USF Libraries
USF Digital Collections

Palladium doped nano porous silicon to enhance hydrogen sensing

MISSING IMAGE

Material Information

Title:
Palladium doped nano porous silicon to enhance hydrogen sensing
Physical Description:
Book
Language:
English
Creator:
Luongo, Kevin
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Room temperature
Gas sensor
Nanoparticle
Labview
Palladium diffused
Dissertations, Academic -- Electrical Engineering -- Masters -- USF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: A mass manufacturable impedance based, palladium doped porous silicon sensor, was fabricated for hydrogen detection. The sensor was built using electrochemical etching to produce mesoporous silicon. Four nanometers of palladium was defused directly into the porous silicon and another four nanometers of Pd was deposited on the defused surface to enhance sensing. The sensor was tested in a sealed chamber in which the impedance was measured while hydrogen in nitrogen was ranged from 0-2 percent. Unlike conventional hydrogen sensors this sensor responded at room temperature to changes in hydrogen concentration. The electrical impedance response due to adsorption and desorption of hydrogen reacted relatively quickly due to the nanoparticle nature of palladium diffusion in and Pd evaporation on porous silicon.
Thesis:
Thesis (M.A.)--University of South Florida, 2006.
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 Kevin Luongo.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 54 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 - 001795168
oclc - 146315119
usfldc doi - E14-SFE0001542
usfldc handle - e14.1542
System ID:
SFS0025860:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Palladium Doped Nano Porous Sili con to Enhance Hydrogen Sensing by Kevin Luongo 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 Major Professor: Shekhar Bhansali, Ph.D. Yogi Goswami, Ph.D. Sangchae Kim, Ph.D. Date of Approval: March 24, 2006 Keywords: room temperature, gas sensor, nanoparticle, Labview, palladium diffused Copyright 2006, Kevin Luongo

PAGE 2

DEDICATION I would like to dedicate this to my son, Samuel Luongo. It is for him I continue to push myself beyond perceived limitations. Though I may desire, it is not my expectation of him to follow in my field of study or even my path of higher education. I do wish for him the strength to get up when he falls and to continue toward his goals no matter how unlikely success may appear. With out question I could not have comp leted this milestone without the support and patients of my wife, Ha Doan Nguyen. Her loving support has given and continues to give me strength and courage. Doan I am truly grateful, anh yeu em!

PAGE 3

ACKNOWLEDGMENTS I extend my sincere gratitude and appreci ation to the many people who made this masters thesis possible. Special thanks are due to Dr. Bhansali who has taken the time and energy to guide me from an undergradu ate and who continues to guide me through my graduate tenure. I am highly appreci ative to Dr. Vankat for providing initial accommodations for testing while testing cham bers were being established within the BioMEMs department. Many thanks are due to my committee members Dr. Yogi Goswami and Dr. Sangchae Kim for taking time from there own schedules and providing invaluable feedback. Many thanks to Praveen Shekhar for his technical feedback that helped contribute to the completion of this thesis. I w ould also like to acknowledge Altagrace Sine with much appreciation for her crucial role in testing and sample preparation. Recognition should be given to the outstanding BioMEMs students from 2001 to present who all have impacted my unde rstanding of this field. Many thanks go to the Government Department N SF whose financial assistance was vital for this research. Many more persons participated in various wa ys to ensure my research succeeded and I am thankful to them all.

PAGE 4

i TABLE OF CONTENTS LIST OF FIGURES ........................................................................................................................iv ABSTRACT ............................................................................................................................... .....vi CHAPTER 1 INTRODUCTION ............................................................................................1 1.1 Thesis Overview ....................................................................................................................1 1.2 Motivation .............................................................................................................................2 1.3 General Introduction to Hydrogen Sensor Research .............................................................3 1.4 Applications of Hydrogen Sensors ........................................................................................4 1.5 Significance of Current Work ................................................................................................5 CHAPTER 2 HYDROGEN SENSOR CLASIFICATION .....................................................6 2.1 Solid State Semiconductor Sensor .........................................................................................6 2.1.1 MIS-Schottky Barrier .....................................................................................................6 2.1.2 MIS-Capacitor ................................................................................................................7 2.1.3 MIS-Transistor ...............................................................................................................8 2.2 Optical Sensors ......................................................................................................................9 2.2.1 Absorptance Sensor ........................................................................................................9 2.2.2 Reflectance Sensor .........................................................................................................9 2.2.3 Surface Plasmon Resonance ...........................................................................................9 2.2.4 Fiber Optic Hydrogen Sensors .......................................................................................9 2.3 Electrochemical Sensor .......................................................................................................10 2.4 Resistive Hydrogen Sensors ................................................................................................11 2.4.1 Chemiresistors ..............................................................................................................11

PAGE 5

ii 2.4.2 Conducting Polymers ...................................................................................................12 2.4.3 Impedance Sensor ........................................................................................................12 2.4.4 Microhotplate...............................................................................................................13 CHAPTER 3 POROUS SILICON .........................................................................................14 3.1 Introduction .........................................................................................................................14 3.2 Porous Silicon Theory .........................................................................................................15 3.3 Porous Silicon Sensors........................................................................................................17 3.3.1 Contact Potential Difference ........................................................................................17 CHAPTER 4 PALLADIUM .................................................................................................19 4.1 Introduction .........................................................................................................................19 4.2 Diffusion..............................................................................................................................1 9 4.2.1 Diffusion Theory..........................................................................................................19 4.3 Palladium Absorption Principles .........................................................................................20 4.4 Palladium-Hydrogen System ...............................................................................................21 CHAPTER 5 HYDROGEN SENSOR FABRICATION .......................................................25 5.1 Introduction .........................................................................................................................25 5.2 Complete Sensor Process .....................................................................................................26 CHAPTER 6 TEST SETUP AND PROTOCOLS ................................................................33 6.1 Introduction .........................................................................................................................33 6.2 Hardware .............................................................................................................................33 6.3 Software ............................................................................................................................... 37 CHAPTER 7 RESULTS AND DISCUSSION ......................................................................41 7.1 Introduction .........................................................................................................................41 7.2 Morphology .........................................................................................................................41 7.3 Hydrogen Test Response .....................................................................................................48

PAGE 6

iii 7.4 Conclusion ...........................................................................................................................51 7.5 Future Work .........................................................................................................................52 REFERENCES ..............................................................................................................................5 3

PAGE 7

iv LIST OF FIGURES Figure 2.1 Energy Diagrams of Metal-Oxide-Semiconductor Junction [9] .......................7 Figure 2.2 Schematic of Hydrogen Sensitive Pd-MOS Capacitor [9] ................................8 Figure 2.3 General Diagram of a Chemiresistor [14] .......................................................12 Figure 3.1 I-V Characteristics Governing El ectrochemical Dissolution of Silicon [26] ..15 Figure 3.2 Anodic Dissolution of Silicon in HF [26] .......................................................16 Figure 3.3 Doping Density Vs Pore Density [26] .............................................................17 Figure 3.4 CPD Response to Hydrogen in Minutes [30] ..................................................18 Figure 4.1 (a) Substitution Atom Taking up a Vacancy (b) Interstitial Atom Sitting Between Silicon Lattice. .................................................................................20 Figure 4.2 Hydrogen Atoms Occupying the Tetr ahedral Interstitial Site at x=1 [11] ......22 Figure 4.3 Hydrogen Atoms Occupyi ng the Octahedral Sites [11] ..................................22 Figure 4.4 Different Versions of Palladium-Hydride [10] ................................................23 Figure 4.5 PCT Curve of Pd-H System [12] .....................................................................23 Figure 5.1 Process Flow for Hydrogen Sensor Fabrication ..............................................26 Figure 5.2 AJA International Electron Beam Evaporator .................................................27 Figure 5.3 Planetary Rotation ...........................................................................................28 Figure 5.4 (a) Rendering (b) Photo of Etching Jig for Pore Formation ............................29 Figure 5.5 Aluminum Etching Using HF Vapor at Room Temperature ...........................30 Figure 5.6 Dicing Saw Stage ............................................................................................30 Figure 5.7 Open Tube Furnace .........................................................................................32

