Superoxide radical and UV irradiation in ultrasound assisted oxidative desulfurization (UAOD): a potential alternative for green fuels

PDF

Download

Open document

Flip pages

Summarize

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
i

SUPEROXIDE RADICAL AND UV IRRADIATION IN ULTRASOUND ASSISTED
OXIDATIVE DESULFURIZATION (UAOD):
A POTENTIAL ALTERNATIVE FOR GREENER FUELS

by

Ngo Yeung Chan

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(ENVIRONMENTAL ENGINEERING)

May 2010

Copyright 2010 Ngo Yeung Chan

ii
ACKNOWLEDGMENTS

I would like to dedicate my heartfelt gratitude to my advisor (former committee
chairperson) Prof. T. F. Yen (1927 – 2010), for his selfless support, guidance and
edification. I would like to thank also my committee chairperson, Prof. M. Pirbazari, and
the committee members Prof. J. S. Devinny, Prof. J. J. Lee and Prof. K. S. Shing for their
helpful advice and encouragement.

I sincerely express my special thanks to Dr. M. Quinlan for his valuable suggestions and
help. I definitely appreciate all the friendships and helps from my colleagues including Dr.
M. W. Wan, Dr. O. Etemadi, Dr. S. S. Cheng, Dr. C. Y. Yang, and Dr. W. Fan. I would also
thank T. Y. Lin, Y. Tung, and S. Angkadjaja for their excellent assistance.

I would like to acknowledge Eco Energy Solutions Inc., Reno, Nevada and Intelligent
Energy Inc., Long Beach, California for their financial support in this study; the U.S. Navy
for the instrumental support of the ultrasonic device; and the U.S. Army for the
instrumental support of the Horiba Sulfur in Oil Analyser.

iii
Last by not least, I express my special thanks to my parents Yin Man Wong and Ho Yee
Chan, and my brother Ngo Fung Chan for their endless love and support.

iv
TABLE OF CONTENTS

ACKNOWLEDGMENTS ii
LIST OF TABLES ix
LIST OF FIGURES xiii
LIST OF ABBREVIATIONS xviii
ABSTRACT xxii
CHAPTER 1: INTRODUCTION 1
1.1 General Overview 1
1.2 Diesel Fuel 2
1.2.1 Diesel Fuel and Organic Sulfur Compounds (OSCs) 2
1.2.2 Diesel Fuel and Sulfur Regulations 6
1.3 Residual Oil 8
1.3.1 Residual Oil and OSCs 8
1.3.2 Residual Oil and Sulfur Regulations 14
1.4 Current Desulfurization Technologies 18
1.4.1 Hydrodesulfurization (HDS) 19
1.4.2 Adsorptive Desulfurization (ADS) 24
1.4.3 Biodesulfurization (BDS) 28
1.4.4 Oxidative Desulfurization (ODS) 31
1.5 Research Objectives 35
CHAPTER 2: THEORETICAL BACKGROUND 36
2.1 Introduction 36

v
2.2 Ultrasound 37
2.2.1 Fundamentals of Ultrasound 37
2.2.2 Theory on Sonochemistry 38
2.2.3 Sonochemistry in Aqueous Phase 42
2.2.4 History of UAOD and Its Modifications 46
2.3 Oxidant Selection 52
2.3.1 Hydrogen Peroxide 53
2.3.2 Superoxide Anions 55
2.4 Acid Catalysis 57
2.5 Phase Transfer Catalysis 59
2.5.1 Overview of Phase Transfer Catalysis 59
2.5.2 Mechanism of Phase Transfer Catalysis 60
2.5.3 PTC Selection in UAOD Process 65
2.6 Ionic Liquids (ILs) 68
2.6.1 Overview of ILs and RTILs 68
2.6.2 Applications of ILs in Hydrocarbon Separation 74
2.6.3 Applications of ILs in Desulfurizaton 76
2.7 Ultraviolet Irradiation 79
2.7.1 Fundamentals of Photochemistry and UV Radiation 79
2.7.2 UV and Photochemical Reactions in Aqueous System 84
2.7.3 Photocatalysis and Titanium (IV) Oxide 86
2.7.4 Photolysis and Photo-Oxidation Desulfurization 88
CHAPTER 3: MODIFIED OXIDATIVE DESULFURIZATION
USING SUPEROXIDE ON MODEL SULFUR COMPOUND STUDY 90
3.1 Introduction 90
3.2 Materials and Experimental Procedures 93
3.2.1 Chemical Preparation 93
3.2.2 Ultrasonic Reactor 94
3.2.3 Ultraviolet Lamp 95
3.2.4 Analytical Method 96
3.3 Experimental Design, Procedure, Results and Discussion 97
3.3.1 Use of Solid Oxidants in Oxidative Desulfurization 97

vi
3.3.1.1 Solid Oxidants Selection 97
3.3.1.2 Experimental Procedure 98
3.3.1.3 Results and Discussion 99
3.3.2 Effect of Different Phase Transfer Catalysts 104
3.3.2.1 Phase Transfer Catalyst Selection 104
3.3.2.2 Experimental Procedure 106
3.3.2.3 Results and Discussion 107
3.3.3 Effect of Acid Catalysts 110
3.3.3.1 Acid Catalyst Combination 110
3.3.3.2 Experimental Procedure 111
3.3.3.3 Results and Discussion 112
3.3.4 Effect of Ionic Liquid 114
3.3.4.1 Ionic Liquid Dosage 114
3.3.4.2 Experimental Procedure 115
3.3.4.3 Results and Discussion 116
3.3.4.4 Ionic Liquid Selection 118
3.3.4.5 Experimental Procedure 119
3.3.4.6 Results and Discussion 120
3.3.5 Effect of Treatment Methods 121
3.3.5.1 Time of Ultrasonication 121
3.3.5.2 Experimental Procedure 122
3.3.5.3 Results and Discussion 123
3.4 Desulfurization Efficiency on Various Model Sulfur Compounds 125
3.4.1 Identification of Experimental Optimum Conditions 125
3.4.2 Experimental Procedure 127
3.4.3 Results and Discussion 128
3.5 Kinetic Studies of Desulfurization on Model Sulfur Compounds 129
3.5.1 Experimental Procedure 133
3.5.2 Results and Discussion 134
3.6 Preliminary Study on UV Assisted Desulfurization Process 142
3.6.1 Challenge in UAOD Processes 142
3.6.2 Experimental Procedure 143
3.6.3 Results and Discussion 144
3.7 Summary and Conclusion 145

vii
CHAPTER 4: MODIFIED OXIDATIVE DESULFURIZATION
USING KO
2
AND H
2
O
2
ON PETROLEUM FUEL 148
4.1 Introduction 148
4.2 Materials 150
4.3 Experimental Procedure and Analytical Method 151
4.4 Results and Discussion 152
4.4.1 Desulfurization of JP-8 152
4.4.2 Desulfurization of MGO 155
4.4.3 Desulfurization of Sour Diesel 157
4.4.4 Kinetic Studies of Desulfurization on Various Diesel Samples 159
4.4.5 Desulfurization of Heavy-Distillates 163
4.5 Summary and Conclusion 164
CHAPTER 5: QUALITATIVE ANALYSIS USING GC-SCD 166
5.1 Introduction 166
5.2 Materials 171
5.3 Experimental Procedure 172
5.3.1 Preparation of Model Sulfur Compound Solutions 172
5.3.2 Desulfurization of Feedstock 173
5.3.3 Analytical Method 174
5.3.4 Estimation of Retention Time of Different OSCs 176
5.4 Results and Discussion 179
5.4.1 Model Sulfur Compounds Identification 179
5.4.2 Characterization of Untreated Diesel Samples 184
5.4.3 Characterization of Desulfurized Diesel Samples 187
5.4.4 Characterization of Heavy-Distillates 197
5.5 Mechanism of Inorganic Sulfate Formation 203
5.6 Summary and Conclusion 205
CHAPTER 6: CONCEPTUAL MODEL FOR THE MODIFIED
UAOD DESULFURIZATION PROCESS 208
6.1 Introduction 208

viii
6.2 Model Overview 209
6.3 Summary 212
CHAPTER 7: CONCLUSION AND RECOMMENDATIONS 213
7.1 Summary and Conclusion 213
7.2 Recommendations for Future Work 221
REFERENCES 224

ix
LIST OF TABLES
Table 1.1 Major types of OSCs in petroleum 5
Table 1.2 Sulfur content standards for diesel 7
Table 1.3 Classification of residual oil’s components 11
Table 1.4 Standard properties of residual oil 14
Table 1.5 Sulfur content standards for all marine-use fuel oils 17
Table 1.6 Chemical reactions for the Claus process and the contact process 20
Table 2.1 Sonochemistry in a cavitation bubble formed in water 43
Table 2.2 Chemical reactions initiated by ultrasound in water 44
Table 2.3 Comparison of the UAOD process and its modifications 50
Table 2.4 Oxygen-donor Oxidants 54
Table 2.5 Hydration of anion in chlorobenzene-aqueous system 64
Table 2.6 Standard electrode potentials for selected half-reactions 66
Table 2.7 Specific conductivities 69
Table 2.8 General properties of modern ionic liquids 70
Table 2.9 Toxicity of ionic liquids, expressed as EC
50
in µM 71

x
Table 2.10 Ultraviolet Classification 82
Table 2.11 Energy per mole of photons 83
Table 3.1 Specifications of Ultrasonic Reactor VCX-750 listed
in product catalogue 94
Table 3.2 Specifications of UV lamp UVLMS-38 listed in product catalogue 95
Table 3.3 Desulfurization efficiency with 30% wt. H
2
O
2
as oxidant 99
Table 3.4 Desulfurization efficiency with KMnO
4
as oxidant 100
Table 3.5 Desulfurization efficiency with NaO
2
as oxidant 100
Table 3.6 Desulfurization efficiency with KO
2
as oxidant 101
Table 3.7 Effect of surfactants on the UAOD process 105
Table 3.8 Desulfurization of DBT solution with TOAF as PTC 108
Table 3.9 Desulfurization of DBT solution with 18-crown-6 as PTC 109
Table 3.10 Desulfurization of DBT solution with respect to acid dosage 112
Table 3.11 Desulfurization of DBT solution with respect to acid catalyst
applied 113
Table 3.12 Desulfurization of DBT solution with respect to IL dosage 116
Table 3.13 Desulfurization of DBT solution with respect to type of IL 120
Table 3.14 Desulfurization of DBT solution under magnetic stirring and
ultrasound 124

xi
Table 3.15 Selected conditions for desulfurization in model compound
studies with 10 minutes ultrasonication 126
Table 3.16 Desulfurization of various model sulfur compounds 128
Table 3.17 Selected conditions for desulfurization in model compound
studies without ultrasonication 132
Table 3.18 Desulfurization efficiencies with respect to reaction time 134
Table 3.19 Rate constants for various model sulfur compounds using KO
2
138
Table 3.20 Rate constants for various model sulfur compounds using H
2
O
2
138
Table 3.21 Apparent activation energies for oxidation of BT and DBT 140
Table 3.22 Desulfurization of model sulfur compounds with UV 144
Table 4.1 Desulfurization of JP-8 using KO
2
as oxidant 152
Table 4.2 Desulfurization of JP-8 using 30% wt. H
2
O
2
as oxidant 153
Table 4.3 Desulfurization of MGO using KO
2
as oxidant 156
Table 4.4 Desulfurization of MGO using 30% wt. H
2
O
2
as oxidant 156
Table 4.5 Desulfurization of sour diesel using KO
2
as oxidant 157
Table 4.6 Desulfurization of sour diesel using 30% wt. H
2
O
2
as oxidant 157
Table 4.7 Rate constants for various diesel samples using KO
2
as oxidant 160
Table 4.8 Desulfurization of different heavy-distillates using KO
2
as oxidant 163

xii
Table 5.1 Calculated retention time of various BT and DBT derivatives 177
Table 5.2 Calculated retention time of various OSCs 178
Table 5.3 Total sulfur content for various diesel samples in different stages 188
Table 7.1 Chemical cost comparison of different UAOD generations to
desulfurize 10 grams sample with 1000ppm
w
DBT in bench
scale study 219

xiii
LIST OF FIGURES
Figure 1.1 World Marketed Energy Use by Fuel Type 1
Figure 1.2 Schematic diagram of oil refinery 4
Figure 1.3 Schematic diagram of residual oil production 9
Figure 1.4 Separation of petroleum into four major fractions 10
Figure 1.5 Examples of asphaltene formulae 11
Figure 1.6 Examples of OSCs in residual oils 13
Figure 1.7 Schematic diagram of HDS unit 21
Figure 1.8 IUPAC numbering for DBT 22
Figure 1.9 Schematic diagram of PSU-SARS followed by HDS 25
Figure 1.10 Sulfur-specific degradation pathway of DBT 29
Figure 1.11 General process scheme for ODS 31
Figure 1.12 General reactions of BT and DBT in ODS 32
Figure 2.1 The process of sonoluminescence 40
Figure 2.2 Sonochemical reactions in water with Ar, O
2
and O
3
45
Figure 2.3 The Keggin structure of [PW
12
O
40
]
3-
47

xiv
Figure 2.4 Conceptual model of 1
st
generation of UAOD process 48
Figure 2.5 Degradation of specific ionic liquid with ultrasound 51
Figure 2.6 Chemical reactions between OSCs and superoxide anion 56
Figure 2.7 Structures of crown ethers: 15-crown-5 and 18-crown-6 61
Figure 2.8 Starks’ extraction mechanism 62
Figure 2.9 Modified Starks’ extraction mechanism with water of hydration 64
Figure 2.10 Chemical structures of common cations and anions in RTILs 73
Figure 2.11 Schematic diagram of oxidative extractive desulfurization 77
Figure 2.12 Section of Electromagnetic Spectrum 80
Figure 2.13 Organic destruction by UV/H
2
O
2
system 85
Figure 2.14 Radicals generation in Photocatalysis 87
Figure 3.1 Sulfur-in-oil analyzer (SLFA-20) 96
Figure 3.2 Desulfurization efficiencies of BT sample with different oxidants 101
Figure 3.3 Desulfurization efficiencies of DBT sample with different
oxidants 102
Figure 3.4 Desulfurization efficiencies of BT and DBT with different
KO
2
dosage 103
Figure 3.5 Superstructure of alternate reaction micro-kinetic pathway 130

xv
Figure 3.6 Desulfurization efficiencies of various model sulfur compounds
for KO
2
process 136
Figure 3.7 Linear regression of –ln(C
t
/C
o
) vs time for T and 2MT
for KO
2
process 136
Figure 3.8 Linear regression of –ln(C
t
/C
o
) vs time for BT and 2BMT
for KO
2
process 137
Figure 3.9 Linear regression of –ln(C
t
/C
o
) vs time for DBT and 4,6DMDBT
for KO
2
process 137
Figure 3.10 Linear regression of –ln(C
t
/C
o
) vs time at different temperature
for BT 140
Figure 3.11 Linear regression of –ln(C
t
/C
o
) vs time at different temperature
for DBT 141
Figure 3.12 -ln(k’) versus 1/T for model sulfur compounds BT and DBT 141
Figure 4.1 Desulfurization of JP-8 with respect to total reaction time 154
Figure 4.2 Desulfurization removal efficiencies of different diesel fuels
as a function of reaction time 160
Figure 4.3 Desulfurization rates of different diesel fuels as a function
of reaction time 161
Figure 5.1 Major reactions in SCD 168
Figure 5.2 Some of the standard sulfur measurement methods adopted
by ASTM 169
Figure 5.3 Block diagram of GC-SCD setup 175

xvi
Figure 5.4 Experimental retention time versus reference retention time
of model sulfur compounds BT, DBT and 4, 6DMDBT 177
Figure 5.5 GC-SCD chromatogram of sample with BT, DBT and
4, 6DMDBT before oxidation process 180
Figure 5.6 GC-SCD chromatogram of sample with BT, DBT and
4, 6DMDBT after oxidation process 180
Figure 5.7 GC-SCD chromatogram of sample with BTO before
oxidation process 181
Figure 5.8 GC-SCD chromatogram of sample with BTO after
oxidation process 181
Figure 5.9 GC-SCD chromatogram of sample with DBTO before
oxidation process 181
Figure 5.10 GC-SCD chromatogram of sample with DBTO after
oxidation process 182
Figure 5.11 Chemical reactions between OSCs and superoxide anion 183
Figure 5.12 GC-SCD chromatogram of various model sulfur compounds 185
Figure 5.13 GC-SCD chromatogram of various diesel samples 186
Figure 5.14 GC-SCD chromatogram of JP-8 at different stages of the process 190
Figure 5.15 GC-SCD chromatogram of MGO at different stages of the process 191
Figure 5.16 GC-SCD chromatogram of sour diesel at different stages
of the process 193

xvii
Figure 5.17 GC-SCD chromatogram of treated sour diesel at different
stages of the process 195
Figure 5.18 GC-SCD chromatograms of two untreated heavy distillates 197
Figure 5.19 Typical GC-MCD chromatogram of vacuum gas oil 198
Figure 5.20 GC-SCD chromatogram of treated RO-6 at different stages
of the process 199
Figure 5.21 GC-SCD chromatogram of treated IFO at different stages
of the process 200
Figure 5.22 Possible BT destruction pathways by OH radicals 204
Figure 5.23 Photolysis of DBTO 204
Figure 6.1 Conceptual model of 4
th
generation UAOD process 211
Figure 7.1 Schematic diagram of the 4
th
generation UAOD process 220

xviii
LIST OF ABBREVIATIONS

2MBT 2-Methyl Benzothiophene
2MT 2-Methyl Thiophene
4,6DMDBT 4, 6-Dimethyl dibenzothiophene
ADS Adsorptive Desulfurization
AED Atomic Emission Detector
BDS Biodesulfurization
BT(s) Benzothiophene(s)
BTO Benzothiophene Sulfone
CFCs Chlorofluorocarbons
CMFR Completely Mixed Flow Reactor
DBT(s) Dibenzothiophene(s)
DBTO Dibenzothiophene sulfone
DW Deionized Water
EIA Energy Information Administration
EPA Environmental Protection Agency

xix
FID Flame Ionization Detector
FPD Flame Photometric Detector
FUS Focused Ultrasound Surgery
GC Gas Chromatography
HCFCs Hydrochlorofluorocarbons
HDS Hydrodesulfurization
HECD Electroconductivity Detector
IL(s) Ionic Liquid(s)
IMO International Maritime Organization
LSD Low Sulfur Diesel
MCD Microcoulometric Sulfur Detector
MEPC Marine Environment Protection Committee
MS Mass Spectroscopy Detector
NDXRF Non-Dispersive X-Ray Fluorescence
ODS Oxidative Desulfurization
OSC(s) Organic Sulfur Compound(s)
PAHs Polycyclic Aromatic Hydrocarbons

xx
PCE Perchloroethylene
PFPD Pulse Flame Photometric Detector
PM Particulate Matter
POM(s) Polyoxometalate(s)
PTC(s) Phase Transfer Catalyst(s)
QAS(s) Quaternary Ammonium Salts
RTIL(s) Room Temperature Ionic Liquid(s)
SCD Sulfur Chemiluminescence Detector
SECAs Sulfur Emission Control Areas
SO
x
sulfur oxides
S-Zorb SRT S-Zorb Sulfur Removal Technology
T(s) Thiophene(s)
TCE Trichloroethylene
TFA Trifluoroacetic Acid
TMC Transition Metal Catalysts
TOAF Tetraoctylammonium Fluoride
UAOD Ultrasound Assisted Oxidative Desulfurization

xxi
ULSD Ultralow Sulfur Diesel
UV Ultraviolet
VOC(s) Volatile Organic Compound(s)

xxii
ABSTRACT
This study is aimed at improving the current ultrasound assisted oxidative desulfurization
(UAOD) process by utilizing superoxide radical as oxidant. Research was also conducted
to investigate the feasibility of ultraviolet (UV) irradiation-assisted desulfurization. These
modifications can enhance the process with the following achievements:

• Meet the upcoming sulfur standards on various fuels including diesel fuel oils and
residual oils
• More efficient oxidant with significantly lower consumption in accordance with
stoichiometry
• Energy saving by 90%
• Greater selectivity in petroleum composition

Currently, the UAOD process and subsequent modifications developed in University of
Southern California by Professor Yen’s research group have demonstrated high
desulfurization efficiencies towards various fuels with the application of 30% wt.

xxiii
hydrogen peroxide as oxidant. The UAOD process has demonstrated more than 50%
desulfurization of refractory organic sulfur compounds with the use of Venturella type
catalysts. Application of quaternary ammonium fluoride as phase transfer catalyst has
significantly improved the desulfurization efficiency to 95%. Recent modifications
incorporating ionic liquids have shown that the modified UAOD process can produce
ultra-low sulfur, or near-zero sulfur diesels under mild conditions with 70 °C and
atmospheric pressure.

Nevertheless, the UAOD process is considered not to be particularly efficient with respect
to oxidant and energy consumption. Batch studies have demonstrated that the UAOD
process requires 100 fold more oxidant than the stoichiometic requirement to achieve high
desulfurization yield. The expected high costs of purchasing, shipping and storage of the
oxidant would reduce the practicability of the process. The excess use of oxidant is not
economically desirable, and it also causes environmental and safety issues. Post treatments
would be necessary to stabilize the unspent oxidant residual to prevent the waste stream
from becoming reactive or even explosive.

xxiv
High energy consumption is another drawback in the UAOD process. A typical 10 minutes
ultrasonication applied in the UAOD process to achieve 95% desulfurization for 20g of
diesel requires 450 kJ of energy, which is equivalent to approximately 50% of the energy
that can be provided by the treated diesel. This great expenditure of energy is impractical
for industries to adopt.

In this study, modifications of the UAOD process, including the application of superoxide
and selection of catalysts, were applied to lower the oxidant dosage and to improve the
applicability towards heavy-distillates such as residual oil. The results demonstrated that
the new system required 80% less oxidant as compared to previous generations of UAOD
process without the loss of desulfurization efficiency.

The new system demonstrated its suitability towards desulfurizing commercial
mid-distillates including jet fuels, marine gas oil and sour diesel. This process also
demonstrated a new method to desulfurize residual oil with high desulfurization yields.
The new process development has been supported by Eco Energy Solutions Inc., Reno,
Nevada and Intelligent Energy Inc., Long Beach, California.

xxv

A feasibility study on UV assisted desulfurization by replacing ultrasound with UV
irradiation was also conducted. The study demonstrated that the UV assisted
desulfurization process consumes 90% less energy than the comparable process using
ultrasonication.

These process modifications demonstrated over 98% desulfurization efficiency on diesel
oils and more than 75% on residual oils with significantly less oxidant and energy
consumption. Also the feasibility to desulfurize commercial sour heavy oil was
demonstrated. Based on the UAOD process and the commercialized modifications by Wan
and Cheng, the feasible applications of superoxide and UV irradiation in the UAOD
process could provide deep-desulfurization on various fuels with practical cost.

1
CHAPTER 1: INTRODUCTION
1.1 General Overview

Petroleum was discovered several thousands years ago and has been extensively used
starting from the Industrial Revolution. Nowadays, petroleum and its derivative products
provide approximately 37% of the world’s energy consumption (Energy Information
Administration, 2009) and 90% of the vehicular fuel. While sulfur is one of the major
contaminants in petroleum, it generates air pollution that leads to severe environmental
and health consequences. The U.S. Environmental Protection Agency (EPA) and other
similar groups across the world have started proposing regulations to lower sulfur content
in various fuels since 1990s, and as a result stringent sulfur content regulations are
established for diesel and residual oil in the upcoming future.
37%
23%
26%
6%
8%
Liquids
Natural Gas
Coal
Nuclear
Renew ables

Figure 1.1 World Marketed Energy Use by Fuel Type (EIA, 2009)

2
1.2 Diesel Fuel
1.2.1 Diesel Fuel and Organic Sulfur Compounds (OSCs)

Diesel, also known as petrodiesel, is a petroleum product through fractional
distillation of catalytic cracking of crude oil. It is a mixture of hydrocarbons with
typical carbon chain length of 8 to 21, having a boiling point ranged from 200°C to
325°C (Collins, 2007). Depending on its original sources and the refinery conditions,
properties such as sulfur content of a diesel fuel vary. Diesel fuel consists of
approximately 70% of aliphatic hydrocarbons including paraffins and naphthalenes,
with approximately 30% of aromatic hydrogen carbons (Steynberg et al, 2004).

Comparing with gasoline engine, diesel engine is generally operated under oxygen-
rich conditions, which leads to a more complete combustion thus releasing less
carbon monoxide and hydrocarbons. Besides, diesel engine has a lower fuel per
mile consumption, thus releasing less carbon dioxide. However, more soot, or
particulate matter (PM) including PM
2.5
and PM
10
, are emitted in burning diesel fuel.
These particulate matters are identified as contributing to variety of health problems
such as asthma, emphysema and bronchitis, while sulfur compounds in diesel is one
of the major factors causing the emission of these particulate matters. Recent
studies report that emission of diesel particulate matter increases significantly with
sulfur content in the diesel fuel burnt (Saiyasitpanich, 2005). In 1998, the California
Air Resources Board declared diesel PM as a toxic air contaminant and a

3
potential cancer risk; and in 2000, the U.S. EPA identified diesel PM as a likely
human carcinogen.

Emission of sulfur oxides (SO
x
) is another problem on using high sulfur diesel. SO
x

can dissolve in water vapour in the atmosphere resulting in acid rain, which is
known to be harmful to plants, aquatic animals, and infrastructure.

Sulfur compounds, or more specifically organic sulfur compounds (OSCs), are
considered as the most important non-hydrocarbon constituents in petroleum. There
are three major types of OSCs: thiols (R-SH), sulfides (R-S-R’) and thiophenes. In
mid-range distillates such as diesel, the OSCs are primarily cyclic sulfide derivates,
benzothiophene derivates and dibenzothiophene derivates (Speight, 1999). The
chemical structures of the major OSCs in petroleum are listed in Table 1.1.

4

Figure 1.2 Schematic diagram of oil refinery (Beychok, 2005)

5
Table 1.1 Major types of OSCs in petroleum
Thiols RSH
Sulfides RSR’
Cyclic Sulfides (thiacyclanes)

Thiophene

Benzothiophene

Dibenzothiophene

Naphthobenzothiophene

S
S
S

S
S
S

6
1.2.2 Diesel Fuel and Sulfur Regulations

In order to lower the particulate and sulfur oxides emissions, the U.S. EPA has
established regulations to limit sulfur content in diesel fuels. Established in
November 1990 as promulgated by the EPA Clean Air Act Title II, all commercial
motor vehicle diesel fuels (highway diesel fuel) are required to be Low Sulfur
Diesel (LSD) with maximum sulfur content of 500 ppm
w
as of October 1
st
, 1993. In
January 2001, the EPA established the Highway Diesel Rule (the 2007 Highway
Rule) with tightened limit on sulfur content.

Starting from June 1
st
, 2006, refiners in U.S. are required by the U.S. EPA to
produce ultralow sulfur diesel (ULSD) with maximum sulfur content of 15ppm
w
for
highway vehicle uses. By December 1
st
, 2010, all highway diesel fuel in U.S. must
be ULSD (Energy Information Administration, 2001). In California, the use of
ULSD for all highway diesel fuel is required by the California Air Resource Board
since September 2006 (California Air Resource Board, 2003).

Non-road diesel refers to land-based non-road, locomotive and marine engines
diesel use. Before June 2004, sulfur content in non-road diesel fuels was not
regulated by the EPA; except for an industrial specification of 0.5% (5000 ppm
w
)
sulfur. In June 2004, the EPA released the new standards for non-road diesel fuels.
For land-based non-road diesel fuel, the sulfur reductions will be accomplished in

7
two steps: (i) from uncontrolled levels to a 500 ppm
w
cap starting in June, 2007;
and (ii) to 15ppm
w
in June, 2010. Similarly, sulfur content limit for locomotive and
marine diesel fuels have been changed from uncontrolled levels to a 500 ppm
w
cap
starting in June, 2007; and will be further reduced to a 15 ppm
w
in June, 2012 (U.S.
EPA, 2004).

Table 1.2 Sulfur content standards for diesel (U.S. EPA, 2004)
Diesel Type
Maximum Sulfur Content
(ppm
w
)
Implementation
Date
Highway 500 October 1993
Highway 15 June 2006
Land-base non-road 500 June 2007
Land-base non-road 15 June 2010
Locomotive & Marine 500 June 2007
Locomotive & Marine 15 June 2012

8
1.3 Residual Oil
1.3.1 Residual Oil and OSCs

Residual fuel oil, also known as heavy fuel oil or bunker oil, is manufactured from
the residuum obtained from petroleum non-destructive distillation. Residual oil is a
highly viscous material with carbon chain length ranged from 12 to 70. Depending
on the distillation conditions and the nature of the crude oil, residual oil could be a
liquid or a solid in room conditions. Generally, a “liquid” residual oil is produced
from atmospheric distillation, while a “solid” or “almost solid” residual oil is
produced from reduced pressure distillation. Residual oil may be blended with
medium distillate such as heating oil or diesel to reduce its viscosity to acceptable
level.

Similar to crude oil, residual oil is a composition of asphaltenes, resins, aromatic
hydrocarbons and saturated hydrocarbons, while asphaltenes and resins are the
predominant components of residual oil. Asphaltenes are generally combination of
aromatic-naphthenic systems with substitution of different alkyl groups. They are
heteroatomic organic compounds which may contain atoms such as oxygen,
nitrogen, sulfur or metals in addition to carbon and hydrogen (Yen et al, 1994). It is
considered as the most complicated known organic components in petroleum, due
to high molecular weight, chemical structure diversity, tendency to associate and
other properties (Simanzhenkov et al, 2003).

9

Figure 1.3 Schematic diagram of residual oil production (Sunggyu Lee et al, 2007)

The word “asphaltene” was first used by J.B. Boussingault in 1837, to describe the
components of bitumen which are alcohol insoluble, turpentine soluble solid. In
1945 J. Marcusson classified asphaltenes as the insoluble fraction in light gasoline
and petroleum ether. More recently, asphaltenes are widely accepted as fraction
derived from carbonaceous sources such as petroleum and coal. They are soluble in
benzene but insoluble in low boiling point paraffin solvent such as n-pentane.

10

In contrast, resins are the soluble fraction in n-pentane (Priyanto et al, 2001).
Therefore, asphaltenes and resins can be separated with appropriate solvents.
Similar concept to separate components in petroleum is illustrated in Figure 1.4. A
simple classification of residual oil’s components with respect to solubility in
different solvents is illustrated in Table 1.3.

Figure 1.4 Separation of petroleum into four major fractions (Speight, 1999)

11

Figure 1.5 Examples of asphaltene formulae (Yen et al, 1994)

Table 1.3 Classification of residual oil’s components (Yen et al, 1994)
Fraction Solubility Remarks
Gas oils Propane soluble Saturated and aromatic hydrocarbons
Resins
Propane insoluble
Pentane soluble
Combined distillates and resins are also
known as maltene or petrolene
Asphaltenes
Pentane insoluble
Benzene soluble

12
In general, residual oil contains approximately 79% to 88% w/w carbon, 7% to 13%
w/w hydrogen, trace to over 6% w/w sulfur, 2% to 8% w/w oxygen, less than 3%
w/w nitrogen and trace metals such as vanadium, copper, titanium, zinc, calcium,
iron which can be found in crude oil. Due to the nature of non-destructive
distillation process used to produce residual oil, majority of sulfur compounds
which are generally considered as higher molecular weight fractions and metals in
the form of salts or organometallic constituents are concentrated in the residual oil
(Sunggyu Lee et al, 2007).

Among all non-hydrocarbon constitutes, sulfur compounds are considered as the
most important due to the corrosiveness which can severely damage piping and
processing units. The major sulfur species in residual oil are alkyl benzothiophene
derivates, dibenzothiophene derivatives, benzonaphtho-thiophene derivatives and
phenanthro-thiophene derivatives (Speight, 1999). Some of the common OSCs
found in residual oil are listed in Figure 1.6.

13

Figure 1.6 Examples of OSCs in residual oils

S
S
Benzo[b]naphtho[1,2-d]thiophene
Benzo[b]naphtho[2,1-d]thiophene

S
Phenanthro[2,1-b]thiophene
Phenanthro[1,2-b]thiophene

S

14
1.3.2 Residual Oil and Sulfur Regulations

Residual oil is mainly used in marine vessels and power plants for power generation;
and in some commercial or industrial buildings for heating and other processing
purposes. Because residual oil is rarely used in dense areas but remote sites,
specifications on residual oil are generally set based on technical instead of
environmental issues. Among all, kinematic viscosity and sulfur content are the
most critical specifications for residual oil. Kinematic viscosity at 100°C for
residual fuel oil should be in the range of 10 to 55 centistoke (mm
2
/s). For residual
fuel oil with higher viscosity, it is usually blended with lighter distillates such as
diesel to achieve a lower viscosity for handling and processing.

Table 1.4 Standard properties of residual oil
Properties Range
Density (at 15°C) 0.975-1.01 (g/cm
3
)
Kinematic Visocity (100°C) 10-55 (cSt)
Flash Point >60°C
Pour Point 0-45°C
Water Content < 1%

15

In the past, maximum sulfur content in residual oil was limited at the range of 3.5%
to 5% in order to protect the engines or boilers. Using higher sulfur content fuel
could cause severe corrosion on engine, due to the formation of sulfur dioxide and
sulfur trioxide during combustion. With excess air, appropriate temperature and
pressure, sulfur dioxide and sulfur trioxide would be converted to sulfurous acid
and sulfuric acid causing damages. This is also known as “cold end corrosion”
because the conversions of sulfur oxides to the corresponding acids happen in
relatively low temperature locations (lower than 150°C) of engines.

In the United State, there is no sulfur cap on heavy oil or residual oil for land-based
uses such as fossil fuelled electricity generation plant. Instead, sulfur dioxide
emission is regulated by EPA’s Acid Rain Program. In some states such as
Tennessee, SO
2
emission from specific fuel is also regulated in certain locations.
According to Tennessee Air Quality Act, any fuel burning installation in Shelby
County using No. 5 and No. 6 fuel oils, which are considered as residual oils, has a
emission limit of 2.7 lbs SO
2
/10
6
BTU. A sulfur cap of 0.3% on all liquid and
gaseous fuels has also been adopted recently to limit SO
2
emission from all
stationary gas turbines in the State of Tennessee (Environment and Conservation,
2009).

16
While another major use of residual oil is on marine vessels, regulations established
by the Marine Environment Protection Committee (MEPC) of the International
Maritime Organization (IMO) are generally applicable to residual oil. The Protocol
of 1997 (MARPOL
*
Annex VI - Regulations for the Prevention of Air Pollution
from Ships) is the first international agreement that limits the sulfur content in all
fuel oil including residual oil to 4.5% by mass. Sulfur content of fuel oil used in
Sulfur Emission Control Areas (SECAs) has to be lower than 1.5% by mass. The
SECAs regulation is applied only to Baltic Sea (enforced in 2005); North Sea and
English Channel (enforced in 2007). The protocol is active since May 19
th
, 2005
(MEPC, 1997).

