USF Libraries
USF Digital Collections

Nanomaterial sensing layer based surface acoustic wave hydrogen sensors

MISSING IMAGE

Material Information

Title:
Nanomaterial sensing layer based surface acoustic wave hydrogen sensors
Physical Description:
Book
Language:
English
Creator:
Srinivasan, krishnan
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Saw resoantor
Tobacco mosaic virus
Carbon nanotubes
Palladium
Adsorption
Dissertations, Academic -- Electrical Engineering -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: This thesis addresses the design and use of suitable nanomaterials and surface acoustic wave sensors for hydrogen detection and sensing. Nanotechnology is aimed at design and synthesis of novel nanoscale materials. These materials could find uses in the design of optical, biomedical and electronic devices. One such example of a nanoscale biological system is a virus. Viruses have been given a lot of attention for assembly of nanoelectronic materials. The tobacco mosaic virus (TMV) used in this research represents an inexpensive and renewable biotemplate that can be easily functionalized for the synthesis of nanomaterials. Strains of this virus have been previously coated with metals, silica or semiconductor materials with potential applications in the assembly of nanostructures and nanoelectronic circuits.Carbon nanotubes are another set of well-characterized nanoscale materials which have been widely investigated to put their physical and chemical properties to use in design of transistors, gas sensors, hydrogen storage cells, etc. Palladium is a well-known material for detection of hydrogen. The processes of absorption and desorption are known to be reversible and are known to produce changes in density, elastic properties and conductivity of the film. Despite these advantages, palladium films are known to suffer from problems of peeling and cracking in hydrogen sensor applications. They are also required to be cycled for a few times with hydrogen before they give reproducible responses. The work presented in this thesis, takes concepts from previous hydrogen sensing techniques and applies them to two nanoengineered particles (Pd coated TMV and Pd coated SWNTs) as SAW resonator sensing materials.Possible sensing enhancements to be gained by using these nanomaterial sensing layers are investigated. SAW resonators were coated with these two different nano-structured sensing layers (Pd-TMV and Pd-SWNT) which produced differently useful hydrogen sensor responses. The Pd-TMV coated resonator responded to hydrogen with nearly constant increases in frequency as compared to the Pd-SWNT coated device, which responded with concentration-dependent decreases in frequency of greater magnitude upon hydrogen exposure. The former behavior is more associated with acousto-electric phenomena in SAW devices and the later with mass loading. The 99% response times were 30-40 seconds for the Pd-TMV sensing layer and approximately 150 seconds for the Pd-SWNT layer. Both the films showed high robustness and reversibility at room temperature.When the Pd film was exposed to hydrogen it was observed that it produced decreases in frequency to hydrogen challenges, conforming to mass loading effect. It was also observed that the Pd film started degrading with repeated exposure to hydrogen, with shifts after each exposure going smaller and smaller.
Thesis:
Thesis (M.S.E.E.)--University of South Florida, 2005.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by krishnan Srinivasan.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 81 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001709562
oclc - 69396486
usfldc doi - E14-SFE0001325
usfldc handle - e14.1325
System ID:
SFS0025646:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Nanomaterial Sensing Layer Based Su rface Acoustic Wave Hydrogen Sensors by Krishnan Srinivasan A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Electrical Engineering Department of Electrical Engineering College of Engineering University of South Florida Co-Major Professor: Venkat R. Bhethanabotla, Ph.D. Co-Major Professor: Paris H. Wiley, Ph.D. Thomas Weller, Ph.D. Babu Joseph, Ph.D. Date of Approval: October 13, 2005 Keywords: Saw Resoantor, Tobacco Mosa ic Virus, Carbon Nanotubes, Palladium, Adsorption Copyright 2005, Krishnan Srinivasan

PAGE 2

DEDICATION This dissertation is dedicated to my parents, G. Srinivasan and Lakshmi Srinivasan and to my sister Kavita Srinivasa n. Their love, support and patien ce have always been the foundations of my inspiration.

PAGE 3

ACKNOWLEDGEMENTS I wish to express sincere appreciation to my major professor Dr. Venkat Bhethanabotla for his valuable guidance and support during this thesis work. Thanks to Dr Paris H. Wiley, Dr. Thomas Weller and Dr. Babu Joseph for agreeing to serve on my supervisory committee. I would like to thank Stefan Cular and Randy Williams for helping me at every step. I also appreciate Dr. Michael T Harris and Sang-Yup Lee of Purdue University for providing the Pd-TMV materials. Special thanks are extended to Sensors Rese arch Group and my friends for their support and encouragement throughout this research.

PAGE 4

i TABLE OF CONTENTS LIST OF TABLES.............................................................................................................iii LIST OF FIGURES...........................................................................................................iv ABSTRACT......................................................................................................................v ii CHAPTER 1: INTRODUCTION.......................................................................................1 1.1. Sensors.................................................................................................................1 1.2. Gas Sensors..........................................................................................................2 1.3. Hydrogen Sensors................................................................................................3 1.4. Acoustic Wave Gas Sensors................................................................................5 1.5. Thesis Outline......................................................................................................7 CHAPTER 2: THEORY.....................................................................................................8 2.1. Saw Devices.........................................................................................................8 2.2. Two-Port Saw Resonator.....................................................................................8 2.3. Operation............................................................................................................11 2.4. Saw Preturbation Mechanisms...........................................................................12 2.5. Saw Mass Loading.............................................................................................13 2.6. Saw Acousto-electric Effect..............................................................................15 2.7. Saw Sensor Operation........................................................................................17 2.8. Oscillator Based Measurement Method............................................................18 2.9. Type of Measurement........................................................................................19 2.10. Understanding Sensor Response........................................................................20 2.11. Film Preperation.................................................................................................21 2.12. Gas Delivery......................................................................................................21 2.13. Mass Flow Controller Operation.......................................................................22 2.14. Saw Housing......................................................................................................23

PAGE 5

ii 2.15. Sensing Layer.....................................................................................................24 2.16. Carbon Nanotubes..............................................................................................24 2.17. Single Walled Nanotubes (SWNTS).................................................................25 2.18. Multi Walled Nanotubes (MWNTS).................................................................25 2.19. Tobacco Mosaic Virus (TMV)...........................................................................26 CHAPTER 3: EXPERIMENTAL SETUP.......................................................................28 3.1. Gas Dilution System..........................................................................................28 3.2. Calibration of Mass Flow Controllers................................................................29 3.3. Test Cell.............................................................................................................32 3.4. Saw Resonator – PCB Design............................................................................34 3.5. Oscillator Circuit................................................................................................35 3.5.1 Amplifier……………………………………………………….. 36 3.5.2 Filter……………………………………………………………. 36 3.5.3 Attenuator……………………………………………………… 36 3.5.4 Phase Shifter…………………………………………………… 36 3.5.5 Coupler……………………………………………………........ 38 3.6. Automation........................................................................................................38 CHAPTER 4: RESULTS..................................................................................................42 4.1. Tobacco Mosaic Virus.......................................................................................42 4.2. Carbon Nanotubes..............................................................................................51 4.3. Palladium Film...................................................................................................55 4.4. Comparison of Films..........................................................................................59 CHAPTER 5: CONCLUSION.........................................................................................62 REFERENCES.................................................................................................................65 APPENDICES..................................................................................................................68 APPENDIX A PRINTED CIRCUIT FABRICATION..............................................69

PAGE 6

iii LIST OF TABLES Table 1.1. Measurement Format..........................................................................................3 Table 2.1. Measurement Techniques Used by Different Gas Delivery Systems...............22 Table 3.1. MFC Gauge Factor...........................................................................................30 Table 3.2. Scaling Factor for Hydrogen Based on Flow Rates..........................................32 Table 4.1. SAW Resonator Frequency and Atte nuation Measurements after Pd-TMV Coating .............................................................................................................45 Table 4.2. SAW Resonator Frequency and Atte nuation Measurements after Pd-SWNT Coating .............................................................................................................51 Table 4.3. SAW Resonator Frequency and Attenu ation Measurements after Pd Coating .............................................................................................................55 Table 4.4. Comparison of Different Sensor Pa rameters for the Three Sensing Films.......60

PAGE 7

iv LIST OF FIGURES Figure 2.1. (A) DelayLine (B) Two-Port SAW Resonator (C) Single-Port SAW Resonator..........................................................................................................9 Figure 2.2. (A) Lumped Equivalent Ci rcuit for 0 Phase Shift between Input/Output IDT’s at Resonance. (B) Ideal 1:-1 Transformer Included for 180 Phase Shift a Resonance....................................................10 Figure 2.3. Two-Port SAW Resonator...............................................................................10 Figure 2.4. Equivalent Circuit Model Desc ribing the Interaction between SAW and Charge Carriers on Film Overlay.............................................................15 Figure 2.5. Oscillator Block Diagram................................................................................19 Figure 2.6. The Mass Flow Controller...............................................................................23 Figure 2.7. S.E.M. Image of Carbon Nanotubes...............................................................26 Figure 2.8. Microscopic Image of Single Walled Carbon Nanotubes...............................26 Figure 2.9. (A) Armchair (5,5), (B ) Zigzag (10,0) (C) Chiral(10,5)................................27 Figure 2.10. High Resolution T.E.M. Imag e of Multi-Walled Carbon Nanotubes..........27 Figure 2.11. Electron Microscope View of a Single TMV Tube.......................................27 Figure 2.12. T.E.M. View of a Pd Coat ed TMV Sample (Scale Bar200nm)..................27 Figure 3.1. Mass Flow Controller Setup............................................................................28 Figure 3.2. The Complete Experimental Setup..................................................................29 Figure 3.3. Schematic of the Cell Design..........................................................................32 Figure 3.4. The Machined Stainless Steel Cell..................................................................33 Figure 3.5. The Top and Bottom Surface of PCB Designed in AutoCAD. The Bottom Side is One Facing the Cell................................................................34 Figure 3.6. The Final PCB Design (Bottom).....................................................................35 Figure 3.7. The Final PCB Design (Top)...........................................................................35 Figure 3.8. PCB Design for Tw o 180 Degree Phase Shifter Connected in Series............37

PAGE 8

v Figure 3.9. The Final PCB Design of the Phase Shifter....................................................37 Figure 3.10. The Oscillator Circuit....................................................................................38 Figure 3.11. The LABVIEW Program for Writi ng Data Automatically into a Text File.................................................................................................................39 Figure 3.12. The Front Panel Di splay of LABVIEW Program.........................................40 Figure 3.13. The LABVIEW Program to Cont rol the Mass Flow Controllers and Log Sensor Parameters..................................................................................41 Figure 4.1. The Two-Port Characteristic s of SAW RP1239 Two-Port Resonator............42 Figure 4.2. The Two-Port Characteristic s of SAW RP1239 Two-Port Resonator after TMV-Pd Coating....................................................................................43 Figure 4.3. The Two-Port Characteristic s of SAW RP1239 Two-Port Resonator after Si02 Coating............................................................................................44 Figure 4.4. The Response of Pd Co ated TMV to 0.2-2.5 % Hydrogen.............................46 Figure 4.5. The Response Times for Pd Co ated TMV Resonator in Response to 0.2-2.5% Hydrogen.........................................................................................46 Figure 4.6. The Frequency Shifts Produced for Two Different Runs at Various Hydrogen Concentrations...............................................................................47 Figure 4.7. The Sensing Film Response to 2-4.5% Hydrogen...........................................48 Figure 4.8. The Response Times of Se nsing Film for 2-4.5% Hydrogen..........................48 Figure 4.9. The Frequency Shift Produ ced in the Sensing Film 2.5-4.5% Hydrogen.........................................................................................................49 Figure 4.10. Cracking of Palladium Film Upon Hydrogen Sorption/Desorption..............50 Figure 4.11. Device Surface Coated with Pa lladium Nanoparticle-Coated Tobacco Mosaic Virus Film Fr ee of Rearrangement...................................................50 Figure 4.12. The Curve Showing the Respons e of Pd Coated SWNTs Response to Hydrogen.......................................................................................................52 Figure 4.13. Response Times for Pd-SWNTs Coated SAW Resonator to 0.5-2.5% Hydrogen Exposures.....................................................................................52 Figure 4.14. Frequency Shifts for Pd-SWNTs Coated SAW Resonator to 0.2-2.5% Hydrogen Exposures.....................................................................................53 Figure 4.15. Frequency Shifts for Pd-SWN Ts Coated SAW Resonator to 1.5 % Hydrogen.......................................................................................................54