PAGE 8

v Figure 6.1 Test Bed ...........................................................................................................34 Figure 6.2 Mass Flow Cont rollers and Test Bed ..............................................................35 Figure 6.3 Keithley Multimeter (t op), MKS Controller (bottom) ....................................35 Figure 6.4 Four Wire Probe Configuration [31] ...............................................................36 Figure 6.5 Hardware Test Setup .......................................................................................37 Figure 6.6 Testing Module ................................................................................................38 Figure 6.7 Main Body of Code .........................................................................................40 Figure 7.1 SEM of Porous Silicon Substrate at 200k Magnification ................................42 Figure 7.2 SEM of Porous Silicon Substrate at 300k Magnification ................................42 Figure 7.3 Importing an Image .........................................................................................43 Figure 7.4 Calibrate Image Pixels to Real World Units Via SEM Scale Bar ...................44 Figure 7.5 Determine the Pore by Choosing a Proper Threshold Value ..........................44 Figure 7.6 Assessing Information on Spreadsheet ............................................................45 Figure 7.7 Side View of Sample After Pd Diffusion........................................................46 Figure 7.8 Surface Image of Sample After Diffusion .......................................................47 Figure 7.9 XRD Analysis of Sensor.................................................................................47 Figure 7.10 Initial Sensor Reaction During Conditioning ................................................49 Figure 7.11 Sensor Reaction Ending Conditioning ..........................................................49 Figure 7.12 Sensor Reaction After Conditioning .............................................................50 Figure 7.13 Resistances Displaying Pr oportional Relationship to Hydrogen...................51

PAGE 9

vi PALLADIUM DOPED NANO PORO US SILICON FOR ENHANCED HYDROGEN SENSING Kevin Luongo ABSTRACT A mass manufacturable impedance based, palladium doped porous silicon sensor, was fabricated for hydrogen detection. The sensor was built using electrochemical etching to produce mesoporous silicon. Four nanometers of palladium was defused directly into the porous silic on and another four nanometers of Pd was deposited on the defused surface to enhance sensing. The sensor was tested in a sealed chamber in which the impedance was measured while hydrogen in nitrogen was ranged from 0-2 percent. Unlike conventional hydrogen sensors this sensor responded at room temperature to changes in hydrogen concentra tion. The electrical impedance response due to adsorption and desorption of hydrogen reacted relatively quickly due to the nanoparticle nature of palladium diffusion in and Pd evaporation on porous silicon.

PAGE 10

1 CHAPTER 1 INTRODUCTION 1.1 Thesis Overview This thesis focuses on hydrogen sensor theory and development. The purpose of chapter one is to demonstrate the need for a well designed hydrogen sensor by interpreting present technological goals and trends as explicated under the following motivation section of this chapter. Ch apter one includes a ge neral introduction to hydrogen sensor research and a se ction dedicated to current a pplications. Brief analysis and limitations of proposed methodologies are al so included in this section. Chapter two elaborates on the state of the art developments and is accompanied by an explanation on correlating theory. This survey is designed to arm the read er with knowledge of current hydrogen sensor design and development. Chap ter three introduces por ous silicon theory. Chapter four introduces palladium as a hydrogen sensing element. Palladium is used in the fabrication of the hydrogen sensor in this research and will theref ore be discussed at length. Chapter five presents the sensor fabrication process employed during the development phase of the functional hydrogen sens or presented in this work. All relevant process parameters and experimental stages are documented. Chapter six describes the hardware and software used for the sensor testing chamber and data acquisition. Finally Chapter seven includes results obta ined from testing and discussions.

PAGE 11

2 1.2 Motivation The need for a hydrogen sensor with high se nsitivity, fast regeneration, and a fast response time has gained momentum with th e advent of hydrogen fuel cell technology. Hydrogen has become one of the most promisin g gases relating to vari ed applications in transportation, food, metallurgy, electronic and nuclear industries. Focus on hydrogen sensor development is becoming increasingly critical considering that in recent years there has been a large push to develop hydrogen as a fossil fuel alternative. Hydrogen can provide signifi cant power while producing heat and water as the only byproducts, thus allowing for an environmentally friendly fuel. Hydrogen use as a clean energy source is estimated to account for a trillion dollar revolution in the near future. However a small hydroge n leak in any of its applications ca n lead to a fatal accident. Hydrogen when mixed with air between 4.65 and 93.9 percent volume are within explosive limits [1]. Thus it is importa nt to detect small leak s as they can signify potentially hazardous conditions. This calls for robust sensor mechanisms which can detect PPM concentrations. If indeed we as a society plan to use hydrogen as a fuel for personal and/or public transportation then it would be inevitable that hydrogen sensors become an essential safety precaution standard in all hydrogen fueled vehicles and processes. Numerous companies and organizations such as NASA and DOE use large quantities of hydrogen and oversee the development of the technology. Aware of the flammable or explosive nature of hydrogen, these organizations have outlined detailed performance criterion for acceptable hydrogen sensors [2]. The ideal sensor would be affordable, easy-to-fabricate, user friendly and responsive to low concen trations of hydrogen in real time while

PAGE 12

3 eliminating any possible false positives e .g. cross contaminating gasses which induce false alarms. Though false positives may be a major irritant, the possi bility of a sensor failing to detect levels of hydrogen altoge ther is unacceptable and therefore sensor characterization is essential. Another id eal property to decrea se sensor cost and development would be the ability to in tegrate a sensor into Lab on a Chip configurations for porta bility. Portability wi ll lead to field em ployment and system integration when needed. This research ex plores the development of such a sensor. 1.3 General Introduction to Hydrogen Sensor Research Numerous approaches are currently bei ng investigated in the development of hydrogen sensors, these include but are not limited to: sol-gel based, semiconductor based, oxide based, thin-film based and acous tic wave sensors. These devices are relatively high power consumers, show slow response times or do not have the required sensitivity. In order to a ddress these limitations the use of nanostructures for hydrogen sensing has attracted increased interest. There has been overwhelming interest in understanding the fabrication and physical in tricacies of nanostructured materials both from basic scientific standpoint and from an industrial (mass fabrication) standpoint [3, 4]. Nanostructures such as nanowires and qua ntum dots exhibit fascinating optical, thermal, electronic, mechanical and transpor t properties that coul d be utilized in many sensing applications. Within the past fe w years nanowire based sensors for detecting hydrogen have been reported. These sensors respond in real time, however they lack sensitivity and they do not respond to low c oncentrations of hydrogen. Additionally, the techniques used to fabricate these nanowire sensors entail complex procedures, such as

