Thermally Evaporated Tungsten Oxide (WO3) Thin Films for ... · Thermally Evaporated Tungsten Oxide (WO3) Thin Films for Gas Sensing Applications Submitted in fulfilment of the requirements

Thermally Evaporated Tungsten Oxide (WO3) Thin Films for Gas Sensing

Applications

Submitted in fulfilment of the requirements of Doctor of Philosophy

Mohammed Ahsan

School of Engineering Systems Faculty of Built Environment and Engineering

Queensland University of Technology Australia

2012

ii

Statement of Original Authorship

The present thesis reports the results of the work done during the years of my PhD

project. The work has been carried out mostly at Queensland University of

Technology although some experiments were performed also at University of

Queensland, Australian National University, Australian Nuclear Science and

Technology Organization and RMIT University.

I hereby declare that:

All the experiments have been performed during the PhD project.

All the results presented come directly from the experimental activity done.

All the interpretations and observations are based on the results of the

experiments and have been often referred, with the reported reference, to previous

literature results.

For these reasons, I can declare, at the best of my knowledge, that this work is

original and never submitted before in any other academic institution for higher

degree qualification purposes.

Mohammed Ahsan

________________________

28th April 2012

iii

Acknowledgements This work has been done with the invaluable help of many people. I wish to thank

them all.

First of all I wish to thank immensely to Dr. Tuquabo Tesfamichael for

providing me the opportunity to conduct my PhD and for his advice, support and

constructive feedback throughout my PhD. He has been a great support on all fronts

and made my PhD journey a memorable experience. Special thanks to my associate

supervisor Prof. John Bell for encouraging me to join QUT and his advice and

support throughout my PhD project. I wish to thank to my associate supervisor Prof.

Prasad Yarlagadda, for his positive feedback and advices.

I deeply acknowledge the funding provided by Prof. Nunzio Motta through

NIRAP Project “Solar powered Nano Sensors” for purchases and visits during the

project work.

I wish to thank Prof. Barry Wood from UQ, Prof. Wojtek Wlodarski from

RMIT, Dr. Mihail Ionescu from ANSTO and Nina De Caritat from ANU for their

immense help and support during experimental work at these facilities. I extend my

thanks to Peter Hynes, Cristina Theodoropoulos, Lambert Bekessy, Thor Bostrom

and Tony Raftery from AEMF, QUT for their continuous support and guidance

during my PhD. I would also like to extend my thanks to the technical staff of Built

Environment and Engineering Faculty for their support. I acknowledge all the lovely

people from Research Portfolio Office, for their support and guidance during my

PhD. I extend my sincere thanks to all my friends and colleagues in O401 for their

support and encouragement during my entire stay.

iv

Words are not sufficient to thank the most special person in my life, my wife,

Tehmeena whose sacrifice, love, encouragement and immense support made this

PhD an easy journey for me. I owe special thanks and immense love to my beautiful

children Ayman, Akmal and Hamnah who have been away from their dear father for

more than 2 years. This thesis could not have been accomplished with the immense

support of my father Amrullah Sharief, my mother Rabia Begum, my brother Akbar

and my sister Najma. I take this opportunity to thank all my relatives including my

father-in-law Prof. Sofi Ali, brother-in-laws Mansoor, Masood, Moudood and Late

Maqsood for their encouragement and support.

Mohammed Ahsan

v

Abstract

In this thesis, the author proposed and developed gas sensors made of

nanostructured WO3 thin film by a thermal evaporation technique. This technique

gives control over film thickness, grain size and purity. The device fabrication,

nanostructured material synthesis, characterization and gas sensing performance have

been undertaken. Three different types of nanostructured thin films, namely, pure

WO3 thin films, iron-doped WO3 thin films by co-evaporation and Fe-implanted

WO3 thin films have been synthesized. All the thin films have a film thickness of 300

nm. The physical, chemical and electronic properties of these films have been

optimized by annealing heat treatment at 300ºC and 400ºC for 2 hours in air.

Various analytical techniques were employed to characterize these films. Atomic

Force Microscopy and Transmission Electron Microscopy revealed a very small

grain size of the order 5-10 nm in as-deposited WO3 films, and annealing at 300ºC or

400ºC did not result in any significant change in grain size. X-ray diffraction (XRD)

analysis revealed a highly amorphous structure of as-deposited films. Annealing at

300ºC for 2 hours in air did not improve crystallinity in these films. However,

annealing at 400ºC for 2 hours in air significantly improved the crystallinity in pure

and iron-doped WO3 thin films, whereas it only slightly improved the crystallinity of

iron-implanted WO3 thin film as a result of implantation. Rutherford backscattered

spectroscopy revealed an iron content of 0.5 at.% and 5.5 at.% in iron-doped and

iron-implanted WO3 thin films, respectively. The RBS results have been confirmed

using energy dispersive x-ray spectroscopy (EDX) during analysis of the films using

transmission electron microscopy (TEM). X-ray photoelectron spectroscopy (XPS)

revealed significant lowering of W 4f7/2 binding energy in all films annealed at 400ºC

vi

as compared with the as-deposited and 300ºC annealed films. Lowering of W 4f7/2 is

due to increase in number of oxygen vacancies in the films and is considered highly

beneficial for gas sensing. Raman analysis revealed that 400ºC annealed films except

the iron-implanted film are highly crystalline with significant number of O-W-O

bonds, which was consistent with the XRD results. Additionally, XRD, XPS and

Raman analyses showed no evidence of secondary peaks corresponding to

compounds of iron due to iron doping or implantation. This provided an

understanding that iron was incorporated in the host WO3 matrix rather than as a

separate dispersed compound or as catalyst on the surface.

WO3 thin film based gas sensors are known to operate efficiently in the

temperature range 200ºC-500 ºC. In the present study, by optimizing the physical,

chemical and electronic properties through heat treatment and doping, an optimum

response to H2, ethanol and CO has been achieved at a low operating temperature of

150ºC. Pure WO3 thin film annealed at 400ºC showed the highest sensitivity towards

H2 at 150ºC due to its very small grain size and porosity, coupled with high number

of oxygen vacancies, whereas Fe-doped WO3 film annealed at 400ºC showed the

highest sensitivity to ethanol at an operating temperature of 150ºC due to its

crystallinity, increased number of oxygen vacancies and higher degree of crystal

distortions attributed to Fe addition. Pure WO3 films are known to be insensitive to

CO, but iron-doped WO3 thin film annealed at 300ºC and 400ºC showed an optimum

response to CO at an operating temperature of 150ºC. This result is attributed to

lattice distortions produced in WO3 host matrix as a result of iron incorporation as

substitutional impurity. However, iron-implanted WO3 thin films did not show any

promising response towards the tested gases as the film structure has been damaged

vii

due to implantation, and annealing at 300ºC or 400ºC was not sufficient to induce

crystallinity in these films.

This study has demonstrated enhanced sensing properties of WO3 thin film

sensors towards CO at lower operating temperature, which was achieved by

optimizing the physical, chemical and electronic properties of the WO3 film through

Fe doping and annealing. This study can be further extended to systematically

investigate the effects of different Fe concentrations (0.5 at.% to 10 at.%) on the

sensing performance of WO3 thin film gas sensors towards CO.

viii

Table of Contents

Abstract ........................................................................................................................ v

Table of Contents ...................................................................................................... viii

List of Figures ............................................................................................................. xi

List of Tables .............................................................................................................. xv

List of Publication – Research Activities ................................................................. xvii

CHAPTER 1 : INTRODUCTION ........................................................................ 19

1.1 Motivation .................................................................................................... 19

1.2 Conductometric gas sensors ......................................................................... 20

1.3 Parameters expressing sensing properties .................................................... 21

1.4 Aims and objectives of the project ............................................................... 23

1.4 Thesis Organization ..................................................................................... 25

CHAPTER 2 : LITERATURE REVIEW ........................................................... 27

2.1 Introduction .................................................................................................. 27

2.2 Construction principles of conductometric metal oxide gas sensors ........... 27

2.2.1 Receptor function ................................................................................. 27

2.2.2 Transduction function .......................................................................... 28

2.3 Limitations of existing metal oxide gas sensors .......................................... 29

2.4 Basic characteristics of a metal oxide conductometric sensors ................... 31

2.5 Basic mechanisms of gas sensing in semiconductor metal oxide sensors ... 34

2.6 Role of additives in gas sensing ................................................................... 39

2.7 Importance of the dimension in gas sensing ................................................ 40

2.8 Structural properties of tungsten oxide ........................................................ 43

2.9 Metal oxide based gas sensors ..................................................................... 46

2.10 Tungsten oxide (WO3) based gas sensors ................................................... 47

2.11 Deposition techniques of nanostructured metal oxide films ....................... 53

2.11.1 Thin Film Processes ............................................................................. 54

ix

2.11.1.1 Solgel ................................................................................................ 54

2.11.1.2 Spray pyrolysis ................................................................................. 55

2.11.1.3 Chemical vapour deposition (CVD) ................................................. 55

2.11.1.4 Physical vapour deposition (PVD) ................................................... 55

2.11.2 Thermal evaporation ................................................................................... 56

2.12 Summary ..................................................................................................... 59

CHAPTER 3 : EXPERIMENTAL METHODS .................................................. 60

3.1 Introduction .................................................................................................. 60

3.2 Deposition of nanostructured WO3 thin films by thermal evaporation ....... 60

3.2.1 Material ................................................................................................ 60

3.2.2 Substrate ............................................................................................... 61

3.2.3 Film deposition .................................................................................... 62

3.2.4 Ion implantation ................................................................................... 62

3.2.5 Post deposition heat treatment of the films .......................................... 63

3.3 Characterization of nanostructured WO3 thin films ..................................... 63

3.3.1 Transmission Electron Microscopy (TEM) ......................................... 64

3.3.2 Atomic Force Microscopy (AFM) ....................................................... 64

3.3.3 Rutherford Backscattered Spectroscopy (RBS) ................................... 65

3.3.4 X-Ray Photoelectron Spectroscopy (XPS) .......................................... 65

3.3.5 X-Ray Diffraction (XRD) .................................................................... 66

3.3.6 Raman spectroscopy ............................................................................ 68

3.4 Gas sensing characterization ........................................................................ 68

3.5 Summary ...................................................................................................... 73

CHAPTER 4 : THIN FILM CHARACTERIZATION ....................................... 74

4.1 Introduction .................................................................................................. 74

4.2 Characterization of Structural Properties .................................................... 75

4.2.1 AFM and TEM analysis ....................................................................... 75

x

4.2.2 GIXRD analysis ................................................................................... 80

4.2.3 Summary and outlook .......................................................................... 84

4.3 Characterization of chemical and electrical properties ............................... 86

4.3.1 Raman spectroscopy ............................................................................. 86

4.3.2 XPS analysis ......................................................................................... 89

4.3.3 Summary and outlook .......................................................................... 99

4.4 Characterization of compositional properties ........................................... 100

4.4.1 Rutherford backscattered spectroscopy .............................................. 100

4.4.2 Energy dispersive x-ray spectroscopy (EDX) .................................... 106

4.4.3 Summary and outlook ........................................................................ 107

CHAPTER 5 : GAS SENSOR CHARACTERIZATION ................................. 108

5.1 Introduction ................................................................................................ 108

5.2 Response to Hydrogen ............................................................................... 108

5.3 Response to Ethanol (C2H5OH) ................................................................. 120

5.4 Response to Carbon monoxide (CO) ......................................................... 129

5.5 Summary and outlook ................................................................................ 133

CHAPTER 6: CONCLUSIONS AND FUTURE WORK ....................................... 137

6.1 Conclusions ................................................................................................ 138

6.2 Recommendations for Future Work ........................................................... 140

BIBLIOGRAPHY .................................................................................................... 141

xi

List of Figures

Figure 2-1: A typical response curve of a conductometric gas sensor. ...................... 32

Figure 2-2: Schematic diagram showing the surface layer and corresponding electron

band structure, adapted from [48]. ............................................................................. 37

Figure 2-3: Schematic view of physical and electron band structure model for a

polycrystalline material [50]. ..................................................................................... 38

Figure 2-4: Schematic illustration of intergrain constriction situations: (a) Open neck

(b) Closed neck and (c) Conduction limited by point of contact. .............................. 39

Figure 2-5: Schematic of additive role in gas sensing [40]. ....................................... 40

Figure 2-6: The effect of particle size on gas sensitivity of SnO2 sensor exposed to

CO and H2, adapted from [56]. .................................................................................. 41

Figure 2-7: Influence of grain size on gas response of In2O3 film to NO2, adapted

from [57]. ................................................................................................................... 42

Figure 2-8: Tungsten oxide (WO3) crystal structure. ................................................. 44

Figure 2-9: A schematic of the classification of processes used for deposition of

nanostructured films. .................................................................................................. 54

Figure 2-10: A typical sketch of thermal evaporation chamber. ................................ 57

Figure 2-11: Thornton Zone diagram [186]. .............................................................. 59

Figure 3-1: Interdigitated Pt electrodes on Si substrate. ............................................ 62

Figure 3-2: Schematic of X-ray diffraction. ............................................................... 67

Figure 3-3: Schematic diagram of gas chamber setup. .............................................. 71

Figure 3-4: Schematic of sensor testing by flow through method. ............................ 72

Figure 4-1: AFM semicontact mode image (a) and TEM image (b) of as-deposited

nanostructured WO3 film. .......................................................................................... 76

Figure 4-2: AFM semicontact mode image (a) and TEM image (b) of nanostructured

WO3 film annealed at 300ºC for 2 hours in air. ......................................................... 76

Figure 4-3: TEM image of nanostructured WO3 film annealed at 400ºC for 2 hours in

air. .............................................................................................................................. 76

Figure 4-4: AFM semicontact mode image (a) and TEM image (b) of as-deposited

nanostructured Fe-doped WO3 film. .......................................................................... 77


Fe-doped WO3 film annealed at 300ºC for 2 hours in air. ......................................... 78

xii

Figure 4-6: TEM image of nanostructured Fe-doped WO3 film annealed at 400ºC for

2 hours in air. .............................................................................................................. 78


Fe-implanted WO3 film. ............................................................................................. 79


Fe-implanted WO3 film annealed at 300ºC for 2 hours in air. ................................... 80

Figure 4-9: TEM image of nanostructured Fe-implanted WO3 film annealed at 400ºC

for 2 hours in air. ........................................................................................................ 80

Figure 4-10: GIXRD pattern of nanostructured WO3 film, (a) as-deposited, (b)

annealed at 300ºC in air for 2 hours and (c) annealed at 400ºC in air for 2 hours. .... 81

Figure 4-11: GIXRD pattern of nanostructured Fe-doped WO3 film, (a) as-deposited,

(b) annealed at 300ºC for 2 hours in air and (c) annealed at 400ºC for 2 hours in air.

.................................................................................................................................... 83

Figure 4-12: Comparison of GIXRD patterns of nanostructured WO3 and Fe-doped

WO3 films annealed at 400ºC for 2 hours in air. ........................................................ 83

Figure 4-13: GIXRD pattern of nanostructured Fe-implanted WO3 film, (a) as-

deposited, (b) annealed at 300ºC for 2 hours in air and (c) annealed at 400ºC for 2

hours in air. ................................................................................................................. 85

Figure 4-14: Raman spectra of nanostructured WO3 films. ....................................... 87

Figure 4-15: Raman spectra of nanostructured Fe-doped WO3 films. ....................... 87

Figure 4-16: Raman spectra of nanostructured Fe-implanted WO3 films. ................. 89

Figure 4-17: XPS wide spectra of nanostructured WO3 films, (a) as-deposited, (b)

annealed at 300ºC for 2 hours in air and (c) annealed at 400ºC for 2 hours in air. .... 90

Figure 4-18: XPS wide spectra of nanostructured Fe-doped WO3 films, (a) as-


hours in air. ................................................................................................................. 91

Figure 4-19: XPS wide spectra of nanostructured Fe-implanted WO3 films, (a) as-


hours in air. ................................................................................................................. 92

Figure 4-20: W 4f core level high resolution spectra of nanostructured WO3 films. 93

Figure 4-21: W 4f core level high resolution spectra of nanostructured Fe-doped

WO3 films. .................................................................................................................. 94

xiii

Figure 4-22: W 4f core level high resolution spectra of nanostructured Fe-implanted

WO3 films. ................................................................................................................. 95

Figure 4-23: O 1s core level high resolution spectra of nanostructured WO3 films. . 96

Figure 4-24: O 1s core level high resolution spectra of nanostructured Fe-doped WO3

films. .......................................................................................................................... 97

Figure 4-25: O 1s core level high resolution spectra of nanostructured Fe-implanted

WO3 films. ................................................................................................................. 99

Figure 4-26: RBS spectrum of (a) as-deposited and (b) 300ºC annealed

nanostructured WO3 films. ....................................................................................... 101

Figure 4-27: RBS depth profile of (a) as-deposited and (b) 300ºC annealed

nanostructured WO3 films. ....................................................................................... 102

Figure 4-28: RBS spectrum of (a) as-deposited and (b) 300ºC annealed

nanostructured Fe-doped WO3 film. ........................................................................ 103

Figure 4-29: RBS spectrum (a) and depth profile (b) of nanostructured Fe-doped film

annealed at 300ºC for 2 hours in air. ........................................................................ 104

Figure 4-30: RBS spectrum (a) and corresponding depth profile (b) of nanostructured

Fe-implanted WO3 film. ........................................................................................... 105

Figure 4-31: EDX spectrum of for 2 hours in air. ................................................... 106

Figure 4-32: EDX spectrum of Fe-implanted WO3 film annealed at 400ºC for 2 hours

in air. ........................................................................................................................ 107

Figure 5-1: Dynamic response (a) and sensitivity (b) to H2 measured for a

nanostructured WO3 film annealed at 300ºC for 2 hours in air. .............................. 110


nanostructured WO3 film annealed at 400ºC for 2 hours in air. .............................. 111


nanostructured Fe-doped WO3 film annealed at 300ºC for 2 hours in air. .............. 114




nanostructured Fe-implanted film annealed at 300ºC for 2 hours in air. ................. 118


nanostructured Fe-implanted WO3 film annealed at 400ºC for 2 hours in air. ........ 119

xiv

Figure 5-7: Dynamic response (a) and sensitivity (b) to ethanol measured for a

nanostructured WO3 film annealed 400ºC for 2 hours in air. .................................. 122





Figure 5-10: Comparison of ethanol sensitivities of nanostructured WO3 and Fe-

doped WO3 films annealed at 400ºC for 2 hours in air at an operating temperature of

150ºC. ....................................................................................................................... 127


nanostructured Fe-implanted WO3 film annealed at 400ºC for 2 hours in air. ........ 128

Figure 5-12: Dynamic responses of nanostructured Fe-doped WO3 films to CO, (a)

annealed at 300ºC for 2 hours in air and (b) annealed at 400ºC for 2 hours in air. . 131

Figure 5-13: Sensitivity of annealed (at 300ºC and 400ºC) nanostructured Fe-doped

WO3 films upon exposure to CO. ............................................................................. 132

xv

List of Tables

Table 2-1: Some selective publications of metal oxides used for gas sensing. ......... 46

Table 2-2: Summary of published results on gas sensing characteristics of WO3. .... 48

Table 3-1: Physical properties of WO3 and Fe powders. ........................................... 61

Table 3-2: Target gases and their concentrations. ...................................................... 69

Table 4-1: Comparison of the positions of O-W-O stretching vibration mode peaks

observed for nanostructured WO3 and Fe-doped WO3 films annealed at 400ºC for 2

hours in air. ................................................................................................................ 88

Table 4-2: Comparison of W 4f7/2 and O 1s peak positions of nanostructured WO3

and Fe-doped WO3 films annealed at 400ºC for 2 hours in air. ................................. 98

Table 5-1: Optimum operating temperature of WO3 thin film based gas sensors upon

exposure to H2. ......................................................................................................... 109

Table 5-2: Response and Recovery times of annealed (at 300ºC and 400ºC)

nanostructured WO3 films upon exposure to H2. ..................................................... 111


nanostructured Fe-doped WO3 films upon exposure to H2. ..................................... 115


nanostructured Fe-implanted WO3 films upon exposure to H2. ............................... 119

Table 5-5: Optimum operating temperature of nanostructured WO3 thin film based

gas sensors upon exposure to ethanol. ..................................................................... 121

Table 5-6: Response and Recovery times of 400ºC annealed nanostructured WO3

films upon exposure to ethanol. ............................................................................... 122


nanostructured Fe-doped WO3 films upon exposure to ethanol. ............................. 126

Table 5-8: Response and Recovery times of 300ºC annealed nanostructured Fe-

implanted WO3 film upon exposure to ethanol. ....................................................... 129

Table 5-9: Optimum operating temperature of nanostructured WO3 thin film based

gas sensors upon exposure to CO. ........................................................................... 130


nanostructured Fe-doped WO3 films upon exposure to CO. ................................... 132

Table 5-11: Optimum operating temperature of the films upon exposure to H2,

ethanol and CO in the tested temperature range of 100ºC-300ºC. ........................... 133

xvi

Table 5-12: Comparison of microstructural properties of 400ºC annealed films. ... 135

xvii

List of Publication – Research Activities

M. Ahsan, T. Tesfamichael, J. Bell, M. Ionescu and M.G. Blackford,

“Microstructural characterization of electron beam evaporated tungsten

oxide films for gas sensing applications”, 16th AINSE Conference on

Nuclear and Complementary Techniques of Analysis - 25 - 27 November

2009 at Lucas Heights, Sydney.

M. Ahsan, T. Tesfamichael, A. Ponzoni, and G. Faglia, “Sensing

Properties of E-Beam Evaporated Nanostructured Pure and Iron-Doped

Tungsten Oxide Thin Films”, Sensor Letters, 2011, Volume 9, pp.759-

762.

M. Ahsan, T. Tesfamichael, J. Bell, N. Motta, W. Wlodarski, “Iron doped

nanostructured WO3 thin film based conductometric hydrogen sensor”,

EMRS 2011 Fall Meeting, September 19-23, 2011, Warsaw University of

Technology, Poland.

M. Ahsan, T. Tesfamichael, M. Ionescu and J. Bell, “Annealing effect on

iron doped tungsten oxide thin films prepared by thermal evaporation”,

Materials Innovation in Surface Engineering 2011, 18-20 October 2011,

Melbourne, Australia.

M. Ahsan, T. Tesfamichael, M. Ionescu, J. Bell and N. Motta, “CO

sensitive nanostructured WO3 thin films doped with Fe”, NanoS-E3 -

Nanostructures for Sensors, Electronics, Energy and Environment, 12-16

September 2011, Kingscliff, Australia.

xviii

M. Ahsan, T. Tesfamichael, M. Ionescu, J. Bell and N. Motta, “Low

temperature CO sensitive nanostructured WO3 thin films doped with Fe”

Sensors and Actuators B, Volume 162, Issue 1, 2012, pp 14-21.

M. Ahsan, T. Tesfamichael, J. Bell and N. Motta, “Affect of annealing on

the sensing properties of nanostructured tungsten oxide (WO3) thin films

towards hydrogen and ethanol”, Sensors and Actuators B (Submitted).

T. Tesfamichael, A. Ponzoni, M. Ahsan, G. Faglia, “Gas Sensing

Characteristics of Fe-Doped Tungsten Oxide Thin Films”, Sensors and

Actuators B (Accepted).

19

CHAPTER 1 : INTRODUCTION

1.1 Motivation

Human living standards have grown remarkably due to the industrial revolution.

Industrialization also has a negative aspect on human health due to emission of gases

that pollute environment and pose risk to public health. Flammable gases also need

to be monitored to protect against unwanted incidence of explosion or fire.

Applications that ensure human health and safety, protect the environment, monitor

manufacturing processes and optimize the performance of control systems have

increased the demand for gas sensors in recent years [1-3]. The requirements of

these applications demand immediate, near continuous analysis to detect target gases.

Online analysis of gas mixtures is fundamental to quality control in industrial

production and assuring safety in the work place. In situ sensing of various gases is

important in many technological fields, such as the oil and petrochemical industry,

water treatment plants and biogas applications [4]. Modern alcohol breath analysers

based on ethanol sensors are replacing traditional detecting methods for the

identification of drunken drivers. Developments leading to electronic noses which

mimic the human olfactory system by using an array of electronic chemical sensors

and appropriate pattern-recognition electronics will further expand the requirements

for gas sensors [5, 6]. Poor indoor air quality due to emission of toxic gases, volatile

organic compounds and microbial contaminants can reduce the health and comfort of

occupants of a building. Hence monitoring of toxic and hazardous gases is a subject

of growing importance in both domestic and industrial built environments. To be

useful in providing timely feedback that enables effective control of these gasses,

20

sensors that are capable of continuously monitoring the levels and types of gases are

needed. The threshold limit values (TLV) for acceptable levels for some toxic gasses

can be as low as 20 ppb (1 ppm = 1 ml/m3 = 2.66 x 10-3 mg/m3) which requires high

performance sensing devices [7]. Together, this suite of applications indicates a

growing demand for the development of reliable and low power gas sensors with

excellent sensitivity and selectivity towards specific gases.

1.2 Conductometric gas sensors

Gas sensors operate on the principle of conversion of gas concentration into a

measurable signal. Gas sensor devices that have been developed so far include mass

sensitive sensors, optical sensors, electrolytic sensors and solid state sensors [8].

Among the solid-state gas sensors, semiconductor metal oxide gas sensors have

received the most attention as they show good potential for continuous monitoring of

gases. These sensors offer a wide variety of advantages over the traditional analytical

instruments which include lower cost, easier manufacturing, smaller size, short

response and faster recovery.

Conductometric metal oxide sensors are based on the principle of a change in

electrical resistance due to interaction between the surface and the target gas. The

change in sensor’s resistance which is linked to surface reactions depends on many

factors such as specific material reactivity, microstructure, free charge concentration,

sensitive layer morphology, etc. Most of the work on “conduction type” in these

sensors was performed on n-type semiconducting metal oxides such as SnO2, ZnO

and WO3. Unfortunately, the available conduction models [9] are not suitable to p-

type materials. Tin oxide (SnO2) is one of the most widely studied sensing material

since Taguchi built the first SnO2 sensor for Figaro Sensors in 1970 [10]. However,

recent research has shown other metal oxides such as TiO2, ZnO and WO3 also have

21

potential for high sensing properties. This is evident from the increasing number of

publications addressing the sensing properties of these metal oxides.

1.3 Parameters expressing sensing properties

It is important at this point to briefly explain sensing parameters in order to

develop a clear understanding of the effect of various factors on these parameters.

The response of a gas sensor is characterized by the following parameters [8]:

1. Sensitivity: It is defined as the ratio of ‘resistance change’ of a sensor upon

exposure to target gas to resistance in target gas for n-type materials and

reducing gas. For p-type materials and reducing gas, it is the ratio of

resistance change of sensor in target gas to that in air. Many authors use the

term ‘sensitivity’ to indicate response amplitude. However, in a true sense,

sensitivity of a gas sensor is defined as the derivative of the response to the

gas concentration [11]. Therefore, keeping in view the general trend of using

the definitions, the term ‘sensitivity’ would be used to indicate the response

amplitude of the sensors in this study.

2. Selectivity: It is expressed in terms of a dimension that compares the

concentration of the corresponding interfering gas that produces the same

sensor signal. It is expressed as ratio of sensitivity towards interfering gas to

sensitivity towards desired gas.

3. Response time: It is the time interval over which resistance of the sensor

material attains a fixed percentage (usually 90%) of final value when the

sensor is exposed to full scale concentration of the gas. A small value of

response time is highly desirable in applications such as detection of

flammable or combustible gases to prevent fire.

22

4. Recovery time: It is the time interval over which sensor resistance reduces to

10% of the saturation value when the target gas is switched off and the

sensor is placed in synthetic (or reference) air. A sensor should have a small

recovery time so that it can be ready for next detection.

5. Long term stability: It is the ability of a sensor to maintain its sensing

properties when operated continuously during its lifetime. Sensors are

designed to have long term stability that lasts up to several years without

showing a drift in sensor performance.

6. Short term stability: It is the ability of a sensor to reproduce its

characteristics during a certain period of time at working conditions, which

may include high temperature and presence of known target gas [12].

The operating temperature also plays an important role in sensing performance of

metal oxides. Most of the metal oxide sensors are operative only at high temperatures

in the range 200ºC-500ºC, as they are thermally activated in this temperature range.

At low temperatures, metal oxides do not show any response to target gases. The

overall gas sensing phenomenon is a combination of the following three functions

[13]:

Receptor function: It depends on the ability of the oxide surface to interact with

the target gas. It is largely affected by the surface properties of the oxide. Dopants

and defects such as interstitial cations or anion vacancies can play in important role

in enhancing the conductivity and hence sensitivity towards target gas [14, 15].

Transducer function: It depends on the ability to transport electrons through grain

boundaries. The electronic band structure and hence conductivity also depend on

particle size and porosity [16-20].

23

Utility function: It concerns the kinetic factor as a result of difference between

the diffusion rate and reaction rate. Film porosity plays an important role in order for

the target gas to penetrate the entire volume of the metal oxide layer and balance the

reaction rate with diffusion.

1.4 Aims and objectives of the project

The aim of the current research was to develop tungsten oxide (WO3)

conductometric gas sensors and optimize their sensing properties such as sensitivity,

selectivity, response and recovery times, and operating temperatures towards a few

selected gases (H2, ethanol and CO) by depositing suitable nanostructured thin films

using thermal evaporation technique, post-deposition heat treatment and

characterization of physical, chemical and electronic properties of the films. To

achieve this goal, gaps in the literature were identified. Based on the literature review

(Chapter 2), the following conclusions have been derived from available studies:

1. As with any semiconducting metal oxide, WO3 based gas sensors show good

performance only at elevated temperatures (>200ºC).

