STRUCTURAL, OPTICAL AND ELECTRICAL CHARACTERIZATION OF INDIUM TIN OXIDE AND ALUMINUM NITRIDE THIN FILMS FOR GAS SENSING APPLICATION

THESIS SUBMITTED TO BHARATHIDASAN UNIVERSITY, TIRUCHIRAPALLI FOR THE AWARD OF THE DEGREE OF

DOCTOR OF PHILOSOPHY IN PHYSICS

by V.VASANTHI PILLAY, M.Sc., M. Phil Ref : 25874/Ph.D.1/P.T./Physics/Oct.2006/Confir./Date:26.05.2009

Under the guidance of Dr. K. VIJAYALAKSHMI, M.Sc., M. Phil., B.Ed., Ph.D.,

PG and Research Department of Physics Bishop Heber college (Autonomous) (Nationally Reaccredited at the A+ level by NAAC) (Recognized by UGC as college with potential for excellence) Tiruchirappalli – 620 017

April - 2012 PG and Research Department of Physics Bishop Heber College (Autonomous) (Nationally Reaccredited at the A+ level by NAAC) (Recognized by UGC as college with potential for excellence) Tiruchirappalli- 620 017

Dr. K. VIJAYALAKSHMI, M.Sc., M.Phil., B.Ed., Ph.D., Office : 0091-431-2770136 Assistant Professor Resi : 0091-431-2770755 Mobile : +919994647287 Email : [email protected]

CERTIFICATE

This is to certify that the thesis entitled “STRUCTURAL, OPTICAL AND

ELECTRICAL CHARACTERIZATION OF INDIUM TIN OXIDE AND ALUMINUM

NITRIDE THIN FILMS FOR GAS SENSING APPLICATION” submitted by V.Vasanthi pillay, (Ref: 25874/Ph.D.1/P.T./Physics/Oct.2006/Confir./Date:26.05.2009) is a bonafide record of research work done by her under my guidance in the Department of Physics,

Bishop Heber College, Tiruchirappalli, and that the thesis has not previously formed the basis for the award to the candidate of any degree, diploma, associateship, fellowship or any other similar title. The thesis is the outcome of personal research work done by the candidate under my overall supervision.

Station : Tiruchirappalli (Dr. K. VIJAYALASKHMI) Date :

DECLARATION

I hereby declare that the thesis entitled “STRUCTURAL, OPTICAL AND

ELECTRICAL CHARACTERIZATION OF INDIUM TIN OXIDE AND

ALUMINUM NITRIDE THIN FILMS FOR GAS SENSING APPLICATION” submitted by me for the degree of Doctor of Philosophy is the record of work carried out by me under the guidance and supervision of Dr. K. Vijayalakshmi, Assistant

Professor, Department of Physics, Bishop Heber College, Tiruchirappalli and had not formed the basis for the award of any degree, diploma, associateship, fellowship or any other similar titles of this or any other University and other similar institution of higher learning.

Place :Tiruchirappalli (V.VASANTHI PILLAY) Date :

PREFACE

The main purpose of this work is to study and develop new materials for gas sensing elements starting from the knowledge in thin film production using magnetron sputtering. The structure of this report is summarized below.

The first chapter gives a brief introduction to the need and importance of gas sensor in the contest of current status, its implication at many different levels in our society, necessity of hydrogen and nitrogen gas sensor etc. The general approach to gas sensor, the classification of gas sensor and the characteristics of gas sensor have been discussed in this chapter. The second chapter describes the material used for gas detection. The properties of ITO and

AlN are discussed in detail in this chapter.

The third chapter explains dc/rf sputtering process and sputtering chamber used for thin film preparation in detail. The various instrumentation techniques used to characterize ITO and AlN thin films are also discussed in this chapter. The fourth chapter explains the optimization of ITO thin film, and the structural, optical and electrical characterization of the prepared ITO film are discussed in this chapter.

Fifth chapter explains the optimization of AlN thin films, and the structural, optical and electrical properties are discussed in this chapter. Sixth chapter presents the detailed fabrication of gas sensing chamber used to sense various gases. The gas sensing mechanism and the measurement of sensitivity are also explained in this chapter.

Seventh chapter describes the fabrication of ITO and AlN gas sensing devices, and response of the devices for various gases. Finally, summary and main conclusion of the research work and some directions for future research that may be carried out are discussed.

ACKNOWLEDGEMENT

First and Foremost, I thank GOD Almighty for having bestowed me to complete this research work.

I am thankful to the Secretary, Bishop Heber College, Tiruchirappalli, and the

Principal Dr. D. Paul Dhayabaran, for having granted me permission to do Ph.D. programme in Bishop Heber College, Tiruchirappalli.

I am thankful to the Secretary, Sujatha Degree and PG College for Women,

Hyderabad, and the Principal Dr. Aruna Singh, for their support in the completion of the thesis.

I wish to place on record my heartful thanks to Dr. K. Vijayalakshmi, Assistant

Professor, Department of Physics, Bishop Heber College, Tiruchirappalli, for reposing full confidence in my ability to work and write on the subject, for her unceasing and steadfast encouragement and guidance. I wish to gratefully thank her for the care and affection, without which this thesis would not have seen the light of the day.

I owe my sincere thanks to Dr. C. Ravidhas, Head, Department of Physics, Bishop

Heber College, Tiruchirappalli, for providing laboratory facilities to carry out this research work and for his valuable suggestions and support throughout the course of the programme.

My profound thanks to Member of Doctoral Committee, Dr. K.Ramamurthi,

Professor, School of Physics, Bharathidasan University, Tiruchirappalli, for his periodic evaluation and candid appraisal of the research work. I wish to express my gratitude to Dr. S. Jayakumar, Head, Department of Physics,

PSG College of Technology, Coimbatore, for providing equipments at the thin film laboratory, to carry out this research work.

I earnestly thank Dr. A.Chandrabose, Head, Department of Physics, NIT,

Tiruchirappalli, for providing laboratory facilities to carry out this research work.

Special thanks are extended to Dr. S. Sanjiv Kumar, Scientist, National center for characterization and composition of materials, Hyderabad, for his help in characterization of the samples.

I must thank profusely Dr. S. Prasanna, Assistant Professor and E. Nitya,

Associate Professor, PSG College of Technology, Coimbatore, for all their help, support, interest and valuable hints.

I would like to thank all the Faculty Members of the Department of Physics,

Bishop Heber College, Tiruchirappalli, for their encouragement and support during the course of my research work.

I am deeply indebted to my parents and my husband for their moral support to complete my research.

V. VASANTHIPILLAY

LIST OF TABLES

Table 1.1 Examples of applications of gas sensor.

Table1.2 Types of solid state gas sensors with the corresponding physical change used as gas detection principle.

Table 2.1 Mechanical, Thermal and Electrical properties of AlN.

Table 4.1 Summary of deposition parameters for ITO thin films.

Table 4.2 Deposition rate of ITO thin films for different rf power

Table 4.3 Structural parameters of ITO thin films deposited at different rf power Table 4.4 Electrical parameters of ITO thin films deposited at different rf power Table 4.5 Structural parameters of ITO thin films deposited at an rf power of 150 W Table 4.6 Electrical parameters of ITO thin films deposited at an rf power of 150 W Table 5.1 Summary of deposition parameters for AlN thin films Table 5.2 Deposition rate of AlN thin films for different nitrogen concentration

Table 5.3 Deposition rate of AlN thin films for different deposition time

Table 5.4 Structural parameters of AlN thin films deposited on Si(100) for different deposition time

LIST OF FIGURES

Figure 1.1 Signal processing in living organisims and in intelligent machines. Figure 1.2 Lennard-Jones model of physisorption and chemisorptions. (a) Physisorption of a molecule (b) Chemisorption of a

molecule. Activation energy Ea, dissociation energy Ediss,

desorption energy Edes Figure 1.3 Brunauer‘s five types of adsorption isotherms. Figure 1.4 Characteristics of several Langmuir isotherms for different b (T). Figure 2.1 Crystal Structure of Indium tin oxide. Figure 2.2 Two indium sites of ITO. Figure 2.3 Carrier concentrations as function of the tin doping level. Figure 2.4 Energy band diagrams of In2O3 and ITO. Figure 2.5 Crystal structure of the Aluminum Nitride. Figure 2.6 Geometry of crystallographic planes (002), (102) and (101) in Hexagonal AlN lattices. Figure 3.1 Schematic diagram of rotary pump. Figure 3.2 Schematic diagram of a diffusion pump Figure 3.3 Schematic diagram of penning gauge. Figure 3.4 Schematic diagram of vacuum coating unit. Figure 3.5 Photograph of the dc/rf magnetron sputtering unit. Figure 3.6 Photograph of the Rutherford backscattering spectrometer Figure 3.7 The geometry of the incident X-rays impinging the sample satisfies the Bragg Equation. Figure 3.8 Schematic of the detection of diffracted X-rays by diffractometer Figure 3.9 Photograph of the X-ray diffraction unit Figure 3.10 Schematic diagram of Scanning electron microscope

Figure 3.11 Photograph of the Scanning electron microscope

Figure 3.12 Schematic diagram of FTIR spectrometer system Figure 3.13 Photograph of FTIR spectrometer Figure 3.14 Schematic diagram of UV-VIS spectrophotometer Figure 3.15 Photograph of UV-VIS spectrophotometer Figure 3.16 Photograph of Precision LCR Meter Figure 3.17 Schematic diagram of LCR measurement set up (a) with specimen (b) without specimen Figure 4.1 The dependence of deposition rate of ITO thin films on rf power

Figure 4.2 XRD pattern of ITO thin films prepared at different rf power: (a) 50 W (b) 100 W (c) 150 W (d) 200 W

Figure 4.3 Surface morphology of ITO films deposited on glass substrate at different rf power : (a) 50 W (b) 100 W (c) 150 W (d) 200 W

Figure 4.4 Variation of sheet resistance of ITO thin films deposited at different rf power

Figure 4.5 Variation of resistivity of ITO thin films deposited at different rf power

Figure 4.6 Variation of carrier concentration of ITO thin films deposited at different rf power

Figure 4.7 Variation of mobility of ITO thin films deposited at different rf power Figure 4.8 Absorbance spectra of ITO films deposited on glass substrate for different rf power : (a) 50 W (b) 100 W (c) 150 W (d) 200 W

Figure 4.9 Transmittance spectra of ITO films deposited on glass substrate for different rf power : (a) 50 W (b) 100 W (c) 150 W (d) 200W

Figure 4.10 XRD pattern of ITO thin films prepared at an rf power of 150W (a) before annealing and (b) after annealing

Figure 4.11 SEM pictures of the ITO thin films deposited on glass substrate at an rf power of 150W (a) before annealing and (b) after annealing

Figure 4.12 Transmission spectra of ITO thin films deposited at an rf power of 150 W (a) before annealing and (b) after annealing

Figure 4.13 Band gap of ITO thin films deposited at an rf power of 150W (a) before annealing and (b) after annealing

Figure.5.1 The principle mass sensitive depth microscopy in the case of a thin film B on a substrate A before and after reaction

Figure 5.2 The RBS spectra of AlN thin films for different nitrogen concentration

Figure 5.3. XRD pattern of AlN thin films prepared on glass substrate at different nitrogen concentration

Figure 5.4 The dependence of deposition rate of AlN thin films on nitrogen concentration

Figure 5.5 XRD pattern of AlN films prepared on Si (100) substrate at (a) 2 min (b) 4 min (c) 6 min (d) 8 min of deposition time

Figure 5.6 The dependence of deposition rate of AlN thin films on deposition time

Figure 5.7 FTIR spectra of AlN thin films prepared on Si (100) substrate for different deposition time

Figure 5.8 Surface morphology of AlN thin films deposited on Si (100) substrate at (a) 2 min (b) 4 min (c) 6 min (d) 8 min of deposition time Figure 5.9 Reflectance spectra of AlN / Si thin films prepared at different deposition time

Figure 5.10 Band gap of AlN / Si thin films deposited at different deposition time

Figure 5.11 The variation of capacitance of AlN thin films deposited on Si (100) substrate as a function of frequency

Figure 5.12 The variation of dielectric constant of AlN films deposited on Si (100) substrate as a function of frequency

Figure 5.13 The variation of dissipation factor of AlN films deposited on Si (100) substrate as a function of frequency

Figure 6.1 Schematic of the gas sensing set-up used for sensitivity measurement Figure.6.2 Diagram of set up used to measure the electrical parameters and heat the samples Figure 6.3 Typical thin film resistor gas sensor Figure 6.4 (a) Schematic drawing of the Si-based capacitor (b) Simplified equivalent circuit of MOS capacitor Figure 6.5 The C-V characteristics of an n-doped semiconductor capacitor with a metal Figure 6.6: Energy-band diagrams for an ideal MIS diode (n-type semiconductor) at (a) accumulation, (b) flat band conditions, (c) depletion and (d) inversion Figure 7.1 Schematic representation of the ITO sensor device Figure 7.2 SEM image of the lay up of ITO film sensor Figure 7.3 Variation in sensitivity of ITO towards NO2 with operating temperature.

Figure 7.4 Variation in sensitivity of ITO device with concentration of NO 2 Figure 7.5 Schematic diagram of band bending after chemisorption of charged species Figure 7.6 Variation of resistance of ITO sensor with operating temperature towards 900 ppm of ethanol

Figure 7.7 Variation of sensitivity ITO sensor with operating temperature towards enthanol (a) before annealing and (b) after annealing

Figure 7.8 Variation of sensitivity ITO sensor with concentration of ethanol (a) before annealing and (b) after annealing

Figure 7.9 Variation of sensitivity ITO sensor towards H2 with operating temperature

Figure 7.10 Variation in sensitivity ITO sensor with concentration of H2

Figure 7.11 Schematic representation of the Pd/AlN/Si device Figure 7.12 Schematic representation of the Al/AlN/Si device Figure 7.13 Schematic illustration of the principle of gas sensitive MOS capacitor Figure 7.14 Variation of capacitance of Pd/Al/AlN device with voltage Figure 7.15 Variation of capacitance of AlAl/AlN device with voltage ABBREVIATIONS AND SYMBOLS

IUPAC - International Union for Pure and Applied Chemistry ppm - Parts per million ppb - Parts per billion ITO - Indium Tin Oxide AIN - Aluminum Nitride PVD - Physical Vapour Deposition CVD - Chemical Vapour Deposition SAW - Surface Acoustic Wave QMB - Quart Crystal Micro Balance TCO - Transparent Conducting Oxide LCD - Liquid Crystal Display MIS - Metal Insulator Semiconductor MOS - Metal Oxide Semiconductor SMO - Semi Conducting Metal Oxide MISFET - Metal Insulator Semiconducting Field Effect Transistor CTE - Coefficient of Thermal Expansion SCCm - Standard Cubic Centimeter MFC - Mass Flow Controller XRD - X-ray Diffraction SEM - Scanning Electron Microscopy FTIR - Fourier Transform Infrared MEMS - Micro electromechanical systems RBS - Rutherford Backscattering Spectrometer ΔT - Heat of temperature ΔE - Energy gain P - Partial pressure q - Loading of absorbate KA - Adsorption constant

Rads - Rate of adsorption qs - Saturation capacity θ - Fractional surface converge K - Boltzmann constant Pa - Pascal - Frequency Å - Armstrong μF - Micro Farad nm - Nanometer. Ω/□ - Ohm / square

Rsh - Sheet resistance - Resistivity - Mobility n - Carrier concentration α - Absorption coefficient - Dielectric constant tan δ - Dissipation factor C - Capacitance

ε0 - Permittivity of vacuum KHz - Kilo Hertz

CONTENTS

Title Page No. Preface Acknowledgement List of Tables List of Figures List of Abbreviations and Symbols Introduction 1

CHPATER-1 GAS SENSOR-STATE OF THE ART 4 1.1 Sensor and Sensor science 5 1.2 The Term Sensor 7 1.3 Applications and Implications of gas sensors 8 1.4 General approach to gas sensors 11 1.4.1 Solid state gas sensors – State of the art 12 1.4.2 Classification of sensors 14 1.5 Chemical gas sensor- State of the art 15 1.5.1. Classification of chemical sensors 16 1.6 Manufacturing processes 21 1.7 Mechanism of gas detection 23 1.7.1 Sorption process 24 1.7.2 Adsorption and desorption 25 1.7.3 Adsorption theorems 27 1.7.4 Langmuir Isotherm 29 1.8 Characteristics of gas Sensors 33

1.9 Need for NO2 gas sensor 36 1.10 Need for Ethanol gas sensor 38

1.11 Need for H2 gas sensor 38

CHAPTER- 2 MATERIALS AND THEIR PROPERIES 41 2.1 Material properties of Indium Tin Oxide 42 2.1.1 Crystal Structure of Indium tin Oxide 43 2.1.2 Electrical properties of ITO 46 2.1.3 Optical properties of ITO 49 2.1.4 Applications of Indium tin oxide 51 2.2 Material properties of Aluminium nitride 54 2.2.1 Crystal Structure of Aluminium Nitride 55 2.2.2 Thermal and Electrical properties 58 of Aluminium Nitride 2.2.3 Physical properties 59 2.2.4 Chemical properties 60 2.2.5 Applications of Aluminium Nitride 62

CHAPTER- 3 EXPERIMENTAL BACKGROUND AND 65 PROCEDURES 3.1 Sputtering 66 3.1.1 Plasma Process 67 3.1.2 Magnetron sputtering 69 3.1.3 Growth kinetics of thin films 71 3.1.4 Vacuum coating unit 77 3.2 Analytical Instruments and technique 87 3.2.1 Rutherford back scattering Spectrometry 87 3.2.2 X-ray powder diffraction 89 3.2.3 Scanning electron microscopy 94 3.2.4 Fourier transform infrared spectrophotometer 99 3.2.5 Ultraviolet- Visible spectrophotometer 103 3.2.6 LCR meter 107

CHAPTER- 4 PREPARATION AND CHARACTERIZATION 109 OF INDIUM TIN OXIDE THIN FILMS 4.1 Optimization of Indium tin oxide thin films 111 4.1.1 Effect of rf power on thickness and 114 deposition rate of ITO thin films.

4.1.2 Effect of rf power on the structure of ITO 115 thin films 4.1.3 Effect of rf power on the surface morphology 119 of ITO thin films 4.1.4 Effect of rf power on the electrical and 120 optical properties of ITO 4.1.5 Effect of annealing on the structure and 122 morphology of ITO thin film 4.1.6 Effect of annealing on the electrical and 125 optical properties of ITO thin film 4.2 Conclusion 128

CHAPTER- 5 PREPARATION AND CHARACTERIZATION 129 OF ALUMINUM NITRIDE THIN FILMS 5.1 Optimization of Aluminum Nitride thin films 131 5.1.1 Rutherford backscattering spectrometer 133 analysis of AIN thin films 5.1.2 Influence of nitrogen concentration on 135 the crystal orientation and deposition rate of AIN thin films.

5.1.3 Influence of deposition time on the 137 preferential orientation and deposition rate of AIN/Si thin films

5.1.4 Influence of deposition time on the 140 vibrational modes of AIN/Si thin films 5.1.5 Influence of deposition time on the 142 morphology of AIN/Si thin films

5.1.6 Influence of deposition time on the optical 143 properties of AIN/Si thin films

5.1.7 Influence of deposition time on the 144 electrical properties of AIN/Si thin films 5.1.7.1 Dielectric constant and dielectric loss 145 of AIN/Si thin films 5.2 Conclusion 148 CHAPTER- 6 FABRICATION OF GAS TESTING SYSTEM 150 6.1 Design of gas sensing unit 151 6.2 Measurement of sensitivity 154 6.3 Gas sensing mechanism of MOS device 157 6.4 Gas sensing mechanism of MIS device 161 CHAPTER- 7 FABRICATION AND CHARACTERIZATION 166 OF ITO AND AIN SENSORS 7.1 Fabrication and gas sensors measurement of ITO Sensor 167 7.1.1 Response of ITO sensor to Nitrogen dioxide gas 168 7.1.2 Response of ITO sensor to Ethanol gas 172 7.1.3 Response of ITO sensor to Hydrogen gas 176 7.1.4 Conclusion 179 7.2 Fabrication and gas sensing measurement of AlN sensor 180 7.2.1 Response of Pd/AIN/Si device to Hydrogen gas 184 7.2.2 Response of of AI/AIN/Si device to Hydrogen gas 186 7.2.3 Conclusion 187 Summary and Conclusion 188 REFERENCES 192 List of Publications / Presentation 207

INTRODUCTION

Research and development of gas sensing devices is in the focus of activity of scientists and engineers in many countries in the last two decades.

Such detectors can be used for different applications like continuous monitoring of the concentration of noxious gases in the environment and premises, detection of toxic dangerous gases, drugs, smoke and fire, energy saving as well as various technological processes in industry. Nitrogen dioxide is a hazardous gas which affects the environment and human health. It causes injury to plants and is also a deep lung irritant that can produce pulmonary edema and fatality if inhaled at high concentrations. Hence there is a great interest in developing chemoresistive gas sensor based on metal oxides for the detection of low concentration of NO2 for air quality monitoring. However, many gas sensing micro-systems have not yet reached commercial viability because of high consumed electric power and working temperatures, inaccuracies, and the inherent as well as non-specific characteristics of the sensing elements. So, suitable semiconductor materials currently needed, which have the required surface and bulk properties and high sensitivity, stability and selectivity. A wide variety of materials including SnO2 and In2O3 have been tried for NO2 detection, but the sensing properties combined in a binary form are not completely known and demands further exploration. Owing to the excellent stability of ITO as compared to the problem of drift, common with tin oxide materials, use of ITO is very promising for NO2 detection. The resistivity is in the range of 10-3 to 10-4 Ωcm making it easy to be interfaced with opto electronic circuits.

Monitoring of the hydrogen is an undesirable contaminant, for example, in chemical industry. Both leak detection and concentration monitoring are important applications of hydrogen sensors. At present, commercial hydrogen detectors are not suitable for widespread use, particularly in transportation, because they are too bulky, expensive, and some are dangerous. The sensor working at high temperatures becomes itself a possible trigger of explosion, due to high enough input of electric energy for sensor operation. From the standpoint of the safety with the global environment, the present work has been planned with a necessity to develop new hydrogen sensors working at or near room temperature. The group II – V semiconductor nitride sensors, due to their small dimensions, low cost and high compatibility with microelectronic processing, has been widely investigated to meet the requirement of H2 detection. AlN is one of the pioneering and promising H2 sensing materials due to several advantages including higher electron saturation velocity, wide bandgap, higher breakdown electric field, and superior thermal and chemical stability. The characteristics of AlN films are greatly influenced by their microstructure. Growth of smooth surface and defect less structure were the goal of device development. A large variety of AlN nanostructure sensors have been synthesized to detect H2 gases. AlN/Si device with a Pd gate is sensitive to hydrogen. Pd has high hydrogen solubility and the palladium–hydrogen system has been studied extensively. Moreover, Pd/AlN/Si films are more compatible with current microfabrication processes.

In the present work field effect gas sensors (MIS capacitors) based on wide band gap Aluminum nitride (AlN), and resistive gas sensors such as

Indium tin oxide (ITO) have been developed for the detection of reducing and oxidizing gases. The investigation includes (a) Optimization of ITO films and studying its structural, optical and electrical properties. (b) Investigating the influence of deposition parameters and annealing on the structural, optical and electrical properties of ITO film. (c) Optimization of AlN thin films and studying its structural, optical and dielectric properties. (d) Investigating the influence of deposition time on structural, optical and electrical properties of the

AlN film (e) Fabricating a gas sensing device and studying the gas sensing properties of ITO and AlN towards NO2 ethanol and H2 gases.

CHAPTER 1

GAS SENSORS - STATE OF THE ART

The atmospheric air we live in contains numerous kinds of chemical species, natural and artificial, some of which are vital to our life, while many others are harmful more or less. The vital gases like oxygen and humidity should be kept at adequate levels in living atmospheres, while hazardous gases should be controlled to be under the designated levels. As for lower hydrocarbons and hydrogen, which are used as fuels, their explosion after leakage into air is a major concern, and 1/10 of lower explosion limit for each gas is taken as an alarming level. For toxic gases, offensive odors, volatile organic compounds (VOCs) and other air pollutants, their standards have been legislated by various laws based on the strength of toxicity or offensiveness of each gas [1].

The atmospheric pollution can cause major disasters within a short period of time, since this type of pollution can diffuse rapidly over large areas. The emitting sources of the atmospheric pollution can be a stationary or mobile source. Stationary sources include houses, workplaces, thermal power stations, whereas mobile sources include automobiles, trains, ships, and so forth.

Stationary emitting sources also cover many diverse fields, such as oil refineries, chemical plants, metal refineries, and grocery plants. In particular, with the tremendous increase in automobiles and consumption of fossil fuel, the atmospheric environment has deteriorated significantly. To prevent or minimize the damage caused by atmospheric pollution, monitoring and controlling systems are needed that can rapidly and reliably detect and quantify pollution sources within the range of the regulating standard values.

1.1 SENSOR AND SENSOR SCIENCE

Sensors belong to the modern world like the mobile phone, the compact disc or the personal computer. The term ‗sensor‘ is easily understood. People may imagine a sensor similar to a sensing organ or a tentacle of an ant. A generation ago, the word sensor was not widely used. Today, however, sensors are becoming ubiquitous in our daily lives. Sensors first represented technical sensing organs, i.e. eyes, ears and tentacles, of automatic machines [2]. The human senses can not only see, hear and feel, but also smell and taste. The latter sensations are the results of some kind of chemical analysis of our environment, either of the surrounding air or of liquids and solids in contact with us.

Consequently, chemical sensors can be considered artificial noses or artificial tongues. If sensors are accepted as technical sensing organs, then it might be useful to compare a living organism with a machine.

Figure 1.1 illustrates the similarities between biological and technical systems [3]. In a living organism, the receptor of the sensing organ is in direct contact with the environment. Environmental stimuli are transformed into electrical signals conducted by nerve cells (neurons) in the form of potential pulses. Strong stimuli generate a high pulse frequency, i.e. the process is basically some kind of frequency modulation. Conduction is not the only function of neurons. Additionally, signal amplification and signal conditioning, mainly in the form of signal reduction, take place. In the brain, information is evaluated and, finally, some action is evoked.

Many similarities are seen between living organisms and machines when modern sensors and living organisms that acquire and process signals are compared. As in a living organism, a receptor which is part of the technical sensor system is found. The receptor responds to environmental parameters by changing some of its inherent properties. In the adjacent transducer, primary information is transformed into electrical signals. Frequently, modern systems contain additional parts for signal amplification or conditioning. At the end of the chain, we find a microcomputer working like the central nervous system in a living organism. The above considerations, although simplified, demonstrate that signal processing by electronic amplifiers or by digital computers are indispensable for sensor function, like the indispensability of neurons and the brain for physiological processes in organisms. As a consequence, we should accept the fact that ‗sensor‘ does not mean simply a new expression for well- known technical objects like the microphone or the ion selective electrode.

1.2 THE TERM SENSOR

It would not be sufficient to see sensors merely as some kind of technical

sensing organs. They can be used in many other fields besides just intelligent

machines. According to the definition given by Jacob Fraden in the handbook of

modern sensors [4], which states that ―A sensor is a device that receives a signal

or stimulus and responds with an electrical signal‖. Regardless of such

differences, there is broad agreement about attributes of sensors. Sensors should

[5]:

• Be in direct contact with the investigated subject

• Transform non-electric information into electric signals • Operate continuously or at least in repeated cycles

• Respond quickly

• Be small and cheap

• Be specific, i.e. they should respond exclusively to one analyte, or at least

be selective to a group of analytes

The above list could be extended with, e.g., the postulation of a low detection limit, or a high sensitivity. This means that low concentration values should be detected.

1.3 APPLICATIONS AND IMPLICATIONS OF GAS SENSORS

Gas sensors are essential for many aspects of our everyday life. Some examples of the applications of gas sensors is summarized in Table 1.1.

Table 1.1 Examples of applications of gas sensor [6]

Applications

Automobiles Car ventilation control Filter control Gas-online vapour detection Alcohol breath test

Safety Fire detection Leak detection Toxic/flammable/explosive gas detection Boiler control

Personal gas monitor Indoor air quality Air purifiers Ventilation control Cooking control

Environmental control Weather stations Pollution monitoring

Food Food quality Process quality Packing quality control (off odour) Industrial production

Fermentation control Process control Medicine Breath analysis Diseases detection

Automotive industry: control of the concentration of the gases in the

engine, to guarantee the highest possible efficiency of the combustion

process. The same concept can also be applied to power plants, as the

energy is generated by combustion.

Safety at work: monitoring toxic gases in a working environment, for

instance in a factory where dangerous chemicals are used.

Domestic safety: detection of poisonous gases or smoke in households,

due to accidents such as fires, explosions, etc.

Public security: detection of substances dangerous for the safety of the

general public, such as flammable gases and explosives.

Environment: monitoring toxic gases present in the atmosphere, due to

industrial emissions.

Food quality control: detection of particular molecules, which are formed

when food starts to rot and it is no longer good for consumption.

The correct and appropriate use of gas sensors is important for our society, as it can have implications at many different levels. For instance:

Monitoring toxic gases and automotive emissions is a vital tool in

maintaining cleaner air, especially in big cities. This can decrease the incidence of respiratory system illnesses;

consequently, people will have a better quality of life. This is particularly

true for children and elderly people.

This will also have an economic impact on society, as it will help to

reduce the costs associated with the treatments of these illnesses.

The use of gas sensors in a power plant can make the process of energy

generation more efficient.

This means less fuel will be necessary to produce the same amount of

energy; this will benefit the society economically.

As less CO2 will be produced, there will also be a benefit for the environment.

1.4 GENERAL APPROACH TO GAS SENSORS

Nowadays, there is a great interest in implementing sensing devices in order to improve environment safety and control of gases. There is also a great need of this kind of sensors to carry out the optimization of combustion reactions in the emerging transport industry and in domestic and industrial applications. The most used gas sensor devices can be divided in three big groups, depending on the technology applied in their development: solid state, spectroscopic and optic. Spectroscopic systems are based on the direct analysis of the molecular mass or vibrational spectrum of the target gas [5]. These sensors can measure quantitatively the composition of the different gases with a good precision. Mass chromatography and mass spectrograph may be the most important gas sensor spectroscopic systems. Optical sensors measure absorption spectra after the target gas has been stimulated by light. This kind of sensors requires a complex system: a monochromatic excitation source and an optical sensor for the analysis of the absorbed spectra. While spectroscopic and optic systems are very expensive for domestic use and sometimes difficult to implement in reduced spaces as car engines, the so-called solid state sensors present great advantages due to their fast sensing response, simple implementation and low prices [7]. These solid state gas sensors are based on the change of the physical and/or chemical properties of their sensing materials when exposed to different gas atmospheres.

1.4.1 Solid state gas sensor: State-of-the-art

Solid state gas sensors, based on a variety of principles and materials, are the best candidates to the development of commercial gas sensors for a wide range of applications [8, 9]. The great interest of industrial and scientific world on solid state gas sensors comes from their numerous advantages, like small sizes, high sensitivities in detecting very low concentrations (at level of ppm or even ppb) of a wide range of gaseous chemical compounds, possibility of on- line operation and, due to possible bench production, low cost. On the contrary, traditional analytical instruments such as mass spectrometer, NMR, and chromatography are expensive, complex, and large in size. In addition, most analysis requires sample preparation, so that on-line, real-time analysis is difficult. Solid-state chemical sensors have been widely used. Several physical effects are used to achieve the detection of gases in solid state gas sensors. A characteristic of solid state gas sensors is the reversible interaction of the gas with the surface of a solid-state material. In addition to the conductivity change of gas-sensing material, the detection of this reaction can be performed by measuring the change of capacitance, work function, mass, optical characteristics or reaction energy released by the gas/solid interaction.

