtually remained the same. These results
show that the CO2 participate via reverse water gas shift and reverse boudouard reactions
(some of the well reported pathways).
181
A Curriculum Vitae
Rajesh Koirala
Born 17 May, 1984, Nepal
Education
2016 PhD in Mechanical and Process Engineering, Thesis Flame-made
Catalytic Materials for Alkene Production Particle Technology
Laboratory, ETH Zurich, Switzerland, Advisor: Prof. Sotiris E.
Pratsinis
2012 MSc Chemical Engineering, Thesis: Synthesis of highly durable and high
performing various metal-doped CaO-based nano-sorbents to capture
CO2 at high temperatures Heterogeneous Catalysis Laboratory,
University of Cincinnati, Cincinnati, Ohio, USA. Advisor: Prof.
Panagiotis (Peter) G. Smirniotis.
2009 BSc Chemical Engineering, Texas A&M University, College Station,
Texas, USA.
Experience
2012-16 Research Associate & Teaching Assistant, Particle Technology
Laboratory, ETH Zurich.
2009-12 Research Associate & Teaching Assistant, Heterogeneous Catalysis
Laboratory, University of Cincinnati.
2008-09 Undergraduate Research Assistant, Molecular and Nano-Engineering
Group, Texas A&M University.
Awards
2016 Chemistry Travel Award, The platform chemistry of Switzerland
(SCNAT, SCS and SSFC), 2016, Switzerland for The 16th International
Congress on Catalysis, July 3-8, Beijing, China.
2012 First Place Environmental Division Graduate Student Paper Award, at the
Fall Meeting of the American Institute of Chemical Engineering. AIChE
2012, Oct. 28-Nov. 2, Pittsburgh, PA, USA.
2009-12 University Graduate Scholarship (UGS) at the University of Cincinnati,
Cincinnati, OH, USA.
2008 Artie McFerrin Department of Chemical Engineering Scholarship, Texas
A&M University, College station, TX, USA.
2007 Undergraduate Summer Research Grant, Texas A&M University, College
station, TX, USA.
Languages
English, Nepali, Hindi, German (beginner)
182 Curriculum Vitae
183
7. Publications and Presentations
Refereed Publications
1. Koirala, R., Gunugunuri, K.R., Pratsinis, S.E., and Smirniotis, P.G. Effect of Zirconia
Doping on Calcium Oxide using Flame technology on Sorbent Performance and
Stability for CO2 Capture, Journal of Physical Chemistry C, 115 (2011), 24804-
24812. (Paper Award, First Place, AIChE Environmental Division)
2. Koirala, R., Reddy, G. K. and Smirniotis, P.G. Single Nozzle Flame-Made Highly
Durable Metal Doped Ca-based Sorbents to Capture CO2 at High Temperatures,
Energy & Fuels, 26 (2012), 3103-3109.
3. Boningari, T., Koirala, R., Smirniotis, P.G. Low-temperature selective catalytic
reduction of NO with NH3 over V/ZrO2 prepared by flame-assisted spray pyrolysis:
Structural and catalytic properties, Applied Catalysis B: Environmental, 127 (2012),
255-264.
4. Boningari, T., Koirala, R., Smirniotis, P.G. Low-temperature Catalytic Reduction of
NO by NH3 over Vanadia-based Nanoparticles prepared by Flame-assisted spray
pyrolysis: Influence of various supports, Applied Catalysis B: Environmental, 140-
141 (2013), 289-298.
5. Koirala, R., Reddy, G. K., Lee, J.-Y., Smirniotis, P.G. Influence of Foreign Metal on
the Durability and Performance of Zr/Ca Sorbents during High Temperature CO2
Capture, Separation Science and Technology, 49 (2013), 47-54.
6. Koirala, R. Buechel, R., Pratsinis, S.E., Baiker, A. Oxidative Coupling of Methane on
Flame-made Mn-Na2WO4/SiO2: Influence of Catalyst Composition and Reaction
Conditions, Applied Catalysis A: General, 484 (2014), 97-107.
7. Koirala, R. Buechel, R., Krumeich, F., Pratsinis, S.E., Baiker, A. Oxidative
Dehydrogenation of Ethane with CO2 over Flame-made Ga-loaded TiO2, ACS
Catalysis, 5 (2015), 690-702.
8. Koirala, R., Pratsinis, S.E., Baiker, A. Synthesis of Catalytic Materials in Flame:
Opportunities and Challenges, Chemical Society Reviews, 45 (2016), 3053-3068.
9. Koirala, R., Buechel, R., Pratsinis, S.E., Baiker, A. Silica is Preferred Over Various
Single and Mixed Oxides as Support for CO2-assisted Cobalt-Catalytized Oxidative
Dehydrogenation of Ethane, Applied Catalysis A: General, 527 (2016), 96-108.
184 Publications and Presentations
Conference Presentations
Oral(underlined name is speaker)

