You are on page 1of 98

Removal of Sulfur Contaminants in Natural Gas Streams

A Thesis

Presented to the

Graduate Faculty of the

University of Louisiana at Lafayette

In Partial Fulfillment of the

Requirements for the Degree

Master of Science

Jerry F. Conerly III

Fall 2019
© Jerry F. Conerly III

2019

All Rights Reserved


Removal of Sulfur Contaminants in Natural Gas Streams

Jerry F. Conerly III

APPROVED:

Rafael Hernandez, Chair August Gallo


Professor of Chemical Engineering Professor of Chemistry

Emmanuel Revellame Mary Farmer-Kaiser


Assistant Professor of Industrial Dean of the Graduate School
Technology
Table of Contents

List of Tables ..................................................................................................................... vi

List of Figures ................................................................................................................... vii

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

Chapter 2: Literature Review .............................................................................................4


2.1 Hydrogen Sulfide ...............................................................................................4
2.1.1 Properties and analysis. .......................................................................4
2.1.2 𝐇𝟐𝐒 removal. .......................................................................................4
2.1.2.1 Adsorption. ............................................................................4
2.1.2.2 Reactions................................................................................9
2.2 Mercaptans/Thiol ............................................................................................. 13
2.2.1 Properties and analysis. ..................................................................... 13
2.2.2 Activated carbons. ............................................................................. 14
2.2.3 Magnesium oxide and magnesium oxide/silicon dioxide composite. 14
2.2.4 Biofiltration. ....................................................................................... 15

Chapter 3: Objectives........................................................................................................ 17

Chapter 4: Materials and Methods ................................................................................... 18


4.1 Chemicals ......................................................................................................... 18
4.2 Sample Preparation ......................................................................................... 18
4.3 Reactor Configurations .................................................................................... 18
4.4 Reactor Start Up .............................................................................................. 20
4.5 Product Sampling ............................................................................................. 22
4.6 Reactor Shutdown ............................................................................................ 23
4.7 Analytical Methods .......................................................................................... 23
4.7.1 GC Analysis of gas samples. .............................................................. 23
4.7.1.1 Calibration curve. ................................................................. 26
4.7.1.2 Reaction studies.................................................................... 27
4.7.2 Surface analysis. ................................................................................ 27
4.7.2.1 Surface area and porosimetry measurement. ....................... 27
4.7.2.2 Fourier Transform Infrared Spectroscopy (FTIR)............... 30
4.7.2.3 Thermogravimetric Analysis (TGA). .................................... 32
4.7.2.4 Proton Induced X-ray Emission (PIXE) Spectroscopy. ........ 33

Chapter 5: Results and Discussion ................................................................................... 35


5.1 Catalyst Analysis .............................................................................................. 35
5.1.1 BET surface analysis. ........................................................................ 35
5.1.2 TG analysis. ....................................................................................... 36
5.1.3 FTIR Characterization of spent catalyst. ......................................... 36
5.1.4 PIXE analysis. .................................................................................... 39
5.2 Sulfur Removal ................................................................................................ 42
5.2.1 Coastal experiments........................................................................... 45

iv
5.2.2 Graphite blend experiments. ............................................................. 49
5.2.3 Clay blend experiments. .................................................................... 53
5.2.4 Sulfatreat experiments....................................................................... 56
5.2.5 Anthracite blend experiments. .......................................................... 57
5.2.6 Experimental catalysts comparison to literature. ............................. 61
5.3 Economics ......................................................................................................... 75

Chapter 6: Conclusions ..................................................................................................... 79

Chapter 7: Recommendations and Future Works ........................................................... 81

Bibliography ...................................................................................................................... 82

Appendix A ........................................................................................................................ 86

Abstract ............................................................................................................................. 87

Biographical Sketch .......................................................................................................... 88

v
List of Tables

Table 1: GC Oven Method for Injection Sample Analysis ................................................... 25

Table 2: GC Oven Method for Online Sample Analysis ...................................................... 26

Table 3: BET Characteristics of the Catalysts and Fillers .................................................... 35

Table 4: H2 S Breakthrough Times ...................................................................................... 42

Table 5: Methyl Mercaptan Breakthrough Times ................................................................ 43

Table 6: Ethyl Mercaptan Breakthrough Times ................................................................... 44

Table 7: Maximum Removal of H2 S ................................................................................... 72

Table 8: Maximum Removal of Methyl Mercaptan ............................................................. 73

Table 9: Maximum Removal of Ethyl Mercaptan ............................................................... 74

Table 10: Material Cost per 1000 ft 3 Treated Gas at each Variation ................................... 77

vi
List of Figures

Figure 1: H2 S removal capacities of literature adsorbents ..................................................... 8

Figure 2: Pilot Reactor Setup System ................................................................................. 19

Figure 3: Bench Scale Reactor System ............................................................................... 20

Figure 4: Schematic of Pilot Scale Reactor ......................................................................... 21

Figure 5: Agilent Technologies 6890N Network GC equipped with Agilent Technologies


5975 Inert MS Detector............................................................................................ 25

Figure 6: Chromatograph of Model Gas ............................................................................. 27

Figure 7: Micromeritics ASAP 2020 Surface Area and Porosimetry Analyzer .................... 29

Figure 8: Jasco FT/IR 4000 Series ...................................................................................... 30

Figure 9: Agilent Technologies Cary 640 FTIR .................................................................. 31

Figure 10: FTIR Spectra of Raw Coastal Catalyst ............................................................... 32

Figure 11: SDT 2960 Simultaneous DSC-TGA .................................................................. 32

Figure 12: Linear Accelerator for PIXE Analysis ............................................................... 33

Figure 13: TGA Results of Raw Coastal Catalyst ............................................................... 36

Figure 14: FTIR Spectra of Raw Coastal Catalyst ............................................................... 38

Figure 15: FTIR Spectra of Spent Coastal Catalyst ............................................................. 38

Figure 16: PIXE Elemental Maps of Unreacted Coastal Catalyst ........................................ 39

Figure 17: PIXE Elemental Maps of Spent Coastal Catalyst ............................................... 40

Figure 18: PIXE Analysis of Unreacted Coastal Catalyst .................................................... 41

Figure 19: PIXE Analysis of Spent Coastal Catalyst ........................................................... 41

Figure 20: Coastal H2 S Breakthrough vs Time ................................................................... 47

Figure 21: Coastal Methyl Mercaptan Breakthrough vs Time ............................................. 49

Figure 22: Coastal Ethyl Mercaptan Breakthrough vs Time ................................................ 49

Figure 23: Graphite H2 S Breakthrough vs Time ................................................................. 51

vii
Figure 24: Graphite Methyl Mercaptan Breakthrough vs Time ........................................... 52

Figure 25: Graphite Ethyl Mercaptan Breakthrough vs Time .............................................. 52

Figure 26: Clay H2 S Breakthrough vs Time ........................................................................ 53

Figure 27: Clay Methyl Mercaptan Breakthrough vs Time ................................................. 54

Figure 28: Clay Ethyl Mercaptan Breakthrough vs Time .................................................... 55

Figure 29: Sulftreat H2 S Breakthrough vs Time .................................................................. 56

Figure 30: Sulfatreat Methyl Mercaptan Breakthrough vs Time .......................................... 56

Figure 31: Sulfatreat Ethyl Mercaptan Breakthrough vs Time............................................. 57

Figure 32: Anthracite H2 S Breakthrough vs Time ............................................................... 58

Figure 33: Anthracite Methyl Mercaptan Breakthrough vs Time ........................................ 59

Figure 34: Anthracite Ethyl Mercaptan Breakthrough vs Time ........................................... 60

Figure 35: H2 S Breakthrough at 100 psi 100℃ 50 mL/min................................................. 63

Figure 36: Methyl Mercaptan Breakthrough at 100 psi 100℃ 50 mL/min .......................... 63

Figure 37: Ethyl Mercaptan Breakthrough at 100 psi 100℃ 50 mL/min ............................. 64

Figure 38: H2 S Breakthrough at 100 psi 100℃ 36 mL/min................................................. 64

Figure 39: Methyl Mercaptan Breakthrough at 100 psi 100℃ 36 mL/min .......................... 65

Figure 40: Ethyl Mercaptan Breakthrough at 100 psi 100℃ 36 mL/min ............................. 65

Figure 41: H2 S Breakthrough at 100 psi Amb 50 mL/min ................................................... 66

Figure 42: Methyl Mercaptan Breakthrough at 100 psi Amb 50 mL/min ............................ 66

Figure 43: Ethyl Mercaptan Breakthrough at 100 psi Amb 50 mL/min ............................... 67

Figure 44: H2 S Breakthrough at 100 psi Amb 36 mL/min ................................................... 67

Figure 45: Methyl Mercaptan Breakthrough at 100 psi Amb 36 mL/min ............................ 68

Figure 46: Ethyl Mercaptan Breakthrough at 100 psi Amb 36 mL/min ............................... 68

Figure 47: H2 S Breakthrough at 200 psi Amb 36 mL/min ................................................... 69

Figure 48: Methyl Mercaptan Breakthrough at 200 psi Amb 36 mL/min ............................ 69

viii
Figure 49: Ethyl Mercaptan Breakthrough at 200 psi Amb 36 mL/min ............................... 70

Figure 50: H2 S Breakthrough at 200 psi 100℃ 36 mL/min................................................. 70

Figure 51: Methyl Mercaptan Breakthrough at 200 psi 100℃ 36 mL/min .......................... 71

Figure 52: Ethyl Mercaptan Breakthrough at 200 psi 100℃ 36 mL/min ............................. 71

Figure 53: Natural Gas Price Trends ................................................................................... 75

ix
Chapter 1: Introduction

A significant number of undesirable compounds are found in numerous industrial

effluents. One particularly undesirable compound is hydrogen sulfide (H2 S), and it causes

complications primarily in the energy sector whose products such as natural gas, liquefied

petroleum gas, tail gas, fuel cells and the transportation gases like diesel, gasoline and jet

fuel, either naturally contain, consume, or form the compound in the production process. It is

considered a major problem in industry due to health hazards, catalyst poisoning, and strict

environmental regulations due to environmental detriments such as corrosion, air pollution

and acid rain.

Natural gas production contains H2 S, which is an acid gas. Natural gas containing

more than 5.7 mg/m3 H2 S is labeled sour gas. The Scientific Advisory Board on Toxic Air

Pollutants has set acceptable levels of ambient H2 S to be in the range of 20-100 ppb (Vikrant

et al., 2018). OSHA has set the H2 S allowable exposure limit to be 10 ppm time-averaged

over 8 hours (Tian, Wang, Yu, Li, & Guo, 2017). It is detrimental to pipeline health due to

corrosion such as hydrogen blistering, hydrogen embrittlement, pitting, and stress cracking

(Fontana, 1986).

Hydro-desulfurization (HDS) is the main commercial process in petroleum refining to

convert various sulfur compounds into H2 S and sulfur-free organics using catalytic treatment

with hydrogen. Most catalysts in use are alumina-based that has been impregnated with

cobalt and molybdenum. The HDS reactor is typically operated at elevated temperature

ranges from 573 K to 673 K and elevated pressures ranging from 30 atm to 130 atm. The

converted H2 S is then removed from the gas stream via amine scrubbing and the Claus

1
Process. HDS is considered expensive due to the energy demand and costs of high purity

hydrogen as a feed source (Gupta, Ibrahim, & Al Shoaibi, 2016).

Amine scrubbing has been used for H2 S removal along with CO2 for many years.

Alkanolamines, being basic in nature, are suitable for acid gas absorption, such as H2 S, due

to the hydroxyl and amino functional groups. The amino group provides the needed alkalinity

to the aqueous solution that is essential for acid gas absorption. Chemically, the lone pair on

the amino group allows the alkanolamine to function as a base, thus increasing the pH and

allowing H2 S/CO2 to dissolve in the solution. There are four typical alkanolamines that are

used for acid gas treatment, each used under different conditions. These alkanolamines are

monoethanolamine, diethanolamine, N-methyldiethanolamine, and digycolamine (W.

Agbroko, Karishma, & Benson, 2017).

H2 S that is removed through HDS and amine stripping is processed and cleaned up

through the Claus Process. This is the main industrial process for large quantity removal of

H2 S (W. Nehb, 1994). It works through the two reaction steps:

2H2 S + O2 → SO2 + H2 O + heat

kJ
(ΔH = −519 mol) (1)

2H2 S + SO2 → S + H2 O + heat (2)

The first reaction step, the thermal step, combusts one-third H2 S to form SO2 and

water with combustion temperatures between 980℃ to 1540℃ and pressures usually below

10 psia. The second step is a catalytic reaction that takes place over activated alumina- or

titania-based catalysts. The SO2 produced from reaction (1) is reacted with the remaining H2 S

with the final products obtained being elemental sulfur and water. This process is responsible

for the majority of H2 S removal, leaving a much smaller quantity of H2 S that still has to be

2
removed in tail gas treatment units. The Claus process can be operated in either a straight-

through or split-flow configuration. The straight-through configuration is used for streams

with H2 S concentrations greater than 40%. If the concentration of H2 S is not high enough,

carsul formation, a carbon-sulfur polymer, will take place instead of commercial grade

elemental sulfur due to elemental carbon formation from incomplete hydrocarbon

combustion. This typically occurs with C3 and larger hydrocarbons (Optimized Gas Treating,

2018).