2. Sensors based on WO3 thin films have been used to detect a number of

oxidizing and reducing gases such as NH3, H2, NO2, H2S and ethanol.

However, there is very little evidence in the literature on their sensing

performance towards CO.

3. As in case of any semiconductor metal oxide, selectivity to a specific gas is a

major problem in sensors based on tungsten oxide (WO3).

4. Doping the WO3 thin films with noble metals such as Au, Ag and Pd has

shown improved sensing performance towards various gases which is mainly

attributed to the noble metal catalytic effect on the surface/gas interaction. In

the present study, iron has been used to dope the WO3 thin films. Since iron

24

has a similar atomic radius as W, therefore, it can be introduced as a

substitutional impurity in the WO3 matrix. The influence of Fe doping on

physical, chemical, electronic and gas sensing properties of WO3 thin films

has not been studied extensively.

In an attempt to address the above problems, the following objectives were set:

1. To synthesise nanostructured WO3 thin films by thermal evaporation method.

This method gives control over film thickness, grain size and porosity. These

process parameters are critical in gas sensing.

2. To modify the physical, chemical and electronic properties of the WO3 films

through doping with Iron (Fe) during thermal evaporation. As Fe has a

similar atomic radius as W, it can be introduced as a substitutional impurity in

the WO3 matrix and its influence on physical, chemical, electronic and gas

sensing properties can be investigated. Iron would also be incorporated by

implantation to investigate the effect of different doping techniques.

3. To optimise the film properties such as grain size, crystallinity, stoichiometry

and porosity by post deposition heat treatment. Post deposition heat treatment

also relieves the residual stresses and surface contamination.

4. To characterize the physical, chemical and electronic properties of these films

using various scientific instruments such as: Scanning Electron Microscopy,

Atomic Force Microscopy, Transmission Electron Microscopy, X-ray

Diffraction Analysis, X-ray Photoelectron Spectroscopy, Rutherford

Backscattered Spectroscopy and Raman Spectroscopy. This is valuable in

assessing the microstructural and electronic properties of the films.

25

5. To characterize the sensing properties (sensitivity, selectivity, response and

recovery times) of these films to H2, ethanol and CO in the temperature range

100ºC to 300ºC.

6. Investigate the effect of microstructure, doping and post deposition heat

treatment on the sensing performance and operating temperature of these

films.

1.4 Thesis Organization

This thesis consists of six chapters and is presented as follows:

Chapter 1 gives an overview of author’s motivation for performing this

research and specific objectives.

Chapter 2 presents the literature review on various aspects of gas sensing. It

includes the construction principle and basic characteristics of gas sensors,

gas sensing mechanisms, role of additives, importance of dimensions in gas

sensing and limitations of current metal oxide based gas sensors. The

structural properties of tungsten oxide and brief overview of various

deposition techniques with a particular focus on thermal evaporation are also

described in this chapter.

Chapter 3 outlines the processes and procedures involved in the synthesis and

characterization of tungsten oxide thin films investigated in this research.

This chapter is broadly divided into three sections which describe the thin

film deposition, thin film characterization and gas sensor characterization. In

the first section of this chapter, thermal evaporation of nanostructured

tungsten oxide thin films is described. It is followed by the description of

various analytical techniques such as Transmission Electron Microscopy,

26

Atomic Force Microscopy, Rutherford Backscattered Spectroscopy, X-ray

Diffraction, X-ray Photoelectron Spectroscopy and Raman spectroscopy

which were employed to characterize these films. The last section describes

the details of the conductometric gas sensing setup used in this study.

Chapter 4 focuses on the physical, chemical and electronic characterization of

nanostructured thin films. The characterization outcomes using various

analytical techniques are presented and the effect of heat treatment and

doping are discussed.

Chapter 5 presents the experimental results obtained from the gas sensor

characterization of nanostructured tungsten oxide thin film based gas sensors.

The results are linked to the various physical, chemical and electronic

properties of these films investigated in Chapter 4 and the influence of

various factors such as surface morphology, grain size, crystallinity and

stoichiometry on gas sensing performance are discussed.

Chapter 6 presents the conclusions of this thesis and suggestions for possible

future work.

27

CHAPTER 2 : LITERATURE REVIEW

2.1 Introduction

In Chapter 1, conductometric gas sensors were described and the parameters

expressing the sensor performance have been discussed. In this chapter, the

fundamental characteristics of conductometric metal oxide gas sensors are discussed.

Then, a critical review of the state of the technology of metal oxide gas sensors is

presented with a focus on tungsten oxide.

2.2 Construction principles of conductometric metal oxide gas

sensors

In principle, conductometric metal oxide gas sensors are constructed by two key

functions: Receptor function and Transduction function.

2.2.1 Receptor function

Receptor function transforms chemical information into a form of energy which

can be measured by the transducer [21]. It is determined through various interactions

between the surface and target gas such as adsorption, ion exchange or

electrochemical reaction. In air, oxygen is adsorbed on the oxide grains as negatively

charged ions, inducing a surface space charge layer depleted of electrons (space

charge layer) which leads to a band bending [22]. The band bending and thickness of

space charge layer created due to oxygen species are further decreased when the

metal oxide is exposed to a reducing gas as reducing gases inject electrons to the

conduction band. For oxidizing target gasses such as NO2 the conductivity decreases

as oxidising gases extract electrons. Variation of the height of the potential barrier is

28

believed to be the origin of the conductance response to gases [23]. The receptor

function of metal oxides is largely affected by their intrinsic electronic properties and

any deviation from their stoichiometric chemical properties. Defects such as oxygen

vacancies are inherent in metal oxides. This creates the space charge layer depleted

of electrons and negatively charged oxygen ions on the surface. The mobility of main

carriers is of primary importance for semiconductors, because it provides the

proportionality constant of the change of electrical conductivity when the number of

main carriers changes as a result of gas–solid interactions [24]. As the metal oxides

approach stoichiometry, the conductivity becomes extremely low (high resistance).

WO3 and TiO2 films have very small electron mobility (high resistivity) which

ranges between 0.03 and 0.2 cm2 V−1 s−1 [25, 26]. Dopants and defects such as

interstitial cations or anion vacancies can play an important role in enhancing the

conductivity [14, 15]. Dopants are important to the increased formation of oxygen

vacancies and modifying of the electronic structure and band gap energy of metal

oxides. It has been reported that doping of TiO2 with Fe increases the oxidation

activity of the oxide and this has been related to a higher density of oxygen vacancies

[27].

2.2.2 Transduction function

Transduction function is related to the ability to transport electrons through grain

boundaries. In polycrystalline metal oxide, gas sensing reaction takes place at the

surface of the individual particles and at grain boundaries and it becomes easy for

electrons to conduct through different grains. Experiments on various metal oxide

films exposed to specific gases suggest that the electronic band structure and hence

conductivity are also dependent on film microstructure (i.e. particle size and film

porosity) [16-19, 28]. This shows that the magnitude of the conductivity change

29

depends mainly on the ratio between particle size and Debye length (distance over

which charge separation occurs in a semiconductor). If the grain size is large (>>

Debye length) the depletion of the space charge region between the grain boundaries

controls the conductivity variation. If the radius of the grain is extremely small (less

than the Debye length, e.g. for WO3, is 10 nm), the entire particle is depleted and no

band bending occurs which results in high conductivity-change (or high sensitivity)

when exposed to target gas. In practice nanomaterials of particle size less than 10 nm

favours large gas-active surface area and provide high sensor sensitivity.

2.3 Limitations of existing metal oxide gas sensors

The functions discussed above in Section 2.2 strongly influence the sensitivity of

semiconducting metal oxides. Although most of the semiconductor metal oxide

sensors that have been investigated to date are promising, they do not show response

to gases at lower operating temperatures (100ºC-200ºC) and must be thermally

activated at higher temperatures (200ºC-500ºC). These high operating temperatures

cause grain growth and changes in material properties that lead to long term stability

problems of the sensors. The higher optimum operating temperatures [29, 30] also

demand higher power consumption, which makes them unsuitable for battery

operated sensor devices that would be advantageous in some in situ applications.

Lower operating temperature metal oxide gas sensors with an acceptable sensitivity

would overcome these stability and high power consumption problems. By analogy

with the current high temperature metal oxide gas sensors, a method for enhancing

sensitivity to gas at lower temperatures is modification of the electronic structure of

the metal oxides by using mixed metal oxides such as SnO2-ZnO, Fe2O3-ZnO and

ZnO-CuO [28, 31-33]. Composites of SnO2-ZnO and SnO2-In2O3 have shown

enhanced sensitivity when compared with single oxide sensors when exposed to

30

ethanol [34]. Sensitivity is improved by selective catalytic activity of one component

of the mixed oxide to a particular gas. However, this approach has not shown any

significant reduction in optimum working temperature of the sensors [33].

The catalytic activity of the gas sensor material can also be enhanced by doping

with noble metals such as Pt, Au, Pd and Ag [35-39]. The noble metals chemically

sensitize the metal oxide surface i.e., they activate target gases by enhancing their

spill-over (more surface coverage), so that they react with oxygen adsorbates more

easily. The oxygen supply can also be improved by metal additives, at the surface of

which oxygen molecules from the ambient can be dissociated and migrate to the

surface of metal oxide. In this way, the additive enhances the sensing properties of

the metal oxide.

Metal oxides can also be electronically sensitized to improve the gas response

and lower the operating temperature [40]. Addition of fine particles of some metals

results in a rise of the base resistance in air. The electron concentration in the oxide

surface layer is low, which corresponds to an increase in the space-charge depth as a

result of the transfer of electron from the metal oxide to the metal loaded onto its

surface. When the metal surface is covered with oxygen adsorbates at elevated

temperatures in air, its oxidation state changes (the metal is oxidized).The oxygen

adsorbates extract electrons from this metal, which in turn extracts electrons from the

metal oxide, leading to a further increase in the space charge depth. Consumption of

oxygen adsorbates on the metal, in addition to those on the metal oxide surface, by

reaction with gas, enhances the sensitivity.

Another limitation of current metal oxide sensors is that they are non-selective

i.e. they are sensitive simultaneously to a wide range of gases. Selectivity can be

improved by exploiting the influence of operating temperature on the sensitivity of

31

the sensor [41, 42]. A change in sensor resistance is expected with change in

temperature, as the reactions occurring at the surface of the sensor

(chemisorption/redox reaction) are functions of temperature [9]. The temperature

dependence of the sensitivity of different gases can therefore be exploited to obtain

the selectivity to a particular gas.The selectivity can also be improved by depositing

a diffusion filter layer, such as SiO2 on top of the metal oxide [43]. By doing so, only

small molecules such as H2 are able to reach the surface of sensing material.

However, this has the undesirable effect of reducing the total number of gas

molecules that reach the sensor, hence reducing sensitivity.

One of the recent developments in improving selectivity of metal oxide gas

sensors is the use of a neural network. The idea is inspired from the biological

olfactory system. In human and mammalian noses, there are thousands of receptors

with bad selectivity to different odours, but the brain is able to derive specific odour

identification by processing the signals from all the receptors. The technological

analogue is called ‘electronic nose’ and mimics the natural olfactory process [5, 6].

In this system, an array of sensors with different functionality is employed and their

data are processed by neural networks to determine the gas concentrations.

2.4 Basic characteristics of a metal oxide conductometric sensors

In general, the electrical resistance of a conductometric metal oxide gas sensor

changes upon exposure to the molecules of the target gas. The nature of sensor

material (n-type or p-type semiconductor metal oxide) and the target gas (oxidizing

or reducing) governs the increase or decrease in electrical resistance. For an n-type

semiconductor exposed to reducing gas, the resistance decreases, whereas, upon

exposure to oxidizing gas, the resistance increases. The variation of resistance of

32

sensor with time on exposure and withdrawal of target gas is depicted by a typical

response curve as shown in Fig. 2-1.

Figure 2-1: A typical response curve of a conductometric gas sensor.

The performance of a gas sensor is characterized by the following five parameters

[8]:

1. Sensitivity or Response Amplitude

2. Response time

3. Recovery time

4. Selectivity

5. Long term stability

A brief description of these parameters follows below:

1. Sensitivity(S):

It is defined as the ratio of resistance change of a sensor upon exposure to target

gas to the resistance in target gas for n-type materials [8]. For p-type materials, it is

the ratio of resistance change of sensor upon exposure to target gas to the resistance

in air.

Response time

Recovery time

Gas on

Gas off

Time

Res

ista

nce

33

)1.2()( materialstypenforR

RS

gas

)2.2()( materialstypepforR

RS

air

where ΔR is the change in sensor resistance upon exposure to target gas. Sensors of

high S value are desirable in order to sense low concentration of gases.

As pointed out in the Section 1.3, the term ‘sensitivity’ is often used to indicate

response amplitude. However, in a true sense, sensitivity of a gas sensor is defined as

the derivative of the response to the gas concentration [11]. For the purpose of

simplicity, the term ‘sensitivity’ would be used to indicate the response amplitude of

the sensors in this study.

2. Response time:

This is the time interval over which resistance of the sensor material attains a

fixed percentage (usually 90%) of final value when the sensor is exposed to full scale

concentration of the gas. It is usually expressed as T90, T80, etc. A T80 of 50s

means that the sensor exhibits 80% of saturation value of resistance in 50s.

3. Recovery time:

It is the time interval over which sensor resistance reduces to 10% of the

saturation value when the sensor is exposed to full scale concentration of the gas and

then placed in the clean air. A sensor should have a small recovery time so that it can

be used over and over again.

4. Selectivity:

Most of the chemiresistive sensors exhibit high value of sensitivity to many

gases under similar operating conditions. Thus, selectivity of a sensor towards target

gas is expressed in terms of dimension that compares the concentration of the

corresponding interfering gas that produces the same sensor signal. It is expressed as

34

(2.3) d gasthe desirey towards Sensitivit

ering gasfor interf the sensorof y SensitivitySelectivit

5. Long term stability:

The ability of a sensor to maintain its properties when operated continuously for

long durations is called its stability. Good sensors have long term stability that last up

to several years without showing a drift in sensor performance.

All the above sensing parameters depend on several factors including the following:

Sensing material i.e. intrinsic properties of the metal oxide.

Sensing mechanisms i.e. interaction between the gas and sensor surface

(details in Section 2.5).

Operating conditions i.e. temperature, type of target gas.

Film properties such as microstructural features, film type (thick or thin film),

stoichiometry, etc.

In order to control these parameters, scientific understanding of gas - sensor

interaction (sensing mechanism) needs to be addressed, which follows.

2.5 Basic mechanisms of gas sensing in semiconductor metal oxide

sensors

Semiconductor metal oxides are used for two different types of gas sensing

applications. Broadly speaking, these applications can be categorized as

Determination of partial pressure of oxygen.

Determination of concentration of a minor constituent (oxygen partial

pressure remains constant).

The materials that have been used commercially for determining partial pressure

of oxygen are TiO2, Cr2O3 and Ga2O3 [44]. If the sensor operates at a high

35

temperature (700ºC and above), the mechanism responsible for the detection is bulk

conduction. The oxygen partial pressure and electrical conductivity are related as

[44]

(2.4)1/m2

B

A OPTk

Eexpσσ

where, is the electrical conductivity,

* is a constant,

EA is activation energy for conduction,

KB is Boltzman’s constant,

P[O2] is oxygen partial pressure,

m is the oxygen vacancy constant dependent on dominant type of bulk defect

involved in the reaction between sensor and oxygen and

T is absolute temperature.

The second application of semiconductor metal oxide gas sensors involves

situations where the oxygen partial pressure is constant and concentration of minor

constituent gases such as H2, CO, CH4 and H2S are to be determined.

For an n-type metal oxide semiconductor, possible gas sensing mechanisms are

[45]:

Surface reactions with adsorbed gases.

Ion exchange.

Direct gas adsorption.

For semiconductor materials, the observed sensor effects are dominated by direct

gas adsorption and surface reactions with preadsorbed molecules. The sensor

characteristics will vary depending on whether the sensor material is n-type or p-type

and whether the interacting species are reducing or oxidizing gases.

36

For n-type sensor material and reducing gas, there will be preadsorbed oxygen

ions on the surface of the sensor. The gas will then react with oxygen ions to form

neutral molecules, leading to electron transfer to the sensor material and

consequently decrease the resistance. Following is the detailed description:

In air environment, oxygen molecules adsorb onto the surface of metal oxide

layer to form O2-, O- and O2- species by extracting electrons from the conduction

band depending on the temperature [46] and type of metal oxide (n-type or p-type).

These oxygen adsorbates play an important role in detecting gaseous species.

It has been experimentally confirmed by TPD, FTIR and ESR techniques that

dominant oxygen species are [47]:

Molecular (O2-) below 150ºC.

Atomic (O-) between 150ºC and 300ºC.

Atomic (O2-) above 300ºC.

The oxygen adsorption can be described by the following rate equations:

)6.2()()( 221 adsOgO k

)7.2()()( 222 adsOadsOe k

)8.2()(2)(2 32 adsOadsOe k

)9.2()()( 24 adsOadsOe k

The reaction of the target gas X to be detected can be represented by:

)10.2()()( 225 egXOadsOX k

)11.2()()( 6 egXOadsOX k

)12.2(2)()( 22 7 egXOadsOX k

In an n-type semiconductor metal oxide, electrons are transferred to the surface

and then ionize the oxygen adsorbates to form O2- and O- which results in a negative

37

charge being developed on the surface. The surface layer is therefore depleted of

electrons, and so called depletion layer is created [48].

Fig. 2-2 shows the schematic of the surface layer and the corresponding electron

band structure. The conduction band energy levels of the bulk and the surface are

represented by Ecb and Ecs, respectively. The valence band energy levels for the bulk

and the surface are represented by Evb and Evs, respectively, and Ef denotes the Fermi

energy. Et is trap energy level of electrons at surface states due to adsorbed oxygen,

eVs is the height in energy of the band bending at the surface, d is the distance from

the surface and dd is the depth of depletion. The difference between Ecb and Evb is the

band gap energy Eg.

Figure 2-2: Schematic diagram showing the surface layer and corresponding electron band structure,

adapted from [48].

If the surface belongs to an ideal bulk material (large d) without grain

boundaries, the influence of the depletion layer is of little or no importance for the

conduction along the surface. However, in case of polycrystalline film, each interface

between grains gives rise to a band bending as depicted in Fig. 2-3. At the

intergranular contact, the conduction is restricted by this Schottky potential barrier

Evs Evb

EF

Ecb

Ecs eVs

d

Surface

O2-

O-

Et

Depletion layer (for n-type

semiconductor)

dd

38

due to depletion layer and the electrons have to overcome the energy barrier, eVs.

The change of the barrier height makes the electrical resistance of the material

dependent on the gaseous atmosphere [49]. The resistance and hence, gas sensitivity

in this case are not dependant on particle size.

Figure 2-3: Schematic view of physical and electron band structure model for a polycrystalline

material [50].

Fig. 2-4 shows three situations with different influence on the depletion layer

[50]. Fig. 2-4a demonstrates the situation where the area of the depletion zone at the

contact is less than the contact area. The depletion layer extends into the grain to a

depth marked by the dashed line. The resistivity is about that of the undepleted

region in the center of the neck. Fig. 2-4b illustrates a closed neck, where, the

depletion layers from the surfaces overlap resulting in a higher resistance in the

center of the constriction. The above two situations imply that the porous structure

would respond to the gas in the same way as a thin film sensor. Fig. 2-4c illustrates a

Barrier

eVs

O-

O2-

O-

O- O2

-

O- O2-

O- O-

O2- O-

O- O2- O

- O- O2

- O-

O-

O-

O2-

O- O- O- O2

- O- O- O-

Electronic current

Conduction band electrons

Adsorbed oxygen

Depletion layer

Physical Model

Band Model

39

situation where conduction is limited by the point of contact. This type of situation is

applicable to a porous material.

Figure 2-4: Schematic illustration of intergrain constriction situations: (a) Open neck (b) Closed neck and (c) Conduction limited by point of contact.

Wang [51] investigated the transition from open neck, neck controlled to point of

contact limited conduction of metal oxide gas sensors. He showed that the sensitivity

to adsorbed gas increased rapidly as the grain size became smaller than 40 nm, at

which point the conduction mechanisms for neck controlled and point of contact

limited conduction exist together.

2.6 Role of additives in gas sensing

Appropriate amounts of metal additives such as Au, Pt, Pd, etc. when added to

metal oxide sensors have shown improvement for various kinds of gases.

Enhancement in sensor response and a decrease in operating temperature for

maximum sensor response have also been achieved, in addition to decrease in

response time and better selectivity. The main idea of using metal additives is to

enhance the reaction rate of the gases when they come in contact with the sensor

surface.

40

Shimizu et al [52] identified that metal additives can lead to two different

sensitization mechanisms: Electrical sensitization and Chemical sensitization. These

mechanisms are shown in Fig. 2-5.

Figure 2-5: Schematic of additive role in gas sensing [40].

In chemical sensitization, the metal cluster catalyzes the reaction and reaction

products are subsequently spill over the semiconducting metal oxide surface. These

reaction products then cause the gas sensing response [40].

Electronic sensitization occurs due to the alignment of Fermi energies of the

metal oxide and the additive. This is similar to the Schottky barrier influencing the

surface charge region in the semiconductor material. Oxidation or reduction of metal

additive controls the band bending of the metal oxide thereby controlling the sensing

mechanism [53].

2.7 Importance of the dimension in gas sensing

Nanomaterials are defined as those that have at least one of their dimensions

≤100nm. Thus we may visualize them as structures produced by reducing one, two,

or three dimensions of a bulk material, thereby resulting in 2D nanolayers, 1D

nanowires or 0D nanoclusters [54, 55]. At such small length scales, most of the

Electrical sensitization Chemical sensitization

O-

e-

O2

O- 2R+O2 2RO

O-

O- O- O-

R RO

Noble Metal Cluster

41

atoms are surface atoms, thus significantly increasing the effective number of sites

available for reactions. Increase in surface area to volume ratio with decrease in grain

size is very important in the context of gas sensing. Thus reducing the grain size

plays an important role in applications that involve surface reactions such as

catalysis, chemical gas sensing, etc.

Fig. 2-6 demonstrates the grain size dependence of sensitivity of SnO2 films

exposed to CO and H2 [56]. It can be observed that grain size reduction below 10 nm

results in a drastic increase in sensitivity.

Figure 2-6: The effect of particle size on gas sensitivity of SnO2 sensor exposed to CO and H2, adapted from [56].

Fig. 2-7 shows the response of electron beam evaporated In2O3 film to NO2 as a

function of grain size [57]. A remarkable increase in sensitivity is observed upon

grain size reduction from 20 nm to 5 nm. The depletion layer depth (Debye length)

plays an important role if the grain size is very small. For most nanostructures,

Debye length is of the order of the grain diameter (considering spherical particles,

nanowires or nanotubes) or their width in case of nanobelts and other flat

nanostructures. Under such conditions the surface chemical processes strongly

influence the electronic properties.

42

Figure 2-7: Influence of grain size on gas response of In2O3 film to NO2, adapted from [57].

A 2 to 3 order increase in sensitivity was observed in the case of In2O3 films

when the grain size was decreased from 60-80 nm to 10-50 nm [58].

For grains large enough to have a bulk region unaffected by the surface

phenomena, i.e. when the grain diameter d >> λD (Debye length), the surface charge

carrier density ns, is given by [9]

)13.2(exp

Tk

qVnn

B

sbs

where nb is the concentration of free charge carriers (electrons), qVs is activation

energy, kB is Boltzman's constant and T is temperature in Kelvin.

When the grain size d ≤ λD is comparable to depth of depletion layer (Debye length),

the activation energy ΔE is related to Debye length as [9]

)14.2(4

DB

DTkE

where D is the grain diameter.

43

If ΔE is comparable to the thermal energy then a homogeneous electron

concentration is attained in the grain and leads to the flat band case. For grain sizes

lower than 10 nm, complete depletion of charge carriers occurs inside the grain and a

flat band condition results in a wide range of temperatures. Also, as the gas sensing

mechanisms involve adsorption processes, the physical properties and the shape of

the material determine the response of the nanosensor. Higher area/volume ratio

favors gas adsorption (and change in conductivity), decrease the response time and

increase the sensitivity of the device. Additionally, the time taken for gas molecules

to diffuse into and out of the volume of nanostructures is minimized [59].

Because of the advantages nanostructured material based sensors have over the

same sensor built with bulk material, nanoparticles and thin films of metal oxides

(less than 1000 nm thick) have been used to detect a wide variety of gases. As a

result, nanotechnology offers sensor devices with improved sensing properties.

2.8 Structural properties of tungsten oxide

Amorphous tungsten oxide film has large open pores and constitutes clusters

which are built from 3-8 WO6 octahedra [60]. These octahedra are linked together at

corners or edges by W-O-W bonds or water bridges [61, 62]. Random packing of the

clusters results in open structure or voids which are usually filled with molecular

water taken from the air [61]. The ionic conduction of amorphous WO3 film is

carried out by proton transport through water bridges in pores. On the other hand, the

electronic conduction is done by clusters linked through W-O-W bonds.

Tungsten oxide exhibits a cubic perovskite-like structure based on the corner

sharing of WO6 octahedra, with the O atoms at the corner of each octahedron [63]. A

schematic view of WO3 crystal structure is shown in Fig. 2-8. The symmetry of

tungsten oxide is lowered by two distortions: tilting of WO6 octahedra and

44

displacement of tungsten from the center of its octahedron [64]. These distortions

result in a number of temperature dependant phases of WO3, which are listed below

[64-68]:

Monoclinic ε – WO3 phase below -50ºC.

Triclinic δ – WO3 phase (from -50 to 17ºC).

Monoclinic γ – WO3 phase (from 17 to 330ºC) stable at room temperature.

Orthorhombic β – WO3 phase (from 330-740ºC).

Tetragonal α – WO3 phase above 740ºC.

In addition to the above phases, a metastable hexagonal WO3 phase has also been

reported around 400ºC [69].

Figure 2-8: Tungsten oxide (WO3) crystal structure.

The electrical and optical properties of tungsten oxide are strongly dependent on

the crystalline structure. It has been demonstrated that resistivity decreases with

increasing temperature [70]. Migas et al [71] investigated the dispersion of bands

near the gap region of different phases of WO3. They found that the dispersion of

bands near the gap region is identical in all phases except the hexagonal WO3 phase,

x

y z

W

O

Absent

Corner sharing arrangement of WO6

45

which exhibited flat band structure in the last valence band and the first conduction

band.

One of the elementary defects in tungsten oxide structure, as in most metal

oxides, is the lattice oxygen vacancy, where an oxygen atom is absent from the

normal lattice site. The octahedral changes from corner sharing to edge sharing

lattice with the formation of crystallographic shear planes built up of edge and plane

sharing structures, which corresponds to partial replacement of W6+ to W5+ and W4+

ions. This leads to formation of non-stoichiometric tungsten oxides WO3-x and

strongly influences its electronic properties [16]. The influence of oxygen vacancies

on electronic properties of different phases of WO3 has been investigated in terms of

bonding-antibonding interactions [72]. Changes in band gap and positions of the

valence band maximum and the conduction band minimum have been revealed for

different phases of WO3 due to presence of oxygen vacancies [73]. From an

electronic point of view, an oxygen vacancy increases the electronic density on the

metallic (W) adjacent cations. This raises the Fermi level into the conduction band

where it partly crosses some bands and partly stays in the energy gap. It has been

observed that the energy gap shrinks by about 0.5 eV if an oxygen vacancy is formed

[71]. The absorption coefficient also shifted to the lower energy range if an oxygen

vacancy is present. This leads to formation of donor-like states slightly below the

edge of conduction band of the oxide and it acquires semiconducting properties [74].

In nanostructured WO3, the bandgap generally increase with reducing grain size [75].

This has been experimentally observed as a blue shift of the optical absorption

bandedge as the nanostructure dimensions are reduced.

The effect of doping on band gap shifts of WO3 has also been reported in the

literature [71, 76]. Hwang et al showed that Mg-doped WO3 has a band gap similar

46

to undoped WO3 [77]. Doping with Cu, Ag, Mo is also reported without much

success [71, 78, 79]. However, doping with 3.4% nitrogen significantly reduced the

band gap from 3.0 eV to 2.2 eV [80].

2.9 Metal oxide based gas sensors

The history of semiconducting metal oxide based gas sensors dates back to over

5 decades. Most of the metal oxides have been commercially in use for sensing

various gases. Table 2-1 is an attempt to encompass some selected metal oxides and

corresponding target gas based on the literature survey.

Table 2-1: Some selective publications of metal oxides used for gas sensing.

Reference Additives Target gas Metal Oxide

[81-90] Pt, Ag, Pd, Os, Fe, Au, In,

Ru, Bi2O3, CeO2, CuO

CO, CH4, SO2, N2O, CO2, NO2,

CH3OH, C2H5OH, H2, H2S, NH3,

LPG

SnO2

[91-96] La, Pt, Cr2O3, WO3 CH3OH, C2H5OH, O2, H2, NH3, NO2 TiO2

[97-103] Al, Sn, Cu, Pd, Fe2O3 NH3, H2, NO2, CH4, CO, H2S,

CH3OH, C2H5OH, LPG ZnO

[104-106] Ti NH3, CO, NO2 MoO3

[107-110] Au, Zn, Pt, Pd, RuO2 CH4, Benzene, toulene, CO, NO2 Fe2O3

[111-114] Al, SiO2/Si Humidity, CH4, NH3 Al2O3

[115, 116] TiO2 NO2, O2, NH3, Humidity Cr2O3

[117] ZnFe2O4 Ethanol CdO

[118-121] SnO2, Pd, Ta2O5, WO3, NiO O2, CO, CH4, NO, NH3 Ga2O3

[122-127] Mg, Zn, Mo, Re, Au, Pd NO2, NH3, H2S, O3 WO3

Tungsten oxide (WO3) is a well-known n-type semiconductor with a band gap of 2.6

-3.6 eV [128, 129] that has been used not only in catalytic/photocatalytic [130],

electrochromic applications [131] but also in solid state gas sensors. The number of

research publications where tungsten oxide is used for gas sensing applications has

increased dramatically during recent years leading this material to be the second

most studied metal oxide sensor for gas sensing applications after SnO2. As the main

47

focus of this PhD research is tungsten oxide, in the following section, the state of the

research of nanostructured tungsten oxide sensors will be explored.

