INVESTIGATION OF ZEOLITE SYSTEMS: FOCUS ON FENTON CHEMISTRY OXIDATIVE STRESS FROM ASBESTOS LIKE MINERALS AND ZEOLITE-BASED DISSOLVED OXYGEN SENSING
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
By
Toni Ann Ruda
*****
The Ohio State University 2007
Dissertation Committee: Approved by Prof. Prabir K. Dutta, Advisor
Prof. Susan Olesik ______Prof. W. James Waldman Advisor Graduate Program in Chemistry Prof. Bernard LaLonde
ABSTRACT
The research in this dissertation has focused on studying zeolite systems in two
areas. The first goal of the research was to examine particle properties leading to
hydroxyl radical related toxicity. The second goal was to create an efficient optical
oxygen sensor for intracellular monitoring of oxygen. Asbestos is known to cause a variety of respiratory health problems; however, the exact mechanism leading to these
problems is unknown. Typically, asbestos participates in Fenton chemistry producing
hydroxyl radicals which lead to toxicity in vivo. During these times of oxidative stress in
vivo, the intracellular oxygen is under flux. Being able to monitor the intracellular oxygen concentration can provide critical information related to the condition of the cell.
Initially, studies were performed focusing on Fenton chemistry and oxidative
stress from asbestos like minerals. Researchers are striving to elucidate an exact
mechanism of asbestos toxicity because it has led and is still leading to multiple
respiratory health problems. The physicochemical characteristics of asbestos
contributing to respiratory health problems are not fully defined. The goal of this
research was to correlate particle toxicity with physicochemical characteristics to help
eventually elucidate the mechanism of asbestos toxicity. After the asbestos is inhaled,
the particles interact with lung lining fluid, which contains antioxidants. The antioxidants
ii have the ability to reduce ferric iron on the asbestos particles. The particles are then
phagocytosed by macrophages, and subsequently an oxidative burst is induced in an
effort to remove the inhaled particles. During this oxidative burst, hydrogen peroxide is
produced inducing Fenton chemistry with the ferrous iron. Mimicking the oxidative
burst process with minerals having properties similar to that of asbestos, but different
toxicities, can provide great insight as to the critical physicochemical characteristics
related to toxicity. Mordenite, a benign aluminosilicate, and erionite, a highly
carcinogenic aluminosilicate, were iron ion-exchanged, exposed to antioxidants, and then exposed to hydrogen peroxide to induce oxidative burst. Monitoring the production of hydroxyl radicals determined if differences between the physicochemical properties of the mineral aluminosilicates resulted in differing Fenton activity. The conclusion of this
study was that the coordination of the iron on the mineral surface plays an important role
in Fenton activity.
During the oxidative burst, a flux of oxygen within the cell occurs. Being able to
monitor the hydroxyl radical production, along with the oxygen consumption, would be useful in understanding the full mechanism of what happens during particle inhalation.
Furthermore, efficient monitoring of intracellular oxygen concentration is an important area of research that can provide insight as to cellular function and status. This leads into
the second and main component of this research, creation of efficient optical oxygen with a novel matrix to prevent probe leaching and provide linear Stern-Volmer plots. The probe utilized is tris(2,2’-bipyridyl) ruthenium(II), and the matrix is siliceous zeolite-Y.
The goal of this research was to improve the loading level of a sensor previously
iii established in our research group then demonstrate the sensor’s abilities in the following
areas: (1) monitoring dissolved oxygen in solution via a glucose oxidase assay as an in
situ measurement via emission quenching, (2) monitoring dissolved oxygen in
macrophage cells during an oxidative burst via confocal microscopy, and (3)
demonstrating the ability to immobilize the ruthenium loaded zeolite on the end of a fiber
optic and discussing future optimization to improve the sensing parameters. The
synthetic scheme utilized to load the tris(2,2’-bipyridyl) ruthenium(II) inside of the
siliceous zeolite-Y supercages was altered by changing the ruthenium precursor and some
of the experimental conditions to improve the loading level. The results showed an improved loading level of the ruthenium complex in the zeolite, which did not leach and
gave linear Stern-Volmer plots during oxygen quenching experiments. In the realms of
intracellular monitoring of dissolved oxygen as well as fiber optic sensing, parameters
have been established demonstrating the working ability of the ruthenium loaded zeolite
providing a basis for future optimization.
iv
I would like to dedicate this work to the three most important people in my life…
My parents, Michael and Tanya Ruda… for always providing support, encouragement, and unconditional love.
My soulmate, Brendan Eberenz… for believing in my ability and loving me completely.
v
ACKNOWLEDGMENTS
I would like to thank Dr. Prabir K. Dutta, my advisor, for all of his support throughout my studies. Dr. Dutta is a true inspiration, always ready to bring ideas to the table, ask challenging questions, and provide support for his students. His blunt honesty, although sometimes difficult to take, has made me stronger. I especially appreciate that he is an excellent listener, whether I am talking about my ideas related to research, lack of ideas related to research, or ideas related to ways that I believe would make the lab run more efficiently. Thanks for all your guidance Dr. Dutta.
My collaborator, Dr. James Waldman, also deserves a gigantic thanks. His help
was integral to several parts of my research. Dr. Waldman’s biological knowledge
amazes me. I appreciate that he answered my most basic biological questions without
making me feel completely ignorant. Also, a huge help in the biological realm were
Amber Nagy, a graduate student under Dr. Waldman, and Elizabeth Wheeler, who ran
the confocal fluorescence microscope. I truly appreciate the time and assistance
everyone provided. Thank you for all the biological knowledge.
Team Dutta has played an important role in my research progress, and also in my
daily life. I would like to thank current team members: Dr. Supriya Sabbani; Dr. Rajesh
Cheruvallil for trying to bestow his infinite wisdom on me, even when he has to repeat
vi something multiple times; Bill Schumacher for his support at IGERT functions; Brian
Peebles for always telling me jokes even though he often has to explain them to me and
also for being a good listener; Haoyu Zhang for his extreme kindness and helpfulness
both in and out of the lab; Dedun Adeyemo for taking over my job (good luck) and
providing me an outlet for wedding dress conversation; Kevin Cassidy for updating me
on the crazy news happenings in the world; Julia Rabe for offering support while I was
writing. I would also like to thank Team Dutta members who left within the past year or
so: Dr. Yanghee Kim for sharing her huge amount of wisdom with me, as well as
enjoying conversation about important topics such as fashion and what stores are having
the best sales; John Spirig for interesting jokes and news stories, always taking care to be
tasteful when I’m around. Finally, I would like to thank past Team Dutta members that I
have had the pleasure of working with: Dr. Ramasamy Ramamoorthy; Dr. Marla
DeLucia; Dr. Joe Trimboli; Dr. John Doolittle; Dr. Bob Kristovich; Dr. Nick Szabo; Dr.
Jiun-Chan Yang; Dr. Kefa Onchoke; and Dr. Hyunjung Lee; as well as all the Team
Dutta members that came before me (especially those that left behind well written
notebooks). Thank you Team Dutta.
My final thanks go to my mom, dad, and soon-to-be husband. I will do my best to
put into words how much my parents and Brendan mean to me, although it will never match the intensity of the feelings in my heart.
Mom and dad, you have truly played a profound role in my life; encouraging me
to enroll in graduate school and providing the support I needed to finish graduate school.
Dad, thank you for constantly emphasizing how important an education is by reminding
vii me how little shopping I would be able to do without a good job. That has definitely
been a motivating factor for me to get out of school and have more time to go shopping!
Mom, you have supported me on a daily basis through phone conversations, helping me to rationalize how to balance this time in my life so that I get my work done efficiently
but don’t miss out on the fun life has to offer. I appreciate you always reminding me that
my hard work will pay off in the future, and that just having my degree will open more
doors than I can imagine. Knowing that you both are ALWAYS on my side, supporting
me, and having the willingness to do absolutely anything for me is amazing. You both
provided an infinite amount of love and encouragement, especially when I was stressed
and feeling down. I can’t thank you both enough for all you have done for me. The love
and support you both provided to me was so necessary while in graduate school and
while I was trying to complete this work. I am so incredibly lucky to truly have the best
parents ever. I love you so much, and thank you for everything.
Brendan, my soulmate and soon-to-be husband, thank you for dealing with me
during the times of extreme stress and crankiness I’ve had while finishing my research.
At times when I wouldn’t even want to deal with myself you still embrace me with
encouragement and love. You have been incredible at picking up so much of the slack I
left while working on my dissertation, both in taking care of the house chores and the
wedding plans. I can’t express how amazing it is to have met you, my soulmate, while
being in graduate school. At times when research looked bleak, I knew I could look
forward to seeing you. You always lift me up, reminding me of how lucky I am and how
viii incredible my life has become. Being able to share the great accomplishment of
obtaining my doctorate with you is truly special to me, and I can’t wait to share in that
same excitement with you when you obtain your doctorate. I also am so looking forward
to marrying you on November 3rd, 2007; I can’t wait to share everything life has to offer with you. I love you more than words can say!
.
ix
VITA
April 8th, 1979...... Born – Uniontown, Pennsylvania
1997...... Uniontown Area High School Uniontown, Pennsylvania
2000...... Colgate-Palmolive Corporation Internship Piscataway, New Jersey
2001...... B.S., Chemistry (A.C.S. certified) Washington and Jefferson College Washington, Pennsylvania
2001 – 2002...... Teaching Asst. The Ohio State University
2002 – 2006...... IGERT Fellow The Ohio State University
2006-2007...... Teaching Asst. The Ohio State University
PUBLICATIONS
Research Publications
1. Ruda, Toni A. Dutta, Prabir K. “Fenton Chemistry of Fe-III Exhanged Zeolitic Minerals Treated with Antioxidants.” Environ. Sci. Technol. 2005, 39, 6147-6152.
FIELDS OF STUDY
Major Field: Chemistry
x
TABLE OF CONTENTS
P a g e
Abstract.…………………………………………………………………...………………ii
Dedication.………………………………………………………………….……………..v
Acknowledgments .……………………………………………………………..………..vi
Vita .…………………………………………………………………..…………………..x
List of Tables.………………………………………………………..………….…..…..xvi
List of Figures .………..………………………………………………………………..xvii
Chapters:
1. Introduction.……………………………………………………………………………1
1.1 Mineral Particles: Regulations, Morbidity, and Mortality…………………….....1
1.2 Mineral Toxicity………………………………………………………………… 2
1.2.1 Size and Dimension…………………………………………..………….3 1.2.2 Surface Charge…………………………………………………………...4 1.2.3 Biopersistence and Biodurability………………………………………...5 1.2.4 Chemical Composition……………………………………………….…..5
1.3 Lung Physiology / Particle Inhalation and Lines of Defense……………..……..6
1.3.1 Particle Inhalation: Basics………………….………………….………..6 1.3.2 Lung Lining Fluid Antioxidants…………………………………………8
1.4 Toxic Particles of Interest…………………………………….………….………9
xi 1.4.1 Asbestos………………………………………………...………………..9 1.4.1.1 Asbestos: Definition and Background…………….………………10 1.4.1.2 Asbestos Mechanism of Toxicity…………………….…………….11
1.4.2 Zeolites…………………………………………..…………..………….12 1.4.2.1 Erionite……………………………………………..……..………..13 1.4.2.2 Mordenite……………………………………………..……………15
1.5 The Fenton Reaction and The Haber Weiss Cycle…….……………………….16
1.5.1 Detection of Hydroxyl Radical Production…………….……………….18
1.6 Oxidative Burst and Oxygen Concentration……………………………………19
1.6.1 Oxidative Burst…………………………………………………………19 1.6.2 Detection of Oxidative Burst……………………………...……………20 1.6.3 Importance of Intracellular Oxygen Concentration……………….……21 1.6.4 Extracellular Oxygen versus Intracellular Oxygen Concentration…..…21
1.7 Oxygen Sensors…………………………………………………………….…..22
1.7.1 Typical/Conventional Oxygen Sensors……………………..………….23 1.7.2 Typical Probes……………………………...……………………….….24 1.7.3 Typical Matrices…………………………………………………..……25 1.7.3.1 Novel Zeolite Matrix………………………………………………26
1.8 Tris(2,2’-bipyridyl) Ruthenium Photochemistry and Quenching…………..…..29
1.8.1 Oxygen Sensing/Quenching………………………………………….…31
1.9 Intracellular Optical Direct Dissolved Oxygen Monitoring………..….……….34
1.9.1 Fiber Optic……………………………………………...………………35 1.9.2 Oxygen Sensitive Dye…………………………………….….…………35 1.9.3 Phospholipid Spheres………………………………………..………….36 1.9.4 PEBBLE……………………………………………..………………….37
1.10 Goals of This Thesis…………………………………………………………..39
1.11 References………………………………………….………………………….54
2. Fenton Chemistry: Treatment of FeIII-exchanged Zeolite Minerals Treated with Antioxidants………………………………………………………………………….61
xii 2.1 Introduction…………………………………………………………...……….61
2.2 Experimental Section………………………………..…………….……………63 2.2.1 Iron Incorporation into Zeolites…………….…………………………..63 2.2.2 Zeolite Crystallinity……………………………….……………………64 2.2.3 Surface Iron Measurements………………………………………...…..64 2.2.4 Exposure of Zeolites to Antioxidants…………………………………..65 2.2.5 Detection of Hydroxyl Radicals………………………….……...……..65 2.2.6 XPS Studies of Surface Iron……………………………....……………66
2.3 Results……………………………………………………….……...………….67
2.3.1 Ion-exchange process………………………………..……..………….67 2.3.2 Surface Iron………………………………….…….…………………..68 2.3.3 Fenton Reactivity………………….……………..…………………….70
2.4 Discussion……………………………………………….………………..……73
2.5 Conclusion…………………………………………….……………………….80
2.6 References………………………………………………………..……………91
3. Entrapment of Ionic Species in Siliceous Zeolite: Ship-in-a-Bottle Synthesis of Tris(2,2’-bipyridyl) ruthenium(II) and its Emission Quenching by O2 in Different Environments…………………………….…………………………………..………94
3.1 Introduction………………………...……...……………………….……...…..94
3.2 Experimental (methods, techniques and materials studied)………...….….…..98
3.2.1 Synthesis of Siliceous Zeolite-Y……………………………..….……..98 3.2.2 Synthesis of Ru(bpy)Cl2(CO)2…………………………………....……99 3.2.3 Synthesis of Ru(bpy)3-Siliceous Zeolite-Y……………...………..…..100 3.2.4 PDMS - Ru(bpy)3-Siliceous Zeolite-Y Membrane…………..…….…101 3.2.5 Leaching…………………………………………………...…….....…102 3.2.6 Emission Quenching………………………………………..………...102 3.2.7 Monitoring Emission in Live Cells……………………...………..…..104 3.2.7.1 Isolation and differentiation of human monocyte-derived macrophages (MDM)……………………………………..….....104 3.2.7.2. Treatment of MDM with ruthenium sensor particles….………..105 3.2.7.3 Luminol Assay: Oxidative Burst Optimization………………..105 3.2.7.4 Confocal Fluorescence Microscopy………...………..…..……..106
3.3 Results………………………………………………………………………..107
xiii
3.3.1 Synthesis…………….………………………………………………….107 3.3.1.1 Synthesis of Siliceous Zeolite-Y………………………………..107 3.3.1.2 Synthesis of Ru(bpy)Cl2(CO)2………………………………….109 3.3.1.3 Synthesis of Ru(bpy)3-Siliceous Zeolite-Y………….….……....112 3.3.1.4 Synthesis of PDMS - Ru(bpy)3-Siliceous Zeolite-Y Film..….…113
3.3.2 Sample Optimization………………………………………………..…114 3.3.2.1 PDMS Film Stability - Leaching Experiment…………….....….114 +2 3.3.2.2 Emission Quenching of High versus Low Sample Ru(bpy)3 Loading…………..……………………………………………..114
3.3.3 Emission Quenching…………………….……....…………………....115 3.3.3.1 Calibration…………………………………….…….……...…..116 3.3.3.2 Glucose Oxidase Assay………………………….….………….116 3.3.3.3 Intracellular Oxygen Monitoring…….……………..……….….117
3.4 Discussion………………………………..……………………………………120
3.4.1 Synthesis of Siliceous Zeolite-Y….…………………….………...…..120 3.4.1.1 Dealumination Treatments……………………………..……….120 3.4.1.2 Analysis of Zeolite Dealumination………..………...... …..122
+2 3.4.2 Synthesis of Ionic Ru(bpy)3 Within Hydrophobic Siliceous Zeolite-Y……………………………………………….……….……124 3.4.3 PDMS Film Quality: 3.4.4 Leaching……..…………………..……..127 3.4.4 Oxygen Quenching……………………….…………….………….…129 3.4.5 Oxygen Quenching: Stern-Volmer Analysis….…………...….……..130 3.4.6 Live Cell Experiments………………………….…………….……....133
3.5 Conclusions…………………………………………………………………..136
3.6 References………………………………………………………………..…..162
4. Fiber Optic Dissolved Oxygen Sensor: Optimization of Fiber Configuration Utilizing Tris(2,2’-bipyridyl) ruthenium(II) Inside of Siliceous Zeolite-Y as Oxygen Quenching Probe……………………………………………….…...... …..166
4.1 Introduction………………………………………….………………………..166
4.2 Experimental….…………………………………………………....…………170
4.2.1 Materials…………………………………………………………….…170 4.2.2 Fiber Optic Preparation………….…………………………….………171
xiv 4.2.3 Sensor Preparation……………………………..………..…………….172 4.2.3.1 Zeolite Sensor Only Inside of Fiber……………...………..……172 4.2.3.2 PDMS and Sensor Inside of Fiber….…………..………………172 4.2.3.3 PDMS and Sensor on Outside of Fiber………….………….…..173
4.2.4 Physical Characterization of the Sensors………………..……………173 4.2.5 Experimental Set-up…………………..………………………...…….173
4.3 Results & Discussion………………………………………………….……..174
4.3.1.1 Fiber Etching Results…………………………..……………....174 4.3.1.2 Fiber Etching Process………………………….………..……..175 4.3.1.3 Etched Fiber Optic Oxygen Sensors……………….….……….178
4.3.2 Parameters of Fiber Optic Oxygen Sensors………..………..….…….179 4.3.2.1 Stern-Volmer Linearity…...……………………………..……...181 4.3.2.2 Sensitivity…………………………………………………….…184 4.3.2.3 Response Time……...………………………………………..…187
4.4 Conclusions…………………………………………………………………..190
4.5 References…………………………………………………………………….204
Bibliography……………………………………………………………………………207
xv
LIST OF TABLES
Table Page
2.1 Comparison of hydroxyl radical production in the presence of different antioxidants…………………………………………………………..….81
3.1 Representation of the quenching seen during the gas saturated water quenching experiments……………………………………………..….138
3.2 Summary of the quenching data from the glucose oxidase assay and a sample calculation using Henry’s law…………………………………139
3.3 Table comparing the oxygen concentration calculated stiochiometrically to the concentration of oxygen determined by using the gas saturated water Stern-Volmer plot equation……………………………………. 140
4.1 Table showing properties of fiber optic dissolved oxygen sensors in order to compare literature, commercial based sensors, and the sensors configured in this research……………………………………………..191
xvi
LIST OF FIGURES
Figure Page
1.1 Correlation between death and inhalable particulate matter...... 40
1.2 A depiction of the cascade of event that occur after a mineral dust is inhaled...... 41
1.3 Flow chart to represent critical components of particle inhalation……...42
1.4 Relationship between particle diameter and regional deposition in the lung……………………………………………………………………...43
1.5 Diagram of the respiratory tract to show the path particles encounter upon inhalation………………………………………………………….…….44
1.6 Detailed view of the alveolar area in the lung…………………………..45
1.7 Hydrophilic, non-enzymatic antioxidants during their function…….…..46
1.8 Chart comparing number of asbestos related deaths to other causes of mortality……………………………………………………………...….47
1.9 Zeolite structures which are formed from the sodalite, or β, cage subunits………………………………………………………………….48
1.10 Mechanism by which luminol is able to detect reactive oxygen species………………………………………………………………….. 49
1.11 Diagram of a Clark cell………………………………………………….50
1.12 Probes that are commonly used in optical oxygen sensing……………...51
1.13 Depiction of intrazeolite water………………………………………..…52
1.14 Energy level diagram representing tris(2,2’-bipyridyl) ruthenium(II) decay pathways………………………………………………………….53
xvii
2.1 Absorption spectrum illustrated Fenton chemistry………………….…..82
2.2 Surface iron calibration curve made with methanesulfinic acid sodium salt………………………………………………………………….……83
2.3 X-ray powder diffraction patterns of erionite and mordenite………..….84
2.4 X-ray powder diffraction pattern of erionite after iron ion-exchange……………………………………………………...……..85
2.5 NMR spectra of Dab-4Br…………………………………………..……86
2.6 Absorption spectra for surface iron analysis…………………………….87
2.7 X-ray photoelectron spectra of Fe(III) exchanged erionite and mordenite………………………………………………………………..88
2.8 Total hydroxyl radical production and hydroxyl radical production normalized to surface iron in the presence of glutathione (GSH), ascorbic acid (AA) or hydrogen peroxide only (Blank)……………………….…89
2.9 The aluminosilicate ring structures in mordenite and erionite……..…...90
3.1 Reactor used during delaumination of zeolite-Y………………………141
3.2 Sample set-up for fluorescence sensing experiments utilizing a PDMS film in a cuvette………………………………………………………..142
3.3 SSNMR 29Si spectra of synthesized siliceous zeolite-Y and commercial siliceous zeolite-Y from the Tosoh Corporation……………….……....143
3.4 XRD of siliceous zeolite-Y synthesized in the laboratory.…………….144
3.5 SEM of siliceous zeolite-Y, from Tosoh corporation, after calcination at 500oC…………………………………………………………………..145
3.6 An IR spectra of Ru(bpy)Cl2(CO)2 synthesized in the laboratory……..146
3.7 An NMR spectrum of one of the RuCl2(CO)2(bpy) samples synthesized in the laboratory…………………………………………………………..147
3.8 Diffuse-reflectance UV absorption spectra of tris(2,2’-bipyridyl) ruthenium (II) loaded in siliceous zeolite-Y………………………..….148
xviii 3.9 Overall reaction synthetic scheme for loading tris(2,2’-bipyridyl) ruthenium II inside of siliceous zeolite-Y……………………..……….149
3.10 Pictures of tris(2,2’-bipyridyl) ruthenium(II) siliceous zeolite-Y powders and PDMS films synthesized at different loading levels……….……...150
3.11 Pictures of the PDMS film utilized before and after the leaching experiment was performed……………………………………………..151
3.12 Plot showing trend of the leaching seen over the ten week period ………………………………………………………….………………152
3.13 Luminescence intensity quenching of tris(2,2’-bipyridyl) ruthenium(II) siliceous zeolite-Y in water………………………………………….…153
3.14 Absorption and emission spectra showing the peak shift correlating to different loading levels of the ruthenium complex in the zeolite……...154
3.15 Stern-Volmer plot with gas saturated water solutions………………....155
3.16 Confocal image of a macrophage after 24 hour exposure to the ruthenium- loaded zeolite…………………………………………………………..156
3.17 Luminol assay performed to show intensity of reactive oxygen species produced by phorbol 12-myristate 13-acetate and zymosan…………...157
3.18 Visual of the confocal images after the elapsed time with no induction of oxidative burst. …………………………………………………..….…158
3.19 Visual of the confocal images after the elapsed time after induction of oxidative burst……………………………………………………...…..159
3.20 Analysis of confocal experiment: plot of response to oxidative burst induction for first dulplicate run………………………………...……..160
3.21 Analysis of confocal experiment: plot of response to oxidative burst induction for first dulplicate run……………………………………….161
4.1 A SEM of the fiber optic used in this research…………………..……..192
4.2 Picture of sample holder and set-up used for fiber optic quenching experiments……………………………………………………………..193
4.3 SEM of fiber after HF etching for about three hours…………………...194
xix 4.4 Pictures of the fiber optic after HF etching overnight, then sonicating in DI water………………………………………….…………………………195
4.5 Picture of an etched fiber optic with the ruthenium-loaded zeolite placed inside of the etched portion, then a very thin PDMS layer placed on the outside of the fiber to prevent zeolite loss from the fiber optic core...…196
4.6 Pictures of an etched fiber optic filled with the ruthenium-loaded zeolited PDMS in the space where the core was etched.………………….…..…197
4.7 Pictures of a fiber optic coated with the ruthenium-loaded zeolite embedded in PDMS on the outside of the fiber optic…………………..198
4.8 Stern-Volmer plot for an etched fiber optic with the ruthenium-loaded zeolite placed inside of the etched portion with thin PDMS overlay layer ………………………………………………………………………..…199
4.9 Stern-Volmer plot for an etched fiber optic filled with the ruthenium- loaded zeolited PDMS in the space where the core was etched………..200
4.10 Stern-Volmer plot for fiber optic coated with the ruthenium-loaded zeolite embedded in PDMS on the outside of the fiber optic…………………..201
4.11 Response curve for Fiber 1 (dye loaded zeolite inside of etched fiber core with thin PDMS layer on outside) between 100% oxygen and 100% nitrogen………………………………………...…….………………....202
4.12 Response curve of a 1x10-5 M solution of tris(2,2’-bipyridyl) ruthenium(II) chloride in deionized water between 100% oxygen and 100% nitrogen.…………………………………………………………….…..203
xx
CHAPTER 1
INTRODUCTION
1.1 Mineral Particles: Regulations, Morbidity, and Mortality
In 1997, the Environmental Protection Agency (EPA) proposed that particulate
matter with a diameter of 2.5 μm and less (PM2.5) needs to be regulated (Green et al.,
2002). Inhalation of particulate matter has been correlated to health problems such as
those seen during the London Fog disaster in 1952. The disaster was associated with
high concentrations of airborne particulate matter due to an atmospheric temperature
inversion that caused a dormant air mass to settle over the city trapping the high
concentrations of particulate matter. The increased particulate matter caused an increase
in deaths of somewhere between 5,000 to 12,000 over three days. Figure 1.1 shows that
the number of deaths correlated to the increase in air particulate matter. Although the population is not typically exposed to such high levels of particulate matter, the low levels that the population is exposed to daily may still have a deleterious affect that validates investigation. The problem with placing regulations on all PM2.5 is cost,
estimated to be $8-150 billion annually (Green et al., 2002). In order to combat the costs,
1 it was decided that focus should be placed on the toxic PM2.5 since not all PM2.5 is toxic.
Therefore, is being placed on determining which factors are related to toxicity/bioactivity of particles after inhalation.
Inhalation of mineral particles often leads to a wide range of health effects from relatively benign to extremely toxic. Studying the health effects of mineral dusts is important because humans are exposed to mineral dusts from both anthropogenic and natural resources. Therefore, not only is exposure to mineral dust occupational but it can also be found in rural environments (Guthrie, 1992). Since the actual mechanism of mineral toxicity is still largely unknown, it will be easier to uncover if differences in the toxicity and carcinogenicity of various minerals can be related to fundamental differences in crystal structure/chemistry (Guthrie, 1992).
1.2 Mineral Toxicity
Toxicity of particles, which lead to a diseased state of the lung, is determined by a variety of factors related to the process a particle goes through after inhalation. The particles are first deposited in the respiratory tract via impact, sedimentation, or interception. At this point the shape, dimension, and dosage have a large influence on the affinity of the particles for their target cells. Then the particle is cleared either via cilia or phagocytosis by an alveolar macrophage, which also depends on the size and shape of the particle. The geometry and dimensions of these minerals may govern their deposition and clearance kinetics, biological reactivity, and dissolution in the lung; chemical and
surface properties, including sorption, oxidation/reduction reactions, and charge, also play
2 important roles in biopersistence, cellular responses, and pathogenicity. Chemical
composition and surface are especially relevant in regard to the activity of the iron on or
in the particle. Finally, an oxidative burst occurs where, again, surface composition is
important and may lead to hydroxyl radical production. Hydroxyl radicals can have a
deleterious affect damaging DNA, lipids, proteins, and carbohydrates. Biopersistence
and biodurability are also important factors related to particle toxicity because they relate
to retention of the particle in the lungs. The longer particles are retained the more
deleterious affects they may create. More specific discussion of the factors leading to
particle toxicity and lung disease follow. Figure 1.2 describes a cascade of events that
may occur after a particle is inhaled. Figure 1.3 gives detail related to the different steps
of particle inhalation and the critical factors during each step (van Zijverden, 2007;
Fubini, 1997).
1.2.1 Size and Dimension
Particle size and shape is critical in assessing toxicity due to its influence on the
following: particle deposition, particle translocation, surface area, reactive sites, particle-
cell interactions, and catalysis. Particle deposition, translocation, and clearance are critical to particle toxicity and are directly affected by the size and dimension of the particle, dictating where it is able to travel in the airways. Figure 1.4 shows the relationship between particle size and particle deposition in the lung. Stanton and co- workers proposed that the fiber aspect ratio, length:width, comprised the most significant feature related to toxicity (1981). Archer then did a study stating that longer fibers did
3 show an increased toxicity due to “frustrated phagocytosis” (1979). During “frustrated phagocytosis” long fibers can not be engulfed by macrophages. Clearance, by means of phagocytosis, translocation, or dissolution, is limited by the size and shape of the particle.
Larger particles will be cleared by ciliated cells, never making it into the bronchioles or alveoli depending on their size. Very small particles that reach the alveoli will be easily phagocytosed and removed by macrophages. Finally, the aspect ratio, or length to width ratio, of particles is also found to be related to toxicity. Spherical particles have an aspect ratio of 1:1. Fibers, typically defined as any particle with an aspect ratio great than 3:1, can become difficult for the lungs to clear causing “frustrated phagocytosis” to occur during clearance especially in the case of longer fibers (Guthrie, 1997; Mossman and
Churg, 1998).
1.2.2 Surface Charge
The interaction of particles with the cells/cell membranes is one of the first steps in the process of particle toxicity. The surface charge of the particles directly relates to their interaction with cells/cell membranes. Besides surface charge, polarity, acid-base sites, hydrophobicity, and hydrogen bonding potential are also crucial factors related to particle-cell interactions. A relationship between the surface charge of asbestos and haemolytic activity was found to exist, showing that particles with more surface charge have a higher haemolytic activity (Light and Wei, 1977). Another study showed that chrysotile fibers with more surface charge was related to an increased cytotoxicity due to electron transfer from the asbestos fiber to the cell biomembrane (Valentine et al.,1983).
4 1.2.3 Biopersistence and Biodurability
Biopersistence and biodurability are both related to the period of time a particle is
retained in the lung. Although these are sometimes used interchangeably, the difference
is that biopersistence is related to particle clearance due to physiology of the lungs, and
biodurability relates to the ability of the particle to be cleared via biological fluids in an
effort to dissolve/destroy the particle. Biopersistent and biodurable particles are
potentially more fibrogenic and carcinogenic because they are not able to be efficiently
cleared. Fibers that are biodurable and biopersistent will continue to accumulate in the
lung and have been shown to lead to chronic inflammation, fibrosis, and tumors
(Hesterberg et al., 1998).
1.2.4 Chemical Composition
Surface chemistry is one of the most important factors relating to particle toxicity.
Focusing solely on factors such as particle shape and size can not explain particle toxicity
fully, since long dust fibers like talc do not lead to disease states (Guthrie, 1992). Fiber
induced toxicity is typically related to the formation of radicals, which form due to metals
on the surface of the fibers. After the particle is phagocytosed, the macrophage will
undergo an oxidative burst producing oxidative species in an attempt to destroy the
foreign matter. The oxidative burst is a defense mechanism. During the oxidative burst,
hydrogen peroxide is produced. In the presence of iron the hydrogen peroxide can decompose to form hydroxyl radicals. The reaction of the hydrogen peroxide with iron is known as the Fenton reaction, and is related to increases in mutagenicity and
5 inflammatory responses. Early evidence of this was determined when iron bound to
inhalable crocodilite asbestos was found to be a possible factor in asbestos
carcinogenicity (Hardy and Aust, 1995). Weitzman et al. then showed asbestos catalyzes
•- hydroxyl and superoxide radical (O2 ) generation via an iron dependent pathway known
as the Fenton reaction (1984). The Fenton reaction has attained much focus when looking at particle toxicity. The reaction itself will be discussed in more detail later. The radicals formed are linked to intracellular oxidative stress, as a product of the Fenton reaction can lead to DNA damage, lipid peroxidation, oxidation of nucleotide bases, and oxidative protein damage (Hardy and Aust, 1995). Although not all fibers have iron on their surface, they may be able to pick up iron in the body by competing with ferritin during iron transfer processes.
1.3 Lung Physiology / Particle Inhalation and Lines of Defense
1.3.1 Particle Inhalation: Basics
An inhaled particle will first encounter the upper respiratory tract where it is taken
in through the mouth and/or nose. The particle will then travel through the larynx, the
trachea, then into one of the two main bronchi which each lead to a lung. The respiratory
tract, from the nose to the bronchi, are covered with ciliated epithelial cells. Cilia, a line
of lung defense, are short hair-like cellular appendages with the purpose of moving
mucus towards the throat so that it is coughed or swallowed. The mucus captures larger
inhaled particles, and with the help of cilia clears them to prevent irritation. The
bronchus divides into smaller and smaller air ducts in the lung, with the narrowest ducts
6 known as bronchioles. Alveolar ducts are found at the terminal bronchioles, which open into the alveoli. The alveoli are composed of epithelial cells of two types: Type I, epithelial tissue cells and Type II, rounded, cuboidal, secretory epithelial cells. A layer of surfactant covers the epithelial cell layer, and alveolar macrophages, or “dust cells”, also reside in the alveoli, both acting as lines of defense. The macrophages phagocytize inhaled particles, undergo oxidative burst in an effort to destroy and clear the material, and begin to signal proinflammatory cytokines in order to start the process of preventing particle toxicity. The physiology of the respiratory system has major implications on particle deposition and clearance (Fubini and Arean, 1999). Figure 1.5 shows the physiology of the lung to provide of better idea of particle deposition. Figure 1.6 shows a more detailed picture of the alveoli. As a reminder, Figure 1.4 correlates particle diameter to regional deposition in the lung.
Respiratory tract lining fluids cover the respiratory tract epithelial cells from the
nasal mucosa to the alveoli. The lung lining fluid (LLF) is an interface between the
underlying respiratory tract epithelial cells and the external environment. Antioxidants in
the LLF often provide an initial defense against inhaled environmental toxins. The main
components present in the LLF are mucin, uric acid, protein (mainly albumin), ascorbic acid, and reduced glutathione (GSH). The LLF fluid is known to have much higher levels of reduced glutathione and ascorbic acid than plasma. The composition of LLF antioxidants varies throughout the different levels of the respiratory tract (Cross et al.,
1994). Debate exists in literature as to if antioxidants act as prooxidants, promoting oxidative damage (Proteggente et al., 2000; Carr and Frei, 1999; Ruda and Dutta, 2005;
7 Nappi and Vass, 1997). Results in the literature are still contradictory, and overall it
seems that in vitro models lean towards the fact that antioxidants may be acting as
prooxidants, but human and animal studies that have been performed show antioxidants
acting as antioxidants (Valko et al., 2005).
Another critical component of the LLF is transition metal ions. Transition metal ions in normal LLF may be limited because of the presence of antioxidant metal ion- binding proteins; however, oxidative stress itself can often cause metal ion release (Cross et al., 1994). The availability of these metal ions allows them to interact with the inhaled
particle, and may then be related to oxidative damage.
1.3.2 Lung Lining Fluid Antioxidants
Hydroxyl radical production, in an aqueous solution of quartz particles and
hydrogen peroxide, was found to be enhanced significantly by reduced GSH, cysteine,
AA, and selected catechols. Inhaled mineral particles interact with the LLF in transit to
the alveolar space. LLF is made up by surfactants and proteins and is rich in GSH and
AA, both of which act as antioxidant body defenses (Fenoglio et al., 2000). The
properties of the thiols and ascorbic acid, being either an antioxidant or pro-oxidant, are
complex and may be related to their intracellular concentrations (Nappi and Vass,1997).
Ascorbate and glutathione are two of the most important non-enzymatic scavengers. The mechanism for both ascorbate and glutathione are shown in Figure 1.7 (Chaudière and
Ferrari-iliou, 1999). The mechanisms show the electron being given up, which can then
reduce iron and promote the Fenton reaction.
8 Ascorbate acts as a potent water-soluble antioxidant in biological fluids by
scavenging physiologically relevant ROS and reactive nitrogen species (RNS). Evidence
for the interaction of ascorbate with radicals and oxidants has been found in extracellular
fluids such as plasma and respiratory tract lining fluid. The reduction of transition metal
ions by ascorbate could have deleterious effects via reaction with hydrogen peroxide to
produce hydroxyl radicals. The reaction in vivo has been a matter of controversy because of the unknown availability of catalytic metal ions in vivo (Carr and Frei, 1999). The
iron-mediated oxidative metabolism of GSH may impart a pro-oxidant function as seen in
the reaction from Figure 1.7 (Nappi and Vass, 1997).
1.4 Toxic Particles of Interest
1.4.1 Asbestos
Asbestos is often the focus of mineral inhalation studies because of its toxicity.
Frequently, lung problems are uncovered much later after exposure because asbestos has a latency period of twenty plus years (Guthrie, 1992). Worldwide, about 125 million
people are exposed to asbestos at their workplace. Asbestos is attributed to 90,000 deaths
per year globally due to asbestos-related lung cancer, mesothelioma, and asbestosis due
to occupational exposures. Figure 1.8 shows the relationship between asbestos related
deaths and other causes of death in the United States. Asbestos related diseases are still
problematic, even in countries that have banned asbestos use in the early 1990s, due to its
long latency period. Therefore, stopping the use of asbestos will not decrease the number
of asbestos related deaths for decades (World Health Organization, 2007).
9 1.4.1.1 Asbestos: Definition and Background
Asbestos has been used in a variety of applications because of its desirable properties such as high tensile strength, good chemical and thermal stability, and resistance to acidic conditions (EPA, 2007; Mossman and Gee, 1989). Exposure to asbestos has been shown to lead to various problems such as asbestosis, mesothelioma, and lung cancer (EPA, 2007). The effects often go unnoticed for 15-40 years. Asbestos is a crystalline mineral fiber composed of a hydrated silicate; it encompasses six different minerals: chrysotile, crocidolite, amosite, anthophyllite, tremolite, and actinolite with the first three being the most common types (Mossman and Gee, 1989; EPA, 2007).