PAGE 9

vi Figure 4.16. Frequency Shifts for Pd-SWN Ts Coated SAW Resonator to 1 % Hydrogen.......................................................................................................54 Figure 4.17. The Curve Showing the Res ponse of Pd Film on Exposure to 0.22.5% Hydrogen..............................................................................................56 Figure 4.18. Frequency Shifts for Pure Pd Coated SAW Resonator to 1-2.5% Hydrogen Exposures.....................................................................................57 Figure 4. 19. Response Times for Pure Pd Coated SAW Resonator to 1-2.5% Hydrogen Exposures.....................................................................................57 Figure 4.20. Response Times for three differe nt runs of Pure Pd Coated SAW Resonator to 1-2.5% Hydrogen Exposures...................................................58 Figure 4.21. Frequency Shifts for three diffe rent runs of Pure Pd Coated SAW Resonator to 1-2.5% Hydrogen Exposures...................................................58 Figure 4.22. The Response Times Observed fo r the Sensing Films for 63% of Full Response........................................................................................................59 Figure 4.23. The Frequency Shifts Observed for the Sensing Films for 63% of Full Scale Response.......................................................................................60

PAGE 10

vii Nanomaterial Sensing Layer Based Surfa ce Acoustic Wave Hydrogen Sensors Krishnan Srinivasan ABSTRACT This thesis addresses the design and use of suitable nanomaterials and surface acoustic wave sensors for hydrogen detection and sensing. Nanotechnology is aimed at design and synt hesis of novel nanoscale materials. These materials could find uses in the desi gn of optical, biomedical and electronic devices. One such example of a nanoscale biological system is a virus. Viruses have been given a lot of attention for assembly of na noelectronic materials. The tobacco mosaic virus (TMV) used in this research represen ts an inexpensive and renewable biotemplate that can be easily functionalized for the synthe sis of nanomaterials. Strains of this virus have been previously coated with metals, silica or semiconductor materials with potential applications in the assembly of nanostr uctures and nanoelectronic circuits. Carbon nanotubes are another set of well-characteri zed nanoscale materials which have been widely investigated to put their physical and chemical properties to use in design of transistors, gas sensors, hydr ogen storage cells, etc. Palladi um is a well-known material for detection of hydrogen. The processes of absorption and desorption are known to be reversible and are known to produce chan ges in density, elas tic properties and conductivity of the film. Desp ite these advantages, pallad ium films are known to suffer from problems of peeling and cracking in hyd rogen sensor applications. They are also required to be cycled for a few times w ith hydrogen before they give reproducible responses. The work presented in this thesis, take s concepts from previous hydrogen sensing techniques and applies them to two nanoengi neered particles (Pd coated TMV and Pd coated SWNTs) as SAW resonator sensing materials. Possible sensing

PAGE 11

viii enhancements to be gained by using these na nomaterial sensing laye rs are investigated. SAW resonators were coated with these two different nano-structured sensing layers (PdTMV and Pd-SWNT) which produced differently useful hydrogen se nsor responses. The Pd-TMV coated resonator responded to hydrog en with nearly constant increases in frequency as compared to the Pd-SWN T coated device, which responded with concentration-dependent decreases in fr equency of greater magnitude upon hydrogen exposure. The former behavior is more a ssociated with acousto-e lectric phenomena in SAW devices and the later with mass lo ading. The 99% response times were 30-40 seconds for the Pd-TMV sensing layer and approximately 150 seconds for the Pd-SWNT layer. Both the films showed high robustnes s and reversibility at room temperature. When the Pd film was exposed to hydrogen it was observed that it produced decreases in frequency to hydrogen challenges, conformi ng to mass loading effect. It was also observed that the Pd film started degradi ng with repeated exposure to hydrogen, with shifts after each exposure going smaller and smaller.

PAGE 12

1 CHAPTER 1: INTRODUCTION 1.1. Sensors A Sensor can be defined as a biological or gan or a technological device capable of detecting and/or sensing a chemical com pound or physical conditions. Most sensors convert these physical parameters into a meaningful format, usually an electrical signal which can be easily interpreted. There are tw o types of sensors, biological sensors and artificial or man made sensor s. Human body can be said to be a biological sensor, some of which include eyes, nose, ears, tongue etc. All these sensors res pond to some form of external stimuli. For example, when we en ter a bright rooms from a dark one, human eyes shrink making it easy to adjust to higher intensity of light, or the nose, which is highly sensitive to different types of smells, thus making it easier for us to distinguish among them. As we head toward the future developing newer products and technologies there is a desire to develop better sens ors capable of doing whole range of sensing applications some of which imitate biological sensors. Such sensors are called artificial sensors. These sensors are either direct indicating such as a mercury thermometer or an electrical meter, or they may be coupled with an indicator to make the value sensed to be represented in a more meaningful format. Mo st artificial sensors are classified on the basis of energy they detect. For example: 1. Light sensors: Capable of converting light energy into el ectrical energy. Photocell, photodiode, image sensor are some of light sensors. Gas Sensors and liquid flow meter: Capable of detect ing gas and liquid fl ow. Flow sensor, anemometer, gas meter and flow meter ar e some examples of such sensors. 2. Temperature sensor: Capable of detecting ei ther direct temperature of changes in temperature. Mercury thermometer, ther mocouple and thermostats are examples of such sensors.

PAGE 13

2 3. Motion Sensors: Capable of detecting and measuring linear or angular motion. Tachometer, radar gun, speedometer are examples of such sensors. 4. Pressure sensor: Capable of measuring direct or change s in pressure. Barometer, barograph, pressure gauge are examples of such sensors. 5. Chemical Sensors: Capable of detecting certain chemical reactions and detecting presence of chemicals. 6. Orientation Sensor: Capable of detecting orientation in two or three dimensional space. Gyroscope, artificial horizon, ring laser gyroscope are example of such sensors. These are some commonly used sensors with hun dreds of more different type of sensors existing in industry. 1.2. Gas Sensors Gas sensors are devices that interact with gases and give data in a more meaningful format. Important measurement sp ecifications to consider when looking for gas sensors as detectors include the response time, the distance, and the flow rate. The response time is the amount of time required for process the signal from the instant the gas comes in contact. Distance is the maxi mum distance between the position of sensors and gas leak to which the device is sensitive, while flow rate in the necessary flow rate of the gas across the sensor to which it is sensi tive. Some of the commonly used gas sensors are for ammonia, aerosols, arsine, bromin e, carbon dioxide, car bon monoxide, chlorine, chlorine dioxide, diborane, dust, fluorine, germane, halocarbons or refrigerants, hydrocarbons, hydrogen, hydrogen chloride, hydrogen cyanide, hydrogen fluoride, hydrogen selenide, hydrogen sulfide, mercury va por, nitrogen dioxide nitrogen oxides, nitric oxide, organic solvents, oxygen, ozone, phosphine, silane, sulf ur dioxide and water vapor. The measurements made by such instru ments could include percent volume, trace, leakage, consumption, density, spectra, lowe r explosive limit or flammable limit.

PAGE 14

3 Table 1.1. Measurement Format Gas Property Format(Units) Leakage ml/min Trace Concentrationppm Consumption ml/L/hr Gas Spectra Chromatograph Density Mg/m3 Common outputs from gas sensor s include analog voltage, pulse signals, analog currents and switch or relays. Some key factors to be considered for gas sensors include temperature and operating humidity. Market for Gas Sensors In the year 2004, the global market for gas sensors and gas metering equipment was estimated to be about at $2.8 billion and is expected to grow at average annual growth rate of 5.9% to reach $3.8 billion in 2009. Of these, the sensor monitoring market is estimated to grow at an average annual gr owth rate of 6.3% to reach $1.8 billion in 2009 from $1.3 billion in 2004. The secondary inst rumentation is also expected to reach $512 million in 2009, rising at an average annual growth rate of 7.2% [1]. 1.3. Hydrogen Sensors Fossil fuel has been in use for a long time sustaining economies the world over, but with increasing consumption and price, and ever decreasing reserves, it has led us to explore for newer alternatives of fuel. Hydroge n is an important industrial fuel being used as feedstock for chemical, food, metallurgi cal, and electronic industries with current production exceeding 500 billion Nm3 [2]. Hydrogen is a clea n, reliable, and affordable energy supply that can sustain world econom y and environment. There has been ever growing research in utiliz ing hydrogen energy for trans portation, electric power generation, and portable electronic devices su ch as mobile phones and laptop computers

PAGE 15

4 [3]. With prototype hydrogenfuelled cars on the road there is an increasing need for large scale production, distribution and network for hydrogen gas. An economy powered by hydroge n has many advantages: 1. It is a clean technology. The only bypr oduct is water thus eliminating environmental dangers such as oil spills. 2. If the hydrogen comes from the electrolysis of water, then hydrogen adds no greenhouse gases to the environment. There is a perfect cycle wherein electrolysis produces hydrogen from water, and th e hydrogen recombines with oxygen to create water and power in a fuel cell. However, a non-polluting energy source is necessary for he electrolysis, such as solar photoelectric. 3. Hydrogen can be produced anywhere that you have electricity and water, thus enabling distributed production. Though with numerous advantages offe red by an economy sustained by hydrogen there are also a number of problems that n eed to be addressed. Cost and performance issues associated with hydrogen energy system s will need to be addressed in tandem with customer awareness and acceptance [3]. Key issues include consumer safety, convenience and affordability. Developing hydrogen as a major energy carrier, however, will require solutions to many challenges in the areas of infrastructure, technology, and economics. Cost, safety, and reliability issu es will influence th e planning, design, and development of central versus distributed production and delivery. One major problems of hydrogen is if mixed with air in the ratio 4.65:93.9 vol% it could be explosive [4, 5]. Different types of hydrogen sensors [4]: 1. Piezoelectric sensors: Such sensors allow transduction between electrical and acoustic energies based on piezoelectric cr ystals. Commonly used sensors in this category include quartz crystal microba lance (QCM) and surface acoustic wave sensor (SAW). Whenever there is absorp tion of hydrogen by palladium film on be

PAGE 16

5 these devices there is an attenuation of signal and change in frequency which can be calibrated to determine the amount of gas absorption. 2. Fiber optic sensors: In such sensors the palladium coated fiber produces a change in light guided within it upon of hydr ogen absorption. The response can be measured in terms of absorbance, re flectance, luminescence, or scattering 3. Electrochemical sensors: Solid state elec trochemical cells are used for continuous measurements of flowing hydrogen. Sin ce the hydrogen provides a proton path, equivalent to electrolytic ionic conducti on the measured, current can be used to monitor hydrogen/proton concentration. 4. Semiconductor Sensors: The semiconductor chemical sensor is based on the metal oxide semiconductor (MOS) junction princi ple. MOS sensors are of two types, MOS capacitors and MOS transistors. H ydrogen molecules when adsorbed on a metal surface act as dipoles and they give rise to macroscopic measurable voltage drop. The dipole layer shifts the energy leve ls at the metal-insulator interface. The voltage is recorded as a function of ti me for various hydrogen concentrations. 5. Resistive sensors: Such sensors are sensit ive to change in resistance of a metal film upon the absorption of hydrogen. Such sensors use very simple films of Pd, tin oxide, etc where change in resistance is related to the hydrogen concentration. The oxide sensors often operate at very high temperatures. 1.4. Acoustic Wave Gas Sensors Acoustic wave sensors have been in research for a long time. These are constructed primarily on two platforms, su rface or bulk acoustic wave (SAW or BAW) devices used as delay-lines or resonators and thickness shear mode (TSM) bulk wave resonators. These devices are typically made on a piezoelectric substrate. This substrate is coated with a gas absorbing film. The abso rption of target gas molecules by the film cause the signal to be attenuated and the fr equency to change. Th e use of BAW devices was first reported in year 1964 by King [6]. He coated the device with stationary phase materials and incorporated it into gas liq uid chromatograph instrument as detector. Similar concept was used for many years in designing NH3, SO2, CO2 sensors. The first