When streams do not have high H2 S concentrations, a split-flow configuration is

used. With split-flow configuration, streams containing 20%-40% H2 S can be treated. In this

configuration 2/3 of the gas stream is fed to the catalytic reactor while the other 1/3 is fed to

the burner. Regardless of configuration used, the stream leaving the Claus reactor, while

being purer, still needs further treatment. SCOT (Shell Claus off-gas treating) is a common

tail gas treatment improvement to the Claus process. H2 is added to the catalytic reactor to

reduces gases such as COS, SO2 , and CS2 to H2 S which is then recycled to the Claus unit.

There have also been improvements to the Claus process such as the Superclaus process

which can offer a sulfur recovery up to 99.5%. However, this is highly selective to H2 S alone

and no other sulfur compounds (de Angelis, 2012).

Due to other sulfur impurities found in natural gas streams, there has been a shift to

find better methods compared to HDS, amine scrubbing, and the Claus Process to remove

sulfur contaminants that can decompose H2 S and other small-chain yet problematic organic

compounds such as thiols. In this study, an iron-based catalyst was characterized, tested to

determine sulfur removal, and compared to a local industry standard. This industry standard

was then compared to the catalyst dispersed among filler materials.

3
Chapter 2: Literature Review

2.1 Hydrogen Sulfide

2.1.1 Properties and analysis. Hydrogen sulfide is a colorless gas that is extremely

toxic and flammable. It naturally occurs in crude petroleum, natural gas, volcanic gases and

hot springs with the odor of rotten eggs that is common for sulfur compounds. When in

solution, it is highly corrosive to equipment. It is also poisonous to most industrial catalysts.

According to OSHA, the odor threshold for H2 S is from 0.01 to 1.5 ppm with rapidly

worsening health effects at higher concentrations. When exposed, these symptoms include

nausea, headache, insomnia, cough, fatigue, dizziness, eye irritation, loss of smell,

drowsiness, staggering, unconsciousness, and death. Unconsciousness occurring within 1 to 2

breaths of air containing 700 to 1,000 ppm H2 S and near instantaneous death at

concentrations higher than 1,000 ppm (Occupational Safety and Health Administration,

2006).

H2 S is usually monitored via infrared sensors for in-situ measurements. It is typically

analyzed by gas chromatography (GC) through the use of a flame photometric detector, a

chemiluminescence detector, a thermal conductivity detector, or a mass spectrometer.

2.1.2 𝐇𝟐 𝐒 removal. Removal of sulfur compounds in natural gas is commonly

achieved with adsorbents, the Claus process, and/or absorption into aqueous solutions.

2.1.2.1 Adsorption.

2.1.2.1.1 Zeolites/clays. Zeolites have been utilized for H2 S removal. Even though the

cost of these adsorbents is relatively low, their adsorption capacities are considered poor

compared to activated carbons and metal oxides (J. Wang, Guo, Parnas, & Liang, 2015). Due

to its low cost, zeolites have been used as a support while having been modified with metals

4
and metal oxides via impregnation and ion exchange. A study conducted by “(Micoli,

Bagnasco, & Turco, 2014)” showed that adding Cu or Zn improves 13X zeolite’s H2 S

adsorption. Almost all modifications yielded at least twice the breakthrough time of the

unmodified zeolite. The ion exchange modification showed a much longer breakthrough time

when compared to unmodified 13X zeolite and the metal impregnated 13X zeolite. It was

theorized that the adsorption increase found in the ion exchange sample is due to the initial

cations being replaced in large quantities by Cu2+ or Zn2+ ions while the impregnated samples

contain Cu and Zn in their respective oxide forms instead of direct replacement. The metal

content between the types of modifications confirm this. For Cu, the ion exchanged zeolite

has roughly 2.5 times the Cu content when compared to the impregnated zeolite. For Zn, the

ion exchanged zeolite has 11 times the Zn content when compared to the impregnated

zeolite. In the case of the ion exchange modification, it was assumed that H2 S was

chemisorbed on the 13X zeolite by an acid-base reaction such as:

H2 S + NaZ ↔ HZ + NaHS (3)

Where Na can be substituted with other cations such as Cu2+ and Zn2+ to give similar

reactions. The H2 S adsorption onto the impregnated zeolites is due both to the cations in the

sorbents’ cavities and the reactions between CuO or ZnO and H2 S (Micoli et al., 2014).

Na-A zeolites produced from kaolin, a type of clay, were modified by impregnating

varying amounts of zinc oxide: 10, 20 and 30 wt %, in an attempt to increase the hydrogen

sulfide capacity (Abdullah, Mat, Somderam, Abd Aziz, & Mohamed, 2018). Experimental

results showed that H2 S adsorption capacity was increased in all three cases. The baseline

Na-A zeolite H2 S capacity was 0.58 mg S/ g adsorbent. The 10 wt % ZnO addition showed

an increase to 3.55 mg S/ g adsorbent, and the 30 wt % loading showed a capacity of 8.13 mg

5
S/ g adsorbent. The 20 wt % ZnO loading showed the highest H2 S capacity of 15.75 mg S/ g

adsorbent and thus considered the optimal loading (Abdullah et al., 2018). It is theorized that

the reasons for the 20 wt % ZnO loading is higher than the 30 wt % ZnO loading are that

there could be pore blockage, damage/collapse of pore structure, and an aggregation of ZnO

particles on the surface due to exceeding the critical dispersion capacity of the zeolite,

possibly creating a less reactive aggregate of active sites. Other studies conducted by Li et al.

showed similar results when modifying SBA-16, MCM-48 and KIT-6 with ZnO and ran at

room temperature. An increased H2 S adsorption capacity was observed when ZnO was added

to all three of these 3-D mesoporous structures (Li, Sun, Shu, & Zhang, 2016).

Another study investigated modifying the clay subtype kaolin with iron chlorides,

copper chlorides and iron oxide to determine the effects had on H2 S adsorption. The kaolin

was modified by ion exchange through dropwise dispersion. In the H2 S breakthrough tests,

the iron oxide showed the most effective removal of 90% at 95 minutes followed closely by

iron with 85% at 100 minutes. The copper modification yielded approximately 75% removal

at 100 minutes while unmodified kaolin had 66% removal at this time. The three different

depositions onto kaolin increased its BET surface area, effectively enhancing H2 S adsorption

capacity. This is reflected by the increased experimental breakthrough time when compared

to kaolin (Louhichi, Ghorbel, Chekir, Trabelsi, & Khemakhem, 2016).

2.1.2.1.2 Activated carbon. Activated carbons have been utilized as sorbents for

sulfur removal but typically have a low sulfur adsorption capacity while unmodified.

Modifications such as surface oxidation, metal impregnation, and ion exchange have been

investigated by several researchers (Figueiredo, Pereira, Freitas, & Órfão, 1999; Ning et al.,

2018). Results in these studies have shown that modifications added more adsorption sites

6
and acid groups on the activated carbon surface. In a study by Cui, Turn and Reese, coconut

shell activated carbon was used as the base material for modifications through metal

impregnation and oxidation (Cui, Turn, & Reese, 2009). The impregnation modifications

were metal from Zn, Cu, Fe, K and Na sources with approximately 0.4 mmol of metal loaded

per gram of virgin AC. Virgin AC had the lowest H2 S capacity out of all the sorbents

investigated, while Cu and Zn sorbents never experienced breakthrough during the conducted

test. All modifications resulted in an increase H2S sorption capacity. The most effective

modification for H2 S removal was the Cu impregnation followed by the Zn. The alkaline

metals, K and Na, showed slightly better results when compared to Fe and oxidation

modifications (Cui et al., 2009).

Hydrogen sulfide removal has also been examined using non-thermal plasma

modification on Fe-walnut shell AC, particularly NH3 -plasma, in hopes that it could

introduce amino-functional groups to possibly enhance desulfurization through an increased

adsorption or oxidation capacity. Ning, Liu et al. utilized dielectric barrier discharge to

generate non-thermal plasma to introduce these amino surface modifications to the AC

through transforming the discharge gas. It was shown that H2 S removal was improved when

the treatment input voltage and time were in the ranges of 5.6-6.8 kV and 5-10 min,

respectively. However, H2 S removal was shown to decrease when treatments extended past

these ranges (Ning et al., 2018).

A study conducted by Xiao et al. evaluated the effects of impregnating sodium

carbonate onto coal-based AC on H2 S oxidation (Xiao, Wang, Wu, & Yuan, 2008). Xiao et

al. proved that both the unmodified and modified AC have catalytic activity for H2 S

oxidation. It was observed that water had a beneficial effect on the oxidation process by

7
changing the relative humidity in the experiments. The highest sulfur capacity of this

impregnated AC was 420 mg S/g. This experiment was conducted with a relative humidity of

80% at 30℃. Due to the Na2 CO3 impregnation the pH was made more basic. This gave

hydrosulfide ion concentration a more preferred environment due to the acidity of H2 S, thus

improving the H2 S oxidation allowing full utilization of the AC pores. The unmodified AC

deactivated rapidly. It was postulated that this rapid deactivation was caused by the formation

of sulfuric acid produced as a byproduct of H2S oxidation. According to Xiao, this suppresses

H2 S dissociation. The impregnation of Na2 CO3 onto the unmodified and impregnated ACs

provided a basic environment instead of an acidic one, avoiding deactivation until the pores

are saturated (Xiao et al., 2008).

Figure 1: H2 S removal capacities of literature adsorbents (Elyassi et al., 2014; Fang, Zhao,

Fang, Huang, & Wang, 2013; Guo et al., 2007)

8
2.1.2.2 Reactions.

2.1.2.2.1 Metal oxides reactions. Due to its high selectivity for H2 S the iron oxide

process has been a prevalent means of removing H2 S since implemented in England during

the nineteenth century (Crynes, 1977). The iron oxide catalyst removes H2 S by forming iron

sulfides and water through the following reaction:

2Fe2 O3 + 6H2 S → 2Fe2 S3 + 6H2 O (4)

ΔH = −22 kJ (25 °C and 1 atm) (5)

The process is typically operated with 2 towers in parallel. A tower is operated until

the bed is saturated with sulfur and the sweetened gas stream begins to show H2 S. Once this

point is reached, the tower is removed from service and regenerated by circulating gas with a

small portion of air contained throughout the bed. While the primary tower bed is undergoing

regeneration, the sweetening process is changed over to the parallel tower. The regeneration

of the catalyst is achieved by exposing the spent catalyst to an oxygen stream and can be

repeated several times before disposal. The regeneration reaction is shown in reaction (6).

2Fe2 S3 + 3O2 → 2Fe2 O3 + 6S (6)

ΔH = −198 kJ (25 °C and 1 atm) (7)

The H2 S removal reaction is recommended to be carried out at room or slightly

elevated temperatures; the recommended temperature range is 10°C-49°C. High temperatures

could cause the iron oxide structure to dehydrate. This leads to highly reduced reaction rates

and possible decomposition of Fe2 S3 to FeS2 and Fe8 S9 . These are vastly more difficult to

regenerate. The water produced in reaction (3) does provide some of the needed moisture to

allow the catalyst to work optimally. The elemental sulfur formed from the oxidation of

ferric sulfide will eventually cover the majority of the catalyst’s surface after numerous

9
regeneration cycles. This will lead to a loss of activity of the oxide’s surface and can cause

excessive pressure drop and bed channeling. The regeneration reaction is highly exothermic

(Kouichi Miura et al., 1992). Due to this, the spent catalyst can be pyrophoric when exposed

to air or low oxygen content streams so care should be exercised when change-out occurs.

A mechanistic study by Davydov et al. confirmed a possible, more descriptive set of

reactions for the ferric oxide-H2 S interactions set by Wiȩckowska (Davydov, Chuang, &

Sanger, 1998). Due to the thermodynamically unstable Fe2 S3 product formation, it was

proposed that Fe2 S3 reacted further to form pyrite (FeS2 ) and greigite (Fe3 S4 ). Wiȩckowska

proposed that iron oxides only perform desulfurization when they are in their hydrated form

as shown in reaction (3) or as Fe(OH)3 as shown by the following reactions:

2Fe(OH)3 + 3H2 S → 2FeS + S + 6H2 O (8)

2Fe(OH)3 + 3H2 S → Fe2 S3 + 3H2 O (9)

Fe(OH)3 is proposed to be part of the catalyst after the initial regeneration of the

spent catalyst. The regeneration of the ferric sulfide and pyrite are shown as follows:

3
2FeS + 3H2 O + O2 → 2Fe(OH)3 + 3S (10)
2

3
Fe2 S3 + 3H2 O + O → 2Fe(OH)3 + 3S (11)
2 2
3
Fe2 S3 + O2 → Fe2 O3 + 3S (12)
2

The overall desulfurization and regeneration reactions are shown in the equation:

2H2 S + O2 → 2H2 O + 2S (ΔH = −443.1 kJ) (13)

In this reactions (9) and (10), it is shown that water is needed to regenerate the spent

catalyst back to ferric hydroxide or hydrated ferric oxide. Due to the highly exothermic

10
process, like in the initial mechanism, it is noted that the oxygen or air used in the

regeneration be used carefully (Wiȩckowska, 1995).