Organic (as conducting polymers [10], porphyrins and phtalocyanines

[11, 12]) or inorganic (as semi conducting metal oxides [13, 14]) materials, deposited in the form of thick or thin films, are used as active layers in such gas sensing devices. The read-out of the measured value is performed via electrodes, diode arrangements, transistors, surface wave components, thickness-mode transducers or optical arrangements. Indeed, although the basic principles behind solid state gas sensors are similar for all the devices, a multitude of different technologies have been developed. Hence, nowadays the number of different solid state based gas sensors is really very large. Due to the large variety of sensors, a rich fabric of interdisciplinary science ranging from solid state physics, chemistry, electronics, biology, etc., governs the modern gas-sensing devices. A steering technology is that of micromachining which for chemical sensors has led to the development of gas sensor devices with small power consumption and short time constants, greater portability and easy integration with electronics.

1.4.2 Classification of sensors

Sensor classification schemes range from very simple to complex.

Depending upon the classification purpose, different classification criteria may be selected. Broad classification based on the sensor input stimulus are:

Electric sensors

Magnetic

Electromagnetic

Acoustic

Chemical

Optical

Heat, Temperature

Mechanical

Radiation

Biological

1.5 CHEMICAL GAS SENSOR: STATE-OF-THE-ART

A chemical gas sensor can be described as a device, which upon exposure to a gaseous chemical compound or a mixture of chemical compounds, alters one or more of its physical properties (e.g. mass, electrical conductivity, or capacitance) in a way that can be measured and quantified directly or indirectly.

According to the definition of a gas sensor, given by the International Union of

Pure and Applied Chemistry (lUPAC),―A chemical sensor is a device that transforms chemical information, ranging from the concentration of a specific sample component to total composition analysis, into an analytically useful signal” [15]. The chemical information mentioned above, may originate from a chemical reaction of the analyte or from a physical property of the system investigated". Typically, chemical sensors consist of two main parts, a receptor and a transducer. The receptor transforms chemical information into a form of energy, which can be measured by the transducer. The transducer converts this energy into a useful, typically electrical, analytical signal.

1.5.1 Classification of chemical sensors

For the purpose of presenting the state-of-the-art of chemical sensor technologies, it is necessary to put some order in the available devices. As in the case of general sensors, considering the detection principles, chemical sensors can be classified into electrical, magnetic, thermal, optical, mechanical or radiation sensors [16]. A classification could also be tried having in mind the materials or the technologies, however, the former is preferable since it can give a broader view of this field of research. In the sensors based on electrical properties the output signal is either generated by a reaction involving charge transport or is modulated by that reaction. This implies that there must be an electric current flow through the active sensor material in order to make a measurement, and thus, at least two electrodes. One of the first devices used to sense a chemical quantity was the electrochemical cell. Over the years several modifications have been introduced in the original configuration of the cell in order to widen the measuring scope and optimize the performance of the device.

An electrochemical cell is usually built with two or more electronic conductors, called the electrodes, immersed in an electrolyte, which is an ionic conductor.

The electrolyte might be a solid, a liquid or a gas.

A more recent example is the field effect chemical sensor. This is a solid state device similar in its working principle to the JFET or the MOSFET devices. In these devices a conducting channel exists between the source and the drain electrodes, that is modulated by the potential of a third electrode that is called the gate. In field effect chemical sensors the potential generated at the gate depends on the chemical quantity being measured. This quantity can therefore be determined by observing the current flowing from the source to the drain of the device. In this same class of chemical sensors based on electric properties are included, the thin film chemoresistive sensors. These are very promising devices, because they can be made very small and cheap and are compatible with silicon technology, which opens the possibility to integrate the sensors in the processing electronic circuit [17].

Chemical sensors based on thermal properties measurements are also an important class. These are mostly used to detect flammable gases because of the exothermic character of their reaction with oxygen. The most common of these devices have a catalytic layer that promotes the reaction at low temperature, and a temperature sensor that measures the temperature variation caused by the heat exchange. Both the presence of a substance and its concentration in atmosphere can be determined by this kind of sensors. When there is a temperature difference between the sensor and the surrounding atmosphere, the temperature variation can be related to a change of the thermal conductivity of the atmosphere and therefore also used to monitor its composition. Obviously, not only exothermic but also endothermic reactions can be monitored using such a device. Optical characteristics provide good fingerprints to distinguish different substances and are widely used in materials characterization. Recent advances in optoelectronics and fiber-optic techniques have brought some promising new ways to use these properties in chemical sensors [18]. A fiber-optic chemical sensing device consists usually, a light source, a fiber coupler to lead the light into the fiber, the light guide, a decoupler, where the returning light is separated from the exciting light and a light detection and amplification system. The measured properties may be, for instance: absorbance, reflectance, fluorescence, light scattering or refractive index.

Finally, there are two types of devices that may be included in the mechanical sensors class, that are the bulk acoustic wave sensors (BAW) and the surface acoustic wave sensors (SAW). The advantages of SAW technology over BAW include the compatibility with planar silicon technology and the possibility of using higher frequencies and potentially higher sensitivity.

Acoustic wave sensors working principle originates in the mutual dependence of electrical and mechanical properties, in a piezoelectric crystal. Typically, in

BAW gas sensors a material that adsorbs a given gas is deposited over the piezoelectric crystal, and the variation of the oscillation frequency of the crystal is related to the adsorbed mass. In order to get an unequivocal relation between the adsorbed mass and the detected substance it is necessary that the sensitive material is highly selective. Since adequate selectivities are more easy to get with organic than with inorganic materials, these sensors are generally built with the former type of materials. The fact that there is no need for a flux of energy to go through the sensitive material is also very convenient if organic materials are employed, in as much as they are usually more unstable than inorganic materials. In the case of SAW sensors, the influence of the sensitive over layer on the propagation of the acoustic wave is higher than in BAW, and thus the phase velocity and amplitude of the acoustic wave have a non negligible dependence on the elastic, piezoelectric, dielectric and conductive properties, as well as the mass of the mentioned over layer. These properties have also been studied as a step towards achieving better selectivity or sensitivity of the devices. The most common types of changes utilized in chemical gas sensors are shown in Table 1.2 along with the classes of sensor devices used to detect these changes.

Table 1.2 Types of solid state gas sensors with the corresponding physical change used as gas detection principle [6].

Physical changes Type of devices

Electrical conductivity/resistivity, Conductivity sensors (metal oxide gas sensors, conducting polymers) Δζ/ΔR

Work function, ΔΦ Field effect gas sensors: diodes, transistors, capacitors.

Mass, Δm Piezoelectric sensors: Quartz crystal microbalances (QMB), surface acoustic wave (SAW), micro cantilevers

Optical parameters: reflection, absorption interferometry, Optical sensors (fiber optic or thin film) fluorescence, refractive index or optical path length

Heat of temperature, ΔT Catalytic gas sensors: Seebeck effect, pellistors, semistors

Electromotive force or electrical Electrochemical gas sensors current in a solid state (potentiometric or amperometric) electrochemical cell

1.6 MANUFACTURING PROCESSES

The mechanism, along with the materials used in the design of chemical sensors, will obviously guide the selection of techniques employed to produce the devices. Mechanical, electrical, thermal and optical properties of the materials have to be considered when designing a new device. A good database of the available materials and their characteristics is then needed in order to get better and faster results in this research. To give some examples, it may be mentioned among the organic materials: electronic conductors like Pb- and Cu- phthalocyanines or ion conductors like Nafion. Other organic compounds are used for their piezoelectric properties, as polyvinylidine or for the high electrical conductivity and optical transmission, as polypyrrole [19]. Oxides also show different properties: SnO2 and TiO2 are electron conductors, while ZrO2 and CeO2 are ion conductors. LiTaO3 shows good pyroelectric properties, and quartz is a widely used piezoelectric material [20]. Some metals like Pt and Pd may be used because of the catalytic properties, and a ceramic material like alumina because of the electrical insulation and thermal conductivity properties.

Different preparation procedures have to be employed when the materials are organic or inorganic; when produced in single crystal form or in polycrystalline form; when in bulk or with a thin or thick film structure; whether pure or doped with one or more selected elements. It has also been shown previously that a sensor has usually a complex structure. Accordingly, it is important to have in mind not only the techniques used to prepare the surface where the interactions with the material being sensed take place, but also the techniques to produce and assemble all the other sensor elements that carry, transport or protect the generated signal from environmentally adverse conditions.

Integrated circuit technology, for instance, involves several processing steps. First it is needed a polished semiconductor wafer, usually Si or GaAs.

Then a film is grown on the wafer under ultra-high vacuum or in a carefully controlled atmosphere. After the film formation, the wafer often undergoes diffusion or ion implantation for impurity doping. The final step consists in the formation of a pattern mask on the film surface using a lithographic process, followed by etching of unwanted film or substrate parts specified by the mask.

These steps are usually repeated several times, using different masks, to produce a microstructured circuit device [21]. Over the years, several modifications have been introduced to the basic thick film techniques, leading to a higher dimensional control but also reducing the process simplicity which is one of its major advantages. Thin film techniques involve both over layer growth and surface modification techniques. The process mechanisms of these techniques have the specific characteristic of working at the atomic scale, which explains the better dimensional control usually achieved.

The processes used in the deposition of inorganic layers can be divided into three different categories: physical, chemical and hybrid methods. The physical processes include the physical vapour deposition (PVD) processes, such as: thermal evaporation, sputtering deposition and ion plating. Chemical processes include not only chemical vapour deposition (CVD), but also deposition from solutions, such as: cathodic deposition, electroless plating or homogeneous precipitation. Finally, there are the methods that combine both physical and chemical processes, like plasma enhanced CVD or laser activated deposition.

1.7 MECHANISM OF GAS DETECTION

In the gas sensing mechanism the most important processes are surface interactions. These may change surface properties and thus affect the measured resistance. If the current that flows through the sample takes a path far away from the interface between the sensing layer and the atmosphere, the measured resistance cannot be used to detect gas concentration variations. Therefore in the following discussion it is supposed that the current flowing through the sensing layer is significantly disturbed by surface interactions. This section describes some of the most common interactions taking place on a solid-gas interface and how they are related to some common thermodynamic variables, such as, temperature and pressure. In order to affect the current flowing through the sensing layer, gas molecules have to interact with the layer surface. This usually starts through adsorption and desorption processes. These interactions have characteristic lifetimes and energies, that will control adsorption rate, desorption rate and coverage of each atmospheric compound. 1.7.1 Sorption processes

When a gas or vapour is brought into contact with a solid, it will be partly taken up by the solid until equilibrium of the concentrations in the two phases is reached. The molecules either enter the inside of the solid, called absorption or dissolution, or remain on the outside attached to the surface, named adsorption.

The general term for all possible processes is sorption. It concludes adsorption, absorption, clustering of sorbate molecules, and incorporation into microvoids of the sorbent. Several modes of sorption may occur simultaneously depending on the nature of the sorbate and sorbent, the ambient conditions, namely pressure and temperature, and the concentration of the sorbate in the gas phase.

The amount of sorbed molecules as a function of the partial pressure in the gas phase at a fixed temperature is modelled by a sorption isotherm.

1.7.2 Adsorption and desorption

A solid material usually exhibits a heterogeneous distribution of surface energy. Gas, vapor, or liquid molecules may become bound to the surface if they approach sufficiently close to interact. The solid is called the adsorbent; the gas or vapor molecule prior to being adsorbed is called the adsorptive and while bound to the solid surface, the adsorbate. The adsorbed amount depends on the pressure of the adsorbate, the temperature and the properties of the adsorbent.

The phenomenon of adsorption plays an important role in both heterogeneous catalysis and in adsorption and membrane applications. Two main types of adsorption processes exists; physisorption and chemisorption. Chemisorption involves the creation of bonds between the adsorbent and the adsorbate and resembles chemical reactions. Most of the reactions being catalyzed by a solid are believed to involve an intermediate step with chemisorption of at least one of the reactants.

Physisorption is caused by weak intermolecular forces such as van-der-

Waals, induced dipoles and dipole-dipole interactions and adsorption energy usually not exceeding 80 kJ/mole, with typical energies being considerably less.

Physisorption resembles condensation of vapours rather than actual chemical reactions as in chemisorption. Physisorption is the main phenomenon used in adsorptive separation processes. Further, physisorption is used to determine the specific surface area as well as pore sizes and pore size distributions of the adsorbent. Chemical bonds involving chemical reactions are stronger than the ones involving van-der-Waal forces. The heat of adsorption is a direct measure of the bond strength between the surface and the adsorbate. This parameter will yield information on which adsorption process is taking place. For physisorption the heat of adsorption is approximately 2-3 times smaller than the heat of vaporization [22]. chemisorption is restricted to forming a monolayer, whilst in physisorption both monolayers as well as multilayers may form.

Chemisorption is usually explained using the Lennard-Jones-Model [23].

Figure 1.2 details the potential energies in the case of physisoption (Ephys) and chemisorption (Echem) as a function of the distance r from the surface. If a gaseous molecule approaches the surface it will first be physisorbed, gaining ΔE equal to Ephys. Upon a further approach towards the surface the molecule encounters a growing energy barrier, tending towards an infinite energy for a finite distance r. By spending the activation energy Ea the gaseous molecule can dissociate, thereby allowing a further approach to the surface. This stronger interaction with the surface (chemisorption) results in a higher energy gain ΔE equal to Echem than during physisorption. This energy gain Echem depends strongly on the individual surface sites available and their reactivity. The most reactive sites will therefore be occupied with gaseous molecules during thermodynamic equilibration.

Figure 1.2 Lennard-Jones model of physisorption and chemisorptions (a) Physisorption of a molecule, (b) Chemisorption of a molecule. Activation energy Ea, dissociation energy Ediss, desorption energy Edes.

1.7.3 Adsorption isotherms

However, the chemisorption energy not only depends on the number of reactive sites (high potential gain in Δ Echem) but also on the ambient gas concentration ρgas and temperature T (probability of molecules overcoming the energy barrier Ea). As for chemisorption, desorption also requires the molecule to overcome an energy barrier Edes = Echem + Ea. Therefore chemisorption and desorption are both activated processes requiring an activation energy supplied either thermally or by photoexcitation, contrary to physisorption which is a slightly exothermic process [24].

Adsorption isotherms are usually classified according to Brunauer

(Figure. 1.3). Microporous materials usually show type I behaviour. The type I isotherm is also referred to as a Langmuir type of isotherm, with a steep increase in surface coverage at low partial pressures. At higher partial pressures the isotherms starts to level off towards a distinct saturation limit when the pore system is completely filled with the adsorbate. Type II isotherms represent multilayer adsorption on non-porous solids. Type IV isotherms are typical for porous materials containing mesopores where capillary condensation occurs in the mesopores. Types III and V are rare and occurs in systems where the forces of adsorption are relatively weak.

Figure 1.3 Brunauer’s five types of adsorption isotherms [25]

At low partial pressures of the adsorbate there will be a low surface coverage, and the adsorbed molecules may be regarded as isolated from the neighbours.

Assuming that the surface is uniform, the relationship between the partial pressure and the amount adsorbed on the surface will be linear. This relationship is often referred to as Henry‘s law because of the similarity to the limiting behaviour of gases dissolved in liquids. For a Henry constant expressed in pressure the Henry law is written as:

q = kH P …. (1.1)

-1 -1 The loading of the adsorbate is q (mol kg ), kH is the Henry constant (mol kg

Pa-1) and P is the partial pressure of the adsorbate in gas phase. At higher partial pressures molecules adsorbed at adjoining sites will interact with each other.

The surface will begin to reach monolayer coverage alt. These factors will affect the amount adsorbed so that the linear relationship between the partial pressure and the surface coverage according to Eq 1.1 is no longer valid. To model this behavior a number of adsorption models have been proposed, a common model is the so-called Langmuir model or the Langmuir isotherm.

1.7.4 Langmuir isotherm

This simplest physically plausible isotherm is based on three assumptions:

Adsorption cannot proceed beyond the point at which the adsorbates are

one layer thick on the surface (monolayer).

All adsorption sites are equivalent.

The ability of a molecule to adsorb at a given site is independent of the

occupation of neighboring sites.

The Langmuir isotherm describes the surface coverage (θ) of an adsorbed gas and is dependent upon the pressure (P) at a fixed temperature [26]. It can easily be derived by assuming that the rate of adsorption is proportional to the rate of molecular collisions with vacant sites:

Rads = KA P (1-θ) …. (1.2)

In the expression, KA is the adsorption constant; P is the partial pressure of the adsorbate. Further, if qs is the saturation capacity of the surface and by letting θ be the fractional surface coverage, θ = q/qs, this implies that 1-θ is the fraction of vacant sites. The disengagement of an atom or molecule from a surface is termed ―desorption‖. At low temperatures a molecule adsorbed on a surface could remain for an indefinite time. However, as the surface temperature increases the molecule will receive energy, and so one of the following processes can occur:

The molecules can react with the surface, The molecules can break down to two or more other molecules (either in

gaseous or solid state),

The molecules can desorb.

The last option is the desorption process, which could also occur in the second option if one of the products is a gas. The rate constant of the desorption process with respects to the change of the surface coverage by desorption is given by:

Rdes = KD θ …. (1.3)

In the expression, the desorption constant is denoted KD. At equilibrium the rate of adsorption and desorption rates in equations (1.2) and (1.3) are equal, so that

KA P (1-θ) = KD θ …. (1.4)

By setting b = kA/kD and by solving for θ, the following expression is obtained:

…. (1.5)

At high partial pressures q→qs and θ →1 while at low partial pressures bP <<

1. As a consequence, at low partial pressures, Henry‘s law is valid. The adsorption equilibrium constant, b, is easily obtained from experimental data.

Eq. (1.5) is linearised to:

…. (1.6)

A plot of P/q as a function of P should yield a straight line with slope equal to

1/qs and the intercept with the y-axis equal to 1/ (qsb). Figure 1.4 shows the characteristics of several Langmuir isotherms for different values of b (T). A surface with high desorption relative to the adsorption rate has a small surface coverage (b (T) is small). With an increase in desorption rate the surface coverage will rise.

Figure 1.4 Characteristics of several Langmuir isotherms for different b (T) [26]

When considering an adsorption process of two gases (gas A and gas B) on a surface it is necessary to modify Eq. (1.5) to

bAPA A = … (1.7) 1+bAPA+bBPB

bBPB B = … (1.8) 1+bAPA+bBPB

These equations allow to express the surface coverage of two gases, which react

with each other on the surface in the following way:

A ads + Bads → ABads ABg … (1.9)

This model suitably describes the surface coverage of oxygen in terms of

oxygen partial pressure and the concentration of gaseous co-reactants. Examples

of such reactions are:

CO + O− → CO +e− 2 …. (1.10) and

NO +O− → NO +e− 2. ….. (1.11)

1.8 CHARACTERISTICS OF GAS SENSORS

In order to characterize sensor performance a set of parameters is used [3]. The

most important parameters and their definitions are listed below.

 Sensitivity: Sensitivity is a change of measured signal per analyte concentration

unit, i.e., the slope of a calibration graph. This parameter is sometimes confused

with the detection limit.

 Selectivity: Selectivity refers to characteristics that determine whether a sensor

can respond selectively to a group of analytes or even specifically to a single

analyte.

 Stability: Stability is the ability of a sensor to provide reproducible results for a

certain period of time. This includes retaining the sensitivity, selectivity,

response, and recovery time.  Detection limit: Detection limit is the lowest concentration of the analyte that

can be detected by the sensor under given conditions, particularly at a given

temperature.

 Dynamic range: Dynamic range is the analyte concentration range between the

detection limit and the highest limiting concentration.

 Linearity: Linearity is the relative deviation of an experimentally determined

calibration graph from an ideal straight line.

 Resolution: Resolution is the lowest concentration difference that can be

distinguished by sensor.

 Response time: Response time is the time required for sensor to respond to a

step concentration change from zero to a certain concentration value.

 Recovery time: Recovery time is the time it takes for the sensor signal to return

to its initial value after a step concentration change from a certain value to zero.

 Hysteresis: Hysteresis is the maximum difference in output when the value is

approached with an increasing and a decreasing analyte concentration range.

 Life cycle: life cycle is the period of time over which the sensor will

continuously operate.

 Working temperature is usually the temperature that corresponds to maximum

sensitivity.

All of these parameters are used to characterize the properties of a particular material or device. An ideal chemical sensor would possess high sensitivity, dynamic range, selectivity and stability; low detection limit; good linearity; small hysteresis and response time; and long life cycle. Investigators usually make efforts to approach only some of these ideal characteristics, disregarding the others. On one hand, this is because the task of creating an ideal sensor for some gases is extremely difficult, if at all possible. On the other hand, real applications usually do not require sensors with all perfect characteristics at once. For example, a sensor device monitoring the concentration of a component in industrial process does not need a detection limit at the ppb level, though the response time at range of seconds or less would be desirable. In case of environmental monitoring applications, when the concentrations of pollutants normally change slowly, the detection limit requirements can be much higher, but response time of a few minutes can be acceptable.

1.9 NEED FOR NO2 GAS SENSOR

Nitrogen dioxide (NO2) is one of several nitrogen oxides. Nitrogen oxides represent a mixture of gases designated by the formula NOX. The mixture includes nitric oxide (NO), nitrogen dioxide (NO2), nitrogen trioxide (N2O3), nitrogen tetroxide (N2O4), and nitrogen pentoxide (N2O5). Nitrogen dioxide is a paramagnetic bent molecule with C2v point group symmetry. It is a colorless brown liquid at room temperature and a reddish-brown gas above 70º F, poorly soluble in water [27].

In recent years much attention has been focused on the role of NO2 containing gases and their detection and measurement in urban air quality due to automotive and industrial emissions. Nitrogen oxide NO which is found in the combustion exhaust of automobiles and oil-burning power plants is a poisonous, odourless, colorless gas. Most of combustion processes result in the formation of nitric oxides, commonly considered inside the global label NOx. NO2 is the main gas to be detected because NO is easily oxidized to NO2 in atmosphere.

There is a complex relationship between emissions for NOx and the resulting concentrations of NO2. It transforms in the air to form gaseous nitric acid and toxic organic nitrates.

Nitrogen dioxide is heavier than air, so that exposure in poorly ventilated, enclosed, or low-lying areas can cause asphyxiation. Nitrogen oxide NOx (NO or NO2) can cause diseases of respiratory system and leads to pulmonary edema and death. People with bronchitis or asthma are particularly sensitive to the gas, and lungs may become inflamed, leading to breathing difficulties. The monitoring of the gas in environments where NO2 is at particularly high risk concentrations is desirable as one would like to reduce the risks to human health as much as possible. NO2 is often found at higher levels indoors compared with outdoors. Mainly this occurs in settings where gas stoves and kerosene heaters are being used. In this sense, the emission control and effective methods to detect nitrogen oxides are highly demanded today to prevent environmental and health problems [28]. Therefore, to optimize combustion and reduce emissions, it is necessary for the precise monitoring of exhaust gasses in boilers, heaters, gas stoves, combustion furnaces, vehicles or automobile engines and their direction by controlling initial reactant mixtures. Therefore the development of portable fast-response sensors that are robust, small sized, long lifetime, quick in response and with sufficient sensitivity for the detection of nitrogen dioxide in low concentrations, such as few ppm, in the ambient is necessary and demanded also in order to prevent irreversible changes in the global atmosphere.

1.10 NEED FOR ETHANOL SENSOR

Sensor materials for detection of alcohols, especially ethanol, are being actively developed. Ethanol is the most important alcohol owing to its varied applications. The interest in ethanol is connected with its wide application in chemistry, medicine, and the food industry. Precise quantitative detection of ethanol vapors is required for determination of quality of wines and human health. It is a solvent for various alcohol soluble active ingredients and also an alternative to automotive fuels. Ethanol is widely used as well in food industry, brewing process control, medical and clinical applications and bio- technological processes. Scientific research on ethanol sensor has made considerable progress but still, there is a need and scope of improving the selectivity of the sensor for different levels of ethanol vapour [29]. Hence, there is a great interest in developing chemoresistive gas sensor based on metal oxides for the detection of low concentration of ethanol.

1.11 NEED FOR H2 GAS SENSOR With the increasing concern about the global climate change, more attention is paid to hydrogen as a clean energy source. Since hydrogen is a renewable, abundant, efficient energy source and provides zero emission, it is the near-future fuel. Hydrogen has many potential applications as an energy carrier. Changes in gasoline and diesel fuel specifications, prompted by environmental legislation, have led to the greater use of hydrogen for improving the grade of gasoline. The use of hydrogen is extending quickly in fuel refinement, like the breaking down by hydrogen (hydrocracking), and in sulphur elimination. Huge quantities of hydrogen are consumed in the catalytic hydrogenation of unsaturated vegetable oils to obtain solid fat. Hydrogenation is used in the manufacture of organic chemical products. Enormous quantities of hydrogen are used as rocket fuels, in combination with oxygen or Fluro, and as a rocket propellent propelled by nuclear energy. Hydrogen can be burned in internal combustion engines. Hydrogen fuel cells are being looked into as a way to provide power, and research is being conducted on hydrogen as a possible major future fuel. Due to its chemical properties, hydrogen poses unique challenges in the plant environment. Hydrogen gas is colorless, odorless, and not detectable by human senses. Hydrogen has a large diffusion coefficient (0.61cm2/s) and wide combustion range (4-75%) and small ignition energy

(0.02 mJ in air, methane‘s value is 0.3 mJ) [30]. A hydrogen leak should be avoided because hydrogen when mixed with air in the ratio larger than ~4 vol.

% is explosive.

There are several hazards associated with hydrogen, ranging from respiratory ailment, component failure, ignition, and burning. High concentrations of this gas can cause an oxygen-deficient environment.

Individuals breathing such an atmosphere may experience symptoms which include headaches, ringing in ears, dizziness, drowsiness, unconsciousness, nausea, vomiting and depression of all the senses.

In all these applications some precautions are required for the safe use of hydrogen. For this reason, it has become very important to develop highly sensitive hydrogen detectors with a large variety of specifications to prevent accidents due to hydrogen leakage. Such detectors should allow continuous monitoring of the concentration of gases in the environment in a quantitative and selective way. Therefore, hydrogen sensors would form an integral part of any such systems incorporating hydrogen.

CHAPTER 2

MATERIALS AND THEIR PROPERTIES

Interest in transparent conductors can be traced back to 1907 when reports of transparent and conductive cadmium oxide (CdO) films first appeared. Since then there has been a growing technological interest in materials with these unique properties as evidenced by not only their increased numbers but also the large variety of techniques that have been developed for their deposition. It is now known that non-stoichiometric and doped films of oxides of tin, indium, cadmium, zinc and their various alloys exhibit high transmittance and nearly metallic conductivity is achievable [31]. Of all the transparent conducting oxide films, tin-doped indium oxide (ITO) an n-type, highly degenerate, material is now being used for gas sensing applications, due to their properties such as good adhesion to substrates, chemical stability and easy patterning ability compared to the other TCO materials.

Recently, many semiconductor-type sensors have been fabricated successfully and reported widely [31-35]. Although the fabrication technology of low-cost Si-based semiconductor gas sensors is mature, they could not be operated under higher environmental temperature due to their relatively small bandgap (1.12 eV) [36, 37]. Materials with wide bandgap such as AlN , as GaN

[38,39] and InN [40,41], were proposed to overcome this disadvantage. AlN has a high breakdown field, can operate at high temperatures in excess of 400 0C and has decent thermal conductivity in bulk, low-defect wafers making it an effective material choice for gas sensing application. The physical and/or chemical changes induced by gases when they come in contact with sensor device mainly depend on the properties of the sensing materials. In the present chapter the properties of ITO and AlN materials used for gas detection are discussed in detail.

2.1 MATERIAL PROPERTIES OF INDIUM TIN OXIDE

Tin doped indium oxide (ITO), with reported transmittance and conductivity as high as 95% and 104 W-1cm-1, respectively, is among the most popular of thin films which have found a host of electronic, opto-electronic and mechanical applications. Indium tin oxide (ITO) thin films have been extensively used as transparent conducting electrodes in flat-panel displays

(FPDs), solar cells [42, 43], and organic light-emitting diodes (OLEDs) [44, 45] because they have high electrical conductivity, high optical transparency, and smooth surface morphology. ITO thin film can also be used as gas sensor, to detect NOx gases and carbon tetrachloride [46-48].Recent research has shown that (ITO) offers new advantages in the design of metal oxide-based gas sensors. ITO has high gas sensitivity, because its conductivity increases remarkably when exposed to reducing gases.

However, although it has been found for about 40 years that ITO has both excellent electrical and optical properties [49,50], the theoretical understanding of ITO has been limited, mainly because of the complex crystal structures of

ITO, which exhibit a bixbyite structure, with a unit cell containing 40 atoms and two non-equivalent cation sites. In the present chapter, the structure, properties and application of ITO are discussed briefly.

2.1.1 Crystal structure of Indium tin oxide

Indium tin oxide (ITO or tin-doped indium oxide) is a solid solution of indium (III) oxide (In2O3) and tin (IV) oxide (SnO2), typically 90% In2O3, 10%

SnO2 by weight. It is transparent and colorless in thin layers, while in bulk form it is yellowish to grey. The unique properties of ITO come from its structure and composition. Indium oxide has a cubic bixbyite structure which is shown in

Figure 2.1. One unit cell contains 16 units of In2O3. Therefore, for defect free

In2O3 crystal, there are 80 atoms in one unit cell. The lattice constant is reported to be 10.118 Å [51]. The theoretical density is 7.12 g/cm3. Two kinds of non- equivalent indium sites present in In2O3 crystal structure is shown in Figure 2.2

[52]. In Figure 2.2 (a), the separation between indium and oxygen atoms is 2.18

Å, and the oxygen atoms are positioned at the corners of the cube with two body-diagonally opposite corners unoccupied. In Figure 2.2 (b), the In-O separations are 2.13, 2.19 and 2.23 Å; where the oxygen atoms occupy positions at the corner of the cube with two face-diagonally opposite corners unoccupied.

Based on the description above, both indium sites can be viewed as an incomplete body centered cubic structure with an indium atom located at the center and oxygen atoms at the corners. One fourth of the anions are missing.

To vary the material properties, In2O3 can be extrinsically doped with tin.

Indium tin oxide is essentially formed by subsititutional doping of In2O3 with

Sn which replaces the In3+ atoms from the cubic bixbyte structure of indium oxide [53]. Sn thus forms an interstitial bond with oxygen and exists either as

SnO or SnO2; accordingly it has a valency of +2 or +4 respectively. This valency state has a direct bearing on the ultimate conductivity of ITO. The lower valence state results in a net reduction in carrier concentration, since a hole is created which acts as a trap and reduces conductivity. On the other hand,

4+ predominance of the SnO2 state means Sn , which acts as an n-type donor releasing electrons to the conduction band. However, in ITO, both substitutional tin and oxygen vacancies contribute to the high conductivity, and the material can be represented as In2-xSnxO3-2x.

Figure 2.1 Crystal Structure of Indium tin oxide

Figure 2.2 Two indium sites of ITO [52]

ITO has lattice parameter close to that of In2O3. However, if the doping level is extremely high, the tin atoms may enter interstitially and distort the lattice structure.

2.1.2 Electrical properties of ITO

Indium oxide is a wide gap semiconductor. The band gap of In2O3 was reported to be 3.75eV [53]. Generally, In2O3 crystal structure is not ideal since oxygen vacancies are present. The accurate formula of indium oxide is In2O3-x

(VO)x, where Vo is donor-like doubly-charged oxygen vacancy, and x is the oxygen vacancy portion and depends on the oxidation state. The typical value of x is less than 0.01. Under room temperature, the excitation of electrons from valence band to conduction band is negligible. Oxygen vacancies dominate the conduction mechanism of In2O3 and the free charge carrier concentration has been reported to be in the range 1019-1020 cm-3. Extrinsic doping can vary the electrical properties of indium oxide significantly. If In2O3 is doped with tin atom; tin atoms will replace indium atoms and form tin oxide in either SnO or

SnO2 depending on the valence. If SnO is formed, tin acts as an acceptor since it accepts an electron. Otherwise, when SnO2 is formed, it acts as donor since it gives off an electron. Usually, SnO2 dominates, and tin atoms act as donors.