Asbestos is broken down into two mineral groups, serpentine and amphibole which differ in their crystalline properties (EPA, 2007). Serpentines have a sheet or layered structure.
The only member of the serpentine group is chrysotile, which is the most common type of asbestos making up approximately 90%-95% of all asbestos contained in buildings in the United States (EPA, 2007). In contrast, amphiboles have a chain-like structure.
Asbestos is found in a variety of materials that have been in use and/or are still in use today such as (EPA, 2007): cement pipes, laboratory hoods/table tops, elevator brake shoes, cement wallboard, laboratory gloves, HVAC duct insulation, cement siding, fire curtains, breaching insulation, caulking/putties, adhesives, pipe insulation (corrugated air-cell, block, etc.), construction mastics (floor tile, carpet, ceiling tile, etc.), heating and electrical ducts, textured paints/coatings, roofing shingles, ceiling tiles and lay-in panels, spray-applied insulation, electrical cloth, blown-in insulation, fireproofing materials. Obviously asbestos exposure can happen readily (EP, 2007).
10 In the 1970s both the EPA and the Occupational Safety and Health
Administration (OSHA) began to regulate asbestos. The passage and enactment of regulations by OSHA helped reduce Americans asbestos exposure in the workplace
(Mossman and Gee, 1989). OSHA intertwined rules with the EPA to protect the health of workers and implement asbestos regulations. Two different laws/regulations implemented by the EPA are the Toxic Substances Control Act (TSCA) and the National
Emission Standards for Hazardous Air Pollutants for Asbestos, under Section 112 of the
Clean Air Act (NESHAP establishes work practices to minimize release of asbestos fibers during activities involving processing, handling and disposal of asbestos when a building is being demolished or renovated). These were established in the 1970s to protect the general public from asbestos exposure (EPA, 2007).
1.4.1.2 Asbestos Mechanism of Toxicity
Although regulations have been made in order to prevent people from being exposed to asbestos, it is important to elucidate the actual mechanism of asbestos toxicity because of its long latency period and the fact that people are still becoming sick from exposure. Asbestos toxicity may be related to a variety of properties, similar to the particle properties previously discussed: fiber morphology, dosage, biodurability, surface charge, and surface chemistry.
One of the main focuses when studying asbestos toxicity is Fenton chemistry, which stems from iron on the surface of the asbestos. The iron on the asbestos mineral fibers, such as crocidolite and amosite, can be bioactive and cause the production of
11 hydroxyl radicals, leading to the detrimental lung diseases often associated with asbestos
(Eborn and Aust, 1995). Martra et al. determined that the coordination of the iron to the asbestos (or inhaled particle) attributes to the iron bioactivity (2003).
To fully understand asbestos toxicity, it is necessary to understand what happens
in your body when the fibers are inhaled. Especially in the case of asbestos, where our
focus is on the surface iron and how it interacts with various compounds after inhalation.
Minerals are often used as replacements for asbestos so their health effects need to be known, and the toxicity mechanisms need to be understood (Guthrie, 1992).
1.4.2 Zeolites
The word zeolite originated from the Greek “zein” meaning “to boil” and “lithos”
meaning “a stone”. The word “zeolite” originated in the 18th century when Axel Fredrik
Cronstedt observed natural mineral stones would bounce around upon heating due to
water evaporation (Zeolyst, 2007). Since asbestos is composed of a variety of fibers, causing it to have a complex structure, it would be beneficial to use models for asbestos
to correlate surface iron to bioactivity. Studying surface iron/bioactivity relationships
with zeolites will allow us to have a better understanding of what is contributing to
asbestos toxicity and model bioactivity after a potentially toxic fiber is inhaled. The
model fibers used in our research are the zeolites known as erionite and mordenite.
Zeolites are aluminosilicates whose silicon (Si) atoms are each tetrahedrally coordinated to oxygen (O) atoms. Figure 1.9 shows the basics of zeolites. The Si atoms are occassionally replaced by aluminum (Al) atoms, giving an overall negative charge to
12 the mineral, which is balanced by counterions (Eborn and Aust, 1995). The cations are
easily ion-exchanged. Under ambient conditions, the zeolite will fill its void volume with
water giving it the following formula for a unit cell,
. Mx/n[(AlO2)x(SiO2)y] wH2O
where n = cation valence, y/x = a ratio usually in the range of 1-5, and w = number of water molecules.
Many zeolite structures exist, which vary greatly in dimension and pore size. One
of the zeolites focused on later in this research is zeolite-Y. Figure 1.9 shows how
zeolite-Y is constructed from the sodalite or β cages. Zeolite-Y consists of the sodalite
cages having a tetrahedral arrangement, forming supercages. The supercages have 13 Å
diameters, and 7.4 Å windows.
1.4.2.1 Erionite
Erionite is a naturally occurring zeolite mineral with a white, fibrous, wool-like
appearance found in volcanic and sedimentary rocks. Significant amounts of erionite are
found in Turkey. In the United States, erionite is found in Oregon, Nevada, Utah,
California, and Arizona (National Toxicology Program, 2007). Erionite was first
described by A.S. Eakle in 1898. The chemical structure of erionite is as follows:
NaK2MgCa1.5 (Al8Si28O72) 28H2O
13 No iron present in the framework of erionite. Erionite is able to attain iron through
weathering and erosion processes (Eborn and Aust, 1995). Furthermore, erionite is able
to attain iron in the body since there are free iron pools in vivo. It has been shown that
low molecular mass, chelatable, redox active iron is present in normal human pulmonary
epithelial lining fluid recovered from bronchoalveolar lavage. The presence of pro-
oxidant iron in normal epithelial LLF may contribute to the known susceptibility of the
lung to oxidative insult (Gutteridge et al., 1996). The iron associated with erionite is believed to be a large contributor to its toxicity. Erionite is the most carcinogenic mineral known in both laboratory animals and man (Eborn and Aust, 1995).
Malignant pleural mesothelioma is a frequent cause of death in three villages of
the Cappadocian region of central Anatolia in Turkey: Sarihidir, Tuzkoy, and Karain
(Metintas et al., 1999). Turkey has the highest prevalence of endemic asbestos-related
pulmonary disease (Karakoca et al., 1997). Epidemiological data suggests the exposure
to erionite increases the risk of mesothelioma, even at a much lower exposure than
required for amphibole asbestos (Guthrie, 1992). The Cappadocia region of Turkey
inhabitants had a high incidence of pleural and peritoneal mesothelioma, which was then
related to erionite in the volcanic tuff. Wagner et al. showed that erionite caused
essentially 100% incidence of mesothelioma in exposed rats (1985).
Eborn and Aust showed that erionite not treated with an iron solution is unable to
induce the formation of single strand breaks in DNA (1995). Nejjari et al. found that
erionite is also unable to catalyze the formation of ROS and 8-hydroxydeoxyguanosine in
vitro (1993). After erionite is exposed to iron solutions it shows a high activity both in
14 ROS production as well as DNA single strand breaks (Eborn and Aust, 1995). Erionite
has been found to be more carcinogenic then asbestos (crocidolite). The increased
toxicity is likely due to the erionite having more bioactive iron, which can more readily
produce hydroxyl radicals simply because its coordination to the surface of the erionite.
1.4.2.2 Mordenite
Mordenite is a naturally occurring zeolite mineral, which crystallizes as fibrous
aggregates, masses, and vertically striated prismatic crystals. Mordenite can be colorless, white, faintly yellow, or faintly pink. The well developed crystals are hairlike, being very long, thin, and delicate. Mordenite was described in 1864 by Henry How, who named it after the small community of Morden, Nova Scotia, Canada where it was first found.
Mordenite is found in volcanic rock. Mordenite has been found in Iceland, India, Italy,
Oregon, Washington, Nevada, and Idaho (Amethyst Galleries, 2007). The structure of mordenite is as follows:
(Ca,Na2,K2)Al2Si10O24·7(H2O)
Mordenite does not have iron in its structure either, but it is able to acquire iron through the same processes as erionite. Overall, mordenite is known to be relatively benign.
Little data exists related to human exposure of mordenite. Based on the toxicity
of other fibrous particles, it is anticipated that mordenite would also be toxic to humans.
Mordenite is fibrous; however, it has also been described as a mixture of fibrous and
15 nonfibrous. The nonfibrous component of mordenite represents impurities (Guthrie,
1992). In vivo experiments have shown that mordenite is indeed fibrogenic but not
carcinogenic; however, it must be noted that this experiment was done with an impure
sample which may effect the results (Guthrie, 1992). Palekar et al. found that mordenite
was not cytotoxic to Chinese hamster lung cells (1988).
1.5 The Fenton Reaction and The Haber Weiss Cycle
In 1876 a chemist, H. J. H. Fenton, described the reaction between metal ions,
iron being the most common, and hydrogen peroxide which was causing oxidative
damage. The reaction is known as the Fenton reaction. The reaction with iron follows:
2+ 3+ . - Fe + H2O2 Æ Fe + OH + OH
Iron-catalyzed Fenton Reaction
During phagocytosis of inhaled particles the iron is provided by the particle and the
hydrogen peroxide by the oxidative burst. The iron can either be already associated with the particle, it can be picked up before inhalation due to outside processes, or it can be picked up inside the body after inhalation since free iron has been found in areas of inflammation or infection.
Two mechanisms for the Fenton reaction have been proposed. Although the
Fenton reaction was discovered over a century ago, there is still debate about the actual
mechanism. The first mechanism mentioned in 1931 by Haber and Willstätter, then
16 studied extensively by Haber and Weiss, involves free hydroxyl radical formed by the metal-catalyzed decomposition of hydrogen peroxide (Winterbourne, 1995). The second and third reactions shown in the mechanism represent the Haber-Weiss cycle, where hydrogen peroxide is consumed and oxygen is produced. The reactions proceed as follows:
Initial Reaction (Fenton reaction):
2+ 3+ - • Fe + H2O2 Æ Fe + HO + HO (1)
which is then followed by chain Reactions 2 and 3,
• •- + HO + H2O2 Æ H2O + O2 + H (2)
•- + • O2 + H + H2O2 Æ O2 + HO + H2O (3)
while chain termination is caused by Reaction 4:
2+ • + 3+ Fe + HO + H Æ Fe + H2O (4)
The second mechanism, introduced by Bray and Gorin, involves the formation of a
highly reactive high-valent iron complex (ferryl-oxo) (Buda et al., 2003). The
mechanism follows:
17
Initial Reaction (Fenton reaction):
2+ 2+ Fe + H2O2 Æ FeO + H2O (5)
which is then followed by:
2+ 2+ FeO + H2O2 Æ Fe + H2O + O2 (6)
which terminates with:
2+ 2+ 3+ - FeO + Fe + H2O Æ 2Fe + 2HO (7)
Overall, the more common thought is that hydroxyl radical production does occur.
Literature provides evidence for the fact that Fe(II)-EDTA in the presence of hydrogen peroxide does produce hydroxyl radicals via electron spin resonance (ESR) spectroscopy and analysis of aromatic hydroxylation products (Gutteridge, 1984; Halliwell and
Gutteridge, 1992; Croft et al., 1992).
1.5.1 Detection of Hydroxyl Radical Production
A variety of ways to detect hydroxyl radical production exist. The methods involve spin-trapping EPR spectroscopy, hydroxylation of aromatic organic compounds leading to fluorescent products, and HPLC of hydroxylated products or methanesulfinic acid (MSA) which are produced by the reaction of hydroxyl radical with aromatic compounds or dimethyl sulfoxide (DMSO) respectively (Cheng et al., 2002; Yildiz and
Demiryurek, 1998; Mason and Knecht, 1994; Floyd et al., 1984; Jahnke, 1999).
18 One of the more popular methods involves trapping hydroxyl radicals with probes
that yield chemically detectable products after reaction with the hydroxyl radical. DMSO
is often used when analyzing biological systems because it can be used in high
concentrations and remain nontoxic (Babbs and Steiner, 1990). When hydroxyl radicals
oxidize DMSO, MSA is produced. MSA can be detected by both HPLC and a
spectrophotometric method (Scaduto, 1995). The spectrophometric method was utilized
in this research and described further in Chapter 2.
1.6 Oxidative Burst and Oxygen Concentration
As mentioned, when a particle is inhaled and travels into the alveoli, macrophages act as a line of defense by phagocytosing the particle and undergoing an oxidative burst.
The oxidative burst will now be described in more detail, with focus drifting away from
Fenton chemistry and towards intracellular oxygen concentrations.
1.6.1 Oxidative Burst
During macrophage phagocytosis, particles are engulfed into phagosomes by
invagination of the surface membrane. The phagosome moves deeper into the cell and
fuses with a lysosome, forming a phago-lysosome. NADPH oxidase, an enzyme which
•- converts oxygen to O2 , is assembled and activated in the phago-lysosome and begins the process of molecular oxygen consumption for ROS formation known as the oxidative or respiratory burst. When the NADPH oxidase is activated, beginning the oxidative burst,
•- a variety of mechanisms convert O2 into other ROS such as hydroxyl radicals, singlet
19 oxygen, peroxide, and hydrogen peroxide. The following also happens during the
oxidative burst: (1) macrophages release RNS, fibroblast growth factors, and cytokines,
(2) the concentration of Ca+2 changes, which plays an important role in signal
transduction pathways, and (3) intracellular pH drops (Iles and Forman, 2002).
1.6.2 Detection of Oxidative Burst
Because the lifetime of the hydroxyl radical is so short its production can not be directly measured (reactivity k = 10-9 M-1 s-1 with most organic molecules). A variety of
ways exist to determine if an oxidative burst is occurring. Intracellular detection of
oxidative burst is typically done with a dye having a reduced “dihydro” form. In the
presence of ROS (especially H2O2) this dye reacts to form a florescent species. Many
modified and unmodified forms of dichlorodihydrofluorescein are utilized. Fluorescence
microscopy is commonly used with dyes (like 8-hydroxypyrene-1,3,6-trisulfonic acid) or
dichlorofluorescein diacetate and its derivatives. Flow cytometry is also utilized (Patel et
al., 1987; Lehmann et al., 2000; Bass et al., 1983).
Extracellular detection of oxidative burst if often detected with cytochrome c,
•- •- which measures O2 . The most direct way to detect O2 production is via
chemiluminescence. Luminol (5-amino-2,3-dihydro-1,4-phthalazinedione) reacts with an oxidizing agent exhibiting chemiluminescence. The mechanism consists of luminol being oxidized to the luminol radical, then being oxidized to α-hydroxy hydroperoxide (α
-HHP), then finally the α –HHP decomposing and exhibiting chemiluminescence (Figure
1.10) [Rose and Waite, 2001].
20 1.6.3 Importance of Intracellular Oxygen Concentration
During oxidative stress when ROS are being produced there is a change in the oxygen concentration, since oxygen is being consumed in the process. Because intracellular oxygen concentration is related to important cellular functions, it is important to monitor the oxygen levels for insight to physiological and pathological
processes. Furthermore, the detection of oxygen concentration is extremely important in
tumor cells providing information related to radiation and chemotherapy response. Ji et
al. notes that developing a real time sensor for intracellular monitoring would be very useful for studies to determine how oxygen imbalance affects normal and cancerous cells
(2001). The end goal is for our sensor to have intracellular oxygen monitoring capabilities.
1.6.4 Extracellular Oxygen versus Intracellular Oxygen Concentration
Galbraith et al. has shown that there is a relationship between oxidative stress and
oxygen concentration (Galbraith et al., 2002). In their study, rat alveolar macrophages
were stimulated with zymosan to induce an oxidative burst. The purpose of their study
was to determine the role of NF-κB and protein kinase C during immune system
activation. They monitored the oxygen consumption, determining that oxygen
consumption increases as the oxidative burst is occurring, which is a measure of
extracellular oxygen (Galbraith et al., 2002).
21 As reported by James et al., no direct correlation exists between extracellular and intracellular oxygen concentrations, which were monitored during macrophage oxidative burst (1998). The oxygen level inside of macrophages needs to be maintained so that mitochondrial functions can carry out normally and ROS species can be produced. The
NADPH-oxidase consumes large amounts of oxygen when activated upon membrane stimulation. The macrophages were stimulated with zymosan, which induces oxidative burst. ROS production by the NADPH-oxidase system was shown to be regulated by the
availability of oxygen. Providing different extracellular oxygen environments by varying
concentrations of perfused oxygen as well as cell concentration drew a correlation
between extracellular oxygen and oxidative burst. The main conclusion of this study
shows that the extracellular oxygen concentration does affect the extent of oxidative
burst.
1.7 Oxygen Sensors
Oxygen is an extremely important gas found in our environment that plays
important roles in medicinal, environmental, and analytical chemistries, in which
determining the concentration of oxygen is important. Oxygen sensors have a wide range
of applications such as monitoring combustion processes, pressure sensitive paints,
sterilization and fermentation processes, biological oxygen demand, and intracellular
oxygen concentrations.
22 1.7.1 Typical/Conventional Oxygen Sensors
Oxygen sensors typically fall into one of two categories, electrochemical or optical. The electrochemical sensor most commonly used is called the Clark electrochemical cell invented by Josiah Latimer Clark in 1873 (Figure 1.11). The Clark cell is a voltametric sensor based on the following redox reactions:
+ - O2 + 4H + 4e Æ 2H2O
Ag + Cl- Æ AgCl(s) + e-
The cell is made of a platinum disk cathode and a silver anode. The assembly is in a larger tubular housing that holds a buffered solution of potassium chloride. The bottom of the housing has a thin membrane that is oxygen permeable, and typically made of
Teflon or polyethylene (Clark, 1959). The disadvantages to electrochemical sensors are: local consumption of oxygen during detection, requires stirring, long term drift of the sensor, slow response, and poor stability (Mitchell, 2006).
Optical sensors have become a focus in the last 20 years because they are able to combat many of the disadvantages associated with electrochemical sensors. Optical sensors are based simply on a luminescent probe in some sort of matrix, and the luminescent signal is monitored as a function of oxygen quenching. The main disadvantages associated with an optical sensor are: leaching of the dye from the support matrix and toxicity of the dye. The disadvantages of the optical dye can be easily combated via a novel sensor matrix (Cao et al., 2004; Cheng and Aspinwall, 2006).
23 1.7.1.1 Typical Probes
A variety of probes have been used for oxygen sensing. The options are typically
an organic dye or an organometallic compound. The organic dyes, which are usually
polycyclic aromatic compounds, are not used often because they require UV excitation,
and they are known to easily photodecompose (Amao, 2003).
As far as organometallic compounds are concerned metalloporphyrin complexes,
which commonly center around platinum or palladium, or a transition metal polypyridine complex, commonly centered around ruthenium, osmium, rhenium, or iridium, are used.
Molecules with ruthenium and platinum are ideal because they are photostable, produce
long electronic excited state lifetime, have high quantum yields, and are readily quenched
by oxygen. Downfalls of platinum and palladium complexes are that they often undergo
oxidation when illuminated in the presence of oxygen (Amao, 2003).
Some typical probes can be seen in Figure 1.12. Overall, the ideal characteristics
of a probe are as follows: high quenchability by oxygen, large Stokes shift, high
quantum yield, long lifetime, and visible excitation. In our research we decided to use
tris(2,2’-bipyridyl) ruthenium(II) because it meets the above ideal requirements, and it
also meets a critical size criteria that will be discussed after revealing the matrix of
choice.
24 1.7.1.2 Typical Matrices
The type of matrix utilized for the sensor is of key importance. Requirements for
an ideal matrix consist of having good oxygen permeability, as well as high solubility of
the probe in the matrix. If solubility is low, some methodology should be utilized for
nearly permanent encapsulation of the probe in the matrix. Typically, polymers and sol-
gel compounds are utilized for matrices.
The oxygen permeable polymers that are often used are polystyrene, poly
(methylmethracrylate), fluoropolymers, and cellulose derivatives (Amao, 2003). Silica
sol-gels have also been investigated as a matrix for ruthenium complex immobilization
(McDonagh et al., 1998). PDMS polymers are the most common matrices that have been
studied due to their high oxygen permeability and inertness in biological systems.
However, PDMS is hydrophobic and ruthenium complexes are polar, making it difficult
to immobilize the ruthenium complex in the PDMS without leaching of the dye. A
variety of solutions have been studied to solve this problem, such as ionic binding and the incorporation of fillers to which the ruthenium complex may bind. Covalently linking the dye to the polymer backbone has also been utilized (Xu et al., 1994). A general disadvantage of using a polymer matrix are the fact that the photo-excited state of the trapped organic dye may react with the matrix to decrease the dye and matrix stability.
Inorganic glasses from sol-gel processes can overcome this problem. Hence, sol-gel support matrices are also commonly used. The advantages of sol-gel matrices are that they are chemically inert, mechanically stable, optically transparent, and highly porous
(Yeh et al., 2006). The heterogeneous environment provided by the polymer and sol-gel
25 type matrices cause the probe to be in different environments, contributing to a
distribution of lifetimes and a non-linear Stern-Volmer relationship described in more
detail later when discussing oxygen quenching.
1.7.1.3 Novel Zeolite Matrix
A zeolite matrix is able to solve the problems presented previously of the sol-gel
and polymer matrices. Because the zeolite offers supercages with smaller windows, a
molecule can be synthesized inside of the supercage and will hence be permanently
entrapped. A variety of techniques have been utilized in order to encapsulate dyes into
microporous solids: (1) inclusion of the dye in the synthesis gel, (2) simple surface
adsorption, and (3) the “ship-in-a-bottle” synthesis, which allows for permanent
encapsulation of the guest molecule (Yao et al., 2001; Sarkar et al., 1994; Herron et al.,
1985). The research presented in this thesis focuses on the “ship-in-a-bottle” synthesis.
“Ship-in-a-bottle” seems to have been coined by Herron when describing nickel complexes entrapped inside of zeolite-X (Herron et al., 1985). Typically, the precursor is introduced via ion-exchange techniques since zeolites often have an overall charge.
However, you can also utilize a non-polar molecule as a precursor by introducing it via vapor phase, in a solvent, or via a solid state reaction after the zeolite has been dehydrated or if it is highly hydrophobic (Chen et al., 2005; Payra and Dutta, 2003;
Kaupp, 2003). A “ship-in-a-bottle” technique was utilized in order to construct
26 2+ Ru(bpy)3 inside of the supercages of zeolite-Y by DeWilde et al. (1980). DeWilde et
3+ al. performed this experiment by ion-exchanging Ru(NH3)6 into zeolite-Y. Excess 2,2’-
2+ bipyridine was then added under to form the Ru(bpy)3 complex inside of the zeolite-Y.
Utilizing a siliceous zeolite-Y is useful in our research since we would like to monitor oxygen quenching of our probe in an aqueous medium. Intrazeolitic water is believed to prevent dissolved oxygen from diffusing into the pores of the zeolite supercages in order to access the probe (Coutant et al., 2003; Payra and Dutta, 2003).
Figure 1.13 demonstrates what this would look like. Therefore, by creating a siliceous zeolite, we create a zeolite that has no overall charge and hence no interaction with water to prevent diffusion of oxygen through the pores.
A variety of dealuminatoin methods have been utilized in order to create siliceous
zeolite-Y, which include treatment with acids, volatile halides, chelating agents, and
steam (Hriljac et al., 1993). Often times a combination of these methods will be utilized
to produce a highly siliceous zeolite with the least defects possible. One of the most
common dealumination methods involves the hydrothermal treatment of ammonium-
exchanged zeolite-Y (Kerr, 1967). The dealumination method tried in this research
involves flowing silicon tetrachloride gas over zeolite-Y; the method is simple and can be
performed in one step (Anderson and Klinowski, 1986; Beyer et al., 1985). The silicon
tetrachloride gas allows for a high amount of dealumination, as well as the ability to
dealuminate the sample to different extents. Characterization of the dealuminated
zeolite-Y is typically done via powder X-ray diffraction and solid state nuclear magnetic
resonance spectroscopy (SSNMR) [Hriljac et al., 1993].
27 Besides the advantage of permanent entrapment of the probe in the zeolite matrix,
another advantage of utilizing a zeolite matrix is known as “the cage effect”, which has
been intensively studied by Maruzewski et al. (1993, 1995). The “cage effect” is
important because one of the main limitations of using ruthenium polypyridyl complexes
is the photoinitiated decomposition arising from the population of the low-lying dd states.
The decomposition of the ruthenium complex can actually be diminished if the dd ligand
field can be destabilized, which will lower the probability of populating it. Placing the
ruthenium complex in a rigid medium like a zeolite can reduce the photolability
(Maruszewski and Kincaid, 1995). The cage of the zeolite actually alters the energy level
of the dd state, which is related to decomposition or ligand loss of the ruthenium
complex. The argument can be rationalized by realizing that the population of the dd
state equates to elongation of the Ru-N bonds. In a rigid matrix, like a zeolite, these
distortions are prohibited which raises the energy level of the dd state, and hence demotes
the promotion of electrons into the dd state that cause decomposition. Lumpkin and co-
workers initially proposed this to explain similar behavior of tris(2,2’-bipyridine)
ruthenium(II) in a rigid cellulose acetate matrix (1990). Therefore, decreased photolability arises from the destabilization of the dd state, as well as the increased
chance of the 2,2’-bipyridine ring enclosure to recombine with the ruthenium due to
encapsulation and the forced close proximity.
28 1.8 Tris(2,2’-bipyridyl) Ruthenium Photochemistry and Quenching
2+ Much research has involved studying Ru(bpy)3 , and α–diimines in general because of their useful photochemical properties. Tris(2,2’-bipyridyl) ruthenium(II)
2+ 2+ (Ru(bpy)3 ) is octahedral with D3 symmetry. Luminescence of Ru(bpy)3 was first
noted in 1959 by Paris and Brandt after they recorded an absorption and emission spectra
of the complex in solution.
2+ The photochemistry of Ru(bpy)3 , which will now be briefly discussed, is shown
2+ in Figure 1.14. Ru(bpy)3 shows a broad absorption around 450nm with an absorptivity
of around 15,000 L mol -1 cm-1, which represents a dπÆπ*(bpy) metal to ligand charge transfer (MLCT) transition (Demas and Taylor, 1979). In other words, an electron is promoted from the ruthenium atom to the bipyridyl ligands. The reaction is as follows
(Caspar and Meyer, 1983):
light
2+ .- 2+* Ru(bpy)3 Æ [Ru(III)(bpy)2bpy ]
(dπ)6 (dπ)5 (π*)1
Within less than one picosecond after absorption into the singlet MLCT state (1MLCT)
an intersystem crossing into the three closely space emissive triplet MLCT states
(3MLCT) occurs efficiently (Φ~1) (Demas and Taylor, 1979). Deactivation of this state can occur via radiative decay, nonradiative decay, or thermal population of the dd state
(Caspar and Meyer, 1983).
29
Emission light
.- 2+* 2+ [Ru(III)(bpy)2bpy ] Æ Ru(bpy)3
The lifetime of the excited state molecule is about 800ns at room temperature in
water, and decays while emitting a red photon (around 620nm) with a quantum yield of
about 0.042 (Meyer, 1986; Roundhill, 1994).
As mentioned, the excited state which provides the luminescence of the ruthenium
complex has been determined to actually be a manifold of three closely spaced (100cm-1)
3MLCT states in thermal equilibrium (Figure 1.14) [Hager, 1975]. The state which has
the highest energy, and the most singlet character, is expected to dominate both the radiative and non-radiative decay. The three states are also responsible for the electron
and energy transfer reactions of interest.
Two other exited states exist, which are thermally accessible from the low lying
emitting states. These states are a fourth 3MLCT state as well as a 3dd state. In solution,
the dd state dominates the temperature dependent effects on emission. The dd state, if
accessible, is very important in the chemistry occurring because it has a very short
lifetime and leads to ligand loss when populated.
The dd state has been shown to be made thermally inaccessible by lengthening the
Ru-N bonds. The Ru-N bond can be lengthened via appropriate choice of host matrix,
which causes an increased energy of the state. As the energy of the state increases it
becomes less thermally accessible until it becomes completely inaccessible. When the dd
30 state is no longer available the fourth 3MLCT state dominates the temperature dependent
properties of the complex and provides an efficient pathway for decay of the Ru complex
excited state. Theory predicts that this state should have a greater amount of singlet
character and therefore a shorter lifetime than the ensemble of lower lying 3MLCT states.
Placing the ruthenium complex inside of a zeolite allows us to have the benefit of a
crystalline, homogeneous matrix as well as a decrease in photodecomposition
(Maruszweski et al., 1993; Lumpkin et al., 1990).
1.8.1 Oxygen Sensing/Quenching
Two types of quenching mechanisms are generally known, static and dynamic.
These quenching mechanisms describe interaction between the fluorescent dye molecule and the quenching molecule. During dynamic, or collisional, quenching the quencher diffuses to the fluorescent dye molecule at some point during the lifetime of the excited state. The fluorophore then returns to the ground state without emission of light. During static quenching the fluorescent molecule and the quencher form a non-fluorescent complex, which is independent of excited state population. We are focused on dynamic quenching, since that is the mechanism believed to prevail when a dye molecule is quenched by oxygen (McEvoy et al., 1996; Hughes and Douglass, 2006).
Oxygen sensing is quantified by determining the luminescence or lifetime
quenching caused due to the presence of oxygen. The interaction of the excited state
ruthenium complex with oxygen results in a deactivation pathway. In the presence of
2+* oxygen, there is an energy transfer process that occurs from Ru(bpy)3 to molecular
31 2+ oxygen to form Ru(bpy)3 and singlet oxygen (Demas et al., 1999). Luminescence
intensity and lifetime measurements are related to oxygen concentration through the
Stern-Volmer equations which follow:
Io/I = 1 + Ksv[Q] = 1 + kqτo[Q]
τo/ τ = 1 + Ksv[Q]
Ksv = kq* τo
Ksv represents the Stern-Volmer quenching constant. Q is the quenching species
concentration, I is the luminescence emission intensity in the presence of the quencher, Io is the emission intensity in the absence of the quenching species, τ is the luminescence lifetime in the presence of the quencher, τo is the luminescence lifetime in the absence of
the quencher, and kq is bimolecular quenching constant. The Stern-Volmer relationship
is linear in aqueous solutions, making calibration easy. The linearity is due to the fact
that we have a homogeneous system, and hence a single species is quenched
bimolecularly. The straight line represents the inverse relationship of oxygen
concentration to emission intensity, with a slope equal in magnitude to the Stern-Volmer
constant (Hartmann et al., 1995).
However, in most applications (i.e. fiber optics, etc.) the sensor is cast into a
heterogeneous matrix, such as a sol-gel or polymer, that cause the Stern-Volmer relationship to deviate from linearity. The deviation from linearity is related to the fact
the dye molecules are in a heterogeneous environment, which causes them to show
32 different responses due to their more complex decay. The heterogeneous environment
can be viewed as changing the decay mechanism in two ways. First, the dye complexes
may exist in two totally different environments, one quenchable and one unquenchable.
Secondly, the dye complexes exist in two totally different, quenchable environments that
exhibit different quenching rate constants (Carraway et al., 1991). Carraway et al. states
that the second model is the more realistic model. The Stern-Volmer equation used to
describe the second situation is as follows:
Io/I = 1 / ((f01/1+KSV1[Q]) + (f02/1+KSV2[Q]))
Where f01 and f02 represent the fraction of emission from the dye complex in each specific
environment, unquenched. The KSV1 and KSV2 are the Stern-Volmer constants for each of
the dye molecules in the two different environments.
Bacon and Demas demonstrated that tris(4,7-diphenyl-1,10-phenanthroline)
ruthenium(II) perchlorate immobilized in a silicone rubber matrix shows a downward
curvature of the Stern-Volmer plot (1987). Douglas and Eaton placed Pt and Pd octaethylporphyrins in three different polymers (ethyl cellulose, cellulose acetate butyrate polymer, or polyvinylchloride) with all cases revealing non-linear Stern-Volmer plots
(2002). Sol-gel sensors were then tried. Tang et al. demonstrated that tris(4,7-diphenyl-
1,10-phenanthroline) ruthenium(II) immobilized in n-octyltriethoxysilane / tetraethylorthosilane composite films demonstrated a non-linear Stern-Volmer plot
(2003). Overall, with the polymer and sol-gel matrices it is very common to see
33 nonlinear Stern-Volmer plots. Meier et al. showed a non-linear Stern-Volmer plot when
tris(2,2’-bipyridyl) ruthenium(II) was immobilized inside of commericial zeolite-Y,
which was then distributed in a silicone polymer (1995). The non-linearity in this case is
likely due to the intrazeolitic water influencing the environment of the ruthenium probes.
Utilizing a siliceous zeolite should allow each probe to see the same environment, have the same reponse, and demonstrate a linear Stern-Volmer plot.
1.9 Intracellular Optical Direct Dissolved Oxygen Monitoring
Very few techniques exist that monitor intracellular oxygen concentration. In
fact, most techniques monitor extracellular oxygen concentrations, then make
correlations to intracellular oxygen concentrations. It is assumed that intracellular
oxygen is similar to that of extracellular oxygen due to the fact that oxygen diffuses
freely across cell membranes. However, utilizing extracellular oxygen concentrations to
make conclusions about intracellular oxygen concentrations is not so simple. It has been
shown that there are differences between intracellular and extracellular oxygen
concentrations in cells that vary nonlinearly (Morse and Swartz, 1985; Glockner et al.,
1989; James et al., 1998; Grinberg et al., 1998). The actual mechanism to relate
intracellular and extracellular oxygen concentrations is not yet elucidated. Direct
intracellular oxygen monitoring is important for this reason.
34 1.9.1 Fiber Optic
Fiber optic optical oxygen sensors allow for easy placement of the sensor for
intracellular monitoring. The typical configuration of a fiber optic sensor is to have the
sensor dye immobilized in a polymer matrix at the end of the fiber. Another advantage of
a fiber optic sensor is that the polymer serves the purpose of protecting the sensing dye
from intracellular macromolecules. Interactions of the sensing dye with the
macromolecules can prohibit accurate sensing.
The main disadvantage of fiber optic sensors for intracellular monitoring is their
size, which is typically on the microns scale. The fibers readily lead to cell perturbation
and damage. Although fiber tips can now be made down to 20nm, there is still physical
distortion of the cell due to the large insertion volume of the fiber (Buck et al., 2004).
Besides the limitation of the size and their destructiveness to the cell, they are also limited by their low throughput and longer response times due to their size (Ji et al.,
2001; Buck et al., 2004).
1.9.2 Oxygen Sensitive Dye
Oxygen sensitive dyes can be independently used to monitor oxygen
concentration by looking at their lifetime or emission quenching. For example, tris(1,10-
phenanthroline)ruthenium chloride (Ru(phen)3) has been utilized for the measurement of
the effect of external hypoxia on the molecular oxygen level in J774 murine
macrophages, looking at the emission quenching of the dye by molecular oxygen. The
measurements were not very sensitive because the emission of the dye in a cellular
35 environment was not stable due to interaction with cellular components like proteins,
DNA, and ROS. Furthermore, when utilizing a dye alone for intracellular monitoring care must taken to ensure that the dye does not cause any toxicity (Asiedu et al., 2001).
The advantage of these sensors is their size. As opposed to the fiber optic
sensors, there is no concern about the physical perturbation of the cell. However, there
are multiple disadvantages of directly loading a dye into cells for intracellular
monitoring. As mentioned, there are issues with toxicity of the dye as well as
interferences from cellular components. Because there is not matrix to separate dyes and
prevent interactions between dye molecules, multiple dyes can not be utilized in order to
sense multiple analytes. A matrix surrounding the dye molecules would solve these main
disadvantages, which is where research is currently focusing (Aylott, 2003).
1.9.3 Phospholipid Spheres
Phospholipid vesicles, also known as liposomes, can entrap a sensor in a
phospholipid membrane. The entrapment of the oxygen sensitive dye limits interactions
between the dye and cellular compartments, as well as the toxicity of the dye. Traditional
phospholipid sensors have no chemical or environmental stability. When liposomes are
incubated with cells, they fuse with the cellular membrane in order to deliver intracellular
components. The loss in structure from fusion with cellular membranes is the main
disadvantage of these intracellular oxygen sensors (McNamara et al., 2001). In order to
combat the instability of the spheres, polymerizable phospholipids have been utilized.
The phospholipids are difficult to work with and still show leaching of the dye (Cheng
36 and Aspinwall, 2006; Regen et al., 1980). The method with more promise involves
polymerizing monomer units in the bilayer region of the sphere (Cheng and Aspinwall,
2006; Hotz and Meier, 1998). These sensors still need tested in cells to determine their
stability. Furthermore, it is reported that the sensors can be made in the size range of 50-
500nm, which is larger than the size of PEBBLE nanosensors, which will be introduced
in the next section (Cheng and Aspinwall, 2006). Note, there are also lipobead sensors
which consist of a dye entrapped inside of a polymer then coated with a phospholipid
membrane (McNamara et al., 2001). However, these sensors tend to have diameters on
the order of microns, which is not ideal.
1.9.4 PEBBLE
PEBBLE (probes encapsulated by biologically localized embedding) sensors were
developed to overcome the poor throughput of fiber optic sensors, and to combat the size
issue related to cellular damage. Xu et al. were the first to show the PEBBLE sensor operating as an intracellular oxygen sensor ratiometrically (2001). Xu et al. formed the
PEBBLE sensors from sol-gel and utilized tris(4,7-diphenyl-1,10-phenanthroline) ruthenium(II) chloride as the oxygen sensor along with Oregon Green 488-dextran as an oxygen-insensitive dye for the purpose of a ratiometric measurement. Intracellular oxygen measurements were performed with rat C6 glioma cells by changing the oxygen concentration of the Dulbecco’s phosphate buffered saline (Xu et al., 2001). The average intracellular oxygen concentration in cells that were exposed to air saturated buffer was
7.9 ± 2.1 ppm, while the concentration for the cells exposed to nitrogen saturated buffer
37 for 2 minutes was less than 1.5 ppm. Overall, the response of the sensor was reversible,
stable, and the first demonstration of real-time intracellular analysis oxygen
concentrations (Xu et al., 2001).