MARPOL Annex 13, also known as the Revised MARPOL Annex VI, was adopted
in October 10
th
, 2008. A new set of global sulfur caps for all fuel oil will be applied
gradually so as to further reduce SO
x
emission from ships. This new regulations
will be accomplished in two steps: (i) instead of 4.5%, a 1.5% cap will be effective
starting from January 1
st
, 2012; and (ii) a progressive reduction on sulfur level to
0.5%, which will be effective from January 1
st
, 2020. Similarly, the new regulations
for SECAs will first be reduced from the current 1.5% to 1% sulfur cap strating
from July 1
st
, 2010; then be further reduced to a 0.1% sulfur cap starting from
January 1
st
, 2015 (MEPC, 2008).

* MARPOL 73/78 is the major international convention for preventing pollution of the marine
environment by ships from operational or accidental causes

17
Table 1.5 Sulfur content standards for all marine-use fuel oils
Area of Effect Maximum Sulfur Content (%)
Implementation
Date
Global 4.5 May 2005
Global 3.5 January 2012
Global 0.5 January 2020
SECAs 1.5 May 2005
SECAs 1 July 2010
SECAs 0.1 January 2015

18

1.4 Current Desulfurization Technologies

Before the introduction of tightened sulfur standards, blending with low sulfur containing
fuels was a common practice to lower the sulfur content in high sulfur containing fuels
for technical purposes. Nowadays, various technologies on desulfurization have been
developed. It is, however, an important issue to identify if the technologies developed are
able to produce ultra-low sulfur diesel and other low sulfur fuels so as to meet the sulfur
standards set locally and globally; and if these technologies are cost effective.

Up to now, hydrodesulfurization is the main stream in the desulfurization technology
adopted commercially. Due to the tightening of sulfur regulation, traditional
hydrodesulfurization faces its limitation to produce ultralow sulfur fuel. Development of
new technologies becomes a necessity in the refinery industry. As a result, alternative
desulfurization processes including adsorptive desulfurization, biodesulfurization and
oxidative desulfurization are widely discussed as possible technologies to produce
ultralow sulfur diesel.

19
1.4.1 Hydrodesulfurization (HDS)

Hydrodesulfurization, or hydrotreating, is a conventional refinery process for
desulfurization. This is one of the most common desulfurization technologies which
have been applied on naphtha desulfurization since 1950s. Hydrodesulfurization is a
catalytic hydrogenolysis which would result in breaking the C-S chemical bond and
forming C-H and H
2
S.

Traditional HDS reaction takes place in a fixed-bed reactor under high temperature
and high pressure, typically in the range of 290°C to 455°C and 150psi to 3000psi,
respectively. The oil feedstock and hydrogen gas are pumped to the reactor at high
temperature and pressure with the presence of metal catalyst, for instance, cobalt-
molybdenum supported by alumina (CoMo/Al
2
O
3
) or nickel-molybdenum
supported by alumina (NiMo/Al
2
O
3
), producing desulfurized hydrocarbons and
hydrogen sulfide. The HDS chemical equations for mercaptans and sulfides are
listed as below:

S H RH H RSH
O Al CoMo
s
2
/
2
3
+    →  + (Eq. 1.1)

S H H R RH H RSR
O Al CoMo
s
2
/
2
' '
3
+ +    →  + (Eq. 2.1)

20
The mixture of hydrocarbons, hydrogen gas and hydrogen sulfide from the reactor
would then pass through gas separator so as to separate the fuel from hydrogen and
hydrogen sulfide. The mixture of hydrogen and hydrogen sulfide from the gas
separator would be treated by amine gas so as to purify hydrogen gas for reusing in
the reactor. Hydrogen sulfide can be oxidized to sulfur dioxide by air, which can be
further converted to elemental sulfur through the Claus process or sulfuric acid
through the contact process. Chemical equations for these two processes are listed
in Table 1.6.

Table 1.6 Chemical reactions for the Claus process and the contact process
Process Chemical Equations

Claus Process:
Overall reaction: O H S O S H
2 2 2
2 2 2 + → +
Hydrogen sulfide oxidation: O H SO O S H
2 2 2 2
2 2 3 2 + → +
Catalytic conversion: O H S SO S H
TiO
2 2 2
2 3 2
2
+  →  +

Contact Process:
Sulfur dioxide oxidation:
3 2 2
2 2 SO O SO → +
Oleum formation:
7 2 2 4 2 3
O S H SO H SO → +
Sulfuric acid formation:
4 2 2 7 2 2
2 SO H O H O S H → +

21
Fuel obtained from the gas separator would be further treated in stripper distillation
unit with reflux so as to remove sour gas including hydrogen, hydrogen sulfide,
methane, ethane, propane and other volatile organic compounds. Similarly, this gas
mixture would be treated by amine gas to recover hydrogen sulfide. The remaining
fraction can be used for other purposes such as refinery fuel gas.

Figure 1.7 Schematic diagram of HDS unit

Although HDS has been used to produce low sulfur fuel for decades, the upcoming
tightened regulations will be a new challenge on this process. HDS has been proven
to desulfurize mercaptans, sulfides and thiophene. However, the more condensed

22
derivatives including benzothiophenes (BTs) and dibenzothiophenes (DBTs) are
more difficult to treat by HDS. The reactivity of the one- to three-ring OSCs
decreases accordingly: Thiophenes > Benzothiophenes > Dibenzothiophenes
(Girgis et al., 1991).

It is found that low sulfur diesel fuel produced from HDS contains approximately
500 ppm
w
sulfur, mostly alkyl derivatives of DBTs which are considered as
refractory compounds and cannot be easily desulfurized by HDS (Ma et al., 1994).
4-alkyl DBTs, 6-alkyl DBTs and 4, 6-alkyl DBTs have very low reactivity which
can be explained by combinations of electronic density, bond order, and spatial,
geometric and steric hindrance around the sulfur atom (Hans Schulz et al., 1999). 4,
6-dimethyl-dibenzothiophenes are well known for its high stability against HDS.
Thus, traditional HDS is not applicable to produce ULSD required by the new
sulfur regulations.

Figure 1.8 IUPAC numbering for DBT

1
2
3
4 5 6
7

S
9
8

23
Modifications on HDS, including increment of hydrogen, catalyst dosage, and also
operating pressure, have been investigated. In general, hydrogen usage is the major
operational cost for HDS. In order to lower the sulfur content from 500ppm
w
to
15ppm
w
, an addition of 25% to 45% of hydrogen gas is required (Energy
Information Administration, 2001). This implies that the operational cost will be
doubled. On the other hand, doubling catalyst dosage can only lower the sulfur
content by 100 ppm
w
(Whitehurst et al., 1998). Thus, a dramatic increase in catalyst
dosage is required to produce ULSD.

Suggested by the National Petroleum Council, operating pressure has to be
increased from 1100psi to 1200psi in order to produce diesel with less than 30
ppm
w
sulfur. This requires a specific thick-walled reactor to withstand such a high
pressure, and thus increasing the capital and operational costs. Beside the huge
increase in cost, safety issue is also a big concern for HDS which requires high
operating temperature and pressure with the use of hydrogen gas. Reactor wall
failure and even explosion can be resulted if uncontrollable “hot-spots”
phenomenon (Speight, 1994), or any other operation errors happen.

24
1.4.2 Adsorptive Desulfurization (ADS)

Adsorption of OSCs in fuel is another possible desulfurization technology
developed recently. In adsorptive desulfurization process, OSCs are adsorbed into a
specified solid adsorbent so as to produce none- or low-sulfur fuel. Depending on
the interaction between OSCs and the adsorbent, adsorptive desulfurization can be
classified into direct adsorption desulfurization and reactive adsorption
desulfurization. In direct adsorption desulfurization, OSCs are physically adsorbed
on the adsorbent surface. Spent adsorbent can be regenerated by washing with a
desorbent, usually a solvent, and the sulfur compounds can be concentrated simply
by distillation.

PSU-SARS developed at Pennsylvania State University is an example of direct
adsorption desulfurization. This process is basically a composition of direct
adsorption desulfurization and hydrodesulfurization. Fuel feedstock is first treated
with selective adsorption of sulfur compounds by specific transition metal
compounds such as nickel phosphides. The spent adsorbent is regenerated by
solvent washing and the sulfur rich fraction is further concentrated by evaporation
so as to recycle the solvent. The concentrated sulfur fraction is then treated by
hydrodesulfurization unit (Ma, 2001). Figure 1.9 shows the schematic diagram of
PSU-SARS followed by HDS.

25

Figure 1.9 Schematic diagram of PSU-SARS followed by HDS (Ma, 2001)

Reactive adsorption desulfurization, on the other hand, is based on chemical
interaction between OSCs and the adsorbents. In this process, sulfur portion of an
OSC molecule is fixed on the adsorbent, and the sulfur-free hydrocarbon portion is
released to the fuel. Regeneration of sent adsorbent can be done by either oxidation
or reduction of sulfur portion. Depending on the regeneration process selected,
elemental sulfur, hydrogen sulfide or sulfur oxides would be generated (Zhou ed.,
2007).
ADS HDS

26
The S-Zorb Sulfur Removal Technology (S-Zorb SRT) is a representative reactive
adsorption desulfurization process announced by ConocoPhillips Company in 2000.
Fuel feedstock is vaporized at 380-420°C in the presence of hydrogen gas and
injected to the adsorption reactor. Sulfur portion in the OSCs cleaved from the
molecules are adsorbed on the adsorbent, leaving the hydrocarbon portion in the
fuel stream.

Hydrogen gas is primarily used to prevent coke building up on the adsorbent only,
thus hydrogen consumption is relatively low comparing with HDS. Spent adsorbent
is regenerated by an oxidation process, which convert the adsorbed sulfur to sulfur
dioxide. S-Zorb SRT can be used to lower the sulfur content in gasoline and diesel
fuels to a level of 5 ppm
w
with a relatively low capital cost comparing with
hydrodesulfurization units.

Several studies have also demonstrated that adsorptive desulfurization can produce
desulfurized diesel or gasoline with sulfur content less than 30 ppm
w
(Liu et al.
2007 and Tang et al, 2009). However, all these processes require specially prepared
or synthesized adsorbents which are not commercially available at this stage.

Commercially available adsorbents such as activated carbon, activated alumina or
zeolites are reported to be not applicable on adsorptive desulfurization (Takahashi et
al., 2002). Besides, it is predicted that ADS is not a cost effective process to be

27
applied on untreated fuel with high sulfur content. Factors affecting desulfurization
including adsorption capacity, durability, regenerability and selectivity on sulfur
compounds are also a major concern in developing ADS (Kwon et al., 2008).

28
1.4.3 Biodesulfurization (BDS)

Biodesulfurization, an enzymatic process involving the use of bacteria as
biocatalysts to remove OSCs in fuels, is an innovative technology, which uses
bacteria as the catalyst to remove sulfur from the feedstock.

Theoretically, biodesulfurization can be conducted in aerobic or anaerobic
conditions. In aerobic conditions, OSCs are stepwise oxidized and eventually
forming sulfate salts in the presence of sulfur-specific desulfurization microbes such
as R. Rhodochrous, R. erythropolis D-1, Gordona CYKS1 and Rhodococcus UM3.
OSCs are first oxidized to the corresponding sulfoxide, and then to sulfone,
followed by sulfinate, and finally to desulfurized organic portion and inorganic
sulfate ions (McFarland, 1999). While OSCs in fuel are converted to water-soluble
sulfate which can be easily removed, the desulfurized organic potion would stay in
the fuel. Thus, the fuel value would not be degraded. Figure 1.10 shows the general
degradation pathway of DBT in the presence of sulfur-specific desulfurization
microbes.

29

Figure 1.10 Sulfur-specific degradation pathway of DBT

On the other hand, anaerobic biodesulfurization causes the reduction of OSCs to
H
2
S in the presence of sulfur-reducing bacteria such as Desulfovibrio sapovorans.
Although it has been demonstrated as a possible desulfurization pathway in model
compound studies (Armstrong et al., 1995), no significant reduction in sulfur
content in real fuel oil samples is observed under anaerobic biodesulfurization
(McFarland, 1999). The industries are therefore more interested in aerobic
biodesulfurization, and in most cases, the word “biodesulfurization” refers to
aerobic biodesulfurization.

S

S
O
Dibenzothiophene
monooxygenase
Dibenzothiophene
monooxygenase

S
O O
Dibenzothiophene -5, 5-
dioxide monooxygenase
S
OH
-
O
O
-
Dibenzothiophene
monooxygenase
OH
+ SO
4
2-

Dibenzothiophene
Dibenzothiophene
sulfoxide
Dibenzothiophene
sulfone
2-(2-Hydroxyphenyl )
benzenesulfinate
2-Hydroxybiphenyl

30
Recent researches have successfully demonstrated that various sulfur-specific
desulfurization microbes can achieve 57% reduction in BT content (Kirimura et al.,
2002), 90% reduction in DBT content (Li et al., 2006), 50% reduction in some of
the alkyl derivatives of DBTs (Rashidi et al., 2006), higher than 80%
desulfurization in HDS treated diesel (Li et al., 2003) and higher than 45%
desulfurization in heavy oil (Yu et al., 2006).

Nevertheless, there are some factors limiting the applicability of BDS. Biological
and enzymatic reactions are highly sensitive to environmental conditions such as
operating temperature, solvent used, toxin and nutrient availability. Besides, a pre-
HDS treatment is usually required to provide a relatively low sulfur fuel. Up to now,
biodesulfurization has only been tested in bench scale study. Pilot-scale tests with
more detailed designs and cost estimation have not been developed.

31
1.4.4 Oxidative Desulfurization (ODS)

Oxidative desulfurization is considered as the latest unconventional desulfurization
process which involves chemical oxidation of divalent organic sulfur compounds to
the corresponding hexavalent sulfur, also known as sulfone. The physical and
chemical properties of sulfones, for instance boiling points, polarity and solubility
in various solvents, are significantly different from the original sulfur compounds.
In general, sulfones have higher boiling points and increased polarity which leads to
higher solubility in polar solvent. Therefore, sulfones can be easily separated from
fuels through distillation, solvent extraction or adsorption.

Figure 1.11 General process scheme for ODS

OSCs
Oxidation
Phase
Separation
Sulfone
Separation
Fuel
Feedstock
Oxidant &
Catalysts
Recovered
Catalysts
Recovered
Sulfones
Clean
Fuel

32
Theoretically, ODS can be performed by various type of oxidants. Nitric acid and
nitrogen oxides were two popular oxidants used to remove both organic sulfur
compounds and organic nitrogen compounds in 1980s (Tam et al., 1990). Due to
poor selectivity, low yield and loss in heating value for the treated oil, these
oxidants have not been widely used.

Recently, more studies are focused on hydrogen peroxide and organic peroxides as
oxidants in ODS. With application of specific catalysts such as transition metal
complexes and also phase transfer catalysts, hydrogen peroxide can be activated to
effectively oxidize OSCs to sulfones under mild conditions (Yen et al., 2000;
Zapata et al., 2005).

Figure 1.12 General reactions of BT and DBT in ODS (Gatan et al., 2004)

S
S
S
O
S
O O
+ [O]
catalyst
+ [O]
catalyst
+ [O]
catalyst
+ [O]
catalyst
O
S
O O
S

33
After the oxidation process, sulfones separation can be achieved by liquid-liquid
extraction using polar solvents or by adsorption using, for instance, silica gel or
alumina. Spent solvent can be purified by distillation so as to obtain clean solvent
and concentrated sulfones. N, N-dimethylformamide, dimethyl sulfoxide, sulfolane,
methanol and acetonitrile are some of the polar solvents which can be used for
liquid-liquid extraction.

Among all the solvents listed, N, N-dimethylformamide is the most effect solvent to
remove sulfones. Nevertheless, the oil recovery rate would be significantly lowered
when this solvent is used for extraction (Otsuki et al., 2000). It has also been
reported that some of the hydrocarbons such as naphthalene would be extracted
from fuel during liquid-liquid extraction, resulting in reduction in heating value and
fuel quality (Mei et al., 2003).

Instead of liquid-liquid extraction, sulfone adsorption with alumina could be applied
to increase selectivity and reduce loss of valuable hydrocarbons (Etemadi et al.,
2007). The major advantages of ODS include low capital cost, low reactor
temperatures and pressures, short reaction time, no emissions, and no hydrogen
requirement. It has been estimated by pilot plant studies that ODS could be operated
in less than half of the cost of a new high-pressure hydrotreater (Energy Information
Administration, 2001).

34
The Ultrasound Assisted Oxidative Desulfurization (UAOD) process is one of the
most promising ODS systems with greater than 95% oxidation yield of organic
sulfur in short period of time under mild conditions (Wan et al., 2007). However,
the UAOD process cannot successfully produce ultra-low sulfur diesels. The
modified UAOD process with the application of ionic liquids developed later
demonstrated a greater than 98% desulfurization on various diesels which can meet
the ULSD standards (Cheng et al., 2008).

Nonetheless, both the UAOD and the modified UAOD processes require high
dosage of 30% wt. hydrogen peroxide solution and high energy consumptions. Both
systems required 100 fold more oxidant than stoichiometic requirements. Lowered
oxidant concentration would significantly reduce the OSCs oxidation yield (Wan et
al., 2007). On the other hand, a typical 10 minutes ultrasonication using a probe
ultrasonic reactor (model number VCX-750) applied in the UAOD and modified
UAOD process to desulfurize 20g diesel requires 450 kJ of energy, which is
equivalent to approximately 50% of chemical energy in 20g treated diesel,
assuming a energy density of 45 MJ/kg in diesel (Gibilisco, 2006).

35
1.5 Research Objectives

With the tightened environmental regulations on diesel and other fuel oil, traditional HDS
alone is not adequate to meet the upcoming sulfur limits. Oxidative desulfurization has
been proven to be one of the feasible alternatives. Ultrasound Assisted Oxidative
Desulfurization and the modified UAOD system developed recently have demonstrated
more than 95% desulfurization on varies diesels which meets the ULSD standards (Wan
et al., 2007; Cheng et al., 2008). However, high dosage of 30% wt. hydrogen peroxide
solution and high energy consumptions are required in those processes.

In order to improve ODS efficiency and its applicability to heavy oil such as residual oil,
this research investigates the use of alternative oxidants such as superoxide of alkali
metals. Experiments were carried out to optimize ODS process as well as evaluate the
desulfurization effectiveness in different fuel oil including residual oil. Furthermore,
studies on an alternative enhancement technology, namely, ultraviolet assisted oxidative
desulfurization, has also been performed to evaluate its feasibility and its potential in
reducing energy consumption.

36
CHAPTER 2: THEORETICAL BACKGROUND
2.1 Introduction

In order to improve the Ultrasound Oxidative Desulfurization process so as to develop a
technology to produce ultralow sulfur fuels with reasonable chemical consumption, a
series of process modifications have to be considered.

The UAOD process includes the application of ultrasonication, hydrogen peroxide as
oxidizing agent, acid catalysis, and phase transfer catalysis. Based on the UAOD process,
the following elements were considered in the modifications: ultrasonication; oxidant
selection; acid catalysis; phase transfer catalysis; application of ionic liquid; and photo-
catalysis. The basic concepts of these six elements listed are discussed in this chapter.

37
2.2 Ultrasonication
2.2.1 Fundamentals of Ultrasound

Sonic wave is periodic vibration with frequency in between 15 Hz to 20 kHz which
is audible to the average human. Ultrasound, on the other hand, is the radio wave
with frequency higher than 20 kHz but lower than 100 MHz (Berlan et al., 1996).
Nowadays, ultrasound is commonly applied to various industries including
chemical synthesis, biotechnology and environmental engineering. Ultrasound
frequency level is inversely proportional to the power output. Typically, low-power,
high frequency ultrasound with frequency higher than 1 MHz is considered as non-
destructive ultrasound. Ultrasound in this frequency range does not affect the
medium it travels through. Thus, it is commonly used in medical sonography or
other medical diagnosis.

On the other hand, high-power, low frequency ultrasound with frequency ranged
from 20 kHz to 100 kHz does alter the medium it travels through. This type of
ultrasound is mostly applied on sonochemical reactions. It has been demonstrated
that ultrasound in this range is applicable to improve mixing, increase rate of
chemical reactions, promote emulsification, and others (Thompson et al., 1999).
Ultrasound with frequency ranged from 100 kHz to 1 MHz has medium power
intensity. It is usually applied on biomedical treatments such as focused ultrasound
surgery (FUS) or ultrasound based physical therapy (Baker et al., 2001).

38
2.2.2 Theory on Sonochemistry

The Chemical and mechanical effects of high-power, low frequency ultrasound
were first identified in 1930s from a phenomenon called acoustic cavitation or
ultrasonic cavitation (Suslick et al., 1999). Cavitation is a unique phenomenon in
which a relatively low energy of an acoustic field is concentrated in very small
volumes, resulting in a relatively high energy density locally. It is defined as the
pulsation, oscillation, growth, splitting and other motion of bubbles and their
interaction due to a first reduced, then an increased pressure produced in a liquid
(Margulis, 1995).

When the negative pressure produced at the rarefaction period of a sound wave
exceeds the van der Waals force among molecules in the liquid, it would initiate the
formation of cavitation bubbles at gas or solid particles (Mason, 1999). These
cavitation bubbles would grow to an equilibrium size or resonance size when the
resonance frequency of the bubbles equals the ultrasound frequency applied. When
the cavitation bubble grows to size greater than the resonance size, the bubble
would collapse and generate a local high pressure greater than 1000 atm and high
temperature up to 5000K instantaneously (Storey et al., 2001).

39
Sonochemical activity and occurrence of acoustic cavitation could be affected by
factors including ultrasound frequency, ultrasonic power, temperature, reactor
pressure, solvent properties and sparge gas applied (Beckett et al., 2000).

Sonoluminesence is another phenomenon happening when ultrasound is applied in
a liquid. It is the light emitted when cavitation bubbles collapse, which implies the
existence of local high temperature. This phenomenon was first discovered by H.
Frenzel and H. Schultes by putting an ultrasound transducer in photographic
developer fluid in 1934 (Crum et al., 1994). This discovery was later recognized as
multiple-bubble sonoluminescence (MBSL). The concept of sonoluminescence is
illustrated in Figure 2.1.

Another type of sonoluminescence, single-bubble sonoluminescence (SBSL), was
discovered by F. Gaitan and L. Crum in 1989 (Gaitan et al., 1992). The peak
temperature could be ranging from 6000K to approximately 20000K. Production of
active chemical species such as hydrogen peroxide and hydroxyl radicals are
reported (Didenko et al., 2002).

Different theories, including the electrical theories, the mechanochemical theory,
the chemiluminescence theory, the hot spot theory, the shock wave theory and
others, have been developed to explain these phenomena caused by ultrasound
based on light emission mechanisms. Among all, researchers are more interested in

40
the electrical theories, the hot spot theory and the shock wave theory. It was first
proposed that the light emitting phenomenon was a result of charge separation in
cavitation bubbles, the charge fluctuations when bubbles collapse and other
electrical microdischarge (Levshin et al., 1937). Electrical theories have a common
assumption that emitting bubble has asymmetric charge distribution, which is found
later to be contradicted with systematic studies on SBSL (Ohl, 2000).

Figure 2.1 The sonoluminescence process (Lohse, 2002)

41
The hot spot theory is described that the energy for light emission causing
sonoluminesence is supplied by thermal energy generated during cavitation bubble
collapse. The hot spot is believed either a black body where the radiation and matter
are approximately in equilibrium (Noltingk et al., 1950), or Bremsstrahlung, also
known as free-free transitions caused by accelerating unbounded electrons (Yasui,
1999). The hot spot theory generally suggested that an instantaneous increase of
temperature to a range higher than 10000K, and that of pressure up to 1800 atm
would occur during bubble collapse, would be responsible for the sonoluminesence.

The shock wave theory proposed by Jarman in 1960s suggests that mircroshocks
propagated within the imploding bubbles induce high temperature and pressure
causing sonoluminesence (Taylor et al., 1970). This theory is now considered as an
alternative to the hot spot theory, and the possibility of microscale explosive
shockwave synthesis at the final stage of bubble collapse has been discussed in
various studies (Greenspan et al., 1993; Crum et al., 1998; Young, 2005).

42
2.2.3 Sonochemistry in Aqueous Phase

Applying high-power low frequency ultrasound to aqueous system would cause the
formation of cavitation bubbles and sonoluminescence. Although part of the energy
would be released as light and heat, more energy would be involved in chemical
reactions. As listed in Table 2.1, it is demonstrated from a single cavitation bubble
model in aqueous solution that energy for sonochemical reactions is more than 100
fold of the energy for sonoluminescence (Didenko et al., 2002).

Local high temperature and high pressure generated due to collapse of cavitation
would lead to the formation of intermediate radical species including hydroxyl
radical (OH．), hydrogen radical (H．), and hydroperoxyl radical (HO
2
．). These
radicals, especially OH ．, are highly active and are precursors of many other
chemical reactions. Formation of hydrogen peroxide by ultrasonic irradiation is also
of particular interests, because it is usually used as an indirect measurement of OH．
production.

43

Table 2.1 Sonochemistry in a cavitation bubble formed in water (Didenko et al., 2002)
Conditions 3°C 22°C
Number of OH． radicals per cycle 8.2×10
5
6.6×10
5

Number of NO
2
-
ions per cycle 9.9×10
6
3.7×10
6

Number of photons per cycle 7.5×10
4
8.1×10
3

Maximum potential energy of bubble (eV) 7.5×10
10
6.4×10
10

Energy to form OH． radicals (eV per cycle) 4.3×10
6
3.4×10
6

Energy to form NO
2
-
ions (eV per cycle) 4.2×10
6
1.6×10
6

Energy to form photons (eV per cycle) 2.6×10
5
2.7×10
4

Energy efficiency of sonochemistry 1.1×10
-4
7.8×10
-5

Energy efficiency of sonoluminescence 3.5×10
-6
4.3×10
-7

Ultrasound causes formation of radicals similar to ionizing radiation, which split
water molecules into H． and OH． (Yazici et al., 2006). It is also reported that
atomic oxygen can be generated from dissolved oxygen in aqueous solution upon
ultrasound irradiation (Fang et al., 1995). These radicals and atomic oxygen would
further react with dissolved oxygen, water molecules, or other radicals to form
hydrogen peroxide and oxygen eventually.

44
A summary of these chemical reactions according to the initial active species
generated from ultrasound is listed in Table 2.2. It has also been reported that the
presence of ozone (O
3
) would promote generation of free radicals and reactive
oxygen species. The presence of inert gas, argon (Ar) for instance, would promote
sonoluminescence reactions (Beckett et al., 2001). Figure 2.2 illustrates the
chemical reactions during acoustic cavitation with gases including Ar, O
2
and O
3
.

Table 2.2 Chemical reactions initiated by ultrasound in water (Mason et al., 2002)
Radiolysis of H
2
O
⋅ + ⋅   →  OH H O H
ultrasound
2

⋅ →  + ⋅
2 2
HO O H

2 2 2 2 2
O O H HO HO + →  ⋅ + ⋅

2 2
O H OH OH →  ⋅ + ⋅

Radiolysis of O
2

O O
ultrasound
2
2
  → 

⋅ →  + OH O H O 2
2

2 2
O H OH OH →  ⋅ + ⋅

45

Figure 2.2 Sonochemical reactions in water with Ar, O
2
and O
3
(Beckett et al., 2001)

46
2.2.4 History of UAOD and Its Modifications

When ultrasound is applied, very fine cavitation bubble would form. This
phenomenon can greatly improve emulsification in a system with two or more
immiscible liquid phases. This would greatly increase the contact area and thus, the
mass transfer of reactants. Combining the effect of local high temperature and
pressure produced during collapse of cavitation bubble, ultrasound could greatly
increase the rate of reactions, especially in a system involving multiple liquid
phases such as UAOD process (Mei et al., 2003).

UAOD process has been under development since 1990s. At the very beginning of
the UAOD history, desulfurization was observed by applying a 20 kHz ultrasound
to crude oil suspended in basic or acidic conditions. Use of transition metals such as
nickel and vanadium as catalysts with hydrogen peroxide as a chemical assisted
ultrasound method was developed (Lin et al., 1993; Sadeghi et al., 1994; Yen, 1998).

In the first generation of UAOD, or simply the UAOD process, the Venturella type
of transition metal catalysts (TMC), or polyoxometallates with Keggin structure
such as phosphotungstic acid (H
3
PW
12
O
40
), and a quaternary ammonium salt,
tetraoctylammonium bromide, as phase transfer catalyst (PTC) were introduced so
as to produce high effective and selective oxidation with hydrogen peroxide to
improve desulfurization (Mei et al., 2003). Structure of phosphotungstic acid is

47
illustrated in Figure 2.3. In the UAOD process, higher than 98% desulfurization on
selected diesel samples can be achieved in 10 minutes ultrasonication with 30% wt.
hydrogen peroxide as oxidant, together with the selected TMC and PTC. However,
it has been demonstrated that the UAOD process has a low conversion on
benzothiophene (BT) and its derivatives. Brominated byproducts are also identified
in the treated oil samples.

Figure 2.3 The Keggin structure of [PW
12
O
40
]
3-
(Bochet et al., 2009)

48

Figure 2.4 Conceptual model of 1
st
generation of UAOD process (Wan et al., 2007)

Changing phase transfer agent in the second generation of UAOD led to
desulfurization without brominated byproduct. Instead of bromide as anion,
quaternary ammonium salt with fluoride as anion was used as phase transfer
catalyst. This eliminated the formation of byproduct, and also increased the overall
desulfurization efficiency (Wan et al., 2007).

49
The second generation of UAOD demonstrated greater than 95% desulfurization in
various diesel samples within 10 to 20 minutes. A portable, continuous flow
desulfurization unit was customized based on the second generation UAOD (Wan et
al., 2008). The portable UAOD unit can treat a maximum of 52.8 lb diesel per day
with 92% desulfurization. In order to reduce hydrocarbon loss, alumina adsorption
was applied as a post treatment to remove sulfones instead of liquid-liquid
extraction (Etemadi et al., 2007).

The recent development of the third generation of UAOD has included the addition
of room temperature ionic liquid (RTIL) and organic acid, into the system. Ionic
liquid can serve as both extracting reagent and phase transfer catalysts which would
further improve desulfurization efficiency. The third generation of UAOD can
achieve greater than 99.9% desulfurization in various diesel samples by mechanical
mixing for 3 hours with or without an addition of 10 minutes ultrasonication (Cheng
et al., 2008). Ultra-low sulfur diesel with sulfur content less than 15 ppm
w
can be
produced from various diesels by this process.

A simplified comparison of the UAOD process (1
st
generation of UAOD), portable
UAOD (2
nd
generation of UAOD) and the Modified UAOD process (3
rd
generation
of UAOD) is summarized in Table 2.3. With the progressive improvement from the
UAOD process to the modified UAOD process, ULSD can now be produced under
mild conditions. However, energy and oxidant consumption in all three processes

50
are extremely high. These would reduce the applicability of the processes in the
industry.

Table 2.3 Comparison of the UAOD process and its modifications
UAOD Portable UAOD Modified UAOD
Total Reaction Time, min 10 60 120
Sonication Time, min 10 60 10
Power Used, W 750 100 750
Energy Used, kJ 450 360 450
Temperature, °C 70 70 70
[S]:[O] ratio, mol:mol 1:200 1:200 1:100
Desulfurization Yield of BT* 50% 95% > 98%
Desulfurization Yield of MGO 95% 95% > 98%
Brominated Byproduct? Yes No No
Produce ULSD? No No Yes
* based on model compound (BT) studies

Besides, it is reported that some RTILs, for instance imidazolium type ionic liquid,
would undergo thermolysis by ultrasonic irradiation (Oxley et al., 2003). Oxidative
degradation of ionic liquid is also demonstrated in the presence of hydrogen
peroxide, acetic acid and ultrasonication (Li et al., 2007). Selection on RTIL was
studied during the third generation of UAOD. It was found that the typical type of
RTIL, 1, 3-dialkylimidazolium hexafluorophosphate would be degraded to reactive
gases including hydrogen fluoride (HF), phosphorus pentafluoride (PF
5
) and
phosphorus oxyfluoride (POF
3
), which can severely damage glass and other

51
equipment. Ionic liquids with alkylsulfate as anion have been used in the sixth
generation of UAOD instead. Alkylsulfate anion-based ionic liquids are relatively
more stable, and produce non-corrosive byproducts upon degradation (Jess et al.,
2004).

Figure 2.5 Degradation of specific ionic liquid with ultrasound (Li et al., 2007)

52
2.3 Oxidant Selection

Oxidizing agent is one of the key elements in any oxidative desulfurization process. To
achieve oxidative desulfurization, oxidant is required to oxidize organic sulfur
compounds into the corresponding sulfoxides or sulfones with higher polarity. With
higher polarity than other hydrocarbon, sulfoxides or sulfones can be easily removed
from fuel by extraction, adsorption or other post-treatments.

Wide variety of oxidants, such as concentrated nitric acid (Tam et al., 1990), organic
hydroperoxides (Boikov et al., 2008), peroxyacids (Tetsuo et al., 1994), hydrogen
peroxide (Mei et al., 2003), permanganate (Dehkordi et al., 2008), ozone (Otsuki et al.,
1999) and oxygen (Campos-Martin et al., 2004) are considered as possible oxidants for
ODS processes. Among all, hydrogen peroxide is considered as the most promising
oxidant in terms of selectivity, availability, safety, cost effectiveness and environmental
influence (Filippis et al., 2003).

53
2.3.1 Hydrogen Peroxide

Hydrogen peroxide (H
2
O
2
) is considered as “green” reagent which is commonly
used in oxidative desulfurization processes. With the aid of catalysts, hydrogen
peroxide can oxidize OSCs to the corresponding sulfones in ambient conditions.
During reactions or degradation, water and oxygen are the only by-products which
are in general, considered to have no adverse effect on the environment. The
degradation of hydrogen peroxide can be illustrated in the following chemical
equation:

2H
2
O
2
Æ 2H
2
O + O
2
(Eq. 2.1)

Pure hydrogen peroxide is a pale blue to colorless liquid with density higher than
water. Hydrogen peroxide is miscible with water in any portion forming a colorless
solution. Nowadays, hydrogen peroxide is widely used as disinfectant, oxidizing
agent, and even as a propellant. Due to the high reactivity and oxidizing power,
hydrogen peroxide is considered as a reactive oxygen species (Takishima, 1994).

Although dilute (27.5% wt. or lower) hydrogen peroxide are considered stable and
safe for storage, concentrated hydrogen peroxide are corrosive and extremely
reactive. In fact, solution with higher than 35% wt. hydrogen peroxide are
considered as possible cause of spontaneous ignition of combustible materials if

54
contacted. It becomes unstable at elevated temperature and/or pressure. Hydrogen
peroxide at 52% wt. solution or above could cause a significantly higher rate of
spontaneous ignition if contacted with combustible materials. It could also undergo
vigorous self-sustained decomposition or even explosive reaction if it is exposed to
heat or contaminants. At 91% wt. or higher, hydrogen peroxide solutions are used as
rocket propellant which can undergo explosive reactions.

Due to safety reasons, aqueous solution of 30% wt. of hydrogen peroxide is more
commonly used. Although pure hydrogen peroxide has a high active oxygen ratio
(Bregeault, 2003), dilution effect should be considered as it would significantly
reduce the active oxygen ratio. Active oxygen ratios of some common oxidizing
agents are listed in Table 2.4.