PAGE 17

6 use of SAW device for chemical sensing was published in 1979 by Wohltjen and Dessy [7] [8]. They discussed the deposition of sorptive films on the surface. It was soon established that the sensitivity of SAW de vices was higher than BAW device because of their higher frequencies. It was also s hown theoretically that the absolute mass sensitivity of 3e-15 g could be achieved using SAW devices [9]. SAW devices are favored for use in chemical sensing applicat ions because of their small size, low cost, high sensitivity and reliabilit y. Most SAW chemical sensors monitor changes of the SAW phase velocity and attenuation as the vapor interacts with the sensing layer. The shift in the phase velocity and/or attenuation is measured by recording the frequency and insertion loss of the SAW device, respectivel y. Various effects, including mass loading, viscoelastic loading, and acous to-electric coupling contribut e to SAW sensor response. Typical chemical sensors take advantage of one or more of these mechanisms in designing gas sensors. Earlier work by Christofides and Mandeli s [4] has reviewed the theoretical and experimental development of solid state hydroge n detectors. They compared some of the existing technologies and examined the role of Pd in hydrogen absorption. Their review indicates that a Pd-fiber optic sensor was most sensitive while a Pd-photo-pyro-electric sensor to be most economical, had high signal to noise ratio at wide temperature range and was second best in sensitivity at room temperature. More closely related to this research, Jakubik and co-workers [10, 11] have contributed to the understanding of the ma ss and acousto-electric effect response of phthalocyanine layers on SAW delay-line de vices. They have st udied responses of different bi-layer films of phthalocyanines and Pd to hydrogen to optimize the response. Their work showed that the best stability be tween the sensing layer and analyte occurred for copper (II) phthalocyanine and Pd bilayers while best responses were seen using the metal-free phthalocyanine, (hydrogen phthalocy anine) and Pd as the bi-layer sensing film. They showed that though the interacti on between the sensing film and analyte was stable, sensitive, and reversible, the inter action of the layers was highly dependent on temperature in an inverse relationship.

PAGE 18

7 Current work of this group involves an explanat ion of the interaction of bi-layer films and the role of the film thickness with regard to SAW device response. Penza and co-workers [12, 13, and14] have used SAW resonators, rather than delay-lines, coated with diffe rent vapor sensing layers fo r detection of organic and inorganic compounds. They have shown that res onators coated with funtionalized singlewalled carbon nanotubes (SWNTs) and multi-walled carbon nanotubes (MWNTs) when used to detect volat ile organic compounds display sensiti vities that are three to four orders of magnitude larger than polyme r-coated SAW devices. While other work utilizing both SAW delay-lines and resonato rs with polymer and other sensing layers exists in the literature for gas and vapor sensor s, this work of Penza et al. is perhaps the only example of orders of magnitude enhancement in sensitivity when nanomaterial sensing layers are employed. Recently, various metal custer-coated biomolecular nanotubes were synthesized using a genetically engineered Tobacco mosaic virus (TMV) biotemplate. TMV is a wellstudied tubular plant virus 300nm in length and 18nm in di ameter. Monodispersed stable Pd cluster-coated nanotubes were produced from the engineered TMV, exploiting its well-defined tubular structure and the metal binding specific ity conferred by engineered cysteine derived sulfhdryl groups. This viru s-based nanotube is a promising alternative to carbon nanotubes. 1.5. Thesis Outline Chapter 2 presents the background information needed for understanding how SAW sensors work and how to exploit their parameters for designing sensors. Chapter 3 presents the experimental setup used for de signing sensors. The experimental results are presented in Chapter 4, and Chapter 5 contains a discussion of the experimental results, suggestions and future work.

PAGE 19

8 CHAPTER 2: THEORY 2.1. Saw Devices There are two categories of SAW devices, resonators and delay-lines. Delay-lines are twoport devices having two separate sets of Interdigital Transducers (IDTs) deposited on a piezoelectric substrate with centre-centre separati on equal to a whole number of wavelengths. The re sonators on the other hand ca n be either one-port or twoport devices, each having either one or two ID Ts deposited on a piezoelectric substrate, between a pair of reflection gratings [6]. In a resonator, the acoustic waves are launched within a resonant cavity thus causing the si gnal to be reflected ba ck to the generating IDTs [6]. Figure 2.1 shows both delay-line an d resonator structures. Both these devices fall under the category of surface launched ac oustic wave devices. These devices are called surface launched as the wave motion is produced by the transducers on the surface and also picked by the transducers on the same surface, even though bulk waves are propagating in the surface. Acoustic sensors require a piezoelectric medium for generation of waves. The wave emanation ca n be explained in terms of crystal reordering. In a piezoelectric crystal, at equilibrium, inherent cr ystal strain is balanced by the internal polarization force. When disturbed either by appl ying external electric field or stress, de-polarization waves are emitted acting as a restoring force to maintain equilibrium [16]. The magnitude of the for ce emitted is directly proportional to stress applied as well as crystal or ientation and arrangement of ferroelectric domains, both functions of crystal morphology. 2.2. Two-Port Saw Resonator A two-port SAW resonator consists of an input IDT, output IDT and end reflection gratings. These IDTs are constructed on piezoelectric material such as quartz,

PAGE 20

9 lithium niobate, lithium tantalite etc. The cut of the substrate and orientation of the IDTs with respect to crystallographic axis ar e chosen such that it undergoes periodic deformation in response to applied potential. The application of elect rical signal across a pair of interdigital transducers causes the piezoelectric material to be strained thus generating acoustic waves. The input IDT converts the applied electrical signal to mechanical acoustic waves. These waves tr avel across the surface towards the output IDT. The output IDT converts these mechani cal SAW vibrations back to electrical voltage. Figure 2.1. (A) DelayLine (B) Two-Port SAW Resonator (C) Single-Port SAW Resonator

PAGE 21

10 In a two-port SAW resonators there are e qual finger numbers at both the input and output IDT, with finger width e qual to finger spacing. The equi valent circuit is shown in Figure 2.2 [15]. In the Figure 2.2 the shunt capacitor CT represents IDT capacitance while elements Lr, Cr, Rr represent equivalent motional pa rameters for series resonance Figure 2.2. (A) Lumped Equivale nt Circuit for 0 Phase Shi ft between Input/Output IDT’s at Resonance. (B) Ideal 1:-1 Transformer Included for 180 Phase Shift at Resonance Figure 2.3. Two-Port SAW Resonator

PAGE 22

11 condition. The SAW resonator used for the experiment was RP 1239, 315 MHZ. A twoport RP 1239 SAW resonator consists of an input and output port. It is a three pin device having one pin for input, one for output and third pin is common a nd serves as device ground as shown in Figure 2.3{SAW resonator 1239). 2.3. Operation When an alternating voltage is applied across the IDT, the substrate undergoes periodic deformation. This causes an acous tic wave to be launc hed propagating away from the IDT in both directions and perpendi cular to the IDT fingers. The wavelength is determined by the IDT geometry and the sp acing “d’ between the fingers as given by equation 1. l = 4d (1) The acoustic wave propagation velocity is de termined by the acoustic mode, orientation and the type of substrate. This acoustic wave velocity is related to fundamental resonant frequency is given by equation 2. o ov f (2) Where fo is wave frequency vo is acoustic wave propagation velocity l is wavelength For any perturbation mechanism the velocity change can be measured as a change in resonant frequency, since th e wavelength from equation 2 is seen to be fixed by the geometry of the device. The SAW device is hi ghly sensitive to perturbation arising from

PAGE 23

12 many different mechanisms such as changes in mass density, elastic stiffness coefficients, dielectric property of film, electric conductivity [6] give by equation 3. The variation of these characteristics causes the va riation in SAW wave velocity “ vo”. Variation of wave velocity is directly related to the changes in fundame ntal resonant frequency of the device. By using different perturbation approach and appr opriate boundary condition it is possible to get the fractional changes in velo cities which are rela ted to frequency. The commonly used perturbation mechanisms are mass loading and acousto-electric mechanism. ..... 1 p p v T T v v v c c v m m v v v vo o (3) Where, m is mass c is stiffness coefficient is dielectric constant is surface conductivity T is temperature P is pressure. 2.4. Saw Preturbation Mechanisms In a surface acoustic wave device, th e entire energy of surface waves is concentrated just at the surface. This makes the SAW devices highly sensitive to perturbations arising from many different physical para meters. Hence, the wave parameters such as velocity, frequency, amp litude and phase in an acoustic device are very sensitive to the immedi ate environment. Unlike othe r sensing mechanisms, it is more of a surface phenomenon. An incrementa l variation in the surface has pronounced effect on the wave parameters. The waves ar e hence easily and effectively affected by a medium placed along the path on the piezoelectric surface. The propagation of waves due

PAGE 24

13 to the presence of this medium is slight ly disturbed. The amount of disturbance is dependent on thickness of the layer, its density and electrical conductivity. When gas molecules come in contact with this medi um usually a selective absorbent film, an interaction occurs which causes the furthe r disturbance of waves. The fundamental physical and chemical propertie s of these analytes are dire ctly related to the wave disturbance which makes acoustic sensi ng more reliable scheme of detection. 2.5. Saw Mass Loading One of the most common interactions between SAW device and surface film is due to mass loading. This occurs due to changes in mass deposited on device surface. This causes the wave velocity to decrease. Th e degree of effect produced is derived from energy consideration [16].The movement of wa ve along the surface having a thin or rigid film causes an increase in kinetic energy density Uk [16]. This changes in kinetic energy per area of surface is given by 2 2 2 4 zo V yo V xo V s k U (6) Where Vxo, Vyo and Vzo the SAW particle velocity and s is surface mass density. The particle velocities are related to a displacement u by vi = jwui. The power density P (power/area) carried by a wave can be related to wave energy U (energy/volume) stored in a lossless medium as Uv P (7) In a lossless medium P is constant Differentiating (7) would yield o ov v u u (8)

PAGE 25

14 Combining equation (6), (7) and (8) p V p V p V v v vzo yo xo s o o 2 2 24 (9) Since the entire wave energy is concentrated near the surface, as operating frequency increases, surface particle veloci ties increase in proportion to ( P ) 1/2. Thus, the quantity in equation 6 is independent of wave am plitude and is dependent only on substrate material. p zo V p yo V p xo V 2 2 2 (10) Combining all substrate dependent parame ters, the mass induced change in surface velocity is given by s o m of c v v (11) Where mass sensitivity factor cm, is given by p V p V p V v Czo yo xo o m 2 2 22 (12) From equation 11 we can see that the mass loading results in fractional changes in velocity. These fractional changes in veloci ty can be tracked by fractional changes in frequency.

