DEVELOPMENT OF AN ALTERNATIVE METHOD FOR THE DESULFURIZATION
OF FLUID CATALYTICALLY CRACKED GASOLINE
by
Jennifer Anne Hendsbee
B.Sc., University o f Northern British Columbia, 2003

THESIS SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
in
NATURAL RESOURCES AND ENVIRONMENTAL STUDIES
(ENVIRONMENTAL SCIENCE)

THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA
March 2006

Jennifer Anne Hendsbee, 2006

1^1

Library and
Archives Canada

Bibliothèque et
Archives Canada

Published Heritage
Branch

Direction du
Patrimoine de l'édition

395 W ellington Street
O ttaw a ON K1A 0N4
C anada

395, rue W ellington
Ottawa ON K1A 0N4
C an a d a
Your file Votre référence
ISBN: 978-0-494-28396-7
Our file Notre référence
ISBN: 978-0-494-28396-7

NOTICE:
The author has granted a non­
exclusive license allowing Library
and Archives Canada to reproduce,
publish, archive, preserve, conserve,
communicate to the public by
telecommunication or on the Internet,
loan, distribute and sell theses
worldwide, for commercial or non­
commercial purposes, in microform,
paper, electronic and/or any other
formats.

AVIS:
L'auteur a accordé une licence non exclusive
permettant à la Bibliothèque et Archives
Canada de reproduire, publier, archiver,
sauvegarder, conserver, transmettre au public
par télécommunication ou par l'Internet, prêter,
distribuer et vendre des thèses partout dans
le monde, à des fins commerciales ou autres,
sur support microforme, papier, électronique
et/ou autres formats.

The author retains copyright
ownership and moral rights in
this thesis. Neither the thesis
nor substantial extracts from it
may be printed or otherwise
reproduced without the author's
permission.

L'auteur conserve la propriété du droit d'auteur
et des droits moraux qui protège cette thèse.
Ni la thèse ni des extraits substantiels de
celle-ci ne doivent être imprimés ou autrement
reproduits sans son autorisation.

In compliance with the Canadian
Privacy Act some supporting
forms may have been removed
from this thesis.

Conformément à la loi canadienne
sur la protection de la vie privée,
quelques formulaires secondaires
ont été enlevés de cette thèse.

While these forms may be included
in the document page count,
their removal does not represent
any loss of content from the
thesis.

Bien que ces formulaires
aient inclus dans la pagination,
il n'y aura aucun contenu manquant.

Canada

Abstract
Canada recently developed new regulations for the amount of total sulâir in fuel, as of
January 1^, 2005, thereby limiting the total sulfur to 30 ppmw. Current teclinology for the
removal of sulfur involves hydrotreating, is an expensive process, which may reduce the value
o f fossil fuels when operated under deep desulfurization conditions. This thesis reports on
experiments with a variety o f solvents, metal salts, acids, and bases which were unsuccessful
at removing sulfur-containing compounds from FCC gasoline. The thesis also reports on the
success achieved through the use o f Raney nickel. Simply by varying the reaction tune and
temperature, a maximum reduction of 99 % in the sulfur concentration was obtained using
Raney nickel. Finally, experiments were performed that recovered the nickel as nickel oxide.
This allows for further processing in order to convert the nickel oxide into nickel aluminum
alloy.

Table of Contents
ABSTRACT................................................................................................................................. II
TABLE OF CONTENTS.........................................................................................................Ill
LIST OF TABLES.....................................................................................................................VI
LIST OF FIGURES................................................................................................................. VII
NOMENCLATURE..................................................................................................................IX
ACKNOWLEDGEMENT......................................................................................................... X
CHAPTER 1: INTRODUCTION............................................................................................. 1
1.1 I n t r o d u c t io n ...............................................................................................................................................1
1.2 E n v ir o n m e n t a l H a z a r d s a n d C a t a l y s t P o is o n in g P o t e n t ia l o f S u l f u r ..............1

1.2.1 Acid Rain...................................................................................................................... 2
1.2.2 Health Risks.................................................................................................................. 3
1.2.3 Catalyst Poisoning....................................................................................................... 4
1.3 C a n a d a ’s N e w S u l f u r R e g u l a t io n s f o r G a s o l i n e ............................................................. 5
1.4 F l u id C a t a l y t ic C r a c k in g P r o c e s s .............................................................................................. 6
1.4.1 FCC Catalysts............................................................................................................... 7

1.4.2 FCC Reactions..............................................................................................................8
1.4.3 FCC Reactor Types.................................................................................................... 10
1.5 C u r r e n t S u l f u r P r o c e s s in g T e c h n iq u e s ................................................................................ 11

1.5.1 Amine Processing Units............................................................................................ 12
1.5.2 Wet Sulfur Scrubbers.................................................................................................13
1.5.3 The Merox Process....................................................................................................14
1.5.4 The Claus Process...................................................................................................... 15
1.5.5 Hydrotreating.............................................................................................................16
1.5.6 Limestone Fluidized Beds.......................................................................................... 19
1.6 A n a l y t ic a l M e t h o d s .......................................................................................................................... 20

1.6.1 Gas Chromatography................................................................................................ 21
1.6.2 Ultraviolet Fluorescence........................................................................................... 23
1.6.3 Energy Dispersive X-Ray Fluorescence Spectrometry........................................... 24
1.6.4 Inductively Coupled Plasma..................................................................................... 25
1.7 S o l v e n t E x t r a c t io n a n d S o l u b il it y P a r a m e t e r s ............................................................26

1.7.1 Solvent Extraction Methods......................................................................................26
1.7.2 Solubility Parameter and Solvent Relationship....................................................... 2 7
1.8 S o r b e n t E x t r a c t i o n ........................................................................................................................... 29
1.9 R a n e y N ic k e l ........................................................................................................................................... 2 9

1.9.1 Activation o f Raney Nickel........................................................................................30
1.9.2 Structure..................................................................................................................... 32
1.9.3 Desulfurization with Raney nickel............................................................................32

ill

1.10 O b j e c t iv e s o f R e s e a r c h .................................................................................................................. 33
1.11 C o n c l u s i o n ............................................................................................................................................. 34
1.12 L it e r a t u r e C i t e d ................................................................................................................................. 35

CHAPTER 2: CHEMICAL METHODS STUDIED FOR THE DESULFURIZATION
OF FCC GASOLINE................................................................................................................41
2.1 I n t r o d u c t io n ............................................................................................................................................41

2.1.1 Solvent Desulfurization............................................................................................. 42
2.1.2 Adsorption and Precipitate Desulfurization........................................................... 42
2.1.3 Desulfurization using basic and acidic solutions....................................................44
2 .2 M e t h o d o l o g y .......................................................................................................................................... 45

2.2.1 FCC Gasoline Sampling and Storage...................................................................... 45
2.2.2 Solvent Methodology..................................................................................................46
2.2.3 Sorbent Methodology.................................................................................................48
2.2.4 Basic and Acidic Methodology................................................................................. 49
2.3 R e s u l t s ....................................................................................................................................................... 4 9

2.3.1 Solvent Results............................................................................................................49
2.3.2 Sorbent Results.......................................................................................................... 56
2.3.3 Basic and Acidic Results........................................................................................... 57
2 .4 D i s c u s s i o n ................................................................................................................................................. 58

2.4.1 Solvent Desulfurization............................................................................................. 58
2.4.2 Adsorption and Precipitate Desulfurization........................................................... 61
2.4.3 Desulfurization using basic and acidic solutions....................................................62
2.5 C o n c l u s io n ................................................................................................................................................63
2 .6 L it e r a t u r e C it e d ....................................................................................................................................65

CHAPTER 3: ACTIVATION METHODS FOR RANEY NICKEL IN THE
DESULFURIZATION OF GASOLINES..............................................................................67
3.1 I n t r o d u c t io n ............................................................................................................................................67

3.1.1 Activation o f Raney Nickel........................................................................................69
3.1.2 Structure..................................................................................................................... 69
3.1.3 Desulfurization with Raney nickel............................................................................ 70
3 .2 M e t h o d o l o g y .......................................................................................................................................... 70

3.2.1 FCC Gasoline Sampling and Storage.......................................................................70
3.2.2 Raney Nickel Activation............................................................................................ 71
3.2.3 Methodology fo r the Amount o f Raney Nickel........................................................ 73
3.2.4 Methodology o f Desulfurization o f FCC Gasoline..................................................73
3.2.5 Repeatability o f Method 5 ......................................................................................... 74
3.3 R e s u l t s ........................................................................................................................................................75

3.3.1 Results o f the Six Different Activation Methods...................................................... 75
3.3.2 Results fo r the Amount Nickel-Aluminum Alloy......................................................81
3.3.3 Repeatability o f Method 5 .........................................................................................81
3 .4 D i s c u s s i o n ................................................................................................................................................. 82

3.4.1 Raney Nickel Activation............................................................................................ 82
3.4.2 Raney Nickel Amount.................................................................................................84
3.4.3 Repeatability...............................................................................................................84
IV

3 .5 C o n c l u s io n ................................................................................................................................................84
3 .6 L it e r a t u r e C i t e d ....................................................................................................................................86

CHAPTER 4: OPTIMIZATION OF THE REACTION CONDITIONS FOR RANEY
NICKEL FOR THE DESULFURIZATION OF FCC GASOLINE................................. 88
4.1 I n t r o d u c t io n ............................................................................................................................................88
4 .2 M e t h o d o l o g y .......................................................................................................................................... 91

4.2.1 FCC Gasoline Sampling and Storage...................................................................... 91
4.2.2 Preparation o f the Raney Nickel.............................................................................. 91
4.2.3 Time Trial Experiments............................................................................................. 92
4.2.4 Temperature Trial Experiments................................................................................ 92
4.2.5 Sequential Addition Runs.......................................................................................... 93
4.2.6 Method o f Analysis....................................................................................................94
4.2.7 Statistical Analysis..................................................................................................... 94
4.2.8 Hydrocarbon Analysis............................................................................................... 95
4.2.9 Recovery o f Nickel.....................................................................................................95
4.3 R e s u l t s ....................................................................................................................................................... 96

4.3.1 Time Trials................................................................................................................. 96
4.3.2 Temperature Trials....................................................................................................97
4.3.3 Sequential Addition Runs.......................................................................................... 98
4.3.4 Hydrocarbon Analysis Results.................................................................................. 98
4.3.5 Recovery o f Nickel...................................................................................................101
4 .4 D i s c u s s i o n ...............................................................................................................................................102

4.4.1 Effect o f Time on the Removal o f Sulfur................................................................ 102
4.4.2 Effect o f Temperature on the Removal o f Sulfur...................................................104
4.4.3 Effect o f Sequential Addition on the Removal o f Sulfur.......................................104
4.4.4 Hydrocarbon Components Analysis...................................................................... 105
4.4.5 Recovery o f Nickel fo r Possible Re-Use................................................................106
4 .5 C o n c l u s i o n s ...........................................................................................................................................107
4 .6 L it e r a t u r e C i t e d ................................................................................................................................. 109

CHAPTER 5: CONCLUSION.............................................................................................. I l l

List of Tables
T a b l e 1.1: W 1 -W 8 R a n e y n ic k e l p r e p a r a t io n m e t h o d s (A u g u s t in e , 1 9 9 6 )....................31
T a b l e 2.1: T e n S o l v e n t s s e l e c t e d b a s e d o n s o l u b i l i t y p a r a m e t e r s s i m i l a r t o f i v e
S e l e c t e d s u l f u r c o m p o u n d s ( B a r t o n , 1 9 9 1 )...........................................................................50
T a b l e 2.2: S u m m a r y o f t h e s e p a r a t io n b e t w e e n t h e s o l v e n t a n d FCC g a s o l in e . ... 51
T a b l e 2.3: S u m m a r y t a b l e o f t h e f o u r l e v e l s o f r in s in g p e r f o r m e d o n a s a m p l e
e x t r a c t e d w it h 10 % SULFURIC ACID.............................................................................................. 58
T a b l e 3.1: S u m m a r y o f R a n e y N i c k e l A c t i v a t i o n M e t h o d s ...................................................75
T a b l e 3.2: A N O V A t a b l e f o r t h e C o m p a r is o n o f S ix D if f e r e n t A c t iv a t io n M e t h o d s .

.............................................................................................................................................. 76
T a b l e 3.3: S u m m a r y o f t h e D i f f e r e n c e in D e s u l f u r i z a t i o n B a s e d o n t h e A m o u n t o f
N i - A l A l l o y ................................................................................................................................................. 81
T a b l e 3.4: S u m m a r y S t a t is t ic s f o r 15 I d e n t ic a l S a m p l e s ....................................................... 82
T a b l e 4.1: C o m p a r i s o n o f h y d r o c a r b o n c o m p o n e n t c o n c e n t r a t i o n s f o r a v a r i e t y
OF TIME INTERVALS ......................................................................................................................................99

T a b l e 4.2: C o m p a r i s o n o f h y d r o c a r b o n c o m p o n e n t c o n c e n t r a t i o n s f o r a v a r i e t y
OF t e m p e r a t u r e s ..................................................................................................................................... 100
T a b l e 4.3: C o m p a r is o n o f h y d r o c a r b o n c o m p o n e n t c o n c e n t r a t io n s f o r s e q u e n t ia l
ADDITION ....................................................................................................................................................... 100

VI

List of Figures
F ig u r e 1.1: I l l u s t r a t io n o l s il ic a / a l u m in a c a t a l y s t w it h b o t h t h e B r 0N s t l d a c id
(LEFT SIDE) AND LEWIS ACID (RIGHT SIDE) SITES (CAMPBELL, 1 9 8 8 )........................................7
F ig u r e 1.2: U O P s t r a ig h t - r is e r F C C u n it f l o w c h a r t (M e y e r s , 2 0 0 4 )...............................11
F ig u r e 1.3: S c h e m a t ic o f f u e l g a s a m in e t r e a t in g (M e y e r s , 2 0 0 4 )...................................... 12
F ig u r e 1.4: A m in e R e g e n e r a t io n U n it (M e y e r s , 2 0 0 4 )..................................................................13
F ig u r e 1.5: W e t S c r u b b e r a b s o r b e r v e s s e l / s p r a y t o w e r (M e y e r s , 2 0 0 4 )....................... 14
F ig u r e 1.6: M e r o x m e r c a p t a n - e x t r a c t io n u n it (M e y e r s , 2 0 0 4 )............................................15
F ig u r e 1.7: T w o - s t a g e C l a u s u n it (M e y e r s , 2 0 0 4 ) ..........................................................................15
F ig u r e 1.8: H y d r o t r e a t in g f l o w s c h e m e (M e y e r s , 2 0 0 4 ) ...........................................................17
F ig u r e 2.1 : G C /F ID a n a l y s is o f o r ig in a l F C C g a s o l in e ............................................................. 53
F ig u r e 2.2 : A n a l y s is o f F C C g a s o l in e e x t r a c t e d w it h 1- d e c a n o l u s in g G C /F lD

54

F ig u r e 3.1 : S E M im a g e o f R a n e y n ic k e l p r e p a r e d u s in g M e t h o d 2 (2 0 °C )..................... 77
F ig u r e 3.2: S E M im a g e o f R a n e y n ic k e l p r e p a r e d u s in g M e t h o d 5 (8 0 °C )..................... 78
F ig u r e 3.3: S E M im a g e o f R a n e y n ic k e l p r e p a r e d u s in g M e t h o d 6 (1 0 0 °C )...................78
F ig u r e 3.4: E l e m e n t a l A n a l y s is o f R a n e y n ic k e l p r e p a r e d u s in g m e t h o d 2 (2 0 °C ).7 9
F ig u r e 3.5: E l e m e n t a l A n a l y s is o f R a n e y n ic k e l p r e p a r e d u s in g m e t h o d 5 (8 0 °C ).8 0
F i g u r e 3.6: E l e m e n t a l A n a l y s i s o f R a n e y n i c k e l p r e p a r e d u s i n g m e t h o d 6 (1 0 0 °C).

80
F ig u r e 3.7: D e t e r m in a t io n o f S u l f u r C o n c e n t r a t io n in 15 U n iq u e l y P r e p a r e d
S a m p l e s o f R a n e y N ic k e l U s in g A c t iv a t io n M e t h o d 5 a n d 3 .0 g r a m s o f N i -A l
A llo y .............................................................................................................................................................. 82
F ig u r e 4.1 : R e f l u x c o n d e n s e r a p p a r a t u s f o r t e m p e r a t u r e e x p e r im e n t s : a .
CONDENSER, B. ROUND BOTTOM FLASK, C. THERMOMETER....................................................93
F ig u r e 4.2: S u l f u r C o n c e n t r a t io n in F C C G a s o l in e w it h T im e o f T r e a t m e n t a t 2 0 °C.
........................................................................................................................................................................... 97

YU

F ig u r e 4.3: E f f e c t o f T e m p e r a t u r e o n t h e r a t e o f R e d u c t io n in s u l f u r
CONCENTRATION IN FCC GASOLINE OVER 30 MINUTES OF REACTION TIME.......................... 98

F i g u r e 4.4: X R D d i f f r a c t i o n p a t t e r n f o r s p e n t R a n e y n i c k e l r o a s t e d a t 800 °C f o r 6
HRS a n d NiO i n d e x e d d i f f r a c t i o n p a t t e r n ............................................................................. 102

Vlll

Nomenclature
AMU

Aminé Processing Unit

Co(Ni)Mo

Cobalt/NickeFMoIybdenum imbedded into a zeolite catalyst
Carbonium Ion

CEPA

Canadian Environmental Protection Act

FCC

Fluid Catalytieally Cracked

FID

Flame Ionization Detector

GC

Gas Chromatography

LPG

Liquefied Petroleum Gas

Ni-Al

Nickel Aluminum Alloy

Nl-H2Raney

Raney Nickel

ppmw

Parts per Million by Weight

SCO

Sulfur Chemiluminescence Detector

SEM

Scanning Electron Microscope

UVF

Ultraviolet Fluorescence

Vol %

Volume Percent

Wt%

Weight Percent

XRF

X-ray Fluorescence

IX

Acknowledgement
Firstly, I would like to thank my supervisor, Dr. Ron Thring, for his support
throughout my graduate studies at UNBC. Also, I am grateful for the financial support he
provided me from his Natural Sciences and Engineering Research Council o f Canada
(NSERC) Research Grant. 1 would also like to express my gratitude to the other members of
my supervisory committee, namely. Dr. David Dick and Mr. Cory Sieben, whom, despite
their hectic schedules, were both more than willing to meet with me whenever the need arose.
1 would like to acknowledge the Prince George Husky Refinery for their collaboration on this
project, and especially the laboratory staff, namely, Mr. Ken Gallant, Mr. Brad Hensby and
Ms. Katja Otting, for their teehnieal assistance.
In addition, Iwould like to thank Core Laboratories in Calgary for their prompt work
in analysing our samples and reporting the data.
Lastly, I wish to express my appreciation and gratitude to my family and friends for all
their love and support, especially during the past two years of excitement and frustration.

Chapter 1: Introduction
1.1 Introduction
The environmental impaets of the oil and gas seetor are becoming mereasingly
important. The reduction o f sulfur requires new technology. As of January

2005

Canada’s aim was to have the sulfur levels down to 30 parts per million by weight (ppmw) for
gasoline compared to the old regulation of 150 ppmw (CEPA, 2004). In order to meet
increasingly strict regulations there have been several methods developed for the removal of
sulfur from fossil fuels. The following is a brief literature review of the environmental risks,
the important current technologies used, analytical chemical methods and the approach this
thesis investigates.
1.2 Environmental Hazards and Catalvst Poisoning Potential of Sulfur
There are several environmental hazards that sulfur in fossil fuels can create both in the
environment and during processing. The combustion products of sulfur are known to cause
much environmental damage and are also poisonous to processing catalysts and catalytic
converters in vehicles (Ertl et al., 1999). It is estimated that over 100 million tonnes of sulfur
enter the atmosphere per year from human activities alone (Kennedy, 1992). Natural sources
of sulfur include hydrogen sulfide (H2 S) from decaying organic matter and sulphate respiration
(Kennedy, 1992). In addition, the burning of fossil fuels causes sulfur-containing compounds
within the fossil fuels to become oxidized into sulfur dioxide, which is the leading cause of
acid rain (Kennedy, 1992).

1.2.1 Acid Rain
The production of sulfur dioxide leads to the creation of acid rain, through two
mechanisms: aqueous phase oxidation and gas phase oxidation (Schwedt, 2001). In gas phase
oxidation sulfur dioxide is being oxidized to SO3 then forming sulfurons acid in either a gas or
liquid phase (Schwedt, 2001). In the aqueous phase sulfur dioxide can be oxidized by
hydrogen peroxide. The hydrogen peroxide is minor part of the atmosphere, which is formed
by the disproportionation of HO2 radicals (Bunce, 1994). The aqueous phase oxidation
involves sulfur dioxide being hydrolyzed into sulfurons acid, and then it is oxidized further to
sulfuric acid (Schwedt, 2001). These two mechanisms have several steps and typically take
place in cloud droplets, producing acid rain. The reactions for formation of acid rain through
aqueous phase oxidation can be seen in the following reactions 1.1, 1.2, and 1.3, whereas the
reactions for gas phase oxidation can be seen in reactions 1.4 and 1.5 (Bunce, 1994):
2iT02

^2(g)

(^'^)

(1.4)

(1.5)
The acid rain produced has several impacts on the forest and lake ecosystems. Plants
are impacted especially when they adsorb sulfur trioxide as a gas (Kennedy, 1992). This
causes mineral acids to form on the cytoplasm. Another major problem can occur if soils are
exposed to continued acidification from acid rain. This continuous acidification can cause an

imbalance in the natural buffering capacity of a soil causing serious pH declines in the soils
(Kennedy, 1992). This pH decline can lead to weakened vegetation, which is more
susceptible to pathogen and insect attack, crown thinning from leaf or needle loss, nutrient
deficiencies, changes in branching habit, and decline in radial growth (Kennedy, 1992). The
affects of pH decline eventually lead to tree death and forest decline (Kennedy, 1992).
The acidification o f lakes affects the lake biota, ft has been found that at a pH of 5.8,
lake trout were eliminated in some cases (Kennedy, 1992). At a pH below 6 the populations
of small animals (shrimp, etc) can decrease and several plant species are substituted for other,
hardier species (Kennedy, 1992). Therefore, acidification of lakes causes a reduction in the
diversity o f both plant and animal species.
1.2.2 Health Risks
Besides acidification of forest soils and lakes, particulate matter and soil or water
contamination with thiophenes can be a serious health risk. Also, sulfur dioxide is a gas,
which can adsorb to fine particulates in the atmosphere and can irritate respiratory tracts
(Schwedt, 2001). Sulfur dioxide is typically removed from flue gases directly after
combustion. This is typically achieved through the addition of calcium hydroxide (Ca(0H)2),
calcium carbonate (CaCOa) or activated charcoal as seen in reaction 1.6 (Schwedt, 2001).
However, not all sulfur dioxide is removed in this process as the reaction is quite temperature
sensitive (Schwedt, 2001).