Table 2.4 Oxygen-donor Oxidants (Bregeailt, 2003)
Oxidant Active Oxygen (% wt.) By-product
H
2
O
2
(pure) 47.1 H
2
O
O
3
33.3 O
2

HNO
3
25.4 NO
x

t-BuOOH 17.8 t-BuOH
H
2
O
2
(30% wt.) 14.1 H
2
O

55
2.3.2 Superoxide Anions

Superoxide ion, ⋅
−
2
O , is a free radical with one unpaired electron. Many types of
superoxide are considered stable in ambient conditions in the absence of water.
Upon contacting with water, it undergoes reaction forming oxygen and hydrogen
peroxide, as illustrated in the following equation:

− −
+ + → + ⋅ OH O H O O H O 2 2 2
2 2 2 2 2
(Eq. 2.2)

Solid state superoxide is available metal superoxides or organic compounds
superoxides such as tetraalkylammonium superoxides. The stability of metal
superoxides depends on the electropositivity of the metal cation. The
electropositivity decreases along the period, and increases down the group of the
periodic table. Therefore, alkali metal with higher atomic weight gives a greater
stability to its superoxide. Metal superoxides are stable even in high purity at dry
ambient conditions. Thus, it can provide a high active oxygen ratio. For instance,
potassium superoxide has an active oxygen ratio of 45% wt.

Similar to hydrogen peroxide, superoxide anion is also considered as a reactive
oxygen species. In fact, researches have found hydrogen peroxide, the superoxide
anion radical, and the hydroxyl radical the most important reactive oxygen species
(Callahan et al., 2001). Superoxide can generate other reactive oxygen species

56
including hydrogen peroxide, hydroxyl radical, and perhydroxyl radical by a series
of free radical reactions (Foote, 1995).

Although superoxides have not been widely applied to oxidative desulfurization
process, chemical reactions between superoxides and organic sulfur compounds
were recorded in early 1980s. For instance, thiol, alkyl disulfides and aryl disulfides
can be oxidized by superoxide anion to the corresponding sulfinic acids or sulfonic
acids (Oae et al., 1981). More complex organic sulfur compounds such as
thioamides, thioureas and thiouracils can also be oxidized by superoxides forming
amides or other corresponding hydrocarbons, and also elemental sulfur or inorganic
sulfate to achieve desulfurization (Chang et al., 1989, Kim et al., 1990). Some of
the reactions between OSCs and superoxide are illustrated in Figure 2.6.

Figure 2.6 Chemical reactions between OSCs and superoxide anion (Afanas’ev, 1989)
Oxidation of Thiols
2RSH + 3O
2
¯· Æ RSO
2
¯ + RSO
3
¯ + H
2
O

Oxidation of Disulfides
2RSSR’ + 5O
2
¯· Æ 2RSO
2
¯ + 2R’SO
3
¯

Oxidation of N-(2-hydroxy-phenyl)-N-Phenylthioureas

57
2.4 Acid Catalysis

Although oxidative desulfurization can be carried out through noncatalytic oxidation or
catalytic oxidation of OSCs, noncatalytic oxidative desulfurization processes requires
high temperature of approximately 200°C and high pressure (Paniv et al., 2006). In
contrast, catalytic oxidative desulfurization requires relatively mild conditions with
temperature ranging from 25°C to less than 100°C under ambient pressure. Various types
of ODS catalysts including, aldehydes (Murata et al., 2004), transition metal salts (Chen
et al., 2007), polyoxometalate acids (Rosa et al., 2006) and carboxylic acids (Ma et al.,
2001), while organic acids and polyoxometalate acids are the most commonly adopted.

Oxidation catalysis with polyoxometalates is considered to be complicated and diverse.
Basically, polyoxometalate anions are capable of reversible redox reactions to actively
transfer oxygen to the targeting compounds for selective oxidation (Bäckvall, 2004). For
instance, active peroxo polyoxometalates can be formed by the interaction between the
polyoxometalate anion and hydrogen peroxide. The active peroxo polyoxometalates can
be transferred to organic phase easily with the aid of a phase transfer agent.

Polyoxometalates are relatively thermal stable comparing with other transition metal
catalysts. Beside, polyoxometalate acids are well known for the high Brönsted acidity.
Generally, Brönsted acidity of polyoxometalate acids could be stronger than that of
mineral acids such as sulfuric acid by several orders of magnitude. Brönsted acidity

58
of different polyoxometalate acids are in the following sequence: H
3
PW
12
O
40
>
H
4
PVW
11
O
40
> H
4
SiW
12
O
40
> H
3
PMo
12
O
40
> H
4
SiMo
12
O
40
(Borras-Almenar et al., 2003).
Nevertheless, polyoxometalates are not stable towards strong oxidants such as hydrogen
peroxide in aqueous solution. Decomposition of polyoxometalates to peroxometalates
were observed and thus, causing loss of catalyzing ability (Kozhevnikov, 2002).

Carboxylic acids, or more commonly referred to formic acid and acetic acid, are also
applicable in catalytic oxidative desulfurization processes. Coupling with hydrogen
peroxide, peroxyl acids could be generated in situ (Greenspan, 1947). It is suggested that
peroxyl acids are some of the most possible oxidants for selective oxidation of OSCs
(Lanju et al., 2008; Wang et al., 2003).

59
2.5 Phase Transfer Catalysis
2.5.1 Overview of Phase Transfer Catalysis

Phase transfer catalysis is a major enhancement in many chemical reactions
involving multiple phases, especially for those which are immiscible in each other.
Generally, reactions could be greatly inhibited due to reactants separation by
immiscible phases. Introducing a phase transfer catalyst (PTC) would improve the
miscibility of phases, thus increasing chance of reactants contact so as to enhance
reaction rate.

One of the most common uses of PTC is to enhance reactions between organic (oil)
and inorganic (aqueous) phases. Due to the significant improvement in reaction rate,
using PTC would generally increase productivity, improve quality and reduce cost.
Phase transfer catalysis has been applied for many applications, including chemicals
and pharmaceuticals manufacture, petroleum processing, and other industries
(Starks et al., 1994).

60
2.5.2 Mechanism of Phase Transfer Catalysis

It is known that ionic compounds such as salts and other polar compounds with
strong electric dipole are difficult to enter organic phase. Phase transfer catalysis is
based on the ability to improve reaction rate of reagents in different phases of a
reaction mixture by accelerating interfacial transfer. Depending on the ions required
for reaction in organic phase, phase-transfer catalysis can be performed by cation or
anion transfer.

Cation transfer is usually related to metallic cation transfer. Neutral complexants
such as crown ethers and cryptands for inorganic cation transfer. Crown ethers are
considered as some of the most common PTCs for metallic cations transfer. Crown
ethers are heterocyclic compounds with several ether groups. Specific crown ether
is a strong binding agent for specific cation to form stable complex. Basically,
oxygen in the ring structure of crown ether would coordinate with the cation,
locating it at the interior of the ring structure and leaving the exterior of the ring as
hydrophobic. As a result, the cation would be stabilized and become soluble in
organic or nonpolar phase. Affinity of crown ether for a cation depends highly on
the structure and denticity. For instance, 15-crown-5 has high affinity for sodium
cation, while 18-crown-6 has high affinity for potassium cation (Guida et al., 1980).

61

Figure 2.7 Structures of crown ethers: 15-crown-5 and 18-crown-6

Anionic transfer, on the other hand, requires stabilizing the target anion, or in some
case neutral molecule, from the polar phase to non-polar phase. PTC is necessary
because anions and neutral compounds which are soluble in aqueous phases are
generally not soluble in organic phases, whereas the organic reactants in organic
phases are not soluble in aqueous phases. These type of phase transfer catalysts are
usually referred to as salts of onium cation, including ammonium, phosphonium and
arsonium, where quaternary ammonium salts (QASs) are the most common PTCs
for anion transfer (Starks, 1971). For example, tetrahexylammonium chloride,
(C
6
H
13
)
4
N
+
Cl
-
, can be used to catalyze the reaction between cyanide in aqueous
phase and chlorooctane, a reaction which is theoretically not feasible without PTC.
The concept of Starks’ extraction is illustrated in Figure 2.8.

62

Figure 2.8 Starks’ extraction mechanism

Phase transfer catalysts such as QASs are commonly considered as loosen ion pairs
comparing with normal salts such as sodium chloride. The looseness of the ion pair
is a major reason for enhanced reactivity. According to Starks’ extraction
mechanism, phase transfer catalysis is applied on nucleophilic substitution reaction:

RX + Y
-
Æ RY + X
-
(Eq. 2.3)

where Y
-
is the active nucleophile required to be transferred from aqueous phase
into the organic phase. An extraction of the active nucleophile Y
-
can be performed
by addition of quaternary ammonium cation Q
+
, so that an ion pair [Q
+
Y
-
]

63
would be formed and entering the organic phase for the nucleophilic substitution
reaction to happen. Q
+
would recombine with X
-
released and reach equilibrium
between phases, so as to form a cycle of catalytic reactions.

For successful phase transfer catalysis, it is not only required to transfer the
targeting ion to the organic phase, but also activate it or render it to highly active
form. QASs can be used for anion activation due to a longer separation between
cation and anion in comparison to normal metallic salts (Starks et al., 1994).

Quaternary ammonium salts can also reduce the levels of hydration around the
active nucleophile by selecting an appropriate anion (Jones, 2001). The less water
of hydration around the active nucleophile, the more reactive it is in most organic
phase reactions. Thus, the efficiency of quaternary ammonium catalyst is greatly
influenced by the choice of counter anion of the catalyst. The mechanism of
nucleophile activation by reducing level of hydration is illustrated in Figure 2.9.
Table 2.5 listed the degree of hydration of ionic anions.

64

Figure 2.9 Modified Starks’ extraction mechanism with water of hydration

Table 2.5 Hydration of anion in chlorobenzene-aqueous system (Jones, 2001)
Anion Hydration of Anion in Chlorobenzene
F
-
8.5
RCO
2
-
4.0
Cl
-
3.0
Br
-
2.0
I
-
1.1

65
2.5.3 PTC Selection in UAOD Process

In order to enhance organic sulfur oxidation, phase transfer catalysis has been
applied in different desulfurization processes. It has been reported that thiophenol,
and benzyl mercaptan can be oxidized by chromium trioxide to the corresponding
disulfide with 18-Crown-6 as the phase transfer catalyst and dichloromethane as
solvent (Juaristi et al., 1984).

In biphasic reactions with a highly polar solvent such as ethyl acetate, methanol or
water, ammonium salts are more effective PTC compared with crown ethers.
Organic sulfur compounds are generally oxidized to the corresponding sulfoxides or
sulfones under these conditions. For example, tetrabutylanunonium bromide is used
as a phase transfer agent in the oxidation of organic sulfides to sulfoxides with
periodate as oxidant (Venkatachalapathy et al., 1999).

Before the modification of the second generation of the UAOD process,
tetraoctylammonium bromide was used as the phase transfer catalyst. However,
formation of brominated by-products was observed. Bromine formation during the
oxidation process with hydrogen peroxide and acid catalysts could be one of the
possible reasons for the formation of by-products. Equation 2.4 illustrates the
oxidation of bromide to bromine in acidified hydrogen peroxide solution.

66

H
2
O
2
+ 2H
+
+ 2Br
-
Æ 2H
2
O + Br
2
(Eq. 2.4)

According to the standard electrode potentials listed in Table 2.6, the oxidation of
bromide by acidified hydrogen peroxide is a spontaneous reaction. This can be
proved easily by a simple experimental setup which mixes tetraoctylammonium
bromide with acidified hydrogen peroxide solution. A brownish color would be
observed instantaneously upon mixing, indicating the oxidation of bromide ion to
bromine. Bromine formed could undergo substitution reactions or addition reactions
forming brominated by-products.

Table 2.6 Standard electrode potentials for selected half-reactions (Milazzo et al., 1978)
Half-Reaction Electrode Protential (V)
F
2
(g) + 2e
−
→ 2F
−
(aq) +2.87
H
2
O
2
(aq) + 2H
+
+ 2e
−
→ 2H
2
O +1.76
Cl
2
(g) + 2e
−
→ 2Cl
−
(aq) +1.36
Br
2
(aq) + 2e
−
→ 2Br
−
(aq) +1.09

67

Starting from the second generation of the UAOD process, a specific quaternary
ammonium salt, tetraoctylammonium fluoride (TOAF), is used to improve the
transfer and reactivity of hydrogen peroxide to oxidize organic sulfur compounds.
Because fluoride ion cannot be easily oxidized to fluorine by hydrogen peroxide, no
fluorinated by-products would be formed. In addition, fluoride ion has a high
degree of hydration, which suppresses the degree of hydration of the active
nucleophile. The application of TOAF in the UAOD process has significantly
improved the desulfurization efficiency by enabling a better active oxygen transfer
from aqueous phase to organic phase, enhancing oxidation of OSCs to the
corresponding sulfones (Wan et al., 2007).

Ultrasound serves as a co-agent to lower surface tension and enhance emulsification
through micro-bubbles production. Thus, the surface area between reactants would
be increased so as to promote reactions (Thompson et al., 1999). With the
conjunctive use of ultrasound, acid catalysts and phase transfer catalyst, the
desulfurization effectiveness of the UAOD process could be highly increased
(Wang et al., 2007). By applying TOAF in the UAOD process to enhance oxidation
of organic sulfur compounds in a diesel-hydrogen peroxide emulsion, greater than
95% desulfurization can be achieved on different diesels (Mei et al., 2003; Wan et
al., 2007).

68
2.6 Ionic Liquids (ILs)
2.6.1 Overview of ILs and RTILs

Ionic liquids (ILs), in broad definition, are liquid with ions as the major constituents.
In other words, they are the liquid states of any ionic compounds. As defined in
early 1940s, ionic liquids include all molten salts such as molten sodium chloride
(Barrer, 1943). Development of ionic liquids was initiated in electrochemistry.
While most of the electrolytes used were metallic salts in aqueous solution,
limitation in conductivity and side reactions from water, or more specifically,
hydrogen ions and hydroxide ions, usually suppressed the desired electrochemical
reactions (Bockris et al., 1998).

The concept of “zero solvent electrolyte” was developed to eliminate the loss in
conductivity and effectiveness. In room conditions, however, pure metallic salts are
usually in solid states with a relatively low conductivity and applicability. High
temperature was applied to loosen the ionic lattices of salts forming the
corresponding ionic liquids. Specific conductivity of the molten salts can be
increased by a thousand fold compared to the salt crystals. Specific conductivities
of water, aqueous solution of sodium chloride, molten sodium chloride and molten
potassium chloride are listed in Table 2.7.

69
Table 2.7 Specific conductivities (Bockris et al., 1998)
Substance Temperature (K) Specific conductivity (s cm
-1
)
H
2
O
l
291 4×10
-8

NaCl
aq
(5M) 298 0.25×10
-3

NaCl
l
melt 1181 3.903
KCl
l
melt 1145 2.407

With the rising concerns in pollutions and clean technologies, Montreal Protocol
was established to limit the use of volatile organic solvents including
chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons (HCFCs). These are
greatly affected organics synthesis processes in various industries such as
petrochemical and pharmaceutical industries (Hough et al., 2007). The use of ionic
liquids with relatively low melting points as solvents has been considered as
potential substitutes for traditional volatile organic solvents.

Nowadays, ionic liquids are redefined as liquids of fused salts containing only ions
with melting point below 100°C (Wasserscheid et al., 2002). Ionic liquids are
generally neither flammable, nor explosive (Fox et al., 2008). It is demonstrated
that ionic liquids have excellent solvent properties towards various organic and
inorganic chemicals. The general properties of ionic liquids are summarized in
Table 2.8.

70

Table 2.8 General properties of modern ionic liquids (Johnson, 2007)
Properties Descriptions
Melting Point < 100°C
Boiling Point > 200°C
Thermal Stability High
Viscosity < 100 cP
Polarity Moderate
Specific conductivity < 1×10
-2

Vapor pressure Negligible

With high boiling points, high thermal stability and negligible vapor pressure, there
is almost no solvent emission at room temperature. Emission of volatile organic
compounds (VOCs) can be minimized by replacing traditional organic solvents
with ionic liquids and thus, ionic liquids are considered as green solvents
(Huddleston et al., 1998).

Nevertheless, some ionic liquids are found to be toxic (Zhao et al., 2007). Planning
to prevent accidental discharge and design for less toxic ionic liquids could be done
to maintain the potential benefits. Table 2.9 shows the toxicity of three ionic liquids,
including 1-butyl-3-methylimidazolium chloride, 1-butyl-3-methylimidazolium
tetrafluoroborate and 1-butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)-
imide.

71

Table 2.9 Toxicity of ionic liquids, expressed as EC
50
in µM (Matzke et al., 2007)

1-butyl-3-methyl-
imidazolium chloride
1-butyl-3-methyl-
imidazolium
tetrafluoroborate
1-butyl-3-methyl-
imidazolium bis(trifluoro-
methylsulfonyl)imide
Growth inhibition
(duckweed)
660 310 380
Growth inhibition
(wheat)
>3000 1700 110
Growth inhibition (cress) >3000 1900 400
Reproduction inhibition
(green algae)
140 130 50
Reproduction inhibition
(springtail)
>3000 >4400 30

Because the properties of ionic liquids such as melting point, boiling point,
viscosity, and hydrophobicity can be simply adjusted by altering the structures of
ions, ionic liquids are also known as designer solvents (Freemantle, 1998). Due to
the unique properties and their stabilities, ionic liquids can serve as, not only carrier
solvents for organic or inorganic chemical reactions, but also solvents for separation
and co-catalysts. It is demonstrated that ionic liquids can be used to separate
organic compounds, for instance, olefins from paraffins (Munson et al., 2002).

Ionic liquids can also be applied as co-catalysts or catalysts in reactions including
hydrogenation of cyclohexene (Suarez et al., 1996); oxidation of 2, 2-
dimethylchromene (Song et al., 2000); Knoevenagel condensation of various

72
ketones and aldehydes (Ranu et al., 2006); ethylbenzene production (Rogers et al.,
2003); and biocatalytic esterification of carbohydrates (Rantwijk et al., 2003).

Room temperature ionic liquids are ionic liquids which maintain in liquid state at or
below room temperature. The first RTIL ethylammonium nitrate, or [EtNH
3
][NO
3
],
was discovered early in 1914 (Earle et al., 2000). RTILs usually consist of bulky
organic cations such as imidazolium, pyridinium, pyrrolidinium, alkylammonium,
alkylsulfonium and alkylphosphonium derivatives, and anions which could be
inorganic such as halide, tetrafluoroborate, and hexafluorophosphate, or organic
such as fluorinated imide, alkylsulfate and tosylate (Marsh et al., 2002).

In general, RTILs with halide, nitrate, methylsulfate, or trifluoroacetate anions are
miscible in water; RTILs with hexafluorophosphate or bis(trifluoro-
methylsulfonyl)amide anions are immiscible in water. Nonetheless, miscibility in
water of RTILs with tetrafluoroborate and trifluoromethanesulfonate anions
depends on the alkyl chain length on the cation (Poole, 2004). Chemical structures
of some common cations and anions of RTILs are listed in Figure 2.10.

73

Figure 2.10 Chemical structures of common cations and anions in RTILs

Cations
imidazolium pyridinium pyrrolidinium
alkylammonium alkylsulfonium alkylphosphonium
Anions
halide tetrafluoroborate hexafluorophosphate
Bis(trifluoromethyl-
sulfonyl)imide
methylsulfate tosylate

74
2.6.2 Applications of ILs in Hydrocarbon Separation

It has been discussed in several publications that hydrocarbons can be separated by
extraction using ionic liquids (Huddleston et al., 1998; Munson et al., 2002;
Domanska et al., 2007; Arce et al., 2008). Olefins and paraffins have similar
physical properties which make them difficult to be separated. With carbon-carbon
double bonds, however, olefins show different reactivity to other chemicals and
solvents comparing with paraffins.

Separation of olefins from paraffins using ionic liquids can be explained by
Prausnitz and Anderson’s solution thermodynamics (Lei et al., 2006). Mobility of
electron in the functional group of a molecule has great influence on the interaction
towards different solvents. With greater electron cloud mobility, the functional
group is easier to be polarized resulting in a higher solubility in polar solvent
(Safarik et al., 1998).

In general, mobility of electron cloud in C=C bond is greater than in C-C bond, thus
olefins would be relatively more soluble in polar solvents. Based on this principle,
ionic liquids such as 1-butyl-3-methylimidazolium hexafluorophosphate can be
used to extract olefins from paraffins. Similarly, aromatic hydrocarbons such as
benzene, toluene and alkylbenzenes can be extracted from hexane, heptane or other
paraffins using ionic liquids, for instance, 1-butyl-3-methylimidazolium

75
hexafluorophosphate and ethyl(2-hydroxyethyl)dimethylammonium bis(trifluo-
methylsulfonyl)imide (Domanska et al., 2007; Meindersma et al., 2006).

76

2.6.3 Applications of ILs in Desulfurization

It has also been demonstrated that ionic liquids, especially room temperature ionic
liquids, could be applicable in desulfurization technologies. Similar to the
mechanism in separating olefins from paraffins, selected types of ionic liquids
could form relatively strong π complex with OSCs such as thiophene compared
with benzene and paraffins (Huang et al., 2004). Thus, OSCs can be extracted from
other hydrocarbons so as to obtain desulfurized fuel.

Ionic liquids such as 1-ethyl-3-methylimidazolium tetrafluoroborate, 1-butyl-3-
methylimidazolium hexafluorophosphate and 1-butyl-3-methylimidazolium
tetrachloroaluminate have been investigated for this purposed. Extractive
desulfurization using ionic liquids are of particular interest due to the simplicity of
treatment arrangement which can be applied in ambient conditions. With high ionic
liquid to fuel ratio (greater than 5:1), more than 95% desulfurization on various
fuels can be achieved in 30 minutes (Zhang et al., 2004).

Another use of RTILs is a combination of oxidative and extractive desulfurization
process. With water miscible RTILs, OSCs can be extracted from fuel and become
oxidized by oxidant such as hydrogen peroxide to the corresponding sulfones in
RTIL phase (Lo et al., 1004). Desulfurization efficiency can be improved even with

77
less ionic liquid consumption. More than 95% desulfurization can be achieved in 3
hours with low ionic liquid to fuel ratio (1:1). Schematic diagram of this concept is
illustrated in Figure 2.11.

Some ILs with strong Brönsted acidity or basicity, such as N-methylpyrrolidonium
tetrafluoroborate, are considered as possible catalysts to generate hydroxyl radicals
from hydrogen peroxide, so as to improve the oxidizing power (Parvulescu et al.,
2007).

Figure 2.11 Schematic diagram of oxidative extractive desulfurization (Li et al., 2009)

78

In the third generation of UAOD, or the modified UAOD process, ionic liquid is
also utilized for multiple purposes (Cheng et al., 2008). The major mechanism
involved in the modified UAOD process is multiple phase-transfer catalysis. A
selected type of ionic liquid, imidazolium alkylsulfate, can be used as phase transfer
catalyst to transfer active oxygen species from aqueous phase to organic phase as
illustrated in Equation 2.5 to Equation 2.7 (Cheng et al., 2008). Ionic liquid could
also extract OSCs or oxidized OSCs from the fuel to achieve a lower sulfur fuel. It
is demonstrated that the modified UAOD process can achieve more than 99%
desulfurization with various type of diesel fuel oils.

− + − +
↔ +
4 4
RSO IM RSO IM (Eq. 2.5)

[ ]
*
6
*
4
2
− + − +
↔ + RSO IM O RSO IM (Eq. 2.6)

[]
2 4
*
6
RSO RSO IM RS RO IM + → +
− + − +
(Eq. 2.7)

79

2.7 Ultraviolet Irradiation
2.7.1 Fundamentals of Photochemistry and UV Radiation

Photochemistry refers to the study of light induced chemical reactions. Basically, a
photochemical reaction is initiated by absorption of light, or photon by an atom,
molecule or ion converting to the corresponding excited state species. There are two
fundamental principles in photochemistry. The first law, also known as the
Grotthus-Draper law, was established by Chiristian J. Grotthus and John W. Draper
in early 1800s, stating that only light which is absorbed by a system, such as an
atom, molecule or ion, can cause chemical change.

The second law, also known as the Stark-Einstein law, was established by Johannes
Stark and Albert Einstein in early 1900s, stating that for one quantum of light, or a
photon that is absorbed, only the atom, molecule or ion which absorbs the photon
would be excited (Mukherjee, 1978). Due to the complexity of photochemical
reactions, not all photo-excited atoms, molecules or ions would be chemically
active. Loss of energy from the excited state could happen in unproductive ways.
Quantum yield Φ was introduced by Einstein to specify the efficiency of a
particular photochemical reaction. Φ is defined as the number of molecules
decomposed or formed per number of photon absorbed.

80

Figure 2.12 Section of Electromagnetic Spectrum (Willis, 2006)

For any chemical reaction, energy is required to overcome the activation energy
∆E
a
, and also the enthalpy ∆H. In photochemical reactions, energy is provided by
electronic excitation by photon absorption. While the energies of bond dissociation
energy per mole of molecules are generally within 150 kJ to 600 kJ, these can be
provided by one mole of photons from light with wavelengths between 200nm to
800nm (Mukherjee, 1978). Thus, researches in photochemistry are more focused on
a few sections of the light spectrum, or more precisely electromagnetic spectrum
including infrared, visible light and ultraviolet (Murov et al., 1993). An excited
species could undergo different reactions such as photoionization,
photoisomerization, photooxidation, photoreduction, bimolecular quenching and
photo substitution (Gold et al., 1987).

81

Ultraviolet is a type of electromagnetic radiation with wavelength ranging from 10
nm to 400 nm. Depending on the wavelength, UV light can be classified into three
major categories as listed in Table 2.10. The most famous photochemical reactions
related to UV, perhaps, would be the interaction with ozone generation and
decomposition in the stratosphere. The Sun emits UV radiation with a wide range of
wavelengths, including UVA, UVB, and UVC. However, UVA is the only major
UV radiation reaching the Earth surface. Harmful UVB and UVC would be
absorbed in the ozone layer through a series of photochemical reactions:

O
2
+ hv Æ 2O (Eq. 2.8)

O + O
2
+ M Æ O
3
+ M (Eq. 2.9)

O
3
+ hv Æ O + O
2
(Eq. 2.10)

O
3
+ O Æ 2O
2
(Eq. 2.11)

Due to the high energy per mole of photons as suggested in Table 2.11, ultraviolet is
applicable in a wide range of photochemical reactions. While UV light is effective
to achieve bond cleavage and other photochemical reactions, UV light has been
applied to many areas including air purification, water disinfection, organic

82
destruction, food processing and biomedical applications. UV light has also been
used in spectrophotometry and other analytical processes due to UV fluorescence
reactions.

Table 2.10 Ultraviolet Classification (ISO, 2007)
Name Abbreviation Wavelength (nm)
Ultraviolet A UV A 320 – 400
Ultraviolet B UVB 280 – 320
Ultraviolet C UVC 100 – 280

Near Ultraviolet NUV 300 – 400
Middle Ultraviolet MUV 200 – 300
Far Ultraviolet FUV 122 – 200
Vacuum Ultraviolet VUV 10 – 200
Extreme Ultraviolet EUV 10 – 121

83

Table 2.11 Energy per mole of photons (Mukherjee, 1978)
Light Wavelength (nm) Energy per mole of Photons (kJ)
Ultraviolet 200 – 400 299 – 590
Violet 400 – 450 266 – 299
Blue 450 – 500 239 – 266
Green 500 – 570 209 – 239
Yellow 570 - 590 203 – 209
Orange 590 – 620 192 – 203
Red 620 – 750 159 – 192

84
2.7.2 UV and Photochemical Reactions in Aqueous System

While water is usually involved in most of the photochemistry studies, the
photochemical reactions related to water is particularly important. Although it is
suggested that visible light and UVA can dissociate water molecule, significant
photolysis of water can only be achieved by UVC or even extreme UV. It has been
recorded that the quantum yield Φ of water dissociation to hydrogen radicals and
hydroxyl radicals varies from 0.335 to 1, with the corresponding wavelength of
UVC applied from 185 nm to 124 nm (Smith, 1970).

H
2
O + hv Æ (H
2
O)* Æ H· + OH· (Eq. 2.12)

Photolysis of hydrogen peroxide in aqueous solution is another important aspect in
aqueous photochemistry. UV irradiation with wavelength ranged from 200 nm to
280 nm, typically 254 nm, can be applied to dissociate hydrogen peroxide into
hydroxyl radicals, hydrogen radicals and perhydroxyl radicals (Venkatadri et al.,
1993). Instead of ground state hydroxyl radicals (OH·), excited state of hydroxyl
radicals (OH*) can also be formed through a series of radical reactions (Smith,
1970).

The excitation of hydrogen peroxide by UV irradiation has been applied in water
treatment as one of the advanced oxidation process to destroy organic compounds

85
such as trichloroethylene (TCE), perchloroethylene (PCE) and colored organic
compounds. The photochemical reactions of hydrogen peroxide as listed in Figure
2.13.

Figure 2.13 Organic destruction by UV/H
2
O
2
system (Venkatadri et al., 1993)

H
2
O
2
+ hv Æ OH· + OH·
H
2
O
2
↔ HO
2
¯ + H
+

OH· + H
2
O
2
Æ HO
2
· + H
2
O
OH· + HO
2
¯ Æ HO
2
· + OH¯
2HO
2
· Æ H
2
O
2
+ O
2

RH + OH· Æ H
2
O + R· Æ further oxidation

86

2.7.3 Photocatalysis and Titanium (IV) Oxide

Photocatalysis is, in fact, closely related to photoelectrochemistry. During the
development of semiconductor photoelectrochemistry in 1970s, titanium (IV) oxide
(TiO
2
) was used as semiconductor electrodes for processes such as
photoelectrolysis of water and photocell to harvest solar energy. It was found out
later that TiO
2
could be applicable for photocatalytic degradation of pollutants such
as cyanide (Frank et al., 1977). The interest in photocatalysis and their
environmental applications were then realized.

Titanium (IV) oxide is considered as one of the most important photocatalysts due
to its high availability, chemically stable structure, high effectiveness with highly
oxidizing photogenerated holes on surface, and relatively low cost (Kaneko et al.,
2002).

The principle of photocatalysis by TiO
2
or other semiconductors involves the
excitation of the semiconductor particles. Under UV irradiation, the energy state of
electrons of the semiconductor would change, forming conduction band electrons
(e¯
CB
) and valence band holes (h
+
VB
) on the surface of the particles (Robert et al.,
2002).

87
Surface with conduction band electrons and valence band holes are the active sites
for oxidation or other chemical reactions to generate radicals such as hydroxyl,
superoxide, and perhydroxyl radicals from oxygen and water. Radicals generated
can be used to oxidize the target pollutants. Direct oxidation of the target pollutants
at valence band holes is also possible. The pathways for photogeneration of radicals
are illustrated in Figure 2.14.

Figure 2.14 Radicals generation in Photocatalysis (Al-Ekabi et al., 1992)

TiO
2
+ hv Æ e¯
CB
+ h
+
VB

h
+
VB
+ OH¯ Æ OH·
h
+
VB
+ H
2
OH Æ H
+
+ OH·
e¯
CB
+ O
2
Æ O
2
¯·
e¯
CB
+ h
+
VB
Æ heat

88

2.7.4 Photolysis and Photo-Oxidation Desulfurization

Desulfurization using photochemical reactions can be separated into two major
categories: direct photolysis of OSCs and photo-oxidation of OSCs. Similar to other
compounds, organic sulfur compounds can be photo-excited in photochemical
reactions. For instance, alkyl radicals and alkanethiyl radicals can be generated
from dialkyl sulfides through photolysis to cleave C-S bond.

With the aid of trivalent phosphorus compound, sulfur can be abstracted from
alkanethiyl radicals to achieve desulfurization (Coyle, 1991). However, this process
is highly dependent on the target compounds and the wavelength of radiation
applied. Single wavelength radiation may not be applicable on all OSCs. Besides,
instead of C-S bond cleavage, S-H bond or C-H bond cleavage is observed in direct
photolysis of thiols and thiophenes (Bianchini et al., 1997). Thus, direct photolysis
is not commonly applied on desulfurization.

Photo-oxidation of OSCs is another pathway to achieve desulfurization
photochemically. Selected reactants such as water and oxygen could be activated
forming reactive oxygen species and radicals which can further react to oxidize
OSCs (Baba, 1974). Although photo-oxidation can be improved by the addition of
photocatalyst such as TiO
2
, low desulfurization efficiency, for instance less than

89
40% oxidation of DBT in 10 hours, was obtained due to limited reactivity and
solubility of oxygen. Aqueous oxidant such as hydrogen peroxide could be used
instead to achieve better desulfurization efficiency (Matsuzawa et al., 2001).

90
CHAPTER 3: MODIFIED OXIDATIVE DESULFURIZATION USING
SUPEROXIDE ON MODEL SULFUR COMPOUND STUDY
3.1 Introduction

It is known that organic sulfur compounds are slightly more polar comparing with
hydrocarbons with similar structures. In order to separate OSCs from hydrocarbons
effectively, however, the slight difference in polarity is not enough. In oxidative
desulfurization process, the major goal is to support a highly selective and effective
oxidation of organic sulfides to the corresponding sulfones which are significantly more
polar comparing with the organic sulfides.

Based on the first two generations of the UAOD process developed, organic sulfur
compounds oxidation by hydrogen peroxide with application of transition metal catalysts,
or more specifically polyoxometalates (POMs) has been studied (Met et al., 2003, Wan et
al., 2007). It has been demonstrated that benzothiophene, dibenzothiophene and their
derivatives could be oxidized by the UAOD process in mild conditions with considerably
high yield. However, desulfurization efficiencies on thiophene and the derivatives were
comparatively low using the TMC/H
2
O
2
system.

In the third generation of UAOD system, new components including use of organic acid
catalysts such as glacial acetic acid (HAc), and ionic liquids were introduced (Cheng et

91
al., 2008). These modifications were aimed at improving desulfurization efficiency and
selectivity. The modified UAOD process had successfully demonstrated high yield to
produce various sulfones from the corresponding OSCs by HAc/H
2
O
2
oxidation with or
without ultrasonication (Cheng et al., 2009).

Although hydrogen peroxide is known as an effective oxidant in oxidative desulfurization
(ODS) with minimal pollutants to the environment, only low concentration (< 30% wt.)
of hydrogen peroxide solution can be applied due to safety reasons. High concentration
(> 50% wt.) of hydrogen peroxide solution is considered as unstable, which could cause
spontaneous ignition or even explosion due to vigorous self-sustained decomposition.
Low concentration of hydrogen peroxide is relatively safe, but oxidation efficiency would
be significantly lowered due to dilution effect (Wan et al., 2008). Weight and volume of
oxidant required would also be increased when a low concentration oxidant is used.

Alternative oxidants have been investigated to replace hydrogen peroxide (Chan et al.,
2008). In this study, selection of oxidant is based on oxidation effectiveness, availability,
cost, safety, and potential by-products or pollutants formation. Similar to the third
generation of the UAOD process, complementary techniques including ultrasonication,
acid catalyzed oxidation, phase transfer catalysis, mechanical mixing and application of
room temperature ionic liquid mentioned in Chapter 2 are employed in this development.
Preliminary investigation on utilizing UV irradiation to enhance oxidative desulfurization
is also discussed.