Advantages of PEBBLE sensors are that they offer great oxygen sensing
characteristics, such as small size (20-200nm in diameter) for minimal cell perturbation,
fast response, high selectivity, and good reversibility. The sensors are delivered into the
cells by a variety of techniques, including picoinjection, gene gun delivery, liposomal
incorporation, and natural ingestion. The sol-gel glass matrix, or organically modified
silicate matrices sometimes used, allows for high oxygen permeability, mechanical and
chemical stability, as well as optical clarity. Because of the small size of PEBBLE
sensors they can be used to sense oxygen in different cellular organelles. The PEBBLE
sensors are truly noninvasive, allowing for the cells to remain viable (Clark et al., 1998).
The main disadvantage of the PEBBLE sensor is the concern with spectral
overlap of the sensing components when trying to monitor several analytes at once (Park
et al., 2005). It is noted that spatial resolution may be able to overcome this drawback;
however, it is difficult to spatially resolve the nanometer sized PEBBLEs. PEBBLE sensors also have issues limiting their quantitative power due to leaching of the dye from the matrix, which limits their stability and sensitivity. This requires the use of a
technique where the dye is covalently bound to the sol-gel or polymer matrix, ratiometric
measurements to be made, or lifetime measurements to be made instead of emission (Xu
et al., 2001; McNamara et al., 2001).
38 1.10 Goals of This Thesis
The aim of the research presented was to study asbestos toxicity by monitoring hydroxyl radical concentration in the presence of two different antioxidants using ferric iron exchanged mordenite and erionite as model particles. The model particles will allow us to make conclusions as to how the surface structure may or may not play a role in bioactivity. The hydroxyl radicals produced due to the particle bioactivity play an important role intracellularly and are the cause of in vivo oxidative stress. When cells are undergoing oxidative stress they typically experience fluctuations in intracellular oxygen concentration. Intracellular oxygen concentration provides information related to cellular performance and viability. Therefore, synthesizing and testing a dissolved optical oxygen sensor was of interest to eventually provide information as to cell function. The main goal of the sensor was to ensure that the dye did not leach from the matrix and to have good sensitivity. This was done by optimizing the synthesis of tris(2,2’-bipyridyl) ruthenium (II) inside of the supercages of siliceous zeolite-Y. Also, for application purposes the sensor was immobilized on the end of a fiber optic and tested in gas saturated aqueous solutions, as well as placed in cells to determine if oxygen sensing would be possible intracellularly. Overall, fabricating a dissolved oxygen optical sensor that has a linear Stern-Volmer plot, no dye leaching, and high sensitivity was the main goal in order to provide information as to cell condition, which is extremely important in situations where oxidative damage may have occurred in vivo.
39 DATE, 1952 Particles mg/m3 Deaths/Day 12-01 0.25 265 12-02 0.3 295 12-03 0.4 310 12-04 0.35 280 12-05 1.1 405 12-06 1.6 590 12-07 1.7 895 12-08 1.65 905 12-09 1.1 780 12-10 0.4 550 12-11 0.25 530 12-12 0.2 485 12-13 0.4 500 12-14 0.35 460 12-15 0.3 435
deaths per day particles: mg/m3
1 2 3 4 5 6 7 8 9 101112131415 December
Figure 1.1: During the London fog disaster the severe increase in inhalable particulate matter showed deleterious results. The data demonstrates the fact that the inhalable particulate matter increased or decreased with a direct correlation to the number of deaths per day (estimated from Green et al., 2002). The plot has arbitrary units for the y-axis and is presented to show the correlation between particle concentration and number of deaths per day.
40
Alveolar epithelium asbestos ROS, RNS
silica
AM TNF, IL-1
lymphocytes IM MIP, MCP
mast cells cytokines
F
neutrophils
cytokines
Figure 1.2. A depiction of the cascade of event that occur after a mineral dust is inhaled. AM = alveolar macrophage; IM = interstitial macrophage; F = fibroblast; ROS = reactive oxygen species; RNS = reactive nitrogen species; TNF = tumor necrosis factor; IL-1 = interleukin-1; MIP = macrophage inflammatory proteins; MCP = macrophage chemotactic proteins (Mossman and Chung, 1998).
41 Pre-Exposure Phase
Adsorption of chemicals/allergens to the particles
______
DOSE ______
Exposure Phase
Number, Weight, Total Surface Area Æ Size
______
Deposition and Retention Phase
Deposition Pattern Æ
Translocation, Internalization Æ
size Biopersistence Æ suface area hydrophobicity biodurability
Retention (=deposition – clearance)
______
Effector Phase
Biological Response Æ changed macrophage Æ phagocytosis and cytokines
size suface area number adsorption capacity
B Cells general cytoxicity epithelium (adsorption capacity) (adsorption capacity, (surface area) surface area)
“CHANGED MICROENVIRONMENT”
______
Figure 1.3: Flow chart to represent critical components of particle inhalation. The overall changed microenvironment in the lung can be thought of as being caused by the different phases described. Underlined font represents crucial particle characteristics (van Zijverden, 2007). 42
nasal, pharynx, 80 larynx % , 60 alveolar osition p
40 ional De
g tracheo-
Re bronchial
20
0.001 0.01 0.1 1 10 100 Particle Diameter, μm
Figure 1.4: Relationship between particle diameter and regional deposition in the lung (IRCP, 1994).
43
Figure 1.5: Diagram of the respiratory tract to show the path particles encounter upon inhalation (Fischer, 2005).
44
Figure 1.6: Detailed view of the alveolar area showing ciliated cells, clara cells that contribute to the lung lining fluid layer, and macrophage locations (Molson Medical Informatics Sampler, 2007)
45 -e- –H+ - - · -
Ascorbate (AH-) Ascorbyl Radical (A·-) Dehydroascorbate (A)
slow dismutation
GS· GSSG –H· ·- - + O2 ( -e –H ) GS-
O Glutathione (GSH) ·- 2 GSSG
Figure 1.7: Hydrophilic, non-enzymatic antioxidants during their function. Ascorbate is a one-electron donor. GSH prefers to act as a hydrogen donor (Chaudière and Ferrari- iliou, 1999).
46
Figure 1.8: Demonstrates the degree of seriousness with which asbestos toxicity should be taken (Environmental Working Group, 2007).
47
SOD
LTA
sodalite cage
FAU
Structure 6
Figure 1.9: Zeolite structures which are formed from the sodalite, or β, cage subunits. Four and six membered rings of silica and alumina form the sodalite cages. SOD = sodalite, LTA = zeolite A, FAU = faujasite (zeolites X and Y), Structure 6 = Breck structure six (zeolite EMT).
48
- - . + oxidant ─ (H )
luminol luminol radical
superoxide
HO O
C
- CO3 OOH
+ N2 + hν OH+
aminophthalate α-hydroxy-hydroperoxide
Figure 1.10: Mechanism by which luminol is able to detect reactive oxygen species. Three major steps exist in the pathway (Rose and Waite, 2001).
49
F
G
C
B
E
D A
Figure 1.11: Diagram of a Clark cell, which is used for the voltammetric determination of oxygen. The areas are labeled as follows: (A) Pt- (B) Ag/AgCl-electrode (C) KCl electrolyte (D) Teflon membrane (E) rubber ring (F) voltage supply (G) galvanometer (Clark, 1959).
50
phenanthroline 2,2’-bipyridine
M
pyrene metal octaethylporphyrin (metal = Pd, Pt)
Figure 1.12: Some commonly used probes for optical oxygen sensing (Demas et al., 1999)
51
N N
N Ru2+ N 2+ = water N = Ru(bpy) N 3
Crystal defect
N N
N Ru2+ N dry N N wet
partially blocked windows completely blocked window
extrazeolitic water barrier
Figure 1.13: Depiction of intrazeolite water, which then results in problems with oxygen diffusion to the probe and hence a reduction in sensor sensitivity (Coutant et al., 2003).
52
energy
1 MLCT 3dd Ligand kisc loss
hν 3MLCT kq 1 k O2 r k Excited nr Oxygen hν state energy Quenching
transfer 3O GS 2
*arbitrary energy levels
Figure 1.14: Energy level diagram representing tris(2,2’-bipyridyl) ruthenium(II) decay pathways. Luminescence emission and lifetime quenching measurements are taken from the lowest three 3MLCT states, which are in thermal equilibrium (Hager and Crosby, 1975; Maruszewski et al., 1993).
53 1.11 References
Amao, Y. Microchim. Acta. 2003, 143, 1-12.
Amethyst Galleries. The Mineral Mordenite. http://mineral.galleries.com/minerals/ silicate/mordenit/mordenit.htm (accessed 07/25/2007)
Anderson, M.W.; Klinowski, J. J. Chem. Soc. Faraday Transactions I. 1986, 82, 1449-1469.
Archer, V.E. Med. Hypotheses. 1979, 5(11), 1257-1260.
Asiedu, J.K.; Ji, J.; Nguyen, M.; Rosenzweig, N.; Rosenzweig, Z. J. Biomed. Opt. 2001, 6(2), 116-121.
Aylott, J.W. Analyst. 2003, 128, 309-312.
Babbs, C.F.; Steiner, M.G. Method Enzymol. 1990, 186, 137-147.
Bacon, J.R.; Demas, J.N. Anal. Chem. 1987, 59(23), 2780-2785.
Bass, D.A.; Parce, J.W.; Dechatelet, L.R.; Szejda, P.; Seeds, M.C.; Thomas, M. J. Immunol. 1983, 130(4), 1910-1917.
Beyer, H.K.; Belenykaja, I.M.; Hange, F.J. J. Chem. Soc. Faraday Transactions I. 1985, 81, 2889-2901.
Buck, S.M.; Xu, H.; Brasuel, M.; Philbert, M.A.; Kopelman, R. Talanta 2004, 63(1), 41-59.
Buda, F.; Ensing, B.; Gribnau, M.C.M.; Baerends, E.J. Chem. Eur. J. 2003, 9(14), 3436-3444.
Cao, Y.; Koo, Yong-Eun, L.; Kopelman, R.. Analyst. 2004, 129(8), 745-750.
Carr, A.; Frei, B. FASEB J. 1999, 13, 1007-1024.
Carraway, E.R.; Demas, J.N.; DeGraff, B.A. Anal. Chem. 1991, 63, 332-336.
Caspar, J.V.; Meyer, T.J. J. Am. Chem. Soc. 1983, 105, 5583-5590.
Chaudière, J.; Ferrari-iliou, R. Food Chem. Toxicol. 1999, 37, 949-962.
Chen, Y.; Li, X.; Gao, F.; Qiu, S. Mater. Lett. 2005, 59, 2598-2600.
54
Cheng, F.; Jen, J.; Tsai, T. J. Chromatogr. B. 2002, 781(1-2), 481-496.
Cheng, Z.; Aspinwall, C.A. Analyst. 2006, 131, 236-243.
Clark, H.A.; Barker, S.L.R.; Brasuel, M.; Miller, M.T.; Monson, E.; Parus, S.; Shi, Z- Y.; Song, A.;Thorsrud, B.; Kopelman, R.; Ade, A.; Meixner, W.; Athey, B.; Hoyer, M.; Hill, D.; Lightle, R.; Philbert, M.A. Sens. Actuators, B. 1998, 51, 12-16.
Coutant, M.A.; Payra, P.; Dutta, P.K. Microporous Mesoporous Mater. 2003, 60, 79-90.
Croft, S.; Gilbert, B.C.; Smith, J.R. Lindsay; Whitwood, A.C. Free Radic. Res. Commun. 1992, 17(1), 21-39.
Cross, C.E.; van der Vliet, A.; O’Neill, C.A.; Louie, S.; Halliwell, B. Environ. Health Perspec. 1994, 102(10), 185-191.
Demas, J., DeGraff, B.A., Coleman, P.B. Anal. Chem. 1999, 793A-800A.
Demas, J.N.; Taylor, D.G. Inorg. Chem. 1979, 18, 3177-3181.
DeWilde, W.; Peeters, G.; Lunsford, J.H. J. Phys. Chem. 1980, 84, 2306-2310.
Douglas, P.; Eaton, K. Sens. Actuators, B. 2002, 82, 200-208.
Eborn, S.K.; Aust, A.E. Arch. Biochem. Biophys. 1995, 316(1), 507-514.
Environmental Working Group. The Asbestos Epidemic in America. http://www.ewg.org/reports/asbestos/facts/fact1.php (accessed 07/24/2007).
EPA: Environmental Protection Agency. Asbestos and Vermiculite: Basic Information. http://www.epa.gov/asbestos/pubs/help.html (accessed 07/24/2007).
Erionite CAS No. 66733-21-9. http://ntp.niehs.nih.gov/ntp/roc/eleventh/profiles/ s083erio.pdf (accessed 07/25/2007), part of National Toxicology Program http://ntp.niehs.nih.gov/ (accessed 07/25/2007).
Fenoglio, I.; Martra, G.; Coluccia, S.; Fubini, B. Chem. Res. Toxicol. 2000, 13, 971- 975.
Fischer, T.G. Faculty homepage. http://faculty.ircc.edu/faculty/tfischer/images/ lower%20resp.%20tract.jpg (accessed 07/24/2007), part of Teresa G. Fischer's homepage http://faculty.ircc.edu/faculty/tfischer (accessed 07/24/2007).
55
Floyd, R.A.; Watson, J. J.; Wong, P.K. J. Biochem. Bioph. Methods. 1984, 10(3-4), 221-235.
Fubini, B. Environ. Health Persp. 1997, 105(5), 1013-1020.
Fubini, B.; Arean, C.O. Chem. Soc. Rev. 1999, 28, 373–381
Galbraith, W.; Thompson, K.; Haq, T.; Timmins, M. BD Biosciences. http://www. bdbiosciences.com/discovery_labware/Products/drug_discovery/ polystyrene_microplates/oxygen_biosensor_system/pdf/ OBS_innate_immune_system.pdf (accessed 07/26/2007). Innate Immune System Signaling: NR8383 Cell Oxidative Burst Measured by Oxygen Consumption. 2002.
Glockner J F; Swartz H M; Pals, M.A. J. Cell. Physiol. 1989, 140(3), 505-511.
Green, L.C.; Crouch, E.A.C.; Ames, M.R.; Lash, T.L. Regul. Toxicol. Pharm. 2002, 35, 327-337.
Grinberg, O.Y.; James, P.E; Swartz, H.M. Adv. Exp. Med. Biol. 1998, 454, 415-423.
Guthrie, G.D. Environ. Health Persp. 1997, 105(5), 1003-1011.
Guthrie Jr., G.D. Am. Mineral. 1992, 77, 225-243.
Gutteridge, J.M.C. FEBS Lett. 1984, 172(2), 245-249.
Gutteridge, J.M.C.; Mumby, S.; Quinlan, G.J.; Chung, K.F.; Evans, T.W. Biochem. Bioph. Res. Co. 1996, 220, 1024-1027.
Hager, G.D.; Crosby, G.A. J. Am. Chem. Soc. 1975, 97, 7031-7037.
Halliwell, B.; Gutteridge, J.M.C. FEBS. Lett. 1992, 307(1), 108-112.
Hardy, J.A.; Aust. A.E. Chem. Rev. 1995, 95(1), 97-118.
Hartmann, P.; Leiner, M.J.P.; Lippitsch, M.E. Sens. Actuators, B. 1995, 29, 251-257.
Herron, N.; Stucky, G.D.; Tolman, C.A. Inorg. Chim. Acta. 1985, 100, 135-140.
Hesterberg, T.W.; Chase, G.; Axten, C.; Miller, W.C.; Musselman, R.P.; Kamstrup, O.; Hadley, J.; Morscheidt, C.; Bernstein, D.M.; Thevenaz. P. Toxicol. Appl. Pharmacol. 1998, 151, 262-275.
56 Hotz, J.; Meier, W. Langmuir. 1998, 14(5), 1031-1036.
Hriljac, J.A.; Eddy, M.M.; Cheetham, A.K.; Donohue, A.J.; Ray, G.J. J. Solid State Chem. 1993, 106, 66-72.
Hughes, V.A.; Douglas, P. J. Fluorescence. 2006, 16(3), 403-409.
ICRP: International Commission on Radiological Protection [1994]. Human respiratory tract model for radiological protection. Oxford, England: Pergamon, Elsevier Science Ltd., International Commission on Radiological Protection Publication No. 66.
Iles, K.E.; Forman, H.J. Immunol. Res. 2002, 26(1–3), 95-105.
Jahnke, L.S. Anal. Biochem. 1999, 269(2), 273-277.
James, P.E.; Grinberg, O.Y.; Swartz, H.M. J Leukoc. Biol. 1998, 64(1), 78-84.
Ji, J.; Rosenzweig, N.; Jones, I.; Rosenzweig, Z.. Anal. Chem. 2001, 73, 3521-3527.
Karakoca, Y.; Emri, S.; Cangir, A.K.; Baris, Y. I. Indoor Built Environ. 1997, 6(2), 100-105.
Kaupp, G. Cryst. Eng. Comm. 2003, 5(23), 117-133.
Kerr, G.T. J. Phys. Chem. 1967, 71, 4155-4156.
Lehmann A. K; Sornes S; Halstensen A. J. Immunol. Methods. 2000, 243(1-2), 229- 242.
Light, W. G.; Wei, E.T. Nature. 1977, 265, 537-539.
Lumpkin, R.S.; Kober, E.M.; Worl, L.A.; Murtaza, Z.; Meyer, T.J. J. Phys. Chem. 1990, 94, 239-243.
Martra, G.; Tomatis, M.; Fenoglio, I.; Coluccia, S.; Fubuni, B. Chem. Res. Toxicol. 2003, 16, 328-335.
Maruszewski, K.; Kincaid, J.R. Inorg. Chem. 1995, 4, 2002-2006.
Maruszewski, K.; Strommen, D.P.; Kincaid, J.R. J. Am. Chem. Soc. 1993, 115, 8345- 8350.
Mason, R.P.; Knecht, K.T. Methods Enzymol. 1994, 233, 112-117.
57
McDonagh, C.M.; Shields, A.M.; Mcevoy, A.K.; Maccraith, B.D.; Gouin, J.F. J. Sol-Gel Sci. Technol. 1998, 13(1/2/3), 207-211.
McEvoy, A.K.; McDonagh, C.M.; MacCraith, B.D. Analyst. 1996, 121, 785-788.
McNamara, K.P.; Nguyen, T.; Dumitrascu, G.; Ji, J.; Rosenzweig, N.; Rosenzweig, Z. Anal. Chem. 2001, 73(14), 3240-3246.
Meier, B.; Werner, T.; Klimant, I.; Wolfbeis, O.S. Sens. Actuators, B. 1995, 29(1- 3), 240-245.
Metintas, M.; Hillerdal, G.; Metintas, S. Eur. Respir. J. 1999, 13(3),523-526.
Meyer, T.J. Pure Appl. Chem. 1986, 58, 1193-1200.
Mitchell, T. O. 2006. Luminescence Based Measurement of Dissolved Oxygen in Natural Waters. http://www.hydrolab.com/ pdf/Hach%20Env%20White%20Paper%20-%20LDO.pdf (accessed 07/24/2007).
Molson Medical Informatics Sampler. http://www.mmi.mcgill.ca/ mmimediasampler2005/default.htm (accessed 07/24/2007), part of McGill University http://curriculum.mmi.mcgill.ca/ (accessed 07/24/2007).
Morse, P.D.; Swartz H.M. Magn. Res. Med. 1985, 2(2), 114-127.
Mossman, B.T.; Churg, A. Am. J. Respir. Crit. Care Med. 1998, 157(5), 1666-1680.
Mossman, B.T.; Gee, J.B.L. New Engl. J. Med. 1989, 320(26), 1721-1730.
Nappi, A.J.; Vass, E. Biochim. Biophys. Acta. 1997, 1336, 295-301.
Nejjari, A; Fournier, J; Pezerat, H; Leanderson, P. Br. J. Ind. Med. 1993, 50(6), 501-504.
Palekar, L.D.; Most, B.M.; Coffin, D.L. Environ. Res. 1988, 46(2), 142-152.
Paris, J. P.; Brandt, W.W. J. Am. Chem. Soc. 1959, 81, 5001-5002.
Park, E.J.; Reid, K.R.; Tang, W.; Kennedy, R.T.; Kopelman, R. J. Mater. Chem. 2005, 15, 2913-2919.
Patel, A.K; Hallett, M.B; Campbell, A K. Biochem. J. 1987, 248(1), 173-180.
58 Payra, P.; Dutta, P.K. Microporous Mesoporous Mater. 2003, 64, 109-118.
Proteggente, A.R.; Rehman, A.; Halliwell, B.; Rice-Evans, C.A. Biochem. Biophys. Res. Commun. 2000, 277, 535-540.
Regen, S.L.; Czech, B.; Singh, A. J. Am. Chem. Soc. 1980, 102(21), 6638-6640.
Rose, A.L.; Waite, T.D. Anal. Chem. 2001, 73, 5909-5920.
Roundhill, D.M. Photochemistry and Photophysics of Metal Complexes; Plenum Press: New York, 1994.
Ruda, T.A.; Dutta, P.K. Environ. Sci. Technol. 2005, 39(16), 6147-6152.
Sarkar, N.; Das, K.; Nath, D.N; Bhattacharyya, K. Langmuir. 1994, 10(1), 326-329.
Scaduto, R.C. Free Radical Biol. Med. 1995, 18(2), 271-277.
Stanton, M.F.; Layard, M.; Tegeris, A.; Miller, E.; May, M.; Morgan, E.; Smith, A. J. Natl. Cancer Inst. 1981, 67(5), 965-975.
Tang, Y.; Tehan, E.C.; Tao, Z.; Bright, F.V. Anal. Chem. 2003, 75(10), 2407-2413.
Valentine, R.; Chang, M.J.W.; Hart, R.W.; Finch, G.L.; Fisher, G.L. Environ. Health Persp. 1983, 51, 357-368.
Valko, M.; Morris, H.; Cronin M.T.D. Curr. Med. Chem. 2005, 12, 1161-1208. van Zijverden, M. Dissertation. http://igitur-archive.library.uu.nl/dissertations/ 1954471/c8.pdf (accessed 07/24/2007), part of Igitur: Utrecht Publishing and Archiving Services. http://www.igitur.nl/igiturarchief/index.php (accessed 07/24/2007).
Wagner, J.C.; Skidmore, J.W.; Hill, R.J.; Griffiths, D.M. Br. J. Cancer. 1985, 51, 727-730.
Weitzman, S.A.; Graceffa, P. Arch. Biochem. Biophys. 1984, 228(1), 373-376.
Winterbourn, C.C. Toxicol. Lett. 1995, 82/83, 969-974.
World Health Organization 2006. Elimination of asbestos-related disease. http://www.who.int/occupational_health/publications/asbestosrelateddiseases.pdf (accessed 07/24/2007).
59 Xu, H.; Aylott, J.W.; Kopelman, R.; Miller T.J.; Philbert, M.A. Anal. Chem. 2001, 73, 4124-4133.
Xu, W.; McDonough, R.C.; Langsdorf, B.; Demas, J.N.; DeGraff, B.A. Anal. Chem. 1994, 66, 4133-4141.
Yao, Y.; Zhang, M.; Shi, J.; Gong, M.; Zhang, H.; Yang, Y. Mater. Lett. 2001, 48(1), 44-48.
Yeh, T.S.; Chu, C.S.; Lo, Y.L. Sens. Actuators, B. 2006, 119, 701-707.
Yildiz, G.; Demiryurek, A.T. J. Pharmacol. Toxicol. Methods. 1998, 39(3), 179-184.
Zeolyst International. http://www.zeolyst.com/html/faq.asp (accessed 07/24/2007).
60
CHAPTER 2
FENTON CHEMISTRY: TREATMENT OF FeIII-EXCHANGED ZEOLITIC
MINERALS TREATED WITH ANTIOXIDANTS
2.1 Introduction
The development of pulmonary fibrosis, bronchogenic carcinoma, and malignant mesothelioma has been linked to environmental and occupational exposure of mineral particulates, in particular, asbestos (Mossman and Gee, 1989; Mossman et al, 1990).
Lung problems are often uncovered much later after exposure because asbestos has a latency period of more than twenty years (Guthrie, 1992). Asbestos is comprised of a group of naturally occurring hydrated silicates that encompasses six major fiber types each with different chemical compositions, morphologies, and durabilities (Veblen and
Wylie, 1993). The toxicity of asbestos fiber has been related to physico-chemical properties such as shape, dimension, biopersistence, chemical composition, surface charge, and solubility (Mossman and Gee, 1989; Mossman et al, 1990). Asbestos toxicity has been correlated with both framework iron and surface iron, which may or may not be chelatable. These iron species can participate in Fenton chemistry to produce
61 free radicals (Kamp and Weitzman, 1999; Kane et al., 1996). In order to determine what
causes fibers to be toxic, it is important to develop relationships between surface characteristics and bioactivity.
The aluminosilicate minerals erionite and mordenite provide good models for
studying bioactivity because of their contrasting toxicity. Mordenite, a natural mostly
non-fibrous mineral zeolite, is known to have little biological activity (Guthrie, 1992).
Erionite, a natural fibrous mineral zeolite, is known to be the most carcinogenic fiber,
more so than asbestos (Wagner et al., 1985; Eborn and Aust, 1995). These zeolites contain little to no iron in their framework or as ion-exchangeable ions. However, since
- zeolites are crystalline aluminosilicates with interlinked SiO4 and AlO4 tetrahedral
frameworks, they are capable of ion-exchange (Auerbach et al., 2003). The toxicity of
erionite is proposed to be due to ion-exchanged iron that participates in Fenton chemistry
(Eborn and Aust, 1995; Hardy and Aust, 1995). A similar ion-exchange process can also occur on mordenite, which is benign, suggesting that other factors must play a role in toxicity. Fach et al. examined the formation of reactive oxygen species (superoxide assay by luminol chemiluminescence) by macrophages upon phagocytosis of erionite versus mordenite and found that they were of similar magnitude (2002). However, the Fenton reactivity of Fe(II)- exchanged erionite and mordenite were different, with the former showing a factor of two increase in hydroxyl radical production for comparable iron
loading levels. It was proposed that the differences in surface structure cause iron to
62 coordinate differently which influences the redox properties (Fach et al, 2002). Fach et al. also reported that erionite is significantly more cytotoxic and mutagenic than mordenite in the presence of Fe(II) ions (2003).
After inhalation, it is likely that the iron exchanged onto the mineral in the lung is
in the Fe(III) form and will need to be reduced prior to exhibiting Fenton reactivity. The
lung has several lines of defense against inhaled minerals, including the epithelial lung
lining fluid (LLF) (Cross et al., 1994; Slade et al., 1993). When fibers interact with LLF,
the surface reactivity may be modified (Brown et al., 1998; Brown et al., 2000). Two of
the main antioxidants found in LLF, ascorbic acid and glutathione, are able to reduce the
surface iron of inhaled particles (Fubini and Arean, 1999; Gutteridge et al., 1996). The
purpose of this study is to compare the Fenton reactivity of Fe (III) exchanged erionite
and mordenite after treatment with glutathione or ascorbic acid.
2.2 Experimental Section
2.2.1 Iron Incorporation into Zeolites
Iron was exchanged into the zeolites following a method adapted from previously
published procedures (Badran et al., 1977; Evmiridis, 1986). Ground erionite or
mordenite (1 g, Minerals Research, Clarkson, NY) was shaken for 15 minutes with
10 mL of 0.1 M sodium chloride (NaCl) [Aldrich, Milwaukee, WI]. The grinding of the
minerals was necessary to break up the agglomerates. The samples were centrifuged, the
supernatant discarded, and the process repeated twice. A slurry was prepared with the
zeolites using 5 mL water. A solution of 0.1 M ferric chloride (FeCl3) [Fluka, Fairlawn,
63 NJ] and 0.5 M potassium thiocyanate (KSCN) [Aldrich, Milwaukee, WI] was made, and the ferric thiocyanate product was extracted with an equal volume of diethyl ether (J.T.
Baker Chemical, Phillipsburg, NJ). 10 mL of the extract was added to the zeolite slurry and shaken for 60 minutes. The samples were washed with water and dried at 45oC.
2.2.2 Zeolite Crystallinity
X-ray powder diffraction patterns were monitored with a Rigaku Geigerflex
diffractometer using nickel-filtered CuKα radiation (40 kV and 25 mA).
2.2.3 Surface Iron Measurements
Dab-4Br was synthesized following a method previously published (Daniels et al.,
1978). The first step in Dab-4Br synthesis involved dissolving 0.016 mol of 1,4-
diazabicyclo[2.2.2]octane (Dabco) in 16 mL of dimethyl sulfoxide and heating to 45oC with stirring. 0.016 mol of 1,4-dibromobutane was added dropwise, and the temperature was kept below 70oC. The temperature was held at 60oC for 1 hour, then increased to
90oC for 6 hours. The solution was then cooled overnight, washed with ether, methanol,
and ether once more. The solid was allowed to dry, and the product was a white/faint
yellow, nontacky powder. The compound’s structure was confirmed via nuclear
magnetic resonance spectroscopy (NMR) with peaks at 2.1, 3.8, and 4.2 ppm, which
matched the published data (Daniels et al., 1978).
64 A solution of Dab-4Br was made by adding 0.5 g of Dab-4Br to 40 mL of
nanopure water. 8 mL of the Dab-4Br solution was added to 0. 4g of the ion-exchanged
zeolite and shaken for 40 hours. The supernatant (1 ml) was placed in a 5 mL volumetric
flask, and the pH was adjusted with 1.5 mL of a 1 M sodium acetate buffer. 1 mL of 0.03
M ascorbic acid (AA) was added, then 1 mL of 0.5 mM phenanthroline was added. The color was allowed to develop for 2 hours, the solution was diluted to 5 mL and then the absorbance at 512 nm was measured on a Shimadzu UV-2501PC spectrophotometer
(Shimadzu, Columbia, MD).
2.2.4 Exposure of Zeolites to Antioxidants
A 10mL portion of a 200 μM AA (Aldrich, Milwaukee, WI) or glutathione (GSH)
(Sigma-Aldrich, St. Louis, MO) solution was added to 1 g of the dried, iron-exchanged
zeolite. The sample was placed on a wrist shaker for 30 minutes, centrifuged, the
supernatant discarded, and 10mL of a solution of 0.353 mM hydrogen peroxide (Fluka,
Fairlawn, NJ) and 0.934 mM dimethyl sulfoxide (Acros, NJ) were added to the zeolite.
The sample was placed on a wrist shaker for 1 hour. The sample was then centrifuged,
and the supernatant was utilized for hydroxyl radical detection.
2.2.5 Detection of Hydroxyl Radicals
Hydroxyl radicals were detected spectrophotometrically by a previously published
procedure (Babbs and Steiner, 1990). A 2 mL portion of the supernatant from the exposed zeolite was pH adjusted to ~2.5 by adding 0.3 mL of 0.01 N HCl (Fluka,
65 Fairlawn, NJ). To the supernatant, 0.2 mL of 0.02 M Fast Blue BB dye (Fluka, Fairlawn,
NJ) solution was added. The 0.2 mL came from 1mL of the Fast Blue BB dye extracted
with 1 mL chloroform. Note that the Fast Blue BB dye is light sensitive. The dye was
reacted with the solution for 10 minutes in the dark, then 1.5 mL of a 3:1 toluene-butanol
solution was added to stop the reaction. The solution was vortexed for 2 minutes, then
centrifuged for 3 minutes. 1 mL of the upper phase was transferred to a cuvette and 2 mL
of butanol-saturated water was added, vortexed for 30 seconds, and centrifuged for 3
minutes. 0.9 mL of the upper phase was placed in a cuvette with 1.5 mL of 3:1 toluene-
butanol, and 0.1 mL of pyridine-glacial acetic acid. The latter reagent is added to increase color stability. The spectrum was recorded on a Shimadzu UV-2501PC spectrophotometer (Shimadzu, Columbia, MD) from 520-330 nm. The absorbance peak at 425 nm represented the diazo complex of interest. Figure 2.1 is the absorption
spectrum. The band with a maximum around 320 nm (not shown), is photo-oxidized
species from the Fast Blue BB dye. The reference was 2.4 mL of the 3:1 toluene-butanol
solution. A calibration curve as shown in Figure 2.2 was made with methanesulfinic acid
sodium salt so that known concentrations could be determined.
2.2.6 XPS Studies of Surface Iron
X-ray photoelectron spectroscopy (XPS) was taken of Fe(III)-exchanged
mordenite and erionite with a Kratos Ultra Axis X-ray photoelectron spectrometer
(Kratos Analytical, Shimadzu, Chestnut Ridge, NY). An Al Kα source (1486.7 eV) was
66 used to collect the spectrum. The analyzer was operated at a pass energy of 20 eV. The
XPS sample was in pellet form. Binding energies were referenced to the Si 2p peak at
102.8 eV from the silicon (IV) in the minerals.
2.3 Results
2.3.1 Ion-exchange process
Typically, ion-exchange of zeolites is done from aqueous solutions of ions. In a
previous study, it was noted that iron exchange of zeolite-Y with Fe(II) led to loss of crystallinity (Fach et al., 2002). Loss in crystallinity has also been noted by Sato et al., where it is pointed out that the Si:Al of the zeolite framework played a role in how much ion-exchange procedures will influence the crystallinity of the resulting zeolite (2003).
Because the type and valence of the cation influence the overall acidity of the zeolite, it is
an important influence on crystallinity of the zeolite. Also, other factors such as Si:Al
ratio, additives, synthesis method of the zeolite, and the nature of the cationic exchange
alter zeolite acidity. A loss in zeolite crystallinity, or any change in zeolite framework,
will influence the Fenton reactivity. Since in the present study, we are dealing with ion- exchange of a highly charged ion, Fe(III), an alternative procedure of ion-exchange was used to avoid problems with loss of zeolite crystallinity (Evmiridis, 1986). The process
involves contacting an aqueous slurry of the zeolite with an ethereal solution of
(3-n)+ Fe(SCN)n . The iron thiocyanate that phase transfers into water is hydrolyzed into
2+ + 4+ Fe(OH) , Fe(OH)2 , Fe2(OH)2 , or other polymeric species. The smaller aquated iron
species rapidly ion-exchange into the zeolite.
67 The powder diffraction pattern for the initial erionite sample (Figure 2.3a) matches well with the known spectra but does show slight impurity due to sodium aluminum silicate hydrate (Fach et al., 2002). The powder diffraction pattern for the initial mordenite indicates impurities of quartz and feldspar (Figure 2.3b). Powder X-ray diffraction patterns of the erionite after ion-exchange (Figure 2.4) shows no significant difference from the original X-ray diffraction patterns, illustrating no loss in zeolite crystallinity after ion-exchange. Data for mordenite is not shown since it has a higher
Si:Al, making it the least prone to lose crystallinity (Sato et al., 2003).
2.3.2 Surface Iron
Iron can ion-exchange into the porous framework as well as onto the zeolite
surface. Since the reaction with AA and GSH will only occur on the surface, it was
necessary to determine the level of surface iron. The amount of surface iron present on
the zeolites was estimated via ion-exchange with a cationic polymeric molecule, Dab-
4Br, that will not enter the zeolites’ pores because of its size. Dab-4Br was synthesized,
and the structure was verified through NMR. As is shown in Figure 2.5, the peak at
4.1 ppm represents the protons on the Dabco unit, the peak at 3.8ppm represents the
proton on the α–carbon and the peak at 2.0 ppm represents the proton on the β–carbon.
Integration of the peaks shows the area ratio to be 1.0:0.33:0.34 (Dabco unit protons: α–
carbon protons : β–carbon protons), representing the proper ratio of the protons on the molecule. Dab-4Br is shown below.
68 + + NNCH2 4 Br- Br- x
Dab-4Br is composed of cylindrically shaped units with a length of 8.7 Å and a diameter of 6.1 Å (Daniels et al., 1978). Erionite’s surface has a network of 8-membered rings with dimensions of 3.6 x 5.2 Å (Breck, 1974). Mordenite has 4, 8 and 12-membered rings, with the largest having dimensions of 7.0 x 6.7 Å (Breck, 1974). Therefore, Dab-
4Br will only exchange with surface iron and not penetrate into the zeolite.
The released iron was converted to an iron-phenanthroline complex and quantified via absorption spectroscopy. The reaction for the iron complexation is as follows:
2+ N N N 2+ Fe Fe + 3 (C12H8N2) Æ l (1) N N N
69 The absorbance was measured at 512 nm. The absorption spectra is shown in Figure 2.6.
The amount of surface iron for mordenite and erionite were determined to be 1.8x101 μg and 8.8x10-1 μg iron per gram of zeolite, respectively. Note that these numbers cannot be
directly compared with a previous study on iron loaded erionite and mordenite, which
used ferrous ammonium sulfate solutions for ion-exchange (Fach et al., 2002).
X-ray photoelectron spectra (XPS) of erionite and mordenite in the Fe2p region are shown in Figure 2.7. Mordenite shows an increase in the binding energy for iron, i.e., the Fe2p peak is at 711.5 eV for mordenite as compared to 710.8 eV for erionite.
2.3.3 Fenton Reactivity
The Fenton activity of the iron-exchanged mineral fibers involved reduction of
surface iron with GSH or AA, followed by treatment with hydrogen peroxide in the
presence of dimethyl sulfoxide. The methane sulfinic acid that is created is converted to
a diazosulfone compound and quantified by absorption spectroscopy. The overall
process is outlined in reactions (2)-(3) (Babbs and Steiner, 1990).
