Catalytic Aromatization of Paraffin-Rich Oil under Methane ...

University of Calgary

PRISM: University of Calgary's Digital Repository

Graduate Studies The Vault: Electronic Theses and Dissertations

2018-09

Catalytic Aromatization of Paraffin-Rich Oil under

Methane Environment

Jarvis, Jack

Jarvis, J. (2018). Catalytic Aromatization of Paraffin-Rich Oil under Methane Environment

(Unpublished master's thesis). University of Calgary, Calgary, AB. doi:10.11575/PRISM/32934

http://hdl.handle.net/1880/107758

master thesis

University of Calgary graduate students retain copyright ownership and moral rights for their

thesis. You may use this material in any way that is permitted by the Copyright Act or through

licensing that has been assigned to the document. For uses that are not allowable under

copyright legislation or licensing, you are required to seek permission.

Downloaded from PRISM: https://prism.ucalgary.ca

UNIVERSITY OF CALGARY

Catalytic Aromatization of Paraffin-Rich Oil under Methane Environment

by

Jack Jarvis

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF MASTER OF SCIENCE

GRADUATE PROGRAM IN CHEMICAL AND PETROLEUM ENGINEERING

CALGARY, ALBERTA

SEPTEMBER, 2018

© Jack Jarvis 2018

ii

Abstract

Naphtha fractions obtained from petroleum refinement contain an abundant mixture

of hydrocarbons including paraffins, naphthenes, aromatics, and even olefins. n-paraffins

are the largest constituents of such oils and are the most undesirable because of their

poor octane values and low economic value as chemical feeds. Thus, scientific research

aims to convert these components into more valuable components with higher octane

numbers for fuels and/or high value chemical precursors used for chemical synthesis.

Current naphtha reforming processes require an element of hydrocracking to reduce the

number of larger carbon number components but hydrogen is expensive to obtain through

the current process of steam reforming natural gas and so an alternative source of

hydrogen is also desirable. One such source of hydrogen is methane, a naturally,

occurring, and cheap alternative. However, the activation of methane, the most stable of

the hydrocarbons, is difficult to achieve. This research aims at the conversion of naphtha

feeds (rich in n-paraffins) to more valuable benzene, toluene, and xylenes (BTX) whilst

using methane as a hydrogen source through heterogenous catalysis. Catalysts are

screened to gauge those with the highest performance and then the effect of methane is

also probed. This approach was conducted for two different fractions of naphtha as

provided by the petrochemical industry with very different components. A model

compound study was also conducted to enable a more comprehensive understanding of

the processes involved during upgrading.

iii

Acknowledgements

First and foremost, I would like to acknowledge the support provided by my extremely

supportive and driven supervisor, Dr. Hua Song, without whom I would not have been

able to provide the scientific community with the quality of work seen here. Dr. Song’s

financial support has enabled me to study and work here in Canada as a UK international,

an opportunity rarely seen by others. Perhaps most importantly however, has been Dr.

Songs expertise and willingness to guide me through this degree and even elevate my

research experience to the point that a continuing experience is desired by both parties

in the form of a PhD.

Further acknowledgement is given to my group members for their unyielding support

during my studies and for their assistance in my research. Working with this team has

taught me the importance of a group effort when conducting research.

I would also like to express my gratitude towards Dr. Chen and Dr. Park, without

whose expertise I would not have been able to perform my defense.

In addition, I would like to thank Dr. Matthew Clarke for his support during my degree

and for being there to answer my many questions.

Shangdong Chambroad Petrochemicals Co., Ltd were instrumental for without the

financial support my research would not have been able to move forward at the pace it

did.

Thanks are also given to the Canadian Light Source, where the XAS study described

in this thesis was performed. The Canadian Light Source is supported by the Canada

Foundation for Innovation, Natural Sciences and Engineering Research Council of

Canada, the University of Saskatchewan, the Government of Saskatchewan, Western

iv

Economic Diversification Canada, the National Research Council Canada, and the

Canadian Institutes of Health Research.

My final acknowledgement is dedicated to Danielle Austin and Omar Maan, my

dearest friends and colleagues throughout my two years at the University of Calgary.

v

Dedication

To

My dearest Mother and Father

I rarely have the opportunity to thank you and have taken this, my first official piece, to express

my thanks for assisting me both financially and morally throughout my academic career

My brothers

It is impossible to thank you enough for the support you unknowingly provide whilst I am half

the world away

My friends and family

All my love

vi

Table of Contents

Abstract .................................................................................................................... ii

Acknowledgements ................................................................................................ iii

Dedication................................................................................................................ v

Table of Contents .................................................................................................... vi

List of Tables ........................................................................................................... ix

List of Figures ......................................................................................................... x

Nomenclature ......................................................................................................... xv

Chapter One: Introduction ..................................................................................... 1

1.1 Background .................................................................................................... 1

1.2 Motivation ....................................................................................................... 4

1.3 Literature investigation on heavy naphtha ...................................................... 6

1.4 Literature investigation on light naphtha ......................................................... 9

1.5 Literature investigation on ZSM-5 and bimetallic species ............................. 13

1.6 Objectives ..................................................................................................... 16

1.7 Thesis organization ...................................................................................... 17

Chapter Two: Experimental Methods .................................................................. 19

2.1 Preparation of catalyst supports ................................................................... 19

2.2 Metal loading of catalyst supports ................................................................ 20

2.3 Reactor systems ........................................................................................... 21

2.4 Pyridine adsorption via DRIFTS ................................................................... 23

2.5 Ammonia temperature-programmed desorption (NH3-TPD) ......................... 26

2.6 X-ray powder diffraction (XRD) analysis ....................................................... 26

2.7 Thermogravimetric analysis-differential scanning calorimeter (TGA-DSC) ... 28

2.8 X-ray photoelectron spectroscopy (XPS) ...................................................... 29

2.9 X-ray absorption spectroscopy (XAS) ........................................................... 30

2.10 Transmission electron microscopy (TEM)................................................... 32

2.11 Gas chromatography and mass spectrometry ............................................ 33

vii

Chapter Three: Catalytic Aromatization of Naphtha Under Methane Environment: Effect of Surface Acidity and Metal Modification of HZSM-5 .... 35

3.1 Abstract ........................................................................................................ 36

3.2 Experimental ................................................................................................. 37

3.2.1 Catalyst synthesis ........................................................................... 37

3.2.2 Experimental procedure .................................................................. 38

3.2.3 Characterization techniques ........................................................... 39

3.2.4 Performance evaluation .................................................................. 42

3.3 Results and discussion ................................................................................. 45

3.3.1 HZSM-5 support effect .................................................................... 45

3.3.2 Effect of Zn loading ......................................................................... 48

3.3.3 Promoter effect ............................................................................... 50

3.3.4 Effect of external surface coverage................................................. 52

3.3.5 Effect of methane ............................................................................ 55

3.4 Catalyst characterization .............................................................................. 57

3.5 Conclusion .................................................................................................... 65

Chapter Four: Pt-Zn/HZSM-5 as a Highly Selective Catalyst for the Co-aromatization of Methane and Light Straight Run Naphtha .............................. 66

4.1 Abstract ........................................................................................................ 67

4.2 Experimental ................................................................................................. 68

4.2.1 Catalyst synthesis ........................................................................... 68

4.2.2 Experimental procedure .................................................................. 69

4.2.3 Characterization techniques ........................................................... 71

4.2.4 Performance evaluation .................................................................. 73

4.3 Results and discussion ................................................................................. 75

4.3.1 LSR naphtha conversion ................................................................. 75

4.3.2 Catalyst characterization ................................................................. 80

4.4 Conclusion .................................................................................................... 90

Chapter Five: Selective Aromatization and Isomerization of n-Octane from Intermetallic Pt-Zn Nanoparticle Alloys Supported on a Uniform Aluminosilicate ................................................................................................................................ 91

5.1 Abstract ........................................................................................................ 92

viii

5.2 Experimental ................................................................................................. 93

5.2.1 Synthesis of UZSM-5 ...................................................................... 93

5.2.2 Synthesis of Pt-Zn/UZSM-5 ............................................................ 93

5.2.3 Catalyst performance experimental procedure ............................... 94

5.2.4 Catalyst characterization procedures .............................................. 94

5.3 Results.......................................................................................................... 95

5.3.1 n-Octane aromatization and isomerization performance of Pt-Zn/UZSM-5 .............................................................................................. 95

5.3.2 UZSM-5 morphology ..................................................................... 100

5.3.3 Characterization of Pt-Zn/UZSM-5 ................................................ 102

5.4 Discussion and conclusion ......................................................................... 106

Chapter Six: Conclusions and Recommendations .......................................... 109

6.1 Conclusions ................................................................................................ 109

6.2 Recommendations for future work .............................................................. 111

References........................................................................................................... 113

ix

List of Tables

Table 3.1 ................................................................................................................ 43

Table 3.2 ................................................................................................................ 52

Table 3.3 ................................................................................................................ 64

Table 4.1 ................................................................................................................ 88

Table 4.2 ................................................................................................................ 89

Table 5.1 .............................................................................................................. 108

x

List of Figures

Figure 1.1 Percentage of completed scientific studies of different categories focused on naphtha reforming from 1949-2013 (Rahimpour, Jafari, and Iranshahi 2013) ..... 2

Figure 1.2 Structure of ZSM-5 showing the pentasil unit which is linked by oxygen species to form what’s known as pentasil chains (Kalogeras and Vassilikou-Dova 1998) ......................................................................................................................... 3

Figure 1.3 Natural gas production by region (billion cubic metres) demonstrating the growth of natural gas demand (BP Global 2018) ...................................................... 5

Figure 2.1 Wetness Impregnation set-up for metal loading of HZSM-5 .................. 20

Figure 2.2 Reactor set-up for naphtha reforming at 400°C and 30 bar pressures. Note that the pressure release valve as well as PM filters are omitted from this figure for simplification ........................................................................................................... 21

Figure 2.3 Parr instrument company 100-mL batch reactor used for model compound studies. Capable of temperatures and pressures up to 500°C and 5000 psi respectively (Parr Instrument Company 2018) ........................................................ 22

Figure 2.4 Pyridine as a suitable probe for Bronsted and Lewis acid sites due to its ability to its lone pair of electrons that aren’t delocalized and involved in the aromatic ring .......................................................................................................................... 24

Figure 2.5 DRIFTS set-up showing pyridine bubbler used to introduce pyridine to the sample .................................................................................................................... 25

Figure 2.6 Powder x-ray diffraction instrument basic schematic ............................ 27

Figure 2.7 Schematic of a TGA-DSC instrument as used in this research ............. 28

Figure 2.8 Schematic of an XPS instrument used to ascertain electronic states and quantify metal elements .......................................................................................... 29

xi

Figure 2.9 Depiction of terms referring to electron transitions in metals during XAS provided by Wikipedia ............................................................................................. 31

Figure 2.10 Schematic of imaging mode of a TEM instrument ............................... 32

Figure 2.11 A gas chromatography-mass spectrometry instrument using electron ionization for mass spectrometry ............................................................................ 34

Figure 3.1 Distribution of hydrocarbons within naphtha feedstock ......................... 43

Figure 3.2 A comparison of different Si:Al ratio HZSM-5 support catalysts and their effect on various quantifiable values ....................................................................... 45

Figure 3.3a A comparison of different Si:Al ratio ZSM-5 supports and their respective effects on the liquid product distribution of upgraded naphtha ................................ 47

Figure 3.3b A comparison of different Zn loadings (calculated on a weight basis) onto ZSM-5 (80:1) and their respective effects on the liquid product distribution of upgraded naphtha ................................................................................................... 47

Figure 3.3c A comparison of different promoter loadings on Zn/HZSM-5 (80:1) and their respective effects on the liquid product distribution of upgraded naphtha ....... 47

Figure 3.3d A comparison of different weight promoter loadings on Zn/HZSM-5 (80:1) with the external surface sites blocked and their respective effects on the liquid product distribution of upgraded naphtha ................................................................ 47

Figure 3.4 A comparison of different Zn loadings (calculated by weight) on HZSM-5 (80:1) and their effect on various quantifiable values .............................................. 49

Figure 3.5 A comparison of different promoter loadings on Zn/HZSM-5 (80:1) and their effect on various quantifiable values ............................................................... 51

Figure 3.6 The effect of external surface coverage on BTX selectivity and yield for promoter loaded Zn/HZSM-5 catalysts ................................................................... 48

Figure 3.7 A comparison of different environments and their effects on the distribution of components in the liquid product using a Pt-Zn/ZSM-5 catalyst ......................... 50

xii

Figure 3.8 A comparison of different environments and their effects on liquid and BTX yields using a Pt-Zn/ZSM-5 catalyst ........................................................................ 57

Figure 3.9 XRD patterns of HZSM-5 and varying Zn metal loadings as well as their spent counterparts under a methane environment .................................................. 58

Figure 3.10 XRD patterns of Zn/HZSM-5 and varying promoter loadings as well as their spent counterparts under a methane environment .......................................... 59

Figure 3.11 XRD patterns of ZSM-5, Zn/ZSM-5, Pt-Zn/ZSM-5 and their spent counterparts under a Methane Environment ........................................................... 60

Figure 3.12a TEM images of the spent Pt-Zn/HZSM-5 catalyst collected under CH4

................................................................................................................................ 61

Figure 3.12b TEM images of the spent Pt-Zn/HZSM-5 catalyst collected under N2 ................................................................................................................................ 61

Figure 3.12c Fringing of the Pt-Zn/HZSM-5 catalyst under CH4 ............................ 61

Figure 3.13 NH3-TPD profiles of a) pure ZSM-5, b) Zn/HZSM-5 and c) Pt-Zn/ZSM-5

catalysts .................................................................................................................. 63

Figure 3.14 DRIFT spectra of ZSM-5, Zn/ZSM-5 and Pt-Zn/ZSM-5 using pyridine as the molecular probe for adsorption, followed by desorption at 400°C ..................... 64

Figure 4.1 Distribution of components found in LSR naphtha ................................ 75

Figure 4.2 Graph displaying the performance of various HZSM-5 (280:1) catalysts with regards to BTEX selectivity, conversion, and BTEX yield, where BTEX selectivity is calculated as per equation (4) ............................................................................. 76

Figure 4.3 Graph displaying the performance of various HZSM-5 (280:1) catalysts with regards to BTEX liquid product selectivity and total liquid yield ....................... 77

Figure 4.4 Graph displaying the liquid product distribution of various HZSM-5 (280:1) catalysts .................................................................................................................. 78

xiii

Figure 4.5 Gaseous product distribution for varying HZSM-5 (280:1) catalysts on a fixed bed reaction system with LSR naphtha and methane or nitrogen .................. 79

Figure 4.6a TEM image of Pt-Zn/HZSM-5 post reaction under a methane environment ............................................................................................................ 81

Figure 4.6b TEM image of Pt-Zn/HZSM-5 post reaction under a nitrogen environment ................................................................................................................................ 81

Figure 4.6c Pristine Pt-Zn/HZSM-5 lattice fringing ................................................. 81

Figure 4.7 HAADF STEM images of pristine Pt-Zn/ZSM-5 (280:1) with elemental mapping, red corresponds to Pt particle locations with green corresponding to Zn particle locations ..................................................................................................... 81

Figure 4.8 XRD patterns for varying HZSM-5 (280:1) catalysts ............................. 83

Figure 4.9 Pyridine DRIFT spectra of varying HZSM-5 (280:1) catalysts collected at RT ........................................................................................................................... 85

Figure 4.10 NH3-TPD spectra of a) HZSM-5, b) Pt/HZSM-5, c) Zn/HZSM-5, and d) Pt-Zn/HZSM-5 ......................................................................................................... 87

Figure 5.1 Reaction pathways for a) n-octane, b) n-heptane, c) n-hexane, and d) n-pentane, when Pt-Zn/UZSM-5 is utilized as the catalyst ......................................... 96

Figure 5.2 Performance comparison between Pt-Zn/UZSM-5 and the conventional alternative where reactions were conducted as per the methods section ............... 97

Figure 5.3 A comparison of temperature changes when n-octane is employed as the reactant over Pt-Zn/UZSM-5 ................................................................................... 98

Figure 5.4 Demonstrating the isomerisation and aromatisation capability of Pt-Zn/UZSM-5 using different carbon number n-alkanes ............................................ 99

Figure 5.5 XRD patterns of UZSM-5 and CZSM-5, indicating typical MFI crystal structures .............................................................................................................. 100

xiv

Figure 5.6 HAADF-STEM images and analysis of Pt-Zn/UZSM-5 ....................... 101

Figure 5.7 HAADF-STEM images and analysis of Pt-Zn/CZSM-5 ....................... 102

Figure 5.8 Zn2p XPS spectrum of 4 catalysts to gauge the effect of intermetallic contribution to the states of Zn2+ ........................................................................... 103

Figure 5.9 Pt4f XPS spectra of 4 catalysts to gauge the effect of intermetallic contributions to the states of Pt metal and PtO species ........................................ 104

Figure 5.10 XANES spectra on the Zn L-edge region of 4 catalysts with magnifications on the pre-edge region and multiple scattering regions ............... 106

Figure 5.11 Preferred reaction pathway of CZSM-5 and UZSM-5 catalysts when loaded with Pt-Zn nanoparticles. UZSM-5 avoids cracking and oligomerisation to give a far more selective product .................................................................................. 107

xv

Nomenclature

Abbreviation Definition

APTES Aminopropyl-triethyoxylsilane

BTX Benzene, Toluene, Xylene

Cx Carbon number of species

CZSM-5 Conventional ZSM-5

DI Deionized

DSC Differential scanning calorimeter

EDX Energy dispersive x-ray analysis

EXAFS Extended x-ray absorption fine structure

GC Gas chromatography

GR Galvanic replacement

HAADF High-angle annular dark-field

HPLC High performance liquid chromatography

ICP Inductively coupled plasma

IE Ion exchange

IR Infrared

IWI Incipient wetness impregnation

KOH Potassium hydroxide

MFI Mordenite framework inverted

MS Mass spectrometry

PM Particulate matter

RT Room temperature

xvi

STEM Scanning transmission electron microscopy

TEM Transmission electron microscopy

TEOS Tetraethyl orthosilicate

TGA Thermogravimetric analysis

TPAOH Tetrapropylammonium hydroxide

TPD Temperature-programmed desorption

UZSM-5 Uniform ZSM-5

WI Wetness impregnation

XANES X-ray absorption near edge structure

XAS X-ray absorption spectroscopy

XPS X-ray photoelectron spectroscopy

XRD X-ray powder diffraction

ZSM-5 Zeolite Socony Mobil-5

1

Chapter One: Introduction

If the road is easy, you’re likely going the wrong way.

Terry Goodkind

Catalytic naphtha reforming is employed worldwide to meet varying legislations

interested in increasing the octane number of low-octane hydrocarbon containing streams.

Current processes require an element of hydrotreating, where the hydrogen is obtained

from the steam reformation of natural gas, an expensive process. This thesis shows the

technical feasibility of employing methane (the predominant component of natural gas)

as a direct hydrogen source by co-feeding it with various naphtha streams at moderate

temperatures and pressures using heterogenous catalysis.

1.1 Background

Naphtha oil in this context refers to the light oil fraction obtained from the refinement

of crude oil. These streams can contain any number of olefinic, paraffinic, aromatic and

naphthenic C5-C12 hydrocarbons with varying distributions depending on the refinery and

the crude oil source. However, n-paraffins are usually the main constituents (Li et al.

2018). The Octane number of a fuel indicates the antiknocking properties of a fuel, a

factor that effectively determines the ability of a fuel to resist compression before it is

ignited in the engine. n-paraffins have very poor octane numbers and so it is

understandable these components hold little value and are upgraded to either their

isomers or, preferably, aromatics. The most desirable of these higher-octane upgraded

chemicals are benzene, toluene, and xylene (BTX) not only for their higher-octane values

2

to prevent engine knocking but also for their use in the chemical industry as precursors

for polymer synthesis. In fact, naphtha reforming is responsible for approximately 40% of

the worlds production of high-octane gasoline (Nabgan, Rashidzadeh, and Nabgan 2018),

an apt demonstration of the focus of naphtha reforming and the need to refine reforming

processes.

The first real commercial outfit designed for the reforming of naphtha was in 1949 by

UOP using a Pt/Al2O3 catalyst modified with a Cl containing organic (Haensel 1951).

Although selectivity’s towards i-paraffins were appreciable, the catalyst was non-

regenerable and often posed corrosion problems. Since then, researchers have placed a

heavy focus on improving the process with 3 main categories as outlined in Fig. 1.1.

Figure 1.1 Percentage of completed scientific studies of different categories focused on naphtha reforming from 1949-2013 (Rahimpour, Jafari, and Iranshahi 2013)

3

From this alone, it is clear to see how catalyst formulation is the most important factor

when reforming naphtha feedstocks. 1969 saw the formulation of Zeolite Socony Mobil-5

(ZSM-5) (Argauer 1972), the catalyst featured in the work shown here and has an MFI

type framework. The morphology of ZSM-5 makes it very successful at promoting

aromatization through the shape of its pores as well as an increased surface area with

micro and meso porosity.

Figure 1.2 Structure of ZSM-5 showing the pentasil unit which is linked by oxygen species to form what’s known as pentasil chains (Kalogeras and Vassilikou-Dova 1998)

Arguably one of most exciting basic attributes of this support is the ability to fine-tune

the acidity. Made up of SiO2 and Al2O3 species, the Si4+ species can be replaced by Al3+

cations to create an additional positive charge which is necessary to keep the material

neutral. Acid-catalyzed reactions (explained later) can therefore be promoted or

depressed. In recent years, there has been a lot of interest focusing on the methods of

4

zeolite synthesis; using different organic templates to obtain the correct morphology

(Yaripour et al. 2015), bases such as KOH to form hierarchical structures (Meng et al.

2017), and even techniques used to alter the size of zeolite crystals formed. Arguably one

of the most effective ways to alter the performance of the support is the addition of a metal

or various metals to provide further attributes to the catalyst. Depending on the metal

used, the performance can be increased through enhanced dehydrogenation which can

then be further promoted by the addition of a second metal to form intermetallic particles

that alter the electronic properties of the metals in question.