PAGE 13

4 transfer and organized assembly [5-7]. Th ese fabrication methodologi es add to cost and do not seem suitable for commercial production. 1.4 Applications of Hydrogen Sensors Research has indeed exhibited a growing interest in hydroge n sensor technology due to industrys increased desire to promote safety standards. This is a driving force in sensor technology and has led to increased mi niaturization, functionality, bulk fabrication methodology and increased reliab ility. Therefore hydrogen se nsors have been utilized more readily within the past years. A broad list of applications that can benefit from hydrogen sensors has been described else where and is summarized below. [8] Sensing hydrogen buildups in lead acid st orage cells found in most vehicles and other applications. Detecting hydrogen leaks during ammonia, methanol manufacturing, and desulphurization of petroleum products along with many other petrochemical applications where high pr essure hydrogen is used. Detecting impending transformer failu re in electric power plants. Monitoring hydrogen buildup in radioactive waste tanks and in plutonium reprocessing. Detecting hydrogen leaks during space shuttle launches and other National Aeronautics and Space Administ ration (NASA) operations. Molecules and other gases besides hydroge n can be detected using arrays of sensors with different catalytic metals. In the future hydrogen economy, sens ors will be widely needed.

PAGE 14

5 1.5 Significance of Current Work This research has been undertaken in order to design and fabricate a porous silicon based hydrogen sensor which utilize na noparticle palladium and palladium silicide as a sensing element. The goal of this wo rk was to investigate the effect on hydrogen sensing due primarily to the impregnation of palladium deep impurities into a porous silicon template. This work would contribute significantly to exis ting technologies as follows: Ease of fabrication: In this work, hydrogen sensing was accomplished using well known porous silicon etching methods, eva porations and thermal processes which enable mass production capabilities. Reduced cost: The methods mentioned above have been used in IC processing and other industrial applica tions for decades. This primary factor has made equipment readily accessible and, as a result, silicon production has become relatively inexpensive. Lab on a chip: The sensor is derived from a silicon wafer thereby significantly increasing its ability to be integrated w ithin a multifunctional array of sensors and on-board electronics. Speed: The sensing rate is increased sin ce the surface area to volume ratio of the sensing element is optimized as nanostructures. Safety: The introduction of a mass producib le and cost effective sensor with improved sensing rates will allow for an increased implementation of safety.

PAGE 15

6 CHAPTER 2 HYDROGEN SENSOR CLASIFICATION 2.1 Solid State Semiconductor Sensor Solid state sensors are fabr icated as MIS (Metal-Ins ulator-Semiconductor) diodes, capacitors and transistors. Their sub category of MOS (Metal-O xide-Semiconductor) is the most common field deployable sensor because of the known characteristics. 2.1.1 MIS-Schottky Barrier When a metal comes in contact with a semiconductor the Fermi levels automatically align themselves proportionally to an amount equal to the difference between the work function of the metal and th e electron affinity in the semiconductor. The adjustment is not disturbed due to the presence of a sandwiched dielectric layer as shown in figure 2.1. The barrier height is given by: q Bn =q ( m -X) (1) Where, q Bn Barrier Height q m Metal Work Function q X Electron affinity in the semiconductor E C Lowest energy conduction band E V Higher energy Valence band

PAGE 16

Figure 2.1 Energy Diagrams of Metal-Oxide-Semiconductor Junction [9] There are two hypotheses explaining the consequence of passing hydrogen. First being the polarization of hydroge n atoms [9] due to the intrin sic electric field of oxide when the sensor is biased. This in turn redistributes charges in the depletion region decreasing the degree of bending or other words reducing the barrier height. The conducting current is directly proportional to the amount of hydrogen atoms adsorbed at the interface. The second being the formation of excess interfacial charge states on passing hydrogen thereby reducing the Schottky height. Some of the past sensors include the Pt\SiC, Pd\Cu, Pd-ZnO, Pd-CdS and Pd-TiO 2 systems. 2.1.2 MIS-Capacitor The capacitive sensors were characterized by ultra thin insulating materials deposited by special techniques like plasma de position [10], pulsed laser ablation and LB (Langmuir Blodgett) process which produced a high quality dielectric. The capacitance variation near the depletion region determin es the concentration of hydrogen adsorbed. 7

PAGE 17

Figure 2.2 Schematic of Hydrogen Sensitive Pd-MOS Capacitor [9] 2.1.3 MIS-Transistor When the gate region of a conventional MOSFET is coated with palladium, it functions as a hydrogen sensor. The sensitivity of this device corresponds to the effect of hydrogen on the gate metal work function. Several versions based on the same principle have been implemented like the OGFET (Open-Gate Filed Effect Transistor), ADFET (Adsorption Field Effect Transistor) a nd SAFET (Surface-Accessible Field Effect Transistor). Advantages of solid state sensors: Wide sensing range Ease of fabrication Low Cost Disadvantages: Non-linear relationship Unreliable due to possibilities of contamination 8

PAGE 18

9 Suitability of the sensor: Ideally used in isolated chambers for th e detection of very minute hydrogen leaks 2.2 Optical Sensors Optical sensors use visible or invisible light beams to detect objects and make measurements. Many of the hydrogen sensor s use optical properties to determine hydrogen concentrations. These se nsors are classified below. 2.2.1 Absorptance Sensor These are based on the light intensity shif t dictated by the quantity of absorbing gas in the optical path. Using th e Beer-Lambert relationship [9], A i = m a X C o ... (2) Where, A i is the absorbtance; m a is the gas molar absortivity and C o concentration of the absorbing species. 2.2.2 Reflectance Sensor The reflectance property is measured as a function of the composition of an optically excited system. 2.2.3 Surface Plasmon Resonance The interaction of the sens ing layer (Palladium) with the gas causes the coupling of photons to surface plasmons (waves) at the interface. Th is induces a change in the refractive index of the active layer[11]. 2.2.4 Fiber Optic Hydrogen Sensors In this type, the palladium layer acts as chem-optical transducer and a support for the surface plasmon wave. Generally, the valu e of real and imaginary part of complex permittivity of palladium [11] is equal for visible wavelength spectrum. The principle

PAGE 19

10 behind a FOS (Fiber Optic Sensor) is change in the permittivity of palladium on contact with hydrogen. Advantages of optical hydrogen sensor: Capable of detecting up to ppb (parts per billion) concentration of H 2 No electrical interference Small size Disadvantages: Poor selectivity and irreversibility High background noise levels hindering the resolution Suitability: Deployed in very dangerous environment Considerable electromagnetic activity 2.3 Electrochemical Sensor Electrochemical sensors are based on the ionic conductivity in thin films. Basically it is an electrochemical cell set up with two electrode compartments separated by proton conducting electrolyte. The EMF (El ectromotive Force) generated is used for calculating the gas concentration adsorbed in on e of the active electrodes. Potentiometric sensors are generally avoided due to the lack of accuracy at higher concentrations and high power consumptions. Empirically, the EMF generated is found to vary with the following relationship [9]: I = k [H] 0.5 (3) Where ICurrent generated due to the induced EMF