2.10 Tungsten oxide (WO3) based gas sensors

The first report on WO3 for gas sensing was published by Shaver in 1967. He

showed that conductivity of Pt-activated WO3 increased by one order of magnitude

on exposure to H2. In the following years, several reports on WO3 based sensors have

been published. It has been found that WO3 can be used for detecting a variety of

gases such as H2, CH4, NH3, CO, NO, CHOH, O2, H2S, NO2, C2H5OH, O3, (CH3)3N,

SO2, Cl2. A summary of selected publications related to NH3, H2S and NO2 is

presented in Table 2-2. A detailed literature survey and state of the art of WO3 gas

sensors is given below.

The sensing properties of screen-printed WO3 films doped with Bi2O3 to NO

were investigated by Tomchenko et al [132]. The films showed good sensitivity to

NO over a wide range of temperature (200-500ºC). It was demonstrated that a bottom

thick-film layer of about 15µm determines the sensing characteristics of WO3 based

sensors. The authors also investigated fabrication of planar type WO3 based thick

film sensors for online monitoring of NO concentrations.

Bi2O3 doped WO3 sensors showed a good sensitivity to low NO concentrations

(2-300 ppm) at 300ºC. Penza et al [133] investigated NOx gas sensing properties of

WO3 thin films fabricated by reactive rf sputtering on glass substrates. Palladium,

platinum and gold were evaporated as activator layers onto the films. The sensor was

found to possess excellent sensitivity towards NO and NO2 gases in the temperature

range 100-300ºC. The optimum operating temperature was found to be 150ºC and

200ºC for Pt and Au doped WO3 films, respectively. They concluded that the

activator layers act as catalysts and speed up the response to NO and NO2 at low

48

temperatures and improve the selectivity with respect to other reducing gases (CO,

CH4, H2, SO2, H2S, NH3).

Table 2-2: Summary of published results on gas sensing characteristics of WO3.

Material Concentration

(ppm) Temperature

(ºC) Sensitivity* Year (Reference)

Method of preparation

NH3

WO3:Au 50 450 40 1992 ]134[ Powder-dip coating WO3 10 300 1.1 1995 [135] Sputtering WO3 1000 300 6 2000 [136] Drop coating

WO3:Mg 30 350 7.65 2000 [137] Powder-dip coating WO3 100 400 7 2003 [138] Sputtering

WO3 10 200 18.2 2005 [139] Modified thermal

evaporation WO3 100 220 7 2008 [140] Thermal oxidation

WO3:Pt 1000 260 1000 2006 [141] RF sputtering H2S

WO3 200 200 2 1990 [142] Powder dip coating WO3:Au 100 200 30 1993 [143] Sputtering

WO3:Au 100 250 50 2000 [144]

Sputtering

WO3: Au 3.5 250 46 2006 [141] RF sputtering WO3 10 200 10000 2001 [145] Deposition

NO2 WO3 80 300 97 1991 ]146[ Powder-dip coating WO3 100 350 20 1996 [147] Powder evaporation

WO3/TiO2 30 350 200 1999 [148] Powder-printing WO3:Mg 30 300 8 2000 [149] Powder evaporation

WO3 0.2 150 168 2011 [150] Hydrothermal

treatment WO3 100 250 33 2008 [140] Thermal oxidation

WO3: Au 10 150 430 2008 [151] Colloidal chemical

method WO3 0.045 200 0.8 2010 [152] Plasma spray

WO3:Ag 100 260 10 2006 [141] RF sputtering H2

WO3:Pt 200 200 8.5 2011 [153] RF sputtering WO3:Pd 200 200 10 2011 [154] Chemical Synthesis

WO3 10000 450 1.4 2011 [155] Thermal oxidation

WO3 1000 250 15 2010 [156] Electrochemical

anodizing CO

WO3:Pd 200 100 1.6 2011 [154] Chemical route

synthesis WO3 1000 150 0.2 2011 [157] Thermal evaporation

CoOHWO3:SWNT 1000 25 164

(S=ΔmV) 2009 [158] Hydrothermal

CoWO4 800 250 1.2 2010 [159] Pulsed laser deposition

*The sensitivity values presented in this table are absolute values as reported in the literature. Their magnitude cannot be compared as different authors have used different formula to calculate

sensitivity. The main intent of this table is to highlight the published results on WO3 based gas sensors.

Kawasaki et al [160] investigated NOx sensing properties of WO3 films

synthesized by Pulsed Laser Deposition (PLD) method. Their results demonstrated

that PLD technique is an efficient method to produce crystalline WO3 thin films

which are sensitive to NOx gases. It was also observed that substrate temperature (Ts)

49

is an important parameter in producing crystallinity in films and it increases with

increasing Ts. A sensor produced by using nanocrystalline WO3 and thin film

microfabrication technology showed a high degree of sensitivity to low NO2

concentration in the range from 50 to 550 ppb with relatively fast response and

recovery time [161]. The sensitivity was found to depend on surface structure, grain

size and geometrical heterogeneity of the films which were controlled by the

calcination temperature. The sensitivity was also found to depend on NO2 adsorbed

form on the surface which was affected by operating temperature. The optimal

sensing condition was found when the films were calcined at 550ºC for 1 hour and

operating the sensor at 300ºC. Ponzoni et al [139] obtained nanostructured WO3 gas

sensors by modified thermal evaporation technique which consisted of sublimation

from a metallic tungsten wire followed by oxidation in low vacuum conditions and

reactive atmosphere (PO2 = 0.22 mbar) with substrate heated at high temperature

(600ºC). The films were composed of agglomerates with nanometric size and present

high surface roughness and large effective area suitable for gas sensing applications.

Sensing measurements indicated high performance at a working temperature of

100ºC, high response towards sub-ppm concentrations of NO2 compared to lower

ones for NH3 and CO.

Iron-doped nanostructured WO3 thin films prepared by Electron Beam

Evaporation (EBE) technique were investigated towards acetaldehyde by

Tesfamichael et al [162]. Addition of 10 at.% Fe slightly decreased the band gap

energy and subsequent annealing at 300ºC for 1 hour in air further decreased the

band gap energy. The annealed Fe-doped WO3 sensor produced gas selectivity but a

reduced gas sensitivity towards acetaldehyde as compared to WO3 sensor. The NO2

sensing performance of pure and iron-doped WO3 thin films prepared by EBE

50

technique was investigated by Ahsan et al [163]. The pure WO3 films were found to

be highly sensitive to 5 ppm NO2 at lower temperature (150ºC). Doping with Fe was

found to decrease the film resistance significantly but also a reduced sensitivity. The

high sensitivity towards NO2 was attributed to the improved nanostructure obtained

through e-beam evaporation and subsequent annealing at 300ºC for 1 hour in air.

The H2S, N2O and CO sensing performance of Al-doped WO3 nanoparticle films

prepared by advanced gas deposition was investigated by Hoel et al [164]. A

maximum sensitivity towards H2S, N2O and CO was observed at temperatures

130ºC, 250ºC and 430ºC, respectively.

Iron addition lower than 10 at.% to WO3 films prepared by reactive RF

sputtering produced an enhancement in sensor response when exposed to NO2 [165].

Additionally, iron addition was found to be advantageous in sensing ozone, CO and

ethanol. NO2 and humidity sensing characteristics of WO3 thin films prepared by

vacuum thermal deposition and subsequent annealing in the temperature range of

300ºC-600ºC were investigated by Xie et al [166]. It was found that NO2 sensing was

strongly dependant on annealing and working temperature. Different WO3

mesoporous structures obtained by hard template route were used by Rossinyl et al

[167] to investigate their sensing response towards NO2. It was found that WO3 was

sensitive to NO2 even at low concentrations, although differences attributable to

different structures were observed. Introduction of copper as catalytic additive

improved both sensor response and response time [167].

Khatko et al [168] investigated the NO2, NH3 and ethanol sensing performance

of WO3 thin films deposited by reactive rf sputtering with interruptions during the

deposition process. Sensitivity was found to increase with increase in number of

interruptions and interruption time, which was attributed to observed grain size

51

reduction during interruption. In another study, the authors observed that the

response of these sensors to ozone is up to four times higher than that of the sensors

prepared using rf sputtering [169]. A high sensitivity to NO2 at a temperature of 50ºC

for a sensor made of WO3 particles of size ~36 nm was reported by Meng et al [170].

In this study, WO3 nanoparticles were prepared by evaporating tungsten filament

under a low pressure of oxygen gas, namely, by gas evaporation method. The

deposition was carried out under various oxygen pressures and samples were

annealed at different temperatures. The sensitivity was found to increase with

decreasing particle size, irrespective of oxygen partial pressure during deposition and

annealing temperature.

Solis et al [171] investigated the H2S response of nanocrystalline WO3 thick

films prepared by evaporation of tungsten metal by an electric arc discharge in

reactive atmosphere. The structure was found to consist of monoclinic and tetragonal

phases with a mean grain size of 40 nm. The influence of sintering temperature on

H2S sensitivity was studied. These films showed excellent sensing properties upon

exposure to low concentrations of H2S in air at room temperature. The conductance

of films sintered at 300ºC was found to increase by a factor of about 104 when

exposed to 10 ppm of H2S. A further rise in sintering temperature resulted in

decrease in sensor response. This effect was attributed to disappearance of the

tetragonal phase which may point at a specific crystal structure being responsible for

unique gas sensing properties of WO3. Nanoparticles and nanoplatelets of WO3 and

nanowires of WO2.72 were investigated for their H2S sensing characteristics by Rout

et al [172]. The WO2.72 nanowires emerged as good candidate for H2S sensors with

little effect of humidity (upto 60% relative humidity) as well as improved response

and recovery times.

52

The electrical response of WO3 based sensors for ozone detection was reported

by Boulmani et al [173]. Thin films (40 nm thick) of WO3 were deposited by rf

reactive magnetron sputtering on SiO2/Si substrate with Pt interdigitated micro

electrodes. The response towards ozone was found to strongly depend on film

morphology which depends on the oxygen concentration during the deposition

process. The sensor response was also affected by bias voltage, sputtering time and

oxygen concentration during deposition.

WO3 thin films with different effective surface area were deposited under

various discharge gas pressures at room temperature by using reactive magnetron

sputtering and their response towards H2 was investigated by Shen et al [174]. It was

observed that effective surface area and pore volume of WO3 thin films increased

with increasing discharge gas pressure. The peak sensitivity for H2 gas was observed

at 300ºC. The results indicate the importance of achieving high effective surface area

on improving the gas sensing performance. The hydrogen response of WO3

nanotextured thin films coated with a 2.5 nm Pt layer was investigated by Yaacob et

al [175]. The films exhibited gasochromic characteristics when tested in visible-NIR

(400-900 nm) range. The total absorbance in this range increased by 15% upon

exposure to 600 ppm H2 in synthetic air and 60% upon exposure to 10,000 ppm H2 in

synthetic air. The films were found to be highly sensitive with stable and repeatable

responses towards low concentrations of H2 at 100ºC. However, the recovery time

was found to be slow at room temperature.

The effect of cerium oxide additive on WO3 nanoparticles prepared by solgel

method towards Volatile Organic Compound (VOC) gases was investigated by Luo

et al [176]. The highest gas response of Ce-added WO3 samples was found to shift to

lower temperatures compared to pure WO3 samples. Grain boundaries were pinned

53

due to CeO2 which resulted in reduction in grain size and increase in surface area.

Complex impedence spectroscopy analysis indicated that grain boundary resistance

increased and grain boundary capacitance decreased with increasing concentration of

CeO2 which indicates that Ce ions mainly exist at WO3 grain boundaries and help to

improve the microstructure.

Tungsten oxide films have also shown a good sensing performance towards

ethanol [177-181]. The sensitivity towards ethanol has been attributed to the

desorption of oxygen at the surface of grains [181].

Carbon monoxide (CO) sensing characteristics of CoOOH-WO3 doped with Au

and SWCNT was investigated by Wu et al [158]. It was found that mixture with a

CoOOH-WO3 ratio of 2:1 had the highest sensor response at room temperature.

Doping with 1 wt% SWCNT and 0.1 wt% Au in CoOOH-WO3 was found to boost

the CO response by 3.6 times. Azad et al [182] investigated the sensing performance

of WO3 towards 100 ppm CO. The authors achieved sensitivity towards CO by

modulating ambient oxygen partial pressure to create oxygen deprivation on the

metal oxide surface. However, WO3 responded to CO only at 450ºC.

2.11 Deposition techniques of nanostructured metal oxide films

For the purpose of gas sensing, coatings or films of nanostructured metal oxides

are deposited on substrate. This section will briefly highlight the different processing

routes for deposition of metal oxide films. The methods can broadly be classified

into two main categories:

1. Thin film processes: Processes such as solgel, spray pyrolysis, physical

vapour deposition and chemical vapour deposition are classified under this

category.

54

2. Thick film processes: Processes such as screen printing and spin coating are

classified under this category.

Fig. 2-9 shows a schematic of the classification of different processing routes for

nanostructured materials. It is to be noted that there is a huge number of deposition

processes and only a few processes are discussed in the following sections.

Figure 2-9: A schematic of the classification of processes used for deposition of nanostructured films.

2.11.1 Thin Film Processes

2.11.1.1 Solgel

The process involves the hydrolysis of a metal organic compound such as a

metal alkoxide or inorganic salts such as chlorides to produce a colloidal solution

[183]. The hydrolysis can take place with the help of alcohol, acid or base. The sol is

then allowed to age and settle. This is called the gelation step. The sol can then be

coated on the substrate by either spin/dip coating to form a 'xerogel' film.

Alternatively, the solvent from the sol can be evaporated to precipitate particles of

uniform size and then these can be screen printed.

General Processing Routes for Nanostructured materials

Thin films Thick films

Wet process Vapour phase deposition

Solgel

Spray pyrolysis

Spin casting

Screen printing CVD PVD

RGTO

55

2.11.1.2 Spray pyrolysis

It involves atomization of a liquid precursor through a series of reactors, where

the aerosol droplets undergo evaporation, solution condensation within the droplet,

drying, thermolysis of the precipitate particle at higher temperature to form a

microporous particle which then gets sintered to give a dense film [184].

2.11.1.3 Chemical vapour deposition (CVD)

Chemical vapour deposition involves exposing a substrate of choice to a mixture

of volatile precursors that react or decompose on the substrate to give the desired

product. A wide variety of CVD techniques that are currently being in use are

Atmospheric pressure CVD (APCVD), Atomic layer CVD (ALCVD), Low pressure

CVD (LPCVD), Metal organic CVD (MOCVD), Microwave plasma assisted CVD

(MPCVD), Plasma enhanced CVD (PECVD) and Molecular Beam Epitaxy (MBE).

2.11.1.4 Physical vapour deposition (PVD)

Physical vapour deposition (PVD) uses physical means as opposed to chemical

vapour deposition techniques. The various techniques used in this method are

evaporation, sputtering and pulsed laser deposition. Thermal evaporation and

electron beam evaporation are classified under the category of evaporation.

The advantage of using CVD and PVD is that they offer an enormous amount of

control over film thickness, stoichiometry, and microstructure. Therefore, highly

controlled properties can be achieved using these methods. This project uses thermal

evaporation technique to deposit nanostructured WO3 thin films. A more detailed

description of thermal evaporation is given below.

56

2.11.2 Thermal evaporation

Thermal evaporation is a physical vapour deposition process in which the atoms

or molecules from a thermal vaporization source in a vacuum chamber reach the

substrate without collisions with residual gas molecules. This type of PVD process

requires a vacuum of better than 10-5 mbar. A high vacuum (10-7 mbar) or ultra high

vacuum (<10-9 mbar) is required if high purity films are desired. A schematic sketch

of the thermal evaporation chamber is shown in Fig. 2-10.

The system essentially consists of vaporization source, substrate holder,

substrate heater, a shutter and a thickness monitor enclosed in a chamber. Thermal

vaporization requires that the surface and generally a large volume of material must

be heated to a temperature where there is an appreciable vapour pressure. Common

heating techniques for evaporation include resistive heating, high energy electron

beams, low energy electron beams, inductive (rf) heating or arc deposition. The most

common way of heating materials that vaporize below about 1500ºC is by contact to

a hot surface that is heated by passing a current [183]. The evaporator must be able to

contain molten liquids without extensive reaction and also the molten liquid must be

prevented from falling from the heated surface. This is accomplished either by using

a container such as a crucible. The heated surface can be in the form of a wire,

usually stranded, boat, basket, etc., which can withstand high temperatures without

melting.

57

Figure 2-10: A typical sketch of thermal evaporation chamber.

Typical resistive heater materials are W, Mo, Ta, C and BN/TiB2 composite

ceramics. Typically, resistive heating is carried out at low voltage and very high

current (several hundreds of amperes) using AC transformer supplies. The vapour

travels from the source to the substrate in a straight line (line of sight). Physical

masks can be used to intercept the flux, producing defined patterns of deposition on

the surface. Thermally vaporized atoms may not always condense when they

impinge on the surface of substrate, instead they can be reflected or re-evaporate. Re-

evaporation is a function of the surface temperature and the flux of depositing atoms.

Hence, the substrate temperature should be carefully selected during deposition.

Deposition rates in vacuum can vary significantly. The rates can range from less than

one monolayer per second (MLS) (<3 Å/s) to more than 104 MLS (>3 µm/s). Various

factors such as power input to the source, system geometry, material, vacuum

Fixture temperature

monitor

Fixture heater

Substrate heater

(radiant)

Vacuum chamber

Thermocouple

Vacuum Guage High Vacuum

pumping Roughing pump

Shutter

Deposition rate

monitor

Substrate

View port

Vaporization source

Fixture rotation tooling

Gas Inlet

High Vacuum guage

58

condition, angle of deposition, substrate temperature and source to substrate distance

affect the deposition rate. In terms of deposition thickness, a uniform deposit over a

planar surface can be obtained by using multiple sources with overlapping patterns as

shown in Fig. 2-10. However this produces source control and flux distribution

problems [185]. By moving the substrate away, the uniformity over a given area can

be improved, but the deposition rate is decreased. The most common technique to

improve uniformity is to move the substrate in a random manner over the vapour

source using different fixture geometries. Thermally evaporated films have a residual

tensile stress. Compressive stresses exist only when deposition is done at very high

substrate temperatures. The thin film microstructure can be controlled by two factors:

the ratio of growth temperature to melting temperature (of the thin film material) and

the amount of added energy. This is depicted in Fig. 2-11 by the famous structure

zone model of Thornton [186], which is often used for thermal evaporation and

sputtering techniques. In case of thermal evaporation, the arrival energies are low

and the thin film structure would follow the high pressure side of the diagram. One

important feature that is often found in vacuum deposition chambers is the relatively

large distance between the heated source and the substrates. This is to minimize the

radiant heating from the source and allows elaborate fixture motion to randomize the

position of the substrate. In thermal evaporation, the temperature of the heating

filament is sufficient enough to sublimate metal or metallic compounds. Hence

semiconducting metallic oxides such as SnO2, TiO2, WO3, etc can be deposited using

thermal evaporation.

The physical and chemical properties of tungsten oxide thin films are largely

affected by a number of factors such as film thickness, substrate type, substrate

temperature, annealing temperature, annealing time and dopant addition [133, 187-

59

190]. Vacuum thermal evaporation offers the benefit of controlling the film

thickness, substrate temperature and co-evaporation of more than one species. Ultra

high purity films can also be achieved by improving the vacuum in the deposition

chamber.

Figure 2-11: Thornton Zone diagram [186].

2.12 Summary

In this chapter, a literature review on various aspects of gas sensors and state of

the art of WO3 based gas sensors is presented. The chapter starts with introduction to

the construction principle of gas sensors, basic characteristics, mechanisms of gas

sensing and limitations of current gas sensors. This is followed by literature review

on role of additives and importance of dimensions in gas sensing. The structural

properties of tungsten oxide and the state of the art in WO3 based gas sensing are

discussed in detail. This is followed by a brief overview of various deposition

techniques for thin film synthesis with emphasis on thermal evaporation.

60

CHAPTER 3 : EXPERIMENTAL METHODS

3.1 Introduction

This chapter is divided into three main sections:

Deposition and doping techniques (evaporation and implantation), which

discuss the details of WO3 thin film preparation (Section 3.2).

Characterization techniques to investigate the physical, chemical and

electronic properties of WO3 thin films (Section 3.3).

Characterization of gas sensing, which discusses the methodology adopted

for testing the sensor devices (Section 3.4).

It is beyond the scope of this thesis to give complete and detailed information on

each technique. Hence, the intent will be to give main information which is required

to understand the research carried out in this study.

3.2 Deposition of nanostructured WO3 thin films by thermal

evaporation

3.2.1 Material

In the present study, 99.9% purity WO3 and Fe powders of sizes 20 μm and 100

μm, respectively, were obtained from Sigma Aldrich Pty Ltd. Before deposition, the

powders were placed in dessicator in order to avoid any moisture and

decontamination. For the purpose of doping, iron was mixed with WO3 and the

mixture was evaporated. The physical properties of as-received WO3 and Fe powders

are listed in Table 3-1.

61

Table 3-1: Physical properties of WO3 and Fe powders.

WO3 Fe Density 7.16 g/mL at 25ºC 7.86 g/mL at 25ºC

Molecular weight 231.85 g/mol 55.85 g/mol

Melting point 1473ºC 1535ºC

3.2.2 Substrate

A gas sensor basically comprises of a sensitive layer deposited on a substrate

provided with two pre-printed electrodes for the measurement of the electrical

characteristics. The sensor device can be heated by its own heater, which is separated

from the sensing layer and the electrodes by an electrically insulating layer. Alumina,

silicon or glass is often used as substrate material. The electrodes are usually made of

Au or Pt, whereas Pt heating element is used for heating purpose. During the recent

years, interdigitated electrodes are being used more extensively than one finger

electrodes (see Fig. 3-1). In this case, instead of one electrode with two fingers, an

array of electrode fingers is used, thereby covering a larger surface area under the

film and improving the response time.

In this study, WO3 films were deposited by thermal evaporation at room

temperature on silicon substrate with interdigitated Pt electrodes (Electronics Design

Center, Case Western Reserve University, Cleveland, USA). A thin SiO2 film (10 nm

thick) was thermally grown on the substrate before printing interdigitated Pt

electrodes on it. The size of the substrate was 8 mm x 8 mm x 0.5 mm. The electrode

fingers have a line width and line thickness of 100 μm and a height of 300 nm,

respectively. A snapshot of the interdigitated Pt electrode on silicon substrate is

shown in Fig. 3-1.

62

Figure 3-1: Interdigitated Pt electrodes on Si substrate.

3.2.3 Film deposition

A bell jar type PVD unit (Varian Coater with AVT Control System, Australia)

was used to deposit pure and Fe-doped WO3 thin films. Prior to deposition, the

chamber was evacuated to 4 x 10-5 mbar. The substrates were mounted on a substrate

holder which was placed at a distance of 38 cm in line of sight from the evaporation

source. Before the actual deposition, a trial deposition was done to determine the

deposition parameters including the power required to evaporate tungsten oxide

powder. The powder was evaporated at about 650 W using constant voltage of 20 V

and variable current. 99.106 g WO3 powder mixed with 0.894 g of Fe was co-

evaporated during the Fe-doping. Deposition of all films was carried out at 4 x 10-5

mbar, 650 W, and the deposition rate of 35 nm per second. A thickness monitor

adopting crystal oscillator was used to adjust the thickness of the films to 300 nm. To

obtain uniformity of deposition, all the films were deposited under the same

conditions of vacuum, power and deposition rate.

3.2.4 Ion implantation

Implantation, which induce changes in the surface composition, morphology and

chemical bond structure, is an effective method to improve physical and chemical

63

properties of the material [191]. It has been reported that the optical band gap of the

films can be reduced by implantation [192].

In order to implant the WO3 thin films with small amount of iron, the ion

implantation facility at Australian Nuclear Science and Technology Organization

(ANSTO) was employed. Samples of WO3 thin film were mounted on a rotating

sample holder with a good electrical contact, which reduced charging effects during

ion implantation. Implantation was performed in a vacuum chamber, which was

evacuated to a base pressure of about 3 × 10-6 mbar. After achieving the required

vacuum, a high current was applied to the Fe cathode to generate Fe ion beam. The

beam was incident normally onto the sample surface at 40 KeV. Uniformity of the

implantation was assured by sweeping the sample across the ion beam.

3.2.5 Post deposition heat treatment of the films

Post deposition heat treatment of the films was carried out in order to improve

the microstructure and crystalline properties of the deposited films. Samples were

annealed at 300ºC and 400ºC in air for a period of 2 hours with a heating and

cooling rate of 2ºC per minute. It has been reported previously that annealing of WO3

film at temperatures below 300oC has not changed the particle size significantly

[128]. Annealing was performed in tube furnace at Microelectronics and Materials

Technology Center, RMIT University.

3.3 Characterization of nanostructured WO3 thin films

In order to characterize the WO3 thin films, various analytical techniques have

been used. Some of the most important techniques used for characterization are

discussed below:

64

3.3.1 Transmission Electron Microscopy (TEM)

Transmission electron microscopy is used to determine the nanostructure of the

films (grain size, structure and crystallinity). In transmission electron microscopy,

the sample is exposed to a beam of highly accelerated electrons. The electrons

interact with the material when they pass through the sample. As TEM works on

transmission principle, the sample thickness should be very small (less than 100 nm).

After interaction, the beam travels through the TEM column and produces an image

on a fluorescent screen and can be imaged using photographic film.

In the present work, a Jeol 1200 TEM was used at an accelerating voltage of 120

kV. Samples were investigated by scratching the film and placing it on TEM grid.

Three main features that have been investigated by this technique are size and shape

of WO3 nanoparticles and crystalline structure in the film.

3.3.2 Atomic Force Microscopy (AFM)

The surface morphology of the WO3 thin films was studied in ambient

conditions using atomic force mircroscopy. An NT-MDT P47 Solver Scanning Probe

Microscope was used for this purpose. This is a powerful technique with high spatial

resolution to study the surface morphology by scanning the surface with a special tip.

The AFM can be operated in different modes depending on sample type and surface

properties. In the present study, the WO3 film surface was scanned by a silicon tip

(radius of curvature 10 nm) in semi-contact mode over an area ranging from 500 nm2

to 2000 nm2. The mean grain size, grain distribution and surface roughness were

determined by using the Nova NT-MDT Image Analysis Software.

65

3.3.3 Rutherford Backscattered Spectroscopy (RBS)

RBS provides information on the profile of concentration as a function of depth

in a film. A beam of 2-3 MeV He+ ions is directed perpendicularly on the sample

surface. The He+ ions lose energy during collisions with electrons and occasionally

with nuclei. When the positively charged He+ ion comes close to the nucleus of an

atom, it is repelled by positively charged nucleus. These repulsive forces increase

with the mass of the target atom. By measuring the energy of the recoiled ions, the

information on the depth profile of the elements can be determined. In the present

study, the Ion Beam Analysis (IBA) facility at Australian Nuclear Science

Technology Organization (ANSTO) has been used to conduct RBS experiments. A

1.8 MeV He+ beam was directed on the film surface under a vacuum of 7 x 10-6

mbar. The concentration profile of constituent elements in WO3 film has been

determined using RBS. The experimental data was fitted using SIMNRA. The

sample was modelled by dividing it into a series of slabs, perpendicular to the normal

of the surface with each slab having a different thickness. The software calculates the

energy of the projectile on each facet of interaction. This is the basis of the calculated

spectrum. Once the spectrum is obtained, the depth profile is presented in terms of

slab concentration and thickness.

3.3.4 X-Ray Photoelectron Spectroscopy (XPS)

X-ray photoelectron spectroscopy (XPS) provides both elemental and chemical

state information of a material. When the sample is illuminated with monochromatic

x-rays, photoelectrons are emitted from the surface. The kinetic energy of these

emitted electrons is characteristic of the element from which the photoelectron

originates. The position and intensity of the peaks in the energy spectrum provide the

desired chemical state and quantitative information. The chemical state of an atom

66

alters the binding energy (BE) of a photoelectron which results in a change in the

measured kinetic energy (KE). The BE is related to the measured photoelectron KE

by the following equation.

)1.3(KEhBE

where h is the photon (x-ray) energy. These chemical shifts help to determine the

chemical or bonding information of elements.

In the present study, XPS analysis was performed using Kratos AXIS Ultra XPS

incorporating a 165 mm hemispherical electron energy analyser and using

monochromatic Al K X-rays (1486.6 eV) at 150 W (15 kV, 10 mA) incident at 45o to

the sample surface. Photoelectron data was collected at take off angle of 90o. Survey

(wide) scans were taken at analyser pass energy of 160 eV and multiplex (narrow)

high resolution scans at 20 eV. Survey scans were carried out over 1200-0 eV

binding energy range with 1.0 eV steps and a dwell time of 100 ms. Narrow high-

resolution scans were run with 0.05 eV steps and 250 ms dwell time. Base pressure

in the analysis chamber was maintained at 1.3 x 10-9 mbar and at 1.3 x 10-8 mbar

during sample analysis. Depth profiling of the film was also carried out by etching

the surface with an Ar source at a rate of 10 nm per second.