Both tin and oxygen vacancies contribute to the conductivity of ITO.

The doping level is critical for electrical properties. As the tin concentration increases, the carrier concentration increases until a saturation level is reached [54]. An increase in the tin concentration above this saturation level causes a decrease in the free carrier concentration. This is due to an increased probability of the occupation of adjacent cation positions by two or more tin atoms, which can deplete the active tin concentration. L.Gupta et al.

[55] has studied the carrier concentration as a function of tin doping level which is well illustrated in Figure 2.3. The highest carrier concentration, which corresponds to the lowest resistivity, occurs when the tin doping level is about

10%. The solid solubility of tin in indium oxide is approximately 8%. Beyond this range, tin oxide phase will be formed and distort the lattice structure. This will eventually cause a decrease of free carrier concentration.

In order to get conductivity > 103 ohm-1cm-1 and avoid the formation of tin oxide phase, the doping level is typically chosen to be 8-10%. This doping level will produce degenerative ITO [56]. Figure 2.4 shows the energy diagram of

In2O3 and ITO. ITO has metal like electrical properties because the carrier concentration is typically around 1020 to 1021cm-3. For heavily doped indium oxide, the contribution of oxygen vacancies to the conductivity is negligible.

Figure 2.3 Carrier concentrations as function of the tin doping level [56]

Figure 2.4 Energy band diagram of In2O3 and ITO [57]

2.1.3 Optical properties of ITO

The direct optical bandgap of ITO films is generally greater than 3.75 eV although a range of values from 3.5 to 4.06 eV have also been reported in the literature [53]. The high optical transmittance, of these films is a direct consequence of their being a wide bandgap semiconductor. The fundamental absorption edge generally lies in the ultraviolet of the solar spectrum and shifts to shorter wavelengths with increasing carrier concentration, N. This is because the band gap exhibits N2/3 dependence due to the Moss-Burstein shift [57]. The band structure of ITO is assumed to be parabolic as shown in Figure 2.4.

The conduction band is curved upwards, the valence band is curved downwards and the Fermi level is located at mid bandgap for the undoped material; addition of Sn dopants results in the formation of donor states just below the conduction band. As the doping density is increased, these eventually merge with the conduction band at a critical density, nc, which was calculated to be 1×1020 to 1×1021 cm−3 by Gupta et al.[55]. Free electron properties are exhibited by the material when the density of electrons from the donor atoms exceeds this value. All reported values of carrier concentration are greater than nc

[58]. Hence all ITO films are expected to be degenerate in nature. Once the material becomes degenerate, the mutual exchange and coulombic interactions shift the conduction band downwards and the valence band upwards - effectively narrowing it from Eg to Eg' as shown in Figure 2.4 earlier. The bandgap increase by the Burstein-Moss shift is partially compensated by this effect. The reported value for the refractive index of ITO is 1.96 [59]. The transmittance of ITO films is also influenced by a number of minor effects which include surface roughness and optical inhomogeneity in the direction normal to the film surface. Inadvertently grown dark brown (effectively translucent) metallic films of ITO have also been reported [60]. This opaqueness has been attributed to unoxidised Sn metal grains on the ITO surface as a result of instability due to absence of sufficient oxygen during deposition [61].

The theoretical treatment of the dielectric function of ITO was derived by

Bender et al. [62] from the imaginary part of the dielectric constant; the absorption coefficients of the films were derived. Two absorption edges was reported, one at photon energies below 1eV (far infrared region), which is due to the free electrons, and one at photon energies above 4 eV which is caused by the excitation of the valence electron into the conductance electron. In the region between 1 to 4 eV, the imaginary part of the dielectric constant is negligible which corresponds to a high transmission region in the visible and NIR region.

2.1.4 Applications of Indium tin oxide Indium tin oxide is one of the most widely used transparent conducting oxides because of its two chief properties, its electrical conductivity and optical transparency, as well as the ease with which it can be deposited as a thin film.

As with all transparent conducting films, a compromise must be made between conductivity and transparency, since increasing the thickness and increasing the concentration of charge carriers will increase the material's conductivity, but decrease its transparency [63]. Indium tin oxide is a conductive oxide of the indium element. Global production of refined indium compounds including indium tin oxide increased by 14 percent between 2003 and 2004, due to the rise in demand for LCD products.

ITO is mainly used to make transparent conductive coatings for liquid

crystal displays, flat panel displays, plasma displays, touch panels,

electronic ink applications, organic light-emitting diodes, solar cells,

antistatic coatings and EMI shieldings. In organic light-emitting diodes,

ITO is used as the anode (hole injection layer).

ITO has been used as a conductive material in the plastic

electroluminescent lamp of toy Star Wars type light sabers.

ITO is also used for various optical coatings, most notably infrared-

reflecting coatings (hot mirrors) for architectural, automotive, and sodium

vapor lamp glasses. Other uses include gas sensors, antireflection

coatings, electrowetting on dielectrics, and Bragg reflectors for VCSEL

lasers.

ITO was used as a sensor coating in the later Kodak DCS cameras,

starting with the Kodak DCS 520, as a means of increasing blue channel

response. It is reportedly used as a sensor coating in the Canon

400D/XTi and Sony Alpha DSLR-A100. ITO thin film strain gauges can operate at temperatures up to 1400 °C and

can be used in harsh environments, e.g. gas turbines, jet engines, and

rocket engines.

ITO wire pressed between two pieces of glass can be used to create a

hermetic seal that will remain viable even at cryogenic temperature.

 Conductive property The main advantage of indium tin oxide is its ability to conduct electricity

in small quantities. A very thin layer provides continuous conductivity

over the entire surface it covers. Indium tin oxide is also stable; it does

not react with several cleaning agents such as acetone, toluene and

xylene.

 Infrared Reflection In addition to its electric conducting properties, indium tin oxide

coatings are also able to reflect infrared radiation. This allows it to reduce

the amount of radiation transmitted to the liquid crystals, thus regulating

its temperature. As a result of this property, indium tin oxide coatings are

applied to glass windows and doors to prevent fog and frost, as with

aircraft and commercial refrigerator

 Solar Panels The conductive and transparent properties of indium tin oxide are used in

solar panels. Conductors inside the panels are made of indium tin oxide

and collect energy from plastic solar cells.  Display Screens Some screens, including plasma displays, flat panel televisions and

computers, checkout displays and items with touch-screen technology,

including the iPhone and other mobile phones, also take advantage of

indium tin oxide's transparency and conductivity.

 Windshields Indium tin oxide is lightly sprayed on the surface on many aircraft and

high-end automobile windshields to provide a thin coating that protects

against lower temperatures, effectively deicing the windshield and

creating a frost-free travel experience.

 Energy-Efficient Windows Using the same principles as the deiced windshields, indium tin oxide is

also used to thinly coat the outside of many energy-efficient windows.

The properties of indium tin oxide create a barrier that keeps cooler air

outside.

2.2 MATERIAL PROPERTIES OF ALUMINUM NITRIDE

Aluminum nitride, which is a III-V family compound, has the structure of hexagonal wurtzite. As it features a wide bandgap, high electrical resistivity, high resistance to breakdown voltage, a high acoustic propagation rate and a low transmission loss, AlN film has a wide application potential in microelectronic devices, especially in surface acoustic wave SAW devices [64-

66]. Since the acoustic velocity of AlN film is approximately two-fold that of ZnO and CdS, the central frequency is able to be enhanced twice without decreasing the interdigital electrode width [67]. AlN film is the optimal material for GHz grade SAW devices [68]. As the gate silicon oxides of contemporary

MOS IC shrink to thicknesses of several nanometers, indispensable quantum- mechanical currents across the gate oxide film severely hinder the IC functioning. Materials with high dielectric constants have been considered as potential substitutes for silicon oxide. Compared with low-permittivity gate dielectrics, more charge is induced for the same applied gate voltage.

Accordingly, they can thus be thicker, which would reduce tunneling currents.

Bulk AlN has a dielectric constant of 10 εo compared to 3.9 ε0 for SiO2 [69].

Besides, Aluminum nitride (AlN) is one of the blue and UV light emitting nitrides, which gives it an edge over other high-k dielectrics. Preparing AlN-Si metal-insulator-semiconductor (MIS) structures with acceptable qualities will thus open the path to co-integration of light emitting nitrides with silicon microelectronics. In addition, it is established that AlN- Si interfaces are thermally stable. The electrical properties, mainly fixed dielectric charge of

AlN-Si MIS capacitors and AlN-Si MISFET [70] have been investigated; however, the specific resistivity of AlN flms in these structures was the only parameter that was estimated. The concentration and nature of AlN film defects play an important role in the electro-physical properties of AlN-Si MIS structures. Also, a good understanding of these properties is essential for establishing optimal conditions for AlN deposition by different methods.

2.2.1 Crystal structure of Aluminum Nitride

Aluminum nitride has the wurtzite structure, 6-mm point group and P63mc space group. Reported lattice parameters range from 3.110 to 3.113 Å for the ‗a‘ lattice and from 4.978 to 4.982 Å for the c lattice. The c/a ratio thus varies between 1.600 and 1.602. The variation in c/a ratio, deviating from that of the ideal wurtzite crystal, is probably due to lattice instability and ionicity [71].

Figure 2.5 shows the crystalline lattice structure of AlN. In this structure, each

Al atom is surrounded by four N atoms, forming a distorted tetrahedron with three Al-N(I) (I=1,2,3) bonds named B1 and one A1-N0 bond in the direction of the c-axis, named B2. The stacking sequence of the (0001) plane is ABABAB in the <0001> direction as shown in Figure 2.6. The bond lengths of B1 and B2 are

0 0.1885 and 0.1917 nm, respectively. The bond angle for N0-Al-N1 is 107.7 and

0 that for N1-Al-N2 is 110.5 . Similarly, a tetrahedron is formed with the N atoms as the corner. Two tetrahedrons form a triangular prism with the C3v symmetry.

In the AlN cell, the atoms of Al and N form four sp3-hybridized orbits. The Al atom has three semi full orbits and one empty orbit; the N atom has three semi- full orbits and one full orbit. The bond B2 is formed by the coupling of the Al empty orbit and the N full orbit. Consequently, the ionic character of the B2 bond is greater.

The bond energy of B2 is relatively smaller than that of the three equivalent B1 bonds and is easy to break, so that the energy required for sputtering particles to be deposited in the direction of the c-axis must be exceeded. The plane (100) is composed of the bond B1, while planes (002) and

(101) consist of the bonds B1and B2 together. AlN is generally reported to be non-polymorphous [72].

Figure 2.5 Crystal structure of the aluminum nitride [72]

Figure 2.6 Geometry of crystallographic planes (002), (102) and (101) in hexagonal AlN lattices

However, several reports suggested the occurrence of a metastable zincblende polytype of AlN, with a lattice parameter a=4.83 Å. A pressure

5 induced rocksalt (space group Oh ) phase has been observed experimentally. The equilibrium transformation pressure was reported to be less than 14 GPa. The rocksalt structure has an a-axis value of 4.043~4.045 Å at room temperature

2.2.2 Thermal and electrical properties of Aluminum Nitride

Aluminum nitride has attractive thermal and electrical properties for high density and high power integrated circuit applications. Table 2.5 shows the mechanical, thermal and electrical properties of AlN. The thermal conductivity of single crystal AIN is estimated to be 320 W/m·K, but that of sintered polycrystalline AIN is generally found to be about 80 - 200 W/m·K. [73-76]. Its coefficient of thermal expansion (CTE) is 4.4 x 10-6/ 0C, which is very close to that of silicon ( 3.2 x 10-6/ 0C) these two properties make it possible that AIN can be used in high power, high density circuits as substrate materials rather than other ceramic materials, which exhibit lower thermal conductivities and/or higher CTEs.

Because in high power, high density circuits, large amounts of heat are produced during operation, if a substrate material does not have high thermal conductivity and CTE to that of the Si chips, either the substrates or Si chips may fail during operation. In addition, AIN's electrical properties, such as its electrical resistivity (1014Ω.cm), dielectric constant (8.9), and breakdown voltage (15 kV/mm) are equally as good as the other ceramic materials used for this type of application.

Slack et al. [77] measured the thermal conductivity of high purity, single crystal AIN over a wide temperature range and obtained a value of 319 W/mK at

300 K. One reason for the decrease in thermal conductivity of AIN ceramics is oxygen. It has been reported by Kurokawa et al. [78] that the thermal conductivity of AIN ls drastically affected by oxygen and other metallic impurities such as Si,

Fe and Mg. When oxygen substitutes for nitrogen in the AIN lattice, it creates aluminum vacancies due to the difference in valency between oxygen and nitrogen. This creation of vacancies leads to phonon scattering and thereby lowers the thermal conductivity far below that of pure AIN single crystals (320 W/m·K)

[79]. Furthermore, if there are any other metallic impurities (e.g. Mg, Si, Be) [78,

79] in the AIN lattice, phonon impurity interactions occur and cause the thermal conductivity to be reduced still further. Therefore, to obtain high thermal conductivity AIN, it is very important to produce AlN thin films which contain extremely low amounts of oxygen and metallic impurities.

2.2.3 Physical Properties

Pure aluminum nitride is a III-V insulating compound which crystallizes with the wurtzite structure. There are no other known polytypes. The theoretical density is 3.25 g/cm3 and it is colourless and optically transparent when pure. [54].

The melting point or decomposition point of AIN has been reported to be 2800 oC under 100 atm. pressure of N2 [80]. Elsewhere [81], it was reported that at atmospheric pressure, AIN decomposes lnto aluminum and nitrogen at 2450 oC.

The flexure strength of AIN at room temperature (35-50 MPa) is much higher for

Al2O3 (31-32 MPa) and BeO (25 MPa). On the other hand, the Vickers hardness of

AIN (1200 MPa) is half of that of Al2O3 (2300 – 2700 MPa). These factors mean that AIN is stronger than Al2O3 and BeO but more easily machinable than Al2O3.

2.2.4 Chemical properties The material is stable at very high temperatures in inert atmospheres. In air, surface oxidation occurs above 700 °C, and even at room temperature, surface oxide layers of 5-10 nm have been detected. This oxide layer protects the material up to 1370 °C. Above this temperature bulk oxidation occurs. Aluminum nitride is stable in hydrogen and carbon dioxide atmospheres up to 980 °C. The material dissolves slowly in mineral acids through grain boundary attack, and in strong alkalis through attack on the aluminum nitride grains. The material hydrolyzes slowly in water. Aluminum nitride is resistant to attack from most molten salts including chlorides and cryolite.

Table 2.1 Mechanical, Thermal and Electrical properties of AlN [80]

Units of Mechanical SI/metric (Imperial) measurement Density gm/cc (lb/ft3) 3.26 (203.5)

Flexural Strength 2 3 MPa (lb/in x10 ) 320 (46.4)

Elastic Modulus 2 6 GPa (lb/in x10 ) 330 (47.8)

Shear Modulus GPa (lb/in2x106) — —

Bulk Modulus GPa (lb/in2x106) — — Poisson‘s Ratio — 0.24 (0.24)

Compressive Strength MPa (lb/in2x103) 2100 (304.5)

Hardness Kg/mm2 1100 —

Fracture Toughness MPa.m1/2 2.6 —

THERMAL

Thermal Conductivity W/m.°K 140–180 (970–1250) (BTU.in/ft2.hr.°F) Coefficient of Thermal Expansion 10-6/°C (10-6/°F) 4.5 (2.5)

Specific Heat J/Kg.°K (Btu/lb.°F) 740 (0.18)

ELECTRICAL

Dielectric Strength kv/mm (volts/min) 17 (425)

Dielectric Constant @ 1 MHz 9 (9)

Dissipation Factor @ 1 MHz 0.0003 (0.0003)

2.2.5 Applications of Aluminum Nitride

During the period 1985-88, the following concerns arose about the use of

AlN in microelectronic packaging: oxidative stability, hydrolytic stability, surface finish/chemistry, metallization paste systems, cofiring and brazing, thermal conductivity, cost. High-Quality AlN substrates and packaging emerged in the early 1990s. The consistency of AlN products has increased dramatically and processing technologies have been developed to the point that AlN ceramics are now in use worldwide in commercial packages. Metallization systems as consistent and strong as the oxide metallization systems have been developed for nitrides. In addition, co-fired, multilayer AlN packaging has been developed and offers an alternative for power applications. More than 20 companies worldwide are developing AlN technology. In comparison to other electrically insulating ceramic materials only AlN and beryllium oxide (BeO) offer high thermal conductivities. BeO shows higher values than AlN but has been banned by numerous manufacturers because of its toxicity. Thus, the nontoxic AlN continues to replace BeO for most applications. The attractive application of aluminum nitrides are:

Opto-electronics

Dielectric layers in optical storage media

Electronic substrates, chip carriers where high thermal conductivity is

essential Military application

Power electronics (electrical engines)

Microelectronics (LSI circuits, sensor carriers, high frequency modules)

Naval radio systems, defense systems

Railway systems (inverters for drive systems)

Aeronautical systems (telecommunication and research satellites)

Environmental systems (emission control)

 Besides high thermal conductivity, other attractive properties of AlN

which make it an ideal candidate for substrate in electronic packaging are

(1) low coefficient of thermal expansion, (2) high resistivity, (3) low

dielectric constant and low dielectric loss at high frequencies and (4) high

dielectric breakdown strength

 AlN is the preferred choice for substrates, packages and heat-sinks or

coolers, whenever the thermo mechanical properties or power dissipation

capabilities of conventional aluminum oxide materials are insufficient.

Typical examples are highly integrated thick-film and thin-film

components, water-cooled power converters in rail transport systems, and

transmitter and high frequency diode substrates subjected to thermal

cycling in satellite systems.  Conventional metallizing processes for AlN have in recent years been

developed to the extent that they are now being increasingly used for

volume production.

 For power electronics in particular, conducting lines able to carry high

currents are essential. Copper combines the required low electrical

resistance with exceptionally high thermal conductivity.

 Currently there is much research into developing light-emitting diodes to

operate in the ultraviolet using the gallium nitride based semiconductors

and, using the alloy aluminum gallium nitride, wavelengths as short as

250 nm have been reported.

 AlN thin films are used for surface acoustic wave (SAW) devices because

of their low acoustic loss and high ultrasonic velocity.

 Agilent after more than a decade of research now has a RF filter used in

mobile phone called the FBAR. This technology is closely associated

with engineers working in the MEMS field.

CHAPTER 3

EXPERIMENTAL BACKGROUND AND PROCEDURES

Many fabrication methods have been used in the production of metal oxide and metal nitride semiconductor sensors. Factors that must be considered when selecting the production technique include; expense (if the films are expensive, the demand will be low and will have only limited applications), purity, porosity (if the material is highly porous, the surface area available to the gas for interaction will be far higher, giving a higher sensitivity), reliability and reproducibility. Common techniques for making the metal oxide and nitride films for gas sensors are mainly chemical vapour deposition (CVD), screen- printing of ceramic powders, sol-gel techniques and physical vapour deposition

(PVD), electron beam and sputtering. In order to fabricate the high quality ITO thin films, several methods have been reported, such as rf magnetron sputtering

[82-84], and dc magnetron sputtering [85-87]. The effects of the deposition parameters on properties of aluminum nitride thin films were investigated by dc

[88, 89] and rf sputtering [90-95].

In the present work, ITO and AlN thin films have been prepared by magnetron sputtering. Sputtering may be a more desirable process for the deposition of oxide and dielectric materials due to its processing simplicity and scalability in the semiconductor industry. The advantages of reactive magnetron sputtering include inexpensive equipment, low temperature growth conditions and ease of deposition. It is important to be able to control the crystal orientation and crystalline quality of the film for many applications. Properties of films are strongly dependent on deposition parameters such as target power, growth temperature, sputtering pressure and gas compositions [96-102].

3.1 SPUTTERING

Magnetron sputtering is a thin film deposition technique based on the physical sputtering effects caused by the bombardment of a target material with accelerated ions produced within glow discharge plasma. A wide variety of thin film materials, from metals to insulators, may be produced using this technique

[88]. Usually the thin film composition will be determined exclusively by the elements composing the target, since the only elements occurring in the gas phase are the noble gas ions used to generate the plasma and other particles ejected from the target as the result of the sputtering process. In some cases a reactive element may be introduced in the gas phase that will interact with the deposited elements; this latter case is called reactive sputtering. In order to understand the thin film growth processes and the dependence of its characteristics on the system parameters, it is necessary to know how the particles are removed from the target and how they interact on the substrate surface. Therefore in the following sections ion production, sputtering mechanisms and film growth will be briefly explained, before presenting the system used to produce the sensor film layers and the correlations found with the deposition parameters.

3.1.1 Plasma process

Ions used in planar magnetron sputtering systems are generated by glow discharge processes. A glow discharge may be established when a potential is applied between two electrodes in a gas. Once established two or three different regions can be clearly distinguished between the electrodes. Close to the cathode there is a dark region that exists also near the anode although the latter is generally too thin to be observed. Next there is negative glow region and if the separation between the electrodes is higher than a few times the dark space thickness, a positive column develops between the negative glow and the anode.

The positive column is the region that more closely resembles a plasma. The plasma does not take a potential between those of the electrodes, but rather acquires a potential slightly higher than that of the anode. Inside the plasma, as expected, the electric field is very low. The main potential difference is therefore observed at the sheaths next to each of the electrodes. The role of the magnetron is forming electron traps that help to sustain the discharge.

Without the magnetron, the electrons emitted from the cathode by ion bombardment are accelerated nearly to the full applied potential in the cathode dark space and enter the negative glow, where they collide with the gas atoms producing the ions required to sustain the discharge. As the electron mean free path increases with gas pressure decrease, at low pressures the ions will be produced far from the cathode where the chances of being captured by the walls are high. At the same time many primary electrons (the ones emitted from the cathode) reach the anode with high energy, inducing secondary electron emission. Therefore, ionization efficiencies are low and self sustained discharges cannot be maintained in planar diodes at pressures below 1:3Pa

[103]. Currents are also limited since voltage increases the primary electron energy and consequently their mean free path. At high pressures the sputtered atom transport is reduced by scattering, which at some point starts to force a decrease in the deposition rate. On the other hand, with the magnetron system configuration, the primary electron motion is restricted to the vicinity of the cathode and thereby ionization efficiency is increased. This effect is easily obtained imposing a magnetic field parallel to the cathode surface and thus normal to the electric field.

In a planar configuration, the E x B motion causes the discharge to be swept to one side. This difficulty can be overcome using cylindrical cathodes, which allow E x B to close on them. Planar magnetrons can be achieved by placing the magnets, directly behind the cathode, in a configuration such that at least one region in front of the cathode surface has a closed path which is perpendicular to the magnetic field lines that are parallel to the surface.

Although there are many variations in geometry, all have in common a closed path or region in front of a mostly flat cathode surface, where the magnetic field is normal to the electric field. Bounding this region the magnetic field lines enter the cathode surface. Ideally, at the entry points the field lines are perpendicular to the cathode surface. The ionization region is thus confined to an area adjacent to the cathode surface by one or more endless toroidal electron trapping regions, bounded by a tunnel shaped magnetic field.

3.1.2 Magnetron sputtering

When an ion impinges on the cathode surface several processes may take place: the incident ion will either be implanted or reflected, probably as a neutral and with a large loss of energy; a target atom may be ejected (sputtered); the ion impact and the resulting cascade will cause an amount of structural reordering in the surface layers and secondary electrons may be ejected. Of the mentioned processes the most important for the thin film formation is the sputtering process.

The sputtering yield, S, is defined as the number of atoms ejected from the target per incident particle. Important factors that affect the sputtering yield include surface structure, mass of the bombarding ion and incident energy. The sputtering process is rather insensitive to temperature, and in certain cases, there is even a decrease in S with increasing target temperature [104]. Near the threshold energy, the sputtering yield rises rapidly with ion energy increase.

From about 100 eV upwards, the sputtering yield increases almost linearly with ion energy, where the values begin to be large enough to be useful for film deposition. With further increase in ion energy the sputtering yield reaches a maximum and starts to decrease, mainly due to the larger penetration depth of the ions within the target surface. For light ions, such as hydrogen or helium, the maximum of S is reached at a few thousand eV, since this particle penetrate rather easily. For heavy ions, such as xenon or mercury, the maximum may not be attained until values near 50 KeV or higher. The sputtering yield dependence on the atomic number of the target atoms for various inert gas ions shows some kind of periodicity.

A sputtering yield increase is observed, for instance, when the target material changes from group III-B to group I-B transition metals, but a steep decrease follows from a metal of group I-B to the metal of group III-B of the next period (row of the periodic table). This can be qualitatively understood looking at the electronic structure of transition metals. As the d-orbitals are filled by electrons, collisions among atoms within the solid become more elastic, this results in more efficient energy transfer. On the other hand the bond energies between the atoms become weaker and thus the removal of the atoms is easier. Neutral gas atoms or molecules may also initiate target material sputtering, although in plasma-solid interactions these processes have lesser importance due to the extremely small momentum of the neutral particles.

3.1.3 Growth kinetics of thin film

In sputtering deposition as in other standard vacuum deposition processes, the material arrives at the substrate mostly in an atomic or molecular form. Using the kinetic theory of gases, it is possible to estimate the frequency with which gas particles impinge on a surface, υ, when the gas phase pressure is

P

…. (3.1)

In equation 3.1, m is the mass of the gas particles, K the Boltzmann constant and T the temperature. Considering that for air the mean particle mass is 4.8 x

10-23 g, collision frequency will be approximately 3 x 1024 cm-2 s-1 at 25 0C and

1 atm. Given that a perfectly smooth surface of 1 cm3 has about 1015 atoms, when immersed in a gas at 1 atm pressure, each atom on the surface will be hit about 109 times each second. At 5 x 10-3 mbar, collision frequency will be reduced to 1.5 x 1019 cm-2 s-1 which is still a very high frequency. In the case of sputtering the particles that will interest to the film growth will be the ones evaporated from the target and these will have a much lower collision frequency.

The condensed particles may diffuse around the surface, with a motion determined by their binding energy to the substrate, may be incorporated into the lattice or evaporate. Given the high collision frequency of the gas particles inside the deposition chamber these will have a non-negligible influence on the adsorbed particles diffusion. The diffusion process may lead to adsorption, particularly at special site edges or other defects, or the diffusing particle may desorb. During these processes, characteristic activation energies have to be overcome. The corresponding activation energies for adsorption or diffusion depend on the atomic details of each process. Besides adsorption and surface diffusion, nucleation of more than one adsorbed particle might occur.

Interdiffusion of adsorbed particles with the substrate often happen leading to film-substrate interface smoothing. In thermodynamic equilibrium all processes take place in two opposite directions at equal rates. Therefore, in equilibrium no net film growth would be observed, so that layer growth must be a non- equilibrium kinetic process.

The final macroscopic state of the system may not be the most stable one, since it is kinetically determined. In general however, certain parts of the overall process may be kinetically forbidden, whereas others may be in local thermodynamic equilibrium. In this case equilibrium arguments may be applied locally even though the whole growth process is a non-equilibrium one. Given this fact, a global theory of film growth requires a description in terms of rate equations for each of the processes taking place at the surface. Instead of following a more theoretical atomistic approach it is possible to consider the film growth mechanism using a more phenomenological perspective.

Usually, three distinct modes of film growth may be considered: layer by layer growth mode or Frank-Vander Merve mode; island growth mode or

Vollmer-Weber mode and layer-plus island growth mode, that is also called

Stransky-Krastanov mode; each named after the -investigators associated with their initial description. In layer by layer growth mode the interaction between the substrate and the layer atoms is stronger than between neighbouring atoms.

Each new layer starts to grow, only when the last one is completed. If, on the contrary, the interaction between neighboring atoms is stronger than the overlayer-substrate interaction, the particles will rather form aggregates over the surface that grow in size and eventually coalesce during film growth. This makes up the island growth mode. The layer-plus-island growth mode is an intermediate case where the film starts to grow layer by layer in a first stage and only afterwards begins the formation of island agglomerates. In island growth mode, each island is usually a single crystal or contains just a few crystals.

On a polycrystalline or amorphous substrate, the orientation of each island will be random and the resultant film will be polycrystalline. On single crystal substrates, the islands orientations may be constrained to a given direction by the substrate structure, so that growth and coalescence leads to the formation of a single crystal film. This case is usually known as epitaxy. If surface atoms have high mobility, they have greater opportunity of finding low energy positions consistent with crystal growth. Knowing that mobility is increased by surface temperature, it is expected that higher substrate temperature will promote crystal growth. The same effect can be achieved by reducing the deposition rate, which gives more time to the adsorbed species to find an energetically favourable lattice position. Epitaxial growth was also found to be promoted by electron or ion bombardment and increased energy of deposited atoms [100]. The environmental conditions around the substrate during magnetron sputtering deposition deserve also to be mentioned, since they will inevitably affect the film structure obtained. Namely, it was ignored in the above description the effect of the energy of the impinging sputtered atoms, as well as the effect of many other particles that may impact on the surface. In first place there might be contaminants arriving at the substrate. These may result from an internal source, such as outgassing from a heating substrate, or by an external source such as the sputtering gas. If the contaminant atoms are chemically active, the contamination will be particularly effective and can only be minimized by reducing the contaminant partial pressure. If the contaminant source is outgassing, the system may have to be evacuated to a higher vacuum or the sources of outgassing heated in order to reduce the outgassing rate and guarantee a sufficiently low partial pressure of the contaminant during deposition. On the other hand, if the contaminant comes with the sputtering gas there is no way to reduce its partial pressure without affecting simultaneously the sputtering gas pressure. Thus it is very important to use high purity sputtering gases.

The sputtering gas atoms might also become part of the deposited film.

Although the sputtering gas used is usually an inert gas, given its high partial pressure when compared with that of the sputtered atoms it is not surprising that some of these atoms happen to be trapped in the growing film. Inert gas atoms, are only expected to be physisorbed, so if substrate temperature is increased during deposition, they are less likely to be adsorbed and at the same time more likely to be subsequently desorbed. However, there exists an appreciable difference between the interaction of fast gas particles and that of thermal neutrals with the surface. Energetic neutrals may result from the reflection and neutralization of ions that impinge on the target. These neutrals, arriving at the substrate, have much higher probability to be embedded in the growing film than thermal neutrals. In fact, it has been observed that nickel films deposited by sputtering in an argon atmosphere have higher argon content than nickel films produced in a similar atmosphere, but using evaporation [105]. Positive ions may also impact on the substrate due to the sheath voltage drop near its surface.

Negative ions and electrons can only reach the substrate if they have enough energy to cross the space-charge sheath. Once again it is necessary to distinguish fast from slow particles. Fast negative ions and electrons are produced in the target and accelerated in the target sheath. These fast particles can have a major influence on the structure and properties of the growing film and also cause substrate heating. Finally there are also photons arriving at the substrate. Photons can be produced by ion or electron bombardment on any surface or result from relaxation of excited atoms in the glow. In the former case the photons may have high energy, at most as high as the energy of the bombarding particle. Such energies may be, in a sputtering system, of the same order of the accelerating potential at the target, which is usually higher than 200 eV. The main effect of photon bombardment of the substrate will be electron emission, which may also affect the film growth processes occurring at the substrate.

3.1.4 Vacuum coating unit

A vacuum coating unit consists of vacuum chamber, rotary pump, diffusion pump and penning guage.

(a) Rotary oil pump

This pump is most commonly used to produce moderately low pressures of the order of 10-3 torr. It is widely used in laboratories in vacuum coating units. The working is based on the principle of displacement of a gas. A portion of the gas inside a chamber is isolated by a rotating disc and is then compressed till it attains a sufficiently high pressure and gets discharged to the atmosphere.

As a gas consists of fast moving molecules, the residual gas comes and occupies the whole space and in turn, also gets removed. This process continues and the pressure of the chamber keeps on decreasing. The rotary pump consists of a solid cylinder R called Rotor rotating eccentrically inside a hollow cylinder S called Stator (Figure 3.1). The stator has an inlet I connected to the chamber to be evacuated and an outlet O which always open outward only. The rotor is rotated by an electric motor. The space inside the stator S is divided into two air-tight compartments by means of a vane V between the inlet and the outlet.