Catalytic naphtha reforming has a rich history and research is still growing in this area

as the world attempts to better understand the processes involved when isomerizing and

aromatizing n-paraffins. The next subsection will outline the motivation of this research,

considering the reasoning behind using methane as a co-feed.

1.2 Motivation

During the naphtha reforming process, hydrogen partial pressure is achieved to

promote hydrocracking where the lager carbon number components are broken down.

Hydrogen is mostly obtained through reforming, where natural gas (whose major

component is methane) is steam reformed at extreme temperatures (~800°C) and

produces CO2 as a major greenhouse gas contributor (Hart et al. 2014). Understandably,

it would be a great boon to the industry if this step were eliminated, both from an

economical and environmental standpoint. As shown in Fig. 1.3, natural gas saw a very

successful year in 2017 with increases in global production by 4%, nearly double the

average growth rate for a period of 10 years (2.2%). This comes as a product of increased

5

demand given natural gases avenue in many areas including power generation, domestic

heating and fuel, transportation fuel, fertilizer, and many manufacturing processes.

Figure 1.3 Natural gas production by region (billion cubic metres) demonstrating the growth of natural gas demand (BP Global 2018)

Therefore, it is proposed that methane can be used as a direct donor of hydrogen

due to its recent abundance and availability through technological developments which

have enhanced the exploration and recovery of shale gas. Currently, this is not

reasonable at the commercial scale, however, since the discovery of the feasibility of low-

temperature nonoxidative activation of methane in the presence of olefins and higher

carbon number paraffins (Choudhary 1997), the field has boomed. Thus, this research

has been motivated by the prospect of reforming various naphtha feeds in the presence

of methane as an alternative source of hydrogen, with the major target products being

6

BTX for their use as gasoline additives and valuable industrial scale precursors in the

chemical industry.

1.3 Literature investigation on heavy Naphtha

Naphtha fractions obtained from petroleum refinement contain an abundant mixture

of hydrocarbons in the form of paraffins, naphthenes and aromatics in varying

distributions (Saxena, Viswanadham, and Garg 2013), although usually in respective

decreasing order of percent by weight. Unless converted, these fractions can be both

detrimental to the environment and economically unviable for the industry, making them

otherwise undesirable. During current reforming processes, lower carbon number

components of naphtha feeds (≤ 6 carbon atoms) tend to crack further into lower chain

hydrocarbons (Labinger et al. 2015), making them unsuitable for increasing the octane

number of gasoline feeds as per governmental regulations (Rossini 2003)Marques et al.

2005) as well obtaining desirable liquid yields. It is also important to reduce olefin content

(although seen in small quantities in most naphtha mixtures) as these compounds can

result in gum formation in combustion engines which can further result in less efficient

combustion (Lou et al. 2016).

Naphtha reforming requires an element of hydrocracking to reduce the number of

larger carbon number components and is also one of the reactions that requires hydrogen.

One predominant method of acquiring hydrogen for hydrotreating is through the process

of steam reforming natural gas (whose main component is methane) at high temperatures

(700-900°C) along with high amounts of CO2 production (Hart et al. 2014). We are

therefore encouraged to find an effective and inexpensive method for obtaining hydrogen.

7

A cheap, naturally occurring, and abundant alternative source to expensive hydrogen is

methane, a notoriously difficult molecule to activate given its stable C-H bond strengths

of 439 kJ mol-1, a result of its symmetric geometry and electronic configuration (Au, Ng,

and Liao 1999; Xing, Pang, and Wang 2011). Given this challenge, high temperatures

are required, but this inhibits selectivity towards desired products. Therein lies the

problem and a strong desire within the scientific community to activate methane at lower

temperatures so that selectivity is maintained along with affordable economics. This has

indeed been achieved through the use of bifunctional catalysts at low temperatures and

the presence of co-reactants where an increase in liquid product is seen under a methane

environment when compared to nitrogen (He, Gatip, et al. 2017).

Another important aspect of naphtha reforming is the ability to produce useful

aromatic products such as benzene (currently the second largest volume petrochemical),

toluene and xylenes (henceforth called BTX) (Yao, le van Mao, and Dufresne 1990; Song

et al. 2015; Ahuja et al. 2011; Seddon 1990); not to mention hydrogen as a by-product

which is seeing increased interest due to its application in many hydro-consuming

processes such as fuel cells and even its potential as a future energy carrier. Thus, the

prospect of producing these highly valued products used in polymer and other

petrochemical syntheses from otherwise undesirable feeds has been investigated

thoroughly since its conception by Vladimir Haensel in the 1950’s (Sterba and Haensel

1976). In fact, between 1949 and 2013 there were over 600 publications dedicated to the

process of naphtha reforming with 50% of these being dedicated to catalyst formulation

and performance analysis (Rahimpour, Jafari, and Iranshahi 2013).

8

Thanks to the works of many researchers (Fling and Wang 1991; Kwak et al. 1994;

Dautzenberg and Platteeuw 1970), the key reactions that take place during the formation

of aromatics from paraffins and naphthene’s have been elucidated as: dehydrocyclization,

isomerization, dehydrogenation, hydrodealkylation and hydrocracking. With regards to

catalyst formulations, H type ZSM-5 (HZSM-5) has proved to be a promising catalyst for

cracking hydrocarbons within the desired naphtha-range through two mechanistic

pathways. Protolytic cracking is the first and proceeds through the protonation of alkanes

to provide carbenium ions which then collapse to give the final products (Kotrel, Knözinger,

and Gates 2000). This pathway occurs because of the Brönsted acid function of zeolites

which is attained through aluminum substitution with silicon and a resulting proton for

electronic neutrality as well as their Lewis acid function through extra-framework

aluminum (Motz, Heinichen, and Holderich 1997). The second is through hydride transfer

between a reactant and an adsorbed carbenium ion in which a smaller carbenium ion is

then obtained through β-scission (Motz, Heinichen, and Holderich 1997). Ultimately, the

HZSM-5 acid functions provide sites for isomerization as well as cracking, but further

catalyst alterations are needed to provide the necessary tools for desirable reactions.

Thus, a metal function can provide additional pathways that lead to the desired products.

For example, noble metals are very successful at enabling hydrogenation and

dehydrogenation reactions which are vital processes in the aromatization processes of

naphtha feeds. However, coke formation is extensive when these metals are employed

and so this issue needs to be addressed (Choudhary et al. 1996). Zn and Ga metal

addition to HZSM-5 has shown some of the best results for the aromatization of paraffins

(Mériaudeau and Naccache 1996; Lukyanov and Vazhnova 2007; Viswanadham et al.

9

1996; Choudhary, Mulla, and Banerjee 2003; Su et al. 2017; Choudhary et al. 1996) as

well as reduced coking, with evidence of reduced undesirable cracked products and more

aromatics. In addition to this, a promoter can allow for more flexibility for attaining desired

reaction pathways and as a result, higher selectivity’s towards desirable products. The

synergy of Zn with other metal promoters however, is not well established and so is

investigated here with a range of promoters to compliment the limited work already

conducted (U. Mroczek, W. Reschetilowski 1991; He, Wen, et al. 2017; He, Lou, and

Song 2016). Catalyst modification can be taken a step further by investigation of the

effect of specific functions; for example, Ding et. al (Ding, Meitzner, and Iglesia 2002)

investigated the effect of external acid sites by silanation of Mo/HZSM-5 using 3-

aminopropyl-triethoxylsilane (APTES). This synthetic method allowed for effective

blocking of external acid sites with SiO2, formed after calcination and removal of bulky

organosilicon molecules. whilst leaving the acid sites within the internal pores free to

contribute to catalytic reactions.

Thus, this research aims to provide a clear presentation of the effects of support

acidity, Zn loading and promoter on a naphtha feedstock during its aromatization when

placed under a methane environment.

1.4 Literature investigation on light Naphtha

Light straight run naphtha (LSR) produced by petroleum refineries contains an

abundant mixture of many C5-C7 (predominantly C5) hydrocarbons (Chica and Corma

1999) which require conversion to more valuable products for the purpose of increasing

the octane number of gasoline fuels and/or producing valuable chemical feedstocks. Thus,

10

branched paraffins are becoming increasingly important given their increased research

octane number (RON) compared to normal paraffins, especially i-pentane and i-hexane

(~80) compared to n-pentane and n-hexane (~64) (Ghosh, Hickey, and Jaffe 2006).

Isomerized paraffins are not only important for the enhancement of a feedstocks octane

number, but are also more environmentally friendly than olefins when used in gasoline

(Fan et al. 2004). Thus, it is important to minimize olefin formation during the upgrading

of such LSR feedstocks.

It has been indicated by many researchers that the use of bifunctional catalysts can

be a very effective way of achieving such isomerization of normal paraffins since AlCl3

and metal on Al2O3 are not used anymore due to its poor performance (Ono 2003) and

Pt-Cl/Al2O3 encounters reactor corrosion issues (Chen et al. 2006). Pt/Mordenite avoids

these issues and it has even been reported that Pt/beta results in more activity for these

isomerization reactions (Arribas and Martínez 2001). It is widely accepted that

bifunctional mechanisms proceed through the following reactions: dehydrogenation of

alkanes on Pt, protonation of produced olefins on Brönsted acid sites to form carbenium

ions, rearrangement, β-scission, as well as de-protonation and hydrogenation over Pt to

result in the final isomerized forms of normal paraffins, as well as cracked gaseous

products.

Other desirable products from the upgrading of LSR are benzene, toluene, and the

xylenes (hereafter called BTX) which can provide enhanced octane numbers to fuels as

well as serve as solvents and other valuable industrial chemical feeds. Not only does the

aromatization provide BTX, but also hydrogen as a by-product, an increasingly valuable

commodity for its use in hydro-consuming processes such as fuel cells and as a future

11

energy carrier (Cano et al. 2018). Thus, the use of a successful aromatization catalyst

which can facilitate both isomerization and aromatization could potentially enhance the

value of LSR through the production of i-paraffins, BTX and hydrogen as a gaseous by-

product. One such catalyst is ZSM-5, a heterogenous catalyst capable of such reactions

and, furthermore, upon selective metal loading, can increase the selectivity of naphtha

feeds towards BTX in particular (Jarvis et al. 2018).

Deactivation of these catalysts during naphtha reforming occurs predominantly

through coking, an issue which is resolved in industry by applying a hydrogen atmosphere

which effectively reduces the environment to minimize carbon deposition (Lin et al. 2007).

Hydrogen also serves as a source to assist with hydrocracking during these upgrading

reactions, but it comes at a price due to its natural unavailability. A predominant method

of acquiring hydrogen is through the steam reforming of natural gas which consists mainly

of methane (Haryanto et al. 2005). This process is conducted at high temperatures,

900°C, and produces large amounts of CO2 during the recovery of more hydrogen

through the water gas shift reaction (CO + H2O ↔ CO2 +H2) which is an undesirable by-

product acting as a contributor to the greenhouse effect. Thus, we turn our attention to

CH4, an abundant, naturally occurring and cheap source of hydrogen if effectively

activated. Although the prospect of such an attainable hydrogen source seems promising,

methane’s notorious stability, attributed to its stable tetrahedral C-H bonds as well as its

electronic structure obtained from sp3 hybridization, has proven difficult to overcome

(Tang et al. 2014). Nonetheless, due to the large volume of researches focusing on the

conversion of methane and despite the understandably more reactive components

formed, nonoxidative activation of methane at low temperatures has been achieved in the

12

presence of olefins and higher carbon number paraffins as co feeds. Thus, a strong desire

amongst the scientific community continues to drive towards the ability to overcome the

challenge of activating CH4 through novel catalysis so that it can contribute to the

desirable liquid products whilst simultaneously providing the hydrogen needed to inhibit

catalyst deactivation.

Until now, many studies have been conducted on single metal loading onto ZSM-5

with regards to Zn (Abdelsayed, Smith, and Shekhawat 2015; J. Liu et al. 2018; Zhao and

Zhou 2018; Roshanaei and Alavi 2018; Li, Zhang, et al. 2018) and Pt (Samanta et al.

2017; Parsafard, Peyrovi, and Jarayedi 2017; Bai et al. 2018) but little attention has been

paid to their performance as an alloy. The Zn function of HZSM-5 has been shown to

promote the formation of aromatics (Tshabalala and Scurrell 2015) by providing

dehydrogenation sites on the catalyst surface which lead to olefin formation, followed by

aromatic production. Pt on the other hand, although more active, is less selective and

results in the formation of undesirable aromatics (>C9) because of its preferred reaction

pathway of hydrogenolysis which sees C-C bonds cleaved by H2, enabling the formation

of higher aromatics. Thus, this group has recently shown great interest in the effect of

Pt-Zn alloys on the aromatization of hydrocarbons, in this case, LSR, with the intent that

the alloy can promote the direct isomerization over less acidic, porous HZSM-5 so that

the positive morphological characteristics can still be taken advantage of. The idea is that

the less selective effect of sole Pt can be nullified using a Zn-Pt alloy whilst simultaneously

omitting the need to form gas products, which may remain unconverted if sole Zn is used

and we rely on its strong oligomerization tendencies.

13

We therefore provide an in-depth study of the effect of Pt-Zn/HZSM-5, a successful

catalyst for higher carbon number hydrocarbon feed conversion to aromatics (Jarvis et al.

2018), nominally BTX, with the intention of providing a novel solution to converting LSR

normal paraffin constituents to i-paraffins and BTX for increasing gasoline octane number

as well as increasing its value through forming high value added chemical feeds under

methane environment.

1.5 Literature investigation on ZSM-5 and bimetallic species

Petroleum refinery naphtha from distilled crude oil is catalytically reformed to improve

the anti-knocking index of fuel reformates by increasing the content of aromatics and i-

alkanes. These components also provide positive solution characteristics which help to

maximise swelling and reduce leaking of nitrile seals in aircraft fuels (Dewitt et al. 2008).

However, benzene is not a favourable component due to its high freezing point, an

undesirable characteristic of jet fuel, and is too toxic to allow for high concentrations in

gasoline (Shrivastav et al. 2017). As well as fuel applications, aromatics are highly valued

by the chemical production industry because of their role in the large-scale synthesis of

polymers, with the xylene isomers being of significance. Thus, given the high selectivity

towards benzene when concerned with <8 carbon number hydrocarbons and the

predominant presence of n-octane compared to other hydrocarbons in many naphtha

feeds, it has become increasingly important to investigate the aromatisation and

isomerisation of n-octane.

Since its conception at the lab scale in 1965, Zeolite Socony Mobil #5 (ZSM-5) has

found widespread involvement as a fluid catalytic cracking (FCC) additive for the

purposes of octane number improvement and propylene production from otherwise

14

undesirable n-alkanes and alkenes (Biswas and Maxwell 1990;Xin et al. 2017;Lv et al.

2017). A wide array of synthesis techniques on this support have since been investigated

to selectively dope with chosen metals (Jarvis et al. 2018), construct hierarchical porous

systems (Petushkov, Yoon, and Larsen 2011), as well as introduce structure variants(Tao,

Kanoh, and Kaneko 2003). Hydrothermally synthesising ZSM-5 using organic templates

has proven to be effective for controlling the size of crystals (Reding, Mäurer, and

Kraushaar-Czarnetzki 2003), as has the addition of varying transition metals, including

noble metals. Modern day catalysis aims for selectivity’s of 100% so that economic and

environmental targets are realised through the reduction of undesirable side reactions

(An et al. 2014). Until now, the use of ZSM-5 with/without metal doping has shown

selectivity’s well short of 100% towards i-alkanes and aromatics from n-alkanes. This

performance has been attributed to continuous oligomerisation of alkene intermediates

followed by cyclisation and hydrogen transfer over acidic sites on the support to form

higher aromatics (>C8) and polycyclic aromatic hydrocarbons (PAH) (Lubango and

Scurrell 2002). Metal addition, nominally through Zn doping, has been shown to increase

the selectivity towards aromatics through increased alkene intermediate concentration, a

result of its dehydrogenation capabilities (Zhao and Zhou 2018;Roshanaei and Alavi

2018;J. Liu et al. 2018). Although direct aromatisation of these alkenes is also promoted

through Zn presence, selectivity’s are still too poor to act as an effective solution for n-

alkane isomerisation and aromatisation.

Bimetallic alloys have enhanced the performance of catalytic systems in recent years,

a result of electronic and chemical property changes that are distinctly different from those

of their single parent metals (Bai et al. 2018;Samanta et al. 2017). Ligand and strain

15

effects have been shown as contributors to these changes through the formation of

heteroatom bonds and metal-metal bond length variations respectively (F. Liu, Wu, and

Yang 2017). The addition of a support, nominally oxides, to this bimetallic system adds

another aspect that enables the synthesis of a wide range of catalysts whilst furthering

the manufacturers ability to fine-tune the electronic properties. The term for such a

phenomenon between metals and their supports was originally coined as a strong metal

support interaction (SMSI) where, in a similar way to bimetallic effects, metal and support

interact and result in electronic, geometric and bifunctional effects (Narayanan 1985). As

SMSI’s usually require elevated temperature reduction under hydrogen and in the

presence of easily reducible oxides such as TiO2, this term is not suitable for our system

where we use the MFI structure of the effectively non-reducible mixed Al2O3 and SiO2

oxides (Haller and Resasco 1989). Thus, reactive metal support interactions (RMSI’s)

have recently been found to be more suitable for systems that do not necessarily require

extreme temperatures, easily reducible oxide supports, and even occur through thermal

annealing or partial reduction in place of reduction under hydrogen (Armbrüster, Schlögl,

and Grin 2014).

We herein report intermetallic nanoparticles of Pt-Zn supported on uniform

alluminosilicate aggregates (UZSM-5) synthesized without a reducing atmosphere until

experiments are conducted. This unique catalyst exhibits unrivalled direct aromatisation

and isomerisation of n-alkanes, tuneable with temperature because of the higher

activation energy needed to aromatise compared to isomerisation. A comparison has

been made with the conventional Pt-Zn aluminosilicate (CZSM-5) with synthesis shown

in the methods. The comparisons show an increased selectivity when UZSM-5 is

16

employed due to support morphology and interactions between the support and

intermetallic particles.

1.6 Objectives

The primary objective of this research is to formulate a catalyst capable of reforming

various naphtha feedstocks, nominally n-paraffins, under a methane environment at

moderate temperatures and pressures. More detailed objectives are outlines here:

1. Evaluate the composition of naphtha feeds provided by industry

2. Prepare a suitable reactor system to replicate an industrial process at the lab

scale

3. Screen different catalyst formulations and analyse their performance to

ascertain active sites

4. Evaluate the effect of methane by comparing with nitrogen control experiments

5. Conduct model compound studies on prominent n-paraffins as found in the

raw feedstocks

6. Perform deactivation studies on the catalysts to investigate the mode of

deactivation and the extent

17

1.7 Thesis organization

This thesis is organized into 6 chapters including 3 journal pieces. One journal is

published, the second is under peer review, and the third is in its final stages of completion.

I, Jack Jarvis have conducted the majority of experiments and data analysis with

assistance from my group members in terms of final submission and interpretation of

results. These group members are; Dr. Peng He, Aiguo Wang, Dr. Jonathan Harrhy, and

Dr. Qingyin Li. Dr. Song, the principle investigator, provided technical assistance in all

aspects of this work, providing advice and recommended steps. Dr. Ali Darbandi was

instrumental in assisting with TEM measurements made at the foothills campus. XPS

measurements were taken at the nanoFAB facility at the University of Alberta. ICP-MS

measurements were also made at the University of Alberta. HAADF-STEM images taken

at the Manitoba institute for materials (MIM) were conducted by Abdul Khan. XANES

measurements were made by Dr. Peng He at the Canadian light source (CLS).

Chapter 1 gives a brief outline of the history of naphtha reforming, including

background information, objectives, and organization of the thesis. Although an

introduction is provided, more detailed introductions are given in each chapter with

reference to the journals published, under review, or prepared.

Chapter 2 is the section that details the instruments used, the theories behind them,

and the experimental set-ups.

Chapter 3 Investigates the effect of surface acidity and metal loading of ZSM-5 on

naphtha. This chapter is published in Volume 223 and pages 211-221 of “Fuel” as

18

“Catalytic aromatization of naphtha under methane environment: Effect of surface acidity

and metal modification of HZSM-5”

Chapter 4 Demonstrates the effective upgrading of light straight run naphtha (a

naphtha oil containing light paraffins from C5-C7) over Pt-Zn/ZSM-5. This chapter is

submitted to “Fuel” and is titled “Pt-Zn/HZSM-5 as a highly selective catalyst for the co-

aromatization of methane and light straight run naphtha”

Chapter 5 Provides a model compound study through n-octane as well as the

synthesis of a recently novel designed catalyst with exemplary aromatization of n-

paraffins to their respective carbon number aromatics. This chapter is currently in the final

stages of submission to a journal and is titled “Selective aromatization and isomerization

of n-octane from intermetallic nanoparticle alloys supported on a uniform aluminosilicate”

Chapter 6 provides a conclusion to this work and delivers recommendations for future

studies to be undertaken during my PhD studies.

19

Chapter Two: Experimental Methods

There is nothing like looking, if you want to find something. You certainly usually

find something, if you look, but it is not always quite the something you were after.

J.R.R. Tolkien

2.1 Preparation of catalyst supports

In most cases, catalyst supports (ZSM-5) were purchased from Zeolyst International

in various SiO2:Al2O3 ratios (23, 30, 50, 80, 280) in their ammonium form. To convert to

their desired hydrogen forms (HZSM-5) the purchased supports were calcined in air at

600°C with the following ramping rate: 5°C/min to 120°C, hold for 1 hour, 5°C/min to

300°C, hold for 1 hour, and finally, 5°C/min to 600°C and held for 3 hours.