PAGE 26

15 o of f v v (13) These fractional changes in frequency are tracke d using an oscillator circuit thus tracking sensor response. 2.6. Saw Acousto-electric Effect The acousto-electric effect in SAW aris es from generation of a layer of bound charges at the surface accompanying the mechanical wave. The deposition of a conductive film causes the redistribution of charge carriers. This redistribution compensates the layer of bound charges developed due to passing surface wave. Figure 2.4 [16] shows the equivalent circuit model for SAW acousto-electric response. Figure 2.4. Equivalent Circuit Model De scribing the Intera ction between SAW and Charge Carriers on Film Overlay

PAGE 27

16 The current generated pe r unit area of surface, Io is given as P k K Is o o) ( 22 2 2 (14) Where K2 is Electro mechanical coupling coefficients o and s are air and substrate dielectric permittivity k is acoustic wave number P is power density Displacement currents generated in the subs trate and air arise from capacitances of s k o and k s, respectively. Shunt conductance k 2 accounts for the conduction currents in the film overlay. To study the eff ect of velocity and attenua tion arising from SAW film acoustoelectric coupling two cases are considered. 1. Without conductive film – In the absence of conductive film the energy generated by the wave gets stored in the evanescent electric field. The power flow in this case is given as, ) ( 22 1 s o o Tk j I P (15) 2. With the conductive film, the power flow becomes ) ( 22 2 2 s o s o Tk j k I P (16)

PAGE 28

17 The overall response due to acousto-electri c is the difference in two powers given by s s s o T T TkC j k k I P P P 2 2 2 1 22 (17) Where Cs = o + s Putting equation (14) in (17) gi ves the complex power flow, PT The splitting of real and imaginary parts of power results in separating the effect of fractional changes in velocity and attenuation. 2 2 2 2) ( 2s o s s oc v K v v (18) 2 2 2) ( 2s o s o S sc v v c K k (19) The magnitude of acousto-elect ric response is proportional to K2 thus being dependent on substrate. 2.7. Saw Sensor Operation The SAW device can be perturbed by a num ber of mechanisms as described in equation 3. This interaction occurring due to different mechanisms such as changes in mass density, elastic stiffness coefficients dielectric propert y of film, electric conductivity causes the variati on in SAW wave velocity “ Vo”. This variation in Vo produces a phase shift between the applied and the received signals thus causing the frequency of minimum impedance to be shifted away from resonant frequency “ f ’. Further, dampening of the acoustic waves may also result in decrease of amplitude of the received signals. Thus, perturbation can resu lt in both phase and at tenuation change [6].

PAGE 29

18 These changes are usually measured by tracking changes in frequency, phase, attenuation or electrical impedance. Two methods are us ed to track these cha nges in the device: 1. The device either becomes a part of a frequency controlling element of an oscillator circuit 2. An external signal may be applied to the device with its phase and amplitude compared to the reflected or transmitted signal The former is more commonly used as it’s simpler and equipments required are cheaper. The other method is ge nerally used in understanding device characteristics or understanding perturbation response mechanisms The oscillator approach was used to design the sensor and the design approach is desc ribed in detail in the oscillator section of chapter 3. 2.8. Oscillator Based Measurement Method There are two common methods of designi ng oscillator circuit for SAW filters. One method is a feed-back loop method and ot her is a negative resistance method. The feedback method is more commonly used for SAW delay-lines and SAW two-port resonators as they provide bett er control. In order for the SAW resonator to oscillate two essential conditions have to be met 1. The overall gain in the loop must be one or higher( at least 0 dB) 2. The overall phase shift in the loop must be an integral of 2n where n is an integer Once the above two conditions are satisfi ed, a positive feedback would occur through the closed loop causing the SAW device to oscillate at its resonant frequency. The point to be noted here is that the oscillations occur at a frequency where both these conditions are met, which may or may not be the resonant frequency of the device and may lie in the neighbor hood of the resonant frequency [6]. The basic block diagram for the oscillator circuit is show n in Figure 2.5. The RF amp lifier is chosen high enough so that it can compensate for losses in the feedback loop and provid e a gain of at least 0 dB. The amplifiers should be selected carefully as there are usually limits on the maximum permissible input and output pow er without damage. The choice of filters is done so as to remove any unwanted harmonics and spurious oscillations.

PAGE 30

19 Figure 2.5. Oscillator Block Diagram The phase shifters help in manually controlling the point of zero phase shift change and high gain to provide continuous a nd sustained oscillation. The attenuators are used to reduce the gain at the input of am plifier as they have a maximum permissible input level. One of the largest sources of e rror is the temperature and stability control of the amplifiers and voltage vari able phase shifters All these components rely on a stable DC power source and a temperat ure controlled environment. 2.9. Type of Measurement SAW devices can be perturbed using a number of mechanisms as discussed earlier. These mechanisms depend on the interaction between the SAW device surface, the selective absorbent surface film and the gas. For a good interaction between the gas molecules and the selective ab sorbent surface film, there must be a means to transport these gas molecules to the surface. Also, thes e molecules must interact for some duration followed by which they should be transp orted back from the surface to outlet. Measurements could be made either statically or dynamically. Static measurements involve exposing the target analyte to a fixe d amount of analyte in fixed volume of gas. These are usually used to investigate the properties of thin films deposited on the SAW device. The dynamic case involves the continu ous transportation of carrier across the surface. There are two methods of dynamic measurements, one involving fixed analyte

PAGE 31

20 concentration throughout the meas urement and the other involving a variation of analyte concentration from zero to maximum and back to zero. For accurate measurements they require good control of temperature, and pr essure along with proper RF shielding and electrical connections. 2.10. Understanding Sensor Response The target analyte reaching the surface s hould be absorbed by the sensor surface having a selective absorbent film. The SAW re sponse to the analyte is driven by kinetic phenomenon such as adsorption, desorption, absorption and diffusion. These responses can occur either by weak physical interactions ( H < 40KJ mol-1) or chemical bond formation ( H = 80-400KJ mol-1 approximately) [17]. Physical absorption or physisorption usually occurs under conditions th at favor liquefaction. These interactions are by weak Van der Waals forces or hydrogen bonding which are easily reversible. In physisorption the temperature alters the rate of surface absorption or desorption. Chemical adsorption or chemisorptions is le ss reversible process occurring usually at lower temperature and higher pressures. Some examples include dissociative adsorption process found in metal surface oxide and hydride film formation, and bonding of water molecules to such surface [6]. Some interact ion such as Bronsted-Lowry and Lewis Acid lie between the chemisorptions and physisorptio n. There are also some interactions that are driven by entropic consideration typically in case of highly non polar analytes and sensor coatings which include hydrophobic e ffects. The rates of absorption/desorption and the sensor response are t ypically dependent on temperatur e, film properties and the amount of analyte concentration. Also the av ailable sensor surface and its density also affect the rates of absorption which in turn a ffect the sensor response. At any instant the vapor-liquid phase analyte concentration, CV, would give rise to a corresponding sensor phase analyte concentration, Cs, generating the resulting response [6]. CV is usually less than Cs unless the process in irreversible.

PAGE 32

212.11. Film Preperation SAW coatings fulfill two important ro les. They provide a mechanism for providing selectivity towards the target analyte and the mechanism for improving device sensitivity. This sensing film can be coat ed on the surface of a SAW device using a number of techniques. The c hoice of method depends on ease of using the technique and available instruments for carrying out the same Techniques such as casting, painting, dip coating. spin coating and drop coating are commonly used for coating the surface. Film thickness can usually be monitored by observi ng the frequency shift produced on coating. To get good coating the SAW surface must be thoroughly cleaned [18]. Commonly used techniques such as sonicating in chlorofo rm, washing with strong acid, or plasma cleaning could be used to clean the surface. The type of coating method adopted also effects the type of film obtai ned [19] [20] [21]. Some key techniques of coating are explained below. Spin Coating: Solution is dropped on the rota ting crystal and centrif uge force spreads the solution. Factors such as rotating rate, solu tion viscosity, and solu tion density, directly affect the film thickness and uniformity. Spray Coating: An airbrush containing a so lvent reservoir is drawn upon by a mild flow of carrier gas. The solvent is sprayed th rough a small nozzle. The type and amount of coating is dependent on the skill of the pers on using it. The material is sprayed till the desired coating is achieved. Drop Coating: The solution of a required volume is dropped on the surface using a pipette. The solvent is allowed to evaporate by itself or with the help of gentle heating. This is a simple method for achieving desired thickness. 2.12. Gas Delivery There are two types of flow devices, volum e flow devices and mass flow devices. As the name suggests volume flow devices tend to control the vo lume of gas flow. Volumetric flow (Q) measurement techniques re quire separate measurements of pressure (P) and temperature (T) to calculate density and mass flow (m). These measurements are hence less reliable as they are easily affected by temperature and pr essure changes which

PAGE 33

22 affect the density of gas. Mass Flow de vices on the other hand are not affected by changes in pressure and temperature as they regulate molecular flow. Hence they are more reliable, repeatable and a more accura te method for delivering the analyte or the required material at a desired rate. Mass flow meters are usually calibrated for a specific gas or flow range to provide direct readou t. They can be used for different gases by applying appropriate conversion factors calculated from thermodynamic considerations. Mass flow devices such as thermal mass flow meters (MFMs) and mass flow controllers (MFCs) are commonly used for delivery of gas. Table 2.1 shows the measurement techniques used for some volumetri c and mass flow measuring devices. Table 2.1. Measurement Techniques Used by Different Gas Delivery Systems Flow meter Measurement Technique Parameters Required Turbine Meter Q Need P,T Positive Displacement Q Need P,T Orifice Meter Q Need P,T Rotameter Q Need P,T Thermal MFC/MFM M Measures Mass Directly 2.13. Mass Flow Controller Operation The mass flow controllers are flow m easurement techniques commonly used for regulating the flow of gas. The gases that fl ow in the lines are divided into two parts while entering the MFCs. Most of the gas (m2) flows through the bypass thus creating a pressure drop causing a small part of the fl ow (m1) to go through the sensor tube as shown in Figure 2.6. The sensor tube has tw o resistance coils, causing a fixed amount of heat flow into the tube. Heat gets transferre d from the tube to the gas molecules. Heat transfer between these elements results fr om interaction with the molecules of the flowing gas independent of pressure or temper ature fluctuations. The gas carries this heat from the upstream coil to the downstream coil. The downstream coil has a higher resistance and is maintained at a higher temperature as compared to the upstream coil. These resistors form legs of a Wheatstone bridge whose output vo ltage is directly

PAGE 34

23 calibrated to calculate the mass of gas flowing in the tube. As the gas from bypass and sensor tube leaves the MFC they have to go through a servo controlled Valve. This valve could be electromagnet, where the current in the electromagnet modulates the orifice opening thus maintaining a fixed mass of gas flow. Built-in PID electronics allows the device to maintain continuous proportional control by comparing the measured sensor signal to the commanded flow rate. These valv es could be normally open ones or closed configurations. Most therma l MFMs and MFCs are specified at 1% of full-scale accuracy and 0.25% of fu ll-scale repeatability. Figure 2.6. The Mass Flow Controller 2.14. Saw Housing The manner in which SAW device is mounted and coupled to the sample handling system determines the type of SAW housing to be designed. The key consideration while designing SAW housing are precise temperature, pressure control, sa tisfactory electrical connections and chemical stability of housing material. Some material such as Brass, Aluminum, Stainless Steel and PTFE are co mmonly used to design ing housing. Though brass and aluminum provide good electrical sh ielding, good temperature stability and are easy to machine, they however suffer from corrosion by chemicals used in the housing. PTFE, though is a good material free from corro sion due to chemical attack, suffers from problems of machining as they tend to creep an d deform due to applie d pressure or stress.

PAGE 35

24 Stainless steel is a good material of choice as it is not susceptible to problems discussed previously. 2.15. Sensing Layer For successful detecti on of target analyte the SAW de vice must be coated with a suitable sensing film. The sensing film chos en should provide sele ctivity and enhanced sensitivity to the target analyte which is a key requirement for a gas sensor. Two nanomaterials sensing layer chosen in th is study were Carbon Nanotubes and Tobacco Mosaic Virus. 2.16. Carbon Nanotubes Carbon Nanotubes (CNTs) are thin, cylindric al structures made of carbon similar to honey comb lattice. Out of the many polymorphs of car bon they tend to resemble graphite. Though they are similar to buckyballs as both are made of graphite there exits an enormous structural difference. They struct ural difference arises from the fact that the buckyballs are graphite sheets rolled into balls, while the CNT are graphite sheets rolled in the form of cylinder. This structural difference gives remarkable electronic and mechanical properties to carbon nanotubes. They are known to have a Young’s modulus of over 1 tera Pascal and estimated tensile strength of 200 giga Pascal. Figure 2.7 shows the SEM image of Carbon Nanotubes [22]. The carbon nanotubes can be manufactured in two types: 1. Single walled carbon nanotube (SWNT) 2. Multi walled carbon nanotubes (MWNT) Each of them has distinct material pr operties making them su itable for different applications.