The vane is always tightly pressed against the rotor by means of a spring P. To avoid any leakage, the whole pump is kept immersed in a special of oil called vacuum oil or hydrocarbon oil. The working of the rotary pump can be explained by analyzing, four successive positions of the rotor R as shown in Figure 3.1. In the first position

Figure 3.1(a) the air or gas from the chamber just enters into the space in between the cylinder and the rotor R. As it rotates, the space inside the stator is divided into two parts as in Figure 3.1(b). As the rotor rotates further the volume of the outlet portion decreases and hence air inside these is further compressed. Figure 3.1(c) shows the final stage of compression which opens the valve at the outlet O and the gas is expelled out. Meanwhile, the inlet volume increases, and hence air enters from the chamber. Thus during each rotation of the cylinder the air is drawn from the chamber and is removed from it continuously. Within a few minutes, the pump can produce a low pressure of 10-

3 torr in the chamber.

Fig 3.1(a) Fig 3.1(b) Fig 3.1(c) Fig 3.1 Schematic diagram of rotary pump

(b) Diffusion Pump

These pumps can produce very low pressure down to 10-7 torr. It cannot operate directly from the atmospheric pressure but requires force vacuum or backing vacuum from 10-1 to 10-2 torr. The basic principle used in these pumps is that in a mixture of gases, the diffusion of gas occurs from a region where its partial pressure is higher to that where it is lower, irrespective of the total pressure in the two regions. The diffusion pump is made up of stainless steel and it consists of a cylindrical body having a boiler at the base with pumping fluid (Silicone DC 704). A number of nozzle (combination of nozzles and jet) assembles are arranged inside the cylinder. Various nozzles originate at different parts of the boiler and are designed such that the pressure rise across the various jet streams increases progressively as one goes from the vacuum chamber side to the backing pump side. The temperature of the walls is usually maintained few degrees below room temperature by water cooled coils.

Figure 3.2 shows the schematic diagram of a diffusion pump. The vapour from the boiling fluid passes up the cylindrical chimney. It jets out through the nozzle with high speed being directed at an angle downwards. Molecules of gas which wander are diffused from the chamber towards the jet stream and will be struck by vapour molecules of the heated pumping fluid. The molecules are given a downward motion into the dense part of the vapour jet. The net effect is to compress the gas to the point where they can be removed by a backing pump.

The compression is carried out by stages by using several nozzles in line. As each stage handle the same number of molecules, the pressure decreases from the top most nozzle in the last, the annular spacing between the nozzle and the pump casing increases in the same direction. The pump fluid must be removed to the boiler after it has performed its function and hence the water cooling of the pump casing.

Figure 3.2 Schematic diagram of a diffusion pump

(c) Penning gauge

The penning gauge is widely used for Ultra High Vacuum (UHV) pressure measurements. In the penning gauge, the filament has dispensed width.

It consists of three electrodes kept inside a bulb and are connected to air tight seal. This bulb is placed inside a permanent magnet. The opening of the bulb is connected to the vessel whose pressure is to be measured. The electric field between the electrodes is verified due to the short distance of separation between them. Here the electrons are produced by field emission. The magnetic field is applied perpendicular to the electric field. So the electrons travel in an elliptical path. These electrons produce more ionization due to their collision with the gas molecules. The ion current is registered by the milli ammeter in the external electric circuit. This ion current is the measure of the pressure.

Figure 3.3 Schematic diagram of penning gauge

Figure 3.3 shows the schematic diagram of the penning gauge. The electrons are generated at the cathode and are accelerated towards the anode.

They pass through the anode ring and are then repelled by the other cathode and they oscillate back and forth eventually winding up the anode ring. This process will lead to ionization by collision. These ions impinge upon the cathode releasing electrons which will maintain the discharge. The effective path of the electron increases since they take a helical path due to the magnetic field. The long paths of the electrons, result in the production of many more ions than if they were able to go directly from the cathode to anode. The positive ions thus produced are collected by the cathode and the ion current is measured. The ion current is proportional to the gas pressure in the chamber. The micro ammeter is calibrated in terms of pressure. (d) Vacuum chamber

The schematic of the vacuum coating unit is shown in Figure 3.4. The chamber is pumped to vacuum using a diffusion pump which is able to reduce the pressure below 10-3 Pa. If the chamber is kept open during a long time period it will take longer to reach high vacuum pressures because of outgassing.

Heating the chamber walls often accelerates evacuation, by promoting outgassing. The diffusion pump is backed by a rotary pump that is also used to do primary vacuum inside the chamber before connecting the diffusion pump.

In order to change the substrates the chamber has to be open, so that usually the frequency of experiments is dependent on the time required to reach high pressures inside the chamber. The sputtering gases are fed into the chamber using two mass flow meters. One is calibrated for argon, other for oxygen. The maximum flow reached by each mass flow meter is 200 sccm for the argon device, and 20 sccm for N2. The control and measurement is performed using a power supply and readout system that permits to set the flow of the mass flow controller. The substrates are positioned on a stainless steel holder, with suitable holes that work as masks for the deposition.

Figure 3.4 Schematic diagram of vacuum coating unit

The substrate holder can be fixed at different distances above the target and is maintained at the same potential as the anode and the chamber walls. It is possible to insulate the substrate holder electrically in order to apply a voltage bias to control the charged particles that impinge on the substrates. The substrate heating is performed using a metallic base weaved with an electrical resistance. The temperature is measured with a thermocouple inserted into a hole made in one side of the metallic base and controlled by a temperature controller.

The setup used to heat and measure the substrate temperature does not make possible to know the actual substrate temperature, because there is an unknown temperature gradient between the substrate surface and the point where the thermocouple is placed. However, if given enough time for the heat transfer fluxes to stabilize, the temperature conditions can be reasonably reproduced in repeated experiments and the effect of temperature upon the films‘ characteristics may be studied. Nevertheless, if the substrates pressure or the sputtering power is changed, for instance, it is not possible to maintain a fixed temperature on the substrate over repeated experiments, since the temperature gradient is very likely to change.

Figure 3.5 shows the photograph of the dc/rf magnetron sputtering unit.

The dc power source is able to operate at a maximum voltage of 500V and to deliver a maximum current of 5A. The user may set the voltage, the current and the power between 5% and 100% of respective range. The push on button arrangement with an LED display provision on the unit was used both to set the source parameters and read the voltage and current values during operation. Rf frequency of 13.56 Hz is the maximum frequency prescribed for the unit. Peak to peak voltage greater than 1000V, current densities of 1 mA/cm2, discharge pressure of 0.5 to 10 m torr are the typical value range for rf magnetron sputtering. Pressure inside the chamber is measured using a Pirani gauge in the pressure range from 105 Pa to 10-1 Pa and using a Penning gauge in ranges from 10-1 Pa to 10-5 Pa. The Penning gauge is cleaned periodically to ensure repeatability. If the electrodes of this device are covered by an oxide layer the measurements stray toward lower pressures.

Figure 3.5 Photograph of the dc/rf sputtering unit

3.2 ANALATICAL INSTUREMENTS AND TECHNIQUE [106]

3.2.1 Rutherford backscattering spectrometry (RBS)

Rutherford Backscattering Spectrometry (RBS) is a widely used nuclear method for the near surface layer analysis of solids. A target is bombarded with ions at an energy in the MeV-range (typically 0.5–4 MeV), and the energy of the backscattered projectiles is recorded with an energy sensitive detector, typically a solid state detector. RBS allows the quantitative determination of the composition of a material and depth profiling of individual elements. RBS is quantitative without the need for reference samples, nondestructive, has a good depth resolution of the order of several nm, and a very good sensitivity for heavy elements of the order of parts-per-million (ppm). The analyzed depth is typically about 2 μm for incident He-ions and about 20 μm for incident protons.

An RBS instrument generally includes three essential components:

2+ An ion source, usually alpha particles (He ions) or, less commonly,

protons.

A linear particle accelerator capable of accelerating incident ions to high

energies, usually in the range 1-3 MeV.

A detector capable of measuring the energies of backscattered ions over

some range of angles.

Figure 3.6 shows the photograph of the RBS instrument used to take the

RBS spectra. Tandem accelerator starts with a source of He- ions and position the positive terminal at the center of the acceleration tube. A stripper element included in the positive terminal removes electrons from ions which pass through, converting He- ions to He++ ions. The ions thus start out being attracted to the terminal, pass through and become positive, and are repelled until they exit the tube at ground. A typical tandem accelerator with an applied voltage of

750 kV can achieve ion energies of over 2 MeV.

Figure 3.6 Photograph of Rutherford backscattering spectrometer Ions which reach the detector lose some of their energy to inelastic scattering from the electrons, and some of these electrons gain enough energy to overcome the band gap between the semiconductor valence and conduction bands. This means that each ion incident on the detector will produce some number of electron-hole pairs which is dependent on the energy of the ion. The energy loss of a backscattered ion is dependent on two processes: the energy lost in scattering events with sample nuclei, and the energy lost to small-angle scattering from the sample electrons. The first process is dependent on the scattering cross-section of the nucleus and thus on its mass and atomic number.

For a given measurement angle, nuclei of two different elements will therefore scatter incident ions to different degrees and with different energies, producing separate peaks on an N(E) plot of measurement count versus energy. These peaks are characteristic of the elements contained in the material, providing a means of analyzing the composition of a sample by matching scattered energies to known scattering cross-sections. Relative concentrations can be determined by measuring the heights of the peaks.

3.2.2 X-ray Powder Diffraction

X-ray powder diffraction (XRD) is a rapid analytical technique primarily used for phase identification of a crystalline material and can provide information on unit cell dimensions. The analyzed material is finely ground, homogenized, and average bulk composition is determined. Max von Laue, in

1912, discovered that crystalline substances act as three-dimensional diffraction gratings for X-ray wavelengths similar to the spacing of planes in a crystal lattice. X-ray diffraction is now a common technique for the study of crystal structures and atomic spacing. X-ray diffraction is based on constructive interference of monochromatic X-rays and a crystalline sample. These X-rays are generated by a cathode ray tube, filtered to produce monochromatic radiation, collimated to concentrate and directed towards the sample. The interaction of the incident rays with the sample produces constructive interference (and a diffracted ray) when conditions satisfy Bragg's Law (nλ= 2d sinθ) (figure 3.7). This law relates the wavelength of electromagnetic radiation to the diffraction angle and the lattice spacing in a crystalline sample. These diffracted X-rays are then detected, processed and counted. By scanning the sample through a range of 2θ angles, all possible diffraction directions of the lattice should be attained due to the random orientation of the powdered material. Conversion of the diffraction peaks to d-spacings allows identification of the mineral because each mineral has a set of unique d-spacings. Typically, this is achieved by comparison of d-spacings with standard reference patterns.

Figure 3.7 The geometry of the incident X-rays impinging the sample satisfies the Bragg Equation

X-ray diffractometer consist of three basic elements: an X-ray tube, a sample holder, and an X-ray detector. X-ray are generated in a cathode ray tube by heating a filament to produce electrons, accelerating the electrons toward a target by applying a voltage, and bombarding the target material with electrons.

When electrons have sufficient energy to dislodge inner shell electrons of the target material, characteristic X-ray spectra are produced. These spectra consist of several components, the most common being Kα and Kβ. Kα consists, in part of Kα1 and Kα2. Kα1 has a slightly shorter wavelength and twice the intensity as

Kα2. The specific wavelengths are characteristic of the target material (Cu, Fe, Mo, Cr). Filtering, by foils or crystal monochrometers, is required to produce monochromatic X-rays needed for diffraction. Kα1and Kα2 are sufficiently close in wavelength such that a weighted average of the two is used. Copper is the most common target material for single-crystal diffraction, with CuKα radiation

= 1.5418 Å. These X-rays are collimated and directed onto the sample. As the sample and detector are rotated, the intensity of the reflected X-rays is recorded.

When the geometry of the incident X-rays impinging the sample satisfies the

Bragg equation, constructive interference occurs and a peak in intensity occurs.

A detector records and processes this X-ray signal and converts the signal to a count rate which is then out putted to a device such as a printer or computer monitor.

The geometry of an X-ray diffractometer is such that the sample rotates in the path of the collimated X-ray beam at an angle θ while the X-ray detector is mounted on an arm to collect the diffracted X-rays and rotates at an angle of 2θ.

The instrument used to maintain the angle and rotate the sample is termed a goniometer. For typical powder patterns, data is collected at 2θ from ~5° to 70°, angles that are preset in the X-ray scan. Figure 3.8 shows the schematic of the detection of diffracted X-rays by diffractometer. The photograph of the XRD unit used for recording the XRD patterns is shown in Figure 3.9.

Figure 3.8 Schematic of the detection of diffracted X-rays by diffractometer

Figure 3.9 Photograph of X-ray diffraction unit

3.2.3 Scanning electron microscopy

The scanning electron microscope (SEM) uses a focused beam of high- energy electrons to generate a variety of signals at the surface of solid specimens. The signals that derive from electron-sample interactions reveal information about the sample including external morphology (texture), chemical composition, and crystalline structure and orientation of materials making up the sample. In most applications, data are collected over a selected area of the surface of the sample, and a 2-dimensional image is generated that displays spatial variations in these properties. Areas ranging from approximately 1 cm to

5 microns in width can be imaged in a scanning mode using conventional SEM techniques (magnification ranging from 20X to approximately 30,000X, spatial resolution of 50 to 100 nm).

Accelerated electrons in an SEM carry significant amounts of kinetic energy, and this energy is dissipated as a variety of signals produced by electron-sample interactions when the incident electrons are decelerated in the solid sample. These signals include secondary electrons (that produce SEM images), backscattered electrons (BSE), diffracted backscattered electrons

(EBSD that are used to determine crystal structures and orientations of minerals), photons (characteristic X-rays that are used for elemental analysis and continuum X-rays), visible light (cathode luminescence--CL), and heat.

Secondary electrons and backscattered electrons are commonly used for imaging samples: secondary electrons are most valuable for showing morphology and topography on samples and backscattered electrons are most valuable for illustrating contrasts in composition in multiphase samples (i.e. for rapid phase discrimination). X-ray generation is produced by inelastic collisions of the incident electrons with electrons in discrete orbital‘s (shells) of atoms in the sample. As the excited electrons return to lower energy states, they yield X- rays that are of a fixed wavelength (that is related to the difference in energy levels of electrons in different shells for a given element). Thus, characteristic

X-rays are produced for each element in a mineral that is "excited" by the electron beam. SEM analysis is considered to be "non-destructive"; that is, x- rays generated by electron interactions do not lead to volume loss of the sample, so it is possible to analyze the same materials repeatedly.

Essential components of all SEM include the following:

Electron Source ("Gun")

Electron Lenses

Sample Stage

Detectors for all signals of interest

Display / Data output devices

Infrastructure Requirements

Power Supply

Vacuum System

Cooling system Vibration-free floor

Room free of ambient magnetic and electric fields

In a typical SEM, an electron beam is thermionically emitted from an electron gun fitted with a tungsten filament cathode. Tungsten is normally used in thermionic electron guns because it has the highest melting point and lowest vapour pressure of all metals, thereby allowing it to be heated for electron emission, and because of its low cost. The electron beam, which typically has an energy ranging from 0.2 keV to 40 keV, is focused by one or two condenser lenses to a spot about 0.4 nm to 5 nm in diameter. The beam passes through pairs of scanning coils or pairs of deflector plates in the electron column, typically in the final lens, which deflect the beam in the x and y axes so that it scans in a raster fashion over a rectangular area of the sample surface. When the primary electron beam interacts with the sample, the electrons lose energy by repeated random scattering and absorption within a teardrop-shaped volume of the specimen known as the interaction volume, which extends from less than

100 nm to around 5 µm into the surface.

The size of the interaction volume depends on the electron's landing energy, the atomic number of the specimen and the specimen's density. The energy exchange between the electron beam and the sample results in the reflection of high-energy electrons by elastic scattering, emission of secondary electrons by inelastic scattering and the emission of electromagnetic radiation, each of which can be detected by specialized detectors. The beam current absorbed by the specimen can also be detected and used to create images of the distribution of specimen current. Electronic amplifiers of various types are used to amplify the signals, which are displayed as variations in brightness on a cathode ray tube. The raster scanning of the CRT display is synchronized with that of the beam on the specimen in the microscope, and the resulting image is therefore a distribution map of the intensity of the signal being emitted from the scanned area of the specimen. The image may be captured by photography from a high-resolution cathode ray tube, but in modern machines is digitally captured and displayed on a computer monitor and saved to a computer's hard disk.

Figure 3.11 shows the photograph of the Scanning electron microscopy used to take SEM image.

Figure 3.10 Schematic diagram of the scanning electron microscope.

Figure 3.11 Photograph of scanning electron microscope 3.2.4 Fourier Transform Infrared spectrophotometer FT-IR stands for Fourier Transform Infrared, the preferred method of infrared spectroscopy. In infrared spectroscopy, IR radiation is passed through a sample. Some of the infrared radiation is absorbed by the sample and some of it is passed through (transmitted). The resulting spectrum represents the molecular absorption and transmission, creating a molecular fingerprint of the sample.

Like a fingerprint no two unique molecular structures produce the same infrared spectrum. This makes infrared spectroscopy useful for several types of analysis.

The interferometer produces a unique type of signal which has all of the infrared frequencies ―encoded‖ into it. The signal can be measured very quickly, usually on the order of one second or so. Thus, the time element per sample is reduced to a matter of a few seconds rather than several minutes. Figure 3.12 and Figure

3.13 shows the schematic diagram and photograph of FTIR-interferometer system used to take the FTIR spectra.

Most interferometers employ a beamsplitter which takes the incoming infrared beam and divides it into two optical beams. One beam reflects off of a flat mirror which is fixed in place. The other beam reflects off of a flat mirror which is on a mechanism which allows this mirror to move a very short distance

(typically a few millimeters) away from the beam splitter.

Figure 3.12 Schematic diagram of FTIR-spectrometer system

Figure 3.13 Photograph of Fourier transform infrared spectrometer

The two beams reflect from their respective mirrors and are recombined when they meet back at the beamsplitter. Because the path that one beam travels is a fixed length and the other is constantly changing as its mirror moves, the signal which exits the interferometer is the result of these two beams

―interfering‖ with each other. The resulting signal is called an interferogram which has the unique property that every data point (a function of the moving mirror position) which makes up the signal has information about every infrared frequency which comes from the source. This means that as the interferogram is measured; all frequencies are being measured simultaneously. Thus, the use of the interferometer results in extremely fast measurements.

Because the analyst requires a frequency spectrum (a plot of the intensity at each individual frequency) in order to make identification, the measured interferogram signal cannot be interpreted directly. A means of ―decoding‖ the individual frequencies is required. This can be accomplished via a well-known mathematical technique called the Fourier transformation. This transformation is performed by the computer which then presents the user with the desired spectral information for analysis. The normal instrumental process is as follows:

Source: Infrared energy is emitted from a glowing black-body source.

This beam passes through an aperture which controls the amount of

energy presented to the sample (and, ultimately, to the detector). Interferometer: The beam enters the interferometer where the ―spectral encoding‖ takes place. The resulting interferogram signal then exits the interferometer.

Sample: The beam enters the sample compartment where it is transmitted through or reflected off from the surface of the sample, depending on the type of analysis being accomplished. This is where specific frequencies of energy, which are uniquely characteristic of the sample, are absorbed.

Detector: The beam finally passes to the detector for final measurement.

The detectors used are specially designed to measure the special interferogram signal.

Computer: The measured signal is digitized and sent to the computer where the Fourier transformation takes place. The final infrared spectrum is then presented to the user for interpretation and any further manipulation.

Because there needs to be a relative scale for the absorption intensity, a background spectrum must also be measured. This is normally a measurement with no sample in the beam. This can be compared to the measurement with the sample in the beam to determine the ―percent transmittance.‖ This technique results in a spectrum which has all of the instrumental characteristics removed. Thus, all spectral features which are present are strictly due to the sample. A single background measurement can be used for many sample measurements because this spectrum is

characteristic of the instrument itself.

3.2.5 Ultraviolet - Visible spectrophotometer

Molecules containing π-electrons or non-bonding electrons (n-electrons) can absorb the energy in the form of ultraviolet or visible light to excite these electrons to higher anti-bonding molecular orbitals [106] . The more easily excited the electrons (i.e. lower energy gap between the HOMO and the

LUMO) the higher the wavelength of light it can absorb. The instrument used in ultraviolet-visible spectroscopy is called a UV/Vis spectrophotometer. The UV- visible spectrophotometer is configured to measure reflectance. In this case, the spectrophotometer measures the intensity of light reflected from a sample (I), and compares it to the intensity of light reflected from a reference material (Io).

The ratio I / Io is called the reflectance, and is usually expressed as a percentage

(%R). The basic parts of a spectrophotometer are a light source, a holder for the sample, a diffraction grating in a monochromator or a prism to separate the different wavelengths of light, and a detector. The radiation source is often a

Tungsten filament (300-2500 nm), a deuterium arc lamp, which is continuous over the ultraviolet region (190-400 nm), Xenon arc lamps, which is continuous from 160-2,000 nm; or more recently, light emitting diodes (LED) [100] for the visible wavelengths. The detector is typically a photomultiplier tube, a photodiode, a photodiode array or a charge-coupled device (CCD). Single photodiode detectors and photomultiplier tubes are used with scanning monochromators, which filter the light so that only light of a single wavelength reaches the detector at one time. The scanning monochromator moves the diffraction grating to ―step-through‖ each wavelength so that its intensity may be measured as a function of wavelength. Fixed monochromators are used with CCDs and photodiode arrays. As both of these devices consist of many detectors grouped into one or two dimensional arrays, they are able to collect light of different wavelengths on different pixels or groups of pixels simultaneously. In a double-beam instrument, the light is split into two beams before it reaches the sample. One beam is used as the reference; the other beam passes through the sample. The reference beam intensity is taken as 100%

Transmission (or zero Absorbance), and the measurement displayed is the ratio of the two beam intensities. Some double-beam instruments have two detectors

(photodiodes), and the sample and reference beam are measured at the same time. In other instruments, the two beams pass through a beam chopper, which locks one beam at a time. Figure 3.14 shows the schematic diagram of UV- visible spectrometer.

Figure 3.14 Schematic diagram of UV-VIS spectrophotometer

The detector alternates measuring between the sample beam and the reference beam in synchronism with the chopper. There may also be one or more dark intervals in the chopper cycle. In this case, the measured beam intensities may be corrected by subtracting the intensity measured in the dark interval before the ratio is taken. Samples are typically placed in a transparent cell, known as a cuvette. Cuvettes are typically rectangular in shape, commonly with an internal width of 1 cm. (This width becomes the path length, L, in the

Beer-Lambert law.) Test tubes can also be used as cuvettes in some instruments.

The type of sample container used must allow radiation to pass over the spectral region of interest. The most widely applicable cuvettes are made of high quality fused silica or quartz glass because these are transparent throughout the UV, visible and near infrared regions. Glass and plastic cuvettes are also common, although glass and most plastics absorb in the UV, which limits their usefulness to visible wavelengths. Figure 3.15 shows the photograph of UV-Visible spectrometer used to take the transmission and absorbance spectra.

Figure 3.15 Photograph of UV-Visible spectrophotometer 3.2.6 LCR Meter

Impedance measurements are a basic means of evaluating electronic components and materials. Every material has a unique set of electrical characteristics that are dependent on its dielectric or insulation properties.

Accurate measurements of these properties can provide valuable information to ensure an intended application or maintain a proper manufacturing process.

When making these measurements, connection of the material to the measuring instrument (an LCR Meter) is one of the major challenges faced, special fixtures are generally required depending on the material type. The most common piece of test equipment for holding a variety of solid materials is the LD-3, a liquid

tight, three terminal connection cell with electrode spacing adjustable by a

precision micrometer shown in Figure 3.16.

Figure 3.16 Photograph of Precision LCR Meter

Specimen Air Cxm and Dxm Ca and Da (a) (a) (b)

Figure 3.17 Schematic diagram of LCR measurement set up (a) with specimen (b) without specimen

First the sample is inserted in the cell and the electrodes closed with the micrometer until they just touch the sample lightly, Fig 3.17 (a). The micrometer spacing hm is then recorded for this specimen. The capacitance value, Cxm and dissipation factor, Dxm are then measured with the LCR meter.

The specimen is then removed from the cell, micrometer readjusted to and the measurements repeated in air as Ca and Da, Figure 3.17(b).

CHAPTER 4

PREPARATION AND CHARACTERIZATION OF INDIUM TIN OXIDE THIN FILMS

Indium tin oxide (ITO) is a well-known transparent conducting oxide thin film used in various opto-electronic device fabrication technologies, due to its high optical transmittance and low electrical resistivity [107-109]. So far, various deposition methods have been used to deposit ITO films and extensive studies have been carried out to understand the correlation between the deposition parameters and film properties [107-112]. Studies on the structural properties of ITO films deposited by various methods such as evaporation, reactive thermal deposition, spray pyrolysis, sputtering and e-beam evaporation pointed out that, there is a change in the type of crystallographic planes that are preferentially parallel to the film surface [(222), (400) and (440) planes], depending on the deposition conditions [107-115]. The reason for this change was proposed by many researchers such as: oxygen vacancies and its induced stressed growth in reactively deposited ITO films [116]; difference in the energy of the sputtered atoms in RF magnetron sputtered ITO films [117]; the ‗Sn‘ content of the film etc. [118]. Since no change in the basic crystalline structure has been identified in relation to change in crystalline texture, further investigation is needed in order to understand the crystal growth phenomena, which could be useful for particular application requirements [119,120]. Since the early 1960s, gas-sensitive resistors have existed commercially with a demonstrated capability to detect combustible gases at the ppm (parts per million) level. However, recent thin film gas sensors demonstrate great promise at reliably detecting concentrations at the ppb (parts per billion). A number of semi-conducting oxides such as ZnO, SnO2 and In2O3 have been used for different gas sensors [121-123]. Most of these gas sensors are based on a variation in resistance, when the film sensors are exposed to target gases specifically; the use of ITO follows as the most recent solid-state gas sensor development [124]. Recent research has shown that the transparent and conducting Sn doped In2O3 (ITO) offers new advantages in the design of metal oxide-based gas sensors [125]. ITO has high gas sensitivity, because its conductivity increases remarkably when exposed to reducing gases.

In the present chapter, the optimization of ITO films have been discussed in detail. In order to develop a metal oxide thin film gas sensor, ITO films have been deposited onto glass substrate by rf magnetron sputtering. The dependence of structural, optical and electrical properties on the deposition condition is investigated. Finally, optical, electrical and structural properties of the ITO films were studied. Transmission was measured with Perkin Elmer Lambda 40

UV/VIS spectrometer, sheet resistance with four point probe, and thickness by

Mitutoyo surface profilometer. The crystallographic orientation of the ITO films was determined by an X'pert PRO X-ray diffractometer (XRD) using Cu Kα irradiation. The morphology of ITO films was investigated by Hitachi S-3000H scanning electron microscopy (SEM) operating at 20 kV. The effect of rf sputtering power and annealing on the crystalline orientation, surface morphology, electrical and optical parameters have been analyzed to obtain the best suitable film for ITO sensor device fabrication.

4.1 OPTIMIZATION OF INDIUM TIN OXIDE THIN FILMS

There are a few parameters that can affect the ITO properties. These parameters are the power distributed onto the target, the gas flow (argon and oxygen), the distance between the anode and cathode, the base pressure, annealing temperature, annealing time and the deposition time. It is a tedious task to do design of experiment with all seven parameters. To simplify the optimization process, the parameter number was reduced. Some initial deposition process showed that the base pressure did not affect the film properties significantly. So the base pressure was held at 1.33 X10-3 Pa prior to all the sputtering processes. The distance between the anode and the cathode is fixed; oxygen and argon flow and deposition time was also kept constant for this system.

The sensitivity, selectivity and response of the thin films to gas strongly depend on the conditions of deposition, which determine the structure, the crystallinity and the composition [126,127]. Along with other parameters, the rate of deposition, substrate temperature and the nature of the substrate are perhaps the most critical parameters determining the structure and properties of thin solid films produced by techniques of vacuum evaporation and sputtering.

Therefore to optimize the growth parameters during deposition, in order to get the reproducible film and, facilitate the fabrication of device, a study on growth and structural properties of ITO thin films for different rf power has been carried out.

ITO thin films were prepared by rf magnetron sputtering. The ITO was prepared using In2O3 containing 15% SnO2 (Alfa Aesar product). The ITO films were deposited on Corning 7059 glass substrates to study the optical and electrical properties of the film. The distance between target and substrate was approximately 7 cm. Initially, all the substrates were ultrasonically cleaned in isopropyl alcohol (propan-2-ol) and dried to remove all the organic contaminants. A base pressure of 1.33 X10-3 Pa was maintained prior to the deposition. Before sputtering, the target was presputtered for about 10 minutes with a shutter covering the target, in order to remove the surface oxide layer.

Pre-sputtering of ITO target play an important role for preparation of homogeneous films. Then, the shutter was opened and the deposition process was started. The deposition process was carried out under an oxygen partial pressure of 1 Pa using mass flow controller. The melting point of indium tin target is low, so the power should be less than 200 W to avoid target melting.

However, if the power is too low, the deposition rate is expected to be low, since the deposition rate is strongly related with the rf power. Hence, the rf power was varied from 50 to 200 W, in steps of 50 W. V.S. Vaishnav et al. [128] has studied the variation of substrate temperature for the ITO film. They have reported that the crystallite size increases with increase in substrate temperature up to 648 K. They found that at a temperature of 648 K, the crystallite size was maximum. The electron diffraction pattern showed a diffused ring pattern at low temperature and became sharper with increase in temperature, at 648 K displayed a large number of distinct and well- defined rings, proving the polycrystalline nature of ITO film. Dutta and Ray

[129] have studied the variation in structural properties of magnetron sputtered indium tin oxide films with deposition parameters. They found that the degree of crystallinity improved and the grain size increased as the substrate temperature was increased. This was attributed to the increase in mobility of the adatoms on the deposition surface at higher temperature. At still higher temperature, the crystallites seemed to be smaller, due to the re-evaporation of the film.

Hence, in the present work substrate temperature of 648 K was chosen as the suitable temperature for the growth of ITO films. The substrate temperature was set at 648 K using a thermocouple and deposition was done for 25 min. The films were subsequently annealed at 700 K for 1 h. The deposition parameters used for the preparation of ITO thin films in the present study are summarized in Table 4.1.

4.1.1 Effect of rf power on thickness and deposition rate of ITO thin films

The thickness measurement of ITO films was performed using SJ 301

Mitutoyo Stylus profilometer. The samples were partially masked in order to obtain a step height between the glass slide and the film. The thickness of indium tin oxide thin films was measured at five different positions and averaged. The film thickness measured for different rf power at 25 minutes of deposition time was calculated, and the results are shown in Table 4.2.

The rf power has a direct influence on the deposition rate and generation of heat due to the electron bombardment intrinsic to this technique [130]. After experimenting with various rf power for ITO deposition, a marked influence was observed on the electrical properties of the deposited films. During each deposition, at various powers, the plasma was ‗tuned‘ to maximize the forward power and reduce the reflected power to zero. Moreover, a reduction in the oxygen flow rate or the oxygen partial pressure, (pO2), in the reactive plasma caused a proportional decrease in the deposition rate at a given rf power. Figure

4.1 shows the ITO deposition rate vs. rf sputtering power, at a pO2 of 1 Pa.

Since, the ITO deposition rate was directly proportional to the rf power at a given chamber pressure and partial pressure of oxygen, the dependence of the film properties on the deposition rate will be the same as their dependence on the rf power.