In chapter 5 we will see a catalyst synthesized using a hydrothermal technique using

the following procedure: Firstly, aluminum nitrate nonahydrate (Al(NO3)3·9H2O)

purchased from Alfa Aesar was added to a 1.0 M aqueous solution of

Tetrapropylammonium hydroxide (TPAOH) purchased from Sigma Aldrich and stirred at

room temperature until complete dissolution of Al(NO3)3·9H2O occurred. Tetraethyl

orthosilicate (TEOS) purchased from Merck KGaA was added dropwise under stirring to

the TPAOH and Al(NO3)3·9H2O solution, as prepared above, using a Pasteur pipette. The

resultant solution, upon complete addition of TEOS, was then stirred at room temperature

for approximately 1 hour where supersaturation occurred. The resultant gel was applied

to an autoclave and crystallized at 170°C for 72 hours. The crystals were then dried for

12 hours at 88°C and then calcined as per the Zeolyst International material above. Molar

20

ratios can be altered by changing the amounts of materials applied during synthesis. In

this case, a ratio of Al2O3:80SiO2:30TPAOH:1323H2O was calculated.

2.2 Metal loading of catalyst supports

There are many different techniques used to dope the catalyst supports with the

desired metals including Incipient wetness impregnation (IWI), wetness impregnation

(WI), ion exchange (IE), galvanic replacement (GR), and even incorporation of the metal

species into the zeolite framework itself. However, it was found that WI provided the best

dispersion of metals onto the catalyst as well as a reliable relationship between calculated

loading and actual loading. This process is outlined in the subsequent chapters

throughout this thesis, with a brief description shown here. As illustrated in Fig. 2.1, metal

salts (e.g. zinc nitrate hexahydrate and Tetraammineplatinum(II) nitrate) were added

simultaneously if bimetallic loading was desired to 30 mL of deionized water and stirred

at room temperature. Upon complete dissolution, 5 g of pre-calcined support was added

to the mixture and left for 24 hours. The mixture was then heated at 88°C until complete

evaporation of water. The dried catalyst was then dried further at 88°C for 12 hours before

being calcined as previously mentioned. By using the WI technique, metals were loaded

onto the framework and the nitrate salts removed during calcination.

Figure 2.1 Wetness Impregnation set-up for metal loading of HZSM-5

21

2.3 Reactor systems

Figure 2.2 Reactor set-up for naphtha reforming at 400°C and 30 bar pressures. Note that the pressure release valve as well as PM filters are omitted from this figure for simplification

A closed system as per Fig. 2.2 was achieved to be able to apply pressure with a

pressure release valve included to release pressure at a constant rate to maintain the

desired pressure and constant gas flow. The high-pressure liquid chromatography (HPLC)

pump was used in conjunction with a 3-necked round bottomed flask (rbf) to hold the

liquid feedstock. The use of a 3-necked rbf enabled the application of a He gas bubbler

to effectively degas the feedstock before being applied to the pump. The liquid then met

the reactant gas flow and entered the reactor at 400°C. The catalyst bed was applied to

the centre of the fixed bed column and dispersed with inert quartz wool. Working with the

gas flow, the product in the gas phase was then put through a condenser at -20°C to trap

the liquid products. Any gas products were put through the outlet and analyzed in-situ via

22

micro-GC. The liquid product was analyzed via GC-MS and liquid yields calculated by

comparison of the condenser weight before and after the reaction. Typical reactions were

run for 1 hour unless specified otherwise in remaining chapters. This reactor system

(4100C) was provided by Xi’an Sino-Green Hi-Tech Co., Ltd. More details of the specific

reactions are provided in later chapters as they occur.

In chapter 5, the reactions were conducted on a 100-mL batch reactor system as

provided by Parr instrument company and is shown in Fig. 2.3.

Figure 2.3 Parr instrument company 100-mL batch reactor used for model compound studies. Capable of temperatures and pressures up to 500°C and 5000 psi respectively (Parr Instrument Company 2018)

In a typical reaction, catalyst is loaded into the cylinder along with the desired

feedstock. The head piece is then applied and sealed using a graphene seal,

23

compression seals and finally, a compression ring. The reactor is then purged using

Nitrogen three times at 30 bar. The pressure was then released to 1 atm and the reactant

gas added to the desired initial pressure. The reactor could then be weighed and placed

in the stand, attached to the magnetic drive for agitation and heated at a rate of 20°C to

the desired temperature. Upon reaction completion (usually 1 hour) the system is allowed

to cool before weighing again. When at RT, the gas products are analyzed via micro-GC

and applied through a pressure valve so that the GC is not subject to high pressure. The

gas is then fully expunged to 1 atm and the reactor weighed again. The headpiece is

finally removed and the liquid products extracted using carbon disulfide (CS2) for 30 mins

under sonication. The CS2 diluted product is then applied to the GC-MS to analyze the

molar composition.

2.4 Pyridine adsorption via DRIFTS

Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) is commonly

employed in catalysis as a surface sensitive technique capable of analyzing materials

adsorbed on the surface of the catalyst. This technique can even be employed at reaction

temperatures with reactants. Using this, we can witness the bonding occurring on the

catalyst at varying temperatures as different interactions present varying wavenumbers

depending on the bonding species and their modes. Perhaps more commonly in the case

of ZSM-5 however, is the use of pyridine adsorption to provide qualitative information of

the acid sites present on the catalyst.

HZSM-5 contains the two acid types (Brönsted and Lewis) where Brönsted acid refers

to proton donors as situated on the OH sites when Al3+ species replace Si4+, and Lewis

24

acid refers to the acceptance of an electron pair from a donator, such as extra framework

Al species and metal species loaded onto the catalyst. The advantage of using pyridine

over other probe molecules is its ability to identify both acid types, this is because the

nitrogen in the heteroatomic compound has a lone pair of electrons that are able to

provide the electrons donated to Lewis acids. As this lone pair is not delocalized on the

ring, it is simple enough for the nitrogen to also accept a proton as provided by the catalyst,

hence this molecules ability to identify Brönsted acid sites (shown in Fig. 2.4).

Figure 2.4 Pyridine as a suitable probe for Bronsted and Lewis acid sites due to its ability to its lone pair of electrons that aren’t delocalized and involved in the aromatic ring

Pyridine adsorption was conducted using a Hg/Cd/Te (MCT) detector attached to a

Nicolet iS50 as provided by Thermo Scientific. Details of the analysis are provided in

subsequent chapters, but the instrument layout is provided here in Fig. 2.5. Potassium

Bromide (KBr) as a dispersion medium is not necessary for these studies as the bulk

catalyst material ranges between 100-400 nm, far below the necessary size for effective

DRIFTS measurements (Accardo et al. 2014).

25

Figure 2.5 DRIFTS set-up showing pyridine bubbler used to introduce pyridine to the sample

The IR beam reaches the sample and can either be reflected from the surface,

adsorbed or even penetrate the sample before scattering occurs. Scattered light is then

sent to the detector where absorption according to specific bonding can be identified. This

way, qualitative information on the acid sites present in the catalyst can be observed. As

well as this, DRIFTS is non-destructive, therefore, samples can be cleaned after their

analysis, avoiding waste of valuable material.

26

2.5 Ammonia temperature-programmed desorption (NH3-TPD)

Although DRIFTS provides qualitative information on the acid sites present in the

catalyst samples, ammonia temperature-programmed desorption (NH3-TPD) is used to

calculate the quantity and strength of acid sites. This is important because usually in the

case of metal loaded catalysts, the distribution of acid sites does not change dramatically,

the strength and total acidity however, is affected. There are many possible probes, with

ammonia being the preferred molecule due to its small size, enabling it to enter ZSM-5

pores. Once ammonia is sufficiently adsorbed onto the catalyst, the sample is flushed to

remove any physisorbed species, leaving chemisorbed ammonia on the varying acid sites.

A gradual temperature increase coupled with a thermal conductivity detector (TCD)

means that at certain points, the temperature is enough to remove ammonia from the

surface, resulting in a change in thermal conductivity and ultimately a different signal. In

the case of ZSM-5, there are usually two peaks that occur, indicative of strong and weak

acid sites which can then be quantified.

2.6 X-ray powder diffraction (XRD) analysis

The use of x-rays is important to identify the crystallographic structure of many

materials and is very applicable to MFI structures as well. Simply put, x-rays are applied

to the sample and scatter off the atoms therein, in the case of ZSM-5, the atoms are

ordered systematically and provide information on their arrangement through the way in

which the x-ray beam is diffracted. This type of analysis is very useful for confirming that

the material you have made is ZSM-5, among other things. This includes the observation

of intensity changes in the diffraction patterns which are indicative of the number of atoms

27

in the planes associated with those peaks. When metal loadings are high enough, XRD

can also be used to confirm the presence of said metals in the structure as these metal

oxides will appear as additional peaks in the patterns. Powder diffraction is preferred to

single crystal samples as this avoids the need to reorient to obtain d-spacings of each

plane. When powder diffraction is used, the particles are randomly orientated, therefore,

an x-ray beam can be applied from 3-60 degrees (the range where MFI crystal structure

patterns occur) and rely on the fact that parallel alignment of the different atomic planes

will occur if enough powder is present. Given that catalyst samples are ordinarily in

powder form, this technique makes the most sense to our system as grinding is not

necessary. The instrument used for this research was a Rigaku Multiflex diffractometer

using Cu Kα irradiation at 20 Kv and 40 mA within 2θ.

Figure 2.6 Powder x-ray diffraction instrument basic schematic

28

2.7 Thermogravimetric analysis-differential scanning calorimeter (TGA-DSC)

In this technique, a mass of sample material (spent catalyst) is observed as a function

of temperature which is elevated at a controlled rate. This can be used to monitor

oxidations and/or reductions of varying species occurring due to thermal treatment. In this

research, this method was used to measure the mass reduction as a result of temperature

where the sample could be heated up to 800°C (far above reaction temperatures) in air

to oxidize carbon species (coke) on the surface and remove it as CO2. The mass loss

because of this can be used in the mass balance to calculate the amount of coke formed

during the reaction.

Figure 2.7 Schematic of a TGA-DSC instrument as used in this research

The instrument used in this research was used to calculate coke yields and thus,

assess the level of deactivation of the catalyst in question. The equipment used was a

simultaneous thermal analyzer STA 6000 provided by PerkinElmer.

29

2.8 X-ray photoelectron spectroscopy (XPS)

When metals are loaded onto catalyst supports, especially in the case of bimetallic

loading, it is important to assess the oxidation states of metal species as well as the state

of the support itself. As another surface technique, XPS can be used to quantitively gauge

the metal species on the surface of the catalyst. When the sample is irradiated with low

energy x-rays, the material is simultaneously measured for kinetic energy. This kinetic

energy comes from the photoelectric effect where electrons with binding energy’s

indicative of the species that they come from are emitted by photons. This, in turn provides

information on the electron energy distribution of species within the sample.

Figure 2.8 Schematic of an XPS instrument used to ascertain electronic states and quantify metal elements

30

2.9 X-ray absorption spectroscopy (XAS)

XAS builds on the information gained through XPS but is even more sensitive and

can provide more detailed information on the samples such as local geometric and

electronic structures of the species analysed. Due to the specific beams required, the

work was conducted at the Canadian light source (CLS), the only synchrotron radiation

facility available in Canada. This facility can provide highly tunable x-ray beams to suit

the specific requirements of the samples used. In XAS, the electromagnetic radiation of

the x-rays which have an oscillating electric field, strike the electrons of an atom of focus

which are then excited to a higher energy or enter a continuum. This is not dissimilar to

XPS, however, in this case, the x-ray energy is at wavelengths on the same order of

magnitude as atom-atom separation, hence the ability to deduce local structures. There

are typically two regions of interest when observing XAS; X-ray absorption near edge

structure (XANES) and Extended x-ray absorption fine structure (EXAFS). The energy

used for our work was that of soft x-rays capable up to 3k eV and so only the XANES

region of the Zn L-edge was obtainable. The XANES allows the user to witness oxidation

states of species, their bound state transitions, and even changes in bonding and

geometries. XAS uses special terms to show excitation of electrons and is shown in Fig.

2.9.

31

Figure 2.9 Depiction of terms referring to electron transitions in metals during XAS provided by Wikipedia

Zn 2p L-edge spectra as obtained in this research correspond to the temporary

excitation of electrons from the 2p3/2 and 2p1/2, in other words, the L1 and L2 edges.

32

2.10 Transmission electron microscopy (TEM)

When catalyst samples are suspended in ethanol (EtOH) and dispersed, they are

then placed on a copper grid to obtain a very thin sample layer. The sample is then

subjected to a beam of electrons which interact with the sample under vacuum, the

subsequent image can then be magnified and focused. This technique allows the user to

obtain information on the bulk crystal size, metal nanoparticle size and distribution as well

as confirm the typical lattice fringing observed for crystal structures. When coupled with

energy dispersive x-ray analysis (EDX), we are able to confirm the elemental species

present through x-ray excitation. High-angle annular dark field scanning transmission

electron microscopy (HAADF-STEM) can also be used in place of TEM to enable a clearer

image of nanoparticles, a handy technique given that metal nanoparticles loaded onto

ZSM-5 often result in sizes less than 2 nm.

Figure 2.10 Schematic of imaging mode of a TEM instrument

33

2.11 Gas chromatography and mass spectrometry

When identifying product molecules from reactions, these two forms of analysis are

perhaps the most important. GC refers to gas chromatography and GC-MS refers to gas

chromatography-mass spectrometry.

Gas chromatography was conducted on a 490 micro-GC provided by Agilent

Technologies and was used to analyze the gas products. GC works by vaporizing

components without breaking them down, the vaporized species are then passed over a

column (4 columns in our case) using a carrier gas (usually He or Ar) where the analyte

species interact with the column in the form of adsorption. The time of adsorption is

related to the component of interest. Thus, different species elute at different times,

resulting in a series of peaks on the resultant spectra which can be assigned to specific

molecules. The peak area ratios are in tandem with molar % after calibration.

GC-MS was performed on a GC Claus 690 and MS Clarus SQ 8T provided by

PerkinElmer. The difference here is that the liquid product was injected directly into the

instrument after dilution so as not to saturate the column. The sample is vaporized without

breaking down the molecules and passed over a Paraffin, Olefin, Naphthene, Aromatic

(PONA column) capable of adsorbing these species and causing them to elute at different

times. However, the liquid product composition can be extremely complicated (over 200

components) and so the system is coupled with a MS which is able to ionize the species

and organize them based on their mass:charge ratio. Given that ionized species will occur

in particular (patterns) based on the preferred ionized forms, it is possible to uniquely

identify each component, even separating isomers.

34

Figure 2.11 A gas chromatography-mass spectrometry instrument using electron ionization for mass spectrometry

35

Chapter Three: Catalytic Aromatization of Naphtha Under Methane

Environment: Effect of Surface Acidity and Metal Modification of HZSM-5

This chapter is adapted from the following publication:

Jack Jarvis, Ashley Wong, Peng He, Qingyin Li, and Hua Song. 2018. “Catalytic

Aromatization of Naphtha under Methane Environment: Effect of Surface Acidity and

Metal Modification of HZSM-5.” Fuel 223: 211–21.

https://doi.org/10.1016/j.fuel.2018.03.045.

36

3.1 Abstract

The reformation of naphtha to obtain valuable chemical intermediates such as BTX

has attracted much attention. Here we report a novel catalytic process whereby a

bimetallic heterogeneous catalyst in the form of Zn-Pt/HZSM-5 is applied to a flow reactor

system to reform a complex naphtha feed under a methane environment. Different levels

of acidity of the HZSM-5 support are tested with a superior performance witnessed using

a SiO2:Al2O3 ratio at 80. Further testing is conducted using Zn as a dehydrogenating

component with improved performance witnessed when Zn is loaded to a certain degree.

A metal promoter is then combined with Zn to compare the effect of different promoters

with superior performance witnessed using Ga and Pt as promoters with Zn. BTX

selectivity and BTX yield are reported as 86.44% and 34.22% respectively under the Zn-

Pt/HZSM-5 catalyst. External site coverage improves performance for all bimetallic

catalysts with the exception of Zn-Pt/HZSM-5, suggesting that Pt might not promote the

migration of metal functions to the internal pores during synthesis. XRD spectra

demonstrate that Pt addition results in a more intact crystal structure after reaction when

compared to Zn alone. NH3-TPD and DRIFT analyses show a reduced amount of strong

acidic sites upon metal loading with an increase in the number of Lewis acid sites and

reduced Brönsted sites. This change in acidity could be one of the reasons for an

improved performance when Zn-Pt/HZSM-5 is employed. The effect of methane is also

witnessed over Zn-Pt/HZSM-5 with improved selectivity, yield (BTX and liquid) and active

metal dispersion when compared to a nitrogen environment, suggesting the possible

incorporation of methane into the products, BTX in particular. Ultimately, the catalyst

employed here opens an avenue for further research into the possible industrial

37

application of naphtha aromatization to form valuable chemical products under methane

environment.

3.2 Experimental

3.2.1 Catalyst synthesis

MFI Zeolite support material in the cationic ammonium form with SiO2/Al2O3 molar

ratios of 23, 30, 50, 80 and 280 were obtained from Zeolyst International. All supports

were converted to their hydrogen forms by calcining in air at 10°C/min to 100°C, holding

for 8 hours, ramping again at a rate of 10°C/min to 300°C, holding for 20 mins and finally

ramping at a rate of 10°C/min until 550°C was reached and holding for 3 hours. These

catalysts were then loaded with varying amounts of Zn; 1%, 3%, 5% and 7%, calculated

on a mass basis. The loadings were conducted using the wetness impregnation method:

The appropriate mass of crystalline Zn nitrate hexahydrate [Zn(NO3)3.6H2O], 99% purity,

from Alfa Aesar was added to deionized water (15 cm3) and thoroughly stirred, dried

catalyst (5 g) was then added to the mixture with stirring maintained for 8 hours at room

temperature (RT) post addition of catalyst support. The catalysts were then calcined

using the previously mentioned calcination parameters. Promoter metals (Ag, Ga, Ir, Pd,

Pt and Ru) were added at the same time as the Zn salt and the aforementioned procedure

undertaken. Promoter salts used were obtained from Sigma Aldrich: Silver nitrate (≥99%),

palladium(II) nitrate dihydrate (99.8%), tetraammineplatinum(II) nitrate (99.995%) and

ruthenium(III) chloride hydrate (99.98%), and Alfa Aesar: Ga(III) nitrate hydrate (99.9%)

and iridium(III) chloride (63.9% min).

When coating the promoted catalysts with silica, the technique used is that of Ding et

al. (Ding, Meitzner, and Iglesia 2002) with some minor alterations: A solution of

38

anhydrous ethanol (100 g) from Commercial Alcohols and (3-

aminopropyl)triethyoxysilane (APTES) (1.5 g), (99%) from Sigma Aldrich was added to a

200 (cm3) beaker and stirring applied. Promoter loaded catalyst (5 g) was then added

and vigorous stirring maintained for 4 hours. The gel-like mixture was subsequently

stirred at 75°C until the ethanol was mostly evaporated. The residuum (coated catalyst

with small amounts of ethanol remaining) was then placed in an oven (80°C) overnight to

complete the evaporation. The resulting catalysts were then calcined as above when

obtaining the H-type zeolites.

3.2.2 Experimental procedure

Reactions were carried out in a fixed bed microreactor with a 1.46 mm i.d. and a length

of 54 cm. The fixed bed microreactor system-4100C was obtained from the Xi’an Sino-

Green Hi-Tech Co., Ltd. Catalyst (1 g) was added to the reactor and packed with ceramic

beads to achieve comparable catalyst volumes. The naphtha feedstock was pumped into

the system at a rate of 0.055 cm3/min to achieve a 3.3g•h-1 WHSV for all reactions and

joined by methane gas (100 Standard Cubic Centimeters per Minute, SCCM) or nitrogen

gas (100 SCCM) depending on the desired reaction. The feedstocks then entered the

pressurized (5 MPa) reactor at 400°C to undergo conversion for 1 hour. A condenser at

-20°C was then applied to the system to capture the resulting liquid products. The liquid

products were analyzed by a Gas Chromatography-Mass Spectrometer provided by

PerkinElmer (PerkinElmer GC Claus 680 and MS Clarus SQ 8T) and equipped with a

paraffin/olefin/naphthene/aromatic (PONA) column provided by Agilent (HP-PONA). The

temperature program was as follows: Hold at 35oC for 15 mins, ramp up to 70oC at a

39

heating rate of 1.5oC/min, and subsequently elevate to 150oC at 3oC/min and maintain for

30 mins. The system was then ramped at a rate of 3oC/min to 250oC and held for 30 mins.

Gas products were fed to a 490 micro-GC from Agilent Technologies and analyzed. The

system is made up of 4 channels and is equipped with thermal conductivity detectors

capable of precisely analyzing hydrogen, oxygen, nitrogen, methane, and carbon

monoxide in the first channel; carbon dioxide, acetylene, ethylene, and ethane in the

second channel; and paraffins between propane and hexane as well as olefins between

propylene and n-pentene in the third and fourth channels. The channels were equipped

with a 10-m molecular sieve 5A column, 10-m PPU column, a 10-m alumina column and

an 8-m CP-Sil 5 CB column respectively. argon is employed as the carrier gas for the first

channel with helium for the remaining three.

3.2.3 Characterization techniques

A plethora of techniques need to be employed to characterize a small selection of the

catalysts that were tested under the reaction conditions mentioned above. X-ray Powder

Diffraction (XRD), Transmission Electron Microscopy-Energy Dispersive X-ray

Spectroscopy (TEM-EDX), CO Pulse Chemisorption, Ammonia Temperature

Programmed Desorption (NH3-TPD), Diffuse Reflectance Infrared Fourier Transform

(DRIFT) and Inductively Coupled Plasma-Mass Spectrometry (ICP-MS) were thus used

to obtain information on the effects that different catalysts and environments (CH4 and N2)

had on the reactions carried out as per section 2.2 with characterizations also being

carried out on fresh and spent catalysts.

40

XRD was used to examine the crystalline compositions and was conducted on a

Rigaku Multiflex Diffractometer with Cu Kα irradiation at 20 kV and 40 mA in the 2θ region

between 5° and 60°.