PAGE 20

11 H-Concentration of the ionic species Amperometric sensors are preferred due to the linearity and accuracy. The general configuration [12] of the cell is as follows Pt [O 2 ]\HX\Pd [H 2 ]. The platinum electrode is exposed to any ambient gas while the Pd electrode is exposed to hydrogen, the electrodes are short circuited to meas ure the current flowing thru the cell which determines the hydrogen concentration. A solid electrolyte [13] is preferred over a liquid electrolyte for spillage reasons. Advantage of EC sensors: Response time is very high Low power consumption Simpler Design Disadvantage of EC sensor: The target signal is weak and needs some kind of amplifier Electrolyte is susceptible to degradation by environment Suitability of the sensor: For high temperature applications 2.4 Resistive Hydrogen Sensors Resistive sensors encompass multiple sensor subtypes that depend on the change in conductivity or the inverse resistivity. Some of these subtypes are defined below. 2.4.1 Chemiresistors Chemiresistors are based on the principle of change in resistance of the sensing layer upon absorption of gas. It consists [14] of a pair of contacts deposited on an insulating substrate and a selective laye r whose conductivity is modulated by the

PAGE 21

interaction with the analyte. Potential diffe rence is measured across the electrodes when a constant current is applied th rough the circuit. The measur ed voltage is indicative of the response due to adsorption of the gas. Figure 2.3 General Diagram of a Chemiresistor [14] 2.4.2 Conducting Polymers These are the most common material used as a chemiresistor. The effect of interaction of gas species w ith the polymer is considered equivalent to doping of the polymer irrespective of the type of polymer (n-polymer or p-polymer). Doping the polymer varies the surface conductivity whic h is measured as a response. Hydrogen sensors have already been developed using polyaniline based chemistry [15]. 2.4.3 Impedance Sensor They are based on the change in electrica l properties of the thin films due to absorption, adsorption, desorption and molecula r rearrangement due to the reaction of gases on their surface or bulk. These sensors are targeted to be fabricated as high surface to volume ratio structures as the response beco mes stronger suited for an ideal gas sensor. 12

PAGE 22

13 Hydrogen sensors based on palladium nanowire arrays [16], titania nanotubes [17] and porous silicon structures [18] have been reported earlier. 2.4.4 Microhotplate This is a conductometric gas sensor with in-situ temperature control for studying various properties during gas adsorption. Temperature depe ndent processing, and related interfacial phenomenon at microscale can be studied using these thermally isolated, individually addressable, mi crohotplate structures [19]. Advantage: Ease of fabrication Inexpensive Accurate Fast (better response) Disadvantage: Performance affected due to spurious signals Very sensitive to the surrounding environment Possible degradation of sensing layers over time Suitability: For applications requiring ve ry accurate, linear, isolated sensing with short span of sensor life.

PAGE 23

14 CHAPTER 3 POROUS SILICON 3.1 Introduction Porous silicon (PS) can be dated back to 1956 in a study associated with the electrolytic shaping of germanium. Uhlir noticed discolorati on on anodized silicon and believed them to be a suboxide of silicon [20]. The first conclusive reports on the electrochemically etched, porous nature of silic on electrodes can be credited to Watanabe and Sakai in 1971 [21]. First model of pore formation based on a breakdown of the depletion layer was proposed in 1972 by Th eunissen [22]. In 1977 Arita and Sunohara proved the porous silicon layer had the same cr ystal orientation as the bulk silicon, this allowed them to conclude that localized di ssolution generates pores [23]. In 1983 it was discovered pore diameters may be 2nm small [2 4]. Photoluminescence (PL) from porous silicon was discovered in 1984 by Pickering et al. [25]. It wasn t until the ear ly 90s that the chemical sensing properties of porous silicon gained momentum. Porous silicon is broken down into three classifications as outlined by IUPAC. The classification is determined by the pore di ameter (d) defined as the distance between two opposing walls also referred to as inside diameter (ID). Class 1) microporous: d < 2nm Class 2) mesoporous: 2nm < d < 50nm Class 3) macroporous: d > 50nm

PAGE 24

This research investigation utilizes me soporous silicon in the production of the hydrogen sensor due to relative ease in formation and high porosity profile. 3.2 Porous Silicon Theory Porous silicon is formed by the electrochemical dissolution of silicon in an HF electrolyte. The anodic dissolution is de pendent upon the composition and pH of the electrolyte. Pore formation is largely dependent on silicon s crystallographic orientation (e.g. <100>, <110>, <111>), doping level (resis tivity), type of dopant and the current density in relation to the critical current density (J ps ). It should be noted, only the <100> orientation will produce pore formation nor mal to the substrate and was the only orientation used by the author. J ps is the critical current density above which pore formation stops and electro polishing of the substrate begins. The I-V characteristics governing pore formation are illustrated in figure 3.1. Conditions in which pores form are highlighted in the shaded area. Figure 3.1 I-V Characteristics Governing El ectrochemical Dissolution of Silicon [26] 15

PAGE 25

Mesopores are initiated by form ing etch pits in silicon w ith the help of standard anisotropic silicon wet etching mechanism [26]. Mesoporous silicon formation also occurs under certain anodic conditions in both n type and p type substrates in presence of the aqueous electrolyte HF [ 26]. The pore formation is at tributed to the tetravalent dissolution of silicon characterized by an anod ically formed oxide step (one and two of figure 3.2) and then by the chemical dissol ution of the formed oxide in aqueous HF solution (steps three and four of the figur e 3.2 [26]). By increas ing the doping level of the substrate the depletion re gion decreases which enables ch arge carriers to pass the space charge region (SCR) via tunneling. Tunneling is the dominant mechanism for charge transfer with dopi ng densities in excess of 1018 cm Under anodic bias at constant current density J 3 ps silicon dissolution reaches steady state condition between charge transfer and mass transport. It can be seen from figure 3.3 that the pore density is highly dependant on doping de nsity and type of doping. Figure 3.2 Anodic Dissolution of Silicon in HF [26] The electrical transport is mainly affected by the disordered structure of the Si skeleton which restricts the conductive paths to a highly constrained ge ometry. Electrical resistivity in PSi is five orders of magnitude higher than in intrinsic Si.[27] 16

PAGE 26

Figure 3.3 Doping Density Vs Pore Density [26] 3.3 Porous Silicon Sensors As mentioned in the introduction the use of porous silicon for gas sensing began to build momentum in the early 90s. Porous silicon has been used to detect an array of target gasses such as nitrogen dioxide, carbon monoxide, methane, vapor and hydrogen [27-30]. There are many different sensing mechanisms when it comes to the utilization of porous silicon. Some mechanisms are discusse d below. For the sake of this thesis we will concentrate on porous silicon hydrogen sensors for comparative study. 3.3.1 Contact Potential Difference In a study done by V. Polishchuk et al porous silicon was modified by a 20 nm deposition of palladium. The effect of hydr ogen was monitored using contact potential difference variations (CPD). Their measuremen ts were performed at room temperature. 17

PAGE 27

Figure 3.4 CPD Response to H ydrogen in Minutes [30] As can be seen from figure 3.4 the CPD me thod is a tool for hydrogen detection. Their sensors are characterized by se nsitivity in the range 200 ppm H 2 concentrations, the response is in the minut es range and the recovery takes tens of minutes when operated at room temperature.[30] 18