• • CH3-S(=O)CH3 + OH Æ CH3-S(=O)OH + CH3 (2)
dimethyl sulfoxide methane sulfinic acid (MSA)
MSA (aq phase) + diazonium salt (fast blue BB dye; aq phase) Æ (3)
H+ + diazosulfone (organic phase; yellow; 425 nm)
70 Where:
fast blue BB dye
OCH2CH3 O
+ - C NH N2 Cl .1/2 ZnCl2
CH CH O 3 2
diazosulfone
OCH CH O 2 3
C NH N2 SO2 CH3
CH CH O 3 2
The concentration of antioxidant was chosen to be 200μM in order to simulate the typical
concentration in LLF (Greenwell et al., 2002). Figure 2.8a shows the total hydroxyl radical generated, and Figure 2.8b shows the hydroxyl radical production normalized to the amount of surface iron. The error bars in Figure 2.8 are the result of two measurements. Data from Figure 2.8b is summarized in Table 1, along with the percent conversion of Fe(III) to Fe(II). The conversion assumes a 1:1 stoichiometry between hydroxyl radical and iron; that is, the assumption is made that each incorporated iron ion
71 produces a hydroxyl radical. Several trends are apparent. First, for both erionite and
mordenite, the amount of hydroxyl radical follows the order: AA > GSH > no
antioxidant. Note that in the absence of antioxidant, the Fe(III) can also be reduced to
Fe(II) by the hydrogen peroxide. The reactions of ferric iron with hydrogen peroxide, to
produce ferrous iron are as follows (Dunford, 2002):
3+ III 2+ + Fe + H2O2 Æ [Fe OOH] + H (4)
[FeIIIOOH]2+ Æ Fe2+ + OOH• (5)
Furthermore, the iron(III)hydroperoxo complex may produce the hydroxyl radical via the
following reaction:
[FeIIIOOH]2+ Æ [FeIVO]2+ + OH• (6)
The reaction of ferric iron with hydrogen peroxide is known to be optimal at a pH 1-3
(Gallard et al., 1999). Second, the surface-iron normalized data shows that erionite produces significantly more hydroxyl radicals than mordenite. Third, the amount of hydroxyl radical relative to the surface iron varies from about ~ 1% in the case of mordenite with no antioxidant added to ~ 106% with AA treated erionite. Thus, in the case of erionite all of the surface iron is being reduced by AA.
72 2.4 Discussion
The extreme carcinogenicity of erionite in vivo has been correlated with its ability to acquire iron in the lung (Eborn and Aust, 1995). As noted, the amount of surface iron acquired by mordenite is about two orders of magnitude larger than that of erionite.
Typically, erionite contains more total iron than mordenite due to the fact that mordenite has more structural defects blocking the channels (Fach et al., 2002). Also, the Si:Al is
3.6 for erionite and 5.1 for mordenite leading to the fact that erionite has higher ion- exchange capacity. Because the structural defects in mordenite block channels, the surface iron ion-exchange will not be affected by the defects. Therefore, the higher level of surface iron seen with mordenite needs an alternate explanation. Firstly, the mode of unfractionated erionite is 10 μm while the mode of unfractioned mordenite is 3.7 μm
(Fach et al., 2002). The mordenite has an overall large external surface area attributing to the larger amount of surface iron measured. Secondly, taking into account the pore sizes of erionite and mordenite, the Dab-4Br has a diameter of 6.1 Å which can more readily access mordenite’s outer-most, largest pores which are 7.0 x 6.7 Å while the largest erionite pores are 3.6 x 5.2 Å. This allows for more iron to be exchanged from the surface of mordenite, since some of the iron is likely to be coming from inside of the outer most pores.
As noted, the quantification of iron can not be compared to previous data on iron exchanged mordenite and erionite because the exchange procedure utilized ferrous ions
(Fach et al., 2002). The mechanisms for ferrous and ferric iron may be very different.
Ferrous ions appear to bind through a process of ion-exchange while ferric ions may bind
73 through a precipitation or crystallization process (Shen, 2000). Eborn and Aust noted
that when performing ion-exchange of erionite with ferric and ferrous iron, the amounts
of ferric iron exchanged onto the erionite were greater than that of ferrous iron (1995).
Because of the different binding mechanisms, a direct comparison between ferric and
ferrous iron ion-exchange can not be made.
In a previous study, the bioactive properties of mordenite and erionite, focusing
primarily on the oxidative burst upon phagocytosis by rat lung macrophages and their
Fenton reactivity by an assay similar to that described in this paper, was compared (Fach et al., 2002). It was reported that for the same mass of mineral loading, the oxidative burst increased with decreasing particle size, but was relatively independent of the mineral chemical composition. On the other hand, the Fenton reactivity for erionite was higher than that of mordenite by about a factor of two for comparable Fe(II) loadings.
The conclusion from this study was that the different mineral structures led to different
surface coordination of the iron.
. Erionite and mordenite, with compositions (Na2, Ca)3.5 K2[Al9Si27O72] 27H2O and
. Na8[Al8Si40O96] 24H2O, respectively, contain no iron (Hardy and Aust, 1995; Breck,
1974). However, when these mineral fibers are inhaled, they can imbibe iron via ion- exchange (Hardy and Aust, 1995). Though levels of iron in vivo are typically low due to
sequestration by metal binding proteins, tissue injury can lead to metal ion release (Carr
and Frei, 1999). Bronchoalveolar lavage (BAL) has also shown the presence of
74 chelatable and redox-active iron in the LLF (Gutteridge et al., 1996). The iron acquired
in biological systems will most likely be in the Fe(III) form. Antioxidants present in the
LLF can reduce iron, promoting Fenton reactivity as shown in reactions (3) through (4).
reductant + Zeolite-Fe(III) Æ oxidized reductant + Zeolite-Fe(II) (3)
• - Zeolite-Fe(II) + H2O2 Æ OH + OH + Zeolite-Fe(III) (4)
Considerable controversy exists regarding the mechanism of reaction (4), especially
concerning the generation of free hydroxyl radicals. Alternative mechanisms propose the
formation of Fe(IV) or Fe(V) containing species, which can act as the oxidant (Lloyd et
al., 1997; Winterbourn, 1995; Dunford, 2002; Kremer, 2003). In this study, we represent the active oxidant in Fenton chemistry as hydroxyl radicals.
Debate regarding which molecules play the role of reductant in reaction (3) also
exists (Chaudière and Ferrari-iliou, 1999; Young and Woodside, 2001). The dominant
antioxidants in the LLF that can reduce Fe(III) are AA and GSH (Hardy and Aust, 1995;
Brown et al., 2000). AA plays an important biological role as a co-substrate in the
reduction of transition metal ions for hydroxylase and oxygenase enzymes. In vitro
studies have shown that AA can exhibit a pro-oxidant role in biological fluids depending
on the nature of the iron complex, e.g. in the presence of iron-EDTA, but not if present as
ferrous ammonium sulfate (Winterbourn, 1981; Minetti et al., 1992).
75 Both AA and GSH can reduce iron; however, AA should be able to more easily reduce ferric iron. Reduction potentials describe the ability of a species to be reduced, or acquire electrons. As can be seen from the reduction potentials, ascorbic acid would be more likely to reduce iron than glutathione. Note that the ascorbate is completely in its conjugate base form of AH- at physiological pH (Chaudière and Ferrari-iliou, 1999). The reactions below show the antioxidants acting as prooxidants to reduce Fe(III) to Fe(II).
Fe(III) + e- Æ Fe(II) Eo = ~0.7V
. - (AH /AH ) Em7 = ~0.3V
- • AH + Fe(III) Æ AH + Fe(II) Erxn = ~0.4V
Fe(III) + e- Æ Fe(II) Eo = ~0.7V
. (GS /GSH) Em7 = ~0.9V
• GSH + Fe(III) Æ Fe(II) + GS Erxn = ~ -0.2V
Besides reducing Fe(III) to Fe(II), AA can also react with mineral surfaces and bring about dissolution, as has been noted for silica and asbestos (Hardy and Aust, 1995;
Fenoglio et al., 2000; Martra et al., 2003). Since our goal was to compare surface reactivity of erionite and mordenite, the reaction with the antioxidants was limited to 30 minutes, and the antioxidant was removed prior to studying Fenton reactivity. Another
76 reason to remove the antioxidants from the system prior to the Fenton reaction was to
avoid recycling the redox states of iron and the reaction of the thiyl and ascorbate radicals
with oxygen which promotes superoxide formation.
As shown in Table 1, AA is more effective in promoting Fenton chemistry than
GSH and hydrogen peroxide for both minerals. AA is a stronger reducing agent than
GSH since it will reduce the GSH radical. Therefore, AA is expected to reduce a larger
fraction of the zeolites’ surface iron. The production of hydroxyl radicals from a
2+ Fe /EDTA/H2O2 system in the presence of GSH and AA ascorbic acid has been reported
(Nappi and Vass, 1997).
In the present study, erionite was about an order of magnitude more Fenton
reactive than mordenite for all reductants, normalized to the surface iron levels. This is
quite striking, especially considering that the Fenton reactivity of Fe(II) showed only a
factor of two increase for erionite (Reaction (4)) [Fach et al., 2002]. The rationale for
normalizing Fenton activity with respect to surface iron is that the iron present on the
minerals’ surface is in the molecular form as coordinated Fe3+. As long as the Fe3+ is isolated, the normalized reactivity provides a direct comparison of how the different iron coordination on the two mineral surfaces is influencing the Fenton activity. Another reason for examining normalized reactivity is that all iron on the minerals was introduced via ion-exchange (i.e. no framework iron). The high Fenton activity of erionite is also consistent with the report that, normalized to the iron content, erionite was about 200 times more reactive than crocidolite for DNA single-strand breaks (Eborn and Aust,
1995; Hardy and Aust, 1995).
77 Both natural mineral samples have impurities, quartz and feldspar for mordenite
and an aluminosilicate for erionite. Are these impurities relevant in the Fenton chemistry?
First, the level of impurities in erionite are quite small, as evidenced from the X-ray
diffraction patterns in Figure 2.3 (impurities are marked as ‘S’), and thus expected to play
a minor role. The major impurity in mordenite is quartz and should have no ion-exchange
binding capacity.
We are proposing that the Fenton chemistry is occurring on the surface of the
minerals. After the treatment with GSH or AA, the solution is removed prior to the
hydroxyl radical assay using the solid, and thus all solublized iron is removed. The
importance of surface and mobilized iron from crocidolite asbestos in hydroxyl radical
generation has been noted (Ghio et al., 1992; Shukla et al., 2003; Lund and Aust, 1990).
Asbestos toxicity has also been correlated with specific surface iron binding sites. The
bioactivity of the iron is dependent on the coordination geometry (Shukla et al., 2003).
For Reaction (3), the ligands determine the reduction potential of iron and will be
the controlling feature for the efficacy with which reductants can reduce Fe(III). The ligand environment also plays a critical role in Reaction (4). For an inner sphere mechanism of Reaction (4), availability of ligation sites through which hydrogen peroxide can bind to Fe(II) is necessary. For example, with ligands such as EDTA and citrate, which upon iron complexation still leave available coordination sites, the complexes are redox active. On the other hand, strongly coordinated chelating ligands, such as desferrioxamine B and ferrozine render the iron redox inactive (Winterbourn,
1995).
78 The XPS data provide insight on the coordination environments for Fe(III)-
exchanged erionite and mordenite. For mordenite, the peaks are shifted to higher binding
energies, e.g. for the Fe 2p3/2 by 0.8 eV. The shift indicates that in mordenite, the Fe(III)
is coordinated by stronger electron-withdrawing ligands. Amongst the minerals, the closest parallel is the comparison between Fe2O3 and goethite, FeOOH. The presence of
hydroxyl ligands in goethite leads to increased binding energy for both Fe2p and Fe3p
peaks, e.g. the Fe2p by about 0.85 eV (McIntyre and Zetaruk, 1977). The higher metal
(2p) binding energy is observed also for other metal oxides and hydroxides, such as those
of copper, cobalt, zinc, and chromium (McIntyre and Cook, 1975). The possible ligands
on the zeolite mineral surface are water, hydroxide, O from T-O-T (T= Si, Al) framework
and broken T-O bonds on the surface. A simple correlation between binding energy and
reducibility cannot be made because, as mentioned above, the coordination environment
also determines the reduction kinetics.
Figure 2.9 is a surface structure rendering of mordenite and erionite showing the different rings, 4-, 5-, 8-, and 12-membered rings for mordenite and 4-, 6- and 8-
membered rings for erionite. Certainly, the coordination environment and ligands of
these mineral surfaces are different, leading to different redox characteristics. “Free”
metal ions are thought to be very low in vivo due to their sequestration by various metal
binding proteins. Bleomycin-detectable iron in BAL from normal humans (control) was
found to be 0.132 μmol/L ± 0.019 (7.392 μg/L), whereas total iron content in the BAL of
a normal human was 0.212 μmol/L ± 0.038 (11.872 μg/L) (Gutteridge et al., 1996). At
these low levels of iron in vivo, the amounts of iron on the mineral surfaces is expected to
79 be small, and the present study suggests that, under these circumstances, erionite will exhibit enhanced Fenton activity compared to equivalent levels of iron on mordenite.
2.5 Conclusion
The focus of this study has been to compare the Fenton reactivity of Fe(III)- exchanged erionite and mordenite, two zeolitic minerals well known for very different toxicities upon inhalation. Since the Fenton reaction requires the presence of Fe(II), which is most likely to occur in biological systems by antioxidants like AA and GSH in the LLF fluid reducing Fe(III), we studied the Fenton reaction of Fe(III)-erionite and mordenite in the presence of these antioxidants. The Fenton reactivity as measured by the reaction of dimethyl sulfoxide with the hydroxyl radicals (produced from hydrogen peroxide) shows an order of magnitude increase for erionite as compared to mordenite if normalized to the amount of surface iron per gram of zeolite. The extent of Fenton reactivity increased in the order AA > GSH > no antioxidant (note that hydrogen peroxide was present in all cases). The higher Fenton reactivity of Fe(III)- erionite over mordenite must be related to the surface structure which leads to different coordination states and therefore, different redox properties.
80
No Antioxidant Glutathione Ascorbic Acid μmol OH./ Percent μmol OH./ Percent μmol OH./ Percent μg Fe per g Conversion μg Fe per g Conversion μg Fe per g Conversion zeolite of Fe(III) zeolite of Fe(III) zeolite of Fe(III) Mordenite 1.9 x 10-4 1% 3.2 x 10-4 2% 1.4 x 10-3 8% ± 2.5 x 10-6 ± 2.9 x 10-5 ± 7.2 x 10-5 Erionite 2.6 x 10-3 15% 1.2 x 10-2 67% 1.9 x 10-2 106% ± 3.8 x 10-5 ± 6.2 x 10-4 ± 1.2 x 10-3
81
Table 2.1: Comparison of hydroxyl radical production in the presence of different antioxidants.
81 0.6
0.5
0.4
0.3 Absorbance 0.2
0.1
0.0 340 360 380 400 420 440 460 480 500 Wavelength, nm
Figure 2.1: Absorption spectrum illustrated Fenton chemistry. The presence of hydroxyl radical production is noted with the peak at 425 nm.
82 UV-VIS Cal Curve for Surface Iron (baseline corrected by ave of 650-750nm)
0.16
0.14 Standards 0.12 Mordenite 0.10 Erionite
0.08
0.06
0.04
Abs at 512nm 0.02
0.00
-0.02
-0.04 0.000 0.002 0.004 0.006 0.008 0.010 0.012 0.014 0.016
Conc, mM
Figure 2.2: Surface iron calibration curve made with methanesulfinic acid sodium salt.
83 (A) E E
E E
E E E E E Intensity E EE E E E E E E E S EE E
10 20 30 40 50 2 Theta
(B)
M M M
M F F M M M M Intensity M F MM M
10 20 30 40 50
2 Theta Figure 2.3: X-ray powder diffraction patterns: (A) erionite (E) with sodium aluminum silicate hydrate (S) and (B) mordenite (M) with quartz (Q) and feldspar (F) impurities. 84
10 20 30 40 2 Theta
Figure 2.4: X-ray powder diffraction pattern of erionite after iron ion-exchange showing no loss in zeolite crystallinity.
85
(A)
(ii) (iii) (iii) (ii) + + NNCHC 2 – CH2 – CH2 – CH2
Br - Br - x (i)
(B)
(i)
(ii) (iii)
Figure 2.5: NMR spectra of Dab-4Br. The protons in the structure (A) are correlated with the NMR spectra (B).
86
UV-VIS Spectra for Surface Iron Analysis Erionite and Mordenite
0.12 MORDENITE
0.10 M041-1 M041-2 0.08 E041-1 E041-2
0.06 intensity
0.04
0.02 ERIONITE
0.00 300 400 500 600 700 800
wavelength, nm
Figure 2.6: Absorption spectra for surface iron analysis. Mordenite and erionite were each run in duplicate.
87
725.1 711.5
2p1/2 2p3/2
Mordenite
710.8 724.3 Intensity Erionite
740 730 720 710 Binding Energy, eV
Figure 2.7. X-ray photoelectron spectra of Fe(III) exchanged erionite and mordenite (Fe 2p region).
88 (A) Total Hydroxyl Radical Production
0.03 Erionite Mordenite AA 0.025 AA 0.02
0.015 GSH GSH gram sample 0.01 mol hydroxyl radical /
μ Blank Blank 0.005
0
(B) Hydroxyl Radical Production Relative to Surface Iron
Erionite Mordenite
0.02 AA
0.015 GSH
0.01
AA mol hydroxyl radical / μ 0.005 GSH Blank
g surface iron per gram of sample g surface iron per gram of sample Blank μ 0
Figure 2.8. Comparison of (A) total hydroxyl radical production and (B) hydroxyl radical production normalized to surface iron in the presence of glutathione (GSH), ascorbic acid (AA) or hydrogen peroxide only (Blank).
89
(A)
Mordenite
(B)
Erionite
Figure 2.9. The aluminosilicate ring structures in mordenite (A) and erionite (B).
90 2.6 References
Auerbach, S.M.; Carrado, K.A.; Dutta, P.K. Handbook of Zeolite Science and Technology; Marcel Dekker, Inc: New York, 2003.
Babbs, C.F.; Steiner, M.G. Method Enzymol. 1990, 186, 137-147.
Badran, A.H.; Dwyer, J.; Evmerides, N.P.; Manford, J.A. Inorg. Chim. Acta. 1977, 21, 61-64.
Breck, D.W. Zeolite Molecular Sieves; Wiley-Interscience Publication: New York, 1974.
Brown, D.M.; Beswick, K.S.; Bell, K.S.; Donaldson, K. Ann. Occup. Hyg. 2000, 44(2), 101-108.
Brown, D.M.; Roberts, N.K; Donaldson, K. In Vitro. 1998, 12, 15-24.
Carr, A.; Frei, B. FASEB J. 1999, 13, 1007-1023.
Chaudière, J.; Ferrari-iliou, R. Food Chem. Toxicol. 1999, 37, 949-962.
Cross, C.E.; van der Vliet, A.; O’Neill, C.A.; Louie, S.; Halliwell, B. Environ. Health Persp. 1994, 102(10), 185-191.
Daniels, R.H.; Kerr, G.T.; Rollmann, L.D. J. Am. Chem. Soc. 1978, 100, 3097-3100.
Dunford, H. B. Coordin. Chem. Rev. 2002, 233/234, 311-318.
Eborn, S.K.; Aust, A.E. Arch. Biochem. Biophys. 1995, 316(1), 507-514.
Evmiridis, N. P. Inorg. Chem. 1986, 25, 4362-4369.
Fach, E.; Kristovich, R.; Long, J.F.; Waldman, W.J.; Dutta, P.K.; Williams, M. V. Environ. Int. 2003, 29, 451-458.
Fach, E.; Waldman, W.J.; Williams, M.; Long, J.; Meister, R.K.; Dutta, P.K. Environ. Health Persp. 2002, 110, 1087-1096.
Fenoglio, I.; Martra, G.; Coluccia, S.; Fubini, B. Chem. Res. Toxicol. 2000, 13, 971- 975.
Fubini, B.; Arean, O. Chem. Soc. Rev. 1999, 28, 373-381.
91
Gallard, H.; De Laat, J.; Legube, B. Water Research. 1999, 33(13), 2929-2936.
Ghio, A.; Zhang, J.; Piantadosi, C. Arch. Biochem. Biophys. 1992, 298(2), 646-650.
Greenwell, L.; Moreno, T.; Jones, T.; Richards, R. Free Radical Bio. Med. 2002, 32(9), 898-905.
Guthrie Jr., G.D. Am. Mineral. 1992, 77, 225-243.
Gutteridge, J.M.C.; Mumby, S.; Quinlan, G.J.; Chung, K.F.; Evans, T.W. Biochem. Bioph. Res. Co. 1996, 220, 1024-1027.
Hardy, J.A.; Aust, A.E. Chem. Rev. 1995, 95, 97-118.
Kamp, D.W.; Weitzman, S.A. Thorax. 1999, 54, 638-652.
Kane, A.B.; Boffetta, P.; Wilbourn, J.D., Eds. Mechanisms of Fibre Carcinogenesis; IARC: Lyon, France, 1996.
Kremer, M.L. J. Phys. Chem. A. 2003, 107, 1734-1741.
Lloyd, R.V.; Hanna, P. M.; Mason, R. P. Free Radical Bio. Med. 1997, 22, 885-888.
Lund, L.G.; Aust, A.E. Arch. Biochem. Biophys. 1990, 278(1), 61-64.
Martra, G.; Tomatis, M.; Fenoglio, I.; Coluccia, S.; Fubuni, B. Chem. Res. Toxicol. 2003, 16, 328-335.
McIntyre, N.S.; Cook, M.G. Anal. Chem. 1975, 47(13), 2208-2213.
McIntyre, N.S.; Zetaruk, D.G. Anal. Chem. 1977, 49(11), 1521-1529.
Minetti, M.; Forte, T.; Soriani, M.; Quarisima, V.; Menditto, A.; Ferrari, M. Biochem. J. 1992, 282, 459-465.
Mossman, B.T.; Gee, J.B.L. New Engl. J. Med. 1989, 320(26), 1721-1730.
Mossman, B.T.; Bignon, J.; Corn, M.; Seaton, A.; Gee, J.B.L. Science. 1990, 247, 294-301.
Nappi, A, J.; Vass, E. Biochem. Biophys. Acta. 1997, 1336, 295-301.
92 Sato, K.; Nishimura, Y.; Matsubayashi, N.; Imamura, M.; Shimada, H. Microporous and Mesoporous Mater. 2003, 59, 133-146.
Shen, Z.; Bosbach, D.; Hochella, Jr., M.F.; Bish, D.L.; Williams, Jr., M.G.; Dodson, R.F.; Aust, A.E. Chem. Res. Toxicol. 2000, 13, 913-921.
Shukla, A.; Gulumian, M.; Hei, T.; Kamp, D.; Rahman, Q.; Mossman, B.T. Mossman, B.T., Ed. Free Radical Bio. Med. 2003, 34(9), 1117-1129.
Slade, R.; Crissman, K.; Norwood, J.; Hatch, G. Exp. Lung Res. 1993, 19, 469-484.
Veblen, D.R.; Wylie, A.G. In Reviews in Minerology; Gutherie, G.D.; Mossman, B.T. Eds; Bookcrafters, Inc.: Chelsea, MI. 1993, 28, 61-137.
Wagner, J.C.; Skidmore, J.W.; Hill, R.J.; Griffiths, D.M. Br. J. Cancer. 1985, 51, 727-730.
Winterbourn, C.C. Toxicol. Lett. 1995, 82/83, 969-974.
Winterbourn, C.C. Biochem. J. 1981, 198, 125-131.
Young, I.S.; Woodside, J.V. J. Clin. Pathol. 2001, 54, 176-186.
93
CHAPTER 3
ENTRAPMENT OF IONIC SPECIES IN SILICEOUS ZEOLITE: SHIP-IN-A-
BOTTLE SYNTHESIS OF TRIS(2,2’-BIPYRIDYL) RUTHENIUM(II) AND ITS
EMISSION QUENCHING BY O2 IN DIFFERENT ENVIRONMENTS
3.1 Introduction
Optical oxygen sensors are used in a variety of applications such as monitoring dissolved oxygen in water, soil, industrial fermentation and sterilization processes, pressure sensitive paints, and intracellular systems (Amao, 2003). Monitoring dissolved oxygen in various difficult environments is the final application focus of this research, especially intracellular monitoring.
Currently, intracellular monitoring is usually performed via a fiber optic device, which has a probe immobilized in a sol-gel or polymer matrix allowing for efficient oxygen diffusion. The main disadvantages of these sensors are (1) the fiber optic itself often causes cellular damage due to its size (Xu et al., 2001), (2) the probe can leach from the support matrix, especially when a hydrophobic support matrix is used with an ionic dye, which will cause a drift in signal and may lead to toxicity during intracellular
94 monitoring (Cao et al., 2004), and (3) the matrix does not provide a homogeneous
environment for the dye; therefore, Stern-Volmer plots are typically non-linear
(Carraway et al., 1991).
Besides fiber optic probes, phospholipid vesicles, also known as liposomes, can
entrap a sensor in a phospholipid membrane. Traditional phospholipid sensors severely
lack chemical and environmental stability. The loss in structure from fusion with cellular
membranes is the main disadvantage of these intracellular oxygen sensors (McNamara et
al., 2001). In order to combat the instability of the spheres, polymerizable phospholipids
have been utilized. The polymerizable phospholipids are difficult to work with and still
show leaching of the dye (Cheng & Aspinwall, 2006; Regen et al., 1980). The method
with more promise involves polymerizing monomer units in the bilayer region of the
sphere (Cheng & Aspinwall, 2006; Hotz & Meier, 1998). These sensors still need tested
in cells to determine their stability.
Finally, PEBBLE (probes encapsulated by biologically localized embedding)
sensors are being utilized to monitor intracellular dissolved oxygen concentrations.
PEBBLE sensors offer small size (20-200 nm in diameter) with minimal cell
perturbation, fast response, high selectivity, and good reversibility. PEBBLE sensors
typically consist of a probe immobilized in a sol-gel matrix or an organically modified
silicate matrix that forms a nanometer sized sphere which is introduced to the cells
allowing them to remain viable (Clark et al., 1998). The main disadvantage of PEBBLE sensors relates to their inability for reliable quantitation due to leaching of the dye from the matrix, which limits their stability and sensitivity (Xu et al., 2001).
95 The research performed seeks to improve upon the currently used methods for
intracellular monitoring. Zeolites are crystalline aluminosilicates which can serve as a
support matrix. Siliceous zeolite-Y is comprised only of silicon and oxygen making it
highly hydrophobic. Also, it has pores which allow for permanent entrapment of
molecules. Other work previously determined that utilizing siliceous zeolite-Y as a
support matrix provided enhanced oxygen sensing when compared to conventional
zeolite-Y because the intrazeolitic water in conventional zeolite-Y blocks oxygen
transport to the sensitizer (Coutant et al., 2003). Solubility and transport of oxygen
through the support matrix used in optical oxygen sensors is crucial in synthesizing a sensitive sensor. Utilizing a siliceous zeolite as a support matrix allows for dye molecules to be securely entrapped and not leach from the support matrix. Futhermore, the zeolite matrix will provide a crystalline, homogeneous environment for the sensitizer molecules. Hence, it is likely that the gas molecules will readily diffuse to the probe allowing for a homogeneous response and a linear Stern-Volmer plot. Therefore, the support matrix used in this research is siliceous zeolite-Y.
2+ Tris(2,2’-bipyridyl) ruthenium(II) (Ru(bpy)3 ) is known to serve as a dye for
optical oxygen sensing due to its long excited state lifetime, good quantum yield, and
large Stokes shift. The ruthenium compound has a diameter of about 12 Å, while the
supercages of the siliceous zeolite-Y have diameters of about 13 Å with 7 Å windows.
Therefore, a “ship-in-a-bottle” synthetic scheme is necessary in order to build the
ruthenium compound inside of a zeolite supercage.
96 2+ Loading Ru(bpy)3 into conventional zeolite-Y (Si:Al = 2.6) is a simple
procedure involving ion-exchange of a ruthenium precursor then reacting excess 2,2’-
bipyridine in order to construct the final ruthenium complex in the supercages (Meier et
al., 1995). However, an ion-exchange “ship-in-a-bottle” synthesis can not be used with siliceous zeolite-Y because in the absence of aluminum the zeolite is neutral and hence not able to ion-exchange species. Entrapment of the sensitizer inside the supercages of siliceous zeolite-Y has been performed using a “ship-in-a-bottle” technique, but by impregnation of the ruthenium precursor into the zeolite. However, the problem with this method was obtaining a high loading level of sensitizer molecules per supercage. The loading level of the sensitizer in the siliceous zeolite-Y achieved in the past was about 1 ruthenium complex in 10,000 supercages (Payra & Dutta, 2003).
The objectives of this research were as follows: (1) to devise a new synthetic
strategy seeking to improve the loading level of an ionic sensitizer molecule in hydrophobic siliceous zeolite-Y, (2) utilization of a glucose oxidase assay to monitor
changes in dissolved oxygen concentration, (3) to examine the ability to detect changes in
dissolved oxygen concentration using gas saturated water experiments as a calibration for a glucose oxidase assay, (4) to present supporting data for detecting oxygen concentrations in live macrophage cells.
97 3.2 Experimental (methods, techniques and materials studied)
3.2.1 Synthesis of Siliceous Zeolite-Y
The dealumination method utilized was adapted from Beyer et al. and Ray and
Nerheim (1985, 1988). A fixed bed vertical reactor was created and utilized in the dealumination procedure (Figure 3.1). The reactor allowed for a thermocouple to
monitor the temperature of the zeolite bed. Commercial zeolite-Y from Union Carbide
was cleaned via sodium ion-exchange and calcination at 500oC to remove organics. The
zeolite was then placed into the reactor and dehydrated by activation at 400oC (ramped at
10oC/minute) under flowing dry nitrogen for 6 hours. The temperature was then lowered
to 200oC (ramped at 10oC/minute). Nitrogen was flowed through a reaction vessel filled
with silicon tetrachloride (SiCl4), and the SiCl4 saturated nitrogen gas was flowed over
the zeolite for 4 hours. The sample was purged with dry nitrogen for at least 4 hours at
200oC. The zeolite sample was then cooled to room temperature, washed with distilled
water until no more chloride was present, and dried at 80oC in a vacuum oven.
o After the SiCl4 treatment, the sample was stirred with 3.8 M HCl at 90 C for 3
hours. The sample was then centrifuged and washed several times with distilled water.
After the acid leaching, the zeolite was placed in a reactor and heated to 700oC while
steam passed through the sample for 16 hours. Then nitrogen was passed over the sample
until cooled to room temperature.
The siliceous zeolite-Y was characterized via solid state nuclear magnetic
resonance spectroscopy (SSNMR), X-ray powder diffraction (XRD), scanning electron
microscopy (SEM), and Brunauer-Emmett-Teller (BET) for surface area measurements.
98 Siliceous zeolite-Y was also obtained from a commercial source (HSZ-390HUA).
HSZ-390HUA, Tosoh Corporation, Tokyo, Japan, was reported to have a SiO2/Al2O3 ratio of >200. The commercial siliceous zeolite-Y was characterized via the same methods as the synthesized zeolite. Parameters of the commercially available and synthesized siliceous zeolite-Y were compared.
3.2.2 Synthesis of Ru(bpy)Cl2(CO)2
The synthesis of the ruthenium precursor utilized, RuCl2(CO)2(bpy), is adapted
from Haukka et al. (1995). The reagents and solvents used were p.a. grade and the synthesis was performed in an inert atmosphere. 1 g of Ru(CO) 3Cl2 was mixed with
80mL dry tetrahydofuran (THF), then refluxed for 2.75 hours. 0.75 g 2,2’-bipyridine was mixed with 20 mL of THF, then was slowly added to the [Ru(CO) 3Cl2] 2 solution via
syringe/septa. The solution was refluxed for 45 minutes, cooled to room temperature,
then placed in a refrigerator for a 2-3 day precipitation. After the precipitate was
collected, a second precipitation was performed and collected after 2-3 more days in the
refrigerator. Recrystallization was performed with chloroform and ethanol. The
precursor was characterized via infrared (IR) and NMR spectroscopy.
99 3.2.3 Synthesis of Ru(bpy)3-Siliceous Zeolite-Y
Two main methods of loading the ruthenium complex into the supercages of the
siliceous zeolite-Y were tried in order to achieve high loading levels.
Method 1:
The method utilized by Payra & Dutta was adapted (2003). The first step
involved loading the above ruthenium precursor into the siliceous zeolite-Y supercage.
The siliceous zeolite-Y was activated on a vacuum line at 500oC. The activated Si-ZY
was then placed in acetonitrile with the ruthenium precursor, and the solvent was
evaporated overnight in the vacuum oven. In order to form the tris(2,2’-bipyridyl)
ruthenium complex excess 2,2’-bipyridine was added. The excess 2,2’-bipyridine was
added to the above powder with ethanol as a solvent, and the solvent was evacuated in a
vacuum oven overnight. The sample was then placed on a vacuum line at 150oC for
2+ about 3 days. Synthesis of Ru(bpy)3 inside of the siliceous zeolite-Y supercage was then performed by solid state reactions performed under vacuum at elevated temperatures.
Method 2:
The ruthenium precursor was dissolved in acetonitrile with the activated zeolite
and dried (rotovap) to promote the ruthenium precursor’s mobility into the pores of the
zeolite. The zeolite was then placed under vacuum and heated to 150oC, again to act as
an activation step and promote precursor mobility into the pores. A vacuum was pulled,
100 the tube was sealed off of the vacuum line, and the sample was heated in an oven to
200oC with an excess amount of 2,2’-bipyridine. The 2,2’-bipyridine was found to sublime at this temperature, which allows it to have great mobility to get into the pores and form the tris(2,2’-bipyridyl) complex.
The progress for both synthetic methods was monitored via diffuse reflectance.
Additions of 25% more 2,2’-bipyridyl were added until the diffuse reflectance spectra showed no change in absorbance around 450nm. A general idea of how well the synthetic procedures worked was determined from diffuse reflectance absorption spectroscopy (DRUV). After utilizing that data to determine which scheme provides the most promising results, different loading levels were made.
The ruthenium precursor was added to the siliceous zeolite-Y in ratios of 1 ruthenium complex per 10 supercages and 5 ruthenium complexes per zeolite supercage.
Elemental analysis of the ruthenium loaded siliceous zeolite-Y was performed by
Galbraith laboratories via ICP-OES for the most promising samples.
3.2.4 PDMS - Ru(bpy)3-Siliceous Zeolite-Y Membrane
The polydimethyl siloxane (PDMS) [RTV 615A and RTV 615B] used was from
GE Silicones, Wilton, CT, USA. Component A contains the prepolymer with long
PDMS chains and terminal vinyl groups. Component B is the cross-linker consisting of several SiH groups per polymer chain. The procedure was adapted from Alvaro et al. and
2+ Payra and Dutta (1998, 2003). The Ru(bpy)3 - siliceous zeolite-Y was dehydrated under vacuum at 110oC for 12 hours. Typically, 0.01 g dehydrated zeolite was mixed with
101 1.6 g of dry hexane and sonicated for 4-6 hours. 0.15 g of RTV 615A and 0.01 g of dry
hexane was added to the dry hexane treated siliceous zeolite then sonicated for 5 hours.
0.015 g of RTV 615B was then added and sonicated for an additional 30 minutes. The
mixture was cast onto a glass plate and kept in a vacuum at 60oC overnight. Film
thickness was examined via profilometry using a Veeco Dektak3 ST.
3.2.5 Leaching
The higher loading level sample, synthesized at 5 ruthenium complex per
supercages, was tested for leaching of the ruthenium complex from the zeolite or the
polymer film. After making the PDMS film, it was placed in water in the dark at room temperature. The fluorescence of the solution was tested nine times over a ten week period to determine if any of the ruthenium complex or zeolite had leached from the
PDMS film. A SPEX Fluorolog spectrophotometer was utilized, and the spectra was taken by exciting the sample at 450nm and collecting emission spectra at 500-800 nm.
3.2.6 Emission Quenching
Emission was monitored with a SPEX Fluorolog spectrophotometer. Two
different samples were tried for calibration of the glucose oxidase assay. First, gas
saturated water experiments were performed by pressing the Ru-loaded zeolite powder
into a pellet. The pellet was then placed diagonally into a cuvette leaning against a glass
slide. Second, the Ru-loaded zeolite powder was dispersed in a PDMS film and then cast
on a glass slide. The PDMS films were then placed diagonally in a cuvette, filled with
102 deionized (DI) water, so that the sample was immobilized in a vertical position. The
sample was excited at 450 nm using a 450 nm band pass filter and an emission scan was
performed collecting data from 500-800 nm using right angle mode. Refer to Figure 3.2
to see a partial view of the sample set-up.
Gas was bubbled through the water for 15 minutes, then a measurement was
taken. The gases used were oxygen (99.99%), zero grade air, a 60/40 mixture of
oxygen/nitrogen, and nitrogen (99.999%). The area of the peak emission at 610 nm was
analyzed in order to determine percent quenching.
The glucose oxidase assay, which was utilized to monitor changes in dissolved
oxygen concentration, involved the PDMS film on the glass slide being placed in a
cuvette filled with a saturated solution of glucose oxidase (EC 1.1.3.4 Type II-S from
Aspergillus niger with a specific activity of 47,200 units per gram of solid) made at a
concentration of 4.7 U/mL in phosphate buffer at pH = 6.5 (SIGMA, St. Louis, MO).
Small volumes of a 17.5 mM β-D-glucose (SIGMA, St. Louis, MO) were added in increments to produce solutions of the following glucose concentrations: 0.25 mM,
0.50 mM, 0.75 mM, 1.0 mM. The different concentrations of glucose correspond to different amounts of dissolved oxygen being consumed. Again, the area of the peak emission at 610 nm was analyzed in order to determine percent quenching.
103 3.2.7 Monitoring Emission in Live Cells
3.2.7.1 Isolation and differentiation of human monocyte-derived macrophages (MDM)
[performed by Prof. Waldman and his research group]
Peripheral blood mononuclear cells (PBMCs) were separated by Ficoll-Hypaque
(Histopaque, Sigma) density gradient centrifugation from buffy coats purchased from the
American Red Cross as previously described (Waldman et al., 1992). (Because complete
donor anonymity is a strict condition of this arrangement, no Institutional Review Board human subjects protocol is required, as specified by the NIH and OSU IRB guidelines.)