Zinc and calcium oxides have also been studied and used to remove H2 S through the

following reactions:

ZnO + H2 S → ZnS + H2 O (14)

CaO + H2 S → CaS + H2 O (15)

H2 S removal by zinc oxide adsorption in a packed-bed reactor at 636 K and 1 atm

was studied by Kim and Park (K. Kim & Park, 2010). Inlet concentrations of 100, 500 and

800 ppmv were added to a gas mixture consisting of 22% CH4 , 18.7% H2 , 8.8% CO2 , and

5.5% CO which was balanced with steam. Breakthrough threshold was chosen to be at 2

ppmv H2 S detected in the effluent stream. It was observed that sulfur removal at each

different inlet concentration was independent of space velocity, which was varied from 6000,

8000, and 12000 hr −1 . The breakthrough times linearly decreased as the inlet concentration

was increased. It was found that while the H2 S loading capacities were approximately

constant at each space velocity, the H2 S capture capacity increased when inlet concentration

was changed. At 100 ppmv, the capacity calculated was 10.9-12.5 g S/ 100 g sorbent; 500

ppmv, 24.8-26.4 g S/ 100 g sorbent; and 800 ppmv; 28.3-30.4 g S/ 100 g sorbent. It was

theorized that the non-steam components alter the ZnO sulfidation mechanism and its

equilibration to explain the increased H2 S capacity when the sorbent is exposed to higher

space velocities (K. Kim & Park, 2010).

2.1.2.2.2 Biological methods. Oxidation has been used for odor control in gas

streams in numerous industries and in wastewater treatment for hydrogen sulfide and other

11
odorants. In this instance, hydrogen sulfide is converted to sulfuric acid in the presence of

bacteria and oxygen as shown by the following overall reaction:

H2 S + 2O2 → H2 SO4 (16)

Many different types of bacteria such as Thiobacillus, Thermothrix, Thiothrix, and

Beggiato have been studied and utilized for this purpose along with removing other inorganic

and organic sulfur compounds (Arellano-García, Le Borgne, & Revah, 2018; Awe, Zhao,

Nzihou, Minh, & Lyczko, 2017; Zhu, Li, Wu, & Dumont, 2017). Hydrogen sulfate, and

sulfuric acid once protonated, can be produced from hydrogen sulfide through the sulfite-

oxidase pathway utilized in Thiobacillus microorganisms (Vikrant et al., 2018). This

mechanism is shown in the equations below:

H2 S + 2O2 → SO4 2− + 2H + (17)

1
HS − + O2 → S 0 + H2 O (18)
2

1
S 0 + H2 O + O → SO4 2− + 2H + (19)
2 2

1 1
S2 O3 2− + H2 O + O2 → SO4 2− + H + (20)
2 2

The overall products will be converted into sulfuric acid. Due to this sulfuric acid

formation and associated pH reduction, microbial activity, mass transfer rate into the biofilm

and removal efficiency can be negatively affected (Ben Jaber et al., 2016; Jaber et al., 2014).

Removal efficiencies up to 96% at a pH of 1.2 with an H2 S concentration of 250 ppm were

observed. However, this efficiency decreased to 78% when H2 S loading was increased to 360

ppm, lowering the pH to 0.5 (Ben Jaber et al., 2016). This is attributed to sulfate buildup and

lower H2 S mass transfer at more acidic conditions (Ben Jaber, Couvert, Amrane, Le Cloirec,

12
& Dumont, 2017). Upping the watering flow rate and using basic packing media can counter

problems associated with acidic pH conditions (i.e. reduce sulfate accumulation).

2.2 Mercaptans/Thiol

2.2.1 Properties and analysis. Methanethiol (methyl mercaptan) is a colorless gas at

room temperature that has a highly malodorous smell akin to that of rotten cabbage and

rotten eggs and is flammable. It occurs naturally in the human body, particularly in the brain

and blood, and is present in natural gas in certain US regions, coal tar and some crude oils. It

is produced, desired and undesired, in multiple industries such as plastics, animal feeds, pulp

and paper, jet fuel, pesticides and natural gas. At high concentrations methyl mercaptan is

toxic and affects the central nervous systems. It is absorbed rapidly through inhalation and

builds up in the vascular system.

Ethanethiol (ethyl mercaptan) is a clear liquid at room temperature with a low boiling

point of 95-97 ℉ that is naturally found in natural gas and petroleum wells. Due to this low

boiling point, it is typically in the gas phase at bulkhead when removed from the petroleum

or natural gas stream. It is predominantly used as an odorant added in small concentrations to

odorless gaseous products to help warn of gas leakage. Ethyl mercaptan’s scent, similar to

skunks and cooked cabbage, can be detected by scent at extremely low threshold

concentrations of approximately 8.7 ppt according to ACS (American Chemical Society,

2007). It is also used as an adhesive stabilizer in some industries. In addition to being highly

flammable, it is also toxic. Vapor inhalation can cause muscular weakness, convulsions and

respiratory paralysis. Pulmonary irritation may occur at high concentrations. The liquid

irritates the eyes and skin while ingestion can cause nausea and irritation of the mouth and

13
stomach (U.S. National Library of Medicine, 2005). Both mercaptans can also be

decomposed into H2 S which is also a cause for concern.

Methyl and ethyl mercaptan are usually monitored via infrared sensors for in-situ

measurements. It is typically analyzed by gas chromatography (GC) through the use of a

flame photometric detector, a chemiluminescence detector, a thermal conductivity detector,

or a mass spectrometer.

2.2.2 Activated carbons. Like H2 S, activated carbons have been utilized to remove

the low chain mercaptans as well but still show some of the same problems; mainly low

removal. A study conducted by Zhao et. al showed that metal addition modification greatly

enhanced the catalytic activity when compared to the unmodified coconut shell based

activated carbon. They found that the Cu-AC and Ni-AC modified catalysts had the best

catalytic activity when removing methyl mercaptan followed by Al-AC, then Fe-AC and

finally Zn-AC (Zhao et al., 2015). In the study conducted by Cui et al., Fe-impregnated AC

showed to have the most effective removal of methyl and ethyl mercaptans when compared

to Zn, Cu, K and Na impregnated AC and oxidized AC. In the case of methyl mercaptan, Fe

and Cu showed approximately the same removal efficiency followed closely by Zn and then

oxidation. The alkaline impregnations showed a slight increase in methyl mercaptan capacity

when compared to virgin AC. Fe showed the highest affinity for ethyl mercaptan capacity

followed by Zn and Cu and then oxidation. Again, the alkaline impregnations showed a slight

increase in ethyl mercaptan uptake capacity when compared to virgin AC (Cui et al. 2008).

2.2.3 Magnesium oxide and magnesium oxide/silicon dioxide composite. Similar

to H2 S, metal oxides have also been studied in the removal of low chain mercaptans such as

methyl mercaptan. In a study conducted by Kim et al., different preparations of magnesium

14
oxides were synthesized and tested in the desulfurization of municipal gas using methyl

mercaptan as the model contaminant. The feed gas was methane containing 291 μmol/mol

methyl mercaptan. The preparations methods were polyol-meditation thermolysis (PMT),

hydrothermal and aerogel. The aerogel MgO showed the highest adsorption capacity when

compared to the PMT and hydrothermal MgO at both temperatures studied, 278K and 288K.

The sorption capacities of the hydrothermal and aerogel MgO were approximately 6 times

and 180 times, respectively greater than the PMT MgO at 278K. This was magnified at 288K

where the hydrothermal MgO showed 15 times removal and the aerogel MgO showed 476

times the removal of PMT MgO. The aerogel MgO was also slightly more favorable when

compared to AC in terms of adsorption capacity and mass transfer rate even though AC has a

higher surface area (Y.-H. Kim, Tuan, Park, & Lee, 2014).

2.2.4 Biofiltration. Biofiltration has been utilized for sulfur contaminant removal,

primarily H2 S, but it has also been tested for other compounds such as ethyl mercaptan. One

such study conducted by Ben Jaber et al. delved into the effectiveness of this technique for

ethyl mercaptan removal alongside H2 S and dimethyl disulfide contaminants present in the

stream (Jaber et al., 2014). Three different columns inoculated with activated waste sludge

upon pine bark and composted wood mulch were used. Each column was watered with a

different solution at a fixed rate of 0.344 L/day. The effects of nutrient concentration were

the main factor observed in this study. The first column operated without a nutrient supply

while different nutrient solutions were supplied to the second and third columns. The second

solution consisted 0.12 g/L K 2 HPO4 , 0.08 g/L (NH4 )2 SO4 , and 0.39 g/L Na2 CO3 , while the

third consisted of deionized water, 0.48 g/L (NH4 )2 SO4 , and 1.97 g/L Na2 CO3 . Ethyl

mercaptan removal was highly affected by the different nutrient solutions. In Column 1, only

15
30% was removed in the pine bark portion of the bed and only 72% on the total bed height of

the column without nutrients present. However, there was 90-97% ethyl mercaptan removal

when nutrients were present in the watering solution. An overall improvement was shown for

ethyl mercaptan removal with the high nutrient supply due to the neutral/alkaline pH of the

system. Without the nutrient supply and the simultaneous removal of H2 S, the biofilter bed is

highly acidic, leading to ethyl mercaptan microbial inhibition (Jaber et al., 2014).

16
Chapter 3: Objectives

The overall objectives of this study are to understand the characterization, and

catalytic and removal behavior of an iron-based catalyst for H2 S, methyl mercaptan, and

ethyl mercaptan removal and compare these results to a commercial standard. The other

major objective is to blend the catalyst with different filler materials in order to save on costs

while still exceeding the commercial standard’s sulfur removal. The catalyst was provided by

Coastal Chemical LLC. Methyl and ethyl mercaptan removal were studied because these

compounds can easily degrade into H2 S in downstream processes, leading to toxicity,

emissions and corrosion issues. Many characterization methods for solid and gas matrices

were applied in this study ranging from Brunauer-Emmett-Teller (BET) surface analysis,

Fourier Transform Infrared Spectroscopy (FTIR), X-ray fluorescence (XRF), X-ray powder

diffraction (XRD), thermogravimetric analysis (TGA), gas chromatography-mass

spectroscopy (GC-MS), and Proton Induced X-ray Emission (PIXE) Spectroscopy. The

specific objectives of this study were to:

1. Characterize the raw catalyst

2. Determine adsorption vs catalytic activity

3. Determine and optimize parameters for most effective removal

4. Determine product formation

5. Characterize the spent catalyst

6. Compare the catalyst to a commercial standard used for sulfur removal

7. Test sulfur removal using a 1:9 catalyst/filler ratio and compare removal to the

commercial standard at different parameters, all of which were chosen by Coastal

Chemical LLC.

17
Chapter 4: Materials and Methods

4.1 Chemicals

Both catalysts (Coastal and Sulfatreat) and the three supplied filler materials

(anthracite, clay and graphite) used in the experiments were provided by Coastal Chemical

Co., LLC (Broussard, LA). The model gas used in the flow reactor system was obtained from

CSI Gas, LLC (Humble, TX) while the hydrogen sulfide (H2 S), methyl mercaptan (CH3 SH),

and ethyl mercaptan (C2 H6 S) standards used for the analytical works were obtained from

AccuStandard (New Haven, CT) and AirGas (Lafayette, LA)

4.2 Sample Preparation

The Coastal catalyst was sent through a series of sieve trays to obtain a desirable

particle size, between 425 μm and 850 μm (US Mesh 40). In early testing, some of the

Coastal catalyst samples were heat-treated at 300℃ to remove water and any trace

hydrocarbons potentially present on the surface of the raw catalyst to study the effect of these

compounds had on sulfur removal. It was determined that the heat treatment had a negative

effect on the removal of the three target sulfur compounds. The Sulfatreat catalyst, initially in

pellet form, was crushed with a pestle and mortar to obtain a similar particle size to the

Coastal catalyst and also to fit in the reactor bed. The filler materials were each mixed with

1:9 weight proportion of Coastal catalyst to filler material. The mixtures were then run

through a grinder to attain smaller and more uniform pieces. These ground mixtures were

then separately sieved to collect the 425 μm to 850 μm particle sizes. Glass wool, 1g of

catalyst/mixtures, followed by more glass wool were inserted in a steel tube serving as the

reactor bed in the system.

4.3 Reactor Configurations

18
Figure 2: Pilot Reactor System

All pilot scale experiments were carried out using a continuous flow packed column

reactor. The initial systems are depicted in Figures 2 and 4 showing the gas flow piping,

heating, valves, pumps, flow meters, and sampling ports. This system was later modified

with a T-joint at the base of the reactor to add a slip stream for flow to be sent to the Agilent

490 Micro GC for analysis. All of the gas flow lines used were 1/8 th stainless steel tubing.

The reactor tubing was a ¼ in stainless steel tubing.

19
Figure 3: Bench Scale Reactor System

All bench scale experiments were carried out using a continuous flow packed column

reactor shown in Figure 3. The reactor effluent was sent to the Agilent 6890N GC System

connected to an Agilent 5975 MS detector for analysis.