+

( 1.6)

There have been recent studies reported on the toxicities of thiophenes in the
environment. These studies particularly have been looking into the toxicological effects of
thiophenes in petroleum mixtures. The research is ongoing but it has been found that some of
the higher molecular weight thiophenes have mutagenic and carcinogenic potential (Kropp
and Fedorak, 1998). However, more research is needed on this topic to determine
degradation potential of the thiophenes and the ability o f biological organisms or the
environment to degrade organic suhur compounds. Also, more research is necessary to
determine the specific health effects of these contaminants.
1.2.3 Catalyst Poisoning
Finally, the poisoning o f catalysts in refineries and vehicles causes many problems. In
refineries, the poisoning of refinery catalysts causes many processing problems. For example
some sulfur compounds, such as hydrogen sulfide and sulfur dioxide, can be corrosive to
equipment and damage or reduce activity of processing catalysts, which is similar to the
affects seen in catalytic converters of vehicles (Krishnan and Sotirchos, 1994).
The poisoning of catalysts in vehicles involves the ability o f sulfur oxides to reduce the
activity of wash coat oxides such as aluminum oxides and cesium oxides that surround the
noble metal portion o f the catalyst (Ertl et al., 1999). Adsorption and/or chemisorption of
sulfur oxides on to wash coat oxides can occur at low temperatures, deactivating the catalysts
(Ertl et ah, 1999). Oxides of sulfur also have the ability to poison the noble metal portion of
the catalyst in the catalytic converters (Dupain et al., 2003). However, this reaction can
usually be reversed once the catalyst is operating at a higher temperature.

Catalyst deactivation in vehicles can also occur from the precious metals being
poisoned by sulfur oxides (Ertl et al., 1999). The sulfur oxides come from the combustion of
sulfur-containing compounds in the fuel. The result of the catalyst poisoning is a decrease in
catalyst activity for the reactions, which destroy carbon monoxide and nitric oxides. The
reactions involving the elimination of carbon monoxide and nitric oxides can be seen in
reactions 1.7, 1.8, 1.9, and 1.10 (Ertl et al., 1999). This will reduce a vehicle’s pollution
control efficiency and increase other, potentially more harmful, pollutants entering the
atmosphere (Gokeler et a l, 2002).

(1.7)

CO-k

CO, -k

(1.8)

C0 + # 0 -> l# 2 + C 0 2

(1.9)

77, + # 0 ^ ^ # 2 + 77,0

( 1. 10)

1.3 Canada’s New Suhur Regulations for Gasoline
With the variety o f environmental and health concerns around suhur, the regulation’s
of sulfur in gasoline fall under the Canadian Environmental Protection Act (CEPA). The
CEPA was revised on April 30**' 2004, to include new suhur regulations (CEPA, 2004). The
new regulations stated that as of January E* 2005, sulfur levels in gasoline would need to be
below 30 ppmw when calculated using a pool average (CEPA, 2004). The CEPA states that
the pool average is the volume-weighted average concentration of suhur in gasoline produced
at that refinery. This means, for example, that if the fluid catalytieally cracked (FCC) gasoline

stream makes up 40% of the total gasoline produeed at that refinery then it will contribute
40% o f the sulfur concentration to the final product. Therefore, the FCC gasoline can have a
sulfur concentration higher than 30 ppmw, as long as when it is added to the rest of the
gasoline produeed at the refinery the concentration is below 30 ppmw. The new Canadian
regulations and regulations around the world result from the growing concern over sulfur in
fossil fuels and are leading towards suhur-ffee fuels in the future.
1.4 Fluid Catalvtic Cracking Process
The idea o f fluid catalytic cracking is to crack lower-value, higher molecular weight
stocks to produce higher-value, lower molecular weight products (Matar and Flateh, 2001).
These higher value products are typically gasoline, distillates and C 3 / C 4 olefins (Meyers,
2004). The main feeds for catalytic cracking are heavier refinery streams and excess refinery
gas oils which typically contain higher concentrations of basic and polar molecules and
asphaltenes (Matar and Flateh, 2001). The crackability of any one of these feeds is a function
of the proportions of paraffmic, naphthenie, and aromatic species (Meyers, 2004). Three
basic process functions affect the product yield of FCC units: operating conditions, feedstock
properties and catalyst characteristics (Sertie-Bionda et ah, 2000).
Due to the strict regulations surrounding suhur, it is important to determine the most
contaminated fuels. The gasoline from reforming or isomerization units is typically made from
distillate feeds, which contain almost no suhur (Leflaive et ah, 2002). However, the FCC
feedstock is a combination of a variety of heavier feeds, which contain between 0.5 - 1.5 wt
percent of sulfur (Leflaive et a l, 2002). The amount of sulfur in the feed is a result of the
geographical origin o f the crude oil (Meyers, 2004). However, Leflaive et al (2002)

determined that in the presence o f H2 S, olefms and diolefms the FCC catalyst could transform
these compounds into thiophenic compounds. Also, long alkyl chain thiophenes can be
cracked by FCC catalysts into thiophene and short alkyl chain thiophenes, which are then
stable under FCC conditions (Leflaive et al., 2002). Therefore, as FCC gasoline typically
makes up between 30-40% of the gasoline pool, it is important to reduce its sulfur
concentration as much as possible (Leflaive et al., 2002).
1.4.1 FCC Catalysts
FCC units continually circulate a zeolite catalyst in a fluidized bed reactor, which
rapidly reacts in the vapour phase with the feed (Meyers, 2004). The catalysts used in
cracking have improved from synthetic amorphous silica/alumina to silica/alumina catalysts
with incorporated zeolites (Matar and Flateh, 2001). These zeolite catalysts have both Lewis
and Br0nsted acid sites and the presence of holes in the crystal lattice which make them
superior to amorphous silica-alumina catalysts (Matar and Hatch, 2001). An example of both
the Brensted acid (left side) and the Lewis acid (right side) form of silica-alumina can be seen
in Figure 1.1.

” ''P e

Oe''“

H

Figure 1.1: Illustration of silica/alumina catalyst with both the Bronsted acid (left side)
and Lewis acid (right side) sites (Campbell, 1988).

The Lewis acid sites are a result of the bonding of the silica and alumina structure.
The other potential site is a Br0 nsted acid site. The Br0nsted acid site can be created through
the interaction of a Lewis acid site with a hydroxide ion (Campbell, 1988). This creates a
negative charge on the aluminum atom while the proton can bond with an oxygen atom
forming a partial bond (Campbell, 1988). This partial bonding of the hydrogen atom allows it
to be easily donated thus creating a Bronsted acid site (Campbell, 1988). The greater acid site
density and higher adsorption power along with smaller pores creates higher activity and
better selectivity (Matar and Hatch, 2001). Lewis acid and Bronsted acid sites are locations
for the reactions in the FCC unit.
1.4.2 FCC Reactions
There have been several attempts to model the kinetic reactions occurring in FCC
units; this has become important in order to optimize fuel processing in the FCC unit (Dupain
et al., 2003). This is especially important as it is not feasible to change the process conditions
in a commercial FCC unit in order to test each scenario (Pareek et al., 2003). There are three
main reactions produced during catalytic cracking, which will produce carbonium ions. The
carbonium ions are more selective towards specific bonds than free radicals, which are formed
in thermal cracking and cause random bond breaking (Meyers, 2004). As shown below, the
first reaction involves the abstraction of a hydride ion by a Lewis acid site (Matar and Hatch,
2001 );

^

A

0

The second possible reaction is between a Br0nsted acid site and an olefin as seen in
reaction 1.12 (Matar and Hatch, 2001):

Vv""
v\/\,

yy""
y\A/

Finally, in reaction 1.13, the reaction of a carbonium ion formed in either of the above
equations with another abstraction o f a hydride ion (Matar and Hatch, 2001):
+ RCH^CH, - ^ R H + RC^HCH, (1.13)
Once these carbonium ions are formed they can react in four potential ways. The
molecules containing the C ion can crack into smaller molecules, react with another molecule,
isomerize into a different form or react with the catalyst to stop the reaction (Meyers, 2004).
It is most important to note that the reactions proceed to form the most stable carbonium ion,
so isomerization of secondary to tertiary carbonium ions occurs frequently (Meyers, 2004).
These reactions occur when the feedstock contacts the hot regenerated catalyst
(Meyers, 2004). The feedstock becomes vaporized and is converted by the catalyst through
the above reaetions to lower boiling fractions, such as gasoline, light cycle oil and dry gas
(Meyers, 2004). As suhur is the focus of this thesis it is important to note that sulfur
compounds do not affect the crackability o f a feed, however the cracked sulfur compounds
show up in the liquid products. Sulfur also exits the FCC units as HiS and suhur oxides
causing air pollution problems if not captured (Meyers, 2004).

1.4.3 FCC Reactor Types
These reaetions typieally oeeur in one of two main types of reaetors: fluidized bed or
moving bed. The fluidized bed is more eommon and involves the eatalyst typieally being a
porous powder with an average particle size of 60 pm (Matar and Hatch, 2001). In this
process the feed is preheated and held at a temperature of about 450 - 520 °C. A concurrent
upward flow of heated feed and hot regenerated catalyst occurs in the riser with an
approximate pressure o f 10 - 20 psig (Matar and Hatch, 2001). The vapours are then
separated irom the catalyst using cyclones, which are passed into a ffactionator for separation
o f different product streams (Matar and Hatch, 2001). The separated catalysts are
regenerated due to coke deposits forming on the catalysts surface during FCC thus making it
less active (Meyers, 2004). Typically, the catalysts are combusted with air to strip the coke
from the catalyst surface (Meyers, 2004). This regenerated catalyst is then returned to the
riser to go through the process again. In the moving bed process, the eatalyst is in the form of
beads rather than a powder. These beads descend through the feed by gravity action to the
regeneration zone (Matar and Hatch, 2001). A general schematic of a straight - riser unit can
be seen in Figure 1.2.

10

Flue Gas to
C O B o lk f

Gas and
Gasotiné

I
Pressurefteducing
Chamber

Reactor

^ U g N C y c le
141
Catalyst
^ Z ^ Stripper

Fractlonator
Heavy Cycle
il
Clarified Oil
Slurry
Settler

Regenerator

I

Charge

Figure 1.2: UOP straight-riser FCC unit flow chart (Meyers, 2004).
1.5 Current Sulfur Processing Techniques
There are currently several commercial processes to remove sulfur from refinery feeds
either before they reach the FCC unit or after they have been treated in the FCC unit. Most of
these processes are designed for the recovery of H2 S and mercaptans. There are very few
processes that remove the heavier sulfur compounds such as thiophenes. The many processes
for H 2 S and mercaptan recovery include amine-processing units, wet sulfur scrubbers, Claus
process, Merox process and limestone beds. One of the only processes in operation at
refineries for the removal of other sulfur species is hydrotreating. While hydrotreating has the
ability to reduce sulfur concentrations to the new regulation levels, the problem is that with
deep desulfurization, the production of light hydrocarbons can increase the amount of
hydrogen consumed and decrease the yield of liquid fuel (Kemsley, 2003). Also, the olefins,
which are responsible for the high octane number, can be destroyed decreasing the octane

11

value of the gasoline causing more additives to be needed (Kemsley, 2003). A brief overview
of these methods follows.
1.5.1 Amine Processing Units
The amine processing units (AMU) remove H 2 S from the recycled gas streams and
fuel gas/liquefied petroleum gas (LPG), which is produeed during processing of crude oil
(Meyers, 2004). The amines selected for these processes are typieally monoethanolamine
(MEA), diethanolamine (DBA), or methyldiethanolamine (MDEA). Typieally, DBA is used in
a 25 to 33 wt % solution in water (Meyers, 2004).

SO U R FUEL GAS

GAS
RECOVERY
UNIT

AMINE
TREATING

REFINERY
FUEL G AS

AMINE
TREATING

MERCAPTAN
REMOVAL

NAPHTHA

SO UR LPG

LPG
PRODUCT

"MEROX
" MERICHEM

Figure 1.3: Schematic of fuel gas amine treating (Meyers, 2004).
The process for amine treating involves multiple amine absorbers with a common
amine regeneration unit (Meyers, 2004). A circulating amine stream removes the H2 S from
the recycled gas stream and the off-gases from the EPG recovery units as seen above in Figure
1.3 (Meyers, 2004). The contaminated amine then flows to the regenerator where steam
strips the H 2 S from the amine as seen in Figure 1.4. The amine is then cooled before being
returned to the absorbers (Meyers, 2004). This stripping procedure produces sour water.

12

which is further treated with steam to vaporize the H 2 S, this vaporized H 2 S flows to the suhur
plant where the Claus process concentrates and oxidizes it into elemental sulfur (Meyers,
2004). The elean water is then cooled in order to be reused in the refinery.

TREATED
STREAM

CONDENSER
COOLER

À

/

\

TO
SRU

I, ^(iACCUMl)

O;. CARBON
^
BED
FILTER

^ I ^^PURQE
FILTER

SOURU

\_

ABSORBERS

REGEN

1 2 7 '° "

F E E lT i

IY

GAS

TO FUEL GAS

REBOILER

TANK
FLASH
DRUM

HYDROCARBONS

Figure 1.4: Amine Regeneration Unit (Meyers, 2004).
1.5.2 Wet Sulfur Scrubbers
Suhur scrubbers are used to remove 80% from the air emissions (Meyers, 2004). The
SOx emissions come from the sulfur being attached to the coke on FCC catalysts, which is
then oxidized and emitted with the flue gas. The wet sulfur scrubbers use a caustic soda
(NaOH) to react with SO2 , which is then removed as a soluble salt, while SO3 forms a sulfuric
acid mist in the presence o f NaOH. The soluble salts and acid mist are then filtered into a
spray tower where condensation occurs, collecting the soluble salts and acid mist (Meyers,
2004). Finally, the flue gas is sent through a droplet separator, which causes the gases to
spiral down a tower with the centrifugal force causing any remaining water droplets to run
down the sides of the tower. This proeess has been shown to be 92 % efficient at removing
harmful sulfur oxide eontaining particles and can be seen in Figure 1.5 (Meyers, 2004).

13

However, this thesis is more concerned with the sulfur in the gasoline products rather than in
the air emissions.
6r

;
^

■"»» V
t.

—

I.

Î

Figure 1.5: Wet Scrubber absorber vessel/spray tower (Meyers, 2004).
1.5.3 The Merox Process
The Merox process uses a catalytic procedure to remove mercaptans or convert them
to lesser disulfides as the schematic in Figure 1.6 shows (Meyers, 2004). A caustic solution is
used for mercaptan removal from LPG, treating gases, and light-gasoline fractions. Once the
mercaptans are dissolved in the caustic solution, air is injected into the stream. The air
converts the mercaptans to disulfides, which are then separated from the caustic stream
(Meyers, 2004). The disulfides are then sent to a processing unit for conversion to elemental
sulfur.

14

ExtracW
Product

Excess

Extractor

A ir

Disulfide

Air

Ridi

Merox

•

Caustic %
Caalyst
liÿectx»

Figure 1.6: Merox mercaptan-extraction unit (Meyers, 2004).
1.5.4 The Claus Process
The Claus process partially oxidizes H 2 S to create suhur (Matar and Hatch, 2001).
This process is present at almost every refinery and uses the hydrogen sulfide feeds from the
above-mentioned methods to produce sulfur. There are two parts to the Claus process, a
burning section that oxidizes sulfur, and a reactor which causes a reaction between H2 S and a
bauxite catalyst as seen in Figure 1.7 (Matar and Hatch, 2001).

AEMEATER

REHEATER

1

ACIDGA8

Æ 31
jA iR

BLOWER

r PQ

r

f COND

T

TAILGAS

COND I

Figure 1.7: Two-stage Claus unit (Meyers, 2004).

15

The burner section involves the oxidation of hydrogen sulfide into sulfur dioxide and
the partial oxidation o f hydrogen sulfide into sulfur (Matar and Hatch, 2001). In the reactor
the unchanged H 2 S reacts with a bauxite catalyst in the presence of oxygen to produce sulfur.
The sulfur in the reactor is then removed through condensation (Matar and Hatch, 2001).
This process produces approximately 90 - 95 % of the world’s sulfur, which is then sold and
used in many industries, such as fertilizer production (Matar and Hatch, 2001).
1.5.5 Hydrotreating
Hydro treating can be used to pretreat feeds before entering the FCC unit or can be
used to treat the gasoline fraction after the FCC process (Dupain et al., 2003). This is one of
the only methods commercially available to remove heavier sulfur compounds such as
thiophenes. Hydrotreating reactions typically occur in the liquid phase with Figure 1. 8
showing a schematic of the process (Meyers, 2004). The feed is saturated with hydrogen gas,
which then flows through the catalyst pores and adsorbs to the catalysts surface. The reaction
then occurs on the catalyst surface where sulfur-containing molecules are cracked into smaller
molecules (Meyers, 2004). The reactions are very exothermic and, in the case of sulfur,
involve breaking the two sulfur-carbon bonds and adding four hydrogen atoms. The removed
sulfur then forms H 2 S, which can diffuse out of the catalyst pores and be removed from the
hydrotreater (Meyers, 2004).

16

Make-# Mydmgm

Siw r

OO

Figure 1.8: Hydrotreating flow scheme (Meyers, 2004).
Perfecting hydrotreating catalysts has been extensively researched to increase its ability
to remove contaminants and give the catalyst a long life, which are both accomplished by
decreasing the poisoning o f catalysts. The hydrotreating catalysts are typically Co(Ni)Mo or
NiW supported by aluminum oxide (Coulier et ah, 2002). There has been some research on
the addition of chelating agents to improve the effectiveness of the catalyst preparation. The
addition of nitrilotriacetic acid (NTA) and ethylenediaminetetraacetic acid (EDTA) stabilizes
Ni and Co until a temperature is reached where Mo or W have reacted with sulfur to form
sulfides (Kishan et ah, 2001). Medici and Prins (1996) also showed that by adding chelating
agents during the impregnation step and leaving out the calcination step for Si0 2 -supported
hydrotreating catalysts, the activity is just as strong as y-Ah O3 - supported hydrotreating
catalysts.

17

The development o f new hydrotreating catalysts other than the typical Co(Ni)Mo
catalysts is fast growing. Oyama et al, (2002) have developed a new type of catalyst from the
transition metal phosphides group, such as NhP and FczP. The catalyst, NizP, has shown
promise as 98 % of the sulfur was removed compared to only 78 % with NiMo (Oyama et a l,
2002). These metal phosphide catalysts are also promising due the lack o f a layered structure,
allowing the entire surface area to be exposed and due to the moderate preparation
temperatures and inexpensive precursors (Oyama et a l, 2002).
In terms o f catalyst poisoning, both H2 S and NH 3 are present in fossil fuels, which can
inhibit the hydrotreating catalysts by competitive adsorption with unsaturated hydrocarbons
(Blanchin et a l, 2001). The most important property about Co(Ni)Mo hydrotreating catalysts
is the ability to perform the conversion of suhur-containing compounds into hydrocarbons
even in the presence of H 2 S and NH 3 (Hensen et a l, 2003). The inhibiting effect of H2 S is
much more significant at high rather than low partial pressures (Blanchin et a l, 2001).
Much of the research into poisoning of catalysts involves the effects of the support on
the sulfidation o f the metals. Coulier et al, (2002) determined that the presence of tungsten
prevented cobalt and nickel from interacting with the support. This allowed a higher degree
of sulfidation starting at lower temperatures and more catalyst poisoning (Coulier et a l,
2002). So a strong interaction with the support is necessary to reduce catalyst poisoning.
Also, Hensen et al, (2003) showed that the poisoning effect of H2 S could be reduced by
supporting the Co(Ni)Mo catalysts on materials with high Bronsted acidity and a strong
metal-support interaction. These materials include carbon or aluminum oxide supports.