4.1.2 Effect of rf power on the structure of ITO thin films

XRD analysis were carried out using RIGAKU X-ray diffractrometer employing Cu Kα (1.5406Å) radiation; θ -2θ scans were performed with step size of 0.05, at a scan speed of 3 sec per step in the range of 10-800. The diffraction planes in θ-2θ method are parallel to the sample surface.

Accordingly, this method is very helpful in studying the preferred orientation in thin film materials, when compared to grazing angle incidence method [131].

Figure 4.2 shows the XRD pattern of the indium tin oxide thin films deposited on glass for different rf power, keeping all other deposition parameters constant.

The lattice constant (a) and the crystal grain size (D) were calculated for cubic structure from the (222) and (400) peaks of the XRD spectra. The lattice constant was calculated with the following expression:

= …. (4.1)

Here, h, k, l are the miller indices of the lattice plane, and d is the space between two adjacent (222) or (400) planes. The grain size (D) was calculated by using the Debey-Scherrer formula from the full-width at half maximum (β).

D = …. (4.2) where, k is the constant 0.94, λ is the wavelength of the X-ray used, β is the

FWHM in radian and ‗θ‘ is the diffraction angle obtained from 2θ of the XRD pattern. The dislocation density (δ) and number of crystallite N was determined using the following equation:

δ = …. (4.3)

N = …. (4.4) where, ε is the strain, and t the thickness of the film. The structural parameters for the ITO films deposited at different rf powers are summarized in Table 4.3.

The XRD pattern of the film deposited with 50 W shows no characteristic peaks, but a broad diffraction pattern showing the amorphous nature of the film was observed. The film deposited at 100 W shows a weak (222) plane at

2θ=30.580 (Figure 4.2 (b)), which suggest the initiation of crystallization process at 100 W. At an rf power of 150 W, the intensity of (222) plane increased, whereas the intensity of (400) plane decreased (Figure 4.2 (c)). Further increase in rf power to 200 W resulted in films with less intense (222) and (400) peak

(Figure 4.2 (d)). At lower sputtering power of 50 W, the energy acquired by the sputtering inter Ar+ ions, which are colliding on the target, may be low due to relatively lower discharge voltage. This leads to the generation of sputtered atoms with low energy and consequently, the adatoms mobility gets reduced to

a very low value [132]. Hence, ITO films deposited with 50 W show non crystalline/amorphous structures. With increasing rf power the films acquired

(222) orientation from amorphous structure. The mechanism of ITO film formation by sputtering suggests a two step growth process with varying rf power.

From the above observation it was found that in growing rf sputtered ITO films, (222) preferred orientation is invariably dominant during the initial growth stage under the optimized deposition conditions. This is due to the fact that, as soon as sputtering starts, indium atoms reach the substrate first to initiate nucleation process and then settle down at the lowest energy (111) plane, which is also considered as the densely packed plane of indium lattice with face- centered tetragonal structure. Hence, all the ITO films deposited under various rf power invariably contain (222) oriented peaks, but with varying intensity. The results are in agreement with the values reported by V.V. Vidya et al. [133], who studied the variation of rf power on growth mechanism of ITO films in the range of 50-350 W.

M.J. Chuang et al [134] has investigated the influence of the rf power and working pressure on the properties of indium tin oxide (ITO) thin films, prepared by long-throw rf magnetron sputtering technique at room temperature.

A structural change from amorphous to mixed amorphous/polycrystalline structure at (222) and (400) texture with increasing rf power was reported.

Amanullah et al. [135] reported that a slight shift of (222) peak towards higher

2θ value occurs with increase in tin concentration. Thilakan et al [136] reported that the change in preferred orientation takes place between (222) and (400) planes when deposition rate is increased. They also report that higher substrate temperature always helps the film to crystallize along (222) preferred planes. Yi et al. [137] reported that the XRD peaks become broader and shift to smaller 2θ side with decreasing substrate temperature. They also found that polycrystalline

ITO films crystallized by

Table 4.1: Summary of deposition parameters for ITO thin films

Parameters Values

Deposition Process R.F sputtering Magnetron Disc Area of the target, ( inch) 2 Material of the target ITO Rf power, (W) 50 - 200 Substrate glass Substrate temperature, (Kelvin) 648

Working gases Ar+ O2

Working gas ratio Ar:O2, (sccm) 8:12 Total Pressure in the chamber, 1.0 x 10-4 (m bar) Distance between the cathode and 7 anode, (cm) Deposition time, (minutes) 25

Table 4.2 Deposition rate of ITO thin films for different rf power

rf power Thickness Deposition (µ m) (watt) rate ( Å/min)

50 0.1800 30

100 0.5400 90

150 0.7200 120

200 1.1400 190

Deposition rate (Å/min) rate Deposition

rf power (W) Figure 4.1 The dependence of deposition rate of ITO thin films on rf power

Figure 4.2 XRD pattern of ITO thin films prepared at different rf power: (a) 50 W (b) 100 W (c) 150 W (d) 200 W

Table 4.3 Structural parameters of ITO thin films deposited at different rf power Dislocation Number of rf Interplanar Crystallite density d crystallites power distance d size D hkl x 10 14 x 10 15 (Watt) (Å) (nm) (lines/m2) (lines/ m2)

50 2.9194 44 5.2379 1.1144

100 2.9016 50 3.9560 1.3243 222

150 2.9132 51 3.8440 2.9852

200 2.8946 57 3.0765 3.2451

50 2.5296 28 1.3039 3.7124

100 2.5163 34 8.6251 4.1823 444

150 2.5436 37 7.3040 7.9821

200 2.5116 47 4.5854 5.3332

thermally annealing have larger grain size and smaller uniform traits than those of the as deposited polycrystalline films.

4.1.3 Effect of rf power on the surface morphology of ITO thin films The surface morphology of ITO films was analyzed by SEM. Figure 4.3 shows the ITO film morphology for four different rf power, deposited on glass substrate, at a substrate temperature of 648 K. At an rf power of 50 W, the film exhibited a smooth surface, showing an amorphous nature of the film formed at low rf power. The amorphous nature of the film is due to insufficient thermal energy; for the diffusion of adatoms on the substrate surface for the nucleation.

At an rf power of 100 W, an early stage of growth takes place and, island with a random crystallographic orientation was formed on the substrate. Ad atoms that are landing on an island, which offers a high mobility can reach the edge of the islands and will contribute to lateral growth. Figure 4.3(c) shows a nanoscale system with a non-uniform morphology with bright flower-shaped agglomerates of grains at an rf power of 150 W. When the rf power is increased to 200 W the film exhibited dense layers with fine grains. The grain size revealed from SEM pictures was found to increase with the increase in the rf power, which is also confirmed by the increase in the crystallite size revealed from x-ray diffraction data, summarized in Table 4.3

4.1.4 Effect of rf power on the electrical and optical properties of ITO thin film

The sheet resistance of the ITO film deposited on glass slides for different rf power was measured at five points on each sample and averaged. The results of depositions using an oxygen partial pressure of 1 Pa show that the sheet resistances Rsh of the ITO films decreases with increasing rf power used, as shown in Figure 4.4. Films grown at 100 W showed typical Rsh of about 45 kΩ/sq, at 150 W this was 12 kΩ/sq, while at 200 W the Rsh was 35 Ω/sq. It should be noted that these values were obtained without post deposition annealing. A. K. Kulkarni et al. [138] studied the effect of grain size on sheet resistance of ITO samples. They found a decrease in Rsh with the increasing peak intensity of orientation, due to the increase in grain size, with improved film texture. Ma et al. [139] reported that the grain size increases with increase in substrate temperature. They also reported that the change in conductivity and carrier concentration is due to change in oxygen content.

The variation of resistivity ρ, carrier concentration n and mobility µ with rf power is summarized in Table 4.4. The electrical parameters show a direct dependence on rf power, reflecting the crystalline quality of the ITO films.

During the initial growth stages of ITO film formation at low rf power less than

100 W, amorphous films are deposited, with most of the Sn atoms present in the interstitial regions. When rf power is increased the Sn atoms diffuse from interstitial and grain boundary regions into the Indium cation sites causing a decrease in resistivity and an increase in carrier concentration as seen in Figure

4.5 and 4.6, respectively. Thus, the growth process accelerating the Sn diffusion towards Indium sites enhances the carrier concentration. Figure 4.7 shows the variation in mobility of ITO films deposited at different rf powers. Highest mobility of 3.2 cm2 V-1s-1 is observed for the ITO films deposited at 200 W. The observed carrier concentration values show a variation from 1017 to 1019 for films deposited at 50 W to 200 W. This result confirms the transformation of amorphous ITO films deposited at 50 W to well crystallized films at 200 W.

However, films deposited with rf power of 200 W was reported to be difficult to etch [140, 141]. Hence the rf power of 150 W, which yielded films with low sheet resistance and appreciable growth rate was used in all subsequent work.

The absorbance and transmittance measurements were carried out using

UV VIS spectrophotometer. Figure 4.8 and 4.9, shows the measured optical absorbance and transmittance spectra ITO films deposited at different rf power over the wavelength range 300 to 1100 nm. It is observed that above

Figure 4.3 Surface morphology of ITO films deposited on glass substrate at different rf power: (a) 50 W (b) 100 W (c) 150 W (d) 200 W

Table 4.4 Electrical parameters of ITO thin films deposited at different rf power

Sheet Carrier rf power Resistivity Mobility Resistance concentration -3 2 -1 -1 (watt) 18 -3 (ρ x10 ) Ώ-cm cm V s (Rsh) Ώ/sq (n x10 ) cm

50 70 16.422 9.01 1.8

100 45 4.4491 7.82 2.5

150 12 7.3402 3.12 2.9

200 35 3.4042 1.92 4.0

Figure 4.4 Variation of sheet resistance of ITO thin films deposited at different rf power

Figure 4.5 Variation of resistivity of ITO thin films deposited at different rf power

Figure 4.6 Variation of carrier concentration of ITO thin films deposited at different rf power

Figure 4.7 Variation of mobility of ITO thin films deposited at different rf power

Figure 4.8 Absorbance spectra of ITO thin films deposited on glass substrate for different rf power: (a) 50 W (b) 100 W (c) 150 W (d) 200 W

Figure 4.9 Transmittance spectra of ITO thin films deposited on glass substrate for different rf power: (a) 50 W (b) 100 W (c) 150 W (d) 200 W 80% transmittance is achieved from 600 nm wavelength. The films formed at lower rf power of 50 W, exhibited low optical transmittance of about 72%. The low optical transmittance at lower rf power may be due to insufficient energy available for reaction of indium with oxygen to form ITO. It is also confirmed by

XRD pattern that amorphous films were formed at low rf power. The increase of optical transmittance for rf power > 50 W was due to decrease of scattering centers of light, because of the formation of nearly stoichiometric, hence improved the optical transmittance. It is also observed from the figure that the transmittance sharply decreases below threshold wavelength, due to fundamental absorption of

ITO.

4.1.5 Effect of annealing on the structure and morphology of ITO thin film

Film properties can be modified by annealing in either oxidizing or reducing atmospheres, as well as vacuum or nitrogen ambient [142]. Therefore, it is necessary to understand the effects of annealing on the structure and morphology of the film. Because annealing helps to improve the electrical properties of the films, it is necessary to employ low but prolonged annealing. After optimizing the film for various rf power, the film deposited at 150 W under a pressure of 1Pa, and at a deposition time of 25 minutes have been annealed at 700 K in a muffle furnace and the effect of annealing on the structural, electrical and optical characteristics of

ITO was investigated.

The X-ray diffraction patterns obtained for as deposited and annealed ITO thin films prepared at the same sputtering conditions are compared. Figure 4.10 (a) and (b) show the diffraction pattern of the films before and after annealing, respectively. All diffraction peaks are indexed with JCPDS card file (006-0416)

[142], and corresponding peaks are identified. The cell parameter was calculated as

10.24 Å. The diffraction patterns show the presence of (2 2 2) phase in the films.

Studies on ITO films have shown that the different energy state of sputtered particles brings about different texture formation [143,144]. Extremely high energetic sputtered particles promote the (4 0 0) orientation in the big diffraction angle 2θ greater than 34° region, while relatively energetic particles promote the (2

2 2) orientation in the small diffraction angle 31–32° region. However, the (2 2 2) orientation has been reported as the preferred orientation for ITO film sputtered from metal alloy or oxide targets by many works [112, 113].

It was observed that there is an overall increase in the peak heights upon annealing the film. This is because of the improved crystallinity upon post- deposition annealing, probably because of the incorporation of Sn−2 in the In+3 sites in the presence of the oxygen. Thus, annealing provided restructuring of the film leading to better polycrystalline nature. A.M. Gheidari et al. [145] studied the effect of annealing on the rf sputtered ITO films on glass substrate from 200 0C to

450 0C. The films showed the variation of (222) and (400) peak intensity with annealing temperature. Karasawa and Miyata [146] reported that annealing enhances the grain growth and the enlargement and shrinkage of lattice parameters is attributed [147] to the interstitial and substitutional Sn atom.

The mean crystallite size was determined using Scherrer's formula [144], by measuring FWHM of (2 2 2) peak, and it was found to be 56 nm. The observed interplanar spacing was compared with the standard data and the calculated values of crystallite size, dislocation density and number of crystallites for (222) and

(400) planes of ITO are presented in Table 4.5.

It can be seen that the crystallite size calculated for (4 0 0) plane are smaller than that for (2 2 2) orientated film. Moreover, the size of (2 2 2) grains was found to increase with annealing, and takes the maximum value of 56 nm. M.J. Alam and

D.C. Cameron [148] has studied the effect of annealing ITO in different atmosphere, and reported that annealing in different atmosphere has no effect on the crystallinity of the films. L.R. Cruz et al. [149] reported that films annealed at

473 K are amorphous, while those annealed above 523 K are crystalline. G.

Gonçalves et al. [150] compared the post deposition annealing effect between ITO and IZO films. They reported that annealing had no significant effect on XRD patterns and grain size, as the processing temperature of these samples was relatively higher than the annealing temperature.

Fig. 4.11 (a) and (b) show the SEM photographs of ITO film before and after annealing, respectively. The film surfaces were found to be smooth and adherent to the substrate. The SEM analysis before annealing shows a nanoscale system with a non-uniform morphology with bright flower-shaped agglomerates of grains. After annealing the film showed a fine, uniform and closely packed distribution of polycrystalline nanoclusters free of micro cracks. The effect of annealing improved the structural homogeneity and the degree of crystallinity of the film which is evident from the SEM pictures.

4.1.6 Effect of annealing on the electrical and optical properties of ITO thin film

The electrical characteristics of the as deposited and annealed ITO film is summarized in Table 4.6. The table shows that the mobility and the carrier concentration increases with annealing, leading to a decrease in the sheet resistance and resistivity value. It is observed that, after annealing to a temperature of 700 K, the carrier concentration increased from 1018 cm−3 to 1020 cm−3, due to the incorporation of more tin in the film, while the Hall mobility increased from 2.9

2 −1 −1 2 −1 −1 cm V s to 14.7 cm V s resulting in Rsh decreasing from 12 KΩ/sq to 8 Ω/sq.

This significant reduction in the Rsh and hence resistivity ρ, may be attributed to the large increase in the mobility, and suggest that this may be associated with the increase in oxygen vacancies and grain size of the film. A.M. Gheidari et al. [145] studied the effect of annealing on electro-optical properties of the films. It was reported that the resistivity of the films takes its lowest value at 400 0C due to larger carrier mobility, improved crystallinity and increase in grain size [151-156].

H. Morikawa et al. [157] has studied the effect of heat treatment on electrical properties of ITO. They reported a two step decrease in resistivity by heat treatment in vacuum at around 393 K and 503 K. On heat treatment at 503 K, all the film was crystallized and the resistivity markedly decreased, owing to the increase in carrier concentration. Many carriers were probably released from the crystallized area. These facts imply that the dopant Sn must be inactive in the amorphous ITO and become active with crystallization.

Figure 4.12 (a) and (b), show the optical transmittance spectra of as deposited and annealed ITO film deposited at an rf power of 150 W. The transmittance was found to be 85% and 92% for as deposited and annealed films, respectively. Upon annealing, the percentage transmittance was improved, due to decrease in the degree of roughness of the surface of the films and in the density of structural defects in the films. The improvement in optical transmission can be attributed to either the decrease in thickness or the improvement in perfection and stoichiometry of the film. High temperature annealing leads to films with a steeper optical absorption curve, which indicates a better crystallinity of the films and lower density of structural defects near the band edge. Barlow et al [158] reported that the shift in the absorption edge of the ITO film after annealing is related to the increase in crystalline nature of the film associated with the decrease of defect concentration in the film.

Figure 4.13 (a) and (b) show the variation of (hαν) 2 with energy for the as deposited and annealed ITO film, respectively. The extrapolated bandgap for the film was found to be 3.58 and 3.7 eV, respectively, which is in good agreement with the values reported in the literature, within the bounds of experimental error. The

Burstein–Moss shift towards higher energies is clearly seen in the result. The band gap broadening arising from this shift is related to the high concentration of free electrons, which partially fill the conduction band, preventing the transition to the lowest levels. The bandgap shift in ITO films has also been reported by other authors

[159]. In a general way, it can be stated that annealing substantially affected the optical properties of ITO film. Post-deposition annealing at temperatures above 200

0C has been proven effective to promote grain growth or crystallinity of ITO thin films [126-129]. Hence, from

Figure 4.10 XRD pattern of ITO thin films prepared at an rf power of 150W (a) before annealing and (b) after annealing

Table 4.5 Structural parameters of ITO thin film deposited at an rf power of 150 W

dhkl (Å) Dislocation No. of Diffraction Crystallite density x 1014 crystallites angle (2θ) Stage (hkl) size lines/m2 x1015/m3 degree Observed Standard

30.4679 222 2.9315 2.9245 51 3.844 2.985

As deposited

35.2559 400 2.5436 2.5327 37 7.304 7.982

30.4546 222 2.9298 2.9245 56 3.188 2.083 Annealed

34.9656 400 2.5390 2.5327 43 5.408 4.917

Figure 4.11 SEM pictures of the ITO thin films deposited on glass substrate (a) before annealing and (b) after annealing

Table 4.6 Electrical parameters of ITO thin film deposited at an rf power of 150 W

Resistivity Carrier 1 Sheet resistance Mobility Stage ρ x 10-3 concentration R (Ω/sq) μ (cm2V-1s-1) sh (Ω cm) n (cm-3 )

As deposited 19x103 119 15x1018 2.9

Annealed 8.9 6.5 44x1020 14.7

151

Figure 4.12 Transmission spectra of ITO thin films deposited at an rf power of 150 W (a) before annealing and (b) after annealing

Figure 4.13 Band gap ITO thin films deposited at an rf power of 150 W (a) before annealing and (b) after annealing

152

the above studies it was observed that changes in the optical and electrical characteristics with the annealing temperature have been associated with changes in the local ordering of the material during crystallization and also to oxygen-vacancy creation and/or annihilation that depends on the annealing atmosphere.

4.2 CONCLUSION

From the above investigation, it was observed that the ITO film deposited at an rf power of 150 W, yielded films with low sheet resistance and appreciably growth rate. Sheet resistance Rsh decreased drastically after annealing at 700 K, due to the large increase in carrier concentration and mobility. Optical studies revealed that the average transmittance was above 85%, for the as-deposited and annealed films. XRD patterns confirm the polycrystalline nature of the film and the intensity of (4 0 0) diffraction peak improved after annealing. SEM micrograph of the annealed film shows a uniform and closely packed distribution of polycrystalline nanoclusters free of microcracks.

153

CHAPTER 5

PREPARATION AND CHARACTERIZATION OF ALUMINUM NITRIDE THIN FILMS

In recent times the group (III-V) nitrides (AlN, GAN, and InN) have become the subject of intense worldwide attention due to their unique properties. Among these nitrides, AlN and its alloys are the most attractive material investigated over the past several years [160,161]. Growth of high quality epitaxial wurtzite AlN thin film on different substrates have been of great interest due to their peculiar features such as high electrical resistivity

(1011-1013ohm.cm), high thermal conductivity, high hardness (11-15 GPa), wideband gap 4.0-6.2 eV and high velocity of acoustic waves etc. The combination of these characteristics makes AlN thin films a promising material for many electronic, optoelectronic, acoustic devices, such as surface acoustic wave (SAW) [162,163]. The characteristics of AlN films are greatly influenced by their microstructure; growth of smooth surface and defectless structure were the goal of device development. Various deposition techniques and substrates have been employed in an attempt of achieving high quality growth; the most frequently used substrates are Al2O3, SiC, Si (111) and Si

(100). The later is of special interest because of the possibility to integrate

154 with the contemporary Si device technology for different functions, e.g. UV detection or emission and Si-ICs on a common substrate.

In the past years, several methods have been developed to prepare AlN thin films, such as chemical vapor deposition (CVD), molecular beam epitaxy

(MBE), pulsed laser deposition (PLD), ion-beam assisted deposition and reactive sputtering [164-167]. However, these methods are expensive and they require high temperatures to reach satisfying properties, and are often incompatible with microelectronic processes. Among other techniques, the reactive magnetron sputtering process is an attractive deposition technique, because it presents advantages of being low temperature and low cost methods, and it allows fine-tuning of the material characteristics [168-170].

In the present chapter, optimization of AlN thin films has been discussed in detail. For the development of MIS gas senor device, with AlN as the insulator, AlN films were grown on Si (100) substrate by dc reactive magnetron sputtering. The films prepared for different deposition time have been analyzed by different characterization techniques to study the structural, optical and electrical properties of film. MIS capacitors were fabricated using the AlN films and the dependence of dielectric parameters with deposition time has been investigated.

155

5.1 OPTIMIZATION OF ALUMINUM NITRIDE THIN FILMS

AlN films were deposited with a 99.5% pure Al target, using dc magnetron sputtering system operated at 60 W dc cathode power; in pure Ar and N2 gas mixture, which was introduced into the chamber by separate mass flow controllers. The AlN films were grown on glass substrate for different nitrogen concentration of 5%, 10%, 30%, 50% and 70%, and the N2

concentration was optimized at 50%. After optimizing the nitrogen concentration at 50%, AlN films were grown on Si substrate for different deposition time, keeping the other deposition parameters constant. The deposition of AlN films can take place in a wide range of temperatures from room temperature up to 400 C. High temperature deposition has the disadvantage of producing degradation of the substrate, incorporation of impurities and thermal damage to the growing film [171]. Hence, deposition of AlN films at low temperature has become increasingly important and value.

Therefore, AlN films were grown on (100) oriented silicon substrate at room temperature. The native oxide on the silicon wafer was removed through an etching step in dilute HF solution. After this initial cleaning process, high purity Ar gas was introduced into the chamber and the chamber was evacuated to below 1 X10-3 Pa. Then prior to each run the target was pre sputtered with argon gas for 5 min with the target shutter closed. During pre sputtering dc

156 power and Ar pressure was kept constant at 60 W and 1 X10-3 Pa, respectively

Then nitrogen was introduced into chamber and reactive sputtering was initiated. A summary of deposition parameters and ranges used in AlN thin films is listed in Table 5.1.

The thickness measurement of AlN films was performed using a Stylus profilometer. The samples were partially masked in order to obtain a step height between the silicon wafer and the film. The thickness of aluminum nitride thin films was measured at five different positions and averaged. The chemical composition and the elemental depth profiles were studied by RBS using 2.4 MeV 4He+ ions, in random geometry. The type of crystalline structure, orientation and grain size was examined by X-ray diffraction

(XRD). The vibrational phonon modes of AlN films were studied using

Fourier transform infrared spectrometer (FTIR); the microstructure and surface morphology of layers were inspected by Scanning Electron

Microscope (SEM). Optical studies were carried out by UV-visible spectrometer and band gap was calculated from the obtained reflectance spectra. Further, MIS capacitors were fabricated on silicon substrates and variation of dielectric parameters with deposition time is studied.

157

Table 5.1: Summary of deposition parameters for AlN thin films

Parameters Values

Deposition Process D.C sputtering

Magnetron Disc Area of cathode (target), inch 2

Material of the target Al Discharge voltage, V 510

Discharge current, m A 117 Substrate Silicon (100)

Substrate temperature, Kelvin 300 Working gases Ar+ N2 Initial pressure in chamber, mbar 1.3X10-4 -5 Partial pressure of active (N2), m bar 5 X10 Total Pressure in the chamber, m bar 1 X10-3 Distance between the cathode and 4 anode, cm Deposition time, minutes 2, 4, 6, 8

158

5.1.1 Rutherford backscattering spectrometer analysis of AlN thin films

Among the various micro analytical techniques used today to investigate the physical properties of solid surfaces, microanalysis using a 2.4

MeV charged particle beam [172] is a tool, particularly well suited to study physical and chemical phenomena taking place in the near-surface region of solids, and which implies quantitative determination and depth profiling of very small amounts of elements. Rutherford backscattering of charged particles (RBS), mainly 4He+ ions, permits the determination and depth profiling, with a depth resolution of 100-300 Å in the regions of a solid up to

1 µm. Figure 5.1 illustrates the principle mass sensitive depth microscopy in the case of a thin film B on a substrate A before and after a reaction which leads to the inter diffusion and to the complete mixing of A and B. A rapid comparison of the two RBS spectra indicate the nature of the surface atoms and give an insight into the A and B depth distributions. More generally in the case of a sample with ‗i‘ different species, the ideal energy spectrum can be defined as,

N (E) dE = ∑i Ci (xi) d xi (dζi /d Ω) [E (xi)] n Ω where N (E) dE is the number of α particle detected between E and E+ dE; Ci

(xi) the concentration of the ‗i‘ species at a depth xi, n the number of incident

159

α and Ω the solid angle of detection. Based on the plateau heights in Figure

5.1 the precise compositions of the Ax By mixtures can be directly obtained from the spectra by the ratio of the plateau heights HA and HB corrected from cross-section effects [173]. Moreover, as the area under each peak or each part of the spectrum is proportional to the number of atoms cm-2, information can be rapidly obtained from a backscattering spectrum without a complex analysis.

Since composition ratios can be determined directly from RBS spectra, and the kinetics of transformation of thin films can also be followed, the RBS spectra were recorded for the AlN films deposited on glass substrates for different nitrogen concentration. Figure 5.2 shows the film composition of

AlN films using high precision RBS spectrometer deposited for different nitrogen concentration. The depth elemental distribution is calculated in atom/cm2 as seen by the backscattered particles, using the atomic density of the considered material. The film composition was found to be free of contamination and with a homogenous depth elemental distribution. The good

-3 quality AlN thin film was achieved in Ar+N2 atmosphere of about 1x10 m bar and at a nitrogen concentration of 50% , in this case the average atomic ratio Al:N:Zr in the sample was determined to be 0.6958:0.3061:0.0020.

160

Figure 5.1 The principle mass sensitive depth microscopy in case of a thin film B on a substrate A before and after reaction

Figure 5.2 The RBS spectra of AlN thin films for different nitrogen concentration.

161

5.1.2 Influence of nitrogen concentration on the crystal orientation and deposition rate of AlN thin films

Properties of films are strongly dependent on deposition parameters such as target power, growth temperature, sputtering pressure and gas compositions [174-176]. It is important to be able to control the crystal orientation and crystalline quality of the film for many applications [13-20]. In the present work, the effects of deposition conditions on the crystal orientation and microstructure of the aluminum nitride films have been investigated.

The crystalline structure of AlN is discussed in Chapter 2. The (002),

(101) and (100) planes of the wurtzite structure are shown in Figure 2.7. For the (002) orientation, the c-axis is normal to the substrate and the plane parallel to the substrate is the close-packed basal plane, with either all aluminum or nitrogen atoms. On the other hand, for (100) or (110) orientations, where the c-axis is parallel to the substrate, the atoms are loosely packed and the plane consist of equal number of aluminum and nitrogen atoms. For the (101) orientation with aluminum or nitrogen atoms, loosely packed structure than that for the (002) plane would be formed. The adatoms need a sufficient kinetic energy to rearrange themselves and form the (002) orientation. This energy can be supplied by changing nitrogen concentration

162 and deposition time. Hence in the present work, AlN films have been deposited for different nitrogen concentration.

Figure 5.3 shows the XRD pattern of the aluminum nitride thin films deposited on glass substrate for different nitrogen concentration, for a deposition time of 2 minutes, keeping all other deposition parameters constant. For the nitrogen concentration of 5% and 10%, the films show amorphous nature due to low deposition rate. At the nitrogen concentration of

30%, (002) plane with weak intensity was obtained. At 50%, the peak intensity of (002) reflection increased, while (100) reflection just appeared.

Further increase in nitrogen concentration to 70% resulted in films with weak intensity of (100) and (002) peak. The deposition rate of AlN films for different nitrogen concentration are calculated and summarized in Table 5.2.

The dependence of deposition rate on nitrogen content is shown in Figure 5.4.

The deposition rate was highest at 50% of nitrogen, where the XRD showed c- axis orientation. The deposition rate of AlN film deposited with the high nitrogen content of 70% decreased slightly, which may be attributed to the inefficient sputter yield of the nitrided target.

The plasma composition is an important parameter for making stoichiometric films. The crystal orientation of the AlN films changed dramatically with nitrogen content. W. T. Lim et al. [177] reported that the

163

Figure 5.3 XRD pattern of AlN thin films prepared on glass substrate at different nitrogen concentration

164

Table 5.2 Deposition rate of AlN thin films for different nitrogen concentration

Nitrogen Thickness Deposition rate concentration (Å) ( Å/min) (%) 5 120 2

10 180 3

30 360 6

50 540 9

70 480 8

Deposition rate (Å/min) rate Deposition

Nitrogen Concentration (%)

Figure 5.4 The dependence of deposition rate of AlN thin films on nitrogen concentration

165

(002) peak intensity of AlN increases, with increasing nitrogen content above

50% of nitrogen in RF sputtering. On the contrary, Okano et al [178] reported that the c-axis orientation has been improved with decreasing N2 concentration. Kumar et al. [179] also obtained highly oriented c-axis AlN films at low N2 concentrations. These conflicting results may be due to the different types of sputtering power and system design. Therefore the results cannot be compared directly with reported results. There will be changes in the mean free path of sputtered particles, ionization cross-sections, target surface state and ionization energy as well as plasma chemical composition when the relative content of nitrogen gas is increased. These factors may influence the crystal orientation of the deposited films.

5.1.3 Influence of deposition time on the preferential orientation and deposition rate of AlN/Si thin films

After optimizing the AlN films for different nitrogen concentration, the films have been prepared on Si (100) substrate for different deposition time, and the effect of deposition time on the crystalline orientation of aluminum nitride films was investigated. The XRD pattern of the aluminum nitride thin films, prepared at the nitrogen concentration of 50%, for different deposition time are shown in Figure 5. 5. All peaks correspond to the wurtzite hexagonal

166 phase of aluminum nitride. The deposition time was varied from 2 min to 8 min, keeping all other parameters constant. Very weak (002), (102) and (101) peaks and a strong (100) appeared at the low deposition time of 2 min (Figure

5.5 (a)). At the deposition time of 4 min, the intensity of (002) peak increased, while the intensity of (102) and (103) peaks decreased. When the deposition time increased to 6 min the intensity of (002) peak is enhanced, while the intensity of (100), (102) and (101) peak decreased (Figure 5.5 (c)). Further increase in deposition time to 8 min caused the strong (002) peak to increase in intensity (Figure 5.5 (d)), while the (102) and (103) peaks disappeared completely.

This dependence of crystal orientation on deposition may be explained by the kinetic energy of sputtered atoms. Ideally, deposition time should be high so that material can easily travel through vacuum from sputtering target to substrate. Since plasma is maintained by collision between electrons and neutral gas atoms, a high time is required to exit the plasma. Sputtered atoms travel in a line-of-sight mode at low deposition time, in which the energy imparted at the substrate may cause damage of the growing films as is evident in the low intensities of several peaks in XRD (Figure 5.3 (a) and Figure 5.5

(a)). The variation in orientation of the AlN films was thought to be related with deposition rate. The (002) orientation of AlN films was obtained at

167 higher deposition time. At high deposition time adatoms on the surface have a higher probability of rearrangement and a longer time to form a low-energy configuration such as the close-packed (001) atom plane.