1. Design of Refinery Crude Processing Unit Using Aspen Properties, Aspen Plus and
Aspen Icarus, Senior Year Project, Sponsored by Fluor, Texas A&M University,
USA, 2009 (Group Presentation).
2. Koirala, R., Gunugunuri, K.R., and Smirniotis, P.G. Flame Spray Synthesis of
Durable Doped-CaO Nanosorbents for CO2 Capture-Study of Sulfur Tolerance,
AIChE Annual Conference, Salt lake city, UT, USA, (7-12/11/2010).
3. Koirala, R., Gunugunuri, K.R., and Smirniotis, P.G. Flame Made Zr-doped CaO
Sorbents for High Temperature CO2 Capture, AIChE Annual Conference,
Minneapolis, MN, USA, (16-21/10/2011).
4. Gunugunuri, K.R., Koirala, R., and Smirniotis, P.G. Modified Ferrites as Catalysts
for High Temperature Water Gas Shift Reaction, AIChE Annual Conference,
Minneapolis, MN, USA, (16-21/10/2011).
5. Koirala, R., Gunugunuri, K.R., and Smirniotis, P.G. Flame Synthesized CaO-based
Sorbents for CO2 Capture: Influence of Metal ions as Dopants, AIChE Annual
Conference, Philadelphia, PA, USA, (Oct. 28-Nov.2, 2012).
6. Boningari, T., Koirala, R., and Smirniotis, P.G. Low-temperature Selective Catalytic
Reduction of NO with NH3 over V/ZrO2 Prepared by Aerosol Technique: Structure
and Catalytic Properties, NAM23, Louisville, KY, USA, (2-7/06/2013).
7. Koirala, R., Buechel, R., Pratsinis, S. E., Baiker, A. Oxidative Coupling of Methane
on Flame-made Mn-Na2WO4/SiO2: Influence of Catalyst Composition and Reaction
Conditions, NAM Annual Meeting 2015, Pittsburgh, PA, USA (13-19/06/2015).
Poster
1. Koirala, R., Seminario, G. Theoretical Study of Biological Hydrogen Formation in
Clostridium Pasteurianum towards Biomimetic, USRG, Dwight Look College of
Engineering, Texas A&M University, College Station, TX, USA (08/2008).
doping on Calcium Oxide using Flame Technology on Sorbent Performance and
Stability for CO2 Capture, UCEAO 5th Annual Conference, OH, USA, (26-
27/04/2011).
doping on the Structure and Stability of CaO-based Sorbents for CO2 Capture during
Extended Operating Cycles, 7th MRC Graduate Symposium 2012, ETH Zrich,
Switzerland (07/06/2012).
4. Koirala, R., Gunugunuri, K.R., Pratsinis, S.E., and Smirniotis, P., Structure and
Stability of Zr-doped CaO-based Sorbents for CO2 Capture during Extended Operating
Cycles, Partec 2013, Nuernberg, Germany (23-25/04/2013).
5. Koirala, R. Buechel, R., Pratsinis, S. E., Baiker, A. Oxidative Coupling of Methane
Conditions, MaP Graduate Symposium, Zrich, Switzerland (05/06/2014).
185
6. Koirala, R. Buechel, R., Krumeich, F., Pratsinis, S. E., Baiker, A. Oxidative

Dehydrogenation of Ethane with CO2 over Flame-made Ga-loaded TiO2, SCS
Annual Meeting, Zrich, Switzerland (11/09/2014).
7. Koirala, R., Buechel, R., Pratsinis, S. E., Baiker, A. Oxidative Coupling of Methane
Conditions, SCS Annual Meeting, Zrich, Switzerland (11/09/2014).
8. Koirala, R. Buechel, R., Krumeich, F., Pratsinis, S. E., Baiker, A. Oxidative
Dehydrogenation of Ethane with CO2 over Flame-made Ga-loaded TiO2, NAM24,
Pittsburgh, PA, USA (14-19/06/2015).
9. Koirala, R. Buechel, R., Pratsinis, S. E., Baiker, A. Cobalt-based Catalysts for
Oxidative Dehydrogenation of Ethane, MaP Graduate Symposium 2016, Zurich,
Switzerland (09/06/2016).
10. Koirala, R. Buechel, R., Pratsinis, S. E., Baiker, A. Flame-made Single and Mixed
Oxides Supported Cobalt Catalysts for Oxidative Dehydrogenation of Ethane, 16th
ICC, Beijing, China (03-08/07/2016).