18

Carbon has been shown to be a much better support than aluminum oxides as there is a 30 %
increase in the catalytic activity (Glasson et a l, 2002).
1.5.6 Limestone Fluidized Beds
Limestone has been used for many years to remove H2 S from coal through reaction
1.14 (Fenouil and Lynn, 1995):
+H

CaS^^.^ +

(1-14)

This reaction takes place in the gasifier during coal processing at high temperatures
and pressures and is carried out by simply adding limestone or its precalcined form (Fenouil
and Lynn, 1995). The particles must react directly with H2 S to cause desulfurization
(Krishnan and Sotirchos, 1994). Fenouil and Lynn (1995) discovered that a large excess of
limestone is needed compared to theoretical values in order to remove H 2 S. Therefore,
Fenouil and Lynn (1995) used sorption into a moving bed of limestone particles.
The kinetic study of the moving bed of limestone found that at temperatures below
670 °C the reaction between CaCOs and H 2 S is the rate-limiting step for sulfidation (Fenouil
and Lynn, 1995). However, at temperatures above 670 °C the CaS product forms a layer
around the limestone grains and prevents CO2 and H 2 O from dififiising out (Fenouil and Lynn,
1995). Therefore, this is the rate-limiting step at higher temperatures, not the CaCOs and H2 S
reaction shown in reaction 1.14. Fenouil and Lynn (1995) also found in another study that the
calcination of limestone at high temperatures could compete with the CaCOs and H2 S
reaction.
The competition reaction is as follows in reaction 1.15 (Fenouil and Lynn, 1995):

19

CaCO; ^

CaO+CO^ (115)

This reaction forms lime, which can then reaet with H 2 S. The difference between precalcined
limestone and regular limestone is the pore structure present within the preealcined limestone
(Krishnan and Sotirehos, 1994). This pore structure allows diffusion to occur so reactions
can continue at the interface between CaS and the coal gas within the pores (Krishnan and
Sotirehos, 1994). This reaction may occur continuously on precalcined limestone but the
reaction with regular limestone only occurs at the surface of the reacting particles (Krishnan
and Sotirchos, 1994). However, as with the initial study, the reaction was ultimately
controlled by the diffusion through the CaS layer (Fenouil and Lynn, 1995).
Fenouil and Lynn (1995), and Nakazato et al. (2003), both used fluidized bed
reactors. Nakazato et al. (2003) found that the H2 S removal in a fluidized bed reactor was
dependent upon the limestone particle diameter. As the particle diameter decreased the
efficiency increased (Nakazato et al , 2003). Also, it was determined that the H 2 S removal
eflfieieney increased with temperature until approximately 1073 K, where the efficiency then
decreased with increasing temperature (Nakazato et al , 2003). Therefore, the limestone
appeared to be good at removing H 2 S under certain reaction conditions. However, it does not
appear to remove any other types of suhur-containing compounds. With each of these sulfur
removal methods, it is important to have a reliable way to measure the sulfur contained in
fossil fuels.
1.6 Analvtical Methods
There have been many methods developed in order to analyze for sulfur in petroleum
products. Sulfur is one of the major sources of eoneern for heteroatoms in petroleum

20

products next to nitrogen (Hsu, 2003). The methods developed for analyzing sulfur inelude
gas chromatography with a variety of detectors, ultraviolet fluorescence, energy-dispersive Xray fluorescence spectrometry, and inductively coupled plasma spectroscopy. Each of these
methods has its advantages and disadvantages, which will briefly be described below.
1.6.1 Gas Chromatography
Refineries and analytical laboratories have used gas chromatography to analyze for
sulfur extensively. There are several different types of detectors that can be used. These
include: flame photometric detector (FPD), electrolytic conductivity (Hall) detector, atomic
emission detector (AED), electron capture sulfur detector (ECD) and a universal sulfur
chemiluminescence detector (SCO) (Hsu, 2003). Each detector has associated pros and cons
for sulfur analysis.
The FPD is used frequently but the response is adversely affected by co-eluting water
and hydrocarbons (Hsu, 2003). Flame photometric detectors are typically problematic when
large concentrations o f hydrocarbons mask certain sulfur species present at lower
concentrations (Hsu, 2003). The AED is also useful for identifying and quantifying sulfur.
However it is quite costly and requires time-consuming calibrations (Hsu, 2003). The ECD
and electrolytic conductivity detectors are rarely used for sulfur analysis.
The SCD is currently one of the best detectors for sulfur. It is superior to the FPD as
there is no interference from hydrocarbons or water vapour, and is very sensitive to lower
sulfur concentrations (Hsu, 2003). The SCD operates under the following reactions (Hsu,
2003):
RSR^ +O 2

SO + Other products

(1.16)

21

+

(1.17)

When the sulfur compounds pass through the GC column and the FID SO is produced,
it emits a blue chemiluminescence of SO2 when it reacts with ozone (Flsu, 2003). The light is
emitted between 260 to 480 nm (Hsu, 2003). The FID is not capable of deterrnining
individual sulfur compound concentrations so an SCD can be used. There has also been an
ASTM method developed for the use of SCD to measure sulfur eoneentrations. The ASTM
method D5623 is the standard test method for sulfur eompounds in light petroleum liquids by
gas ehromatography and sulfur seleetive deteetion (Annual book of ASTM standards, 2003).
This method gives the reprodueibüity and repeatability of the standard test under the seleeted
operating conditions (Annual book of ASTM standards, 2003).
A method published by Yin and Xia (2004) uses both GC/FPD and GC/MS to identify
sulfides and thiophenes. In order to separate the thiophenes, the sample is washed with silver
nitrate and then passed through a GC/FPD (Yin and Xia, 2004). The sulfides are extracted
using a solvent, which causes a separation of the sulfides into the aqueous phase from the oil
phase. The aqueous phase is diluted with water and then extracted with petroleum ether (Yin
and Xia, 2004). This produces the crude sulfur compounds, which are further purified using
vacuum distillation. This phase is then diluted with water and extracted with hexane to
produce purified sulfur compounds (Yin and Xia, 2004). These purified sulfur compounds are
then analyzed using GC/MS and GC/FPD.
Recently, a two-dimensional GC x GC has been combined with a SCD to separate the
different sulfur groups and identify sulfur compounds (Hua et a l, 2003). The two­

22

dimensional GC x GC analysis uses the retention time on the x-axis and the polarity-based
retention time on the y-axis (Hua et a l, 2003). This reduces the necessity of a highly efFieient
column and highly stable retention times in the GC (Hua et al., 2003). So far this has been
applied mostly to diesel fuels, as diesel fuels typically have more thiophenes and
dibenzothiophenes than gasoline.
1.6.2 Ultraviolet Fluorescence
In the method involving ultraviolet fluorescence, a hydrocarbon sample is injected into
the instrument where a temperature of over 1000 °C causes the sample to combust (Annual
book of ASTM standards, 2003). The sulfur is oxidized into sulfur dioxide, which is then
exposed to an ultraviolet light. The sulfur dioxide becomes excited and fluoresces. The
photomultiplier tube then detects this fluorescence and the signal is a measure of the sulfur
concentration in the sample.
For this method, as the sulfur concentration varies from 1 mg/kg S to 400 mg/kg S,
the repeatability increases from 0.2 mg/kg S to 16.0 mg/kg S. Using equation e. 1.1, the
repeatability can be calculated for sulfur concentrations less than 400 mg/kg S (Annual book
of ASTM standards, 2003):

r = 0.1788(^)''""

(e.1.1)

Where; X = the sulfur concentration in mg/kg. Similarly, the reproducibility increases
from 0.6 mg/kg S to 51.9 mg/kg S. However, the reproducibility for less than 400 mg/kg S
can be calculated using equation e. 1.2 (Annual book of ASTM standards, 2003):
0.75
;g = 0.5797(;ir)"'"

(e.l.2)

23

This method is considered acceptable for samples whose boiling range is tfom 25 °C to
400 °C (Annual book of ASTM standards, 2003). Therefore, gasoline has acceptable
precision under this method with samples containing 1.0 to 8000 mg/kg of total sulfur. This
method is also the one accepted by the Canadian Environmental Protection Act (2004).
1.6.3 Energy Dispersive X-Ray Fluorescence Spectrometry
The energy-dispersive X-ray fluorescence spectrometry method places the
hydrocarbon sample in a beam emitted from an X-ray source (Annual book of ASTM
standards, 2003). The radiation emitted from sulfur is then measured and a count is made that
is then compared with the counts from the calibration standards (Annual book of ASTM
standards, 2003). This comparison gives the sulfur concentration in mass percent.
This method is applicable for non-leaded gasoline and gasoline-oxygenated blends and
for samples with a sulfur range of 48 to 1000 ppmw (Annual book of ASTM standards,
2003). The repeatability can be determined by equation e.1.3:
r = 12.30(%-H0)"'

(e.1.3)

Where; X is the sulfur concentration in mass percent (Annual book of ASTM
standards, 2003). The reproducibility of this method is determined by equation e. 1.4:

;; = 36.26(W-H0)'" (e.1.4)
However, it has been found that X-ray fluorescence methods are not suitable for lowlevel sulfur concentrations and are therefore not commonly used in industry; instead UV
fluorescence, chemiluminescence detectors or FIDs are typically employed (Gokeler et al.,
2002).

24

1.6.4 Inductively Coupled Plasma
Inductively coupled plasma (ICP) methods have been developed for fossil fuels and
other petroleum products. Also, there are other methods, which have been developed for ICP
spectrometers coupled with other types of instrumentation. These coupled instruments
include ICP-IDMS (isotope dilution mass speetrometry) and DIHEN-ICP-IDMS (direct
injection high-eflficiency nebulizer).
The ICP method typically uses a microwave digestion and has been used for sulfur
determination in coal. Laban and Atkin (2000) developed a microwave digestion method.
Hydrochloric acid was used to digest the sulfate sulfur, nitric acid for the digestion of pyritic
sulfur and a combination of nitric, hydrochloric, hydrofluoric, and boric acid for the digestion
o f organic sulfur (Laban and Atkin, 2000). The solutions produced are analyzed using ICPAES at a wavelength of 182.037 nm. The method proved to be good for coal when
determining total sulfur, as the reproducibility is good with coefficients of variation less than 5
% in all cases. The method was determined to be optimal for a total sulfur range of 300 to
5000 ppmw (Laban and Atkin, 2000).
Similarly, a method for sulfur in gasoline and other fuel samples was developed by
Heilmann et al (2004) using a microwave digestion followed by analysis with an ICP-IDMS or
a DIHEN-ICP-IDMS. The microwave digestion involved using concentrated nitric acid
(Heilmann et a l, 2004). The sample and nitric acid were placed in a quartz vessel, which was
then placed inside the Teflon vessel. Distilled water and hydrogen peroxide were then added
to the Teflon vessel to prevent venting of the quartz vessel and therefore, preventing sample
loss (Heilmann et a l, 2004). The digested samples were then analyzed using both ICP-IDMS

25

and DIHEN-ICP-IDMS. The ICP-IDMS method has good precision and accuracy.
However, due to the long analysis time, it may not be suitable for routine analysis (Heilmann
et a l, 2004). The DIHEN-ICP-IDMS method is fast and accurate. However, close attention
needs to be paid to the preparation of the transparent micro-emulsion and optimization of
ICP-MS measurement conditions (Heilmann et a l, 2004). The DIHEN-ICP-IDMS method
has a detection limit o f 20 pg/g while the detection limit of the ICP-IDMS is 10 pg/g
(Heilmann et a l, 2004). Therefore, either method would be suitable under the new Canadian
regulatory limits.
1.7 Solvent Extraction and Solubilitv Parameters
Solvents have been used for years to extract components from mixtures. There are
several conventional solvent extraction processes such as mixer-settler arrangements or
continuous countercurrent contacting equipment (Alonso et al., 2001). In order to be able to
efficiently extract a substance, the solvent requires similar properties to the solute being
extracted (Reichardt, 1988). There are several properties, which can be used to assess the
extractability of different solvents. These properties include partition coefficients and
solubility parameters (Reichardt, 1988). In particular, solvents used in hydrocarbon
processing, for example, methyldiethanolamine (MDEA), are very selective for the removal of
H 2 S (Jou et a l, 1997).
1.7.1 Solvent Extraction Methods
Mixer-settler arrangements are the most common solvent extraction techniques. A
mixer-settler apparatus typically contains both the solvent and the solution containing the
solute. The two solutions are mixed for some period of time then allowed to settle so that the

26

more dense liquid settles to the bottom and can be decanted. However, there are several
properties that need to be addressed with mixer-settler arrangements. When using a mixersettler arrangement the solvent must be dispersed throughout the mixture, there must be a
density difference between the fluids and émulsifications should be avoided (Alonso et a l,
2001). Non-dispersive solvent extraction can overcome many of these disadvantages. In nondispersive solvent extraction, the two liquids flow on opposite sides of a porous interface so
that mixing o f the two liquids does not occur (Alonso et a l, 2001). However, non-dispersive
solvent extraction involves much more complex equipment, whereas mixer-settlers can be as
simple as shake flask or sonication methods (Fitzpatrick and Dean, 2002). There are several
ways to predict which solvents may work for the particular solute that is being extracted.
1.7.2 Solubility Parameter and Solvent Relationship
Fitzpatrick and Dean (2002) determined that solvents with similar solubility properties
could form mixtures; therefore being capable of predicting which solvents can be used for
analyte recovery. The greater the difference in chemical properties between any two solvents
the less soluble they will be in one another (Reichardt, 1988). In order for a solvent to be
effective at extracting a solute there must be a high degree of separation, high selectivity, no
tendency to emulsion formation and rapid separation (Reichardt, 1988). The degree of
separation can be determined by either the density difference of the two solvents or through
the ratio of the partitioning coefficients (Reichardt, 1988). Also, the rapid separation of the
phases requires a large density difference and a low viscosity. The lower the viscosity the
more rapid the separation will occur between the phases (Reichardt, 1988). However, in
order to match the chemical properties, solubility parameters can be used to narrow the list of
solvents to be selected for a particular solute.

27

There are several different solubility parameters, for example Hoy’s cohesion
parameter, Hansen’s parameter, and Hildebrand’s solubility parameter. Each o f these
measures of solubility is slightly different. For example, Hildebrand solubility parameter (ôt),
total solubility, is a measure of the internal energy of cohesion (Fitzpatrick and Dean, 2002).
Therefore, solvents with similar solubility parameters form mixtures and therefore a solvent
and solute with similar solubility parameters should also form a mixture (Fitzpatrick and Dean,
2002). Hansen took the total solubility and based it on three separate components. The three
components are hydrogen-bonding ability (5h), dispersion coefficient (5d), and polarity
coefficient (ôp) (Fitzpatrick and Dean, 2002). Hansen determined that these three components
could be used to determine the total solubility of a solvent through equation e.1.5 (Barton,
1991):
(e.1.5)
Hoy determined these individual components of solubility through a variety of
methodologies. For example. Hoy calculated the total solubility by measuring the change in
vapour pressure at a number of temperatures (Fitzpatrick and Dean, 2002). These solubility
parameters have been determined for a variety of different solvents and can be found in the
CRC Handbook o f Solubility Parameters and Other Cohesion Parameters (1991). Fitzpatrick
and Dean (2002) determined that by using the Hildebrand solubility parameter and the
individual components an appropriate extraction solvent can be chosen. It should be
mentioned, however, that solvents could have retarding effects on hydrodesulfurization
catalysts (Ishüiara and Kabe, 1993). Therefore, care should be taken to remove all solvents
from the feed before further processing. Also, solvent extractions typically are based on

28

hydrogen bonds or van der Waals interactions, whereas a stronger molecular interaction may
be required to remove the desired solute.
1.8 Sorbent Extraction
There are other extraction processes besides solvents, such as sorbent extractions.
Almost all adsorption processes in the industrial world are based on van der Waals
interactions between two solids (Takahashi et a l, 2002). However, chemical complexation
bonds are stronger and therefore more selective that van der Waals interactions (Takahashi et
a l, 2002). For this reason there has been much work done in identitying sorbents that have
chemical complexation bonds rather than simply van der Waals interactions.
Takahashi et a l (2002) have identified copper and silver zeolites, which can be used to
remove thiophenes at temperatures near that of room temperature. Similarly, Ma et a l (2002)
developed a deep desulfurization method using seleetive adsorption onto a transition metal
supported on silica gel. This method placed the adsorbent into a column, where the fuel was
then allowed to flow down through the adsorbent. The method showed good removal o f the
sulfur compounds; however the study was only in the preliminary stages. Other examples, of
sorbents being used are limestone, which has shown good results for H2 S removal in coal
processing (Nakazato et a l, 2003). However, in order to remove some solutes even strong
chemical interactions are necessary, some forms of catalysis also use adsorption to remove
contaminants.
1.9 Ranev Nickel
Raney nickel is one o f the most common metal catalysts and was discovered by
Murray Raney in 1927 (Augustine, 1996). Raney nickel is prepared by the addition of sodium

29

hydroxide to a powdered nickel-alurninum alloy. This reaction causes the nickel to form a
porous structure, which gives it the name of spongy nickel (Freel et a l, 1969). The porous
structure creates a high surface area, which increases the reaction surface (Augustine, 1996).
Raney nickel can be used to effectively remove sulfur and has been used in analytical methods
or purification methods for years.
1.9.1 Activation o f Raney Nickel
The activation o f Raney nickel with sodium hydroxide has been extensively studied
with a variety of activation procedures identified. The nickel-aluminum alloy typically
contains 50 % nickel and 50 % aluminum (Augustine, 1996). In the case of Raney nickel, the
activation process is the process in which sodium hydroxide is added to the alloy to remove
the aluminum and produce hydrogen (Augustine, 1996). This causes the nickel-aluminum
alloy to become active Raney nickel. The hydrogen adsorbs to the nickel surface; however,
there is not a one to one relationship as not all the Raney nickel surface is metallic and
accessible (Fouilloux, 1983). This reaction can be seen in reaction 1.19 (Devred et a l, 2003):
2 ( A z - J / ) ( , ) + 2 ( 9 / / +6//2(9(,) ^ 2A/-//22j„„^^+ 2yl/((9//)4(„^^) + / / 2 (g)

(1.19)

The amount and temperature of the sodium hydroxide determines the type of Raney
nickel and are designated as W1-W8 Raney nickel (Augustine, 1996). The preparation
procedures for these types of Raney nickel are summarized in Table 1.1, which was developed
by Augustine (1996).

30

Tabic 1.1: W1-W8 Raney nickel preparation methods (Augustine, 1996).
Type

Addition
Temp.

NaOH: Alloy
Ratio

Digestion Temp,
and Time

Washing Proeess

Relative
Activity

W1

0°C

1: 1

1 1 5 ° - 120 °C

Neutralize with H2 O

Least Active

4 hours

Ethanol wash

Approx. = W8

Steam bath

H 2 O wash to neutral

<W4: >W1

8-12 hours

Ethanol wash

Most common

50 °C

H 2 O wash several times

Quite Active

50 minutes

Continuous wash with H 2 O

>W2: <W7

W2

W3

25"C

-20°C

4: 3

4: 3

Ethanol wash without
contact with air
W4

50°C

4: 3

50 °C

Same as W3

Same as W3

Same as W3

Same as W3

Continuous H 2 O wash under
H2 atmosphere

Most active

50 minutes
W5

50"C

4: 3

50 °C
50 minutes

W6

50°C

4: 3

50 °C
50 minutes

Ethanol wash without
contact with air
W7

W8

50°C

o“c

4: 3

1: 1

50 °C

Three décantations with H2 O

Very active

50 minutes

Ethanol wash without
contact with air

<W6: >W4

1 0 0 ° - 105 °C

Continuous H 2 O was

Least active

4 hours

Dioxane wash

Approx. = W1

Distil dioxane portion from
catalyst

However, not all types o f Raney nickel are elassified in this manner. Further research
done with Raney nickel has developed other methods for the activation of Raney nickel. For
example, Granatelli (1959) used a 20: 1 ratio of sodium hydroxide to the alloy and allowed the
reaetion to occur at room temperature and overnight. Beigi et al (1999) used a ratio that
ranged from 3 to 12: 1 of sodium hydroxide to the alloy and allowed the reaction to occur

31

overnight at room temperature. Srivastava et al (1985) used a 30; 1 ratio and the digestion
was gradually heated to 100 °C. Finally, Bartok et al (1987) used a 20: 1 ratio and the
digestion occurred at 80 °C for 45 minutes. It should also be mentioned that each of these
procedures has a slightly different washing proeess, but all involve washing with H2 O and then
further rinsing with ethanol or isopropanol. The activation process can have remarkable
affects on the structure o f Raney nickel.
1.9.2 Structure
The structure of Raney nickel incorporates a large surface area, however, it is difficult
to reproduce the surface area even if the same alloy and activation procedure is employed
(Fouilloux, 1983). The preparation temperature has an impact on the porous structure and
therefore, the surface area (Fouilloux, 1983). At low preparation temperatures the pores have
a narrower diameter, whereas at higher preparation temperatures the pores are more
cylindrical and create a higher pore volume (Fouilloux, 1983). FouiUoux (1983) determined
that the more intense the alkali reaction the larger the pore diameter and the greater the pore
volume. In general, Raney nickel is a sponge-Uke particle that is approximately 100 Â in
diameter with a smaller microporous structure. The microstrueture has pore diameters
between 20 and 80 A (Fouilloux, 1983). It is this porous structure that makes Raney nickel a
good catalyst or reactant for the removal of sulfur.
1.9.3 Desulfurization with Raney nickel
There are two reaction mechanisms by which Raney nickel can be used to remove
sulfur. The two mechanisms are as follows (Lieber and Morritz, 1953):
3RSR: + 3 M - F/jtoey ^

+RR-^ R R +

(1.20)

32

Typically, there is also a sulfur-to-sulflir cleavage for disulfides ( e.g. R-S-S-R’)
removing both the sulfur atoms and introducing hydrogen. This was discovered since both
ethane and methane are formed after the reaction with Raney nickel (Lieber and Morritz,
1953). With the thiols, Raney nickel causes a cleavage of the carbon-sulftir bond and the
formation of a carbon-hydrogen bond (Pettit and Tamelen, 1962). A similar cleavage is seen
with thiophenes during a hydrogenolysis reaction where the sulfur is removed and replaced
with hydrogen (Pettit and Tamelen, 1962). It has been noted that the simple hydrogenolysis
reaction does not always go to completion, so it is therefore necessary to treat the solution
twice (Papa et a l, 1949). Thus, from the above cited literature, the effectiveness of Raney
nickel for removing sulfur makes it a promising option for desulfurization of FCC gasoline.
However, no such study has yet been reported to date.
1.10 Objectives o f Research
This thesis was designed to determine if an alternative method could be developed for
removing sulfur-containing compounds from FCC gasoline, the gasoline fraction produced
from fluid catalytic cracking. In order to meet this purpose there are four objectives.
1.

The first objective is to find a solvent or catalyst that will specifically target
sulfur compounds. This will take an experimental approach for a variety of
solvents, metals salts and Raney-type nickel.

2.

The second objective is to reduce sulfur concentrations as much as possible
with a goal of 30 ppmw or lower.

33

3.

The third objective is to reduce the suhur concentrations without making any
major changes to the composition of the FCC gasoline.

4.

The fourth objective is to determine if the reactants can be recycled.