PAGE 36

252.17. Single Walled Nanotubes (SWNTs) SWNTs are sheet of graphite rolled in th e form of cylinder. They usually have diameters in the range of 1-2nm with 25microns in length. Figure 2.8 shows a microscopic image of SWNTs [23]. They can be more easily twisted, flattened and bent into small circles around sharp bends wit hout breaking, unlike MW NTs. They can be either metallic or semi-conducting. They also have very high tensile strength which is about 100 times of steel at a si ze that is 1/6 of it. There are 3 main configurations for SWNTs. 1. Armchair 2. Zigzag 3. Chiral Figure 2.9 [24] shows all the three configuration of SWNT s. Each of them differs, in the manner in which they are rolled up along the diameter. This difference gives them different properties. These properties can be determined using a vector called chiral vector. This is given by Ch = na1 + na2 (20) Where, na1 and na2 are unit vectors in 2 dimensi onal hexagonal lattice and n, m are integers. Depending on the value of chiral vect or and chiral angle (which is the angle between Ch and na1) they can be metallic and semiconductor. Armchair varieties are usually metallic having n=m with 0 chiral an gle. Chiral type SWNTs has a chiral angle anywhere between 0-30 , while the zigzag variety has a chiral angle of 30. 2.18. Multi Walled Nanotubes (MWNTs) Multi Walled Nanotubes (MWNT) can be cons idered as a collection of concentric SWNTs as shown in Figure 2.10 [24]. They ar e usually 2-20nm in di ameter with 100 nm or more in length. They can have anywhere between 5-20 graphite la yers. Due to layered structure they usually have a mix of semi -conducting and metallic properties. A typical

PAGE 37

26 MWNT is shown in Figure 2.10. Although it is easier to produce sign ificant quantities of MWCNTs than SWNTs, their structures are less well understood than single-wall carbon nanotubes because of their greater complex ity and variety. Today, both SWNT and MWNT are being extensively used in the production of high strength composite materials, sensors, optical devices, hydroge n fuel cells with the type of device determining the type of nanotube desired. 2.19. Tobacco Mosaic Virus (TMV) Tobacco mosaic virus is known to cause plant disease in more than 150 types of herbaceous, dicotyledonous plants including many vegetables, flowers, and weeds [25]. These viruses don’t kill the crop but lower their yield and quality of the crop. Their infection on plants is usually characterized by intermingled patche s of normal and light green or yellowish colors on th e leaves [26]. They are about TMV is 300 nm long, 18 nm in diameter and with a 4 nm in diameter axial channel [27, 28]. They have been previously coated with metals, silica or semi-conducting materials to form nanorod assemblies. Figure 2.11 [28] and Figure 2.12 [29] shows the TEM image TMV tube and of a TMV Pd coated TMV tubes respectively. Figure 2.7. S.E.M. Image of Carbon Nanotubes22 Figure 2.8. Microscopic Image of Single Walled Carbon Nanotubes23

PAGE 38

27 Figure 2.9. (A) Armc hair (5,5), (B) Zigzag (10,0) (C) Chiral(10,5)24 Figure 2.10. High Resolution T.E.M. Image of Multi-Walled Carbon Nanotubes24 Figure 2.11. Electron Microscope View of a Single TMV Tube28 Figure 2.12. T.E.M. View of a Pd Coated TMV Sample 29(Scale Bar200nm)

PAGE 39

28CHAPTER 3: EXPERIMENTAL SETUP 3.1. Gas Dilution System An in-house developed custom-automated gas dilution system [30] was utilized for these experiments. The system was desi gned keeping in mind several factors. First there was a need for a controllable concentra tion analyte gas, and a purge gas. These conditions where met by using 4 MKS Type 1479A mass flow controllers (MFCs) connected to a MKS Type 247 4-channel r eadout. The four flow controllers have different maximum flow rates to allow for maximum accuracy and controllability across a wide range of gas concentra tions from high ppb to pure wh ile maintaining a steady flow rate, depending upon the concentration of th e analyte gas cylinder concentration. The MFCs in this work were used to mix hydroge n with nitrogen to obtain different range of concentrations from 0 volume% to 5 volume%. Figure 3.1. Mass Flow Controller Setup

PAGE 40

29 These gases pass through a set of solenoid valv es which were selected in operation for the concentrations required. The gases flow ing through the mass fl ow controllers are mixed in the mixing chamber with carrier gas. The gases are then passed through a temperature controlled cell. Stainless steel tu bing is chosen to connect the cell and the solenoid valves while PTFE tubing is used between the cell and the exhaust. The selection of MFC, solenoids and the concen tration is done using a LABVIEW (version 7) program through a National Instruments PCI Data Acquisition (DAQ) card. The fully automated system has a cycle time of approximately 5 seconds [19]. Figure 3.2. The Complete Experimental Setup 3.2. Calibration of Mass Flow Controllers The gas flow controller has to be calibrated for it to detect the flow rates of different gases as they are ge nerally calibrated for nitrogen. There are a set of scaling potentiometers on the rear panel of the mass flow controller rea douts. These 10 turn potentiometers are used to adjust the full scale voltage signals from the MFC, which

PAGE 41

30 correspond to the flow rate, to a level that en ables the digital panel meter to display the flow rate and set point directly, in sccm. Th e factor is called S caling Control Factor. Scaling Control Factor = Gauge Factor Gas Correction Factor The Gauge factor is the factory set value which scales the +5VDC output signal to the appropriate full scale range for the MFC, so that the panel meter reads 1000 counts. Table 3.1. MFC Gauge Factor MFC Gauge Factors MFC Flow Range(sccm) Gauge Factor 1,10,100,1000,10K 100 2,20,200,2000,2K 200 5,50,500,5000,5K 50 Gas Correction factor (GCF) A gas correction factor is used to indicate he ratio of flow rates of different gases which will produce the same output voltage from mass flow controllers. GCFX = (0.3106) (S) (dx) (Cpx) GCFx depends on specific heat, density, and molecular structure of gases. dx is Standard density of ga s g/l ( at 0c and 760mm Hg) Cpx is Specific heat of gas x, cal/g C 0.3106 is (Standard density of nitrogen) (Specific heat capac ity of nitrogen) S is Molecular structure correction factor Where S equals: 1.030 for Monatomic gas 1.000 for Diatomic gas

PAGE 42

31 0.941 for Triatomic gas 0.880 for polyatomic gas x x xCp d GCF S 0.3106 Applying temperature correction factor as th ese values of gas co rrection factors were calculated at 0 (273 Kelvin). Temperature corrected GCF = s xT T GCF Where Tx = Reference temperature at (K) Ts = 273.15K (0C) For hydrogen, x x xCp d GCF S 0.3106 GCFx depends on specific heat, density, and molecular structure of gases. dx = 0.0899 Cpx = 0.3106 S = 1.000 for Diatomic gas (H2) GCFx = 1.0 Temperature corrected GCF = s xT T GCF Tx = 297.15K (24 C) Ts = 273K (0 C) Temperature corrected GCF = 1.08

PAGE 43

32 Table 3.2. Scaling Factor for Hydrogen Based on Flow Rates MFC Flow rate (sccm) GCF Guage Factor SCF= GCF*Gauge Factor 500 1.08 50 54.80 10 1.08 100 109.70 100 1.08 100 109.70 2000 1.08 200 219.330 3.3. Test Cell The SAW device was housed in a cell ma de of Stainless Steel. The key considerations while designing the cell were pr ecise control of temperature and chemical stability of housing material. Also, the cell volume was required to be small so as minimize the time required for the material to in teract with the sensing material and thus achieve quicker stability. The flow should also be streamlin ed. Based on the requirements the cell design was designed in Microsof t Visio as shown in Figure 3.3. Figure 3.3. Schematic of the Cell Design

PAGE 44

33 The cell was made of 312L stainless steel Stainless steel was used as it does not creep or deform due to applied pressure. This is important as the PCB would be removed very often for changing the SAW device or increasing the coating on the surface of the device. The cell has an inlet and outlet for a llowing the gases to and from the cell. The input of the cell was fed from the solenoids connected to the mass fl ow controllers while the output fed to exhaust. Conn ections between the cell and th e solenoid connected to the MFC were achieved with help of National Pipe Taper (NPT) from Swagelok. Female connectors are commonly used on the tubing ends while male couplings on the casing. The inlet and outlet were threaded for connec ting ” NPT male thread connector capable of handling pressures up to 10000psi. These ma le connectors were c onnected to the inlet and outlet lines using ” Figure 3.4. The Machined Stainless Steel Cell female NPT threads, available at the tube endings. The cell has a small circular opening in the top to house the SAW resonator. Th e SAW resonator was supported on a printed circuit board. The printed circuit board suppor ting the resonator was tightly held on two the cell with help of two Allen screws. The machined cell is shown in Figure 3.4.

PAGE 45

343.4. Saw Resonator – PCB Design A commercial two-port SA W resonator RP1239, 315MHZ was used to design the sensor. The two-port resonator consists of three terminals, one input, output and ground terminal that is common and serves as devi ce ground. The commercia l resonator is shown in Figure 2.3. The resonator was connected to the printed circuit board directly. The design of the printed circuit boa rd was done in-house. Severa l key factors were kept in mind before designing the board. Since the SAW device operates at 315 MHz it was ensured that the there were no bends or curves since RF effects could play a key role in changing the characteristics of the board, such as adding additional phase or changing the characteristic impedance of the board as was seen in one of the earlier PCB designs. This copper board to be etched was a commercial pr esensitized board. A presensitized board is a double sided copper board having positive photoresist precoated on both sides. Figure 3.5. The Top and Bottom Surface of PC B Designed in AutoCAD. The Bottom Side is One Facing the Cell The design of the board was made in AutoCAD as shown in Figure 3.5 keeping all the parameters in mind. This design was later transferred on a transparency acting as a mask to be used on a presensitized copper board The process of development of board is described in detail in the Appendix A. Th e SAW device was directly soldered on the

PAGE 46

35 board to avoid any loose connections whic h were encountered while designing the board with sockets. The SAW device input and output were connected to the oscillator with help of SMA connector. The top and bottom vi ews of the PCB design is shown in Figure 3.6 and Figure 3.7. Figure 3.6. The Final PCB Design (Bottom) Figure 3.7. The Final PCB Design (Top) 3.5. Oscillator Circuit A feedback loop oscillator circuit was designed with the SAW resonator as the frequency controlling element. The design of the feedback loop ensured that there was

PAGE 47

36 positive feedback through the loop providing an overall phase shift of 2n where n is an integer, thus causing the SAW resonator to oscillate at its resonant frequency. 3.5.1 Amplifier To provide the necessary gain for driv ing the SAW resonator a high gain, broad band monolithic, Minicircuits ZFL-500LN amplifier was chosen. The amplifier provided a 24 dB gain in 0.1-500 MHz bandwidth. Th e input voltage driving the amplifier was 15V with current of 60mA. The maximum i nput and output power was +5 dBm. It provided a low noise Figure (N.F.) < 2.9 dB and 3rd order intercept point of 14 dB. The input and output VSWR was 1.5:1 and 1.6:1 respectively with input and output impedance of 50 ohm. The amplifier was chos en in a manner to have SMA connectors at the end. The output of the amplifier was c onnected to the input of the SAW resonator board. 3.5.2 Filter The low pass VLF-320, filter was used in th e oscillator circuit to avoid oscillation at spurious frequencies. The low pass filter was chosen to have a cutoff at 320 MHz with a loss of less than 1dB in the pass band. Th e filter was connected to the output of the SAW resonator board. 3.5.3 Attenuator A set of fixed attenuator was chosen in order to reduce the ove rall gain in the circuit so as to prevent the maximum permi ssible input to the amplifier. MiniCircuits VAT series attenuators of 3, 6, 10 dB were used as and when required in the circuit depending on the amount of at tenuation required. These fixed attenuators had SMA connector at both input and output. 3.5.4 Phase Shifter To provide the requir ed phase shift of 2n a phase shifter was introduced in the circuit. Two voltage variable MiniCircuits JSPHS-446, phase shifter were used in order

PAGE 48

37 to provide a minimum phase sh ift of 360. The phase shifters used we re surface mount devices. Figure 3.8. PCB Design for Two 180 Degree Phase Shifter Connected in Series Figure 3.9. The Final PCB De sign of the Phase Shifter The PCB for the phase shifters were ma de in house using the PCB fabrication technique explained in Appe ndix A. The input and out put of the board had SMA connectors. The AutoCAD layout of the boar d in shown in Figure 3.8, while the completed board is shown in Figure 3.9. The phase shifters were chosen to have no more than 1dB of loss in the required bandwidth.