The deposition rate of AlN/Si thin films are calculated for the different deposition time and the values are summarized in Table 5.3. The variation in deposition rate of the aluminum nitride thin films grown at different deposition time is shown in Figure 5.6. The deposition rate increases with deposition time, showing a maximum value at 8 minutes of deposition time, due to fewer collisions and scattering between the sputtered gas molecules.

The difference in growth rate of the planes decides the orientation. The equilibrium form of wurtzite is made of {100} and {001} planes. The nucleation on the substrate will be random at the initial stage of crystal growth. But only some nucleations survive whose fastest growth directions are closest to the normal to the substrate. This geometrical selection causes preferential orientation. When the growth rate of the (100) plane is faster than that of the (001) plane, the nucleation with (100) planes normal to the substrate grows and it becomes the (100) preferred orientation.

Yumoto et al. [180] explained the preferential orientation of AlN films by an extended periodic bond chain (PBC) theory. The periodic bond chain model was originally proposed by Alexander to interpret the orientation

168 mechanism of AlN films from the relationship of angles between the crystalline planes [181]. In this theory, Al-N dimers are formed when the collisions of Al and N atoms occur between the Al target and the substrate.

Under these conditions, the (100) orientation tends to be obtained. Xu et al.

[182] controlled the crystal orientation by changing experimental conditions, such as target power, sputtering pressure and the distance between target and substrate. They have reported that, longer distance and a higher sputtering pressure are conducive to the growth of AlN (100) films, whose c-axis is parallel with the substrate. It was explained by the decrease in energy of particles due to collisions between sputtered particles. The micro structural properties and the lattice disorders are studied from the analysis of XRD peaks, and are listed in Table 5.4. The mean crystallite size was found to be

~60 nm. By substituting the measured crystallite size, the strain (ε) and dislocation density (δ) in the films was determined.

5.1.4 Influence of deposition time on the vibrational modes of AlN/Si thin films

FTIR spectroscopy is a fast, inexpensive and an effective technique for routine analysis of AlN films. FTIR stands for Fourier Transform Infrared, the preferred method of infrared spectroscopy. In infrared spectroscopy, IR radiation is passed through a sample. Some of the infrared radiation is

169 absorbed by the sample and some of it is passed through (transmitted). The resulting spectrum represents the molecular absorption and transmission, creating a molecular fingerprint of the sample. Like a fingerprint no two unique molecular structures produce the same infrared spectrum. This makes infrared spectroscopy useful for several types of analysis. Figure 5.7 shows the characteristic vibration modes of the lattice deposited at 2, 4, 6, 8 min of deposition time on Si (100) substrate. It was reported that the crystalline AlN

-1 exhibits characteristic modes at ~611, ~670, ~890, ~ 912 cm arising from A1

(To), E1 (To), A1 (Lo) and E1 (Lo) vibration modes, respectively, which are IR active [183]. A strong FTIR transmittance peak around 596 cm-1 was observed for films deposited at 2 min. With increase in deposition time, split peak

-1 profiles were found. Two peaks around 610 and 888 cm were attributed to A1

(To) and A1 (Lo) phonon modes in films deposited at 6 min and 8 min, respectively. A peak shift towards higher wave number side was also found with increasing deposition time. The shift in the peak position from their characteristic position is due to the residual stress in the AlN film induced from sputtering process. Moreover, as the deposition time increases the absorption peak becomes narrower, due to purer phases, less defects, or more stress homogeneity. As shown in Figure 5.7, the peak at 821cm-1 is attributed to a phonon mode of the silicon substrate as reported earlier [184].

170

5.1.5 Influence of deposition time on the morphology of AlN/Si thin films

The surface morphology of the films was analyzed by SEM. Figure 5.8 shows the AlN film morphology for four different deposition times. A granular morphology was observed. The film is composed of regular smooth round grains, both the surfaces are uniform. Based on the deposition conditions, the formation of different mosaic structures is considered to be due to different nitrogen partial pressure during growth [185]. The morphologies can be controlled by fine-tuning the nitrogen content in the sputtering gas. At

50% nitrogen partial pressure, the ion bombardment is more, which leads to the formation of small round grains as shown in Figure 5.8 (a). At the deposition time of 4 minutes, smooth surface was observed (Figure 5.8(b)).

The excessive N species suppress the diffusion of Al species on the growing surface by forming strong Al-N bonds, and a (100) plane is formed. As reported earlier, the observed morphologies are believed to be due to the growth mechanism [186]. At the early stages of growth, island with a random crystallographic orientation was formed on the substrate. Ad atoms that are landing on an island, which offers a high mobility can reach the edge of the islands and will contribute to lateral growth resulting in c-axis oriented film as seen in Figure 5.8 (d). It is clear from the SEM pictures that the crystalline

171

Figure 5.5 XRD pattern of AlN thin films prepared on Si (100) substrate at (a) 2 min (b) 4 min (c) 6 min (d) 8 min of deposition time

172

Table 5.3 Deposition rate of AlN thin films for different deposition time

Deposition Deposition Thickness time rate (n m) (Å/min) ( Å/min)

2 49 8.167

4 81 13.5

6 102 17

8 114 19

Deposition rate (Å/min) rate Deposition

Deposition time (min)

Figure 5.6 The dependence of deposition rate of AlN thin films on deposition time

173

Table 5.4 Structural parameters of AlN thin films deposited on Si(100) for different deposition time

Dislocation Time Thickness d spacing Crystallite size Strain density -4 14 (min) (nm) (dhkl ) Å (D) nm (ε) x10 (δ) x10 lines/m2

2 49 2.7137 61.40 4.48 2.8492

4 81 2.7105 61.40 4. 82 2.9491

6 102 2.7121 57.47 5.47 3.3807

8 114 2.7121 57.47 5.78 3.8147

174

Figure 5.7 FTIR spectra of AlN thin films deposited on Si (100) substrate for different deposition time

175

Figure 5.8 Surface morphology of AlN thin films prepared on Si (100) substrate at (a) 2 min (b) 4 min (c) 6 min (d) 8 min of deposition time

176 growth and surface roughness of the films increased with increasing deposition time, and hence the extended deposition time of 8 min resulted in better crystalline structure.

5.1.6 Influence of deposition time on the optical properties of AlN/Si thin films

The optical reflectance of the films has been recorded employing a

Perkin-Elmer Lamba-950 spectrometer. The UV-visible spectrophotometer is configured to measure reflectance. The optical energy band gap of AlN films is determined by analyzing the optical data with the expression for the reflectance and photon energy using the Tauc relation

n/2 αhν =A (hν-Eg) …. (5.1)

where, n is the constant, which is equal to one for direct band gap semiconducting material and four for indirect band gap semiconducting material. hν is the photon energy, and α the absorption coefficient, which can be written in terms of reflectance as 2αt = ln[(Rmax-Rmin)/ (R-Rmin)], where t is the thickness of the sample and R the reflectance for any intermediate photon energy.

Figure 5.9 shows the reflectance spectra of AlN thin film on Si substrate for different deposition time. A fall in reflectance is observed from

177

Rmax to Rmin due to absorption of light by the material. Figure 5.10 shows, the plot between (αhν) 2 and hν. The extrapolation of linear portion of the plot on the energy axis (αhν)2= 0 gives the value of the energy band gap of the thin film. It is found that the energy band gap of the AlN film varies from 4.35 to

5.3 eV. It is clear that the energy band gap of AlN films increases with increase in the deposition time. This variation in band gap is attributed to the internal strain in the films which increases with the deposition time as summarized in the Table 5.1

5.1.7 Influence of deposition time on the electrical properties of AlN/Si thin films

Impedance measurements are a basic means of evaluating electronic components and materials. Every material has a unique set of electrical characteristics that are dependent on its dielectric or insulation properties.

Accurate measurement of these properties can provide valuable information to ensure an intended application or maintain proper manufacturing process. For electrical measurements, metal-insulator-semiconductor (MIS) structures were formed by sputter deposition through a metal mask of Al dots on AlN film as top electrode, and a continuous Al film on Si wafer backside as bottom electrode and the dielectric properties of Al-AlN-Si (MIS) capacitor was studied. The variation of capacitance with deposition time was studied in the

178 frequency range from 10 KHz - 1MHz, using impedance analyzer at room temperature. The capacitance C as a function of the frequency for different deposition time of AlN films are shown in Figure 5.11. The overall capacitance is given by the geometric expression

C= …. (5.2) where, C is the capacitance, ε is the dielectric permittivity, εo the permittivity of vacuum, A the electrode area and d the film thickness. The capacitance of the thin insulating film almost remains constant in the frequency range from

10 KHz - 1MHz, with its value varying from 0.8X10-8 to 1.8X10-7 µF. It is also observed that, as the deposition time was increased the capacitance decreased, with least value at 8 min of deposition time.

5.1.7.1 Dielectric constant and dielectric loss of AlN/Si thin films

The dielectric constant measurement, also known as relative permittivity, is one of the most popular methods of evaluating insulators such as rubber, plastics, and powders. It is used to determine the ability of an insulator to store electrical energy. The complex dielectric constant consists of a real part (k‘), which represents the storage capability and an imaginary part

(D), which represents the loss. Dielectric constant measurements can be performed easier and faster than chemical or physical analysis techniques making them an excellent material analysis tool. The dielectric constant is

179 defined as the ratio of the capacitance of the material to the capacitance of air, or k‘ = Cx/Co, where Cx is the capacitance with a dielectric material and Co the capacitance without material, or vacuum. The k‘ value of dry air is 1.00053, which for most measurement applications is usually close enough to the value of a vacuum, which is 1.0000. Thus if a material is to be used for insulating purposes only, it would be better to have a lower dielectric constant, or as close to air as possible. To the contrary, if a material is to be used in electrical applications for storage of electrical charge, the higher the dielectric constant the better. More charge is stored when a dielectric is present than if no dielectric (air) is present. The dielectric material increases the storage capacity of the plate capacitor; hence the dielectric constant of any solid or liquid would be greater than 1.

The dielectric constant of AlN films was found to be between 6.0 and

6.8, which is similar with reported values for AlN [187]. The real part of dielectric constant increases with increasing deposition time and decreases with increasing frequency (Figure 5.12). The observed dependence of the dielectric constant on the dielectric film thickness (deposition time) for thinner films (less than ~1000) is attributed to defects such as voids, stresses, inhomogeneity, grain boundaries, discontinuities etc, which are normally present in vacuum-deposited films, rather than to the non-stoichiometry of the

180 deposits resulting from an excess of oxygen or metal atoms. Some of these defects are removed by self-annealing or aging processes, but others require more thermal energy such as is provided by an annealing process. As the films become thicker, the density of voids decreases, resulting in a higher value of dielectric constant, which evidently becomes thickness independent. Adam et al. [188] reported the dependence of dielectric constant on film thickness. The dielectric constant was between 4 and 11 for thicker layers (≥100 Å), and decreased to values between 2 and 6 for thicknesses below 100 Å. Dimitrova et al. [189] reported low dielectric constants (6.8 - 7.1) due to the presence of nitrogen vacancies in the AlN lattice.

Dissipation factor (D) is defined as the ratio of an insulating materials resistance to its capacitive reactance at a specified frequency. It measures the inefficiency or loss of the material, which is always greater than zero, but usually much smaller than the dielectric constant. D measurements are an excellent means of quality control, which can yield indication of contamination or deterioration. The dielectric dissipation factor of an insulating material is the tangent of the loss angle δ. In a perfect dielectric, the voltage wave and the current are exactly 90° out of phase. As the dielectric becomes less than 100% efficient, the current wave begins to lag the voltage in direct proportion. The amount the current wave deviates from being 90° out

181

Figure 5.9 Reflectance spectra of AlN / Si thin films prepared at different deposition time

Figure 5.10 Band gap of AlN / Si thins films prepared at different deposition time

182

Figure 5.11 The variation of capacitance of AlN thin films deposited on Si (100) substrate as a function of frequency

183

Figure 5.12 The variation of dielectric constant of AlN thin films deposited on Si (100) substrate as a function of frequency

Figure 5.13 The variation of dissipation factor of AlN thin films deposited on Si (100) substrate as a function of frequency

184

of phase with the voltage and is defined as the dielectric loss angle. The tangent of this angle is known as the loss tangent or dissipation factor [190].

Hence, a good dielectric film should have minimum dissipation factor. Figure

5.13 shows the variation of the dissipation factor with frequency for different deposition time. The dissipation factor varies from 0.0011 to 0.004 and is independent of the deposition time. It almost remains constant for all the films with different deposition time. The refractive index (η) of the sputter- deposited AlN is dependent on the sputtering conditions, since it is related to the composition and density of the films.

For polycrystalline AlN films obtained by dc reactive sputtering, the refractive index varies from 1.9 to 2.1 [191]. In the present work, refractive index determined from the dielectric measurement of ε‘ and δ, revealed that they are in the range of 1.1-1.2, for the AlN samples prepared at a deposition time of 8 minutes. The decrease in the refractive index values observed in the present study is attributed to the presence of impurities such as oxygen or nitrogen vacancies in the film [192].

5.2. CONCLUSION

The evolution of preferred orientation and morphology of AlN films deposited on Si (100) substrate at 50% nitrogen concentration was studied.

185

The depth elemental distributions studied using 2.4 MeV 4He+ Rutherford

Backscattering spectroscopy confirms that the films are stoichiometric. The prepared films at 50% N2 flow presents a composition close to AlN. The XRD analysis of the films revealed that, at a deposition time of 8 min, the coated film favored the formation of highly oriented (002) preferential plane with enhanced crystal quality, which can provide good piezoelectric response. The band gap increased with the increase in the deposition time and the values of refractive index were in the range of 1.1-1.2, for the samples prepared at 8 min of deposition time. The MIS structures fabricated using AlN at different deposition time showed a significant improvement of electrical characteristics with deposition time as we go from 2 min to 8 min.

186

CHAPTER 6

FABRICATION OF GAS TESTING SYSTEM

Many factors must be addressed when designing a gas sensor; such as the material‘s sensitivity and specificity to the gas in question or if the sensitivity of the material is appropriate for the application. The same gas sensor may not be appropriate in two different environments, a carbon dioxide sensor for the inside of a car exhaust should be designed for high concentrations, whereas one car cabin air quality should be far more sensitive to carbon dioxide at lower concentrations. The sensor would not be accurate enough to register concentration changes of 10 ppm if its sensitive range is

1,000–10,000 ppm. The surrounding pollutant gases will also affect the sensor gas response. Certain gases will change the charge carrier concentration despite them not being the target analyte.

In this chapter, the detailed procedure for the fabrication of gas sensing unit has been described with ITO and AlN thin films as gas sensing devices.

The measurement of sensor resistance when exposed to different gases for various concentrations and different operating temperature is discussed. The mechanism of gas detection of MOS and MIS devices has been explained in detail for ITO and AlN sensors.

187

6.1 DESIGN OF GAS SENSING UNIT

The gas sensing unit consists of a vacuum – tight Borosil glass test chamber of diameter 30 cm and of height 50 cm, with the top base made of removable and of O – ring lid. It has an inlet for allowing the test gas to flow in and an air admittance valve to allow atmospheric air after evacuation.

Another third port is provided for vacuum gauge connection. The main dynamic techniques used for the preparation of gas mixtures are: volumetric pumps, sonic orifices, mass flow controllers, diffusion and permeation. Each method has its own range of applications and requires different equipment to be performed. In present work, mass flow controllers were used to produce different concentrations of gases in a reference atmosphere. Dry synthetic air is used throughout the experiment, as reference gas for the gas sensing measurements. Dry air was chosen in order to simplify the experimental setup.

If water vapour was introduced inside the system, several precautions would have to be taken to avoid condensation. A strict control of gas temperature and pressure has then to be ensured inside the system in order to use an atmosphere with variable humidity. Since humidity is known to increase sensitivity to reducing gases, measured gas response values in dry air will most probably be lower than the ones obtained in real working conditions, where water vapour is almost always present [193].

188

The schematic of the experimental setup used for gas sensing measurement is shown in Figure 6.1. In order to guarantee that the atmosphere inside the system is not contaminated, all the gas lines between the outlet valves of the gas containers and the outlet valve of the chamber may be evacuated using a mechanical vacuum pump. The chamber could be evacuated using an oil free pump to a base pressure of 1.3 Pa. The test gas could be injected inside the chamber using the mass flow controller, and dry air was used as diluting gas which afforded different concentrations of tested gas.

A gas mixing manifold was incorporated to control the mixing ratios of the test and carrier gases prior to being injected into the test chamber. The mixing gas manifold is fed by zero air and test gas through a flow meter and needle valve arrangement. This arrangement of mixing scheme was done to ensure that the gas mixture entering the test chamber is premixed thereby giving the real sensitivity. A multi-pin feed through at the base of the chamber allows for the electrical connections to be established between the sensor and the heater assembly. The heater assembly consists of a hot plate and a Ni-Cr thermocouple inside the chamber in order to control the operating temperature of the sensor. The electrometer was used to register the variation of the sensor resistance exposed to various gas concentrations.

189

Figure 6.1 Schematic of the gas sensing set-up used for sensitivity measurement

Fig.6.2: Diagram of set up used to measure the electrical parameters and heat the samples

190

Using this setup it is not possible to know the temperature at the sample surface. However, if gas flow is kept constant and enough time is allowed in order to achieve equilibrium of heat exchanges, experimental conditions can be reasonably reproduced. Temperature can be varied between 30 0C and

8000C. The temperature is controlled by varying the current flow through the heater and measured with an accuracy of ±1 0C, using the temperature controller. Figure 6.2 shows the diagram of the set up used to measure the electrical parameters and heat the samples. Commonly, in resistive gas sensors, both the heater and the temperature sensor are placed on the substrate.

An arrangement of this sort permits a more accurate surface temperature evaluation but increases sample preparation effort and cost, therefore in the set up shown in Figure 6.1, the heater and the thermocouple are placed below the substrate.

6. 2. MEASUREMENT OF SENSITIVITY

Measurement of thin film resistance or capacitance requires that proper electrical contacts are done on the layer surface. In order to complete this task a number of choices have to be made. In first place, it is necessary to decide whether the contacts on the film surface should be permanent or permit separation and rejoining [194,195]. After this the material and size of the conductor have to be chosen, and finally the means to effect the connection is

191 selected. Since the film resistance has to be measured at temperatures higher than ambient, it has to be checked that the contacts withstand the working temperatures without considerable change. The contact resistance of the electrical connector should also be low and independent of the flowing current, particularly if thin film resistance is low. A permanent connection may be obtained by deposition of a metallic layer over the tested layer. Good contacts provide a uniform contact resistance over the contact-film interface and therefore a uniform electric field over the thin film layer. Sometimes several layers have to be deposited in order to achieve the needed characteristics: namely a bottom layer to prevent diffusion into the tested film, a chemically stable top layer, generally using a noble metal, and eventually a medium layer that provides stable adhesion between the top and bottom layers

[196]. Ideally, these layered structures should be fabricated under the same production process under vacuum in order to prevent development of oxide layers in the interface by exposure to the atmosphere.

In the present work, the ITO films prepared under the best deposition conditions found were used to fabricate the gas sensor in the desired geometry with two thick gold pads on two ends of the film to take out electrical contacts. Operating temperatures are fixed by adjusting the supply voltage given to the flat panel heater and the values of temperature are measured using

192 the thermocouple. Electrical measurements are performed in the temperature range from room temperature to 400o C in air ambient. This response is considered as a reference response for the calculation of sensitivity. After noting the reference response, the test gas was injected inside the chamber, using mass flow controllers. After injecting the test gas, all the valves are closed to avoid the leakage of test gas.

The resistance is once again measured for different temperatures in gas air ambient. The resistance response of each sensor structure will be transformed into a sensitivity value using the formula [197,198];

for oxidizing gases S =Ra / Rg ; and

for reducing gases S = (Ra - Rg) / Rg where, Rg is the sensor resistance influenced by the gas, Ra the sensor resistance in the air. The sensitivity of the metal oxide sensor depends on the film, at which the sensor is operated, with thinner films being more sensitive to gases. The sensitivity can be improved by adding a catalytic metal to the oxide; however, excessive doping can reduce sensitivity.

The grain size of the oxide also affects the sensitivity and selectivity to particular gases as the grain boundaries act as scattering centers for the electrons [199]. The ratio of the grain size (D) to the electron depletion layer thickness (L), ranging below D/2L=1, (the depletion region extends over the

193 entire grain) governs the sensitivity of the sensor [200]. Thin film metal oxide sensors saturate quickly, which can reduce the sensitivity range in which the sensor can operate. The sensitivity of metal oxide sensors is very dependent on the operating temperature, for example, the sensitivity, Ra/Rg (per 10 ppm

NO2), for In2O3 gas sensors range from approximately 1.2 to 1.6 with the operating temperature varying from 350 to 4500C [201].

For the AlN sensor, variation of capacitance of the device with voltage in dry air is measured, and this is considered as a reference. The variation of capacitance with voltage of the sensor is once again measured at room temperature in gas ambient for different gases. Comparing the two graphs, the change in voltage ΔV was used to calculate the sensitivity. The factors that affect the sensitivity of MIS devices are operating temperature, composition and structure of the catalytic metal [202 - 204].

6.3 GAS SENSING MECHANISM OF MOS DEVICE

Semi conducting metal oxide sensors are one of the most widely studied groups of chemoresistive gas sensors. These sensors are designed to react with one class of gases, whereby the semiconducting metal oxide undergoes reduction and oxidation. This process causes the MOS sensors to exchange electrons with the target gas at a certain characteristic rate, thereby affecting the sensor‘s resistance and yielding a certain signal. Despite the simplicity of

194 semiconducting metal oxide measurements for use as gas sensors, the detection mechanism is complex. This complexity is due to the various parameters that affect the function of the solid state gas sensors. These include the adsorption ability, electrophysical and chemical properties, catalytic activity, thermodynamic stability, as well as the adsorption/desorption properties of the surface [202, 205-213]. However, it is believed that gas sensing by semiconducting metal oxide devices involve two major key functions as receptor and transducer functions [214, 215]. The former involves the recognition of a target gas through a gas-solid interface, which induces an electronic change of the oxide surface, while the latter is based on the transduction of the surface phenomenon into an electrical resistance change of the sensor. When a sensor is heated to a high temperature in the absence of oxygen, free electrons easily flow through the grain boundaries of the SMO film. In an oxygen atmosphere, oxygen is adsorbed onto the SMO surface, forming a potential barrier at the grain boundaries. The interaction of atmospheric oxygen with the SMO surface forms charged oxygen species, which trap electrons from the bulk of the material. The layer of charged oxygen at the surface repels other electrons from interacting with the bulk of the film, creating a region depleted of electrons, which results in an increased potential barrier at the grain boundaries. This impedes the flow of electrons

195 and thus increases the resistance. When the sensor is exposed to an atmosphere containing a reducing gas, the SMO surface adsorbs the gas molecules and lowers the potential barrier, allowing the electrons to flow easily and thus reducing the electrical resistance. In this manner, the sensors act as variable resistors whose value is a function of gas concentration.

Figure 6.3 Typical thin film resistor gas sensor

The reaction of SMO materials with gases and the result of the conductometric changes were introduced in the early 1950's by Brattein et al.

[216] and Heiland [217]. There are two types of metal oxide sensors; n-type

(zinc oxide, tin dioxide, titanium dioxide or iron (III) oxide) which respond to reducing gases and p-type (nickel oxide, cobalt oxide) which respond to oxidizing gases [218]. Thermal or photolytic excitation of an n-type (n = negative electron) or donor semiconductor results in an excess of electrons in its conduction band which increases the reactivity with oxidising molecules.

196

An excited p-type (p = positive hole) or acceptor semiconductor shows an electron deficiency in its valence band which promotes reactions with reducing compounds.

The n-type sensor operates as follows: oxygen in the air reacts with the surface of the sensor and traps any free electrons on the surface or at the grain boundaries of the oxide grains. This produces large resistance in these areas due to the lack of carriers and the resulting potential barriers produced between the grains inhibit the carrier mobility. However, if the sensor is introduced to a reducing gas like H2, CH4, CO, C2H5 or H2S the resistance drops because the gas reacts with the oxygen and releases an electron. This lowers the potential barrier and allows the electrons to flow, thereby increasing the conductivity. Equations (1.5) and (1.6) describe the reactions occurring at the surface:

O2 (s) …. (6.1)

R(g) (s) RO(g) …. (6.2) where, e is an electron from the oxide. R (g) is the reducing gas s and g are the surface and gas, respectively [219]. P-type sensors respond to oxidising gases like O2, NO2, and Cl2 as these gases remove electrons and produce holes, i.e. producing charge carriers. The main advantages of metal oxide sensors are

197 fast response and recovery times, which mainly depend on the temperature and the level of interaction between the sensor and gas.

6.4 GAS SENSING MECHANISM OF MIS DEVICE

For investigation of MIS capacitor as gas sensor a theoretical understanding of its functionality is necessary. The Si field effect MIS capacitor is made up of three different parts; a semiconductor, an insulator

(nitride) and a catalytic metal (Figure 6.4). The semiconductor is either n-or p- doped. The semiconductor capacitor is different from ordinary capacitors, in that it has minority and majority carriers that highly influence the behavior of the component.

Figure 6.4 (a) Schematic drawing of the Si-based capacitor (b) Simplified equivalent circuit of MOS capacitor

The semiconductor in itself contributes to the capacitance of the entire device, which means that the total capacitance is the semiconductor

198 capacitance, CS, in series with the nitride capacitance, CN. This can be described according to Eq. 6.3

C tot = …. (6.3)

The capacitance of this type of device is normally studied in a C-V measurement. During such a measurement, a high-frequency voltage is applied to the device, and its capacitance is measured. The amplitude of the alternating voltage is normally about 100 mV. A bias voltage is then added to the high-frequency signal and varied over time, which gives the dependence of the capacitance on the bias voltage. A plot of this dependence is called a C-

V plot and is shown in Figure 6.5, when using an n doped semiconductor. The shape of the C-V plot can be explained through studying the properties of the component and how they are affected by different bias voltages. If an n- doped semiconductor is used, positive voltages on the metal will attract majority carriers, which are the negatively charged electrons in an n-type material, in the semiconductor close to the insulator. This mode is called accumulation (Figure 6.5). The majority carriers easily move back and forth in the semiconductor. This means that they follow the alternating voltage closely and that the entire semiconductor is conductive. In such a case the total capacitance according to Eq. 6.3 becomes close to the capacitance of the

199 oxide. The capacitance of the oxide is independent of the applied voltage, and thus the total capacitance is constant at high positive voltages. [219,220].

Figure 6.5 The C-V characteristics of an n-doped semiconductor capacitor with a metal

When the applied voltage is decreased, the density of majority carriers at the oxide semiconductor interface will decrease as well. At a negative voltage a depletion layer will be formed. In the depletion layer the concentration of majority carriers is lower than in the bulk. The depletion layer has the same effect on the semiconductor as introducing an insulator layer in the material, since carriers cannot move so easily across it, and the total capacitance of the MIS structure thereby decreases.

In the CV plot shown in Figure 6.5 it is seen that depletion is initiated at a positive voltage rather than at a negative. The reason for this is that the

200 discussion above assumes that the capacitor is ideal, that is, the difference in

Fermi level between the metal and the semiconductor is neglected as well as nitride charges and surface states. The effect of these phenomena is to shift the

C-V curve, which causes the behaviour observed in Figure 6.5 [219, 220]. At even larger negative voltages the region closest to the insulator becomes not only depleted of majority carriers, but the negative voltage attracts so many minority carriers that a state called weak inversion is reached. In this state the density of minority carriers in the region is higher than the density of majority carriers. If the voltage is decreased even further, so that the density of minority carriers in the inversion region becomes higher than the density of majority carriers in the bulk, strong inversion has been reached. The size of the depletion layer is then not affected by further decreasing the bias voltage.

This is due to the efficient shielding of changes in the applied electric field by a large concentration of charges in the inversion layer of the depletion region

[219, 220]. At strong inversion the density of carriers in the inversion region is almost the same as in the case of accumulation, although the minority carriers do not move as easily as the majority carriers did. This means that the carriers will not follow the alternating voltage, and the depletion layer will remain even for a large negative bias. Band diagrams for an ideal MIS junction at accumulation, at band conditions (no band bending), depletion and

201 inversion are shown in Figure 6.6. Weak inversion is reached when Efs of the semiconductor is lower than Ei.

Figure 6.6: Energy-band diagrams for an ideal MIS diode (n-type semiconductor) at (a) accumulation, (b) flat band conditions, (c) depletion and (d) inversion [220]

The presence of a dipole layer at the metal-insulator interface, caused by gases across the sensor surface, adds charge to the applied voltage, which means that there will be a shift in the C-V plot. The sensor signal is normally taken as the shift in voltage at a constant capacitance.

202

CHAPTER 7

FABRICATION AND CHARACTERIZATION OF ITO AND ALN SENSORS

Thin film fabrication methods have been used for production of metal oxide gas sensors. Thin film metal oxide sensors are small, and relatively inexpensive to fabricate, have lower power consumption than thick film sensor and can be integrated directly into the measurement circuitry. There are various electrode designs but the interdigitated structure appears to be the most common approach. A heating element is printed onto the back of the substrate to provide the high temperatures required for metal oxides to operate as gas sensors, typically 200 – 500 0C. Film thickness ranges from 10 to 300 mm for thick film and 6-1000 nm for thin film [221]. Catalytic metals are sometimes applied on top of the oxides to improve sensitivity to certain gases.

The same preparative methods are used to apply the catalytic metal [222-224].

In the present chapter, fabrication of ITO thin film as a resistive semiconducting metal oxide gas sensor has been described. The sensor response for three different gases; nitrogen dioxide, ethanol and hydrogen, for different concentration and for different operating temperatures has been discussed. The fabrication of Aluminum nitride thin films as hydrogen sensor,

203 focusing on the interaction mechanism of hydrogen with the MIS device is described. The influence of hydrogen on Pd/AlN/Si and Al/AlN/Si structures have been discussed in detail in this chapter.

7.1 FABRICATION AND GAS SENSING MEASUREMENT OF ITO SENSOR

After having optimized the deposition parameters for ITO films the device was fabricated as a resistive type gas sensor. The substrate used was 4 cm x 4 cm size glass microscope slide. The indium tin oxide thin film of thickness about 0.4 µm grown by rf magnetron sputtering at an rf power 150

W was used to fabricate the device. Two thick gold pads of 7 mm X 4 mm were deposited on two sides of the sensor for ohmic contacts to permit electrical measurements. The sensor area was kept as 25 mm X 25 mm. The schematic representation of the ITO sensor device is shown in Figure 7.1. The

SEM image of the layup of the ITO thin film sensor is shown in Figure 7.2.

ITO thin film gas sensors are of a variable resistance type, the resistivity of which changes in the presence of the gas. Measurement of resistance was carried out using high impedance electrometer (Keithley, Model No. 614).

The substrate could be kept at a chosen temperature using a Pt-heater and the values of temperature are measured by the Ni–Cr thermocouple, with an accuracy of ±1 °C using a temperature controller. The resistance of the sensor

204

Figure 7.1 Schematic representation of the ITO sensor device

Figure 7.2 SEM image of the lay up of ITO film sensor

205

was measured for different temperatures in gas–air ambient. The change in temperature and resistance of sensor were simultaneously measured when different concentrations of gas were introduced in to the test chamber.

Gas sensors based on ITO utilize selective chemical sensitivities of their surface to different adsorbed gases [225]. This causes changes to the electrical resistance of the sensor. Semiconducting metal oxides owe their conductivity to their deviation from stoichiometry. Defects such as cation or anion vacancies play an important role in their conductivity. ITO thin film provides electronic defects that increase the influence of oxygen partial pressure on electrical conductivity. Because oxygen vacancies on metal-oxide surfaces exist as electrically and chemically active, these vacancies function as n-type donors increasing the electrical resistivity of ITO. Upon adsorption of charge, accepting molecules at the vacancy sites, namely from oxidizing gases such as NO2, these electrons effectively deplete themselves from the conduction band of the semiconductor. This leads to a further increase in the electrical resistance of n-type ITO.