PAGE 28

19 CHAPTER 4 PALLADIUM 4.1 Introduction In this research, palladium is diffused in to porous silicon in order to increase adhesion between porous silicon and a thin film of palladium. The thin film of palladium is the active sensing element for hydrogen dete ction. The response of palladium due to the absorption of hydrogen can be well understood if the physics and chemistry behind their interaction are und erstood. This chapter concentrat es on diffusion prin ciples as well as the effect of hydrogen absorption on the atomic structure of palladium. 4.2 Diffusion Diffusion occurs as a natural result of ra ndom motion of particles from a region of high concentration to low concentration. Di ffusion is a standard method through which impurities such as boron and phosphorous have b een introduced into silicon to control the majority carrier concentration and in turn resistivity. Palladium is considered to be a deep impurity and also has the capability of diffusing into silic on at high enough temperatures. 4.2.1 Diffusion Theory The diffusion process follows Fi cks fist law of diffusion, J = -D N / x (3)

PAGE 29

where J is the particle flux of the impurity, N is the concentration and D is the diffusion coefficient. The diffusion coefficient ha s an Arrhenius dependence on temperature. Where, D = D exp (E / kT) (4) 0 A in this case D is a constant depending of the material and the dopant and E is the activation energy. 0 A Diffusion of a material into a crystallin e entity occurs in the form of atomic substitution or intersti tial phenomenon, or both. Substitutio n is described as atoms of the diffused material replacing atoms of the substr ate lattice (figure 4.1a) whereas interstitial is a result of atoms pertaining to the diffused material residing between the atomic lattice (figure 4.1b). Figure 4.1 (a) Substitution Atom Taking up a Vacancy (b) Interstitial Atom Sitting Between Silicon Lattice. 4.3 Palladium Absorption Principles There exist two phenomena in literat ure explaining gas absorption on metal surfaces, namely physisorption and c hemisorption. While the occurrence of physisorption is due to the attraction of ga s species onto a metal surface by Van der 20

PAGE 30

21 Waals force, the latter is due to exchange of electrons with metal surface creating high density energy states. The unde rlying requirement for both mechanisms to exist is the presence of unfilled (unsatisfied valence or co-ordination number) d-orbital of the active material. Materials that satisfy the above unfilled orbitals incl ude some metal species and semiconductor metal oxides. Nobl e metals like Pt, Pd, Rh and Ir have partially [9] filled d-orbital while the semiconductor metal oxide s possess lattice defects due to imbalance of un-reacted oxygen. The palladium-hydrogen interaction involves an exchange of surface electrons to satisfy the partially filled d-orbital ther eby inducing a change in the electrical conductivity of the system. 4.4 Palladium-Hydrogen System The closed shell atomic structure of palladium is given by 4d 10 5s 0 [10] while the open shell configuration is written as 4d 9 5s 1 The open shell confi guration of palladium permits 0.6 unoccupied states per atom. As soon as hydrogen comes in contact with the palladium surface, it gets completely ionized to satisfy the d-orbital. In other words, hydrogen absorption reaches completion on x [H\Pd] = 0.6, where x denotes the atomic percentage of hydrogen to palladium. Figures 4.1, and 4.2 [11] show the arrangement of hydrogen atoms on palladium at their tetrahedral and octahedral interstitials respectively. (T refers to tetrahedral site and O the octahedral sites)

PAGE 31

Figure 4.2 Hydrogen Atoms Occ upying the Tetrahedral Inters titial Site at x=1 [11] Figure 4.3 Hydrogen Atoms Occupying the Octahedral Sites [11] The existence of two types of pallad ium hydrides with varied morphology and different texture is still in debate. The fo rmation of the hydrides depends on the pressure, temperature and atomic ratio (PCT) of hydrogen to palladium. Experimental observations based on simulati ons [11] and enthalpy calcul ations have revealed the occupation of hydrogen in the interstitial sites of palladium leading to the PdH structures [10] as shown in figure 4.3. Hydrogen in the form of solid solution, embedded in 22

PAGE 32

palladium, influences the resistance of the pare nt atom by disturbing its lattice periodicity. Moreover, the scattering of the hydrogen atoms with palladium is also considered to be the most viable reason for the increase in resistance. Figure 4.4 Different Versions of Palladium-Hydride [10] Figure 4.5 PCT Curve of Pd-H System [12] The resistance critically depends on the phase transformation of the palladiumhydrogen system with phase identificati on through the PCT curve in figure 4.4. Resistance variation is more of an at omically induced surface phenomenon which depends on morphology and reaction kinetics of the palladium hydride formation. Materials/Systems possessing large surface to volume ratio or better yet a two 23

PAGE 33

24 dimensional film with almost no thickness would be an ideal c hoice for gas sensor application. This condition would allow for only surface eff ects which would neglect the scattering caused by interstitial hydrogen. Th e added electrons in this case would decrease the overall resistance of the palladium film. In this context porous silicon is used as a template and will be discussed.

PAGE 34

25 CHAPTER 5 HYDROGEN SENSOR FABRICATION 5.1 Introduction This chapter will guide the reader thr ough a detailed description of the process sequence adopted to develop a palladium ba sed hydrogen sensor. The process steps are well characterized and compa tible with standard MEMS and IC processing techniques making them ideal choice for further integra tion. The overall process (from start to finish) was designed to be simple by nature meaning there are no intricate or complex steps involved in fabrication. The entire fabrication process can be completed without lithography allowing high throughp ut and if applicable mass production with minimal equipment. The entire fabrication module can be broken into three major categories:Thin film deposition: This step is used multiple times throughout the sensor fabrication. All depositions were done us ing electron beam evaporation due to its high vacuum capabilities and tunable deposition parameters. Substrate formation: In this case, porous silicon is the substrate of choice and was formed from a single crystalline silic on wafer via electrochemical etching. Diffusion: this process allows impregnati on of palladium into the porous silicon substrate. The process flow (figure 5.1) is presented with detailed description below.

PAGE 35

5.2 Complete Sensor Process Figure 5.1 Process Flow for H ydrogen Sensor Fabrication 26

PAGE 36

Step A) Choice of initial substrate: A 2 Double side polished degenerately doped (p++) <100> silicon wafer with approximate thickness of 230-260 m and a resistivity of 0.001-0.003 ohms-cm was chosen to be the substrate. These wafers were subjected to RCA clean prior to processing. Step B) Evaporation of ohmic layer: Th e silicon substrate was placed in an AJA International Electron Beam Evaporator (figure 5.2) with planetary rotation (figure 5.3). A base pressure of 10 Torr was attained before aluminum was evaporated on one side of the silicon substr ate for an ohmic contact. (This layer is used to help distribute charge evenly throughout the substrate). Thickness was monitored using a Maxtek 350 quartz crysta l monitor (QCM) also seen in figure 5.2. The thickness was observed to be about 1000 6 Figure 5.2 AJA International El ectron Beam Evaporator 27

PAGE 37

Figure 5.3 Planetary Rotation Step C) Electrochemical etching: Th e substrate was placed in a specially fabricated Teflon jig (figure 5.4a and 5.4b). In this case the substrate was placed aluminum side facing the base of the jig and the silicon side facing the chamber of the jig. Between the aluminum side of the substrate and the base of the jig aluminum foil was placed for anodi c contact to an alli gator clip. An Oring was accommodated between the silicon substrate and the chamber to prevent leaks. The electrolyte solution comp rising of a composite mixture of (24%)HF/(50%)ethanol/(26%)H TM 2 O by volume was added to the chamber. The lid of the jig also made of Teflon had a platinum wire mesh on it to act as a cathode. The cathode is subm ersed in the electrolyte and a hole in the lid allows for the release of hydrogen that evolved during porous silicon et ching. The setup was connected to a Keithley I-V source for a constant current of 200 mA which equated to a current density of 13 mA cm after accounting for the area of the OTM 2 28