PAGE 49

383.5.5 Coupler A MiniCircuits ZX30-9-4 coupler was us ed to sample the signal from an oscillator loop with a minimu m drain of power. Coupler is a three port device having an input, output and coupled port. The output port with a 4 dB loss was connected to input of the amplifier. The coupled port having a loss of 9 dB wi th respect to the input was connected to the frequency count er. The coupler was chosen so that the minimum loss in the coupler at the direct port (from input port to output port). By placing the coupling network after the SAW and filters, much of the harmonic content from the saturated amplifier will be suppressed. The complete oscillator circuit is shown in Figure 3.10. Figure 3.10. The Oscillator Circuit 3.6. Automation A fully automated in house developed instrument control and data acquisition program with little modifications was used to accurately log sensor parameters and to reliably generate gas from the Mass Flow Controllers. The Data acquisition program was written using a LABVIEW (version 7) prog ram. The program consisted of two main parts:

PAGE 50

39 1. Logging the sensor response 2. Generating the desired concentration of gas Figure 3.11, 3.12 and 3.13 show the labview program used for data logging and instrument control. All experimental values we re logged to a text f ile or excel sheet for processing and analysis. Figure 3.11. The LABVIEW Program for Writin g Data Automatically into a Text File

PAGE 51

40 Figure 3.12. The Front Panel Display of LABVIEW Program

PAGE 52

41 Figure 3.13. The LABVIEW Program to Control the Mass Flow Controllers and Log Sensor Parameters

PAGE 53

42CHAPTER 4: RESULTS 4.1. Tobacco Mosaic Virus An RF Monolithics, RP1239, 315 MHz, TO -39 package two-port SAW resonator was uncapped and used for the experiment. The S21 characteristics of the uncapped resonator as measured with Agilent 8753ES SParameter Network anal yzer is shown in Figure 4.1. The Figure shows the resonant fr equency and corresponding attenuation. The SAW resonant frequency is 315.010750 MHz w ith attenuation of -3.5755 dB at the resonant frequency. Figure 4.1. The Two-Port Characterist ics of SAW RP1239 Two-Port Resonator

PAGE 54

43 The uncapped SAW resonator was then sputter coated with 1000 of Si02. The Si02 was used as an insulating layer to prevent the me tallic nanoparticles from shorting the IDTs. The choice of Si02 was justified in its ability to act both as an insulating and guiding layer thus promoting the propagation of waves thr ough the medium [31, 32, 33, 34]. The Si02 coating caused an attenuation of the signal and a corresponding decrease in the resonant frequency of the device. Figure 4.2 shows the change in S21 characteristics of the device on Si02 coating. The SiO2 coated SAW device was later drop coated with 20 l of 0.015 mg/mL Pd coated TMV. The Pd coated TMV samples were dispersed in 4DIMETHYLAMINOAZOBENZENE DIMETHYLAMINE-BORANE (DMAB) solution. Figure 4.2. The Two-Port C haracteristics of SAW RP1239 Tw o-Port Resonator after Si02 Coating

PAGE 55

44 Figure 4.3. The Two-Port Characteristics of Si02 Coated SAW RP1239 Two-Port Resonators after TMV-Pd Coating The coating of the sensing layer or Pd coated TMV particle s produced a further attenuation of signal and decrease in resonant frequency of the device as seen from the S21 characteristics of the device in Figure 4. 3. These attenuation and frequency changes have been tabulated in Table 4.1. The coat ed device was then soldered on the PCB and placed in the temperature controlled stainless st eel cell. The seal was ensured to be leak proof. The resonator was also connected to th e oscillator circuit to ensure that the fractional changes in veloci ty occurring due to differe nt perturbation mechanisms discussed earlier can be tracked by changes in frequency. The coating of the sensing layer or Pd coated TMV particles produced a furt her attenuation of si gnal and decrease in resonant frequency of the device as seen from the S21 characteristics of the device in

PAGE 56

45 Figure 4.3. These attenuation and frequency ch anges have been tabulated in Table 4.1. The coated device was then soldered on the P CB and placed in the temperature controlled stainless steel cell. The seal was ensured to be leak proof. The resonator was also connected to the oscillator circuit to ensu re that the fractional changes in velocity occurring due to different perturbation mech anisms discussed earlier can be tracked by changes in frequency. Table 4.1. SAW Resonator Frequency and Attenuation Measurements after Pd-TMV Coating Process Frequency change Attenuation change 1000 of SiO2. 0.65625 MHz -7.7245 dB 20 l Pd-TMV coating 0.05250 MHz -5..907 dB The SAW device with the nanoparticle se nsing layer was exposed to varying concentration of hydrogen. The varying c oncentrations of hydrogen were achieved by mixing hydrogen with nitrogen using MFCs and a mixing cham ber, with the flow rate fixed at 1000sccm. The flow rate was fixed to prevent seeing any flow related effects [6]. The entire process was automated using a LABVIEW program. The behavior of sensing film of exposure to 0.2-2. 5% hydrogen is seen from Figure 4.4. The figure shows responses of the sensing film to varying volum e% hydrogen. It is seen that the resonant frequency of the device increases on exposure to hydrogen. This behavior is contrary to mass loading wherein the absorption /adsorp tion of analyte results in decrease in frequency. This type of be havior is related to acoscu to-electric phenomenon wherein a layer of bound charges developed in the pi ezoelectric material by passing SAW is redistributed by the presence of a conduc ting layer. This re distribution causes compensation of bound charges developed due to the passing surf ace wave [16]. This phenomenon usually results in producing an increase in frequency. The corresponding response times were calculated to be around 30s ec. It was observed that the process of absorption/desorption of hydrogen from the sensing film was reversible.

PAGE 57

46 Figure 4.4. The Response of Pd C oated TMV to 0.2-2.5 % Hydrogen 0 5 10 15 20 25 30 35 40 45 00.511.522.53Volume% hydrogen Response time (Seconds) Run1 Run2 Figure 4.5. The Response Times for Pd Coated TMV Resonator in Response to 0.2-2.5% Hydrogen

PAGE 58

47 The response times for 99% of full respons e for two different runs of 0.2-2.5% hydrogen are shown in Figure 4.5. It is seen from the grap h that the response times are have an average value of 30 seconds. The values at 0.2% were very small to be discernable, due to which their values have not been graphed. The calibration curve is shown in Figure 4.6 which shows a concentr ation independent response. The sensing film shows good response times for 99% of full response. 0 50 100 150 200 250 00.511.522.53Volume% hydrogenFrequency shift (Hertz) Run1 Run2 Figure 4.6. The Frequency Shifts Produced for Two Different Runs at Various Hydrogen Concentrations The sensing film was also tested to 2.5-4% hydrogen in nitrogen. These tests were carried out to test below the lower explosi on limits of hydrogen (4.65% in air). The response of the sensing layer on hydrogen exposure has been plotted in Figure 4.7. The responses reaffirm the fact that the analyte absorption causes the frequency of the device to increase. It is also seen from the res ponses that the process absorption/desorption of hydrogen from the sensing film is reversible The response times and frequency shifts produced by 2.5-4.5% hydrogen concentrations have also be en plotted in Figure 4.8 and Figure 4.9.

PAGE 59

48 Figure 4.7. The Sensing Film Response to 2-4.5% Hydrogen 20 25 30 35 40 45 50 55 22.533.544.55Volume% hydrogen Response time (Seconds) Figure 4.8. The Response Times of Sensing Film for 2-4.5% Hydrogen

PAGE 60

49 125 135 145 155 165 175 185 195 205 22.533.544.55Volume% HydrogenFrequency shift (Hertz) Figure 4.9. The Frequency Shift Produced in the Sensing Film 2.5-4.5% Hydrogen From the response curves shown previously it can be seen that the responses were repeatable over several runs It was also observed that the process of absorption and desorption of hydrogen in the sensing film was reversible. Previous work by our group on nano-structured thin film Pd SAW sensors has shown that repeated absorption and desorption of hydrogen in the sensing film cau ses the palladium film to crack and peel [30], which was not seen to be the case with palladium nanoparticle-coated Tobacco Mosaic Virus sensing layers. The cracked and peeled Pd film upon hydrogen sorption /desorption is shown in Fi gure 4.10 [30] and compared with the hydrogen exposed (repeatedly) Pd-TMV layer in Figure 4.11. Also, the Pd film based SAW sensors show effects of aging after few days of not bei ng exposed to hydrogen [35]. The devices have to be exposed to hydrogen fo r some time before they st art responding in the desired manner. The Pd-TMV coated devices were seen to respond without any need for activation cycling after weeks of testing. The sensor noise was calculated to be 17Hz approximately which was defined as standard deviation of the mean frequency calculated over a 10 minute interval. Sin ce the response frequency shif t magnitudes are small, Pt

PAGE 61

50 coated TMV samples were also tested for their response to hyd rogen. These devices showed no response to hydrogen challenges. Figure 4.10. Cracking of Palladium F ilm Upon Hydrogen Sorption/Desorption Figure 4.11. Device Surface Coated with Palladium Nanoparticle-Coated Tobacco Mosaic Virus Film Free of Rearrangement

PAGE 62

514.2. Carbon Nanotubes One more RF monolithic RP 1239 SAW two-port resonator was uncapped and sputter coated with 1000 of Si02. The Si02 coated resonator was then drop coated with 10 L sample of 0.167 mg/mL SWNTs dispersed in 1% SDS in D2O solution. This resonator was evaporated with 27 of palla dium using electron be am evaporation. To evaluate the changes in resonant frequenc y and attenuation due to addition of SiO2 and sensing layer the S21 parameters were obtained from Agilent 8753ES S-parameter Network Analyzer. The SAW resonant fre quency was found to be 315.010750 MHz with attenuation of -3.5755 dB at th e resonant frequency. The changes due to various coating have been tabulated in Table 4.2. Table 4.2. SAW Resonator Frequency and Attenuation Measuremen ts after Pd-SWNT Coating It was seen that the drop coating of SWNTs and evaporation of caused an attenuation of the signal and a corresponding d ecrease in the resonant frequency of the device. The resonator was fixed on a PCB and pl aced in the Stainless Steel cell to be in contact with the incoming gas. This resonator wa s also made a part of oscillator loop to track the fractional changes in frequency. The resonator with the sensing film wa s exposed to 0.2-2.5% volume% hydrogen in nitrogen at a flow rate of 1000sccm. The re sponse of the sensing film has been plotted in Figure 4.12. In accordance with a ma ss-loading response mechanism, frequency decreases were observed upon hydrogen exposures [ 16]. These shifts are seen to be quite large, keeping in mind that only a 27 laye r of Pd was deposited on the SWNTs. Figure Process Frequency change Attenuation change 1000 of SiO2 0.64315 MHz -7.200 dB 10 l SWNT solution 0.40000 MHz -6.432 dB 27 of Palladium 0.10000 MHz -2.138 dB

PAGE 63

52 4.13 shows the response times, defined as time taken to reach nearly 99% of complete response, for 0.2-2.5% of hydrogen exposures. Figure 4.12. The Curve Showing the Response of Pd Coated SWNTs Response to Hydrogen 50 70 90 110 130 150 170 190 210 00.511.522.53Volume% HydrogenResponse time (Seconds) Figure 4.13. Response Times for Pd-SWNTs Coated SAW Resonator to 0.5-2.5% Hydrogen Exposures

PAGE 64

53 These are seen to be about 175 seconds. Wh ile these are larger than for the PdTMV films, it should be noted that the large and immediate response reflected in the 66% response time (see Figure 4.15) allows for fast-responding hydrogen sensors using PdSWNTs to be constructed. The device calib ration curve is shown in Figure 4.14, which clearly indicates a concentra tion dependant frequency shift, unlike for the Pd-TMV layer tested 0.00 2000.00 4000.00 6000.00 8000.00 10000.00 12000.00 0.000.501.001.502.002.503.00Volume% Hydrogen Frequency shift (Hertz) Figure 4.14. Frequency Shifts for Pd-SWNTs Coated SAW Resonator to 0.2-2.5% Hydrogen Exposures The Pd coated SWNT film was later exposed to several runs of fixed concentrations to see if there is any change in the film char acteristics in terms of amount of shift produced and the response times. From Figure 4.15 and Figure 4.16 it is seen that repeated exposures to 1 and 1. 5% produces no appreciable cha nge in terms of amount of shift produced and the response times. The freq uency shifts are seen to be around 6 KHz and 8 KHz at 1% and 1.5 % hydrogen respectiv ely. The comparison of these frequency shifts to the calibration curve (Figure 4.14) show no appreciable change in the amount of shifts. This shows th at film was robust and did not su ffer from problems of peeling, degradation and cracking fr om repeated exposures.