4.4 Reactor Start Up

After the pilot-scale reactor bed depicted in Figures 2 and 4 was loaded with catalyst,

it was connected into the system. It was pressurized at either 100 psig or 200 psig with a

nitrogen gas flow rate of either 36 mL/min or 50 mL/min controlled with a needle valve

before Flow Meter 2. Once the operating pressure for that particular run was reached, Valves

1 and 18 were closed to block in the system. With these valves closed, the system was leak

checked using Snoop and was then left for an hour to determine if there were any small leaks

that the Snoop did not find by measuring the pressure drop on the pressure gauge. Once the

20
system was determined to be leak free, the reactor was wrapped with a heating coil for

temperature requirements above ambient temperature. The rheostats for the pre-heater and

the reactor were then turned on and adjusted. The pre-heater rheostat was maintained at

approximately 50 V, heating the stream up to 150 °C. This vaporized any moisture in the

tubing before the reactor. The reactor rheostat was increased to the appropriate voltage to aim

for 100 °C and 200 °C, respectively. The correct voltages for the rheostats were determined

through trial and error.

Figure 4: Schematic of Pilot Scale Reactor (S. Sovine, 2012)

Once the reactor was brought up to and stabilized at the desired temperature, the

system block valve before Flow Meter 18 was opened and the nitrogen was allowed to vent.

21
The feed valve, Valve 2, was then opened to send the feed gas into the reactor system. This is

noted as time zero. The flow rate was adjusted to the desired flow, and the temperature was

allowed to equilibrate.

In the bench-scale experiments, the reactor was loaded with 1 gram of material.

Depending on the parameters, a heater coil was wrapped around the reactor and a

thermocouple. These were then wrapped with insulation, and the temperature was raised and

held at 100°C via a rheostat. While the reactor system was brought up to temperature, a

nitrogen flow was introduced to the system and blocked in. This allowed for leak testing

along all connections from the gas tank all the way to the downstream mass flow controller.

Once the temperature has stabilized and all connections checked and corrected for leaks, the

nitrogen was purged from the reactor. Then, the online GC-MS sequence was prepared. Like

the pilot system, the model gas flow was then introduced into the system at either 36 mL/min

or 50 mL/min depending upon manual settings on the mass flow controller. The GC-MS

sequence was immediately started once the flow was stabilized at the correct pressure.

4.5 Product Sampling

While the pilot-scale reactor was used, sampling began an hour after startup by

withdrawing 0.5 mL of gas. Initially, the effluent gas was tested every 4 hrs. After 12 hrs, the

sampling time was changed to every 2 hrs. After an additional 6 hrs, sampling was conducted

hourly until breakthrough. Breakthrough was defined differently between H2 S and the

mercaptans due to concentration differences and compound sensitivity with the MS detector.

H2 S breakthrough was defined to be reached at a concentration threshold of 10% of the

original concentration while the mercaptans were defined when concentrations reached 40%

of their original concentrations. To determine concentration and breakthrough of the sulfur

22
contaminants in bulk methane, gas samples were taken from the gas port after Flow Meter 1

shown in Figure 4. A gas-tight syringe was used to take samples. The syringe was plunged

into the septum of the gas port and flushed numerous times prior to sample collection to

ensure a representative sample.

During the bench scale experiments conducted in the reactor system shown in Figure

3, the reactor effluent stream was directly introduced into the GC-MS. This online method

utilized a valve as the connection for the effluent stream to pass into the GC-MS column.

This valve served as the “injection” of the effluent sample. When closed, the effluent gas was

diverted and passed to the exhaust line that was fed into a vent for safety.

4.6 Reactor Shutdown

An experimental run was deemed complete once breakthrough was achieved. When

this occurred, the shutdown procedure was initiated. The inlet gas and rheostat for controlling

the reactor temperature were turned off. Once the pressure due to the test gas was reduced in

the line, the nitrogen valve was opened to purge the system of remaining test gas. The system

was allowed to purge for approximately 2 hours. Once the reactor came down to room

temperature and the nitrogen was finished purging, the nitrogen source valve was closed.

When the system pressure dropped to atmospheric pressure, the packed bed reactor was

removed from the system. The catalyst was removed as quickly as possible from the reactor

segment and into a closed scintillation vial for further analysis. The reactor was then washed

with solvent and dried in an oven until the next experiment.

4.7 Analytical Methods

4.7.1 GC Analysis of gas samples. Two different GC units were utilized to analyze

the concentrations of the test gas downstream of the pilot scale reactor. Initially, samples

23
were injected into an Agilent Technologies 6890N Network Gas Chromatograph with a 30 ft,

0.320 mm inner diameter GS-GASPRO column equipped with an Agilent Technologies 5975

Inert Mass Selective detector and a thermal conductivity detector. The MS filament

temperature was set at 170℃ and scanned from 30 m/z to 80 m/z. The data and chemicals

were quantified and identified through the use of the Enhanced MSD Productivity

ChemStation’s library and data analysis programs. Later, an Agilent Technologies 490 Micro

GC was installed downstream of the reactor to better integrate and automate the effluent

analysis. Two different columns were used in this Micro GC. A PoraPLOT U column was

used in the first channel while a CP-Sil 13 CB for TBM column was used in the second

channel. The PoraPLOT U column was used for H2 S detection while the CP-Sil 13 was used

for the mercaptan detection due to the columns’ selectivity for the analytes and separation

efficiencies with respective column lengths and thicknesses.

Two of the original eight parameter variations, 200 psi, 100℃, and 50 mL/min and

200 psi, ambient, and 50 mL/min, were not included in this document. Both of these

experiments were conducted in either the pilot reactor system that did not utilize the

automated sampling for Agilent Technologies 6890N GC/5975 Inert MS. One variation was

manually sampled using a different time schedule and GC-MS method for detection. The

other variation used an inline Agilent 490 Micro GC for detection. This GC was provided by

Coastal and had been already used for other experiments in removing higher chain and

heavier sulfur compounds. Due to how this GC worked, the gasifier maxed out a temperature

that was insufficient for gasifying these heavier compounds which lead to inaccurate data to

show in the chromatograph.

24
Figure 5: Agilent Technologies 6890N Network GC equipped with Agilent Technologies

5975 Inert MS Detector

Originally, gas samples were extracted from the gas sampling port downstream of the

pilot scale reactor via a 1.0 mL gas syringe. The syringe was introduced into the sample port

and flushed numerous times to ensure a representative sample. After the sample was drawn

but before removal from the port, the valve of the syringe was closed to prevent any air from

entering the system and to prevent any of the sample to escape. The sample was then injected

into the GC-MS-TCD and analyzed using the method in Table 1.

Table 1: GC Oven Method for Injection Sample Analysis

Temperature Rate Increase Holding Time

40℃ --- 4 min

180℃ 10℃/min 2 min

In the bench-scale experiments, gas sampling was automated by connecting a GC-MS

directly to the reactor effluent stream, downstream of the mass flow controller. A

pneumatically controlled gas sample valve was connected to a split/split-less injector. Until a

25
sample injection was made, the valve controlling access to the GC column was closed. When

an injection was to be made, the valve was opened for a specified amount of time. Once

injected, the sample was analyzed using the method in Table 2. Once the designated time had

passed, the valve was then closed.

Table 2: GC Oven Method for Online Sample Analysis

Temperature Rate Increase Holding Time

40℃ --- 30 sec

130℃ 25℃/min

300℃ 12℃/min 2 min

4.7.1.1 Calibration curve. Calibration of the products was first carried out by

injecting different volume amounts of gas standards: 500 ppm H2 S, 50 ppm CH3 SH, and 50

ppm C2 H6 S in bulk CH4 , into the Agilent Technologies 6890N Network Gas Chromatograph

and recording the peak area counts at these volumes. A calibration curve was prepared from

known volumes injected and the area counts. Once the experimental setup changed to include

automated sampling, the calibration of the products was conducted by repeatedly flowing the

gas standards through an empty reactor bed every 26-30 min to obtain the average area count

for the gas standards’ unmodified concentrations. Calibration curves relate the average area

counts with the concentrations of the gas standards. The curve is used to later convert area

counts of test samples collected during experiments to concentrations. A total ion

chromatograph of the model gas, and what is being replicated through the online calibration,

is shown in Figure 6.

26
Figure 6: Chromatograph of Model Gas

4.7.1.2 Reaction studies. The results from the GC analysis were analyzed using the

calibration data to provide the actual concentrations of H2 S, methyl mercaptan and ethyl

mercaptan entering and exiting the reactor system. These concentrations were used to

C
generate breakthrough graphs of Cf vs time for different reaction parameters evaluated for the
i

catalysts/fillers.

Due to the unknown speciation of the Coastal catalyst and that sorbents are used as

filler for the Coastal catalyst, sulfur removal was calculated using the following equation:

(Ci − Cf ) ∗ 𝑡 ∗ Q
q= (21)
M

Where q is the amount removed (g/g), Ci and Cf (ppm) are the individual initial and

final concentrations of H2 S, methyl mercaptan and ethyl mercaptan, respectively, Q is the

volumetric flow rate of the model gas (mL/min), t is the time of the reaction (hr), and M is the

mass of the catalyst loaded (g). This equation was used to track the amount of sulfur removal

while varying the temperature, pressure, flow rate, and catalyst in the reactor.

4.7.2 Surface analysis.

4.7.2.1 Surface area and porosimetry measurement. The surface area, pore volume,

particle size and pore size were delivered upon completion through equations developed by

Brunauer, Emmett and Teller to explain the physical adsorption of gas molecules on the

27
catalyts’ solid surface. The BET theory is an extension of the Langmuir theory except that it

focuses on multilayer adsorption with some assumptions. The general BET equation is:

1 𝑐−1 𝑝 1
= ( )+ (22)
𝑣 [(𝑝0 ⁄𝑝) − 1] 𝑣𝑚 𝑐 𝑝0 𝑣𝑚 𝑐

and c, the BET constant is expressed as

𝐸1 − 𝐸𝐿
𝑐 = exp ( ) (23)
𝑅𝑇

With p and 𝑝0 , the adsorbates’ equilibrium and saturation pressures at the temperature of

adsorption, v is the adsorbed gas quantity and 𝑣𝑚 is the monolayer adsorbed gas quantity. In

the BET constant expression, 𝐸1 is the heat of adsorption for the first layer, 𝐸𝐿 is the heat of

adsorption for the second and higher layers and is equal to the heat of liquefaction or heat of

vaporization.
𝑝
This process can also be plotted as a straight line with 𝑝 on the x-axis and the left
0

side of equation 22 on the y-axis. The slope (A) and y-intercept (I) can be used to calculate

𝑣𝑚 and c by the following equations:

1
𝑣𝑚 = (24)
𝐴+𝐼

𝐴
𝑐 =1+ (25)
𝐼

These values are then used to calculate the total surface area (𝑆total ) and specific

surface area (𝑆BET ), given by

𝑣𝑚 𝑁𝑠
𝑆total = (26)
𝑉

𝑆total
𝑆BET = (27)
𝑎

28
Where N is Avogadro’s number, s is the adsorption cross section of the adsorbing species, V

is the molar volume of adsorbate gas, and a is the mass of the adsorbent.

Figure 7: Micromeritics ASAP 2020 Surface Area and Porosimetry Analyzer

To run an analysis on the Micromeritics ASAP 2020 surface area and porosimetry

analyzer (Figure 7), a small sample of approximately 0.1 to 0.25 g was first placed in a tared

sample vial. The sample vial was then degassed at 350℃ for at 600 min with an empty

sample vial in the other port to ensure gas flow continued in the system. After degassing, the

sample was allowed to cool. While the sample cooled, the dewar was filled with liquid

nitrogen. The cooled, degassed sample vial was weighed to get the degassed sample weight

and screwed into the analysis port. The dewar was then placed on the elevator, and the

method was started. The elevator raised up until the dewar was in place over the vial in the

analysis port, and then the cover was placed over the analysis port and dewar unit. The gas

29
flow and liquid nitrogen levels were monitored throughout the entire analysis, about a 7 hr

process. The samples were run through method completion. The sample was then removed

and disposed of.

4.7.2.2 Fourier Transform Infrared Spectroscopy (FTIR). An attempt to measure

the surface functional groups was made before anything was known about the chemistry of

the two catalysts and the blends. This was carried out using two different spectrometers. The

two machines used were a Jasco Ft/IR-4700 and an Agilent Technologies Cary 640 FTIR and

are shown in Figures 8 and 9.

Figure 8: Jasco FT/IR 4000 series

30
Figure 9: Agilent Technologies Cary 640 FTIR

Precursory imagery of the raw catalyst in Figure 10 showed similarities to the IR

spectra of iron oxide and calcium oxide. Both iron and calcium were confirmed to be the two

major components of the catalyst in XRF analysis, and the spectra agree with this finding.

The spectra band at approximately 520 is consistent with Fe2 O3 while the peaks at 874 and

1424-1458 match up to FTIR standards of CaO found in literature.

31
Figure 10: IR spectra of Raw Coastal Catalyst

4.7.2.3 Thermogravimetric Analysis (TGA).

Figure 11: SDT 2960 Simultaneous DSC-TGA

32
The catalyst was submitted to thermal treatment to find out water content and

possible contamination of compounds such as hydrocarbons. Once the catalyst weight was

recorded and placed inside the TGA’s heating chamber, the TGA was heated from ambient

temperature to 1000℃ at a heating rate of 20℃/min. The TGA continuously recorded mass

and the concurrent temperature during this heating rate. This data was used to plot a TGA

curve, allowing a better understanding of reactions, degradations, and product

formation/removal.