3.3.5 X-Ray Diffraction (XRD)

X-ray diffraction is widely used to characterize crystalline materials. The

technique gives information about crystallinity, orientation, composition, phases,

internal lattice strain and particle size. A schematic of X-ray diffraction is shown in

Fig. 3-2. The diffraction is represented by Bragg’s law as follows:

)2.3(2 rhkl nSind

Where dhkl is interplanar distance,

67

is the angle of reflection,

nr is the order of reflection and

is the wavelength of the incoming wave.

Figure 3-2: Schematic of X-ray diffraction.

For thin film analysis, grazing incident x-ray diffraction analysis (GIXRD) is

performed. In GIXRD measurement, the angle of incidence of the x-rays with respect

to the sample surface is kept fixed at a very low value. This value has to be higher

than the critical angle of total external reflection to ensure penetration of the x-rays

into the thin film. These measurements are performed in a parallel beam geometry,

which makes it non-sensitive to sample height displacement. Since the penetration

depth is significantly reduced due to low angle of incidence, x-ray beam is confined

within the thin film. The GIXRD technique is employed in this project as the film

thickness is very small (300 nm) and it is important to confine the x-ray beam within

the film.

In the present study, XRD analysis of WO3 thin films was performed on

PANanalytical XPert Pro Multi Purpose Diffractometer (MPD). A Cu K radiation of

wavelength 1.540 Å was used. The incident angle was kept at 2o and the 2 range

was kept between 10o to 85o with a step size of 0.05o.

dhklsindhkl

Incoming x-rays

Scatterd x-rays

68

3.3.6 Raman spectroscopy

Raman spectroscopy is a powerful tool to monitor the intensity and wavelength

of light that is scattered inelastically from molecules or crystals. It is used to

determine the chemical structure and physical state of the film. Raman effect occurs

when light strikes molecules and interacts with the electrons of the bonds of those

molecules. Subsequently, Raman scattering occurs due to these interactions and the

wavelength of the scattered light is shifted with respect to the wavelength of the

incident light. The change in wavelength is determined by analysing the spectrum of

the scattered light.

From the Raman spectrum, the wave numbers of Raman shifts are plotted

against their respective intensities generated from the interaction between phonons

and molecular vibrations. These phonon modes are related to the chemical bonding.

Therefore, information on molecules, bonds or defects can be obtained. Thus, this

technique can be applied to the identification of the type of bonds as well as

crystallinity in WO3 films.

In the present study, Raman measurements were performed using an

Oceanoptics QE 6500 spectrometer. A 532 nm line from an argon ion laser was used

as the excitation source. To avoid local heating of the samples, a power as low as

about 5 mW was incident onto the samples. A Raman shift between wavenumbers

200 cm-1 and 1200 cm-1 has been measured.

3.4 Gas sensing characterization

WO3 gas sensors prepared by the author were exposed to hydrogen (H2), ethanol

(C2H5OH) and carbon monoxide (CO). These gases at relatively low concentrations

(TLV: CO: 25 ppm (toxicity); ethanol: 200 ppm (toxicity), 6.7% (explosive limit);

69

H2: 4% (explosive limit)) [8] are considered harmful, not only to human life but to

the environment.

The primary physical hazards associated with hydrogen gas are its flammability

and potential for explosions. It forms flammable mixtures in air over a wide range of

concentrations and very low energy is needed to ignite hydrogen-air mixtures [193].

Hence, there is a need to detect hydrogen leaks to avoid any possible explosion.

Carbon monoxide is a major industrial gas that has significant fuel value. However,

CO is a toxic chemical, which is harmful to human health. Inhalation of CO can

cause interruptions in normal supply of oxygen in the bloodstream and influence

vital functions of the body [194]. Therefore, it requires early detection to prevent

health hazards. Ethanol is a major breakdown product of foodstuff where bacteria or

fungi develop. Sensitive ethanol sensors are necessary if the onset of spoilage is to be

detected as early as possible and remedial measures can be taken [195]. Stable

ethanol sensors can also be used as breathalysers to monitor human breath [196]. In

the present study, the WO3 sensor responses to various concentrations (10-10,000

ppm) of the above mentioned gases at various operating temperatures (100ºC to

300ºC) were measured. The relative humidity (RH) in the testing chamber during

the whole measurement was kept at zero (0% RH). All the gases were diluted in

synthetic air to achieve the desired concentrations. For all the experiments, the total

flow was adjusted to 200 sccm. Table 3-1 lists the gases and their concentrations

used in the present study.

Table 3-2: Target gases and their concentrations.

Target Gas Concentrations (ppm)

Hydrogen (H2) 600, 1250, 2500, 5000, 7500, 10000 Ethanol (C2H5OH) 12, 33, 67, 185

Carbon monoxide (CO) 50, 100, 250, 500, 1000

70

The sensor performance is typically expressed in terms of response parameters

such as response amplitude and response and recovery times. Many authors use the

term ‘sensitivity’ to indicate response amplitude. However, in a true sense,

sensitivity of a gas sensor is defined as the derivative of the response to the gas

concentration [11]. Therefore, keeping in view the general trend of using the

definitions, the term ‘sensitivity’ would be used to indicate the response amplitude of

the sensors in this study.

The sensitivity for reducing gases such as H2, ethanol and CO, denoted as

Sreducing is defined as the ratio of change in resistance to resistance in gas:

)3.3(gas

reducing R

RS

where Rair is the resistance in air under stationary conditions and Rgas represents the

resistance after the sensor is exposed to the target gas during a definite time.

Equation 3.3 can be applied ton-type material such as WO3 and reducing gases such

as H2, ethanol and CO. For the same type of material and oxidising gases, the

sensitivity is defined as:

)4.3(air

oxidizing R

RS

Fig. 3-3 shows the schematic diagram of the gas chamber setup. The gas cell was

built using 20 mm thick machined Teflon blocks and a fused quartz lid. The total cell

volume is approximately 30 mm2. The small volume of the gas chamber ensures

short gas equilibrium times within the chamber, and hence, the response times can be

considered to be the true response of the gas sensors. The sensor device was mounted

on a planar alumina micro-heater. The external power supply controlled the

71

operating temperature upto 400ºC. This gas chamber was then enclosed by an

environmental chamber to maintain 0% humidity and ambient temperature.

The response curve was recorded under a continuous flow of known amount of

target gas. This method is known as Flow through method [8]. Fig. 3-4 shows the

concept of the flow through method. The concentration of the gas is controlled by

mixing it with synthetic air using mass flow controllers. For recovery measurements,

the mass flow controller of the target gas is switched off. The sensor response can be

recorded repeatedly as a function of different concentrations.

Figure 3-3: Schematic diagram of gas chamber setup.

Fused quartz lid

Teflon

Printed circuit board

Gas Inlet

Gas Outlet

O-ring seal

Heater

Sensor

72

Figure 3-4: Schematic of sensor testing by flow through method.

In the present study, a sequence control computer was utilized to computerize

the pulse sequence of the target gas concentrations. Initially, synthetic air was passed

through the chamber at testing temperature until the stable baseline resistance was

observed. Then a sequence of target pulse for 10 minutes followed by synthetic air

pulse. This procedure was continued until a stable baseline was observed after

alternate pulses. This was followed by the experimental sequence of pulses and data

was recorded. Each sensor was tested at temperatures between 100ºC to 300ºC at

intervals of 50ºC under various concentrations of mentioned gases, and optimum

operating temperature of each sensor was determined. This was followed by two full

range tests for each sensor and corresponding gas at the optimum operating

temperature.

Chamber

Exhaust

Resistance measurement

and heater power

Sensor

Mass Flow Controller

Mass Flow Controller

Car

rier

gas

Tes

t gas

Control PC

73

3.5 Summary

This chapter describes the processes and procedures used in the deposition and

characterization of pure, Fe-doped and Fe-implanted WO3 thin films. Thermal

evaporation and ion implantation techniques were used to synthesize these films as

described in Section 3.2 of this chapter. This is followed by Section 3.3 which

describes the various analytical techniques employed to investigate the physical,

chemical and electronic properties of these films. The last section (Section 3.4)

describes the gas sensing setup and procedure employed in this study to investigate

the gas sensing performance of these films.

74

CHAPTER 4 : THIN FILM CHARACTERIZATION

4.1 Introduction

The physical, chemical and electronic properties of nanostructured metal oxide

thin films are strongly dependant on the deposition techniques, deposition

parameters, composition of the metal oxide and post-deposition processes (e.g. heat

treatment). These also directly affect the gas sensing properties and performance of

the metal oxides gas sensors. Thus, to understand the various properties of

nanostructured tungsten oxide thin films synthesized in this research work, it is

necessary to characterize their physical, chemical and electronic properties.

This chapter is aimed at presentation and discussion of results obtained by

various characterization techniques that have been employed to investigate the

various properties of thermally evaporated WO3 thin films. Following are film

properties characterized using different techniques:

1. Structural properties of films characterized by:

- Atomic Force Microscopy (AFM).

- Transmission Electron Microscopy (TEM).

- X-ray diffraction (XRD).

2. Chemical and Electronic properties of films characterized by:

- Raman spectroscopy.

- X-ray photoelectron spectroscopy.

3. Compositional properties of film characterized by:

- Rutherford backscattered spectroscopy (RBS).

- Electron Diffraction (EDX).

75

In this work, the author has extensively employed the above techniques to

characterize nanostructured tungsten oxide thin films. The characterization results

are subsequently linked with their gas sensing properties in Chapter 5. For the

purpose of identification, the various films are designated as follows:

WO3 film (pure WO3 film).

Fe-doped WO3 film.

Fe-implanted WO3 film.

4.2 Characterization of Structural Properties

4.2.1 AFM and TEM analysis

The AFM surface topography of as-deposited WO3 film (Fig. 4-1a) shows a

nanostructured surface with well defined grains of mean size and roughness of ~13 nm

and 0.5 nm, respectively, as obtained by NT-MDT Nova Image Analysis Software.

The TEM image of the film (Fig. 4-1b) reveals a dense structure and grain size of

about 10-15 nm. The broad and hazy ring around the center spot in selected-area

diffraction pattern (SADP) indicates highly amorphous nature of the WO3 film (inset in

Fig. 4-1b). Annealing of the as-deposited WO3 film at 300ºC for 2 hours in air resulted

in the reduction of the grain size from 13 nm to 10 nm as shown by the AFM (Fig. 4-

2a), but no change was observed in roughness. The TEM image (Fig. 4-2b) shows a

very compact and dense film. The halo diffraction pattern (inset in Fig. 4-2b) indicates

that the film is still amorphous. Annealing of WO3 films at 400ºC in air for 2 hours

resulted in crystalline grain growth with well defined grains of mean size of ~ 5 nm

(Fig. 4-3). The film appears to be loosely packed after annealing at 400ºC. The bright

spots in the diffraction pattern (inset in Fig. 4-3) shows that the film is highly

crystalline.

76

Figure 4-1: AFM semicontact mode image (a) and TEM image (b) of as-deposited nanostructured WO3 film.

Figure 4-2: AFM semicontact mode image (a) and TEM image (b) of nanostructured WO3 film

annealed at 300ºC for 2 hours in air.

Figure 4-3: TEM image of nanostructured WO3 film annealed at 400ºC for 2 hours in air.

77

The AFM surface topography of as-deposited Fe-doped WO3 film is shown in Fig.

4-4. The surface reveals well defined grain boundaries with an average grain size of 15

nm (Fig. 4-4a). However, the grains appear to be densely packed as compared to the

pure WO3 film (Fig. 4-1a). Addition of iron also resulted in an increase in roughness to

0.6 nm as compared to 0.5 nm for as-deposited WO3 film. The TEM image (Fig. 4-4b)

reveals a compact structure and a mean grain size of about 15 nm, which is consistent

with the AFM results. The halo pattern in the SADP indicates the amorphous nature of

Fe-doped WO3 film (inset in Fig. 4-4b).

Figure 4-4: AFM semicontact mode image (a) and TEM image (b) of as-deposited nanostructured Fe-

doped WO3 film.

The surface topography of Fe-doped WO3 film annealed at 300ºC for 2 hours in

air is shown in Fig. 4-5. Annealing at 300ºC decreased the grain size from 15 nm to

10 nm (Fig. 4-5a) and surface roughness from 0.6 to 0.5 nm. Doping with iron and

subsequent annealing at 300ºC results in a topography whcih is similar to the pure

WO3 film annealed at 300ºC (Fig. 4-2a). The TEM image of this film (Fig. 4-5b)

shows a densely packed structure. The film appears to be significantly amorphous, as

it is evidenct from dominant continuous rings (inset in Fig. 4-5b).

78

Figure 4-5: AFM semicontact mode image (a) and TEM image (b) of nanostructured Fe-doped WO3

film annealed at 300ºC for 2 hours in air.

Fig. 4-6 shows the TEM image of Fe-doped WO3 film annealed at 400ºC for 2

hours in air. A mean grain size of the order of 5-10 nm is observed. Annealing at

400ºC induced significant crystallinity in the film, indicated by bright spots in the

diffraction pattern (inset in Fig. 4-6). The film appears to have less dense structure

upon annealing at 400ºC, which was not observed in as-deposited or 300ºC annealed

Fe-doped WO3 films. The crystallinity of this film is similar to that of 400oC annealed

WO3 film but with a slightly larger grain size.

Figure 4-6: TEM image of nanostructured Fe-doped WO3 film annealed at 400ºC for 2 hours in air.

79

The surface topography of Fe-implanted WO3 film shows a nanostructured

surface with a grain size of less than 10 nm and a mean roughness of 0.2 nm (Fig. 4-

7a). The image was obtained in phase contrast mode as the semicontact topography

mode was unable to resolve the surface. The film shows a densely packed surface

and a very small grain size. The TEM image (Fig. 4-7b) also shows a very dense and

compact film characterized by a highly amorphous nature, as observed by the

diffraction pattern (inset in Fig. 4-7b).

Figure 4-7: AFM semicontact mode image (a) and TEM image (b) of nanostructured Fe-implanted

WO3 film.

Upon annealing the Fe-implanted WO3 film at 300ºC for 2 hours in air, a slight

improvement in surface topography, characterized by small clusters consisting of few

grains (Fig. 4-8) is observed. The grain size is less than 10 nm and roughness is

reduced to 0.1 nm. The TEM image (Fig. 4-8b) shows a dense and compact film with

a highly damaged surface. The film appears to be predominantly amorphous, as

observed from the diffraction pattern (inset in Fig. 4-8b). Fig. 4-9 shows the TEM

image of Fe-implanted WO3 film annealed at 400ºC for 2 hours in air. The film

appears to be very dense and highly amorphous as indicated by the inset in Fig. 4-9.

The grain size could not be quantified from TEM analysis in this case.

80

Figure 4-8: AFM semicontact mode image (a) and TEM image (b) of nanostructured Fe-implanted

WO3 film annealed at 300ºC for 2 hours in air.

Figure 4-9: TEM image of nanostructured Fe-implanted WO3 film annealed at 400ºC for 2 hours in air.

4.2.2 GIXRD analysis

The GIXRD patterns of WO3 films before and after annealing (at 300ºC and

400ºC) are shown in Fig. 4-10. The as-deposited and 300ºC annealed films do not

show any diffraction peak (Fig. 4-10a,b), which indicates that as-deposited WO3 film

is highly amorphous and annealing this film at 300ºC for 2 hours in air did not induce

any crystallinity in the film. This is consistent with the TEM observations. However,

81

significant crystallinity is observed in the film after annealing at 400ºC for 2 hours in

air (Fig. 4-10c).

0

50

100

150

200

10 20 30 40 50 60 70 80

Inte

ns

ity

(AU

)

2(degree)

(a) as-deposited WO3

0

50

100

150

200

10 20 30 40 50 60 70 80

Inte

nsi

ty (

AU

)

2 (degree)

(b) WO3 annnealed @ 300oC

0

200

400

600

800

1000

10 20 30 40 50 60 70 80

Inte

nsi

ty (

AU

)

2 (degree)

(c) WO3 annealed @ 400oC(2

00)

(111

)

(220

)

(22

1)

(112

)

(420

)

(312

)

Figure 4-10: GIXRD pattern of nanostructured WO3 film, (a) as-deposited, (b) annealed at 300ºC in air

for 2 hours and (c) annealed at 400ºC in air for 2 hours.

The peaks observed at 2 = 24.112º, 28.538º, 34.361º, 41.615º, 49.843º, 55.684º,

61.941º are closely associated with the monoclinic WO3 phase [197]. It should be

noted that the lattice parameters of orthorhombic WO3 phase are very similar to

monoclinic phase, and thus, these two phases cannot be distinguished within the

accuracy of GIXRD data. The two intense peaks usually observed at 2=24.278° and

34.117° belong to (2 0 0) and (2 2 0) monoclinic planes, corresponding to d=3.663 Å

82

and 2.626 Å, respectively [198]. The lattice parameters were found to be a = 7.375 Å,

b = 7.375 Å and c = 3.903 Å and its unit lattice volume is about 212.38 Å3.

The GIXRD patterns of Fe-doped WO3 films before and after annealing at 300ºC

and 400ºC for two hours in air are shown in Fig. 4-11. The as-deposited and 300ºC

annealed Fe-doped WO3 films do not show any diffraction peak (Fig. 4-11a,b),

which indicates that as-deposited Fe-doped WO3 film is highly amorphous and

annealing the Fe-doped WO3 film at 300ºC for 2 hours in air did not induce

crystallinity in the film. This result is consistent with the TEM observations.

However, after annealing at 400ºC, significant crystallinity is indicated by the

diffraction peaks in GIXRD pattern (Fig. 4-11c). The peak positions of 400ºC

annealed Fe-doped WO3 film closely match with that of 400ºC annealed WO3 film

(Fig. 4-12). The lattice parameters were found to be a = 7.372 Å, b = 7.372 Å and c =

3.897 Å and its unit lattice volume is about 211.87 Å3. The observed matching of

GIXRD pattern of WO3 and Fe-doped WO3 film can be explained from a

crystallographic viewpoint. The ionic radius of W6+ (0.62 Å) is similar to that of Fe3+

(0.64 Å). Moreover, the W6+ is octahedrally coordinated with O2-. In iron oxides, the

crystal field stabilization energy of Fe3+ is higher for octahedral orientation than for

tetrahedral orientation [199]. Therefore, Fe3+ can fulfil the same coordination as that

of W6+. Consequently, Fe-doped WO3 film shows the same crystal structure as that

of WO3 film. Similar crystal structures were also observed between pure ZnO and

Fe-doped ZnO by Han et al [200]. The observed shift in peak positions of Fe-doped

WO3 film compared to WO3 film, although very little (0.02 Å), can be attributed to

the ionic radius of W6+ and Fe3+.

83

0

50

100

150

200

10 20 30 40 50 60 70 80In

ten

sit

y (A

U)

2 (degree)

(a) as-deposited Fe-doped WO3

0

50

100

150

200

10 20 30 40 50 60 70 80

Inte

ns

ity

(AU

)

2 (degree)

(b) Fe-doped WO3 annealed @ 300 oC

0

200

400

600

800

1000

10 20 30 40 50 60 70 80

Inte

ns

ity

(AU

)

2 (degree)

(c) Fe-doped WO3 annealed @ 400oC(200)

(111

)

(220

)

(221

)

(112

)

(420

)

Figure 4-11: GIXRD pattern of nanostructured Fe-doped WO3 film, (a) as-deposited, (b) annealed at

300ºC for 2 hours in air and (c) annealed at 400ºC for 2 hours in air.

0

500

1000

1500

2000

10 20 30 40 50 60 70 80

Inte

ns

ity

(AU

)

2 (degree)

Fe-doped WO3 annealed at 400 oC

WO3 annealed at 400 oC

Figure 4-12: Comparison of GIXRD patterns of nanostructured WO3 and Fe-doped WO3 films


84

The ionic radius of Fe3+ is slightly greater than that of W6+ and this can cause slight

distortion in the crystal lattice when WO3 is doped with Fe, and consequently a shift

the diffraction peaks. Such distortions can also produce a number of defects in the

film, making it a better candidate for gas sensing.

The GIXRD patterns of Fe-implanted WO3 films before and after annealing are

shown in Fig. 4-13. The implanted film appears to be highly amorphous (Fig. 4-13a)

as indicated by the diffraction pattern, where no characteristic peaks of WO3 are

observed. Upon annealing at 300ºC, there is no change in crystallinity of the film

(Fig. 4-13b). However, after annealing at 400ºC, the two characteristic peaks of

monoclinic planes of WO3 are observed (Fig. 4-13c). These two peaks observed at

2=24.025º and 33.875º belong to (2 0 0) and (2 2 0) monoclinic planes of WO3,

respectively [198]. Annealing at 400ºC appears to have induced a slight crystallinity

in the film, however, the crystallinity achieved in pure WO3 and Fe-doped WO3 films

after annealing at 400ºC could not be achieved in case of Fe-implanted WO3 film.

4.2.3 Summary and outlook

The as-deposited pure, Fe-doped and Fe-implanted WO3 thin films synthesized

by thermal evaporation are amorphous. The mean grain size in pure and Fe-doped

films is between 10-15 nm. The grain size of Fe-implanted film could not be

quantified due to its highly amorphous and damaged film structure. AFM, TEM and

GIXRD analyses have shown that annealing of these films at 300ºC for 2 hours in air

could not induce any crystallinity, indicating that the films are predominantly

amorphous. However, the grain size of the pure WO3 film reduced from 13 to 10 nm

and that of Fe-doped films from15 nm to 10 nm. Both the pure and Fe-doped WO3

films showed no change in porosity after annealing at 300oC. Annealing at 400ºC in

air for 2 hours significantly improved the crystallinity and porous structure in the

85

pure and Fe-doped WO3 thin films without any significant change in grain size.

However, Fe-implanted WO3 thin film annealed at this temperature (400ºC) remains

essentially amorphous. This is attributed to the highly damaged film structure caused

by implantation.

0

50

100

150

200

250

10 20 30 40 50 60 70 802 (degree)

Inte

nsi

ty (

AU

) (a) Fe-Implanted WO3

0

50

100

150

200

10 20 30 40 50 60 70 802 (degree)

Inte

nsi

ty (

AU

)

(b) Fe-Implanted WO3 annealed @ 300oC

050

100150200250300350400

10 20 30 40 50 60 70 80

Inte

nsi

ty (

AU

)

2 (degree)

(c) Fe-Implanted WO3 annealed @ 400oC

(200)

(220)

Figure 4-13: GIXRD pattern of nanostructured Fe-implanted WO3 film, (a) as-deposited, (b) annealed

at 300ºC for 2 hours in air and (c) annealed at 400ºC for 2 hours in air.

86

4.3 Characterization of chemical and electrical properties

4.3.1 Raman spectroscopy

The Raman spectra of as-deposited and annealed WO3 films are shown in Fig. 4-

14. Two characteristic bands are associated with WO3. The first band lies between

200-500 cm-1 and is associated with O-W-O bending vibration modes. The second

band lies in the range 600-1000 cm-1 and is associated with W-O stretching vibration

modes. The as-deposited WO3 film exhibited week and broad Raman band centred at

315 cm-1 and 799 cm-1. These features are characteristic of amorphous materials and

are usually assigned to O-W-O deformation modes and O-W-O stretching vibration

modes of monoclinic WO3 phase, respectively [201]. The amorphous nature of as-

deposited WO3 thin films is consistent to the results obtained by TEM and GIXRD

observations.

The film annealed at 300ºC also appears to be amorphous with a slight

broadening of peak at 315 cm-1. However, the crystallinity of this film increased after

annealing at 400ºC, as shown by the sharp peaks at 707 cm-1 and 799 cm-1 which are

characteristic of O-W-O stretching vibration modes [70], corresponding well with the

results obtained by GIXRD and TEM analysis.

The Raman spectra of as-deposited and annealed Fe-doped WO3 films are

shown in Fig. 4-15. The as-deposited Fe-doped WO3 thin film exhibited very weak

and broad Raman bands centered at 320 cm-1 and 804.4 cm-1, which are associated to

O-W-O deformation vibration modes and O-W-O stretching vibration modes,

respectively [201]. These weak and broad bands are indicative of the amorphous

nature of as-deposited Fe-doped WO3 thin film which is also confirmed by TEM and

GIXRD observations.

87

2000

2500

3000

3500

4000

4500

5000

5500

6000

200 400 600 800 1000 1200

WO3 annealed @ 300oC


as-deposited WO3

Inte

ns

ity

(AU

)

Raman shift (cm-1)

Si

Si

O-W-Odeformationmodes

O-W-Ostretchingmodes

799

707

315

Figure 4-14: Raman spectra of nanostructured WO3 films.

2400

2600

2800

3000

3200

3400

3600

200 400 600 800 1000 1200

as-deposited Fe-doped WO3

Fe-doped WO3 annealed @ 300oC


Inte

ns

ity

(AU

)

Raman shift (cm-1)

804.

4

712.

6320

394

O-W-Odeformationmodes

O-W-Ostretchingmodes

Si

Si

Figure 4-15: Raman spectra of nanostructured Fe-doped WO3 films.

Upon annealing at 300ºC for 2 hours in air, the intensity of these bands increased

slightly, indicating the onset of crystallinity growth of the film. Upon annealing at

400ºC for 2 hours in air, the peak intensity of O-W-O stretching modes at 712.6 cm-1

and 804.4 cm-1 increased significantly, indicating that the film is highly crystalline,

which corresponds well with the TEM and XRD observations. The O-W-O

stretching vibration mode peak positions of WO3 and Fe-doped WO3 films annealed

88

at 400ºC are compared in Table 4.1. A slight blue shift of about 5.5 cm-1 is observed

for both the peaks after doping with Fe. Such shifts are caused by shortening of O-

W-O bonds [202], which corresponds to slightly smaller cell parameters of Fe-doped

WO3 film as compared with WO3 film. The GIXRD analysis has shown that the

lattice parameters of Fe-doped WO3 film are slightly smaller than WO3 film.

However, the octahedral orientation of WO3 has been retained after doping with Fe,

indicating that the preferred oxidation state of Fe is Fe3+. This is evident from similar

XRD patterns of WO3 and Fe-doped WO3 films annealed at 400ºC and absence of

any Raman peaks associated with Fe in 400ºC annealed Fe-doped WO3 film.

Table 4-1: Comparison of the positions of O-W-O stretching vibration mode peaks observed for nanostructured WO3 and Fe-doped WO3 films annealed at 400ºC for 2 hours in air.

Raman Peak position (cm-1)

WO3 annealed at 400ºC 707 799 Fe-doped WO3 annealed at 400ºC 712.6 804.4

Blue Shift (cm-1) 5.5 5.5

Figure 4-16 shows the Raman spectra of Fe-implanted WO3 films. The as-

deposited film appears to be highly amorphous as no characteristic Raman peaks are

observed. Annealing the Fe-implanted film at 300ºC did not induce any crystallinity

in the film. However, after annealing at 400ºC, characteristic Raman peaks at 792

cm-1 and 706 cm-1 are observed. However, the intensity of these peaks is smaller than

those observed for WO3 and Fe-doped WO3 thin films annealed at 400ºC. This

indicates that the film is essentially amorphous even after annealing at 400ºC as

confirmed by GIXRD analysis.

89

2000

2500

3000

3500

4000

4500

5000

5500

6000

200 400 600 800 1000 1200

Fe-Implanted WO3

Fe-Implanted WO3 annealed @ 300oC


Inte

ns

ity

(AU

)

Raman shift (cm-1)

792

706

Si

Si

Figure 4-16: Raman spectra of nanostructured Fe-implanted WO3 films.

4.3.2 XPS analysis

Fig. 4-17 shows the XPS spectra obtained by wide survey scans on the surface of as-

deposited and annealed (300ºC and 400ºC) WO3 films between binding energies 0

and 1200 eV. Survey scan information is useful in identification of elements present

on the film surface. Peaks of O, N, C and W are observed in all the films. Presence of

carbon and nitrogen on the surface is attributed to atmospheric contamination. The C

peak measured at binding energy of 284.80 eV coincides with C 1s binding energy

reported in literature [203] and is, thus, used as a point for binding-energy reference.

Fig. 4-18 shows the XPS spectra of as-deposited and annealed (at 300ºC and 400ºC)

Fe-doped WO3 films between binding energies 0 and 1200 eV. Characteristic peaks

of O, N, C and W are observed as in the case of the pure WO3 film. However, no

characteristic peak of Fe was observed on the surface of all the films. The sensitivity

of XPS is only about 0.01 at.% over a depth of 10 nm. From RBS analysis, which is

discussed in the next section, the amount of Fe is only about 0.016 at.% over 10 nm

90

depth, which is much less than the critical amount that XPS can detect. This might

indicate the non-uniform deposition of Fe in the WO3. Hence, no Fe was observed on

the surface of the Fe-doped film as analysed using XPS.

0

5x104

1x105

1.5x105

2x105

2.5x105

3x105

020040060080010001200

as-deposited WO3

Co

un

ts/s

eco

nd

(A

U)

Binding energy (eV)

W 4dC 1s

N 1s

W 4f

O 1s (a)

0

5x104

1x105

1.5x105

2x105

2.5x105

3x105

020040060080010001200

Co

un

ts/s

eco

nd

(A

U)

Binding energy (eV)

W 4dC 1s

N 1s

W 4f

O 1sWO3 annealed @ 300oC (b)

0

5x104

1x105

2x105

2x105

020040060080010001200

Co

un

ts/s

eco

nd

(A

U)

Binding Energy (eV)

O 1s

N 1sC 1s

W 4d

W 4f

WO3 annealed @ 400oC (c)

Figure 4-17: XPS wide spectra of nanostructured WO3 films, (a) as-deposited, (b) annealed at 300ºC for

2 hours in air and (c) annealed at 400ºC for 2 hours in air.