To promote monocyte differentiation into the macrophage phenotype, PBMCs were aspirated from the plasma/Histopaque interface, washed three times in phosphate buffered saline solution (PBS, buffered to pH = 7.4), suspended in RPMI 1640 (GIBCO), and supplemented with 10% pooled normal human serum. The suspended cells were transferred to Teflon plates where they were incubated for 5-7 days at 37oC in a humidified
atmosphere of 5% CO2/95% air. The cells were then transferred to Lab-Tek II 4-well
sterile chambered borosilicate coverglass slides (Nalge Nunc International) for confocal
microscopy, or Optilux 96-well microtiter plates (Falcon) for luminol assay, and
incubated 24 hours prior to removal of non-adherent cells as previously described
(Waldman et al., 2007). Monocyte-derived macrophages (MDM) prepared in this
manner routinely marked 90-95% positive for CD14 with undetectable levels of CD3+ T cell contamination as determined by immunofluorescence flow cytometry.
104 3.2.7.2 Treatment of MDM with Ruthenium Sensor Particles
Tris(2,2’-bipyridyl) ruthenium(II) loaded siliceous zeolite-Y particles were
sterilized by steam autoclave and suspended in phosphate buffered saline (PBS) at a
concentration of 5mg/mL. Following 2-3 days incubation of MDM in chamber slides,
sensor particles were sonicated and added to cell cultures at a concentration of 6 μg/cm2
surface area. Slides were incubated for 24 hours to allow internalization of particles.
Residual sensor was removed. DAPI was applied at 1:20,000 in medium and incubated
for at least 2 hours. DAPI was washed off after incubation and fresh media was applied.
To induce oxidative burst, zymosan (50 μg/cm2, Sigma) was added immediately prior to
observation by laser scanning confocal microscopy.
3.2.7.3 Luminol Assay: Oxidative Burst Optimization
To determine the optimal stimulant to induce oxidative burst, phorbol 12- myristate 13-acetate or zymosan, luminol assays were adapted from those described by
Long et al. (2005). Following 2-3 days incubation of MDM in chamber Optilux
microtiter plates, sensor particles were sonicated and added to cell cultures at a
concentration of 6 μg/cm2 surface area. Plates were incubated for 24 hours to allow
internalization of particles, then washed free of uninternalized particles with PBS. To
each of triplicate wells was added 100 μL serum-free culture medium, 100 μL luminol
(5-amino-2,3-dihydro-1,4-phthalazinedione, sodium salt, Sigma) for a final concentration of 0.5 mM, and either phorbol 12-myristate 13-acetate or zymosan. Triplicate control wells contained culture medium and luminol but no stimulant. All solutions were at 40C
105 and plates were kept on ice during addition of reagents. Luminescence was measured
immediately on a Top Count Scintillation and Luminescence counter (Packard),
o incubated for 10 minutes at 37 C in a humidified atmosphere of 5% CO2/95% air, then
counted again. This process was repeated over a period of 80 minutes. Luminescence indices for each time point were determined by dividing the mean luminescence values of
the triplicate control wells from those generated by the stimulant.
3.2.7.4 Confocal Fluorescence Microscopy
All luminescent images were obtained via a Leica TCS SP2 confocal fluorescence
microscope. The tris(2,2’-bipyridyl) ruthenium (II) loaded zeolite was excited at 488 nm
with an argon ion laser, while the DAPI was excited at 350 nm with a UV laser. Cells
were warmed to 37oC on the microscope stage and held under a total carbon dioxide
(CO2) atmosphere. Emission was monitored between 600-650 nm for the ruthenium
complex, 400-480 nm for the nuclear stain, and collected via a photomultiplier tube.
Confocal fluorescence microscopy was first used to determine if the sensor was internalized by the macrophages by scanning through the z-axis. The emission of the
phagocytosed ruthenium-loaded zeolite particles was monitored under varying dissolved
oxygen concentrations to determine a response. The intracellular dissolved oxygen
concentrations were varied by holding the cells under an air atmosphere and then a CO2
atmosphere. When the total CO2 atmosphere was used, a 10 minute exposure time was
initially given before adding the zymosan. The total CO2 atmosphere was maintained
during all measurements. A series of measurements typically lasted one hour per well
106 when including the initial 10 minute saturation period for the CO2 atmosphere experiments. Also, an experiment was performed using autoclaved zymosan, and one using unautoclaved zymosan. The zymosan (50 μg/cm2) was added immediately before the scan had started in order to induce oxidative burst. Scans were taken every 4 minutes for 20-28 minutes. Data was analyzed with the Leica Confocal Software, where specific cells, known as regions of interest, were selected and the luminescence intensity per cell was provided for both the nuclear stain and the ruthenium complex. For data analysis, all completely visible cells (i.e. cells that were fully in the frame being monitored and did not move out of frame during the scan) in the region where the scan was performed were selected and used in the final data analysis.
3.3 Results
3.3.1 Synthesis
3.3.1.1 Synthesis of Siliceous Zeolite-Y
Siliceous zeolite-Y was synthesized in the laboratory successfully. The BET surface area measurement determined the surface area to be 640 m2/g. The surface area before dealumination was 768 m2/g. The SSNMR can be seen in Figure 3.3A, which depicts a single peak centered around -107.8ppm. The XRD spectra showed a crystalline structure after dealumination as can be seen in Figure 3.4. SEM of the lab synthesized siliceous zeolite-Y exhibited a particle size of close to 1μm (data not shown).
107 The siliceous zeolite-Y from Tosoh was determined to have a surface area of
775 m2/g from a BET surface area measurement. As seen in Figure 3.3B, the SSNMR of the Tosoh sample displays one strong peak at -107.8 ppm. A single peak around
-107 ppm is expected for tetrahedrally bound silicon with no aluminum present
(Si(OSi)4)) (Engelhardt & Michel, 1987). The absence of additional peaks, as well as a broad baseline around the main peak of interest, shows the absence of hydroxyl group
defect sites, so no amorphous material is present. The XRD shows that the siliceous
zeolite is crystalline (data not shown). The SEM of the commercial siliceous zeolite-Y in
Figure 3.5 shows the particle size to be about 0.5 μm.
After performing the analysis of both siliceous zeolite-Y samples, it was decided
to use the commercially available siliceous zeolite-Y for future experiments. The loss of
surface area for the siliceous zeolite-Y sample dealuminated in the laboratory could be
attributed to a loss in crystallinity. Furthermore, the dealumination set-up often
encountered problems, causing the process to take up significant amounts of time.
108 3.3.1.2 Synthesis of Ru(bpy)Cl2(CO)2
Two-dimensional structures of the three Ru(bpy)Cl2(CO)2 stereoisomers are depicted below.
N N 2+ -Cl Ru Cl-
O C C O
(I) trans complex (trans Cl ligands, cis CO ligands)
N N 2+ O C Ru Cl-
C O Cl-
(II) cis complex (cis Cl ligands, cis CO ligands)
109 N N 2+ O C Ru C O
-Cl Cl-
(III) trans complex (trans CO ligands, cis Cl ligands)
The trans(CO) complex (III) is not discussed here because it is not a likely product (de
Klerk-Engels et al., 1993). Characterization of Ru(bpy)Cl2(CO)2 using IR and NMR
spectroscopy in literature reports two different sets of data. The differences have been
thought to be due to an impurity of RuCl3(CO)(bpy) (Collomb-Dunand-Sauthier, 1991);
however, the more likely possibility is the presence of both the trans (I) and cis (II)
isomer of the complex (Kelly et al., 1982; Haukka et al., 1995; Chardon-Noblat et al.,
2000; Black et al., 1982). The trans complex (I) has the formation of the Cl ligands being
trans to each other and the CO ligands being cis to each other, while the cis complex (II)
has both the Cl and CO ligands cis to each other.
An IR spectra (in CH2Cl2) of one of the synthesized batches of Ru(bpy)Cl2(CO)2
can be seen in Figure 3.6, showing two ν(CO) stretching bands at 2065(vs) and
2003(vs) cm-1. Multiple batches were created, and the IR spectra taken for other batches
of synthesized ruthenium complex resulted in similar spectra that sometimes varied by
110 -1 1 1-2 cm . Our results of H NMR for bpy (in CDCl3) also varied between synthetic
batches. A typical spectra is shown in Figure 3.7, with peaks at δ 9.8(d), 8.9(d), 8.1(t),
7.8(t), 7.6(t) and several overlapped peaks from 8.0 to 8.3 ppm (as seen in Figure 3.7).
The IR results show bands at frequencies typical of CO stretching, and two bands would be expected for both trans (I) and cis (II) conformations due to the bipyridyl group causing different environments for each CO. Chardon-Noblat et al. points out that the trans complex (I) and cis complex (II) both have similar IR bands; the trans complex IR bands are at 2063 and 1999 cm-1, and the cis complex bands are at 2067 and 1998 cm-1
(2000). Haukka et al. determined that the trans complex (I) and the cis complex (II) both
have very similar IR spectra also, with the bands at 2067 and 2003cm-1 for the cis
complex (II) and 2066 and 2003 cm-1 for the trans (I) complex (1995).
1 Haukka et al. showed the H NMR (in CDCl3) analysis for two different precipitates (1995). The first precipitate was mainly the trans complex (I) and has peaks at: 9.2 (d), 8.3 (d), 8.1 (t), 7.7 (t). The second precipitate was mainly the cis complex (II) and shows peaks at: 9.7 (d), 8.8 (d), 8.2 (2 overlapped doublets), 8.1 (t), 8.0 (t), and 7.7
(t), and 7.5 (t). The NMR for both precipitates contained a variable amount of each other.
The synthesis was performed over 5 times in our laboratory, with each having its own
unique mixture of the trans (I) and cis (II) complexes. The proton NMR of one of the
synthesized ruthenium complex products shows its most distinct peaks at 9.8, 8.9, 8.1,
7.8, and 7.6 ppm, which represent the cis complex (II). Overall, the synthesized
ruthenium product exhibits properties of both the cis and trans complex, and varied from
batch to batch depending on how well the recrystallization was performed.
111 3.3.1.3 Synthesis of Ru(bpy)3-Siliceous Zeolite-Y
Method 1:
This synthetic scheme was tried multiple times. Method 1 involves the use of
solvents in the steps of adding the ruthenium precursor as well as the excess 2,2’-
bipyridine. Furthermore, the steps during synthesis were performed on a dynamic
vacuum line. The differences between the trials dealt with the step where excess 2,2’-
bipyridine was added to the sample. Often, longer exposure times and higher amounts of
the 2,2’-bipyridine was added to see if the diffuse reflectance spectra would show some
improvement, denoted as an increase in the metal to ligand charge transfer band at 450
nm. As can be seen from Figure 3.8A, the peak at around 450nm was small after trying
to optimize the reaction, measuring about 0.1 Kubelka-Munk units. Optimization
entailed the addition of the ruthenium precursor to the zeolite and heating to 150oC on the vacuum line for 24 hours, then adding the excess 2,2’-bipyridine on the vacuum line at
190oC for 1-2 days and repeating the addition of the 2,2’-bipyridine every 1-2 days until
the peak at 450 nm stopped increasing. The small absorption at 450 nm implies that the
complex was not formed at a high amount since that absorption represents the metal to
ligand charge transfer band of the complex. Therefore, a different method was tried.
Method 2:
The synthetic scheme for the synthesis of the ruthenium complex inside of the
siliceous zeolite-Y supercages can be seen in Figure 3.9. Overall, this synthetic scheme
provided the best results for a higher loading of the ruthenium complex inside of the
112 zeolite supercages. This scheme avoided using a dynamic vacuum line. The synthesis
was monitored via diffuse reflectance (Figure 3.8B, C). The peak at 450 nm increased significantly to 0.4 and 1.9 Kubelka-Munk units when compared to Method 1 at 0.1
Kubelka-Munk units. The peak around 285 nm represents the 2,2’-bipyridyl ligand transitions, and the peak with an absorbance of 45 0nm represents the metal to ligand charge transfer (MLCT) band (Demas and Taylor, 1979). Elemental analysis of
+2 Ru(bpy)3 in the siliceous zeolite-Y was done in duplicate with one result being a
loading level of less than 1 ruthenium complex per 17 supercages and the other being less
than 1 ruthenium complex per 25 supercages for the lower loading level sample (B). The
higher loading level sample showed a loading level of 1 ruthenium complex per 7
supercages (A). Figure 3.10 shows the variation in color of the zeolite powder samples
depending on the loading level.
3.3.1.4 Synthesis of PDMS - Ru(bpy)3-Siliceous Zeolite-Y Film
2+ The Ru(bpy)3 loaded siliceous zeolite-Y powder sample was pressed into a
pellet and utilized for emission quenching measurements. The emission quenching (data
not shown) had problems with reproducibility and stability. To combat the problem of
poor reproducibility/stability seen with the pellet the powder was then dispersed in a
PDMS film, which was anticipated to allow for better gas diffusion and hence more reproducible/stable results. Siloxane polymers, such as PDMS, are extensively used when creating optical oxygen sensors due to high permeability to oxygen and biological
inertness (Xu et al., 1994; Demas et al., 1999; Bossi et al., 1999; Klimant and Wolfbeis,
113 1995; Carraway et al., 1991). Films with thickness varying from 50-150μm via
profilometry experiments were made. Figure 3.10 shows a photograph to depict
differences in the PDMS films made with both the high (~1 complex per 7 supercages)
and low (~1 complex per 25 supercages) loading level of ruthenium loaded siliceous
zeolite.
3.3.2 Sample Optimization
3.3.2.1 PDMS Film Stability - Leaching Experiments
Leaching experiments determined that the maximum amount of leaching was seen
after four weeks and correlates to 1.36% of the zeolite leaching into the aqueous solution.
Figure 3.11 shows that during the testing the PDMS film actually swelled. Figure 3.12 shows the leaching trend over time. A maximum occurs after four weeks, which will be discussed later. Treatment of the ruthenium loaded zeolite after synthesis consists of an extensive Soxhlet extraction, as well as stirring in a 0.1 M solution of NaCl overnight.
Because a Soxhlet extraction was performed for several weeks, it is concluded that no ruthenium complex was adsorbed onto the zeolite particle which could then leach into the solution giving a positive response.
+2 3.3.2.2 Emission Quenching of High versus Low Sample Ru(bpy)3 Loading
A comparison between the two loading levels (less than 1 ruthenium complex per
25 supercages and 1 ruthenium complex per 7 supercages) of the ruthenium complex in
the siliceous zeolite-Y was performed by investigating the quenching between oxygen
114 and nitrogen saturated water of the probe encapsulating zeolite in a PDMS film. The
average of a series of quenching experiments performed for the higher loading level
sample (A) was 33.0% with a standard deviation of 2.43 and a 7.36% error, while the
average quenching seen for the lower loading level sample (B) was 46.4% with a
standard deviation of 0.87 and a 1.88% error (Figure 3.13). In other words, the higher
loading level sample shows poorer quenching which may be expected for a higher loading level sample (Sykora et al., 1999; Laine and Lanz, 1996). Because the higher loading level sample may experience problems due to intermolecular interactions, the
lower loading sample (B) was chosen for use in the glucose oxidase and gas saturated
water experiments. Confirmation of this is found in Figure 3.14 where the absorption and
emission maximum value has shifted in the higher loading level sample, implying that
intermolecular interactions are taking place in the higher loading level sample.
3.3.3 Emission Quenching
+2 Emission quenching of intrazeolitic Ru(bpy)3 by dissolved oxygen was
examined. Calibration was performed with gas saturated water solutions while a glucose
oxidase assay was utilized to check the validity of the calibration. The sample was in the form of a PDMS film during testing.
115 3.3.3.1 Calibration
Calibration was performed by monitoring the emission quenching of the PDMS film in gas saturated water. The results expressing percent quenching (standard deviation) are shown in the Table 3.1. The overall quenching seen between a solutions of nitrogen degassed water and oxygen degassed water is about 41%. The Stern-Volmer plot is shown in Figure 3.15 and is linear. The R2 value is 0.98. Furthermore, the Stern-
Volmer constant (Ksv) is 0.89 atm-1.
3.3.3.2 Glucose Oxidase Assay
A method to test the calibration is necessary, and the glucose oxidase assay was used for testing the gas saturated water calibration since it allows for small changes in oxygen concentration compared to gas saturated water solutions. In the presence of the catalyst glucose oxidase, β-D-glucose is oxidized to D-glucono-1,5-lactone and hydrogen peroxide using molecular oxygen as the electron acceptor as seen in the following reaction:
glucose oxidase β-D-glucose + O2 D-glucono-1,5-lactone + H2O2
O O O O O O (1) O O O O O O
116 Experimental results are summarized in Table 3.2, which shows the concentration
of glucose, the concentration of oxygen in solution, and the percent quenching. The
concentration of oxygen in solution was determined by using Henry’s law. Henry’s law
is as follows:
c = p k
where p is the partial pressure of the solute above the solution, k is the Henry’s Law
constant, and c is the concentration of the solute in the solution. For oxygen, Henry’s
Law constant is 1.3x10-3 mol atm-1 (dm3)-1. Table 3.2 shows a sample calculation. The
total amount of quenching, determined by analyzing the area of the emission curves, seen between 1.0mM and 0mM glucose was 41.7%. Using the equation from the gas saturated water Stern-Volmer plot, the concentrations of oxygen that would occur with the glucose oxidase assay were estimated. The percent error between the Stern-Volmer predicted value and the value calculated by using stiochiometric calulations based on amount of glucose is determined. The data between the two experiments is compared in Table 3.3 and does compare well, showing an average percent error of less than 5%.
3.3.3.3 Intracellular Oxygen Monitoring
Human monocyte-derived macrophages were utilized for intracellular monitoring
experiments. Peripheral blood mononuclear cells were separated, differentiated, washed,
and transferred to Teflon plates for incubation. After incubation the cells were
117 transferred to either chambered borosilicate coverglass slides for confocal microscopy or
96-well microtiter plates for luminol assays. The macrophage cells were exposed to the
ruthenium loaded zeolite, allowing them to be taken in via phagocytosis. The
fluorescence from zeolite particles phagocytized by the macrophages was observed with
confocal fluorescence microscopy. After scanning through the z-plane it was determined that the particles were truly phagocytosed by the cell. Figure 3.16 shows the image obtained with the confocal microscope when exciting the nuclear stain simultaneously
with the ruthenium complex.
In order to determine the sensing ability of the ruthenium loaded zeolite in cells,
the cells were placed under conditions which increase oxygen consumption and
sometimes decrease available oxygen. The phagocytosed ruthenium-loaded zeolite
monitored intracellular oxygen concentrations. The extracellular environment is
comprised of the media surrounding the cells, which was either air saturated or perfused with CO2 for the entire experiment. In the case of the total CO2 atmosphere, the cells
were kept under a CO2 atmosphere so that oxygen was not available to the cells
extracellularly. An oxidative burst was induced, which is known to consume oxygen.
Therefore, as time progresses the amount of intracellular oxygen would decrease, and in
the case of the CO2 atmosphere would not be available to replenish oxygen to the cell.
In order to optimize the oxidative burst conditions, thereby maximizing the
amount of reactive oxygen species being formed, a luminol assay was utilized. Luminol
detects extracellular hydroxyl radicals and superoxide. 12-phorbol 13-myristate acetate
was compared with zymosan to determine which one was more effective. As can be seen
118 in Figure 3.17, zymosan shows a more intense oxidative burst by a factor of about 3.
Zymosan was utilized at a concentration of 1 mg/mL (78 μg/cm2 for the luminol assay;
50 μg/cm2 for the confocal experiments). A luminol assay was also performed with
autoclaved zymosan which showed an oxidative burst about half as intense as unautoclaved zymosan (data not shown).
Confocal fluorescence microscopy was first performed under a total air
atmosphere. Overall, no increase in signal was seen when looking at the data with the absence and presence of added zymosan (data not shown). Then the cells were held under a total CO2 atmosphere, and intracellular oxygen was monitored in the absence and
presence of added zymosan, both autoclaved and unautoclaved zymosan. The
unautoclaved zymosan led to cellular malfunction noted by the background of the image becoming polluted with the DAPI stain, and hence nuclear components of the cells (data not shown). The autoclaved zymosan did not lead to such cell failure and is therefore the only data described within. Figure 3.18 shows the confocal images taken initially and at the end of 20 minutes before the addition after the addition of zymosan. Figure 3.19 shows data after the addition of zymosan. Figure 3.20 and Figure 3.21 plot the change in intensity over time. A ratio between the ruthenium emission and the DAPI emission is taken in order to normalize for cell movement. Each data point represents the average of
6-10 cells (regions of interest as described in the experimental section). The experiment was done in triplicate; however, only two datasets are shown. This is because one of the datasets was collected from a well that was exposed to a 100% carbon dioxide atmosphere for about 1 hour longer than the other wells, which led to loss in cell activity.
119 The large error bars are attributed to cell movement, which was seen during
measurements. During cell movement, especially depending on the loading level, the
illumination may fall on different cellular planes that exhibit different amounts of sensor
and stain being illuminated. Overall, the intensity increase for the cells exposed to
zymosan in a total carbon dioxide atmosphere is about 2.7 fold, on average between the
two samples seen in Figures 3.20 and 3.21.
3.4 Discussion
3.4.1 Synthesis of Siliceous Zeolite-Y
3.4.1.2 Dealumination Treatments
Increasing the Si:Al is beneficial in order to minimize intrazeolitic water from
preventing oxygen’s access to the intrazeolitic sensitizer. Other research has utilized
zeolites that are dealuminated because of their thermal stability and improvement in
catalytic activity, which is beneficial in the petroleum cracking industry (Hriljac et al.,
1993). Three main types of dealumination have been utilized: (1) hydrothermal treatment of an ammonium-exchanged form of the zeolite, (2) chemical treatment of the zeolite in solution (i.e. acid, chelating agent, or salt) or in the vapor phase (i.e. SiCl4 or
other halides, F2), and (3) a combination of the above treatments.
Dealumination utilizing hydrothermal treatment typically involves having the
zeolite in the presence of steam at 500oC. During this treatment acidic protons are
formed by deammoniation, splitting the Si-O-Al bonds, and removing the tetrahedral
aluminum from a framework position into a non-framework position. This is believed to
120 leave a vacancy in the framework, while the aluminum is in interstitial cationic positions.
The silicon that reoccupies the tetrahedral framework positions may then come from not
only the surface or amorphorous parts of the structure, but also from the bulk causing
decreased crystallinity. During solution based chemical dealumination procedures,
aluminum is extracted from the zeolite in a soluble form and hydroxyl groups are formed
at the vacancies.
In 1980 a new dealumination method was introduced, which involved treatment
with SiCl4 vapor, the framework aluminum is filled in with silicon from the SiCl4, and the product, aluminum chloride (AlCl3), is volatilized or forms a stable
tetrachloroaluminum complex which can be removed from the zeolite by washing. The
reaction is as follows:
Na[AlO2(SiO2)x]+SiCl4 NaAlCl4 +(SiO2)x+1
SiCl Cl 4 Cl Cl + Si Na NaCl Na(AlCl4) O- O O Si Al Si Al Si Si
Treatment with SiCl4 does not cause vacancies in the framework, as in the hydrothermal
or acid extraction treatments. This allows for high levels of dealumination while having
nearly no loss in crystallinity.
121 Dealumination performed by combining the silicon tetrachloride treatment with
thermal treatment is known to produce a highly siliceous zeolite, nearly defect free and
still crystalline (Beyer, 2002). After chemical treatment, acid washing was performed in
order to remove any extraframework aluminum formed, which will increase the
hydrothermal stability of the zeolite (Beyer, 2002). After acid leaching, steaming was
performed to heal defect sites by removing hydroxyl nests and helping to reincorporate
any amorphous silica formed (Ray and Nerheim, 1988; Hriljac et al., 2003). Because
high quality siliceous zeolite-Y is available for purchase, experiments were performed
using the commercial powder, being sure to use the same sample batch for all
experiments.
3.4.1.2 Analysis of Zeolite Dealumination
A variety of methods were used in order to provide information as to the
effectiveness the dealumination process. The main issue with dealumination is loss in crystallinity. Information as to how well the zeolite structure stayed intact can be provided with XRD as well as BET measurements. The XRD in Figure 3.4 shows little to no amorphous material formation in the zeolite dealuminated in the laboratory. A loss in surface area after dealumination was noted, from 768 m2/g to 640 m2/g, alluding to the
fact that some of the structure was destroyed. Hence, the laboratory dealuminated zeolite
was not utilized in this research.
122 SSNMR is one of the most useful techniques in monitoring the dealumination of
zeolites. Both 27Al and 29Si SSNMR can provide valuable information related to the
extent of dealumination. When looking at the spectra for 27Al SSNMR, a single strong
peak at 61ppm represents the tetrahedral framework aluminum. After dealumination and
washing, the intensity decreases and another peak forms around 0ppm representing non-
framework octahedral aluminum. The SSNMR spectra for 29Si typically displays four
-5 peaks, which depend on the number of AlO4 tetrahedrally linked to the silicon. The
peaks represent Si(3Al), Si(2Al), Si(1Al), and Si(0Al) with chemical shifts of about -
88.5, -93.7, -99.2, and -105.0 ppm respectively. After dealumination only one peak around -107.0 ppm is expected. Note that the shift in Si(0Al) from -105.0 ppm to
-107.0 ppm is significant and arises from aluminum atoms (Klinowski et al., 1983). In aluminous zeolite-Y, each silicon atom has 0-4 aluminum atoms as its nearest neighbors, while the dealuminated sample has none. However, aluminum atoms present in the second-nearest tetrahedron coordination shell cause the signal to appear at -105.0 ppm.
When no aluminum atoms exist in the second-nearest tetrahedron coordination shell the peak will appear at -107.0 ppm. Dealuminated samples may exhibit a weak peak around
-112 ppm, sometimes identified as a rise in the baseline, representing amorphous silica
(Anderson and Klinowski, 1986; Klinowski et al., 1983). The Tosoh sample used does
not show the presence of amorphous silica (Figure 3.3).
123 2+ 3.4.2 Synthesis of Ionic Ru(bpy)3 Within Hydrophobic Siliceous Zeolite-Y
2+ In order to synthesize Ru(bpy)3 inside of the zeolite supercages a “ship-in-a- bottle” technique is used since the size of the windows is about 7 Å and the diameter of the ruthenium complex is about 12 Å. Although the term “ship-in-a-bottle” was not coined until 1985, the technique had been used previously to get metal complexes into zeolite supercages (Herron et al., 1985; Quayle et al., 1982). As mentioned, the synthesis
2+ of Ru(bpy)3 inside of the supercages of aluminum containing zeolite-Y can be readily
2+ done via an ion-exchange method. The complex Ru(bpy)3 was synthesized inside the
3+ pores of zeolite-Y by ion-exchange with an aqueous slurry of Ru(NH3) and then heated
2+ with excess bipyridine in order to form the Ru(bpy)3 complex inside of the zeolite supercage (DeWilde et al., 1980).
The ship-in-a-bottle synthesis, when used in conjunction with siliceous zeolite-Y, requires the use of a neutral ruthenium complex as a precursor. Previous attempts have used [Ru(NH3)5Cl]Cl2 in water/CH3CN (allowing for uptake into the zeolite as a neutral species) followed by reaction with 2,2-bipyridyl at 200oC (Coutant et al., 2003).
Ru(bpy)Cl3 has also been used as a precursor in siliceous zeolite-Y due to its good solubility in organic solvents. The Ru(bpy)Cl3 precursor can only be characterized via
TLC, making it difficult to know the exact purity of the product. Ru(bpy)Cl3 has been found in non-monomeric forms, and dimers and polymers will not fit through the windows of the supercages. The loading of the final ruthenium complex in the supercages will be quite low if this precursor is utilized (Kelch and Rehahn, 1997;
Thummel et al., 1987). Therefore, it was very important to pick a neutral ruthenium
124 precursor that is able to be synthesized purely, and well characterized after synthesis, to
increase the loading level of the tris(2,2’-bipyridyl)ruthenium in the zeolite supercages.
Dichlorodicarbonyl (2,2’-bipyridyl) ruthenium, RuCl2(CO)2(bpy) is neutral, can be
synthesized in non-polymeric forms, and lends itself to a ship-in-a-bottle synthesis with
siliceous zeolite-Y.
The loading level of the sensitizer inside of the zeolite needs to be optimized. A
higher loading level will lead to better sensitivity; however, if the loading level is too
high intramolecular interactions between the ruthenium complexes will take place leading
to different spectroscopic properties. Sykora et al. studied various loading levels of
ruthenium tris(2,2’-bipyridyl) in zeolite-Y, and described the zeolite in a shell system to
describe the number of empty supercages between ruthenium complexes (1999).
Focusing on one supercage of a zeolite, A, it can be stated that it is connected to four
tetrahedrally arranged supercages via 7 Å windows, B, which are all equidistant from A.
The four neighboring supercages cages would be considered shell B. Sykora et al.
determined that loading levels higher than 1 complex per 14 supercages was nonideal
(1999). The loading of 1 complex per 14 supercages represents 25.65% of the B cages
being occupied, and at loadings higher than this the interaction between the ruthenium
complexes in neighboring cages has a significant effect on spectral data. The ideal
loading level can be easily achieved with conventional zeolite-Y ion-exchange “ship-in-
a-bottle” synthesis.
125 As noted, the loading level was improved to about 1 ruthenium complex per 25 supercages over previous synthetic attempts which showed a loading of 1 ruthenium
complex per 10,000 supercages (Payra and Dutta, 2003). The main reasons for
improvement of the loading level are as follows: (1) better choice of precursor, (2)
utilizing commercially produced siliceous zeolite-Y, and (3) utilizing a solid state
reaction scheme.
Because the new precursor was synthesized with no dimers or polymers formed,
the loading level was improved. Although the product formed is impure, the cis and trans
isomers do not decrease the ability of the precursor molecules to enter the supercages.
Also, the Tosoh zeolite sample is more defect free as compared to samples synthesized in
the laboratory. The evidence for this comes from the SSNMR. As seen in Figure 3.3B,
the half width of the Tosoh siliceous zeolite-Y is about 0.6 ppm. The SS-NMR published
by Payra et al. shows a peak with a half width approximately double of the Tosoh
siliceous zeolite-Y (2003). The base of the peak in SS-NMR relates to the amorphous
silica/defect sites throughout the zeolite-Y (Engelhardt and Michel, 1987). Another
measure of the quality is the Si/Al ratio which was listed as being >100 by Payra and
Dutta, and for Tosoh commercial zeolite is >176 (2003).
Finally, instead of using a solvent based synthetic scheme for incorporation of the
sensitizer, a solid state method was utilized. Advantages of using a solid state reaction scheme, where the reaction occurs between a gas and a solid or a solid and a solid, include: efficiency of reaction, the ease of synthesis and manipulation of chemicals, the avoidance of solvents that are often volatile and/or toxic, and avoidance of chemical
126 waste (Kaupp, 2003). The solvent based reaction scheme did require the use of elevated
temperatures for extended amounts of time, which often led to the solvent totally evaporating, even with the use of a condenser. However, the higher temperatures used with the solid-state synthetic scheme, compared to a solvent based scheme, promotes the efficiency of the reaction. Furthermore, when no solvent is present there is no competing force to keep the ruthenium precursor in solution, so promoting the precursor into the supercages may happen more readily.
3.4.3 PDMS Film Quality: Leaching
Placing zeolite-Y into organic polymer membranes has been done previously in
literature (Vankelecom et al., 1995; Barrer and James, 1960). The organic polymer
membranes show problems of zeolite leaching. The main difference between siloxane
and organic polymers is the fact that the backbone of the siloxane polymer gives the
polymer a greater hydrophobicity overall, especially when hydrocarbon functional groups
are attached. This provides the opportunity for hydrogen bonding of embedded
compounds. PDMS membranes are very popular for pervaporation, gas separation, vapor
permeation, and dialysis applications (Gevers et al., 2006; Vankelecom et al., 1994; Gao
et al., 2003). Because of problems experienced with PDMS membrane swelling in their
applications, modifications such as the addition of zeolites have been performed and
successful in preventing swelling of the membrane. The addition of zeolites to the
PDMS film is believed to help prevent swelling because the polymer chains adsorb to the silanol groups on the zeolite. Adsorption between the polymer and zeolite increases the
127 effective cross-linking density and decreasing the amount of swelling. Therefore, siloxane polymers such as PDMS do not show the same problems of zeolite leaching as the organic polymers (Alvaro et al., 1998).
Adnadjevic et al. demonstrated the successful use of siliceous zeolite-Y incorporated in a PDMS membrane for pervaporation studies (1997). Siliceous zeolite-Y has been reported as being especially useful in helping to prevent PDMS swelling because it is able to disperse well through the PDMS matrix, and the surface silanol groups can interact with the PDMS to increase crosslinking (Gevers et al., 2006). As seen in Figure 3.11, swelling did occur at the end of the 10 week period and was particularly noted during week five when the film had fallen off of the glass slide.
Because the addition of siliceous zeolite-Y is anticipated to increase hydrophobicity, and hence decrease swelling, the PDMS film quality is believed to attribute to the swelling seen in the experiment. Because PDMS has not been soaked in water, a swelling comparison with and without zeolite filler can not be made. Voids in the polymer network lead to decreased strength in the PDMS film and increased swelling (Gevers et al., 2006). Also, filler aggregates lead to voids in the polymer and increased swelling
(Gevers et al., 2006). Vankelecom et al. noted that voids appeared in PDMS membranes that had zeolite fillers due to sedimentation of the zeolite in the PDMS at the start of the curing (1994). Films created in our laboratory were noted to have some air bubbles in the
PDMS film; therefore, during the curing processes it is likely that voids were created in the polymer network causing defects leading to increased film swelling.
128 The leaching seen is either due to the ruthenium complex leaching out of the
zeolite, or the ruthenium loaded zeolite leaching out of the PDMS film. Because an
extensive Soxhlet extraction had been performed, all of the ruthenium complexes are
securely entrapped inside of the zeolite supercages. The leaching increased during the
first 5 weeks, then began to decrease for the last 5 weeks. This can be attributed to the
agglomeration of the ruthenium complex loaded zeolite particles leaching out of the
PDMS, agglomerating, then settling out of solution.
3.4.4 Oxygen Quenching
The glucose oxidase assay described has been utilized previously in literature,
mainly in glucose sensing to determine the response of glucose sensors (Pasic et al.,
2007; Li and Walt, 1995; Rosenzweig and Kopelman, 1996). Typically, an oxygen
sensor is utilized along with the glucose sensor, and the change in oxygen concentration
is used as an indirect measure of the glucose concentration (Moreno-Bondi and Wolfbeis,
1990). Moreno-Bondi and Wolfbeis utilized a fiber optic sensor based on the quenching of tris(1,10-phenanthroline)-ruthenium(II) adsorbed on silica gel, then incorporated into a silicone matrix on the tip of a fiber optic (1990). The glucose oxidase is also immobilized on the end of the fiber optic, so upon interaction with glucose the changes in oxygen are monitored and correlated to a glucose concentration. In this research, the glucose oxidase assay was used in order to demonstrate that dissolved oxygen sensing can occur in a more biological media than water, as well as a demonstration of the smaller changes in oxygen concentration that can be monitored.
129 Because using the glucose oxidase assay to induce changes in dissolved oxygen
concentration relies on chemical reactions, it was important for us to have a calibration
where the concentration of dissolved oxygen is well known. The gas saturated water
sensing experiment was used as a calibration in order to have well known dissolved
oxygen concentrations, as well as to have an external calibration method. External
calibration is necessary if future applications focus solely on in vivo measurements such
as intracellular oxygen measurements.
3.4.5 Oxygen Quenching: Stern-Volmer Analysis
Note that the Stern-Volmer plot in Figure 3.15 is linear, which is not typical when
sensitizers are immobilized in a polymer or sol-gel matrix. Bacon and Demas
demonstrated that tris(4,7-diphenyl-1,10-phenanthroline) ruthenium(II) perchlorate
immobilized in a silicone rubber matrix shows a downward curvature when looking at the
Stern-Volmer plot (1987). Douglas and Eaton placed Pt and Pd octaethylporphyrins in
three different polymers (ethyl cellulose, cellulose acetate butyrate polymer, or
polyvinylchloride) with all cases revealing non-linear Stern-Volmer plots (2002). Tang et al. demonstrated that tris(4,7-diphenyl-1,10-phenanthroline) ruthenium(II) immobilized in either n-triethoxysilane / tetraethylorthosilane composite films
demonstrated a non-linear Stern-Volmer plot (2003).
Meier et al. show a non-linear Stern-Volmer when tris(2,2’-bipyridyl)
ruthenium(II) was immobilized inside of commercial zeolite-Y, which was then
distributed in a silicone polymer (1995). The non-linearity in this case is likely to due the
130 intrazeolitic water influencing the environment of the ruthenium probes. Utilizing a siliceous zeolite should allow each probe to see the same environment, have the same response, and demonstrate a linear Stern-Volmer plot.
The curved Stern-Volmer plots represent the dye molecules in different environments, having different responses and more complex decays. The curved Stern-
Volmer plots can be fit utilizing a two-site model which was developed by Demas and
Taylor as follows (1979):
Io/I = 1 / ((f01/1+KSV1[Q]) + (f02/1+KSV2[Q]))
Where f01 and f02 represent the fraction of emission from the dye complex in each specific environment, unquenched. The KSV1 and KSV2 are the Stern-Volmer constants for each of the dye molecules in the two different environments.
The linear Stern-Volmer plot seen in this research is interpreted as the sensitizer being homogeneously distributed throughout the crystalline zeolite environment, and therefore a single species quenched bimolecularly by oxygen. The polymer and sol-gel matrices do not provide this homogeneous environment. The Stern-Volmer equations demonstrate the relationship between lifetime/emission and amount of quenching present.
131 They are as follows:
Io/I = 1 + Ksv[O2] = 1 + kqτo[Q]
τo/ τ = 1 + Ksv[O2]
-1 Ksv = kq* τo = kq(kr + knr)
where Ksv is the Stern-Volmer quenching constant, calculated in the present study from
-1 Figure 3.15 to be 0.89 atm . I represents the luminescence emission intensity, Io is the emission intensity in the absence of the quenching species, τo is the luminescence lifetime
in the absence of the quencher, and [Q] is the quenching species concentration. The
linear Stern-Volmer plot represents the inverse relationship of oxygen concentration to
emission intensity, with a slope equal in magnitude to the Stern-Volmer constant
(Hartmann et al., 1995).