7.1.1 Response of ITO sensor to Nitrogen dioxide gas

The ITO sensor was operated at different concentrations of nitrogen dioxide in the temperature range 300–700 K. Figure 7.3 shows the variation in sensitivity of the ITO sensor with operating temperature in the presence of

206

NO2 gas. The gas concentration was kept constant at 50 ppm. It was observed that the response of ITO sensor increased with temperature. At 600 K, it was maximum and then decreases rapidly at 700 K. This reflects that the surface redox state of the elements changes with temperature, with the highest at 600

K. The ratio of measured resistance before and after exposing the sample surface to nitrogen dioxide gives the sensitivity of the sensor towards NO2.

The maximum sensitivity was found to be 28, which is higher than the response of single oxide SnO2, In2O3 and Ga2O3 sensors, reported in literature

[226-228]. Also, the mixing of two different metal phases should be a factor that imparts higher dispersivity and defectiveness to the material, hence increasing the response towards NO2. The sensor working temperature that regulates the adsorption/desorption processes at equilibrium and the competition of chemisorptions between NO2 and atmospheric oxygen O2 for the same active surface sites, play an important role in determining the specific interaction between the gas and sensor [229].

L. Francioso et al. [230] have investigated the nanocrystalline SnO2,

In2O3 and In2O3–SnO2 (molar ratio 1:1) thin films prepared by modified sol– gel method. Their performances in the detection of nitrogen dioxide (2–20 ppm in dry air) by electrical characterization in controlled atmosphere have been reported. They showed that all the sensors showed high responses to

207

NO2. In general, the best performances in terms of response, sensitivity and low detection limit were found in In2O3–SnO2 based sensor. Gas-sensing nanocomposites of SnO2–In2O3 have been reported by Aifan C et.al [231], using a chemically controlled co-precipitation method. They reported that nanocomposites exhibited high sensitivity and selectivity for the detection of

NOx, and the sensitivity depended on the composition of the composites, calcination temperature and operating temperature. They also showed that the incorporation of In2O3 as a secondary component suppressed the grain growth of SnO2, resulting in the increased sensitivity. Particularly, additives of Pd and

Al2O3 as dopants and as surface coating, greatly improved the sensitivity and selectivity of the optimized nanocomposite.

Figure 7.4 shows the sensitivity of ITO thin film sensors to varying concentrations of NO2 at two different working temperatures. At a working temperature of 400 K, ITO sensor had good response over a wide range of

NO2. The sensitivity was linear from 5 ppm to 200 ppm. At 600 K, the sensor showed enhanced response to NO2, at concentrations lower than 50 ppm. At a higher concentration, the sensitivity increased slightly. NO2 molecules are active oxidizing species and may decompose on adsorption sites. If there are no stable adsorption sites for NO2 on the ITO thin film sensor, some NO2 molecules will decompose and desorb from the sensor, which inhibits the

208 further adsorption of NO2. At low temperature, a direct ionosorption at the active film surface takes place as

NO2, gas → NO2, ads …. (7.1)

At higher temperature, , ads ions tend to dissociate into two independent atomic , ads and NO2 molecules start to compete with oxygen for the available surface sites active for chemisorptions [232]. In this situation, with a considerable surface coverage of , ads ions, NO2 reacts with , ads ions forming nitrato complexes , ads,

NO2, gas → N , ads …. (7.2)

Therefore, higher working temperature will accelerate decomposing and desorbing of NO2, and we get a higher sensitivity at low concentration. By suitably changing working temperature, the ITO gas sensor can adapt applications in both low and high NO2 concentration.

Cantalini et al. [233] produced films of WO3 of 150 nm thickness.

Varying the concentration of the NO2 gas between 0.2 and 5.0 ppm at 623 K operating temperature, the films demonstrated good sensitivity at all concentration. The largest responses to NO2 gas were found at working temperature of 473 K, using the film annealed at 773 K. Comini et al. [234] showed a film formed by magnetron sputtering of tin oxide to be sensitive to

209 nitrogen dioxide. The response was reported to be obvious at just 1ppm and are large at 5 ppm and 10 ppm. Mixed oxide (Sn–In)O + Pt with various composition Sn/In = 1:1, 2:1, 5:1 and 10:1 for sensing materials were prepared by Kap-Duk Song et.al [235]. It was reported that the gas sensing characteristics had linearity for wide range (0.5–20 ppm) of gas concentration.

7.1.2 Response of ITO sensor to Ethanol gas

The most widely used metal oxide materials for alcohol detection are SnO2, ZnO, WO3, TiO2 and Zr2O3. Palladium-doped ZnO nanoparticles, produced by flame spray pyrolysis of organometallic precursors and mixed into an organic paste to form thick sensing films, were tested for detection of ethanol vapors in the 25-250 ppm range in dry air at 400°C. A lower concentration of ethanol was detected by CeO2-doped ZnO thin films, fabricated using the dip-coating method [236,237]. But owing to the excellent stability of indium tin oxide (ITO), which is free from drift common in tin oxide materials, use of ITO is very promising for ethanol detection. To study the sensing properties of ITO senor for ethanol gas, the chamber was evacuated to a base pressure of 1.3 Pa using an oil free vacuum pump. Ethanol was then injected inside the chamber using a calibrated digital micro-pipette, where in, it was allowed to evaporate naturally and come in contact with the sensor kept at elevated temperature. A typical 0.4 ml of ethanol corresponds

210 to 10 ppm as reported by Shurmers et al [238]. Before the measurement the samples were preheated up to 700K and then cooled to 650K. An electric fan was installed in the container, which improve uniformity of the gas. The change in resistance of the sensor was measured for different temperatures.

The sensitivity for reducing gases is defined using the relation S = (Ra - Rg) /

Rg, where Rg and Ra the resistance of the sensor in the testing gas and in the air, respectively.

The factors influencing the gas sensing properties of ITO are essential trapping of electrons at adsorbed molecules and band bending induced by these charged molecules which is responsible for a change in conductivity.

The negative charge trapped in these oxygen species causes an upward band bending and thus a reduced conductivity compared to the flat band situation.

As shown in Figure 7.5, when O2 molecules are adsorbed on the surface of metal oxides, they would extract electrons from the conduction band Ec and trap the electrons at the surface in the form of ions. This will lead a band bending and an electron depleted region. The electron-depleted region is so called space-charge layer, of which thickness is the length of band bending region. Reaction of these oxygen species with reducing gases or a competitive adsorption and replacement of the adsorbed oxygen by other molecules decreases and can reverse the band bending, resulting in an increased

211 conductivity. O− is believed to be dominant at the operating temperature of

573–450°C [239], which is the working temperature for most metal oxide gas sensors.

Figure 7.6 shows the variation in resistance Rg of ITO thin film gas sensor with change in operating temperature upon exposure to ethanol vapours. The ethanol concentration was kept constant at 900 ppm. It was observed that the resistance of the sensor Rg decreases with an increase in the temperature and at 650 K, it was minimum. This proves 650 K to be the optimum temperature of the ITO sensor for ethanol detection. The reason for a decrease in the resistance may be due to the oxidation of the ethanol vapour upon coming in contact with the oxide semiconductor surface, which liberates free electrons and H2O. Ethanol vapors react with the chemisorbed oxygen and re inject the carrier thereby reducing the resistance of the material.

The mechanism of the ethanol detection is due to an adsorption of the ambient oxygen onto the sensitive element of the sensor kept at elevated temperature. The reaction kinetics is as follows [240].

→ …. (7.3)

→ …. (7.4)

+ → 2 …. (7.5)

212

Oxygen adsorption takes place at high working temperatures of 650 K. In fact, at lower temperatures, the surface reactions proceed too slowly to be useful, at the same time as at higher temperatures the increased promotion of electrons in to the conduction band tends to obscure the effects of the contaminant reducing gases. The next step is a reaction between ethanol and ionic oxygen species.

C2H5 H → CH3CHO H2 …. (7.6)

Figure 7.7 shows the variation in the sensitivity of sensor S, with concentration of ethanol at a working temperature of 650 K for as deposited and annealed films. The concentration was varied from 200 ppm to 1400 ppm.

As it is seen, the sensitivity of the sensor increased with an increase in the concentration, and the response seems to be quite linear in the entire concentration range. The effect of post deposition annealing induces a surface stabilization of the ITO film and increases the sensitivity as the concentration of ethanol increases. Figure 7.8 shows how the response of the sensor changes with operating temperature for as-deposited and annealed film. The concentration of ethanol was 900 ppm. It was observed that the closely packed morphology of the film annealed at 700 K, show enhanced response due to the interaction of ethanol gas with ITO films, which strengthens the output signal and sensitivity of the sensor.

213

Figure 7.3 Variation of sensitivity of ITO sensor towards NO2 with operating temperature

Figure 7.4 Variation of sensitivity of ITO sensor with concentration of NO2

214

Figure 7.5 Schematic diagram of band bending after chemisorptions of charged species

Figure 7.6 Variation of resistance of ITO sensor with operating temperature towards 900 ppm of ethanol

215

Figure 7.7 Variation of sensitivity of ITO sensor towards ethanol with operating temperature (a) before annealing and (b) after annealing

Figure 7.8 Variation of sensitivity of ITO sensor with concentration of ethanol (a) before annealing and (b) after annealing

216

7.1.3 Response of ITO sensor to Hydrogen gas

Hydrogen attracts more and more attention as a variable clean fuel and thus there are extensive research activities in hydrogen storage, generation, and fuel cells. Since hydrogen is explosive when its concentration in air is more than 4 vol %, sensitive, selective, and stable sensors are needed to measure hydrogen concentration. Factors influencing gas sensitivity in thin- film gas sensors are the nature of gas-sensing materials, microstructure and thickness of the films, fabrication condition, operating temperature, etc. [241,

242].

Figure 7.9 shows the variation in sensitivity of the ITO sensor with operating temperature in the presence of H2 gas. The gas concentration was kept constant at 1000 ppm. It was observed that the sensitivity of ITO sensor increased with temperature and at 400 K, it was maximum and then decreases rapidly, with minimum at 700 K. The maximum response was found to be 1.6 at 400 K, which is considered to be the optimum temperature at which ITO can be operated as a hydrogen gas sensor. Figure 7.10 shows the variation in the sensitivity S, with concentration of hydrogen gas at a working temperature of 300 K. The concentration was varied from 1000 ppm to 7000 ppm. As it is seen, the response of the sensor decreased with an increase in the concentration. It is maximum at 1000 ppm and gradually decreases, becomes

217 minimum at 5000 ppm and almost remains constant up to 7000 ppm. The mechanism of hydrogen detection is described as follows [243]. There are two stages of gas sensing detection. Atmospheric oxygen gets adsorbed on the surface, removing a carrier from the conduction band of n-type semiconductor, becomes O–2 or O– species. This phenomenon reduces the overall conductance. The reducing gas such as hydrogen will react with chemi-adsorbed oxygen. This process will reinject the carrier and increase the sensor conductance. The interaction between a semiconductor surface and H2

(with reducing properties) can be explained in terms of the reaction of H2 molecules with the preadsorbed O2.

2H2 (g) 2H2 O (g) …. (7.7)

H2 (g) H2 O (g) …. (7.8)

H2 (g) H2 O (g) …. (7.9)

As seen from the above equations, the interaction between chemisorbed O2 and the reducing gas (in this case H2) results in a decrease of surface chemisorbed O2, and in an increase of electrical conductance. An important factor to be taken in to account is the presence of water vapor at the sensor surface, since the formation of hydroxyls is in competition with the O2 adsorption; the OH− group does not influence directly the surface reactions but affects the reaction rate. Water adsorbing on the metal oxide surface will not

218

Figure 7.9 Variation of sensitivity of ITO sensor towards H2 with operating temperature

Figure 7.10 Variation of sensitivity of ITO sensor with concentration of H2

219 donate electrons to sensing layers. Moreover, it will lower the sensitivity of metal oxide sensors for some reasons as follows [244]. The reaction between the surface oxygen and the water molecules conduces to a decrease in baseline resistance of the gas sensor, and results in a decrease of the sensitivity.

Secondly, the adsorption of water molecules leads to less chemisorption of oxygen species on the ITO surface, due to the decrease of the surface area that is responsible for the decrease in sensor response.

On the other hand, water molecules also act as a barrier against H2 adsorption. The superficial migration of the H2 atoms on the ITO surface becomes difficult, thus decreasing the response. Water adsorption will significantly lower the sensitivity of metal oxide gas sensors, as shown in

Figure 7.9 and 7.10. Furthermore, prolonged exposure to humid environments leads to the gradual formation of stable chemisorbed OH─ on the surface

[245], causing a progressive deterioration of the sensitivity of the sensor.

The sensitivity of ITO sensor towards NO2, H2 and ethanol for different operating temperature shown in Figure 7.3, 7.7 and 7.9, respectively, reveal that different gases have similar shapes. The response increase and reach their maximum at a certain temperature, and then decreased rapidly with increasing temperature. This tendency is commonly observed in many reports [246, 247-

251]. The shape observed resulted from the competition between slow kinetics

220 at low temperatures and enhanced desorption at high temperatures. Typically, temperature dependence of a metal-oxide sensor signal to the presence of a given analyte possesses a bell shape with a maximum at a certain temperature

[249]. This dependence arises due to several reasons. First of all, as mentioned above, the charge of oxygen species adsorbed at the oxide's surface depends on temperature [252]. Second, since the oxidation reaction is an activated process, its rate increases with temperature. Finally, all adsorption, desorption, and diffusion processes are temperature dependent [250]. Thus, temperature modulation leads to response patterns that are characteristic of the species present in a gas mixture [251].

7.1.4 CONCLUSION

The gas sensing properties of ITO film towards NO2, ethanol and H2 was investiaged. The ITO thin film sensor showed enhanced sensitivity to

NO2 gas at 600 K especially, at concentrations lower than 50 ppm. The sensor show a linear response to ethanol gas in the concentration range of 200 to

1400 ppm. Post-deposition annealing of the film at 700 K, ehanced the response of the ITO senor to ethanol gas. The sensor showed low response to

Hydrogen gas when compared to nitrogen dioxide and ethanol. The maximum response was found to be 1.6 at 400 K and 1000 ppm of H2.

221

7.2. FABRICATION AND GAS SENSING MEASUREMENT OF AIN SENSOR After having optimized the deposition parameters for AlN films, two different MIS devices were fabricated as gas sensor with Al and Pd as catalytic layer respectively. Metal-insulator-semiconductor (MIS) structures were formed by sputter deposition through a metal mask of Al and Pd dots of

1mm in diameter and 150 nm thick on AlN film as top electrode and a continuous Al film of 150 nm thick, on Si wafer backside as bottom electrode.

The thicknesses of AlN thin film as insulating layer was approximately

50 nm. Figure 7.11 and Figure 7.12 show the schematic representation of

Pd/AlN/Si and Al/AlN/Si devices, respectively. First a 50 nm AlN film was deposited by magnetron sputtering technique onto an n-type Si (1 0 0) wafer.

This is followed by a circular 1mm diameter and 150 nm thick Pd dot deposited onto the AlN layer using magnetron sputtering as a top electrode.

An Al film of 150 nm in thickness was deposited onto the back side of the Si to form a back contact. In the similar, way another MIS structure was fabricated with the same values of the thickness, but with Al as top electrode.

The response of both the device towards hydrogen gas sensing is investigated.

The MIS structure mounted on a heater was housed within an airtight chamber. The test gas was injected inside the chamber using two mass flow controllers one for H2 and other for dry air. The C-V characteristic of the

222 sensor in air and vacuum was measured using an impedance analyzer at 1MHz frequency at room temperature. Hydrogen is reported to be able to alter the effective charge at the metal/semiconductor interface or metal/dielectric interface, resulting in changes in barrier height [253,254]. This effect is utilized in semiconductor based gas sensors fabricated on various semiconductor materials such as Si SiC [255] and GaN [256].

Figure 7.11 Schematic representation of the Pd/AlN/Si device.

Figure 7.12 Schematic representation of the Al/AlN/Si device.

223

The interaction of hydrogen with semiconductor devices has long been studied, and intensive research led to a model, which attributes the reaction mechanism of the devices to hydrogen, to the formation of a hydrogen- induced dipole layer at the metal/dielectric/semiconductor interface [257-262].

Figure 7.13 shows the schematic illustration of the principle of gas sensitive

MOS capacitor. The molecular hydrogen adsorbs on Pd or Pt surface and dissociates. Hydrogen atoms diffuse through Pd or Pt, and adsorb at the metal–oxide interface forming a dipole layer. The dipole layer is responsible for the work function change, for example, showing a voltage shift in the C-V characteristics of the device. The hydrogen atoms at the metal insulator interface decrease the flat band voltage of the capacitor (the voltage at which there is no band bending in the semiconductor), which gives a voltage drop,

ΔV, resulting in a shift, ΔV along the voltage axis, of the C-V curve of a capacitor. The field effect devices also respond to other hydrogen-containing molecules like hydrocarbons, which deliver hydrogen atoms to the metal- oxide interface, provided that the molecules are dissociated on the metal surface [263 - 265]. Oxygen molecules are also dissociated on the catalytic metal surface. Oxygen and hydrogen gas adsorbed on the surface forms water which leaves the surface.

224

Figure 7.13 Schematic illustration of the principle of gas sensitive MOS capacitor

In the present work, the interaction mechanism of hydrogen with AlN thin films focusing on the metal/semiconductor interfaces has been investigated. The electrical properties were evaluated by measuring capacitance-voltage (C-V) characteristics, obtained from an Impedance

Analyzer operated at 1MHz. The (C-V) characteristics of MOS devices studied at high frequencies revealed a low capacitance in the case of inversion and a high capacitance in the case of accumulation. The interaction of gas molecules with MIS-Si devices gives rise to changes in the (C-V) curves.In the case of reducing gas H2, the C-V curve is shifted toward negative voltages.

During exposure to gas, the capacitance was held constant by controlling the bias voltage. The gas response R, which is defined as the difference between the steady state voltages measured at the test gas exposure and in dry air is described by the following relation.

225

R= ΔV= VV-Vg …… (7.12)

where, VV and Vg are the voltage of the sensor in vacuum and in the presence of gas, respectively. The voltage shift obtained is positive in the case of reducing gases and negative in the case of oxidizing gases.

7.2.1 Response of Pd/AlN/Si device to Hydrogen gas

The response of Pd/AlN/Si device reveals that, there is a substantial difference between the capacitance-voltage (CV) curves in dry air and in the presence of H2 gas. A voltage shift of approximately 0.8 V is observed in the capacitance –voltage curve, as seen in Figure 7.14. In Pd/AlN/Si device, gas- phase H2 dissociates on the surface of the catalytic metal to form H atoms.

These atoms then rapidly diffuse through the metal film to the metal-insulator interface, where they are preferentially trapped in stabilized adsorption sites.

The layer of interfacial hydrogen created by this process exists in a dipole layer, creating an additional voltage drop across the MIS sensor that can be measured as a voltage shift in the capacitance-voltage (C-V) curve of a capacitor [264]

The Pd/AlN/Si structure behaves as a voltage dependent capacitor. In the sensing process, the hydrogen molecules are dissociated into hydrogen atoms on the Pd outer surface. At low concentration, the hydrogen atoms, most

226 probably as protons surrounded by an excess electron charge, diffuse through the metal to the insulator interface, where they are deposited onto available sites as protons. The compensating negative charge will remain behind, within the conduction band of the metal gate. Because the diffusion through the Pd is very fast, it is assumed that there is a chemical equilibrium between the hydrogen on the outer surface of the metal gate and the protons at the insulator interface.

When negative voltage is applied to the gate, the conduction electrons in the n-type semiconductor are repelled from the semiconductor/insulator interface, forming a capacitive depletion layer, thus the total capacitance of the device will be a function of the applied voltage. The presence of the extra charge from the proton layer generates an additional electric field across the

AlN thus changing the potential of the AlN/Si interface, which in turn will modulate this capacitance as it decreases the thickness of the depletion layer.

This is the source of the electrical response of the device. Thus, at constant capacitance the presence of hydrogen in the surrounding gas causes a voltage shift. However, the Pd/AlN/Si device did not show any shift upon exposure to

NO2, indicating that the hydrogen could be a probe in order to investigate the interface properties of the devices.

227

7.2.2 Response of Al/AlN/Si to Hydrogen gas

Figure 7.15 shows the variation in capacitance of Al/AlN/Si device in dry air and in the presence of hydrogen gas. The results shown in Figure 7.15 reveal that the hydrogen-induced dipole layer is not formed in the metal–dielectric interface for MIS Al-AlN-Si device with a AlN dielectric. These C-V characteristics are quite anomalous and have not been reported yet. Because no work function change is observed as the work function change should be reflected in the flatband voltage shifts in the C-V curves. Therefore, it is obvious that the interface between the metal and the semiconductor plays a critical role in the interaction of hydrogen with semiconductor devices.

Although the detailed reaction mechanism of MIS (Al-AlN-Si) device to hydrogen is unknown at present, the plausible mechanisms are described as follows. First, the hydrogen-induced changes in the electrical properties may be related to the number of adsorption sites for hydrogen at the metal/dielectric/semiconductor interface of the devices [265]. Second, hydrogen could be contained in the SiO2 and AlN dielectrics during the sputter deposition, and some of the hydrogen molecules trapped at the metal/dielectric interface may affect the hydrogen detection sensitivity. Third, the AlN is well known to have a high density of trap states and hence potentially suffer from charge trapping instabilities [266].

228

Figure 7.14 Variation of capacitance of Pd/AlN/Si device with voltage

Figure 7.15 Variation of capacitance of Al/AlN/Si device with voltage

229

7.2.3 CONCLUSION

Gas sensing properties of Pd/AlN/Si and Al/AlN/Si towards H2 was compared. Al/AlN/Si device showed no response to hydrogen gas, whereas

Pd/AlN/Si device showed a voltage shift of 0.8 V in capacitance-voltage (C-V) curve, thereby giving an indication of the concentration of hydrogen in the gas to which the device is exposed. The presence of donor levels in the bandgap of the AlN layer with a wide range of energy is responsible for the observed shift in the CV curve of the Pd/AlN/Si. Furthermore, the Pd/AIN/Si device could not detect even 1000 ppm NO2, suggesting that the realization of chemical sensitivity of AIN to NO2 is not possible.

230

SUMMARY AND CONCLUSION

Metal oxide and nitride sensors, got remarkable positions in science and technology, since they allow producing fast, reliable, low cost and low maintenance devices. In the present work, field effect gas sensors (MIS capacitors) based on wide band gap Aluminum nitride (AlN), and resistive gas sensor made of indium tin oxide (ITO) have been developed for the detection of oxidizing and reducing gases. Optimization of individual ITO and AlN thin films with regards to the attainment of desirable levels of crystallinity, preferred orientation was done by dc and rf magnetron sputtering, respectively. A comprehensive study of the properties of these films using various characterization techniques helped to understand the sensor performance. Optimized thin film gas sensing layers were incorporated to form gas sensing device. A portable and low cost gas sensor chamber was fabricated and successfully utilized to monitor the electrical response and gas sensing performance of the material as function of gas concentration. With the results achieved from this study, the following conclusions can be made:

 ITO thin films deposited by rf sputtering revealed that the structural,

electrical and optical properties of the film greatly depended on

231

deposition conditions. The change in the resistivity ρ of the film with

annealing is very similar to the corresponding change in its sheet

resistance. The average transmittance was above 85% for the as-

deposited and annealed films. XRD patterns confirm the polycrystalline

nature of the film and the intensity of (4 0 0) diffraction peak improved

after annealing. This suggests that there is minimal structural change in

the ITO films followed by annealing. SEM micrograph of the annealed

film shows a uniform and closely packed distribution of polycrystalline

nanoclusters free of microcracks.

 By promoting the devation from the stoichiometry and decreasing the

resistance of the films, the performance of ITO material as gas sensor

was improved. The ITO film senor showed enhanced response at 600 K,

especially, at concentrations lower than 50ppm NO2. The ITO sensor

show a linear response to ethanol gas in the concentration range of 200

to 1400 ppm. Post-deposition annealing of the film at 700 K, ehanced

the response of the ITO senor to ethanol gas. Very low sensitivity is

observed towards Hydrogen gas when compared to NO2 and ethanol.

The maximum response was found to be only 1.6 at 400 K and 1000 ppm

of H2. The results suggest ITO films to be an excellent material for low

concentration, highly sensititive and selective NO2. The similar shape of

232

the sensitivity curves observed for ITO sensor towards NO2, H2 and

ethanol, resulted from the competition between slow kinetics at low

temperature and enchanced disorption at high temperature.

 AlN films prepared by dc reactive magnetron sputtering on Si (100)

substrate for different deposition time revealed that, films prepared at 8

min favored the formation of highly oriented (002) preferential plane

with enhanced crystal quality, which can provide good piezoelectric

response. The band gap increased with the increase in the deposition time

and the values of refractive index were in the range of 1.1-1.2, for the

samples prepared at 8 min of deposition time. The MIS structures

fabricated using AlN for different deposition time showed a significant

improvement of electrical characteristics with deposition time as we go

from 2 min to 8 min.

 The gas sensing properties of Al/AlN/Si device showed no response to

hydrogen gas, whereas Pd/AlN/Si device showed a voltage shift of 0.8 V

in capacitance-voltage (C-V) curve. However, the Pd/AlN/Si devices

could not detect even 1000 ppm NO2, suggesting that the realization of

chemical sensitivity of AlN to NO2 is not possible.

233

 Future work in this research will involve the development of an

electronic nose with oxide semiconducting material, which could detect

various gases. The more detailed interaction mechanism of hydrogen

with nitride – based semiconductor devcies should be studied.

Invesigations varying the gases, measurement temperature, device

structures and measurement methods may lead to the more specific

elucidation of the interaction mechanism.

234

REFERENCES

[1] Noboru Yamazo, Sens. Actuat. B 108 (2005) 2. [2] Dordrecht, Sensors and Sensory Systems for an Electronic Nose, Kluwer Academic Publishers, Boston, 1992. [3] Peter Grundler, Chemical Sensors an Introduction for Scientists and Engineers, Springer, Berlin and Heidelberg, New York, 2007. [4] Jacob Farden, Hand book of Modern Sensors: Physics, Designs, Applications, 3rd ed. Springer, Berlin and Heidelberg, New York, 2010. [5] J. Janata A. Bezegh, Chemical sensors, Anal. Chem. 60(12) (1988) 62. [6] S. Capone, A. Forleo, L. Francioso, R. Rella, P. Siciliano, J. Spadavecchia, D. S. Presicce, J. Optoelectron. Adv. M. 5(5) (2003) 1335. [7] J. T. Woetsman, E. M. Logothetis, Controlling Automotive Emissions, The Industrial Physicist 20-24, Ed. American Institute of Physics, 1995. [8] P. T. Moseley, Meas. Sci. Technol. 8 (1997) 223. [9] I. Lundstrom, Sens. Actuat. B 11 (1996) 35. [10] G. Harsanyi, Sensor. Rev. 20(2) (2000) 98.

[11] Teruhisha Kudo, Mutsumi Kimura, Kenji Hanabusa, Hirofusa Shirai,

Takehiko Sakaguchi, J. Porphyr. Phthalocya. 3 (1999) 65.

[12] W. Hu, Y. Liu, Y. Xu, S. Liu, S. Zhou, P. Zeng, D. B. Zhu, Sens. Actuat. B 56 (1999) 228.

[13] H. Meixner, U. Lampe, Sens. Actuat. B 33 (1996) 198. [14] A. Vancu, R. Ionescu, N. Bârsan, Thin Film Resistive Sensors, P. Ciureany, S. Middelhoek (ed.), IOP Publishing Ltd. 1992, p. 437.

235

[15] A. Hulnnicki, S. Glab, F. Ingman, Pure and Appl. Chem. 63 (1991) 1247. [16] G. Sberveglieri (ed.), Gas Sensors - Principles, Operation and Developments, Kluwer Academic Publishers, Dordrecht, 1992.

[17] J. Chou (ed.), Hazardous Gas Monitors-A Practical Guide to Selection, Operation and Applications, SciTech Publishing, 2000. [18] J. W. Gardner, P. N. Bartlett, Electronic noses - Principles and Applications, Oxford University Press, 1999. [19] H. Meixner, J. Gerblinger, U. Lampe, M. Fleischer, Sens. Actuat. B 23 (1995) 119. [20] T. Takeuchi, Sens. Actuat. B 14 (1988)109. [21] Emery Lightner Moore and Ramon Perez de Paula, Optical fibers and Integrated Optics (1) 2008 p.217-245. [22] D. Rutven, Principles of adsorption and adsorption processes, Wiley- Interscience, New York, 1984. [23] W. Gopel, Solid state chemical sensors: Atomistic models and research trends, Sensor Actuator 16 (1989) 167. [24] D. Shaw, Introduction to colloid and surface chemistry, Butterworth- Heineman, Oxford, 1992. [25] R. I. Mase, Principles of adsorption and reaction on solid surfaces, Wiley, New York, 1996. [26] I. Langmuir, J. Am. Chem. Soc. 40 (1918) 1361. [27] P. J Barth, B. Muller, U. Wagner, A. Bittinger, Eur. Respir. J. 8 (1995) 1115. [28] M. Wegmann, A. Fehrenbach, S. Heimann, H. Fehrenbach, H. Renz, H. Garn, U. Herz, Expt. Tox. Path. 56 (2005) 341.

236

[29] Fang Song, Huilan Su, Jie Han, Di Zhang, Zhixin Chen, Nanotechnology 20 (2009) 495502. [30] Y. D. Wang, C. L. Ma, X. H. Wu, X. D. Sun, H. D. Li, Sens. Actuat. B 85 (2002) 270. [31] C. G. Grangvist, A. Hultaker, Thin Solid Films, 411 (2002) 1. [32] I. Lundstrom, S. Shivaraman, C. Svensson, L. Lundkvist, Appl. Phys. Lett. 26 (1975) 55. [33] H. I. Chen, C. K. Hsiung, Y. I. Chou, Semicond. Sci. Tech. 18 (2003) 620. [34] F. Solzbacher, C. Imawan, H. Steffes, E. Obermeier, M. Eickhoff, Sens. Actuat. B 78 (2001) 216. [35] Y. Gurbuz, W. P. Kang, J. L. Davidson, D.V. Kerns, IEEE. T. Electron. Dev 46 (1999) 914. [36] P. Salomonsson, E. Jobson, B. Haggendal, J. Nytomt, C. Carlsson, M. Glavmo, A. Baranzahi, Sens. Actuat. B 43 (1997) 52. [37] C. Chang, G. C Chi, W. Wang, L. Chen, K. H Chen, F. Ren, S. J. Pearton J. Electron. Mater 35 (2006) 738. [38] C.Y Chang, T. W Lan, G. C Chi, L. C Chen, K. H Chen, J. J Chen, S. Jang, F .Ren, S. J. Pearton, Electrochem. Solid-State Lett. 9 (2006) G155. [39] R. Mehandru , B. Luo , B. S. Kang, J. Kim, F. Ren, S. J. Pearton, C. C Fan, G. T. Chen, J. I. Chyi, Solid-State Electron. 48 (2004) 351. [40] W. I Park, G.C Yi, M.Y Kim, S.J Pennycook, Adv. Mater. 15 (2003) 526. [41] T. P. Nguyen, Le. Rendu, N. N. Dihn, M. Fourmigue, C. Meziere, Synth. Met 138 (2003) 229.