91

0

5x104

1x105

2x105

2x105

3x105

3x105

020040060080010001200

Co

un

ts/s

eco

nd

(A

U)

Binding energy (eV)

W 4dC 1sN 1s

W 4f

O 1sas-deposited Fe-doped WO3 (a)

0

5x104

1x105

2x105

2x105

3x105

020040060080010001200

Co

un

ts/s

eco

nd

(A

U)

Binding energy (eV)

W 4dC 1s

N 1s

W 4fO 1s

Fe-doped WO3 annealed @ 300oC (b)

0

5x104

1x105

2x105

2x105

3x105

020040060080010001200

Co

un

ts/s

eco

nd

(A

U)

Binding Energy (eV)

W 4fO 1s

N 1s

W 4d

C1s

Fe-doped WO3 annealed @ 400oC (c)

Figure 4-18: XPS wide spectra of nanostructured Fe-doped WO3 films, (a) as-deposited, (b) annealed at

300ºC for 2 hours in air and (c) annealed at 400ºC for 2 hours in air.

The XPS spectra of Fe-implanted WO3 films are shown in Fig. 4-19. Characteristic

peaks of O, N, C, W and Fe are observed in all the films.

92

0

5x104

1x105

2x105

2x105

3x105

020040060080010001200

Co

un

ts/s

eco

nd

(A

U)

Binding energy (eV)

W 4dC 1s

N 1s

W 4f

O 1sFe-Implanted WO3

Fe 2p

(a)

0

1x105

2x105

3x105

4x105

020040060080010001200

Co

un

ts/s

eco

nd

(A

U)

Binding energy (eV)

W 4dC 1sN 1s

W 4f

O 1s


Fe 2p

(b)

0

1x105

2x105

3x105

4x105

020040060080010001200


Co

un

ts/s

eco

nd

(A

U)

Binding Energy (eV)

O 1s

C 1sN 1s

W 4fW 4d

Fe 2p

(c)

Figure 4-19: XPS wide spectra of nanostructured Fe-implanted WO3 films, (a) as-deposited, (b)

annealed at 300ºC for 2 hours in air and (c) annealed at 400ºC for 2 hours in air.

For the as-deposited WO3 film, the core level spectra of W 4f are observed at

binding energy Eb of 35.74 eV and 37.88 eV corresponding to W 4f7/2 and W 4f5/2,

respectively (Fig. 4-20). The Eb value for W 4f7/2 reported in literature is 35.8 eV

[204]. The measured value of 37.88 eV is in good agreement with those of WO3

powder, electron beam evaporated and electrodeposited WO3 films [205]. The W 4f

peak shapes get sharper with increasing annealing temperature, which indicates that

93

the surface becomes cleaner due to desorption of surface contaminants by annealing.

The broadening of peaks is associated with change in stoichiometry of the sample

surface, with the formation of different oxides such as WO2 or WO [206]. The W

4f7/2 peak of metallic tungsten is located at 31.50 eV [207]. The W 4f7/2 peaks located

at +4.5, +3 and +1.5 eV from the metallic tungsten W 4f7/2 peak are attributed to

W6+, W5+ and W4+ electronic states, respectively [208]. No significant change in

binding energy of W 4f7/2 is observed when the WO3 film is annealed at 300ºC.

However, annealing at 400ºC resulted in lowering of W 4f7/2 binding energy by 0.3

eV, indicating presence of mixed tungsten states [209]. The downshift of W 4f7/2

peak can be explained by the fact that, if an oxygen vacancy exists in the film, the

electronic density near its adjacent W atom increases. The 4f level binding energy is

expected to be at lower binding energy as the screening of its nucleus is higher

because of increased electronic density [210]. Oxygen vacancies play an important

role as adsorption sites for gaseous species and eventually, a minor shift of the

binding energy may imply greatly enhanced gas sensitivity [56].

0

1x104

2x104

3x104

4x104

5x104

6x104

323436384042

as-deposited WO3

WO3

annealed @ 300oC


Co

un

ts/s

eco

nd

(A

U)

Binding Energy (eV)

W 4f5/2

35.7437.88

W 4f7/2

Figure 4-20: W 4f core level high resolution spectra of nanostructured WO3 films.

94

0

1x104

2x104

3x104

4x104

5x104

6x104

7x104

32343638404244




Co

un

ts/s

ec

on

d (

AU

)

Binding Energy (eV)

35.8

35.7

35.63

W 4f5/2

W 4f7/2

Figure 4-21: W 4f core level high resolution spectra of nanostructured Fe-doped WO3 films.

The W 4f core level high resolution spectra of as-deposited and annealed (at

300ºC and 400ºC) Fe-doped WO3 films is shown in Fig. 4-21. For the as-deposited

Fe-doped WO3 film, the W 4f7/2 and W 4f5/2 peaks are observed at 35.80 eV and

37.95 eV, respectively, closely matching with the values reported in literature [205] .

Similarly to the pure WO3 thin film, the W 4f peaks get sharper with increasing

annealing temperature, which indicates that the surface becomes cleaner due to

desorption of surface contaminants by annealing. The W 4f7/2 peak in Fe-doped WO3

film is located at +4.3 eV with respect to the 4f7/2 peak of metallic tungsten located at

31.5 eV. This indicates the formation of mixed W states upon doping with Fe [208].

Compared with the pure WO3 film (W 4f7/2 peak located at +4.2 eV from metallic

tungsten 4f7/2 peak), doping with Fe appears to have altered the stoichiometry

towards nominal WO3 (+4.3 eV). No significant change in binding energy of W 4f7/2

is observed when this film is annealed at 300ºC. However, annealing at 400ºC

resulted in lowering of W 4f7/2 binding energy by about 0.2 eV, indicating the

95

presence of mixed tungsten states and a corresponding increase in number of oxygen

vacancies on the surface [210].

0

1x104

2x104

3x104

4x104

5x104

6x104

7x104

323436384042

Fe-Implanted WO3



Co

un

ts/s

eco

nd

(A

U)

Binding Energy (eV)35

.83 35

.58

35.70

W 4f7/2W 4f5/2

Figure 4-22: W 4f core level high resolution spectra of nanostructured Fe-implanted WO3 films.

The W4f core level high resolution spectrum of Fe-implanted WO3 films is

shown in Fig. 4-22. The Eb value of W 4f7/2 (35.83 eV) obtained for this film closely

matches with the literature reported value [204]. The W 4f peak shapes get sharper

with increasing annealing temperature, indicating that the surface is cleaner due to

desorption of surface contaminants such as carbon by annealing, leading to change in

stoichiometry of the sample surface. A significant change is observed in binding

energy of W 4f7/2 after annealing at 300ºC (about 0.25 eV), indicating the presence of

mixed tungsten states (W6+, W5+, W4+, etc) [209]. However, after annealing at 400ºC,

the W 4f7/2 peak downshifted only by about 0.13 eV with respect to the implanted

film. Annealing at 400ºC appears to have oxidized the implanted film accompanied

with a composition approaching the stoichiometry.

96

0

1x104

2x104

3x104

4x104

5x104

6x104

7x104

8x104

524526528530532534

as-deposited WO3



Co

un

ts/s

ec

on

d (

AU

)

Binding Energy (eV)

O 1s

530.7

530.4

532.9

Figure 4-23: O 1s core level high resolution spectra of nanostructured WO3 films.

The O 1s core level high resolution spectrum of the as-deposited WO3 film (Fig.

4-23) shows Eb of 530.7 eV. The main maximum did not change upon annealing at

300ºC, which is similar to the trend observed for W 4f7/2 peak. However, annealing at

400ºC lowered the binding energy Eb by 0.3 eV, which is same as the downshift

observed for W 4f peak after annealing at 400ºC. This is most likely based on a shift

of the Fermi level, corresponding to band bending due to desorption of surface

contaminants during annealing at 400ºC [211]. A small shoulder centred at about

532.9 eV is observed in the as-deposited and 300ºC annealed film. This shoulder

transformed into a peak when the film is annealed at 400ºC. Such feature is

characteristic of substoichiometric monoclinic tungsten oxides [212]. The formation

and increasing intensity of this feature is in the sequence WO3WO2. XPS analysis

reveals that after annealing at 400ºC, the film surface is free from contamination and

has mixed W states. This indicates the presence of oxygen vacancies in the film,

which is highly beneficial for gas sensing.

97

0

2x104

4x104

6x104

8x104

1x105

524526528530532534




Co

un

ts/s

eco

nd

(A

U)

Binding Energy (eV)

O 1s

530.

75

30.5

530.4

Figure 4-24: O 1s core level high resolution spectra of nanostructured Fe-doped WO3 films.

The O 1s core level high resolution spectra of as-deposited and annealed Fe-

doped WO3 films are shown in Fig. 4-24. The O 1s binding energy peak is observed

at Eb of 530.70 eV for as-deposited Fe-doped WO3 film, which is the same as that

observed for WO3 film. The peak intensity increased with annealing temperature,

which is attributed to the desorption of surface contaminants such as C during

annealing at 400ºC. Upon annealing at 300ºC, there is no significant shift in O 1s

peak position. However, when the Fe-doped WO3 film is annealed at 400ºC, the

binding energy of O 1s lowered by 0.3 eV as in the case of the pure WO3 film, in

which the lowered binding energy was ascribed to the shift in Fermi level [211]. The

W 4f7/2 and O 1s peak positions of pure WO3 and Fe-doped WO3 films annealed at

400ºC are compared in Table 4-2. The peak positions of both W 4f7/2 and O 1s did

not change significantly upon doping with Fe. This indicates that both the WO3 and

Fe-doped WO3 films have similar tungsten states after annealing at 400ºC.

98

XPS analysis reveals that the surface of Fe-doped WO3 film after annealing at

400ºC is free from surface contaminants and has mixed W states, similarly to the

WO3 film after annealing at 400ºC. This indicates the presence of oxygen vacancies

in both the films, which is highly beneficial for gas sensing.

Table 4-2: Comparison of W 4f7/2 and O 1s peak positions of nanostructured WO3 and Fe-doped WO3 films annealed at 400ºC for 2 hours in air.

XPS Peak position (eV)

W 4f7/2 O 1s

WO3 annealed at 400ºC 35.56 530.30

Fe-doped WO3 annealed at 400ºC 35.63 530.40

The O 1s core level high resolution spectrum of Fe-implanted WO3 films is

shown in Fig. 4-25. The Fe-implanted WO3 film shows an Eb value of 530.7 eV with

a significant shift of about 0.2 eV upon annealing at 300ºC. This shift, which is

similar to that observed in W 4f7/2, is due to the shift in Fermi level corresponding to

band bending caused by the desorption of surface contaminants such as C during

annealing [211]. However, after annealing at 400ºC, there is no change in O 1s peak

position, which indicates that annealing of Fe-implanted WO3 film at 400ºC shifted

the stoichiometry towards nominal WO3. The shift in peak position after annealing at

400ºC is similar to the change observed in W 4f7/2 peak (0.13eV) after 400ºC

annealing. Therefore, Fe-implanted WO3 film annealed at 400ºC has the same

tungsten states as in Fe-implanted WO3 film.

99

0

2x104

4x104

6x104

8x104

1x105

526528530532534

Fe-Implanted WO3



Co

un

ts/s

eco

nd

(A

U)

Binding Energy (eV)

530.7

530

.7 530.

5

O 1s

Figure 4-25: O 1s core level high resolution spectra of nanostructured Fe-implanted WO3 films.


Raman analysis has revealed that the as-deposited and 300ºC annealed pure, Fe-

doped and Fe-implanted WO3 films are amorphous. However, after annealing at

400ºC for 2 hours in air, the pure and Fe-doped WO3 thin films become crystalline,

whereas weak and broad peaks obtained in the Fe-implanted film annealed at 400ºC

are indicative of its amorphous nature. These results are in confirmation with TEM

and XRD characterization as discussed in Section 4.2. Iron addition by co-

evaporation and subsequent annealing at 400ºC slightly blue shifted the peaks by 5.5

cm-1. Both the Raman and GIXRD analyses reveal that iron is incorporated as a

substitutional impurity in the WO3 host matrix resulting in slightly smaller cell

parameters. This can lead to more distortions (defects) in the film structure and is

potentially beneficial for gas sensing.

XPS analysis reveals that annealing the films at 300ºC for 2 hours in air induced

a slight substoichiometry in all the films, and annealing at 400ºC for 2 hours in air

resulted in surface decontamination as observed by the high intensity of W and O

100

peaks. Moreover, the peak positions in both pure and Fe-doped WO3 thin films

shifted to lower binding energies indicating mixed tungsten states, correspondingly,

increase in number of oxygen vacancies. Both these films have similar tungsten

states after annealing at 400ºC. However, the Fe-implanted WO3 thin film annealed

at 400ºC has similar tungsten states as the as-deposited Fe-implanted WO3 thin film.

4.4 Characterization of compositional properties

4.4.1 Rutherford backscattered spectroscopy

The depth profile of the films was measured by using Rutherford Backscattered

Spectroscopy. The RBS spectra of as-deposited and 300ºC annealed WO3 films are

shown in Fig. 4-26. The as-deposited WO3 film spectrum (Fig. 4-26a) exhibits a

typical staircase structure with each step associated with an element in the film and

substrate. Well separated W peak with a high intensity is due to the higher mass

(atomic weight) of W compared with O and other trace elements such as N. The He+

particles are scattered with much higher recoil energy in the film than from the

substrate (Si) in this elastic scattering process. Upon annealing the WO3 film in air at

300ºC for 2 hours, the W peak becomes slightly sharper and more intense (Fig. 4-

26b) than the peak of the as-deposited film (Fig.4-26a).

Fig. 4-27 shows the depth profiles of the as-deposited and 300ºC annealed WO3

film. The depth profile of WO3 film (Fig. 4-27a) indicates the presence of O, N and

W. A high amount of nitrogen concentration is observed in the film. This can be due

to surface contamination of the film. Moreover, RBS is limited in resolution for light

elements with similar masses such as B, C, N and , hence other elements may

contribute to the high concentration of nitrogen. Presence of oxygen can be

attributed to adsorbed oxygen from the environment in addition to the lattice oxygen

101

existing in the film. The depth profile of 300ºC annealed WO3 film (Figure 4-27b)

shows a slight decrease in the amount of O and increase in the amount of W as

compared to the corresponding values for the as-deposited film (Figure 4-27a). This

reduced amount of oxygen may be attributed to desorption of surface contaminants

from the top few atomic layers.

0

500

1000

1500

2000

2500

3000

3500

4000

50 100 150 200 250 300 350 400

400 600 800 1000 1200 1400 1600 1800

ExperimentalSimulation

Co

un

ts (

AU

)

Channel

as-deposited WO3

Energy (KeV)

O+

N in

WO

3, SiO

2

O in

WO

3

Si i

n s

ub

stra

te

Si i

n S

iO2

Si s

ub

stra

teSiO

2

WO

3

He

W(a)

0

500

1000

1500

2000

2500

3000

3500

4000

50 100 150 200 250 300 350 400

400 600 800 1000 1200 1400 1600 1800

Experimental

Simulation

Co

un

ts (

AU

)

Channel


Energy (KeV)

O+

N i

n W

O3, S

iO2

O in

WO

3

Si i

n s

ub

stra

te

Si i

n S

iO2

W

Si s

ub

str

ate

SiO2

WO

3

He

(b)

Figure 4-26: RBS spectrum of (a) as-deposited and (b) 300ºC annealed nanostructured WO3 films.

102

0

0.2

0.4

0.6

0.8

0 50 100 150 200 250 300 350

Co

nc

entr

ati

on

(x

10

0 a

t%)

Depth (x1015 at/cm2)

O

W

N

Si

as-deposited WO3

(a)

0

0.2

0.4

0.6

0.8

0 50 100 150 200 250

Co

nc

entr

ati

on

(x1

00 a

t%)

Depth (x 1015 at/cm2)

OW

N

Si

WO3 annealed @ 300oC (b)

Figure 4-27: RBS depth profile of (a) as-deposited and (b) 300ºC annealed nanostructured WO3 films.

The RBS spectra of as-deposited and 300ºC annealed Fe-doped WO3 films are

shown in Fig. 4-28. The intensity of W peak in the as-deposited Fe-doped WO3 film

(Fig. 4.26a) is smaller than the other films (as-deposited WO3, Fig. 4.24a, 300oC

annealed WO3, Fig. 4.24b, and 300oC annealed Fe-doped WO3, Fig. 4.26b).

103

0

500

1000

1500

2000

2500

3000

3500

4000

50 100 150 200 250 300 350 400

400 600 800 1000 1200 1400 1600 1800


Co

un

ts (

AU

)


Energy (KeV)

Channel

O+

N in

WO

3, S

iO2

O in

WO

3

Si i

n s

ub

stra

te

Si i

n S

iO2

W

Si

su

bs

trat

e

SiO2

WO

3

He

(a)

0

500

1000

1500

2000

2500

3000

3500

4000

50 100 150 200 250 300 350 400

400 600 800 1000 1200 1400 1600 1800


Co

un

ts (

AU

)

Fe-doped WO3 annealed at 300 oC

Energy (KeV)

Channel

O+

N in

WO

3, S

iO2

O i

n W

O3

Si i

n s

ub

stra

te

Si i

n S

iO2

W

Si s

ub

stra

te

SiO2

WO

3

He

(b)

Figure 4-28: RBS spectrum of (a) as-deposited and (b) 300ºC annealed nanostructured Fe-doped WO3

film.

The depth profiles of the as-deposited and 300ºC annealed Fe-doped WO3 films

are shown in Fig. 4-29. The depth profile of as-deposited Fe-doped WO3 film shows

O, W, N and Fe (Fig. 4-29a) and the amount of Fe in the film is only about 0.5 at.%.

This film also appears to be contaminated on the surface, hence a high amount of

nitrogen is observed owing to the poor resolution of RBS for lighter elements. AFM,

104

GIXRD and Raman analyses have shown that addition of Fe resulted in slight

increase in grain size and distortions in the crystal structure. Addition of Fe appears

to have slightly changed the stoichiometry of the film (change in amount of O and

W). The depth profile of 300ºC annealed Fe-doped WO3 film (Fig. 4-29b) shows that

the amounts of W and O are similar to those in the 300ºC annealed WO3 film. AFM

analysis also shows the same grain size of 10 nm in both the WO3 and Fe-doped WO3

films annealed at 300ºC. This finding indicates that the stoichiometry (Table 4-3) of

WO3 is improved by annealing and the film is similar to 300ºC annealed WO3 film.

0

0.2

0.4

0.6

0.8

0 50 100 150 200 250 300

Co

nc

entr

atio

n (

x100

at%

)


O

W

Fe

N

Si


(a)

0

0.2

0.4

0.6

0.8

0 40 80 120 160

Co

nc

entr

ati

on

(x

100

at%

)

Depth (x 1015 at/cm2)

O

W

FeSi

N

Fe-doped WO3 annealed at 300oC (b)

Figure 4-29: RBS spectrum (a) and depth profile (b) of nanostructured Fe-doped film annealed at

300ºC for 2 hours in air.

105

The RBS spectrum and corresponding depth profile of Fe-implanted WO3 film is

shown in Fig. 4-30. The spectrum (Fig. 4-30a) exhibits a typical staircase structure

with each step associated with an element in the sample. Well separated peaks of W

and Fe are observed in the spectrum. The depth profile (Fig. 4-30b) reveals the

presence of O, N, Fe and W. This film also shows high concentration of nitrogen as

surface contamination owing to the poor resolution of RBS for lighter elements.

From the RBS depth profile, the amount of Fe was estimated to be 5.5 at.%.

0

500

1000

1500

2000

2500

3000

3500

4000

0 50 100 150 200 250 300 350 400

0 400 800 1200 1600


Co

un

ts (

AU

)

Fe-Implanted WO3

Energy (KeV)

Channel

O+

N in

WO

3, S

iO2

Si i

n S

iO2

Si i

n s

ub

stra

te

O i

n W

O3

Si

su

bs

tra

te

SiO2

WO

3

He

W

Fe

(a)

0

0.2

0.4

0.6

0.8

0 40 80 120 160

Co

nc

entr

ati

on

(x

100

at%

)


O

W

Fe

N

Si

Fe-Implanted WO3

(b)

Figure 4-30: RBS spectrum (a) and corresponding depth profile (b) of nanostructured Fe-implanted

WO3 film.

106

4.4.2 Energy dispersive x-ray spectroscopy (EDX)

Energy dispersive x-ray spectroscopy (EDX) was carried out during TEM analysis to

determine the elemental composition of the films. However, it must be understood

that the depth of penetration in EDX is of the order of a few microns, and, hence, it is

difficult to determine the accurate composition of the thin films 300 nm thick

synthesized in this research. Nevertheless, the presence of various elements in the

films can be confirmed. Fig. 4-31 shows the EDX spectrum of Fe-doped WO3 film

annealed at 400ºC. Peaks corresponding to W, O, C, Fe and other contaminants are

revealed. From the EDX analysis, the amount of Fe was found to be about 0.7 at.%,

which is similar to the amount obtained by RBS (0.5 at.%). Fig. 4-32 is EDX

spectrum of Fe-implanted WO3 film annealed at 400ºC, showing, the presence of W,

O, C, Fe and other contaminants. The amount of Fe obtained by EDX was about 3

at.%, which is lower than the concentration obtained by RBS (5.5 at.%). It is to be

noted that the concentration obtained by RBS is more reliable than those obtained by

EDX.

Figure 4-31: EDX spectrum of for 2 hours in air.

Fe-doped WO3 film annealed @ 400ºC

107

Figure 4-32: EDX spectrum of Fe-implanted WO3 film annealed at 400ºC for 2 hours in air.


EDX analysis has revealed the presence of W, O, C and Fe in Fe-doped and

Fe-implanted WO3 films annealed at 400ºC. The XPS analysis could not reveal

the presence of Fe in Fe-doped film due to its low sensitivity (0.1 at.% per 10

nm). To accurately determine the elemental composition of the films, RBS

analysis was carried out. The results from both RBS analysis indicate that the

amount of Fe is about 0.5 at. % in Fe-doped WO3 thin film and 5.5 at.% in Fe-

implanted WO3 thin film. RBS analysis has revealed desorption of surface

contaminants after annealing at 300ºC. Iron incorporation in WO3 film slightly

changed the stoichiometry which is restored by annealing the film at 300ºC.

Fe-Implanted WO3 film annealed @ 400ºC

108

CHAPTER 5 : GAS SENSOR CHARACTERIZATION

5.1 Introduction

In this chapter, the gas sensing performance of the nanostructured tungsten oxide

thin films is presented. Nanostructured pure WO3, Fe-doped WO3 and Fe-implanted

WO3 thin films were deposited onto conductometric transducers by thermal

evaporation and subsequently annealed at 300ºC and 400ºC for 2 hours in air to

develop novel devices for gas sensing. These sensor devices have been tested

towards various concentrations of H2, ethanol and CO in a dry air in the temperature

range of 100–300ºC. The experimental setup and parameters for gas sensing

measurements have been detailed earlier in Section 3.4. Chapter 4 has provided

information about the physical, chemical and electronic properties of nanostructured

WO3 thin films which is a first step to evaluate their suitability for gas sensing.

In this chapter, the sensing performance of nanostructured thin film devices

based on WO3 towards various gases is presented. This will provide important static

and dynamic parameters such as sensitivity, optimum operating temperature,

response and recovery times and baseline resistance stability of these films. The gas

sensing performance will be linked to the physical, chemical and electronic

properties of the films to develop an understanding of the factors that contribute to a

better sensor performance.

5.2 Response to Hydrogen

The as-deposited and annealed pure, Fe-doped and Fe-implanted WO3 thin film

based gas sensors were exposed to H2 in the temperature range of 100–300ºC at

increments of 50ºC to determine the optimum operating temperature of these sensors.

109

Table 5-1 shows the optimum operating temperature obtained for various sensors

upon exposure to H2. The as-deposited films did not show any response towards H2

in the selected temperature range. It has been shown experimentally that amorphous

films have poor sensor response characteristics [213]. The amorphous nature of the

as-deposited films seems to lead these films being not suitable for gas sensing.

Table 5-1: Optimum operating temperature of WO3 thin film based gas sensors upon exposure to H2.

Optimum operating temperature (ºC) for H2

as-deposited WO3 - WO3 annealed at 300ºC 280 WO3 annealed at 400ºC 150 as-deposited Fe-doped WO3 - Fe-doped WO3 annealed at 300ºC 250 Fe-doped WO3 annealed at 400ºC 200 as-deposited Fe-implanted WO3 - Fe-implanted WO3 annealed at 300ºC 200 Fe-implanted WO3 annealed at 400ºC 100

An optimum response to H2 was observed for 300ºC annealed and 400ºC

annealed WO3 films at different temperatures. The dynamic responses and

sensitivities of 300ºC and 400ºC annealed WO3 films exposed to H2 at optimum

temperatures are shown in Figs. 5-1 and 5-2, respectively. Table 5-2 shows the

response dynamics (response and recovery times) of these films.

The 300ºC annealed film is found to be not stable as observed from its baseline

resistance as well as response (Fig. 5-1a). This is due to the amorphous nature of the

film. The 400ºC annealed film shows a stable response and a stable baseline

resistance (Fig. 5-2a), which is attributed to its crystalline properties (Fig. 5-2a). The

response dynamics (response and recovery times) of both the films exhibit some

fluctuation at high concentrations due to saturation caused by increased coverage of

the film surface at high H2 concentration which requires longer time to purge out H2

from the surface when synthetic air is injected between two cycles.

110

4

6

8

10

12

14

16

18

20

500 1500 2500

Re

sis

tan

ce

(K

Oh

ms

)

Time (seconds)

WO3 annealed @ 300oC exposed to H

2

Operating temperature: 280 oC

600

pp

m

1000

0 p

pm

7500

pp

m

5000

pp

m

2500

pp

m

120

0 p

pm

(a)

0

0.1

0.1

0.2

0.2

0.3

0.3

0 2000 4000 6000 8000 10000 12000

Se

nsi

tiv

ity

(

R/R

ga

s)

H2 concentration (ppm)


2


(b)

Figure 5-1: Dynamic response (a) and sensitivity (b) to H2 measured for a nanostructured WO3 film


111

0

1

2

3

4

5

6

7

8

0 500 1000 1500 2000 2500 3000 3500

Re

sis

tan

ce (

M o

hm

s)

Time (seconds)


2

Operating temperature: 150oC

600 12

00

120

0

250

0

5000

7500 10

000

(a)

0

2

4

6

8

10

12

14

0 2000 4000 6000 8000 10000 12000

Se

nsi

tiv

ity

(

R/R

air)

H2 Concentration (ppm)


2


(b)

Figure 5-2: Dynamic response (a) and sensitivity (b) to H2 measured for a nanostructured WO3 film


Table 5-2: Response and Recovery times of annealed (at 300ºC and 400ºC) nanostructured WO3 films upon exposure to H2.


WO3 film annealed at 300ºC (Operating temperature: 280ºC)


Response time (seconds)

Recovery time (seconds)

Response time (seconds)

Recovery time (seconds)

600 32 28 152 44 1200 28 24 152 72 2500 28 28 164 80 5000 24 32 176 56 7500 20 64 156 72

10000 40 44 140 80

112

The resistance does not saturate within a few tens of seconds after introduction of H2.

The WO3 film annealed at 300ºC showed an optimum response at an operating

temperature of 280ºC (Fig. 5-1a) with a sensitivity S=0.3 to 10000 ppm H2 (Fig. 5-

1b). However, when the film is annealed at 400ºC, an optimum response is obtained

at much lower operating temperature of 150ºC (Fig. 5-2a) with a high sensitivity

S=10 to 10,000 ppm H2 (Fig. 5-2b). Moreover, the sensitivity of the 400ºC annealed

film is much greater than the 300ºC annealed film for all concentrations of H2 (Fig.

5-2b). However, the response dynamics of the 400ºC annealed WO3 film is found to

be slow at lower operating temperature of 150ºC (Table 5-2). The response and

recovery time to 10000 ppm H2 are 140 s and 80 s at 150ºC for the 400°C annealed

film, and 40 s and 44 s at 280°C for the 300°C annealed film. It has been reported

that the response amplitude (sensor signal) of WO3 based gas sensors decreases

drastically with increasing temperature from 150°C to 300°C and that the sensor

indicates slow response and long recovery times at low operating temperatures of

150°C [214].

WO3 is an n-type semiconductor material and commonly operates as a gas

sensor in the temperature between 200°C -500°C [215]. When it is exposed to a

reducing gas such as H2, the oxygen adsorbates on the film surface interact with the

gas and release electrons back to the film, causing a drop in film resistance. This

behaviour is observed for the 300°C annealed WO3 film when exposed to H2 at

280ºC. However, the opposite behaviour (i.e. an increase in resistance) is observed

for the 400ºC annealed WO3 film upon exposure to H2 at 150ºC. Such behaviour

cannot be explained by merely considering the microstructural properties such as

grain size and porosity of the film.

113

In polycrystalline materials, surface barriers which electrons have to overcome

for taking part in the conduction are formed at the intergranular surfaces. The height

of surface barrier depends on the concentration of charge carriers (oxygen

adsorbates) at the surface, and, therefore, overall resistance changes can be correlated

with changes in surface band bending. The overall resistance, and, hence, surface

band bending increased exponentially when the polycrystalline WO3 surface was

exposed to increasing concentrations of oxygen [216]. The increase in resistance

observed for the 400ºC annealed WO3 film exposed to H2 at an operating

temperature of 150ºC might arise from various forms of oxygen adsorbates (O-, O2-

and O2-) on WO3 surface, which depend on temperature. At 150ºC, the most

dominant form of adsorbed oxygen is O2- [217]. Upon exposure to H2, the O2

-

species dissociates into O- with the formation of water, as per the following equation.