92
A series of experiments based on BT and DBT model sulfur compounds were conducted
to accomplish optimum reaction conditions. In this chapter, the effects of type and
amount of acid catalyst, phase transfer catalyst, ionic liquid, oxidant, and treatment
methods applied are discussed.

93
3.2 Materials and Experimental Procedures
3.2.1 Chemical Preparation

Model sulfur compounds used in this study including thiophene (T), 2-methyl
thiophene (2MT), benzothiophene (BT), 2-methyl benzothiophene (2MBT),
dibenzothiophene (DBT), 4, 6-dimethyl dibenzothiophene (4,6DMDBT) were
obtained from Sigma-Aldrich Co., Allentown, Pennsylvania. Solvents for model
sulphur compounds solutions preparation including toluene and n-decane were
obtained from VWR Inc., West Chester, Pennsylvania. Acetonitrile was obtained
from VWR Inc. Oxidants used including 30% wt. hydrogen peroxide (H
2
O
2
)
solution was obtained from VWR Inc., while sodium superoxide, potassium
superoxide and potassium permanganate were obtained from Sigma-Aldrich Co.
Acid catalysts including glacial acetic acid and trifluoroacetic acid were obtained
from Sigma-Aldrich Co. Phase transfer catalysts including 18-Crown-6,
tetraoctylammonium bromide and tetraoctylammonium chloride were obtained
from Sigma-Aldrich Co., while tetraoctylammonium fluoride was synthesized by
halogen exchange process (Dermeik et al., 1989). Photocatalyst used including
titanium (IV) oxide was obtained from Sigma-Aldrich Co. Ionic liquids including 1-
butyl-3-methylimidazolium hexafluoro-phosphate [BMIM][PF
6
], 1-ethyl-3-
methylimidazolium ethylsulfate [EMIM][EtSO
4
], 1,2,3-trimethyl-imidazolium
methylsulfate [TMIM][MeSO
4
], and tributylmethyl-phosphonium methyl-sulfate
[TMBP][MeSO
4
] were obtained from Sigma-Aldrich Co.

94

3.2.2 Ultrasonic Reactor

Similar to the previous generation of the UAOD process, a probe ultrasonic reactor,
model number VCX-750, manufactured by Sonic & Materials Inc., Newtown,
Connecticut was used in this study. Probe type ultrasonic reactor can provide high
intensity (100 watts/cm
2
) ultrasound irradiation as a point source, producing ultra-
fine emulsion to enhance mixing. This reactor can support variable power output,
integrated temperature control and remote processing for different purposes.
Specifications of this ultrasonic reactor are listed in Table 3.1.

Table 3.1 Specifications of Ultrasonic Reactor VCX-750 listed in product catalogue
Dimensions (H×W×D) 235 mm × 190 mm × 340 mm
Power Output 750 Watts
Frequency 20kHz
Sealed Converter Piezoelectric Lead Zirconate Titanate Crystal (PZT)
Standard Probe Size Diameter: 13 mm; Length 136 mm
Probe Material Titanium Alloy Ti-6Al-4V
Processing Capacity 10 ml – 250 ml

95
3.2.3 Ultraviolet Lamp

An ultraviolet lamp, model number UVLMS-38, manufactured by UVP Ltd.,
Upland, California was used in this study. This UV lamp can provide UVA, UVB
and UVC at the wavelength or 365 nm, 302 nm and 254 nm respectively. Instead of
normal glassware, fused quartz glass reactor was used in the UV related
experiments to minimize UV light blockage. Specifications of this ultraviolet lamp
are listed in Table 3.2.

Table 3.2 Specifications of UV lamp UVLMS-38 listed in product catalogue
Dimensions (L×W×D) 376 mm × 96 mm × 64 mm
Power Output 8 Watts
Frequency 254nm / 302 nm / 365 nm

96
3.2.4 Analytical Method

In accordance with ASTM D4294 and ISO 8754, non-dispersive X-ray fluorescence
(NDXRF) was used to determine the total sulfur content of the samples. Sulfur-in-
Oil Analyzer (SLFA-20), manufactured by Horiba Inc., Irvine, California was used
to measure total sulfur content of samples based on this method. Sulfur-in-Oil
Analyzer is applicable to measure samples with total sulfur content ranged from 0
to 5 wt%, with a lower detection limit of 20 ppm
w
.

Figure 3.1 Sulfur-in-oil analyzer (SLFA-20)

97

3.3 Experimental Design, Procedure, Results and Discussion
3.3.1 Use of Solid Oxidants in Oxidative Desulfurization
3.3.1.1 Solid Oxidants Selection

As discussed in Chapter 2.3, dilution of oxidants with water would lower the
reaction rate and increase the total reactant volume. It usually happens when an
aqueous oxidant, such as hydrogen peroxide, is used in the process. Previous work
has demonstrated that lowering H
2
O
2
concentration would significantly retard the
conversion of sulfur to sulfone (Wan et al., 2007). While high concentration of
H
2
O
2
is unstable and highly reactive, other pure oxidants for instance, solid
oxidants, are considered to improve the oxidative desulfurization process.

Selection of solid oxidant was primarily based on the oxidation potential. Three
solid oxidants, potassium permanganate, sodium superoxide, and potassium
superoxide were selected for this study. Potassium permanganate is known as a
strong oxidizing agent, especially in acidified condition. On the other hand,
superoxide is highly reactive radical anion. It is also a precursor of other reactive
oxygen species such as singlet oxygen, hydrogen peroxide, and hydroxyl radical.

98
3.3.1.2 Experimental Procedure

Either BT or DBT model sulfur compound was dissolved into a solvent mixture
with 30% wt. toluene and 70% wt. n-decane to make a stock solution with
approximately 1000 ppm
w
sulfur content. Instead of a pure solvent, a solvent
mixture was used to simulate real petroleum products.

Known amount of selected oxidant, either 30% wt. hydrogen peroxide, potassium
permanganate, sodium superoxide or potassium superoxide, was first mixed with 5
grams of 1-butyl-3-methylimidazolium hexafluoro-phosphate [BMIM][PF
6
], and
0.1 gram of tetraoctylammonium fluoride to produce mixture A; 10 grams of stock
solution of model sulfur compound were mixed with 3 grams of acetic acid to
produce mixture B.

Mixture A was then slowly added to mixture B, and the resulting mixture was
heated up to 70°C with continuous magnetic stirring. Emulsion formed after the
treatment process was separated by centrifugation. Oil phase was collected on the
top and was extracted with acetonitrile to remove sulfones. The acetonitrile-
extracted oil phase was analyzed by the Sulfur-in-Oil Analyzer.

99
3.3.1.3 Results and Discussion

Table 3.3 to Table 3.6 show the desulfurization of BT and DBT with selected
oxidants: 30% wt. hydrogen peroxide (H
2
O
2
), potassium permanganate (KMnO
4
),
sodium superoxide (NaO
2
), and potassium superoxide (KO
2
), respectively, at fixed
oxidant to sulfur ratios with magnetic stirring at 70°C for 3 hours. In high sulfur to
oxidant mole ratio (1:4), greater than 85% desulfurization efficiency for both BT
and DBT samples could be achieved for all solid oxidants. In contrast, less than
70% desulfurization for both BT and DBT samples were obtained when 30% wt.
hydrogen peroxide was used.

Table 3.3 Desulfurization efficiency with 30% wt. H
2
O
2
as oxidant
Mixing
Time
30% wt. H
2
O
2

Applied
Initial Sulfur
Content
Sulfur to
Oxidant Ratio
Final Sulfur
Content
Sulfur Removal

hr g mmol ppm
w
mmol mmol : mmol ppm
w
%
3 5.5 48 1012 0.32 1:150 < 20 > 98
3 3.6 32 1012 0.32 1:100 < 20 > 98
3 1 9.6 1012 0.32 1:30 98 90
3 0.36 3.2 1012 0.32 1:10 193 81
BT
3 0.15 1.28 1012 0.32 1:4 354 65

3 5.5 48 1006 0.31 1:150 < 20 > 98
3 3.6 32 1006 0.31 1:100 < 20 > 98
3 1 9.6 1006 0.31 1:30 81 92
3 0.36 3.2 1006 0.31 1:10 169 83
DBT
3 0.15 1.28 1006 0.31 1:4 322 68

100

Table 3.4 Desulfurization efficiency with KMnO
4
as oxidant

Table 3.5 Desulfurization efficiency with NaO
2
as oxidant

Mixing
Time
KMnO
4
Applied
Initial Sulfur
Content
Sulfur to
Oxidant Ratio
Final Sulfur
Content
Sulfur Removal

hr g mmol ppm
w
mmol mmol : mmol ppm
w
%
3 1.5 9.6 1012 0.32 1:30 < 20 > 98
3 0.5 3.2 1012 0.32 1:10 85 92
BT
3 0.2 1.28 1012 0.32 1:4 147 85

3 1.5 9.6 1006 0.31 1:30 < 20 > 98
3 0.5 3.2 1006 0.31 1:10 83 91
DBT
3 0.2 1.28 1006 0.31 1:4 139 86
Mixing
Time
NaO
2
Applied
Initial Sulfur
Content
Sulfur to
Oxidant Ratio
Final Sulfur
Content
Sulfur Removal

hr g mmol ppm
w
mmol mmol : mmol ppm
w
%
3 0.55 9.6 1012 0.32 1:30 < 20 > 98
3 0.18 3.2 1012 0.32 1:10 59 94
BT
3 0.07 1.28 1012 0.32 1:4 108 90

3 0.55 9.6 1006 0.31 1:30 < 20 > 98
3 0.18 3.2 1006 0.31 1:10 58 94
DBT
3 0.07 1.28 1006 0.31 1:4 88 91

101

Table 3.6 Desulfurization efficiency with KO
2
as oxidant

1:30 1:10 1:4
0
10
20
30
40
50
60
70
80
90
100
H2O2
KMnO4
NaO2
KO2
Sulfur to Oxidant Ratio
Sulfur Removal, %

Figure 3.2 Desulfurization efficiencies of BT sample with different oxidants

Mixing
Time
KO
2
Applied
Initial Sulfur
Content
Sulfur to
oxidant Ratio
Final Sulfur
Content
Sulfur Removal

hr g mmol ppm
w
mmol mmol : mmol ppm
w
%
3 0.68 9.6 1012 0.32 1:30 < 20 > 98
3 0.23 3.2 1012 0.32 1:10 39 96
BT
3 0.09 1.28 1012 0.32 1:4 73 93

3 0.68 9.6 1006 0.31 1:30 < 20 > 98
3 0.23 3.2 1006 0.31 1:10 48 95
DBT
3 0.09 1.28 1006 0.31 1:4 81 92

102

Figure 3.3 Desulfurization efficiencies of DBT sample with different oxidants

Among the three solid oxidants used, relatively low desulfurization efficiencies
were obtained when potassium permanganate was used, as illustrated in both Figure
3.2 and 3.3. It was demonstrated that high desulfurization efficiencies could be
achieved on both BT and DBT samples when either sodium superoxide or
potassium superoxide was used as oxidant, where potassium superoxide
demonstrated slightly higher desulfurization efficiencies. Besides, sodium
superoxide is relatively unstable comparing with potassium superoxide as discussed
in Chapter 2.3. Thus, potassium superoxide was selected as the alternative oxidant
to replace hydrogen peroxide used in the UAOD process.

103

91
92
93
94
95
96
97
98
99
0 0.2 0.4 0.6 0.8 1
KO
2
dosage (g)
Desulfurization (%)
BT
DBT

Figure 3.4 Desulfurization efficiencies of BT and DBT with different KO
2
dosage

Under the same conditions, the desulfurization of BT and DBT samples with
various amounts of potassium superoxide is illustrated in Figure 3.4.
Desulfurization efficiency was optimized by using 0.7 gram of KO
2
to oxidize 10
grams solution with 1000 ppm
w
model sulfur compounds, either BT or DBT. The
sulfur to oxidant mole ratio at the optimal point was 1:30. Comparing with 30% wt.
H
2
O
2
, using KO
2
as oxidant for oxidative desulfurization process can significantly
reduce oxidant consumption, the weight of oxidant requirement, and hence the
volume of reactor without losing desulfurization efficiency.

104
3.3.2 Effect of Different Phase Transfer Catalysts
3.3.2.1 Phase Transfer Catalyst Selection

As discussed in Chapter 2.5, reactions involving reactants in two or more
immiscible phases are hindered by inefficient contact and interfacial transfer of
reactants. Utilization of phase transfer catalysts can significantly improve interfacial
mixing so as to increase rate of reactions. In the UAOD process, QASs are applied
as PTCs to increase oxidation rate of OSCs in organic phase by hydrogen peroxide
solution. As demonstrated in Table 3.7, only cationic surfactants including QAS are
effective PTCs in the UAOD process.

In the conditions with no surfactant, or with either anionic or nonionic surfactants,
no or low rate oxidation of OSCs including BT and DBT would be resulted (Wan et
al., 2007). Tetraoctylammonium fluoride is identified as the best PTCs in the
UAOD process by supporting a high oxidation rate of OSCs with no by-product
formation. Thus, TOAF has been used as PTC for the UAOD process.

105
Table 3.7 Effect of surfactants on the UAOD process (Wan et al., 2007)
Type Surfactant Desulfurization
Tetraoctylammonium Bromide (TOAB) +
Tetrabutylammonium Bromide (TBAB) +
Methyltributylammonium Chloride (MBAC) +
Methyltributylammonium Hydroxide (MBAH) +
Cationic
Tetramethylammonium Fluoride (TMAF) +
Anionic 1-Octanesulfonic Acid, Sodium Sat –
Nonionic Tween 80 –
Control No Surfactant –
Note: + indicates the system undergoes reaction in the given conditions
– indicates no or eligible reaction in the given conditions

In the application of solid oxidant in the UAOD process, another type of phase
transfer catalyst, crown ether, is also considered due to the ability to stabilize
specific metallic cation. For instance, 18-crown-6 shows the ability to form stable
complex with potassium cation in organic solvent. Increase in solubility of
potassium superoxide in dimethyl sulfoxide has been demonstrated with the
application of 18-crown-6. (Suzuki et al., 1979)

106
3.3.2.2 Experimental Procedure

DBT model sulfur compound was dissolved into a solvent mixture with 30% wt.
toluene and 70% wt. n-decane to make a stock solution with approximately 1000
ppm
w
sulfur content.

Using a sulfur to oxidant ratio of 1:30, 0.7 gram of KO
2
was first mixed with 5
grams of [BMIM][PF
6
] and known quantity of the selected PTA, either TOAF or
18-crown-6, to produce mixture A; 10 grams of stock solution of model sulfur
compound were mixed with 3 grams of acetic acid to produce mixture B.

Mixture A was then slowly added to mixture B, and the resulting mixture was
heated up to 70°C with continuous magnetic stirring. Emulsion formed after the
treatment process was separated by centrifugation. Oil phase was collected and was
extracted with acetonitrile to remove sulfones. The acetonitrile-extracted oil phase
was analyzed by the Sulfur-in-Oil Analyzer.

107
3.3.2.3 Results and Discussion

Table 3.8 and Table 3.9 show the desulfurization of DBT with selected phase
transfer catalyst, including TOAF and 18-crown-6. Comparing with 18-crown-6,
TOAF is a more effective phase transfer catalyst for oxidative desulfurization
process using KO
2
as oxidant. Although 18-crown-6 was able to improve the
desulfurization process, its effectiveness is considerably lower than TOAF.
Formation of tetraalkylammonium superoxide could be a possible pathway to
enhance desulfurization with TOAF (Afanas’ev, 1989).

Considering the results using TOAF as PTC with 1 hour reaction time shown in
Table 3.9, desulfurization efficiency could be improved from 45% to 80% by
adding 0.05 gram of TOAF. Doubling the dosage of TOAF could result in a slight
increase of desulfurization efficiency, from 80% to 86%. However, further
increasing the dosage of TOAF from 0.1 to 0.5 gram did not show a significant
improvement. It suggests that optimal phase transfer condition can be achieved at
0.1 gram of TOAF dosage.

Based on the experimental result, anionic transferrer is a better phase transfer
catalyst for this process. Tetraoctylammonium fluoride is considered as an effective
PTC due to its stability towards strong oxidizing condition as well as the ability to
activate the oxidant in organic phase. Tetraoctylammonium superoxide could

108
possibly be synthesized as an intermediate product. This could help transferring the
oxidant to the organic phase so as to achieve oxidative desulfurization. Although
higher dosage of PTC is expected to yield better desulfurization efficiency, the
difference is not significant when the PTC dosage is greater than 0.1 gram. In this
experiment, 0.1 gram of tetraoctylammonium fluoride is considered as the optimum
dosage.

Table 3.8 Desulfurization of DBT solution with TOAF as PTC

Mixing Time
(hr)
TOAF Applied
(g)
Initial Sulfur Content
(ppm
w
)
Final Sulfur Content
(ppm
w
)
Sulfur Removal
(%)
1 0 1006 551 45
2 0 1006 413 59
3 0 1006 372 63

1 0.05 1006 198 80
2 0.05 1006 107 89
3 0.05 1006 58 94

1 0.1 1006 141 86
2 0.1 1006 49 95
3 0.1 1006 < 20 > 98

1 0.5 1006 110 89
2 0.5 1006 38 96
3 0.5 1006 < 20 > 98

109

Table 3.9 Desulfurization of DBT solution with 18-crown-6 as PTC

Mixing Time
(hr)
18-Crown-6
Applied (g)
Initial Sulfur Content
(ppm
w
)
Final Sulfur Content
(ppm
w
)
Sulfur Removal
(%)
1 0 1006 551 45
2 0 1006 413 59
3 0 1006 372 63

1 0.05 1006 463 54
2 0.05 1006 390 61
3 0.05 1006 333 67

1 0.1 1006 447 56
2 0.1 1006 374 63
3 0.1 1006 314 69

1 0.5 1006 401 59
2 0.5 1006 340 66
3 0.5 1006 298 71

110

3.3.3 Effect of Acid Catalysts
3.3.3.1 Acid Catalyst Combination

Acetic acid is one of the most common organic acid catalysts used in oxidative
desulfurization processes. Recently, it is suggested that trifluoroacetic acid (TFA)
could be used as an acid catalyst to improve oxidation of OSCs (Wang et al., 2003;
Yazu et al., 2004).

Trifluoroacetic acid is a strong carboxylic acid with more than a thousand folds
acidity than acetic acid. In the third generation of the UAOD process, an acid
catalyst solution with 20% wt. of TFA and 80% wt. of acetic acid were used to
achieve ultralow sulfur diesel with theoretically 0 ppm
w
sulfur content (Cheng et al.,
2008).

In order to test for the catalytic effect in this oxidative desulfurization process with
the application of superoxide, experiments were conducted under different ratios of
acetic acid and trifluoroacetic acid combination.

111

3.3.3.2 Experimental Procedure

DBT model sulfur compound was dissolved into a solvent mixture with 30% wt.
toluene and 70% wt. n-decane to make a stock solution with approximately 1000
ppm
w
sulfur content. 0.7 gram of KO
2
was first mixed with 5 grams of [BMIM][PF
6
]
and 0.1 gram of TOAF to produce mixture A; 10 grams of stock solution of model
sulfur compound were mixed with known amount of two acid catalysts: acetic acid
and TFA to produce mixture B.

Mixture A was then slowly added to mixture B, and the resulting mixture was
heated up to 70°C with continuous magnetic stirring. Emulsion formed after the
treatment process was separated by centrifugation. Oil phase was collected and was
extracted with acetonitrile to remove sulfones. The acetonitrile-extracted oil phase
was analyzed by the Sulfur-in-Oil Analyzer.

112
3.3.3.3 Results and Discussion

Table 3.10 shows the desulfurization of DBT solution with different amounts of
acetic acid added. Increasing the acid catalyst dosage from 1 gram to 2 grams would
result in a slight increase in desulfurization efficiency. However, no significant
difference in desulfurization efficiency was observed by further increasing the acid
catalyst dosage to 3 grams.

Table 3.10 Desulfurization of DBT solution with respect to acid dosage

Mixing Time
(hr)
Acetic Acid
Applied (g)
Initial Sulfur Content
(ppm
w
)
Final Sulfur Content
(ppm
w
)
Sulfur Removal
(%)
1 1 1006 141 86
2 1 1006 49 95
3 1 1006 < 20 > 98

1 2 1006 102 90
2 2 1006 < 20 > 98
3 2 1006 < 20 > 98

1 3 1006 103 90
2 3 1006 < 20 > 98
3 3 1006 < 20 > 98

113
Table 3.11 Desulfurization of DBT solution with respect to acid catalyst applied

Based on the result, 2 grams of acid catalyst was considered as the optimum dosage
for the system. Table 3.11 shows the desulfurization of DBT with 2 grams of acid
catalyst in various combinations of acetic acid and trifluoroacetic acid.

Addition of 0.1 gram of TFA could result in a slight increase of desulfurization
efficiency, from 90% to 94%. However, further increasing the dosage of TFA did
not give a significant improvement in desulfurization efficiency. It may be
postulated that addition of a small portion, for instance 10% of TFA, in the acid
catalyst would be enough for optimization.

Mixing Time
(hr)
Acetic Acid/
TFA Applied (g)
Initial Sulfur Content
(ppm
w
)
Final Sulfur Content
(ppm
w
)
Sulfur Removal
(%)
1 2 / 0 1006 141 90
2 2 / 0 1006 < 20 > 98

1 1.9 / 0.1 1006 59 94
2 1.9 / 0.1 1006 < 20 > 98

1 1.8 / 0.2 1006 53 95
2 1.8 / 0.2 1006 < 20 > 98

1 1.5 / 0.5 1006 47 95
2 1.5 / 0.5 1006 < 20 > 98

114
3.3.4 Effect of Ionic Liquid
3.3.4.1 Ionic Liquid Dosage

As discussed in Chapter 2, ionic liquid is applicable in oxidative and extractive
desulfurization process. Organic sulfur compounds can be extracted from fuel to
ionic liquid phase. Especially for water miscible ionic liquid, it provides a better
opportunity for the oxidant such as hydrogen peroxide to react with the OSCs
forming the corresponding sulfones. More than 95% desulfurization can be
achieved in 3 hours with 1:1 ionic liquid to fuel ratio.

Although ionic liquid is known as the new generation of green solvent and catalyst,
most of the ionic liquids available are expensive. High dosage of ionic liquid would
increase the operating cost of the process, making it less economically feasible.

In order to find the optimal dosage of ionic liquid, experiments were conducted
with different dosage of ionic liquids including [BMIM][PF
6
] and [EMIM][EtSO
4
].

115
3.3.4.2 Experimental Procedure

DBT model sulfur compound was dissolved into a solvent mixture with 30% wt.
toluene and 70% wt. n-decane to make a stock solution with approximately 1000
ppm
w
sulfur content. 0.7 gram of KO
2
was first mixed with known amount of the
selected ionic liquid and 0.1 gram of TOAF to produce mixture A; 10 grams of
stock solution of model sulfur compound were mixed with 1.9 grams acetic acid
and 0.1 gram of TFA to produce mixture B.

Mixture A was then slowly added to mixture B, and the resulting mixture was
heated up to 70°C with continuous magnetic stirring for 3 hours. Emulsion formed
after the treatment process was separated by centrifugation. Oil phase was collected
and was extracted with acetonitrile to remove sulfones. The acetonitrile-extracted
oil phase was analyzed by the Sulfur-in-Oil Analyzer.

116
3.3.4.3 Results and Discussion

The desulfurization efficiency for DBT solution with different dosage of ionic
liquid is listed in Table 3.12. The result demonstrated that the desulfurization
efficiency increases when the dosage of ion liquid increases. Both ionic liquids
exhibited similarity in enhancing the desulfurization process.

Table 3.12 Desulfurization of DBT solution with respect to IL dosage

IL
IL dosage
(g)
Initial Sulfur Content
(ppm
w
)
Final Sulfur Content
(ppm
w
)
Sulfur Removal
(%)
0 1006 153 85
1 1006 36 96
3 1006 24 98
[BMIM][PF
6
]
5 1006 < 20 > 98

0 1006 153 85
1 1006 42 96
3 1006 23 98
[EMIM][EtSO
4
]
5 1006 < 20 > 98

117

By applying 1 gram of the [BMIM][PF
6
], the desulfurization efficiency was
increased by 11%. Similarly, the desulfurization efficiency was increased by 11%
by applying 1 gram of the [EMIM][EtSO
4
].

It is observed that further increase of ionic liquid dosage could improve the
desulfurization efficiency. When the dosage of ionic liquid increased from 1 gram
to 3 grams, the desulfurization efficiency increased from 96% to 98%. When the
dosage increased to 5 grams, the total sulfur content of the desulfurized sample was
less than 20 ppm
w
. Although the improvement was marginal, this would
nonetheless be applicable to situations where sulfur content less than 20 ppm
w
is
desired. Therefore, the dosage of ionic liquid was selected to be 5 grams.

118

3.3.4.4 Ionic Liquid Selection

As discussed in Chapter 2.2, some RTILs would undergo thermolysis by ultrasonic
irradiation, or oxidative degradation in the presence of hydrogen peroxide, acetic
acid and ultrasonication. One of the mostly employed RTIL, 1,3-
dialkylimidazolium hexafluorophosphate [BMIM][PF
6
] would be degraded to
fluorinated reactive gases which could damage glassware and other equipment.

Ionic liquids with alkylsulfate as anion are relatively more stable and relatively low
in toxicity, and thus are considered as possible substitutions of [BMIM][PF
6
] used
in the previous experiments. Three alternative ionic liquids including 1-ethyl-3-
methylimidazolium ethylsulfate [EMIM][EtSO
4
], 1,2,3-trimethyl-imidazolium
methylsulfate [TMIM][MeSO
4
], and tributylmethylphosphonium methylsulfate
[TMBP][MeSO
4
] were investigated in this study.

119
3.3.4.5 Experimental Procedure

DBT model sulfur compound was dissolved into a solvent mixture with 30% wt.
toluene and 70% wt. n-decane to make a stock solution with approximately 1000
ppm
w
sulfur content. 0.7 gram of KO
2
was first mixed with 5 grams of selected
ionic liquid, either [BMIM][PF
6
], [EMIM][EtSO
4
], [TMIM][MeSO
4
] or
[TMBP][MeSO
4
], together with 0.1 gram of TOAF to produce mixture A; 10 grams
of stock solution of model sulfur compound were mixed with 1.9 grams acetic acid
and 0.1 gram of TFA to produce mixture B.

Mixture A was then slowly added to mixture B, and the resulting mixture was
heated up to 70°C with continuous magnetic stirring for 3 hours. Emulsion formed
after the treatment process was separated by centrifugation. Oil phase was collected
and was extracted with acetonitrile to remove sulfones. The acetonitrile-extracted
oil phase was analyzed by the Sulfur-in-Oil Analyzer.

120
3.3.4.6 Results and Discussion

Table 3.13 shows the desulfurization of DBT solution with different ionic liquids
applied with 3 hours reaction time. Both systems with ionic liquid [EMIM][EtSO
4
]
and [TMIM][MeSO
4
], respectively, demonstrated similar desulfurization efficiency
comparing with the system using [BMIM][PF
6
]. With relatively high stability,
[EMIM][EtSO
4
] and [TMIM][MeSO
4
] are possible substitutes for [BMIM][PF
6
].
System using [TMBP][MeSO
4
] has a slightly lower desulfurization efficiency, and
thus is not considered as substitute to [BMIM][PF
6
]. Considering cost and ease of
application, [EMIM][EtSO
4
] will be used in these studies.

Table 3.13 Desulfurization of DBT solution with respect to type of IL

IL State at 25°C
Initial Sulfur Content
(ppm
w
)
Final Sulfur Content
(ppm
w
)
Sulfur Removal
(%)
[BMIM][PF
6
] Liquid 1006 < 20 > 98
[EMIM][EtSO
4
] Liquid 1006 < 20 > 98

[TMIM][MeSO
4
] Solid 1006 < 20 > 98
[TMBP][MeSO
4
] Solid 1006 48 95

121
3.3.5 Effect of Treatment Methods
3.3.5.1 Time of Ultrasonication

Although it is believed that ultrasonication time would generally increase oxidative
desulfurization efficiency, prolonged ultrasonication would also increase energy
consumption and cost. Besides, it is suggested that the reaction would generally
approach steady state quickly within the first few minutes of ultrasonication (Mei et
al., 2003).

A combination of magnetic stirring and ultrasonication was used starting with the
third generation of the UAOD system so as to improve desulfurization efficiency
and reduce operational cost (Cheng et al., 2008). Based on literature, 10 minutes
ultrasonication was applied in the combination of different magnetic stirring time to
evaluate the effect of ultrasonication on desulfurization system using KO
2
as
oxidant.

122

3.3.5.2 Experimental Procedure

DBT model sulfur compound was dissolved into a solvent mixture with 30% wt.
toluene and 70% wt. n-decane to make a stock solution with approximately 1000
ppm
w
sulfur content. 0.7 gram of KO
2
was first mixed with 5 grams of
[EMIM][EtSO
4
] and 0.1 gram of TOAF to produce mixture A; 10 grams of stock
solution of model sulfur compound were mixed with 2 grams acetic acid to produce
mixture B.

Mixture A was then slowly added to mixture B, and the resulting mixture was
heated up to 70°C with continuous magnetic stirring followed by 10 minutes
ultrasonication. Emulsion formed after the treatment process was separated by
centrifugation. Oil phase was collected and was extracted with acetonitrile to
remove sulfones. The acetonitrile-extracted oil phase was analyzed by the Sulfur-in-
Oil Analyzer.

123

3.3.5.3 Results and Discussion

Table 3.14 shows the desulfurization efficiency of DBT solutions with different
reaction time under continuous magnetic stirring, followed by ultrasound irradiation.
It was observed that 51% desulfurization could be achieved by 10 minutes
ultrasonication alone. However, the desulfurization efficiency was increased only
from 51% to 53% by doubling ultrasound irradiation to 20 minutes. While
ultrasound could enhance OSCs oxidation, it could also increase the rate of
degradation of oxidant and other catalysts including ionic liquid. Prolonged
ultrasonication would not effectively enhance desulfurization in this system and
therefore, the optimum reaction time under ultrasonication was considered as 10
minutes.

Comparing the results from experiment with magnetic mixing alone and that with
magnetic mixing followed by 10 minutes ultrasonication, experiments with 10
minutes ultrasonication demonstrated higher desulfurization efficiency, especially
in shorter total reaction time. Higher than 98% desulfurization can be achieved in
the system with 110 minutes of mixing followed by 10 minutes ultrasound
irradiation. However, similar results can be achieved with 180 minutes of mixing
alone. Thus, it could be more energy-effective to use magnetic mixing alone.

124

Table 3.14 Desulfurization of DBT solution under magnetic stirring and ultrasound

In our study, it is demonstrated that it requires 180 minutes to desulfurize 1000
ppm
w
DBT solution by mechanical mixing alone. With the aid of 10 minutes
ultrasonication, it requires only 120 minutes to achieve the same goal. It is believed
that ultrasonication is a possible enhancement method in the new oxidative
desulfurization with the application of superoxide. It can provide high local
temperature and pressure in microenvironment. As a result, reaction rate is
increased and thus improving productivity.

Mixing
Time (min)
Ultrasonication
(min)
Total Reaction
Time (min)
Initial Sulfur
Content (ppm
w
)
Final Sulfur
Content (ppm
w
)
Sulfur
Removal (%)
0 0 0 1006 1006 0
30 0 30 1006 368 63
60 0 60 1006 141 86
120 0 120 1006 49 95
180 0 180 1006 < 20 > 98

0 10 10 1006 494 51
20 10 30 1006 274 73
50 10 60 1006 93 91
110 10 120 1006 < 20 > 98
170 10 180 1006 <20 > 98

0 20 20 1006 476 53

125
3.4 Desulfurization Efficiency on Various Model Sulfur Compounds
3.4.1 Identification of Experimental Optimum Conditions

It is known that there are various types of organic sulfur compounds in petroleum
fuel. Thiophene, benzothiophene, dibenzothiophenes, and their derivatives are the
major OSCs in middle distillates such as diesel, and heavier distillates. Among all, 4,
6-dimethyldibenzothiophene is well known for its refractory characteristic in
traditional HDS process.

Although the optimal conditions, including the selection of oxidant, catalysts, ionic
liquid, and their dosage in our process to desulfurize DBT has been discussed
earlier in this chapter, it is important to test if the process can desulfurize other
OSCs. In order to investigate the applicability of the process to various OSCs in
petroleum products, different model sulfur compounds, including T, 2MT, BT,
2MBT), DBT, and 4,6DMDBT were selected to study their treatability under our
process’s optimal conditions. Table 3.15 provides a list of the selected optimum
desulfurization conditions from this chapter.

126

Table 3.15 Selected conditions for desulfurization in model compound studies with 10
minutes ultrasonication
Solution of Target Model Sulfur Compound (1000ppm
w
) 10 grams
Oxidant: KO
2
0.7 gram
Phase Transfer Catalyst: TOAF 0.1 gram
Acid: 95% HAc and 5% TFA 2 grams
Ionic Liquid: [EMIM][EtSO
4
] 5 grams
Reaction Temperature 70°C
Magnetic Stirring Time 110 minutes
Ultrasonication Time 10 minutes

127

3.4.2 Experimental Procedure

The selected model sulfur compound, either T, 2MT, BT, 2MBT, DBT or
4,6DMDBT was dissolved into a solvent mixture with 30% wt. toluene and 70% wt.
n-decane to make a stock solution with approximately 1000 ppm
w
sulfur content.

A dosage of 0.7 gram KO
2
was added to 5 grams of selected ionic liquid and 0.1
gram of TOAF to produce mixture A; 10 grams of stock solution of model sulfur
compound were mixed with 1.9 grams acetic acid and 0.1 gram of TFA to produce
mixture B.

Mixture A was slowly added to mixture B, and the resulting mixture was heated up
to 70°C with continuous magnetic stirring for 110 minutes followed by
ultrasonication for 10 minutes. Emulsion formed after the treatment process was
separated by centrifugation. Oil phase was collected and was extracted with
acetonitrile to remove sulfones. The acetonitrile-extracted oil phase was analyzed
by the Sulfur-in-Oil Analyzer.

128

3.4.3 Results and Discussion

Table 3.16 shows the desulfurization efficiencies on various model sulfur
compounds under the designed conditions. The process demonstrated higher than
97% desulfurization on all model sulfur compounds in this study. Higher than 98%
desulfurization on BT, 2MBT, DBT and 4,6DMDBT were achieved, while slightly
lower desulfurization efficiencies (97%) on T and 2MT were observed. The high
desulfurization efficiencies on various model sulfur compounds suggested that the
designed process could be an effective alternative desulfurization process.