TEM analysis was conducted on an FEI Tecnai F20 system with the FEG operating

at 200 kV and a 4k Gatan CCD camera. The sample was first dispersed in ethanol and

supported on honey carbon on a 200 mesh Cu grid before the TEM images were recorded.

CO Pulse Chemisorption was conducted on an INESORB-3010 Chemisorption

Analyzer provided by SinoGreen Hi-Tech. The catalyst (~0.2 g) was loaded to the system

and then flushed with pure Argon gas (20 SCCM) for 30 mins. The flow was then switched

to a 5% CO in Ar gas mixture (20 SCCM) and left for a further 30 minutes where the

temperature was then increased at a rate of 3°C/min to 200°C and held for 90 mins before

changing the gas flow to pure argon (20 SCCM) for 30 mins to remove any excess CO.

The system was then cooled to room temperature under pure argon (20 SCCM) and the

catalyst sample was subsequently subjected to 20 pulses of a 5% CO in argon gas

mixture (0.803 cm3). Metal dispersion was then calculated using the software with CO to

metal adsorption ratio at literature reported values (Canton et al. 2002), (Balakrishnan,

Sachdev, and Schwank 1990) and peak information obtained from the Chemisorption

Analyzer.

NH3-TPD was also conducted on the Chemisorption Analyzer mentioned above with

helium and nitrogen as the carrier gases. The catalyst samples (0.2 g) were loaded to

the system and degassed at 100°C for 60 mins with helium (20 SCCM), followed by a

temperature ramp of 10°C/min to 600°C where they were then held for 100 mins. The

system was then cooled to 120°C in the pure helium (20 SCCM) stream, after which, the

41

gas was switched to a continuous flow of 0.5% NH3 in nitrogen gas mixture (30 SCCM)

for 60 mins at 120°C to achieve chemisorption of the ammonia species with minimized

physisorption. The system is then purged for another 60 mins under a helium gas flow

(20 SCCM) to remove any remaining physisorbed ammonia. Desorption of the ammonia

species was then achieved through heating of the system to 600°C at a rate of 10°C/min

where it was then held for 30 mins. Peak deconvolution was then conducted to obtain

the acidity (µmol NH3/g cat.) of different acid sites.

DRIFT analysis of the catalyst samples was conducted on a Nicolet iS50 from Thermo

Scientific equipped with a Hg/Cd/Te (MCT) detector. The fresh catalyst samples were

applied to the closed system and heated to 500°C in a flow of pure nitrogen (30 SCCM)

and held for 15 mins. Post pre-treatment, the system was cooled to room temperature

(RT) and industrial nitrogen (35 Standard Cubic Feet per Hour, SCFH) was also applied

to purge air from the system surrounding the capsule. After running tests to ensure all

CO2 was removed from the system, background spectra were obtained and then pyridine

adsorption was acquired by flowing nitrogen gas (30 SCCM) through a pyridine bubbler

for 30 mins, followed by a removal of excess pyridine by purging with pure nitrogen (30

SCCM) for a further 30 mins. The spectra were then collected using the Kubelka-Munk

model with 512 scans at a resolution of 4 cm-1 at atmospheric pressure. The resulting

spectra then enabled qualification of the Brönsted and Lewis acid sites on the fresh

catalysts.

ICP analysis was performed using a Perkin Elmer Elan 6000 after successful sample

digestion in HF and HNO3.

42

3.2.4 Performance evaluation

The yield of liquid product was obtained by weighing the condenser pre-and post-

reaction and from there, the gas yield was calculable by subtraction. Molar composition

was obtained using the GC-MS and GC spectra mentioned earlier and following that, the

mass composition was calculable. Of course, there are a variety of ways to assess the

performance from this point, however, given that BTX is currently of great interest to the

community, the following quantifiable definitions have been selected and assessed:

𝐶𝑜𝑛𝑣𝑒𝑟𝑠𝑖𝑜𝑛 (%) = ((𝑀𝑎𝑠𝑠 𝑜𝑓 𝐿𝑖𝑞𝑢𝑖𝑑 𝐹𝑒𝑑−𝑀𝑎𝑠𝑠 𝑜𝑓 𝑈𝑛𝑟𝑒𝑎𝑐𝑡𝑒𝑑 𝐿𝑖𝑞𝑢𝑖𝑑)

𝑀𝑎𝑠𝑠 𝑜𝑓 𝐿𝑖𝑞𝑢𝑖𝑑 𝐹𝑒𝑑) × 100% (1)

𝐿𝑖𝑞𝑢𝑖𝑑 𝑌𝑖𝑒𝑙𝑑 (%) = (𝑀𝑎𝑠𝑠 𝑜𝑓 𝐿𝑖𝑞𝑢𝑖𝑑 𝑝𝑟𝑜𝑑𝑢𝑐𝑒𝑑

𝑀𝑎𝑠𝑠 𝑜𝑓 𝐿𝑖𝑞𝑢𝑖𝑑 𝐹𝑒𝑑) × 100% (2)

𝐵𝑇𝑋 𝑌𝑖𝑒𝑙𝑑 (%) = (𝑀𝑎𝑠𝑠 𝑜𝑓 𝐵𝑇𝑋 𝑖𝑛 𝑃𝑟𝑜𝑑𝑢𝑐𝑡

𝑀𝑎𝑠𝑠 𝑜𝑓 𝐿𝑖𝑞𝑢𝑖𝑑 𝐹𝑒𝑑) × 100% (3)

𝐵𝑇𝑋 𝑆𝑒𝑙𝑒𝑐𝑡𝑖𝑣𝑖𝑡𝑦 = (𝑀𝑜𝑙𝑒𝑠 𝑜𝑓 𝐵𝑇𝑋 𝐶𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡

((𝑇𝑜𝑡𝑎𝑙 𝐶# 𝑅𝑒𝑎𝑐𝑡𝑒𝑑 𝐶# 𝐶𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡⁄ )×𝑇𝑜𝑡𝑎𝑙 𝑀𝑜𝑙𝑒𝑠 𝑅𝑒𝑎𝑐𝑡𝑒𝑑)) × 100% (4)

𝐵𝑇𝑋 𝑌𝑖𝑒𝑙𝑑#2 (%) = (𝑀𝑠𝑠𝑠 𝑜𝑓 𝐵𝑇𝑋 𝑖𝑛 𝑃𝑟𝑜𝑑𝑢𝑐𝑡−𝑀𝑎𝑠𝑠 𝑜𝑓 𝐵𝑇𝑋 𝑖𝑛 𝐹𝑒𝑒𝑑

𝑀𝑎𝑠𝑠 𝑜𝑓 𝐿𝑖𝑞𝑢𝑖𝑑 𝐹𝑒𝑑−𝑀𝑎𝑠𝑠 𝑜𝑓 𝐵𝑇𝑋 𝑖𝑛 𝐿𝑖𝑞𝑢𝑖𝑑 𝐹𝑒𝑒𝑑) × 100% (5)

The calculations were made using compositional analysis through GC-MS and GC for

liquid and gas samples respectively, where each catalyst was tested in triplicate to obtain

a standard deviation for each quantifiable value as per the above calculations. The

composition of the naphtha feed provided by Chambroad Petrochemicals is shown in Fig.

3.1 and thus, it should be noted that the feedstock contains 14.4% BTX by weight which

is considered when applied to the summary table as shown in Table 3.1. It should also

be noted that given the method used to calculate yield, it will not be possible to obtain

100% liquid yield due to the expulsion of hydrogen through aromatization. However, this

is negligible and will not affect yield calculations to a significant degree.

43

Figure 3.1 Distribution of hydrocarbons within naphtha feedstock

Table 3.1 BTX Selectivity, corrected selectivity and BTX yields for varying catalysts

Catalyst Selectivity(%)

/±0.77 Selectivity1(%)

/±0.77 Yield(%) /±0.94

Yield2(%) /±0.94

HZSM-5 (23:1) 55.16 20.75 10.43 -4.66

HZSM-5 (30:1) 49.52 25.89 14.09 -0.39

HZSM-5 (50:1) 54.45 30.20 15.13 0.83

HZSM-5 (80:1) 65.31 41.50 20.10 6.64

HZSM-5 (280:1) 77.06 43.93 17.14 3.18

44

0.9%Zn/HZSM-5(80:1) 67.23 43.02 20.99 7.67

2.5%Zn/HZSM-5 (80:1) 73.20 21.69 21.74 8.55

4.1%Zn/HZSM-5 (80:1) 84.47 59.96 28.54 16.50

5.84%Zn/HZSM-5 (80:1) 71.13 43.57 18.71 5.01

Ru-Zn/HZSM-5 (80:1) 80.03 50.89 21.34 8.09

Ag-Zn/HZSM-5 (80:1) 79.97 52.92 23.39 10.48

Pd-Zn/HZSM-5 (80:1) 82.37 56.38 24.33 11.58

Ir-Zn/HZSM-5 (80:1) 83.55 57.11 26.48 14.09

Ga-Zn/HZSM-5 (80:1) 82.18 59.18 31.21 19.61

Pt-Zn/HZSM-5 (80:1) 86.44 63.88 34.22 23.14

Pt-Zn/HZSM-5 (80:1) N2 82.47 58.17 26.07 13.62

Ru-Zn/HZSM-5 (80:1) SiO2 85.46 58.13 24.41 11.67

Ag-Zn/HZSM-5 (80:1) SiO2 88.17 64.00 32.17 20.74

Pd-Zn/HZSM-5 (80:1) SiO2 85.32 63.34 34.85 23.88

Ir-Zn/HZSM-5 (80:1) SiO2 84.47 57.49 25.56 13.01

Ga-Zn/HZSM-5 (80:1) SiO2 88.37 62.92 30.33 18.59

Pt-Zn/HZSM-5 (80:1) SiO2 89.34 64.90 31.08 19.47

1 Selectivity calculated with inclusion of BTX in feedstock

2 Yield calculated with inclusion of BTX in feedstock as per equation (5)

45

3.3 Results and discussion

3.3.1 HZSM-5 support effect

HZSM-5 is well established within the field of catalysis as a successful support for

heterogeneous applications involved with hydrocarbon isomerization reactions (Ono

1992), not to mention its synthetic flexibility with regards to different SiO2:Al2O3 ratios

within the framework, leading to well controlled acidity (Ono 1992). To investigate the

effect of acidity on aromatization of complex naphtha feedstocks, five different levels of

acidity were synthesized to gauge which support would provide the best results in terms

of BTX yield and selectivity. The ratios used (SiO2:Al2O3) were 23:1, 30:1, 50:1, 80:1 and

280:1, leading to reduced acidity as aluminum is replaced in the framework by silicon.

Fig. 3.2 shows the effect of HZSM-5 acidity on the quantifiable values calculated as per

section 3.31.

Figure 3.2 A comparison of different Si:Al ratio HZSM-5 support catalysts and their effect on various quantifiable values

46

Liquid yield shows a negative trend with increased acidity, an understandable

observation given that the increased aluminum content provides more Brönsted sites for

proton donation and thus enables the formation of more gaseous products as a result of

isomerization and cracking. Protolytic paraffin cracking can aid with an explanation for

this in the form of a more acidic catalyst protonating paraffins to form pentavalent

carbonium ions, which in turn collapse to form smaller paraffins as well as olefins. It must

be noted that this mechanism as proposed by Haag and Dessau and now widely accepted,

is only relevant when olefins are in small quantities (Kotrel, Knözinger, and Gates 2000),

as in the case of the naphtha feedstock used here. At any rate, despite showing the

greatest conversions, HZSM-5 (23:1 and 30:1) also demonstrate the poorest BTX

selectivity’s and yields due to the formed gas products contributing to the calculations.

Fig. 3.3a demonstrates a small quantity of olefins in all cases but a complete reduction

of naphthene content to zero when less acidic supports are employed, suggesting that

aromatization of cyclic alkanes is inhibited upon increased support acidity under a

methane environment. A general trend is almost evident with increasing ratios of BTX to

other aromatics in the liquid fraction when the acidity is continually reduced, even as far

as the HZSM-5 (280:1) catalyst. It would seem that the HZSM-5 (80:1) catalyst however,

is equipped with the desired number of acidic sites for obtaining BTX as a product with

catalysts of lower and higher alumina ratios falling short when concerned with BTX yield.

This could be because a reduced acidity does not provide enough acidic sites for

Protolytic cracking to occur to a significant degree, whereas higher acidities provide too

much cracking, resulting in a kinetically favorable reaction towards gaseous products. As

a result of this, a plateau can be seen with regards to superior performance from the

47

HZSM-5 (80:1) catalyst under a methane environment, encouraging further catalyst

modifications to be conducted on this support.

Figure 3.3 a) A comparison of different Si:Al ratio ZSM-5 supports and their respective effects on the liquid product distribution of upgraded naphtha. b) A comparison of different Zn loadings (calculated on a weight basis) onto ZSM-5 (80:1) and their respective effects on the liquid product distribution of upgraded naphtha. c) A comparison of different promoter loadings on Zn/HZSM-5 (80:1) and their respective effects on the liquid product distribution of upgraded naphtha. d) A comparison of different weight promoter loadings on Zn/HZSM-5 (80:1) with the external surface sites blocked and their respective effects on the liquid product distribution of upgraded naphtha.

48

3.3.2 Effect of Zn loading

Metal addition to HZSM-5 has been extensively studied due to the increased stability,

neutralized acidity, and improved selectivity towards BTX that can be observed if the

metals chosen are suitable. Zn and Ga in particular, have been of significant interest to

those in the field of aromatization, with different pathways to those of conventional

zeolites observed upon the study of a model compound. However, Zn is the focus for this

Naphtha feedstock due to its enhanced dehydrogenation capabilities (Tshabalala and

Scurrell 2015) as well as methane dehydroaromatization success along with an oxidized

active phase that allows the avoidance of lengthy reactivation (Ono 1992). To observe

which loading is suitable for this feedstock, Zn was loaded onto HZSM-5 (80:1) at nominal

loadings of 1%, 3%, 5% and 7% (actual loadings 0.9, 2.5, 4.1 and 5.84 respectively were

achieved through ICP-MS analysis) and the aforementioned reactions undertaken with

graphical displays shown in the form of performance (Fig. 3.4) and liquid product

distribution (Fig. 3.3b). A marked increase in performance is instantly witnessed upon

the loading of Zn to the catalyst in all cases, probably owing to zincs attributes in the

activation of methane and dehydrogenation of paraffinic components. Recalling that

overall catalyst performance is gauged based on BTX selectivity and yield, it is clear to

see that 4.1%Zn loading by weight onto HZSM-5 (80:1) yields the most promising results

with an impressive improvement made even more noticeable when comparing BTX and

non-BTX aromatic fractions. However, understandably, an increase of Zn loading beyond

4.1% by weight to 5.84% impedes the overall catalyst performance with results similar to

that of un-modified Zeolites witnessed in section 3.1. Clearly, Zn has a positive impact

on performance but when loading is increased too dramatically, too many acidic sites are

49

blocked and higher metal loadings even cause a deformation in the crystal structure of

the support (witnessed in the supplementary data), ultimately altering the effectiveness

initially achieved with the unique morphology of MFI zeolites. As a result of 4.1%Zn’s

superior performance when compared to other loadings under methane, 4.1%Zn is

loaded in situ with other metal promoters to identify any positive synergetic effects

between these metals in the next section.

Figure 3.4 A comparison of different Zn loadings (calculated by weight) on HZSM-5 (80:1) and their effect on various quantifiable values

50

3.3.3 Promoter effect

Promoters and their contribution to aromatization reactions have been studied but only

a fragment of the combinations have been investigated. As a result, a much wider range

of metal promoters are investigated here with Zn as a base metal on the HZSM-5 catalysts

with Fig. 3.5 and Fig. 3.3c showing quantifiable results and liquid product distribution

respectively. On addition of most metals (Ru, Ag, Pd and Ir) the performance is actually

reduced when compared to that of 4.1%Zn/HZSM-5 (Fig. 3.4) with only Ga and Pt

showing an increase in BTX yields, no doubt a result of their well-established attributes

as seen in literature (Mikhailov et al. 2007). However, when concerned with the poorer

performing metal promoters, a likely result is due to increased loading of metals as

mentioned earlier with the 5.84% by weight loading of Zn to the catalyst. Perhaps this is

what is being seen here and the performance of the catalyst is thus impeded. In the

cases of Pt and Ga however, perhaps these metals are efficient dispersion promoters

(Mikhailov et al. 2007) leading to increased dispersion of both Zn species as well as Ga

and Pt themselves during catalyst synthesis. A higher metal dispersion is indeed

witnessed in the case of Pt (Table 3.2) and is witnessed in literature (Schanke 1996).

51

Figure 3.5 A comparison of different promoter loadings on Zn/HZSM-5 (80:1) and their effect on various quantifiable values

52

Table 3.2 Metal loadings and dispersion for varying catalysts

3.3.4 Effect of external surface coverage

External surface sites were blocked using APTES, a bulky organosilicon, to conduct

a preliminary investigation on the role that external sites have to play (Ding, Meitzner, and

Iglesia 2002) when different metal promoters are employed. Interestingly, when looking

at Fig. 3.6 the deactivation of the external sites has an adverse effect on Pt with a positive

effect on the others, albeit a small increase in some cases. Of particular note is the

increase of performance on the Pd catalyst, where originally the performance was

mediocre at best, upon blocking the external sites, the performance increases drastically.

On the other hand, the Pt and Ga catalysts which gave promising results when free of

blocking, seem to be negatively affected or only affected in a minor way, by the blocking

of external sites. Perhaps Ga and Pt were not well distributed within the pores of the

Catalyst

Zn

Loading

(Wt. %)

Promoter

Loading

(Wt. %)

Metal

Dispersion

(%)

Active Metal

Site

(m2/g)

Ru-

Zn/HZSM-5 3.46 0.09 3.8 1.37

Ag-

Zn/HZSM-5 3.86 0.45 0.6 0.48

Pd-

Zn/HZSM-5 3.06 0.51 6.5 2.50

Ir-

Zn/HZSM-5 4.16 0.28 1.2 0.51

Pt-

Zn/HZSM-5 3.90 0.57 7.2 1.97

53

catalyst, preferring to locate on the external surface and hence upon blocking of these

external sites, these metal species are rendered ineffective. In fact, it has been suggested

by Lou et al (Lou et al. 2016) that Pd may prefer to diffuse into the internal pores of the

catalyst, allowing the assumption that more effective reactions are permitted to occur if

the external surface sites are blocked. At any rate, the performance is not increased

dramatically in favor of BTX selectivity and BTX yield and so the extra costs associated

with catalyst synthesis may not be worthwhile for industrial applications. Thus, Pt-

Zn/HZSM-5 (80:1) is investigated further using varying characteristic techniques as well

as 4.1%Zn/HZSM-5 (80:1) and HZSM-5 (80:1) to establish the positive synergetic effect

that Pt and Zn have for these reactions. As previously mentioned, the enhanced

dispersion of metallic functions as promoted by Pt is one reason for improved

performance as well as the findings made by Mroczek et. al (U. Mroczek, W.

Reschetilowski 1991) where Pt-Zn/HZSM-5 resulted in the highest aromatization activity

in their investigation on the aromatization of ethane. It was concluded that Pt and Zn can

exhibit a stronger interaction between the corresponding metallic sites and so show a

higher activity towards aromatization. Despite this, the blocking of external sites deserves

future investigation using model compounds on the reaction system used here to try and

elucidate the reason for the interesting results observed.

54

Figure 3.6 The effect of external surface coverage on BTX selectivity and yield for promoter loaded Zn/HZSM-5 catalysts

55

3.3.5 Effect of methane

All reactions up to this point were conducted under a methane environment, therefore,

it was necessary to conduct a control experiment using a nitrogen environment to gain a

better understanding of the effect of methane presence. This was done over the Pt-

Zn/HZSM-5 (80:1) catalyst and the results are shown in Figs. 3.7 & 3.8. When looking

at the effect on liquid product distribution, fewer polyaromatics are formed under a

methane environment, possibly a result of a reduced environment upon CH4 activation

and increased generation of H2 which might stabilize the formed monoaromatics for

preventing its further condensation to produce polyaromatics. As well as this, it is a

possibility that methane incorporation finds preference on the benzylic sites of benzene

due to a slight increase in toluene and xylene components in the liquid product after

reaction under methane compared to nitrogen and a slight reduction in benzene formation

under methane, an observation made by other researchers during mechanistic studies

(He, Gatip, et al. 2017). As well as this, Fig. 3.8 demonstrates an increased liquid yield

upon performing the reaction under a methane environment, a likely result of methane

incorporation into the liquid products and an attractive reality for future industrial

applications.

Methane clearly enhances the performance of the catalyst in all aspects and is even

more successful using Pt-Zn/HZSM-5 than any other catalyst used here. Nonetheless,

a catalyst structure/performance relationship must be obtained to make more solid

conclusions about the reasoning for the superior performance of Pt-Zn/HZSM-5 over this

naphtha feedstock under a methane environment when compared to Zn/HZSM-5 and

their HZSM-5 support. Thus, XRD, TEM-EDX, CO Pulse Chemisorption (Table.3.2),

56

NH3-TPD, DRIFT and metal quantification techniques were employed to conduct these

investigations.

Figure 3.7 A comparison of different environments and their effects on the distribution of components in the liquid product using a Pt-Zn/ZSM-5 catalyst

57

Figure 3.8 A comparison of different environments and their effects on liquid and BTX yields using a Pt-Zn/ZSM-5 catalyst

3.4 Catalyst characterization

XRD patterns are displayed for all catalysts in Figs. 3.9 & 3.10, with XRD patterns for

Pt-Zn/HZSM-5, 4.1%Zn/HZSM-5, HZSM-5 and their spent counterparts shown in Fig.

3.11. Unfortunately, the metal loading is too low or dispersion is too high to witness the

characteristic diffraction peaks of Pt, Zn, and their oxidized forms, however, all catalysts

(fresh and spent) contain the peaks associated with HZSM-5. Although loading of metals

does reduce the crystallinity of the catalyst, the Pt promoter reduces the crystallinity to a

lesser extent when compared to the catalyst where only Zn is loaded, suggesting that Pt

can help to maintain crystallinity of these planes and thus, might increase performance.