)1.5(222 OOHOH

Insitu Raman analysis of WO3 films annealed at 300ºC and 400ºC has shown

that the rate of water desorption above 100ºC is much faster than the rate of water

formation on the film surface when WO3 film is exposed to H2 [218]. At 150ºC, the

high concentration range of H2 (600 ppm – 10000 ppm) produces more O- species on

the surface, leading to increase in surface barrier height, consequently increasing the

resistance. Hence, this can be a reason why an increase in resistance was observed at

lower operating temperature.

The high sensitivity to H2 at 150ºC observed for the 400ºC annealed WO3 film is

attributed to its very small grain size (5 nm) and porous structure (see Fig. 4-3). The

film is also composed of mixed W states, as observed in XPS analysis, resulting in

increased number of oxygen vacancies on the surface, when compared with the as-

deposited or 300ºC annealed film.

114

The dynamic responses and sensitivities to H2 at the optimum temperature of

250ºC measured for 300ºC and 400ºC annealed Fe-doped WO3 films are shown in

Figs. 5-3 and 5-4, respectively. Table 5-3 shows the response dynamics (response

and recovery times) of these films. The response dynamics (response and recovery

times) of these films also exhibit variation similar to that observed for the pure WO3

films.

120

160

200

240

280

500 1000 1500 2000 2500 3000

Re

sis

tan

ce

(K

oh

m)

Time (seconds)

Fe-doped WO3 annealed @ 300oC exposed to H

2


600

pp

m

1200

pp

m

2500

pp

m

5000

pp

m

7500

pp

m

100

00 p

pm

(a)

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0 2000 4000 6000 8000 10000 12000

Se

nsi

tiv

ity

(

R/R

ga

s)



2

Operating temperature: 250 oC (b)

Figure 5-3: Dynamic response (a) and sensitivity (b) to H2 measured for a nanostructured Fe-doped


115

0

0.5

1

1.5

2

0 1000 2000 3000 4000 5000 6000

Re

sis

tan

ce

(M

Oh

ms

)

Time (seconds)


2


600

pp

m

120

0 p

pm

2500

pp

m

5000

pp

m

7500

pp

m

1000

0 p

pm

(a)

2

2.5

3

3.5

4

4.5

5

5.5

6

0 2000 4000 6000 8000 10000 12000

Se

ns

itiv

ity

(R

/Rg

as)


Fe-doped WO3 annealed @ 400oC exp to H

2


(b)

Figure 5-4: Dynamic response (a) and sensitivity (b) to H2 measured for a nanostructured Fe-doped


Table 5-3: Response and Recovery times of annealed (at 300ºC and 400ºC) nanostructured Fe-doped WO3 films upon exposure to H2.


Fe-doped WO3 film annealed at 300ºC

(Operating temperature: 250ºC)


(Operating temperature: 200ºC) Response time

(seconds) Recovery time

(seconds) Response time


(seconds) 600 56 64 84 228 1200 52 64 76 236 2500 66 60 68 244 5000 52 52 64 232 7500 84 40 60 148

10000 44 52 100 172

116

The response of the 300ºC annealed Fe-doped WO3 film is not stable (Fig. 5-3a)

due to its amorphous nature. The response is also characterized by spikes especially

at higher concentration, which is due to saturation of the film surface at higher H2

concentrations. This film shows an optimum response at an operating temperature of

250ºC (Fig. 5-3a) with a sensitivity S=0.26 to 10000 ppm H2 (Fig. 5-3b). The 400ºC

annealed Fe-doped WO3 film shows a stable response (Fig. 5-4a) which is attributed

to its crystalline properties, coupled with small grain size of 5-10 nm and porosity

compared to 300oC annealed Fe-doped WO3 film which is amorphous and dense film

with a grain size of 10 nm. The optimum response of this film is observed at 200ºC

(Fig. 5-4a) with a high sensitivity S=5.5 to 10000 ppm H2 (Fig. 5-4b). Moreover, the

sensitivity of this film is much higher than the 300ºC annealed Fe-doped WO3 film

while the response dynamic are slower for all concentrations of H2. It was mentioned

earlier that the response amplitude (sensor signal) of WO3 based gas sensors

decreases drastically as the operating temperature increases from 150ºC to 300ºC and

the sensor is characterized by slow response and long recovery times at low

operating temperatures [214]. At an operating temperature between 200ºC-300ºC, the

most dominant oxygen species on the film surface is O- [217]. The conduction

mechanism is governed by the following equation:

)2.5(22 eOHOH

Upon interaction with H2, the electron concentration in the film increases,

causing a drop in resistance. The higher sensitivity of 400ºC annealed Fe-doped WO3

film compared to that of the 300ºC annealed Fe-doped WO3 film is due to the

presence of high number of oxygen vacancies in the 400ºC annealed film, as

observed from XPS and Raman analysis. The baseline resistance of the as-deposited

and annealed (at 300ºC and 400ºC) Fe-doped WO3 films is also higher than the as-

117

deposited and annealed (at 300ºC and 400ºC) WO3 films in the tested temperature

range. The increase in resistance of Fe-doped film is due to the acceptor nature of Fe

in the host matrix [219]. Doping with lower valency cations, such as Fe3+ increases

the Debye length through a reduction in the carrier concentration, which increases

the film resistance [220] . XRD and Raman analysis have shown that Fe is

incorporated in the host WO3 matrix, rather than as a catalyst on the surface (no

evidence of Fe on film surface from XPS observations). The sensitivity of 400ºC

annealed WO3 is almost double than that of 400ºC annealed Fe-doped WO3 film to

10,000 ppm H2. This is due to smaller grain size (5 nm vs 10 nm) and larger porosity

of the 400ºC annealed WO3 film (Fig. 4-3) than the 400ºC annealed Fe-doped WO3

film (Fig. 4-6). Additionally, the 400ºC annealed WO3 contained higher number of

oxygen vacancies than the 400ºC annealed Fe-doped WO3 film, leading to higher

sensitivity.

The dynamic response and sensitivities of Fe-implanted WO3 films annealed at

300ºC and 400ºC are shown in Figs. 5-5 and 5-6, respectively. The 300ºC annealed

film shows an optimum response to H2 at an operating temperature of 200ºC (Fig. 5-

5a). However, the response is not stable and a drifting baseline resistance is

observed, which is due to its amorphous nature [213]. A maximum sensitivity

S=0.13 is obtained for 7500 ppm H2 (Fig. 5-5b), with response and recovery times of

32 s and 64 s, respectively (Table 5-4).

118

80

85

90

95

100

105

110

0 500 1000 1500 2000 2500 3000

Re

sis

tan

ce (

K O

hm

s)

Time (seconds)

Fe-Implanted WO3 annealed @ 300oC exposed to H

2


600

pp

m

1200

pp

m

250

0 p

pm

500

0 p

pm

100

00 p

pm

750

0 p

pm

(a)

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0 2000 4000 6000 8000 10000 12000

Se

nsi

tiv

ity

(

R/R

gas

)


Fe-Implanted WO3 annealed @ 300

oC exposed to H

2

Operating temperature: 200oC (b)

Figure 5-5: Dynamic response (a) and sensitivity (b) to H2 measured for a nanostructured Fe-implanted

film annealed at 300ºC for 2 hours in air.

119

60

65

70

75

80

0 500 1000 1500 2000 2500 3000

Res

ista

nce

(K

oh

ms)

Time (seconds)


2


600

pp

m

1200

pp

m

2500

pp

m

5000

pp

m

7500

pp

m

1000

0 p

pm

(a)

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0 2000 4000 6000 8000 10000 12000

Se

ns

itiv

ity

(R

/Ra

ir)



2


Figure 5-6: Dynamic response (a) and sensitivity (b) to H2 measured for a nanostructured Fe-implanted


Table 5-4: Response and Recovery times of annealed (at 300ºC and 400ºC) nanostructured Fe-implanted WO3 films upon exposure to H2.


Fe-implanted WO3 film annealed at 300ºC


Fe-implanted WO3 film annealed at 400ºC





(seconds) 600 52 100 92 143

1200 44 56 120 196 2500 36 56 144 196 5000 36 64 136 212 7500 32 64 116 244

10000 36 72 97 136

120

The optimum response for the 400ºC annealed Fe-implanted film was observed

at a lower operating temperature (100ºC), characterized by a drifting baseline

resistance (Fig. 5-6a) and a lower sensitivity S=0.048 to 7500 ppm H2 (Fig. 5-6b),

with response and recovery times of 116 s and 244 s, respectively (Table 5-4). Both

the annealed films have a dense structure characterized by low porosity, and hence

lower sensing characteristics. However, the presence of higher number of oxygen

vacancies in the 300ºC annealed film, as observed from XPS analysis (Fig. 4-24)

improved the sensitivity of this film, although the response is characterized by noise,

which is attributed to its amorphous nature [213]. In case of 400ºC annealed Fe-

implanted WO3 film, a p-type sensor response is observed (Fig. 5-6a) which is

similar to that observed in case of 400ºC annealed WO3 film. At an operating

temperature of 100ºC, the dominant oxygen adsorbate is O2- [47]. Upon exposure to

H2, the O2- species dissociates into O- with the formation of water, as per equation

5.1. The high concentration range of H2 (600 ppm – 10000 ppm) produces more O-

species on the surface, leading to increase in surface barrier height, consequently

increasing the resistance. Hence, this can be a reason why an increase in resistance

was observed at lower operating temperature. At 100ºC, the rate of water desorption

from the surface is lower than rate of water formation [218]. Accumulation of water

on the surface of the film would increase the film resistance, hence, an increase in

baseline resistance is observed upon successive exposure cycles of H2.

5.3 Response to Ethanol (C2H5OH)

The as-deposited and annealed pure, Fe-doped and Fe-implanted WO3 thin film

based gas sensors were exposed to ethanol in the temperature range 100ºC-300ºC at

increments of 50ºC to determine the optimum operating temperature of these sensors.

Table 5-5 shows the optimum operating temperature obtained for various sensors

121

upon exposure to ethanol. The as-deposited films did not show any response to

ethanol in the selected temperature range. The highly amorphous nature of the as-

deposited films seems to make these films unsuitable for gas sensing [213]. The WO3

film annealed at 300ºC also did not show any response to ethanol. This may also be

attributed to the amorphous nature of the film. The dynamic response and sensitivity

of 400ºC annealed WO3 film exposed to ethanol at 150ºC are shown in Fig. 5-7. The

response and recovery times are shown in Table 5-6. Similarly to the response

dynamics upon exposure to H2, a variation in response and recovery times is also

observed upon exposure to ethanol at higher concentrations.

Table 5-5: Optimum operating temperature of nanostructured WO3 thin film based gas sensors upon exposure to ethanol.

Optimum operating temperature (ºC) for ethanol

as-deposited WO3 - WO3 annealed at 300ºC - WO3 annealed at 400ºC 150 as-deposited Fe-doped WO3 - Fe-doped WO3 annealed at 300ºC - Fe-doped WO3 annealed at 400ºC 150 as-deposited Fe-implanted WO3 - Fe-implanted WO3 annealed at 300ºC 250 Fe-implanted WO3 annealed at 400ºC -

The film exhibits a stable baseline resistance (Fig. 5-7a), due to its crystalline

properties. The response amplitude increased with increasing ethanol concentration,

reaching 0.2 for 185 ppm ethanol (Fig. 5-7b). The response and recovery times for

185 ppm ethanol are 180 s and 288 s, respectively (Table 5-6).

122

300

350

400

450

500

0 500 1000 1500 2000 2500 3000 3500

WO3 annealed @ 400oC exposed to ethanol


Re

sis

tan

ce

(K

oh

ms

)

Time (seconds)

12 p

pm

30 p

pm

60

pp

m

185

pp

m

(a)

0

0.05

0.1

0.15

0.2

0.25

0 50 100 150 200

Se

nsi

tiv

ity

(

R/R

ga

s)

Ethanol Concentration (ppm)

WO3 annealed @ 400oC exposed to ethanol


(b)

Figure 5-7: Dynamic response (a) and sensitivity (b) to ethanol measured for a nanostructured WO3

film annealed 400ºC for 2 hours in air.

Table 5-6: Response and Recovery times of 400ºC annealed nanostructured WO3 films upon exposure

to ethanol.



Response time (seconds) Recovery time (seconds)

12 120 296 30 144 260 60 164 292

185 180 288

123

As mentioned earlier, the dominant oxygen species on the film surface at 150ºC is

O2-, the conduction mechanism takes place by the following reaction.

)3.5(3252 eCOOHCHOOHHC

The above reaction increases the concentration of electrons, which decrease the

surface band bending, resulting in drop in resistance. Hence, a decrease in resistance

is observed upon exposure to ethanol. With increasing concentration of ethanol, a

further drop in resistance is expected, which is confirmed by the dynamic response

(Fig. 5-7a).

The sensitivity of 400ºC annealed WO3 film at 150ºC is higher to H2 than

ethanol. It is due to the porous structure of the film as well as small size of H2

molecule compared to ethanol. Hydrogen can easily diffuse through the surface and

interact with more surface area, hence, higher sensitivity is achieved.

Fig. 5-8 shows the dynamic response and sensitivity of 300ºC annealed Fe-

doped WO3 film exposed to ethanol at an optimum operating temperature of 250ºC.

It can be observed that the film shows an unstable response and a drifting baseline

resistance, due to its amorphous nature [213]. Moreover, the response behaviour is

typical of a p-type material. At 250ºC, the dominant oxygen species on the film

surface is O- [217]. Reaction of ethanol molecules with surface-adsorbed oxygen

produces water molecules with electrons according to the following equation [221]:

)4.5(2352 eOHCHOCHOOHHC ads

The above reaction is expected to result in production of free charges and

consequently decrease the film resistance. However, water molecules further interact

with the oxide surface and neutralize the effect of free charges according to the

following mechanism [221].

124

)5.5(0lattlattlatt2 V(W/Fe)OH2O2(W/Fe)OH

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 500 1000 1500 2000 2500 3000 3500

Res

ista

nce

(M

oh

ms

)

Time (seconds)

Fe-doped WO3 annealed @ 300oC exposed to ethanol

Opearating temperature: 250oC

12

pp

m

30

pp

m

60 p

pm

185

pp

m

(a)

0

0.2

0.4

0.6

0.8

1

0 50 100 150 200

Se

nsi

tiv

ity

(R

/Rai

r)



Operating temperature: 250oC(b)

Figure 5-8: Dynamic response (a) and sensitivity (b) to ethanol measured for a nanostructured Fe-

doped WO3 film annealed at 300ºC for 2 hours in air.

The water molecule reacts with W/Fe and O at the lattice positions, which results in

the creation of a oxygen vacancy along an OH-(W/Fe) complex. These vacancies

neutralize the free charges produced through above reactions (equations 5.4 and 5.5)

and, consequently, increase the surface barrier height, provided that the rate of

vacancy formation is higher than the rate of adsorption on the surface. Below 200ºC,

125

the physisorbed water molecules can easily get desorbed from the surface, but the

temperature of desorption of chemisorbed water molecules is reported to be as high

as 500ºC to 600ºC [222]. This can cause a drift in baseline resistance as well as a

drop in sensitivity with increasing concentration of ethanol, which is observed in Fig.

5-8.

Fig. 5-9 shows the dynamic response and sensitivity of the 400ºC annealed Fe-

doped WO3 film upon exposure to ethanol at an optimum operating temperature of

150ºC. The film exhibits a stable baseline resistance (Fig. 5-9a), due to its crystalline

properties. The sensitivity increased with increasing ethanol concentration, reaching

S=5 for 185 ppm ethanol (Fig. 5-9b). The response and recovery times for 185 ppm

ethanol are 64 s and 108 s, respectively (Table 5-7).

The dominant oxygen species on the film surface at 150ºC is O2-, the conduction

mechanism takes place by the following reaction.

)6.5(3252 eCOOHCHOOHHC

The above reaction increases the concentration of electrons, which decrease the

surface band bending, resulting in drop in resistance. Hence, a decrease in resistance

is observed upon exposure to ethanol. With increasing concentration of ethanol, a

further drop in resistance is expected, which is confirmed by the dynamic response

(Fig. 5-9a).

126

0

0.4

0.8

1.2

1.6

2

2.4

0 500 1000 1500 2000 2500 3000

Res

ista

nc

e (

M o

hm

s)

Time (seconds)



12

pp

m

30

pp

m

60 p

pm

185

pp

m

(a)

0

1

2

3

4

5

6

7

0 50 100 150 200

Se

ns

itiv

ity

(R

/Rg

as)



Operating temperature: 150oC(b)


doped WO3 film annealed at 400ºC for 2 hours in air.

Table 5-7: Response and Recovery times of annealed (at 300ºC and 400ºC) nanostructured Fe-doped WO3 films upon exposure to ethanol.

Ethanol Concentration

(ppm)








(seconds) 12 68 108 48 52 30 40 100 60 56 60 32 96 76 84

185 32 116 64 108

127

The sensitivities of the 400ºC annealed WO3 and Fe-doped WO3 films to ethanol

at the operating temperature of 150ºC are compared in Fig. 5-10. The sensitivity of

the 400ºC annealed Fe-doped WO3 film is higher than the 400ºC annealed WO3 film

at the operating temperature of 150ºC. At this temperature, both the films show the

same magnitude of baseline resistance. However, the 400ºC annealed Fe-doped WO3

film contains more distortions, consequently contains higher number of oxygen

vacancies, thus enabling higher sensitivity at the same operating temperature of

150ºC.

0

1

2

3

4

5

6

0 50 100 150 200



Sen

siti

vit

y (

R/R

gas

)


Target gas: ethanol


Figure 5-10: Comparison of ethanol sensitivities of nanostructured WO3 and Fe-doped WO3 films

annealed at 400ºC for 2 hours in air at an operating temperature of 150ºC.

An optimum response to ethanol was obtained only by the Fe-implanted WO3

film annealed at 300ºC at an operating temperature of 250ºC. Fig. 5-11 shows the

dynamic response and sensitivity of the 300ºC annealed Fe-implanted WO3 film

upon exposure to ethanol. The response and recovery times are shown in Table 5-8.

The dynamic response is characterized by a slightly drifting baseline resistance and

noise. This is due to the amorphous nature of the film [213]. A maximum sensitivity

128

S=0.65 is obtained to 60 ppm ethanol at 250ºC. At this temperature, reaction of

ethanol molecules with surface adsorbed oxygen produces water molecules with

electrons according to equation (5.4) [221], resulting in production of free charges

and consequently decrease the film resistance. However, water molecules further

interact with the oxide surface and neutralize the effect of free charges according to

equation 5.5 [221].

16

24

32

40

48

56

64

72

80

0 500 1000 1500 2000 2500 3000

Fe-Implanted WO3 annealed @ 300oC exposed to

Re

sis

tan

ce (

K O

hm

s)

Time (seconds)

12

pp

m

30 p

pm

60 p

pm

185

pp

m


ethanol

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

0 50 100 150 200

Fe-Implanted WO3 annealed @ 300oC exposed to

Se

nsi

tiv

ity

(

R/R

air)



ethanol


implanted WO3 film annealed at 400ºC for 2 hours in air.

129

Table 5-8: Response and Recovery times of 300ºC annealed nanostructured Fe-implanted WO3 film upon exposure to ethanol.

Ethanol Concentration (ppm) Fe-implanted WO3 film annealed at 300ºC (Operating temperature: 250ºC)

Response time (seconds) Recovery time (seconds) 12 92 200

30 80 228 60 52 187 185 48 228

The water molecules react with W/Fe and O at the lattice positions which results in

the creation of a vacancy along an OH-(W/Fe) complex. These vacancies neutralize

the free charges produced through reactions (5.4) and (5.5) and consequently,

increase the surface barrier height, if the rate of vacancy formation is higher than rate

of adsorption on the surface. Between 100ºC-200ºC, the physisorbed water

molecules can easily get desorbed from the surface, but the temperature of desorption

of chemisorbed water molecules is reported to be as high as 500ºC to 600ºC [222].

This can cause a drift in baseline resistance with increasing concentration of ethanol,

which is observed in Fig. 5-11. The higher recovery times are associated with the

accumulation of water molecules on the film surface which prolong the recovery of

the film surface upon removal of ethanol from the chamber.

5.4 Response to Carbon monoxide (CO)

Various WO3 thin films were exposed to carbon monoxide in the temperature

range 100ºC-300ºC at increments of 50ºC to determine the optimum operating

temperature of these sensors. Table 5-9 shows the optimum operating temperature

obtained for various sensors upon exposure to CO.

130

Table 5-9: Optimum operating temperature of nanostructured WO3 thin film based gas sensors upon exposure to CO.

Optimum operating temperature (ºC) for CO as-deposited WO3 - WO3 annealed at 300ºC - WO3 annealed at 400ºC - as-deposited Fe-doped WO3 - Fe-doped WO3 annealed at 300ºC 150 Fe-doped WO3 annealed at 400ºC 150 as-deposited Fe-implanted WO3 - Fe-implanted WO3 annealed at 300ºC - Fe-implanted WO3 annealed at 400ºC -

It is shown in Table 5-9 that sensitivity to CO is achieved only by the Fe-doped

WO3 thin film sensor. It is widely accepted that pure WO3 thin film sensors are not

sensitive to CO. There is very little evidence of CO sensing performance of pure

WO3 thin films available in literature [158, 223]. CO-sensing characteristics of a

CoOOH-WO3 composite doped with Au and SWCNT was investigated by Wu et. al

[130]. In the present study, after doping with Fe, a response to CO has been

observed. Moreover, no response to CO was detected by the Fe-implanted WO3

films.

Fig. 5-12 shows the dynamic response of the annealed Fe-doped WO3 films

upon exposure to CO. The sensitivity of the films is shown in Fig. 5-13. The

dynamic response of the 300ºC annealed Fe-doped WO3 film shows a response

characterized with a noise and a drifting baseline (Fig. 5-12a) at an operating

temperature of 150ºC, which is due to the amorphous nature of the film [213]. The

400ºC annealed film shows a stable response at the same operating temperature. Its

sensitivity is also much higher (S=0.2) than 300ºC annealed Fe-doped WO3 film

(S=0.05) to 1000 ppm CO at an operating temperature of 150ºC (Fig. 5-13). The film

is also characterized by short response time of 64 s compared with 108 s for the

300ºC annealed film (Table 5-10). The crystalline nature, loosely packed film

structure and small grain size of the 400ºC annealed Fe-doped WO3 film contribute

131

to its improved sensing performance to CO. However, as evident from Table 5-10

that there are some variations in the response kinetics with increasing CO

concentration. This is attributed to delay in recovery of high amount of CO from the

film surface at operating temperature of 150ºC.

1.5

1.6

1.7

1.8

1.9

2

2.1

2.2

2.3

0 500 1000 1500 2000

Re

sis

tan

ce (

M o

hm

s)

Time (seconds)

100

pp

m

500

pp

m

250

pp

m

100

0 p

pm

Fe-doped WO3 annealed @ 400oC exposed to CO

Operating temperature: 150 oC (a)

4.8

4.9

5

5.1

5.2

5.3

5.4

5.5

5.6

0 500 1000 1500 2000

Re

sis

tan

ce

(M o

hm

s)

Time (seconds)

100

pp

m

250

pp

m

500

pp

m

100

0 p

pm

Fe-doped WO3 annealed @ 300oC exposed to CO


Figure 5-12: Dynamic responses of nanostructured Fe-doped WO3 films to CO, (a) annealed at 300ºC

for 2 hours in air and (b) annealed at 400ºC for 2 hours in air.

132

0

0.05

0.1

0.15

0.2

0.25

0 200 400 600 800 1000 1200



Se

nsi

tiv

ity

(R

/Rg

as)

CO concentration (ppm)


Figure 5-13: Sensitivity of annealed (at 300ºC and 400ºC) nanostructured Fe-doped WO3 films upon

exposure to CO.

As mentioned earlier, the most dominant species on the film surface at an

operating temperature of 150ºC is O2-. The conduction mechanism is governed by the

following equation [224].

)6.5(22 22 eCOOCO

Upon exposure to CO, carbon dioxide is formed with the production of free charge

carriers, which decreases the surface barrier height and causes a drop in resistance of

the film.

Table 5-10: Response and Recovery times of annealed (at 300ºC and 400ºC) nanostructured Fe-doped WO3 films upon exposure to CO.

CO Concentration (ppm)








(seconds) 100 64 148 140 88 250 64 140 44 36 500 68 104 48 48 1000 108 76 64 80

133

The response of rather non-response WO3 films towards CO is attributed to the

acceptor nature of Fe in the host matrix [219], which makes it easier for

measurement of the sensor signal specially at low temperatures. Hence, WO3 film,

which otherwise is not sensitive to CO, shows a response at low temperature upon

doping the film with Fe.

5.5 Summary and outlook

The optimum operating temperatures obtained for various films upon exposure

to the H2, ethanol and CO in the temperature range 100ºC-300ºC are shown in Table

5-11. Gas sensors based on as-deposited WO3 thin film did not show any response to

H2, ethanol and CO due to their amorphous nature.

Table 5-11: Optimum operating temperature of the films upon exposure to H2, ethanol and CO in the tested temperature range of 100ºC-300ºC.

Optimum temperature (ºC) H2 ethanol CO

as-deposited WO3 - - - as-deposited Fe-doped WO3 - - - as-deposited Fe-implanted WO3 - - - WO3 annealed at 300ºC 280 Fe-doped WO3 annealed at 300ºC 250 150 Fe-implanted WO3 annealed at 300ºC 200 250 WO3 annealed at 400ºC 100 150 Fe-doped WO3 annealed at 400ºC 200 150 150 Fe-implanted WO3 annealed at 400ºC - - -

The microstructural properties and gas sensing characteristics of various WO3

thin films are compared in Table 5-12. All the films have a very small grain size

ranging between 5-15 nm. No crystallinity is observed in any of the films after

annealing at 300ºC for 2 hours in air and their gas sensing response was poor.

However, the pure and Fe-doped WO3 films annealed at 400ºC exhibit high

crystallinity and stable sensor response. The Fe-implanted WO3 film annealed at

400ºC shows amorphous nature and a similar gas sensing characteristic to that of the

134

as-deposited WO3 film. This indicates that annealing the WO3 and Fe-doped WO3

films at 400ºC for 2 hours in air improved the crystalline properties without

significantly altering the grain size and achieved excellent sensor response as

compared with other films.

135

Table 5-12: Comparison of microstructural properties of 400ºC annealed films.

AFM XRD and TEM RBS XPS Baseline resistanc

e at 150ºC

(k)

Sensitivity (S) at optimum operating temperature (ºC)

Mean grain size (nm)

Crystalline Properties

Fe (at.%)

W 4f7/2

peak position

(eV)

S Target

Gas Concentratio

n (ppm) Temperatur

e (ºC)

as-deposited WO3 13 Amorphous - 35.74 15.6 - - - - as-deposited Fe-doped WO3

15 Amorphous 0.5 35.70 746 - - - -

as-deposited Fe-implanted WO3

8 Amorphous 5.5 35.83 163 -

WO3 annealed at 300ºC 10 Amorphous 35.71 128 0.3 H2 10,000 280

Fe-doped WO3 annealed at 300ºC

10 Amorphous 0.5 35.80 165 0.25 H2 10,000 250 0.25 ethanol 185 250 0.05 CO 1000 150

Fe-implanted WO3 annealed at 300ºC

8 Amorphous 5.5 35.58 43 0.12 H2 10,000 200

0.65 ethanol 185 250

WO3 annealed at 400ºC 5 Crystalline - 35.56 172 10 H2 10,000 150

0.22 ethanol 185 150

Fe-doped WO3 annealed at 400ºC

10 Crystalline 0.5 35.63 729 5.5 H2 10,000 200 5 ethanol 185 150

0.2 CO 1000 150 Fe-implanted WO3 Imp annealed at 400ºC

5 Amorphous 5.5 35.70 37 0.05 H2 10,000 100

136

However, the Fe-implanted film shows poor gas sensing response as it was highly

damaged and annealing at 300ºC or 400ºC is not sufficient to crystallize this film.

The W 4f 7/2 binding energies of 400ºC annealed films has the maximum downshift

as compared to as-deposited and 300ºC annealed WO3 films, indicating high number

of oxygen vacancies in this film. However, no significant change is observed in W

4f7/2 binding energy of Fe-implanted film, thus, its stoichiometry is essentially

similar to that of pure WO3.This indicates that 400ºC annealed WO3 and Fe-doped

WO3 films have higher number of oxygen vacancies as compared to as-deposited

WO3 film and showed better sensing performance towards the tested gases (H2,

ethanol and CO).

Highest sensitivity to H2 is demonstrated by the 400ºC annealed pure WO3 film

at an operating temperature of 150ºC, which is due to the very small grain size (5

nm) and high porosity of the film, combined with high number of oxygen vacancies.

The highest sensitivity to ethanol is shown by the 400ºC annealed Fe-doped WO3

film at operating temperature of 150oC. Response to CO is also observed by the

400ºC annealed Fe-doped WO3 thin film at an operating temperature of 150ºC.