Table 3.16 Desulfurization of various model sulfur compounds

Model Sulfur Compounds
Initial Sulfur Content
(ppm
w
)
Final Sulfur Content
(ppm
w
)
Sulfur Removal
(%)
T 985 23 97
2MT 1010 26 97
BT 1012 < 20 > 98
2MBT 994 < 20 > 98
DBT 1006 < 20 > 98
4,6DMDBT 1007 < 20 > 98

129
3.5 Kinetic Studies of Desulfurization on Model Sulfur Compounds

Kinetic studies of oxidative desulfurization of model sulfur compounds in carboxylic acid
/ H
2
O
2
have been demonstrated in recent research articles (Dhir et al., 2009; Huang et al.,
2007; Te et al., 2001; Yan et al., 2007). The reaction rates depend on various factors,
including concentration of organic sulfur compounds, concentration of oxidant and
concentration of catalysts. Micro-kinetic pathway of OSC oxidation could include
multiple reactions as illustrated in Figure 3.5. In general, the reaction could be written as

Activation of Catalyst: O H MO M O H
2 2 2
+ ↔ + (Eq. 3.1)

Oxidation of OSC: M RSO MO RS 2 2
2
+ ↔ + (Eq. 3.2)

Overall Reaction: O H RSO O H RS
M
2 2 2 2
2 2 + → ← + (Eq. 3.3)

where M is the catalyst applied. In such a case, the overall rate equation can be expressed
as

β α
] [
2 2
O H kC
dt
dC
r = − = (Eq. 3.4)

130
where r is the rate of desulfurization of the selected OSC, k is the reaction rate constant,
C is the concentration of the selected OSC. In cases with highly excess amount of
hydrogen peroxide, change in concentration of oxidant would not be significant and thus,
[H
2
O
2
] can be considered as a constant. Putting k’ = k[H
2
O
2
]
β
, the rate equation can be
simplified to

α
C k
dt
dC
r ' = − = (Eq. 3.5)

Figure 3.5 Superstructure of alternate reaction micro-kinetic pathway (Dhir et al., 2009)

131
Because desulfurization using potassium superoxide is not yet widely applied, very
limited information on its reaction pathway and kinetics can be found in the literature.
With excessive dosage of KO
2
, OSC oxidation should follow the pseudo first order
reaction kinetics. Thus, the rate equation could be rewritten as

C k
dt
dC
r ' = − = (Eq. 3.6)

Upon integration,

t k
C
C
o
t
' ln − = (Eq. 3.7)

where C
o
is the initial OSC concentration, C
t
is the OSCs concentration at time t. The
effect of temperature on the reaction rate constant can be expressed by the Arrhenius
Equation listed below:

RT E
a
Ae k
/
'
−
= (Eq. 3.8)

RT
E
A k
a
− = ln ' ln (Eq. 3.9)

where A is the Arrhenius frequency factor, E
a
is the activation energy of the reaction, R is
the ideal gas constant and T is the temperature in K.

132

In the previous section, studies on various model sulfur compounds including T, 2MT,
BT, 2MBT, DBT, 4,6DMDBT under optimum conditions have been discussed.

In this section, desulfurization kinetic studies are performed on the model sulfur
compounds in carboxylic acid / KO
2
mixture to estimate the rate constant k’. Effect of
temperature on reaction rate for two model compounds, BT and DBT are also studied.
Ultrasonication was not applied in this study. Table 3.17 provides a list of the selected
desulfurization conditions.

Table 3.17 Selected conditions for desulfurization in model compound studies without
ultrasonication
Solution of Target Model Sulfur Compound (1000ppm
w
) 10 grams
Oxidant: KO
2
0.7 gram
Phase Transfer Catalyst: TOAF 0.1 gram
Acid: 95% Acetic Acid and 5% TFA 2 grams
Ionic Liquid: [EMIM][EtSO
4
] 5 grams
Reaction Temperature 30 to 70°C
Magnetic Stirring Time 0 to 180 minutes

133

3.5.1 Experimental Procedure

The selected model sulfur compound, either T, 2MT, BT, 2MBT, DBT or
4,6DMDBT was dissolved into a solvent mixture with 30% wt. toluene and 70% wt.
n-decane to make a stock solution with approximately 1000 ppm
w
sulfur content.

A dosage of 0.7 gram KO
2
was added to 5 grams of selected ionic liquid and 0.1
gram of TOAF to produce mixture A; 10 grams of stock solution of model sulfur
compound were mixed with 1.9 grams acetic acid and 0.1 gram of TFA to produce
mixture B.

Mixture A was slowly added to mixture B, and the resulting mixture was heated up
to the designated temperature with continuous magnetic stirring up to 180 minutes.
Emulsion formed after the treatment process was separated by centrifugation. Oil
phase was collected and was extracted with acetonitrile to remove sulfones. The
acetonitrile-extracted oil phase was analyzed by the Sulfur-in-Oil Analyzer.

134
3.5.2 Results and Discussion

Table 3.18 and Figure 3.6 show the desulfurization efficiencies on various model
sulfur compounds as a function of time under the selected conditions listed in Table
3.17. With high dosage of KO
2
, it is assumed that the desulfurization of the model
sulfur compounds follow the pseudo first order reaction kinetics. Thus, linear
relation should be demonstrated by plotting -ln(C
t
/C
o
) versus t with the slope equals
to the reaction rate constant k’, as illustrated in Figure 3.7 to Figure 3.9. The
reaction rate constants for each model sulfur compounds are summarized in Table
3.19. The reaction rate constants using 30% H
2
O
2
as oxidant are listed in Table 3.20
for comparison.

Table 3.18 Desulfurization efficiencies with respect to reaction time

Mixing Time
(min)
T (°C)
Initial Sulfur
Content (ppm
w
)
Final Sulfur
Content (ppm
w
)
Sulfur
Removal (%)
0 70 985 985 0
30 70 985 380 61.4
60 70 985 178 81.9
120 70 985 71 92.8
T
180 70 985 28 97.2

135
Table 3.18 (continued)

Mixing Time
(min)
T (°C)
Initial Sulfur
Content (ppm
w
)
Final Sulfur
Content (ppm
w
)
Sulfur
Removal (%)
0 70 1010 1010 0
30 70 1010 411 59.3
60 70 1010 187 81.5
120 70 1010 88 91.3
2MT
180 70 1010 31 96.9
0 70 1012 1012 0
30 70 1012 341 66.3
60 70 1012 112 88.9
120 70 1012 35 96.5
BT
180 70 1012 <20 > 98
0 70 994 994 0
30 70 994 358 64
60 70 994 151 84.8
120 70 994 42 95.8
2MBT
180 70 994 21 97.9
0 70 1006 1006 0
30 70 1006 368 63.4
60 70 1006 141 86
120 70 1006 49 95.1
DBT
180 70 1006 < 20 > 98
0 70 1007 1007 0
30 70 1007 388 61.5
60 70 1007 167 83.4
120 70 1007 58 94.2
4, 6 DMDBT
180 70 1007 27 97.3

136

0 20406080 100 120 140 160 180 200
0
10
20
30
40
50
60
70
80
90
100
T
2MT
BT
2MBT
DBT
4, 6 DMDBT
Time, min
Sulfur Rem oval, %

Figure 3.6 Desulfurization efficiencies of various model sulfur compounds for KO
2
process

20 40 60 80 100 120 140 160 180 200
0
0.5
1
1.5
2
2.5
3
3.5
4
T
Linear Regression for T
2MT
Linear Regression for 2MT
Time, min
-ln(Ct/C0)

Figure 3.7 Linear regression of –ln(C
t
/C
o
) vs time for T and 2MT for KO
2
process

137

20 40 60 80 100 120 140 160 180 200
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
BT
Linear Regression for BT
2MBT
Linear Regression for 2MBT
Time, min
-ln(Ct/C0)

Figure 3.8 Linear regression of –ln(C
t
/C
o
) vs time for BT and 2MBT for KO
2
process

20 40 60 80 100 120 140 160 180 200
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
DBT
Linear Regression for DBT
4, 6 DMDBT
Linear Regression for 4, 6
DMDBT
Time, min
-ln(Ct/C0)

Figure 3.9 Linear regression of –ln(C
t
/C
o
) vs time for DBT and 4,6DMDBT for KO
2
process

138

Table 3.19 Rate constants for various model sulfur compounds using KO
2

Model Sulfur Compound k', min
-1
R
2

T 0.0190 0.991
2MT 0.0184 0.987
BT 0.0229 0.935
2MBT 0.0212 0.970
DBT 0.0211 0.978
4,6 DMDBT 0.0197 0.978

The reaction rate constants are within the range of 0.018 to 0.022 min
-1
for all 6
model sulfur compounds tested. The reaction rate constants indicate the oxidation
efficiencies of the tested model sulfur compounds marginally diminish in following
order: BT > 2MBT > DBT > 4, 6DMDBT > T > 2MT.

Table 3.20 Rate constants for various model sulfur compounds using H
2
O
2
(Cheng, 2007)
Model Sulfur Compound k', min
-1
R
2

T 0.0196 0.997
2MT 0.0199 0.995
BT 0.0276 0.998
2MBT 0.0218 0.997
DBT 0.0228 0.997
4,6 DMDBT 0.0230 0.979

139

From Table 3.19 and 3.20, the reaction rate constants using KO
2
as oxidant are
comparable to those using 30% H
2
O
2
as oxidant. With all R
2
values greater than
0.93, the plots demonstrate good linear relation and thus, it is valid to assume that
the reactions follow the pseudo first order reaction kinetics. However, R
2
value for
the linear regression of -ln(C
t
/C
o
) versus time in the experiments using KO
2
as
oxidant are relatively lower than the those using 30% H
2
O
2
as oxidant.

The reaction rate constants for BT and DBT at different temperatures can be found
by conducting the same experimental conditions listed in Table 3.17. Linear
regression of –ln(C
t
/C
o
) versus time at different temperature for the desulfurization
of BT and DBT are illustrated in Figure 3.10 and Figure 3.11, respectively. Using
the reaction rate constants at different temperature, the activation energy E
a
and the
Arrhenius frequency factor A can be estimated from the Arrhenius equation by
plotting -ln(k’) versus 1/T, as illustrated in Figure 3.12. The value of E
a
and A for
BT and DBT are summarized in Table 3.21. The apparent activation energies also
demonstrated that BT with lower activation energy has a higher reactivity as
compared to DBT.

140

Table 3.21 Apparent activation energies for oxidation of BT and DBT
k', min
-1

Temperature, K 303 323 343
E
a
, kJ/mol A, min
-1

BT 0.0061 0.0155 0.0229 28.77 601.85
DBT 0.0050 0.0098 0.0211 31.01 1085.72

Figure 3.10 Linear regression of –ln(C
t
/C
o
) vs time at different temperature for BT

20 40 60 80 100 120 140 160 180 200
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
BT 30C
Linear Regression
for BT 30C
BT 50C
Linear Regression
for BT 50C
BT 70C
Linear Regression
for BT 70C
Time, min
-ln(C
t
/C
o
)
R
2
= 0.965
R
2
= 0.992
R
2
= 0.988

141

Figure 3.11 Linear regression of –ln(C
t
/C
o
) vs time at different temperature for DBT

Figure 3.12 -ln(k’) versus 1/T for model sulfur compounds BT and DBT

20 40 60 80 100 120 140 160 180 200
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
DBT 30C
Linear Regression
for DBT 30C
DBT 50C
Linear Regression
for DBT 50C
DBT 70C
Linear Regression
for DBT 70C
Time, min
-ln(C
t
/C
o
)
R
2
= 0.978
R
2
= 0.997
R
2
= 0.929
2.80E-03 2.90E-03 3.00E-03 3.10E-03 3.20E-03 3.30E-03 3.40E-03
3
3.5
4
4.5
5
5.5
BT
Linear Regression for BT
DBT
Linear Regression for DBT
1/T
-lnk'
R
2
= 0.995
R
2
= 0.962

142
3.6 Preliminary Study on UV Assisted Desulfurization Process
3.6.1 Challenge in UAOD Processes

It has been demonstrated that different generation of UAOD processes can achieve
high oxidation rates with various OSCs, and thus, yielding high desulfurization
efficiencies. However, power consumption in ultrasonication is considerably high.
For instance, power output for ultrasound reactor VCX-750 is 750 watts. This could
be the major drawback of UAOD systems. Besides, ultrasonication could cause
decomposition of various catalysts including phase transfer catalyst, trifluoroacetic
acid, and ionic liquids. Therefore, alternative activation methods with lower energy
consumption are of particular interest.

UV irradiation is known to improve oxidation in water treatment systems. Recent
researches have also demonstrated the feasibility in utilizing UV to enhance
oxidative desulfurization. However, most UV related desulfurization processes
require use of high pressure mercury UV lamp with energy output greater than 200
watts, with irradiation time longer than 1 hour (Matsuzawa et al., 2001). This would
result in even higher energy consumption comparing with ultrasonication. Therefore,
a low power UV lamp with power output at 8 watts was used in this preliminary
study to illustrate the feasibility of UV assisted desulfurization process with low
energy consumption. 30% wt. hydrogen peroxide was used in this point of the study
for better comparison.

143
3.6.2 Experimental Procedure

Selected model sulfur compound, either BT or DBT, was dissolved into a solvent
mixture with 30% wt. toluene and 70% wt. n-decane to make a stock solution with
approximately 1000 ppm
w
sulfur content. 1 gram of H
2
O
2
was first mixed with 1
gram of selected ionic liquid and 0.1 gram of TOAF to produce mixture A; 10
grams of stock solution of model sulfur compound were mixed with 1.9 grams
acetic acid and 0.1 gram of TFA to produce mixture B.

Mixture A was then slowly added to mixture B, and 0.1 gram of a photo-catalyst,
titanium (IV) oxide was added to the mixture. The resulting mixture was kept at
ambient condition using continuous magnetic stirring for 60 minutes with or
without UV irradiation.

The experiments were performed in dark to eliminate the interference by other light
sources. Emulsion formed after the treatment process was separated by
centrifugation. Oil phase was collected and was extracted with acetonitrile to
remove sulfones. The acetonitrile-extracted oil phase was analyzed by the Sulfur-in-
Oil Analyzer.

144
3.6.3 Results and Discussion

Table 3.22 shows that desulfurization for BT and DBT were not significant in 60
minutes mixing alone. In contrast, higher than 50% desulfurization can be achieved
for both BT and DBT solutions by applying UV irradiation. The result is
comparable to system with 10 minutes ultrasonication.

Table 3.22 Desulfurization of model sulfur compounds with UV

The power output of the ultrasound device used is 750W and the power output used
of UV lamp is 8W. Considering the energy consumption, 10 minutes ultrasonication
requires more than 15 times energy comparing with 60 minutes UV irradiation.
More experiments are required to test for the applicability of the system on other
model compounds and fuel samples.

Mixing
(min)
UV
Ultrasound
(min)
S
initial
(ppm
w
)
S
final
(ppm
w
)
S Removal
(%)
BT 60 Off 0 1012 971 4
DBT 60 Off 0 1006 968 4
BT 60 On 0 1012 461 54
DBT 60 On 0 1006 418 58
BT 50 Off 10 1012 384 62
DBT 50 Off 10 1006 360 64

145

3.7 Summary and Conclusion

This study aimed at developing alternative oxidation processes to achieve high
desulfurization. Use of solid type oxidant can significantly reduce the volume and weight
of required oxidant as compared with systems using 30% wt. hydrogen peroxide solution
as oxidant.

Potassium superoxide is found to be effective as the alternative oxidant due to the high
effectiveness in oxidizing OSCs, high purity, high stability for storage under ambient
conditions and relatively low adverse environmental impacts comparing with other solid
oxidants. By using potassium superoxide, the weight of oxidant required can be reduced
to one-fifth of the weight when hydrogen peroxide is used.

In the reaction system with multiphase, phase transfer catalyst was demonstrated to
promote reaction as a result of interfacial exchange. With the application of
tetraalkylammonium salts, an intermediate product tetraalkylammonium superoxide
could be formed so as to transfer superoxide radical to organic phase. PTC counter ion
selection is also important for successful phase transfer catalysis. It is suggested that
fluoride should be used as the PTC counter ion to maximize the efficiency.

146
Acid catalysts including acetic acid and trifluoroacetic acid are also used to improve
reaction rate and selectivity. It has been demonstrated that higher acidity of acid catalysts
can promote oxidative reactions using hydrogen peroxide as oxidant, but it is relatively
not significant in system using superoxide as oxidant. Trifluoroacetic acid could be
decomposed forming acid fumes upon ultrasonication, and explosive reaction may also
occur in extreme cases. Thus, minimum amount of trifluoroacetic acid should be used in
the system. Trifluoroacetic acid would only be used up to 5% of the total weight of acid
catalyst to enhance oxidative desulfurization in this system.

Addition of ionic liquids can improve desulfurization efficiency of the system. Although
1-butyl-3-methylimidazolium hexafluorophosphate is known as the most common ionic
liquid used in oxidative desulfurization, its instability in ultrasonication and oxidants
systems could cause operational problems. Corrosive gases including hydrogen fluoride,
phosphorus pentafluoride and phosphorus oxyfluoride could be generated as
decomposition byproducts of this ionic liquid. High stability ionic liquids such as 1-ethyl-
3-methylimidazolium ethylsulfate and 1,2,3-trimethylimidazolium methylsulfate could be
used to alleviate formation of decomposition byproduct.

It is demonstrated that the modified process with solid oxidant is applicable to desulfurize
various model sulfur compounds. Higher than 98% oxidation and desulfurization can be
achieved on benzothiophene, 2-methyl benzothiophene, dibenzothiophene, and 4, 6-
dimethyl dibenzothiophene, while 97% oxidation and desulfurization can be achieved on

147
thiophene and 2-methyl thiophene. As shown in the kinetic studies, the oxidation
efficiencies of the model sulfur compounds decrease in the following order: BT > 2MBT
> DBT > 4, 6DMDBT > T > 2MT.

A preliminary investigation on UV assisted oxidative desulfurization conducted in this
study demonstrated a possible way to lower the power consumption in UAOD processes.
Higher than 50% desulfurization on both BT and DBT can be achieved in 1 hour under
UV irradiation with application of 30% wt. hydrogen peroxide, phase transfer catalyst,
acid catalysts, photo-catalyst, and ionic liquid. This result is comparable to a system with
same reactants under ultrasonication for 10 minutes. It was demonstrated that energy
consumption can be lowered by roughly 95% by using low power UV irradiation instead
of ultrasonication. Further studies could be conducted to investigate the applicability of
UV irradiation in desulfurization other OSCs and fuel samples.

148
CHAPTER 4: MODIFIED OXIDATIVE DESULFURIZATION USING
KO
2
AND H
2
O
2
ON PETROLEUM FUEL
4.1 Introduction

As discussed in Chapter 1, alkyl benzothiophene derivatives and alkyl dibenzothiophene
derivates are the major organic sulfur compounds found in mid-distillates such as diesel,
and also heavy distillates such as residual oil. These two groups of OSCs, especially with
alkyl substituent at 4- and/or 6-postion are considered refractory to traditional
hydrodesulfurization process. Thus, these groups of OSCs are the major sulfur
compounds found in hydrodesulfurized fuels.

In previous generations of the UAOD processes, high desulfurization efficiencies have
been demonstrated on benzothiophene and dibenzothiophene using 30% wt. hydrogen
peroxide as oxidant under mild conditions (Wan et al., 2007; Etemadi et al., 2007). The
third generation of the UAOD process, or the modified UAOD process, has successfully
demonstrated the production of ultra-low sulfur diesel using 30% wt. H
2
O
2
as oxidant.
However, this process is found not to be applicable to heavy oil desulfurization mainly
due to heavy deposition of asphaltenes.

Besides, concentrated hydrogen peroxide is a potential hazard due to its spontaneous
ignition and explosive properties. Diluted hydrogen peroxide solutions are commonly
used in the oxidative desulfurization process. However, dilution would increase the

149
volume and weight of oxidant required together with a reduction of the reaction
efficiency. These effects would increase the overall cost of the process.

The modification using potassium superoxide (KO
2
) as oxidant, discussed in Chapter 3,
has demonstrated high desulfurization efficiencies for thiophene, benzothiophene,
dibenzothiophene and their derivatives. The results are comparable to the third generation
of the UAOD process (modified UAOD process). In this study, modified desulfurization
process using KO
2
as oxidant was applied to petroleum fuels including diesel and
residual oil. The process can be operated under mild conditions with the enhancement
from phase transfer catalysis and acid catalysis (Chan et al., 2009). The process consists
of two steps: selective oxidation of OSCs in fuel with or without ultrasound irradiation
followed by separation of oxidized OSCs from fuels by solvent extraction or solid
adsorption.

In this chapter, modified desulfurization process using KO
2
as oxidant was applied to
some of the commercial fuels including marine gas oil (MGO), jet propellant 8 (JP-8),
sour diesel and also residual fuel oils to investigate the feasibility of desulfurizing
commercial fuels.

150
4.2 Materials

Three diesel fuels, including jet propellant 8 (JP-8) with 782 ppm
w
sulfur content
received from U.S. Army Research Laboratory, Adelphi, Maryland; marine gas oil (MGO)
with sulfur 1631 ppm
w
content received from Navy Station, Long Beach, California; and
sour diesel with 8117 ppm
w
sulfur content received from Golden Eagle Oil Refinery Inc.,
Woods Cross, Utah; were used as the mid-distillate feedstock. Two residual oil samples,
including residual oil no. 6 (RO-6) received from Eco Energy Solutions Inc., Reno,
Nevada; and intermediate fuel oil (IFO) received from Intelligent Energy Inc., Long
Beach, California; were used as the heavy-distillate feedstock.

Oxidants used including 30% wt. H
2
O
2
solution was obtained from VWR Inc., while
potassium superoxide, acid catalysts (glacial acetic acid and trifluoroacetic acid), and
phase transfer catalysts (tetraoctylammonium bromide and tetraoctylammonium chloride)
were obtained from Sigma-Aldrich Co., while tetraoctylammonium fluoride was
synthesized by halogen exchange process (Dermeik et al., 1989). Ionic liquids including
1-butyl-3-methyl-imidazolium hexafluorophosphate [BMIM][PF
6
], 1-ethyl-3-
methylimidazolium ethylsulfate [EMIM][EtSO
4
], and 1,2,3-trimethyl-imidazolium
methylsulfate [TMIM][MeSO
4
] were obtained from Sigma-Aldrich Co. Acetonitrile,
solvent used for sulfones extraction, was obtained from VWR Inc.

151
4.3 Experimental Procedure and Analytical Method

Known amount of oxidant, either potassium superoxide or 30% wt. hydrogen peroxide,
was first mixed with 5 grams of selected ionic liquid, either 1-butyl-3-methylimidazolium
hexafluorophosphate [BMIM][PF
6
], 1-ethyl-3-methyl-imidazolium ethylsulfate
[EMIM][EtSO
4
] or 1,2,3-trimethylimidazolium methyl-sulfate [TMIM][MeSO
4
], and 0.1
gram of tetraoctylammonium fluoride to produce mixture A; 10 grams of the selected
mid-distillate feedstock, either JP-8, MGO or sour diesel; or heavy-distillate feedstock,
either RO-6 or IFO, were mixed with 2 grams of acetic acid to produce mixture B.

Mixture A was slowly added to mixture B, and the resulting mixture was heated up to 70°
C with continuous mixing for desired time period. Emulsion formed after the treatment
process was separated by centrifugation. Oil phase was collected and extracted with
acetonitrile. The acetonitrile-extracted oil phase was analyzed by the Sulfur-in-Oil
Analyzer.

The total sulfur concentration of the fuel samples was determined according to ASTM
D4294 and ISO 8754. Sulfur-in-Oil Analyzer (SLFA-20), manufactured by Horiba Inc.
was used to measure total sulfur content of samples based on non-dispersive X-ray
fluorescence (NDXRF). Sulfur-in-Oil Analyzer is applicable to samples with total sulfur
as high as 5% wt., with a lower detection limit of 20ppm
w
.

152
4.4 Results and Discussion
4.4.1 Desulfurization of JP-8

Table 4.1 shows the desulfurization of JP-8 using potassium superoxide with the
application of different ionic liquids. Table 4.2 shows the desulfurization of JP-8
using 30% wt. hydrogen peroxide using different ionic liquids for comparison. The
weight of oxidant used was fixed at 0.5 gram. The reaction time was 180 minutes.

Table 4.1 Desulfurization of JP-8 using KO
2
as oxidant

As shown in Table 4.1, higher than 98% desulfurization efficiencies for JP-8 were
achieved by using potassium superoxide as oxidant, regardless of the type of ionic
liquid used. Thus, all ionic liquids including [BMIM][PF
6
], [EMIM][EtSO
4
], and
[TMIM][MeSO
4
] were considered applicable to this desulfurization process. As
Ionic Liquid KO
2
[S]
Initial
[S] : [KO
2
] [S]
Final
[S] Removal
g mmol ppm
w
mmol mol : mol ppm
w
%
[BMIM][PF
6
] 0.5 9.6 782 0.24 1 : 29 < 20 > 98
[EMIM][EtSO
4
] 0.5 9.6 782 0.24 1 : 29 < 20 > 98
[TMIM][MeSO
4
] 0.5 9.6 782 0.24 1 : 29 < 20 > 98

153
discussed in Chapters 2 and 3, [BMIM][PF
6
] could decompose to corrosive gases
under ultrasonication and/or oxidation conditions. Therefore, [EMIM][EtSO
4
], and
[TMIM][MeSO
4
] would be considered as better options of ionic liquids applied to
this desulfurization process.

Table 4.2 Desulfurization of JP-8 using 30% wt. H
2
O
2
as oxidant

Comparing with systems using 30% wt. hydrogen peroxide (Figure 4.2), systems
using potassium superoxide afforded higher desulfurization efficiencies. One of the
possible reasons could be a higher availability of oxidant per mass of potassium
superoxide as compared to 30% wt. hydrogen peroxide system.

Ionic Liquid H
2
O
2
(30%) [S]
Initial
[S] : [H
2
O
2
] [S]
Final
[S] Removal
g
H
2
O
2
mmol
ppm
w
mmol mol : mol ppm
w
%
[BMIM][PF
6
] 0.5 4.4 782 0.24 1 : 18 101 87
[EMIM][EtSO
4
] 0.5 4.4 782 0.24 1 : 18 92 90
[TMIM][MeSO
4
] 0.5 4.4 782 0.24 1 : 18 112 86

154
Another set of experiments were conducted using 0.5 gram of potassium superoxide
as oxidant and [EMIM][EtSO
4
] as the ionic liquid. Desulfurization efficiencies as a
function of reaction time are shown in Figure 4.1. It is demonstrated that the
desulfurization efficiency approached to 100% at 100 minutes.

0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160 180
Time (min)
Desulfurization (%)

Figure 4.1 Desulfurization of JP-8 with respect to total reaction time

155

4.4.2 Desulfurization of MGO

Table 4.3 shows the desulfurization of MGO using potassium superoxide with using
different ionic liquids. Table 4.4 shows the desulfurization of MGO with 30% wt.
hydrogen peroxide using different ionic liquids. The weight of oxidant used was 0.5
gram for both systems. The reaction took place at 70°C with reaction time of 180
minutes.

Similar to desulfurization of JP-8, higher than 98% desulfurization efficiencies was
achieved for MGO by using potassium superoxide as oxidant, regardless of the type
of ionic liquid used. However, under the same conditions, the maximum
desulfurization efficiency achieved ffor 30% wt. H
2
O
2
was only 82%. A possible
explanation could be the availability of 1.5 times more oxidant in the superoxide
system as compared to hydrogen peroxide system.

As described in the previous chapters, corrosive gases could be generated from
[BMIM][PF
6
] under strong oxidizing condition. Without losing desulfurization
efficiency, alternative ionic liquids such as [EMIM][EtSO
4
] and [TMIM][MeSO
4
]
should be considered for application to this desulfurization process.

156

Table 4.3 Desulfurization of MGO using KO
2
as oxidant

Table 4.4 Desulfurization of MGO using 30% wt. H
2
O
2
as oxidant

Ionic Liquid KO
2
[S]
Initial
[S] : [KO
2
] [S]
Final
[S] Removal
g mmol ppm
w
mmol mol : mol ppm
w
%
[BMIM][PF
6
] 0.5 9.6 1600 0.5 1 : 14 < 20 > 99
[EMIM][EtSO
4
] 0.5 9.6 1600 0.5 1 : 14 < 20 > 99
[TMIM][MeSO
4
] 0.5 9.6 1600 0.5 1 : 14 < 20 > 99
Ionic Liquid H
2
O
2
(30%) [S]
Initial
[S] : [H
2
O
2
] [S]
Final
[S] Removal
g
H
2
O
2
mmol
ppm
w
mmol mol : mol ppm
w
%
[BMIM][PF
6
] 0.5 4.4 1600 0.5 1 : 8.8 358 77
[EMIM][EtSO
4
] 0.5 4.4 1600 0.5 1 : 8.8 294 82
[TMIM][MeSO
4
] 0.5 4.4 1600 0.5 1 : 8.8 309 81

157

4.4.3 Desulfurization of Sour Diesel

Tables 4.5 and 4.6 show the desulfurization of sour diesel using potassium
superoxide and 30% wt hydrogen peroxide, respectively, using different ionic
liquids. The weight of oxidant used was fixed at 0.5 gram for both systems. The
reaction temperature was maintained at 70°C with a reaction time of 180 minutes.

Table 4.5 Desulfurization of sour diesel using KO
2
as oxidant

Table 4.6 Desulfurization of sour diesel using 30% wt. H
2
O
2
as oxidant
Ionic Liquid KO
2
[S]
Initial
[S] : [KO
2
] [S]
Final
[S] Removal
g mmol ppm
w
mmol mol : mol ppm
w
%
[BMIM][PF
6
] 0.5 9.6 8100 2.53 1 : 2.8 491 94
[EMIM][EtSO
4
] 0.5 9.6 8100 2.53 1 : 2.8 428 95
[TMIM][MeSO
4
] 0.5 9.6 8100 2.53 1 : 2.8 432 95
Ionic Liquid H
2
O
2
(30%) [S]
Initial
[S] : [H
2
O
2
] [S]
Final
[S] Removal
g
H
2
O
2
mmol
ppm
w
mmol mol : mol ppm
w
%
[BMIM][PF
6
] 0.5 4.4 8100 2.53 1 : 1.7 4733 42
[EMIM][EtSO
4
] 0.5 4.4 8100 2.53 1 : 1.7 4348 46
[TMIM][MeSO
4
] 0.5 4.4 8100 2.53 1 : 1.7 4197 48

158

As can be seen in Table 4.5, a maximum of 95% desulfurization efficiency can be
achieved by using potassium superoxide as oxidant, with either [EMIM][EtSO
4
], or
[TMIM][MeSO
4
]. Lower desulfurization efficiency (42%) was obtained in the
system using [BMIM][PF
6
] as ionic liquid.

In comparison, for the system using 30% wt. hydrogen peroxide (Table 4.6), the
maximum efficiency was only 48%. It is thus shown that superoxide is highly
superior to hydrogen peroxide in desulfurization of sour diesel.

These studies indicate that superoxide results in higher desulfurization efficiency as
compared to hydrogen peroxide. Two major reasons can be advanced: 1) with the
equal mass consideration, the number of moles of oxidant provided by 30% wt.
H
2
O
2
is 55% less than number of mole of oxidant provided by pure KO
2
; and 2)
water molecules in H
2
O
2
solution could hinder the reaction rate due to dilution
effect (Wypych, 2001). Among the three diesel fuels; i.e., JP-8, MGO, and sour
diesel, the greatest difference in desulfurization efficiencies between the two
oxidants (superoxide versus hydrogen peroxide) occurred for sour diesel.

159

4.4.4 Kinetic Studies of Desulfurization on Various Diesel Samples

Kinetic studies of oxidative desulfurization reported by different research groups
(Tam et al., 1990, Te et al., 2001; Huang et al., 2007; Yan et al., 2007; Dhir et al.,
2009) indicate that oxidation rate using excess amount of oxidant follows pseudo
first order reaction kinetics. In the studies reported herein, desulfurization of model
sulfur compounds reported in Chapter 3, section 3.5, using excess amount of
potassium superoxide follows pseudo first order reaction kinetics. The experimental
rate constants were obtained from the rate equations previously applied to model
sulfur compounds including T, 2MT, BT, 2MBT, DBT and 4,6DMDBT.

While sulfur content in diesel is mainly composed of thiophenes, benzothiophenes,
and dibenzothiophenes and their derivatives, it is expected that the desulfurization
of JP-8, MGO and sour diesel would also follow the pseudo first order reaction
kinetics.

Desulfurization kinetic studies were conducted for the three diesel fuels including
JP-8, MGO and sour diesel, using KO
2
oxidant with sulfur to oxidant ratios of 1:29,
1:14 and 1:2.7, respectively. Figure 4.2 illustrates the desulfurization efficiencies as
a function of reaction time for the three diesel fuels using 0.5 gram of potassium
superoxide as oxidant and [EMIM][EtSO
4
] as ionic liquid. The linearized plots of -

160
ln(C
t
/C
o
) versus time are shown in Figure 4.3 and the reaction rate constants are
summarized in Table 4.7.

Table 4.7 Rate constants for various diesel samples using KO
2
as oxidant
Diesel k', min
-1
R
2

JP-8 0.0231 0.970
MGO 0.0202 0.949
Sour Diesel 0.0151 0.815

0 20 40 60 80 100 120 140 160 180 200
0
20
40
60
80
100
120
JP -8
MGO
Sour Diesel
Time, min
[S] removal, %

Figure 4.2 Desulfurization removal efficiencies of different diesel fuels as a function of
reaction time

161

0 20 40 60 80 100 120 140 160 180
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
JP -8
Linear Regression for
JP -8
MGO
Linear Regression for
MGO
Sour Diesel
Linear Regression for
Sour Diesel
Time, min
-ln(Ct/Co)

Figure 4.3 Desulfurization rates of different diesel fuels as a function of reaction time

The apparent first order rate constant, k’, for JP-8, MGO and sour diesel were
0.0231 min
-1
, 0.0212 min
-1
and 0.0151 min
-1
, respectively. The correlation
coefficient, R
2
, for JP-8 and MGO were 0.970 and 0.949, respectively while R
2
for
sour diesel was significantly lower (0.815).

As can be observed in Figure 4.3, the reaction rate for sour diesel is significantly
lower as compared to those for JP-8 and MGO. A possible explanation for this
phenomenon is the comparatively lower oxidant to sulfur ratio in the sour diesel

162
system, in which the concentration KO
2
became significantly low at longer reaction
time. As a result, the assumption of pseudo first order kinetics was no longer valid.

Future studies should consider the determination of superoxide concentration during
the reaction process. Chemiluminescence determination and direct
spectrophotometrical determination of superoxide concentration are some of the
possible methods (Afanas’ev, 1989; Endo et al., 2002; Rose et al., 2008; Wali et al.,
2002). Also, more tests should be conducted to understand the actual reaction
pathway of KO
2
is also required.

163
4.4.5 Desulfurization of Heavy-Distillates

Table 4.8 shows the desulfurization of heavy-distillates, RO-6, and IFO using
potassium superoxide with a fixed “sulfur to oxidant” mole ratio and different ionic
liquids. The reaction took place in 3 hours mechanical mixing.