PAGE 65

54 Figure 4.15. Frequency Shifts for Pd-SWNTs Coated SAW Resonator to 1.5 % Hydrogen Figure 4.16. Frequency Shifts for Pd-SWNTs Coated SAW Resonator to 1 % Hydrogen

PAGE 66

55 From the response curves shown previously it can be seen that the responses were repeatable over several runs It was also observed that the process of absorption and desorption of hydrogen in the sensing film was reversible. The film was robust since repeated runs at varying concentrations and fixed concentrations didn’t seem have produced any appreciable changes in te rms of response times and corresponding frequency shifts. Also the film did not s how any signs of peeling or degradation. 4.3. Palladium Film One more RF monolithic RP 1239 SAW two-port resonator was uncapped and sputter coated w ith 1000 of Si02. The Si02 coated resonator was coated with 27 of palladium by e-beam evaporation process. To evaluate the changes in resonant frequency and attenuation due to addition of SiO2 and Pd sensing layer the S21 parameters were obtained from Agilent 8753ES S-parameter Ne twork Analyzer. The changes due to various coating have been ta bulated in Table 4.3. The unc oated SAW resonant frequency was found to be 315.010750 MHz with attenu ation of -3.5755 dB at the resonant frequency Table 4.3. SAW Resonator Frequency and Atte nuation Measurements after Pd Coating It was seen that the evaporation of Pd caused an attenuation of the signal and a corresponding decrease in the resonant freque ncy of the device. This resonator was then fixed on a PCB and placed in the Stainless Steel cell to be in contact with the incoming gas. This resonator was also made a part of oscillator loop to track the fractional changes in frequency. The resonator with the pure Pd film wa s exposed to 0.2-2.5% volume% hydrogen in nitrogen at a flow rate of 1000 sccm. The re sponse of the sensing film has been plotted Process Frequency change Attenuation change 1000 of SiO2 0.96 MHz -6.8885 dB 27 of Palladium 0.4 MHz -1.64 dB

PAGE 67

56 in Figure 4.17. In accordance with a ma ss-loading response mechanism, frequency decreases were observed upon hydrogen exposures [11]. Figure 4.18 shows the response times for pure Pd sensing layer, defined as ti me taken to reach nearly 99% of complete, for 0.2-2.5% of hydrogen exposures. It is seen from the plots that the process of absorption and desorption of hydrogen from the Palladium film is reversible, though recovery was slow. Figure 4.19 shows corre sponding calibration curve. The calibration curve also shows that the responses obtaine d for pure palladium sensing film are about 1/3 of Pd coated SWNTs se nsing film. Also the pure Pd sensing film shows poor responses at lower conc entrations (<1%). Figure 4.17. The Curve Showing the Response of Pd Film on Exposure to 0.2-2.5% Hydrogen

PAGE 68

57 80 90 100 110 120 130 140 0.511.522.53Volume% hydrogenResponse times (Seconds) Figure 4.18. Frequency Shifts for Pure Pd C oated SAW Resonator to 1-2.5% Hydrogen Exposures 0 500 1000 1500 2000 2500 3000 3500 4000 0.511.522.53Volume% hydrogenFrequency shift (Hertz) Figure 4. 19. Response Times for Pure Pd Coated SAW Resonator to 1-2.5% Hydrogen Exposures

PAGE 69

58 100 105 110 115 120 125 130 135 140 145 0.511.522.53Volume% hydrogenResponse times (Seconds) Run1 Run2 Run3 Figure 4.20. Response Times for three differen t runs of Pure Pd Coated SAW Resonator to 1-2.5% Hydrogen Exposures 0 500 1000 1500 2000 2500 3000 3500 4000 00.511.522.53Volume% hydrogenFrequency shift (Hertz) Run1 Run2 Run3 Figure 4.21. Frequency Shifts for three differen t runs of Pure Pd Coated SAW Resonator to 1-2.5% Hydrogen Exposures

PAGE 70

59 A few more runs were done by exposing the Pd coated sensing film to hydrogen to understand the behavior of the film to repeated hydrogen exposur es. It can be seen from the plot 4.20 and 4.21 that though the re sponse times are same, the size of shifts is going small. This shows the degradation of film by repeated exposur es, some of which were not seen in Pd coated SWNTs or Pd coated TMV sensing films. The degradation of film can be attributed to change in density of Palladium f ilm at high concentration [36, 37, and 38]. Given the area is fixed the film does not have enough area to expand causing it to crack and peel. This type of be havior was not seen in nanostruc tures. It could be justified as nanomaterials are disperse d in such a way that there is enough room for expansion, which is the reason w hy they don’t peel of. 4.4. Comparison of Films The three different sensing films were compared for their response times. The response times were compared for time taken to reach 63% of full scale shift. 0 20 40 60 80 100 120 140 00.511.522.53Volume% Hydrogen Response time (Seconds) Pd Coated TMVs Pd Coated SWNTs Pure Pd Figure 4.22. The Response Times Observed for the Sensing Film s for 63% of Full Response

PAGE 71

60 It was observed that the lowest response times were seen to be for Pd coated TMV virus and highest for Pd coated SWNTs as shown in figure 4.22. Also th e amount of frequency shifts produced for 63% of full scale shift is also plotted in figure 4.23. It is sent that the highest frequency shifts were observed for Pd coated SWNTs while lowest for Pd coated TMV sensing film. The table 4.4 compares th e various sensor parameters for the three films. 0 1000 2000 3000 4000 5000 6000 7000 8000 00.511.522.53Volume% hydrogenFreqeuncy shift (Hertz) Pd Coated TMVs Pd Coated SWNTs Pure Pd Figure 4.23. The Frequency Shifts Observed fo r the Sensing Films for 63% of Full Scale Response Table 4.4. Comparison of Different Sensor Parameters for the Three Sensing Films Pd coated TMV Pd coated SWNTs Pure Pd Reversible Yes Yes Yes Response Mechanism Acousto-elect ric Mass loading Mass Loading

PAGE 72

61 Table 4.5. Continued Average Response time-63% of full scale shift (Seconds) 19.09091 113.1943 77.47727 Frequency shift63% of full scale shift(Hertz) 1% 93.54545 3627.91 1242.182 1.5% 115.1818 5370.27 1568 2% 94.18182 6342.00 2036.364 Repeatability (~50-60 cycles over 8 days) Yes Yes No Reproducibility Yes(6-8 different devices were tested) Yes (2 devices were tested) N.A. Direction of shifts +ive* -ive** -ive** Cycling (requires exposure for few cycles before responding correctly) No No Yes Film robustness Yes Yes No The resonant frequency increases after analyte absorption by the sensing film ** The resonant frequency decreases after analyte absorption by the sensing film

PAGE 73

62CHAPTER 5: CONCLUSION In summary we tested three sensing films to see the response to hydrogen challenges. The two nonmateria l sensing films, Pd coated TMV and Pd coated SNWTs were compared with pure Pd film to obser ve if there is any advantage of using nanomaeterial gas sensing films. The Pd-TMV film produced small, but easily measurable responses with fast response and recovery times for nearly full response. Frequency increases were observed upon hydrogen exposure, which is unexpected and generally associated with an electroacoustic response mechanism. The Pd-SWNT la yers produced much la rger shifts, for a very small coating thickness of 27 of ev aporated Pd on the SWNTs. The frequency decreases observed upon hydrogen exposure are the expected behavior for a mass-loaded SAW sensor response. Shifts were very large and were in the kHz range. The 99% response time was somewhat larger at 150 – 200 seconds, than for the Pd-TMV layer. This behavior of nanomaterials was compar ed to 27 Pd film on SAW resonator. The sensing layer responded by pr oducing a decrease in freque ncy in accordance with mass loading effect for hydrogen challenges and pr evious literature. Thes e shifts were also seen to be in KHz range. The process of absorption and desorp tion was reversible for all three nanomaterials under room conditions. The SWNT sensing film showed highest frequency shifts for any given concentration, with the smallest shifts observed for Pd coated TMV layer. Both TMV and SWNTs film showed no change in the responses sizes, response times or film behavior due to repeated e xposure to hydrogen. The behavior was however not observed in pure Pd film wherein repeated exposure resulted in the size of shifts decreasing with no appreciable changes in response times. Th is shows a degradation of sensing film with time and use. At high concen trations the Pd film density changes which causes the film to expand. As the film doesn ’t have enough room for expansion they tend

PAGE 74

63 to crack and peel. This behavior is not seen in nanostructures as they are dispersed in such a manner that there is enough room for expansion. Both Pd coated TMV and Pd coated SWNTs showed good repeatilibty when tested for over 50 cycles over a span of 8 days which was not seen with pure Pd film s. Also the TMV and SWNT sensing film showed reproducibility. For same level of Pd (27 ) in pure Pd and SWNT film the amount of shift was seen to be three times higher for SWNT as comp ared to pure Pd films. At concentrations below 10000ppm the pure Pd film did not give definite responses which may be related to the need of cycling for pure Pd films. The SWNT film had no problems in giving clear and distinguishable response at lower concen trations. But the response times of SWNT sensing film was seem to be higher than Pd film. This can be explained by the behavior of SWNTs. SWNTs are known to store hydrogen which may be causing the response times to increase as it might take longer th an usual for the both Pd and the SWNTs to achieve equilibrium. SWNTs are capable of storing hydrogen in the 5-10 wt% range and if catalytic metal species are present SWNTs can adsorb up to 8 wt% hydrogen All these nanomaterials were tested for 2000-25000ppm of hydrogen. Today with commercial hydrogen sensors available from 0-5000ppm, future work would require testing these devices for lower ppm levels (0-2000ppm). These commercial sensors have response times of 15 seconds. Though the TMV sensing material produced ~ 20sec of response times comparable to commercial sens ors suitable electronics would have to be designed. When compared to commercial hydrogen sensors costing 100$, the sensors designed using the concept discussed in this th esis could be manufact ured at a fraction of the cost. Considering the SAW resonator itself costs only 50cents, th e entire associated electronics could be designed for under 5$ if large scale manufact uring is adopted. Also, in future, in order to achieve bi gger responses at lowe r concentration we would need to increase the coating on these de vices. This would require us to evaluate the coating thickness experimentally to see if hi gher coating could ge nerate a higher shift without comprising a lot on response time. Also the nanomaterials sensing film need to be characterized better, part icularly the TMV samples. The TMV samples which produce

PAGE 75

64 an increase in frequency more in accordance with acousto-electric phenomenon can be better evaluated by coating them on SAW de lay line configuration or Lithium Niobate wafer which produces a good response to aco uscto-electric phenomenon. Temperature effects on the nanomaterials also need to be studied if commercialization is to be achieved as it should be able to work in a wi de range of temperature with no appreciable change in response characteristics. The su rface morphology before and after exposure need to be evaluated using TEM and AF M to understand the structural changes undergone by the nanomaterials. TEM and AFM measurements on the nanomaterial films will also reveal the mechanisms by which these films are more stable to repeated hydrogen exposures. TMV and SWNT nanomaetrial s coated with different metals could be studied to understand their re sponses with different gases.