137

CHAPTER 6: CONCLUSIONS AND FUTURE WORK

This thesis presents the evolution of the author’s PhD research. The research

program commenced with the aim of optimizing the physical, chemical and

electronic properties of WO3 thin film for improved gas sensing performance. The

specific aim of this project was to optimize the operating temperature of the sensor

device and improve its sensitivity towards H2, ethanol and CO. Although WO3 thin

film gas sensors have shown excellent performance in detecting various gases, the

operating temperature of these films is still very high (300ºC-500ºC) and thus low

operating temperature is desirable. In addition, very little evidence is available in

literature on the CO sensing performance of WO3 thin film gas sensors which is

worthy of investigation through modification of the pure film by doping.

Thermal evaporation technique was used to develop pure nanostructured WO3

thin films. Iron doping of the thin films was performed through two methods: by co-

evaporation during thermal evaporation and by ion implantation. The films were

annealed at 300ºC and 400ºC in air for 2 hours to improve their film properties. The

properties of the films were characterized using AFM, TEM, XRD, RBS, XPS and

Raman to understand and evaluate their suitability for gas sensing. A number of

factors such as the target gas, operating temperature, crystallinity, stoichiometry and

presence of oxygen vacancies on the film surface influenced the sensing performance

The developed sensors were tested towards various gas concentrations within the

TLV range of H2 (600-10,000 ppm), ethanol (12-185 ppm) and CO (50-1000 ppm) in

the temperature range of 100ºC-300ºC and relative humidity of 0%. The gas sensing

properties, namely, sensitivity, response and recovery times, and baseline resistance

were evaluated. The results were thoroughly discussed by understanding the gas

138

sensing mechanism of the films at various temperatures. The sensing mechanism is

largely dictated by the target gas and the dominant oxygen species at the specific

operating temperature, which can lead to an opposite sensor response, as observed in

the present study. The concluding major findings of this research and potential for

future work are summarised in the following sections.

6.1 Conclusions

The major findings of this research program are summarized below:

Highly amorphous nanostructured WO3, Fe-doped WO3 and Fe-implanted

WO3 thin films (300 nm thick) with a grain size less than 15 nm have been

synthesized using thermal evaporation technique. To the best of author’s

knowledge, this was the first attempt to dope WO3 film with Fe using thermal

evaporation technique.

The as-deposited WO3 films showed a highly unstable response towards H2

and ethanol, owing to their amorphous nature. These films did not show any

response towards CO.

Doping with small amount (0.5 at.%) of iron increased the film resistance and

considerably improved the sensing performance. The film characterization

revealed that Fe was incorporated as a substitutional impurity in the WO3

matrix, rather than as a catalyst on the film surface.

Upon implantation with 5.5 at.% of Fe, the film became highly amorphous,

however, no additional compounds were revealed from characterization,

indicating that implantation did not induce any chemical changes in the film,

however, the morphology and grain structure were highly distorted.

139

Annealing at 300ºC for 2 hours in air showed an onset of the crystalline

properties of pure and Fe-doped WO3 films and induced sub-stoichiometry in

these films. These films showed a response towards H2, ethanol and CO,

however, the response was characterized by noise and a drifting baseline.

Annealing at 400ºC for 2 hours significantly improved the crystalline

properties and altered the stoichiometry in the WO3 and Fe-doped WO3 films,

which increased the number of oxygen vacancies in the films. An increase in

number of oxygen vacancies is considered to be highly beneficial for gas

sensing.

The nanostructured WO3 film annealed at 400ºC showed maximum response

to H2 and ethanol at an optimum operating temperature of 150ºC and no

response to CO.

The Fe-doped WO3 film annealed at 400ºC showed maximum response to

H2 at an optimum operating temperature of 200ºC. A response to both ethanol

and CO was observed at an optimum operating temperature of 150ºC.

The Fe-implanted WO3 film annealed at 400ºC showed maximum response to

H2 at 100ºC, but the sensitivity is very low compared to other films and

sensor was characterized by a drifting baseline. This film did not show any

response to ethanol and CO in the temperature range 100ºC-300ºC.

Upon comparison the sensing performance of all the films towards various

gases, firstly, it is observed that the author has been able to achieve a lower

operating temperature of 150ºC towards various gases by depositing

nanostructured thin films and doping of the films with Fe and subsequent heat

treatment. Secondly, doping the WO3 film with Fe has demonstrated a

140

response towards CO, which otherwise is non-sensitive to CO as per

available literature reports and author’s knowledge.

6.2 Recommendations for Future Work

This thesis has presented advances in the field of nanostructured WO3 thin film

based gas sensors. Throughout the course of this research, several areas of interest,

which have tremendous research potential, have been identified. In this section, some

proposals will be made for possible future development of the current research.

These proposals for future work are listed below:

In the present study, the films were annealed at 300ºC and 400ºC for 2 hours

in air. There was no increase in grain size after annealing the film at 400ºC,

however, the films were completely crystalline. This project can be extended

to investigate the effect of annealing temperature between 300-600ºC on

change in grain size and gas sensing behaviour. This will allow to achieve an

optimum annealing temperature of the WO3 based films for improved gas

sensing properties.

The Fe-doped WO3 thin film gas sensor has demonstrated to be a potential

candidate for low temperature CO gas sensing. However, in the present study,

only a fixed Fe concentration (0.5 at.%) of co-evaporated tungsten oxide

films have been investigated. The study can be extended to vary the iron

concentrations (0.5 at.% - 10 at.%) and systematically investigate the

optimum effect of Fe doping on the sensing performance and operating

temperature of the tungsten oxide thin films towards CO.

141

BIBLIOGRAPHY

1. Tuller, H.L. and Micak, R., "Advanced sensor technology based on oxide thin film MEMS integration", Journal of Electroceramics, 2000. 4: p. 415-425.

2. Moseley, P.T. and Tofield, B.C., "Solid State Gas Sensor", ed. A. Hilger. 1987, Bristol and Philadelphia.

3. Huang, X.J. and Choi, Y.K., "Chemical sensors based on nanotructured materials", Sensors and Actuators B: Chemical, 2007. 122: p. 659-671.

4. Sinclair, I.R., "Sensors and Transducers". 2001, Oxford, UK: Newnes. 5. Zampolli, S., Elmi, I., Ahmed, F., Passini, M., Cardinali, G.C., Nicoletti, S.,

and Dori, L., "An electronic nose based on solid state sensor arrays for low-cost indoor air quality monitoring applications", Sensors and Actuators B: Chemical, 2004. 101(1-2): p. 39-46.

6. Sobanski, T., Szczurek, A., Nitsch, K., Licznerski, B.W., and Radwan, W., "Electronic nose applied to automotive fuel qualification", Sensors and Actuators B: Chemical, 2006. 116(1-2): p. 207-212.

7. Amoore, J.E. and Hautala, E., "Odor as an ald to chemical safety: Odor thresholds compared with threshold limit values and volatilities for 214 industrial chemicals in air and water dilution", Journal of Applied Toxicology, 1983. 3(6): p. 272-290.

8. Kaur, M., Aswal, D.K., and Yakhmi, J.V., "Chemiresistor Gas Sensors: Materials, Mechanisms and Fabrication", in Science and Technology of Chemiresistor Gas Sensors, D.K.Aswal and S.K.Gupta, Editors. 2007, Nova Science Publishers, Inc. . p. 33-94.

9. Barsan, N. and Weimar, U., "Conduction Model of Metal Oxide Gas Sensors", Journal of Electroceramics, 2001. 7(3): p. 143-167.

10. Taguichi, N., Japan Patent 45-38200. 1962. 11. D'Amico, A. and Di Natale, C., "A contribution on some basic definitions of

sensors properties", Sensors Journal, IEEE, 2001. 1(3): p. 183-190. 12. Bochenkov, V.E. and Sergeev, G.B., "Sensitivity, selectivity and stability of

gas sensitive metal oxide nanostructures", in Metal Oxide Nanostructures and Their Applications, U. Ahmad and Y.B. Hanh, Editors. 2010, Americal Scientific Publishers. p. 31-52.

13. Yamazoe, N. and Shimanoe, K., "Overview of Gas Sensor Technology", in Science and Technology of Chemiresistor Gas Sensors, D.K. Aswal and S.K. Gupta, Editors. 2007, Nova Science Publishers, Inc. p. 1-32.

14. LeGore, L.J., Lad, R.J., Moulzolf, S.C., Vetelino, J.F., Frederick, B.G., and Kenik, E.A., "Defects and morphology of tungsten trioxide thin films", Thin Solid Films, 2002. 406(1-2): p. 79-86.

15. Rani, S., Roy, S.C., and Bhatnagar, M.C., "Effect of Fe doping on the gas sensing properties of nano-crystalline SnO2 thin films", Sensors and Actuators B: Chemical, 2007. 122(1): p. 204-210.

16. Jiménez, I., Arbiol, J., Dezanneau, G., Cornet, A., and Morante, J.R., "Crystalline structure, defects and gas sensor response to NO2 and H2S of tungsten trioxide nanopowders", Sensors and Actuators B: Chemical, 2003. 93(1-3): p. 475-485.

142

17. Moulzolf, S.C., Ding, S., and Lad, R.J., "Stoichiometry and microstructure effects on tungsten oxide chemiresistive films", Sensors and Actuators B: Chemical, 2001. 77(1-2): p. 375-382.

18. Ponce, M.A., Aldao, C.M., and Castro, M.S., "Influence of particle size on the conductance of SnO2 thick films", Journal of the European Ceramic Society, 2003. 23(12): p. 2105-2111.

19. Rothschild, A. and Komem, Y., "The effect of grain size on the sensitivity of nanocrystalline metal-oxide gas sensors", Journal of Applied Physics, 2004. 95(11 I): p. 6374-6380.

20. Yoon, D.H. and Choi, C.M., "Microstructure and CO gas sensing properties of porous ZnO produced by starch addition", Sensors and Actuators B: Chemical, 1997. 45(3): p. 251-257.

21. Hulanicki, A., Glab, S., and Ingman, F., "Chemical sensors: definitions and classification", Pure and Applied Chemistry, 1991. 63(9): p. 1247-1250.

22. Szuber, J., "The effect of oxygen adsorption on the electronic properties of the polar GaAs(111) surface after thermal cleaning in ultrahigh vacuum", Thin Solid Films, 1989. 170(2): p. 219-226.

23. Mizsei, J. and Lantto, V., "Simultaneous response of work function and resistivity of some SnO2-based samples to H2 and H2S", Sensors and Actuators B: Chemical, 1991. 4(1-2): p. 163-168.

24. Yamazoe, N., Sakai, G., and Shimanoe, K., "Oxide Semiconductor Gas Sensors", Catalysis Surveys from Asia, 2003. 7(1): p. 63-75.

25. Gavrilyuk, A.I., "Nanosized WO3 thin film as a multifunctional hydrogen material for achieving photolysis in CuCl films via hydrogen photosensitization", Solar Energy Materials and Solar Cells, 2010. 94(3): p. 515-523.

26. Tang, H., Prasad, K., Sanjines, R., Schmid, P.E., and Levy, F., "Electrical and optical properties of TiO2 anatase thin films", Journal of Applied Physics, 1994. 75(4): p. 2042-2047.

27. Roldan, A., Boronat, M., Corma, A., and Illas, F., "Theoretical Confirmation of the Enhanced Facility to Increase Oxygen Vacancy Concentration in TiO2 by Iron Doping", The Journal of Physical Chemistry C, 2010. 114(14): p. 6511-6517.

28. Yoon, D.H., Yu, J.H., and Choi, G.M., "CO gas sensing properties of ZnO-CuO composite", Sensors and Actuators B: Chemical, 1998. 46(1): p. 15-23.

29. Barbi, G.B., Santos, J.P., Serrini, P., Gibson, P.N., Horrillo, M.C., and Manes, L., "Ultrafine grain-size tin-oxide films for carbon monoxide monitoring in urban environments", Sensors and Actuators B: Chemical, 1995. 25(1-3): p. 559-563.

30. Kunt, T.A., McAvoy, T.J., Cavicchi, R.E., and Semancik, S., "Optimization of temperature programmed sensing for gas identification using micro-hotplate sensors", Sensors and Actuators B: Chemical, 1998. 53(1-2): p. 24-43.

31. Costello, B.P.J.d.L., R.J.Ewen, P.R.H.Jones, N.M.Ratcliffe, and R.K.M.Wat, "A study of the catalytic and vapour-sensing properties of zinc oxide and tin dioxide in relation to 1-butanol and dimethyldisulphide", Sensors and Actuators B: Chemical, 1999. 61(1-3): p. 199-207.

32. Haeng, Y.J. and Choi, G.M., "Electrical and CO gas sensing properties of ZnO-SnO2 composites", Sensors and Actuators B: Chemical, 1998. 52(3): p. 251-256.

143

33. Zhu, C.L., Chen, Y.J., Wang, R.X., Wang, L.J., Cao, M.S., and Shi, X.L., "Synthesis and enhanced ethanol sensing properties of [alpha]-Fe2O3/ZnO heteronanostructures", Sensors and Actuators B: Chemical, 2009. 140(1): p. 185-189.

34. de Lacy Costello, B.P.J., Ewen, R.J., Ratcliffe, N.M., and Sivanand, P.S., "Thick film organic vapour sensors based on binary mixtures of metal oxides", Sensors and Actuators B: Chemical, 2003. 92(1-2): p. 159-166.

35. Ruiz, A.M., Cornet, A., Shimanoe, K., Morante, J.R., and Yamazoe, N., "Effects of various metal additives on the gas sensing performances of TiO2 nanocrystals obtained from hydrothermal treatments", Sensors and Actuators B: Chemical, 2005. 108(1-2): p. 34-40.

36. Haridas, D., Sreenivas, K., and Gupta, V., "Improved response characteristics of SnO2 thin film loaded with nanoscale catalysts for LPG detection", Sensors and Actuators B: Chemical, 2008. 133(1): p. 270-275.

37. Shimizu, Y., Matsunaga, N., Hyodo, T., and Egashira, M., "Improvement of SO2 sensing properties of WO3 by noble metal loading", Sensors and Actuators B: Chemical, 2001. 77(1-2): p. 35-40.

38. Wang, D., Ma, Z., Dai, S., Liu, J., Nie, Z., Engelhard, M.H., Huo, Q., Wang, C., and Kou, R., "Low-Temperature Synthesis of Tunable Mesoporous Crystalline Transition Metal Oxides and Applications as Au Catalyst Supports", The Journal of Physical Chemistry C, 2008. 112(35): p. 13499-13509.

39. Lu, Y., Li, J., Han, J., Ng, H.T., Binder, C., Partridge, C., and Meyyappan, M., "Room temperature methane detection using palladium loaded single-walled carbon nanotube sensors", Chemical Physics Letters, 2004. 391(4-6): p. 344-348.

40. Hahn, S., "SnO2 thick film sensors at ultimate limits: Performance at low O2 and H2O concentrations; size reduction by CMOS technology". 2002, Eberhard-Karls-Universitat Tubingen.

41. Morrison, S.R., "Selectivity in semiconductor gas sensors", Sensors and Actuators, 2005. 12(4): p. 425-440.

42. Watson, J., "The stannic oxide gas sensor", Sensor Review, 1994. 14(1): p. 20-23.

43. Fleischer, M., Kornely, S., Weh, T., Frank, J., and Meixner, H., "Selective gas detection with high-temperature operated metal oxides using catalytic filters", Sensors and Actuators B: Chemical, 2000. 69(1-2): p. 205-210.

44. Anders, H., "Electrical properties of nanocrystalline WO3 for gas sensing applications". 2004, Uppsala Universitet, Sweden.

45. Romppainen, P. 1988, University of Oulu, Finland. 46. Esser, P. and Göpel, W., ""Physical" adsorption on single crystal zinc oxide",

Surface Science, 1980. 97(2-3): p. 309-318. 47. Barsan, N., Schweizer-Berberich, M., and Göpel†, W., "Fundamental and

practical aspects in the design of nanoscaled SnO2 gas sensors: a status report", Fresenius' Journal of Analytical Chemistry, 1999. 365(4): p. 287-304.

48. Romppainen, P. and Lantto, V., "The effect of microstructure on the height of potential energy barriers in porous tin dioxide gas sensors", Journal of Applied Physics, 1988. 63(10): p. 5159-5165.

49. Ihokura, H., "SnO2 based inflammable gas sensors". 1983, Kyushu University, 52-57, Japan.

144

50. Williams, D.E., in Solid State Gas Sensors, P.T. Moseley and B.C. Tofield, Editors. 1987, Adam Hilger: Bristol.

51. Wang, X., Yee, S.S., and Carey, W.P., "Transition between neck-controlled and grain-boundary-controlled sensitivity of metal-oxide gas sensors", Sensors and Actuators B: Chemical, 1995. 25(1-3): p. 454-457.

52. Shimizu, Y. and Egashira, M., Basic aspects and challenges of semiconductor gas sensors, in MRS Bulletin. 1999. p. 18-24.

53. Zhang, G. and Liu, M., "Effect of particle size and dopant on properties of SnO2-based gas sensors", Sensors and Actuators, B: Chemical, 2000. 69(1): p. 144-152.

54. Lieber, C.M., "One-dimensional nanostructures: Chemistry, physics & applications", Solid State Communications, 1998. 107(11): p. 607-616.

55. Suryanarayana, C., "The structure and properties of nanocrystalline materials: Issues and concerns", Journal of Metals, 2002. 54(9): p. 24-27.

56. Yamazoe, N., "New approaches for improving semiconductor gas sensors", Sensors and Actuators B: Chemical, 1991. 5(1-4): p. 7-19.

57. Gurlo, A., Ivanovskaya, M., Bârsan, N., Schweizer-Berberich, M., Weimar, U., Göpel, W., and Diéguez, A., "Grain size control in nanocrystalline In2O3 semiconductor gas sensors", Sensors and Actuators B: Chemical, 1997. 44(1-3): p. 327-333.

58. Korotcenkov, G., Brinzari, V., Cerneavschi, A., Ivanov, M., Golovanov, V., Cornet, A., Morante, J., Cabot, A., and Arbiol, J., "The influence of film structure on In2O3 gas response", Thin Solid Films, 2004. 460(1-2): p. 315-323.

59. Kalantar-Zadeh, K. and Fry, B., "Nanotechnology enabled sensors". 2007, New York, USA: Springer, US.

60. Gabrusenoks, J.V., Cikmach, P.D., Lusis, A.R., Kleperis, J.J., and Ramans, G.M., "Electrochromic colour centres in amorphous tungsten trioxide thin films", Solid State Ionics, 1984. 14(1): p. 25-30.

61. Arnoldussen, T.C., "Model for electrochromic tungstic oxide microstructure and degradation", Journal of The Electrochemical Society, 1981. 128(1): p. 117-123.

62. Ramāns, G.M., Gabrusenoks, J.V., and Veispāls, A.A., "Structure of Tungstic Acids and Amorphous and Crystalline WO3 Thin Films", physica status solidi (a), 1982. 74(1): p. K41-K44.

63. Cox, P.A., "Transition Metal Oxides, An introduction to their electronic structure and properties". 1995, Oxford (UK): Clarendon Press.

64. Woodward, P. and Sleight, A., "Ferroelectric tungsten oxide", Journal of Solid State Chemistry, 1997. 131: p. 9-17.

65. Vogt, T., Woodward, P.M., and Hunter, B.A., "The High-Temperature Phases of WO3", Journal of Solid State Chemistry, 1999. 144(1): p. 209-215.

66. Salje, E., "The orthorhombic phase of WO3", Acta Crystallographica B, 1977. 33(2): p. 574-577.

67. Kehl, W.L., Hay, R.G., and Wahl, D., "The structure of tetragonal tungsten trioxide", Journal of Applied Physics, 1952. 23(2): p. 212.

68. Diehl, R., Brandt, G., and Saije, E., "The crystal structure of triclinic WO3", Acta Crystallographica B, 1978. 34(4): p. 1105-1111.

69. Gerand, B., Nowogrocki, G., Guenot, J., and Figlarz, M., "Structural study of a new hexagonal form of tungsten trioxide", Journal of Solid State Chemistry, 1979. 29(3): p. 429-434.

145

70. Salje, E. and Viswanathan, K., "Physical properties and phase transitions in WO3", Acta Crystallographica A, 1975. 31(3): p. 356-359.

71. Migas, D.B., Shaposhnikov, V.L., Rodin, V.N., and Borisenko, V.E., "Tungsten oxides. I. Effects of oxygen vacancies and doping on electronic and optical properties of different phases of WO3", Journal of Applied Physics, 2010. 108(9).

72. Chatten, R., Chadwick, A.V., Rougier, A., and Lindan, J.D.P., "The oxygen vacancy in crystal phases of WO3", Journal of Physical Chemistry B, 2005. 109: p. 3146-3156.

73. Wijs, G.A.d., Boer, P.K.d., Groot, R.A.d., and Kresse, G., "Anomalous behavior of the semiconducting gap in WO3 from first-principles calculations", Physical Review B, 1999. 59(4): p. 2684.

74. Nowotny, J. and Dufour, L., "Surface and near surfaces chemistry of oxides materials". 1988: Elsevier: Amsterdam.

75. Zheng, H., Ou, J.Z., Strano, M.S., Kaner, R.B., Mitchell, A., and Kalantar-zadeh, K., "Nanostructured Tungsten Oxide – Properties, Synthesis, and Applications", Advanced Functional Materials, 2011. 21(12): p. 2175-2196.

76. Huda, M.N., Yan, Y., Moon, C.-Y., Wei, S.-H., and Al-Jassim, M.M., "Density-functional theory study of the effects of atomic impurity on the band edges of monoclinic WO3", Physical Review B, 2008. 77(19): p. 195102.

77. Hwang, D.W., Kim, J., Park, T.J., and Lee, J.S., "Mg-doped WO3 as a novel photocatalyst for visible light-induced water splitting", Catalysis Letters, 2002. 80(1-2): p. 53-57.

78. Erbs, W., DeSilvestro, J., Borgarello, E., and Grazel, M., Journal of Physical Chemistry B, 1984. 88: p. 4001.

79. Maruthamuthu, P. and Ashokkumar, M., "Hydrogen generation using Cu(II)/WO3 and oxalic acid by visible light", International Journal of Hydrogen Energy, 1988. 13(11): p. 677-680.

80. Paluselli, D., Marsen, B., Miller, E.L., and Rocheleau, R.E., "Nitrogen doping of reactively sputtered tungsten oxide films", Electrochemical and Solid-State Letters, 2005. 8(11): p. G301-G303.

81. Carbajal-Franco, G., Tiburcio-Silver, A., DomInguez, J.M., and Sánchez-Juárez, A., "Thin film tin oxide-based propane gas sensors", Thin Solid Films, 2000. 373(1-2): p. 141-144.

82. Esfandyarpour, B., Mohajerzadeh, S., Khodadadi, A.A., and Robertson, M.D., "Ultrahigh-sensitive tin-oxide microsensors for H2S detection", IEEE Sensors Journal, 2004. 4(449-454).

83. Gaidi, M., Chenevier, B., and Labeau, M., "Electrical properties evolution under reducing gaseous mixtures (H2, H2S, CO) of SnO2 thin films doped with Pd/Pt aggregates and used as polluting gas sensors", Sensors and Actuators B: Chemical, 2000. 62(1): p. 43-48.

84. Khanna, A.K. and Rajesh, B.S.S., "Cu-doped SnO2 thin films as hydrogen sulfide gas sensor", Applied Physics Letters, 2003. 82(24): p. 4388-4390.

85. Lancok, J., Santoni, A., Penza, M., Loreti, S., Menicucci, I., Minarini, C., and Jelinek, M., "Tin oxide thin films prepared by laser-assisted metal-organic CVD: Structural and gas sensing properties", Surface and Coatings Technology, 2005. 200(1-4): p. 1057-1060.

146

86. Law, M., Kind, H., Messer, B., Kim, F., and Yang, P., "Photochemical Sensing of NO2 with SnO2 Nanoribbon Nanosensors at Room Temperature", Angewandte Chemie International Edition, 2002. 41(13): p. 2405-2408.

87. Niranjan, R.S., Sainkar, S.R., Vijayamohanan, K., and Mulla, I.S., "Ruthenium: tin oxide thin film as a highly selective hydrocarbon sensor", Sensors and Actuators B: Chemical, 2002. 82(1): p. 82-88.

88. Salehi, A. and Gholizade, M., "Gas-sensing properties of indium-doped SnO2 thin films with variations in indium concentration", Sensors and Actuators B: Chemical, 2003. 89(1-2): p. 173-179.

89. Sharma, R.K., Tang, Z., Chan, P.C.H., Sin, J.K.O., and Hsing, I.M., "Compatibility of CO gas sensitive SnO2/Pt thin film with silicon integrated circuit processing", Sensors and Actuators B: Chemical, 2000. 64(1-3): p. 49-53.

90. Song, S.-K., Cho, J.-S., Choi, W.-K., Jung, H.-J., Choi, D., Lee, J.-Y., Baik, H.-K., and Koh, S.-K., "Structure and gas-sensing characteristics of undoped tin oxide thin films fabricated by ion-assisted deposition", Sensors and Actuators B: Chemical, 1998. 46(1): p. 42-49.

91. Benkstein, K.D. and Semancik, S., "Mesoporous nanoparticle TiO2 thin films for conductometric gas sensing on microhotplate platforms", Sensors and Actuators B: Chemical, 2006. 113(1): p. 445-453.

92. Chaudhari, G.N., Bende, A.M., Bodade, A.B., Patil, S.S., and Sapkal, V.S., "Structural and gas sensing properties of nanocrystalline TiO2:WO3-based hydrogen sensors", Sensors and Actuators B: Chemical, 2006. 115(1): p. 297-302.

93. Hossein-Babaei, F., Keshmiri, M., Kakavand, M., and Troczynski, T., "A resistive gas sensor based on undoped p-type anatase", Sensors and Actuators B: Chemical, 2005. 110(1): p. 28-35.

94. Ruiz, A.M., Cornet, A., and Morante, J.R., "Performances of La-TiO2 nanoparticles as gas sensing material", Sensors and Actuators B: Chemical, 2005. 111-112: p. 7-12.

95. Sharma, R.K., Bhatnagar, M.C., and Sharma, G.L., "Mechanism of highly sensitive and fast response Cr doped TiO2 oxygen gas sensor", Sensors and Actuators B: Chemical, 1997. 45(3): p. 209-215.

96. Zheng, L., Xu, M., and Xu, T., "TiO2-x thin films as oxygen sensor", Sensors and Actuators B: Chemical, 2000. 66(1-3): p. 28-30.

97. Cheng, X.L., Zhao, H., Huo, L.H., Gao, S., and Zhao, J.G., "ZnO nanoparticulate thin film: preparation, characterization and gas-sensing property", Sensors and Actuators B: Chemical, 2004. 102(2): p. 248-252.

98. Gong, H., Hu, J.Q., Wang, J.H., Ong, C.H., and Zhu, F.R., "Nano-crystalline Cu-doped ZnO thin film gas sensor for CO", Sensors and Actuators B: Chemical, 2006. 115(1): p. 247-251.

99. Hazra, S.K. and Basu, S., "Hydrogen sensitivity of ZnO p-n homojunctions", Sensors and Actuators B: Chemical, 2006. 117(1): p. 177-182.

100. Nunes, P., Fortunato, E., Lopes, A., and Martins, R., "Influence of the deposition conditions on the gas sensitivity of zinc oxide thin films deposited by spray pyrolysis", International Journal of Inorganic Materials, 2001. 3(8): p. 1129-1131.

101. Shishiyanu, S.T., Shishiyanu, T.S., and Lupan, O.I., "Sensing characteristics of tin-doped ZnO thin films as NO2 gas sensor", Sensors and Actuators B: Chemical, 2005. 107(1): p. 379-386.

147

102. Wang, C., Chu, X., and Wu, M., "Detection of H2S down to ppb levels at room temperature using sensors based on ZnO nanorods", Sensors and Actuators B: Chemical, 2006. 113(1): p. 320-323.

103. Wang, H.T., Kang, B.S., Ren, F., Tien, L.C., Sadik, P.W., Norton, D.P., Pearton, S.J., and Lin, J., "Hydrogen-selective sensing at room temperature with ZnO nanorods", Applied Physics Letters, 2005. 86(24): p. 243503-243503-3.

104. Comini, E., Faglia, G., Sberveglieri, G., Cantalini, C., Passacantando, M., Santucci, S., Li, Y., Wlodarski, W., and Qu, W., "Carbon monoxide response of molybdenum oxide thin films deposited by different techniques", Sensors and Actuators B: Chemical, 2000. 68(1-3): p. 168-174.

105. Imawan, C., Solzbacher, F., Steffes, H., and Obermeier, E., "Gas-sensing characteristics of modified-MoO3 thin films using Ti-overlayers for NH3 gas sensors", Sensors and Actuators B: Chemical, 2000. 64(1-3): p. 193-197.

106. Mutschall, D., Holzner, K., and Obermeier, E., "Sputtered molybdenum oxide thin films for NH3 detection", Sensors and Actuators B: Chemical, 1996. 36(1-3): p. 320-324.

107. Althainz, P., Schuy, L., Goschnick, J., and Ache, H.J., "The influence of morphology on the response of iron-oxide gas sensors", Sensors and Actuators B: Chemical, 1995. 25(1-3): p. 448-450.

108. Neri, G., Bonavita, A., Galvagno, S., Siciliano, P., and Capone, S., "CO and NO2 sensing properties of doped-Fe2O3 thin films prepared by LPD", Sensors and Actuators B: Chemical, 2002. 82(1): p. 40-47.

109. Neri, G., Bonavita, A., Rizzo, G., Galvagno, S., Capone, S., and Siciliano, P., "Methanol gas-sensing properties of CeO2-Fe2O3 thin films", Sensors and Actuators B: Chemical, 2006. 114(2): p. 687-695.