PAGE 38

ring. The source required approximately 1.1 volts to maintain this constant current. These conditions were maintained for 60 minutes. (a) (b) Figure 5.4 (a) Rendering (b) Photo of Etching Jig for Pore Formation Step D) Cleaning the Sample: Once the porous silicon etching was complete the HF solution was safely removed and the sample was rinsed in deionized (DI) water. The wafer was then placed alumin um side down on top of an open wide mouth HF bottle (figure 5.5). This is done in order to remove the aluminum from the backside of the wafer using the HF vapor which forms inside the bottle. Aluminum is highly reactive with HF and only takes about a minute for the aluminum to be removed by the vapor at room temperature. The sample is rinsed with DI water and dried using nitrogen gas. Step E) Dicing: The porous silicon subs trate was placed in Kulicke Soffa Dicing saw (figures 5.6) and diced into 1 cm square samples. 29

PAGE 39

Figure 5.5 Aluminum Etching Using HF Vapor at Room Temperature Figure 5.6 Dicing Saw Stage 30

PAGE 40

31 Step F) Evaporation of Diffusion layer: The porous silicon samples were placed in the electron beam evaporator and processed under the same conditions as step B above. This time using palladium instead of aluminum. The deposition rate was about 1A/s. The final palladium thic kness was maintained to be 4nm on the porous side of the samples. Step G) Diffusion: The samples were pl aced in an open tube furnace (figure 5.7). Nitrogen was passed at a rate of 30 cm 3 for 20 minutes before the furnace was turned on. The furnace was set at 900 0 C and allowed to ramp up over time. Once the desired temperature was reached the furnace was allowed to maintain 900 0 C for the duration of 60 minutes. The furnace was then turned off and allowed to cool over time with the samples inside. NOTE: Porous silicon can be very brittle a nd any stress may cause severe damage to the sample. The samples were allowed to ra mp up and ramp down with the furnace to minimize rapid heating and cooling which could be detrimental to the samples structure.

PAGE 41

Figure 5.7 Open Tube Furnace Step H) Evaporation of Sensing Layer: Again the samples were placed inside the AJA electron beam evaporator under the same vacuum conditions set before in step B. Again 4nm of palladium was evaporated at a rate of 1 A/s on the porous/diffusion side of the samples. NOTE: The reasoning behind th e diffusion layer is to incr ease adhesion with this Pd deposition while increasing the pr obability for an ohmic contact. 32

PAGE 42

33 CHAPTER 6 TEST SETUP AND PROTOCOLS 6.1 Introduction A mass manufacturable impedance based palladium doped porous silicon sensor was fabricated for hydrogen detection. In or der to quantify the hydrogen sensor, a testing apparatus had to be assembled and a testing protocol established. The testing apparatus was built in house using standard software and hardware components as well as custom fabricated hardware to facilitate special needs. 6.2 Hardware In order to isolate the hydrogen sensor from unwanted interactions such as fluctuating ambient conditions a housing chamber (test bed) had to be custom fabricated (figure 6.1). The test bed contains sealed el ectrical contacts which interfaced the sample to be tested. In order to achieve simple tr ansfer of samples (removable, replaceable and reusable without damage), the chamber was bu ilt with a removable lid. Machined holes penetrated the removable lid while spring loaded pins (contacts) prec isely fit the holes. This resulted in the formation of a feed through array seen in figure 6.1. The pins allowed for a tight seal while the spring load ed force maintained contact and permitted variation in sample height. The passage for gas flow was accomplished by fitting the chamber with a gas feed through for both the inlet and outlet. The gas input and output orientation was structured for a unidirectional gas flow. Keeping in mind the explosive nature of hydrogen, the

PAGE 43

hydrogen test gas would have to be vented in a safe manner. Therefore the exhaust of the chamber was led to a pre approved house vent for safety. Figure 6.1 Test Bed In order to control the per centage of hydrogen in the sensor environment an inert carrier gas was passed until steady state conditi ons were met. Then hydrogen was bled into the chamber to achieve variable percentage of hydrogen in nitrogen. In order to precisely monitor the gas mixtures tw o MKS 1176 mass flow controllers were implemented (figure 6.2). The nitrogen flow controller was built for ranges between 0 and 2000 sccm while the hydrogen flow controll er was configured for ranges between 0 and 20 sccm. The rate of error associated with each controller is a bout 0.1 percent of full scale. The mass flow controllers were orient ed upstream with respect to the chamber and controlled by the MKS 146C multicontroll er shown in the bottom of figure 6.3. 34

PAGE 44

Figure 6.2 Mass Flow Cont rollers and Test Bed Figure 6.3 Keithley Multimeter (t op), MKS Controller (bottom) 35

PAGE 45

The hydrogen sensor is an impedance based sensor where the parameter of interest is either re sistance or resistivity. To make measurements the Keithley 2400 I V (figure 6.3) source meter was used in a four wire probe configuration (figure 6.4). This configuration was used in order to find true resistance between two points on the sensor and to filter the impedance eff ects of the probes and wires. Figure 6.4 Four Wire Probe Configuration [31] Once the overall hardware set up was co mplete, desktop computer was used to transfer and collect data. Figure 6.5 illustrates fina l hardware configuration. 36

PAGE 46

Figure 6.5 Hardware Test Setup 6.3 Software It is well known that most errors accu mulated during experimental analysis are associated with human error. In order to help eliminate error and to retrieve data at high rates of speed; it is advant ageous to employ software. In the experimental setup Labview software, a graphical user interface (GUI) was utilized on a Windows based operating sy stem. Labview has the ability to be programmed by the user in order to implem ent predetermined commands as well as interact with real world devices In this case Labview was used to send or retrieve data from our Keithley source meter and MKS cont rollers. The idea behi nd our program was to have one screen in which al l user determined values coul d be entered, including on off switching (figure 6.6). 37

PAGE 47

Figure 6.6 Testing Module The following is the program outline: Set up communication protocols and por ts from the PC to the MKS 146C controller and the Keithley 2400 source meter. Excep t for the communication port the rest of the communication protoc ols are embedded in the program and is accomplished without input from the user. Set up initial testing conditions for th e MKS 146C. The 146C requires the range in sccm of each mass flow controller it is connected to as we ll as the associated channel number (1 of 4 channels are possible). The range of the mass flow controllers can be entered in the main us er page, for example Range 1 in figure 6.6 is the range of the mass flow controller plugged into the fi rst channel of the MKS 146C controller. 38

PAGE 48

39 Set up initial testing conditions for the Keithley. The source meter requires information on how testing will proceed, for example whether it run in four wire or two wire mode. These testing parameters are embedded in the program and are not accessible to the user. Determine path and file name of saved data. When the program is initialized the data procured will be saved in Microsoft notepad. In order for this to occur the user has to input the path of the folder in which the in formation will be saved as well as the file name. Once the program is started and data is being captured the user can change the rate of flow for any of the four gas contro ller channels at any time. If the user has a predetermined percent value of one gas in to another the user only has to set the flow rate of the first gas and the percen t value of the second gas. For example 100 sccm is the total desired flow rate. Enter 100 under set point one. This will flow 100 sccm of gas one. If the user wanted to run 10 percent of gas two into gas one then enter the value 10 under the % of channel 2 string in figure 6.6. Gas one will automatically be readjusted to 90 sccm and gas two will be adjusted to 10 sccm maintaining a total 100 sccm. This was done to maintain constant pressure within our test bed which will insure results of testing is purely a function of gas concentration and not nominal pressure changes should the se nsor be pressure sensitive. The main body of the LabView code is imaged below for viewing (figure 6.7). The top box in the code is responsible for m onitoring the flow rate from the MKS and the