58

As well as this, when only Zn is loaded onto the catalyst the peaks of the spent one are

reduced significantly, suggesting that Zn migration during the reaction may be occurring

and is affecting the structure of the catalyst. This is not the case when Pt is loaded as a

promoter, where the crystal structure remains similar before and after the reaction. As

well as this, Pt addition seems to provide more stability by comparing the diffraction

patterns before and after reactions when compared to HZSM-5 and 4.1%Zn/HZSM-5

catalysts. The notable decrease in lower angle peaks (approximately 7.5° and 9.5°) upon

loading of 4.1%Zn and Pt-Zn is likely a result of significantly decreased micropores within

the zeolite structure, a possible result of metal species blocking the inner pores

(Tshabalala and Scurrell 2015;Mikhailov et al. 2007).

Figure 3.9 XRD patterns of HZSM-5 and varying Zn metal loadings as well as their spent counterparts under a methane environment

59

Figure 3.10 XRD patterns of Zn/HZSM-5 and varying promoter loadings as well as their spent counterparts under a methane environment

60

Figure 3.11 XRD patterns of ZSM-5, Zn/ZSM-5, Pt-Zn/ZSM-5 and their spent counterparts under a Methane Environment

TEM is an effective microscopic technique used here (Fig. 3.12) to witness the

morphology of Pt-Zn/HZSM-5 after reactions under methane and nitrogen. In both cases,

fringing is witnessed in the images, showing that the crystalline form of the catalyst stays

intact after the reactions. Small particles can also be seen when referring to Fig. 3.12a

and 3.12b with average particle sizes of 1.7 nm for the catalyst under a methane

environment and 2.9 nm for the catalyst under a nitrogen environment. The small

particles are assumed to be that of Zn and/or Pt with smaller particle sizes attributed to

less agglomeration under a methane environment when compared to nitrogen, a likely

explanation for increased performance under methane. As well as this, a greater

61

dispersion of particles is seen under methane compared to nitrogen, leading to the

conclusion that despite platinum’s ability to improve dispersion on the catalyst, methane

also plays a role during the reaction that reduces metal agglomeration and enhances

dispersion, leading to enhanced performance.

NH3-TPD data is shown in Fig. 3.13 and Table 3.3 and allows the quantification of

acidic sites for the three catalysts in question. Two distinct peaks are observed in the

case of HZSM-5 at the low temperature region (210°C) and high temperature region

(420°C) and are attributed to desorption from weak acidic sites and strong acidic sites

respectively (Lónyi and Valyon 2001). Upon metal loading however, the peak associated

Figure 3.12 TEM images of the spent Pt-Zn/HZSM-5 catalyst collected under a) CH4 and b) N2 environments as well

as c) fringing of the Pt-Zn/HZSM-5 catalyst under CH4

62

with strong acidic sites is depleted and in fact shifted towards lower temperatures (360°C)

whereas the peak associated with weak acid sites is intensified with the addition of a

shoulder peak at approximate temperatures of 260°C. This indicates that some of the

strong acidic sites associate with the metal species to form weaker acidic sites, an effect

that clearly enhances the performance of the catalysts compared to HZSM-5. Again,

referring to Figs. 3.4 and 3.5 as well as the discussion made in section 3.42 and 3.43

metal loading in both cases (Zn and Pt-Zn) is shown to increase the catalyst performance

in terms of BTX yields, perhaps this is due to the reduction in strong acidic sites which

are replaced by weaker acidic sites in the form of Lewis sites provided by oxidized metal

species. In addition to this, it is suggested by Treesukol et al. (Treesukol et al. 2005) that

Pt loaded on HZSM-5 would not be stable without the presence of Brönsted acid sites

due to the redistribution of electron density in the Pt atom. This would help support the

maintained crystallinity of Zn-Pt/HZSM-5 during reaction compared to Zn/HZSM-5, as the

Brönsted acid sites help prevent agglomeration of Pt atoms but not in the case of Zn alone.

Furthermore, it is suggested that framework oxygens of the zeolite framework provide

electron density to the Pt atoms. This would in turn increase the acidity of strong acid

sites within the catalyst as is witnessed in Table 3.3 and Fig. 3.13. However, the nature

of ammonia does not allow us to discern the contribution of Brönsted and Lewis acid sites

and so to support this data. Therefore, DRIFT spectra (Fig. 3.14) were also obtained

using pyridine as the molecular probe. Peaks at 1445 cm-1 and 1614 cm-1 are attributed

to Lewis acid sites (Fermoso et al. 2016) with the peak at 1597 cm-1 being assigned to H-

bonded pyridine and the peak at 1492 cm-1 is an indication of both Brönsted and Lewis

acid sites (He, Gatip, et al. 2017). The peak at 1542 cm-1 is only evident on the HZSM-5

63

catalyst and is a result of Brönsted acid sites (Emeis 1993). This supports the NH3-TPD

data through an increase in Lewis acid sites when compared to Brönsted sites upon metal

loading, perhaps indicating the interaction of Zn species with Brönsted sites to form ZnO

species and/or [ZnOZn]2+(Hernando et al. 2017). Lewis acid site depression in the case

of Zn/HZSM-5 when compared to Pt-Zn/HZSM-5 is attributed to a poorer metal dispersion

in the case of Zn/HZSM-5, leading to the less concentrated formation of Lewis acid sites.

Figure 3.13 NH3-TPD profiles of a) pure ZSM-5, b) Zn/HZSM-5 and c) Pt-Zn/ZSM-5

catalysts

64

Table 3.3 The acidity of varying catalysts (µmol NH3/g cat.)

Catalyst Weak Acidic

Sites

Medium

Acidic Sites

Strong

Acidic Sites

Total

HZSM-5 (80:1) 546 0 504 1050

Zn/HZSM-5

(80:1) 307 533 403 1242

Pt-Zn/HZSM-5

(80:1) 249 493 442 1185

Figure 3.14 DRIFT spectra of ZSM-5, Zn/ZSM-5 and Pt-Zn/ZSM-5 using pyridine as the molecular probe for adsorption, followed by desorption at 400°C

65

3.5 Conclusion

Naphtha upgrading to valuable chemical intermediates is an important process for

industries to provide environmental and economic stability. The feasibility of this venture

is demonstrated in the form of a bimetallic heterogenous catalyst (Zn-Pt/HZSM-5) whose

performance is exemplary for the formation of BTX from a complex naphtha feed under

a methane environment. XRD, NH3-TPD, CO chemisorption and DRIFT analyses enable

the characterization of Pt-Zn/HZSM-5 so as to elucidate the reason for this superior

performance compared to conventional HZSM-5 and Zn/HZSM-5 by showing Pt’s ability

to promote metal dispersion, maintain catalyst crystallinity throughout reactions as well

as increase stability and provide the catalyst with the suitable acidic sites to enhance

selectivity to BTX. As well as this, methane as an alternative source of hydrogen, an

attractive prospect compared to current reformation techniques, is demonstrated here

probably in the form of methane incorporation into liquid products and the monoaromatics

stabilizing effect.

66

Chapter Four: Pt-Zn/HZSM-5 as a Highly Selective Catalyst for the Co-

aromatization of Methane and Light Straight Run Naphtha

This chapter is adapted from the following manuscript:

Jack Jarvis, Peng He, Aiguo Wang, and Hua Song. 2018. “Pt-Zn/HZSM-5 as a Highly

Selective Catalyst for the Co-aromatization of Methane and Light Straight Run Naphtha.”

Submitted to Fuel

67

4.1 Abstract

Selectivity of catalysts has become increasingly important in the petrochemical

industry in order to reduce waste, maximize efficiency, and reduce costs. Reported here

is a catalyst in the form of Pt-Zn/HZSM-5 with low acidity to reduce over cracking and

maximize selectivity of benzene, toluene, and xylene (BTX) during the aromatization of

naphtha under methane environment. The Pt-Zn alloy when coupled with HZSM-5

support is capable of BTX selectivity up to 94% with no formation of detrimental

polyaromatics and olefins when reacted under a methane environment during the

conversion of light straight run (LSR) naphtha. Such a performance is attributed to the

appropriate distribution of acid sites (seen via NH3-TPD and pyridine DRIFT) with regards

to strength and type because of loading Pt and Zn as well as Zn’s ability to reduce coke

formation as seen via TGA-DSC. TEM imaging shows that a methane environment

effectively suppresses metal particle agglomeration compared to nitrogen with XRD

showing Pt’s ability to maintain catalyst crystallinity during moderate temperature and

moderate pressure as well as reaction time of 1 hour. It has been seen here that by

employing Zn and Pt in tandem over HZSM-5, catalytic performance can be greatly

increased when compared to sole loading of Zn or Pt, thus a positive synergetic effect is

witnessed. Such a selective and coke resistant catalyst shows great promise for the future

of LSR reformation to valuable products with methane as a very feasible co-reactant

capable of contributing to positive liquid products.

68

4.2 Experimental

4.2.1 Catalyst synthesis

Previous studies conducted on heavier feedstocks such as standard petroleum

naphtha, whose composition contains larger hydrocarbon chains as well as some heavier

aromatics than the LSR naphtha shown here, showed that ZSM-5 acidity had varying

effects on the liquid product distribution. The most promising SiO2/Al2O3 ratio was that of

80:1 ZSM-5 (Jarvis et al. 2018), however, after conducting preliminary tests on LSR, it

became well established that a lower acidity was preferable. This is likely because the

components in this feedstock require isomerization and aromatization whereas a heavier

feed would require more of a cracking element to break down the larger components.

Thus, fewer acid sites were employed on all the catalysts used here by synthesizing

HZSM-5 (280:1).

280:1 molar ratio SiO2/Al2O3 MFI zeolitic support composite in its ammonium form was

obtained from Zeolyst International. The support was then converted to its hydrogen form

by calcination in air at 10 °C·min-1 and held at 110 °C for 8 hours before ramping at the

same rate again and holding at 300 °C for 3 hours. The support was finally subjected to

600 °C after ramping at 10°C·min-1 and held for a further 3 hours before allowing to cool

down naturally. To apply metal loadings of Zn, Pt as well as Pt and Zn together, the

catalysts were subjected to wetness impregnation where the calculated amounts of Zn

and Pt salts in the form of crystalline Zn nitrate hexahydrate [Zn(NO3)3.6H2O], 99% purity,

from Alfa Aesar and tetraammineplatinum(II) nitrate (≥50.0% Pt basis), from Sigma

Aldrich respectively were added simultaneously to reverse osmosis deionized (DI) water

(25 mL). The resulting solution was then stirred at room temperature (RT) at 500 rpm.

69

After complete dissolution, 5 g of the previously calcined catalyst support was added with

stirring and the resulting suspension remained stirring for another 3 hours where the

solution was then subjected to heating at 88°C until complete evaporation of water. The

resulting solid was then transferred to an 88°C oven and left to dry to completeness for

12 hours where it was then calcined using the same parameters as above. This technique

with relevant metal salts was used to obtain the following catalysts: 5%Zn/HZSM-5,

1%Pt/HZSM-5, and 5%Zn-1%Pt/HZSM-5.

4.2.2 Experimental procedure

Reactions were carried out in a fixed bed microreactor with a 1.46 mm i.d. and a length

of 54 cm with the catalyst bed volume being 9.3 mL. Xi’an Sino-Green Hi-Tech Co., Ltd

provided the fixed bed microreactor system-4100C. In a typical run, 1.6 g of catalyst was

placed in the reactor with a comparable quartz wool bed and packed with ceramic beads

to obtain reproducible catalyst bed volumes. The feedstock was prepared according to a

known industrial composition containing i-pentane, n-pentane, i-hexane, n-hexane and n-

heptane (discussed in section 2.4) and was pumped into the system at a rate of 1.6 g·h-1

to achieve a weight hourly space velocity (WHSV) of 1 g of feedstock per 1 g of catalyst

per hour for all reactions. The liquid was then joined by methane or nitrogen gas at a flow

rate of 50 standard cubic Centimeters per minute (SCCM) depending on the desired

environment of the reaction and consequently fed to the reactor. The rector was

pressurized to 3 MPa, ramped to 400 °C at a rate of 20 °C·min-1 and allowed to stabilize

for 30 mins before addition of liquid feedstock. Reactions were then carried out for 1 hour

under these conditions. The liquid products were collected via a post reactor condenser

70

system functioning at -18 °C. The reactor system was then shut down after purging with

methane/nitrogen for 30 mins and the liquid products were analyzed via Gas

Chromatography-Mass Spectrometry as supplied by PerkinElmer (PerkinElmer GC Claus

680 and MS Clarus SQ 8T). The column used was a paraffin/olefin/naphthene/aromatic

(PONA) one as provided by Agilent (HP-PONA). This allows the differentiation between

the aforementioned components as well as the different elution times of different carbon

number components to allow for a comprehensive and in depth understanding of the

matrix being analyzed. The column returns different response factors depending on the

species involved (paraffin, olefin, naphthene or aromatic) and so a calibration using

known standards was first conducted. The temperature program consisted of holding at

35 °C for 15 mins and then ramping at a rate of 1.5 °C·min-1 to 75 °C where it was

subsequently ramped to 150 °C at a rate of 3 °C·min-1 and then held for 30 mins. Finally,

the system was held for 30 mins at 250 °C after ramping at a rate of 3 °C·min-1. A 490

micro-GC from Agilent Technologies was used to analyze the gas product composition

on a molar basis during reactions. The micro-GC consists of 4 channels and is equipped

with thermal conductivity detectors. The detectors are capable of accurately analyzing

CO, H2, O2, N2 and CH4 in channel 1; CO2, C2H2, C2H4 and C2H6 in channel 2; as well as

paraffins and olefins between C3 and C6 in channels 3 and 4. A 10-m molecular sieve 5A

column, 10-m PPU column, 10-m alumina column and an 8-m CP-Sil 5 CB column were

equipped to the channels 1-4 in respective order. Carrier gases used were He and Ar for

channels 1 and 2-4 respectively.

71

4.2.3 Characterization techniques

A wide array of characterization techniques has been employed in this work as a

means for elucidating the contribution of catalysts in different environments from both a

pristine (pre- reaction) and spent (post reaction) perspective. Transmission Electron

Microscopy (TEM), Diffuse Reflectance Infrared Fourier Transform (DRIFT), X-ray

Powder Diffraction (XRD), Ammonia Temperature Programmed Desorption (NH3-TPD),

and Thermogravimetric Analysis-Differential Scanning Calorimetry (TGA-DSC) were the

techniques used to analyze catalysts resulting from the procedures given in sections 2.1

and 2.2.

TEM analysis was conducted on a FEI Tecnai F20 system with the FEG operating at

200 kV and a 4k Gatan CCD camera. Approximately 0.1 g of catalyst was first dispersed

in ethanol and supported on holey carbon on a 200 mesh Cu grid before the TEM images

were subsequently acquired. HRTEM images were obtained for fresh Pt-Zn/HZSM-5 and

the analysis was performed on FEI Talos F200X, a high-resolution scanning/ transmission

electron microscope (S/TEM) operated between 80 and 200 KV and equipped

with HAADF (high angle annular dark field) detector for Z contrast imaging and SuperX

EDS detector for compositional analysis.

A Hg/Cd/Te (MCT) detector attached to a Nicolet iS50 from Thermo Scientific was

used to perform DRIFT analysis of the catalyst samples. The analysis was performed by

placing the catalyst sample in a closed system and heating to 500 °C in a 30 SCCM flow

of N2 where it was then held for 15 mins. After this pretreatment, the system was cooled

to ambient RT whilst purging air from the surrounding environment using 35 standard

cubic feet per hour (SCFH) of industrial N2. Confirmation of complete CO2 removal was

72

achieved through rigorous testing followed by obtaining background spectra before

conducting the final tests themselves. Pyridine, an effective probe for this system given

its chemical attributes was applied to a bubbler and N2 gas subsequently flowed through

at a rate of 30 SCCM for 30 mins. Excess pyridine was removed after this by purging for

a further 30 mins with N2 whilst bypassing the pyridine bubbler. The Kubelka-Munk model

with 512 scans was then used at a resolution of 4 cm-1 and at 1 atm to collect the resulting

spectra. Bronsted and Lewis acid presence could then be qualified for the fresh catalysts.

XRD was used to examine the crystalline compositions through diffraction patterns

and was conducted on a Rigaku Multiflex Diffractometer with Cu Kα irradiation at 20 kV

and 40 mA in the 2θ region between 3° and 60°.

NH3-TPD was performed using the Chemisorption Analyzer provided by Sino-Green

Hi-Tech via use of He and N2 gases as carriers. 0.1-0.2 g of catalyst samples were loaded

to the system followed by a thorough degassing at 100°C for 1 hour under He flow at 20

SCCM. The temperature was then elevated to 600°C at a ramping rate of 10°C·min-1 and

held for 100 mins. The pure He stream was maintained at a flow of 20 SCCM whilst

allowing the system to cool to 120 °C. Post this process, a continuous flow of N2 gas

containing 0.5% NH3 was applied in place of He at a flow rate of 30 SCCM for 1 hour and

at 120°C to attain minimum physisorption of NH3 species whilst achieving chemisorption.

The system was then purged for another hour under He at 20 SCCM to remove any

physically adsorbed ammonia. Subsequent desorption occurred through finally heating

the system to 600°C at a ramping rate of 10°C·min-1 and holding for 30 mins.

Quantification of acid sites was then achieved through peak deconvolution.

73

TGA-DSC was used to assess the coke formation and stability of the catalysts used

and thus, assess the extent of deactivation. The instrument used was a simultaneous

thermal analyzer STA 6000 manufactured by PerkinElmer. Approximately 0.02 g of

catalyst sample was applied to the instrument and held at 30 °C for 1 min. The sample

was then heated to 800 °C at a rate of 20 °C/min under a 30 mL/min flow of extra dry air.

The resulting spectra enabled an accurate calculation of the mass of organics and coke

formed during the reactions.

4.2.4 Performance evaluation

well as the amount of gas produced which was calculated using a mass flow meter

post-reactor system. These values were then summed, and the gas law applied to obtain

the total moles of gas present in the reactor and expelled from the system. The micro-GC

as mentioned in section 4.33 then enabled the calculation of moles of gaseous

components both produced and remaining in the reactor due to the assumption of a

homogenous gas mixture throughout the process. Gaseous mass product as a result of

converted feed was then easily calculable. The GC-MS instrument described in section

2.2 allowed for the calculation of molar composition to which a coefficient was applied

given the different response factors acquired through careful instrumental calibration.

Component masses were then calculable and the selectivity of BTX was given as the

selectivity as well as mole % of BTX in the liquid product. Both equations are given here

as equation (4) and (5) respectively. Conversion, BTX yield and total liquid yield are given

as per the reaction equations shown in equations (1), (2), and (3).

74

𝐶𝑜𝑛𝑣𝑒𝑟𝑠𝑖𝑜𝑛 (%) = ((𝑀𝑎𝑠𝑠 𝑜𝑓 𝐿𝑖𝑞𝑢𝑖𝑑 𝐹𝑒𝑑−𝑀𝑎𝑠𝑠 𝑜𝑓 𝑈𝑛𝑟𝑒𝑎𝑐𝑡𝑒𝑑 𝐿𝑖𝑞𝑢𝑖𝑑)

𝑀𝑎𝑠𝑠 𝑜𝑓 𝐿𝑖𝑞𝑢𝑖𝑑 𝐹𝑒𝑑) × 100% (1)

𝐵𝑇𝑋 𝑌𝑖𝑒𝑙𝑑 (%) = (𝑀𝑎𝑠𝑠 𝑜𝑓 𝐵𝑇𝑋 𝑖𝑛 𝑃𝑟𝑜𝑑𝑢𝑐𝑡

𝑀𝑎𝑠𝑠 𝑜𝑓 𝐿𝑖𝑞𝑢𝑖𝑑 𝐹𝑒𝑑) × 100% (2)

𝐿𝑖𝑞𝑢𝑖𝑑 𝑌𝑖𝑒𝑙𝑑 (%) = (𝑀𝑎𝑠𝑠 𝑜𝑓 𝐿𝑖𝑞𝑢𝑖𝑑 𝑝𝑟𝑜𝑑𝑢𝑐𝑒𝑑

𝑀𝑎𝑠𝑠 𝑜𝑓 𝐿𝑖𝑞𝑢𝑖𝑑 𝐹𝑒𝑑) × 100% (3)

𝐵𝑇𝑋 𝑆𝑒𝑙𝑒𝑐𝑡𝑖𝑣𝑖𝑡𝑦 (%) = (𝑀𝑎𝑠𝑠 𝑜𝑓 𝐵𝑇𝑋 𝑃𝑟𝑜𝑑𝑢𝑐𝑒𝑑

𝑀𝑎𝑠𝑠 𝑜𝑓 𝑅𝑒𝑎𝑐𝑡𝑒𝑑 𝐿𝑖𝑞𝑢𝑖𝑑) × 100% (4)

𝐿𝑖𝑞𝑢𝑖𝑑 𝑃𝑟𝑜𝑑𝑢𝑐𝑡 𝐵𝑇𝑋 𝑆𝑒𝑙𝑒𝑐𝑡𝑖𝑣𝑖𝑡𝑦 (%) = 𝑚𝑜𝑙 % 𝑜𝑓 𝐵𝑇𝑋 𝑖𝑛 𝐿𝑖𝑞𝑢𝑖𝑑 𝑃𝑟𝑜𝑑𝑢𝑐𝑡 (5)

Each reaction was conducted in triplicate to obtain a suitable error associated with

each quantifiable value with the original composition of the replicate LSR provided in Fig.

4.1. Paraffinic isomers were only considered in the selectivity calculation if they were not

present in the original feedstock mixture.

75

Figure 4.1 Distribution of components found in LSR naphtha

4.3 Results and discussion

4.3.1 LSR naphtha conversion

The replicated LSR naphtha as shown in Fig. 4.1 was designed to recreate a typical

feedstock found in petrochemical industry and the catalyst is designed to convert the n-

paraffins to their isomers and/or valuable aromatics (nominally BTX). Thus, we see that

the main normal paraffins desired to be converted during this reaction are that of C5 and

C6 with a small amount of C7.