Further, a series of experiments using 30% wt. H
2
O
2
were also conducted to
desulfurize the two heavy-distillates. It was observed, however, that the phases
could not be easily separated after treatment by the modified UAOD process due to
heavy precipitation and conglomeration. It was thus decided that the modified
UAOD process using 30% wt. H
2
O
2
is not applicable to desulfurize these heavy-
distillates.

Table 4.8 Desulfurization of different heavy-distillates using KO
2
as oxidant

Ionic Liquid KO
2
[S]
Initial
[S] : [KO
2
] [S]
Final

[S]
Removal

g mmol ppm
w
mmol mol : mol ppm
w
%
[BMIM][PF
6
] 0.7 9.9 12187 3.8 1 : 2.6 6037 51
[EMIM][EtSO
4
] 0.7 9.9 12187 3.8 1 : 2.6 5578 54
RO-6
[TMIM][MeSO
4
] 0.7 9.9 12187 3.8 1 : 2.6 5711 53
[BMIM][PF
6
] 1.6 22.1 27223 8.5 1 : 2.6 6479 76
[EMIM][EtSO
4
] 1.6 22.1 27223 8.5 1 : 2.6 5710 78
IFO
[TMIM][MeSO
4
] 1.6 22.1 27223 8.5 1 : 2.6 5358 80

164
4.5 Summary and Conclusion

This study aimed as developing alternative oxidation processes to achieve high
desulfurization efficiencies on various commercial fuels including diesels and residual
oils with relatively low consumption of oxidant.

Use of solid type oxidant can significantly reduce the volume and weight of required
oxidant comparing with systems using 30% wt. hydrogen peroxide solution as oxidant. In
low oxidant to sulfur ratio, potassium superoxide showed significantly more effective
comparing with 30% wt. hydrogen peroxide.

As illustrated in Table 4.3, higher desulfurization efficiencies could be observed in
systems using [EMIM][EtSO
4
] and [TMIM][MeSO
4
] instead of [BMIM][PF
6
]. Therefore,
[EMIM][EtSO
4
] and [TMIM][MeSO
4
] are considered as better ionic liquid for this
system.

It was demonstrated that high desulfurization efficiencies on commercial diesels
including JP-8, MGO and sour diesel could be achieved. Greater than 98%
desulfurization efficiencies were observed for JP-8 and MGO. 95% desulfurization was
achieved on sour diesel using potassium superoxide as oxidant. This is possibly due to
the high initial sulfur content of sour diesel. With lower oxidant to sulfur ratio, the
desulfurization efficiency on sour diesel was relatively low.

165

The pseudo first order rate constant for JP-8, MGO and sour diesel obtained from the
kinetic studies are 0.0231 min
-1
, 0.0212 min
-1
and 0.0151 min
-1
, respectively. Lower
reaction rate for sour diesel was possibly due to low oxidant to sulfur ratio applied.

In this study, it was also demonstrated that 50% to 75% desulfurization efficiencies could
be achieved with heavy-distillates by using potassium superoxide as oxidant. The process
could yield residual oil with sulfur content of less than 1%, satisfying the 2010 sulfur cap
of 1% regulated by MARPOL. Nonetheless, in order to meet the 5000ppm
w
surfur cap
goal in 2020, further desulfurization of the heavy distillates is still required.

Desulfurization efficiencies of RO-6 and IFO were relatively low compared to the diesel
fuels and the model compound including T, 2MT, BT, 2MBT, DBT and 4,6DMDBT. This
could be caused by the complex structure of RO-6 and IFO hindering the desulfurization
efficiency. Furthermore, there could be other refractory sulfur compounds which are not
susceptible to oxidation or difficult to be extracted.

166
CHAPTER 5: QUALITATIVE ANALYSIS USING GC-SCD
5.1 Introduction

As previously discussed, it is widely accepted that organic sulfur compounds in
petroleum products can cause serious pollution to the environment. Upon combustion,
OSCs would be transformed to sulfur oxides, which are known to be the key contaminant
causing acid rain and health hazards. Some of the polycyclic aromatic sulfur compounds
could also be possible mutagens or carcinogens. Thus, desulfurization technologies are
expected to significantly lower the sulfur concentration in various petroleum products.

In order to investigate the reactivity and understand the oxidation pathway of organic
sulfur compounds, identification of OSCs in petroleum products is of great importance in
the industry. Also, understanding the distribution of OSCs and identifying the refractory
compounds are highly desirable for the development of desulfurization technologies.

Gas chromatography (GC) equipped with appropriate detectors is one of the most
common methods to measure and characterize volatile sulfur, nitrogen and phosphorus
compounds. There are several detectors which can be used to detect sulfur containing
compounds, including Atomic Emission Detector (AED), Hall Electroconductivity
Detector (HECD), Flame Photometric Detector (FPD), Pulse Flame Photometric Detector
(PFPD), Flame Ionization Detector (FID) and Mass Spectroscopy Detector (MS). These
detectors are applicable to identify a wide range of elements and compounds (Lee,

167
1991; Mistry et al., 1994; Ellzy et al., 1998). However, most of these detectors are not
able to provide accurate results or a linear response to the concentration of sulfur
containing compounds (Andari et al., 1996). Although AED shows a higher linear
dynamic range comparing with other detectors, it is much more expensive and difficult to
operate. Besides, its selectivity towards different elements is questionable (Mistry et al.,
1994).

Sulfur Chemiluminescence Detector (SCD) is a more recent detector which can serve
better for both qualitative and quantitative analyses of organic sulfur compounds. SCD is
based on a two-step process (Benner, 1989). Sulfur containing compounds are first
reacted in a high temperature furnace (> 800°C) with hydrogen rich hydrogen/air mixture
forming sulfur monoxide. Sulfur monoxide formed is transferred to a reaction chamber
where sulfur monoxide is oxidized by ozone forming sulfur dioxide (SO
2
*) in an excited
state. Upon decay, the exited state of SO
2
emits UV radiation with a peak wavelength of
350 nm. The chemiluminescent emission is detected by a photomultiplier tube. Because
the intensity of light emitted is proportional to the concentration of sulfur, SCD can be
used to measure sulfur both quantitatively and qualitatively. A block diagram of GC-SCD
with the major chemical reactions involved is illustrated in Figure 5.1.

168

Figure 5.1 Major reactions in SCD (GAS Inc., 2004)

Sulfur chemiluminescence detector provides higher sensitivity for any organic sulfur
compounds with the lower detection limit in ppb level. It provides linear response with
respect to concentration, which can be used to measure sulfur concentration. While SCD
is specified to detect sulfur only, there is no interference by co-eluting species of other
hydrocarbons and thus, providing a better chromatograph to identify sulfur compounds.
High stability, selectivity and absence of quenching are also some of the advantage of
SCD. GC-SCD is one of the standard test methods to characterize organic sulfur
compounds approved by ASTM as stated in ASTM D 5504, ASTM D 5623, and ASTM
D 7011.

169

Figure 5.2 Some of the standard sulfur measurement methods adopted by ASTM (GAS
Inc., 2004)

In order to investigate the oxidation pathway of organic sulfur compounds in the new
system, a series of experiments were conducted using the modified desulfurization
process with KO
2
as oxidant on different samples, including a premixed solution of

170
model sulfur compounds including benzothiophene, dibenzothiophene, and 4, 6-dimethyl
dibenzothiophene; a solution of benzothiophene sulfone; a solution of dibenzothiophene
sulfone; and also some of the commercial fuels including JP-8, MGO, sour diesel and
also residual fuel oils.

During the experiment, it was found that the intensity of model sulfur compounds were
significantly reduced after the oxidation process even if the total sulfur concentration did
not change. A test was done to examine if there was inorganic sulfur compounds, for
instance, sulfate in the sample after oxidation. Based on the results, it is believed that
sulfate could be one of the products. While sulfate is not detectable by GC-SCD and
could deposit on the GC column, samples after oxidation process were washed with water
to lower down sulfate and other ions concentrations.

In this chapter, qualitative analyses on the sulfur species using HP-6890 Series gas
chromatograph equipped with Sievers model 355B Sulfur Chemiluminescence Detector
purchased from Agilent Technologies Inc., Santa Clara, California are discussed. Also,
the oxidation mechanism of organic sulfur compounds using potassium superoxide is
proposed.

171

5.2 Materials

Model sulfur compounds used in this study including benzothiophene, benzothiophene
sulfone (BTO), dibenzothiophene, dibenzothiophene sulfone (DBTO), 4, 6-dimethyl
dibenzothiophene were obtained from Sigma-Aldrich Co. Solvents for model sulfur
compounds solutions preparation (toluene and n-decane) and solvent used for sulfones
extraction (acetonitrile) were obtained from VWR Inc. Oxidant potassium superoxide,
acid catalysts including glacial acetic acid and trifluoroacetic acid, and phase transfer
catalyst tetraoctylammonium bromide were obtained from Sigma-Aldrich Co., while
tetraoctylammonium fluoride was synthesized by halogen exchange process (Dermeik et
al., 1989). Ionic liquid 1-butyl-3-methylimidazolium hexafluoro-phosphate [BMIM][PF
6
],
was obtained from Sigma-Aldrich Co.

Three diesel fuels, including JP-8 with 780 ppm
w
sulfur content received from Army
Research Laboratory; MGO with sulfur 1600 ppm
w
content received from Navy Station;
and sour diesel with 8100 ppm
w
sulfur content received from Golden Eagle Oil Refinery
Inc., were used as the mid-distillate feedstock. Two residual oil samples, RO-6 received
from Eco Energy Solutions Inc.; and IFO received from Intelligent Energy Inc., were
used as the heavy-distillate feedstock.

172

5.3 Experimental Procedure
5.3.1 Preparation of Model Sulfur Compound Solutions

Sample 1: Known amount of BT, DBT and 4, 6 DMDBT model sulfur compounds
were dissolved into a solvent mixture with 30% wt. toluene and 70% wt. n-decane
to make a stock solution with approximately 300 ppm
w
BT, 300 ppm
w
DBT, and
400 ppm
w
4, 6DMDBT, giving a total of 1000 ppm
w
sulfur content.

Sample 2: Known amount of BTO model sulfur compound was dissolved into a
solvent mixture with 30% wt. toluene and 70% wt. n-decane to make a stock
solution with approximately 1000 ppm
w
sulfur content.

Sample 3: Known amount of DBTO model sulfur compound was dissolved into a
solvent mixture with 30% wt. toluene and 70% wt. n-decane to make a stock
solution with approximately 1000 ppm
w
sulfur content.

173
5.3.2 Desulfurization of Feedstock

To avoid potential contamination of sulfur compounds from other reactants, sulfur
containing ionic liquids including 1-ethyl-3-methyl-imidazolium ethylsulfate
[EMIM][EtSO
4
] and 1,2,3-trimethylimidazolium methyl-sulfate [TMIM][MeSO
4
]
were not used in this study. Instead, 1-butyl-3-methylimidazolium hexafluoro-
phosphate [BMIM][PF
6
] was employed. It is known that ionic liquid can extract a
portion of sulfone from oil samples. To minimize the extraction effect of sulfur
compounds by ionic liquid, the dosage of ionic liquid is reduced to 1 gram without
significant loss of desulfurization efficiency with a 180 minutes reaction time.

Based on the optimal condition discussed in Chapter 3, Section 3.4, 0.7 gram of
potassium superoxide was first mixed with 1 gram of selected ionic liquid
[BMIM][PF
6
] and 0.1 gram of tetraoctyl-ammonium fluoride (TOAF) to produce
mixture A; 10 grams of the selected model sulfur compounds solution as described
in Chapter 5, Section 5.3.1, or selected fuel sample such as JP-8, MGO, sour diesel,
RO-6 or IFO, were mixed with 2 grams of acetic acid to produce mixture B.
Mixture A was then slowly added to mixture B, and the resulting mixture was
heated up to 70°C and stirred for 180 minutes. Emulsion formed after the treatment
process was separated by centrifugation. Oil phase was collected and extracted with
acetonitrile. Total sulfur content of acetonitrile-extracted oil phase was analyzed by
the Sulfur-in-Oil Analyzer.

174
5.3.3 Analytical Method

The total sulfur concentration of the fuel samples was determined according to
ASTM D4294 and ISO 8754. Sulfur-in-Oil Analyzer (SLFA-20), manufactured by
Horiba Inc. was used to measure total sulfur content of samples based on non-
dispersive X-ray fluorescence. Sulfur-in-Oil Analyzer is applicable for samples
with total sulfur as high as 5% wt., with a detection limit of 20ppm
w
.

Identification of organic sulfur compounds was carried out by HP-6890 Series GC
equipped with Sievers 355B SCD (GC-SCD). The GC capillary column was
equipped with a 30m × 0.32mm HP-5 fused silica column with 0.25µm film
thickness.

To separate different compounds, a specified column temperature profile based on
Andari et al. (1996) was applied. The column temperature started at 35°C and was
gradually increased to 70°C at a rate of 10°C per minute. Then, the temperature was
increased to 280°C at a rate of 3°C per minute. The column temperature was kept at
280°C for another 2.5 minutes to vaporize the remaining hydrocarbon. Ultra high
purity helium (carrier gas) was supplied at a constant pressure of 60 psig. Ultra high
purity hydrogen (fuel) was supplied at constant pressures of 40 psig.

175
The SCD furnace temperature was set at 800°C with air flow of 5.8mL per minute
and hydrogen flow of 100mL per min. Ultra high purity oxygen was supplied to the
ozone generator with flow rate of 60 mL/min with a constant pressure of 60 psig at
25°C. The injected volume of sample was 1 µL with appropriate dilution to 50 ppm
w

to 100 ppm
w
total sulfur concentrations.

Figure 5.3 Block diagram of GC-SCD setup

176
5.3.4 Estimation of Retention Time of Different OSCs

An initial test was conducted to illustrate the retention time of three model
compounds namely BT, DBT and 4, 6 DMDBT. Although similar operating
conditions were used as reported by Andari et al. (1996), the retention times of the
model sulfur compounds reported in this study (Table 5.2) were of significantly
different. Andari et al. used an earlier GC model HP 5890 series II using a methyl
silicon fused capillary column, while this study used a HP-6890 Series GC with
HP-5 fused silica capillary column. Both columns are very similar except for their
polarities, which could explain the difference in retention time.

Based on Cheng et al. (2005), it would be reasonable to assume that the retention
times for different columns with similar properties follow a linear regression. A plot
of experimental retention time from this study versus a retention time reported by
Andari et al for the three model sulfur compounds is shown in Figure 5.4. The
estimated retention time for several organic sulfur compounds are presented in
Tables 5.1 and 5.2.

177

Figure 5.4 Experimental retention time versus reference retention time of model sulfur
compounds BT, DBT and 4, 6 DMDBT

Table 5.1 Calculated retention time of various BT and DBT derivatives
Retention Time, min

Andari et al., 1996 Estimated
C2 benzothiophenes 37.68 – 39.45 16.27 – 17.70
C3 benzothiophenes 41.11 – 43.45 19.05 – 20.94
C4 benzothiophenes 43.83 – 48.73 21.25 – 25.22
C5 benzothiophenes 49.25 – 51.33 25.64 – 27.32
C6 benzothiophenes 51.79 – 55.80 27.69 – 30.94
C7 benzothiophenes 56.31 – 57.78 31.35 – 32.54
C8 benzothiophenes 58.18 – 59.87 32.86 – 34.23
C1 di-benzothiophene 56.09 – 57.52 31.17 – 32.33
C2 di-benzothiophene 59.52 – 61.64 33.73 – 35.67
C3 di-benzothiophene 62.17 – 64.95 36.09 – 38.34
C4 di-benzothiophene 65.47 – 68.27 38.77 – 41.03
C5 di-benzothiophene 69.44 – 69.78 41.98 – 42.25
Retention Time, min (Andari et al., 1996)
Retention Time, min (from this study)
R
2
= 0.997

178
Table 5.2 Calculated retention time of various OSCs
Retention Time, min

Andari et al., 1996 From this study Estimated
thiophene 5.52 1.12 1.06
2-methyl thiophene 9.06 2.20 2.26
2-ethyl thiophene 22.29 3.37
2, 5-dimethyl thiophene 23.47 3.47
3-tert-butyl thiophene 25.75 3.60
2, 3, 4-tri-methyl thiophene 25.99 3.73
1, 3-dithiacyclopentane 26.73 3.89
1-heptane thiol 26.99 4.05
1, 4-butane dithiol 28.58 4.24
2-methyl 5-propyl thiophene 30.28 4.48
2-n-butyl thiophene 31.31 4.57
2, 5-diethyl thiophene 33.85 4.66
3-n-butyl thiophene 34.16 4.75
3, 4-diethyl thiophene 35.16 4.96
2, 3, 4, 5 tetra-methyl thiophene 39.45 5.29
n-butyl sulfide 45.68 5.96
ethyl phenyl thiophene 25.75 6.62
1, 5-pentane dithiol 25.99 6.81
1-octane thiol 26.73 7.41
benzyl methyl sulfide 26.99 7.62
benzothiophene 28.58 8.89 8.91
1-nonane thiol 30.28 10.28
1, 6-hexane dithiol 31.31 11.12
5-methyl benzothiophene 33.85 13.17
3-methyl benzothiophene 34.16 13.42
1-decane thiol 35.16 14.23
3, 5-dimethyl benzothiophene 39.45 17.70
di-phenyl sulfide 45.68 22.75
benzyl phenyl sulfide 50.11 26.33
di-benzothiophene 52.25 28.08 28.06
di-benzyl sulfide 54.54 29.92
4, 6-dimethyl dibenzothiophene 59.61 34.01 34.02
thianthrene 59.72 34.11
2, 8-dimethyl dibenzothiophene 61.1 35.23

179
5.4 Results and Discussion
5.4.1 Model Sulfur Compounds Identification

Figures 5.5 and 5.6 depict the GC-SCD chromatograms of BT, DBT and 4, 6
DMDBT before and after the oxidative desulfurization treatment using KO
2
as
oxidant. It can be observed that the retention times for the oxidized products of BT,
DBT and 4,6DMDBT identified as BTO, DBTO, and 4, 6 DMDBTO, respectively,
shifted to the right.

Nevertheless, the intensity of the peaks are significantly reduced even the total
sulfur concentration measured by Sulfur-in-Oil Analyzer did not show significant
difference. One of the hypotheses is that superoxide might be able to oxidize
organic sulfur compounds to inorganic sulfate.

Due to the fact that GC-SCD is not capable of identifying sulfate ion, experiments
were conducted to examine if there was inorganic sulfate present in the sample after
oxidation. The sample was extracted with deionized water (DW) to possibly
partition sulfate ion. Addition of 0.5 gram barium chloride and precipitation of
barium sulfate indicated the presence of sulfate ion. Measurement of total sulfur
concentration before (1037 ppm
w
) and after water extraction (568 ppm
w
) indicate
that nearly 54% of total sulfur was extracted.

180

Figure 5.5 GC-SCD chromatogram of sample with BT, DBT and 4, 6 DMDBT before
oxidation process

Figure 5.6 GC-SCD chromatogram of sample with BT, DBT and 4, 6 DMDBT after
oxidation process

To determine the possible oxidation of sulfones, oxidation studies were conducted
for BTO and DBTO separately. The chromatograms BTO and DBTO before and
after the oxidation process are shown in Figure 5.7 through Figure 5.10.
BT DBT
4, 6 DMDBT
min
uV
BTO
DBTO
4, 6 DMDBTO
min
uV

181

Figure 5.7 GC-SCD chromatogram of sample with BTO before oxidation process

Figure 5.8 GC-SCD chromatogram of sample with BTO after oxidation process

Figure 5.9 GC-SCD chromatogram of sample with DBTO before oxidation process

BTO
min
uV
BTO
DBTO
min
uV
min
DBTO
uV

182

Figure 5.10 GC-SCD chromatogram of sample with DBTO after oxidation process

Figure 5.8 shows that the BTO peak was significantly reduced after oxidation.
Furthermore, the chromatogram shows the presence of other organic sulfur
compounds. The major peak was identified as DBTO. Some other smaller peaks
could be alkyl benzothiophenes, alkyl dibenzothiophenes, or other higher molecular
weight OSCs.

It was also observed that the peak intensity of DBTO after oxidation (Figure 5.10)
is much lower compared to that before oxidation (Figure 5.9). However, the total
sulfur content measured by the Sulfur-in-Oil Analyser did not show significant
change. Therefore, each sample was extracted with water and precipitated with
barium chloride. The total sulfur concentration of each sample was measure before
and after water extraction. The results indicate that the total sulfur concentration for
BTO was reduced from 983 ppm
w
to 129 ppm
w
(87%) and for DBTO from 1080
ppm
w
to 393 ppm
w
(64%). It can thus be postulated that the oxidation process bas
relatively less impact on DBTO than BTO.
min
DBTO
uV

183
As suggested by Oae et al. (1981), Chang et al. (1989), Afanas’ev (1989) and Kim
et al. (1990), superoxides could oxidize organic sulfur compounds to sulfonic acids,
amides, and also inorganic sulfate. Figure 5.11 illustrates some of the chemical
reactions between OSCs and superoxide suggested by Afanas’ev (1989).

In either the UAOD system or the modified UAOD system, OSCs are
stoichiometrically converted to sulfones as the final oxidized products. However, in
this study, 13% of BT and 36% of DBT were converted to correcponding sulfones,
and the rest (87% and 64%, respectively) were converted to sulfate which can be
extracted by DW.

Figure 5.11 Chemical reactions between OSCs and superoxide anion (Afanas’ev, 1989)

184
5.4.2 Characterization of Untreated Diesel Samples

As discussed in Chapter 5.3.4 and Chapter 5.4.1, some of the model sulfur
compounds and their retention times were identified by GC-SCD analysis or
estimated through linear regression. The calculated retention time listed on Table
5.1 and Table 5.2 were used to predict the possible species of unknown organic
sulfur compounds presented in the samples. This is especially important to identify
the refractory compounds in the oxidative desulfurization process with the
application of superoxide. Those compounds are some of the most common OSCs
found in diesel and other petroleum fuel oil. By identifying the retention times of
different OSCs in the fuel samples through GC-SCD analysis, it is possible to
characterize the sulfur content in the samples.

Figure 5.12 shows the GC-SCD chromatogram of various model sulfur compounds
including thiophene, 2-methyl thiophene, benzothiophene, benzothiophene sulfone,
dibenzothiophene, 4, 6-dimethyl dibenzothiophene and dibenzothiophene sulfone.
From the chromatograph, the retention times of different OSCs are listed as follow:
T at 1.06 minutes, 2MT at 2.26 minutes, BT at 8.89 minutes, BTO at 23.31 minutes,
DBT at 28.08 minutes, 4,6-DMDBT at 34.01 minutes, and DBTO at 38.92 minutes.

185

Figure 5.12 GC-SCD chromatogram of various model sulfur compounds

The OSCs compositions of different diesel fuels, including JP-8, MGO and sour
diesel were separated, as illustrated in the GC-SCD chromatograms (Figure 5.13).
As discussed in Chapter 1, the OSCs in diesel samples are primarily comprised of
cyclic sulfide derivates, thiophene derivates, benzothiophene derivates and
dibenzothiophene derivates. The OSCs constituents shown in these chromatograms
may be identified with reference to Tables 5.1 and 5.2.

min

186

Figure 5.13 GC-SCD chromatogram of various diesel samples

It can be observed in Figure 5.13, JP-8 contains mainly BT and its derivates,
together with a small fraction of sulfides, thiols and thiophenes; MGO contains
mainly DBTs, some BTs, and a small fraction of sulfides, thiols and thiophenes;
and sour diesel contains mainly DBTs and some BTs.
Sulfides,
Thiols,
Thiophenes Benzothiophenes Dibenzothiophenes
Sulfides,
Thiols,
Thiophenes Benzothiophenes Dibenzothiophenes
Sulfides,
Thiols,
Thiophenes Benzothiophenes Dibenzothiophenes
JP-8
[S] = 782 ppm
w

Dilution × 10
MGO
[S] = 1631 ppm
w

Dilution × 20
Sour Diesel
[S] = 8117 ppm
w

Dilution × 100
min
min
min
uV
uV
uV

187

5.4.3 Characterization of Desulfurized Diesel Samples

Oxidative desulfurization with the application of potassium superoxide was applied
to each of the diesel sample including JP-8, MGO and sour diesel as stated in
Chapter 5.3.2. The sulfur to oxidant ratios applied for JP-8, MGO, and sour diesel
were 1:29, 1:14 and 1:2.7 respectively.

To prevent damaging the GC-SCD by acids, salts and other ions, the samples were
washed with DW after oxidation, so as to remove ions and other water soluble
compounds from the sample. The samples were then injected into the GC-SCD to
determine the change in sulfur species composition after oxidation. The oxidized
diesel samples were extracted with acetonitrile to remove sulfone formed. The
extracted diesel samples were injected into the GC-SCD again to show the
remaining sulfur species.

The GC-SCD chromatograms for JP-8, MGO and sour diesel at different stages are
illustrated in Figures 5.14, 5.15 and 5.16, respectively. Total sulfur contents in each
stage were measured by the Sulfur-in-Oil Analyser, as recorded in Table 5.3.

188
Table 5.3 Total sulfur content for various diesel samples in different stages
Original After Oxidation
After Washing
(with DW)
After Extraction
(with acetonitrile)

[S], ppm
w
[S], ppm
w
∆[S], %
*
[S], ppm
w
∆[S], %
*
[S], ppm
w
∆[S], %
*

JP-8 782 724 7.4 358 54.2 < 20 > 98
MGO 1631 1547 5.1 1252 23.2 28 98.3
Sour
Diesel
8117 7869 3.1 7166 6.8 554 93.2
* ∆[S] represents percentage change in sulfur content

As can be seen in Table 5.3, the total sulfur content in all three diesel samples were
significantly reduced by washing with DW after oxidation. This indicates the
formation of water soluble sulfur compounds after oxidation. Among the three
diesel fuels, JP-8 exhibited the highest removal of sulfur content, while sour diesel
showed the lowest. There are two possible reasons for this result. The applied
oxidant to sulfur ratio for JP-8 was the highest, thus further oxidation would have
been possible. Furthermore, it was demonstrated in the model sulfur compound
studies that higher percentage of BT can be oxidized to sulfone as compared with
DBT under identical treatment conditions. The sulfur contents in JP-8 are mostly
BT derivatives, while in sour diesel there are mostly DBT derivatives and thus more
water soluble sulfur compounds (sulfate) could be formed from JP-8. Thus, higher
degree of sulfur removal can be achieved by washing the oxidized JP-8.

189
As illustrated in the GC-SCD chromatogram (Figure 5.14-a), the major OSCs in
original JP-8 are benzothiophene derivates (C1BT to C4BT) such as 3-methyl
benzothiophene and 3, 5-dimethyl benzothiophene, together with sulfides, thiols,
and thiophenes derivates. As observed in Figure 5.14-b, the chromatogram peaks
for oxidized JP-8 are shifted to the right showing that OSCs are oxidized to
molecules with longer retention time, which are expected to be the corresponding
sulfones of the original OSCs.

Two major peaks can be seen in the chromatogram of extracted JP-8. They are
identified as DBT and 4, 6DMDBT. In the chromatogram of untreated JP-8 (Figure
5.14-a), these two peaks were practically non-existent, however, it seems that these
two compounds are formed during the oxidation process (Figure 5.14-b), and are
not removed during the extraction process.

In our studies, ultralow sulfur diesel can be produced from JP-8 by a single
desulfurization process with the application of superoxide. A further polished
product, theoretically “zero sulfur” diesel, could possibly be achieved by a post
treatment. Due to the fact that the amount of ionic liquid applied is less than the
optimal condition (Table 3.17), the desulfurization efficiency could be further
improved by simply increasing the dosage of ionic liquid.

190

Figure 5.14 GC-SCD chromatogram of JP-8 at different stages of the process
3, 5DMBT
BT
di-phenyl sulfide
DBT
C2BT
C3BT
C4BT
C5BT
C6BT
C1BT
sulfides,
thiols,
thiophenes
DBT
4, 6DMDBT
JP8 (Original)
[S] = 782 ppm
w
Dilution × 10
JP8 (Oxidized)
[S] = 358 ppm
w
Dilution × 3
JP8 (Extracted)
[S] < 20 ppm
w
No Dilution
3MBT
BTO DBTO
3, 5DMBT
4, 6DMDBT
4, 6DMDBT
DBT
min
min
min
(a)
(b)
(c)
uV
uV
uV

191

Figure 5.15 GC-SCD chromatogram of MGO at different stages of the process
T
di-phenyl sulfide
DBT
C2BT
C3BT
C4BT
C5BT
C6BT
C1BT
sulfides,
thiols,
thiophenes
DBT 4,6DMDBT
2MT
BT
5MBT
3,5DMBT 4,6DMDBT
2,8DMDBT
C7BT,
C1DBT
C8BT,
C2DBT
C3DBT
C4DBT
C5DBT
BTO DBTO
MGO (Original)
[S] = 1631 ppm
w
Dilution × 16
MGO (Oxidized)
[S] = 1252 ppm
w
Dilution × 12
MGO (Extracted)
[S] = 28 ppm
w
No dilution
4,6DMDBTO
min
min
min
(a)
(b)
(c)
uV
uV
uV

192
MGO contains a variety of sulfur species, from thiophenes to dibenzothiophene
derivates, as illustrated in its GC-SCD chromatogram shown in Figure 5.15-a. The
majority of OSCs in original MGO are dibenzothiophene derivates (C1DBT to
C5DBT), together with benzothiophene derivates (C1BT to C8BT).

The peaks of the chromatogram for oxidized MGO (Figure 5.15-b) are shifted to the
right showing that OSCs are oxidized to molecules with longer retention time, and
are expected to be the corresponding sulfones of the original OSCs. However, the
peaks’ intensities of the smaller sulfones, such as thiophene sulfone and
benzothiophene sulfones, are much lower than expected. This observation concurs
with the chromatograms in the model sulfur compound studies and the JP-8 diesel
study shown in Figure 5.6 and Figure 5.14-b, respectively, can be hypothesized that
smaller OSCs such as thiophenes and benzothiophenes can be oxidized to inorganic
sulfur compounds such as sulfate ion, which can simply be removed by washing
with water.

It is observed that the sulfur species remaining the extracted MGO (Figure 5.15-c)
are mainly substituted DBT derivates. The total sulfur content in the extracted
MGO is 28 ppm
w
(Figure 5.15-c). Although this amount is higher than the ULSD
requirement, the desulfurization efficiency can be improved by increasing ionic
liquid dosage, as demonstrated in Chapter 3, Section 3.3.4, so as to achieve the
ULSD standard.

193

Figure 5.16 GC-SCD chromatogram of sour diesel at different stages of the process
T
BT
3MBT
3,5DMBT
DBT
4,6DMDBT
2,8DMDBT
C2BT
C3BT
C4BT
C5BT
C6BT
C1BT
sulfides,
thiols,
thiophenes
C7BT,
C1DBT
C8BT,
C2DBT
C3DBT
C4DBT
C5DBT
Sour Diesel (Original)
[S] = 8117 ppm
w
Dilution × 80
di-phenyl sulfide
BTO DBTO
Sour Diesel
(Oxidized)
[S] = 7166 ppm
w
Dilution × 70
4,6DMDBT
Sour Diesel (Extracted)
[S] = 554 ppm
w
Dilution × 5

4,6DMDBTO
(a)
(b)
(c)
min
min
min
uV
uV
uV

194
Similar to MGO, sour diesel contains a variety of sulfur species as shown in Figure
5.16-a. Sour diesel contains mainly dibenzothiophene derivates (C1DBT to
C5DBT), together with highly substituted benzothiophene derivates (C3BT to
C8BT). Only trace amount of straight chain sulfides, thiols or thiophenes can be
found in the sour diesel sample.

For the oxidized sour diesel, peaks of its chromatogram shifted to the right (Figure
5.16-b). It shows that OSCs are oxidized to the corresponding sulfones. The
chromatogram peaks’ intensities of sulfones in the oxidized sample are similar to
those of the original sulfur compounds in the untreated sample. As listed in Table
5.3, only 7% of the sulfur content in the oxidized sour diesel can be removed by
washing with DW. This is possibly due to the fact that the oxidant to sulfur ratio
applied was too low and as a result further oxidation of sulfones yielded
insignificant amount of sulfate.

The final sulfur content in the primary treated sour diesel was 554 ppm
w
. As
observed in the chromatogram (Figure 5.17-a), the remaining sulfur contents are
presumably highly substituted DBT and other larger sulfur compounds. A secondary
treatment with same the procedure (oxidation followed by extraction) was applied
to the treated sour diesel to remove the remaining sulfur compounds. The
chromatograms of sour diesel at different stages of the secondary treatment are
illustrated in Figures 5.17-a to 5.17-c.

195

Figure 5.17 GC-SCD chromatogram of treated sour diesel at different stages of the
process
4,6DMDBT
C2BT
C3BT
C4BT
C5BT
C6BT
C1BT
sulfides,
thiols,
thiophenes
C7BT,
C1DBT
C8BT,
C2DBT
C3DBT
C4DBT
C5DBT
Sour Diesel (Exacted)
[S] = 554 ppm
w
Dilution × 5
T
BT
3MBT
3,5DMBT
DBT
2,8DMDBT
Sour Diesel (reoxidized)
[S] = 462 ppm
w
Dilution × 5
BTO DBTO
4,6DMDBTO
4,6DMDBT T BT DBT
Sour Diesel (re-extracted)
[S] < 20 ppm
w
No dilution
(a)
(b)
(c)
min
min
min
uV
uV
uV

196
As illustrated in Figures 5.17-a and 5.17-b, the chromatogram peaks shift to the
right after the reoxidation process. It shows that the highly substituted DBTs
remaining in the primary treated sour diesel can be oxidized to the corresponding
sulfones in the secondary treatment process as shown in Figure 5.17-c. The sulfur
content of the secondary treated sour diesel can be lowered to less than 20 ppm
w
.
Thus, two consecutive treatments of the oxidative desulfurization process with the
application of KO
2
as oxidant can result end product which would be in compliance
with ULSD.

From the chromatogram in Figure 5.17-c, it can be seen that there are several OSCs
remaining the secondary treated sour diesel. These OSCs are possibly highly
substituted DBTs (C2DBT or above). Further treatment may be required to produce
“zero sulfur” diesel.

The chromatogram peaks’ intensities of sulfones in the oxidized sour diesel sample
(Figure 5.16-b) are similar to those of the original sample (Figure 5.16-a). As listed
in Table 5.3, only 6.8% of the sulfur content in the oxidized sour diesel can be
removed by washing with DW. This is possibly because of low oxidant to sulfur
ratio which caused further oxidation of sulfones to sulfate insignificant. Besides,
further oxidation of highly substituted highly substituted OSCs may prove difficult
due to their steric hindrance effect (Kimintarachat et al., 2006).

197
5.4.4 Characterization of Heavy-Distillates

Generally speaking, the major sulfur species in heavy-distillate are alkyl
benzothiophene derivates, dibenzothiophene derivatives, benzonaphtho-thiophene
derivatives and phenanthro-thiophene derivatives. Having GC-SCD chromatograms,
some of the OSCs in the two heavy distillate samples, RO-6 and IFO, could be
identified with reference to Table 5.1 and Table 5.2.