PAGE 49

resistance from the Keithley. The top box also saves the data in Microsoft notepad. The bottom box in the code is responsible for th e flow rate instructions to the MKS Figure 6.7 Main Body of Code 40

PAGE 50

41 CHAPTER 7 RESULTS AND DISCUSSION 7.1 Introduction The objective of this thesis is to introduce the fabrication of a novel nanostructured palladium f unctionalized hydrogen sensor. This chapter includes a discussion on physical nature (morphology) of the sensor module. This chapter also includes the results obtained from testing the sensor response to different concentrations of hydrogen in a nitrogen carrier. The correlation between morphology and sensing characteristics are discussed. Different st eps described in the process sequence were tested in hydrogen to identify and isolate the background signal. 7.2 Morphology Referring back to step A of figure 5.1 in chapter five the substrate started out as a single crystal silicon substrat e. The morphology of the s ilicon did not change until porous silicon etching (step C). The result of the etching process yielded a nanoporous substrate that becomes the template of the sensor. The pores of the substrate were imaged with a Hitachi scanning elect ron microscope (figure 7.1 and 7.2).

PAGE 51

Figure 7.1 SEM of Porous Silicon Substrate at 200k Magnification Figure 7.2 SEM of Porous Silicon Substrate at 300k Magnification 42

PAGE 52

The images of the pores were later analyzed using National Instruments vision assistant particle analysis tool which is demonstrated in figure 7.3 through figure 7.6. Figure 7.3 shows the importing of an image; figure 7.4 illustrates the image calibration from pixels to real world units. The calibra tion is done by converting the pixels of the image into unit equivalents. In this case the calibration tool utilizes the SEM scale bar as reference. Figure 7.5 shows the selection of an appropriate threshold. The threshold is the pixel grayscale difference between what a pore is (darker valu e) and what a pore sidewall is (lighter value). From this point each pore which is now represented as a particle will be analyzed for center of mass total area and percent area over the entire image. Figure 7.6 reveals the assessment of data using a spreadsheet. Figure 7.3 Importing an Image 43

PAGE 53

Figure 7.4 Calibrate Image Pixels to Real World Units Via SEM Scale Bar Figure 7.5 Determine the Pore by Choosing a Proper Threshold Value 44

PAGE 54

Figure 7.6 Assessing Information on Spreadsheet From the spreadsheet the percent area over the image area fo r each particle was added in order to determine the total porosity to be approximately 49 percent of area. This value seems to concur with observed value reported elsewhere [26]. The average pore diameter is approximately 35 nanometers. The pore diameter fluctuates between a few nanometers to a litt le over fifty nanometers. The next major morphology change occurred during the palladium diffusion step. Once this step was complete more scanning electron micrographs we re taken. Figure 7.7 is a cross sectional micrograph of the porous silicon sample after diffusion. From this image the pore depth can be observed as 34.8 um and the diffusion depth to be contrasted by the dark shaded band at the top of the sample to be approximately 1.49 um. The image in figure 7.8 shows the top surface of the sample after diffusion. This image shows the presence of agglomerations after th e diffusion process. It has been suggested that palladium silicide can form at temperatures as low as 700 degrees C [32]. This may indicate that the agglomeration formation is results from palladium silicide reflow. 45

PAGE 55

However, the charged regions which appear as bright areas in the image suggest some dielectric formation. In order to determin e the surface content of the sensor an XRD (figure 7.9) was performed. The XRD suggest s the formation of PdO and not palladium silicide. Due to the close d-spacing value between PdO and Pd 2 Si, Pd 2 Si can not be fully ruled out as noted by the black mark at 38.265 on the XRD where it would be, the possibility of palladium silicide is less lik ely taking into consideration the thickness of the Pd monolayer and process temperature fo r a palladium disilicide formation. [32]. Figure 7.7 Side View of Sample After Pd Diffusion 46

PAGE 56

Figure 7.8 Surface Image of Sample After Diffusion Figure 7.9 XRD Analysis of Sensor 47

PAGE 57

48 7.3 Hydrogen Test Response In this section, the impedance response of the sensor is plotte d against the varying hydrogen concentration. All measurements we re taken at room temperature. Before results of the working sample are presented, short synopses of test ed intermediates are given below. Plain <100> silicon had no visibl e impedance reaction to hydrogen Porous silicon had no visible impedance reaction to hydrogen 4 nm of palladium deposited on porous silicon had no visible impedance reaction to hydrogen 4 nm of palladium diffused porous silicon had no visible impedance reaction to hydrogen This work began with the successful test ing of two initial samples; extension of which lead to the development of batch fabr icated samples for furt her exploration. The following results are gathered from the first two samples. Figures 7.10 and 7.11 are graphs illustrating the range of sensor A (lef t) and inverted hydrogen concentration (from high concentration to low con centration) in nitrogen (right) on the Y-axis Vs time in seconds on the X-axis during conditioning. The scale bar pertaining to hydrogen concentration is inverted in or der to quickly correlate the ch ange in sensor response to percent change in hydrogen concentration from the graph. Figure 7.12 is a graph illustrating range of sensor and inverted hydr ogen concentration Vs time in seconds after conditioning. From all three graphs an inverse relationship of range of sensor (impedance) Vs hydrogen concentration is es tablished. Also note the response of the knee of both curves appear to be well aligne d indicating the credibil ity of the measured

PAGE 58

data. A closer analysis of figure 7.12 yields a two second delay between the knee of the gas change and the knee of the sensor change in both the adsorption a nd desorption phase. This two second delay is the reaction time. Figure 7.10 Initial Sensor Reaction During Conditioning Figure 7.11 Sensor Reaction Ending Conditioning 49

PAGE 59

Figure 7.12 Sensor Reaction After Conditioning A few months after the successful imple mentation of a real time hydrogen sensor, additional sub units were fabric ated using the same process parameters. The second set of sensors however displayed different reaction characteristics than the initial set. The impedance (Range of Sensor) for the seco nd set of samples displayed a direct proportionality to change in hydrogen con centration unlike the ear lier batch. Figure 7.13 illustrates this relationship. Even though the re action is opposite from that of the first sensor set, the sensor seems to share some common features observed earlier. The most notable feature is the two second delay between the change in gas c oncentrations and the change in the sensor impeda nce (reaction time) for both adso rption and desorption cycles. Also noteworthy is the consistent shape of the curve. At this juncture, more samples were made under the similar conditions with no minor variations a nd the sensor response was found to exhibit a direc tly proportion relationship. 50