1.11 Conclusion
We hope this research will contribute towards finding a solution to one of the major
environmental issues concerning the burning of fossil fuels. The literature review performed
here has given many insights into the processing of petroleum products and the options
currently available for sulfur removal. The problem with sulfur in the environment is that it
not only causes a risk to humans, it is also a risk to the plants and animals. FCC gasoline is
the largest contributor to sulfur in gasoline and therefore needs to be the focus of suhur
removal efforts. Currently, hydrotreating is the most common method used to remove suhur
from FCC gasoline but there are several problems associated with using hydrotreating for
deep desulfurization. Also, there are several methods available for analyzing for sulfur.
However, a more appropriate one for FCC gasoline will need to be determined. As catalysts
seem to be the current method as well as the most studied approach for removing sulfur from
fuels, perhaps more research needs to be carried out into solvents or sorbents for suhur
removal. Finally, there has already been some research accomplished using Raney nickel for
sulfur removal. However, most o f this work has been for analytical or purification purposes
only. Thus, there is potential for Raney nickel to be used in FCC gasoline sulfur removal.
This approach, if successful, could be potentially used as an alternative to the expensive
hydrotreater.

34

1.12 Literature Cited
Alonso, A.I., Lassahn, A., and G.Gruhn. 2001. Optimal design of non-dispersive solvent
extraction processes. Computers and Chemical Engineering. 25: 267-285.

Annual Book of ASTM Standards. 2003. Petroleum Products. Lubricants, and Fossil Fuels.
Volume 3. ASTM International. West Conshohoeken, PA. Pp. 1-1368.

Augustine, R.L. 1996. Heterogeneous Catalvsis for the Svnthetie Chemist. Marcel Dekker
Inc. New York, USA. Pp. 213-540.

Bartok, M., Wittmann, G., Gondos, G., and G.V. Smith. 1987. Homogeneous and
heterogeneous catalytic asymmetric reactions. 1. Asymmetric hydrogenation of the
proehiral C=C bond on a modified Raney Ni catalyst. Journal o f Organic Chemistry
52:1139-1141.

Barton, A.F.M. 1991. CRC Handbook of Solubilitv Parameters and Other Cohesion
Parameters. 2"‘*edition. CRC Press. Boca Raton, Florida. Pp. 1-739.

Beigi, A.A.M., Teymouri, M., Eslami, M., and M. Farazmand. 1999. Determination of trace
sulfur in organic eompounds by activated Raney nickel desulfurization method with nondispersive gas detection system. Analyst 124: 767-770.

Blanchin, S., Galtier, P., Kasztelan, S., Kressmann, S., Penet, H , and G. Perot. 2001. Kinetic
modeling o f the effect of H 2 S and of NH 3 on toluene hydrogenation in the presence of
NiMo/AfOs hydrotreating catalyst. Discrimination between homolytic and heterolytic
models. Journal o f Physical Chemistry A. 105:10860-10866.

Bunce, N. 1994. Environmental Chemistry. 2"^ edition. University of Guelph. Ontario. Pp. 1376.

Campbell, l.M. 1988. Catalvsis at Surfaces. Chapman and Hall, London. Pp 145-224.

Canadian Environmental Protection Act, 2004. http://laws.justice.gc.ca/en/C-15.31

Coulier, L., Kishan, G., van Venn, J.A.R., and J.W. Niemantsverdriet. 2002. Influence of
support-interaction on the sulfidation behaviour and hydrodesulfurization activity of
35

A^Os-supported W, CoW, and NiW model catalysts. Journal o f Physical Chemistry B
106: 5897-5906.

Devred, F., Hoffer, B.W., Sloof, W.G., Kooyman, P.J., van Langeveld, A.D., and H.W.
Zandbergen. 2003. The genesis of the active phase in Raney-type catalysts: the role of
leaching parameters. Applied Catalysis A: General 244: 291-300.

Dupain, X., Gamas, E.D., Madon, R., Kelkar, C.P., Makkee, M., and J.A. Moulijn. 2003.
Aromatic gas oil cracking under realistic FCC conditions in a mieroriser reactor. Fuel. 82:
1559-1569.

Dupain, X., Rogier, L.J., Gamas, E.D., Makkee, M., and J.A. Moulijn. 2003. Cracking
behaviour of organic suhur compounds under realistic FCC conditions in a microriser
reactor. Applied Catalysis A: General. 238:223-238.

Ertl, G., Knozinger, H., and J. Weitkamp. 1999. Environmental Catalvsis. Wiley-VCH.
Germany. P pl7-118.

Fenouil, L.A, and S. Eynn. 1995. Study of calcium-based sorbents for high-temperature H 2 S
removal. 2. Kinetics of H2 S sorption by calcined limestone. Industrial & Engineering
Chemistry Research. 34: 2334-2342.

Fitzpatrick, E.J., and J.R. Dean. 2002. Extraction solvent selection in environmental analysis.
Analytical Chemistry. 74(1): 74-79.

Fenouil, L.A., and S. Lynn. 1995. Study of calcium-based sorbents for high-temperature H 2 S
removal. 1. Kinetics o f H 2 S sorption by uncalcined limestone. Industrial & Engineering
Chemistry Research. 34: 2324-2333.

Fouilloux, P. 1983. The nature o f Raney nickel, its adsorbed hydrogen and its catalytic activity
for hydrogenation reactions (review). Applied Catalysis 8: 1-42.

Freel, J., Pieters, W.J.M., and R.B. Anderson. 1969. The structure of Raney Nickel I. Pore
Structure. Journal o f Catalysis 14: 347-256.

Glasson, C., Geantet, C., Eacroix, M., Labruyere, F., and P. Duffesne. 2002. Beneficial effect
of carbon on hydrotreating catalysts. Journal o f Catalysis. 212: 76-85.
36

Gokeler, U., Ofifermanns, U., and H. Muller. 2002. Consider online sulflir measurement for
clean fuels. Hydrocarbon Processing 81: 93-94.

Granatelli, L. 1959. Determination of mierogram quantities of sulfur by reduction with Raney
nickel. Analytical Chemistry 31(3):434-436.

Heilmann, J., Boulyga, S.F, and K.G. Heumann. 2004. Accurate determination of sulfur in
gasoline and related fuel samples using isotope dilution ICP-MS with direct sample
injection and microwave-assisted digestion. Analytical Bioanalytical Chemistry 380: 190197.

Hensen, E.J.M., de Beer, V.H.J., van Veen, J.A.R, and R.A. van Santen. 2003. On the sulfur
tolerance of supported Ni(Co)Mo sulfide hydrotreating catalysts. Journal o f Catalysis.
215:353-357.

Hsu, C.S. 2003. Analvtical Advances for Hvdroearbon Research. Kluwer Academic/Plenum
Publishers. New York, USA. Pp. 1-463.

Hua, R., Li, Y., Liu, W., Zheng, J., Wei, H., Wang, L, Lu, X., Kong, H., and G. Xu. 2003.
Determination of sulfur-containing compounds in diesel oils by comprehensive twodimensional gas chromatography with a sulfur chemiluminescence detector. Journal o f
Chromatography A. 1019: 101-109.

Ishihara, A. and T. Kabe. 1993. Deep desulfurization of light oil. 3. effects of solvents on
hydrodesulfurization of dibenzothiophene. Industrial and Engineering Chemistry
Research. 32: 753-755.

Jou, F., Otto, F.D., and A.E. Mather. 1997. The solubility of mixtures of H 2 S and CO2 in an
MDEA solution. The Canadian Journal o f Chemical Engineering. 75: 1138-1141.

Kemsley, J. 2003. Targeting sulfur in fuels for 2006. Chemical and Engineering News Oct:
40-41.

Kennedy, l.R. 1992. Acid Soil and Acid Rain. 2"‘‘ edition. Research Studies Press Ltd.
England. Pp. 150-180.

37

Kishan, G., Coulier, L., van Veen, J.A.R., and J.W. Niemantsverdriet. 2001. Promoting
synergy in CoW sulfide hydrotreating catalysts by chelating agents. Journal o f Catalysis.
200: 194-196.

Krishnan, S.V., and S.V. Sotirehos. 1994. Experimental and theoretical investigation of
factors affecting the direct limestone-H2 S reaetion. Industrial & Engineering Chemistry
Research. 33: 1444-1453.

Kropp, K.G. and P.M. Fedorak. 1998. A review of the occurrence, toxicity, and
biodégradation of condensed thiophenes found in petroleum. Canadian Journal o f
Microbiology 44: 605-622.

Laban, K.L., and B.P. Atkin. 2000. The direct determination of the forms of sulphur in coal
using microwave digestion and i.c.p-a.e.s analysis. Fuel 79: 173-180.

Lefiaive, P., Lemberton, J.L., Perot, G., Mirgain, C., Carriat, J.Y., and J.M. Colin. 2002. On
the origin o f suhur impurities in fluid catalytic cracking gasoline—reactivity of thiophene
derivatives and o f their possible precursors under FCC conditions. Applied Catalysis A:
General. 227: 201-215.

Lieber, L. and F.L. Morritz. 1953. The uses of Raney nickel. Advances in Catalysis 5: 417 455.

Ma, X., Sun, L., and C. Song. 2002. A new approach to deep desulfurization of gasoline,
diesel fuel and jet fuel by selective adsorption for ultra-clean fuels and for fuel cell
applications. Catalysis Today. 77: 107-116.

Matar, S. and L.F. Hatch. 2001. Chemistry of Petrochemical Processes. 2"‘*edition. Gulf
Professional Publishing. USA. Pp. 69-88.

Medici, L. and R. Prins. 1996. The influence of chelating ligands on the sulfidation o f Ni and
Mo in NiMo/SiOz hydrotreating catalysts. Journal o f Catalysis. 163: 38-49.

Meyers, R.A. 2004. Handbook of Petroleum Refining Processes. 3’^‘*edition. McGraw-Hill.
USA. Pp. 1-15.40.

38

Nakazato, T., Lin, Y., Kusumoto, M., Nakagawa, N., and K. Kato. 2003. H 2 S removal by fine
limestone particles in a powder-particle fluidized bed. Industrial & Engineering Chemistry
Research. 42: 3413-3419.

Oyama, S.T., Wang, X., Requejo, F.G., Sato, T., and Y. Yoshimura. 2002.
Hydrodesulfurization of petroleum feedstoeks with a new type of nonsulfide hydrotreating
catalyst. Journal f Catalysis. 209: 1-5.

Papa, D., Schwenk, E., and H.F. Ginsberg. 1949. Reduetions with nickel-aluminum alloy and
aqueous alkali. Part VII. Hydrogenolysis of sulfur compounds. Journal o f Organic
Chemistry 14: 723-731.

Pareek, V.K., Adesina, A.A, Srivastava, A., and R. Sharma. 2003. Modeling of nonisothermal FCC riser. Chemical Engineering Journal. 92: 101-109.

Pettit, G.R. and E.E. van Tamelen. 1962. Organic Reactions. New York. Pp. 356-509.

Reichardt, C. 1988. Solvents and Solvent Effects in Organic Chemistry. VCH, Germany. Pp
426-427.

Schwedt, G. 2001. The Essential Guide to Environmental Chemistry. John Wiley & Sons Ltd.
West Sussex, United Kingdom. Pp. 48-72.

Sertic-Bionda, K., Kuzmie, V., and M. Jednacak. 2000. The influence of process parameters
on catalytic cracking LPG fraction yield and composition. Fuel Processing Technology
64: 107-115.

Srivastava, S., Minore, J., Cheung, C.K., and W.J. le Noble. 1985. Reduction of Aromatic
Rings by 2-propanol with Raney nickel catalysis. Journal o f Organic Chemistry 50: 394396.

Takahashi, A., Yang, F.H., and R.T. Yang. 2002. New sorbents for desulfurization by H complexation: thiophene/benzene adsorption. Industrial and Engineering Chemistry
Research 41: 2487-2496.

39

Yin, C., and D. Xia. 2004. A study of the distribution of sultur compounds in gasoline
produced in China. Part 3. Identification of individual sulfides and thiophenes. Fuel 83:
433-441.

40

Chapter 2: Chemical Methods Studied for the Desulfurization of FCC
Gasoline
2.1 Introduction
Sulfur in fuels has been an environmental concern, particularly the transformation
sulfur undergoes during combustion (Zannikos et al., 1995). Current Canadian legislation
allows 30 ppmw as the pool average for gasoline (CEPA, 2004). There are several different
methods used to desulfurize fuels. However, chemical methods are preferred (Karaca and
Yildiz, 2005). The most acceptable method for sultur removal currently is hydrotreating
(Zannikos et al., 1995).
Hydrotreating removes sultur from gasoline by transforming it into hydrogen sulfide
(Meyers, 2004). The problem with hydrotreating is that in order to meet current legislation,
deep desulfurization processes need to be undertaken. Deep desulfurization refers to
technology used to reduce sulfur content to less than 0.05 wt % sulfur (Ishihara and Kabe,
1993). When conventional hydrotreating technology is used for deep desulfurization, there is
a significant reduction in the octane number. This reduction in octane is due to the saturation
of olefins from fluid catalytie craeking (Ma et al., 2002). This deep desulfurization uses large
amounts of hydrogen, which cannot be produced alone by the catalytic reformers, as done
with traditional hydrotreating (Zannikos et al., 1995). Therefore, other processes are needed
to reduce the fuel sulfur levels without the addition o f hydrogen.
In fluid catalytically cracked gasoline (FCC gasoline) H 2 S, sulfides and mercaptans
appear only in trace amounts. The most abundant suhur-containing compounds in FCC
gasoline are thiophenes, benzothiophenes, and tetrahydrothiophenes. Therefore, other

41

methods may be necessary and more efficient at removing these heavier sulfur-containing
compounds. One characteristic of these heavier sulfurons compounds that is different from
other compounds in a gasoline mixture is that they are slightly more polar than the
hydrocarbon molecules (Zannikos et al., 1995). This increases the potential of their selective
removal from the gasoline.
2.1.1 Solvent Desulfurization
The purpose of this technique is to take advantage of the similar solubility
characteristics between the solvent and the desired solute to be removed or separated.
Solvents have been used in refining technologies for many years. Several solvents are
currently used for sulfur removal. However, they are typically used for the removal of H 2 S,
sulfides and mercaptans. Processes that currently employ solvents include amine-treating
units, the Merox process and sulfur scrubbers. For example, methyldiethanolamine (MDEA)
is very selective for H2 S removal in the amine treating units (Jou et ah, 1997).
In order to be able to recover an analyte, such as the sulfur compounds in gasoline, the
solvent needs properties similar to the analyte being extracted. The two most common
properties include partition coefficients and solubility parameters (Reichardt, 1988). Our
present study will focus on solubility parameters. Solvents with similar solubility properties
can form mixtures; therefore being capable of predicting which solvents can be used for
analyte recovery (Fitzpatrick and Dean, 2002). In order for a solvent to be effective at
extracting an analyte, there must be a high degree of separation, high selectivity, no tendency
to form emulsions and rapid separation (Reichardt, 1988). The extent of the separation can be
determined by the density difference of the two solvents or the ratio of the partitioning

42

coefficients (Reichardt, 1988). In order to match the chemical properties, a comparison of
solubility parameters can be used to narrow the list of solvents to be used for a particular
analyte.
There are several different solubility parameters with one of the most common being
the Hildebrand solubility parameter. The Hildebrand solubility parameter (ôt), or total
solubility, is a measure of the internal energy of cohesion (Fitzpatrick and Dean, 2002). The
total solubility is based on three separate components: hydrogen-bonding ability (5 h ) ,
dispersion coefficient (ôd), and polarity coefficient (ôp) (Fitzpatrick and Dean, 2002). These
solubility parameters have been determined for a variety o f different solvents and can be found
in the literature (Barton, 1991). Solvent extractions typically are based on hydrogen bonding
or van-der-Waals interactions; however, it may be necessary for stronger chemical bonding
rather than just a physical interaction to remove the desired solute.
2.1.2 Adsorption and Precipitate Desulfurization
Other chemical properties of a compound can also be used to separate it from
mixtures, including its polarity. Typically, stronger chemical reactions occur with sorbent
extractions. Therefore, solvent extractions are less selective and effective than most sorbent
extractions.
As FCC gasoline is a complex mixture, it may be important to look at the differences
between the sulfur compounds in the gasoline and the rest of the mixture. For example,
thiophene has a lone pair o f electrons on the sulfur; therefore, it should be possible to cause a
reaction between the lone pair of electrons and an acid or metal salt. Adsorption
desulfurization is the ability of a solid sorbent to selectively adsorb the sulfur containing

43

compounds; where as precipitate desulfurization is based on the formation o f an insoluble
precipitate (Babich and Moulijn, 2003). The efficiency of precipitate desulfurization is low.
However, sorbent desulfurization shows a lot of promise.
There is evidence that metal oxides react with alkanethiols by reacting the thiol with
the metal replacing the oxygen and producing water as seen in the following reaction (Nehlson
et al., 2003).

f 6 0 + 2Æ9/f

(2.1)

In this reaction, the thiols are transformed into insoluble metal thiolates. These thiolates can
then be filtered from the hydrocarbon stream (Nehlson et al., 2003).
Takahashi et al. (2002) identified copper and silver zeolites, which could he used to
remove thiophenes at room temperature. Similarly, Ma et al. (2002) developed a deep
desulfurization method using selective adsorption onto a transition metal. There have also
been examples o f activated carbon and other zeolites, which cause sulfur to form a monolayer
on the surface o f the sorbent (Mikhail et al., 2002). Limestone has been used as an absorbent
for coal processing in order to remove H2 S (Nakazato et ah, 2003). Finally, charcoal,
petroleum coke, cement kiln dust and clay minerals have also been investigated as sorbents
with the acidic clays being shown to be good adsorbents (Mikhail et ah, 2002).
2.1.3 Desulfurization using basic and acidic solutions
Typically, basic and acid solutions have been used to remove sulfur from coal. There
are, however a few instances where it has been attempted for liquid fuels (Mukherjee and
Borthakur, 2001). There have been several compounds studied, including chemicals such as
nitric acid, sulfuric acid, hydrogen peroxide, and sodium hydroxide (Karaca and Yildiz, 2005).

44

In addition, sodium hydroxide is used to help sweeten fuels by removing H2 S (Meyers, 2004).
There are also publications, which discuss the use of heated sodium hydroxide to desulfurize
coal (Mukherjee and Borthakur, 2001). Murkherjee and Bothakur (2001) also determined
that by following a sodium hydroxide treatment with a hydrochloric acid treatment, greater
desulfurization could be achieved. A more recent publication discusses desulfurization by
leaching the fuel with sulfuric acid (Karaca and Yildiz, 2005). This article concludes that
H 2 S O 4

is effective at removing thiols, sulfides and disulfides from fuel oils.
The purpose o f our research is to determine if other solvents, sorbents or basic/acidic

solutions can be used to remove sulfurous components from FCC gasoline. This work will
focus on extractions carried out mostly under ambient conditions.
2.2 Methodologv
2.2.1 FCC Gasoline Sampling and Storage
The FCC gasoline was sampled after the fluid catalytic cracking unit and Merox
processing unit at Husky Energy’s refinery in Prince George. A tap sampling method was
used, where the tap was allowed to run for several minutes before the FCC gasoline samples
were collected. The FCC gasoline was sampled into 1-litre amber, glass bottles with
approximately 100 mL of headspace. Two 1-litre samples were taken on August 4, 2004 and
four 1-litre samples were taken on August 27, 2004. The FCC gasoline samples were
transported to UNBC and stored in a refrigerator at about 4 *^C for no longer than six months,
after which a new sample would be collected. As storage could cause volatile portions of the
FCC gasoline to be lost a baseline sample was always taken and analyzed for the sulfur
concentration with each batch of experiments performed.

45

2.2.2 Solvent Methodology
The ten solvents investigated were carefully selected based on their Hildebrand
solubility parameter (Barton, 1991). The solubility parameters were as closely matched to the
solubility of five difierent sulfur compounds. The five sulfur compounds are: carbon disulfide,
ethyl mercaptan, diethyl sulfide, thiophene, and tetrahydrothiophene. These five compounds
were selected based on a GC/SCD analysis performed on the FCC gasoline confirming the
presence of these five components. The solubility parameter of a selected, potentially the
most compatible, solvent was matched up directly with one of the above sulfur compounds.
Once the best solvents were determined, a separatory funnel extraction was performed.
There were two different separatory funnel methods employed. The first method was
a one-time addition of the solvent, while the second method was a sequential addition o f the
solvent. The volume o f solvent was 5, 10, or 15 millilitres depending on the available volume.
Once the solvent had been added, the separatory funnels were shaken vigorously for one
minute. At this point, if two layers appeared, the lower layer was decanted from the
separatory funnel and the remaining FCC gasoline layer was emptied into a 125 mL amber
bottle for storage at 4 °C. The sulfur analysis was performed on these samples within 48
hours. If two layers were not present then the entire sample was emptied into a 125 mL flask,
stoppered and placed in the freezer at a temperature o f -10 °C to cheek the effect of
temperature on the separation of the solvent and FCC gasoline. Samples remained in the
freezer for 48 hours. Once placed in the freezer, if two layers formed, the FCC gasoline was
decanted from the sample and placed in a 4 "C refrigerator until analysis, which was
performed within 48 hours. The FCC gasoline samples that separated from the solvent were
analyzed using a Horiba energy-dispersive X-ray spectrometry system.

46

The analysis was performed at Husky Energy’s refinery in Prinee George. An energydispersive X-ray fluoreseence (XRF) system was operated according to the ASTM method D
6445 (Annual book of ASTM standards, 2003). The instrument was a Horiba Sulfur-in-Oil
Analyzer (model # SLFA-1800). The instrument measured the sample for 100 seconds and
this was repeated three times for each sample. A check standard was performed first in order
to check the operation of the Horiba. If the check standard passed within 5 % of the actual
value, the samples were then analyzed. If the standard did not pass within 5 % of the actual
value, the instrument was recalibrated and another check standard was performed. Each
sample was analyzed in triplicate with the average being reported in mass percent. A check
standard was done every six to ten samples with a check standard being completed at the end
o f that day’s analysis. An F-test and Student T-test were performed on the data when
necessary using Microsoft Excel.
In addition, promising samples were analyzed using gas chromatography equipped
with a flame ionization detector (GC/FID). This was done in order to monitor any major
changes in the composition of hydrocarbons. The GC/FID used was a Hewlett Packard
HP6890 series GC system with a capillary column HPl (100.0 m in length, 250.0 pm
diameter, and 0.50 pm film thickness). There was a split injection method using an injection
volume of 0.2 pE. The system operated on hydrogen gas with a flowrate of 30 mL / min and
a helium flow of 30 mL / min. The oven heat up was stepped up at a rate of I °C per minute
until 50 °C, then 2 °C per minute until 130 °C and finally at a rate of 4 °C per minute until a
final temperature o f 180 °C. The total run time was equal to 137.5 minutes.