PAGE 76

65REFERENCES 1. D. E. Gobina, "Market Research Re port: Gas Sensors and Gas Metering Applications and Markets," pp. 0-318, 2005. 2. C. G. Padro, "US department of Energy’s Hydrogen program," National Renewable Energy Lab. 3. "National Hydrogen Energy Roadmap: Produc tion, delivery, storage, conversion, applications, public education and outrea ch," United States Department of Energy, Washington, DC April 2-3 2002. 4. C.Christofedes and A. Mandelis, So lid State Sensors for trace hydrogen gas Detection," Journal of Applied Physics vol. 68, pp. R1-R30, 1990. 5. A. Katsuki and K. Fukui, "H2 Sel ective gas sensor based on SnO2," Sensors and Actuators vol. B, pp. 30-37, 1998. 6. M. Thompson and D. C. Stone, Surface launched Surface Acoustic Wave sensors Chemical Sensing and Thin film characterization vol. 144: Wiley Interscience, 1997. 7. H. Wohlten and R. Dessy, Analytical Chemistry pp. 1458, 1979. 8. H. Wohlten and R. Dessy, Analytical Chemistry pp. 1458, 1979. 9. H. Wohlten, Sensors and Actuators vol. 5, pp. 307, 1984. 10. C. K. Campbell, Surface Acoustic Wave Device s for Mobile and Wireless Communication : Academic Press, 1998. 11. Jakubik, W. P., M. W.Urbanczyk, et al (2002). "Bilayer Structure of Hydrogen detection in a surface acoustic wave sensor system." Se nsors and Actuators B(82): 265-271. 12. Jakubik, W. P., M. W. Urnanczyk, et al (2003). "Palladium and Phthalocyanine bilayer films for hydrogen detection in a surface acoustic wave sensor system." Sensors and Actuator s B 96(1-2): 321-328. 13. Penza, M., E.Milella, et al. (1998). Monitoring of NH3 gas by LB polypyrrolebased SAW sensor." Sensors and Actuators B(47): 218-224.

PAGE 77

66 14. Penza, M., F. Antolini, et al. (2004). "Carbon Nanotubes as SAW chemical sensors." Sensors and Actuators B(100): 47-59. 15. Penza, M., G. Cassano, et al. (2002). "Identification and quantification of individual volatile organic compounds in a binary mixture by SAW multisensor array and pattern recognition." Measurement Science and Technology 13: 846858. 16. J. D.S. Ballantine, S. J. Martin, A.J.Ricco, G.C.Frye, and R.M.White, Acoustic Wave Sensors. Theory, Design and Physico-Chemical Applications: Academi c Press, 1997. 17. M.J.Jaycock and G.D.Parfitt, Chemistry of Interfaces. Chichester, UK: Ellis Horwood, 1981. 18. J.W.Grate and R.A.McGill, Analytical Chemistry, pp. 2162, 1995. 19. J.W.Grate, S.W.Wenzel, and R.M.White, Analytical Chemistry, vol. 63, pp. 1552, 1991. 20. A.W.Snow, W.R.Barger, M.Klusty, H.Wohltjen, and N.L.Jarvis, Langmuir, vol. 2, pp. 513, 1986. 21. D.L.Bartley and D.D.Dominguez, Analytical Chemistry, vol. 62, pp. 1649, 1990. 22. SES. Research, "Single Walled Nanotubes. from http://www.sesres.com/nanotubes.asp. (September, 2005) 23. "TEM image of SWNTs produced by the CoMoCat method." from http://members.cox.net/hongweizhu/research.htm#2. (September, 2005) 24. "Examples for armchair, zigzag and chiral nanotube a) (5, 5)-, b) (9, 0)and c) (10, 5)-nanotube." from www.ipc.uni-karlsruhe.de/ mik/seite_195.html. (September, 2005) 25. "High resolution TEM image of a MWNTs." from http://133.5.181.45/ago/images/tem_hr_mwnt.jpg. (September, 2005) 26. F. L. Pfleger and R. J. Zeyen, "Tomato-Tobacco Mosaic Virus Disease." 27. A.A. Balandin and V. A. Fonoberov, "Vibrational Modes of Nano-Template Viruses," Journal of Biomedical Nanotechnology, vol. 1, pp. 90-95, 2005. 28. The American Phytopathological Society, 2001 from http://www.apsnet.org/education/feature/tmv/images/tmvdraw.htm. (Septem ber, 2005)

PAGE 78

67 29. S. Y. Lee, E. Royston, J. Culver, and M. T. Harris, "Improved metal cluster deposition on a genetically engineered tobacco mosaic virus template," Nanotechnology, pp. S435–S441, 2005. 30. A.Chaudhari, "Development of Su rface Acoustic Wave Sensors Using Nanostructured Palladium for Hydrogen De tection," in Department of Chemical Engineering, vol. Masters. Tampa: University of South Florida, 2004. 31. J. Freudenberg, Von Schickfus, M., H unklinger, S., "A SAW immunosensor for operation in liquid using a SiO2 protective layer," Sensors and Actuators, vol. B, pp. 147–151., 2001. 32. F. Josse, Bender, F., Cernosek, R.W., "Guided shear horizont al surface acoustic wave sensors for chemical and biochemi cal detection in liquids," Analytical Chemistry, pp. 5937–5944, 2001. 33. G. Kovacs, Vellekoop, M.J., Haueis, R., Lubking, G.W., Venema, A., "A Love sensor for (bio)chemical sensing in liquids ," Sensors and Actuators, vol. A, pp. 38–43, 1994. 34. B. Jakoby, Vellekoop, M.J., "Analysis a nd optimisation of Love-wave liquid sensors.," Ferroelectr. Frequn ecy Control, pp. 1293–1301, 1998. 35. R. C. Hughes, W. K. Schubert, T. E. Zipperian, and T. A. Plut, "Thin film palladium and silver alloys and laye rs for metal-insulator semiconductor sensors.," Journal of Applied Physics, vol. 62, pp. 1074-1081, 1987. 36. A.Fabre, E. Finot, J. Demo ment, S. Contreras, In situ measurement of elastic properties of PdHx, PdDx and PdTx, J. Alloys Compd.356-357 (2003) 372-376. 37. V.I. Anisimkin, I.M. Kotelyanskii, V. I. Fedosov, C. Caliendo, P.Verardi, E. Verona, Analysis of the different cont ributions to the response of SAW gas sensors, in: Proceedings of the IEEE Ultrasonics Symposium, 1995, pp. 515-518. 38. K. Yamanaka, S. Ishikawa, N. Nakaso, N. Takeda, T. Mihara, Y. Tsukahara, Ball SAW device for hydrogen gas sensor, in: Proceedings of the IEEE Ultrasonics Symposium, 2003, pp. 299302.

PAGE 79

68 APPENDICES

PAGE 80

69APPENDIX A PRINTED CIRCUIT FABRICATION Procedure 1. Make the layout for obtain ing the pattern in AutoCAD or any other suitable designer 2. Obtain (print) the layout on a transparency. NOTE: The PCB has positive photo resist, so exposed areas are removed while developing. 3. Process to prepare developer: Ingredients: NaOH – (6-9) gms DI water – 1000 ml = 1 litre Beaker Mix the NaOH and DI water and stir it using magnetic stirrer till all the NaOH crystals are dissolved. Note: The higher the amount of NaOH the faster the developing time. 4. Then take the PCB and with the cutter cut into suitable dimension 5. Then take the PCB and place the transparency pattern on it 6. Place the entire system on the exposing modu le with the side to be exposed facing the light through the transparency. 7. Close the box and set the number of lamps to 5 and set the timer to 3. 8. Then after the time elapses open a nd put the PCB in solution (developer) 9. It takes anywhere from 1-2 minutes fo r developing which can be clearly seen by the removal of the green photo resist thus exposing the Copper. NaOH solution which is clear also turns green due to this. 10. Now the areas that where not exposed were will still have photo resist on them. 11. Now place the entire PCB in the solution of ferric chloride which causes the copper in unwanted areas to be removed (the areas expos ed).The entire process may take anywhere from 20-30 minutes with ferric chloride at (3040) C. The end of process is indicated by the board becoming transparent. 12. Then put it in acetone to remove the photo resist.

PAGE 81

70 13. Then cut the board into required dimension. 14. Drill holes where it’s required and then complete the circuit.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001709562
003 fts
005 20060614112247.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 060526s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001325
035
(OCoLC)69396486
SFE0001325
040
FHM
c FHM
049
FHMM
090
TK145 (Online)
1 100
Srinivasan, krishnan.
0 245
Nanomaterial sensing layer based surface acoustic wave hydrogen sensors
h [electronic resource] /
by krishnan Srinivasan.
260
[Tampa, Fla.] :
b University of South Florida,
2005.
502
Thesis (M.S.E.E.)--University of South Florida, 2005.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 81 pages.
520
ABSTRACT: This thesis addresses the design and use of suitable nanomaterials and surface acoustic wave sensors for hydrogen detection and sensing. Nanotechnology is aimed at design and synthesis of novel nanoscale materials. These materials could find uses in the design of optical, biomedical and electronic devices. One such example of a nanoscale biological system is a virus. Viruses have been given a lot of attention for assembly of nanoelectronic materials. The tobacco mosaic virus (TMV) used in this research represents an inexpensive and renewable biotemplate that can be easily functionalized for the synthesis of nanomaterials. Strains of this virus have been previously coated with metals, silica or semiconductor materials with potential applications in the assembly of nanostructures and nanoelectronic circuits.Carbon nanotubes are another set of well-characterized nanoscale materials which have been widely investigated to put their physical and chemical properties to use in design of transistors, gas sensors, hydrogen storage cells, etc. Palladium is a well-known material for detection of hydrogen. The processes of absorption and desorption are known to be reversible and are known to produce changes in density, elastic properties and conductivity of the film. Despite these advantages, palladium films are known to suffer from problems of peeling and cracking in hydrogen sensor applications. They are also required to be cycled for a few times with hydrogen before they give reproducible responses. The work presented in this thesis, takes concepts from previous hydrogen sensing techniques and applies them to two nanoengineered particles (Pd coated TMV and Pd coated SWNTs) as SAW resonator sensing materials.Possible sensing enhancements to be gained by using these nanomaterial sensing layers are investigated. SAW resonators were coated with these two different nano-structured sensing layers (Pd-TMV and Pd-SWNT) which produced differently useful hydrogen sensor responses. The Pd-TMV coated resonator responded to hydrogen with nearly constant increases in frequency as compared to the Pd-SWNT coated device, which responded with concentration-dependent decreases in frequency of greater magnitude upon hydrogen exposure. The former behavior is more associated with acousto-electric phenomena in SAW devices and the later with mass loading. The 99% response times were 30-40 seconds for the Pd-TMV sensing layer and approximately 150 seconds for the Pd-SWNT layer. Both the films showed high robustness and reversibility at room temperature.When the Pd film was exposed to hydrogen it was observed that it produced decreases in frequency to hydrogen challenges, conforming to mass loading effect. It was also observed that the Pd film started degrading with repeated exposure to hydrogen, with shifts after each exposure going smaller and smaller.
590
Adviser: Dr Venkat Bhethanabotla.
Co-adviser: Dr Paris H. Wiley
653
Saw resoantor.
Tobacco mosaic virus.
Carbon nanotubes.
Palladium.
Adsorption.
690
Dissertations, Academic
z USF
x Electrical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1325