110. Ryabtsev, S.V., Shaposhnick, A.V., Lukin, A.N., and Domashevskaya, E.P., "Application of semiconductor gas sensors for medical diagnostics", Sensors and Actuators B: Chemical, 1999. 59(1): p. 26-29.

111. Dickey, E.C., Varghese, O.K., Ong, K.G., Gong, D., Paulose, M., and Grimes, C.A., "Room temperature ammonia and humidity sensing using highly ordered nanoporous alumina films", Sensors, 2002. 2: p. 91-110.

112. Nahar, R.K., "Study of the performance degradation of thin film aluminum oxide sensor at high humidity", Sensors and Actuators B: Chemical, 2000. 63(1-2): p. 49-54.

113. Ryzhikov, A., Labeau, M., and Gaskov, A., "Al2O3(M = Pt, Ru) catalytic membranes for selective semiconductor gas sensors", Sensors and Actuators B: Chemical, 2005. 109(1): p. 91-96.

114. Sberveglieri, G., Anchisini, R., Murri, R., Ercoli, C., and Pinto, N., "An Al2O3 sensor for low humidity content: characterization by impedance spectroscopy", Sensors and Actuators B: Chemical, 1996. 32(1): p. 1-5.

115. Du, J., Wu, Y., and Choy, K.-L., "Controlled synthesis of gas sensing Cr2-xTixO3 films by electrostatic spray assisted vapour deposition and their structural characterisation", Thin Solid Films, 2006. 497(1-2): p. 42-47.

116. Neri, G., Bonavita, A., Rizzo, G., and Galvagno, S., "Low temperature sol-gel synthesis and humidity sensing properties of Cr2-xTixO3", Journal of the European Ceramic Society, 2004. 24(6): p. 1435-1438.

117. Xiangfeng, C., Xingqin, L., and Guangyao, M., "Effects of CdO dopant on the gas sensitivity properties of ZnFe2O4 semiconductors", Sensors and Actuators B: Chemical, 2000. 65(1-3): p. 64-67.

148

118. Fleischer, M., Seth, M., Kohl, C.D., and Meixner, H., "A study of surface modification at semiconducting Ga2O3 thin film sensors for enhancement of the sensitivity and selectivity", Sensors and Actuators B: Chemical, 1996. 36(1-3): p. 290-296.

119. Lampe, U., Simon, E., Pohle, R., Fleischer, M., Meixner, H., Frerichs, H.P., Lehmann, M., and Kiss, G., "GasFET for the detection of reducing gases", Sensors and Actuators B: Chemical, 2005. 111-112: p. 106-110.

120. Ogita, M., Higo, K., Nakanishi, Y., and Hatanaka, Y., "Ga2O3 thin film for oxygen sensor at high temperature", Applied Surface Science, 2001. 175-176: p. 721-725.

121. Ogita, M., Yuasa, S., Kobayashi, K., Yamada, Y., Nakanishi, Y., and Hatanaka, Y., "Presumption and improvement for gallium oxide thin film of high temperature oxygen sensors", Applied Surface Science, 2003. 212-213: p. 397-401.

122. Bendahan, M., Boulmani, R., Seguin, J.L., and Aguir, K., "Characterization of ozone sensors based on WO3 reactively sputtered films: influence of O2 concentration in the sputtering gas, and working temperature", Sensors and Actuators B: Chemical, 2004. 100(3): p. 320-324.

123. Bhandarkar, V.B., Katti, V.R., Kaur, M., Gadkari, S.C., and Gupta, S.K., "RF sputtered tungsten oxide films as NH3 sensors", Asian Journal of Physics, 2005. 14(1-2): p. 11-15.

124. Ivanov, P., Hubalek, J., Malysz, K., Prásek, J., Vilanova, X., Llobet, E., and Correig, X., "A route toward more selective and less humidity sensitive screen-printed SnO2 and WO3 gas sensitive layers", Sensors and Actuators B: Chemical, 2004. 100(1-2): p. 221-227.

125. Kim, Y.S., Ha, S.C., Kim, K., Yang, H., Choi, S.Y., Kim, Y.T., Park, J.T., Lee, C.H., Choi, J., Paek, J., and Lee, K., "Room-temperature semiconductor gas sensor based on nonstoichiometric tungsten oxide nanorod film", Applied Physics Letters, 2005. 86(21): p. 1-3.

126. Marquis, B.T. and Vetelino, J.F., "A semiconducting metal oxide sensor array for the detection of NOx and NH3", Sensors and Actuators B: Chemical, 2001. 77(1-2): p. 100-110.

127. Xu, C.N., Miura, N., Ishida, Y., Matsuda, K., and Yamazoe, N., "Selective detection of NH3 over NO in combustion exhausts by using Au and MoO3 doubly promoted WO3 element", Sensors and Actuators B: Chemical, 2000. 65(1-3): p. 163-165.

128. Sivakumar, R., Gopalakrishnan, R., Jayachandran, M., and Sanjeeviraja, C., "Preparation and characterization of electron beam evaporated WO3 thin films", Optical Materials, 2007. 29(6): p. 679-687.

129. Smith, W., Zhang, Z.Y., and Zhao, Y.P., "Structural and optical characterization of WO3 nanorods/films prepared by oblique angle deposition", Journal of Vacuum Science and Technology B, 2007. 25(6): p. 1875.

130. Lietti, L., Alemany, J.L., Forzatti, P., Busca, G., Ramis, G., Giamello, E., and Bregani, F., "Reactivity of V2O5-WO3/TiO2 catalysts in the selective catalytic reduction of nitric oxide by ammonia", Catalysis Today, 1996. 29(1-4): p. 143-148.

131. Granqvist, C.G., "Handbook of Inorganic Electronic Materials". 1995, Amsterdam: Elsevier.

149

132. Tomchenko, A.A., Khatko, V.V., and Emelianov, I.L., "WO3 thick-film gas sensors", Sensors and Actuators B: Chemical, 1998. 46(1): p. 8-14.

133. Penza, M., Martucci, C., and Cassano, G., "NOx gas sensing characteristics of WO3 thin films activated by noble metals (Pd, Pt, Au) layers", Sensors and Actuators B: Chemical, 1998. 50(1): p. 52-59.

134. Mearkawa, T., Tamaki, J., Miura, N., and Yamazoe, N., "Gold-loaded tungsten oxide sensor for detection of ammonia in air", Chemistry Letters, 1992(4): p. 639-642.

135. Sberveglieri, G., Depero, L., Groppelli, S., and Nelli, P., "WO3 sputtered thin films for NOx monitoring", Sensors and Actuators B: Chemical, 1995. 26(1-3): p. 89-92.

136. Llobet, E., Molas, G., Molinas, P., Calderer, J., Vilanova, X., Brezmes, J., Sueiras, J.E., and Correig, X., "Fabrication of highly selective tungsten oxide ammonia sensors", Journal Name: Journal of the Electrochemical Society; Journal Volume: 147; Journal Issue: 2; Other Information: PBD: Feb 2000, 2000: p. Medium: X; Size: page(s) 776-779.

137. Wang, X., Miura, N., and Yamazoe, N., "Study of WO3-based sensing materials for NH3 and NO detection", Sensors and Actuators B: Chemical, 2000. 66: p. 74-76.

138. Wöllenstein, J., Plaza, J.A., Cané, C., Min, Y., Böttner, H., and Tuller, H.L., "A novel single chip thin film metal oxide array", Sensors and Actuators B: Chemical, 2003. 93(1-3): p. 350-355.

139. Ponzoni, A., Comini, E., Ferroni, M., and Sberveglieri, G., "Nanostructured WO3 deposited by modified thermal evaporation for gas-sensing applications", Thin Solid Films, 2005. 490(1): p. 81-85.

140. Siciliano, T., Tepore, A., Micocci, G., Serra, A., Manno, D., and Filippo, E., "WO3 gas sensors prepared by thermal oxidization of tungsten", Sensors and Actuators B: Chemical, 2008. 133(1): p. 321-326.

141. Stankova, M., Vilanova, X., Calderer, J., Llobet, E., Brezmes, J., Gràcia, I., Cané, C., and Correig, X., "Sensitivity and selectivity improvement of rf sputtered WO3 microhotplate gas sensors", Sensors and Actuators B: Chemical, 2006. 113(1): p. 241-248.

142. Barrett, E.P.S., Georgiades, G.C., and Sermon, P.A., "The mechanism of operation of WO3-based H2S sensors", Sensors and Actuators B: Chemical, 1990. 1(1-6): p. 116-120.

143. Smith, D.J., Vatelino, J.F., Falconer, R.S., and Wittman, E.L., "Stability, sensitivity and selectivity of tungsten trioxide films for sensing applications", Sensors and Actuators B: Chemical, 1993. 13(1-3): p. 264-268.

144. Lin, H.-M., Hsu, C.-M., Yang, H.-Y., Lee, P.-Y., and Yang, C.-C., "Nanocrystalline WO3-based H2S sensors", Sensors and Actuators B: Chemical, 1994. 22(1): p. 63-68.

145. Solis, J.L., Saukko, S., Kish, L., Granqvist, C.G., and Lantto, V., "Semiconductor gas sensors based on nanostructured tungsten oxide", Thin Solid Films, 2001. 391(2): p. 255-260.

146. Akiyama, M., Tamaki, J., Miura, N., and Yamazoe, N., "Tungsten Oxide-Based Semiconductor Sensor Highly Sensitive to NO and NO2", Chemistry Letters, 1991. 20(9): p. 1611-1614.

147. Sun, H.-T., Cantalini, C., Lozzi, L., Passacantando, M., Santucci, S., and Pelino, M., "Microstructural effect on NO2 sensitivity of WO3 thin film gas

150

sensors Part 1. Thin film devices, sensors and actuators", Thin Solid Films, 1996. 287(1-2): p. 258-265.

148. Lee, D.-S., Han, S.-D., Huh, J.-S., and Lee, D.-D., "Nitrogen oxides-sensing characteristics of WO3-based nanocrystalline thick film gas sensor", Sensors and Actuators B: Chemical, 1999. 60(1): p. 57-63.

149. Lee, D.-S., Lim, J.-W., Lee, S.-M., Huh, J.-S., and Lee, D.-D., "Fabrication and characterization of micro-gas sensor for nitrogen oxides gas detection", Sensors and Actuators B: Chemical, 2000. 64(1-3): p. 31-36.

150. You, L., Sun, Y.F., Ma, J., Guan, Y., Sun, J.M., Du, Y., and Lu, G.Y., "Highly sensitive NO2 sensor based on square-like tungsten oxide prepared with hydrothermal treatment", Sensors and Actuators B: Chemical, 2011. In Press, Corrected Proof.

151. Xia, H., Wang, Y., Kong, F., Wang, S., Zhu, B., Guo, X., Zhang, J., Wang, Y., and Wu, S., "Au-doped WO3-based sensor for NO2 detection at low operating temperature", Sensors and Actuators B: Chemical, 2008. 134(1): p. 133-139.

152. Zhang, C., Debliquy, M., Boudiba, A., Liao, H., and Coddet, C., "Sensing properties of atmospheric plasma-sprayed WO3 coating for sub-ppm NO2 detection", Sensors and Actuators B: Chemical, 2010. 144(1): p. 280-288.

153. Zhang, C., Boudiba, A., Navio, C., Bittencourt, C., Olivier, M.-G., Snyders, R., and Debliquy, M., "Highly sensitive hydrogen sensors based on co-sputtered platinum-activated tungsten oxide films", International Journal of Hydrogen Energy, 2011. 36(1): p. 1107-1114.

154. Boudiba, A., Zhang, C., Navio, C., Bittencourt, C., Snyders, R., and Debliquy, M., "Preparation of highly selective, sensitive and stable hydrogen sensors based on Pd-doped tungsten trioxide", Procedia Engineering, 2010. 5(0): p. 180-183.

155. Papadimitropoulos, G., Vasilopoulou, M., and Davazoglou, D., "Porous Hot-Wire Metal Oxides Thin Films in Hydrogen Sensing", Procedia Engineering, 2011. 25(0): p. 300-303.

156. Calavia, R., Mozalev, A., Vazquez, R., Gracia, I., CanÃ©, C., Ionescu, R., and Llobet, E., "Fabrication of WO3 nanodot-based microsensors highly sensitive to hydrogen", Sensors and Actuators B: Chemical, 2010. 149(2): p. 352-361.

157. Ahsan, M., Tesfamichael, T., Ionescu, M., Bell, J., and Motta, N., "Low temperature CO sensitive nanostructured WO3 thin films doped with Fe", Sensors and Actuators B: Chemical, 2011. 162(1): p. 14-21.

158. Wu, R.J., Chang, W.C., Tsai, K.M., and Wu, J.G., "The Novel CO sensing material CoOOH-WO3 with Au and SWCNT performance enhancement", Sensors and Actuators B: Chemical, 2009. 138(1): p. 35-41.

159. KÃ¤rkkÃ¤nen, I., Kodu, M., Avarmaa, T., Kozlova, J., Matisen, L., MÃ¤ndar, H., Saar, A., Sammelselg, V.i., and Jaaniso, R., "Sensitivity of CoWO4 thin films to CO", Procedia Engineering, 2010. 5(0): p. 160-163.

160. Kawasaki, H., Namba, J., Iwatsuji, K., Suda, Y., Wada, K., Ebihara, K., and Ohshima, T., "NOx gas sensing properties of tungsten oxide thin films synthesized by pulsed laser deposition method", Applied Surface Science, 2002. 197-198: p. 547-551.

161. Wang, S.-H., Chou, T.-C., and Liu, C.-C., "Nano-crystalline tungsten oxide NO2 sensor", Sensors and Actuators B: Chemical, 2003. 94(3): p. 343-351.

151

162. Tesfamichael, T., Arita, M., Bostrom, T., and Bell, J., "Thin film deposition and characterization of pure and iron-doped electron-beam evaporated tungsten oxide for gas sensors", Thin Solid Films, 2010. 518(17): p. 4791-4797.

163. Ahsan, M., Tesfamichael, T., Ponzoni, A., and Faglia, G., "Sensing Properties of E-Beam Evaporated Nanostructured Pure and Iron-Doped Tungsten Oxide Thin Films", Sensor Letters, 2011. 9(2): p. 759-762.

164. Hoel, A., Reyes, L.F., Heszler, P., Lantto, V., and Granqvist, C.G., "Nanomaterials for environmental applications: novel WO3-based gas sensors made by advanced gas deposition", Current Applied Physics, 2004. 4(5): p. 547-553.

165. Comini, E., Pandolfi, L., Kaciulis, S., Faglia, G., and Sberveglieri, G., "Correlation between atomic composition and gas sensing properties in tungsten-iron oxide thin films", Sensors and Actuators B: Chemical, 2007. 127(1): p. 22-28.

166. Xie, G., Yu, J., Chen, X., and Jiang, Y., "Gas sensing characteristics of WO3 vacuum deposited thin films", Sensors and Actuators B: Chemical, 2007. 123(2): p. 909-914.

167. Rossinyol, E., Prim, A., Pellicer, E., Rodríguez, J., Peiró, F., Cornet, A., Morante, J.R., Tian, B., Bo, T., and Zhao, D., "Mesostructured pure and copper-catalyzed tungsten oxide for NO2 detection", Sensors and Actuators B: Chemical, 2007. 126(1): p. 18-23.

168. Khatko, V., Vallejos, S., Calderer, J., Llobet, E., Vilanova, X., and Correig, X., "Gas sensing properties of WO3 thin films deposited by rf sputtering", Sensors and Actuators B: Chemical, 2007. 126(2): p. 400-405.

169. Vallejos, S., Khatko, V., Calderer, J., Gracia, I., Cané, C., Llobet, E., and Correig, X., "Micro-machined WO3-based sensors selective to oxidizing gases", Sensors and Actuators B: Chemical, 2008. 132(1): p. 209-215.

170. Meng, D., Yamazaki, T., Shen, Y., Liu, Z., and Kikuta, T., "Preparation of WO3 nanoparticles and application to NO2 sensor", Applied Surface Science, 2009. 256(4): p. 1050-1053.

171. Solis, J.L., Saukko, S., Kish, L.B., Granqvist, C.G., and Lantto, V., "Nanocrystalline tungsten oxide thick-films with high sensitivity to H2S at room temperature", Sensors and Actuators B: Chemical, 2001. 77(1-2): p. 316-321.

172. Rout, C.S., Hegde, M., and Rao, C.N.R., "H2S sensors based on tungsten oxide nanostructures", Sensors and Actuators B: Chemical, 2008. 128(2): p. 488-493.

173. Boulmani, R., Bendahan, M., Lambert-Mauriat, C., Gillet, M., and Aguir, K., "Correlation between rf-sputtering parameters and WO3 sensor response towards ozone", Sensors and Actuators B: Chemical, 2007. 125(2): p. 622-627.

174. Shen, Y., Yamazaki, T., Liu, Z., Meng, D., Kikuta, T., and Nakatani, N., "Influence of effective surface area on gas sensing properties of WO3 sputtered thin films", Thin Solid Films, 2009. 517(6): p. 2069-2072.

175. Yaacob, M.H., Breedon, M., Kalantar-zadeh, K., and Wlodarski, W., "Absorption spectral response of nanotextured WO3 thin films with Pt catalyst towards H2", Sensors and Actuators B: Chemical, 2009. 137(1): p. 115-120.

152

176. Luo, S., Fu, G., Chen, H., Liu, Z., and Hong, Q., "Gas-sensing properties and complex impedance analysis of Ce-added WO3 nanoparticles to VOC gases", Solid-State Electronics, 2007. 51(6): p. 913-919.

177. Khatko, V., Gorokh, G., Mozalev, A., Solovei, D., Llobet, E., Vilanova, X., and Correig, X., "Tungsten trioxide sensing layers on highly ordered nanoporous alumina template", Sensors and Actuators, B: Chemical, 2006. 118(1-2): p. 255-262.

178. Labidi, A., Gillet, E., Delamare, R., Maaref, M., and Aguir, K., "Ethanol and ozone sensing characteristics of WO3 based sensors activated by Au and Pd", Sensors and Actuators, B: Chemical, 2006. 120(1): p. 338-345.

179. Labidi, A., Lambert-Mauriat, C., Jacolin, C., Bendahan, M., Maaref, M., and Aguir, K., "dc and ac characterizations of WO3 sensors under ethanol vapors", Sensors and Actuators, B: Chemical, 2006. 119(2): p. 374-379.

180. Sriyudthsak, M. and Supothina, S., "Humidity-insensitive and low oxygen dependence tungsten oxide gas sensors", Sensors and Actuators, B: Chemical, 2006. 113(1): p. 265-271.

181. Lyashkov, A.Y., Tonkoshkur, A.S., and Makarov, V.O., "Gas sensing properties of WO3-based ceramics to ethanol vapors", Sensors and Actuators B: Chemical, 2010. 148(1): p. 1-5.

182. Azad, A.-M. and Hammoud, M., "Fine-tuning of ceramic-based chemical sensors via novel microstructural modification: I: Low level CO sensing by tungsten oxide, WO3", Sensors and Actuators B: Chemical, 2006. 119(2): p. 384-391.

183. Feng, G., Shu, F.W., Chun, F.S., Meng, K.L., Yong, X.Q., Guang, J.Z., Dong, X., and Duo, R.Y., "Synthesis and luminescence properties of SnO2 nanoparticles", Chemical Physics Letters, 2003. 372(3-4): p. 451-454.

184. Messing, G.L., Zhang, S.-C., and Jayanthi, G.V., "Ceramic powder synthesis by spray pyrolysis", Journal of the American Ceramic Society, 1993. 76(11): p. 2707-2726.

185. Dobrowolski, J.A., Ranger, M., and Wilkerson, R.L., "Measure the angular evaporation characteristics of sources", Journal of Vacuum Science and Technology A, 1983. 1: p. 1403.

186. Thornton, J., "Influence of substrate temperature and deposition rate on structure of thick sputtered Cu coatings", J. Vac. Sci. Technol., 1975. 12(4): p. 830.

187. Cantalini, C., Pelino, M., Sun, H.T., Faccio, M., Santucci, S., Lozzi, L., and Passacantando, M., "Cross sensitivity and stability of NO2 sensors from WO3 thin film", Sensors and Actuators B: Chemical, 1996. 35(1-3): p. 112-118.

188. Kim, T.S., Kim, Y.B., Yoo, K.S., Sung, G.S., and Jung, H.J., "Sensing characteristics of dc reactive sputtered WO3 thin films as an NOx gas sensor", Sensors and Actuators B: Chemical, 2000. 62(2): p. 102-108.

189. Lee, D.-S., Nam, K.-H., and Lee, D.-D., "Effect of substrate on NO2-sensing properties of WO3 thin film gas sensors", Thin Solid Films, 2000. 375(1-2): p. 142-146.

190. Penza, M., Tagliente, M.A., Mirenghi, L., Gerardi, C., Martucci, C., and Cassano, G., "Tungsten trioxide (WO3) sputtered thin films for a NOx gas sensor", Sensors and Actuators, B: Chemical, 1998. 50(1): p. 9-18.

191. Sanghera, H.K. and Sullivan, J.L., "Study of low energy high dose nitrogen ion implantation in iron - effect of ion energy and current density", Nuclear

153

Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms, 1999. 152(1): p. 65-79.

192. Narayanan, K.L., Rajaraman, R., Valsakumar, M.C., Vijayakumar, K.P., and Nair, K.G.M., "Irradiation-induced persistent photoconductivity in CdS films prepared by chemical bath deposition", Optical Materials, 2001. 17(1-2): p. 355-358.

193. NDE Associates, I. Detection of hydrogen attack. 2011 [cited 2011]; Available from: http://www.nde.com/hydrogen.htm.

194. Buckley, N.A., Isbister, G.K., Stokes, B., and Juurlink, D.N., "Hyperbaric oxygen for carbon monoxide poisining: a systematic review and critical analysis of the evidence", Toxicology Review, 2005. 24(2): p. 75-92.

195. Ivanov, P., Llobet, E., Vilanova, X., Brezmes, J., Hubalek, J., and Correig, X., "Development of high sensitivity ethanol gas sensors based on Pt-doped SnO2 surfaces", Sensors and Actuators B: Chemical, 2004. 99(2-3): p. 201-206.

196. Ho, J.J., Fang, Y.K., Wu, K.H., Hsieh, W.T., Chen, C.H., Chen, G.S., Ju, M.S., Lin, J.J., and Hwang, S.B., "High sensitivity ethanol gas sensor integrated with a solid-state heater and thermal isolation improvement structure for legal drink-drive limit detecting", Sensors and Actuators, B: Chemical, 1998. B50(3): p. 227-233.

197. Diaz-Reyes, J., J.E., F.-M., Gutierrez-Arias, J.M., Morin-Castillo, M.M., Azucena-Coyotecatl, H., Galvan, M., Rodriguez-Fragoso, P., and Mendez-Lopez, A. "Optical and structural properties of WO3 as a function of the annealing temperature". in 3rd WSEAS International Conference on Sensors and Signals (SENSIG'10). 2010. Portugal: WSEAS Press.

198. Moulzolf, S.C., LeGore, L.J., and Lad, R.J., "Heteroepitaxial growth of tungsten oxide films on sapphire for chemical gas sensors", Thin Solid Films, 2001. 400(1-2): p. 56-63.

199. Cornell, R.M. and Schwertmann, U., "The iron oxides: Structure, properties, reactions, occurences and uses". 2nd ed. 2003: Wiley-VCH, Verlag GmbH & Co.

200. Han, N., Chai, L., Wang, Q., Tian, Y., Deng, P., and Chen, Y., "Evaluating the doping effect of Fe, Ti and Sn on gas sensing property of ZnO", Sensors and Actuators B: Chemical, 2010. 147(2): p. 525-530.

201. Lethy, K.J., Beena, D., Kumar, R.V., Pillai, V.P.M., Ganesan, V., and Sathe, V., "Structural, optical and morphological studies on laser ablated nanostructured WO3 thin films", Applied Surface Science, 2008. 254: p. 2369-2376.

202. Guery, C., Choquet, C., Dujeancourt, F., Tarascon, J.M., and LassÃ¨gues, J.C., "Infrared and X-ray studies of hydrogen intercalation in different tungsten trioxides and tungsten trioxide hydrates", Journal of Solid State Electrochemistry, 1997. 1(3): p. 199-207.

203. Saw, K.G. and Plessis J, D., "The X-ray photoelectron spectroscopy C 1s Diamond Peak of chemical vapor deposition diamond from a sharp interfacial structure", Materials Letters, 2004. 58: p. 1344-1348.

204. Morar, J.F., Himpsel, F.J., Hughes, G., Jordan, J.L., McFeely, F.R., and Hollinger, G., "High resolution photoemission investigation: The oxidation of W", Journal of Vacuum Science and Technology A: Vacuum, Surfaces and Films, 1985. 3(3): p. 1477-1480.

154

205. Yao, J.N., Chen, P., and Fujishima, A., "Electrochromic behavior of electrodeposited tungsten oxide thin films", Journal of Electroanalytical Chemistry, 1996. 406(1-2): p. 223-226.

206. Maffeis, T.G.G., Penny, M.W., Cobley, R.J., Comini, E., Sberveglieri, G., and Wilks, S.P., "XPS characterization of vacuum annealed nanocrystalline WO3 films", e-Journal of Surface Science and Nanotechnology, 2009. 7: p. 319-322.

207. Regragui, M., Addou, M., Outzourhit, A., Bernéde, J.C., El Idrissi, E., Benseddik, E., and Kachouane, A., "Preparation and characterization of pyrolytic spray deposited electrochromic tungsten trioxide films", Thin Solid Films, 2000. 358(1-2): p. 40-45.

208. Hollinger, G. 1976, Thesis, Universite Claude Bernad, Lyon. 209. Santucci, S., Cantalini, C., Crivellari, M., Lozzi, L., Ottaiano, L., and

Passacantando, M., "X-ray photoemission spectroscopy and scanning tunneling spectroscopy study on thermal stability of WO3 thin films", Journal of Vacuum Science and Technology, 2000. 18(4): p. 1077.

210. Granqvist, C.G., "CRC Handbook of Solid State Electrochemistry", ed. P.J. Gellings and H.J.M. Bouwmesster. 1997, Cleveland and Ohio: CRC Press, Inc.

211. Maffeis, T.G.G., Yung, D., LePennec, L., Penny, M.W., Cobley, R.J., Comini, E., Sberveglieri, G., and Wilks, S.P., "STM and XPS characterisation of vacuum annealed nanocrystalline WO3 films", Surface Science, 2007. 601(21): p. 4953-4957.

212. Khyzhun, O.Y., "XPS, XES and XAS studies of the electronic structure of tungsten oxides", Journal of Alloys and Compounds, 2000. 305(1-2): p. 1-6.

213. Sanders, B.W., Yao, J., and Post, M.L., "Thin Films of SrFeO2.5+X - Effect of preferred orneitation on oxygen uptake", MRS Proceedings, 1994. 343: p. 463.

214. Polleux, J., Gurlo, A., Barsan, N., Weimar, U., Antonietti, M., and Niederberger, M., "Template-free synthesis and assembly of single-crystalline tungsten oxide nanowires and their gas-sensing properties", Angewandte Chemie-International Edition, 2006. 45(2): p. 261-265.

215. Su, P.G. and Pan, T.T., "Fabrication of a room-temperature NO2 gas sensor based on WO3 films and WO3/MWCNT nanocomposite films by combining polyol process with metal organic decomposition method", Materials Chemistry and Physics, 2011. 125: p. 351-357.

216. Simion, C.E. and Tomescu-Stanoiu, A. "Differences in the gas sensing properties readout with N and P-type MOX materials". in International Semiconductor Conferences. 2011.

217. Gopel, W. and Schierbaum, K.D., "SnO2 sensors: current status and future prospects", Sensors and Actuators B, 1995. 26(27): p. 1.

218. Ou, J.Z., Yaacob, M.H., Breedon, M., Zheng, H.D., Campbell, J.L., Latham, K., Plessis, J.d., Wlodarski, W., and Kalantar-zadeh, K., "In situ Raman spectroscopy of H2 interaction with WO3 films", Physical Chemistry Chemical Physics, 2011. 13(16): p. 7330-7339.

219. Chakraborty, S., Sen, A., and Maiti, H.S., "Selective detection of methane and butane by temperature modulation in iron doped tin oxide sensors", Sensors and Actuators B: Chemical, 2006. 115(2): p. 610-613.

220. Galatsis, K., Cukrov, L., Wlodarski, W., McCormick, P., Kalantar-Zadeh, K., Comini, E., and Sberveglieri, G., "p- and n-type Fe-doped SnO2 gas sensors

155

fabricated by the mechanochemical processing technique", Sensors and Actuators B, 2003. 93: p. 562-565.

221. Seiyama, T., "Chemical Sensor Technology". Vol. 1. 1988, Tokyo: Kodansha Ltd./Elsevier.

222. Yamazoe, N., Fuchigami, J., Kishikawa, M., and Seiyama, T., "Interactions of tin oxide surface with O2, H2O AND H2", Surface Science, 1979. 86: p. 335-344.

223. Hübner, M., Simion, C.E., Haensch, A., Barsan, N., and Weimar, U., "CO sensing mechanism with WO3 based gas sensors", Sensors and Actuators B: Chemical, 2010. 151(1): p. 103-106.

224. Lu, G., Ocala, L.E., and Chen, J., "Room-temperature gas sensing based on electron transfer between discrete tin oxide nanocrystals and mutliwalled carbon nanotubes", Advanced Materials, 2009. 21: p. 2487-2491.

Documents

Thermally Evaporated Tungsten Oxide (WO3) Thin Films for ... · Thermally Evaporated Tungsten Oxide (WO3) Thin Films for Gas Sensing Applications Submitted in fulfilment of the requirements