47

2 2.3 Sorbent Methodology
Three different aqueous solutions were seleeted for the next portion of the experiment
based on the papers by Nehlson et al (2003) and Ma et al (2002). In these publications,
promising results were shown with similar solutions. The three solutions were prepared in
water at: 5 wt. % barium nitrate, 5 wt. % lead nitrate, and 10 wt. % zinc chloride. The first
set of extractions performed was by sequential addition. First, 10 mL of the solution was
added into a separatory funnel containing 50 mL of FCC gasoline. The solutions were shaken
for approximately one minute, and then allowed to settle for about 15 minutes. The aqueous
solution was then decanted and the procedure was repeated two more times, for a total of 30
mL o f aqueous solution. The remaining FCC gasoline was then placed in the refrigerator until
sulfur analysis o f the FCC gasoline could be performed, which was usually within 48 hours.
Another method was also used for the 5 wt. % barium nitrate and 5 wt. % lead nitrate
solutions. In this method, 30 mL o f the solution was added to 50 mL of FCC gasoline and
then it was stirred vigorously for one hour on a stirring plate. The samples were then allowed
to settle for 15 minutes and the aqueous solution was decanted. The remaining FCC gasoline
was placed in the refrigerator; analysis was performed within 48 hours.
Again, the analysis was performed at Husky Energy’s refinery in Prince George.
Some samples in the sorbent experiments were analyzed using an XRF system, which was
operated according to the ASTM method D 6445 (Annual book of ASTM standards, 2003).
Other samples were analyzed using the Ultraviolet fluorescence (UVF) method. The
Ultraviolet fluorescence method used an Antek Elemental Analyzer 9000 operated according
to the ASTM method D 5453 (Annual book of ASTM standards, 2003). The oxidative

48

furnace was operated at 1075 °C. A cheek standard was performed first in order to cheek the
operation of the Antek. If the check standard passed within 5 wt. % of the actual value, the
samples were then analyzed. If the check standard failed then the instrument was recalibrated
and another check standard was performed to make sure the new calibration was accurate.
Each sample was analyzed in triplicate with the average being reported in parts per million. A
cheek standard was analyzed after every six to ten samples with a check standard being
completed at the end o f that day’s analysis.
2.2.4 Basic and Acidic Methodology
The solutions in this portion were chosen based on publications by Karaca and Yildiz
(2005), and Murkherjee and Bothakur (2001), which showed promising results with similar
solutions. This methodology was exactly the same as the sequential method attempted for the
sorbents in the previous section. A basic solution o f 5 wt. % sodium hydroxide and an acidic
solution of 10 vol. % sulfuric acid were used for this portion of the experiments. The samples
were analyzed using either XRF or UVF described in the previous section.
2.3 Results
2.3.1 Solvent Results
First, the ten solvents identified to be tested are listed in Table 2.1. These solvents
were chosen from the CRC Handbook of solubility parameters and other cohesion parameters
(Barton, 1991). They were purchased from a variety of suppliers with each having 97 % or
better purity. The table also shows that the ten solvents had similar solubility parameters to
one of the five sulfur compounds: carbon disulfide, ethyl mereaptan, diethyl sulfide, thiophene

49

and tetrahydrothiophene. As seen, the solubility parameters of the five sulfur eompounds do
not differ by more than 3.1 units.
Table 2.1: Ten solvents selected based on solubility parameters similar to five selected sulfur
Sulfur
Compound
Carbon
Disulfide

Ethyl
Mercaptan

Diethyl
Sulfide

Solvent

Supplier and Purity

20.3
Diethyl maleate

Aldrich, 97 %

20.3

2,3-Butanedione

Aldrich, 97 %

20.3

Diethylene glycol monomethyl
ether

Alfa Aesar, 99 %

20.3
18.6

Butyraldéhyde

Aldrich, 99 %

18.6

Diethyl pimelate

Alfa Aesar, 98 %

18.6

3-(Dimethylarnino)propylarnine

Aldrich, 99 %

18.6
17.4

Heptylbenzene

Aeros Organics, 97 %

17.4
20.0

Thiophene
Dimethyl pimelate
Tetrahydro­
thiophene

Solubility
Parameter

Aldrich, 99 %

20.0
20.5

1-Decanol

Baker, 99 %

20.5

o-dichlorobenzene

Lancaster, 99 %

20.5

The results o f the individual solvent tests can be separated into two categories: those
that showed a separation, and those that did not show a separation. A summary of the results
for the ten different solvents can be seen in Table 2.2. Table 2.2 shows that only 1-decanol
separated and this separation occurred only at low temperatures when the 1-deeanol froze.
Approximately half o f the sample was frozen on the bottom of the sample container. It was
determined that the frozen section was the solvent as 1-deeanol has a melting point of 7 °C.

50

The remaining liquid was deeanted and approximately 34 mL was reeovered. The sample was
analyzed using XRF. The original FCC gasoline contained a sulfur concentration of 208 ppm
while the 1-decanol extracted sample contained a sulfur concentration of 191 ppm. This was
an 8 % reduction in the sulfur concentration.
Table 2.2: Summary o f the separation between the solvent and FCC gasoline.
Solvent
Separation Observations
diethyl maleate

None

None

dimethyl pimelate

None

None

diethyl pimelate

None

Amber Colour

1-decanol

Froze

None

diethylene glycol monomethyl
ether

None

None

3-dimethylaminopropylamine

None

None

Butyraldéhyde

None

Yellow Colour

2,3 butanedione

None

Red precipitate

1,2 dichlorobenzene

None

None

Heptylbenzene

None

None

Since 1-decanol showed a reduction in the sulfur concentration, further tests were
performed. The first test performed was to determine if one-time addition or sequential
addition was more effective for removing sulfur. The one-time addition used 15 mL of 1decanol; the sequential addition used 5 mL aliquots of 1-decanol for a total of 15 mL. The
samples were placed in the freezer to allow the 1-decanol to freeze and separate. The samples
were partially frozen within 24 hours of extraction, however if left for another 24 hours the
sample became completely frozen. In both samples, approximately 35 mL of FCC gasoline
was recovered from the frozen samples. The samples were analyzed using XRF to determine

51

the sulfur concentrations. The initial concentration o f the FCC gasoline was 236 + 10 ppm.
The one-time addition sample had a concentration o f 227.8 ± 6.8 ppm, while the sequential
addition had a concentration o f 224.7 ± 6.5 ppm. Analysis showed that the two samples were
within the standard deviations o f the original sample therefore, it was concluded that no sulfur
was removed.
An F-test was performed that determined the F-value to be 1.09 and the F-critieal
value to be 5.05 for a 95 % confidence interval with 5 degrees of freedom for each value. As
the F-value is less than the F-critical value, no distinction in the variances can be determined.
Next, a Student T-test determined that there was also no significant difference between the
two mean values within a 95 % confidence interval.
The final analysis using 1-decanol involved adding 30 mL of 1-decanol to 50 mL of
FCC gasoline, covering it and stirring using a stirrer plate for one hour. This resulted in 66
mL of sample being recovered from the initial 80 mL. The sample was placed in the freezer
overnight; similarly, if the sample were in the freezer for too long the entire sample would
freeze solid. The analysis was again performed using XRF. The original FCC gasoline had a
concentration of 268 ± 7.7 ppm while the 1-decanol extracted sample had a concentration of
200 ± 7.7 ppm. This was approximately a 25 % reduction in sulfur concentration. However,
as the entire sample freezes when left long enough, it is possible that the sample was simply a
mixture of FCC gasoline and 1-decanol and the sulfur concentration decreased due to dilution.
In order to investigate if any 1-deeanol remained in the sample a GC/FID analysis was
performed. Figure 2.1 is the original hydrocarbon analysis of FCC gasoline using GC/FID
while Figure 2.2 is the GC/FID analysis of the FCC gasoline extracted with 1-decanol. It was

52

apparent from the chromatograms that there was a large unidentified peak at the end of the
analysis in Figure 2.2 compared to the lack of a peak in Figure 2.1.

27,5 -

175

125
-o

mj
7.5

105

1075

110

115

117.5

Figure 2.1: GC/FID analysis of original FCC gasoline.

53

l,z u im u o ^ / M u ^ r u ^ u i . u ;
dA

r 5-

1-decanol

~ 5 -

100

102

104

106

108

110

114

Figure 2.2: Analysis of FCC gasoline extracted with 1-decanol using GC/FID.
The rest of the ten solvents did not show any significant separation, however some
changes in three of the samples were observed. The solution o f diethyl pimelate and FCC
gasoline, when placed in the freezer overnight, showed a colour change from colourless to
amber, but there was still no separation. A small sample of this mixture was placed in the
fumehood to evaporate to see if the colour change was the result of a non-volatile compound.
Once the sample evaporated, there was no evidence of a precipitate, so 25 mL of the partially
frozen sample was heated at a temperature of 25 °C and there was still no change. No change
was observed until the sample was heated to 130 °C when the sample became a brighter
yellow colour. At a temperature of 150 °C, a portion of the FCC gasoline should have boiled

54

off and as the boiling point of the solvent is 192 °C, the solvent, assuming ideal behaviour,
should have remained in the liquid solution as its boiling point was not reached. This would
not allow the diethyl pimelate to change into the gaseous phase.
Next, the butyraldéhyde extracted sample was placed in the freezer and twelve hours
later, it appeared to be more yellow than the original FCC gasoline. Twenty-five millilitres of
the sample was heated in the tumehood. The boiling point of butyraldéhyde is 75 °C so the
sample was heated to 70 ‘’C. At approximately 51 °C, the sample began to appear more
yellow in colour.
Finally, 2,3-butanedione showed no separation, however a small reddish precipitate
was visible on the sides o f the separatory funnel after shaking. Fifty millilitres of additional
FCC gasoline was used in an attempt to rinse the crystals from the sides of the separatory
funnel into the sample beaker. This, however, did not work so acetone was used to rinse the
crystals into a separate beaker; the crystals dissolved in the acetone. Both the original sample
and the acetone containing the crystals were placed in the freezer. After 12 hours, more
precipitate was visible on the bottom of the sample container. There was no change in the
acetone sample so it was transferred onto a watch glass and placed in the fumehood to
evaporate the acetone. After approximately 10 minutes in the fumehood, the sample became a
pinkish colour and was more viscous. After 20 minutes, the sample was still pink with
noticeable granules suspended in the liquid. After 2 hours, a syrup-like liquid was left that
was red in color. The sample was left for two days and it was found that there was a sticky
red residue left on the watch glass. Although this residue could be re-dissolved in methanol,
no further analysis into its identity or structure was attempted at this time. These results

55

indicate that another chemical approach is needed to remove sulfur-containing compounds
from FCC gasoline rather than just by using solvents.
2.3.2 Sorbent Results
First, the layer eontaining barium nitrate separated from the FCC gasoline. The
sample was analyzed using XRF. The original FCC gasoline had a sulfur eoncentration of 230
± 5 ppm and the barium nitrate extracted sample had a sulfur concentration o f 236 ± 12 ppm.
No significant difference between the mean value of the original FCC gasoline and the barium
nitrate extracted sample was found.
The layer containing lead nitrate separated from the FCC gasoline so it was analyzed
using XRF. The original FCC gasoline had a sulfur concentration of 230 ± 5 ppm and the lead
nitrate extracted sample had a sulfur eoncentration of 243 ± 6 ppm. The student T-test again
showed that there is no signifieant difference between the means of the two samples at a 95 %
confidence limit.
The samples involving a one-time addition of the solution and stirring for one hour
showed interesting results. When the samples were analyzed after the one-time addition the
sulfur eoncentration was much higher in all cases than the original concentration. The original
concentration of the FCC gasoline was 258 ± 8 ppm however, the concentration of sulfur in
the extracted samples was 498 ± 3 ppm for the barium nitrate sample and 504 + 8 ppm for the
lead nitrate sample. Using a student T-test, these values are significantly different than the
original concentration based on a 95 % confidence level. The large values are not logical;
however, the XRF method is known to typically have problems such as matrix interference.
To test this, a sample o f the barium nitrate extracted sample was sent to Core Laboratories in

56

Calgary for analysis using a gas chromatography / sulfur chemiluminescence detector. The
sample showed an increase in sulfur concentration to 310 ± 10 ppm from 268 ± 10 ppm. It
was not significantly different than the original FCC gasoline. The matrix interference
appeared to be less with GC/SCD than with the XRF.
The sample extracted with a 10 wt. % zinc chloride solution showed a separation from
the FCC gasoline. The FCC gasoline originally had a concentration of 243 ± 12 ppm while
the extracted sample had a sultur concentration of 242 ± 8 ppm. The UVF method was used
as matrix interference is not a problem and the method became available as well. The Student
T-test showed that there is no significant difference between the mean of the original sample
and the mean of the extracted sample.
2.3.3 Basic and Acidic Results
A separation occurred for the sequential addition of sodium hydroxide, the sodium
hydroxide solution was then separated from the FCC gasoline. The sample was placed in the
refrigerator with analysis being performed within 48 hours. The sample was analyzed using
XRF. The original FCC gasoline showed a sulfur concentration of 230 ± 5 ppm while the
extracted sample showed a concentration of 241 ± 3 ppm. When a Student t-test was
performed, there was no significant difference between the mean o f the original FCC gasoline
and the sodium hydroxide extracted sample.
The solution of 10 vol. % sulfuric acid also separated from the FCC gasoline. In Table
2.3, the results from a variety of sulfuric acid tests are given. The difference in the samples
was the type of rinse performed after the initial extraction with sulfuric acid. Once the original
sample was analyzed using UV fluorescence, it was apparent that not all the sulfuric acid had

57

been removed from the sample. Therefore, further rinsing was needed to ensure all the
sulfuric acid had been removed from the FCC gasoline, including a rinse with a solution of
acid cadmium chloride that has been shown to remove hydrogen sulfide from gasoline samples
(UOP, 1980).
Table 2.3: Summary table of the four levels of rinsing performed on a sample extracted with
10 % sulfuric acic
Rinsing Method
Mean
Standard
Student T10 %
Deviation
test Value
Sulfuric Acid
1.

20 mL 10 wt. %
NaOH

254.36

6.70

0.12

2.

50 mL 10 wt. %
NaOH

251.72

5.31

0.10

3.

50 mL 10 wt. %
NaOH

248.44

9.11

0.05

4.

20 mL 100 wt. % acid
cadmium chloride

237.80

10.46

0.04

The initial sample was rinsed with 20 mL of 10 wt. % NaOH and then it was analyzed.
After analysis, it was necessary for additional rinses to be performed on the sample. Table 2.3
shows the decrease in sulfur concentration with each subsequent rinse. The Student T-test
showed that there is no difference between the mean values of the four different rinses for a
95 % confidence interval. Therefore, there was no reduction in the sulfur concentration with
the use of a sulfuric acid solution.
2.4 Discussion
2.4.1 Solvent Desulfurization
The ten solvents chosen for this experiment showed less than 3.1 units difference m
solubility parameters between all solvents and all five sulfur compounds. According to

58

Reichardt (1988) two liquids will be miscible in each other if their Hildebrand solubility
parameter differs by no more than 3.4 units. Therefore, the ten solvents should have been
effective for any of the five sulfur compounds. However, this is only true for pure mixtures.
As FCC gasoline contains over 200 compounds, it is considered to be a complex solvent
mixture, which is miscible in many other solvents as seen in Table 2.2.
The only solvent, which showed a promising result, was 1-decanol. The separation,
which occurred due to the high melting point of 1-decanol allowed for further tests to be
performed. Although a decrease in the sulfur concentration was determined for both the one­
time addition and the sequential addition, there was no statistical difference in the values as
shown by the Student T-test. However, as was stated above, the entire sample would freeze if
it remained in the freezer for an extended period of time.
A gas chromatography analysis showed that the 1-decanol extracted sample had a
large peak at near 115 minutes. This large peak is not present in the original FCC gasoline
and is most likely the 1-decanol solvent or possibly a new compound formed through the
extraction process. Therefore, the reduction in sulfur concentration seen in all the
experiments using 1-decanol was most likely a result of the sample being diluted by the 1decanol, with a portion o f the FCC gasoline remaining in the frozen 1-decanol. Therefore, 1decanol was not effective at desulfurizing FCC gasoline.
The three solvents, which showed signs of a reaction during the extraction, can be
seen in Table 2.2. Diethyl pimelate showed a colour change when left overnight in the
freezer. It was determined that the color change was not a result of a precipitate as when the
sample was evaporated there was no precipitate found. However, a distillation would need to

59

be performed to determine if the sulfur is remaining in the solvent. The solvent butyraldéhyde
also showed a colour change when placed in the freezer overnight, which was most likely due
to a similar mechanism as the diethyl pimelate. Therefore, potentially more work could be
done with these two solvents to determine if there is a possible distillation or other separation
method, which could occur in order to separate the solvent from the FCC gasoline.
Finally, the solvent 2,3-butanedione, formed a red precipitate when it was mixed with
the FCC gasoline. In order to identify the red, viscous compound it was analyzed using
GC/MS, however it was not possible to identify, as it was not volatile enough. However, it
was observed during the experiments that the caustic solution added to FCC gasoline during
the Merox process does form a red precipitate with certain fuels. Therefore, it is possible that
the red precipitate observed was just a reaction between the caustic solution in the FCC
gasoline and the solvent.
The fundamental objective that was being searched for was a separation between the
solvent and the FCC gasoline. However, since gasoline is a complex mixture, there was no
separation for the ten solvents with the exception o f 1-decanol at cooler temperatures. In
each case of solvent-solvent extraction, it appears that simple van-der-Waals forces are not
enough to remove the sulfur from the FCC gasoline. However, a ‘solvent cocktail’ has been
shown to remove sulfur compounds from refinery streams and therefore a mixture of solvents
may be required for desulfurization o f FCC gasoline (Babich and Moulijn, 2003). As there
were no promising results seen with the use of the solvents selected, it is possible that a
stronger interaction may be needed to remove sulfur-containing compounds from FCC
gasoline. Therefore, some sorbent reactions were attempted which should have stronger

60

interactions/reactions as they are based on the reaction of the lone pair of electrons on
thiophenes with the sorbent.
2.4.2 Adsorption and Precipitate Desulfurization
There have been publications where metal salts or metal oxides have been used to
convert thiols into metal thiolates (Nehlson et al., 2003). Therefore, three solutions with
metal ions in aqueous solutions were attempted. However, in all three cases there was no
reduction in the sulfur levels observed. It is interesting to note that the energy-dispersive Xray spectrometry consistently measured the extracted samples higher than the original
samples. This was especially apparent in the metal samples that were stirred for one hour. A
potential reason for this is that the energy-dispersive X-ray system can have problems such as
matrix interference when solvents are used and are not fully removed from the samples before
analysis (ASTM, 2003). Therefore, if some of the metal solution remained in the sample, the
matrix is different from the FCC gasoline matrix, which it is calibrated for (ASTM, 2003).
This difference in the matrix may cause false values to be reported. This can occur when the
sample is exposed to the X-ray the solvent may emit radiation at the same wavelength as
sulfur causing the instrument to believe the solvent is also sulfur, thus giving a false reading.
For this reason a sample was sent to be analyzed using GC/SCD at Core laboratories.
However, a significant difference could not be established using a Student T-test.
Finally, in order to prevent this problem in the future, a UV fluorescence method was
made available. This method does not have matrix interference problems, as the sample is
combusted so everything in the sample is converted into CO2 , H2 O, or SO2 . Therefore, when

61

the sample containing zinc chloride was analyzed, the values were very similar with no
significant difference being identified.
The reaction of the aqueous metal salts was not effective at removing sulfur from the
FCC gasoline. Although other papers have shown metal compounds to be effective, it may be
important that the metal contain an oxide component. A publication by Nehlson et al (2003)
used a variety of metal oxides and metal salts and they concluded that oxides or hydroxides of
Pb, Hg(II), and Ba were the most effective. Therefore, the sample extracted using barium
nitrate should have shown a promising result, but did not possibly due to the lack of oxides or
hydroxides. It should be noted in Nehlson et al (2003) that the experiments did not use an
actual fuel sample but instead a mixture of 1-octanethiol was used. Therefore, the complex
mixture of FCC gasoline may not show the same results. An attempt was then made to react
the more polar sulfur compounds with a basic or acidic solution.
2.4.3 Desulfurization using basic and acidic solutions
The final experiment o f a separatory funnel extraction was attempted using a basic
solution of sodium hydroxide and an acidic solution of sulfuric acid. Sodium hydroxide has
been used for many years to remove hydrogen sulfide using the Merox process (Meyers,
2004). As this sample was analyzed before the UV fluorescence method was available, a
similar matrix interference problem was found. However, the Student T-test showed that
there was still no significant difference evident between the original FCC gasoline’s sulfur
concentration and that of the extracted sample. There is a publication about the use of sodium
hydroxide in reducing the sulfur concentration in coal (Mukherjee and Borthakur, 2001). In
coal samples it was found that the sodium hydroxide solution needed to be heated to 190-200

62

°C in order to remove sulfur (Mukherjee and Borthakur, 2001). However, at these
temperatures, most o f the FCC gasoline would all have been boiled off, as the final boiling
point of the FCC gasoline is 210.5 °C.
Finally, the experiments with sulfurie aeid determined that much of the sulfuric acid
remained in the sample after decanting the aqueous solution. This was shown by the large
increases in sulfur concentration in the extracted FCC gasoline samples. As these samples
were analyzed using the UV fluorescence method, there was no matrix interference problem.
Therefore, a stronger rinsing method was required in an attempt to remove all of the
remaining sulfuric acid. As seen in Table 2.3, with each rinse of sodium hydroxide, there was
a decrease in the sulfur concentration, but none were significantly different from the original
FCC sample. The final rinse using an acidic cadmium chloride solution has been known to
remove HaS from gasoline samples before analysis (UOP Laboratory Test Methods, 1980).
2.5 Conclusion
In conclusion, it is apparent that the ten above-mentioned solvents are all somewhat
soluble in FCC gasoline and therefore cannot be used in a simple shake flask extraction
method for sulfur. This is most likely due to the complex mixture of FCC gasoline and its
solubility in most organic solvents. Perhaps trials with aqueous solutions would be more
effective as aqueous solutions are not as soluble in FCC gasoline. The experiments with the
metal salts showed more promise because separation occurred between the aqueous salt
solutions and the FCC gasoline. However, the metal salt extractions showed no signifieant
reduction in sulfur concentration. The basic and acidic solutions also separated from the FCC
gasoline; however there was again no significant difference in the sulfur concentration.