Fig. 4.2 shows the selectivity (equation (4)), conversion, and BTX yield results for the

reactions mentioned in section 2.2 using metal loading variations of HZSM-5 (280:1),

including no metal loading for comparison and the effect of methane and nitrogen

environments. Although sole Pt or sole Zn loading onto HZSM-5 have both shown

promise for the aromatization of paraffins in recent years, the improvements are often

76

‘small’ stepping stones with regards to improved performance. Fig. 4.2 immediately

shows us the marked improvement when both Pt and Zn are loaded in the form of yield

and selectivity. Upon initial inspection, the use of both metals under a methane

environment improve the selectivity drastically as well as the BTX yield. Not surprisingly,

a lack of metal loading, i.e. HZSM-5 (280:1) shows the worse performance, with sole Pt

and sole Zn showing slight improvements. As expected, sole Pt, with the exception of no

metal loading, shows the worse selectivity, likely a result of its strong capacity for

hydrogenolysis. Selectivity in Fig. 4.2 is defined as per equation (4) and so shows

selectivity in the form of total selectivity with the inclusion of gaseous products.

Figure 4.2 Graph displaying the performance of various HZSM-5 (280:1) catalysts with regards to BTEX selectivity, conversion, and BTEX yield, where BTEX selectivity is calculated as per equation (4)

77

For a clearer picture with regards to liquid selectivity and liquid yield, that is to say,

total amount of liquid recovered including feedstock for the latter, Fig. 4.3 and Fig. 4.4

are both provided where selectivity in the case of Fig. 4.3 is given as mol % of BTX in

liquid product as per equation (5). As expected, sole Pt gives the poorest selectivity,

barely surpassing that of conventional ZSM-5 with no metal loading, with sole Zn’s

promotion of oligomerization of gaseous C3 species as well as hydrogenation capacity

accounting for an increase in total liquid yield and increased formation of BTX. The

increased formation of BTX when only Zn is employed when compared to sole Pt

employment is likely a result of increased oligomerization of gaseous olefinic species, C2

in particular.

Figure 4.3 Graph displaying the performance of various HZSM-5 (280:1) catalysts with regards to BTEX liquid product selectivity and total liquid yield

78

Figure 4.4 Graph displaying the liquid product distribution of various HZSM-5 (280:1) catalysts

Fig. 4.5 shows gas product compositions and agrees with this assumption by

demonstrating a significant decrease in ethylene content when comparing sole Zn to sole

Pt loading, likely a result of a lack of Zn presence for effective oligomerization. Gas

products also show a higher mol % of hydrogen when Pt-Zn are loaded onto HZSM-5

which agrees with the higher liquid selectivity witnessed during the formation of BTEX

than seen for other catalysts.

79

Figure 4.5 Gaseous product distribution for varying HZSM-5 (280:1) catalysts on a fixed bed reaction system with LSR naphtha and methane or nitrogen

As well as witnessing the effects of metal loading on the catalyst, the presence of

methane as an atmosphere compared to nitrogen is also demonstrated in Figs. 4.2 to

Fig. 4.5. Despite reducing feedstock conversion; BTX selectivity, yield and even total

liquid yield understandably increase, leading to a much more desirable range of products

being in their liquid form as opposed to undesirable gaseous species. This is displayed in

Fig. 4.2 with the liquid yield of Pt-Zn/HZSM-5 under a nitrogen environment being much

lower than that when CH4 is employed, suggesting a positive effect of methane’s reductive

atmosphere as well as potential CH4 incorporation into the liquid products.

80

In summary, the Pt-Zn catalyst has shown a marked improvement in performance in

all aspects, regardless of environment, suggesting a potential synergetic effect and/or a

highly superior alloy that potentially takes the most positive attributes of the metals

employed. This could be through Pt acting as a dispersion promoter or even a result of

Pt effectively assisting dehydrogenation to olefins, with Zn oligomerizing, followed by

aromatization thanks to the morphology of ZSM-5 and the presence of dehydrogenating

metals.

4.3.2 Catalyst characterization

Fig. 4.6 shows TEM images of spent Pt-Zn/HZSM-5 under a methane environment,

a) and a nitrogen environment, b) where particle agglomeration can be seen with the

naked eye. An average particle diameter of 1.4 nm and 2.1 nm under methane and

nitrogen respectively was found, suggesting that the presence of methane inhibits particle

agglomeration, thus providing another explanation for the enhanced performance under

a methane environment due to a better distribution of active sites throughout the reaction.

Lattice fringing is seen, c), supporting XRD data which shows a maintained crystalline

structure during synthesis. Fig. 4.7 demonstrates the presence and dispersion of Zn and

Pt on fresh Pt-Zn/HZSM through HAADF elemental mapping.

81

Figure 4.6 TEM images of a) Pt-Zn/HZSM-5 post reaction under a methane environment, b) Pt-Zn/HZSM-5 post reaction under a nitrogen environment, and c) pristine Pt-Zn/HZSM-5 lattice fringing

Figure 4.7 HAADF STEM images of pristine Pt-Zn/ZSM-5 (280:1) with elemental mapping, red corresponds to Pt particle locations with green corresponding to Zn particle locations

82

XRD patterns are shown in Fig. 4.8 and all cases exemplify the MFI crystal structure

before and after reaction. However, upon closer inspection, it is evident that HZSM-5 with

no metal loading, experiences no significant change after reaction, likely a result of no

metal presence to migrate and cause crystal deformations during the reaction. Sole Pt

loaded catalysts experience an intensity reduction in the 010 crystal plane post reaction,

accompanied by an increase in intensity of 501, 151, and 303 crystal planes, suggesting

a difference in the number of atoms in the aforementioned planes (Ding et al. 2014). It is

therefore possible that Pt migration from the 010 crystal plane towards the 501, 151, and

303 crystal planes is occurring or that the migration of Pt during the reaction is shifting

other atoms in the framework. The direct opposite effect is seen for the Zn/HZSM-5

catalyst where 501, 151, and 303 peaks are greatly reduced post reaction and the 010

plane intensity is increased. When both Pt and Zn are employed (green diffractograms)

a nitrogen environment seems to slightly reduce 501, 151, and 303 crystal planes

whereas a methane environment increases these planes with a slight reduction in the 010

plane. Given that the presence of nitrogen as opposed to methane results in the formation

of more gaseous products, it is possible that the presence of metals on the active 501,

151, and 303 planes inhibits the formation of these undesired gaseous products. Perhaps

methane activation through metal interaction, in particular CH4-Pt interactions as seen on

Pt clusters by Xiao et al. (Xiao and Wang 2007) and which indeed see a reduced

activation energy, is promoted over the aforementioned planes and increases the stability

of Pt atoms at these positions with migration from the 010 plane, a result of a lack of

methane activation at this plane.

83

Figure 4.8 XRD patterns for varying HZSM-5 (280:1) catalysts

DRIFT spectra for adsorbed pyridine on HZSM-5, Zn/HZSM-5, Pt/HZSM-5, and Pt-

Zn/HZSM-5 are displayed in Fig. 4.9 and enable us to establish the nature of the acid

sites present on the catalyst surface. Three forms of acid sites are investigated here;

including Brönsted, Lewis, and physiosorbed H-bonded sites and so the region

84

investigated lies between 1440-1620 cm-1. Peaks observed at 1441 and 1451 cm-1

demonstrate Lewis acid sites (Sayed, Kydd, and Cooney 1984), i.e. electron pair

acceptors arising from molecularly coordinated pyridine (C-C stretch). The peaks at 1485

and 1492 cm-1 show us the contribution of both Brönsted and Lewis acid sites (Peng He

et al. 2017). Brönsted acid sites, nominally proton donors, can usually be observed at

approximately 1550 cm-1 (Emeis 1993), however, given the relatively high SiO2/Al2O3 ratio

employed for these catalysts (280:1) it is possible that typical Brönsted sites are difficult

to detect as a result of such a small amount of Al used to replace Si in the framework. We

are encouraged to recall that this is a desired attribute of our catalysts given the need for

less catalytic cracking. That being said, bands at 1586 and 1602 cm-1 are indicative of H-

bonded pyridine, a sign of Brönsted acidity (B. S. Liu et al. 2011). The small shoulder

peak noted at 1613 cm-1 is that of pyridine adsorption on Lewis acid sites (Lønstad Bleken

et al. 2013). Intensities see little to no change given that all experiments were conducted

on the same support and at RT with the exception of Pt loading resulting in a reduced

peak intensity compared to Zn, which could be a result of Pt’s high dispersion, and uniform

distribution leading to the occupation of more acid sites. However, more notable changes

come from the comparison of Zn containing catalysts where all peaks are enhanced along

with shoulder peaks in some cases (1451, 1492, and 1613 cm-1) all of which are only

present when Zn is employed. Since all peaks can be attributed to Lewis acid sites, it is

suitable to conclude that Zn presence has increased the Lewis acidity compared to

Bronsted acidity in all cases. This is understandable given the higher loading of Zn and

that metal replacement of Si sites would result in such a phenomenon.

85

Figure 4.9 Pyridine DRIFT spectra of varying HZSM-5 (280:1) catalysts collected at RT

To further acidic inspection of the catalysts, and to even quantify the acidity of the

catalysts used, NH3-TPD was employed with results shown in Fig. 4.10 and Table 4.1.

Two distinct peaks are witnessed when no metal is loaded and when sole Pt is loaded

onto HZSM-5, with peaks at approximately 150 and 350 °C showing desorption from weak

and strong acidic sites respectively. When comparing spectra a) and b) it is clear that

upon Pt loading, the contribution made by weaker acid sites is reduced significantly and

strong acid sites conversely increased. In previous work (Jarvis et al. 2018) as well as

later on in this piece, it has been concluded that an increased performance through

86

selectivity and yield is closely associated with metal loading resulting in strong acid sites

associating themselves with metal species to form weaker acidic sites. This helps explain

why sole Pt loaded onto HZSM-5 results in one of the worst increases in performance

when compared to Zn/HZSM-5 and Pt-Zn/HZSM-5. As seen in our previous work (Li, He,

et al. 2018), and upon inspection of Fig. 4.10 c) and d), strong acid sites have been

dissociated and resulted in the formation of medium acid sites when Zn is loaded. This

phenomenon is enhanced when sole Zn is employed as opposed to Pt-Zn, an expected

occurrence given Pt’s tendency to promote dispersion and thus result in a reduced effect

on acidity. The addition of Pt to Zn has increased the contribution of strong acid sites in

comparison to sole Zn loading which suggests that framework oxygens of the support

help provide electron density to Pt. The reader is also referred to Table 4.2 where

quantified values of acidity can be inspected in the form of x 10-5 mol NH3/g cat

(Rodríguez-González et al. 2007).

It is also important to probe the deactivation tendency of catalysts to assess their

ability to perform for long durations in a reactor system given that less time spent

regenerating catalysts is desired. Given the lack of sulfur containing components in this

system, the main focus of catalyst deactivation is that of carbon deposits, coke. Thus,

TGA-DSC analysis has been conducted so as to gauge the level of deactivation imposed

on the catalysts. Table 4.2 shows the rate of coke formation on the catalysts as obtained

from the liquid feedstock. Not surprisingly, all catalysts under both CH4 and N2

environments exhibited very low coke yields because of little to no formation of

polyaromatics (known precursors to coke). Another contributor to such impressively low

deactivation is the low acidity of the support used, minimizing cracking levels. Pt is

87

notorious for its promotion of coke formation, given its less selective tendencies and so it

is not surprising to see that it is responsible for higher levels of coke formation than all

other catalysts, including non-metal. Sole Zn loading shows the lowest coke yield which

has been seen by other groups (Rasouli and Yaghobi 2018). Interestingly, although the

Pt-Zn catalyst results in higher levels of coking than sole Zn, the performance is clearly

not inhibited, suggesting that the level of coke is not significant enough to block active

sites on the Pt-Zn catalyst.

Figure 4.10 NH3-TPD spectra of a) HZSM-5, b) Pt/HZSM-5, c) Zn/HZSM-5, and d) Pt-Zn/HZSM-5

88

Table 4.1 Performance data on various catalysts during their upgrading of LSR. BTX selectivity is given in terms of liquid product selectivity and total selectivity, columns 2

and 3 respectively, with equations provided in section 4.34

Catalyst BTX

Selectivity

(%)

BTX

Selectivity

(%)

BTX

Yield

(%)

Conversion

(%)

HZSM-5

(280:1)

82 16 11 67

Zn/HZSM-5

(280:1)

87 20 13 66

Pt/HZSM-5

(280:1)

84 19 13 70

Pt-Zn/HZSM-5

(280:1)

94 36 22 62

Pt-Zn/HZSM-5

(280:1), N2

93 23 20 90

89

Table 4.2 The acidity of varying catalysts via NH3-TPD (x 10-5 mol NH3/g cat.) and coke

yield (μg of coke formed·h-1) calculated via TGA-DSC

Catalyst Weak

Acidic Sites

Medium

Acidic

Sites

Strong

Acidic

Sites

Total μg coke

/ (g

cat.h)

HZSM-5 (280:1) 3.5 0 7.1 10.6 5.9

Zn/HZSM-5

(280:1)

18.0 11.6 7.4 37.1 1.0

Pt/HZSM-5

(280:1)

4.6 0 32.6 37.2 8.3

Pt-Zn/HZSM-5

(280:1)

11.1 9.1 11.0 31.3 1.7

Pt-Zn/HZSM-5

(280:1), N2

- - - - 2.4

90

4.4 Conclusion

LSR feedstocks can contain large quantities of undesirable normal paraffins whose

octane numbers are not sufficiently high to provide useful characteristics to combustion

fuels. With respect to the model feedstock provided here by Chambroad petrochemicals,

it has been shown that it is feasible to upgrade such paraffins to more desirable i-paraffins

as well as BTX. Of which, the latter not only contributes to boosting fuel octane numbers

but also can be used as industrial chemical feedstocks of a higher value than their

upgraded feeds. A catalyst in the form of Pt-Zn/HZSM-5 with low acidity has been used

to isomerize and aromatize this LSR into the aforementioned products with superior

performance with respect to total liquid yield (63%), BTX selectivity (94%) as well as BTX

yield (22%). No polyaromatics or olefins are produced during the reaction, meaning that

any other products formed (of which there are few) will not be detrimental to the product

mixture during transportation or utilization. Not only is this experimental data promising

for the future of LSR reforming, but it has also been conducted under a CH4 environment

which has shown to reduce coking, enhance liquid yield and increase BTX selectivity,

likely a result of methane incorporation into the products, when compared to N2. Such

results mean great promise for the future of light oil reformation with potential use of CH4,

the main component of natural gas as opposed to expensive and non-abundant H2

91

Chapter Five: Selective Aromatisation and Isomerisation of n-Octane from

Intermetallic Pt-Zn Nanoparticle Alloys Supported on a Uniform Aluminosilicate

This chapter is provided as a summary of an initial study into a highly selective catalyst

and is not yet submitted to a journal:

Jack Jarvis, Peng He, Aiguo Wang, Jonathan Harrhy, Shijun Meng, and Hua Song.

2018

92

5.1Abstract

ZSM-5 is a leading candidate for present and future applications in the petrochemical

industry, however, selectivity’s towards desirable products (i-alkanes and BTX) are not

as high as they should be. Reported here is a bimetallic-support interaction through Pt-

Zn alloy nanoparticles and uniform ZSM-5 bulk particles to provide a system that is

selective up to 96% towards BTX and i-octane formation from n-octane conversion. This

combination provides an increased amount of external surface alloy nanoparticles whose

concentration increases as the reaction proceeds through particle migration. Both Pt and

Zn on UZSM-5 exhibit a stronger interaction with the support as opposed to Pt and Zn

loaded on CZSM-5, resulting in slightly reduced oxidation states that prove invaluable

when in working in tandem with the morphological properties of the support.

93

5.2 Experimental

5.2.1 Synthesis of UZSM-5

UZSM-5 was synthesised by hydrothermal synthesis. Al(NO3)3·9H2O 98% from Alfa

Aesar was added to 1.0 M Tetrapropylammonium hydroxide (TPAOH) from Sigma Aldrich

and stirred at room temperature (RT) until a clear solution was obtained. Tetraethyl

orthosilicate (TEOS) from Merck KGaA was then added dropwise to the above solution

whilst maintaining stirring at RT. Upon completion of TEOS addition, the solution was left

to stir at RT until supersaturation after approximately 1 hour. The resulting supersaturated

gel was applied to an autoclave and treated in a furnace at 170 °C for 72 hours. Amounts

were calculated to obtain molar ratios of Al2O3:80SiO2:30TPAOH:1323H2O in the gel. The

resultant crystals were then applied to an 88°C oven for 12 hours and subsequently

calcined in air at a rate of 5°C/min, held at 120°C for 1 hour, ramped at the same rate

again, held at 300°C for 1 hour and finally ramped at the same rate to 600°C and held for

3 hours. The resultant catalyst was then ready for reaction.

5.2.2 Synthesis of Pt-Zn/UZSM-5

Pt-Zn/UZSM-5 was synthesized by wetness impregnation (WI) where

tetraammineplatinum (II) nitrate from Sigma Aldrich and Zn nitrate hexahydrate from Alfa

Aesar were added to deionised water (DI) and stirred at RT. Prepared UZSM-5 support

was then added to the solution aimed at a 5%Zn-1%Pt/UZSM-5 final product. The mixture

was stirred at RT for 3 hours and then the DI water removed by heating at 88 °C. The

resulting catalyst was then dried at 88°C for 12 hours once DI water was removed and

calcined as before. The same procedure was carried out for Pt-Zn/CZSM-5 synthesis

94

using ZSM-5 (80:1) as provided by Zeolyst international. The same procedures were also

used for Zn/UZSM-5 and Pt/UZSM-5 synthesis.

5.2.3 Catalyst performance experimental procedure

In a typical experiment, 0.5 g of catalyst was loaded to a 100 mL Parr batch reactor

system with 3 g of n-octane (98% from Sigma Aldrich) applied to a vial to ensure reactant

contact with the catalyst in the gas phase. The reactor was then pressurised to 430 psig

with CH4 (99.9998%) from Praxair after sufficient purging of air. The system was then

heated at a rate of 20°C/min until 400°C where the temperature was held for 1 hour. Upon

completion, the system was cooled naturally. The liquid product was extracted and diluted

with CS2 and applied to a carefully calibrated gas chromatography-mass spectrometer

(GC-MS) as supplied by PerkinElmer (GC Claus 680 and MS Clarus SQ 8T). Gas

products were analysed using a 490 micro-GC as supplied by Agilent Technologies.

5.2.4 Catalyst characterisation procedures

HAADF-STEM images were obtained on a FEI Talos F200X, a high-resolution

scanning/ transmission electron microscope (S/TEM) operated between 80 and 200KV

and equipped with HAADF (high angle annular dark field) detector for Z contrast imaging

and SuperX EDS detector for compositional analysis. XRD images were procured using

a Rigaku Multiflex Diffractometer with Cu Kα irradiation at 20 kV and 40 mA in the 2θ

region between 3° and 60°. XAS was conducted with an SM beamline (10ID-1) at the

Canadian Light Source (CLS), which is equipped with a 35 nm outermost-zone plate

(CXRO, Berkeley Lab). The diffraction-limited spatial resolution for the zone plate is 30 nm.

95

Image sequence (stack) scans over a range of photon energies were acquired for the

same sample region at the Zn L-edge. XPS spectra were obtained using a Kratos Axis

spectrometer with monochromatized Al Kα (hυ = 1486.71 eV) with a base pressure of

∼5 × 10−10 Torr in the analysis chamber where binding energies were then referenced to

the C1S energy at position 284.1 eV.

5.3 Results

5.3.1 n-Octane aromatization and isomerization performance of Pt-Zn/UZSM-5

Experimental procedures for the reported reactions are described in the methods

section. Contrary to CZSM-5, UZSM-5 provides exemplary selectivity towards same

carbon number aromatics and i-alkanes when both supports are loaded with Pt-Zn and

compared. Fig. 5.1 demonstrates the selective reaction pathways observed in the liquid

phase where only the same carbon number isomers and aromatics of the reacted n-

alkanes are witnessed.

96

Figure 5.1 Reaction pathways for a) n-octane, b) n-heptane, c) n-hexane, and d) n-pentane, when Pt-Zn/UZSM-5 is utilized as the catalyst

A direct comparison of Pt-Zn/UZSM-5 and Pt-Zn/CZSM-5 is made (Fig. 5.2) and the

product selectivity is made clear with a strong preference towards xylene formation

(cyclisation and aromatisation of n-octane reactant). Pt-Zn/CZSM-5 on the other hand,

has a much broader product distribution with no formation of i-alkanes, showing the

typical occurrence of over aromatisation with CZSM-5 where cracking occurs, followed

by subsequent oligomerisation and aromatisation to form aromatics with higher carbon

numbers (undesirable products in the petrochemical industry).

7

97

Figure 5.2 Performance comparison between Pt-Zn/UZSM-5 and the conventional alternative where reactions were conducted as per the methods section *i-octane and BTX selectivity

Given the higher energy typically required to perform aromatisation as opposed to

isomerisation, the performance of Pt-Zn/UZSM-5 was tested at lower temperatures to see

if the same levels of total selectivity (i-alkane and BTX) are witnessed. Interestingly, the

total selectivity remains similar with an understandable trend towards preferred i-octane

formation at lower temperatures (Fig. 5.3), demonstrating the ability of this catalyst to

promote i-alkane and aromatic formation within the temperature ranges shown. The

relatively insignificant change witnessed for total selectivity suggests that the only

inhibiting factor here was the temperature requirement for formation of aromatics. The

98

performance of this catalyst shows exemplary same-carbon number aromatic and i-

alkane formation for n-octane conversion and is tested further for other n-alkanes,

nominally n-heptane and n-hexane (Fig. 5.4). As shown, the same trend is witnessed for

other n-alkane conversions. However, aromatisation is preferred to isomerisation with

increasing carbon number. This could be due to the increased stability through methyl

group presence promoting hyperconjugation of the resonating structures. The

performance of Pt-Zn/UZSM-5 has been shown and compared to Pt-Zn/CZSM-5 and we

now move to ascertain the reasoning behind such a selective catalyst.