There are reports of linear Stern-Volmer plots when using an organically
modified silicate (ORMOSIL) with an embedded ruthenium probe (Pang et al., 2007;
Tang et al., 2003; Chen et al., 2002; Koo et al., 2004). However, the advantage that zeolites have over the ORMOSIL matrices is that besides being able to provide linear
Stern-Volmer plots, the probe is permanently encapsulated inside of the zeolite supercage
so leaching is not a concern.
132 3.4.6 Live Cell Experiments
As mentioned in the introduction, intracellular monitoring can be performed via a
fiber optic probe, a sensor trapped in phospholipid vesicles, and PEBBLE sensors. Fiber
optic sensors have severe problems with sensor size causing cell perturbation (Buck et
al., 2004). The phospholipid vesicle sensors lack chemical and environmental stability;
although modifications have been done to remedy this, testing in live cells still needs to
be performed (McNamara et al., 2001; Cheng and Aspinwall, 2006). PEBBLE sensors
are of great interest at this time because they are small (nanometer scale) and have fast response. Xu et al. performed intracellular oxygen measurements with rat C6 glioma cells by changing the oxygen concentration of the Dulbecco’s phosphate buffered saline
(2001). The response of the sensor was reversible, stable, and the first demonstration of real-time intracellular analysis oxygen concentrations.
The ruthenium-loaded zeolite shows potential for intracellular monitoring. The
main advantage that the ruthenium loaded zeolite offers over the above sensors is the fact
that the probe is permanently encapsulated in a zeolite, and hence there is no dye leaching
to influence the sensor stability. Furthermore, in the future the zeolite utilized can be
nanometer-sized, which will increase response time and decrease cell perturbation issues
that may be encountered similar to the PEBBLE sensors.
Experiments were performed with macrophage cells. The advantage of using
these cells is that they utilize phagocytosis to take in particles, which provides a means to introduce our sensor into the cell. Also, it provides a means for introduction of zymosan,
which induces oxidative burst. Other cells would require injections methods, such as
133 picoinjection. Because these cells undergo oxidative burst, we can monitor the physiological relevance of oxygen in the cells when they are under oxidative stress.
Furthermore, studies performed earlier have utilized macrophages when studying hydroxyl radical production from phagocytosis of asbestos like minerals (Ruda and
Dutta, 2005). In the future, it would be ideal to combine these studies so that we can monitor the production of reactive oxygen species while monitoring the consumption of oxygen.
A luminol assay was performed in order to determine optimal conditions for inducing the oxidative burst. Because oxidative burst can be monitored via production of extracellular reactive oxygen species, the luminol assay proposed works well. In deciding between using phorbol 12-myristate 13-acetate and zymosan to induce the oxidative burst, the luminol assay showed an oxidative burst three-fold more intense for zymosan (Figure 3.17). Because the two stimuli induce oxidative burst via different mechanisms, it makes sense that one will perform better than the other, depending on the cell. Phorbol 12-myristate 13-acetate is a chemical stimulus, where induction of oxidative burst involves the activation of protein kinase C (PKC) [Andre et al., 1988;
Pasmans et al., 2001]. Zymosan is a particulate stimulus; after binding to the macrophage and being phagocytosed NADPH-oxidase is activated inducing an oxidative burst (Pasmans et al., 2001). Zymosan is independent of PKC when inducing oxidative burst (Andre et al., 1988).
134 The goal of the confocal luminescence experiment performed was to simply show
that the ruthenium loaded zeolite will respond to changes in intracellular oxygen
concentration. The experiments performed in a total air atmosphere showed no response
with the addition of zymosan. The reasoning for this is due to the fact that the
extracellular oxygen in the air provides a constant supply of oxygen to the intracellular
compartments of the macrophages, which is monitored by our ruthenium-loaded zeolite.
In fact, James et al. monitored the concentrations of oxygen in intracellular compartments
of macrophages under different environmental concentrations of oxygen with and without
the induction of a zymosan induced oxidative burst via electron paramagnetic resonance
(EPR) measurements (1998). Cells that were stimulated with zymosan having a perfused
oxygen concentration of 210μM showed an intracellular oxygen concentration of
175.0 μM; cells with a perfused oxygen concentration of 60μM showed an intracellular
oxygen concentration of 31.5 μM. The cells with the intracellular oxygen concentration
of 175.0 μM oxygen have a 1.2 fold greater amount of perfused oxygen; the cells with
31.5 μM intracellular oxygen show a 1.9 fold greater amount of perfused oxygen. In
other words, the intracellular environment is directly influenced by the perfused oxygen
concentration outside of the cell. Due to the influence of the environment outside of the
cell, a total CO2 atmosphere was utilized to limit the amount of oxygen available to the
cell. Applying a total CO2 atmosphere while inducing oxidative burst caused a decrease
in intracellular oxygen concentration that was monitored based on luminescence
quenching. Furthermore, inducing the oxidative burst increased the rate of consumption of the intracellular oxygen, which was not resupplied by an external environment.
135 Despite the large error bars, the overall trend showed that as the oxygen
concentration decreased the ruthenium-loaded zeolite was able to detect changes in the intracellular dissolved oxygen. The increase in intensity for the cells exposed to zymosan in a total carbon dioxide atmosphere is about 2.7 fold, on average between the two
samples seen in Figures 3.20 and 3.21. As shown in the emission quenching plot in
Figure 3.13 for a ruthenium-loaded zeolite embedded in a PDMS matrix, a 1.9 fold increase is seen between nitrogen and oxygen saturated solution. Although the concentrations of oxygen in the macrophages are unknown, the increase seen in the signal
is encouraging and shows that the ruthenium-loaded zeolite may have use for intracellular
monitoring.
Assay optimization needs to be performed in the future, then other parameters
such as reversibility and response time, can be monitored. In the future, an assay which
allows for a different injection of the sensor into the cell (i.e. picoinjection) may provide
better results. The different injection method allows for the sensors to be in the cellular
cytoplasm and not in the phagolysosome vesicles that they are found in after
phagocytosis, which may or may not experience the same changes in dissolved oxygen
during oxidative burst.
3.5 Conclusions
In summary, we have described a novel method for loading ionic probes into
neutral matrices. The loading level of tris(2,2’-bipyridyl) ruthenium(II) in siliceous
zeolite-Y has been able to be improved up to 1 ruthenium complex per 7 supercages,
136 compared to a maximum of 1 ruthenium complex per 10,000 supercages in the past. The
improvement is mainly attributed to systematic choice of the ruthenium precursor, which
can be synthesized at a high purity. The ruthenium loaded zeolite exhibited the ability to
be quenched by oxygen with a linear Stern-Volmer plot calibration with gas saturated
oxygen solutions. The sensor sensitivity was also tested via a glucose oxidase assay to
monitor small changes in dissolved oxygen concentration. Finally, the ruthenium loaded
zeolite was incorporated into macrophages and shown to respond to changes in dissolved
intracellular oxygen concentrations.
Acknowledgements
This material is based upon work supported by the National Science Foundation under
IGERT Grant No. 0221678.
137
Change in [O2] % Quenching (std. dev.)
0-20% 19.0% (1.0)
20-60% 15.8% (0.76)
60-100% 13.7% (2.2)
0-100% 41.1% (1.7)
Table 3.1: Representation of the quenching seen during the gas saturated water quenching experiments. The percent quenching, correlated to the oxygen concentrations in shown in the left column, is shown in the right hand column with the standard deviation in parentheses.
138 (A)
[Glucose] in Solution, mM [Oxygen] in Solution, ppm Quenching , %
0.00 40.3
0.25 32.3 6.3
0.50 24.3 8.6
0.75 16.3 8.7
1.0 8.30 12.7
(B) Henrys Law: c = p k where p = the partial pressure of the solute above the solution (1.0atm (atmospheric pressure) – 0.031atm (water vapor pressure) = 0.97atm) k = the Henry’s Law constant [1.3x10-3 mol atm-1 (dm3)-1 for oxygen] c = concentration of the solute in the solution
Therefore, for a 100% oxygen environment, at 25oC, we have: -3 -1 3 -1 c(O2) = 0.97 atm * 1.3x10 mol atm (dm ) = 0.00126 mol/L * 32.0 g/mol = 40.3 mg/L =40.3 ppm
Table 3.2: Summary of the quenching data from the glucose oxidase assay showing the correlation between the concentration of glucose in solution, the oxygen concentration in solution, and the amount of quenching seen when changing between dissolved oxygen concentrations (A). A sample calculation showing how to determine the concentration of oxygen in solution based on Henry’s Law is given (B).
139
Stern-Volmer calculated Stiochiometrically % Error [O2], ppm O2 Calculated [O2], ppm O2 0.480 40.1 40.3 4.03 33.7 32.3 4.17 25.4 24.3 8.14 17.7 16.3 7.54 9.17 8.30
Table 3.3: Table comparing the oxygen concentration calculated stiochiometrically based on glucose/glucose oxidase reactions to the concentration of oxygen determined by using the gas saturated water Stern-Volmer plot equation.
140
N2
Teflon Tubing
Swage Lock Teflon Thermocouple N2
Reactor
141 Tempmaster Stir Bar (connected to reactor by heating coil line) Drierite Vacuum SiCl4 Outlet (hood)
Figure 3.1: Reactor used during dealumination of zeolite-Y.
141 (A) Meter Clamp Sample in cuvette Gas flowed from tank into cuvette via Teflon Tube teflon tubing O2
N2
mix
(B)
Figure 3.2: Sample set-up for fluorescence sensing experiments utilizing a PDMS film in a cuvette. After the film is placed diagonally in the cuvette (A) two syringes are inserted through a septa cap, one flowing gas and one acting as a vent. (B) shows the fluorimeter used and the placement of the sample holder.
142
(A) Zeolite Y#102
-101 -103 -105 -107 -109 -111 -113 -115 chemical shift (ppm)
(B) Dutta 021405 commercial Tosoh SiZY
-101 -103 -105 -107 -109 -111 -113 -115 chemical shift (ppm)
Figure 3.3: SSNMR 29Si spectra of synthesized (A) siliceous zeolite-Y and commercial (B) siliceous zeolite-Y from the Tosoh Corporation.
143
Figure 3.4: XRD of siliceous zeolite-Y synthesized in the laboratory.
144 (A)
(B)
Figure 3.5: SEM of siliceous zeolite-Y, from Tosoh corporation, after calcination at 500oC with (A) and (B) presenting different scale bars (2 μm and 1 μm respectively).
145
100
95
90
85 % T
80
75 2065 20032003 2066
70 2100 2080 2060 2040 2020 2000 1980 1960
-1 Wavenumber, cm
Figure 3.6: An IR spectra of Ru(bpy)Cl2(CO)2 synthesized in the laboratory.
146 b c a d e f h g
1 = trans (Cl) cis (CO) 2 = cis (Cl) cis (CO)
d e 8.21 h a c f g b 9.83 8.88 8.10 7.77 7.55
PPM 9.8 9.6 9.4 9.2 9.0 8.8 8.6 8.4 8.2 8.0 7.8 7.6
Figure 3.7: An NMR spectrum of one of the RuCl2(CO)2(bpy) samples synthesized in the laboratory. The NMR represents complex 2, which is in the cis (Cl) cis (CO) conformation. The shades of the circles, as well as the letters, correlate the hydrogen atoms shown in the structure to the NMR peaks.
147
DRUV of Sample A (A) 1.0 0.8 ~0.1units
unk 0.6
0.4
ubel ka-M 0.2 K 0.0 200 300 400 500 600 700 Wavelength, nm
DRUV of Sample A 1.4 (B) 1.2 ~0.4 units 1.0 0.8 0.6 0.4
Kubelka-Munk 0.2 0.0
200 300 400 500 600 700 800 Wavelength, nm
DRUV of Sample C 10
(C) 8 ~1.9 units
6 4 2 Kubelka-Munk 0
200 300 400 500 600 700 800 Wavelength, nm
Figure 3.8: Diffuse-reflectance UV absorption spectra of tris(2,2’-bipyridyl) ruthenium (II) loaded in siliceous zeolite-Y utilizing Method 1 at a loading level of 1 ruthenium complex per 5 supercages (A) and Method 2 at a loading level of 1 ruthenium complex per 10 supercages (B) and 5 ruthenium complexes per supercage (C).
148 1. THF, reflux 2.5-3hr Ru(CO)3Cl2 Ru(bpy)(CO)2Cl2 2. 2,2’-bipyridine in THF Acetonitrile + vacuum, Si-ZY + Ru 150oC precursor 24 hours
Under vacuum; N N 200oC; 3-5 2+ Ru N Ru(bpy)(CO) Cl N 2 2 N N
Figure 3.9: Overall reaction synthetic scheme (Method 2) for loading tris(2,2’-bipyridyl) ruthenium II inside of siliceous zeolite-Y.
149 (A) Powders
(i) 5:1 (ii)1:1 (iii)1:5 (iv) 1:10
(B) PDMS Films
(i) 1:10 (ii) 5:1
Figure 3.10: (A): Tris(2,2’-bipyridyl) ruthenium(II) siliceous zeolite-Y powders synthesized at different loading levels seen in (i)-(iv). (B): PDMS films loaded with tris(2,2’-bipyridyl) ruthenium(II) siliceous zeolite-Y with two loading levels seen in (i) and (ii). All ratios represent the loading level that the tris(2,2’-bipyridyl) ruthenium(II) was synthesized in the siliceous zeolite-Y (i.e. 5:1 = 5 ruthenium complexes per zeolite supercage).
150 (A)
(B)
Figure 3.11: PDMS film utilized in leaching experiments with the picture seen in (A) taken before the experiment began and the picure seen in (B) taken at the end of the experiment, ten weeks later.
151 12
10
8
6
4
2 Ru(bpy)3+2 peak area / water peak area 0 1234568910 time, weeks
Figure 3.12: Plot showing trend of the leaching seen over the ten week period. The y-axis is the ratio of the emission peak of the tris(2,2’-bipyridyl) ruthenium(II) emission area to the water Raman peak area.
152
A 1. 2
1
0.8
0.6
0.4
0.2
Normalized Intensity Intensity Normalized 0
500 540 580 620 660 700 740 780
Wavelengt h, nm
Figure 3.13: Luminescence intensity quenching with the top spectra representing nitrogen saturated water and the bottom spectra representing oxygen saturated water for the sample with the loading level of less than 1 ruthenium complex per 25 supercages.
153 (A) Diffuse DRUVReflectance of 010B UV-Vis and Absorption 012
Blue = Lower Loading010B (synthesized at 012 1 Ru complex:10 supercages)
Pink = Higher Loading (synthesized at 5 Ru complexes: supercage) KM (arbitrary)
200 250 300 350 400 450 500 550 600 650 700 750 800 Wavelength, nm
(B) EmissionEmission of 012Spectra and 010B
Blue = Lower010B Loading (synthesized012 at 1Ru complex:10 supercages)
Pink = Higher Loading (synthesized at 5 Ru complexes: supercage) intensity (arbitrary)
500 560 620 680 740 800 Wavelength, nm
Figure 3.14: Absorption (A) and emission (B) spectra showing the peak shift correlating to different loading levels of the ruthenium complex in the zeolite, which implies is an interaction between the ruthenium complexes at higher loading levels.
154 1.8 1.7 1.6 1.5 1.4 1.3 Io/I 1.2 1.1 1 0.9 0.8 0 0.2 0.4 0.6 0.8 1 p(O2), atm
Figure 3.15: Stern-Volmer plot with gas saturated water solutions. R2=0.98.
155
Figure 3.16: Confocal image of a macrophage after 24 hour exposure to the ruthenium- loaded zeolite. The red area represents luminescence from the ruthenium-loaded zeolite and the blue area is DAPI, a nuclear stain.
156 ROS Generation: Phorbol Ester vs. Zymosan A
100
80
60
40 Luminol Index Luminol
20
0 0102030405060708090 Time (Minutes)
Mean Phorbol Ester (7 ng/uL) Mean Zymosan A (100ng/uL
Figure 3.17: Results of luminol assay performed to show intensity of reactive oxygen species produced by phorbol 12-myristate 13-acetate and zymosan.
157 (A) (B)
(C) (D)
Figure 3.18: Visual of the confocal images after the elapsed time with no induction of oxidative burst. (A) and (B) images are taken at time = 0 minutes. (A) depicts the luminescence of the ruthenium loaded zeolite while (B) depicts the luminescence of the DAPI stain. (C) and (D) images are taken at time = 20 minutes. (C) depicts the luminescnce of the ruthenium loaded zeolite while (D) depicts the luminescence of the DAPI stain.
158
(A) (B)
(C) (D)
Figure 3.19: Visual of the confocal images after the elapsed time after induction of oxidative burst. (A) and (B) images are taken at time = 0 minutes. (A) depicts the luminescence of the ruthenium loaded zeolite while (B) depicts the luminescence of the DAPI stain. (C) and (D) images are taken at time = 20 minutes. (C) depicts the luminescnce of the ruthenium loaded zeolite while (D) depicts the luminescence of the DAPI stain.
159 (A)
4.5 4 3.5 3 2.5 2 Signal 1.5
1 Zymosan 0.5 No Zy mos an
average of all cells: Ru/DAPI Ru/DAPI cells: all of average 0 048121620 time, minutes
(B) (C)
NO ZYM OSAN ADDED ZYM OSAN ADDED
5 9 8 4 7 6 3 5 2 4 3 1 2 Ru/DAPI Signal Ru/DAPI
Ru/DAPI Signal Ru/DAPI 1 0 average ofaverage cells:all
average of all cells: 0 5 10 15 2 0 2 5 0 -1 0 5 10 15 2 0 2 5 time, minutes time, minutes
Figure 3.20: Plot of response to oxidative burst induction for first dulplicate run (A). The y-axis is the ratio of the ruthenium complex mean energy to the DAPI mean energy. Each data point represent the average of 6-10 cells that were in the frame. The lower two plots represent the error bars for each data point. (B) represents the error bars for the data points where there was no addition of zymosan, and (C) represents the error bars for the data points where there was addition of zymosan.
160 (A)
12
10
8
6 Signal 4
Zymosan 2 No Zy mos an
average of all cells: Ru/DAPI Ru/DAPI cells: all of average 0 0 4 8 1216202428 time, minutes
(B) (C)
NO ZYM OSAN ADDED ZYM OSAN ADDED
10 20 8 15 6 4 10 2 5
0 Signal Ru/DAPI Ru/DAPI Signal Ru/DAPI
-2 0 5 10 15 20 25 ofaverage cells: all 0 average of all cells: -4 0 102030 time, minutes time, minutes
Figure 3.21: Plot of response to oxidative burst induction for second duplicate run (A). The y-axis is the ratio of the ruthenium complex mean energy to the DAPI mean energy. Each data point represent the average of 6-10 cells that were in the frame. The lower two plots represent the error bars for each data point. (B) represents the error bars for the data points where there was no addition of zymosan, and (C) represents the error bars for the data points where there was addition of zymosan.
161 3.6 References
Adnadjevic, B.; Jovanovic, J.; Gajinov, S. J. Membrane Sci. 1997, 136, 173-179.
Alvaro, M.; Garcıa, H.; Corrent, S.; Scaiano, J.C. J. Phys. Chem. B. 1998, 102, 7530-7534.
Amao, Y. Microchim. Acta. 2003, 143, 1-12.
Anderson, M.W.; Klinowski, J. J. Chem. Soc. Faraday Trans. 1. 1986, 82, 1449- 1469.
Andre, P.; Capo, C.; Mege, J.L.; Benoliel, A.M.; Bongrand, P. Biochem. Biophys. Res. Commun. 1988, 151(2), 641-648.
Bacon, J.R; Demas, J.N. Anal Chem. 1987, 59(23), 2780-2785.
Barrer, R.M.; James, J.D. J. Phys. Chem. 1960, 64(l), 417-427.
Beyer, H.K. Molecular Sieves. 2002, 3, 203-255.
Beyer, H.K.; Belenykaja, I.M.; Hange, F. J. Chem. Soc. Faraday Trans. 1. 1985, 81, 2889-2901.
Black, D.; Deacon, G.B.; Thomas, N.C. Aust. J. Chem. 1982, 35, 2445-2453.
Bossi, M.L.; Daraio, M.E.; Aramendia, P.F. J. Photoch. Photobio. A. 1999, 120, 15- 21.
Buck, S.M.; Xu, H.; Brasuel, M.; Philbert, M.A.; Kopelman, R. Talanta. 2004, 63(1), 41-59.
Cao, Y.; Koo, Yong-Eun L.; Kopelman, R. Analyst. 2004, 129(8), 745-750.
Carraway, E.R.; Demas, J.N.; DeGraff, B.A.; Bacon, J.R. Anal. Chem. 1991, 63, 337-342.
Chardon-Noblat, S.; Da Costa, P.; Deronzier, A.; Haukka, M.; Pakkanen, T.A.; Ziessel, R. J. Electroanal. Chem. 2000, 490, 62-69.
Chen, X.; Zhong, Z.; Li, Z.; Jing, Y.; Wang, X.; Wong, K. Sens. Actuators B. 2002, 87(2), 233-238.
Cheng, Z.; Aspinwall, C.A. Analyst. 2006, 131, 236-243.
162 Clark, H.A.; Barker, S.L.R.; Brasuel, M.; Miller, M.T.; Monson, E.; Parus, S.; Shi, Z- Y.; Song, A.;Thorsrud, B.; Kopelman, R.; Ade, A.; Meixner, W.; Athey, B.; Hoyer, M.; Hill, D.; Lightle, R.; Philbert, M.A. Sens. Actuators B. 1998, 51, 12-16.
Collomb-Dunand-Sauthier, M.L.; Deronzier, A. J. Electroanal. Chem. 1991, 319, 347-353.
Coutant, M.A.; Payra, P.; Dutta, P.K. Microporous and Mesoporous Mater. 2003, 60(1-3), 79-90.
de Klerk-Engels, B.; Fruhauf, H.W.; Vrieze, K.; Kooijman, H.; Spek, A.L. Inorg. Chem. 1993, 32, 5528-5535.
Demas, J.N; DeGraff, B.A.; Coleman, P.B. Anal. Chem. 1999, 23, 793A-800A.
Demas, J.N.; Taylor, D.G. Inorg. Chem. 1979, 18, 3177-3179.
DeWilde, W.; Peeters, G.; Lunsford, J.H. J. Phys. Chem. 1980, 84, 2306-2310.
Douglas, P.; Eaton, K. Sens. Actuators B. 2002, 82(2-3), 200-208.
Engelhardt, G.; Michel, D. High-resolution solid-state NMR of silicates and zeolites; John Wiley & Sons: Engelhardt, Germany, 1987.
Gao, Z.; Nahrup, J.S.; Mark, J.E.; Sakr, A. J. Appl. Polym. Sci. 2003, 90, 658-666.
Gevers, L.E.M.; Vankelecom, I.F.J.; Jacobs, P.A. J. Membrane Sci. 2006, 278, 199- 204.
Hartmann, P.; Leiner, M.J.P.; Lippitsch, M.E. Sens. Actuators B. 1995, 29(1-3), 251- 257.
Haukka, M.; Kiviaho, J.; Ahlgren, M.; Pakkanen, T. Organometallics, 1995, 14, 825- 833.
Herron, N.; Stucky, G.D.; Tolman, C.A. Inorg. Chim. Acta. 1985, 100, 135-140.
Hotz, J.; Meier, W. Langmuir 1998, 14(5), 1031-1036.
Hriljac, J.A.; Eddy, M.M.; Cheetham, A.K.; Donohue, J.A.; Ray, G.J. J. Solid State Chem. 1993, 106(1), 66-72.
James, P.E.; Grinberg, O.Y.; Swartz, H.M. Journal of Leukocyte Biology. 1998, 64, 78-84.
163 Kaupp, G. Cryst. Eng. Comm. 2003, 5(23), 117-133.
Kelch, S.; Rehahn, M. Macromolecules. 1997, 30, 6185-6193.
Kelly, J.M.; O’Connell, C.M.; Vos, J.G. Inorg. Chim. Acta. 1982, 64, L75-L76.
Klimant, I.; Wolfbeis, O. Anal. Chem. 1995, 67, 3160-3166.
Klinowski, J.; Thomas, J.M.; Fyfe, C.A.; Gobbi, G.C.; Hartman, J.S. Inorg. Chem. 1983, 22, 63-66.
Koo, Y.E.L.; Cao, Y.; Kopelman, R.; Koo, S.M.; Brasuel, M; Philbert, M.A. Anal. Chem. 2004, 76, 2498-2505.
Laine, P; Lanz, M.; Calzaferri, G. Inorg. Chem. 1996, 35, 3514-3518.
Li, L.; Walt, D.R. Anal. Chem. 1995, 67, 3746-3752.
Long, J.F.; Waldman, W.J.; Kristovich, R.; Williams, M.; Knight, D.; Dutta, P.K. Environ. Health Persp. 2005, 113: 170-174.
McNamara, K.P.; Nguyen, T.; Dumitrascu, G.; Ji, J.; Rosenzweig, N.; Rosenzweig, Z. Anal. Chem. 2001, 73(14), 3240-3246.
Meier, B.; Werner, T.; Klimant, I.; Wolfbeis, O.S. Sensor. Actuat. B-Chem. 1995, B29(1-3), 240-245.
Moreno-Bondi, M.C.; Wolfbeis, O.S. Anal Chem. 1990, 62, 2377-2380.
Pang, H.L.; Kwok, N.Y.; Chow, L.M.C.; Yeung, C.H.; Wong, K.Y.; Chen, X.; Wang, X. Sens. Actuators B. 2007, 123(1), 120-126.
Pasic, A.; Koehler, H.; Klimant, I.; Schaupp, L. Sens. Actuat.B. 2007, 122, 60-68.
Pasmans, F.; De Herdt, P.; Nerom, A.V.; Haesebrouck, F. Dev. Comp. Immunol. 2001, 25, 159-168.
Payra, P.; Dutta, P.K. Microporous and Mesoporous Mater. 2003, 64, 109-118.
Quayle, W.H.; Lunsford, J.H.; Peeters, G.; De Roy, G.L.; Vansant, E.F. Inorg. Chem. 1982, 21, 2226-2231.
Ray, G.J.; Nerheim, A.G. Zeolites. 1988, 8, 458-463.
Regen, S.L.; Czech, B.; Singh, A. J. Am. Chem. Soc. 1980, 102(21), 6638-6640. 164
Rosenzweig, Z.; Kopelman, R. Sens. Actuators B. 1996, 35-36, 475-483.
Ruda, T.A.; Dutta, P.K. Environ. Sci. Technol. 2005, 39(16), 6147-6152.
Sykora, M; Kincaid, J.R.; Dutta, P.K.; Castagnola, N.B. J. Phys. Chem. B, 1999, 103, 309-320.
Tang, Y.; Tehan, E.C.; Tao, Z.; Bright, F.V. Anal Chem. 2003, 75, 2407-2413.
Thummel, R.P.; Lefoulon, F.; Chirayil, S. Inorg. Chem. 1987, 26, 3072-3074.
Vankelecom, I.F.J.; Merckx, E.; Luts, M.; Uytterhoeven, J.B. J. Phys. Chem. 1995, 99, 13187-13192.
Vankelecom, I.F.J.; Scheppers, E.; Heus, R.; Uytterhoeven, J.B. J. Phys. Chem. 1994, 98, 12390-12396.
Waldman, W.J.; Adams, P.W.; Orosz, C.G.; Sedmak, D.D. Transplantation. 1992, 54, 887-896.
Waldman, W.J.; Kristovich, R.; Knight, D.A.; Dutta, P.K. Chem. Res. Toxicol. 2007, 20(8), 1149-1154.
Xu, H.; Aylott, J.W.; Kopelman, R.; Miller, T.J.; Philbert, M.A. Anal. Chem. 2001, 73, 4124-4133.
Xu, W.; McDonough, R.C.; Langsdorf, B.; Demas, J.N.; DeGraff, B.A. Anal. Chem. 1994, 66, 4133-4141.
165
CHAPTER 4
FIBER OPTIC DISSOLVED OXYGEN SENSOR: OPTIMIZATION OF FIBER
CONFIGURATION UTILIZING TRIS(2,2’-BIPYRIDYL) RUTHENIUM(II)
INSIDE OF SILICEOUS ZEOLITE-Y AS OXYGEN QUENCHING PROBE
4.1 Introduction
Fiber optic sensors are used for sensing a variety of physical and chemical parameters such as pressure, temperature, viscosity, humidity, pH, and chemical concentrations (Pantano and Walt, 1995). We will focus on measuring dissolved molecular oxygen in this research, which is valuable in a variety of applications. The applications are involved in environmental, industrial, automotive, medical, and biological fields. For example, being able to sense dissolved oxygen in bodies of water is of great importance, as it provides information as to the water quality and possible need for remediation to increase the viability of local flora and fauna. Focusing on industrial processes, it is important to monitor oxygen concentrations carefully to prevent corrosion of pipelines or assist in the synthesis of chemicals by photooxidation. Medical and biological fields utilize oxygen monitoring when it comes to issues such as blood gas analysis and intracellular monitoring (Costa-Fernandez et al., 1998).
166
Fiber optic sensors have often replaced the conventional electrochemical sensors
that have been used for oxygen sensing. Because fiber optic sensors are based on optical
oxygen sensing, they do not consume oxygen or require stirring during measurements like electrochemical sensors. Fiber optic sensors allow for selective, in situ, and real-time chemical sensing, which summarizes their main advantages. They are geometrically flexible so they can probe remote locations. Also, because they lack electrical connections between sample and sensor they are immune to electrical interferences. No metallic components are utilized, so the sensor can be used for monitoring under extreme conditions such as in in vivo nuclear magnetic resonance (NMR) systems (Jorge et al.,
2004). Finally, fiber optic sensors do have the ability to be miniaturized down to the sub- micron scale.
One main disadvantage of fiber optic sensors is in the field of intracellular monitoring and relates to sensor size, typically on the micron scale, which can easily lead to cell perturbation and damage. Although fiber tips can now be made down to 20nm, there is still physical distortion of the cell due to the large insertion volume of the fiber
(Buck et al., 2004). Besides the limitation of the size and their destructiveness to the cell, they are also limited by their low throughput and longer response times due to their size
(Ji et al., 2001; Buck et al., 2004).
A fiber optic is comprised of a light-guiding dielectric material that functions via
total internal reflection (Pantano and Walt, 1995). The fiber has a core of a high
refractive index material and an outer region of lower refractive index material, known as
cladding. When light enters at a critical angle it can be confined via total internal
reflection at the interface of the core-cladding. A variety of options exist when selecting 167
a fiber optic for sensing applications. Fiber optics can be composed of a variety of
compounds, typically pure or modified silicas, polymers, and fused silicas. Fiber optics
can be single- or multi- mode. Single-mode fibers have the advantage of higher
bandwidth and less attenuation (modal dispersion) over multi-mode fibers; however,
multi-mode fibers have the advantage of a higher numerical aperture and cheaper
equipment (Avantes). Numerical aperture characterizes fiber input and output behaviors,
which depends on the difference in the refractive indices of the core and cladding. The
numerical aperture defines what is known as the acceptance cone. Half of the cone angle
is the acceptance angle, which is the maximum angle to the fiber axis for light acceptance
and propagation along the fiber. Finally, a distinction between high and low –OH fibers
exists. Because water causes strong absorption peaks in the NIR, its interference is
removed by utilizing a low –OH fiber optic (called VIS-NIR fiber). Higher –OH
contents in a fiber optic are used in the UV/VIS wavelength range because of their low
absorption in the UV (called UV-VIS fiber).
When it comes to dissolved oxygen fiber optic sensing, many forms of fiber
sensors have been utilized. Fiber optic sensors are created, typically, by immobilizing a
fluorescent dye in a polymer or sol-gel matrix, which is then immobilized at the end of an
optical fiber (distal end). The distal end is then placed in the desired location for sensing,
and the other end of the fiber (proximal end) is connected to the system where analysis
will be performed.
Because the fiber optic sensor is an optical sensor, the probe utilized for sensing
follows the requirements discussed in Chapter 3. Because of the novel immobilization of the probe in siliceous zeolite-Y supercages to prevent leaching, we are restricted to using 168
tris(2,2’-bipyridyl) ruthenium(II) which snugly fits in the zeolite-Y supercage. The immobilization of the probe in a support matrix is a critical component. Typically, immobilization involves physical adsorption or trapping of the probe in a polymer or sol- gel matrix which leads to dye leaching and sensor instability. Our research has overcome those problems by permanently entrapping the probes in zeolite supercages. Therefore, the next step is immobilization of the tris(2,2’-bipyridyl) ruthenium(II) entrapped in the siliceous zeolite-Y to be immobilized on the end of a fiber optic.
Currently, few publications exist referencing zeolites immobilized on the end of
fiber optics. One method utilized for zeolite immobilization on a fiber optic is to grow the zeolite on the fiber’s distal end. MFI type zeolite (also called silicalite) is referenced
as being grown on a fiber optic. Growing a zeolite on the end of a fiber is not an option in this research because we use a siliceous zeolite-Y. Siliceous zeolite-Y can only be achieved by dealumination processes, not direct synthesis (Anderson and Klinowski,
1986). Pradhan et al. has immobilized zeolite-Y on the end of fiber optics via a sol-gel process (2000, 2002). The problem with using this procedure is that is involves an annealing step which occurs at 480oC, a temperature at which the ruthenium complex
would not be stable. Finally, an immobilization method has been described where
covalent linkers are utilized (Kulak et al., 2000). The downfall of this method is that it
also requires an annealing step at 450oC for micron sized zeolites.
The goal of this research was to utilize our novel matrix entrapping the ruthenium
complex probe, which prevents probe leaching as described in Chapter 3, and determine
the best way to immobilize this sensor on the end of a fiber optic. Optimization of the
zeolite immobilization, and hence the overall sensing parameters of the fiber optic, is for 169
future work; the primary focus of this chapter is to simply show that the tris(2,2’- bipyridyl) ruthenium(II) loaded siliceous zeolite-Y can be immobilized on the end of a fiber optic in a manner where it does respond to changes in oxygen concentration. Three different immobilization methods were tried, and the percent emission quenching by oxygen was analyzed in order to determine which method is most sensitive. Stern-
Volmer plots and response times were also utilized to determine which immobilization method seems likely to lead to the most optimized sensor in the future.
4.2 Experimental
4.2.1 Materials
The optical fiber used for the sensing experiments was purchased from Thorlabs
(Newton, NJ). The step index multimode fiber has a 200 µm core. The fiber has a numerical aperature of 0.48 and is comprised of a tefzel jacket, a hard polymer cladding, and a pure silica core. The fiber was a high –OH fiber. Scanning electron microscope
(SEM) images of the fiber can be seen in Figure 4.1. The probe utilized was tris(2,2’- bipyridyl) ruthenium(II) immobilized inside the supercages of siliceous zeolite-Y
(synthesized as described in Chapter 3). The only difference from the sample that was used in Chapter 3 is that this zeolite was synthesized at a loading level of 1 ruthenium complex per 5 zeolite supercages. As you may recall, the loading level used in the quenching experiments in Chapter 3 was synthesized at a loading level of 1 ruthenium complex per 10 supercages. The polydimethylsiloxane (PDMS) [RTV 615A and RTV
615B] used was from GE Silicones, Wilton, CT. Component A contains the prepolymer
170
with long PDMS chains and terminal vinyl groups. Component B is the cross-linker that
consists of several Si-H groups per polymer chain. Hexane used was purchased from
Mallinckrodt (Hazelwood, MO).
4.2.2 Fiber Optic Preparation
The optical fiber was either used with the distal end as is, or with the core etched
out of the distal end. When the optical fiber was used with the distal end fully intact, it
was cut and polished with a series of polishing films (63 µm, 9 µm, 1 µm, then 3/10 µm).
The etching of the fiber optic core was performed by placing the distal end of the optical
fiber in a solution of 49% hydrofluoric acid (HF). Fibers were typically placed in the HF
solution, which was filled to about 13mm, and left in the solution overnight in a hood.
The next day the HF solution had evaporated and approximately 5mm is etched from the fiber core. After etching, the fiber was thoroughly rinsed with water. The fiber was then sonicated for 30 minutes to remove any core residual fluorosilicates. The cleaned fiber was viewed under an optical microscope (Nikon Eclipse E600). Etched fibers utilized in sensor preparation were cleaved so that the etched depth was approximately 1.5 mm.
4.2.3 Sensor Preparation
4.2.3.1 Zeolite Sensor Only Inside of Fiber
An etched fiber optic was packed with the ruthenium loaded siliceous zeolite.
Loading was simply done by tapping the powder into the end of the etched out fiber as tightly as possible. After packing the fiber a thin PDMS coating was placed on the end to
ensure zeolite would not fall out. The PDMS was mixed as instructed (1 part 615B to 10 171
parts 615A), a dab was placed on the distal end of the fiber, the fiber was hung with the
distal end pointing upward to ensure a very thin film was spread over the tip, and the
fiber was dried in the dark overnight.
4.2.3.2 PDMS and Sensor Inside of Fiber
An etched fiber optic was loaded with PDMS that was mixed with the ruthenium complex loaded siliceous zeolite-Y. The fiber was placed into the mixture of PDMS, prepared as instructed by the manufacturer (1 part 615B to 10 parts 615A), and the ruthenium loaded siliceous zeolite-Y. Difficulties were experienced in loading the
PDMS into the optical fiber. Therefore, varying amounts of hexane were added to the mixture to see if decreasing the viscosity would encourage the PDMS mixture to go inside of the fiber. Also, the fiber was placed in the PDMS solution and sonicated for varying amounts of time.
4.2.3.3 PDMS and Sensor on Outside of Fiber
The procedure for preparation of the PDMS was adapted from literature (Payra
2+ and Dutta, 2003; Alvaro et al., 1998). The Ru(bpy)3 - siliceous zeolite-Y was added to
about 0.1 g of RTV 615A and sonicated for 30 minutes. Then, 0.01 g of RTV 615B was added and sonicated for an additional 15 minutes. The mixture was placed on the end of
an in-tact fiber optic, which was hung with the distal end pointing down, in the dark
overnight.