63

Therefore, in this work, a simple shake flask extraction using a single solvent or solution is not
effective for desulfurization of FCC gasoline. This study shows that intermoleeular forces
such as van-der-Waals interactions are normally not strong enough to remove sulfur from
FCC gasoline. Due to time constraints and limited resources, there is room for future research
with each of these compounds, particularly secondary solvents. However, it may be more
suited to a person with significant knowledge of solvent chemistry. A stronger chemical
interaction seems to be required to desulfurize FCC gasoline at ambient conditions, perhaps
through a catalytic reaction or through reactive adsorption.

64

2.6 Literature Cited
Annual Book of ASTM Standards. 2003. Petroleum Products. Lubricants, and Fossil Fuels.
Volume 3. ASTM International. West Conshohocken, PA. Pp. 1-1368.

Babich, I.V., and J.A. Moulijn. 2003. Science and technology of novel processes for deep
desullurization of oil refinery streams: a review. Fuel 82: 607-631.

Barton, A.F.M. 1991. CRC Handbook of Solubüitv Parameters and Other Cohesion
parameters. 2"^ edition. CRC Press. Boca Raton, Florida. Pp. 1-739.

Canadian Environmental Protection Act, 2004. http://laws.iustice.gc.ca/en/C-15.31

Fitzpatrick, L.J., and J.R. Dean. 2002. Extraction solvent selection in environmental analysis.
Analytical Chemistry. 74(1): 74-79.

Ishihara, A. and T. Kabe. 1993. Deep desulfurization of light oil. 3. effects of solvents on
hydrodesulfurization of dibenzothiophene. Industrial and Engineering Chemistry
Research. 32: 753-755.

Jou, F., Otto, F.D., and A.E. Mather. 1997. The solubility of mixtures of H2 S and CO2 in an
MDEA solution. The Canadian Journal o f Chemical Engineering. 75: 1138-1141.

Karaca, H. and Z. Yildiz. 2005. Desullurization of fuel by leaching using H2 O2 and H 2 SO4 .
Petroleum Science and Technology. 23: 285-298.

Ma, X., Sun, L., and C. Song. 2002. A new approach to deep desullurization of gasoline,
diesel fuel and jet fuel by selective adsorption for ultra-clean fuels and for fuel cell
applications. Catalysis Today. 77: 107-116.

Meyers, R.A. 2004. Flandbook o f Petroleum Refining Processes. 3‘^‘*edition. McGraw-Flill.
USA. Pp. 1-15.40.

Mikhail, S., Zaki, T., and L. Khalil. 2002. Desulfurization by an economically adsorption
technique. Applied Catalysis A: General. 227: 265-278.

65

Mukherjee, S. and P.C. Borthakur. 2001. Chemical demineralization/desulphurization of high
sulphur coal using sodium hydroxide and acid solutions. Fuel. 80: 2037-2040.

Nakazato, T., Lin, Y., Kusumoto, M., Nakagawa, N., and K. Kato. 2003. H 2 S removal by fine
limestone particles in a powder-particle fluidized bed. Industrial & Engineering Chemistry
Research. 42: 3413-3419.

Nehlson, J.P., Benzinger, J.B., and I.G., Kevrekidis. 2003. Removal of alkanethiols from a
hydrocarbon mixture by a heterogeneous reaction with metal oxides. Industrial and
Engineering Chemistry Research 42: 6919-6923.

Reichardt, C. 1988. Solvents and Solvent Effects in Organic Chemistrv. VCH, Germany. Pp
426-427.

Takahashi, A., Yang, F.H., and R.T. Yang. 2002. New sorbents for desulfurization by ft
complexation: thiophene/benzene adsorption. Industrial and Engineering Chemistry
Research 41: 2487-2496.

UOP Laboratorv Test Methods. 1980. Trace sulfur in petroleum distillates by the nickel
reduction method. UOP Method 357-80.

Zannikos, P., Lois, E., and S. Stournas. 1995. Desulfurization of petroleum fractions by
oxidation and solvent extraction. Fuel Processing Technology 42: 35-45.

66

Chapter 3: Activation Methods for Raney Nickel in the Desulfurization of
Gasolines
3.1 Introduction
As the previous ehapter determined it is neeessary to have a more selective and
stronger molecular interaction or reaction in order to remove the less reactive sulfur
compounds, such as thiophene. This can be seen in other sulfur removal technologies such as
the Claus process, which uses iron oxides or chromium oxides supported on alumina and
hydrotreating, which uses a variety of supported zeolites (Ertl et al., 1999). In both of these
technologies, catalysts are used to remove the sulfur-containing compounds from the feeds
(Meyers, 2004).
Within a refinery, both water and gas streams can contain H2 S, respectively called
“sour water” and “acid gas” (Meyers, 2004). Sour water is produced when steam is used as a
stripping medium in distillations or from the reduction o f hydrocarbon partial pressure in
catalytic cracking. Acid gas is usually formed in the amine regeneration and sour water
stripping units (Meyers, 2004).
A schematic of the Claus process is shown previously in Figure 1.7. In this process,
acid gas is added into a reaction furnace with air to combust a portion of the H2 S as seen in
reaction 3.1 below. The SO2 produced from the combustion reacts with the uncombusted
H2 S to form elemental sulfur as seen in reaction 3.2 (Meyers, 2004). The second part of the
Claus process involves two or three catalytic reactors. These reactors reheat the acid gas,
converting a portion of it to SO2 ; the stream is then cooled causing any remaining H2 S to be

67

converted to elemental sulfur by reacting with SO2 . The catalyst is typically a bauxite catalyst
(Matar and Hatch, 2001).

^ 2 »^ +

+

^

" ^^2 + ^ 2 ^

^ - ^ + 7/ 2(9

(3 . 1 )

(3.2)

A schematic diagram o f the hydrotreating process can he seen previously in Figure 1.8.
The hydrotreating catalysts are typically Co(Ni)Mo or NiW supported by aluminum oxide
(Coulier et al., 2002). The development of new hydrotreating catalysts other than the typical
Co(Ni)Mo catalysts is fast growing. Oyama et al (2002) have developed a new type of nickel
catalyst, which uses phosphide. The catalyst, Ni]?, has shown promise as 98 % of the sulfur
was removed from petroleum feedstocks compared to only 78 % with NiMo. The lack of a
layered structure, the moderate preparation temperatures, and inexpensive precursors make
this new catalyst promising. Other processing and analytical methods also use nickel-hased
catalysts for desulfurization.
Raney nickel is one of the most common skeletal metal catalysts and is prepared by
adding sodium hydroxide to a powdered nickel-aluminum alloy. This reaction causes the
nickel to form a porous structure so it is sometimes referred to as “spongy nickel” (Freel et
a l, 1969). The porous structure creates a high surface area, increasing the reaction surface
(Augustine, 1996). Raney niekel ean be used to effeetively remove sulfur-eontaining
compounds by cleaving the carbon-sulfur bond and replacing the sulfur with hydrogen, the
result is a nickel sulfide precipitate. Also, Raney nickel has been used as part of an analytical
method or purification method for years.

68

3.1.1 Activation o f Raney Nickel

The activation of Raney nickel with sodium hydroxide has been extensively studied
with a variety of activation procedures identified. These procedures were summarized earlier.
Through the activation process, large amounts of hydrogen are adsorbed through Van der
Waals forces by leaching the aluminum off the nickel (Beigi et al., 1999). The activation
reaction can be seen in Reaction 3.3 (Devred et al. , 2003):
2(M -^/)(,,) + 2 0 H

+6H20^^i^

Q 3)

The amount and temperature of the sodium hydroxide ean greatly affect the activity of
the Raney nickel (Augustine, 1996). The three methods of activation were based on the
information gathered from the literature. This includes the activation methods used by
Granatelli (1959), Beigi et al. (1999), Srivastava et al. (1985), and Bartok et al. (1987). The
activation process can have a variety of effects on the structure of the activated Raney nickel.
3.1.2 Structure

The leaching of Raney nickel creates a large surface area with the preparation
temperature having an impact on the porous structure and therefore, the surface area
(Fouilloux, 1983). At low preparation temperatures the pores have a narrower diameter,
whereas at higher preparation temperatures the pores are more cylindrical and create a higher
pore volume (Fouilloux, 1983). In general, Raney nickel is comprised of sponge-like particles
and, it is this porous structure that makes Raney nickel a high capacity catalyst or rapid
reactant for the removal o f sulfur.

69

3.1.3 Desulfurization with Raney nickel
There are two reaetion meehanisms by which Raney niekel can be used to remove
sulfur. The two mechanisms are as follows (Beigi et ah, 1999):
3RSR +3Ni —H

RSR + Ni -

—y RR + R R-\-R R +3A^/5' + 3//2(g) (3.4)

^ R H + R'H + M S

(3.5)

With the thiols and thiophenes, a cleavage of the carbon-sulfur bond with formation of
a carbon-hydro gen bond is the most common mechanism (Pettit and van Tamelen, 1962).
The effectiveness o f Raney nickel, for removing sulfur makes it a promising option for
desullurization of FCC gasoline.
The purpose of this chapter is to identify the best form of activated Raney nickel to
remove sulfur from FCC gasoline. In order to determine the most effective activated Raney
niekel the chosen method will need to use the smallest amount of Ni-Al alloy possible and the
lowest activation temperature possible to make it cost effective for large scale operations.
3.2 Methodologv
3.2.1 FCC Gasoline Sampling and Storage
The FCC gasoline was sampled as described in Chapter 2 at Husky Energy’s refinery
in Prince George. Two 1-litre samples were taken on March 07, 2005 and six 1-litre samples
were taken on June 14, 2005. The FCC gasoline samples were transported to UNBC and
stored as described in Chapter 2.

70

3.2.2 Raney Nickel Activation
Six activation experiments were studied in order to determine the most elective and
rapid activation method for Raney nickel. One experiment of each of the following leaching
processes was performed. The first experiment was low temperature activation similar to the
process used by Beigi et al (1999). This process used 3.7 ± 0.01 g of nickel-aluminum alloy
(Alfa Aesar / Lancaster, 50/50 %) and 10 mL of 10 % sodium hydroxide solution in a 100 mL
beaker. The solution was allowed to react until the vigorous evolution of hydrogen ceased.
The mixture was covered with Parafikn and allowed to react further for approximately 20 hrs.
After 20 hrs, the mixture was rinsed three times with distilled water and then three more times
with isopropanol, decanting between rinses. 10 mLs of isopropanol was then added to the
activated mixture.
The next three activation experiments were similar except the amount and
concentration of sodium hydroxide was altered. The second experiment used 20 mL of 10 %
sodium hydroxide, the third experiment used 5 mL of 20 % sodium hydroxide and the fourth
experiment used 10 mL of 20 % sodium hydroxide. It should be noted that the Raney nickel
had to be used for the experiments on the same day it was prepared.
The next two experiments were performed at high temperatures. The fifth experiment
was similar to a method used by Bartok et al (1987). This method used 3.0 ± 0.01 g of
nickel-aluminum alloy (Ni-Al) added to 60 mL of 20 % sodium hydroxide that had been
heated to 80 °C. The solution was stirred continuously at 80 °C for 45 minutes. The solution
was then rinsed four times with 40 mL of distilled water and once with 40 mL of isopropanol.
Finally, 10 mL of isopropanol was added to the activated Raney nickel.

71

The sixth experiment and the second high temperature method were similar to a
method used by Srivastava et al (1985). This sixth experiment involved cooling 150 mL of 20
% sodium hydroxide in a refrigerator to a temperature of 4 °C, followed by addition of 3.0 ±
0.01 g of nickel-aluminum alloy while stirring continuously. The mixture was then gradually
heated to 100 °C; the heating process took approximately 35 minutes. The activated Raney
nickel was then removed from the heat and rinsed twice with 5 % sodium hydroxide, four
times with distilled water and once with 20 mL of isopropanol. Finally, 10 mL of isopropanol
was added to the activated Raney nickel.
A Student T-test was used to determine if a significant difference was present between
the original concentration of sulfur and the final concentration of sulfur in the reacted samples.
A one-way ANOVA was then used to determine any significant differences between the six
different activation methods (Miller and Miller, 1993). Finally, an F-test was used to
determine the relative precision of the methods and another Student T-test was performed to
determine which methods had a significant difference from each other.
A scanning electron microscope (SEM) was used to observe any differences in the
surface of Raney nickel, activated under different conditions. A beam of accelerated electrons
(30 kV) is condensed to 50 - 200 A diameters scanning the surface of the Raney nickel. The
interaction between the electrons and atoms produces transitions between energy levels in the
atom. The energy is released as X-ray photons giving a fingerprint of the composition of the
Raney nickel (Reeder, 1983). A Philips XL30 SEM equipped with an energy dispersive
spectrometer (EDS) was used for the analysis. The EDS is capable of chemical analysis with
an element range from boron to uranium. The elemental analysis was performed to determine

72

differences in the composition of elements found in Raney nickel that is activated under
varying conditions. The samples o f Raney nickel were placed on an A1 pin-type stub and
cemented with a carbon-based glue. Then they were coated with a 25 nm layer of gold in a
Denton Desk 11 sputter coater.
3.2.3 M ethodology fo r the Amount o f Raney Nickel

As it was important that the amount of Raney nickel be minimized, a test on the
amount of Raney nickel needed to cause a sufficient reaction was necessary. The method for
this was that a sample of Raney nickel was prepared using the second activation method
described above. The nickel-aluminum alloy solutions were prepared using 1.0, 2.0, 3.0 and
4.0 g of nickel-aluminum alloy. The volume of sodium hydroxide was also decreased or
increased appropriately. A Student T-test was used to determine any significant differences
between the amount o f Ni-Al alloy used and the amount of sulfur removed.
3.2.4 M ethodology o f Desulfurization o f FCC Gasoline

For all of the above methodologies the same reaction procedure was employed using
the FCC gasoline. Each sample of activated Raney nickel had 50 mL of FCC gasoline added.
The samples were then covered and stirred continuously at 20 °C for 30 minutes. At the end
o f the 30 minutes, the samples were filtered through Whatman No.2 filter paper and placed in
a 4 °C refrigerator until the analysis could be completed.
The analysis was performed using UV fluorescence spectrometry according to the
ASTM method D5453 (ASTM, 2003). The instrument was an Antek 9000 series elemental
analyzer with an oxidative furnace that operated at 1075 °C (see chapter 2 for more
information).

73

A check standard was analyzed first in order to check the operation of the Antek
instrument. If it passed the check standard within 5 % of the actual value, the samples were
then analyzed. If it did not pass then it was recalibrated before another check standard was
performed. Each sample was analyzed in triplicate with the average being reported in parts
per million.
3.2.5 Repeatability o f Method 5
Finally, a repeatability test was performed to ensure that the test could produce similar
results during multiple tests. The repeatability test involved using method 5 for the activation
method of Raney nickel. Each batch of Raney nickel was prepared separately using the exact
same method so that the amount o f Raney nickel was consistent among each o f the samples.
Then 50 mL of FCC gasoline was added to each individual sample and reacted at 20 °C for 30
minutes. The individual samples were analyzed using the UV fluorescence method. A
preliminary test o f 5 samples was needed in order to determine an appropriate sample size.
The sample size was determined using the following equation (Caulcutt and Boddy, 1983);

M= (— ) '
c

(e.3.1)

Where, t is the t-statistic value for given sample size, 6' is the standard deviation, and
c is the confidence interval. Once the sample size was determined, the appropriate numbers
o f samples were prepared and analyzed. The values obtained were used to determine the
standard deviation. These values were then placed into the equation for repeatability (Farrant,
1997):
r = i* V 2 *(T

(e.3.2)

74

Where <7^ = the standard deviation obtained when the same method is used under the
same eonditions and t = the eonstant value for a Student T-test.
3.3 Results
3.3.1 Results o f the Six Different Activation Methods
The results from the six different activation methods can be seen in Table 3.1.
Methods 2, 3, 5 and 6 showed a significant difference from the initial sulfur concentration.
Methods 5 and 6 had the largest decrease in sulfur concentration compared to activation
methods 2 and 3. It should be noted that the initial sulfur concentrations were different
between the first four methods and the last two methods, because two different samples of
FCC gasoline were used as we ran out of the first.

Method
Name

Table 3.1: Summary of Raney Niekel Activation Methods.
Significant
Conditions
Initial Sulfur Final Sulfur
Cone, (ppm) Difference
Cone, (ppm)
± 5.0 ppm

± 5.0 ppm

Method
1

3.7 g Ni-Al alloy, 10 mL
10%NaOH 20 °C

163.53

160.22

No

Method
2

3.7 g Ni-Al alloy, 20 mL
10 % NaOH 20 °C

163.53

112.31

Yes

Method
3

3.7 g Ni-Al alloy, 5 mL
20 % NaOH 20 "C

163.53

123.40

Yes

Method
4

3.7 g Ni-Al alloy, 10 mL
20 % NaOH 20 "C

163.53

161.63

No

Method
5

3.0 g Ni-Al alloy, 60 mL
20 % NaOH 80 "C

183.99

52.42

Yes

Method
6

3.0 g Ni-Al alloy, 150
mL 20 % NaOH 100 °C

183.99

49.00

Yes

75

The ANOVA showed that there was a significant difference between the means using
the six different activation procedures as can be seen in Table 3.2, each method had a sample
size o f 3. As the calculated F-value is higher than the critical value, the standard deviations
are therefore significantly different. Since the ANOVA showed there was a significant
difference, further analysis was needed to determine exactly where the significance lies.
Table 3.2: ANOVA table for the Comparison o f Six Different Activation Methods.
Variance
Method 1

1.73

Method 2

1.72

Method 3

3.45

Method 4

0.72

Method 5

0.09

Method 6

0.04

Sample Mean Variance

1030.85

Degrees of Freedom

F-value

Critical
Value
(95 %)

3988.73

3.106

Within-Sample
variance

1.29

12

Between-sample
variance

5154.27

5

An F-test was performed in order to show the more precise methods. The F-test
determined that methods 5 and 6 were more precise than methods 1, 2 and 3 however; the
precision was not significantly different than method 4. Also, the precision between method 5
and 6 was not significantly different. Next, a Student’s T-test was used to determine the
differences between the six methods. The data showed that there was a significant difference
between each of the methods except for when method 1 was compared to method 4. In this
case no significant difference was determined. Even though there was a significant difference

76

between the means o f each of the methods, methods 5 and 6 showed the largest significance
of the six methods.
The differences in the structure of the activated Raney nickel can he seen in the SEM
images of the different methods. Using activation method 2, Figure 3.1 shows that the Raney
nickel is covered with a crystalline material and the structure was cylindrical with rounded
edges. Using activation method 5, Figure 3.2 shows a slightly different structure. In Figure
3.2 the Raney nickel has fewer crystals on the surface. In addition, the structure had sharper
edges, with more surface area visible due to the lack of the crystals. Finally, Figure 3.3 shows
the structure produced when method 6 was used for activation. In Figure 3.3 it can he seen
that there were fewer crystals on the surface than for method 5 and the structures have even
more sharp edges. The elemental analysis shows the chemical differences between the
different types of Raney nickel.

1

Figure 3.1: SEM image of Raney nickel prepared using Method 2 (20 “C).

77

A»

Figure 3.2: SEM image of Raney nickel prepared using Method 5 (80 "C).

Figure 3.3: SEM image of Raney nickel prepared using Method 6 (100 C).
The elemental analysis o f the lower temperature activation method 2 shows that there
was a large portion of aluminum and nickel in the sample as seen in Figure 3.4, the target

78

location can be seen in Figure 3.1. This can be compared to Figures 3.5 and 3.6 from the two
higher temperature activation methods where there was very little aluminum present, the
target positions can be seen in Figure 3.2 and 3.3 respectively. Other than the difference in
the amount of aluminum and oxygen, the elements present were very similar. There are small
differences in the amount of nickel present in each of the Figures 3.4, 3.5 and 3.6. This
difference is simply due to the presence of aluminum. The aluminum peak seen in Figure 3.4
is larger than in Figures 3.5 and 3.6 so it masks the nickel peaks within close proximity. Also,
one can notice that gold appears in the energy spectrum, due to the conductive gold coating
on the Raney niekel.

AI

0

Na

Ni

Ni

Ni

ipu
Au

Au
Au
Au
Au

2.00

4.00

GOO

8 .00

1 0 .0 0

1 2 .0 0

14.00

1G.00

18.00

Energy Dispersive Spectrum

Figure 3.4: Elemental Analysis of Raney niekel prepared using method 2 (20 ®C).

79

Éi
Ni
Ni

Al
Au
Au

0

Ni

'Au

2.00

4.00

6 .00

8.00

1 0 .0 0

1 2 .0 0

1 4.00

1 6 .0 0

1 8.00

Energy Dispersive Spectrum

Figure 3.5: Elemental Analysis of Raney nickel prepared using method 5 (80 ”C).

Ni

Ni
O

O
■ë' , Al
a NI
■§ 0
Nb
Ni

X

Nb

Au
2.00

4.00

6.00

8.00

1 0 .0 0

Au
Au
Au
Au
1 2.0 0

14.00

16.00

18.00

Energy Dispersive Spectrum

Figure 3.6: Elemental Analysis of Raney nickel prepared using method 6 (100 C).