Flame-Made Catalytic Materials For Alkene PDF

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Flame-Made Catalytic Materials For Alkene PDF

Transféré par

Droits d'auteur :

Formats disponibles

Diss. ETH No.

Flame-made Catalytic Materials for Alkene

A thesis submitted to attain the degree of

accepted on the recommendation of

2.2.1 Catalyst preparation 48

4.3 Results and Discussion 114

A.2 Catalytic results 168

FSP. Im Speziellen wird die einfache Kontrolle der Partikeleigenschaften hervorgehoben,

Verschiedene Trgermaterialien mit 4.5 Gewichts% Co wurden alsKatalysatoren mit

1. Synthesis of Catalytic Materials in Flames:

Part of it is published in Chemical Society Reviews, 45 (2016), 3053-3068.

Figure 1.1: Comparison of flame synthesis with classical wet-preparation method

1.2 Experimental conditions of flame synthesis and their

1.2.1 Particle formation mechanism

In liquid-fed flames we distinguish if the supporting fuel is mixed with the

1.2.2 Reactor configurations

periodic table can be used by liquid-fed flame synthesis processes. Both

1.2.3 Flame parameters for catalyst synthesis

Precursor concentration, fuel, mixing, oxidant, entrainment, precursor/dispersion flow

200 300 400

dispersion. Both phenomena favor resistance towards particle growth resulting in

Figure 1.4: Conceptual schematic of effect of precursor/dispersion (P/D) ratio on

(a) (b) (c)

(d) (e) (f)

Figure 1.5: Electron micrographs of the structure of various flame-made catalytic

Temperature strongly affects the characteristics of flame-made catalysts. Flame

of determining flame temperature computationally for a wide range of materials

1.3 Particle properties relevant for catalysis

The chemical composition is a key parameter in the design of catalysts. Flame

Brnsted/Lewis (B/L) acid ratio, -

Content of transition metal in TiO2-SiO2, ppm SiO2 content in Pt/Al2O3, wt.%

CO2 production, ppm

The acidity/basicity of catalysts can be tuned by their chemical composition.

1.3.2 Multi-metal systems

Multicomponent materials such as mixed oxides have been widely investigated in

100 SG-LaMnO3 (a)

1.3.3 Specific surface area

For reactions, however, that are sensitive to structural changes, it is important to

1.3.4 Particle size

1.3.5 Crystal structure and morphology

Correlating catalytic activity with crystal structure and morphology of

Amount of H2, mol

0.8 20Ag/TiO2 (P/D = 8/5)

1.3.6 Spatial distribution of components

Formation of mixed oxides is favored in single nozzle systems due to the

Figure 1.11: (a) Schematic of two-nozzle flame synthesis configurations leading to

better understanding of the particle formation mechanism in multicomponent system

1.3.7 Coating of ceramic supports

Structured ceramic supports have found wide application, particularly in

o-xylene conversion, Xo, -

foam-particles (93 m2/g)

Oxygen flow rate, L/min

160 200 240 280 320 360

1.4 Concluding Remarks

1.5. Ethene production from catalytic conversion of natural gas

1.5.2 Oxidative Coupling of Methane (OCM)

CH4 + O2 C2 H6 + C2 H4 + CO2 + CO + H2 O (1.1)

1.5.3 Oxidative Dehydrogenation of Ethane (ODHE)

1.5.4 Summary and outlook

characterization to gain fundamental understanding at atomic level to tailor the catalyst

Financial support by ETH Zurich Research Grant (ETH-39-12-2) is kindly acknowledged.

(1) Ulrich, G. D. Chem. Eng. News 1984, 62, 22-29.

(15) Thibaut, B. Platinum Met. Rev. 2011, 55, 149-151.