4.7.2.4 Proton Induced X-ray Emission (PIXE) Spectroscopy.

Figure 12: Linear Accelerator for PIXE Analysis

PIXE analysis was used to determine the elemental composition of the catalyst as it is

a non-destructive technique and highly accurate for homogeneous particles. The catalyst was

bombarded with an ion beam from an ion accelerator. The beam’s energy displaced an inner

shell electron which caused a valence electron to drop an energy shell level by emitting

energy. This energy emitted was observable in the X-ray spectrum. Elements have signature

33
energy emissions and transitions which are measured by an energy dispersive detector. One

limitation of PIXE analysis is that elements below sodium do not show up as the elements’

rays will be absorbed before reaching the detector.

34
Chapter 5: Results and Discussion

5.1 Catalyst Analysis

5.1.1 BET surface analysis.

Table 3: BET characteristics of the catalysts and fillers

Catalyst Surface Area Pore Volume Pore Size Average

(m2/g) (𝐜𝐦𝟑 /g) (Å) Particle Size

(Å)

Coastal 201.11 0.38 72.15 285

Sulfatreat 55.78 0.0004 20.88 1035

Clay 124 0.30 96.44 484

Anthracite 22.38 0.036 64.97 2680

Graphite 9.61 0.021 83.54 6002

When the catalysts presented in Table 3 are compared to other catalysts and sorbents

in the literature, such as the popular activated carbons, they are on the lower end of measured

BET surface characteristics. Most of the activated carbons have a minimum surface area of

approximately 700 m2 /g with many peaking over 1000 m2 /g (Guo et al., 2007; Xiao et al.,

2008). Activated carbons (AC) using Fe, Mn, Co, Ce, V, and Cu modifications via

impregnation still showed high BET surface areas in the 800-900 m2 /g range (Fang et al.,

2013). In the literature, mesoporous silica supported adsorbents were synthesized and

modified, greatly changing their surface chemistry. All of the unmodified adsorbents in this

study had roughly doubled BET surface areas when compared to their modified versions.

35
Once modified through impregnation via the incipient wetness method, these adsorbents had

comparable surface areas to the Coastal catalyst (Hussain, Abbas, Fino, & Russo, 2012)

5.1.2 TG analysis. TG analysis was conducted to determine water content, organic

compounds deposited on the catalyst surface, and the temperature region of potential catalyst

degradation. The result in Figure 13 show that approximately 8-10 weight % of the catalyst

weight is lost at 150℃ and exhibits a steady weight loss between 350℃ to 550℃. The first

two peaks of the derivative weight line show the more precise amount lost per ℃. It is

theorized that the first peak on the derivative weight line is water loss off of the catalyst

while the smaller second peak is an adsorbed volatile organic compounds. There is a small

peak between 200℃ to 250℃ that could possibly be attributed to adsorbed contaminants

being released while there is another 2-2.5 weight % loss between 650℃-700℃. It is

possible this weight loss is due to catalyst degradation.

Figure 13: TGA Results of Raw Coastal Catalyst

5.1.3 FTIR Characterization of spent catalyst. FTIR results of the spent catalyst

shown in Figure 15 showed some differences when compared to the raw catalyst’s spectra.

The water peak around 3400-3500 cm−1 disappeared as expected, since the water evaporates

36
at the conditions of the experiment. The CaO peaks also disappeared or was significantly

reduced in intensity. This is most likely due to sulfur product buildup, particularly iron pyrite

and greigite, on the catalyst surface. However, the expected peak for iron pyrite was not

observed because it is detected below 500 cm−1 , which is the minimum limit for the

instrument used in this study. The other speculated iron product based on the literature,

greigite, cannot be identified at this time due to the unavailability of a pre-existing spectra.

One of the peaks representative of CaO in the raw catalyst was also found in the spent

catalyst; however, the intensity of this peak (1424-1458 cm−1 ) was significantly weaker

when comparing transmittance level. The other indication of CaO presence, the peak at 874

was not found in the spent catalyst. The 30-40% transmittance strength difference between

the two spectra (lower for spent catalyst) suggests sulfur product buildup on the catalyst

surface. There is also possibly some regeneration of the catalyst between removal from the

system and the FTIR scanning due to air exposure which provides oxygen required for the

regeneration reactions.

3
2FeS + 3H2 O + O2 → 2Fe(OH)3 + 3S (10)
2

3
Fe2 S3 + 3H2 O + O → 2Fe(OH)3 + 3S (11)
2 2

3
Fe2 S3 + O2 → Fe2 O3 + 3S (12)
2

These reasons can explain why there is some CaO found in the catalyst post-reaction and

why the % transmittance is so low. Another possible explanation is that CaO was not

completely reacted at the reaction conditions of the experiment.

37
Figure 14: FTIR Spectra of Raw Coastal Catalyst

Figure 15: FTIR Spectra of Spent Coastal Catalyst

38
5.1.4 PIXE analysis.

Figure 16: PIXE Elemental Maps of Unreacted Coastal Catalyst

MicroPixe analysis was conducted utilizing a 2 MeV proton beam from a 1.7 MV

Pelletron accelerator with an Amptek Silicon drift detector for X-ray detection and a silicon

surface barrier detector to measure the backscattering spectrum. PIXE mapping of the raw

Coastal catalyst in Figure 16 agrees with the FTIR results that iron oxides and calcium oxide

are the major components of the catalyst. The catalyst granules analyzed showed an Fe-rich

phase with Ca inclusions and trace concentrations of Cr, Mn, Mo, Ti, Zn, and S. Some of

these trace concentrations along with the Fe and Ca mappings are shown in Figure 17. It is

seen in Figure 18 that there are two forms of iron. It is likely that they are hematite and

magnetite, or Fe2 O3 and Fe3 O4 . Overlaying the different element mappings of the spent

catalyst shows the sulfur accumulated from the reactions deposited in the Fe-regions and

very little in the Ca pockets. In Figure 19, it is observed that sulfur is present in the spent

catalyst shown by a characteristic peak around 2.5 keV that is associated with the sulfur

39
element. This peak is not found in the virgin catalyst. The sulfur response is lower than

anticipated considering the run times for the Coastal catalyst experiments. This is likely due

to catalyst regeneration between reactor breakdown and the PIXE analysis and possibly no

reaction between the magnetite and sulfur compounds.

Figure 17: PIXE Elemental Maps of Spent Coastal Catalyst

40
Figure 18: PIXE Analysis of Unreacted Coastal Catalyst

Figure 19: PIXE Analysis of Spent Coastal Catalyst


41
5.2 Sulfur Removal

Table 4: Hydrogen Sulfide Breakthrough Times

𝐇𝟐 𝐒 Coastal Graphite Clay Anthracite Sulfatreat

100 psi 172.5 hr 2.6 hr 18.6 hr 12.5 hr 0.5 hr

100℃

50 mL/min

100 psi 236 hr 5.6 hr 22.5 hr 173 hr 0.5 hr

100℃

36 mL/min

200 psi 246 hr 5.6 hr 34.2 hr 16 hr Immediate

Ambient

36 mL/min

200 psi 276.5 hr 7.4 hr 33.8 hr 28.2 hr 0.5 hr

100℃

36 mL/min

100 psi 121 hr 2.2 hr 13.9 hr 6 hr Immediate

Ambient

50 mL/min

100 psi 175 hr 3 hr 24.2 hr 12.5 hr 0.5 hr

Ambient

36 mL/min

42
Table 5: Methyl Mercaptan breakthrough Times

Methyl Coastal Graphite Clay Anthracite Sulfatreat

Mercaptan

100 psi 132 hr 1.73 hr 15.6 hr 8.2 hr 10.8 hr

100℃

50 mL/min

100 psi 167 hr 4.75 hr 13.4 hr 12.5 hr 0.5 hr

100℃

36 mL/min

200 psi 243 hr 4.33 hr 24.7 hr 10.8 hr Immediate

Ambient

36 mL/min

200 psi 174 hr 6 hr 21.2 hr 17.3 hr Immediate

100℃

36 mL/min

100 psi 101 hr 2.2 hr 13.9 hr 6.5 hr Immediate

Ambient

50 mL/min

100 psi 144 hr 2.2 hr 20.8 hr 4.3 hr 0.5 hr

Ambient

36 mL/min

43
Table 6: Ethyl Mercaptan Breakthrough Times

Ethyl Coastal Graphite Clay Anthracite Sulfatreat

Mercaptan

100 psi 159 hr 3 hr 20 hr 9.1 hr 13.4 hr

100℃

50 mL/min

100 psi 174 hr 5.6 hr 30 hr 13.4 hr 0.5 hr

100℃

36 mL/min

200 psi 245 hr 5.2 hr No 12.5 hr Immediate

Ambient breakthrough

36 mL/min

200 psi 182 hr 7 hr 29.5 hr 17.8 hr Immediate

100℃

36 mL/min

100 psi 104 hr 3.9 hr No No Immediate

Ambient breakthrough breakthrough

50 mL/min

100 psi 149 hr 4.7 hr No 12.5 hr 0.5 hr

Ambient breakthrough

36 mL/min

44
H2 S, methyl mercaptan, and ethyl mercaptan removal were investigated as a function

of time, pressure, temperature, flow rate, and catalyst. Experiments provided the data to

optimize operating conditions for maximum removal efficiency per compound. Removal

efficiencies vs time graphs of the three compounds are shown in Figures 20-52. Removal

results can be observed as a function of time and operating conditions (temperature, pressure,

and flow rate).

5.2.1 Coastal experiments. All runs that utilized the pure Coastal catalyst are shown

in Figures 20-22. In Figure 20, the experiment resulting in the longest breakthrough time

with an effective run time of 276.5 hr removing H2 S was conducted at 200 psi, 100℃, and 36

mL/min followed by the experiment at 200 psi, ambient, and 36 mL/min. This superior

performance (longer breakthrough time) is due to the higher pressure, temperature, and lower

flow rate (longer residence time) which are more desirable reaction conditions when

compared to the other variations. The higher temperature lends more energy that is needed to

meet the activation energy for the reaction to occur. The relationship between the rate

constant, temperature, and activation energy is shown in the Arrhenius equation:


−𝐸𝑎
𝑘 = 𝐴𝑒 𝑅𝑇 (28)

Where k is the rate constant, A is a pre-exponential factor, a constant that is reaction

dependent, 𝐸𝑎 is the activation energy for the reaction (J/mol), R is the universal gas constant

(J/mol*K), and T is absolute temperature (K). Through the ~75 ℃ operating differential, an

approximate 30 hr difference between breakthrough was observed from the 200 psi, 100℃

and 36 mL/min run compared to the 200 psi, ambient, 36 mL/min run, the next variation with

the highest removal. This significant difference in performance indicates that a chemical

reaction is occurring in addition to adsorption. There is also a relationship dependence on

45
pressure as there is a 40 hr run time difference between the 200 psi, 100℃, and 36 mL/min

and 100 psi, 100℃, and 36 mL/min runs. The other major impact is the flow rate. Between

the most effective run and its high flow rate variation, there was approximately an 80 hr run

difference. This parameter exhibited the greatest change between removals when compared

to temperature and pressure changes. The residence time for the two flow rates are 2 seconds

at 36 mL/min and 1.44 seconds at 50 mL/min by use of the following equation:

V
𝜏= (29)
Q

Where 𝜏 is the residence time, V is the reactor volume, and Q is the flow rate. With the runs

within Figure 20, the temperature changes seemed to have the greatest affect alongside flow

rate changes. This is verified by the 80 hr difference in run time. With the energy from the

raised temperature, the activation energies for the three reactions were clearly met. The

pressure changes showed that the pressure increase positively impacted removal. The shortest

run, approximately half of the longest run, was at 100 psi, ambient, and 50 mL/min. The

ambient temperature does not provide the necessary energy to reach the reaction activation

threshold needed to break the compound bonds. The lower pressure correlates with lower

concentrations in and on the catalyst surface during gas phase heterogeneous catalysis. The

H2 S reaction is further affected by the higher flow rate which results in a lower residence

time for reactions with the catalyst. All Coastal runs showed improved H2 S removal

compared to literature results.

46
Figure 20: Coastal H2 S Breakthrough vs Time

The most effective run for methyl mercaptan removal by approximately 70 hours was

at 200 psi, ambient, and 36 mL/min with an effective removal time of 240 hours. The next

were 200 psi, 100℃, and 36 mL/min and 100 psi, 100℃, and 36 mL/min. The least effective

run in terms of breakthrough time was at 100 psi, ambient, and 50 mL/min with a run time of

100 hours. Unlike H2 S, methyl mercaptan does not show temperature dependencies for

removal. Pressure shows the greatest effect on removal. It is likely that physisorption rather

than catalysis (chemi-sorption) is the dominant process in regards to methyl mercaptan

removal. The higher pressure reduces the gas volume and increases its density. This increases

the chance of a contaminant atom to be adsorbed onto the surface of the catalyst. Adsorption

is an exothermic process. When a molecule is adsorbed onto the surface, the molecule’s

freedom of movement becomes restricted which is a decrease in entropy and an increase in

heat. The lower temperature combats this heat increase by lowering gas volatility and

47
increasing the entropy change, maximizing the driving force for adsorption. The most

effective run time drops from 240 hours to 140 hours when the pressure is changed from 200

psi to 100 psi while at ambient temperature and 36 mL/min. Flow rate still plays a role in

removal. The lower flow rate gives a higher residence time, increasing the contact time for

adsorption.