237

[42] Y. H Tak, K. B Kim, H. G. Park, K. H. J. R. Lee, Thin Solid Films, 12 (2002) 411. [43] J Herrero, C. Guillen, Vacuum 67 (2002) 611. [44] M. D. Stoev, J. Touskova, J. Tousek, Thin Solid Films 67 (1997) 299. [45] V. Teixeira, H. N. Cui, L. J Meng, E. Fortunato, R. Martins, Thin Solid Films 70 (2002) 420. [46] M. Kojima, F. Takahashi, K. Kinoshita, T. Nishibe, M. Ichidate, Thin Solid Films 392 (2001) 349. [47] Y. S. Kim, Y. C. Park, S. G. Ansari, J. Y. Lee, B. S. Lee, H. S. Shin, Surf. Coat. Technol. (2011) 78 [48] M. S Hwang, H. J Lee, H. S Jeong, Y. W Seo, S J Kwon, Surf. Coat. Technol 29 (2002) 171. [49] R. Groth, E. Kauer, Philips. Tech. Rev 26 (1965) 105. [50] R. Groth. Phys. Status. Solidi 14 (1966) 69. [51] I. Elfallal, R. D. Rilkington, A. E. Hill, Thin Solid Films 223 (1993) 303. [52] G. B. Gonzalez, T. O. Mason, J. P. Quintana, J. Appl. Phys. 96( 1996) 7. [53] E. Fortunato, D. Ginley, H. Hosono, D. C. Paine, Mater. Res. Soc. Bull. 32 (2007) 242. [54] A. E. Hicho, A. Kachouane, J. L Bubendorrff, M. Troyon, A. Bougrine, Thin Solid Films 458 (2004) 263. [55] L. Gupta, A. Mansingh, P. K. Srivastava, Thin Solid Films 176 (1989) 33. [56] H. Zhou, W. Cai, L. Zhang, Appl. Phys. Lett. 75 (1999) 495. [57] K. L. Chopra, S. Major, D. K. Pandya, Thin Solid Films 102 (1983). [58] Schinich Honda, Michio Watamori, Kenjiro Oura, Thin Solid Films 281 (1996) 206.

238

[59] J. C. Mamifacier, M. de. Munaicia, J. P. Fillard, E. Vicario, Thin Solid Films 70 (2002) 420. [60] G. Neri, A. Bonavita, G. Micali, G. Rizzo, N. Pinna, M. Niederberger, Thin Solid Films 515 (2007) 8631. [61] R. Das, K. Adhilaru, S. Ray, Appl. Surf. Sci. 253 (2007) 6068. [62] M. Bender, J. Trube, J. Stollenwerk, Thin Solid Films 354 (1999) 100. [63] V. Vasu, A. Subramanyam, Thin Solid Films 2 (1995) 971. [64] D. Liufu, K.C. Kao, J. Vac. Sci. Technol., A 16 (1998) 2360. [65] M. A. Dubois, P. Muralt, Appl. Phys. Lett. 74 (1999) 3032. [66] M. T. Wauk, D. K. Winslow, Appl. Phys. Lett. 13 (1968) 286. [67] R. F. Davis, Proc. IEEE 79 (1991) 702. [68] R. Rodriguez-Clemente, B. Aspar, N. Azema, B. Armas, C. Combescure, J. Durand, A. Figueras, J. Cryst. Growth 133 (1993) 59. [69] M. A. Dubois, P. Mulart, Appl. Phys. Lett. 74 (20) (1999) 3032. [70] A. Fathimulla, A. A. Lakkani, J. Appl. Phys, 54 (1983) 4586. [71] C. M. Zetterling, M. Osteling, C. I. Harris, N. Nordell, K. Wongchotigul, M. G. Spencer, Mater. Sci. Forum. 264-268 (1998) 877. [72] A. Bourret, A. Barski, J. L. Roouviere, G. Renaud, A. Barbier, J. Appl. Phys. 83 (1998) 2003. [73] Z. V. Fan, G. Rong, N. Newman, J. Smith, Appl. Phys. Lett. 76(14) (2000) 1839. [74] Y. E Lu, Z. M. Ren, T. C. Chong, B. A. Cheong, S. K. Show, J. P. Wang, J. Appl. Phys. 87(3) (2000) 1540. [75] T. Adam, J. Kolodzey, C. P. Swann, M. W. Tsao, J. F. Rabolt, Appl. Surf. Sci. 175-176 (2001) 428. [76] K. S. Stevens, M. Kinnibourgh, A. F. Schvartzman, A. Ohtani, Appl. Phys. Lett. 66(23) (1995) 31793181.

239

[77] G. A. Slack, T. F. McNelly, J. Cryst. Growth 34 (1976) 263. [78] A. Kurokawa, K. Utsumi, H. Takamizawa, T. Kamata, S. Noguehi, IEEE Tran. Comp. Hib. Manuf. Tech., CHMT 8(2) (1985) 247. [79] K. M. Taylor, C. Lenie, J. Electchem. Soc. 107(4) (1960) 308 [80] Hadis Morkog, Handbook of Nitride Semiconductors and Devices, WILEY-VCH Verlag GmbH & Co., 2008. [81] J. Chaudhuri, R. Thokala, J. H. Edgar, B. S. Sywe, J. Appl. Phys. 77 (1995) 6263. [82] B. H. Lee, I. C. Kim, S.W. Cho, S. H. Lee, Thin Solid Films 302 (1997) 25. [83] R. N. Josh, V. P. Singh, J. C. McClure, Thin Solid Films 257 (1995) 32. [84] W. T. Wu, B. S. Chiou, Thin Solid Films 247 (1994) 201. [85] T. Minami, H. Sonohara, T. Kakumu, S. Takata, Thin Solid Films 270 (1995) 37. [86] T. Karasawa, Y. Miyata, Thin Solid Films 223 (1993) 135. [87] Y. Higuchi, Jpn. Surf. Techno. 88 (1993) 103. [88] M. Ishihara, S. J. Li, H. Yumoto, K. Akashi, Y. Ide, Thin Solid Films 316 (1998) 152. [89] X. H. Xu, H. S. Wu, C J. Zhang, Z. H. Jin, Thin Solid Films 388 (2001) 62. [90] K. Kusaka, D. Taniguchi, T. Hanabusa, K. Tominaga, Vacuum 59 (2000) 806. [91] D. Y. Wang, Y. Nagahata, M. Masuda Y. Hayashi, J. Vac. Sci. Technol., A 14 (1996) 3092. [92] R. N. Tait, A. Mirfazli, J. Vac. Sci. Technol., A 19 (2001) 1586. [93] L. Wu, S. Wu, H. T, J. Vac. Sci. Technol., A 19 (2001) 167.

240

[94] K. Sugiyama, K. Taniguchi, K. Kuwabara, J. Mater. Sci. Lett. 9 (1990) 489. [95] C. C. Cheng, Y. C. Chen, R. C. Horng, J. Vac. Sci. Technol., A 16 (1998) 3335. [96] G. A. Slack, J. Phys. Chem. Solids 34 (1973) 321. [97] C. R. Aita, C. J. Gawlak, J. Vac. Sci. Technol., A 1 (1983) 403. [98] H. C. Lee, G. H. Kim, S. K. Hong, K. Y. Lee, Y. J. Yong, C. H. Chun, J. Y.Lee, Thin Solid Films, 261 (1995) 148. [99] L. Kao, P. J. Shih, C. H. Lai, Jpn. J. Appl. Phys. 38 (1999) 1526. [100] E. J. Bienk, H. Jensen, G. N. Pedersen, G. Sorensen, Thin Solid Films 230 (1993) 121. [101] K. Tominaga, T. Ao, I. Mori, K. Husaka, T. Hanabusa, Jpn. J. Appl. Phys. 35 (1996) 4972. [102] F. S. Ohuchi, P. E. Russell, J. Vac. Sci. Technol., A 5(1987) 1630. [103] John A. Thornton, Alan S. Penfold, Cylindrical magnetron sputtering, Chapter II, Academic Press, 1978 p. 75–113. [104] Mitsuharu Konuma, Film deposition by plasma techniques, volume 10, Atoms and Plasmas, Springer-Verlag, 1992 [105] Brian N. Chapman, Glow discharge processes, John Wiley & Sons, 1980 [106] Skoog, et al. Principles of Instrumental Analysis. 6th ed. Thomson Brooks/Cole, 2007 349-351. [107] D. Borchert, R. Job, G. Grabosch, C. Wolffersdorf, W. R. Fahrner, Proceedings of the Thirteenth European Photovoltaic Solar Energy Conference, 1(1995) p 249. [108] W. F. Wu, B. S. Chiou, Appl. Surf. Sci. 115(1997) 96. [109] J. S. Kim, M. Granstrom, R. H. Friend, J. Appl. Phys. 84 (1998) 6869.

241

[110] P. Thilakan, S. Kalainathan, J. Kumar, P. Ramasamy, J. Electron. Mater. 20(1995) 719. [111] G. Frank, H. Kostlin, Appl. Phys. A 27 ( 982) 197. [112] T. Maruyama, K. Fukui, J. Appl. Phys. 70 (1991) 3848. [113] X. W. Sun, H. C. Hung, H. S. Kwok, Appl. Phys. 68 (1996) 2663. [114] R. H. Tahar, T. Ban, Y. Ohya, Y. Takahashi, J. Appl. Phys. 82 (1997) 865. [115] E. Nishimura, M. Ando, K. Onisawa, M. Takabatake, T. Minemura, Jpn. J. Appl. Phys. 35 (1996) 2788. [116] M. Mizuhshi, Thin Solid Films 70 (1980) 91. [117] C. V. R. V. Kumar, A. Mansigh, J. Appl. Phys. 65 (1989) 1270. [118] M. Tariq Bhatti, Anwar Manzoor Rana, Abdul Faheem Khan, Mater. Chem. Phys. 84 (2004) 126-130 [119] M. Kamei, T. Yagami, S. Takaki, Y. Shigesato, Appl. Phys. Lett. 64 (1994) 2712. [120] C. H. Yi, I. Yasui, Y. Shigesato, Jpn. J. Appl. Phys. 34(1995) 1638. [121] H. Gong, J. Q. Hu, J. H. Wang, C. H. Ong, F. R. Zhu, Sens. Actuat. B 115 (2006) 247. [122] A. Salehi, Sens. Actuat. B 96 (2003) 88. [123] P. Bogdanov, M. Ivanovskaya, E. Comini, G. Faglia, G. S. Berveglieri, Sens. Actuat. B 57 (1999) 153. [124] B. Bessais, N. Mliki, R. Bennaceur, Semicond. Sci. Technol. (1993) 116. [125] N. G. Patel, P. D. Patel, V. S. Vaishnav, Sens. Actuat. B 96 (2003) 180. [126] V. Demarne, R. Sanjines, G. Sberveglieri (Ed), Gas Sensor Principles Operation and Developments, Ch. 3, Kluwer Academic Publishers, Netherlands, (1992) p. 89.

242

[127] Wolfgang Gfpel, Arnaldo D‘Amico, Giogio Sberveglieri (Ed.), Sensors for Domestic Applications, Part I, Proc. of first European School on Sensors (ESS94), World Scientific, (1995) p. 49. [128] V.S. Vaishnava, P. D. Patelb, N.G. Patelc, Thin solid Films 487 (2005) 277. [129] Joydeep Dutta, Swati Ray, Thin Solid Films, 164 (1988) 119. [130] A. Mansingh C. V. R. V. Kumar, J. Appl. Phys. 22 (1989) 455. [131] C. H. L. Weijtens, J. Electrochem. Soc. 138 (1991) 3432. [132] A. Tanusevski, Semicond. Sci. Technol. 18 (2003) 501. [133] V. V. Vidya, V. Malthy, T. Balasubramanian, Curr. Appl. Phy. (2010) 1. [134] Miaoju Chuang, J. Mater. Sci. Technol. 26(7) (2001) 577-583. [135] F. M. Amanullah, K. J. Prathap, V. Hari Babu, Thin Solid Films 254 (1 995) 28 [136] P. Thilakan, J. Kumar, Vacuum 48 (1997) 463. [137] Choong Hoon Yi, ltan Yasui, Yuzo Shigesato, Satoru Takaki, Jpn. J. Appl. Phys. 34(1995) 244. [138] A. K. Kulkarnia, Kirk H. Schulzb, T. S. Lima, M. Khan, Thin Solid Films 345 (1999) 273. [139] Jin Ma, Shu-Ying Li, Jun-qing Zhao, Hong Lie Ma, Thin Solid Films 307(1997) 200. [140] J. C. C. Fan, F. J. Bachner, J. Electrochem. Soc. 122 (1975) 1719. [141] I. Adesida, D. G. Ballegeer, J. W. Seo, A. Katterson, H. Chang, K.Y. Cheng, T. Gessert, J. Vac. Sci. Technol., B 9 (1991) 3551. [142] H. W. Zhang, W. Xu, Vacuum 43 (1992) 835. [143] F. H. Mullins, A. Brunn Schweiller, Solid-State Electron. 19 (1976) 47. [144] B. D. Cullity, Elements of X Ray Diffraction, Addition-Wesley Massachusetts, 1956.

243

[145] A. Mohammadi Gheidari, F. Behafarid, G. Kavei, M. Kazemzad, Mater. Sci. Eng. A 136 (2007) 37. [146] T. Karasawa, Y. Mi yata, Thin Solid Films 223 (1 993) 135. [147] Y. Shigesato, Y. Hayashi, T. Haranoh, Appl. Phys. Lett. 6 (1992) 73. [148] M. J. Alam, D. C. Cameron, Thin Solid Films 420-421 (2002) 76. [149] L. R. Cruz, C .Legnani, I. G. Matoso, C. L Ferreira, H. R. Moutinho, J. Phys. D: Appl. Phys. 39 (2006)184. [150] G. Goncalves, E. Elangovan, P. Barquinha, L. Pereira, R. Martins, E. Fortunato, Thin Solid Films 515 (2007) 8562. [151] Y. Hu, X. Diao, C. Wang, W. Hao, T. Wang, Vacuum 75 (2004) 183. [152] E. Terzini, G. Nobile, S. Loreti, C. Minarini, T. Polichetti, P. Thilakan, Jpn. J. Appl. Phys. 38 (1999) 3448. [153] E. Terzini, P. Thilakan, C. Minarini, Mater. Sci. Eng. B 77 (2000) 110. [154] M. P. R. Panicker, W. F. Essinger, J. Electrochem. Soc. 128 (1981) 1943. [155] C. Guillen, J. Herrero, Thin Solid Films 403 (2003) 431. [156] L. Kerkach, A. Layadi, E. Dogheche, D. Remiens, J. Phys. D: Appl. Phys.39 (2006) 184. [157] Hiroshi Morikawa , Miya Fujita, Thin Solid Films 339 (1999) 309. [158] F. Barlow, M. A. Naby, A. Joshy, A. E. Riad, Sol. Energy Mater. Sol. Cells. 33 (1994) 63. [159] L. Meng, M. P. dos Santos, Thin Solid Films 322 (1998) 56. [160] T. P. Drusedau, J. Blasing, Thin Solid Films 27 (2000) 377. [161] X. H. Ji, S. P. Lau, G. Q. Yu, W. H. Zhong, B. K. Tay, J. Phys. D: Appl. Phys. 37 (2004) 1472. [162] M. Clement, L.Vergara, J. Sangardor, E. Iborra, A. Sanz-Herras, Ultrasonic 42 (2004) 403.

244

[163] T. Palacios, F. Calle, E. Monroy, J. Grajal, M. Eickhoff, O. Ambacher, C. Prieto, Mat. Sci. Eng. B 93 (2000) 154. [164] H. Gong, X. Jiang, J. Cryst. Growth 235 (2000) 1. [165] A. J. Shuskus, T. M. Reeder, E. L. Paradis, Appl. Phys. Lett. 24(4) (1974) 155. [166] M. Ishihara, K.Yamamoto, F.Kokai,Y.Koga, Vaccum 59 (2000) 649. [167] L. Huang, X. D. Wang, K. W. Hipps, U. Mazur, R. Heffron, J. T. Dickinson, Thin Solid Films 279 (1996) 43. [168] M. M. D. Ramos, J. B. Almeida, M. I. C. Ferrerira, M. P. D. Santos, Thin Solid Films 176 (1989) 219. [169] J. Huang, L. Wang, Q. Shen.C. Lin, O. Milae, Thin Solid Films 340 (1999) 137. [170] V. Dimitrova, D. Monova, T. Paskova, T. Z. Uzunov, N. Lvanoc, D. Dechev, Vaccum 51 (1998) 161. [171] Y. F. Lu, Z. M. Ren, T. C. Chong, B. A. Cheong, S. K. Chow, J. P. Wang, J. Appl. Phys. 87(3) (2000) 1540. [172] G. Amsel, J. P. Nadia, E. D‘ Artemare, D. David, E. Girard, J. Moulin, Nucl. Instrum. Methods. 92 (1971) 481. [173] J. Ziegler (ed), New uses of Ion accelerators, Plenum Press, New York (1975).

[174] H. C. Lee, G. H. Kim, S. K. Hong, K. Y. Lee, Y. J. Yong, C. H. Chun and J. Y. Lee, Thin Solid Films 261(1995)148. [175] L. Kao, P. J. Shih, C. H. Lai, Jpn. J. Appl. Phys. 38 (1999) 1526. [176] K. Tominaga, T. Ao, I. Mori, K. Husaka, T. Hanabusa, Jpn. J. Appl. Phys. 35 (1996) 4972.

245

[177] W. T. Lim, B. K. Son, D. H. Kang, C. H. Lee, Thin Solid Films 382 (2001) 56. [178] H. Okano, N. Tanaka, Y. Takahashi, T. Tanaka, K. Shibata, S. Nakano, Jpn. J. Appl. Phys. Lett. 31(1992) 3446. [179] N. Kumar, K. Pourrezaei, R. J. Demria, B. Singh, Mater. Res. Soc. Symp. Proc., 68 (1986) 357. [180] H. Yumoto, M. Ishihara, T. Kaneko, J. Jpn. Assoc. Crst. Growth, 23 (1996) 382. [181] F. Engelmark, G. F. Iruarte, I. V. Katardjiev, M. Ottosson, P. Muralt, S. Berg, J. Vac. Sci. Technol., A 19, (2001) 2664 [182] Xiao-Hong Xu, Hai-Shun Wu, Cong-Jie Zhang, Zhi-Hao Jin, Thin Solid Films 388 (2001) 62. [183] K. Kusaka, D. Taniguchi, T. Hanabusa, K. Tominaga, Vacuum 59 (2000) 806. [184] A. Sanz Hervas, E. Iborra, M. Clement, J. Sangrador, M. Aguilar, Diamond Relat. Mater. 12 (2003) 1186. [185] L. Wu, S. Wu, H. T. Song, J. Vac. Sci. Technol., A 19 (2001) 167. [186] M. B. Assouara, M. E. Hakikia, O. Elmazriaa, P. Alnota, C. Tiusanb, Diamond Relat. Mater. 13 (2004) 1111. [187] Xiufeng Song, Renli Fu , Hong He, Microelectron. Eng. 86(11) ( 2009) 2217. [188] A. Stafiniak, D. Muszynska, A. Szyszka, B. Paskiewicz, K. Ptasinski, S. Patela, R. Paszkiewicz, Marek Tlaczala, Opp. Appl. 39 (2009) 4. [189] V. Dimitrova, D. Manova, E. Valcheva, Mater.Sci. Eng. B 68 (1999) 1. [190] K. Jagannadham, A. K. Sharma, Q. Wei, R. Kalayanraman, J. Narayan, J. Vac. Sci. Technol., A 16 (1998) 2804.

246

[191] Z. Q. Yao, Y. Q. Li, J. X. Tang, W. J. Zhang, S. T. Lee, Diamond Relat. Mater. 17 (2008) 1785. [192] S.Y. Bae, H. W. Seo, D. S. Han, M. S. Park, W. S. Jang, C. W. Na, J. Parka, C. S. Park, J. Cryst. Growth 258 (2003) 296.

[293] A. M. Ruiz, A. Cornet, K. Shinmanoe, J.R Morante, N.Yamazoe, Sens. Actuat. B (2005) 8.

[194] Y. Zhang, X. L. He, J. P. Li, Z. J Miao, F. Huang, Sens. Actuat. B 132 (2008) 67-73.

[195] G. Wang, Y. Ji. X. R. Huang, X. Q Yang, P. I. Gouma, Duley, J. Phys. Chem. B 110(47), (2006) 23777. [196] R. Hao, J. Y Yuan, Peng, Q. Chem. Lett. 6(5) (2004) 249. [197] Zheng Jiao, Minghong Wo, Jianzhong Gu, Xilian Sun, Sens. Actuat. B 94 (2003) 216. [198] K. K. Makhija, Arabinda Ray, R. M. Pater, U. B. Trivedi, Bull. Mater. Sci. 28 (2005) 9. [199] V. E. Henrich, Cox, P.A. The Surface Science of Metal Oxides; Cambridge University PressCambridge, UK, 1994 [200] G. Behr, W. Fliegel, Sens. Actuat. B 26 (1995) 33-7.

[201] H. Steffes, C. Imawan, F. Solzbacher, E. Obermeier, Sens. Actuat. B 78 (2001) 106. [202] E. Comini, C. Baratto, G. Faglia, M. Ferroni, A. Vomiero, G. Sberveglieri, Prog. Mater Sci. 54 ( 2009)1.

247

[203] G. Korotcenkov, Mater. Sci. Eng. B 139 (2007) 1. [204] P. T Moseley, J. O. Norris, D. E. Williams, Eds. Techniques and Mechanisms in Gas Sensing, Adam Hilger, Bristol, UK, 1991. [205] M.J Madou, S.R. Morrison, Chemical Sensing with Solid State Devices Academic Press, Inc., Harcourt Brace Jovanovich Publ., Boston, MA, USA, 1987. [206] G.Sberveglieri E.d Gas Sensors-Principles Operation and Developmens, Kluwer Academic Publishers Dordrecht, Netherlands, 1992. [207] G. Korotcenkov. Sens. Actuat B. 107 (2005) 209. [208] Z. Lu, S. M.Kanan, C. P. Tripp, J. Mater. Chem. 12 (2002) 983. [209] D. Williams, Sens. Actuat. B 57 (1999) 1. [210] Kupriyanov, L.Y., Ed. Semiconductor Sensors in Physico-Chemical Studies; Elsevier:Amsterdam, The Netherlands, 1996. [211] S. R. Morrison, Sens. Actuat. 11(1987) 283. [212] R. Van de Krol, H. L. Tuller, Solid State Ionics 150 (2002) 67. [213] N. Yamazoe, G. Sakai, K. Shimanoe, Catal. Surv. Asia 7 (2003) 63. [214] N. Yamazoe, Sens. Actuat. B. 108 (2005) 2. [215] Faramarz Hossein-Babaei, Mohammad Orvatinia, Sens. Actuat. B 89 (2003) 256. [216] W. H. Brattain W.H, J. Bardeen, Bell. Syst. Tec. J. 1 (1953) 1. [217] G. Heiland, Zum Einfluss von Wasserstoff, Phys. 138 (1954) 459. [218] P. C. Ford, I. M. Lorkovic , Chem. Rev. 102 (2002) 993. [219] M. C. Kung, H. H. Kung, Cat. Rev. - Sci. Eng. 27 (1985) 425. [220] S. M. Sze, Physics of semiconductor devices‖, Publisher: John Wiley and Sons (WIE) Publication ,1981. [221] E. S. Yang, Fundamentals of semiconductor devices, McGraw-Hill, Inc., USA, 1978.

248

[222] K. J. Albert, N.S. Lewis, Chem. Rev. 100 (2000) 2595. [223] M. Penza, G. Cassano, A. Sergi, C. Sterzo, M.V. Russo, Sens. Actuat. B 81(1) (2001) 88. [224] A. Tomchenko, G. P. Harmer, G.P, B.T Marquis, J.W Allen, Sens. Actuat. B 93 (2003) 126. [225] P. C. Ford, I.M. Lorkovic, Chem. Rev. 102 (2002) 993. [226] Xiangdong Lou, ChuanyunPeng, XiaobingWang, WenfeiChu, Vacuum 81 (2007) 883. [227] C. Cantalini, W. Wlodarski, H. T. Sun, M. Z. Atashbar, M. Passacantando, A. R. Phani, S. Santucci, Thin Solid Films 350 (1999) 276. [228] Russell Binions, Claire J. Carmalt, Ivan P. Parkin, Meas. Sci. Technol. 18 (2007) 190. [229] G. Xu, Y.W. Zhang, X. Sun, C.L. Xu, C.H. Yan, J. Phys. Chem. B 109 (2005) 3269. [230] L.Francioso, A. Forleo, S. Capone et al. Sens. Actuat. B 114 (2006) 646. [231] C. Aifan, H. Xiaodong, T. Zhangf, Sens. Actuat. B 115 (2006) 316. [232] M. Schierbaum, U. Weimar, W. Gopel, R. Kowalkowski, Sens. Actuat. B 3 (1991) 205. [233] C. Cantalini, M. Pelino, H. T. Sun, Santyccu, Lozzi, M. Passacantado, thin films Sens. Actuat. B 35 (1996) 112. [234] E. Comini, G. Sberveglien, M. Ferroni, V. Gluidi, G. Martinelli, Sens. Actuat. B 68 (2000) 175. [235] Kap Duk Song, Jung-Il Bang, Sang RokLee et al. Sens. Actuat. B 108 (2005) 211. [236] M. E. Franke, T. J. Koplin, U. Simon, Small 2 (2006) 36.

249

[237] С. Go, С. Xie, S. Cai, Mater. Sci. Eng. В 137 (2007) 53. [238] H. V. Shurmer, J.W. Gardner, Sens. Actuat. B 8 (1992) 1. [239] N. Barsan, M. Schweitzer- Berberich, W. Gopel, J. Fresenius, Anal. Chem. 365 (1999) 287. [240] Jin Ma, Shu-Ying Li, Jun-Qing Zhao, Hong-Lei Ma, Thin Solid Films 307 (1997) 200. [241] P. T. Mosely, Jow Norris, De Williams, Techniques and Mechanisms in Gas Sensing, The Adam Hilger Series On Sensors, IOP publishing Ltd, New York, NY,1991. [242] Faramarz Hossein-Babaei, Mohammad Orvatinia, Sens. Actuat. B 89 (2003) 256. [243] Q. Qi, T. Zhang, X. Zheng, H. Fan, L. Liu, R. Wang, Y. Zeng, Sens. Actuat. B 134 (2008) 36. [244] J. Gong, Q. Chen, M. Lian, N.Liu, R.G. Stevenson, F. Adamic, Sens. Actuat. B 114 (2006 ) 32. [245] E. Traversa, Sens. Act. B 23 (1995) 135. [246] A. Kolmakov, D. O. Klenov, Y. Lilach, S. Stemmer, Moskovits, Nano Lett. 5 (2005 ) 667-673. [247] J. Zing, J. Zhan, Adv. Mater. 20 (2008) 4547. [248] V. V. Malyshev, A. V. Pislyakov, Sens. Actuat. B 134 (2008) 913-921. [249] I. D. Kim, A. Rothschild, B. H. Lee, D. Y. Kim, S. M. Jo, H. L. Tuller, Nano Lett. 6 (2006) 2009. [250] M. Cao, Y. Wang, T. Chen, M. Antonietti, M. A. Niederberger, Chem. Mater. 20, (2008) 5781. [251] N. V. Duy, N. V. Hieu, P. H. Huy, N. D. Chien, M.T. Hhamilselvan, J. Yi Physica E 41 (2008) 258.

250

[252] S. M Chou, F. С Teoh, W. H. Fai, Y. H. Su, M. H. Hon, Sensors 6 (2006) 1420. [253] G. Korotcenkov, M. Ivanov, I. Blinov, J.R. Stetter, Thin Solid Films 515 (2007) 3987. [254] A. T. Winzer, R. Goldhahn, G. Gobsch, A. Dadgar, A. Krost, O. Weidemann, M. Stutzmann, M. Eickhoff, Appl. Phys. Lett. 88 (2006) 024101. [255] I. Lundstrom, H. Sundgren, F. Winquist, M. Eriksson, C. K. Rulcker, A. L. Spetz, Sens. Actuat. B 121 (2007) 247. [256] S. J. Pearton, F. Ren, Y. L.Wang, B. H. Chu, K. H. Chen, C. Y. Chang, W. Lim, J. Lin, D. P. Norton, Prog. Mater Sci. 55 (2010) 1. [257] S. J. Pearton, B. S. Kang, S. Kim, F. Ren, B .P .Gila, C. R. Abernathy, J. Lin, S.N.G. Chu, J. Phys-Condens. Mat 16 (2004) 961-994. [258] I. Lundstrom, T.DiStefano, Solid State Commun. 19 (1976) 871. [259] I. Lundstrom, M.S. Shivaraman, C.Svensson, Surf. Sci. 64, (1977) 497. [260] D. Soderberg, I. Lundtrom, Solid State Commun. 35 (1980) 169-174. [261] L. G. Ekedahl, M. Eriksson, I. Lundstrom, Acc. Chem. Res. 31 (1998) 249. [262] M. C. Petty, Solid- State Electron. 29 (1986) 89. [263] M. Eriksson, A. Salomonsson, I. Lundstrom, D. Briand, A. E. Abom, J. Appl. Phys. 98, (2005) 034903. [264] J. Fogelberg, M. Eriksson, H. Dannetun, L. G. Petersson, J. Appl. Phys. 78 (1995) 988. [265] S. M. Sze, Physics of semiconductor devices, John Wiley and Sons, 1981 [266] M. J. Powell, Appl. Phys. Lett. 43 (1983) 597.

251

LIST OF PUBLICATIONS / PRESENTATION

Publications

1. Influence of Deposition Parameter and Heat Treatment on the NO2 Sensing Properties of Nano Structured ITO Thin Film, K.Vijayalakshmi, C.Ravidhas, V.Vasanthi Pillay and D.Gopalakrishna Thin solid films, 519 (2011) 3378-3382.

2. Effect of heat treatment on the NO2-sensing properties of sputter-deposited indium tin oxide thin films, K. Vijayalakshmi and Vasanthi V. Pillay, Philosophical Magazine Letters, 91(10) 2011, 682–689. 3. Micro Structural Evolution and Growth of Highly A Axis Oriented AlN Films by D.C. Magnetron Sputtering, V. Vasanthi Pillay and K. Vijayalakshmi, International Journal of Nanotechnology and Applications, 5(3) 2011 351-358. 4. Micro Structural Evolution and dielectric properties of 1-D AIN powder synthesis by microwave technique V. Vasanthi Pillay and K. Vijayalakshmi, Super lattices and microstructures, (article in press, 2012).

5. Preparation and characterization of sputter Deposited ITO thin films for

NO2 sensor. Proceedings of the International conference on characterization, consolidation and modeling of Nanomaterials (ICON), Department of Metallurgical Engineering, PSG College of Technology, Coimbatore. (5-6 March 2010).

6. Effect of Annealing on the structural, morphological, electrical and ethanol sensing properties of ITO thin film, (Communicated to the Journal of Applied Sciences).

252

7. Influence of deposition time on the microstructure and dielectric properties of AIN / Si thin film for enhanced hydrogen sensing application (Communicated to the Journal of Microelectronic Engineering).

Papers Presented in the Conferences

1. Properties of multilayered Cd2SnO4-In2O3:Sn transparent conducting oxide film by RF sputtering technique, National Conference on Advanced materials devices and technologies (NCONS), Shri Venkateshwara University, Tirupati. (Feb 20-22,2008). 2. Structural and optical characterization of CTO thin films by rf sputtering technique, National conference on Nano materials preparation characterization and devices, Bishop Heber College Tiruchirappalli, March 2008. 3. Preparation and characterization of Nano Structured ITO thin films by rf sputtering, National conference on thin films and applications, School of Electrical and Electronics Engineering, Sastra University Thanjavur, (March 2009). 4. Optical and structural characterization of Nano structured ITO thin films by rf sputtering, National Conference on Advanced Materials processing Characterization and applications, PSG College of Technology Coimbatore, (Feb 13 - 14, 2009). 5. Effect of Annealing on the structural, morphological, electrical and ethanol sensing properties of ITO thin film, International conference on thin film and applications, School of Electrical and Electronics Engineering, Sastra University, Thanjavur, (March 2012).

253