172
4.2.4 Physical Characterization of the Sensors
The etched sensors were gold coated and viewed with a JEOL JSM-5500 SEM run at 15 kV accelerating current. The prepared fibers were viewed under an optical microscope (Nikon Eclipse E600) to see the effectiveness of the loading of the probe into the fiber. The microscope was also used in the case of the non-etched fiber to have an estimate of the size of the PDMS on the end of the fiber.
4.2.5 Experimental Set-up
The proximal end of the optical fiber was illuminated with 457.9 nm laser light
from a Coherent Innova 90C Argon ion laser (Santa Clara, CA), which went through a
beam splitter to reduce the energy of the laser beam, then a focusing lens. The laser was
run at 27.3 amps, and determined to be 0.010 watts after the beam splitter.
Emission quenching of the fibers by dissolved oxygen was monitored via a SPEX
Fluorolog spectrophotometer with the emission spectra collected from 500-800 nm in right angle mode. The emission spectra were collected with a low pass filter, passing only wavelengths above 550 nm. The area of the peak was analyzed in order to
determine percent quenching.
The distal end of the fiber optic was placed into a cuvette which was filled with
deionized water. The cuvette had a customized cap made out of Teflon, which had three
holes drilled. One hole drilled was for the fiber to go through for support and to prevent
movement of the fiber during measurement, the second hole was for a syringe needle to
be placed through for bubbling of gases through the solution, and the third hole was used
for venting (Figure 4.2). Gas was bubbled through the water for 15 minutes then a 173
measurement was taken. Gases were bubbled through measurements were taken until the intensity no longer changed, in order to signify equilibrium being reached. During the sensing experiments the following gases were utilized: oxygen (99.99%), zero grade air, a 60/40 mixture of oxygen/nitrogen, and nitrogen (99.999%). When the quenching experiments were run, the sensor was cycled through the gases listed so that several trials could be averaged together.
4.3 Results & Discussion
4.3.1.1 Fiber Etching Results
The inner core of the fiber is able to be etched out with 49% HF. The etching procedure was carried out at room temperature in a fume hood. Figure 4.3 depicts SEM images of an etched fiber optic. As shown in Figure 4.4, about 5 mm of the fiber inner core was etched out after exposure to HF overnight. The inner core of the optical fiber ends up forming a cone shape, with the point of the cone coming out towards the etched end of the fiber optic.
4.3.1.2 Fiber Etching Process
Wet etching techniques have been utilized to a wide extent, especially for micro- machining applications (Machavaram et al., 2007). Fused silica can be dissolved or etched via HF treatment following the generalized reaction below:
SiO2 + 6HF Æ 2H2O + H2SiF6
174
Tso and Pask had related HF concentration to the etching rate and found that with
agitation the etching rate increased; therefore, the etching process is diffusion controlled
(1982). Liang and Readey found that the dissolution rate increased with temperature, and the activation energy of the reaction decreased with increasing concentration of HF
(1987).
As noted, typical etching procedures in this research were carried out by allowing
the fiber to sit in the HF solution overnight with its polymer cladding still intact, without
concern about the HF evaporation. Initially, the fiber was placed in the HF solution and
looked at after both 15 and 45 minutes with an optical microscope. The formation of a
cone shape was seen even after 15 minutes (data not shown). As far as chemical etching
is concerned, fiber optics have been etched into cone-shaped tips by removing the
polymer cladding, as well as leaving the polymer cladding in place. Literature refers to
two different methods being utilized for the etching of fiber optics with the cladding still
on, stating that depending on the polymer cladding (HF permeable or HF impermeable)
the steps of the etching process will differ.
Typically in chemical etching processes for optical fibers, the polymer cladding is
removed, and then the fiber tip is etched. The very first chemical etching methods
involved immersing the end of a fiber optic into an etchant solution, then controlling the
etch rate via additives and precise time control. The Turner method was introduced in
1983 (Hoffmann et al., 1995). With the Turner method, the etchant is covered with an oil
layer, and the etching occurs at the interface of the etchant and the oil. During this fiber
etching method, the diameter decreases following the laws of superficial tension. In
175
other words, when etching begins the diameter of the fiber begins to decrease. The fiber
diameter decrease leads to the meniscus height decreasing until the cone shaped tip is
eventually formed. This is a self-terminating process.
A tube etching process was then developed where a fiber tip is etched in its polymer cladding instead of having the cladding stripped off (Stockle et al., 1999;
Lambelet et al., 1998). Depending on the type of polymer, either HF permeable or HF impermeable, the etching process will follow a different formation pathway. In the tube
etching method, the fiber diameter does not change as it does in the Turner method;
therefore, the meniscus height remains constant. The HF permeable membrane etching
begins by having the center of the core form a concave cone. The core of the fiber then decreases in diameter as the HF permeates through the polymer cladding, and the base of the initial cone continues to increase. Eventually, the width of the cone base equals the width of the reduced fiber, forming the point at the end of the etched fiber (Demagh et al.,
2006; Stockle et al., 1999). If the polymer coating is HF impermeable then no thinning of the upper core is seen. The etching begins at the end of the fiber and forms a cone shape with the tip pointing out towards the etched end of the fiber. After the cone shape is formed it is maintained while the tip shortens inside of the tube. The tip quality remains during the etching process, no matter how much core is etched out of the end.
During the tube method etching of fibers with a HF impermeable polymer
cladding, the formation of the cone tip shape is explained by microconvection in the fiber
along with capillary effects (Stockle et al., 1999). Initially, it is expected that the portion
of the fiber nearest the cladding will etch slightly faster due to geometrical constraints
because at the rim of the core, HF supply occurs out of a large volume as compared to the 176
central region. The cone formation is then started, and as soon as the initial taper has
begun convection starts to deliver HF to the region of the cone nearest the wall, which is most etched. Convection is driven by concentration gradients caused by the etching process itself and the gravitational removal of the reaction products. The influence of gravity on the tip formation process was checked by etching the glass fibers when they were placed in the HF at varying angles. The angle of the fiber caused asymmetric tip shapes.
The tube method of etching was utilized in this research to take advantage of the
etched out “cup” formed at the end of the optical fiber. The etched out portion can be
filled with a probe for sensing applications, offering the advantage of having the probe
safely encapsulated and not completely exposed to the sensing environment. HF
impermeable polymer is the type of cladding on the fibers utilized in this research. The
mechanism of etching did follow the mechanism described for the HF impermeable
membrane, as etching was checked at 15 minutes and 45 minutes with an optical
microscope showing the formation of a cone shape, not an overall reduction in the fiber
core diameter.
4.3.1.3 Etched Fiber Optic Oxygen Sensors
Many optical fiber based sensors have been reported with hydrofluoric etching
being utilized in the preparation procedure. The types of sensors include, strain, pressure,
and chemical sensing (Machavaram et al., 2007). Murphy et al. etched a cavity in a
50/125 Germania-doped silica optical fiber with HF (1997). The cavities were then filled
with a dye-loaded sol-gel via dipping. Further detail was not provided because the main 177
application for their work was not single point sensing. The only data provided was a
sensing plot, exhibiting nearly 90% quenching between 100% oxygen and 100% nitrogen
gas; no data was provided for dissolved oxygen sensing. The exact dye and matrix are also unknown.
Bernhard et al. describes etching an optical fiber with a solution containing HF
and using the created microwells for oxygen sensing (2001). The fiber had a total diameter of 1.6 mm which is comprised of about 3000 individually clad hexagonal shaped fibers with diameters of about 22 µm. The microwells were then silanized,
2+ soaked in a prepolymer solution, coated with oxygen sensitive dye (Ru(Ph2phen)3 ) in a polysiloxane copolymer, then mounted on a spin coater in order to coat the microwells.
The thickness of the polymer layer was not given; however, Quah et al. utilized a similar method and found the sensing layer to be about 3 µm thick (2002). Photopolymerization was then performed. Sensing was performed in phosphate buffered saline gas saturated solutions. The Stern-Volmer plot was nonlinear and fit to the two-site Stern-Volmer model described in Chapter 3. The 90% response time change for a 50% oxygen step change was 2.5 seconds, attributed to the thin sensing layer and good oxygen solubility in the matrix. The maximum Io/I was about 4 at 100% oxygen saturation, which compares to other fiber sensors based on the same dye and matrix (i.e. a sensor that was 0.5 µm thick had Io/I = 4 at 100% oxygen saturation and 1 second response time for a 50% oxygen change).
Overall, no sensors are reported utilizing etched fiber optics with the same polymer matrix or dye as utilized in this research, making direct comparison of the sensors impossible. Tris(2,2’-bipyridyl) ruthenium(II) has been utilized as a probe for 178
fiber optic sensors detecting gaseous oxygen concentrations; however, these sensor
parameters can not be directly compared to dissolved oxygen sensor parameters
(MacCraith et al., 1993; Wolfbeis et al., 1988). The next section will therefore describe
optical dissolved oxygen fiber optic sensors utilizing similar probes and matrices,
although the fiber optic is only etched in one reference.
4.3.2 Parameters of Fiber Optic Dissolved Oxygen Sensors
The rest of the discussion will focus on literature based and commercial fiber optic sensors, as well as the three sensor configurations created and tested in this research. Three main parameters will be utilized to determine important parameters of an optimal optical fiber: Stern-Volmer, sensitivity, and response time. Refer to Table 4.2 for a summary of the fiber optic sensors discussed (commercial, literature-based, and from this research). The following literature based and commercial based sensors will be discussed in the following section:
• Rosenzweig et al. created a fiber optic sensor with tris(1,10-phenanthroline)
ruthenium(II) chloride in an acrylamide polymer covalently attached to a
silanized fiber optic tip via photopolymerization (1995).
• Park et al. created a sensor with an all silica multi-mode fiber coated with a
film 2 µm thick and 100 µm in diameter (2005). A pulled optical fiber (pulled
down to 100-500 nm) and an unpulled optical fiber were both tested using
platinum(II) octaethylporphine as the sensing dye and a liquid polymer made
from polyvinyl chloride and the plastisizing agent bis(2-ethylhexyl) sebacate.
179
• Bernhard et al. described a fiber optic sensor created from an etched fiber
which was comprised of about 3000 microwells. The probe used was tris(4,7-
diphenyl-1,10-phenanthroline) ruthenium(II), and the matrix was a
photopolymerizable siloxane (2001).
• The OxyLite E by Oxford Optronics is comprised of a platinum fluorophore
embedded in a silicone matrix at the end of the fiber optic (Oxford Optronix
Ltd.). Griffiths and Robinson reported utilizing the OxyLite sensor, stating
that it was a ruthenium fluorophore in a silicone polymer (1999).
• Ocean Optics produces a dissolved oxygen sensor known as FOXY-LITE.
The sensor utilizes a ruthenium flurorophore in a hydrophobic sol-gel matrix
with a protective overcoat (Ocean Optics).
Furthermore, the fibers created in this research will be discussed. Brief descriptions of each fiber are below:
• Fiber 1 consists of the ruthenium-loaded zeolite being placed into the etched
portion of the optical fiber, then coated with a thin PDMS layer to prevent the
zeolite from falling out. The sensor with the zeolite probe packed on the
inside can be seen in Figure 4.5. The zeolite plug appears tightly packed and
to have a height of about 0.8 mm inside of the fiber optic. The PDMS layer
thickness is difficult to quantitate from the optical microscope image, but
appears to have a thinness of less than 10 µm.
180
• Fiber 2 consists of the ruthenium loaded zeolite immobilized in a PDMS layer
which is in the etched portion of the optical fiber. Refer to Figure 4.6. The
PDMS only made a small plug at the opening of the distal end of the fiber
with a height of about 0.3 mm. The distance from the distal end of the optical
fiber to the inner etched tip is about 0.5 mm. Varying the conditions related to
viscosity and sonication time did not change the width of the plug.
• Fiber 3 consists of the ruthenium loaded zeolite embedded in PDMS
immobilized on the outside of the fiber (Figure 4.7). The diameter of the
PDMS bulb formed at the fiber’s distal end is approximately 1.5 mm wide.
The PDMS bulb on the end of the fiber did not encompass any of the fiber.
The length of the bulb was approximately 2 mm.
In the discussion to follow, 100% oxygen saturated water has been calculated have a
concentration of 40.3 ppm (refer to Chapter 3, Table 3.2). Literature and commercial
based sensors have reported 100% oxygen saturated water to have concentrations in the
range of 40-43 ppm. The range is due to the calculation being performed using different
Henry’s constants, depending on temperature, as well as differing numbers of significant
figures. The concentration of the oxygen saturated water will be given in the
concentration reported by the authors or companies; be aware that concentrations in the
range of 40-43 ppm all refer to 100% oxygen saturated water. Also, when reporting
values of Io/I, Io will always represent the intensity for a 100% nitrogen saturated solution, while I will represent the intensity for a 100% oxygen saturated solution unless otherwise specified, in order to provide a measure of the overall sensitivity.
181
4.3.2.1 Stern-Volmer Linearity
Literature/Commercial Based Sensors:
Rosenzweig et al. created a fiber optic sensor that was linear in the range of
0-12 ppm dissolved oxygen with an R2=0.996 (1995). Fibers created by Park et al.
(2005) showed high linearity in the Stern-Volmer plot over the range of 0-43 ppm
(unpulled fiber had R2=0.992; pulled fiber had R2=0.9886) (2005). Bernhard et al. (2001) did not obtain a linear Stern-Volmer plot. The OxyLite E by Oxford Optronics is reported to sense oxygen from 0-11 ppm with the most sensitive region from 0-3.3 ppm.
Griffith and Robinson (1999) report that the non-linear Stern-Volmer is the main
disadvantage, but that it is linear and sensitive in the range of 0-0.82 mmHg. The Ocean
Optics sensor responds to concentrations of 0-41 ppm of dissolved oxygen. The sensor
has been noted to provide a non-linear Stern-Volmer plot, and to have a higher sensitivity
at lower concentrations (Ocean Optics).
Currently configured sensors:
Fiber 1 had a Stern-Volmer plot with an R2 = 0.96 (Figure 4.8). The standard deviation of each Io/I value for I = nitrogen, I = air, I = 40%nitrogen/60%oxygen, and I = oxygen saturated water is as follows respectively: 0.01, 0.02, 0.03, 0.03. Fiber 2 had a
Stern-Volmer plot with a R2 = 0.94 (Figure 4.9). The standard deviation of each Io/I
value for I = nitrogen, I = air, I = 40%nitrogen/60%oxygen, and I = oxygen saturated
water is as follows respectively: 0.1, 0.2, 0.2, 0.2. Fiber 3 had a Stern-Volmer plot with
182
a R2 = 0.92 (Figure 4.10). The standard deviation of each Io/I value for I = nitrogen, I =
air, I = 40%nitrogen/60%oxygen, and I = oxygen saturated water is as follows
respectively: 0.002, 0.009, 0.02, 0.002.
Comparison/General Improvement:
Fiber 1 offers the best of the three configurations, showing the highest R2.
Although the standard deviation for Fiber 2 is the best, the R2 is significantly smaller and
hence not considered ideal. The nonlinearity is likely to result from light not being able
to aces all the areas to the same extent where the dye is immobilized. Therefore, when compared to Fiber 2, Fiber 1 has the probe physically located closer to the light source, increasing the chance of the dye molecules which are sensing oxygen to be illuminated by the laser light. Fiber 3 probably has the least linearity because, although it does not have an etched core, it does have the largest amount of PDMS for the oxygen, and the light, to travel through (approximately 2 mm).
When comparing Fiber 1 to the literature and commercial based sensors, it should
be noted that it is one of few demonstrating a linear Stern-Volmer plot, especially over
the range of oxygen concentrations used in these experiments. The commercial based
sensors still struggle to obtain linear Stern-Volmer plots over the concentration range of
0-40 ppm O2. When looking at linearity of the commercial and literature based fibers,
the main issues is not related to how much light is hitting the sensor, since all of the
sensors are significantly thinner (100s of microns at the most). The main issue becomes
the matrix itself, which is typically heterogeneous since it is some sort of polymer or sol- gel. The sensor that comes closest was that created by Park et al. showing linearity over 183
the range of oxygen concentrations from 0-40 ppm. The matrix was a polyvinyl chloride
polymer with plasticizer (2005). The plasticizer allows the oxygen to have a higher rate
of diffusion through the polymer, and the polyvinyl chloride was reported to be a
homogeneous matrix (Park et al., 2005; Mills and Chang, 1998).
The improvement that can be achieved in the R2 of Fiber 1 shows a shortcoming
of this sensor. If the probe is in a heterogeneous matrix, then the Stern-Volmer will be
less linear and have a lower R2. Because the probe is immobilized in a zeolite matrix, it
is likely that the nonlinearity arises from issues with oxygen diffusion, especially in
relation to where the laser light reaches in the sensing layer. In this case, the oxygen must diffuse through a thick plug (a height in the etched fiber optic of about 0.8 mm), which is not packed to the same tightness throughout. Shortening the zeolite plug length
in the core to at most hundreds of microns, which is similar to some of the reported
literature, should help to improve oxygen diffusion and the linearity of the Stern-Volmer plot. Furthermore, it is uncertain how far the laser light is reaching when illuminating the probe distal end. Shortening the distance of etched portion of the fiber core will increase the chance of the laser light reaching the sensing layer.
4.3.2.2 Sensitivity
Literature/Commercial Based Sensors:
Rosenzweig et al. created a fiber optic sensor with an Io/I = 3.2 and Ksv =
0.17 ppm-1 between 0-12 ppm (1995). The unpulled fiber and pulled fiber created by
Park et al. had an Io/I estimated at 45 and 77, and Ksv of 1.07 and 1.84 ppm-1,
respectively (2005). Park et al. found that the pulled fiber tip shows a 2-fold increase in 184
sensitivity at higher oxygen concentration possibly due to a decrease in the thickness of
the sensing film, thus reducing the typical range of oxygen diffusion required for
indicator quenching to occur (2005). Bernhard et al. found Io/I to be around 4 and a Ksv
of around 0.075 ppm-1 (2001).
Currently configured sensors:
Fiber 1 showed a percent quenching of 27.52 ± 2.16 between 100% nitrogen and
100% oxygen saturated water, an Io/I = 1.4, and Ksv = 9.57x10-3 ppm-1 (Figure 4.8).
Fiber 2 showed a percent quenching of 24.52 ± 20 between 100% nitrogen and 100% oxygen saturated water, an Io/I = 1.3, and Ksv = 8.59x10-3 ppm-1 (Figure 4.9). Fiber 3
showed a percent quenching of 24.57 ± 0.29 between 100% nitrogen and 100% oxygen
saturated water, an Io/I = 1.3, and Ksv = 8.21x10-3 ppm-1 (Figure 4.10).
Comparison/General Improvement:
Fiber 1 had the highest percent quenching and the largest Io/I (sensitivity)
amongst the three configurations. However, it is noteworthy that the difference between
the best and worse percent quenching is quite small, only about 3 percent. Also, the
differences in the Io/I values were small, only a 0.1 difference. The differences between
the sensitivity of the fibers being small is probably related to the fact that all of the
sensors must overcome the issues of their thick sensing layers. Again, the thick sensing
layer leads to the problem of the illuminated portion not encompassing the dye molecules
that are sensing the oxygen. Fiber 2 showed a very large standard deviation because of a
185
constant drift that was noted in the measurements. The drift can be due to lack of curing
time, which is noted to create instability if not properly cured (Vander-Donckt et al.,
1996).
The Io/I value for Fiber 1, which is the fiber with the best sensitivity, although
narrowly, in 100% oxygen is 1.34 and the overall percent quenching is 27%. Obviously,
this parameter is poor when compared to literature; for example, Park et al. notes a
percent quenching up to 98% and a maximum Io/I of 77 (2005). Sensitivity is largely
related to the probe utilized. Park et al. utilized a platinum(II) octaethylporphine, which
has a higher quantum yield than the ruthenium complex (2005).
The best comparison for sensitivity is to compare Fiber 1 to a solution of tris(2,2-
bipyridyl) ruthenium (II), which exhibits about 72% quenching between 100% nitrogen and 100% oxygen. The PDMS film in Chapter 3 exhibited a quenching of 41%. The higher percent quenching in Chapter 3 is due to the different loading level; the loading level utilized in this research is higher, making it more probable to experience self- quenching as described in Chapter 3.
In order to increase the sensitivity of Fiber 1 several different aspects may be
considered. The first aspect is to consider improving the sensitivity of the sensor keeping
the ruthenium complex as the probe; the second aspect is to improve the sensitivity of the
sensor by switching the probe. The first optimization step would be to utilize a zeolite
that has an optimized loading level with no self-quenching. The next step, which is
crucial, involves altering the thickness of the sensing layer. Park et al. reports that the
sensing layer created was 2 μm thick (2005). The sensing layer in Fiber 1 is about
0.8 mm thick. The sensitivity of Fiber 1 should be able to be greatly improved by 186
making the thickness of the zeolite plug down to a maximum of 100 µm. In this case, the
laser light is able to access all the probe molecules that oxygen will interact with.
Finally, a different probe may be considered to improve the sensitivity, which due to the
constraints of zeolite-Y would require a different material, possibly a mesoporous silica,
to be utilized as the matrix.
4.3.2.3 Response Time
Literature/Commercial Based Sensors:
The fiber optic sensor created by Rosenzweig et al. was had gas bubbled through
the solution for 5 minutes before measurements were performed; however, a millisecond
response time was predicted (no film thickness was stated) (1995). The pulled and
unpulled fibers, with a thickness of 2 μm, created by Park et al. showed a response time
which appeared to be on the order of 10 minutes when viewing the quenching plots for
dissolved gases because 10-15 minutes is required to saturate the solution with the
dissolved gas (2005). Park et al. estimated that the response is actually less than one
second; however, this fast response can not be observed with their current experimental
procedure, which involves bubbling the gas through the distilled water where the fiber
optic end is placed (2005). Bernhard et al. created a fiber that had a thickness on the order of 0-14 μm and showed a response time of 2.5 seconds for a 50% change in oxygen concentration (2001). The experiment was performed by adding a solution already oxygen saturated to the nitrogen saturated solution being monitored. The OxyLite E by
187
Oxford Optronics is reported to have a response time of less than 10 seconds (Oxford
Optronics). The Ocean Optics sensor is used with a protective overcoat, and the response time for dissolved oxygen is 30-60 seconds (Ocean Optics).
Currently configured sensors:
Fiber 1 seemed to stabilize within the first 15 minutes. The sample also showed good reproducibility from trial to trial. A response time trace can be seen in Figure 4.11.
Fiber 2 had a stabilization time that seemed to vary from 15 to 30 minutes, with 20 minutes being the average amount of time. The sample did not show reproducibility from trial to trial. In other words, there seemed to be a constant drift. Stabilization time for Fiber 3 seemed to vary from 10 to 30 minutes, with 20 minutes being the average amount of time. The sample had good reproducibility from trial to trial.
Comparison/General Improvement:
Fiber 1 showed the quickest stabilization time; therefore a response curve was obtained for this fiber. Fiber 2 and Fiber 3 are likely showing a longer equilibration time because they have sensing layers of PDMS on the millimeter scale. When looking at the response curve for a 1x10-5M solution of tris(2,2’-bipyridyl) ruthenium(II) chloride, seen in Figure 4.12, the average T90 going from 100% nitrogen to 100% oxygen was 1.3 minutes and from 100% oxygen to 100% nitrogen was 2.9 minutes. When analyzing the respond for Fiber 1, only the last two full response curves (the last two cycles going from
100% oxygen to 100% nitrogen and visa versa) were utilized because not enough stabilization time was given at the beginning of the experiment. Fiber 1 shows an 188
average time going from 100% nitrogen to 100% oxygen of 3.7 minutes. The response between 100% oxygen to 100% nitrogen averaged 13.8 minutes. The reason for performing the response experiment with the ruthenium complex in solution was to show
the amount of time utilized for mixing of the gases in the cuvette. Therefore, it seems
like for Fiber 1 the average response time going from 100% nitrogen to 100% oxygen is
on the order of 2 minutes, while the response for 100% oxygen to 100% nitrogen is on
the order of 10 minutes. Mills and Chang notes that diffusion based sensors will have a
longer recovery time, 100% oxygen to 100% nitrogen, than response time, 100% nitrogen
to 100% oxygen (1992).
When looking at literature which reports response curves, it is noted that time is
required for diffusion of the gas into the liquid. Therefore, before discussing how to
improve the response time of the sensor, we can discuss how to improve the response
time experiment. Utilizing an experimental set-up where the gases are not bubbled
through the solutions during the measurement, but instead the gas saturated liquid is
quickly pumped into the holder where measurements are being performed, would be
more ideal to obtain an actual response time that eliminates the time needed for mixing
the gas with the water/solution. Improving response time of Fiber 1 can easily be done
by varying one main factor, the thickness of the sensing layer. As mentioned previously,
literature tends to have sensing layers of less than 100 μm. The sensing layer being
0.8mm thick requires significant amounts of time for gas diffusion, especially to areas
where the laser is illuminating the probe.
189
4.4 Conclusions
Overall, tris(2,2’-bipyridyl) ruthenium(II) loaded in siliceous zeolite-Y has been successfully immobilized at the end of a fiber optic for dissolved oxygen sensing. Three different fiber configurations were tried in order to determine which may provide the highest sensitivity with the most linear Stern-Volmer plot. Because the three fiber configurations were tested only in order to demonstrate that a zeolite matrix loaded with a probe could be immobilized on the end of an optical fiber, optimization of the configuration was not carried out. The optimized configuration involved etching the silica core of an optical fiber and loading the core with the ruthenium loaded zeolite because it seemed to offer the best combination of parameters (Stern-Volmer, sensitivity, response time. The optimized sensor showed an R2 = 0.96 and about 27% quenching between oxygen and nitrogen saturated water. These results can be improved by using a thinner sensing layer, thinner PDMS layer, and even changing the outer layer from
PDMS to a more oxygen permeable compound. Overall, the main goal of proving that the fiber optic application is viable for a zeolite matrix was done and optimization is left for future work.
190
PROBE Medium Overall Io/I Ksv Response Time Reference -1 [O2] Range (linear /nonlinear) ppm of Sensing
tris(1,10- acrylamide polymer 0-40ppm 3.2 0.17 5 min. Rosenzweig and phenanthroline) nonlinear (*range 0- <1s predicted Kopelman, 2005 ruthenium(II) chloride 12ppm) platinum(II) polyvinyl chloride 0-43ppm 77 1.84 10 min Park et al. (pulled), octaethylporphine (pulled fiber) linear <1s predicted 1995 tris(4,7-diphenyl-1,10- photopolymerizable 0-40ppm 4 0.075 2.5s for 50% Δ[O2] Bernhardt et al., 2001 phenanthroline) siloxane nonlinear ruthenium(II) ruthenium/platinum silicone polymer 0-11ppm N/A N/A <10s Oxford Optronix Ltd. based fluorophore nonlinear 0-11ppm linear 0-0.82ppm ruthenium hydrophobic sol-gel 0-41ppm N/A N/A 30-60s Ocean Optics flurorophore matrix nonlinear linear -3 191 tris(2,2’-bipyridyl) polydimethylsiloxane 0-40ppm 1.4 9.57x10 T90 N2Æ O2 = 2min Fiber 1 – Chapter 4 ruthenium(II) linear T90 O2ÆN2 = 10min tris(2,2’-bipyridyl) polydimethylsiloxane 0-40ppm 1.3 8.59x10-3 N/A Fiber 2 – Chapter 4 ruthenium(II) linear tris(2,2’-bipyridyl) polydimethylsiloxane 0-40ppm 1.3 8.21x10-3 N/A Fiber 3 – Chapter 4 ruthenium(II) linear
Table 4.1: Table showing properties of fiber optic dissolved oxygen sensors in order to compare literature, commercial based sensors, and the sensors configured in this research. NA means that the information was not available in the publication. Note that the values for oxygen saturated water vary from 40-43 ppm. This is due to the fact that different references performed the calculation at different temperatures and/or utilized different numbers of significant figures in the Henry’s law constant and/or did or did not take into account the vapor pressure of water when performing the calculation.
1 (A)
jacket cladding
core
jacket (B)
cladding
core
Figure 4.1: A SEM of the fiber optic used in this research can be seen in (A). All three layers (core, cladding, and jacket) are displayed. (B) provides a blown up view of (A). 192
(A) (B)
Figure 4.2: A Teflon cap was made to fit into a quartz cuvette and fill nearly one half of the cuvette volume in order to provide support for the fiber optic and help prevent movement during measurements (A). The Teflon cap has three holes drilled into it, one for a syringe to be placed through, one for the fiber optic to be placed through, and one for venting purposes (B).
193 (A)
(B)
Figure 4.3: SEM of fiber after HF etching for about 3 hours. In both (A) and (B) the jacket and the cladding are left intact while the pure silica core is completely etched.
194
(A)
(B)
Figure 4.4: (A) and (B) both depict the fiber optic after HF etching overnight, then sonicating in DI water. A human hair was placed in one of the fiber optics to show the width of the etching, which was the whole core (B). Approximately 5 mm of the core is etched out in both (A) and (B).
195
Figure 4.5: Picture of an etched fiber optic with the ruthenium-loaded zeolite placed inside of the etched portion, then a very thin PDMS layer placed on the outside of the fiber to prevent zeolite loss from the fiber optic core. The ruthenium-loaded zeolite plug appears to have a height of about 0.8 mm inside of the etched fiber.
196 (A)
PDMS layer thickness
(B) farthest point of etching
tip of inner etched cone
PDMS layer thickness
Figure 4.6: Pictures of an etched fiber optic filled with the ruthenium-loaded zeolited PDMS in the space where the core was etched. (A) shows that the width of the PDMS layer does not extent beyond that of the optical fiber. (B) shows that the fiber is etched about 1.5 mm deep. The space from the distal end of the fiber optic to the tip of the cone etched inside is about 0.5 mm. The height of the PDMS layer inside is about 0.3 mm.
197
(A)
(B)
Figure 4.7: Pictures of a fiber optic coated with the ruthenium-loaded zeolite embedded in PDMS on the outside of the fiber optic (A) and (B). The ruler provides some scaling information. The PDMS forms a bulb that is about 1.5 mm wide and 2 mm in length.. 198 1.5
1.4
1.3
1.2 Io/I
1.1
1 R2=0.96
0.9 0 0.2 0.4 0.6 0.8 1 1.2 pO2, atm
Figure 4.8: Stern-Volmer plot for an etched fiber optic with the ruthenium-loaded zeolite placed inside of the etched portion, then a very thin PDMS layer placed on the outside of the fiber to prevent zeolite loss from the fiber optic core. The standard deviation of each Io/I value for I = nitrogen, I = air, I = 40%nitrogen/60%oxygen, and I = oxygen saturated water is shown. R2=0.96
199
1.6
1.5
1.4
1.3 Io/I 1.2
1.1
1 R2=0.94
0.9 0 0.2 0.4 0.6 0.8 1 1.2 pO2, atm
Figure 4.9: Stern-Volmer plot for an etched fiber optic filled with the ruthenium-loaded zeolited PDMS in the space where the core was etched. The standard deviation of each Io/I value for I = nitrogen, I = air, I = 40%nitrogen/60%oxygen, and I = oxygen saturated water is shown. R2=0.94
200
1.4
1.35
1.3
1.25
1.2
1.15 Io/I 1.1
1.05
1
0.95 R2=0.92 0.9 0 0.2 0.4 0.6 0.8 1 1.2 pO2, atm
Figure 4.10: Stern-Volmer plot for fiber optic coated with the ruthenium-loaded zeolite embedded in PDMS on the outside of the fiber optic. The standard deviation of each Io/I value for I = nitrogen, I = air, I = 40%nitrogen/60%oxygen, and I = oxygen saturated water is shown. R2=0.92
201
Response Curve
8.00E+09
7.50E+09
7.00E+09
6.50E+09
6.00E+09 Intensity 5.50E+09
5.00E+09
4.50E+09 1 1200 2399 3598 4797 5996 7195 Time, seconds
Figure 4.11 Response curve for Fiber 1 (dye loaded zeolite inside of etched fiber core with thin PDMS layer on outside) between 100% oxygen and 100% nitrogen. Experiment was performed by bubbling the gases through the solution.
202 Response Curve of Ru(bpy)3+2 in Water
1.90E+06
1.70E+06
1.50E+06 1.30E+06
1.10E+06
Intensity 9.00E+05
7.00E+05 5.00E+05
3.00E+05 0 600 1200 1800 2400 3000 3600 Time, seconds
Figure 4.12: Response curve of a 1x10-5 M solution of tris(2,2’-bipyridyl) ruthenium(II) chloride in deionized water between 100% oxygen and 100% nitrogen. Experiment was performed by bubbling the gases through the solution.
203 4.6 References
Alvaro, M.; Garcia, H.; Corrent, S.; Scaiano, J.C. J. Phys. Chem. B. 1998, 102(39), 7530-7534.
Anderson, M.W.; Klinowski, J. J. Chem. Soc., Faraday Trans. 1986, 82(5), 1449- 1469.
Avantes. Fiber Optics. http://www.avantes.com/ (accessed 9/5/07).
Bernhard, D.D.; Mall, S.; Pantano, P. Anal. Chem. 2001, 73, 2484-2490.
Buck, S.M.; Xu, H.; Brasuel, M.; Philbert, M.A.; Kopelman, R. Talanta. 2004, 63(1), 41-59.
Costa-Fernandez, J.M.; Diaz-Garcia, M.E.; Sanz-Medel, A. Anal. Chim. Acta. 1998, 360(1-3), 17-26.
Demagh, N.; Guessoum, A.; Aissat, H. Meas. Sci. Technol. 2006, 17(1), 119-122.
Force, Inc. Types of Optical Fiber. http://www.fiber-optics.info/articles/fiber- types.htm (accessed 9/5/07).
Garcia-Fresnadillo, D.; Marazuela, M.D.; Moreno-Bondi, M.C.; Orellana, G. Langmuir. 1999, 15, 6451-6459.
Griffiths, J.R.; Robinson, S.P. Br. J. Radiol. 1999, 72, 627-630.
Hoffmann, P.; Dutoit, B.; Salathe, R.-P. Ultramicroscopy. 1995, 61, 165-170.
Ji, J.; Rosenzweig, N.; Jones, I.; Rosenzweig, Z. Anal. Chem. 2001, 73, 3521-3527.
Jorge, P.A.S.; Caldas, P.; Rosa, C.C.; Oliva, A.G.; Santos, J.L. Sens. Actuators, B. 2004, B103(1-2), 290-299.
Jorge, P.A.S.; Rosa, C.C.; Caldas, P.; Oliva, A.G.; Esteves da Silva, J.C.G.; Santos, J.L.; Farahi, F. Fiber Integr. Opt. 2005, 24, 201-225.
Kulak, A.; Lee, Y-J.; Park, Y.S.; Kim, H.S.; Lee, G.S.; Yoon, K.B. Adv. Mater. 2002, 14(7), 526-529.
Lambelet, P.; Sayah, A.; Pfeffer, M.; Philipona, C.; Marquis-Weible, F. Appl. Opt. 1998, 37(31), 7289-7292.
204 Liang, D.; Readey, D.W. J. Am. Ceram. Soc. 1987, 70(8) 570–577.
Liu, N.; Hui, J.; Sun, C.; Dong, J.; Zhang, L.; Xiao, H. Sensors. 2006, 6, 835-847.
Machavaram, V.R.; Badcock, R.A.; Fernando, G.F. Sens. Actuators, A. 2007, 138(1), 248-260.
MacCraith, B.D.; McDonagh, C.M.; O'Keeffe, G.; Keyes, E.T.; Vos, J.G.; O'Kelly, B.; McGilp, J.F. Analyst. 1993, 118(4), 385-388.
Mills, A; Chang, Q. Analyst. 1992, 117(9), 1461-1466.
Murphy, V.; MacCraith, B.D.; Butler, T.; McDonagh, C.; Lawless, B. Electron. Lett. 1997, 33(7), 618-619.
Ocean Optics. Ocean Optics Introduces Handheld Oxygen Sensor System. http://www.oceanoptics.com/corporate/pressreleases/2004_Releases/foxylite_release. asp (accessed 9/5/07).
Oxford Optronix Ltd. OxyLite E™ Specifications. http://www.oxford- optronix.com/pdfs/Technical/OxyLiteE%20Spec.pdf (accessed 9/5/07).
Oxford Optronix Ltd. Frequently Asked Questions. http://www.oxford- optronix.com/pdfs/Technical/TM%20FAQ.pdf (accessed 9/5/07).
Pantano, P.; Walt, D.R. Anal. Chem. 1995, 67(15), 481A-487A.
Park, E.J.; Reid, K.R.; Tang, W.; Kennedy, R.T.: Kopelman, R. J. Mater. Chem. 2005, 15, 2913–2919.
Payra, P.; Dutta, P.K. Microporous Mesoporous Mater. 2003, 64(1-3), 109-118.
Pradhan, A.R.; Macnaughtan, M.A.; Raftery, D. J. Am. Chem. Soc. 2000, 122(2), 404-405.
Pradhan, A.R.; Uppili, S.; Shailaja, J.; Sivaguru, J.; Ramamurthy, V. Chem. Commun. 2002, 6, 596–597.
Quah, S.; Danowski, K.L.; Pantano, P. Biomed. Microdevices. 2002, 4(2), 123-130.
Rosenzweig, Z.; Kopelman, R. Anal. Chem. 1995, 67, 2650-2654.
Stockle, R; Fokas, C.; Deckert, V.; Zenobi, R.; Sick, B; Hecht, B; Wild, U.P. Appl. Phys. Lett. 1999, 75(2), 160-162.
205
Tso, S.T.; Pask, J.A. J. Am. Ceram. Soc. 1982, 65(7), 360–362.
Unger, M.A.; Kossakovski, D.A.: Kongovi, R.; Beauchamp, J.L.; Baldeschwieler, J.D.; Palanker, D.V. Rev. Sci. Instrum. 1998, 69(8), 2988-2993.
Vander Donckt, E.; Camerman, B.; Herne, R.; Vandeloise, H.R. Sens. Actuators, B. 1996, 32, 121–127.
Vasylevska, G.S.; Borisov, S.M.; Krause, C.; Wolfbeis, O.S. Chem. Mater. 2006, 18, 4609-4616.