PAGE 60

Figure 7.13 Resistances Displaying Pr oportional Relationship to Hydrogen 7.4 Conclusion The issue of system impedance behaving both directly and inversely proportional to hydrogen concentration is yet to be resolv ed. It is postulated that the discrepancy might be due to pore reconfiguration during diffusion, or an inconsistent quantity of native oxide formation in the pores. After et ching the pores the oxide could form PdO or SiO during diffusion. A differential signal obtained from sensors made from identical process condition to negate false positives. Pd/Porous Si sensor for hydrogen detecti on has been made and tested in the 0-2 percent range. It has been observed that sensors made with porous Si and Pd nanoparticles demonstrate a significant change in impedance with respect to time when exposed to hydrogen. It is hypothesized that diffused and evaporated Pd on porous Si forms nanoparticles on the surface as well as in the pores. These nanoparticles decrease 51

PAGE 61

52 the adsorption and desorption times, which increases the sensi tivity, sensing, and regeneration times of the sensor. Due to sensitivity, cost effectiveness a nd ease of fabrication, Pd-plated porous Si has the potential of becoming a trul y universal hydrogen sensing device. 7.5 Future Work Future fabrication would be benefited if diffusion is performed under more controlled medium such as vacuum. The use of tungsten or titanium oxygen getters could also help in controlling unwanted oxide fo rmation in the sensor samples. In this work the issue of cross sensitivity or selectivity was not discussed. This is a critical issue to be addressed in any sensor fabrication a nd should be pursued as a part of a continued investigation. Palladium poisoning is a we ll known phenomenon and has limited the life cycle of this sensor to less than five months. Further work should also include the implementation of a masking layer such as Nafion to the finished sensor in order to limit known contaminants.

PAGE 62

53 REFERENCES 1. Weast, R. C., Ed. Handbook of Chemistry and Physics 56th ed.; CRC Press, Inc.: Boca Raton, FL, 1976. 2. NASA requirements document, 4ISE531, private communication. 3. Gregory T.A.Kovacs, Micromachined Transducers Sourcebook Mc-Graw Hill, 1998. 4. Osburn, C.M, Microsensor Engineering Designing Low Power Digital Systems, Emerging Technologies (1996), 1996, pp: 233-328 5. Favier F., Walter E., Zach M., Benter T., Penner R., Science vol. 293, pp. 2227-2231, September 2001. 6. Zach M., Ng K., Penner R., Science vol. 290, pp.2120-2123, December 2000. 7. Chandrasekharan S., Brazzle J. and Frazier A., Journal of Microelectromechanical systems, June 2003, vol. 12, 3, pp 281-288. 8. http://www.sandia.gov/mstc/technologies/mi crosensors/hydrogensensor.html. 9. Andreas Mandelis, Constantinos Christofides, Physics, chemistry, and technology of solid state gas sensor devices, Wiley-Interscience publication, 1993. 10. D.Dwivedi, R. Dwivedi and S. K. Srivastava, Sensors and Actuators B: Chemical Volume 71, Issue 3, 1 December 2000, pp: 161. 11. X Bevenot, A.Trouillet, C Veillas, H Gagnaire and MCement, Measurement Science and Technology Vol.13, 2002, pp: 118-124. 12. Leo B. Kriksunov and Digby D. Macdonald, Sensors and Actuators B: Chemical Volume 32, Issue 1, April 1996, Pages 57-60. 13. C. Ramesh ,G. Velayutham,N. Murugesan ,V. Ganesan ,K.S. Dhathathreyan,G. Periaswami, J. Solid State Electrochem (2003) 7: 511. 14. Jiri Janata, Andras Bezegh; Anal. Chem, 1988 ; 60 (12); 62R-74R. 15. Costa Conn, Stephen Sestak Anthony T. Baker, and Joe Unsworth, Journal of Electroanalysis, 1988, pp:1137-1141. 16. Atashbar, M.Z.; Banerji, D.; Singamaneni, S.; Intelligent Sensing and Information Processing, 2004, pp:258-261.

PAGE 63

54 17. Oomman K. Varghese, Dawei Gong, Maggie Paul ose, Keat G. Ong and Craig A. Grimes, Sensors and Actuators B:Chemical Volume 93, Issues 1-3, 1 August 2003, Pages 338344. 18. V. Polishchuk, E. Souteyrand, J. R. Martin, V. I. Strikha and V. A. Skryshevsky, Analytica Chimica Acta Volume 375, Issue 3, 30 November 1998, Pages 205-210 19. http://www.eere.energy.gov/hydrogenandfuelcells/pdfs/28890ll.pdf 20. A. Uhlir, Bell system tech. J., 1956, 35, 333-347. 21. Y.Wantanabe, T. Sakai, Rev. Electron. Commun. Labs., 1971, 19, 899. 22. M.J.J. Theunissen, J. Electrochem. Soc., 1972, 119, 351-360. 23. Y. Arita, Y. Sunohara, J. Electrochem. Soc., 1977, 124, 185-295. 24. G. Bomchil, R. Herino, K. Baria, J. C. Pfister, J. Electrochem. Soc., 1983, 130, 16111614. 25. C. Pickering, M. I. J. Beale, D. J. Robbins, P. J. Pearson, R. Greef, J. Phys. C: Solid State Phys., 1984, 17, 6535-6552. 26. Lehmann V., Electrochemistry of Silicon Wiley-VCH, 2002. 27. Di Francia, G La Ferrara, V Quercia, LFaglia, G., Journal of Porous Materials Vol. 7, no. 1-3, pp. 287-290. Jan. 2000. 28. Israel Schechter, Moshe Ben-Chorin, Andreas Kux, Analytical Chemistry, Vol. 67, No. 20, October 15, 7995 3727. 29. Pancheri L., Oton CJ, Gaburro Z., Soncini G., Pavesi L, Sensors and actuators. B, Chemical vol. 89, 2003: 237-239. 30. Polishchuk V., Souteyrand E., Martin J., Strikha V., Skryshevsky V., A study of hydrogen detection with palladium modifie d porous silicon, Analytica Chimica Acta vol. 375, pp. 205-210, 1998. 31. Keithley Model 2400 Series SourceMeter Quick Results Guide. 32. B.Y. Tsaur, M-A. Nicolet, Appl. Phys. Lett. 37(8), 15 October 1980.


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 001795168
003 fts
005 20070628142907.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 070628s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001542
040
FHM
c FHM
035
(OCoLC)146315119
049
FHMM
090
TK145 (ONLINE)
1 100
Luongo, Kevin.
0 245
Palladium doped nano porous silicon to enhance hydrogen sensing
h [electronic resource] /
by Kevin Luongo.
260
[Tampa, Fla] :
b University of South Florida,
2006.
3 520
ABSTRACT: A mass manufacturable impedance based, palladium doped porous silicon sensor, was fabricated for hydrogen detection. The sensor was built using electrochemical etching to produce mesoporous silicon. Four nanometers of palladium was defused directly into the porous silicon and another four nanometers of Pd was deposited on the defused surface to enhance sensing. The sensor was tested in a sealed chamber in which the impedance was measured while hydrogen in nitrogen was ranged from 0-2 percent. Unlike conventional hydrogen sensors this sensor responded at room temperature to changes in hydrogen concentration. The electrical impedance response due to adsorption and desorption of hydrogen reacted relatively quickly due to the nanoparticle nature of palladium diffusion in and Pd evaporation on porous silicon.
502
Thesis (M.A.)--University of South Florida, 2006.
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 54 pages.
590
Adviser: Shekhar Bhansali, Ph.D.
653
Room temperature.
Gas sensor.
Nanoparticle.
Labview.
Palladium diffused.
690
Dissertations, Academic
z USF
x Electrical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1542