Figure 5.3 A comparison of temperature changes when n-octane is employed as the reactant over Pt-Zn/UZSM-5

99

Figure 5.4 Demonstrating the isomerisation and aromatisation capability of Pt-Zn/UZSM-5 using different carbon number n-alkanes at 30 bar initial pressure of methane and 400 °C for 1 hour. 1 Aromatic refers to xylene in the case of n-Octane, toluene in the case of n-heptane, and benzene in the case of n-hexane. 2 i-alkane refers to C8 isomers in the case of n-octane, C7 isomers in the case of n-heptane, C6 isomers in the case of n-hexane, and C5 isomers in the case of n-pentane. 3 Total refers to the combination of aromatics and i-alkanes

100

5.3.2 UZSM-5 morphology

Preparation of the supports and their respective metal loadings are summarised in

the methods section. X-ray diffraction (XRD) patterns of both UZSM-5 and CZSM-5

shown in Fig. 5.5 confirm the MFI type framework of the catalysts. Differences in

intensities are attributed to the preferred orientation phenomenon as seen for well-

ordered plate-like crystals, hence the increased intensity of UZSM-5 (Hyett, Green, and

Parkin 2006).

Figure 5.5 XRD patterns of UZSM-5 and CZSM-5, indicating typical MFI crystal structures

101

High-angle annular dark-field-scanning transmission electron microscopy (HAADF-

STEM) images of UZSM-5 (Fig. 5.6a) and CZSM-5 (Fig. 5.7a) show uniform plate-like

crystals and disordered, larger crystals respectively. Elemental mapping (Fig. 5.6b-d and

Fig. 5.7b-d) indicates high dispersion for both supports when loaded with Pt and Zn with

tentative evidence of higher concentrations of Pt accompanied by Zn clusters, suggestive

of intermetallic behaviour. Support bulk sizes are decisively decreased (>2.5 x decrease)

in the case of UZSM-5 due to the controlled synthesis technique (Fig. 5.6e and Fig. 5.7e)

accompanied by the same trend for particle size variations for the supports (Fig. 5.6f and

Fig. 5.7f).

Figure 5.6 HAADF-STEM images and analysis of Pt-Zn/UZSM-5

102

Figure 5.7 HAADF-STEM images and analysis of Pt-Zn/CZSM-5

5.3.3 Characterization of Pt-Zn/UZSM-5

X-ray photoelectron spectroscopy (XPS) of the Zn 2p3/2 region for varying metal

loaded supports (Fig. 5.8), where spent refers to catalyst after reaction, shows typical

peaks of ZnO species in all case. It is typical for wet impregnated ZSM-5 to contain Zn2+

species in the form of isolated Zn2+ and [ZnOZn]2+ cations (Gabrienko et al. 2017).

103

Figure 5.8 Zn 2p XPS spectrum of 4 catalysts to gauge the effect of intermetallic contribution to the states of Zn2+

The small shift towards 1022.3 eV in the case of Pt-Zn/CZSM-5 would indicate

Zn(OH)2 if the shift was more significant, suggesting that these species did not

decompose to ZnO species as expected during the calcination step as seen in the

methods section, thus, this conclusion can be excluded. The binding energy (BE) shift of

Pt-Zn/CZSM-5 should therefore be attributed to stronger binding of Zn2+ species to the

oxygens of the zeolite framework, understandably resulting in a pronounced oxidation

state and increased BE. The reason for this stronger interaction between Zn2+ species

104

and the support framework could be a result of the change in support morphology,

nominally the larger and less uniform bulk sizes of CZSM-5 support. XPS spectra of the

Pt 4f7/2 and Al 2p regions (Fig. 5.9) have been deconvoluted to provide reliable separation

of overlapping regions.

Figure 5.9 Pt 4f7/2 XPS spectra of 4 catalysts to gauge the effect of intermetallic contributions to the states of Pt metal and PtO species

All cases show PtO and Pt metal species, typical of Pt loaded ZSM-5 materials

(Sarkar et al. 2012). There is a clear negative shift of Pt 4f7/2 peaks in the case of UZSM-

5 supports, emphasized by the reduced broadening of the Al 2p peaks when compared

to CZSM-5, which compliments the previously mentioned Zn 2p spectra where a stronger

metal support interaction is the probable cause of such a shift. As a surface technique,

XPS can also be used tentatively as a means for determining the concentration of species

105

at the surface. Interestingly, Pt-Zn/UZSM-5 indicates a reduced amount of Pt on the

external surface compared to Pt/UZSM-5 and Pt-Zn/CZSM-5. Pt concentrations are then

increased after reaction on Pt-Zn/UZSM-5, suggesting Pt presence on the internal pores

of the material which then migrates to the external surface post reaction. This

phenomenon is not witnessed in the case of Pt/UZSM-5, enabling the conclusion that the

presence of Pt-Zn intermetallic species are contributing to the differing concentration of

PtO and metallic Pt on the surface of UZSM-5. This observation is reversed in the case

of the Zn 2p spectra where Zn2+ concentrations are visibly reduced on the surface of the

catalyst in the case of Pt-Zn/CZSM-5. However, Zn migration to the external surface of

the catalyst is also occurring during the reaction, an occurrence witnessed in previous

research in the case of Zn. X-ray absorption near-edge structure (XANES) spectra were

collected with a focus on the Zn L-edge energy region (Fig. 5.10) to gain information on

local atomic structures through the very sensitive technique. The white line intensity of

the CZSM-5 support is higher than that of UZSM-5 supports in both cases (Zn and Pt-Zn

loading) with UZSM-5 white line intensities being almost identical. This suggests the

presence of less-filled d-band Zn species in the case of CZSM-5, hence the increased

oxidation as seen via XPS. The pre- edge could also indicate a higher concentration of

oxidized Zn species to total Zn species in CZSM-5 supports when the increased

intensities are considered. The main edge is significantly reduced in intensity for all cases

when compared to the pre-edge due to the filled d-band of Zn. However, Pt-Zn/UZSM-5

shows a slight intensity increase, perhaps a result of charge transfer from Zn’s d-band to

Pt or potentially the support itself. Zn/CZSM-5 also shows a significant pre- edge energy

shift towards lower photon energy’s, indicative of a change in Zn’s bonding characteristics,

106

nominally an increased bond length, again, in line with XPS observations. This is no

longer witnessed when Pt-Zn is employed in the case of CZSM-5, suggesting an

interaction between Pt and Zn to reduce bond lengths. A similar trend is witnessed for the

multiple scattering region.

Figure 5.10 XANES spectra on the Zn L-edge region of 4 catalysts with magnifications on the pre-edge region and multiple scattering regions

5.5 Discussion and conclusion

UZSM-5 when loaded with Pt-Zn has suggested a reaction pathway where

oligomerisation does not occur (Fig. 5.11). This is because CZSM-5 produces a far more

variable product distribution with differing carbon numbers to that of the reactant, pointing

to the occurrence of cracking, followed by dehydrogenation and oligomerisation to give

107

many carbon number variants. This is clearly not happening in the case of UZSM-5 where

the only products hint at the occurrence of isomerisation, dehydrogenation, and

aromatisation without excessive cracking as is the case of CZSM-5. However, small

amounts of benzene and toluene are seen in the case of UZSM-5, this could be from the

removal of methyl groups from the xylene produced or, more likely, the small occurrence

of oligomerisation for any products that are produced from cracking. In the case of CZSM-

5, the condensation of any two alkenes that are formed to then result in a higher alkene

leads to a random distribution of higher, and lower carbon number, aromatics. This results

in an unpredictable product distribution, an undesirable attribute in an industry which

relies on certainty when using the products formed from naphtha reformation.

Figure 5.11 Preferred reaction pathway of CZSM-5 and UZSM-5 catalysts when loaded with Pt-Zn nanoparticles. UZSM-5 avoids cracking and oligomerisation to give a far more selective product

108

To summarise, a highly selective catalyst has been observed with the unique ability

to convert n-alkanes to their respective i-alkanes and aromatic forms of the same carbon

number. This performance has been attributed to two main factors: the uniform

aluminosilicate support (ZSM-5), and the bimetallic Pt-Zn. UZSM-5 shows more uniform

bulk particle sizes with less variance in available ranges as well as an increased external

surface area supported by Table 5.1. The bimetallic Pt-Zn nanoparticles promote positive

electronic changes and exemplify different properties on the supports (CZSM-5 and

UZSM-5). It is suggested that the external surface of the catalyst might be more selective

for these reactions and so the presence of more Pt-Zn bimetallic nanoparticles on the

external surface in the case of UZSM-5 might be the main contributor to increased catalyst

performance. Through controlled synthesis techniques, the realisation of highly selective

catalysts for the aromatisation and isomerisation of undesirable n-alkanes has been

achieved.

Table 5.1 Nitrogen physisorption of CZSM-5 and UZSM-5 to provide BET surface area’s

BET Surface Area (m2/g)

Catalyst External Microporous Total

Conventional HZSM-5 109 292 401

Hydrothermal HZSM-5 176 171 347

109

Chapter Six: Conclusion and Recommendations

There’s no talent, neither genius without hard work

Dmitri Ivanovich Mendeleev

Given that each chapter has a conclusion specific to its case, this chapter will provide

a broader conclusion as well as recommendations for future work.

6.1 Conclusions

A series of naphtha feedstocks were evaluated with extensive studies conducted on

conventional naphtha and light straight run naphtha, followed by a model compound study

using a variation of the synthesized catalyst for converting n-octane to BTX and its

isomers. After these studies, it is conclusive that a combination of Pt-Zn intermetallic

nanoparticles on ZSM-5 is one of the superior materials for the upgrading of naphtha

feeds. This is because of the electronic changes occurring on these metals as well as a

unique level of acidity provided to the support upon the loading of these metals. These

catalysts and feedstocks were subjected to batch reactor and continuous fixed bed flow

reactor systems to provide an idea of their applicability on an industrial scale. Methane

was compared to nitrogen-controlled experiments and it was seen that methane provides

an improved selectivity through potential incorporation into liquid products, reduced metal

particle agglomeration, and reduced coking.

When heavy naphtha (C6-C12) was employed as the co-feed with methane, the

following conclusions were made:

• Pt promotes metal dispersion, assessed by pulsed CO chemisorption.

110

• Catalyst crystallinity was maintained throughout the reaction, obtained by XRD.

• Pt-Zn presence provided the necessary ratio and strength of acidic sites to

promote the reaction.

• Methane presence increased liquid yield and BTX yield compared to nitrogen

environment, possibly a result of methane incorporation into the liquid products.

Light naphtha in the form of LSR (C5-C7) was also employed as a co-feed with

methane over Pt-Zn/ZSM-5 with the following conclusions:

• Reduced acidity is more effective for the upgrading of LSR compared to

heavier naphtha.

• No formation of polyaromatics or olefins is witnessed, meaning that any

unwanted products are not detrimental to the processes used.

• Methane environment reduces coking, enhances liquid yield, and increases

BTX selectivity compared to nitrogen environment.

• Pt-Zn presence provides desirable distribution and strength of acid sites once

again.

Preliminary testing was conducted on n-octane as a model compound using a novel

synthesis technique where controlled particle sizes are possible. The following

conclusions were made during this work:

• Pt-Zn/UZSM-5 is capable of pure isomerization and aromatization of the model

compound to its same carbon number products, negating the common

111

problem found with CZSM-5 of over aromatization and formation of higher end

aromatics (undesirable).

• Two main factors contribute to this selectivity: The uniform aluminosilicate

support, and the bimetallic synergy of Pt-Zn.

• A possible mechanism where no cracking occurs, making the need for

oligomerization redundant.

• Reduced bond lengths when Pt-Zn is employed on UZSM-5 and compared to

Pt-Zn loading on CZSM-5, suggesting a stronger interaction with the support

in the case of Pt-Zn/UZSM-5.

6.2 Recommendations for future work

1. Although the effect of methane is witnessed in this work, I suggest monitoring

methane through isotopic labelling (D and C13) to trace the evolution of

methane fractions during the reactions. This will be conducted in a batch

reactor system with a predetermined pressure of C13H4, followed by a makeup

of CH4 to desired experimental pressure. C13 NMR can be used to analyze the

liquid products and ascertain the locations of C13 species in the liquid product.

2. All feedstocks used here were provided without sulfur contaminants, although

coke was investigated as a deactivating mechanism on these catalysts, the

effect of sulfur species should also be investigated. H2S is too toxic to be used

in current laboratory setup so suggested alternatives as sulfur poisons come

in the form of diethyl sulfide or dibenzothiophene, or a combination thereof.

Active sites that are affected by coking and sulfur poisoning should be

112

established. For example, it is expected that carbon deposits will form on the

acid sites with sulfur poisoning having a greater effect on the active metal sites.

3. Now that the effect of Pt-Zn has been established as a successful intermetallic

species on ZSM-5, effect of particle size through different synthetic techniques

should be investigated, as well as the effect of varying catalyst bulk sizes.

Single atom alloying is at the forefront of metal loading in heterogeneous

catalysis and has shown great success with Cu-Pt. Implementing the same

technique (Galvanic replacement, GR) is feasible with Zn. However,

maintaining an inert atmosphere to prevent oxidation of Zn is paramount to

achieving single atom alloying with Zn and Pt.

4. The process should be scaled up: similar investigations should be conducted

on a large-scale flow reactor system. Technical issues come when extruding

the catalyst. Items to be taken into consideration include binding material,

pellet size, and pellet shape to optimize surface area and prevent pressure

drops in a larger system.

113

References

Abdelsayed, Victor, Mark W. Smith, and Dushyant Shekhawat. 2015. “Investigation of the Stability of Zn-Based HZSM-5 Catalysts for Methane Dehydroaromatization.” Applied Catalysis A: General 505. Elsevier B.V.: 365–74. https://doi.org/10.1016/j.apcata.2015.08.017.

Ahuja, Ritu, Benudhar Punji, Michael Findlater, Carolyn Supplee, William Schinski, Maurice Brookhart, and Alan S. Goldman. 2011. “Catalytic Dehydroaromatization of N-Alkanes by Pincer-Ligated Iridium Complexes.” Nature Chemistry 3 (2). Nature Publishing Group: 167–71. https://doi.org/10.1038/nchem.946.

An, K., Alayoglu, S., Musselwhite, N., Na, K. & Somorjai, G. A. Designed catalysts from Pt nanoparticles supported on macroporous oxides for selective isomerization of n-hexane. J. Am. Chem. Soc. 136, 6830–6833 (2014).

Argauer, Robert J. 1972. “United States Patent Office,” 19–22.

Armbrüster, M., Schlögl, R. & Grin, Y. Intermetallic compounds in heterogeneous catalysis - A quickly developing field. Sci. Technol. Adv. Mater. 15, (2014).

Arribas, M. A., and A. Martínez. 2001. “Simultaneous Isomerization of N-Heptane and Saturation of Benzene over Pt/Beta Catalysts - the Influence of Zeolite Crystal Size on Product Selectivity and Sulfur Resistance.” Catalysis Today 65 (2–4): 117–22. https://doi.org/10.1016/S0920-5861(00)00576-9.

Au, Chak-Tong, Ching-Fai Ng, and Meng-Sheng Liao. 1999. “Methane Dissociation and Syngas Formation on Ru, Os, Rh, Ir, Pd, Pt, Cu, Ag, and Au: A Theoretical Study.” Journal of Catalysis 185 (1): 12–22. https://doi.org/10.1006/jcat.1999.2498.

Bai, X., Samanta, A., Robinson, B., Li, L. & Hu, J. Deactivation Mechanism and Regeneration Study of Ga-Pt Promoted HZSM-5 Catalyst in Ethane Dehydroaromatization. Ind. Eng. Chem. Res. 57, 4505–4513 (2018).

Bai, Xinwei, Anupam Samanta, Brandon Robinson, Lili Li, and Jianli Hu. 2018. “Deactivation Mechanism and Regeneration Study of Ga-Pt Promoted HZSM-5 Catalyst in Ethane Dehydroaromatization.” Industrial and Engineering Chemistry Research 57 (13): 4505–13. https://doi.org/10.1021/acs.iecr.7b05094.

Balakrishnan, K., A. Sachdev, and J. Schwank. 1990. “Chemisorption and FTIR Study of Bimetallic PtAu/SiO2 Catalysts.” Journal of Catalysis 121 (2): 441–55. https://doi.org/10.1016/0021-9517(90)90252-F.

Bauer, Frank, Karsten Ficht, Marko Bertmer, Wolf-Dietrich Einicke, Thomas Kuchling, and Roger Gläser. 2014. “Hydroisomerization of Long-Chain Paraffins over Nano-Sized Bimetallic Pt–Pd/H-Beta Catalysts.” Catal. Sci. Technol. 4 (11). Royal Society of Chemistry: 4045–54. https://doi.org/10.1039/C4CY00561A.

Biswas, J. & Maxwell, I. E. Recent process- and catalyst-related developments in fluid catalytic cracking. Appl. Catal. 63, 197–258 (1990).

114

BP Global. 2018. “Natural Gas Production | Natural Gas | Statistical Review of World Energy | Energy Economics | BP.” 2018. https://www.bp.com/en/global/corporate/energy-economics/statistical-review-of-world-energy/natural-gas/natural-gas-production.html.

Cano, Zachary P., Dustin Banham, Siyu Ye, Andreas Hintennach, Jun Lu, Michael Fowler, and Zhongwei Chen. 2018. “Batteries and Fuel Cells for Emerging Electric Vehicle Markets.” Nature Energy 3 (4): 279–89. https://doi.org/10.1038/s41560-018-0108-1.

Canton, Patrizia, Giuliano Fagherazzi, Marino Battagliarin, Federica Menegazzo, Francesco Pinna, and Nicola Pernicone. 2002. “Pd/CO Average Chemisorption Stoichiometry in Highly Dispersed Supported Pd/γ-Al2O3 Catalysts.” Langmuir 18 (17): 6530–35. https://doi.org/10.1021/la015650a.

Chen, Xiao Rong, Yu Qiao Du, Chang Lin Chen, Nan Ping Xu, and Chung Yuan Mou. 2006. “Highly Active and Stable N-Pentane Isomerization Catalysts without Noble Metal Containing: Al- or Ga-Promoted Tungstated Zirconia.” Catalysis Letters 111 (3–4): 187–93. https://doi.org/10.1007/s10562-006-0146-3.

Chica, A., and A. Corma. 1999. “Hydroisomerization of Pentane, Hexane, and Heptane for Improving the Octane Number of Gasoline.” Journal of Catalysis 187 (1): 167–76. https://doi.org/10.1006/jcat.1999.2601.

Choudhary, V R. 1997. “Low-Temperature Nonoxidative Activation of Methane over H-Galloaluminosilicate (MFI) Zeolite.” Science 275 (5304): 1286–88. https://doi.org/10.1126/science.275.5304.1286.

Choudhary, V. R., C. Sivadinarayana, A. K. Kinage, P. Devadas, and M. Guisnet. 1996. “H-Gallosilicate (MFI) Propane Aromatization Catalyst Influence of Calcination Temperature on Acidity, Activity and Deactivation Due to Coking.” Applied Catalysis A: General 136 (2): 125–42. https://doi.org/10.1016/0926-860X(95)00277-4.

Choudhary, V.R, S.a.R Mulla, and S Banerjee. 2003. “Aromatization of N-Heptane over H-AlMFI, Ga/H-AlMFI, H-GaMFI and H-GaAlMFI Zeolite Catalysts: Influence of Zeolitic Acidity and Non-Framework Gallium.” Microporous and Mesoporous Materials 57: 317–22. https://doi.org/10.1016/S1387-1811(02)00606-6.

Cordero-Lanzac, Tomás, Andrés T. Aguayo, Ana G. Gayubo, Pedro Castaño, and Javier Bilbao. 2018. “Simultaneous Modeling of the Kinetics for N-Pentane Cracking and the Deactivation of a HZSM-5 Based Catalyst.” Chemical Engineering Journal 331 (August 2017). Elsevier: 818–30. https://doi.org/10.1016/j.cej.2017.08.106.

Dautzenberg, F M, and J C Platteeuw. 1970. “Lsomerization and Dehydrocyclization of Hexanes over Monofunctional Supported Platinum Catalysts.” Journal of Catalysis 19: 41–48.

Dewitt, M. J., Corporan, E., Graham, J. & Minus, D. Effects of aromatic type and concentration in Fischer− Tropsch fuel on emissions production and material compatibility. 2411–2418 (2008).

115

Ding, Weiping, George D. Meitzner, and Enrique Iglesia. 2002. “The Effects of Silanation of External Acid Sites on the Structure and Catalytic Behavior of Mo/H–ZSM5.” Journal of Catalysis 206 (1): 14–22. https://doi.org/10.1006/jcat.2001.3457.

Ding, Wen, Yuyang Cui, Jianjun Li, Yiquan Yang, and Weiping Fang. 2014. “Promoting Effect of Dual Modification of H-ZSM-5 Catalyst by Alkali Treating and Mg Doping on Catalytic Performances for Alkylation of Benzene with Ethanol to Ethylbenzene.” RSC Adv. 4 (91). Royal Society of Chemistry: 50123–29. https://doi.org/10.1039/C4RA07918C.

Emeis, C. A. 1993. “Determination of Integrated Molar Extinction Coefficients for Infrared Absorption Bands of Pyridine Adsorbed on Solid Acid Catalysts.” Journal of Catalysis. https://doi.org/10.1006/jcat.1993.1145.

Fan, Yu, Xiaojun Bao, Gang Shi, Weisheng Wei, and Jian Xu. 2004. “Olefin Reduction of FCC Gasoline via Hydroisomerization Aromatization over Modified HMOR/HZSM-5/Hβ Composite Carriers.” Applied Catalysis A: General 275 (1–2): 61–71. https://doi.org/10.1016/j.apcata.2004.07.046.