80

3.3.2 Results fo r the Amount Nickel-Aluminum Alloy
The results from varying the amount of Ni-Al alloy used during the method 2
activation processes can be seen in Table 3.3. The table shows that each amount of Ni-Al
alloy decreased the sulfur concentration. However, the Student t-values show that 3.0 and
4.0 g of alloy had a much more significant effect on the final sulfur concentrations. However,
there is no significant difference between 3.0 and 4.0 grams.
Table 3.3: Summary of the Difference in Desulfurization Based on the Amount of Ni-Al
Alloy.
TMass of Initial Sulfur
Final Sulfur
Standard TSignificant
Ni-Al
Concentration Concentration Deviation value critical Difference
alloy (g) (ppm)
(ppm)
(ppm)
1

179

168

2.5

7.62

2.92

Yes

2

179

170

2.5

6.24

2.92

Yes

3

179

151

2.5

19.40 2.92

Yes

4

179

152

2.5

18.71

Yes

2.92

3.3.3 Repeatability o f Method 5
As was demonstrated in the previous sections, the most effective method for activation
was method 5, using an optimal amount of 3.0 grams of Ni-Al alloy to get the most
desulfurization. A preliminary test of 5 samples showed that a sample size of 15 samples
would be sufficient for determining the repeatability using equation e.3.1. This is for a 5 ppm
confidence interval and 95 % confidence. Once this was determined, the test was repeated for
a total of 15 samples. The summary statistics are shown in Table 3.4.

81

Table 3.4: Summary Statistics for 15 Identical Samples
Mean
44.00 ppm
Variance

3.99 ppm^

Standard Deviation

2.00 ppm

Repeatability

6.06 ppm

Original FCC Sulfur Concentration

186 ± 4.12 ppm

This shows that trom the original concentration of 186 ppm, there is a 76 % reduction
in the sulfur concentration. In addition, the repeatability of the method is approximately 6
ppm. The scatter of the measurements ean be seen in Figure 3.7.

50.00 -1
48.00

♦

♦

46.00
44.00
42.00

I 40.00
38.00

J

-

36.00 -

w 34.00

-

32.00 30.00

0

10

12

14

16

18

S am p le s

Figure 3.7; Determination of Sulfur Coneentration in 15 Uniquely Prepared Samples of
Raney Nickel Using Activation Method 5 and 3.0 grams of Ni-Al Alloy.

3.4 Discussion
3.4.1 Raney Nickel Activation
The results in Table 3.1 show that the two high temperature activation methods
produce the greatest reduction in sulfur concentration. This is supported by literature where
the activation procedures performed at higher temperatures were reported to produce more

82

reactive Raney nickel (Devred et al., 2003; Lieber and Morritz, 1953; Fouilloux, 1983). This
can be explained by observing the SEM micrographs and elemental analysis results. The
elemental analysis results for the low temperature activation, Figure 3.4, show a large amount
of aluminum remaining on the Raney nickel. This aluminum is in the form of bayerite
{Al^O^

which can be seen in Figure 3.1 as the crystals covering the surface of the

Raney nickel. A crystal of similar structure was reported by Kiyohara et al (2000) for a
variety of aluminas. A similar result was also reported by Devred et al (2003) for low
temperature leaching processes, which did not use an excess of sodium hydroxide. Bayerite
covers the surface of the nickel using up the active sites. This reduces the amount of contact
the FCC gasoline can have with the Raney nickel and its active sites. However, at excess
concentrations of sodium hydroxide and higher activation temperatures, the elemental analysis
shows a lower amount of aluminum or bayerite present (Figures 3.5 and 3.6) this was also
supported by Devred et al (2003). This is supported by the SEM images in Figures 3.2 and
3.3, which show much less o f the bayerite on the surface of the nickel. Therefore, there is a
greater chance for the FCC gasoline to contact the surface of the nickel, making methods 5
and 6 the most effective methods for removing sulfur from FCC gasoline. The Student T-test
shows that they had the largest significant difference when compared to the other methods.
There was only a 3 ppm difference between the 80 °C and 100 °C method. This was
within the standard deviations of the two samples so it is not necessary to heat the sample the
extra 20 °C. Also, the F-test showed that there was no significant difference in the precision
of the two methods. Therefore, method 5 was the most effective activation method for
producing the most reactive Raney nickel for removal of sulfur from FCC gasoline.

83

3.4.2 Raney Nickel Amount

The tests investigating the amount of Ni-Al alloy that was most effective at removing
the sulfur from FCC gasoline showed that 3.0 grams was enough to effectively remove sulfur.
As Table 3.3 shows both 3.0 and 4.0 grams had very significant reductions in sulfur
concentration. However, there was not a significant difference between 3.0 and 4.0 grams.
Therefore, 3.0 grams is the least amount of Ni-Al alloy that could be used to remove the most
amount of sulfur from FCC gasoline based on 50 mL of FCC gasoline.
3.4.3 Repeatability

As the previous sections have determined, method 5 was the best activation method
and 3.0 grams o f Ni-Al alloy was enough to show significant results. Therefore, the
repeatability of this method was tested. The results of 15 identically prepared samples can be
seen in Table 3.4. The method had a standard deviation of 2.0 ppm and had a repeatability of
6.0 ppm.
3.5 Conclusion
In conclusion, the higher temperature methods were more effective at producing
activated Raney nickel for removing sulfur from FCC gasoline. The higher temperature
activations combined with an excess of sodium hydroxide produced less bayerite, allowing
more of the nickel surface to come in contact with the FCC gasoline. In addition, the
difference between the two higher temperature methods, method 5 and 6, was not significant
enough to warrant heating the sodium hydroxide the extra 20 °C. Therefore, method 5 is the
most effective activation method. Next, the amount of Ni-Al alloy necessary to cause the
largest reduction in sulfur concentration was found to be 3.0 grams. The 4.0 grams of Ni-Al

84

alloy caused the same reduction as the 3.0 grams so the lesser amount of 3.0 grams contained
enough Ni-Al alloy to achieve the desired result. Therefore, the final experiment was an
activation method 5 where the sodium hydroxide was heated to 80 °C using 3.0 grams of NiAl alloy. The repeatability of this method was determined to be 6.0 ppm.

85

3.6 Literature Cited
2003 Annual Book of ASTM Standards. 2003. Petroleum Products. Lubricants, and Fossil
Fuels. Volume 3. ASTM International. West Conshohocken, PA. Pp. 1-1368.

Augustine, R.L. 1996. Heterogeneous Catalysis for the Synthetic Chemist. Marcel Dekker
Inc. New York, USA. Pp. 213-540.

Bartok, M., Wittmann, G., Gondos, G., and G.Y. Smith. 1987. Homogeneous and
heterogeneous catalytic asymmetric reactions. 1. Asymmetric hydrogenation of the
prochiral C=C hond on a modified Raney Ni catalyst. Journal o f Organic Chemistry
52:1139-1141.

Beigi, A.A.M., Teymouri, M., Eslami, M., and M. Farazmand. 1999. Determination of trace
sulfur in organic compounds by activated Raney nickel desulfurization method with nondispersive gas detection system. Analyst 124: 767-770.

Caulcutt, R. and R. Boddy. 1983. Statistics for Analytical Chemists. Chapman & Hall.
London. Pp. 1-253.

Coulier, L., Kishan, G., van Venn, J.A.R., and J.W. Niemantsverdriet. 2002. Influence of
support-interaction on the sulfidation behaviour and hydrodesulfurization activity of
Al2 0 3 -supported W, CoW, and NiW model catalysts. Journal o f Physical Chemistry B
106: 5897-5906.

Devred, F., Hofifer, B.W., Sloof W.G., Kooyman, P.J., van Langeveld, A.D., and H.W.
Zandbergen. 2003. The genesis of the active phase in Raney-type catalysts: the role of
leaching parameters. A pplied Catalysis A: General 244: 291-300.

Ertl, G., Knozinger, H., and J. Weitkamp. 1999. Environmental Catalysis. Wiley-VCH.
Germany. Pp 17-118.

Farrant, T. 1997. Practical Statistics for the Analytical Scientist. The Royal Society of
Chemistry. England. Pp. 1-58.

Fouilloux, P. 1983. The nature of Raney nickel, its adsorbed hydrogen and its catalytic activity
for hydrogenation reactions (review). A pplied Catalysis 8: 1-42.

86

Freel, J., Pieters, W.J.M., and R.B. Anderson. 1969. The structure of Raney Nickel I. Pore
Structure. Journal o f Catalysis 14: 347-256.

Granatelli, L. 1959. Determination of microgram quantities of sulfur by reduction with Raney
nickel. Analytical Chemistry 31(3):434-436.

Kiyohara, P.K., Santos, H.S., Coelho, A.C.V., and P.S. Santos. 2000. Structure, surface area
and morphology of aluminas from thermal decomposition of Al(OH)(CH3 COO) 2 crystals.
Academia Brasileira de Ciencias 72(4): 471-495.

Lieber, E. and F.L. Morritz. 1953. The uses of Raney nickel. Advances in Catalysis 5: 417
455.

Matar, S. and L.F. Hatch. 2001. Chemistrv of Petrochemical Processes. 2"'^ edition. Gulf
Professional Publishing. USA. Pp. 69-88.

Meyers, R.A. 2004. Handbook of Petroleum Refining Processes. 3'^‘*edition. McGraw-Hill.
USA. Pp. 1-15.40.

Miller, J.C. and J.N. Miller. 1993. Statistics for Analytical Chemistrv. 3^"^ edition. Ellis
Horwood and Prentice Hall. England. Pp. 1-77.

Oyama, S.T., Wang, X., Requejo, E.G., Sato, T., and Y. Yoshimura. 2002.
Hydrodesulfurization of petroleum feedstocks with a new type of nonsulfide hydrotreating
catalyst. Journal o f Catalysis. 209: 1-5.

Pettit, G.R. and E.E. van Tamelen. 1962. Organic Reactions. New York. Pp. 356-509.

Reeder, R.J. 1983. Carbonates: mineralogy and chemistry. M ineralogical society o f America
reviews in M ineralogy 11.

Srivastava, S., Minore, J., Cheung, C.K., and W.J. le Noble. 1985. Reduction of Aromatic
Rings by 2-propanol with Raney nickel catalysis. Journal o f Organic Chemistry 50:394396.

87

Chapter 4: Optimization of the Reaction Conditions for Raney Nickel for
the Desulfurization of FCC Gasoline
4.1 Introduction
As described in chapter 3 the most active Raney nickel was produced by leaching 3.0
grams of nickel-aluminum (Ni-Al) alloy with an 80

solution of 20 % sodium hydroxide for

45 minutes (Devred et ah, 2003). This method had a repeatability of 6 ppm when reacted for
30 minutes with FCC gasoline. The activation process oxidizes the aluminum and generates
hydrogen gas, which activates the niekel portion of the catalyst (Fouilloux, 1983).
2(M -

-H20Ff-

, 0 -4 2M -

-H2v4Z(0Ff)^(^) -H

(4.1)

Some of the hydrogen remains adsorbed on the nickel, making it a low temperature
hydrogenation catalyst. Therefore, niekel catalysts have been used widely in hydrogenation,
hydrotreating, and in steam-reforming reactions (Wojeieszak et al., 2004). In particular,
Raney nickel has been used in industry for the hydrogenation of organic compounds and the
determination o f trace amounts of sulfur in organic mixtures (Devred et al., 2003; Granatelli,
1959).
This ehapter focuses on optimizing the reaction between Raney nickel and FCC
gasoline in order to remove as much sulfur as possible; as such optimizing the reaction can
reduce the amount of Raney niekel required and therefore reduce costs. In order to further
reduce the costs of Raney niekel, the recovery o f niekel from the spent Raney niekel will also
be assessed and discussed.

88

To the best of our knowledge, the use of a Raney nickel catalyst has never been
attempted for desulfurization o f FCC gasoline although it has been used for the desulfurization
of other fuels and solvents (Augustine, 1996; Granatelli, 1959; Fouilloux, 1983). Beigi et al
(1999) determined that at lower temperatures and with newly prepared catalysts, the
desulfurization occurs according to reaction 4.2, whereas at higher temperatures the
desulfurization occurs according to reaction 4.3. In both cases, the sulfur forms a covalent,
alloy-like, non-stoichiometric compound with the nickel (Beigi et al., 1999).
RSR + Ni —

^ RH + R H + NiS

(4.2)

3RSR ~^3Ni —H2Haney —^ RR + RR + R R +3NiS + 3Hj

(4.3)

These reactions occur at the Raney niekel surface and therefore the Raney nickel is
acting as a heterogeneous catalyst. On the surface, there are two main types of adsorption:
physical adsorption and chemisorption. Physical adsorption involves van der Waals forces,
which cause relatively weak adsorption. Chemisorption involves covalent forces, which are
similar to the forces occurring between bound atoms in molecules (Laidler, 1987).
Chemisorption causes a single layer of adsorbed molecules to cover the surface of the Raney
nickel. Once the surface of Raney nickel is covered with a single layer of reactant it is
saturated and no more molecules can be adsorbed. The only way more molecules can be
absorbed is if they are physically adsorbed to the single layer already formed (Laidler, 1987).
The kinetics of these surface reactions can be affected by temperature. The amount of
a compound adsorbed to a surface is dependent on a variety of factors, including temperature
and pressure (Laidler, 1987). The rate constant (k) is dependent on temperature; therefore

89

temperature will affect the rate at which a reaction occurs (Laidler, 1987). The surface
reactions occur in five steps (Laidler, 1987):
1.

Diffusion o f fhe reactants to the surface
2. Adsorption on the surface
3. Reaction on the surface
4. Desorption of the products

5. Diffusion of the desorbed products into solution
The slowest of these five steps determines the overall rate of the process. In most
cases reactions on the surface and/or desorption of the products are the rate determining steps
(Campbell, 1988). However, there are many factors that affect the reaction process. Olefins
are a large component of FCC gasoline and are a key source of octane rating. During
hydrogenation reactions the olefins can become saturated causing a possible loss in octane.
Olefins affect the recovery of sulfur wifh Raney nickel by forming an addition reaction
between the hydrogen sulfide and an olefin as seen in reaction 4.4 (Beigi et al., 1999):
H - S - H + R = R 4 r^ R -S H + R - S H

(4.4)

Beigi et al (1999) also found that in samples with greater than 10 wt. % of olefins, the double
bonds interact with the active sites of the Raney nickel, interfering in the desulfurization step.
Therefore, olefins are competing with the sulfur-containing compounds to occupy the active
sites of the Raney nickel.
In order to determine the effectiveness of Raney nickel on desulfurization of FCC
gasoline, a variety of experiments were performed. The experiments investigated the effects

90

of time, temperature, and the sequential use of Raney nickel. In addition, analysis was
performed in order to determine any chemical changes in the FCC gasoline due to its reaction
with Raney nickel.
Finally, the ability to recover nickel from the spent Raney nickel was assessed. A
variety of methods have been shown to remove sulfur from nickel. For example, Granatelli
(1959) used a hydrochloric acid solution in order to remove sulfur in the form o f hydrogen
sulfide. Other methods include nickel roasting, which is used extensively in mineral extraction
processes. Nickel roasting involves heating the nickel-containing compound to high
temperatures (600 - 800 °C) causing oxidation.
The objectives o f the work presented in this chapter then are to first determine the
most effective laboratory conditions for Raney nickel to desulfurize FCC gasoline. The
second objective is to determine any effects that Raney nickel may have on the FCC gasoline
composition. Thirdly, if the nickel can be recovered from the nickel sulfide for re-use.
4.2 Methodolosv
4.2.1 FCC Gasoline Sampling and Storage
The FCC gasoline was sampled as described in Chapter 2 at Husky Energy’s refinery
in Prince George. Six 1-litre samples were taken on June 14, 2005 for use with the following
experiments. The FCC gasoline samples were transported to UNBC and stored as described
in Chapter 2.
4.2.2 Preparation o f the Raney Nickel
First, 60 mL o f 20 % sodium hydroxide was heated to 80 °C on a hot plate stirrer this
took approximately 15 minutes. Next, 3.0 ± 0.01 g of niekel-aluminum alloy (50-50 NFAl wt.

91

%, Alfa Aesar) was gradually added over approximately 5 minutes to the heated sodium
hydroxide. The reaction was stirred continuously for 45 minutes at 80 ± 5 °C once all the NiA1 had heen added. The Raney nickel was then rinsed four times with 40 mL of distilled
water, letting the Raney nickel settle between rinses, and then decanting the distilled water and
hydroxide mixture. The Raney nickel was then rinsed once with 40 mL of isopropanol, to
remove any remaining water. Finally, 10 mL of isopropanol was added to the Raney nickel in
order to keep it active and aid in the mixing of the Raney nickel with FCC gasoline (Lieber
and Morritz, 1953; Granatelli, 1959).
4.2.3 Time Trial Experiments
To determine the effect o f time on the removal of sulfur, twelve samples were
prepared by adding 50 mL o f FCC gasoline to the Raney nickel. The reactions occurred at an
ambient temperature o f 20 °C. Duplicate samples were stirred continuously for each of the
following reaction times: 0.5, 30, 60, 90, 120, and 180 minutes. In order to stop the reaction
the samples were filtered, using No. 2 Whatman filter paper, immediately upon removal from
the stirring plate. The filtered FCC gasoline was then stored in 125 mL amber glass bottles at
4 °C for further analysis.
4.2.4 Temperature Trial Experiments
To determine the effect o f temperature on the removal of sulfur, ten samples were
prepared by adding 50 mL of FCC gasoline to the Raney nickel. Duphcate samples were
performed at each of the following temperatures: 5, 20, 30, 40 and 50 °C. The samples were
stirred continuously for 30 minutes.

92

The samples used a reflux condenser with a hot plate stirrer as the heater and mixer as
shown in Figure 4.1. In this set of experiments the FCC gasoline was first heated to the
desired temperature then added to the Raney nickel and held at a constant temperature while
stirring continuously for 30 minutes. In each case, after 30 minutes, the samples were
removed and filtered using No. 2 Whatman filter paper. The filtered FCC gasoline was then
stored in 125 mL amber glass bottles at 4 °C until analysis.

Figure 4.1: Reflux condenser apparatus for temperature experiments: a. condenser, b.
round bottom flask, e. thermometer.

4.2.5 Sequential Addition Runs
The sequential trials involved reacting 50 mL of FCC gasoline with Raney nickel for
30 minutes at 20 °C with continuous stirring. The FCC gasoline was then filtered through No.
2 Whatman filter paper then reacted again for 30 minutes with a fresh batch of Raney nickel.
The FCC gasoline was again filtered and then stored in 125 mL amber glass bottles at 4 °C
until analysis.

93

4.2.6 Method o f Analysis
The samples were analyzed using an Antek 9000 sulfur/nitrogen system. This system
was setup aeeording to the ASTM method D5453 (Annual Book of ASTM Standards, 2003).
The system injects 5 |iL of sample, which is then completely oxidized at a temperature of
approximately 1000 °C. The combustion gases are dried, removing any water formed. The
sample gases then move to the detector module where the fluorescence 80% is recorded. The
fluorescent emission is specific for sulfur dioxide and is also proportional to the amount of
sulfur in the sample.
Two samples o f the original FCC gasoline were analyzed as the baseline for the test
samples. Each of the samples was injected three times. The standard deviation and mean
values were also determined for each o f the samples. The samples were analyzed in triplicate
and average values determined.
4.2.7 Statistical Analysis
The data from the time and temperature trials were subjected to a statistical analysis
consisting of a one-way ANOVA and a Student’s T-test. The one-way ANOVA was
performed first in order to determine if a significant difference existed between the means of
any of the values. The ANOVA was a one-tailed test using a 95 % confidence interval. Once
a significant difference was determined, a Student’s T-test was performed to determine
exactly where the significant differences occurred. The Student’s T-test established
differences at the 99 % confidence interval.

94

4.2.8 Hydrocarbon Analysis
A hydrocarbon analysis was performed on a few samples to determine the effeet that
the reaetion with Raney niekel has on FCC gasoline. The hydrocarbon analysis was
performed using gas ehromatography with a flame ionization detector (GC/FID). A Hewlett
Packard HP 6890 series GC system was used. It was operated with an HP 1 dha column at a
maximum temperature of 280 °C for a run time of 138 minutes, see chapter 2 for more
information.
4.2.9 Recovery o f Nickel
The final step was to assess the recovery of nickel from the nickel sulfide formed
during reaction with FCC gasoline. It was first necessary to determine that the sulfur was
present within the spent Raney nickel. A sample of the spent Raney nickel was reacted with
50 mL of 20 vol. % HCl (Reagant A.C.S, Lot No. 204173) in a 100 mL flask. Once bubbles
started to form in the solution, a GASTEC (Levitt-Safety Limited, model 801) was used to
determine if hydrogen sulfide was being produced. The GASTEC apparatus uses detector
tubes (No. 4LT), whieh ehange colour to indicate the presenee of hydrogen sulfide. The
GASTEC requires 50 mL of air to be drawn and forced through the detector tube.
Once it was determined that the spent Raney nickel contained sulfur, the following
roasting method was used to remove the sulfur, in order to recover the niekel. In this method
3.0 ± 0.01 grams o f spent Raney nickel was added into a ceramic crucible. The crucible was
then inserted into a muffle furnace and stepped up gradually to 800 °C over a period of 6 hrs.
The erucible was then removed from the furnaee and allowed to cool for 18 hrs.

95

X-ray dif&action (XRD) was then used to analyze the sample for different phase
compositions. A Broker D8 X-ray diffractometer, with a 20 X-ray diffraction system and CoKg X-ray source was used for the data collection. The system was operated at 40 kV and 25
mA. The XRD recorded powder diffraction patterns for 180 seconds at three different angles.
The patterns were then compared to a DIFFRAC^'"^ EVA mineral powder diffraction file.
The samples were prepared for XRD analysis using a thin smear method. This
requires mixing the sample with a small amount of distilled water to form a paste (Poppe et
al., 2002). This paste is then smeared onto the corner of glass shdes. The XRD can only
detect compounds that are greater than 1 wt % of the sample.
4.3 Results
4.3.1 Time Trials
The results of the time trial experiments can be seen in Figure 4.2, with the sulfur
concentration on the subordinate and time on the ordinate. The time data follows a power
function with the fitted line having an R-squared value of 0.9942 and a function of equation
e.4.1:
}/ = 78.507x^'"'^

(e.4.1)

The baseline sulfur concentration was 186 ppm indicating there is an immediate
removal of sulfur within the first 30 seconds of contact with the Raney nickel. The data
showed that even at a short reaction time of 30 seconds only 80 ppm remained. The largest
reduction was found at 180 minutes with 32 ppm remaining. However, the same maximum
reduction was found at 120 minutes, so an ANOVA was performed to determine if the
differences were statistically significant.