Wolfbeis, O.S.; Weis, L.J.; Leiner, M.J.P.; Ziegler, W.E. Anal. Chem. 1988, 60(19), 2028-2030.
Xi, C.; Zhong, Z.; Li, Z.; Jiang, Y.; Wang, X.; Wong, K. Sens. Actuators, B. 2002, 87, 233–238.
Zhang, J.; Dong, J.; Luo, M.; Xiao, H.; Murad, S.; Normann, R.A. Langmuir 2005, 21, 8609-8612.
206
BIBLIOGRAPHY
Chapter 1:
1. Amao, Y. Microchim. Acta. 2003, 143, 1-12.
2. Amethyst Galleries. The Mineral Mordenite. http://mineral.galleries.com/minerals/silicate/mordenit/mordenit.htm (accessed 07/25/2007)
3. Anderson, M.W.; Klinowski, J. J. Chem. Soc. Faraday Transactions I. 1986, 82, 1449-1469.
4. Archer, V.E. Med. Hypotheses. 1979, 5(11), 1257-1260.
5. Asiedu, J.K.; Ji, J.; Nguyen, M.; Rosenzweig, N.; Rosenzweig, Z. J. Biomed. Opt. 2001, 6(2), 116-121.
6. Aylott, J.W. Analyst. 2003, 128, 309-312.
7. Babbs, C.F.; Steiner, M.G. Method Enzymol. 1990, 186, 137-147.
8. Bacon, J.R.; Demas, J.N. Anal. Chem. 1987, 59(23), 2780-2785.
9. Bass, D.A.; Parce, J.W.; Dechatelet, L.R.; Szejda, P.; Seeds, M.C.; Thomas, M. J. Immunol. 1983, 130(4), 1910-17.
10. Beyer, H.K.; Belenykaja, I.M.; Hange, F.J. J. Chem. Soc. Faraday Transactions I. 1985, 81, 2889-2901.
207 11. Buck, S.M.; Xu, H.; Brasuel, M.; Philbert, M.A.; Kopelman, R. Talanta 2004, 63(1), 41-59.
12. Buda, F.; Ensing, B.; Gribnau, M.C.M.; Baerends, E.J. Chem. Eur. J. 2003, 9(14), 3436-3444.
13. Cao, Y.; Koo, Yong-Eun, L.; Kopelman, R.. Analyst. 2004, 129(8), 745-750.
14. Carr, A.; Frei, B. FASEB J. 1999, 13, 1007-1024.
15. Carraway, E.R.; Demas, J.N.; DeGraff, B.A. Anal. Chem. 1991, 63, 332-336.
16. Caspar, J.V.; Meyer, T.J. J. Am. Chem. Soc. 1983, 105, 5583-5590.
17. Chaudière, J.; Ferrari-iliou, R. Food Chem. Toxicol. 1999, 37, 949-962.
18. Chen, Y.; Li, X.; Gao, F.; Qiu, S. Mater. Lett. 2005, 59, 2598-2600.
19. Cheng, F.; Jen, J.; Tsai, T. J. Chromatogr. B. 2002, 781(1-2), 481-496.
20. Cheng, Z.; Aspinwall, C.A. Analyst. 2006, 131, 236-243.
21. Clark, H.A.; Barker, S.L.R.; Brasuel, M.; Miller, M.T.; Monson, E.; Parus, S.; Shi, Z-Y.; Song, A.;Thorsrud, B.; Kopelman, R.; Ade, A.; Meixner, W.; Athey, B.; Hoyer, M.; Hill, D.; Lightle, R.; Philbert, M.A. Sens. Actuators, B. 1998, 51, 12-16.
22. Coutant, M.A.; Payra, P.; Dutta, P.K. Microporous Mesoporous Mater. 2003, 60, 79-90.
23. Croft, S.; Gilbert, B.C.; Smith, J.R. Lindsay; Whitwood, A.C. Free Radic. Res. Commun. 1992, 17(1), 21-39.
24. Cross, C.E.; van der Vliet, A.; O’Neill, C.A.; Louie, S.; Halliwell, B. Environ. Health Perspec. 1994, 102(10), 185-191.
25. Demas, J., DeGraff, B.A., Coleman, P.B. Anal. Chem. 1999, 793A-800A.
26. Demas, J.N.; Taylor, D.G. Inorg. Chem. 1979, 18, 3177-3181.
208 27. DeWilde, W.; Peeters, G.; Lunsford, J.H. J. Phys. Chem. 1980, 84, 2306-2310.
28. Douglas, P.; Eaton, K. Sens. Actuators, B. 2002, 82, 200-208.
29. Eborn, S.K.; Aust, A.E. Arch. Biochem. Biophys. 1995, 316(1), 507-514.
30. Environmental Working Group. The Asbestos Epidemic in America. http://www.ewg.org/reports/asbestos/facts/fact1.php (accessed 07/24/2007).
31. EPA: Environmental Protection Agency. Asbestos and Vermiculite: Basic Information. http://www.epa.gov/asbestos/pubs/help.html (accessed 07/24/2007).
32. Erionite CAS No. 66733-21-9. http://ntp.niehs.nih.gov/ntp/roc/eleventh/profiles/s083erio.pdf (accessed 07/25/2007), part of National Toxicology Program http://ntp.niehs.nih.gov/ (accessed 07/25/2007).
33. Fenoglio, I.; Martra, G.; Coluccia, S.; Fubini, B. Chem. Res. Toxicol. 2000, 13, 971-975.
34. Fischer, T.G. Faculty homepage. http://faculty.ircc.edu/faculty/tfischer/images/lower%20resp.%20tract.jpg (accessed 07/24/2007), part of Teresa G. Fischer's homepage http://faculty.ircc.edu/faculty/tfischer (accessed 07/24/2007).
35. Floyd, R.A.; Watson, J. J.; Wong, P.K. J. Biochem. Bioph. Methods. 1984, 10(3- 4), 221-235.
36. Fubini, B. Environ. Health Persp. 1997, 105(5), 1013-1020.
37. Fubini, B.; Arean, C.O. Chem. Soc. Rev. 1999, 28, 373–381
38. Galbraith, W.; Thompson, K.; Haq, T.; Timmins, M. BD Biosciences. http://www.bdbiosciences.com/discovery_labware/Products/drug_discovery/polys tyrene_microplates/oxygen_biosensor_system/pdf/OBS_innate_immune_system. pdf (accessed 07/26/2007). Innate Immune System Signaling: NR8383 Cell Oxidative Burst Measured by Oxygen Consumption. 2002.
39. Glockner J F; Swartz H M; Pals, M.A. J. Cell. Physiol. 1989, 140(3), 505-511.
209 40. Green, L.C.; Crouch, E.A.C.; Ames, M.R.; Lash, T.L. Regul. Toxicol. Pharm. 2002, 35, 327-337.
41. Grinberg, O.Y.; James, P.E; Swartz, H.M. Adv. Exp. Med. Biol. 1998, 454, 415- 423.
42. Guthrie, G.D. Environ. Health Persp. 1997, 105(5), 1003-1011.
43. Guthrie Jr., G.D. Am. Mineral. 1992, 77, 225-243.
44. Gutteridge, J.M.C. FEBS Lett. 1984, 172(2), 1591-1599.
45. Gutteridge, J.M.C.; Mumby, S.; Quinlan, G.J.; Chung, K.F.; Evans, T.W. Biochem. Bioph. Res. Co. 1996, 220, 1024-1027.
46. Hager, G.D.; Crosby, G.A. J. Am. Chem. Soc. 1975, 97, 7031-7037.
47. Halliwell, B.; Gutteridge, J.M.C. FEBS. Lett. 1992, 307(1), 108-112.
48. Hardy, J.A.; Aust. A.E. Chem. Rev. 1995, 95(1), 97-118.
49. Hartmann, P.; Leiner, M.J.P.; Lippitsch, M.E. Sens. Actuators, B. 1995, 29, 251- 257.
50. Herron, N.; Stucky, G.D.; Tolman, C.A. Inorg. Chim. Acta. 1985, 100, 135-140.
51. Hesterberg, T.W.; Chase, G.; Axten, C.; Miller, W.C.; Musselman, R.P.; Kamstrup, O.; Hadley, J.; Morscheidt, C.; Bernstein, D.M.; Thevenaz. P. Toxicol. Appl. Pharmacol. 1998, 151, 262-275.
52. Hotz, J.; Meier, W. Langmuir. 1998, 14(5), 1031-1036.
53. Hriljac, J.A.; Eddy, M.M.; Cheetham, A.K.; Donohue, A.J.; Ray, G.J. J. Solid State Chem. 1993, 106, 66-72.
54. Hughes, V.A.; Douglas, P. J. Fluorescence. 2006, 16(3), 403-409.
210 55. ICRP: International Commission on Radiological Protection [1994]. Human respiratory tract model for radiological protection. Oxford, England: Pergamon, Elsevier Science Ltd., International Commission on Radiological Protection Publication No. 66.
56. Iles, K.E.; Forman, H.J. Immunol. Res. 2002, 26(1–3), 95-105.
57. Jahnke, L.S. Anal. Biochem. 1999, 269(2), 273-277.
58. James, P.E.; Grinberg, O.Y.; Swartz, H.M. J Leukoc. Biol. 1998, 64(1), 78-84.
59. Ji, J.; Rosenzweig, N.; Jones, I.; Rosenzweig, Z.. Anal. Chem. 2001, 73, 3521- 3527.
60. Karakoca, Y.; Emri, S.; Cangir, A.K.; Baris, Y. I. Indoor Built Environ. 1997, 6(2), 100-105.
61. Kaupp, G. Cryst. Eng. Comm. 2003, 5(23), 117-133.
62. Kerr, G.T. J. Phys. Chem. 1967, 71, 4155-4156.
63. Lehmann A K; Sornes S; Halstensen A. J. Immunol. Methods. 2000, 243(1-2), 229-242.
64. Light, W. G.; Wei, E.T. Nature. 1977, 265, 537-539.
65. Lumpkin, R.S.; Kober, E.M.; Worl, L.A.; Murtaza, Z.; Meyer, T.J. J. Phys. Chem. 1990, 94, 239-243.
66. Martra, G.; Tomatis, M.; Fenoglio, I.; Coluccia, S.; Fubuni, B. Chem. Res. Toxicol. 2003, 16, 328-335.
67. Maruszewski, K.; Kincaid, J.R. Inorg. Chem. 1995, 4, 2002-2006.
68. Maruszewski, K.; Strommen, D.P.; Kincaid, J.R. J. Am. Chem. Soc. 1993, 115, 8345-8350.
69. Mason, R.P.; Knecht, K.T. Methods Enzymol. 1994, 233, 112-117.
211 70. McDonagh, C.M.; Shields, A.M.; Mcevoy, A.K.; Maccraith, B.D.; Gouin, J.F. J. Sol-Gel Sci. Technol. 1998, 13(1/2/3), 207-211.
71. McEvoy, A.K.; McDonagh, C.M.; MacCraith, B.D. Analyst. 1996, 121, 785-788.
72. McNamara, K.P.; Nguyen, T.; Dumitrascu, G.; Ji, J.; Rosenzweig, N.; Rosenzweig, Z. Anal. Chem. 2001, 73(14), 3240-3246.
73. Meier, B.; Werner, T.; Klimant, I.; Wolfbeis, O.S. Sens. Actuators, B. 1995, B29(1-3), 240-245.
74. Metintas, M.; Hillerdal, G.; Metintas, S. Eur. Respir. J. 1999, 13(3),523-526.
75. Meyer, T.J. Pure Appl. Chem. 1986, 58, 1193-1200.
76. Mitchell, T. O. 2006. Luminescence Based Measurement of Dissolved Oxygen in Natural Waters. http://www.hydrolab.com/pdf/Hach%20Env%20White%20Paper%20- %20LDO.pdf (accessed 07/24/2007).
77. Molson Medical Informatics Sampler. http://www.mmi.mcgill.ca/mmimediasampler2005/default.htm (accessed 07/24/2007), part of McGill University http://curriculum.mmi.mcgill.ca/ (accessed 07/24/2007).
78. Morse, P.D.; Swartz H.M. Magn. Res. Med. 1985, 2(2), 114-127.
79. Mossman, B.T.; Churg, A. Am. J. Respir. Crit. Care Med. 1998, 157(5), 1666- 1680.
80. Mossman, B.T.; Gee, J.B.L. New Engl. J. Med. 1989, 320(26), 1721-1730.
81. Nappi, A.J.; Vass, E. Biochim. Biophys. Acta. 1997, 1336, 295-301.
82. Nejjari, A; Fournier, J; Pezerat, H; Leanderson, P. Br. J. Ind. Med. 1993, 50(6), 501-504.
83. Palekar, L.D.; Most, B.M.; Coffin, D.L. Environ. Res. 1988, 46(2), 142-152.
84. Paris, J. P.; Brandt, W.W. J. Am. Chem. Soc. 1959, 81, 5001-5002. 212
85. Park, E.J.; Reid, K.R.; Tang, W.; Kennedy, R.T.; Kopelman, R. J. Mater. Chem. 2005, 15, 2913-2919.
86. Patel, A.K; Hallett, M.B; Campbell, A K. Biochem. J. 1987, 248(1), 173-180.
87. Payra, P.; Dutta, P.K. Microporous Mesoporous Mater. 2003, 64, 109-118.
88. Proteggente, A.R.; Rehman, A.; Halliwell, B.; Rice-Evans, C.A. Biochem. Biophys. Res. Commun. 2000, 277, 535-540.
89. Regen, S.L.; Czech, B.; Singh, A. J. Am. Chem. Soc. 1980, 102(21), 6638-6640.
90. Rose, A.L.; Waite, T.D. Anal. Chem. 2001, 73, 5909-5920.
91. Roundhill, D.M. Photochemistry and Photophysics of Metal Complexes; Plenum Press: New York, 1994.
92. Ruda, T.A.; Dutta, P.K. Environ. Sci. Technol. 2005, 39(16), 6147-6152.
93. Sarkar, N.; Das, K.; Nath, D.N; Bhattacharyya, K. Langmuir. 1994, 10(1), 326- 329.
94. Scaduto, R.C. Free Radical Biol. Med. 1995, 18(2), 271-277.
95. Stanton, M.F.; Layard, M.; Tegeris, A.; Miller, E.; May, M.; Morgan, E.; Smith, A . J. Natl. Cancer Inst. 1981, 67(5), 965-975.
96. Tang, Y.; Tehan, E.C.; Tao, Z.; Bright, F.V. Anal. Chem. 2003, 75(10), 2407- 2413.
97. Valentine, R.; Chang, M.J.W.; Hart, R.W.; Finch, G.L.; Fisher, G.L. Environ. Health Persp. 1983, 51, 357-368.
98. Valko, M.; Morris, H.; Cronin M.T.D. Curr. Med. Chem. 2005, 12, 1161-1208.
213 99. van Zijverden, M. Dissertation. http://igitur- archive.library.uu.nl/dissertations/1954471/c8.pdf (accessed 07/24/2007), part of Igitur: Utrecht Publishing and Archiving Services. http://www.igitur.nl/igiturarchief/index.php (accessed 07/24/2007).
100. Wagner, J.C.; Skidmore, J.W.; Hill, R.J.; Griffiths, D.M. Br. J. Cancer. 1985, 51, 727-730.
101. Weitzman, S.A.; Graceffa, P. Arch. Biochem. Biophys. 1984, 228(1), 373- 376.
102. Winterbourn, C.C. Toxicol. Lett. 1995, 82/83, 969-974.
103. World Health Organization 2006. Elimination of asbestos-related disease. http://www.who.int/occupational_health/publications/asbestosrelateddiseases. pdf (accessed 07/24/2007).
104. Xu, H.; Aylott, J.W.; Kopelman, R.; Miller T.J.; Philbert, M.A. Anal. Chem. 2001, 73, 4124-4133.
105. Xu, W.; McDonough, R.C.; Langsdorf, B.; Demas, J.N.; DeGraff, B.A. Anal. Chem. 1994, 66, 4133-4141.
106. Yao, Y.; Zhang, M.; Shi, J.; Gong, M.; Zhang, H.; Yang, Y. Mater. Lett. 2001, 48(1), 44-48.
107. Yeh, T.S.; Chu, C.S.; Lo, Y.L. Sens. Actuators, B. 2006, 119, 701-707.
108. Yildiz, G.; Demiryurek, A.T. J. Pharmacol. Toxicol. Methods. 1998, 39(3), 179-184.
109. Zeolyst International. http://www.zeolyst.com/html/faq.asp (accessed 07/24/2007).
214 Chapter 2:
1. Auerbach, S.M.; Carrado, K.A.; Dutta, P.K. Handbook of Zeolite Science and Technology; Marcel Dekker, Inc: New York, 2003.
2. Babbs, C.F.; Steiner, M.G. Method Enzymol. 1990, 186, 137-147.
3. Badran, A.H.; Dwyer, J.; Evmerides, N.P.; Manford, J.A. Inorg. Chim. Acta. 1977, 21, 61-64.
4. Breck, D.W. Zeolite Molecular Sieves; Wiley-Interscience Publication: New York, 1974.
5. Brown, D.M.; Beswick, K.S.; Bell, K.S.; Donaldson, K. Ann. Occup. Hyg. 2000, 44(2), 101-108.
6. Brown, D.M.; Roberts, N.K; Donaldson, K. In Vitro. 1998, 12, 15-24.
7. Carr, A.; Frei, B. FASEB J. 1999, 13, 1007-1023.
8. Chaudière, J.; Ferrari-iliou, R. Food Chem. Toxicol. 1999, 37, 949-962.
9. Cross, C.E.; van der Vliet, A.; O’Neill, C.A.; Louie, S.; Halliwell, B. Environ. Health Persp. 1994, 102(10), 185-191.
10. Daniels, R.H.; Kerr, G.T.; Rollmann, L.D. J. Am. Chem. Soc. 1978, 100, 3097- 3100.
11. Dunford, H. B. Coordin. Chem. Rev. 2002, 233/234, 311-318.
12. Eborn, S.K.; Aust, A.E. Arch. Biochem. Biophys. 1995, 316(1), 507-514.
13. Evmiridis, N. P. Inorg. Chem. 1986, 25, 4362-4369.
14. Fach, E.; Kristovich, R.; Long, J.F.; Waldman, W.J.; Dutta, P.K.; Williams, M. V. Environ. Int. 2003, 29, 451-458.
215 15. Fach, E.; Waldman, W.J.; Williams, M.; Long, J.; Meister, R.K.; Dutta, P.K. Environ. Health Persp. 2002, 110, 1087-1096.
16. Fenoglio, I.; Martra, G.; Coluccia, S.; Fubini, B. Chem. Res. Toxicol. 2000, 13, 971-975.
17. Fubini, B.; Arean, O. Chem. Soc. Rev. 1999, 28, 373-381.
18. Gallard, H.; De Laat, J.; Legube, B. Water Research. 1999, 33(13), 2929-2936.
19. Ghio, A.; Zhang, J.; Piantadosi, C. Arch. Biochem. Biophys. 1992, 298(2), 646- 650.
20. Greenwell, L.; Moreno, T.; Jones, T.; Richards, R. Free Radical Bio. Med. 2002, 32(9), 898-905.
21. Guthrie Jr., G.D. Am. Mineral. 1992, 77, 225-243.
22. Gutteridge, J.M.C.; Mumby, S.; Quinlan, G.J.; Chung, K.F.; Evans, T.W. Biochem. Bioph. Res. Co. 1996, 220, 1024-1027.
23. Hardy, J.A.; Aust, A.E. Chem. Rev. 1995, 95, 97-118.
24. Kamp, D.W.; Weitzman, S.A. Thorax. 1999, 54, 638-652.
25. Kane, A.B.; Boffetta, P.; Wilbourn, J.D., Eds. Mechanisms of Fibre Carcinogenesis; IARC: Lyon, France, 1996.
26. Kremer, M.L. J. Phys. Chem. A. 2003, 107, 1734-1741.
27. Lloyd, R.V.; Hanna, P. M.; Mason, R. P. Free Radical Bio. Med. 1997, 22, 885- 888.
28. Lund, L.G.; Aust, A.E. Arch. Biochem. Biophys. 1990, 278(1), 61-64.
29. Martra, G.; Tomatis, M.; Fenoglio, I.; Coluccia, S.; Fubuni, B. Chem. Res. Toxicol. 2003, 16, 328-335.
30. McIntyre, N.S.; Cook, M.G. Anal. Chem. 1975, 47(13), 2208-2213.
216
31. McIntyre, N.S.; Zetaruk, D.G. Anal. Chem. 1977, 49(11), 1521-1529.
32. Minetti, M.; Forte, T.; Soriani, M.; Quarisima, V.; Menditto, A.; Ferrari, M. Biochem. J. 1992, 282, 459-465.
33. Mossman, B.T.; Gee, J.B.L. New Engl. J. Med. 1989, 320(26), 1721-1730.
34. Mossman, B.T.; Bignon, J.; Corn, M.; Seaton, A.; Gee, J.B.L. Science. 1990, 247, 294-301.
35. Nappi, A, J.; Vass, E. Biochem. Biophys. Acta. 1997, 1336, 295-301.
36. Sato, K.; Nishimura, Y.; Matsubayashi, N.; Imamura, M.; Shimada, H. Microporous and Mesoporous Mater. 2003, 59, 133-146.
37. Shen, Z.; Bosbach, D.; Hochella, Jr., M.F.; Bish, D.L.; Williams, Jr., M.G.; Dodson, R.F.; Aust, A.E. Chem. Res. Toxicol. 2000, 13, 913-921.
38. Shulka, A.; Gulumian, M.; Hei, T.; Kamp, D.; Rahman, Q.; Mossman, B.T. Mossman, B.T., Ed. Free Radical Bio. Med. 2003, 34(9), 1117-1129.
39. Slade, R.; Crissman, K.; Norwood, J.; Hatch, G. Exp. Lung Res. 1993, 19, 469- 484.
40. Veblen, D.R.; Wylie, A.G. In Reviews in Minerology; Gutherie, G.D.; Mossman, B.T. Eds; Bookcrafters, Inc.: Chelsea, MI. 1993, 28, 61-137.
41. Wagner, J.C.; Skidmore, J.W.; Hill, R.J.; Griffiths, D.M. Br. J. Cancer. 1985, 51, 727-730.
42. Winterbourn, C.C. Toxicol. Lett. 1995, 82/83, 969-974.
43. Winterbourn, C.C. Biochem. J. 1981, 198, 125-131.
44. Woodside, J.V. J. Clin. Pathol. 2001, 54, 176-186.
217 Chapter 3:
1. Adnadjevic, B.; Jovanovic, J.; Gajinov, S. J. Membrane Sci. 1997, 136, 173-179.
2. Alvaro, M.; Garcıa, H.; Corrent, S.; Scaiano, J.C. J. Phys. Chem. B. 1998, 102, 7530-7534.
3. Amao, Y. Microchim. Acta. 2003, 143, 1-12.
4. Anderson, M.W.; Klinowski, J. J. Chem. Soc. Faraday Trans. 1. 1986, 82, 1449- 1469.
5. Andre, P.; Capo, C.; Mege, J.L.; Benoliel, A.M.; Bongrand, P. Biochem. Biophys. Res. Commun. 1988, 151(2), 641-648.
6. Bacon, J.R; Demas, J.N. Anal Chem. 1987, 59(23), 2780-2785.
7. Barrer, R.M.; James, J.D. J. Phys. Chem. 1960, 64(l), 417-427.
8. Beyer, H.K. Molecular Sieves. 2002, 3, 204-208.
9. Beyer, H.K.; Belenykaja, I.M.; Hange, F. J. Chem. Soc. Faraday Trans. 1. 1985, 81, 2889-2901.
10. Black, D.; Deacon, G.B.; Thomas, N.C. Aust. J. Chem. 1982, 35, 2445-2453.
11. Bossi, M.L.; Daraio, M.E.; Aramendia, P.F. J. Photoch. Photobio. A. 1999, 120, 15-21.
12. Buck, S.M.; Xu, H.; Brasuel, M.; Philbert, M.A.; Kopelman, R. Talanta. 2004, 63(1), 41-59.
13. Cao, Y.; Koo, Yong-Eun L.; Kopelman, R. Analyst. 2004, 129(8), 745-750.
14. Carraway, E.R.; Demas, J.N.; DeGraff, B.A.; Bacon, J.R. Anal. Chem. 1991, 63, 337-342.
218 15. Chardon-Noblat, S.; Da Costa, P.; Deronzier, A.; Haukka, M.; Pakkanen, T.A.; Ziessel, R. J. Electroanal. Chem. 2000, 490, 62-69.
16. Chen, X.; Zhong, Z.; Li, Z.; Jing, Y.; Wang, X.; Wong, K. Sens. Actuators B. 2002, 87(2), 233-238.
17. Cheng, Z.; Aspinwall, C.A. Analyst. 2006, 131, 236-243.
18. Clark, H.A.; Barker, S.L.R.; Brasuel, M.; Miller, M.T.; Monson, E.; Parus, S.; Shi, Z-Y.; Song, A.;Thorsrud, B.; Kopelman, R.; Ade, A.; Meixner, W.; Athey, B.; Hoyer, M.; Hill, D.; Lightle, R.; Philbert, M.A. Sens. Actuators B. 1998, 51, 12-16.
19. Collomb-Dunand-Sauthier, M.L.; Deronzier, A. J. Electroanal. Chem. 1991, 319, 347-353.
20. Coutant, M.A.; Payra, P.; Dutta, P.K. Microporous and Mesoporous Mater. 2003, 60(1-3), 79-90.
21. de Klerk-Engels, B.; Fruhauf, H.W.; Vrieze, K.; Kooijman, H.; Spek, A.L. Inorg. Chem. 1993, 32, 5528-5535.
22. Demas, J.N; DeGraff, B.A.; Coleman, P.B. Anal. Chem. 1999, 23, 793A-800A.
23. Demas, J.N.; Taylor, D.G. Inorg. Chem. 1979, 18, 3177-3181.
24. DeWilde, W.; Peeters, G.; Lunsford, J.H. J. Phys. Chem. 1980, 84, 2306-2310.
25. Douglas, P.; Eaton, K. Sens. Actuators B. 2002, B82(2-3), 200-208.
26. Engelhardt, G.; Michel, D. High-resolution solid-state NMR of silicates and zeolites; John Wiley & Sons: Engelhardt, Germany, 1987.
27. Gao, Z.; Nahrup, J.S.; Mark, J.E.; Sakr, A. J. Appl. Polym. Sci. 2003, 90, 658- 666.
28. Gevers, L.E.M.; Vankelecom, I.F.J.; Jacobs, P.A. J. Membrane Sci. 2006, 278, 199-204.
219 29. Hartmann, P.; Leiner, M.J.P.; Lippitsch, M.E. Sens. Actuators B. 1995, B29(1- 3), 251-257.
30. Haukka, M.; Kiviaho, J.; Ahlgren, M.; Pakkanen, T. Organometallics, 1995, 14, 825-833.
31. Herron, N.; Stucky, G.D.; Tolman, C.A. Inorg. Chim. Acta. 1985, 100, 135-140.
32. Hotz, J.; Meier, W. Langmuir 1998, 14(5), 1031-1036.
33. Hriljac, J.A.; Eddy, M.M.; Cheetham, A.K.; Donohue, J.A.; Ray, G.J. J. Solid State Chem. 1993, 106(1), 66-72.
34. James, P.E.; Grinberg, O.Y.; Swartz, H.M. Journal of Leukocyte Biology. 1998, 64, 78-84.
35. Kaupp, G. Cryst. Eng. Comm. 2003, 5(23), 117-133.
36. Kelch, S.; Rehahn, M. Macromolecules. 1997, 30, 6185-6193.
37. Kelly, J.M.; O’Connell, C.M.; Vos, J.G. Inorg. Chim. Acta. 1982, 64, L75-L76.
38. Klimant, I.; Wolfbeis, O. Anal. Chem. 1995, 67, 3160-3166.
39. Klinowski, J.; Thomas, J.M.; Fyfe, C.A.; Gobbi, G.C.; Hartman, J.S. Inorg. Chem. 1983, 22, 63-66.
40. Koo, Y.E.L.; Cao, Y.; Kopelman, R.; Koo, S.M.; Brasuel, M; Philbert, M.A. Anal. Chem. 2004, 76, 2498-2505.
41. Laine, P; Lanz, M.; Calzaferri, G. Inorg. Chem. 1996, 35, 3514-3518.
42. Li, L.; Walt, D.R. Anal. Chem. 1995, 67, 3746-3752.
43. Long, J.F.; Waldman, W.J.; Kristovich, R.; Williams, M.; Knight, D.; Dutta, P.K. Environ. Health Persp. 2005, 113: 170-174.
44. McNamara, K.P.; Nguyen, T.; Dumitrascu, G.; Ji, J.; Rosenzweig, N.; Rosenzweig, Z. Anal. Chem. 2001, 73(14), 3240-3246.
220
45. Meier, B.; Werner, T.; Klimant, I.; Wolfbeis, O.S. Sensor. Actuat. B-Chem. 1995, B29(1-3), 240-245.
46. Moreno-Bondi, M.C.; Wolfbeis, O.S. Anal Chem. 1990, 62, 2377-2380.
47. Pang, H.L.; Kwok, N.Y.; Chow, L.M.C.; Yeung, C.H.; Wong, K.Y.; Chen, X.; Wang, X. Sens. Actuators B. 2007, 123(1), 120-126.
48. Pasic, A.; Koehler, H.; Klimant, I.; Schaupp, L. Sens. Actuat.B. 2007, 122, 60- 68.
49. Pasmans, F.; De Herdt, P.; Nerom, A.V.; Haesebrouck, F. Dev. Comp. Immunol. 2001, 25, 159-168.
50. Payra, P.; Dutta, P.K. Microporous and Mesoporous Mater. 2003, 64, 109-118.
51. Quayle, W.H.; Lunsford, J.H.; Peeters, G.; De Roy, G.L.; Vansant, E.F. Inorg. Chem. 1982, 21, 2226-2231.
52. Ray, G.J.; Nerheim, A.G. Zeolites. 1988, 8, 458-463.
53. Regen, S.L.; Czech, B.; Singh, A. J. Am. Chem. Soc. 1980, 102(21), 6638-6640.
54. Rosenzweig, Z.; Kopelman, R. Sens. Actuators B. 1996, 35-36, 475-483.
55. Ruda, T.A.; Dutta, P.K. Environ. Sci. Technol. 2005, 39(16), 6147-6152.
56. Sykora, M; Kincaid, J.R.; Dutta, P.K.; Castagnola, N.B. J. Phys. Chem. B, 1999, 103, 309-320.
57. Tang, Y.; Tehan, E.C.; Tao, Z.; Bright, F.V. Anal Chem. 2003, 75, 2407-2413.
58. Thummel, R.P.; Lefoulon, F.; Chirayil, S. Inorg. Chem. 1987, 26, 3072-3074.
59. Vankelecom, I.F.J.; Merckx, E.; Luts, M.; Uytterhoeven, J.B. J. Phys. Chem. 1995, 99, 13187-13192.
221 60. Vankelecom, I.F.J.; Scheppers, E.; Heus, R.; Uytterhoeven, J.B. J. Phys. Chem. 1994, 98, 12390-12396.
61. Waldman, W.J.; Adams, P.W.; Orosz, C.G.; Sedmak, D.D. Transplantation. 1992, 54, 887-896.
62. Waldman, W.J.; Kristovich, R.; Knight, D.A.; Dutta, P.K. Chem. Res. Toxicol. 2007, 20(8), 1149-1154.
63. Xu, H.; Aylott, J.W.; Kopelman, R.; Miller, T.J.; Philbert, M.A. Anal. Chem. 2001, 73, 4124-4133.
64. Xu, W.; McDonough, R.C.; Langsdorf, B.; Demas, J.N.; DeGraff, B.A. Anal. Chem. 1994, 66, 4133-4141.
Chapter 4:
1. Alvaro, M.; Garcia, H.; Corrent, S.; Scaiano, J.C. J. Phys. Chem. B. 1998, 102(39), 7530-7534.
2. Anderson, M.W.; Klinowski, J. J. Chem. Soc., Faraday Trans. 1986, 82(5), 1449-1469.
3. Avantes. Fiber Optics. http://www.avantes.com/ (accessed 9/5/07).
4. Bernhard, D.D.; Mall, S.; Pantano, P. Anal. Chem. 2001, 73, 2484-2490.
5. Buck, S.M.; Xu, H.; Brasuel, M.; Philbert, M.A.; Kopelman, R. Talanta. 2004, 63(1), 41-59.
6. Costa-Fernandez, J.M.; Diaz-Garcia, M.E.; Sanz-Medel, A. Anal. Chim. Acta. 1998, 360(1-3), 17-26.
7. Demagh, N.; Guessoum, A.; Aissat, H. Meas. Sci. Technol. 2006, 17(1), 119- 122.
8. Force, Inc. Types of Optical Fiber. http://www.fiber-optics.info/articles/fiber- types.htm (accessed 9/5/07). 222
9. Garcia-Fresnadillo, D.; Marazuela, M.D.; Moreno-Bondi, M.C.; Orellana, G. Langmuir. 1999, 15, 6451-6459.
10. Griffiths, J.R.; Robinson, S.P. Br. J. Radiol. 1999, 72, 627-630.
11. Hoffman, P.; Dutoit, B.; Salathe, R.-P. Ultramicroscopy. 1995, 61, 165-170.
12. Ji, J.; Rosenzweig, N.; Jones, I.; Rosenzweig, Z. Anal. Chem. 2001, 73, 3521- 3527.
13. Jorge, P.A.S.; Caldas, P.; Rosa, C.C.; Oliva, A.G.; Santos, J.L. Sens. Actuators, B. 2004, B103(1-2), 290-299.
14. Jorge, P.A.S.; Rosa, C.C.; Caldas, P.; Oliva, A.G.; Esteves da Silva, J.C.G.; Santos, J.L.; Farahi, F. Fiber Integr. Opt. 2005, 24, 201-225.
15. Kulak, A.; Lee, Y-J.; Park, Y.S.; Kim, H.S.; Lee, G.S.; Yoon, K.B. Adv. Mater. 2002, 14(7), 526-529.
16. Lambelet, P.; Sayah, A.; Pfeffer, M.; Philipona, C.; Marquis-Weible, F. Appl. Opt. 1998, 37(31), 7289-7292.
17. Liang, D.; Readey, D.W. J. Am. Ceram. Soc. 1987, 70(8) 570–577.
18. Liu, N.; Hui, J.; Sun, C.; Dong, J.; Zhang, L.; Xiao, H. Sensors. 2006, 6, 835- 847.
19. Machavaram, V.R.; Badcock, R.A.; Fernando, G.F. Sens. Actuators, A. 2007, 138(1), 248-260.
20. MacCraith, B.D.; McDonagh, C.M.; O'Keeffe, G.; Keyes, E.T.; Vos, J.G.; O'Kelly, B.; McGilp, J.F. Analyst. 1993, 118(4), 385-388.
21. Mills, A; Chang, Q. Analyst. 1992, 117(9), 1461-1466.
22. Murphy, V.; MacCraith, B.D.; Butler, T.; McDonagh, C.; Lawless, B. Electron. Lett. 1997, 33(7), 618-619.
223 23. Ocean Optics. Ocean Optics Introduces Handheld Oxygen Sensor System. http://www.oceanoptics.com/corporate/pressreleases/2004_Releases/foxylite_rele ase.asp (accessed 9/5/07).
24. Oxford Optronix Ltd. OxyLite E™ Specifications. http://www.oxford- optronix.com/pdfs/Technical/OxyLiteE%20Spec.pdf (accessed 9/5/07).
25. Oxford Optronix Ltd. Frequently Asked Questions. http://www.oxford- optronix.com/pdfs/Technical/TM%20FAQ.pdf (accessed 9/5/07).
26. Pantano, P.; Walt, D.R. Anal. Chem. 1995, 67(15), 481A-487A.
27. Park, E.J.; Reid, K.R.; Tang, W.; Kennedy, R.T.: Kopelman, R. J. Mater. Chem. 2005, 15, 2913–2919.
28. Payra, P.; Dutta, P.K. Microporous Mesoporous Mater. 2003, 64(1-3), 109-118.
29. Pradhan, A.R.; Macnaughtan, M.A.; Raftery, D. J. Am. Chem. Soc. 2000, 122(2), 404-405.
30. Pradhan, A.R.; Uppili, S.; Shailaja, J.; Sivaguru, J.; Ramamurthy, V. Chem. Commun. 2002, 6, 596–597.
31. Quah, S.; Danowski, K.L.; Pantano, P. Biomed. Microdevices. 2002, 4(2), 123- 130.
32. Rosenzweig, Z.; Kopelman, R. Anal. Chem. 1995, 67, 2650-2654.
33. Stockle, R; Fokas, C.; Deckert, V.; Zenobi, R.; Sick, B; Hecht, B; Wild, U.P. Appl. Phys. Lett. 1999, 75(2), 160-162.
34. Tso, S.T.; Pask, J.A. J. Am. Ceram. Soc. 1982, 65(7), 360–362.
35. Unger, M.A.; Kossakovski, D.A.: Kongovi, R.; Beauchamp, J.L.; Baldeschwieler, J.D.; Palanker, D.V. Rev. Sci. Instrum. 1998, 69(8), 2988-2993.
36. Vander Donckt, E.; Camerman, B.; Herne, R.; Vandeloise, H.R. Sens. Actuators, B. 1996, 32, 121–127.
224 37. Vasylevska, G.S.; Borisov, S.M.; Krause, C.; Wolfbeis, O.S. Chem. Mater. 2006, 18, 4609-4616.
38. Wolfbeis, O.S.; Weis, L.J.; Leiner, M.J.P.; Ziegler, W.E. Anal. Chem. 1988, 60(19), 2028-2030.
39. Xi, C.; Zhong, Z.; Li, Y.; Jiang, X.; Wang, K.W. Sens. Actuators, B. 2002, 87, 233–238.
40. Zhang, J.; Dong, J.; Luo, M.; Xiao, H.; Murad, S.; Normann, R.A. Langmuir 2005, 21, 8609-8612.
225