Fermoso, Javier, Héctor Hernando, Prabhas Jana, Inés Moreno, Jan Přech, Cristina Ochoa-Hernández, Patricia Pizarro, Juan M. Coronado, Jiří Čejka, and David P. Serrano. 2016. “Lamellar and Pillared ZSM-5 Zeolites Modified with MgO and ZnO for Catalytic Fast-Pyrolysis of Eucalyptus Woodchips.” Catalysis Today 277: 171–81. https://doi.org/10.1016/j.cattod.2015.12.009.

Fling, Jingly, and Ikai Wang. 1991. “Dehydrocyclization of C6C8 N-Paraffins to Aromatics over TiO2ZrO2 Catalysts.” Journal of Catalysis 130 (2): 577–87. https://doi.org/10.1016/0021-9517(91)90137-S.

Gabrienko, A. A. et al. Different Efficiency of Zn2+and ZnO Species for Methane Activation on Zn-Modified Zeolite. ACS Catal. 7, 1818–1830 (2017).

Ghosh, Prasenjeet, Karlton J. Hickey, and Stephen B. Jaffe. 2006. “Development of a Detailed Gasoline Composition-Based Octane Model.” Industrial and Engineering Chemistry Research 45 (1): 337–45. https://doi.org/10.1021/ie050811h.

Haensel. 1951. “United States Patent Office.”

Haller, G. L. & Resasco, D. E. Metal Support Interaction - Group-VIII Metals and Reducible Oxides. Adv. Catal. 36, 173–235 (1989).

Hart, Abarasi, Gary Leeke, Malcolm Greaves, and Joseph Wood. 2014. “Down-Hole Heavy Crude Oil Upgrading by CAPRI: Effect of Hydrogen and Methane Gases upon Upgrading and Coke Formation.” Fuel 119. Elsevier Ltd: 226–35. https://doi.org/10.1016/j.fuel.2013.11.048.

Haryanto, Agus, Sandun Fernando, Naveen Murali, and Sushil Adhikari. 2005. “Current Status of Hydrogen Production Techniques by Steam Reforming of Ethanol: A Review.” Energy and Fuels 19 (5): 2098–2106. https://doi.org/10.1021/ef0500538.

He, P., J. Jarvis, S. Meng, A. Wang, S. Kou, R. Gatip, M. Yung, L. Liu, and H. Song. 2018. “Co-Aromatization of Methane with Olefins: The Role of Inner Pore and External

116

Surface Catalytic Sites.” Applied Catalysis B: Environmental 234. https://doi.org/10.1016/j.apcatb.2018.04.034.

He, Peng, Richard Gatip, Matthew Yung, Hongbo Zeng, and Hua Song. 2017. “Applied Catalysis B : Environmental Co-Aromatization of Olefin and Methane over Ag-Ga / ZSM-5 Catalyst at Low Temperature.” “Applied Catalysis B, Environmental” 211. Elsevier B.V.: 275–88. https://doi.org/10.1016/j.apcatb.2017.04.052.

He, Peng, Yang Lou, and Hua Song. 2016. “Olefin Upgrading under Methane Environment over Ag-Ga/ZSM-5 Catalyst.” Fuel 182. Elsevier Ltd: 577–87. https://doi.org/10.1016/j.fuel.2016.05.126.

He, Peng, Yumeng Wen, Jack Jarvis, Richard Gatip, Danielle Austin, and Hua Song. 2017. “Selective Participation of Methane in Olefin Upgrading over Pd/ZSM-5 and Ir/ZSM-5: Investigation Using Deuterium Enriched Methane.” ChemistrySelect 2 (1): 252–56. https://doi.org/10.1002/slct.201601625.

Hernando, Héctor, Inés Moreno, Javier Fermoso, Cristina Ochoa-Hernández, Patricia Pizarro, Juan M. Coronado, Jiří Čejka, and David P. Serrano. 2017. “Biomass Catalytic Fast Pyrolysis over Hierarchical ZSM-5 and Beta Zeolites Modified with Mg and Zn Oxides.” Biomass Conversion and Biorefinery 7 (3): 289–304. https://doi.org/10.1007/s13399-017-0266-6.

Hyett, G., Green, M. & Parkin, I. P. X-ray diffraction area mapping of preferred orientation and phase change in TiO2 thin films deposited by chemical vapor deposition. J Am Chem Soc 128, 12147–12155 (2006).

Jarvis, Jack, Ashley Wong, Peng He, Qingyin Li, and Hua Song. 2018. “Catalytic Aromatization of Naphtha under Methane Environment: Effect of Surface Acidity and Metal Modification of HZSM-5.” Fuel 223: 211–21. https://doi.org/10.1016/j.fuel.2018.03.045.

Kalogeras, I.M., and A.B. Vassilikou-Dova. 1998. “Electrical Properties of Zeolitic Catalysts.” Defect and Diffusion Forum 164 (January 1998): 1–36. https://doi.org/10.4028/www.scientific.net/DDF.164.1.

Kotrel, S., H. Knözinger, and B. C. Gates. 2000. “The Haag-Dessau Mechanism of Protolytic Cracking of Alkanes.” Microporous and Mesoporous Materials 35–36: 11–20. https://doi.org/10.1016/S1387-1811(99)00204-8.

Kwak, B. S., Sachtler W. M. H., W. O. Haag, W. M. H. Sachtler, and W. O. Haag. 1994. “Catalytic Conversion of Propane to Aromatics: Effects of Adding Ga and/or Pt to HZSM-5.” Journal of Catalysis. https://doi.org/10.1006/jcat.1994.1313.

Labinger, Jay a., David C. Leitch, John E. Bercaw, Mark a. Deimund, and Mark E. Davis. 2015. “Upgrading Light Hydrocarbons: A Tandem Catalytic System for Alkane/Alkene Coupling.” Topics in Catalysis 58: 494–501. https://doi.org/10.1007/s11244-015-0380-2.

Li, Q., F. Zhang, J. Jarvis, P. He, M.M. Yung, A. Wang, K. Zhao, and H. Song. 2018. “Investigation on the Light Alkanes Aromatization over Zn and Ga Modified HZSM-5

117

Catalysts in the Presence of Methane.” Fuel 219. https://doi.org/10.1016/j.fuel.2018.01.104.

Li, Q., P. He, J. Jarvis, A. Bhattacharya, X. Mao, A. Wang, G.M. Bernard, et al. 2018. “Catalytic Co-Aromatization of Methane and Heptane as an Alkane Model Compound over Zn-Ga/ZSM-5: A Mechanistic Study.” Applied Catalysis B: Environmental 236. https://doi.org/10.1016/j.apcatb.2018.05.006.

Li, Qingyin, Fengqi Zhang, Jack Jarvis, Peng He, Matthew M. Yung, Aiguo Wang, Kai Zhao, and Hua Song. 2018. “Investigation on the Light Alkanes Aromatization over Zn and Ga Modified HZSM-5 Catalysts in the Presence of Methane.” Fuel 219 (January). Elsevier: 331–39. https://doi.org/10.1016/j.fuel.2018.01.104.

Lin, Xiuying, Yu Fan, Gang Shi, Haiyan Liu, and Xiaojun Bao. 2007. “Coking and Deactivation Behavior of HZSM-5 Zeolite-Based FCC Gasoline Hydro-Upgrading Catalyst.” Energy and Fuels 21 (5): 2517–24. https://doi.org/10.1021/ef0700634.

Liu, B. S., Y. Zhang, J. F. Liu, M. Tian, F. M. Zhang, C. T. Au, and A. S.C. Cheung. 2011. “Characteristic and Mechanism of Methane Dehydroaromatization over Zn-Based/HZSM-5 Catalysts under Conditions of Atmospheric Pressure and Supersonic Jet Expansion.” Journal of Physical Chemistry C 115 (34): 16954–62. https://doi.org/10.1021/jp2027065.

Liu, F., Wu, C. & Yang, S. Strain and Ligand Effects on CO2 Reduction Reactions over Cu-Metal Heterostructure Catalysts. J. Phys. Chem. C 121, 22139–22146 (2017).

Liu, J., N. He, Y. Zhao, L. Lin, W. Zhou, G. Xiong, H. Xie, and H. Guo. 2018. “The Crucial Role of Skeleton Structure and Carbon Number on Short-Chain Alkane Activation over Zn/HZSM-5 Catalyst: An Experimental and Computational Study.” Catalysis Letters 0 (0). Springer US: 0. https://doi.org/10.1007/s10562-018-2394-4.

Lønstad Bleken, Bjørn-Tore, Lorenzo Mino, Filippo Giordanino, Pablo Beato, Stian Svelle, Karl Petter Lillerud, and Silvia Bordiga. 2013. “Probing the Surface of Nanosheet H-ZSM-5 with FTIR Spectroscopy.” Physical Chemistry Chemical Physics 15 (32): 13363. https://doi.org/10.1039/c3cp51280k.

Lónyi, Ferenc, and József Valyon. 2001. “On the Interpretation of the NH3-TPD Patterns of H-ZSM-5 and H-Mordenite.” Microporous and Mesoporous Materials 47 (2–3): 293–301. https://doi.org/10.1016/S1387-1811(01)00389-4.

Lou, Yang, Peng He, Lulu Zhao, and Hua Song. 2016. “Refinery Oil Upgrading under Methane Environment over PdOx/H-ZSM-5: Highly Selective Olefin Cyclization.” Fuel 183. Elsevier Ltd: 396–404. https://doi.org/10.1016/j.fuel.2016.06.074.

Lubango, L. M. & Scurrell, M. S. Light alkanes aromatization to BTX over Zn-ZSM-5 catalysts: Enhancements in BTX selectivity by means of a second transition metal ion. Appl. Catal. A Gen. 235, 265–272 (2002).

Lukyanov, Dmitry B., and Tanya Vazhnova. 2007. “Aromatization Activity of Gallium Containing MFI and TON Zeolite Catalysts in N-Butane Conversion: Effects of Gallium

118

and Reaction Conditions.” Applied Catalysis A: General 316 (1): 61–67. https://doi.org/10.1016/j.apcata.2006.09.019.

Lv, J. et al. Phosphorus modified hierarchically structured ZSM-5 zeolites for enhanced hydrothermal stability and intensified propylene production from 1-butene cracking. Microporous Mesoporous Mater. 247, 31–37 (2017).

Marques, J. P., I. Gener, J. M. Lopes, F. Ramôa Ribeiro, and M. Guisnet. 2005. “N-Heptane Cracking on Dealuminated HBEA Zeolites.” Catalysis Today 107–108: 726–33. https://doi.org/10.1016/j.cattod.2005.07.003.

Meng, Lingqian, Brahim Mezari, Maarten G. Goesten, Wannaruedee Wannapakdee, Robert Pestman, Lu Gao, Jan Wiesfeld, and Emiel J.M. Hensen. 2017. “Direct Synthesis of Hierarchical ZSM-5 Zeolite Using Cetyltrimethylammonium as Structure Directing Agent for Methanol-to-Hydrocarbons Conversion.” Catalysis Science and Technology 7 (19). Royal Society of Chemistry: 4520–33. https://doi.org/10.1039/c7cy01603d.

Mériaudeau, P., and C. Naccache. 1996. “Gallium Based MFI Zeolites for the Aromatization of Propane.” Catalysis Today 31 (3–4): 265–73. https://doi.org/10.1016/S0920-5861(96)00017-X.

Mikhailov, M. N., I. V. Mishin, L. M. Kustov, and A. L. Lapidus. 2007. “Structure and Reactivity of Pt/GaZSM-5 Aromatization Catalyst.” Microporous and Mesoporous Materials 104 (1–3): 145–50. https://doi.org/10.1016/j.micromeso.2007.01.023.

Motz, J.L., H. Heinichen, and W.F. Holderich. 1997. “Influence of Extra-Framework Alumina in H-[Al]ZSM-5 Zeolite on the Direct Hydroxylation of Benzene to Phenol” 105 (January). Elsevier: 1053–60. https://doi.org/10.1016/S0167-2991(97)80739-2.

Nabgan, Walid, Mehdi Rashidzadeh, and Bahador Nabgan. 2018. “The Catalytic Naphtha Reforming Process: Hydrodesulfurization, Catalysts and Zeoforming.” Environmental Chemistry Letters 16 (2). Springer International Publishing: 507–22. https://doi.org/10.1007/s10311-018-0707-x.

Narayanan, S. Strong Metal Support Interaction. J. Sci. Ind. Res. (India). 44, 580–587 (1985).

Ono, Yoshio. 1992. “Transformation of Lower Alkanes into Aromatic Hydrocarbons over ZSM-5 Zeolites.” Catalysis Reviews 34 (3): 179–226. https://doi.org/10.1080/01614949208020306.

Ono, Yoshio. 2003. “A Survey of the Mechanism in Catalytic Isomerization of Alkanes.” Catalysis Today 81 (1): 3–16. https://doi.org/10.1016/S0920-5861(03)00097-X.

Parr Instrument Company. 2018. “Series 4590 Micro Stirred Reactors, 25-100 ML – Parr Instrument Company.” 2018. https://www.parrinst.com/products/stirred-reactors/series-4590-micro-stirred-reactors/.

Parsafard, N., M. H. Peyrovi, and M. Jarayedi. 2017. “Catalytic Study and Kinetic Modeling of the N-Heptane Isomerization over Pt/Al-HMS/HZSM-5 Hybrid Catalysts.” Energy and Fuels 31 (6): 6389–96. https://doi.org/10.1021/acs.energyfuels.7b00657.

119

Petushkov, A., Yoon, S. & Larsen, S. C. Synthesis of hierarchical nanocrystalline ZSM-5 with controlled particle size and mesoporosity. Microporous Mesoporous Mater. 137, 92–100 (2011).

Rahimpour, Mohammad Reza, Mitra Jafari, and Davood Iranshahi. 2013. “Progress in Catalytic Naphtha Reforming Process: A Review.” Applied Energy 109. Elsevier Ltd: 79–93. https://doi.org/10.1016/j.apenergy.2013.03.080.

Rasouli, Milad, and Nakisa Yaghobi. 2018. “Synthesis and Characterization of Platinum Impregnated Zn-ZSM5 Nanocatalysts for Xylene Isomerization Reactions.” Catalysis Letters 0 (0). Springer US: 0. https://doi.org/10.1007/s10562-018-2439-8.

Reding, G., Mäurer, T. & Kraushaar-Czarnetzki, B. Comparing synthesis routes to nano-crystalline zeolite ZSM-5. Microporous Mesoporous Mater. 57, 83–92 (2003).

Rodríguez-González, Luz, Florian Hermes, Marko Bertmer, Enrique Rodríguez-Castellón, Antonio Jiménez-López, and Ulrich Simon. 2007. “The Acid Properties of H-ZSM-5 as Studied by NH3-TPD And27Al-MAS-NMR Spectroscopy.” Applied Catalysis A: General 328 (2): 174–82. https://doi.org/10.1016/j.apcata.2007.06.003.

Roshanaei, A. & Alavi, S. M. Using two-zone fluidized bed reactor in propane aromatization over Zn/HZSM-5 catalyst. Fuel Process. Technol. 176, 197–204 (2018).

Roshanaei, Abbas, and Seyed Mehdi Alavi. 2018. “Using Two-Zone Fluidized Bed Reactor in Propane Aromatization over Zn/HZSM-5 Catalyst.” Fuel Processing Technology 176 (March). Elsevier: 197–204. https://doi.org/10.1016/j.fuproc.2018.03.020.

Rossini, Stefano. 2003. “The Impact of Catalytic Materials on Fuel Reformulation.” Catalysis Today 77 (4): 467–84. https://doi.org/10.1016/S0920-5861(02)00386-3.

Samanta, Anupam, Xinwei Bai, Brandon Robinson, Hui Chen, and Jianli Hu. 2017. “Conversion of Light Alkane to Value-Added Chemicals over ZSM-5/Metal Promoted Catalysts.” Industrial and Engineering Chemistry Research 56 (39): 11006–12. https://doi.org/10.1021/acs.iecr.7b02095.

Sarkar, B. et al. Pt nanoparticles supported on mesoporous ZSM-5 : A potential catalyst for reforming of methane with carbon dioxide. 51, 1348–1353 (2012).

Saxena, Sandeep K., Nagabhatla Viswanadham, and M. O. Garg. 2013. “Cracking and Isomerization Functionalities of Bi-Metallic Zeolites for Naphtha Value Upgradation.” Fuel 107 (January): 432–38. https://doi.org/10.1016/j.fuel.2013.01.076.

Sayed, Moein B., Ronald A. Kydd, and Ralph P. Cooney. 1984. “A Fourier-Transform Infrared Spectral Study of H-ZSM-5 Surface Sites and Reactivity Sequences in Methanol Conversion.” Journal of Catalysis 88 (1): 137–49. https://doi.org/10.1016/0021-9517(84)90059-9.

Schanke, D. 1996. “Study of Pt-Promoted Cobalt CO Hydrogenation Catalysts.” Fuel and Energy Abstracts. https://doi.org/10.1016/0140-6701(96)88441-4.

Seddon, Duncan. 1990. “Paraffin Oligomerisation to Aromatics.” Catalysis Today 6 (3): 351–72. https://doi.org/10.1016/0920-5861(90)85009-D.

120

Shrivastav, G., Khan, T. S., Agarwal, M. & Haider, M. A. Reformulation of Gasoline to Replace Aromatics by Biomass-Derived Alkyl Levulinates. ACS Sustain. Chem. Eng. 5, 7118–7127 (2017).

Song, Jie, Huaijun Ma, Zhijian Tian, Lijun Yan, Zhusheng Xu, Qianhe Liu, and Wei Qu. 2015. “The Effect of Fe on Pt Particle States in Pt/KL Catalysts.” Applied Catalysis A: General 492. Elsevier B.V.: 31–37. https://doi.org/10.1016/j.apcata.2014.12.017.

Sterba, M.J, and Vladimir Haensel. 1976. “Catalytic Reforming.” Ind. Eng. Chemical 15: 1–17. https://doi.org/10.1021/i360057a002.

Su, Xiaofang, Wang Zan, Xuefeng Bai, Gaoliang Wang, and Wei Wu. 2017. “Synthesis of Microscale and Nanoscale ZSM-5 Zeolites: Effect of Particle Size and Acidity of Zn Modified ZSM-5 Zeolites on Aromatization Performance.” Catal. Sci. Technol. 7. Royal Society of Chemistry: 1–11. https://doi.org/10.1039/C7CY00435D.

Tang, Pei, Qingjun Zhu, Zhaoxuan Wu, and Ding Ma. 2014. “Methane Activation: The Past and Future.” Energy Environ. Sci. 7 (8): 2580–91. https://doi.org/10.1039/C4EE00604F.

Tao, Y., Kanoh, H. & Kaneko, K. ZSM-5 monolith of uniform mesoporous channels. J. Am. Chem. Soc. 125, 6044–6045 (2003).

Treesukol, Piti, Kanokthip Srisuk, Jumras Limtrakul, and Thanh N. Truong. 2005. “Nature of the Metal-Support Interaction in Bifunctional Catalytic Pt/H-ZSM-5 Zeolite.” Journal of Physical Chemistry B 109 (24): 11940–45. https://doi.org/10.1021/jp0511348.

Tshabalala, Themba E., and Michael S. Scurrell. 2015. “Aromatization of N-Hexane over Ga, Mo and Zn Modified H-ZSM-5 Zeolite Catalysts.” Catalysis Communications 72. Elsevier B.V.: 49–52. https://doi.org/10.1016/j.catcom.2015.06.022.

U. Mroczek, W. Reschetilowski, K. Pietzsch and K.H. Steinberg. 1991. “Aromatization of Ethane on Pt and Pd Supported Zn- and Mn-ZSM-5 Zeolites.” Reaction Kinetics and Catalysis Letters 43 (2): 539–44.

Viswanadham, N., A. R. Pradhan, N. Ray, S. C. Vishnoi, Uma Shanker, and T. S R Prasada Rao. 1996. “Reaction Pathways for the Aromatization of Paraffins in the Presence of H-ZSM-5 and Zn/H-ZSM-5.” Applied Catalysis A: General 137 (2): 225–33. https://doi.org/10.1016/0926-860X(95)00287-1.

Xiao, Li, and Lichang Wang. 2007. “Methane Activation on Pt and Pt4: A Density Functional Theory Study.” Journal of Physical Chemistry B 111 (7): 1657–63. https://doi.org/10.1021/jp065288e.

Xin, L. et al. Efficient Conversion of Light Cycle Oil into High-Octane-Number Gasoline and Light Olefins over a Mesoporous ZSM-5 Catalyst. Energy and Fuels 31, 6968–6976 (2017).

Xing, Bin, Xian Yong Pang, and Gui Chang Wang. 2011. “C-H Bond Activation of Methane on Clean and Oxygen Pre-Covered Metals: A Systematic Theoretical Study.” Journal of Catalysis 282 (1). Elsevier Inc.: 74–82. https://doi.org/10.1016/j.jcat.2011.05.027.

121

Yao, Jianhua, Raymond le van Mao, and Louise Dufresne. 1990. “Conversion of N-Butane into BTX Aromatics on New Hybrid Catalysts.” Applied Catalysis 65 (1): 175–88. https://doi.org/10.1016/S0166-9834(00)81596-1.

Yaripour, Fereydoon, Zahra Shariatinia, Saeed Sahebdelfar, and Akbar Irandoukht. 2015. “Conventional Hydrothermal Synthesis of Nanostructured H-ZSM-5 Catalysts Using Various Templates for Light Olefins Production from Methanol.” Journal of Natural Gas Science and Engineering 22. Elsevier B.V: 260–69. https://doi.org/10.1016/j.jngse.2014.12.001.

Zhao, Yong-hua, and Tian-yu Gao Ya-jie Wang Yan-jun Zhou. 2018. “Zinc Supported on Alkaline Activated HZSM-5 for Aromatization Reaction.” Reaction Kinetics, Mechanisms and Catalysis. Springer International Publishing. https://doi.org/10.1007/s11144-018-1426-9.