In terms of ethyl mercaptan removal, the most effective parameter variation was also

at 200 psi, ambient, and 36 mL/min with an effective operating time of 245 hours. The next

highest efficiency parameter is 200 psi, 100℃, and 36 mL/min with approximately 180

hours. The most ineffective parameter set is 100 psi, ambient, and 50 mL/min with 100 hours

of run time. Like methyl mercaptan, there is not a positive impact of adsorption at high

temperature compared to the pressure effects on removal. When the pressure is changed from

200 psi to 100 psi while at ambient temperature and 36 mL/min, the runtime drops by 100

hours. This alongside with the reduced temperature effects suggest that adsorption rather than

reaction is the main sulfur removal process.

48
Figure 21: Coastal Methyl Mercaptan Breakthrough vs Time

Figure 22: Coastal Ethyl Mercaptan Breakthrough vs Time

5.2.2 Graphite blend experiments. The graphite/Coastal blend showed that the most

effective removal parameters in the case of H2 S were the 200 psi, 100℃, and 36 mL/min and

100 psi, 100℃, and 36 mL/min. These two runs showed almost identical 100% efficient run

times at 8 hours. This shows that there is not a significant dependency on pressure while at

49
elevated temperatures. At this temperature and flow rate, the reaction appears to follow a 0 th

order reaction as pressure changes do not have an effect. A 0th order reaction is characterized

as follows:

reaction rate = k (30)

Where k is the rate constant. However, this cannot be confirmed off of these two runs alone.

The following variations: 100 psi, 100℃, and 50 mL/min and 100 psi, ambient, and 50

mL/min showed the worst performance with similar breakthrough times of 2.5 hours. It is

inferred that at the lower pressure, temperature does not have a significant effect on removal.

For optimal removal, the higher temperature is needed to provide the activation energy

requirement for reaction using the graphite filler. The higher flowrate decreased the gas

residence time, allowing for less time along the catalyst bed. It is noted that at this higher

flow and low pressure that while the temperatures were varied, the results were

approximately the same. The surface chemistry of graphite coupled with the lower pressure

and higher flow are likely responsible for this performance. The surface area and pore

volume of graphite are poor compared to the other fillers and the Coastal catalyst. Graphite’s

small surface area has a negligible impact on the number of available active sites for

reaction/adsorption. The lower pressure further inhibits removal as stated earlier while the

higher flowrate reduces the gas residence time.

50
Figure 23: Graphite H2 S Breakthrough vs Time

In the case of methyl mercaptan and ethyl mercaptan, the most effective parameter

variation in terms of breakthrough time is 200 psi, 100℃, and 36 mL/min. This is due to the

higher temperature providing the necessary heat for surpassing the activation barrier, higher

pressure for catalyst penetration through the pores by reducing gas volume and increasing

density, and the increased residence time thanks to the lower flow rate. This is the same as

the H2 S removal; however, the effective breakthrough time is approximately 6 and 6.5 hours,

respectively, compared to the 8 hours for H2 S. With methyl mercaptan, the next two most

effective runs were at 200 psi, ambient and 36 mL/min and 100 psi, 100℃ and 36 mL/min.

Due to the differences between the pressure and temperature in these two runs, one is more

likely an adsorption (200 psi, ambient and 36 mL/min) that is highly dependent on the

limited surface area of graphite while the other is a reaction (100 psi, 100℃ and 36 mL/min)

based on the Coastal catalyst dispersed in the graphite that both have similar removal

effectiveness for methyl mercaptan once mass transfer and surface chemisty is accounted for.
51
The high pressure would aid in adsorption. This happened due to methyl mercaptan and ethyl

mercaptan being larger molecules than H2 S. As such, the catalyst surface and pores are

utilized and spent faster by methyl and ethyl mercaptan than by H2 S.

Figure 24: Graphite Methyl Mercaptan Breakthrough vs Time

Figure 25: Graphite Ethyl Mercaptan Breakthrough vs Time

52
5.2.3 Clay blend experiments. The most effective parameter sets for the clay/Coastal

blend are 200 psi, 100℃, and 36 mL/min with a breakthrough time of 36 hours followed

closely by 100 psi, 100℃, and 36 mL/min and 200 psi, ambient, and 36 mL/min with a

breakthrough time of 35 hours. Of these three runs, the two runs at 100℃ and 36 mL/min are

reaction dominant with pressure not having a noticeable effect on removal. The higher

temperature provided the required energy to meet the activation energy for reaction. Due to

no apparent pressure dependency, it can be assumed that the H2 S reaction is a 0th order

reaction at these conditions. The next two most effective runs were the runs at ambient

temperature and low flow rate. These two runs are adsorption dominant rather than reactive.

While the higher pressure run is comparable to the first two runs, there is a noticeable drop

when the pressure is lowered from 200 psi to 100 psi. Like other runs discussed earlier,

pressure is the major factor in removal via adsorption due to surface contact and pore

penetration. Flow rate had a drastic effect on H2 S removal in this catalyst set at the lower

pressures when compared to their lower flow rate counterparts.

Figure 26: Clay H2 S Breakthrough vs Time

53
In the methyl mercaptan reactions, the most effective parameter set is 200 psi,

ambient, and 36 mL/min followed closely by all other parameters that use 36 mL/min.

Shown in Figure 28, it is noticed that upon initial breakthrough from 100% removal

efficiency, the reactions still operate ranges from 50-80% removal efficiency for several

more hours before trending to a steady accumulation until the catalyst is completely spent. It

is possible that the competing reactions and adsorption sites on the catalyst along with

pressure inconsistencies throughout the run had an impact on removal for the larger

compounds. In regards for methyl mercaptan, adsorption was the more effective route by

utilizing the higher pressure, lower flow rate and ambient temperature for maximum removal.

The higher pressure and low temperature aids in adsorption.

Figure 27: Clay Ethyl Mercaptan Breakthrough vs Time

54
Figure 28: Clay Methyl Mercaptan Breakthrough vs Time

The ethyl mercaptan data for the clay blend shows that no sign of breakthrough at any

point through 52 hours for three parameters sets: 200 psi, ambient and 36 mL/min; 100 psi,

ambient, and 50 mL/min; and 100 psi, ambient, and 36 mL/min. The data for this particular

filler could possibly be warped due to sensitivity limitations for ethyl mercaptan with the GC

analysis method that was used. As per the Method section, the model gas was fed through the

empty reactor bed and into the GC for calibration runs between all runs utilizing the catalysts

and fillers. During the clay runs, there was little to no response for ethyl mercaptan due to an

outside change made to the GC method and detection limits. This sensitivity issue was

resolved in later runs. In the case of there being no detection issues, the adsorption process

showed more promise for ethyl mercaptan removal rather than a catalytic reaction. None of

the ambient temperature runs showed breakthrough during the duration of run times while all

three runs operated at 100℃ showed breakthrough between 20-30 hours. The lower pressure

seemed adequate for ethyl mercaptan removal at both flow rates to ensure 100% removal,

suggesting that clay has a favorable surface chemistry for ethyl mercaptan adsorption. There

55
was no data found in the literature with clay’s performance in regards to ethyl mercaptan

removal.

5.2.4 Sulfatreat experiments.

Figure 29: Sulfatreat H2 S Breakthrough vs Time

Figure 30: Sulfatreat Methyl Mercaptan Breakthrough vs Time

In Sulfatreat runs shown in Figures 29-31, it is observed that breakthrough was

immediate in all cases for each compound except for the 100 psi, 100℃, and 50 mL/min

56
experiment. This happened during duplicate experiments as well. The surface area and pore

volume along with powdered structure upon breaking the Sulfatreat are certainly the reasons

for poor performance. The reason for the effect seen in all three graphs associated with the

100 psi, 100℃, and 50 mL/min experiment is most likely due to blockage of the mass flow

controller downstream of the reactor and which was cleared out later near 13 hours into the

experiment. This is verified by immediate breakthrough shown for all three compounds at

this time.

Figure 31: Sulfatreat Ethyl Mercaptan Breakthrough vs Time

5.2.5 Anthracite blend experiments. For the Anthracite blend, the most effective

parameter variation for H2 S and methyl mercaptan removal via breakthrough time was 200

psi, 100℃, and 36 mL/min with breakthrough occurring at 29 and 16 hours, respectively. As

stated earlier, the higher temperature provided the needed energy to break the activation

energy barrier for the reaction to occur. It can be seen that doubling the pressure from 100 psi

to 200 psi at 100℃ approximately doubles the removal rate. This shows that the pressure is

57
directly proportional to the reaction rate, making this a 1 st order reaction. This is shown as

follows:

reaction rate = kCH2 S (31)

where k is the rate constant and CH2 S is the concentration of H2 S. The next most effective

parameter set is 100 psi, 100℃, and 36 mL/min followed closely by 200 psi, ambient, and 36

mL/min. At the 200 psi, ambient, and 36 mL/min variation, adsorption is the dominant

process and is roughly 33% more effective for H2 S removal than the 100 psi, ambient, and 36

mL/min variation.

Figure 32: Anthracite H2 S Breakthrough vs Time

58
Figure 33: Anthracite Methyl Mercaptan Breakthrough vs Time

The ethyl mercaptan breakthroughs for the clay blend shown in Figure 34 are almost

identical to the other two compounds with the exception of 100 psi, ambient, and 50 mL/min.

This experiment showed no breakthrough for ethyl mercaptan. The other runs follow the

same trend. Theoretically, the 100 psi, ambient, and 50 mL/min run should have been the

first experiment to show breakthrough when compared to the graphs of the other blends.

There is no explanation for this effect at this time other than to adjust the sensitivity of the

GC-MS scans for ethyl mercaptan and redo the experiment at that variation.

59
Figure 34: Anthracite Ethyl Mercaptan Breakthrough vs Time

In Figures 35-52 comparing the blends at specific parameter variations show that clay

was the most effective filler for H2 S, methyl mercaptan, and ethyl mercaptan removal when

compared to the commercially available Sulfatreat among all experiments. The next most

viable filler was anthracite followed by graphite. All three fillers were more effective when

compared to Sulfatreat with the exception of variation 100 psi, 100℃, and 50 mL/min;

however, this result is likely due to a blockage of catalyst downstream of the reactor and

upstream of the GC-MS unit. It is observed that the pore volume and surface area for clay,

anthracite and graphite are generally indicative of how well a filler adsorbent will perform.

Clay had the highest surface area and pore volume followed by anthracite and then graphite.

Sulfatreat had a lower pore size and volume when compared to graphite but a surface area

between anthracite and graphite. Since the initial Sulfatreat pellet was broken apart to utilize

in the reactor, this low pore volume and size hinder sulfur removal to just surface reactions.

While there are discernable differences between the three fillers’ effectiveness, there is a

large gap in performance between these and the Coastal catalyst. The most effective

60
parameter for all four catalysts was 200 psi, 100℃, and 36 mL/min. At this variation, it is

observed that clay came the closest to reaching the Coastal catalyst’s potential when used in

a 9:1 blend followed by anthracite with graphite having the greatest hindrance in aiding

removal. When comparing H2 S removal time ratios between the Coastal catalyst and the

three fillers, clay had the smallest ratio of 8:1. Anthracite had a removal time ratio of

approximately 10:1 while graphite’s was 35:1. Comparisons of breakthrough times for

methyl and ethyl mercaptans between these catalyst and fillers showed similar results.

5.2.6 Experimental catalysts comparison to literature. The maximum amounts of

H2 S, methyl mercaptan, and ethyl mercaptan removed per respective catalyst’s surface area

are shown in Tables 7-9. The Coastal catalyst showed the highest removal for all three

compounds, which were 1.47 g/m2 H2 S at 200 psi, 100℃, and 36 mL/min, 0.13 g/m2

methyl mercaptan at 200 psi, ambient, and 36 mL/min, and 0.13 g/m2 ethyl mercaptan at

200 psi, ambient, and 36 mL/min. Among the fillers, anthracite showed the highest removal

per square meter for hydrogen sulfide and methyl mercaptan while graphite showed the

highest removal per square meter for ethyl mercaptan. The highest H2 S removal for

anthracite occurred at 200 psi, ambient, and 36 mL/min calculated to be 1.25 g H2 S /m2 . The

other two highest removals occurred at 200 psi, 100℃, and 36 mL/min with removals of

0.067 g methyl mercaptan per square meter and a minimum of 0.073 g ethyl mercaptan per

square meter since breakthrough was not achieved.

When compared to the literature removals for H2 S, the Coastal catalyst was at least an

order of magnitude more effective while Sulfatreat was subpar. Out of the three fillers,

anthracite and clay showed the highest removal per gram of catalyst and these were

comparable, if not better, to adsorbents in the literature such as the AC modified variations

61
and zeolites. When compared to modified activated carbons derived from oil-palm shell (Guo

et al., 2007) where it was reported that the various adsorbtion capacities ranged from 46-76

mg H2 S/g, all three fillers showed higher removal capacities with the lowest being graphite.