Figure 5.18 GC-SCD chromatograms of two untreated heavy distillates

Sulfides,
Thiols,
Thiophenes Benzothiophenes Dibenzothiophenes
RO-6
[S] = 12187 ppm
w

Dilution × 120
IFO
[S] = 21233 ppm
w

Dilution × 270
min
min
uV
uV

198
According to the chromatograms of the untreated RO-6 and IFO illustrated in
Figure 5.18, significant portion of the peaks having larger retention time than Ts,
BTs and DBTs. Having longer retention time, these peaks are possibly highly
substituted DBTs (C5DBT) or sulfur containing polycyclic aromatic hydrocarbons
(PAHs) such as benzonaphtho-thiophene derivatives and phenanthro-thiophene
derivatives as demonstrated in a microcoulometric sulfur detector (MCD)
chromatogram shown in Figure 5.19.

Figure 5.19 Typical GC-MCD chromatogram of vacuum gas oil (Drushel, 1969)

The oxidative desulfurization process with the application of KO
2
was applied to
RO-6 and IFO with sulfur to oxidant ratios at 1:2.6. The GC-SCD chromatograms
for RO-6 and IFO are illustrated in Figure 5.20 and Figure 5.21, respectively.

199

Figure 5.20 GC-SCD chromatogram of treated RO-6 at different stages of the process
sulfides,
thiols,
thiophenes
T
DBT
C2BT
C3BT
C4BT
C5BT
C6BT
C1BT
BT
3,5DMBT
4,6DMDBT
2,8DMDBT
C7BT,
C1DBT
C8BT,
C2DBT
C3DBT
C4DBT
C5DBT
3MBT
RO-6 (Original)
[S] = 12187 ppm
w
Dilution × 120
BTO DBTO
4,6DMDBTO
RO-6 (Oxidized)
[S] = 11732 ppm
w
Dilution × 120
DBT
4,6DMDBT
RO-6 (Extracted)
[S] = 6558 ppm
w
Dilution x 70
(a)
(b)
(c)
min
min
min
uV
uV
uV

200

Figure 5.21 GC-SCD chromatogram of treated IFO at different stages of the process
sulfides,
thiols,
thiophenes
DBT
4,6DMDBT
IFO (Extracted)
[S] = 6743 ppm
w
Dilution x 67
T
DBT
C2BT
C3BT
C4BT
C5BT
C6BT
C1BT
BT
5MBT
3,5DMBT
4,6DMDBT
2,8DMDBT
C7BT,
C1DBT
C8BT,
C2DBT
C3DBT
C4DBT
C5DBT
3MBT
IFO (Original)
[S] = 27233 ppm
w
Dilution × 270
BTO DBTO 4,6DMDBTO
IFO (Oxidized)
[S] = 25948 ppm
w
Dilution × 260
(a)
(c)
min
min
min
(b)
uV
uV
uV

201

As illustrated in Figure 5.20-a, the major OSCs in untreated RO-6 are highly
substituted DBTs (C6DBT) or sulfur containing PAHs. It also contains smaller
sulfur compounds including Ts, BTs, and DBTs. As can be seen, upon oxidation
(Figure 5.20-b) and extraction (Figure 5.20-c), most of the peaks for the smaller
OSCs are removed. However, majority of the larger OSCs remain in the sample as
demonstrated in the chromatograms. Some of these compounds could be refractory
which cannot be oxidized or extracted.

Compared with RO-6, IFO contains a larger portion of benzothiophenes and
dibenzothiophenes derivatives. These OSCs are removed upon oxidation and
extraction. However, the remaining portions are not susceptible to oxidation.

As can be seen in Figures 5.20 and 5.21, the sulfur contents in the treated samples
are mainly larger OSCs such as C5DBT, benzonaphtho-thiophene derivatives and
phenanthro-thiophene derivatives. RO-6 appears more difficult to desulfurize due to
high content of these compounds. As discussed in Chapter 4, RO-6 can be
desulfurized by approximately 50%, while IFO can be desulfurized up to 80%.
Comparatively, sour diesel can be desulfurized by more than 95% with the same
oxidant to sulfur ratio. Besides the difference in original sulfur species in the
samples, other properties such as viscosity and presence of inhibitors in heavy
distillates could also reduce the process efficiency.

202

Asphaltene content in heavy distillates could be one of the major factors affecting
the desulfurization efficiency. Asphaltenes are considered as radical inhibitors by
terminating free radical reactions through recombination and disproportionation
(Bukowski et al., 1983). Asphaltenes can also suppress radical chain formation and
the catalytic decomposition of organic sulfur compounds by peroxides and
hydroperoxides (Herrington, 2004).

Upon oxidation, asphaltene could be converted to carboids (Yen, 1974). This could
lead to asphaltene precipitation, flocculation and agglomeration (Margaril et al.,
1971; Wattana et al, 2005). Thus, oil samples could show incompatibility towards
refinery process involving oxidation (Speight, 2004).

203
5.5 Mechanism of Inorganic Sulfate Formation

As discussed in Chapter 4, third generation of UAOD process cannot treat heavy
distillates due to heavy precipitation and agglomeration of asphaltenes during the
treatment process. By utilizing superoxide in the modified UAOD process, asphaltenes
precipitation can be greatly reduced. Formation of carboids is still observed (carboids are
toluene insoluble portion). This could be another reason for the low desulfurization
efficiency of heavy distillates.

Similar to the studies for the model compounds and the diesel samples, chromatogram
peaks of benzothiophene derivatives and some of the dibenzothiophene derivatives are
significantly reduced after oxidation. It is predicted that benzothiophene and
dibenzothiophene derivatives could possibly be further oxidized or reacted forming other
compounds.

One of the possible mechanisms for the disappearance of BT and DBT is further
oxidation of the oxidized products. A few studies demonstrate that BT and DBT can be
oxidized to the corresponding sulfones, and the sulfones can be oxidized to sulfonic acids,
sulfites and sulfates through radical reactions initiated by hydroxyl radicals or photolysis
(Andersson et al., 1992; Kim et al., 2003; Shiraishi et al., 2003). The oxidation
mechanism for BTO and DBTO are illustrated in Figure 5.22 and Figure 5.23,
respectively. By employing the superoxide radial in the modified UAOD process,

204
decomposition of BT and DBT derivatives could follow a similar radical reaction as
described above.

Figure 5.22 Possible BT destruction pathways by OH radicals (Kim et al., 2003)

Figure 5.23 Photolysis of DBTO (Shiraishi et al., 2003)

205
5.6 Summary and Conclusion

With the aid of Gas Chromatography equipped with sulfur chemiluminescence detector,
this study aimed as fingerprinting organic sulfur compounds in different fuel samples and
also the oxidation products by the modified UAOD process with the application of
superoxide. This information could provide a better understanding of the OSCs oxidation
pathway, so as to predict the mechanism of the treatment process. Using GC-SCD to
characterize sulfur species in oil sample is one of the standard test methods approved by
ASTM.

In the model sulfur compounds studies, it was demonstrated that the modified UAOD
process with superoxides can oxidize different organic sulfur compounds, including
benzothiophene, dibenzothiophene and 4, 6 dimethyldibenzothiophene, to the
corresponding sulfones. It was further demonstrated that the resulting water soluble
inorganic sulfur compounds can be extracted and removed.

Two sulfone model compounds, benzothiophene sulfone and dibenzothiophene sulfone,
were used to determine the oxidation by-product by the modified UAOD process with the
application of superoxide. In the oxidation of benzothiophene sulfone, dibenzothiophene
sulfone was recognized as a major by-product. The chromatogram shows several
unidentifiable small peaks. Similarly, investigation of oxidation by-product was
conducted on dibenzothiophene sulfone. It appears that dibenzothiophene sulfone was
partially oxidized to form inorganic sulfate.

206

Three diesel samples, namely JP-8, MGO and sour diesel, were used in the diesel fuel
studies. From the GC-SCD chromatograms, it was found out that JP-8 contains mainly
benzothiophene derivatives, while MGO and sour diesel contain mainly
dibenzothiophene derivatives. In all three cases, benzothiophene derivatives were mostly
removed by the modified UAOD process with superoxide. Less than 28 ppm
w
of
dibenzothiophene derivatives were found in desulfurized JP-8 and desulfurized MGO.
Depending upon the selection of ionic liquids and oxidant dosage, both JP-9 and MGO
were desulfurized to ultralow sulfur diesel standard.

It was demonstrated that with the application of relatively low oxidant to sulfur ratio,
sour diesel was desulfurized to 554 ppm
w
. The remaining sulfur content was mostly
dibenzothiophene derivatives. Further oxidation and extraction (secondary treatment) of
the sour diesel lowered the sulfur content to less than 20 ppm
w
. Only trace amounts of
alkyl substituted dibenzothiophene derivatives remained in the sample.

RO-6 and IFO were used as the heavy-distillate samples in this study. Results indicated
that the major components of heavy-distillates (RO-6 and IFO), mainly dibenzothiophene
derivatives (C5DBT), benzonaphtho-thiophene derivatives and phenanthro-thiophene
derivatives, were not removed in the treatment process. However, the process effectively
removed benzothiophenes and lighter dibenzothiophenes (C1DBT to C4DBT).

207
Increasing superoxide dosage could be possible to increase desulfurization efficiencies of
heavy distillates. However, high dosage of oxidant could affect the stability of asphaltene
colloidal system causing heavy deposition. Further experiments are required to examine
the optimal conditions to desulfurize heavy distillates. More sulfur markers are required
to fingerprint the unidentified compounds.

It was demonstrated that sulfones of benzothiophenes and sulfones of dibenzothiophenes
could be further oxidized to different compounds including sulfur-containing inorganic
compounds such as sulfates and sulfites. Further studies using ion chromatography can
quantitate sulfates and sulfites and other sulfur-containing ions.

In general, due to the formation of sulfonic acid, sulfites and sulfates, the sample’s sulfur
content can be removed by water extraction. This could reduce the use of chemical
solvents such as acetonitrile or adsorbent such as alumina. Further studies should be
conducted to identify the intermediate products to verify the hypothesis.

Due to the column temperature profile setting, some of the heavier fraction of OSCs
including benzonaphtho-thiophene derivatives, phenanthro-thiophene derivatives and
sulfur containing asphaltenes may not be shown in the GC-SCD chromatograms. Further
studies are needed to resolve this problem.

208
CHAPTER 6: CONCEPTUAL MODEL FOR THE MODIFIED UAOD
DESULFURIZATION PROCESS
6.1 Introduction

In the modified UAOD process with the application of superoxide, it is demonstrated that
organic sulfur compounds can be oxidized with the aid of acid catalysis and phase
transfer catalysis in a tri-phase system. Multiple oxidant species including superoxide
radical, hydroxyl radical, hydrogen peroxide, singlet oxygen and peracid are produced by
superoxide radicals with the application of acid catalysis. With both quaternary
ammonium salt and ionic liquids as phase transfer catalysts, multi-ion pairing enhances
the efficiency of active oxygen transfer to the organic phase, leading to the improvement
of overall desulfurization efficiency. Unlike the previous generations of UAOD process,
the modified UAOD process with the application of superoxide can oxidize certain OSCs
to inorganic sulfur compounds.

In this chapter, the mechanism of the modified UAOD process using superoxide is
developed based on the batch studies for various model sulfur compounds, mid-distillates
and heavy distillates discussed in the previous chapters.

209
6.2 Model Overview

The modified UAOD process using superoxide employed the following components:
solid oxidant, phase transfer catalyst, ionic liquid and acid catalyst. The process was
carried out with continuous mixing of three phases: solid phase, polar phase and non-
polar phase. The solid phase includes the solid oxidant – potassium superoxide. The polar
phase includes acid catalysts and ionic liquid. The non-polar phase (or the organic phase)
includes the organic liquid which contains organic sulfur compounds targeted to be
removed. The conceptual model for the process was developed with twelve following
steps (Figure 6.1).

Step 1: In the presence of potassium superoxide, KO
2
, and quaternary ammonium salt,
[Q
+
X
-
], with lipophilic cation and hydrophilic anion, superoxide anion is released as
[Q
+
O
2
-
.
].

Step 2: Anions of acid catalysts, including acetic acid [CH
3
COOH] and trifluoroacetic
acid [CF
3
COOH], are extracted by [Q
+
X
-
] forming [Q
+
CH
3
COO
-
] and [Q
+
CF
3
COO
-
].
The process is known as Stark’s extraction mechanism as discussed in Chapter 2.
Fluoride ion is selected as the counter ion [X
-
] in the quaternary ammonium salt due to its
low lipophilicity and stability towards strong oxidants. The lipophilicities of the anions
are in the following order: MnO
4
-
> ClO
4
-
> SCN
-
> I
-
> ClO
3
-
> NO
3
-
> Br
-
> CN
-
> Cl
-
>
HSO
4
-
> CH
3
COO
-
> F
-
> SO
4
2-
> CO
3
2-
> PO
4
3-
(Dehmlow et al., 1993).

210
Step 3: In the mixture of [Q
+
O
2
-
.
] and acid, hydrogen peroxide and other active oxygen
species are generated.

Step 4: Hydrogen peroxide can react with [Q
+
CH
3
COO
-
] and [Q
+
CF
3
COO
-
] generating
peroxyacetate: [Q
+
CH
3
COOO
-
] and trifluoroperacetate: [Q
+
CF
3
COOO
-
].

Step 5: With tetraoctylammonium ion as the lipophilic cation, the active nucleophile
anions are transferred to the non-polar phase in the form of [Q
+
O
2
-
.
], [Q
+
CH
3
COOO
-
] and
[Q
+
CF
3
COOO
-
] without losing their activities.

Step 6: The active nucleophile anions, [CH
3
COOO
-
] and [CF
3
COOO
-
] in the organic
phase oxidize the organic sulfur compounds to the corresponding sulfones.

Step 7: Superoxide radicals in the organic phase can oxidize some of the organic sulfur
compounds into a radical form of the corresponding sulfone anion, which can be further
oxidized to inorganic sulfur compounds.

Step 8: After oxidation, the reduced form of the ion pairs, [Q
+
CH
3
COO
-
] and
[Q
+
CF
3
COO
-
], return to the polar phase.

Step 9: Ionic liquid provides another mean of active anion extraction, resulting
enhancement of phase transfer catalysis. With appropriate alkyl sulfate ion [RSO
4
-
] as the
counter ion, alkyl persulfate anion [RSO
6
-
.
] can be produced by reacting with superoxide
radical.

211
Step 10: Similar to quaternary ammonium cation, alkyl methylimidazolium [RMIM
+
]
type of ionic liquid can serve as a phase transfer catalyst which can transfer [RSO
6
-
.
] to
the organic phase in the form of [RMIM
+
RSO
6
-
.
].

Step 11: [RMIM
+
RSO
6
-
.
] can oxidize organic sulfur compounds to the corresponding
sulfones or inorganic sulfate as described in Step 6 and 7.

Step 12: The reduced form of [RMIM
+
RSO
4
-
] returns to the polar phase for reactivation
in the catalysis cycle.

Figure 6.1 Conceptual model of 4
th
generation UAOD process

212
6.3 Summary

In this chapter, a conceptual model of the modified UAOD process with the application
of superoxide is developed consisting of twelve steps. They include: (1) extraction of
superoxide radical by phase transfer catalyst; (2) extraction of anion from acid catalysts;
(3) generation of active oxygen species from superoxide; (4) generation of peroxyacetic
acid; (5) transfer of active nucleophiles to organic phase; (6) oxidation of organic sulfur
compounds by peroxyacetic acid; (7) oxidation of organic sulfur compounds by
superoxide radical; (8) regeneration of catalysts in the polar phase; (9) oxidation of ionic
liquid anion to form alkyl persulfate anion; (10) transfer of alkyl persulfate anion to
organic phase; (11) oxidation of organic sulfur by alkyl persulfate anion; and (12)
regeneration of ionic liquid’s ion pair.

213
CHAPTER 7: CONCLUSION AND RECOMMENDATIONS
7.1 Summary and Conclusion

With the tightened environmental regulations on sulfur content of fuels including diesel
and residual fuel oil, an advance desulfurization method is required to produce ultralow
sulfur diesel fuel with sulfur content less than 15 ppm
w
and low sulfur residual fuel oil
with sulfur content less than 5000 ppm
w
. The 3
rd
generation UAOD process is capable of
producing ultralow sulfur diesel at low temperature and ambient pressure. However, the
requirement for high dosage of oxidant and high energy consumption diminish its
economical viability. Additionally, previous generations of UAOD could cause heavy
precipitation of asphaltene, thus are considered not applicable to desulfurize residual fuel
oil. Therefore, this research developed a 4
th
generation UAOD, called the modified
UAOD process with the application of superoxide, summarized in the ensuing paragraphs.

It was demonstrated that oxidation of more than 97% of various model compounds
including thiophene derivatives, benzothiophene derivatives, and dibenzothiophene
derivatives can be achieved by the 4
th
generation UAOD process at low oxidant to sulfur
ratio. Together with a post treatment to remove sulfones by either liquid-liquid extraction
or solid adsorption, the process can be used to produce ultralow sulfur diesel. This
process can also be used to achieve up to 80% desulfurization on residual fuel oil.

214

The 4
th
generation UAOD is designed with the following essential components: oxidation;
phase transfer catalysis; acid catalysis; and ultrasonication. Quaternary ammonium salt
and ionic liquid are used as phase transfer catalysts to extract active oxidant (superoxide)
to the organic phase. Specific type of solid oxidant, such as metallic superoxide, KO
2
, is
used to oxidize organic sulfur compounds. Acetic acid and trifluoroacetic acid are used
together with mixing and/or ultrasonication to enhance the oxidation process. The major
concepts and important points from the experiments are discussed below.

1. Oxidant selection is one of the major issues in the development of the 4
th
generation
UAOD. In the previous generations of UAOD, diluted (0.3% to 30% wt) hydrogen
peroxide was used as oxidant to oxidize organic sulfur compounds. However, high
dosage (greater than 1:100 of sulfur to oxidant ratio) of oxidant was required to achieve
95% desulfurization. In order to reduce the dosage of oxidant, a stable oxidant with high
purity is required. Therefore, solid oxidants including potassium permanganate, sodium
superoxide, and potassium superoxide were investigated. However, in the 4
th
generation
of UAOD process using potassium superoxide, only 1:30 of sulfur to oxidant ratio is
required to achieve 95% desulfurization of various OSCs. At low sulfur to oxidant ratio
of 1:4, potassium superoxide can still achieve 90% oxidization of organic sulfur
compounds.

215
Superoxide can also achieve a faster desulfurization as compared with hydrogen peroxide.
For instance, 80% to 90% desulfurization for various model sulfur compounds including
T, 2MT, BT, 2MBT, DBT and 4,6DMDBT can be achieved using potassium superoxide
within reaction period of 60 minutes,, while it requires 100 minutes to achieve the same
desulfurization efficiencies by hydrogen peroxide (Cheng et al., 2009). It is because free
radical oxidation by superoxide anion is stronger than chemical oxidation by hydrogen
peroxide.

2. Phase transfer catalysts applied includes both quaternary ammonium salt and ionic
liquid. Quaternary ammonium salt can extract superoxide anion from solid phase to
liquid phases. Quaternary ammonium salt reduced the surface tension between phases,
and facilitates oxidant into the organic phase. Selection of counter ion in quaternary
ammonium salt is important because it can affect active nucleophile anions extraction
efficiency. Highly lipophobic ion such as fluoride ion, can effectively suppress the degree
of hydration around active nucleophile anions to achieve better phase transfer. Due to its
stability against strong oxidants, no halogenated by-products would be formed during the
process. Therefore, quaternary ammonium fluoride is selected as one of the phase transfer
agents in the 4
th
generation UAOD.

3. Ionic liquid is the second phase transfer catalyst applied in the process. Similar to
quaternary ammonium salt, ionic liquid can extract active nucleophile anions into organic
phase so as to oxidize organic sulfur compounds. Due to the thermostability,

216
chemostability, and low vapor pressure, ionic liquids are considered as new generation of
green solvents and catalysts. However, certain types of ionic liquids are susceptible to
decomposition with the use of strong oxidant and ultrasonication. For example, one of the
most common types of ionic liquid, hexafluorophosphate [PF
6
] anion based ionic liquids
could be oxidized by strong oxidant such as hydrogen peroxide, or decomposed under
ultrasonication. Corrosive fume of hydrogen fluoride is reported as one of the major
product from [PF
6
] degradation.

Ionic liquids with alkylsulfate as anion are considerably more stable and relatively low in
toxicity, and therefore as a substitute to [PF
6
] based ionic liquid, the following four
compounds were selected in this study: 1-ethyl-3-methylimidazolium ethylsulfate
[EMIM][EtSO
4
], 1,2,3-trimethylimidazolium methylsulfate [TMIM][MeSO
4
], and
tributylmethyl-phosphonium methylsulfate [TMBP][MeSO
4
]. Among these ionic liquids,
[EMIM][EtSO
4
] demonstrated ease of application, excellent catalytic performance and
achieved greater than 98% desulfurization efficiency in the model sulfur compound study.

4. Acetic acid and trifluoroacetic acid applied in the 4
th
generation of UAOD can be
oxidized to peracetic acid and trifluoroperacetic acid, respectively. With the aid of phase
transfer catalysts, the peracids generated were able to oxidize organic sulfur compounds
in the oil phase.

217
5. Ultrasonication used in these desulfurization studies provided local high temperature
and pressure, causing decomposition of OSCs. Desulfurization efficiency of 95% was
achieved after 2 hours of reaction time. Furthermore, 98% efficiency was reached by
increasing the reaction time by 10 minutes.

6. It was demonstrated in the model sulfur compounds studies that the 4
th
generation
UAOD can oxidize various organic sulfur compounds including thiophene (T), 2-methyl
thiophene (2MT), benzothiophene (BT), 2-methyl benzothiophene (2BT),
dibenzothiophene (DBT), and 4, 6-dimethyl dibenzo-thiophene (4,6DMDBT). From the
results of kinetic studies, it was demonstrated that the oxidation efficiencies of the tested
model sulfur compounds decrease in the following order: BT > 2MBT > DBT > 4,
6DMDBT > T > 2MT.

7. It was concluded that high energy consumption by ultrasonication could limit the
applications of the 4
th
generation UAOD. Preliminary studies were conducted using
ultraviolet assisted desulfurization process as an alternative enhancement process to
ultrasonication. The results demonstrated this process can reduce energy consumption by
85%. Ultraviolet irradiation could be a possible alternative to ultrasonication.

8. To demonstrate the applicability of the 4
th
generation UAOD process to real fuels,
various fuels including JP-8, MGO, sour diesel, RO-6 and IFO were tested. It was
demonstrated that ultralow sulfur diesel can be produced from these fuels. The process

218
can achieve 95% desulfurization on mid-distillates even at sulfur to oxidant ratio of 1:2.8.
For heavy distillates, a maximum of 80% desulfurization can be achieved.

9. From the results of gas chromatography with sulfur chemiluminescence detector (GC-
SCD), it was shown that condensed ring structured sulfur compounds including alkyl-
substituted DBTs, benzonaphtho-thiophene derivatives and phenanthro-thiophene
derivatives are the major sulfur constituents in the desulfurized fuel samples.

10. From the GC-SCD chromatograms, it was determined that the 4
th
generation UAOD
can oxidize some of the organic sulfur compounds to compounds other than sulfones. The
most likely products are inorganic sulfur compounds. A possible mechanistic explanation
would be that radical reactions initiated by superoxide anion can convert organic sulfur
compounds such as BT and DBT to inorganic sulfur compounds.

11. A conceptual model for the 4
th
generation UAOD process was developed with the
following twelve steps: (1) extraction of superoxide radical by phase transfer catalyst; (2)
extraction of anion from acid catalysts; (3) generation of active oxygen species from
superoxide; (4) generation of peroxyacetic acid; (5) transfer of active nucleophiles to
organic phase; (6) oxidation of organic sulfur compounds by peroxyacetic acid; (7)
oxidation of organic sulfur compounds by superoxide radical; (8) regeneration of
catalysts in the polar phase; (9) oxidation of ionic liquid anion to form alkyl persulfate
anion; (10) transfer of alkyl persulfate anion to organic phase; (11) oxidation of organic
sulfur by alkyl persulfate anion; and (12) regeneration of ionic liquid’s ion pair.

219

12. Based on bench scale studies, a simplified cost evaluation for different generations of
UAOD processes is illustrated in Table 7.1. It is demonstrated that the total chemical cost
for the 4
th
generation UAOD is considerably less than others.

Table 7.1 Chemical cost comparison of different UAOD generations to desulfurize 10
grams sample with 1000ppm
w
DBT in bench scale study

13. A schematic diagram of the 4
th
generation UAOD process is illustrated in Figure 7.1.
A premixer was used to mix phase transfer catalysts including ionic liquid and quaternary
ammonium salt with potassium superoxide in order to release superoxide anion in the
liquid phase. Fuel oil was subsequently mixed with acid catalysts, superoxide and phase

1st Generation
UAOD
2nd Generation
UAOD
3rd Generation
UAOD
4th Generation
UAOD

Price
($/kg)
Dosage
(g)
Cost
($)
Dosage
(g)
Cost
($)
Dosage
(g)
Cost
($)
Dosage
(g)
Cost
($)
H
2
O
2
(30%) 452 10 4.52 10 4.52 5 2.26
KO
2
718 0.7 0.5
TMC 884 0.1 0.09 0.1 0.09
HAc 30.92 1.4 0.04 1.9 0.06
TFA 216 0.6 0.13 0.1 0.02
QBr 2875 0.1 0.29
QF 3528 0.1 0.35 0.6 2.12 0.1 0.35
IL 245 10 2.45 5 1.23
Total 10.2 4.90 10.2 4.96 17.6 7.00 7.8 2.16

220
transfer catalysts in the main reactor to achieve OSCs oxidation at 70°C and ambient
pressure. A phase separation vessel was used to separate fuel oil from the polar phase
which contained mainly the spent catalysts. Solvent acetonitrile was then used for sulfone
extraction. Desulfurized fuel oil was separated from spent solvent. Spent solvent was
regenerated via distillation. Regenerated solvent was reused for sulfone extraction.

Figure 7.1 Schematic diagram of the 4
th
generation UAOD process

221

7.2 Recommendations for Future Work

Conduct bench-scale completely mixed flow reactor (CMFR) studies to further refine the
4
th
generation UAOD process. Some of the most significant issues are discussed below.

1. Enhanced treatment of heavy distillates. It was shown that the 4
th
generation UAOD
process can achieve greater than 95% desulfurization for various mid-distillates.
However, only a maximum of 80% desulfurization can be achieved for heavy-distillates.
Therefore, it is important to improve desulfurization efficiency for heavy-distillates. A
pre-treatment step to remove asphaltenes from heavy distillates could be one possible
method. To reach this goal, it is necessary to identify and quantify the remaining sulfur
compounds in heavy-distillates in order to implement the required treatment steps.

2. It was proposed that the 4
th
generation UAOD process can oxidize some organic sulfur
compounds such as benzothiophenes and dibenzothiophenes to multiple products
including sulfones and inorganic sulfur compounds. To design an effective post-treatment
step, it is important to qualitatively and quantitatively determine the organic and
inorganic sulfur compounds as well as other by-products at each step of treatment. This
can be accomplished by: i) gas chromatography equipped with sulfur chemiluminescence
detector can be used to detect organic sulfur compounds, ii) ion chromatograph equipped
with pulsed electrochemical detector can be used to detect inorganic sulfate or sulfite

222
ions, and iii) gas chromatography equipped with mass spectrometer can be used to
identify organic products.

3. Investigate the applicability of ultraviolet irradiation for desulfurization of OSCs. It
was demonstrated in the preliminary studies that ultrasonication process could cost 15
times more than that of UV irradiation. Accordingly, UV assisted oxidative
desulfurization process could potentially be one of the next generations of desulfurization
treatment technologies.

Addition of solid catalyst such as titanium (IV) oxide which can increase free radical
generation under UV irradiation should also be considered in the future development.

4. Solar power would be another alternative for desulfurization of OSCs. Demonstrated in
a recent publication (Shiraishi et al., 2003), dibenzothiophene sulfuroxide can be
photolysed to biphenyls and inorganic sulfur compounds with the application of xenon
lamp (λ > 300 nm). Therefore, photo-irradiation could be used as a possible post
treatment step for the 4
th
generation UAOD process.

5. With the addition of potassium superoxide, various potassium salts could be
generated. It is important to investigate the fate of potassium in the process. Recovery of
potassium should also be considered in the future investigations.

223

6. Greener chemicals, including oxidants, solvents and catalysts should be investigated
in the future development.

7. Conduct pilot-scale studies to optimize the process efficiency and assess its
economical viability.

224
REFERENCES

Afanas’ev, I. Superoxide Ion: Chemistry and Biological Implications. Florida: CRC Press,
1989
Al-Ekabi, H., Edwards, G., Holden, W., Safarzadeh-Amiri, A. and Story, J. Chemical
Oxidation. Pennsylvania: Technomic Publishing Company Inc., 1992
Andari, M., Behbehani, H. and Qabazard, H. “Database for Organic Sulfur Compounds
Using GC-SCD Method. Determination of Sulfur Containing Compounds in Straight Run
Gas Oils (SRGO).” Fuel Science & Technology Int’l., 14(7): 897 – 908, 1996
Andersson, J. and Bobinger, S. “Polycyclic Aromatic Sulfur Heterocycles. II.
Photochemcial Oxidation of Benzo[b]thiophene in Aqueous Solution.” Chemosphere,
24(4): 383 – 389, 1992
Arce, A., Earle, M., Katdare, S., Rodriguez, H. and Seddon, K. “Application of Mutually
Immiscible Ionic Liquids to the Separation of Aromatic and Aliphatic Hydrocarbons by
Liquid Extraction: a Preliminary Approach.” Phys. Chem. Chem. Phys., 10: 2538 – 2543,
2008
Armstrong, S., Sankey, B. and Voordouw. G. “Conversion of Dibenzothiophene to
Biphenyl by Sulfate-Reducing Bacteria Isolated from Oil Field Production Facilities.”
Biotechnology Letters, 17(10): 1133-1136, 1995
Baba, Y . “the Photo-Desulfurization of Crude Naphtha by UV Irradiation.” Bulletin of the
Chemical Society of Japan, 47(1): 204 – 209, 1974
Bäckvall, J. Modern Oxidation Methods. Weinheim: Wiley, 2004
Baker, K., Robertson, V. and Duck, F. “a Review of Therapeutic Ultrasound: Biophysical
Effects.” Phys. Ther., 81(7): 1351 – 1358, 2001

225
Barrer, R. “the Viscosity of Pure Liquids II: Polymerised Ionic Melts.” Trans. Faraday
Soc. 39: 59 – 67, 1943
Becker, G. and Colmsjo, A. “Gas Chromatography-Atomic Emission Detection for
Quantification of Polycyclic Aromatic Sulfur Heterocycles.” Analytica Chimica Acta,
376: 265 – 272, 1998
Beckett, M. and Hua, I. “Impact of Ultrasonic Frequency on Aqueous Sonoluminescence
and Sonochemistry.” J. Phys. Chem. A, 105(15): 3796 – 3802, 2001
Benner, R. and Stedman, D. “Universal Sulfur Detection by Chemiluminescence.” Anal.
Chem., 61(11): 1268 – 1271, 1989
Berlan, J. and Mason, T. “Dosimetry for Power Ultrasound and Sonochemsitry.”
Advances in Sonochemistry, 4: 1 –73, 1996
Beychok, M. Fundamentals of Stack Gas Dispersion. California: M.R. Beychok, 2005
Bianchini, C., Casares, J., Osman, R., Pattison, D., Peruzzini, M., Perutz, R. and Zanobini,
F. “C-H Bond Cleavage in Thiophenes by [P(CH
2
CH
2
PPh
2
)
3
Ru]. UV Flash Kinetic
Spectroscopy Discloses the Ruthenium-Thiophene Adduct Which Precedes C-H
Insertion.” Organometallics, 16(21): 4611 – 4619, 1997
Bochet, C., Draper, T., Bocquet, B., Pope, M. and Williams, A. “
182
Tungsten Mossbauer
Spectroscopy of Heteropolytungstates.” Dalton Trans., 5127 – 5131, 2009
Bockris, J. and Reddy, A. Modern Electrochemistry 1: Ionics. New York: Springer, 1998
Boikov, E., Vakhrushin, P. and Vishnetskaya, M. “Oxidative Desulfurization of
Hydrocarbon Feedstock.” Chemistry and Technology of Fuels and Oils, 44(4): 271 – 274,
2008
Borrás-Almenar, J., Coronado, E. Müller, A. and Pope, M. Polyoxometalate Molecular
Science (NATO Science Series II: Mathematics, Physics and Chemistry). The Netherlands:
Kluwer Academic Publishers, 2003

226
Bregeault, J. “Transition-metal complexes for liquid-phase catalytic oxidation: some
aspects of industrial reactions and of emerging technologies.” Dalton Trans., 3289 – 3302,
2003
Bukowski, A. and Milczarska, T. “Asphalts as Inhibitors of Radical Rolymerization.”
Journal of Applied Polymer Science, 28: 1001 – 1009, 1983
California Air Resource Board. California Low Sulfur Diesel Fuel. 2003
Callahan, L., She, Z. and Nosek, T. “Superoxide, Hydroxyl Radical, and Hydrogen
Peroxide Effects on Single-Diaphragm Fiber Contractile Apparatus.” J. Appl. Physiol., 90:
45 – 54, 2001
Campos-Martin, J., Capel-Sanchez, M. and Fierro, J. “Highly Efficient Deep
Desulfurization of Fuels by Chemical Oxidation.” Green Chem., 6: 557 – 562, 2004
Chan, N., Lin, T. and Yen, T. “Superoxides: Alternative Oxidants for the Oxidative
Desulfurization Process.” Energy & Fuels, 22: 3326 – 3328, 2008
Chan, N., Fan, W., and Yen, T., “the Desulfurization Process of Heavy Fuel Oil.” Am.
Chem. Soc., Div. Fuel Chem. (Prepr.), 54(1): 70 – 71, 2009
Chang, H., Yon, G., and Kim, Y. “Facile Synthesis of 2-Substituted Aminobenzoxazole
One Pot Cyclodesulfurization of N-(2-Hydroxylphenyl)-N’-Phenylthioureas with
Superoxide Radical Anion.” Chem. Letters, 1291 – 1294, 1986
Chen, L., Guo, S., and Zhao, D. “Oxidative Desulfurization of Simulated Gasoline over
Metal Oxide-Loaded Molecular Sieve.” Chi. J. Chem. Eng., 15(4): 520 – 523, 2007
Cheng, S. Ultra Clean Fuels via Modified UAOD Process with Room Temperature Ionic
Liquid (RTIL) & Solid Catalyst Polishing. California: University of Southern California.
Libraries, 2008
Cheng, S. and Yen, T. “Use of Ionic Liquids as Phase-Transfer Catalysis for Deep
Oxygenative Desulfurization.” Energy & Fuels, 22: 1400 – 1401, 2008