96

200
180

y = 78.507x°^^®®
R" = 0.9942

Q. 160 140 120 -

100 -

60 40 20 -

0

20

40

60

80

100

120

140

160

180

Time (min)

Figure 4.2: Sulfur Concentration in FCC Gasoline with Time of Treatment at 20 “C.
The ANOVA showed that there was a significant difference between the values at
different reaction times. Further analysis indicated that there were significant differences up to
a reaction time o f 60 minutes, with the most significant occurring within the first 30 minutes.
After 60 minutes there is no longer a significant decrease in the sulfur concentration.
4.3.2 Temperature Trials
The results from the temperature trial experiments can be seen in Figure 4.3. As seen,
there is a decrease in sulfur concentration at all temperatures tested in this work. However,
the decrease in sulfur concentration is much higher at the higher temperatures. The minimum
reduction in sulfur concentration was found at 5 °C with only 53 ppm remaining from a
starting concentration of 213 ppm. The maximum reduction was found at 50 °C with 2 ppm
remaining. The effect of temperature on the rate of the reaction can be seen in the large
concentration decreases in Figure 4.3.

97

70
s

60

a.
3

50

B
O

40

1
I

53

30
18 36
20

O

U
k

a"5
!Zi

10

2 . 53

0

î- - - - - - - - - - - - - —

20

-10

30

...................

40

1- - - - - - - - - - - - - - - - —

^

1

5D

-20

Temperature (degrees Celsius)
Figure 4.3: Effect of Temperature on the rate of Reduction in sulfur concentration in FCC
gasoline over 30 minutes of reaction time.
The ANOVA was performed and determined that there was a significant difference
between the average values determined for the five temperatures used in this study. The
Student’s T-test further confirmed that the signifieant difference was between the values
obtained at 30, 40 and 50 °C. There were no significant differences seen between the results
using the lower temperatures o f 5 and 20 °C.
4.3.3 Sequential Addition Runs
The sequential reaction results showed that with a total of two additions of Raney
nickel, the starting concentration of 213 ± 6.68 ppm for total sulfur compounds could be
lowered to 7 ppm with a standard deviation of 1 ppm. This is equivalent to a 97 % reduction
in the sulfur concentration at 20 °C.
4.3.4 Hydrocarbon Analysis Results
The hydroearbon analysis showed a drop in the paraffins, l-paraffins, naphthenes,
aromaties and multisubstituted aromaties. It should be noted that the multisubstituted

98

aromaties are calculated separately during the GC/FID process so the below values may not
sum to 100 %. The olefin percentage increased independent of the reaction time as can be
seen in Table 4.1. The decrease in percent volume of the paraffins, I-paraffins, naphthenes,
aromaties and multisubstituted aromaties was equivalent to 10 % with the largest drop seen in
the I-paraffins. This drop was equal to the 10 % increase seen in the olefins. Part of the loss
in volume may be a result o f evaporation during the experiments, although no large volume
loss was found.
Table 4.1: Comparison o f hydrocarbon component concentrations for a variety of time
intervals (± 5 %).
Hydrocarbon
Original FCC
0.5
30
60
90
120
180
Gasoline
Components (Vol %)
mins mins mins mins mins mins
Paraffins

5.6

4.0

3.9

3.9

3.9

4.0

4.0

I-paraffins

25.7

20.0

20.4

20.1

20.4

20.3

20.5

Olefins

43.4

54.5

53.5

53.1

53.2

53.5

52.6

Naphthenes

10.4

8.8

8.8

9.6

9.0

9.3

9.4

Aromaties

13.7

12.0

12.1

12.3

12.4

12.0

12.6

Multisubstituted
Aromaties

4.3

3.8

3.9

3.9

4.0

3.8

4.1

The temperature trials showed similar results as to the time trials. There was an
average decrease in the percent volume of all components except the olefins as seen in Table
4.2. The decrease was equivalent to an average of 6 % with an average increase of 10 % for
the total olefins. However, the data shows that the C; olefin compounds were the only olefins
that increased in volume percent.

99

Table 4.2: Comparison of hydrocarbon component concentrations for a variety of
temperatures (± 5 %).
Hydrocarbon
Original FCC
5"C
20 ®C
30 ”C
40 "C
50 "C
Components (Vol %) Gasoline
Paraffins

5.5

4.4

4.4

3.5

4.1

4.4

I-paraffins

27.2

22.3

21.8

18.4

21.0

22.1

Olefins

36.4

47.8

47.6

43.5

47.4

44.3

Naphthenes

9.6

7.8

8.0

7.0

8.2

8.7

Aromaties

18.6

16.8

17.1

15.5

18.1

19.2

Multisubstituted
Aromaties

6.6

6.0

6.1

5.5

6.5

7.0

Finally, Table 4.3 shows the hydrocarbon components for the sequential addition
experiments. As seen in Table 4.3, there is a decrease in each of the hydrocarbon components
except for the olefins, which again showed an increase. The hydrocarbon components showed
a total decrease of 20.4 % with the largest decrease seen in the l-paraffins. The olefin
increase was equivalent to 20.1 %, which is just slightly less than the total decrease in the
other components.
Table 4.3: Comparison of hydrocarbon component concentrations for sequential addition (± 5
Hydrocarbon Components (Vol %)

Original FCC
Gasoline

Sequential
Addition

Paraffins

5.5

3.0

I-paraffins

27.2

16.1

Olefins

36.4

56.5

Naphthenes

9.6

7.1

Aromaties

18.6

15.3

Multisubstituted Aromaties

6.6

5.5

100

4.3.5 Recovery o f Nickel
The GASTEC apparatus showed the production of hydrogen sulfide in the spent
Raney nickel on addition o f acid indicating the presence of sulfur. The detector tubes in both
cases changed from yellow to a pinkish colour as soon as it was exposed to air from the
headspace of the reaction beakers. Also, when the reaction occurs between the spent Raney
nickel and HCl, the solution turns a green colour indicating that the nickel is dissolving into
the solution and forming nickel chloride, implying that there was oxidation of nickel.
Using the nickel roasting method, the spent Raney nickel changed from an almost
black colour to a green colour by the end of the 6 hour roasting process. The spent nickel
also had an average increase in weight of 0.06 grams suggesting formation o f nickel oxide,
which is the product formed during industrial roasting processes. The production of nickel
oxide was further supported by XRD analysis. In Figure 4.4 the diffraction pattern is perfectly
aligned with the index diffraction pattern for nickel oxide. This indicates the sample is largely
nickel oxide as no other diffraction peaks are present.

101

I

2 T h eta -S c a le

Figure 4.4: XRD diffraction pattern for spent Raney nickel roasted at 800 °C for 6 hrs and
NiO indexed diffraction pattern.

4.4 Discussion
4.4.1 Effect o f Time on the Removal o f Sulfur
This study demonstrates that it is possible to remove sulfur from gasoline using Raney
nickel. From the time trial data shown in Figure 4.2, it is evident that as the length of reaction

102

time increases so does the reduction in sulfur concentration until after a certain time. Over 50
% o f the sulfur is removed tfom the FCC gasoline during the first 30 seconds of reaction.
However, in order to remove the maximum amount of sulfur removed, a longer contact time
is necessary. The most significant reduction in sulfur concentration occurs within the first 30
minutes. It can be concluded then that 30 minutes is a sufficient reaction time as at this point,
76 % of the sulfur is removed from the FCC gasoline at 20 °C. However, an extra 4 % can be
removed if left for another 30 minutes, ie. a total reaction time of 60 minutes.
After 60 minutes, there are no significant changes in the sulfur concentration. Without
an in depth investigation into the exact reaction kinetics, there are several possible
explanations for this result. The most probable result is due to the sulfur being adsorbed to
the entire nickel surface available, leaving no additional reaction surface. Finally, it is also
possible that the olefins in FCC gasoline are competing for the reaction sites on the surface of
Raney nickel, making these sites unavailable for sulfur reactions (Campbell, 1988). Existing
research has reported that Cg - Cio olefins strongly inhibit the hydrodesulfurization of
thiophene (Hatanaka and Yamada, 1998).
The concentration of sulfur still decreases after 60 minutes, but by very small amounts
so it is insignificant. The reaction may still be occurring at a much slower rate, the thiophenes
may need a longer reaction time to give the less labile olefins an opportunity to desorb,
making room for more sulfur on the surface of the Raney nickel. Another possible
explanation is that all the highly active sites on the Raney nickel may be full and only the less
reactive sites remain available so the reaction times may be longer. As the reaction slows after
60 minutes, this is sufficient amount of time to remove as much sulfur as possible from the

103

FCC gasoline used in this study. It is apparent that additional work needs to be done in order
to look at the reaction kinetics.
4.4.2 Effect o f Temperature on the Removal o f Sulfur
The temperature runs indicated that the reaction rate is higher at higher temperatures.
This is due to temperatures relation to the rate laws. Interestingly, even at 5 °C, it is seen that
75 % of the sulfur is removed after 30 minutes. There was very little difference in the amount
of sulfur removed up to 30 °C. However, above 30 °C there was a large difference in the
amount of sulfur removed after 30 minutes.
However, the Student’s T-test confirmed that there were only significant differences in
the sulftir concentrations at temperatures higher than 30 °C. Therefore, heating did not make
a significant difference in the removal of sulfur using Raney nickel unless the samples were
heated to above 30 °C. By increasing temperature the bonds in thiophene may be more easily
broken or the desorption of the products off the Raney nickel surface may occur at a faster
rate. These are some possible explanations as to why a greater rate of sulfur removal occurs
at higher temperatures.
4.4.3 Effect o f Sequential Addition on the Removal o f Sulfur
The sequential use of Raney nickel was found to remove a total of 97 % of the sulfur
from FCC gasoline. This method seems to be more effective at removing sulfur than most of
the other experiments, except for the sample at 50 °C. The sample at 50 °C showed over 99
% removal o f sulfur. The 97 % reduction seen during the sequential addition is due to the
amount of Raney nickel being used. This method uses double the quantity o f Raney nickel,
more than in other trials, and a 60 minute reaction time was used. This method is very

104

effective because after tire first 30 minutes a new batch of Raney nickel is added. This new
batch o f Raney nickel is not poisoned by the sulfur or olefins as the previous batch thus
allowing more sulfur to be adsorbed on the surface. In addition, there is the maximum number
of active sites available with a fresh batch of Raney nickel. Therefore, by performing a
sequential addition of Raney nickel, the spent Raney nickel is removed and fresh Raney nickel
is added, allowing for more sulfur to be removed in the 60 minute period than if a single batch
o f Raney nickel were used.
4.4.4 Hydrocarbon Components Analysis
The hydrocarbon components analysis showed that there is a small change in the FCC
gasoline composition due to its reaction with Raney nickel. The reaction results in an increase
in the percent volume of olefins and a decrease in the percent volume of all other components.
Olefins have double bonds and are considered good for fossil fuels as they contribute to the
octane number. The increase in olefin content may not actually be an increase. The analysis
shows each of the components as a percentage of the volume, therefore the resulting increase
in olefins may be due to a decrease in the rest of the components. The decrease in the rest of
the components may be through evaporation or conversion of the sulfur containing
compounds into other such hydrocarbon components, ft is also possible that sulfur is being
removed from aromaties to produce some of the increase in olefin concentration. This could
be due to the cleaving of the sulfur-carbon bond in the thiophene resulting in the breakage of
the ring, forming a C5 olefin. There is definitely a need for more research to determine the
exact effects these changes will have on the end products.

105

4.4.5 Recovery o f Nickel fo r Possible Re-Use
The production o f hydrogen sulfide when the spent Raney nickel is reacted with
hydrochloric acid suggests that NiS is indeed formed during the reaction of FCC gasoline with
Raney nickel. Since hydrochloric acid is effective at liberating the sulfur from nickel in the
form of hydrogen sulfide, this is a possible recovery method for the nickel. As the sulfur is
removed in the form of hydrogen sulfide, the nickel goes partially into solution as nickel II
chloride. An example of this reaction can be seen in reaction 4.5. Once the nickel is in
solution, it may be possible to use an electroplating method to retrieve pure nickel.
(4-5)
Another possible method for the recovery of nickel is through nickel roasting. As we
know that sulfur is present in the spent Raney nickel, it is possible to remove it as SO2 . When
the spent Raney nickel was heated to 800 °C the components of the spent Raney nickel were
oxidized. Oxidation causes the sulfur to form SO2 and the nickel to partially convert into
NiO. The formation of nickel oxide can be seen in the change from a black colour to a green
colour. Also the XRD diffraction pattern shows the sample contains only nickel oxide all
other components are less than 1 wt % as they are not seen in the diffraction pattern. An
example of the reaction can be seen in reaction 4.6, which may also explain the increase in
weight once roasted. As all the nickel may not be in the form of nickel sulfide, some may still
be in the form of Raney nickel which also oxidizes during the roasting process. Therefore,
this may also be another option for nickel recovery. There were conflicting results obtained as
small increases in the mass do not fit with the full conversion of the spent Raney nickel into
NiO, even though the XRD pattern shows only nickel oxide in the roasted samples. Fouilloux

106

(1983) reports that some authors helieve there is hydride formation on Raney niekel, however
XRD data was not eonsistent with this theory. If the spent Raney nickel was hydrated it may
explains the small mass diflferenee. Further research will be needed to determine the exact
course of the reaction.

MS, ,, + Ni -

+10,

>2MO„, + SO,,,, +

(4.6)

4.5 Conclusions
This study demonstrates that Raney niekel is effective at removing sulfur compounds
from FCC gasoline. A variety of reaetion times and temperatures were attempted which
showed that a variety of amounts o f these compounds could be removed. Therefore, the
amount of sulfur, whieh needs to be removed from a FCC gasoline stream, will help determine
the most appropriate method to be employed. For example, if only 80 % of the sulfur needs
to be removed, then allowing the reaetion to occur for 90 minutes at 20 °C is sufficient.
However, if greater than 90 % removal is needed, then the samples should be heated to 40 or
50 °C and allowed to react for 30 minutes. Over 90 % removal of sulfur can also be achieved
by sequential use of fresh Raney nickel. However, the latter, even though highly effective, is
not recommended as the extra amounts of Raney niekel required will most likely make the
process prohibitively expensive. All these processes cause an apparent increase in the olefin
volume, and we have no plausible explanation for this. This may affect the end products or
gasoline requirements so further research is needed.
The experiments to determine if the niekel can be recovered from the spent Raney
niekel were sueeessful. Both the method using hydrochloric acid and the roasting method
removed the sulfur from the spent Raney niekel producing either H2 S or SO2 , respectively.

107

Therefore, these are both viable options for the recovery of nickel, which would then need to
be further processed into Ni-Al.
Future studies could be performed on a larger scale with a detailed economic analysis
to determine if the method is indeed feasible for larger scale operations. The tests involving
the recovery of nickel showed that Raney nickel may not be saturated at this point. Perhaps
further research into the amount of gasoline, which can cause 1.0 gram o f Raney nickel to be
saturated with sulfur, may be useful for economic determinations. Finally, the full impact of
using Raney nickel on the chemical changes it imparts on FCC gasoline should be further
investigated to determine, for example, if the octane value of the fuel is changed and whether
other regulations and requirements are still being met.

108

4.6 Literature Cited
2003 Annual Book of ASTM Standards, Petroleum products, lubricants, and fossil fuels;
ASTM International, West Conshohocken, PA. Vol. 3, Pp.67-75.

Augustine, R.L. 1996. Heterogeneous Catalysis for the Synthetic Chemist. Marcel Dekker
Inc. New York, USA. Pp. 213-540.

Beigi, A. A., M. Teymouri, M. Eslami and M. Farazmand. 1999. Determination of trace sulfur
in organic compounds by activated Raney nickel desulfurization method with nondispersive gas detection system. Analyst 124: 767-770.

Campbell, I.M. 1988. Catalysis at Surfaces. Chapman and Hall, New York, New York. Pp.
11-159.

Devred, F., B.W. Hoffer, W.G. Sloof, P.J. Kooyman, A.D. van Langeveld and H.W.
Zandbergen. 2003. The genesis of the active phase in Raney-type catalysts: the role of
leaching parameters. Applied Catalysis A: General 244: 291-300.

Fouilloux, P. 1983. The nature of Raney nickel, its adsorbed hydrogen and its catalytic activity
for hydrogenation reactions (review). Applied Catalysis 8: 1-42.

Granatelli, L. 1959. Determination of microgram quantities of sulfur by reduction with Raney
nickel. Analytical Chemistry 31(3):434-436.

Hatanaka, S. and M. Yamada. 1998. Hydrodesulfurization of catalytic cracked gasoline. 3.
Selective catalytic cracked gasoline hydrodesulfurization on the Co-Mo/y-Al2 Ü3 catalyst
modified by coking pre-treatment. Industrial and Engineering Chemical Research 37:
1748-1754.

Laidler, K.J. 1987. Chemical Kinetics Chapter 7: Reactions on Surfaces.
HarperCollinsPublishers. New York, New York. Pp. 229-273.

Lieber, L. and F.L. Morritz. 1953. The uses of Raney nickel. Advances in Catalysis 5:417455.

109

Poppe, L.J., V.F. Paskevich, J.C. Hathaway and O.S., Blackwood. 2002. A laboratory manual
for X-ray diffraction. U.S. Geological Survey Open File Report 01-041,
http://pubs.usgs.eov/ol7ol01-041.

Wojcieszak, R., S. Monteverdi, M. Mercy, I. Nowak, M. Ziolek and M.M. Bettahar. 2004.
Nickel containing MCM-41 and AlMCM-41 mesoporous molecular sieves characteristics
and activity in the hydrogenation of benzene; Applied Catalysis A: General 268: 241-253.

110

Chapter 5: Conclusion
The objectives of this research project were to first find a solvent or compound that
would specifically target sulfur compounds within FCC gasoline, second to reduce sulfur
concentrations as much as possible with a goal o f 30 (ppmw), third to reduce the sulfur
concentrations without making any major changes to the composition of the FCC gasoline and
fourth to determine if the reactants could be recycled.
The first objective was investigated through an experimental approach in chapters 2
and 3. Chapter 2 identified a variety of solvents, acids, bases and metals that were
unsuccessful at causing a reduction in the sulfur concentrafion of the FCC gasoline. Through
these experiments it was determined that a strong molecular interaction or reaction was
necessary to remove the less reactive sulfur compounds, such as dibenzothiophenes. There is
room for more research involving the use of solvents. Future research could involve
investigating the use o f secondary solvents or other extraction methods.
Chapter 3 investigated Raney nickel, which is commonly used as a hydrogenation
catalyst and has been shown to remove sulfur in a variety of analytical methods. These
experiments showed that Raney nickel was effective at removing sulfur-containing compounds
from FCC gasoline. The effectiveness of Raney nickel was determined to be dependent on the
activation temperature, sodium hydroxide concentration and the amount of Ni-Al alloy.
Therefore, the most effective form of Raney nickel was determined to result from an
acfivation method using an 80 °C activation temperature, 3.0 grams of Ni-Al alloy, and a 60
mL 20 % sodium hydroxide solution.

Ill

The second objective to reduce the sulfur concentration was investigated in chapter 4.
The experiments within chapter 4 involved using the activated Raney nickel at a variety of
reaction times and temperatures to determine the maximum reduction o f sulfur-containing
compounds. These experiments showed that the sulfur concentration in FCC gasoline could
be reduced to 40 ppmw simply by allowing the reaction to proceed for 90 minutes. Whereas
sulfur concentrations can be reduced to less than 20 ppmw if the reaction mixture is heated to
temperatures above 40 °C as shown in Figure 4.2. Therefore, a goal of less than 30 ppmw of
sulfur in FCC gasoline is possible using Raney nickel. In fact, concentrations below 10 ppmw
can be achieved if the reaction is heated to 50 °C.
The third objective was also briefly investigated in chapter 4. It was determined
through hydrocarbon analysis that there was a higher olefin concentration as seen in Table 4.1
and Table 4.2 after treatment with Raney nickel. This higher olefm concentration may have an
effect on the end products, as olefins are regulated in fossil fuels. The decrease in aromaties
found will cause a decrease in the octane number. The significance of this increase is not
known and further research is necessary to determine exactly why only the C; olefins appeared
to increase.
The fourth objeetive was investigated in ehapter 4, determining that it is possible to
recover the nickel from the spent Raney nickel. Several potential methods for the recover of
nickel are possible. Spent Raney nickel can be reacted with hydrochloric acid to produce H2 S
and a nickel chloride solution. This nickel chloride solution can then be further separated
through an electroplating process. In addition, a nickel roasting method could also be used to
recover the nickel in the form of nickel oxide. This process involves heating the spent Raney

112

nickel to temperatures of approximately 800 °C in order to produee SO2 . Therefore, either
one o f these options could be a feasible method for retrieving the niekel from the spent Raney
niekel.
Although this thesis gives a good starting point there is still mueh researeh needed on
Raney nickel to determine if it is a viable alternative to hydro treating. Although the nickel can
be reeovered from the spent Raney nickel, it will still need to be converted back into Ni-Al
alloy in order to be reused. Therefore, further researeh is needed into the exaet proeess for
reeovery and the cost o f this process. Future researeh is also necessary on a larger scale to
determine the economic feasibility of using Raney nickel in sulfur removal on refinery scales.
In addition, extensive researeh will be needed to determine the exact chemical impacts on the
FCC gasoline through the reaction with Raney nickel. This should inelude the effects on the
octane number, density, and olefin coneentrations.
This research will contribute to finding a possible alternative to hydrotreating that is
more cost effective and potentially more environmentally friendly. The lower levels of sulfur
resulting from the new method will reduee the health problems assoeiated with sulfur dioxide
and the environmental problems associated with aeid rain. The alternative method will also
give the oil and gas sector another way to meet Canada’s strieter regulations by reducing
sulfur levels in FCC gasoline.

113