Cu-ZnO supported on mesoporous silica showed adsorption capacities ranging from 9-80 mg

S/g sorbent (Elyassi et al., 2014). In a study regarding methyl mercaptan removal, it was

observed that virgin AC performed the worst whereas Ni-AC and Cu-AC performed the best

with 100% conversion after 180 minutes followed by Al-AC with 95% conversion, Zn-AC

with 92% conversion and then Fe-AC with 89% conversion. Virgin AC showed less than

45% conversion after only 80 minutes. No removal capacities were included in the published

study outside of removal efficiency vs breakthrough time with an initial concentration of

methyl mercaptan of 500 ppm balanced in nitrogen (Zhao et al., 2015). Removal efficiencies

for Ni-Ac and Cu-AC remained 100% through 180 minutes. Al-AC dropped to 95%

efficiency through 180 minutes while Zn-AC and Fe-AC dropped to approximately 92 and

90% efficiency by this time. Virgin AC fell to 45% removal efficiency by 80 minutes into the

experiment. As the Coastal catalyst is mainly an iron-based catalyst and only dealt with

1/10th of the methyl mercaptan load, it is possible that it is not as efficient when compared to

other metal modified ACs found in the literature. There could also be complications due to

the competing reactions of the three sulfur compounds studied in this work that the other

catalysts did not have. These modified ACs were more effective at removing methyl

mercaptan than the three filler materials when comparing loading weight. A study regarding

the removal of hydrogen sulfide and methyl mercaptan utilizing light expanded clay

aggregates (Leca) showed an affinity for removing these two compounds. Dry Leca showed

removal at 50 mL/min and 180 mL/min for 28 and 20 hours respectively with inlet

62
concentrations of approximately 900-1000 ppb for both compounds, comparable to the clay

and anthracite fillers studied in this work (Tabase, Liu, & Feilberg, 2013).

Figure 35: H2 S Breakthrough at 100 psi 100℃ 50 mL/min

Figure 36: Methyl Mercaptan Breakthrough at 100 psi 100℃ 50 mL/min

63
Figure 37: Ethyl Mercaptan Breakthrough at 100 psi 100℃ 50 mL/min

Figure 38: H2 S Breakthrough at 100 psi 100℃ 36 mL/min

64
Figure 39: Methyl Mercaptan Breakthrough at 100 psi 100℃ 36 mL/min

Figure 40: Ethyl Mercaptan Breakthrough at 100 psi 100℃ 36 mL/min

65
Figure 41: H2 S Breakthrough at 100 psi Amb 50 mL/min

Figure 42: Methyl Mercaptan Breakthrough at 100 psi Amb 50 mL/min

66
Figure 43: Ethyl Mercaptan Breakthrough at 100 psi Amb 50 mL/min

Figure 44: H2 S Breakthrough at 100 psi Amb 36 mL/min

67
Figure 45: Methyl Mercaptan Breakthrough at 100 psi Amb 36 mL/min

Figure 46: Ethyl Mercaptan Breakthrough at 100 psi Amb 36 mL/min

68
Figure 47: H2 S Breakthrough at 200 psi Amb 36 mL/min

Figure 48: Methyl Mercaptan Breakthrough at 200 psi Amb 36 mL/min

69
Figure 49: Ethyl Mercaptan Breakthrough at 200 psi Amb 36 mL/min

Figure 50: H2 S Breakthrough at 200 psi 100℃ 36 mL/min

70
Figure 51: Methyl Mercaptan Breakthrough at 200 psi 100℃ 36 mL/min

Figure 52: Ethyl Mercaptan Breakthrough at 200 psi 100℃ 36 mL/min

71
Table 7: Maximum removal of H2 S

Coastal Sulfatreat Clay Anthracite Graphite

100 psi 1.29 g/𝐦𝟐 0.20 g/𝐦𝟐 0.15 g/𝐦𝟐 0.81 g/𝐦𝟐 0.34 g/𝐦𝟐

100 ℃

50 mL/min

100 psi 1.27 g/𝐦𝟐 0.041 g/𝐦𝟐 0.26 g/𝐦𝟐 0.65 g/𝐦𝟐 0.58 g/𝐦𝟐

100 ℃

36 mL/min

200 psi 1.32 g/𝐦𝟐 0.051 g/𝐦𝟐 0.26 g/𝐦𝟐 0.72 g/𝐦𝟐 0.58 g/𝐦𝟐

Ambient

36 mL/min

100 psi 0.94 g/𝐦𝟐 0.18 g/𝐦𝟐 0.20 g/𝐦𝟐 0.02 g/𝐦𝟐 0.29 g/𝐦𝟐

Ambient

36 mL/min

100 psi 0.91 g/𝐦𝟐 0.059 g/𝐦𝟐 0.16 g/𝐦𝟐 0.38 g/𝐦𝟐 0.27 g/𝐦𝟐

Ambient

50 mL/min

200 psi 1.47 g/𝐦𝟐 0.15 g/𝐦𝟐 0.30 g/𝐦𝟐 1.25 g/𝐦𝟐 0.15 g/𝐦𝟐

100 ℃

36 mL/min

72
Table 8: Maximum Removal of Methyl Mercaptan

Methyl Coastal Sulfatreat Clay Anthracite Graphite

Mercaptan

100 psi 0.1 g/𝐦𝟐 0.03 g/𝐦𝟐 0.015 g/𝐦𝟐 0.045 g/𝐦𝟐 0.021 g/𝐦𝟐

100 ℃

50 mL/min

100 psi 0.09 g/𝐦𝟐 0.005 g/𝐦𝟐 0.011 g/𝐦𝟐 0.042 g/𝐦𝟐 0.049 g/𝐦𝟐

100 ℃

36 mL/min

200 psi 0.13 g/𝐦𝟐 0.004 g/𝐦𝟐 0.022 g/𝐦𝟐 0.050 g/𝐦𝟐 0.044 g/𝐦𝟐

Ambient

36 mL/min

100 psi 0.08 g/𝐦𝟐 0.013 g/𝐦𝟐 0.020 g/𝐦𝟐 0.0017 g/𝐦𝟐 0.020 g/𝐦𝟐

Ambient

36 mL/min

100 psi 0.08 g/𝐦𝟐 0.006 g/𝐦𝟐 0.016 g/𝐦𝟐 0.041 g/𝐦𝟐 0.027 g/𝐦𝟐

Ambient

50 mL/min

200 psi 0.1 g/𝐦𝟐 0.014 g/𝐦𝟐 0.018 g/𝐦𝟐 0.067 g/𝐦𝟐 0.063 g/𝐦𝟐

100 ℃

36 mL/min

73
Table 9: Maximum Removal of Ethyl Mercaptan

Ethyl Coastal Sulfatreat Clay Anthracite Graphite

Mercaptan

100 psi 0.089 g/𝐦𝟐 0.035 g/𝐦𝟐 0.044 g/𝐦𝟐 0.058 g/𝐦𝟐 0.041 g/𝐦𝟐

100 ℃

50 mL/min

100 psi 0.095 g/𝐦𝟐 0.0057 g/𝐦𝟐 0.026 g/𝐦𝟐 0.052 g/𝐦𝟐 0.058 g/𝐦𝟐

100 ℃

36 mL/min

200 psi 0.13 g/𝐦𝟐 0.0029 g/𝐦𝟐 0.046 g/𝐦𝟐 0.059 g/𝐦𝟐 0.054 g/𝐦𝟐

Ambient (minimum)

36 mL/min

100 psi 0.081 g/𝐦𝟐 0.022 g/𝐦𝟐 0.037 g/𝐦𝟐 0.059 g/𝐦𝟐 0.049 g/𝐦𝟐

Ambient (minimum)

36 mL/min

100 psi 0.080 g/𝐦𝟐 0.0043 g/𝐦𝟐 0.038 g/𝐦𝟐 0.22 g/𝐦𝟐 0.054 g/𝐦𝟐

Ambient (minimum) (minimum)

50 mL/min

200 psi 0.11 g/𝐦𝟐 0.010 g/𝐦𝟐 0.025 g/𝐦𝟐 0.069 g/𝐦𝟐 0.073 g/𝐦𝟐

100 ℃

36 mL/min

74
5.3 Economics

Figure 53: Natural Gas Price Trends (U.S. Energy Information Administration, 2019)

Information provided on the U.S. Energy Information Administration website was

used to prepare a preliminary economic analysis. Trending natural gas data from 1985-2019

was used to estimate the cost to reform natural gas from the wellhead to industry use. While

the last published wellhead price was in December 2012, the wellhead data follows the other

two data trends from 2000 to 2019 for commercial and industrial prices. As such, it was

assumed that the latest wellhead price will be approximately $1.5/Mcf less than the industrial

price. This difference is used as the cost to reform sour natural gas to cleaner natural gas that

can be used in industry that has not been marked up like the commercial sale price.

The pricing of gas treatment was calculated by the following equation:

$ catalyst
=$ (32)
MCF Q ∗ t b ∗ conv ∗ 1000

75
Where Q is the volumetric flowrate of the treated gas in mL/min, t b is the breakthrough time

and the cost is adjusted per 1000 ft 3 . Each parameter was costed and is shown in Table 10.

The costs for clay, anthracite and graphite were $77/ton, $93.17/ton, and $1,050/tonne,

respectively (U.S. Energy Information Administration, 2018), (Focus Graphite Inc., 2015) (T.

Wang, 2019).

76
Table 10: Material cost per 1000 ft 3 treated gas (Mcf) at each variation

Coastal Sulfatreat Anthracite Clay Graphite

100 psi $0.36/Mcf $36.57/Mcf $0.08/Mcf $0.04/Mcf $3.78/Mcf

100℃

50 mL/min

100 psi $0.36/Mcf $51.01/Mcf $0.08/Mcf $0.05/Mcf $5.29/Mcf

100℃

36 mL/min

200 psi $0.35/Mcf N/A $0.09/Mcf $0.03/Mcf $2.46/Mcf

Ambient

36 mL/min

200 psi $0.31/Mcf $51.01/Mcf $0.05/Mcf $0.03/Mcf $1.89/Mcf

100℃

36 mL/min

100 psi $0.51/Mcf N/A $0.17/Mcf $0.06/Mcf $4.60/Mcf

Ambient

50 mL/min

100 psi $0.49/Mcf $51.01/Mcf $0.12/Mcf $0.05/Mcf $4.60/Mcf

Ambient

36 mL/min

77
Based on cost per treated gas, the cheapest of the five materials is clay followed

closely by anthracite. Graphite should be excluded due to the poor removal and its high price

per volume of treated gas. It is higher than all except for Sulfatreat. The Coastal material

should still be prioritized over clay and anthracite due to the high amount of gas it treated

when compared to the other two materials. Even though it is more expensive, it outperforms

these two materials to the point where it is still cheaper to operate with a smaller amount of

Coastal catalyst vs the amount of clay or anthracite to achieve the same removal. These three

options are more favorable than the trended data for reforming wellhead natural gas to

industrial grade natural gas while Sulfatreat and graphite would cost more to achieve this

effect at all parameter variations.

78
Chapter 6: Conclusions

The surface chemistry of the unreacted catalysts and surface chemistry and

compositions of the spent catalysts were explored and defined using various analytical

methods such as FTIR, PIXE, BET, and TGA. The breakthrough times of H2 S, methyl

mercaptan, and ethyl mercaptan using the Coastal catalyst, Sulfatreat, clay blend, anthracite

blend, and graphite blend were studied. It was shown that:

1. The Coastal catalyst without fillers was the most effective in removing H2 S, methyl

mercaptan, and ethyl mercaptan by graphing % breakthrough vs time and calculating

amount of sulfur removed. The maximum breakthrough times for H2 S, methyl

mercaptan, and ethyl mercaptan were 276.5 hr, 243 hr and 245 hr, respectively, and

the maximum amounts removed were 296.24 g H2 S, 26.25 g methyl mercaptan, and

26.44 g ethyl mercaptan.

2. The catalyst effectiveness for sulfur removal followed the order from highest to

lowest: Coastal > clay > anthracite > graphite > Sulfatreat.

3. The most effective operating conditions with the six variations used in terms of

longest breakthrough time and highest sulfur removal was 200 psi, 100℃, and 36

mL/min followed closely by 200 psi, ambient, and 36 mL/min and 100 psi, 100℃,

and 36 mL/min.

4. The Coastal catalyst does regenerate in the presence of oxygen, producing a highly

exothermic reaction. This opens up the possibility of multiple uses of the same

catalyst in subsequent experiments.

5. While clay and anthracite are the cheapest of the five materials, much more of these

two materials are needed to compete with the effective runtime of the Coastal

79
catalyst. These three materials are cheaper to operate when compared to the trended

amount to reform wellhead natural gas to industrial grade natural gas.

80
Chapter 7: Recommendations and Future Works

• Run additional experiments of all parameter variations to verify the initial one run

results.

• Add a manual sampling port downstream of the reactor and upstream of the GC-MS

to build in redundancy. Manual samples can be taken and run in a second GC-MS to

verify that there are not any complications with the primary GC-MS and its sampling

control valve.

• Determine full speciation of Coastal catalyst for the development of a more

comprehensive mechanism.

• Determine a full mechanism that covers all species of the Coastal catalyst and which

species are beneficial for sulfur removal.

• Scale-up reaction experiments to observe if flow regimes and mass transfer effects

make a difference in how well the catalysts work.

• Test clay, anthracite, and graphite without the Coastal blend to determine the

adsorption capacity of the fillers.

• Study possible regeneration and the effects on catalyst activity after regeneration of

the five catalysts

• Study the effects on pretreating the catalysts such as heat treating, removing volatiles

and water, to improve catalyst performance.

• Conduct a more robust economic analysis.

81
Bibliography

Abdullah, A. H., Mat, R., Somderam, S., Abd Aziz, A. S., & Mohamed, A. (2018). Hydrogen
sulfide adsorption by zinc oxide-impregnated zeolite (synthesized from Malaysian
kaolin) for biogas desulfurization. Journal of Industrial and Engineering Chemistry,
65, 334-342. doi:https://doi.org/10.1016/j.jiec.2018.05.003

American Chemical Society. (2007). Ethanethiol.

Arellano-García, L., Le Borgne, S., & Revah, S. (2018). Simultaneous treatment of dimethyl
disulfide and hydrogen sulfide in an alkaline biotrickling filter. Chemosphere, 191,
809-816. doi:https://doi.org/10.1016/j.chemosphere.2017.10.096

Awe, O. W., Zhao, Y., Nzihou, A., Minh, D. P., & Lyczko, N. (2017). A Review of Biogas
Utilisation, Purification and Upgrading Technologies. Waste and Biomass
Valorization, 8(2), 267-283. doi:10.1007/s12649-016-9826-4

Ben Jaber, M., Couvert, A., Amrane, A., Le Cloirec, P., & Dumont, E. (2017). Removal of
hydrogen sulfide in air using cellular concrete waste: Biotic and abiotic filtrations.
Chemical Engineering Journal, 319, 268-278.
doi:https://doi.org/10.1016/j.cej.2017.03.014

Ben Jaber, M., Couvert, A., Amrane, A., Rouxel, F., Le Cloirec, P., & Dumont, E. (2016).
Biofiltration of high concentration of H2S in waste air under extreme acidic
conditions. New Biotechnology, 33(1), 136-143.
doi:https://doi.org/10.1016/j.nbt.2015.09.008

Crynes, B. L. (1977). Chemical reactions as a means of separation-sulfur removal: M.


Dekker.

Cui, H., Turn, S. Q., & Reese, M. A. (2009). Removal of sulfur compounds from utility
pipelined synthetic natural gas using modified activated carbons. Catalysis Today,
139(4), 274-279. doi:https://doi.org/10.1016/j.cattod.2008.03.024

Davydov, A., Chuang, K. T., & Sanger, A. R. (1998). Mechanism of H2S Oxidation by
Ferric Oxide and Hydroxide Surfaces. The Journal of Physical Chemistry B, 102(24),
4745-4752. doi:10.1021/jp980361p

de Angelis, A. (2012). Natural gas removal of hydrogen sulphide and mercaptans. Applied
Catalysis B: Environmental, 113-114, 37-42.
doi:https://doi.org/10.1016/j.apcatb.2011.11.026

Elyassi, B., Wahedi, Y. A., Rajabbeigi, N., Kumar, P., Jeong, J. S., Zhang, X., . . . Tsapatsis,
M. (2014). A high-performance adsorbent for hydrogen sulfide removal. Microporous
and Mesoporous Materials, 190, 152-155.
doi:https://doi.org/10.1016/j.micromeso.2014.02.007

82
Fang, H.-b., Zhao, J.-t., Fang, Y.-t., Huang, J.-j., & Wang, Y. (2013). Selective oxidation of
hydrogen sulfide to sulfur over activated carbon-supported metal oxides. Fuel, 108,
143-148. doi:https://doi.org/10.1016/j.fuel.2011.05.030

Figueiredo, J. L., Pereira, M. F. R., Freitas, M. M. A., & Órfão, J. J. M. (1999). Modification
of the surface chemistry of activated carbons. Carbon, 37(9), 1379-1389.
doi:https://doi.org/10.1016/S0008-6223(98)00333-9

Focus Graphite Inc. (2015). Graphite 101.

Guo, J., Luo, Y., Lua, A. C., Chi, R.-a., Chen, Y.-l., Bao, X.-t., & Xiang, S.-x. (2007).
Adsorption of hydrogen sulphide (H2S) by activated carbons derived from oil-palm
shell. Carbon, 45(2), 330-336. doi:https://doi.org/10.1016/j.carbon.2006.09.016

Gupta, A. K., Ibrahim, S., & Al Shoaibi, A. (2016). Advances in sulfur chemistry for
treatment of acid gases. Progress in Energy and Combustion Science, 54, 65-92.
doi:https://doi.org/10.1016/j.pecs.2015.11.001

Hussain, M., Abbas, N., Fino, D., & Russo, N. (2012). Novel mesoporous silica supported
ZnO adsorbents for the desulphurization of biogas at low temperatures. Chemical
Engineering Journal, 188, 222-232. doi:https://doi.org/10.1016/j.cej.2012.02.034

Jaber, M. B., Anet, B., Amrane, A., Couriol, C., Lendormi, T., Cloirec, P. L., . . . Fillières, R.
(2014). Impact of nutrients supply and pH changes on the elimination of hydrogen
sulfide, dimethyl disulfide and ethanethiol by biofiltration. Chemical Engineering
Journal, 258, 420-426. doi:https://doi.org/10.1016/j.cej.2014.07.085

Kim, K., & Park, N. (2010). Removal of hydrogen sulfide from a steam-hydrogasifier
product gas by zinc oxide sorbent: Effect of non-steam gas components. Journal of
Industrial and Engineering Chemistry, 16(6), 967-972.
doi:https://doi.org/10.1016/j.jiec.2010.04.003

Kim, Y.-H., Tuan, V. A., Park, M.-K., & Lee, C.-H. (2014). Sulfur removal from municipal
gas using magnesium oxides and a magnesium oxide/silicon dioxide composite.
Microporous and Mesoporous Materials, 197, 299-307.
doi:https://doi.org/10.1016/j.micromeso.2014.06.026

Li, L., Sun, T. H., Shu, C. H., & Zhang, H. B. (2016). Low temperature H2S removal with 3-
D structural mesoporous molecular sieves supported ZnO from gas stream. Journal of
Hazardous Materials, 311, 142-150.
doi:https://doi.org/10.1016/j.jhazmat.2016.01.033

Louhichi, S., Ghorbel, A., Chekir, H., Trabelsi, N., & Khemakhem, S. (2016). Properties of
modified crude clay by iron and copper nanoparticles as potential hydrogen sulfide
adsorption. Applied Clay Science, 127-128, 123-128.
doi:https://doi.org/10.1016/j.clay.2016.04.007

83
Micoli, L., Bagnasco, G., & Turco, M. (2014). H2S removal from biogas for fuelling
MCFCs: New adsorbing materials. International Journal of Hydrogen Energy, 39(4),
1783-1787. doi:https://doi.org/10.1016/j.ijhydene.2013.10.126

Ning, P., Liu, S., Wang, C., Li, K., Sun, X., Tang, L., & Liu, G. (2018). Adsorption-
oxidation of hydrogen sulfide on Fe/walnut-shell activated carbon surface modified
by NH3-plasma. Journal of Environmental Sciences, 64, 216-226.
doi:https://doi.org/10.1016/j.jes.2017.06.017

Occupational Safety and Health Administration. (2006). Hydrogen Sulfide.

Optimized Gas Treating, I. (2018). Claus Converters, Catalyst Deactivation, and Approach to
Modeling. Retrieved from
https://www.protreat.com/files/publications/228/Contactor%20Vol_12%20No_5%20(
Sulphur%20Converter%20Kinetics).pdf

S. Sovine. (2012). Heterogeneous Catalysis of Glycerol to Propylene Glyco over Copper


Chromite.

Tabase, R. K., Liu, D., & Feilberg, A. (2013). Chemisorption of hydrogen sulphide and
methanethiol by light expanded clay aggregates (Leca). Chemosphere, 93(7), 1345-
1351. doi:https://doi.org/10.1016/j.chemosphere.2013.07.068

Tian, K., Wang, X.-X., Yu, Z.-Y., Li, H.-Y., & Guo, X. (2017). Hierarchical and Hollow
Fe2O3 Nanoboxes Derived from Metal–Organic Frameworks with Excellent
Sensitivity to H2S. ACS Applied Materials & Interfaces, 9(35), 29669-29676.
doi:10.1021/acsami.7b07069

U.S. Energy Information Administration. (2018). Coal Prices and Outlook.

U.S. Energy Information Administration. (2019). Natural Gas.

U.S. National Library of Medicine. (2005). Ethanethiol.

Vikrant, K., Kailasa, S. K., Tsang, D. C. W., Lee, S. S., Kumar, P., Giri, B. S., . . . Kim, K.-
H. (2018). Biofiltration of hydrogen sulfide: Trends and challenges. Journal of
Cleaner Production, 187, 131-147. doi:https://doi.org/10.1016/j.jclepro.2018.03.188

W. Agbroko, O., Karishma, P., & Benson, T. (2017). A Comprehensive Review of H 2 S


Scavenger Technologies from Oil and Gas Streams.

Wang, J., Guo, J., Parnas, R., & Liang, B. (2015). Calcium-based regenerable sorbents for
high temperature H2S removal. Fuel, 154, 17-23.
doi:https://doi.org/10.1016/j.fuel.2015.02.105

Wang, T. (2019). Average bentonite price in the U.S. from 2007 to 2018 (in U.S. dollars per
ton).

84
Wiȩckowska, J. (1995). Catalytic and adsorptive desulphurization of gases. Catalysis Today,
24(4), 405-465. doi:https://doi.org/10.1016/0920-5861(95)00021-7

Xiao, Y., Wang, S., Wu, D., & Yuan, Q. (2008). Catalytic oxidation of hydrogen sulfide over
unmodified and impregnated activated carbon. Separation and Purification
Technology, 59(3), 326-332. doi:https://doi.org/10.1016/j.seppur.2007.07.042

Zhao, S., Yi, H., Tang, X., Gao, F., Zhang, B., Wang, Z., & Zuo, Y. (2015). Methyl
mercaptan removal from gas streams using metal-modified activated carbon. Journal
of Cleaner Production, 87, 856-861. doi:https://doi.org/10.1016/j.jclepro.2014.10.001

Zhu, R., Li, S., Wu, Z., & Dumont, É. (2017). Performance evaluation of a slow-release
packing material-embedded functional microorganisms for biofiltration.
Environmental Technology, 38(8), 945-955. doi:10.1080/09593330.2016.1214624

85
Appendix A

Raw Data Calculations

Due to the information size and number of calculations, the data tables are enclosed

on a disc attached with this work.

86
Conerly, Jerry F., III. Bachelor of Science, Mississippi State University, Spring 2015; Master
of Science, University of Louisiana at Lafayette, Fall 2019
Major: Engineering, Chemical Engineering concentration
Title of Thesis: Removal of Sulfur Contaminants in Natural Gas Streams
Thesis Director: Dr. Rafael Hernandez
Pages in Thesis: 97; Words in Abstract: 214

Abstract

Sulfur contamination and pollution is a growing concern in numerous industries. The

primary pollutant of concern is hydrogen sulfide (H2 S) due to toxicity, corrosion, and

environmental regulations. A proprietary iron-based catalyst was investigated and compared

to a common industrial catalyst for sulfur removal, Sulfatreat, particularly for H2 S, methyl

and ethyl mercaptans. The effects of dispersing the proprietary catalyst in a 1:9 ratio among

three different fillers: anthracite, graphite, and clay, were studied along with variations on

temperature, pressure, and flow rate. The catalyst was characterized using Brunauer-Emmett-

Teller (BET) analysis, proton induced X-ray emission (PIXE) analysis, Fourier transform

infrared spectroscopy, and thermogravimetric analysis (TGA). The reaction results were

observed and generated using gas chromatography-mass spectrometry (GC-MS) and showed

that the optimal operating parameters in terms of breakthrough time were at 200 psi, 100℃,

and 36 mL/min for all catalysts and blends used. When compared to the industrial standard

Sulfatreat, all three blends with the proprietary catalyst showed higher performance. Of the

three blend materials, clay showed the most removal effectiveness of H2 S, methyl mercaptan,

and ethyl mercaptan. Anthracite was the next most effective followed by graphite.

Economically, clay and anthracite are both desirable to use as a filler material to offset the

cost of exchanging Sulfatreat for a 1:9 ratio proprietary catalyst dispersed in filler.

87
Biographical Sketch

Jerry Conerly III was born in McComb, Mississippi, on March 25, 1992, and grew up

in Tylertown, Mississippi. He graduated from Mississippi State University in 2015 with a

Bachelor of Science in Chemical Engineering. He then started to pursue his Master of

Science in Engineering, Chemical Engineering option at the University of Louisiana at

Lafayette in Fall 2015. He completed his M.S. degree at the University of Louisiana at

Lafayette in Fall 2019.

88
ProQuest Number: 22622373

INFORMATION TO ALL USERS


The quality and completeness of this reproduction is dependent on the quality
and completeness of the copy made available to ProQuest.

Distributed by ProQuest LLC ( 2021 ).


Copyright of the Dissertation is held by the Author unless otherwise noted.

This work may be used in accordance with the terms of the Creative Commons license
or other rights statement, as indicated in the copyright statement or in the metadata
associated with this work. Unless otherwise specified in the copyright statement
or the metadata, all rights are reserved by the copyright holder.

This work is protected against unauthorized copying under Title 17,


United States Code and other applicable copyright laws.

Microform Edition where available © ProQuest LLC. No reproduction or digitization


of the Microform Edition is authorized without permission of ProQuest LLC.

ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346 USA

You might also like