227
Cheng, S. and Yen, T. “Ultra Clean Fuel via Modified Ultrasound Assisted Oxidative
Desulfurization Process.” Prepr. A.C.S. Div. Pert. Chem., 54(1): 15 – 17, 2009
Collins, C. “Implementing Phytoremediation of Petroleum Hydrocarbons.” Methods in
Biotechnology, 23: 99 – 108, 2007
Coyle, J. Introduction to Organic Photochemistry. Chichester: John Wiley & Sons Ltd.,
1991
Crum, L., Mason, T., Reisse, J. and Suslick, K. Sonochemistry and Sonoluminescence.
The Netherlands: Kluwer Academic Publishers, 1998
Crum, L. and Roy, R. “Sonoluminescence.” Science, 266: 233 – 234, 1994
Dehkordi, A., Kiaei, Z. and Sobati, M. “Oxidative Desulfurization of Simulated Light
Fuel Oil and Untreated Kerosene.” Fuel Processing Technology, 90(3): 435 – 445, 2009
Dehmlow, E. and Dehmlow, S. Phase Transfer Catalysis (3
rd
). Weinheim: VCH, 1993
Dermeikand, S. and Sasson, Y. “Synthesis of Quaternary Ammonium Fluoride Salts by a
Solid-Liquid Halogen Exchange Process in Protic Solvents.” J. Org. Chem., 54: 4827 –
4829, 1989
Dhir, S., Uppaluri, R. and Purkait, M. “Oxidative Desulfurization: Kinetic Modelling.”
Journal of Hazardous Materials, 161: 1360 – 1368, 2009
Didenko, Y. and Suslick, K. “the Energy Efficiency of Formation of Photons, Radicals
and Ions during Single-Bubble Cavitation.” Nature, 418: 394 – 397, 2002
Domanska, U. Pobudkowska, A. and Krolikowski, M. “Separation of Aromatic
Hydrocarbons from Alkanes Using Ammonium Ionic Liquid C
2
NTf
2
at T = 298.15K.”
Fluid Phase Equilibria, 259: 173 – 179, 2007

228
Drushel, H. “Sulfur Compound Type Distributions in Petroleum Using an In-Line
Reactor or Pyrolysis Combined with Gas Chromatography and a Microcoulometric
Sulfur Detector.” Anal. Chem., 41(4): 569 – 576, 1969
Earle, M. and Seddon, K. “Ionic Liquids: Green Solvents for the Future.” Pure Appl.
Chem., 72(7): 1391 – 1398, 2000
Ellzy, M. and Janes, L. Sulfur Chemiluminescence Detection Compared to Sulfur Flam
Photometric Detection. Maryland: Edgewoood, 1998
Endo, K., Miyasaka, T., Mochizuki, S., Aoyagi, S., Himi, N., Asahara, H., Tsujioka, K.
and Sakai, K. “Development of a Superoxide Sensor by Immobilization of Superoxide
Dismutase.” Sensors and Actuators, B83: 30 – 34, 2002
Energy Information Administration (EIA). The Transition to Ultra-Low-Sulfur Diesel
Fuel: Effects on Prices and Supply. U.S. Department of Energy, SR/OIAF/2001-01, 2001
Environment and Conservation. New Source Performance Standards. Department of
State (TN) Division of Publications, 1200-03-16, 2009
Etemadi, O. and Yen, T. “Aspect of Selective Adsorption among Oxidized Sulfur
Compounds in Fossil Fuels.” Energy and Fuels, 21: 1622 – 1627, 2007
Etemadi, O. and Yen, T. “Selective Adsorption in Ultrasound-Assisted Oxidative
Desulfurization Process for Fuel Cell Reformer Application.” Energy and Fuels, 21:
2250 – 2257, 2007
Etemadi, O. and Yen, T. “Surface Characterization of Adsorbents in Ultrasound-Assisted
Oxidative Desulfurization Process of Fossil Fuels.” J. of Colloid and Interface Science,
313: 18 – 25, 2007
Fang, X., Mark, G. and Sonntag, C. “OH Radical Formation by Ultrasound in Aqueous
Solutions Part I: the Chemistry Underlying the Terephthalate Dosimeter.” Ultrasonics
Sonochemistry, 3: 57 – 63, 1996

229
Filippis, P. and Scarsella, M. “Oxidative Desulfurization: Oxidation Reactivity of Sulfur
Compounds in Different Organic Matrixes.” Energy Fuels, 17(6): 1452 – 1455, 2003
Foote, C. Active Oxygen in Chemistry (Structure, Energetics and Reactivity in Chemistry).
London: Chapman & Hall, 1995
Fox, D., Gilman, H., Morgan, A., Shield, J., Maupin, P., Lyon, R., De Long, H. and
Trulove, P. “Flammability and Thermal Analysis Characterization of Imidazolium-Based
Ionic Liquids.” Ind. Eng. Chem. Res., 47(16): 6327 – 6332, 2008
Frank, S. and Bard, A. “Heterogeneous Photocatalytic Oxidation of Cyanide Ion in
Aqueous Solutions at Titanium Dioxide Powder.” J. Am. Chem. Soc., 99(1): 303 – 304,
1977
Freemantle, M. “Designer Solvents: Ionic Liquids may Boost Clean Technology
Development.” Chemical & engineering news, 76: 32 – 37, 1998
Fukuto, J., Stefano, E., Burstyn, J., Valentine, J. and Cho, A. “Mechanism of Oxidation
of N-Hydroxyphentermine by Superoxide.” Biochemistry, 24: 4161 – 4167, 1985
Galvanic Applied Sciences (GAS) Inc. Sulfur Chemiluminescence Analyzers. 2004
Gaitan, F., Crum, L., Church, C. and Roy, R. “Sonoluminescence and Bubble Dynamics
for a Single, Stable, Cavitation Bubble.” J. Acoust. Soc. Am., 91: 3166 – 3183, 1992
Garcia, C., Becchi, M., Loustalot, M., Paisse, O. and Szymanski, R. “Analysis of
Aromatic Sulfur Compounds in Gas Oils Using GC with Sulfur Chemiluminescence
Detection and High-Resolution MS.” Anal. Chem., 74: 3849 – 3857, 2002
Gatan, R., Barger P. and Gembicki, V. “Oxidative Desulfurization: a New Technology for
ULSD.” Fuel Chemistry (Preprint), 49(2): 577- 579, 2004
Fibilisco, S. Alternative Energy Demystified. New York: McGraw-Hill, 2006

230
Girgis, M. and Gates, B. “Reactivities, Reaction Networks and Kinetics in High-pressure
Catalytic Hydroprocessing.” Ind. Eng. Chem., 30: 2021-2058, 1991
Gokel, G., Cram, D., Liotta, C., Harris, H. and Cook, F. “18-Crown-6.” Org. Synth., 57:
30, 1988
Gold, V., Loening, K., McNaught, A. and Shemi, P. Compendium of Chemical
Terminology. International Union of Pure and Applied Chemistry, 1987
Greenspan, F. “Oxidation Reactions with Aliphatic Peracids.” Ind. Eng. Chem., 39(7):
847 – 848, 1947
Greenspan, H. and Nadim, A. “Sonoluminescence of an Oscillating Gas Bubble.” Physics
of Fluid A: Fluid Dynamics, 5(4): 1065 – 1067, 1993
Guida, W. and Mathre, D. “Phase-Transfer Alkylation of Heterocycles in the Presence of
18-Crown-6 and Potassium Tert-Butoxide.” J. Org. Chem., 45(16): 3172 – 3176, 1980
Herrington, P. “Effect of Concentration on the Rate of Reaction of Asphaltenes with
Oxygen.” Energy & Fuels, 18: 1573 – 1577, 2004
Hough, W., Smiglak, M., Rodriguez, H., Swatloski, R., Spear, S., Daly, D., Pernak, J.,
Grisel, J., Carliss, R., Soutullo, M., Davis, J. and Rogers, R. “the Third Evolution of Ionic
Liquids: Active Pharmaceutical Ingredients.” New Journal of Chemistry, 31: 1429 – 1436,
2007
Huang, C., Chen, B., Zhang, J., Liu, Z. and Li, Y. “Desulfurization of Gasoline by
Extraction with New Ionic Liquids.” Energy and Fuels, 18: 1862 – 1864, 2004
Huang, D., Lu, Y., Wang, Y. And Luo, G. “Catalytic Kinetics of Dibenzothiophene
Oxidation with the Combined Catalyst Quaternary Ammonium Bromide and
Phosphotungstic Acid.” Ind. Eng. Chem. Res., 46: 6221 – 6227, 2007
Huddleston, J., Willauer, H., Swatloski, R., Visser, A. and Rogers, R. “Room Temperature
Ionic Liquids as Novel Media for Clean Liquid-Liquid Extraction.” Chem. Commun.,
1765 – 1766, 1998

231
Jess, A. and Eber, J. “Deep Desulfurization of Oil Refinery Streams by Extraction with
Ionic Liquids.” Electrochemical Society Proceedings, 24: 572-582, 2004
Johnson, K. “What’s an Ionic Liquid.” Interface, 16(1): 38 – 41, 2007
Jones, R. Quaternary Ammonium Salts: Their Use in Phase-Transfer Catalysis.
California: Academic Press, 2001
Juaristi, M., Aizpurua, J., Lecea, B. and Palomo, C. “Reagents and Synthetic Method 41:
Oxidations with Chromium Trioxide under the Influence of Crown Ethers.” Can. J.
Chem., 62: 2941 – 2944, 1984
Kaneko, M. and Okura, I. Photocatalysis Science and Technology. New York: Springer,
2002
Kim, I. and Jung, O. “Sonochemical Reaction Mechanism of a Polycyclic Aromatic
Sulfur Hydrocarbon in Aqueous Phase.” Bull. Korean Chem. Soc., 23(7): 990 – 994, 2002
Kim, I., Yoa, S., Lee, J. and Huang, C. “Reaction Pathways and Kinetic Modeling for
Sonochemical Decomposition of Benzothiophene.” Korean J. Chem. Eng., 20(6): 1045 –
1053, 2003
Kim, Y., Lim, S. and Chang, H. “Activation of Superoxide; Efficient Desulfurization of
Thioamides to the Corresponding Amides using a Peroxyphosphorus Intermediate
Generated from Phenylphosphonic Dicholoride and Superoxide.” Chem. Commun., 36 –
37, 1990
Kirimura, K., Furuya, T., Sato, R., Ishii, Y., Kino, K. and Usami, S. “Biodesulfurization
of Naphthothiophene and Benzothiophene through Selective Cleavage of Carbon-Sulfur
Bonds by Rhodococcus sp. Strain WU-K2R.” Applied and Environmental Microbiology,
68(8): 3867-3872, 2002
Kong, S., Liochev, S. and Fridovich, I. “Aluminum(III) Facilitates the Oxidation of
NADH by the Superoxide Anion.” Free Radical Biology & Medicine, 13: 79 – 81, 1992

232
Kozhevnikov, I. Catalysts for Fine Chemical Synthesis, Catalysis by Polyoxo-metalates.
West Sussex: Wiley, 2002
Kropp, K., Goncalves, J., Andersson, J. and Fedorak, P. “Microbially Mediated
Formation of Benzonaphthothiophenes from Benzo[b]thiophenes.” Applied and
Environmental Microbiology, 60(10): 3624 – 3631, 1994
Kwon, J., Moon, J., Lee, D., Bae, Y., Sohn, H. and Lee, C. “Adsorptive Desulfurization
and Denitrogenation of Refinery Fuels using Mesoporous Silica Adsorbents.”
ChemSusChem, 1(4): 307-309, 2008
Lanju, C., Shaohui, G., Dishun, Z., Jialin, W. and Tong, M. “Oxidation of Thiophenes
over Silica Gel Using Hydrogen Peroxide and Formic Acid.” Energy Sources, 30(4):
370 – 376, 2008
Lee, B., Chi, Y., Lee, J., Choi, I., Song, C., Namgoong, S. and Lee, S. “Imidazolium Ion-
Terminated Self-Assembled Monolayers on Au: Effects of Counteranions on Surgace
Wettability.” J. Am. Chem. Soc., 126: 480 – 481, 2004
Lee, S. “Comparison of the Atomic Emission Detector to Other Element-Selective
Detectors for the Gas Chromatographic Analysis of Pesticide Residues.” J. Agric. Food
Chem., 39: 2192 – 2199, 1991
Lee, S., Speight, J. and Loyalka, S. Handbook of Alternative Fuel Technologies. Florida:
CRC Press, 2007
Levshin, V. and Rzhevkin, S. “Mechanism of Luminescence in Liquids under Ultrasonic
Treatment.” Doklady Akademii Nauk SSSR, 16: 399 – 404, 1973
Li, F., Zhang, Z., Feng, J., Cai, X. and Xu, P. “Biodesulfurization of DBT in Tetradecane
and Crude Oil by a Facultative Thermophilic Bacterium Mycobacterium Goodie.”
Journal of Biotechnology, 127: 222–228, 2007
Li, F., Zhao, D., Li, H. and Liu, R. “Photochemical Oxidation of Thiophene by O
2
in an
Oil/Acetonitrile Two-Phase Extraction System.” Ann. N. Y. Acad. Sci., 1140: 383 – 388,
2008

233
Li, H., Zhu, W., Wang, Y., Zhang, J., Lu, J. and Yan, Y. “Deep Oxidative
Desulfurization of Fuels in Redox Ionic Liquids Based on Iron Chloride.” Green Chem.,
11: 810 – 815, 2009
Li, X., Zhao, J., Li, Q., Wang, L. and Tsang, S. “Ultrasonic Chemical Oxidative
Degradation of 1,3-Dialkylimidazolium Ionic Liquid and Their Mechanistic
Elucidations.” Dalton Trans., 1875 – 1880, 2007
Lin, J and Yen, T. “an Upgrading Process through Cavitation and Surfactant.” Energy
and Fuels, 7: 111 – 118, 1993
Liu, B., Xu, D., Chu, J. and Au, C. “Deep Desulfurization by the Adsorption Process of
Fluidized Catalytic Cracking (FCC) Diesel over Mesoporous Al-MCM-41 Materials.”
Energy & Fuel, 21: 250-255, 2007
Liu, J., Jiang, G., Chi, Y., Cai, Y., Zhou, Q. and Hu, J. “Use of Ionic Liquids for Liquid-
Phase Microextraction of Polycyclic Aromatic Hydrocarbons.” Anal. Chem., 75(21):
5870 – 5876, 2003
Lo, W., Yang, H. and Wei, G. “One-Pot Desulfurization of Light Oils by Chemical
Oxidation and Solvent Extraction with Room Temperature Ionic Liquids.” Green Chem.,
5: 639 – 642, 2003
Lohse, D. “Sonoluminescence: Inside a Micro-Reactor.” Nature, 418: 381 – 383, 2002
Ma, X., Sakanishi, K. and Mochida, I. “Hydrodesulfurization reactivity of Various Sulfur
Compounds in Vacuum Gas Oil.” Ind. Eng. Chem. Res., 35: 2487-2494, 1994
Ma, X., Sun, L. and Song, C. “Deep Desulfurization of Diesel Fuels by a Novel
Integrated Approach.” U.S. AAD Document Control Center, M/S 921-107, 2001
Ma, X., Sun, L., Yin, Z. and Song, C. “New Approaches to Deep Desulfurization of
Diesel Fuel, Jet Fuel, and Gasoline by Adsorption for Ultra-Clean Fuels and for Fuel Cell
Applications.” Am. Chem. Soc., Div. Fuel Chem. (Prepr.), 46: 648 – 649, 2001

234

Magaril, R. and Aksenova, E. “Investigation of the Mechanism of Coke Formation
During Thermal Decomposition of Asphaltenes.” Khimiyai Tekhnologiya Toplivi Masel,
7: 22 -24, 1970
Margulis, M. Sonochemistry and Cavitation. Amsterdam: Gordon and Breach Publishers,
1995
Marsh, K., Deev, A., Wu, A., Tran, E. and Klamt, A. “Room Temperature Ionic Liquids as
Replacement for Conventional Solvents – A Review.” Korean J. Chem. Eng., 19(3):
357 – 362, 2002
Mason, T. Sonochemistry. Oxford: OUP Inc., 1999
Mason, T. and Lorimer, J. Applied Sonochemistry: Uses of Power Ultrasound in
Chemistry and Processing. Weinheim: Wiley-VCH, 2002
Matsumoto, M., Mochiduki, K., Fukunishi, K. and Kondo, K. “Extraction of Organic
Acids Using Imidazolium-Based Ionic Liquids and Their Toxicity to Lactobacillus
Rhamnosus.” Separation and Purification Technology, 40(1): 97 – 101, 2004
Mei, H., Mei, B. and Yen, T. “a New Method for Obtaining Ultra-Low Sulfur Diesel Fuel
via Ultrasound Assisted Oxidative Desulfurization.” Fuel, 82: 405 – 414, 2003
Meindersma, G., Podt, A., Klaren, M. and Haan, A. “Separation of Aromatic and
Aliphatic hydrocarbons with Ionic Liquids.” Chem. Eng. Comm., 193(11): 1384 - 1396,
2006
McFarland, B. “Biodesulfurization.” Current Opinion in Microbiology, 2: 257-264, 1999
Milazzo, G., Caroli, S. and Sharma, V . Tables of Standard Electrode Potentials. New York:
Wiley, 1978
Moschopedis, S. and Speight, J. “Water-Soluble Derivatives of Athabasca Asphaltenes.”
Fuel, 50(1): 34 – 40, 1971

235
Mukherjee, R. Fundamentals of Photochemistry. New Delhi: New Age International,
1978
Munson, C., Boudreau, L., Driver, M. and Schinski, W. “Separation of Olefins from
Paraffins Using Ionic Liquid Solutions.” U.S. Patent Application Publication,
US2002/0063240 A1, 2002
Murata, S., Murata, K., Kidena, K. and Nomura, M. “a Novel Oxidative Desulfurization
System for Diesel Fuels with Molecular Oxygen in the Presence of Cobalt Catalysts and
Aldehydes.” Energy Fuels, 18(1): 116 – 121, 2004
Murov, S., Carmichael, I. and Hug, G. Handbook of Photochemistry. New York: Marcel
Dekker, 1993
Noltingk, B. and Neppiras, E. “Cavitation Produced by Ultrasonics.” Proceedings of
Physical Society, London, 63B: 674 – 685, 1950
Oae, S., Takata, T. and Kim, Y . “Reaction of Organic Sulfur Compounds with Superoxide
Anion – III Oxidation of Organic Sulfur Compounds to Sulfinic and Sulfonic Acids.”
Tetrahedron, 32: 37 – 44, 1981
Ohl, C. “Luminescence from Acoustic-Driven Laser Induced Cavitation Bubbles.” Phys.
Rev. E, 61: 1497 – 1500, 2000
Otsuki, S., Nonaka, T., Takashima, N., Qian, W., Ishihara, A., Imai, T. and Kabe, T.
“Oxidative Desulfurization of Middle Distillate Using Ozone.” J. of Jap. Pet. Ins., 42(5):
315 – 320, 1999
Otsuki, S., Nonaka, T., Takashima, N., Qian, W., Ishihara, A., Imai, T. and Kabe, T.
“Oxidative Desulfurization of Light Gas Oil and Vacuum Gas Oil by Oxidation and
Solvent Extraction.” Energy Fuels, 14(6): 1232–1239, 2000
Oxley, J., Prozorov, T. and Suslick, K. “Sonochemistry and Sonoluminescence of Room-
Temperature Ionic Liquids.” J.A.C.S., 125: 11138 – 11139, 2003

236
MEPC. Annex 13: Amendments to the Annex of the Protocol of 1997 to Amend the
International Convention for the Prevention of Pollution from Ships, 1973, as Modified
by the Protocol of 1978 Relating Thereto. IMO, 2008
MEPC. The Protocol of 1997: Annex VI - Regulations for the Prevention of Air Pollution
from Ships. IMO, 1997
Paniv, P., Pysh’ev, S., Gaivanovich, V. and Lazorko, O. “Noncatalytic Oxidation
Desulfurization of the Kerosene Cut.” Chem. and Tech. of Fuels and Oils, 42(3): 159 -
168, 2006
Park, B., Lee, J., Lee, Y., Hong, J. and Jung, O. “Anion Effects on the Formation of
Cross-Linked Argentophilic Interaction. Synthesis and Structural Properties of AgX
Bearing Bis(S-pyridyl)dimethylsilane (X
-
= CF
3
SO
3
-
, PF
6
-
, and NO
3
-
).” Bull. Chem. Soc.
Jpn., 78, 1624 – 1628, 2005
Parvulescu, V. and Hardacre, C. “Catalysis in Ionic Liquids.” Chem. Rev., 107(6): 2615 –
2665, 2007
Poole, C. “Chromatographic and Spectroscopic Methods for the Determination of Solvent
Properties of Room Temperature Ionic Liquids.” J. of Chromatography A, 1037: 49 – 82,
2004
Preston, W. and Dong, Y. “Physical and Chemical Characterization of Residual Oil-Fired
Power Plant Emissions.” Energy and Fuels, 23: 2544 – 2551, 2009
Priyanto, S., Mansoori, G. and Suwono, A. “Measurement of Property Relationships of
Nano-Structure Micelles and Coacervates of Asphaltene in a Pure Solvent.” Chem. Eng.
Science, 56: 6933 -6939, 2001
Rantwijk, F., Lau, R. and Sheldon, R. “Biocatalytic Transformations in Ionic Liquids.”
TRENDS in Biotechnology, 21(3): 131 – 138, 2003

237
Ranu, B. and Jana, R. “Ionic Liquids as Catalyst and Reaction Medium – a Simple,
Efficient and Green Procedure for Knoevenagel Condensation of Aliphatic and Aromatic
Carbonyl Compounds Using a Task-Specific Basic Ionic Liquid.” Eur. J. Org. Chem., 16:
3767 – 3770, 2006
Rashidi, L., Mohebali, G., Darian, J. and Rasekh, B. “Biodesulfurization of
Dibenzothiophene and its Alkylated Derivatives through the Sulfur-Specific Pathway by
the Bacterium RIPI-S81.” African Journal of Biotechnology, 5(4): 351-356, 2006
Robert, D. and Maloto, S. “Solar Photocatalysis: a Clean Process for Water
Detoxification.” the Science of the Total Environment, 291(1-3): 85 – 97, 2002
Rogers, R., Seddon, K. and Volkov, S. Green Industrial Applications of Ionic Liquids.
The Netherlands: Kluwer Academic Publishers, 2003
Rosa, M., Lamberti, M., Pellecchia, C., Scettri, A., Villano, R. and Soriente, A. “an
Efficient solvent free catalytic oxidation of sulfides to sulfoxides with hydrogen peroxide
catalyzed by a binaphthyl-bridged Schiff base titanium complex.” Tetrahedron Letters,
47(40): 7233 – 7235, 2006
Rose, A., Moffett, J. and Waite, T. “Determination of Superoxide in Seawater Using 2-
Methyl-6-(4-methoxyphenyl)-3,7-dihydroimidazo[1, 2-a]pyrazin-3(7H)-one Chem-
iluminescence.” Anal. Chem., 80: 1215 – 1227, 2008
Ryerson, T., Barkley, R. and Sievers, R. “Selective Chemiluminescence Detection of
Sulfur-Containing Compounds Coupled with Nitrogen-Phosphorus Detection for Gas
Chromatography.” Journal of Chromatorgraphy A, 670: 117 – 126, 1994
Safarik, D. and Eldridge, R. “Olefin/Paraffin Separations by Reactive Absorption: a
Review.” Ind. Eng. Chem. Res., 37: 2571 – 2581, 1998
Saiyasitpanich, P., Lu, M., Keener, T., Liang, F. and Khang, S. “The Effect of Diesel Fuel
Sulfur Content on Particulate Matter Emissions for a Nonroad Diesel Generator.” J. Air
& Waste Manage. Assoc., 55: 993-998, 2005

238
Schulz, H. “Walter Böhringer, Peter Waller and Farid Ousmanov, Gas oil deep
hydrodesulfurization: refractory compounds and retarded kinetics.” Catalysis Today, 49:
87-97, 1999
Sdeghi, K., Lin, J. and Yen, T. “Sonochemical Treatment of Fossil Fuels.” Energy
Sources, 16(3): 439 – 449, 1994
Shiraishi, Y., Tachibana, K., Hirai, T. and Komasawa, I. “Photochemical Production of
Biphenyls from Oxidized Sulfur Compounds Obtained by Oxidative Desulfurization of
Light Oils.” Energy & Fuels, 17: 95 – 100, 2003
Simanzhenkov, V. and Idem, R. Crude Oil Chemistry. New York: Marcel Dekker Inc.,
2003
Smith, D. Photochemistry Vol. 1, London: RSC Publishing, 1970
Song, C. and Roh, E. “Practical Method to Recycle a Chiral (salen)Mn Epoxidation
Catalyst by Using an Ionic Liquid.” Chem. Commun., 10: 837 – 838, 2000
Speight, J. The Chemistry and Technology of Coal. New York: Marcel Dekker Inc., 1994
Speight, J. The Chemistry and Technology of Petroleum. New York: Marcel Dekker Inc.,
1999
Speight, J. The Desulfurization of Heavy Oils and Residua. New York: Marcel Dekker
Inc., 1999
Speight, J. “Petroleum Asphaltenes Part 2: The Effect of Asphaltenes and Resin
Constituents on Recovery and Refining Processes.” Oil and Gas Science and Technology,
59(5): 479 – 488, 2004
Suarez, P., Dullius, J., Einloft, S., Souza, R. and Dupont, J. “the Use of New Ionic
Liquids in Two-Phase Catalytic Hydrogenation Reaction by Rhodium Complexes.”
Polyhedron, 15(7): 1217 – 1219, 1996

239
Starks, C. “Phase-Transfer Catalysis I: Heterogeneous Reactions Involving Anion
Transfer by Quaternary Ammonium and Phosphonium Salts.” J. Am. Chem. Soc., 93(1):
195 – 199, 1971
Starks, C. and Halper, M. Phase-Transfer Catalysis: Fundamentals, Applications, and
Industrial Perspectives. New York: Chapman & Hall, 1994
Steynberg, A. and Dry, M. Fischer-Tropsch Technology, The Netherlands: Elsevier, 2004
Storey, B. and Szeri, A. “A Reduced Model of Cavitation Physics for Use in
Sonochemistry.” Proc. R. Soc. Lond., 457: 1685 – 1700, 2001
Suslick, K., Didenko, Y ., Fang, M., Hyeon, T., Kolbeck, K., McNamara, W., Mdleleni, M.
and Wong, M. “Acoustic Cavitation and Its Chemical Consequences.” Phil. Trans. R. Soc.
Lond. A, 357: 335-353, 1999
Suzuki, H. and Kakinuma, K. “Formation of a TBA Reaction Product from Crown Ether
in the Presence of KO
2
.” J. Biochem., 85: 1547 – 1549, 1979
Thompson, L. and Doraiswamy, L. “Sonochemistry: Science and Engineering.” Ind. Eng.
Chem. Res., 38(4): 1215 – 1249, 1999
Takahama, U. “Oxidation Products of Kaempferol by Superoxide Anion Radical.” Plany
Cell Physiol., 28(5): 953 – 957, 1989
Takahashi, A., Yang, F. and Yang, R. “New Sorbents for Desulfurization by ð-
Complexation: Thiophene/Benzene Adsorption.” Ind. Eng. Chem. Res., 41: 2487-2496,
2002
Takishima, T. Basic and Clinical Aspects of Pulmonary Fibrosis. Florida: CRC Press,
1994
Tam, P., Kittrell, J. and Eldridge, J. “Desulfurization of Fuel Oil by Oxidation and
Extraction. 1. Enhancement of Extraction Oil Yield.” Ind. Eng. Chem Res., 29: 321-324,
1990

240
Tam, P., Kittrell, J. and Eldridge, J. “Desulfurization of Fuel Oil by Oxidation and
Extraction. 2. Kinetic Modeling of Oxidation Reaction.” Ind. Eng. Chem Res., 29: 324-
329, 1990
Tang, H., Li, W., Liu, Q., Guan, L., Song, J., Xing J. and Liu, H. “Adsorptive
Desulfurization of Diesel with Mesoporous Aluminosilicates.” Science in China Series B:
Chemistry, 52(3): 276-281, 2009
Te, M., Fairbridge, C. and Ring, Z. “Oxidation Reactivities of Dibenzothiophenes in
Polyoxometalate/H
2
O
2
and Formic Acid/H
2
O
2
Systems.” Applied Catalysis A, 219: 267 –
280, 2001
Tetsuo, A. and Yamamoto, D. “Oxidative Desulfurization of Liquid Fuels.” ACS Div. of
Fuel Chem., 39(2): 623 – 626, 1994
U.S. EPA. Control of Emissions of Air Pollution from Nonroad Diesel Engines and Fuel;
Final Rule. Federal Register, 69(124), 2004
Vanýsek, P. Handbook of Chemistry and Physics. Florida: CRC Press, 2006
Venkatachalapathy, C., Rajarajan, M., Banu, H. and Pitchumani, K. “Clay-Supported
Tetrabutylammonium Periodate as a Versatile Oxidant for Alcohols and Sulfides.”
Tetrahedron, 55: 4071 – 4076, 1999
Venkatadri, R. and Peters, R. “Chemical Oxidation Technologies: Ultraviolet
Light/Hydrogen Peroxide, Fenton’s Reagent, and Titanium Dioxide-Assisted
Photocatalysis.” Hazardous Waste & Hazardous Materials, 10(2): 107 – 149, 1993
Voronkov, M. and Deryagina, E. “Thermal Transformations of Organic Compounds of
Divalent Sulfur.” Russian Chemical Reviews, 69(1): 81 – 94, 2000
Wan, M. and Yen, T. “Enhance Efficiency of Tetraoctylammonium Fluoride Applied to
Ultrasound-Assisted Oxidative Desulfurization (UAOD) Process.” Applied Catalysis A:
General, 319: 237 – 245, 2007

241
Wan, M. and Yen, T. “Portable Continuous Ultrasound-Assisted Oxidative
Desulfurization Unit for Marine Gas Oil.” Energy and Fuels, 22: 1130 – 1135, 2008
Wang, D., Qian, E., Amano, H., Okata, K., Ishihara, A. and Kabe, T. “Oxidative
Desulfurization of Fuel Oil Part I: Oxidation of Dibenzothiophenes Using Tert-Butyl
Hydroperoxide.” Applied Catalysis A: General, 253: 91 – 99, 2003
Wang, M. and Rajendran, V. “Kinetics for Dichlorocyclopropanation of 1,7-Octadiene
under the Influence of Ultrasound Assisted Phase-Transfer Catalysis Conditions.” J. of
Molecular Catalysis A: Chemical, 273(1-2): 5 – 13, 2007
Wasserscheid, P. and Keim, W. “Ionic Liquids – New “Solutions” for Transition Metal
Catalysis.” Angewandte Chemie, 39(21): 3772 – 3789, 2000
Wasserscheid, P. and Welton, T. Ionic Liquids in Synthesis. Weinheim: Wiley, 2002
Wattana, P. and Fogler, H. “Characterization of Polarity-Based Asphaltene Subfractions.”
Energy & Fuels, 19: 101 – 110, 2005
Willis, P. Watching the Directives: Scientific Advice on the EU Physical Agents
(Electromagnetic Fields) Directive. London: the Science and Technology Committee,
2007
Wypych, G. Handbook of Solvents, New York: ChemTec Publishing, 2001
Yan, X., Lei, J., Liu, D., Guo, L. And Wu, Y. “Oxidation Reactivities of Organic Sulfur
Compounds in Fuel Oil Using Immobilized Heteropoly Acid as Catalyst.” Journal of
Wuhan University of Technology-Mater. Sci., 22(2): 320 – 324, 2007
Yasui, K. “Light Emission Mechanism of Sonoluminesence.” Bussei Kenkyu, 72(3):
337 –344, 1999
Yazici, E., Deveci, H., Alp, I. and Uslu, T. “Generation of Hydrogen Peroxide and
Removal of Cyanide from Solutions Using Ultrasonic Waves.” Desalination, 216: 209 –
221, 2007

242
Yazu, K., Makino, M. and Ukegawa, K. “Oxidative Desulfurization of Diesel Oil with
Hydrogen Peroxide in the Presence of Acid Catalyst in Diesel Oil/Acetic Acid Biphasic
System.” Chemistry Letters, 33(10): 1306 – 1307, 2004
Yen, T. “Correlation between Heavy Crude Sources and Types and Their Refining and
Upgrading Methods.” Petroleum Industry Press, 2: 2137 – 2144, 1998
Yen, T. “Structure of Petroleum Asphaltene and Its Significance.” Energy Sources A, 1(4):
447 – 463, 1974
Yen, T, Mei, H. and Lu, S. “Oxidative Desulfurization of Fossil Fuels with Ultrasound.”
U. S. Patent Application, 2000
Young, F. Sonoluminescence. Florida: CRC Press, 2005
Yu, B., Xu, P., Shi, Q. and Ma, C. “Deep Desulfurization of Diesel Oil and Crude Oils by
a Newly Isolated Rhodococcus erythropolis Strain.” Appl. Environ. Microbiol., 72(1): 54-
58, 2006
Zhang, S., Zhang, Q. and Zhang, Z. “Extractive Desulfurization and Denitrogenation of
Fuels Using Ionic Liquids.” Ind. Eng. Chem. Res., 43: 614 – 622, 2004
Zhao, D., Liao, Y. and Zhang, Z. “Toxicity of Ionic Liquids.” Clean – Soil, Air, Water,
35(1): 42 – 48, 2007
Zapata, B., Pedraza, F. and Valenzuela, M. “Catalyst Screen for Oxidative
Desulfurization using Hydrogen Peroxide.” Catalysis Today, 106: 219 – 221, 2005
Zhou, L. Adsorption: Progress in Fundamental and Application Research. New Jersey:
World Scientific Publishing Co., 2007

Abstract (if available)

Abstract This study is aimed at improving the current ultrasound assisted oxidative desulfurization (UAOD) process by utilizing superoxide radical as oxidant. Research was also conducted to investigate the feasibility of ultraviolet (UV) irradiation-assisted desulfurization. These modifications can enhance the process with the following achievements:•Meet the upcoming sulfur standards on various fuels including diesel fuel oils and residual oils•More efficient oxidant with significantly lower consumption in accordance with stoichiometry•Energy saving by 90%•Greater selectivity in petroleum composition

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Ultra clean fuels via modified UAOD process with room temperature ionic liquid (RTIL) & solid catalyst polishing

PDF

Selective adsorption in ultrasound assisted oxidative desulfurization process with nano-engineered adsorbents: mechanism and characterization

Asset Metadata

Creator Chan, Ngo Yeung (author)

Core Title Superoxide radical and UV irradiation in ultrasound assisted oxidative desulfurization (UAOD): a potential alternative for green fuels

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Civil Engineering (Environmental Engineering)

Publication Date 05/04/2010

Defense Date 03/26/2010

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag green fuel,OAI-PMH Harvest,oxidative desulfurization,radical,Sulfur,superoxide,UAOD

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Pirbazari, Massoud (committee chair), Lee, Jiin-Jen (committee member), Shing, Katherine S. (committee member)

Creator Email ngochan@gmail.com,ngochan@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2992

Unique identifier UC1320732

Identifier etd-Chan-3701 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-316845 (legacy record id),usctheses-m2992 (legacy record id)

Legacy Identifier etd-Chan-3701.pdf

Dmrecord 316845

Document Type Dissertation

Rights Chan, Ngo Yeung

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu