Ph. D. thesis in PDF format ...Physics Department - College of Science - Baghdad University

1. Republic of Iraq
Ministry of Higher Education
& Scientific Research
University of Baghdad
College of Science
Improvement of ZnO and SnO2
Hydrogen Gas Sensors
A thesis
Submitted to the Committee of College of Science,
University of Baghdad
In Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy in
Physics
By
Qahtan Ghatih Hial
B. Sc. 1994
M. Sc. 1997
Supervised By
Dr. Abdulla M. Suhail Dr. Wasan R. Saleh
2011 A D 1432A H
I

2. Supervisor certification
We certify that this thesis was prepared by Mr. Qahtan Ghatih Hial
under our supervision at the Physics Department, College of Science,
University of Baghdad as a partial requirement for the degree of doctor of
philosophy in Physics.
Signature: Suhail Signature: Wasan
Name: Abdulla M. Suhail Name: Wasan R. Saleh
Title: Assist. Professor Title: Assist. Professor
Address: College of Science, Address: College of Science,
University of Baghdad University of Baghdad
Date: November , 2011 Date: November , 2011
In view of the available recommendation, I forward this thesis for
debate by the Examining Committee.
Signature: Raad
Name: Dr. Raad M. S. Al-Haddad
Title: Professor
Address: Collage of Science, University of Baghdad
Date: November 29, 2011
II

3. Examination Committee Certification
We certify that we have read the thesis entitled “Improvement of ZnO
and SnO2 Hydrogen Gas Sensors” as an examining committee, examined the
Student “Qahtan Ghatih Hial” in its contents, and that in our opinion it meets
the standard of a thesis for the degree of Doctor of Philosophy in Phys-
ics/Optoelectronics.
Signature: Raad
Title: Professor
Name: Dr. Raad M. S. Al-Haddad
Date: November 29, 2011
Chairman
Signature:
Title: Professor
Signature: Emad
Title: Professor
Name: Dr. Izzat M. AL-Essa Name: Dr. Emad Kh. Al-Shakarchi
Date: November , 2011 Date: November 27, 2011
Member Member
Signature:M. B. Q. Signature:Bassam
Title: Assist. Professor Title: Assist. Professor
Name: Dr. Mayada Bedry Al-Quzweny Name: Dr. Bassam Ghalib Rasheed
Date: November 28, 2011 Date: November 29, 2011
Member Member
Signature: Suhail Signature:Wasan
Title: Assist. Professor Title: Assist. Professor
Name: Abdulla M. Suhail Name: Wasan R. Saleh
Date: November , 2011 Date: November , 2011
Supervisor Co-Supervisor
Digitally signed
Approved by the Dean of college of Science
Qahta by Qahtan Iliya
Signature:Salih M. Ali DN: cn=Qahtan
Iliya
Title: Professor
Name: Dr. Saleh Mahdi Ali
The Dean of the College of Science
n Iliya Date: 2012.05.19
15:07:50 -07'00'
Date: December 5, 2011
III

4. ABSTRACT
Spray – pyrolyzed palladium – doped metal oxides (zinc oxide
ZnO and tin oxide SnO2) nano films have been prepared on glass sub-
strates and explored as a fast response sensor to hydrogen reducing gas.
Both ZnO and SnO2 sensing films are obtained via chemical spray pyrol-
ysis deposition (SPD) technique at around 450 0C spraying temperature
with atmospheric air as the carrier gas. Zinc chloride ZnCl2 and zinc ace-
tate Zn(CH3COO)2.2H2O starting materials have been exploited in spray-
ing precursor solutions of ZnO thin film whereas, stannous chloride dihy-
drate SnCl2.2H2O is used in obtaining tin oxide SnO2. The SPD technique
has proven its simplicity and reliability in realizing polycrystalline in
nature ZnO films which crystallized along the (002) phase with preferen-
tial orientation along the c – axis of the ZnO hexagonal wurtzite structure
as verified by the XRD structural analysis. The films exhibit high trans-
mission in the visible range of the electromagnetic spectrum with an av-
erage transmittance value of up to 95 %, and present a sharp ultraviolet
cut – off at approximately 380 nm. The transmission but not the estimated
direct band gap Eg increased with decreasing film thickness. Scanning
Electron Microscope SEM and Atomic Force Microscope AFM surface
morphology studies of the ZnO films reveal a uniform distribution of
porous spherical – shaped nanostructure grains of 20 nm diameter. The
electrical characterization of the sprayed thin films shows that they are
highly resistive, but that their properties vary considerably when the
measurements are conducted in vacuum or in air.
For both ZnO and SnO2 metal oxides, the doped sensor exhibit an
increase of the conductance upon exposure to hydrogen gas of various
concentrations and at different operating temperatures, showing excellent
sensitivity.
IV

5. It was found that the sensing mechanism of hydrogen gas in the
present metal oxide sensors is mostly related to the enhancement of ad-
sorption of atmospheric oxygen. The excellent selectivity and the high
sensitivity for hydrogen gas can be achieved by surface promotion of
ZnO/SnO2 metal oxide films. The observed conductance change in Pd –
doped ZnO sensors after exposure to H2 gas (3%) is about two times as
large as that in the undoped ZnO sensors.
The variation of the operating temperature of the film has led to a
significant change in the sensitivity of the sensor with an ideal operating
temperature of about 250 ± 25 0C after which sensor sensitivity decreas-
es. The sensitivity of the ZnO thin films changes linearly with the in-
crease of the gas concentration.
The response – recovery time of Pd:ZnO materials to hydrogen gas
is characterized to be relatively extremely short. ZnO thin films of 20 –
time dipping in palladium chloride solution have the highest sensitivity of
97% and extremely short response time of 3 s, which fit for practice since
it is crucial to get fast and sensitive gas sensor capable of detecting toxic
and flammable gases well below the lower explosion limit (4% by vol-
ume for H2 gas).
For SnO2 sensing elements, the optimum operating temperature is
around 210 0C and 95.744 % sensitivity to 4.5% H2: air mixing ratio.
V

6. Dedicated to
All Those Who Care…
Including…
Her
VI

7. Acknowledgments
It would be impossible to express my thanks on this page to all those who
have supported me, without whose help I could never have come so far. I will attempt,
at least, to satisfy the barest demands of decency by saying a few words here.
Firstly, I would like to thank my advisor Dr. Abdulla M. Suhail for giving me
the opportunity to work on a challenging and interesting project over the past three
years and for his discussions that always challenged me to look at things from a dif-
ferent perspective. I would also like to thank Dr. Wasan R. Al-azawi for her utmost
valuable feedback collaboration on this research and always backing me up.
I am really indebted to the Ministry of Higher Education & Scientific Re-
search, and the Physics Department – College of Science of Baghdad University for
the unceasing generous patronage of the postgraduates.
My sincere appreciation goes to my colleagues at the Electro – optics & Nano-
technology Research Group: Dr. Osama, Dr. Suded, Assistant lecturer Miss Ghaida,
Assistant lecturer Mr. Omar. Also, to my wonderful group postgraduates: Miss Hind,
Miss Hanan, Mrs. Fatin and Mr. Aqil for their continuous encouragement and support.
To the Thin Film Research Group, I would like to express my extreme grati-
tude and indebtedness for collaborating on this research and for all of the material
support provided. Also, the great help of the XRD, AFM at the Material Physics &
Chemistry Research Establishment labs at the Ministry of Science and Technology are
acknowledged. This thesis would not have been possible without their willingness to
work with me.
No gratitude is sufficient to repay the endless love of my parents and family,
who have stood behind me from my first steps through all the moments of skinned
knees and shaken confidence. They read me my first book, and never failed to call
when my studies overwhelmed me. They gave me the ground I could stand on when-
ever the path ahead seemed dim. No son (or brother) could ask for better.
VII

8. Curriculum Vitae
July 1, 1972 ...................................................................................................Born – Iraq.
1994 ..................................................... B.Sc., Physics/Physics – Baghdad University
1997........................................ M.Sc., Physics/Laser Technology – Baghdad University
2002 – 2007.........................................................Assist. Lecturer – Physics Department
2007 – October 31, 2011 .............................. Ph. D. Postgraduate – Physics Department
PUBLICATIONS
Journal Articles
[1] Q. G. Al-zaidi, Abdulla. M. Suhail, Wasan R. Al-azawi, Palladium – doped
ZnO thin film hydrogen gas sensor, Applied Physics Research Vol. 3, No.
1, pp. 89 – 99, (2011).
[2] H. A. Thjeel, A. M. Suhail, A. N. Naji, Q. G. Al-zaidi , G. S. Muhammed,
and F. A. Naum, Fabrication and characteristics of fast photo response high
responsivity ZnO UV detector, Sensors and Actuators A: Physicsl, revised
manuscript submitted for publication.
[3] A. M. Suhail, O. A. Ibrahim, H. I. Murad, A. M. Kadim and Q. G. Al-zaidi,
Enhancement of white light generation from CdSe/ZnS core – shell system
by adding organic pyrene molecules, Journal of Luminescence, revised
manuscript submitted for publication.
:‫إِنِّي رأَيت أَنَّهُ الَ يَكتُب أَحد كتَابا ً فِي يَومه إِالّ قَال فِي غده‬
َِِ َ ِ ِ ِ ٌ َ ْ ُ َ
".‫"لَو غيِّر هَذا لَكانَ أَحْ سن، ولَو زيد هَذا لَكانَ يُستَحسن، ولَو قُدم هَذا لَكان أَفضل، ولَو تُرك هَذا لَكانَ أَجْ مل‬
َ َ َ َ َ َ ْ َ َ َ ِّ َ َ ْ َ َ َ ِ َ َ َ َ َ ُ
.‫وهَذا من أَعظَم العبَر، وهُو دلِيل علَى اِستِيالَء النُقص علَى جُملَة البَشَر‬
ِ ِ َ ِ ِ َ ٌ َ َ ِ ِ ِ ْ ِ َ َ
‫العمـــاد األصفهانـــي‬
Assuming that he’s not dead, Qahtan can best be reached at his “lifetime” email address of:
qahtaniliya@yahoo.co.uk
Mobile No.: +009647702981421
VIII

9. Contents
Page
Abstract ........................................................................................................................ IV
Dedication .................................................................................................................... VI
Acknowledgments...................................................................................................... VII
Curriculum Vitae ...................................................................................................... VIII
List of Tables ............................................................................................................. XII
List of Figures ........................................................................................................... XIII
List of Symbols .......................................................................................................... XX
Chapter 1 Motivation and Project Objectives ................................................................ 1
1.1 Motivation ........................................................................................................... 1
1.2 Gas Sensor Applications ..................................................................................... 4
1.3 Focus of current research .................................................................................... 6
1.4 Thesis Outline...................................................................................................... 6
Chapter 2 Working Principles of Semiconductor Metal Oxide Gas Sensors ................ 8
2.1 Adsorption Mechanisms ...................................................................................... 8
2.2 Non-Stoichiometry in Semiconductors ............................................................. 11
2.3 Gas Sensor Operation: Catalysis and Adsorption ............................................. 13
2.4 Semiconductor Metal Oxide Gas Sensors ......................................................... 21
2.5 Gas Sensor Metrics ............................................................................................ 27
2.5.1 Sensitivity ............................................................................................... 27
2.5.2 Selectivity .............................................................................................. 28
2.5.3 Stability .................................................................................................. 29
2.5.4 Response and Recovery Times ............................................................... 30
2.6 Sensing Mechanism ........................................................................................... 31
2.7 Factors Influencing the Performance ................................................................ 33
2.7.1 Long term effects / Baseline Drift .......................................................... 33
2.7.2 Sensor surface poisoning ........................................................................ 34
2.8 Optimization of Sensor Performance ................................................................ 34
2.8.1 Use of Catalyst ........................................................................................ 34
IX

10. 2.8.1.1 Spill over Mechanism ................................................................ 36
2.8.1.2 Fermi Energy Control ................................................................ 37
2.8.2 Grain size effects .................................................................................... 39
2.8.3 Thickness dependence ............................................................................ 40
2.8.4 Temperature Modulation ........................................................................ 41
2.8.5 Filters for Selectivity............................................................................... 42
2.8.6 AC and DC measurements ...................................................................... 43
2.9 Zinc Oxide ......................................................................................................... 45
2.9.1 Properties of Zinc Oxide ......................................................................... 45
2.9.2 Defects chemistry.................................................................................... 49
2.9.3 Spray pyrolysis deposition technique ..................................................... 55
2.9.3.1 The deposition process and atomization models ........................ 59
2.9.3.2 Deposition parameters ................................................................ 63
I. Substrate temperature............................................................... 63
II. Influence of Precursors ............................................................ 63
III. Spray Rate ................................................................................ 64
IV. Other Parameters ...................................................................... 65
2.9.4 Metal Oxide Gas Sensors ........................................................................ 65
Chapter 3 Experimental Procedure .............................................................................. 78
3.1 Gas Sensor Fabrication ...................................................................................... 78
3.2 Spray pyrolysis experimental set up .................................................................. 80
3.3 Precursor solution .............................................................................................. 81
3.4 The determination of film thickness .................................................................. 83
3.5 Surface modification of ZnO by palladium noble metal ................................... 84
3.6 Al Interdigitated Elecrtodes (IDE) .................................................................... 85
3.7 Gas sensor testing system .................................................................................. 87
3.8 Sensor testing protocol ...................................................................................... 89
3.9 Crystalline structure of the prepared ZnO thin films ........................................ 91
3.10 Thin film surface topography............................................................................ 92
3.11 Optical properties ............................................................................................. 92
3.12 Tin oxide (SnO2) hydrogen gas sensor.............................................................. 94
X

11. Chapter 4 Results and Discussion ................................................................................ 95
4.1 ZnO thin film deposition ................................................................................... 95
4.2 Crystalline structural properties of the ZnO thin film ....................................... 98
4.3 Surface topography and morphology studies ................................................. 100
4.4 Optical Properties ............................................................................................ 105
4.5 Electrical Properties ........................................................................................ 108
4.5.1 Resistance – Temperature Characteristic .............................................. 108
4.5.2 I – V characteristic of the zinc oxide films ........................................... 110
4.5.3 AC Impedance Spectroscopy ................................................................ 112
4.6 Gas Sensing Measurements ............................................................................. 114
4.6.1 Sensing Characteristics of Pure ZnO towards hydrogen gas ................ 114
4.6.2 Sensing characteristics of Pd – doped ZnO towards hydrogen gas ...... 118
4.7 Operation temperature of the sensor ............................................................... 119
4.8 Tin oxide (SnO2) hydrogen gas sensor ............................................................ 122
4.8.1 Crystalline structure and morphology of undoped SnO2 thin film ....... 122
4.8.2 Optical properties of the undoped tin oxide SnO2 thin films ................ 125
4.8.3 Sensing characteristics of pure SnO2 towards hydrogen gas ................ 126
4.8.4 Sensing characteristics of Pd – doped SnO2 towards hydrogen gas ..... 128
4.9 Conclusions and Future work Proposals ......................................................... 132
References ....................................................................................................... 135
XI

12. List of Tables
Table page
1. Table 1.1 ----------------------------------------------------------------------- 5
Examples of application for gas sensors and electronic noses
2. Table 2.1 ----------------------------------------------------------------------- 9
Comparison of physisorption and chemisorption
3. Table 2.2 --------------------------------------------------------------------- 10
Temperature ranges associated with molecular and dissociative oxy-
gen adsorption reactions.
4. Table 2.3 --------------------------------------------------------------------- 46
Zn – O crystal structure data.
5. Table 2.4 --------------------------------------------------------------------- 48
Typical properties of zinc oxide.
6. Table 2.5 --------------------------------------------------------------------- 55
Characteristics of atomizers commonly used for SPD.
7. Table 3.1 --------------------------------------------------------------------- 83
Optimum thermal spray pyrolysis deposition conditions for the prepa-
ration of ZnO thin films
8. Table 4.1 --------------------------------------------------------------------- 96
Spray pyrolysis deposition optimum parameters
9. Table 4.2 -------------------------------------------------------------------- 100
Crystalline structure, Miller indices and d spacings of the as – deposit-
ed ZnO crystal planes
10.Table 4.3 -------------------------------------------------------------------- 100
Crystalline structure, Miller indices and d spacings of the Pd – doped
ZnO crystal planes.
XII

13. LIST OF FIGURES
Figure Page
1. Figure 2.1 -------------------------------------------------------------------- 15
Microstructure and energy band model of a gas sensitive SnO2 thick
film. The potential barriers form as a result of oxygen adsorption.
2. Figure 2.2 -------------------------------------------------------------------- 17
The nature of oxygen species adsorbed on ZnO as reported by several
researchers.
3. Figure 2.3 -------------------------------------------------------------------- 18
The energy barriers in the transformation from reactants (A + B) to
products (C + D). The uncatalyzed reaction is characterized by a large
activation energy (Eg), while the barrier to product formation is low-
ered (Ec) when a catalyst is used.
4. Figure 2.4 -------------------------------------------------------------------- 22
Schematic view of gas sensing reaction in (a) Compact layer and (b)
Porous layer. a: grain boundary model. b: open neck model. c: closed
neck model.
5. Figure 2.5 ------------------------------------------------------------------- 23
Schematic of a compact layer with geometry and energy band repre-
sentation; Z0 is the thickness of the depleted surface layer; Zg is the
thickness of the surface and eVS the band bending. (a) A partly deplet-
ed compact layer (“thicker”) and (b) A completely depleted layer
(“thinner”).
6. Figure 2.6 ------------------------------------------------------------------- 24
Schematic representation of a porous sensing layer with geometry and
energy band for small and large grains. λD Debye length, Xg grain
size.
7. Figure 2.7 -------------------------------------------------------------------- 25
Schematic of a porous layer with geometry and surface energy band
with necks between grains; Zn is the neck diameter; Z0 is the thickness
of the depletion layer and eVS the band bending. (a) a partly depleted
necks and (b) a completely depleted necks.
8. Figure 2.8 -------------------------------------------------------------------- 26
Influence of particle size and contacts on resistances and capacitances
in thin films are shown schematically for a current flow I from left to
right.
XIII

14. 9. Figure 2.9 -------------------------------------------------------------------- 30
Drawing showing how response and recovery times are calculated
from a plot of sensor conductance versus time.
10.Figure 2.10 ------------------------------------------------------------------- 35
Illustration of the catalyst effect. Nano – particles, having higher sur-
face area, act as catalysts. Here, R stands for reducing gas.
11.Figure 2.11 ------------------------------------------------------------------- 36
Mechanism of sensitization by metal or metal oxide additive.
12.Figure 2.12 ------------------------------------------------------------------- 37
Illustration of Spill Over caused by catalyst particles on the surface of
the grain of the polycrystalline particle.
13.Figure 2.13 ------------------------------------------------------------------- 38
An adequate dispersion of the catalysts is required in order to effec-
tively affect the grains of the semi-conducting material to serve the
implied purpose of increase in sensitivity.
14.Figure 2.14 ------------------------------------------------------------------- 39
Schematic models for grain – size effects.
15.Figure 2.15 ------------------------------------------------------------------- 44
Equivalent circuit for the different contributions in a thin film gas sen-
sor; intergranular contact, bulk and electrode contact.
16.Figure 2.16 ----------------------------------------------------------------- 46
T – X diagram for condensed Zn- O system at 0.1 MPa.
17.Figure 2.17 ------------------------------------------------------------------- 47
Many properties of zinc oxide are dependent upon the wurtzite hexag-
onal, close-packed arrangement of the Zn and O atoms, their cohe-
siveness and void space [59].
18.Figure 2.18 ------------------------------------------------------------------- 49
Ellingham diagram of oxides.
19.Figure 2.19 ------------------------------------------------------------------- 50
Various types of point defects in crystalline materials.
20.Figure 2.20 ------------------------------------------------------------------- 60
Schematic diagram of chemical spray pyrolysis unit
21.Figure 2.21 ------------------------------------------------------------------- 62
XIV

15. Spray processes (A, B, C, and D) occurring with increase in substrate
temperature.
22.Figure 3.1 -------------------------------------------------------------------- 78
Schematic of a typical gas sensor structure.
23.Figure 3.2 -------------------------------------------------------------------- 81
Spray pyrolysis experimental set up.
24.Photo plate 3.1 -------------------------------------------------------------- 82
A: experimental set up of the spray pyrolysis deposition SPD. B: Air
atomizer. C: Gemo DT109 temperature controller, and D: Digital bal-
ance with the magnetic stirrer.
25.Figure 3.3 -------------------------------------------------------------------- 85
Vacuum system for the vaporization from resistance – heated sources.
When replacing the transformer and heater with an electron gun, va-
porization by means of an electron beam occurs.
26.Figure 3.4 -------------------------------------------------------------------- 86
A schematic diagram of the IDE masks utilized in this work.
27.Figure 3.5 -------------------------------------------------------------------- 87
Gas sensor testing system.
28.Photo plate 3.2 -------------------------------------------------------------- 88
A photo of the sensor testing system.
29.Figure 3.6 -------------------------------------------------------------------- 91
A schematic diagram of the gas sensor basic measurement electrical
circuit.
30.Photo plate 3.3 -------------------------------------------------------------- 92
LabX XRD – 6000 Shimadzu diffractometer unit.
31.Photo plate 3.4 -------------------------------------------------------------- 93
Ultra 55 SEM unit from ZEISS.
32.Photo plate 3.5 -------------------------------------------------------------- 93
AA3000 Scanning Probe Microscope SPM, tip NSC35/AIBS, from
Angstrom Advance Inc.
33.Photo plate 3.6 -------------------------------------------------------------- 94
Optima sp-3000 plus UV-Vis-NIR spectrophotometer.
34.Figure 4.1 -------------------------------------------------------------------- 96
XV

16. A photo of spray pyrolyzed ZnO thin film on glass samples.
35.Figure 4.2 -------------------------------------------------------------------- 97
Scanning Electron Micrograph photo of spray pyrolyzed ZnO thin
film on glass.
36.Figure 4.3 -------------------------------------------------------------------- 97
Enlarged photos of Al interdigitated electrodes IDE evaporated on
ZnO thin film sample. A: 1 – mm finger spacing IDE on glass, and B:
0.4 – mm finger spacing IDE on silicon.
37.Figure 4.4 -------------------------------------------------------------------- 99
XRD crystal structure of as deposited ZnO thin film prepared from 0.1
M Zinc Chloride aqueous precursor.
38.Figure 4.5 -------------------------------------------------------------------- 99
XRD crystal structure of Pd – doped ZnO thin film prepared from 0.1
M Zinc Chloride aqueous precursor.
39.Figure 4.6 ------------------------------------------------------------------- 101
Scanning Electron Micrograph of ZnO film prepared at a) 400 0C and
the inset b) 200 0C.
40.Figure 4.7 ------------------------------------------------------------------- 102
Scanning Probe Microscope images of zinc oxide thin film spray py-
rolysed on glass substrate at 450 0C spraying temperature with the
precursor of 0.2 M zinc acetate dissolved in 100 mL distilled water.
41.Figure 4.8 ------------------------------------------------------------------- 103
Scanning Probe Microscope images of zinc oxide thin film spray py-
rolysed on glass substrate at 450 0C spraying temperature with the
precursor of 0.2 M zinc acetate dissolved in 100 mL isopropyl alcohol.
42.Figure 4.9 ------------------------------------------------------------------- 104
Granularity cumulation distribution report of ZnO thin film deposited
at 450 0C on glass substrate using 0.2 M zinc acetate in distilled water
precursor solution.
43.Figure 4.10 ------------------------------------------------------------------ 105
Transmission spectra of ZnO thin films of different thicknesses
sprayed on – glass at 400 0C temperature. The precursor was 0.1 M
dissolved in distilled wa-ter except the 189.34 – nm thick sample
which was a 0.2 M zinc acetate dissolved in 3:1 volume ratio isopro-
pyl alcohol and distilled water.
XVI

17. 44.Figure 4.11 ------------------------------------------------------------------ 106
Absorption spectra of ZnO thin films of different thicknesses sprayed
on – glass at 400 0C temperature. The precursor was 0.1 M zinc ace-
tate dissolved in distilled water.
45.Figure 4.12 ------------------------------------------------------------------ 107
Plots of (αhν)2 vs. photon energy hν for ZnO thin films of different
energy gaps and thicknesses.
46.Figure 4.13 ------------------------------------------------------------------ 108
Relationship of energy gap Eg of sprayed ZnO thin films with film
thickness.
47.Figure 4.14 ------------------------------------------------------------------ 109
The variation of resistance of the spray – pyrolyzed deposited zinc ox-
ide film of 668 nm film thickness with temperature.
48.Figure 4.15 ------------------------------------------------------------------ 110
The I – V characteristic in dark and under UV illumination.
49.Figure 4.16 ------------------------------------------------------------------ 111
The effect of vacuum on base line current of a ZnO thin film at 200 0C
and 10 v bias voltage.
50.Figure 4.17 ------------------------------------------------------------------ 112
The I – V characterization of sprayed ZnO film in the temperature
range from RT to 300 0C.
51.Figure 4.18 ------------------------------------------------------------------ 113
The Cole – Cole plot for the impedance spectrum of the films at room
temperature. The inset is the R-C equivalent circuit of the simulation
of the impedance spectrum.
52.Figure 4.19 ------------------------------------------------------------------ 114
Sensing behavior of ZnO thin film at 6 v bias voltage and 210 degrees
temperature to traces of hydrogen reducing gas mixing ratio in air of
3%, 2%, and 1% respectively.
53.Figure 4.20 ------------------------------------------------------------------ 115
The sensitivity dependence of as – deposited ZnO sensor on hydrogen
gas mixing ratio.
54.Figure 4.21 ------------------------------------------------------------------ 116
XVII

18. Transient responses of ZnO thin film (245 nm thick) at 210 0C testing
temperature upon exposure to hydrogen gas of mixing ratios of 1%,
2%, and 3% respectively.
55.Figure 4.22 ------------------------------------------------------------------ 117
Response and recovery time of the sensor as a function of testing gas
mixing ratio at a testing temperature of 210 0C and bias voltage of 6 v.
56.Figure 4.23 ------------------------------------------------------------------ 117
I - V characteristics of undoped ZnO gas sensor to 5%, 3%, and 1%
Hydrogen gas mixture in air and at 200 degrees temperature.
57.Figure 4.24 ------------------------------------------------------------------ 118
The switching behavior of the Pd – sensitized ZnO thin film maximum
con-ductance to hydrogen of 3% H2:air mixing ratio at 200 0C and bias
voltage of 10 v.
58.Figure 4.25 ------------------------------------------------------------------ 119
Effect of the testing temperature on the Pd – sensitized ZnO thin film
maximum conductance to hydrogen of 3% H2:air mixing ratio and bias
voltage of 10 v.
59.Figure 4.26 ------------------------------------------------------------------ 121
The variation of sensitivity with the operating temperature of the Pd –
doped ZnO gas sensor.
60.Figure 4.27 ------------------------------------------------------------------ 122
Transient responses of Pd – sensitized ZnO thin film (668 nm thick) as
exposed to hydrogen gas of mixing ratio of 3% and at three different
testing temperatures of (1) 250, (2) 300, and (3) 350 0C successively.
61.Figure 4.28 ------------------------------------------------------------------ 123
X – ray diffraction (XRD) pattern of SnO2 thin film spray pyrolyzed
on glass substrate at temperature of 450 0C.
62.Figure 4.29 ------------------------------------------------------------------ 124
AFM image of undoped SnO2 thin film deposited at 450 0C on glass
substrate with the precursor being tin dichloride dehydrate dissolved
in isopropyl alcohol.
63.Figure 4.30 ------------------------------------------------------------------ 125
Transmission spectra of undoped SnO2 thin films of different thick-
nesses deposited at 450 0C on glass substrates.
64.Figure 4.31 ------------------------------------------------------------------ 126
XVIII

19. Absorption coefficient versus the photon energy for energy gap esti-
mation of undoped SnO2 thin films of different thicknesses deposited
at 450 0C on glass substrates.
65.Figure 4.32 ------------------------------------------------------------------ 127
Sensitivity behavior of undoped tin oxide SnO2 thin film to different
hydrogen concentrations. The bias voltage was 5.1 v with the tempera-
ture set to 210 0C.
66.Figure 4.33 ------------------------------------------------------------------ 127
Sensitivity versus H2 gas concentration of undoped tin oxide SnO2 thin
film. The bias voltage was 5.1 v with the temperature set to 210 0C.
67.Figure 4.34 ------------------------------------------------------------------ 128
Sensing behavior of Pd – doped SnO2 gas sensor to different H2 : air
mixing ratios. The tests were performed at 210 0C temperature and 10
v bias.
68.Figure 4.35 ------------------------------------------------------------------ 129
Response transient of Pd – doped SnO2 gas sensor to different H2 : air
mixing ratios. The tests were performed at 210 0C temperature and 10
v bias.
69.Figure 4.36 ------------------------------------------------------------------ 130
Sensitivity and Response time as a function of the H2 test gas mixing
ratio. The test was performed at 210 0C and 10 v bias on SnO2 sample
sprayed over the IDE and surface coated with 20 PdCl2 layers sprayed
at 400 0C over the film.
70.Figure 4.37 ------------------------------------------------------------------ 130
Transient responses of SnO2 thin film of 248 nm thick at 150, 175, and
210 0C testing temperature upon exposure to1% H2:air gas mixing ra-
tio.
71.Figure 4.38 ------------------------------------------------------------------ 131
Variation of sensor response current with temperature of Pd - doped
SnO2 thin film exposed to 4.5% hydrogen gas mixing ratio in air and
at 10 v bias voltage.
XIX

20. LIST OF SYMBOLS
ε Static dielectric constant
λ Wavelength nm
μ (electron) mobility cm2.s-1.v-1
ρ Density kg/m3
σ Conductivity Ω.cm
τrec Recovery time s
τres Response time s
Ω Ohm
qVS Surface Potential Barrier eV
AFM Atomic Force Microscope
CVD Chemical Vapor Deposition
CSP Chemical Spray Pyrolysis
DMM Digital Multi Meter
ENC Electro Neutrality Condition
G Conductance (electrical) S
IDE Interdigitated Electrode
kB Boltzmann’s constant J/K
LEL Lower Explosion Limit %
LD Debye length (LD≡λD) nm
Nd Concentration of Donors cm-3
ns Concentration of Electrons cm-3
NO Nitric oxide
N2O Nitrous Oxide
NO2 Nitrogen dioxide
NTCR Negative Temperature Coefficient of Resistance
PID Proportional–Integral–Derivative Controller
ppm Parts Per Million
PTCR Positive Temperature Coefficient of Resistance
R Resistance (electrical) Ω
S Siemens
sccm Standard Cubic Centimeter per Minute
SEM Scanning Electron Microscope
SMO Semiconductor Metal Oxide
t90 Time to accomplish 90% of sensor response change s
TEM Transmission Electron Microscope
VOC Volatile Organic Compound
XRD X-Ray Diffraction
Z0 Depletion Region nm
XX

21. Chapter 1
Motivation and Project Objectives
Introduction
The purpose of this chapter is to provide a general framework and
introduction for the work presented in the current Ph.D. project. This
chapter is divided into four sections, addressing the research motivations,
objectives, focus of current research and thesis outline.
1.1 Motivation
Sensors are devices that produce a measurable change in output in
response to a specified input stimulus [1]. This stimulus can be a physical
stimulus like temperature and pressure or a concentration of a specific
chemical or biochemical material. The output signal is typically an elec-
trical signal proportional to the input variable, which is also called the
measurand. Sensors can be used in all three phases of matter although gas
and liquid sensors are the most common.
The presence of a reducing/oxidizing gas at the surface of certain
metal oxide semiconductors changes its electrical resistance R. It is this
phenomenon that has spurred the use of these materials in the detection of
a gaseous ambient. The theoretical basis for semiconductor gas sensors
arose in 1950, when Carl Wagner proposed a concept to explain the de-
composition of nitrous oxide (N2O) on zinc oxide (ZnO) [2]. He made the
novel assumption that an exchange of electrons was taking place between
the gaseous N2O and the solid ZnO, which possessed a layer of adsorbed
oxygen. A few years later, Brattain et al. found that ambient gas produced
changes in potential between an electrode and a germanium surface [3].
These findings were explained in a theory outlining the existence of do-
1

22. nor and acceptor traps that lead to the generation of a space charge layer
on the surface of the germanium. A working gas sensor was realized in
1962, when Seiyama et al. detailed the use of ZnO thin films in the detec-
tion of such gases as ethanol (C2H6O) and carbon dioxide (CO2) [4]. It
was in that same year that Naoyoshi Taguchi issued a patent for a gas
sensor based on tin oxide (SnO2) [5]. As such, gas sensors based on SnO2
are typically referred to as Taguchi sensors and are commercially availa-
ble through Figaro Engineering Inc. [6].
The Taguchi gas sensor is a partially sintered SnO2 bulk device
whose resistance in air is very high and drops when exposed to reducing
gases such as combustibles (H2, CO, CH4, C3H8) or volatile organic va-
pors and it has enjoyed a substantial popularity because of its ease of
fabrication, low cost, robustness, and their sensitivity to a large range of
reductive and oxidative gases [7]. In addition to research on understand-
ing the fundamentals of the sensing mechanism, the studies on ZnO and
SnO2 sensors have been directed on enhancing the sensor performance
through the addition of noble metals (Pt, Pd, etc.), synthesis of thick and
thin film sensors, and doping with other semiconductors [7-10]. Other
metal oxides such as Fe2O3, TiO2, WO3 and Co3O4 have also been used as
gas sensors. Despite these broad studies in the semiconductor sensor area,
problems such as insufficient gas selectivity, slow response and recovery
times, inability to detect very low gas concentrations, and degradation of
the sensor performance by surface contamination still persist. Thus, there
is a growing need for chemical sensors with novel properties.
The principle mechanism for gas detection in metal oxides in am-
bient air is the ionosorption of oxygen at its surface, which produces a
depletion layer (for n-type semiconductors), and hence reduces conduc-
tivity [11]. Here, ionosorption refers to the process where a species is
2

23. adsorbed and undergoes a delocalized charge transfer with the metal ox-
ide. This can then be used to measure reducing and oxidizing gases, as
they will change the amount of ionosorbed oxygen, and therefore the
conductivity of the metal oxide.
At higher temperatures the adsorption and desorption rates of oxy-
gen are faster, resulting in a greater response (sensitivity) and a lower
response time for the gas sensor. However, the physical properties of the
metal oxides place an upper limit on the temperatures that can be used. If
the temperature is too high, the stability and reliability of the sensors
diminishes because of possible coalescence and structural changes [12].
Furthermore, as temperature increases, the charge – carrier concentration
will increase and the Debye length, LD, will decrease, resulting in less
sensitivity [13]. In most cases, the optimal temperature for metal oxide
gas sensors is between 200 0C and 500 0C [17].
There are two well – known ways for improving the gas sensing
properties of these films. The first is to add noble metals for their catalyt-
ic activity and to dope the film, with many reports showing that it leads to
better sensitivity and stability, e.g. [14, 15]. The second is to reduce grain
size, which has been shown to increase sensitivity [16]. This is because
the depletion layer caused by ionosorption has a greater effect on the
conduction channel of the grain as the grain size decreases. Consequently,
there is great interest in using nanoparticles in gas sensors, since they can
be used to make films with very small grain sizes. These two approaches
have been experimented in the current research to maximize the sensitivi-
ty and enhance the response time of the metal oxide ZnO/SnO2 thin film
– based hydrogen gas sensors. Thus, the surface modification of the ZnO
sensing element with palladium has greatly enhanced the sensor sensitivi-
3

24. ty and response time of minute – grain size ZnO thin films made possible
through using organic solvent other than water in spraying precursor.
1.2 Gas Sensor Applications
Gases are the key measurands in many industrial or domestic activ-
ities. In the last decade the specific demand for gas detection and moni-
toring has emerged particularly as the awareness of the need to protect the
environment has grown. Gas sensors find applications in numerous fields
[17, 18]. Two important groups of applications are the detection of single
gases (as NOx, NH3, O3, CO, CH4, H2, SO2, etc.) and the discrimination
of odours or generally the monitoring of changes in the ambient. Single
gas sensors can, for examples, be used as fire detectors, leakage detectors,
controllers of ventilation in cars and planes, alarm devices warning the
overcoming of threshold concentration values of hazardous gases in the
work places. The detection of volatile organic compounds (VOCs) or
smells generated from food or household products has also become in-
creasingly important in food industry and in indoor air quality, and multi-
sensor systems (often referred to as electronic noses) are the modern gas
sensing devices designed to analyze such complex environmental mix-
tures [19]. In Table 1.1 [17] examples of application for gas sensors and
electronic noses are reported.
Industry currently employs many varieties of gas sensing systems
for monitoring and controlling emissions from their processes. Applica-
tions exist in the steel, aluminum, mineral, automotive, medical, agricul-
tural, aroma and food industries.
Analytical instruments, based on optical spectroscopy and electro-
chemistry, are widely used in the scientific community. These instru-
ments give precise analytical data, however are costly, slow, and cumber-
4

25. some and require highly qualified personnel to operate. Current trends are
to improve low cost, solid state gas sensor performance in order to obtain
high linearity, sensitivity, selectivity and long term stability [19].
Table 1.1: Examples of applications for gas sensors and electronic noses [17].
Applications
Automobiles
 Car ventilation control
 Filter control
 Gasoline vapour detection
 Alcohol breath tests
Safety
 Fire detection
 Leak detection
 Toxic/flammable/explosive gas detectors
 Boiler control
 Personal gas monitor
Indoor air quality
 Air purifiers
 Ventilation control
 Cooking control
Environmental control
 Weather stations
 Pollution monitoring
Food
 Food quality control
 Process control
 Packaging quality control (off-odours)
Industrial production
 Fermentation control
 Process control
Medicine
 Breath analysis
 Disease detection
The only practical way to monitor air quality or provide a mean to
alert a human of potential danger is by direct gas sensing. A gas sensor
can form part of an early warning system, notifying the appropriate au-
thorities or provide the feedback signals to a process control system. To
achieve this, a gas sensor system must be capable of accurate and stable
in-situ real time measurements. Environmental factors such as operating
temperature, vibration, mechanical shock, chemical poisoning, as well as
5

26. various device characteristics (accuracy, resolution, physical size and
cost) must also be taken into consideration.
1.3 Focus of current research
The main focus of the present thesis is on the improvement of semi
conducting metal oxide (SMO) thin film based gas sensors (with special
emphasis on SnO2 and ZnO) and their characterization. An objective
analysis of the various substrates used by several investigators is per-
formed as a part of this work. The gas sensitive zinc oxide and tin oxide
films are deposited by chemical spray pyrolysis deposition technique with
air blast atomization at 400 – 450 0C spraying temperature. The character-
ization of the as deposited film is performed by XRD, absorption, trans-
mission, scanning electron microscope SEM and atomic force microscope
AFM. Too much effort has been spent to maximize the sensitivity S and
reduces the response τres and recovery τrec times of the sensing element
upon exposing to hydrogen reducing gas H2 of various concentrations C
and at different operating temperatures T. The catalytic effect of the pal-
ladium noble metal and grain size effect are exploited to accomplish these
vital objectives. The thesis also describes the development of the gas
sensor test setup which has been used to measure the sensing characteris-
tics of the sensor.
1.4 Thesis Outline
The current thesis is organized into four chapters. The first chapter
is a general introduction where the overview of the field of study and
scope of the work carried out is outlined. The second chapter entitled
“Working principles of semiconductor metal oxide gas sensors” briefly
describes the working principle of the kind of sensors developed and
surveys the various methods used currently to improve the sensor charac-
6

27. teristics. The fabrication and characterization of the sensing element as
well as testing the sensors towards hydrogen reducing gas is dealt with in
the third chapter. Moreover, the development of the gas sensor testing
chamber and the protocol to use it are also detailed in this chapter. The
subsequent chapter (chapter four), a discussion of the experiments carried
out and the results obtained in the development of gas sensors is submit-
ted. At the end of the chapter four, the conclusions and the scope for fu-
ture work are summarized.
7

28. Chapter 2
Working Principles of Semiconductor Metal Oxide Gas Sensors
Background
Background information relevant to gas sensor technology is intro-
duced in this chapter. Adsorption of gases on the oxide surface is dis-
cussed in the first section as it is of fundamental importance in sensors
built using metal oxide materials. Particular emphasis is placed on the
adsorption of oxygen because most sensors operate in air and oxygen is
the dominant adsorbed species in this case. The mechanism where an
oxide transforms gas – surface interactions into a measurable electrical
signal is reviewed with a focus on the effects of particle size on this phe-
nomenon. The current understanding of the gas sensing mechanism and a
brief discussion of theoretical and empirical models proposed for semi-
conductor metal oxide (SMO) gas sensors are discussed. The metrics by
which gas sensor performance is judged are defined in this chapter and an
introduction to SMO gas sensors is presented. Background information is
concluded with a discussion of reported spray pyrolysis deposition tech-
nique for oxide semiconductors.
2.1 Adsorption Mechanisms
Physical adsorption (physisorption) is defined as an adsorption
event where no geometric change occurs to the adsorbed molecule and
van der Waals forces are involved in the bonding between the surface and
adsorbate [11]. Chemical adsorption (chemisorption) is the formation of a
chemical bond between the molecule and the surface during the adsorp-
tion process and requires an activation energy (e.g. ~0.5eV for chemi-
sorption of oxygen on SnO2) [20]. Chemisorption is a much stronger
bond than physisorption and the characteristics of each are summarized in
8

29. Table 2.1. Two types of chemisorption occur on the surface of metal ox-
ides: (1) molecular or associative chemisorption, in which all the atomic
bonds are preserved in the adsorbed molecule; and (2) dissociative chem-
isorption, where bonding within the adsorbed molecule decomposes and
molecular fragments or ions are bound to oxide surface. Molecular chem-
isorption is the most probable type of adsorption for molecules that pos-
sess free electrons or multiple bonds. Gas molecules with single bonds
Table 2.1: Comparison of physisorption and chemisorption [20].
Physisorption Chemisorption
Covalent Bonding
Intermolecular (van der Waals) Force
(adsorbate & surface)
Low Temperature High Temperature
Low Activation Energy High Activation Energy
(<< 0.5 eV) (> 0.5 eV)
Low Enthalpy Change High Enthalpy Change
(ΔH < 20kJ/mol) (50kJ/mol < ΔH < 800kJ/mol)
Reversible Reversible at High Temperature
Adsorbate energy state unaltered Electron density increases at interface
Multilayer formation possible Monolayer surface coverage
tend to react via dissociative chemisorption; however; there is an activa-
tion energy associated with dissociation. The type of chemisorbed oxygen
on the surface of a metal oxide is dependent on the temperature of the
system. Barsan and Weimar compiled results from a survey of the litera-
ture concerning oxygen adsorption on SnO2 and correlated the adsorbed
oxygen species to temperature where techniques such as infrared spec-
troscopy, temperature programmed desorption and electron paramagnetic
resonance were used [21]. Table 2.2 summarizes the temperature ranges
associated with each species of oxygen adsorption.
9

30. Table 2.2: Temperature ranges associated with molecular and dissociative oxygen
adsorption reactions [20].
Temperature Range (°C) Adsorption Reaction(s)
O2 (g) + e− → O− (ads)
2
Room Temp.< T < 175°C
O2 (g) + 2e− → O2− (ads)
2
175°C < T < 500°C O2 (g) + 2e− → 2O− (ads)
O− (ads) + e− → O2− (ads)
T > 500°C
O2 (g) + 4e− → 2O2− (ads)
In the reaction shown in Table 2.2, (g) indicates the gaseous form,
(ads) indicates the molecule or ions that are adsorbed on a surface and e-
is an electron initially in the metal oxide. Oxygen interactions with the
surface of an oxide are of utmost importance in gas sensing. Oxygen is a
strong electron acceptor on the surface of a metal oxide. A majority of
sensors operate in an air ambient; therefore, the concentration of oxygen
on the surface is directly related to the sensor electrical properties. The
conversion to O- or O2- at elevated temperatures are useful in gas sensing
since only a monolayer of oxygen ions are present with these strongly
chemisorbed species [20, 22].
Desorption is the opposite reaction to adsorption where the chemi-
cal bonds are broken, the adsorbed atoms are removed from the surface,
and electrons are injected back into the material. Desorption is achieved
by thermal stimulation up to a specific temperature or by reactions with
other gaseous species. A desorption process that is isothermal occurs
when, for instance, a reducing gas such as carbon monoxide (CO) is in-
troduced into the surrounding atmosphere. Oxygen is consumed in a reac-
tion with the CO to form carbon dioxide (CO2) as written in equation 2.1.
CO(g) + O− (ads) → CO2 (g) + e− (2.1)
10

31. where the extra electron generated is injected back into the metal oxide.
This desorption reaction results in a lower surface coverage of oxygen
adsorbates which influences the electrical properties of the oxide.
2.2 Non – Stoichiometry in Semiconductors
The relevance of non – stoichiometry to the transport properties of
metal oxide semiconductors will now be explored using ZnO as an exam-
ple. It is well – known that ZnO is stoichiometrically deficient in oxygen
traced to either zinc interstitials or oxygen vacancies. To begin with, the
notation of Kröger and Vink [23] must first be introduced.
A defect is characterized by the charge it carries relative to the sur-
rounding crystal lattice [23]. A defect’s superscript denotes this relative
charge, with a dot (˙) being a single positive charge and a prime (′) denot-
ing a negative charge. Neutrals are written with either an x (X) or no su-
perscript. A subscript is used to denote the lattice site of the defect, with i
(i) being used to signify an interstitial atom. Vacancies are represented
⋅⋅
with the letter V, as in VO , which denotes a doubly charged vacancy
occupying an oxygen lattice site. Electrons and holes may be signified by
e and h, respectively.
The equilibrium constant, K(T), for the general chemical reaction
of reactants A, B and products C, D [19]:
aA + bB → cC + dD (2.2)
can be written as:
[C]c [D]d ∆G
K(T) = = e−KT (2.3)
[A]a [B]b
11

32. where ΔG represents the standard change in free energy for the reaction.
An oxygen vacancy, known as a type of Schottky defect, can be generated
in ZnO through the following reaction:
∙∙
1
ZnZn + OO → ZnZn + VO + 2e′ + O2 (g) (2.4)
2
It is conventional practice to denote the left side of (2.4) as “nil”. As seen
in (2.4), a positively charged oxygen vacancy is compensated by the gen-
eration of electrons, thus leading to n-type conductivity of ZnO.
The mass action constant for (2.4) can be written as:
1
∙∙
K R = [VO ](pO2 )2 ⋅ n2 (2.5)
∙∙
where pO2 denotes the partial pressure of oxygen, [VO ] and n represent the
oxygen vacancy and electron concentrations, respectively. To solve for
∙∙
[VO ] or n, one must evoke the electroneutrality condition (ENC), which
states that the concentration of positive defects present in the material
must equal the concentration of negative defects. The ENC for (2.4) is:
∙∙
2[VO ] = 𝑛 (2.6)
∙∙
using (2.6) and solving for [VO ] or n in (2.5) yields:
1
KR 3 1
[VO ] = ( ) (pO2 )−6
∙∙
(2.7)
4
and
1 1 1
n = (2)3 (K R )3 (pO2 )−6 (2.8)
Thus, the logarithmic concentration of oxygen vacancies and electrons in
ZnO, plotted against log (pO2 ), is shown to have what is termed a -1/6
pO2 dependence. The electronic conductivity, 𝜎, is given by:
12

33. 1 1 1
σ = q(2)3 (K R )3 (pO2 )−6 μe (2.9)
where q is the electronic charge and 𝜇 𝑒 is the electronic mobility.
An oxygen deficiency in ZnO may also be realized through the
formation of a Zn interstitial, known as a type of Frenkel defect. This can
be formed through the following reaction:
1
nil = Zn⋅⋅ + 2e′ + O2 (g)
i (2.10)
2
using the ENC for this reaction and substituting it into the proper equa-
tion for the equilibrium constant will also yield a -1/6 pO2 dependence.
2.3 Gas Sensor Operation: Catalysis and Adsorption
The electrical conductivity of a semiconductor is dictated in large
part by the concentration of electrons or holes present in the material. In
certain metal oxide semiconductors, the majority charge carrier concen-
tration changes as a result of an interaction with a gaseous species [4].
The resulting change in conductance may be quite large and provides the
basis for semiconductor gas sensor operation. This behavior is unlike
metals, where the adsorption of a gas may cause small conductance
changes due to a modification of charge carrier mobility [6]. As an exam-
ple, recall the experiments of Wagner on the decomposition of N 2O on
ZnO. If the generation of two electrons proceeds as in (2.4), the decom-
position reaction was proposed as follows [2]:
2e− + N2 O → N2 + O2− (ads. ) (2.11)
O2− (ads. ) + N2 O → N2 + O2 + 2e− (2.12)
The adsorption of oxygen in (2.11) would result in an increase in
ZnO resistivity due to the capture of majority charge carriers. The subse-
13

34. quent reaction between the adsorbed oxygen and N2O in (2.12) acts to
restore the supply of conduction electrons and thus, an increase in con-
ductivity may be observed. It is this simple and reversible change in
charge concentration that drives the use of metal oxide gas sensors.
For a visual perspective, a schematic of an n – type semiconductor
tin dioxide SnO2 thick film with an accompanying band structure model
is shown in figure 2.1 [19]. For conduction to occur, an electron must
pass from one grain to the next. While there exists an ample concentra-
tion of electrons in the bulk of the material, adsorbed oxygen has cap-
tured electrons near the surface of the film. The electrons that bind to the
adsorbed oxygen leave behind positively charged donor ions. An electric
field develops, between these positive donor ions and the negatively
charged adsorbed oxygen ions, which serves to impede the flow of elec-
trons between neighboring grains. The barrier generated by the electric
field has a magnitude of eVS, where e is the electronic charge and VS is
the potential barrier. The magnitude of VS increases as more oxygen ad-
sorbs on the film surface. Utilizing the Boltzmann equation, the concen-
tration of electrons, ns, that possesses ample energy to cross the barrier
and reach a neighboring grain is given by:
eVs
ns = Nd exp (− ) (2.13)
kT
where Nd is the concentration of donors, k is Boltzmann’s constant, and T
is the temperature. Since conductance (or resistivity) is proportional to n s,
an increase in the adsorbed oxygen content will raise eVS and thus, fewer
electrons will cross the potential barrier. This may be empirically moni-
tored as an increase in resistivity. The introduction of a reducing gas will
reverse this effect, lowering the potential barrier and decreasing resistivi-
ty. It is this reducing gas, often termed the analyte, whose presence is of
interest.
14

35. 1 −
Electron – depleted O
region 2 2
O- O-
O- O-
O- O-
-
O O-
Grain Grain (a) In air
O- O-
O- O-
O-
Thermionic
emission
𝑒𝑉 𝑆1
𝑒𝑉 𝑆1 ; 𝐺1 = 𝐺0 exp(− )
𝐾𝑇
EC
EF
CO2
CO
-
O O-
O
O- O-
Grain Grain (b) In the presence
of reducing gas
O- O-
O O
O- O-
CO CO2
CO2
CO
𝑒𝑉 𝑆2
𝑒𝑉 𝑆2 ; 𝐺2 = 𝐺0 exp(− )
𝐾𝑇
𝐺2 𝑒 𝑉 𝑆1 − 𝑉 𝑆2
𝑠𝑒𝑛𝑠𝑖𝑡𝑖𝑣𝑖𝑡𝑦 ⇉ 𝑆 = = exp ( )
𝐺1 𝐾𝑇
Figure 2.1: Microstructure and energy band model of a gas sensitive SnO2 thick film.
The potential barriers form as a result of oxygen adsorption [19].
SnO2
15

36. Using an example of ZnO in the detection of hydrogen, the following
reactions may occur [24]:
2e− + O2 → 2O− (ads. ) (2.14)
2O− (ads. ) + 2H2 → 2H2 O + 2e− (2.15)
The reaction of H2 with adsorbed oxygen on the surface of ZnO
(2.15) will result in a measurable reduction in the resistivity. It should be
noted that adsorbed oxygen may exist in multiple forms. Takata et al.
proposed that oxygen adsorbed on ZnO is transformed with increasing
temperature in the following manner [25]:
O2 → O− → 2O− → 2O2−
2 (2.16)
of these forms, O2 is considered fairly inactive due to its non – dissocia-
tive state. With regards to O− and O−, electron spin resonance (ESR)
2
studies have shown that O− is far more reactive than O− [26]. The nature
2
of adsorbed oxygen on ZnO as reported by several researchers is shown
in figure 2.2.
A catalyst acts to increase the rate at which a chemical reaction ap-
proaches equilibrium, without permanently being altered in the process
[27]. In the reactions detailed in (2.17) – (2.18) and (2.14) – (2.15), ZnO
acts as a heterogeneous catalyst, as it is a phase distinct from the reactants
and products. To illustrate the phenomenon of catalysis, consider the
following reaction:
A+B→ C+D (2.17)
Two possible paths in which this reaction may proceed are shown
in figure 2.3 [27]. In the absence of a catalyst, the reaction of (2.17) is
characterized by a large activation energy, Ea. When a catalyst such as
ZnO or SnO2 is used, the gaseous products adsorb onto the metal oxide
16

37. Figure 2.2: The nature of oxygen species adsorbed on ZnO as reported by several
researchers [25].
surface with an exothermic heat of adsorption ΔH (State I). The reaction
to form adsorbed products then proceeds with a lower activation energy
Ec (State II). It is evident from figure 2.3 that if ΔH is too large, the gase-
ous reactants are strongly adsorbed and Ec may become too large for the
reaction to proceed. An undesirably low activation energy will cause the
reaction to be energetically easier, but will result in fewer products avail-
able for the reaction.
A gas molecule approaching the surface of a solid will be subject
to an attractive potential [28]. This potential is the origin of adsorption
and arises from the multitude of unsatisfied bonds that exist at the surface
of the solid. The adsorbed species is often called the adsorbent and the
solid surface is termed the adsorbate [11]. Physical adsorption, or phy-
sisorption, occurs as a result of electrostatic and van der Waals forces that
exist between the adsorbent and adsorbate. Heats of adsorption for phy-
sisorption tend to be low, with ΔH values typically in the range of 2 – 15
kcal/mole [29]. In the case of chemical adsorption, (termed chemisorp-
17

38. Ea
Energy
A+B
ΔH Ec II
I C+D
Reaction coordinate
Figure 2.3.The energy barriers in the transformation from reactants (A + B) to products
(C + D). The uncatalyzed reaction is characterized by a large activation energy (E a),
while the barrier to product formation is lowered (Ec) when a catalyst is used [27].
tion), the adsorbent forms a chemical bond with the solid surface. Values
for ΔH tend to be higher for chemisorption and are often in the range of
15 – 200 kcal/mole [29]. As chemisorption tends to provide the necessary
catalysis conditions, it is often the adsorption mode of interest when dis-
cussing semiconductor gas sensors.
If the electrical conductivity of a semiconductor is to be used for
gas detection, then changes in conductivity must be proportional to the
concentration of the gaseous analyte. To understand this relationship,
adsorption kinetics must be discussed. The residence time, of an ad-
sorbed atom is given by [28]:
∆H
= exp ( ) (2.18)
T
18

39. where is related to surface vibration time and R is the universal gas
constant. The surface coverage, S, of a gaseous species is dependent on
both and the flux, F, of gas molecules per unit area per second through:
= (2.19)
Typical units for S are molecules per cm2. Relating the gas flux to
the pressure through the kinetic theory of gases will yield:
N ∆H
=( ) exp ( ) (2.20)
√2 T T
where NA is Avogrado’s number, P is the partial pressure of the gas, and
M is the average molar weight of the gaseous species. Experimental
curves of S plotted as a function of P at a given temperature are known as
adsorption isotherms [28]. One particular isotherm derived by Irving
Langmuir is of key interest in the field of semiconductor gas sensors [30].
It is based on two assumptions [27]:
1) Adsorption terminates upon the completion of one monolayer.
2) There exists neither surface heterogeneity nor interaction among
adsorbed species.
While these assumptions are to some degree impractical for real
surfaces, modified isotherms have been developed [29]. Regardless, the
derivations of Langmuir provide a sound qualitative relationship between
surface coverage and gas concentration.
Using assumption 1), any gaseous molecule will reflect off a sur-
face when striking an adsorbed species. Thus, if So denotes a completely
covered surface, then a concentration of S adsorbed molecules will result
in So – S available sites [28]. The fluxes for both reflected molecules, FR,
and adsorbed molecules, FA, are given by:
R =( ) = (1 − ) (2.21)
substitution of FA into (2.19) and subsequent rearrangement will yield:
19

40. a
= = (2.22)
+ +a
The constant a is comprised of the grouping of terms, with the ex-
ception of P, from (2.20). If θ = (S/So), where θ is defined as the degree
of coverage, then (2.22) takes the following form:
b
= (2.23)
1+b
where b = (a/So).
When the degree of surface coverage is proportional to the partial
pressure, changes in the electrical conductivity may be related to gas
concentration. Inspection of (2.23) shows that if bP is small, θ is propor-
tional to P. However, if bP >> 1, then θ approaches unity and the lack of
proportionality makes the gas sensor insensitive to coverage. If a compe-
tition ensues for surface sites between two gas species, A and B, then
(2.23) becomes [27]:
b b
= = (2.24)
1+b +b 1+b +b
If bBPB >> bAPA, then the equations of (2.24) become:
b b
= = (2.25)
1+b 1+b
in the case that bBPB is >> 1, the equations of (2.25) reduce to:
b
1 (2.26)
1+b
Thus, if the conductivity of the gas sensor is strongly dependent on spe-
cies A, then the concentration of either A or B may be measured. Howev-
er, if the conductivity possesses a strong dependence on species B, it will
become independent of the gas concentration as approaches 1.
The rate of a reaction between A and B may be given by [27]:
ae=k (2.27)
20

41. If the rate in (2.27) is much higher than the rate of adsorption of say, A,
then will fall to zero and will increase. As an example, note that the
coverage of oxygen, represented by in (2.26) on an n-type semicon-
ductor is quite low. The coverage of a reducing gas, in (2.26) is quite
high. As the reaction rate between the oxygen and the reducing species
increases, falls to zero, enabling the reducing gas to be detected with a
high degree of sensitivity.
2.4 Semiconductor metal oxide gas sensors
Metal oxide semiconductor gas sensors are, essentially, gas de-
pendent resistors [31]. A broad range of metal oxides are known for their
gas sensing properties, each with a unique sensitivity and selectivity.
Their detection principle is based on a modulation of their electrical con-
duction properties by surface adsorbed gas molecules. The sensitive layer
is deposited onto a substrate with a set of electrodes for measuring re-
sistance changes and heating the sensitive layer; normally 2 – point re-
sistance measurements are accurate enough for gas sensors. The used
metal – oxides are n – or p – type semiconductors, due to the presence of
oxygen – vacancies in the bulk. Generally the conductance or the re-
sistance of the sensor is monitored as a function of the concentration of
the target gases. Additionally the performance of the sensor depends on
the measurement parameters, such as sensitive layer polarization or tem-
perature, which are controlled by using different electronic circuits.
The elementary reaction steps of gas sensing will be transduced in-
to electrical signals measured by appropriate electrode structures. The
sensing itself can take place at different sites of the structure depending
on the morphology. They will play different roles, according to the sens-
ing layer morphology. An overview is given in figure 2.4.
21

42. Gas Product Gas Product desorption
adsorp-
tion
(a) Compact layer
Gas Product
B
C
A
Sensitive layer
(b) Porous layer Electrodes
Substrate
Figure 2.4: Schematic view of gas sensing reaction in (a) Compact layer and (b) Po-
rous layer. A: grain boundary model. B: open neck model. C: closed neck model [32]
A simple distinction can be made between:
 Compact layers; the interaction with gases takes place only at the
geometric surface (Figure 2.5, such layers are obtained with most
of the techniques used for thin film deposition such as pulsed laser
deposition, sputtering etc.) and
 Porous layers; the volume of the layer is also accessible to the gas-
es and in this case the active surface is much higher than the geo-
metric one (Figure 2.6, such layers are characteristic to thick film
techniques and RGTO (Rheotaxial Growth and Thermal Oxida-
tion).
For compact layers, gases only interact with metal oxides at geo-
metrical surfaces, resulting in a surface depletion layer through the film
either partly (a) or completely (b) (Figure 2.5). Whether the sensor oper-
ates under a partly or a completely depleted, the condition is determined
22

43. by the ratio between layer thickness Zg and Debye length λD (or LD). For
partly depleted layers, when surface reactions do not influence the con-
duction in the entire layer (Zg >Z0 see Figure 2.5), the conduction process
takes place in the bulk region (of thickness Zg −Z0, much more conduc-
tive than the surface depleted layer).
Formally two resistances occur in parallel, one influenced by sur-
Gas Product Surface band
bending
Volume not accessible (a)
Z eVS
to gases
Z0
Current flow Zg Conducting
channel
Z Energy
X
eVS (b)
Z
Figure 2.5: schematic of a compact layer with geom- e∆VS
etry and energy band representation; Z0 is
the thickness of the depleted surface lay-
Zg
er; Zg is the thickness of the surface and
eVS the band bending. (a) A partly deplet-
ed compact layer (“thicker”) and (b) A Eb Energy
completely depleted layer (“thinner”)
[32].
face reactions and the other not; the conduction is parallel to the surface,
and this explains the limited sensitivity. Such a case is generally treated
as a conductive layer with a reaction-dependent thickness.
For the case of completely depleted layers in the absence of reduc-
ing gases, it is possible that exposure to reducing gases acts as a switch to
the partly depleted layer case (due to the injection of additional free
charge carriers) [32].
23

44. Large grains Product
Gas
Z
X
Energy
Current
flow
qVS
Eb
Xg> λD
Xg X
2X0
Small grains
Current flow
q∆VS <KB T
Energy Flat band condition
q∆VS
Eb
Xg< λD
X
Extended surface influence
Figure 2.6: Schematic representation of a porous sensing layer with geometry and
energy band for small and large grains. λD Debye length, Xg grain size [33].
24

45. It is also possible that exposure to oxidizing gases acts as a switch
between partly depleted and completely depleted layer cases depending
on the initial state of the sensing film.
Figure 2.6 illustrates the conduction model of a porous sensing lay-
er with geometry and surface energy band for small and large grains. For
large grains, conduction can be hindered by the formation of depletion
layers at surface/bulk regions and grain boundaries; the presence of ener-
gy barriers blocks the motion of charge carriers. In contrast, flat band
conditions dominate in case of small grains, allowing fast conduction for
this case [34]. For porous layers the situation may be complicated further
by the presence of necks between grains (Figure 2.7). It may be possible
to have all three types of contribution presented in figure 2.8 in a porous
Z
(b)
eVS
Zn
X
eVS
Z Z
(a)
eVS
Current Zo
flow
Zn Conduction
Zn- 2Z0
X Channel
Energy
Figure 2.7: schematic of a porous layer with geometry and surface energy band with necks
between grains; Zn is the neck diameter; Z0 is the thickness of the depletion layer and eVS
the band bending. (a) a partly depleted necks and (b) a completely depleted necks [32].
25

46. layer:
I. Surface/bulk (for large enough necks Zn >Z0, figure 2.5).
II. Grain boundary (for large grains not sintered together), in which
conduction can be hindered by the formation of depletion layers at
surface/bulk regions and grain boundaries; the presence of energy
barriers blocks the motion of charge carriers.
III. Flat bands conditions which dominate in case of small grains and
Surface Grain
Model bulk nano crystal Schottky contact
−
boundary
O2
Geometric
Z
X and Metal
I 𝐿 𝐷>1
2
CO effect
EC
Electronic EC
EF EF
EF EF
Band LD − LD LD
O2
EV EV
bulk Surface Metal
Z X X X
Surface
Electrical equiva-
lent circuit
(Low current)
Bulk
Figure 2.8 : Influence of particle size and contacts on resistances and capacitances in thin
films are shown schematically for a current flow I from left to right [35].
small necks, allowing fast conduction for this case.
Of course, what was mentioned for compact layers, i.e. the possible
switching role of reducing gases, is valid also for porous layers. For small
grains and narrow necks, when the mean free path of free charge carriers
becomes comparable with the dimension of the grains, a surface influence
26

47. on mobility should be taken into consideration. This happens because the
number of collisions experienced by the free charge carriers in the bulk of
the grain becomes comparable with the number of surface collisions; the
latter may be influenced by adsorbed species acting as additional scatter-
ing centers [34].
2.5 Gas Sensor Metrics
There are several measures of the performance of a gas sensor. The
“3S” parameters are often cited in the literature as sensitivity, stability,
and selectivity [20]. Sensitivity is the most frequently studied of these
parameters in the literature. Korotcenkov recently pointed out that stabil-
ity in nanoparticle – based sensors may be equally as important as sensi-
tivity when operating at elevated temperatures [36]. Selectivity has also
been extensively examined in the literature since this is a crucial factor
when creating a commercially viable device [37, 38]. Other issues such as
response time, recovery time and re – producibility have been less inten-
sively studied. All of these factors are important for building a microsen-
sor or microsensor array. These parameters are defined here as most of
them will be applied in this research.
2.5.1 Sensitivity
The response of a sensor upon the introduction of a particular gas
species is called the sensitivity (S The most general definition of sensitiv-
ity applied to solid – state chemi – resistive gas sensors is a change in the
electrical resistance (or conductance) relative to the initial state upon
exposure to a reducing or oxidizing gas component). The sensitivity de-
pends on many factors including the background gas composition, rela-
tive humidity level, sensor temperature, oxide microstructure, film thick-
ness and gas exposure time. One of the most common methods is to re-
27

48. port the ratio of the electrical resistance (R) in air to the resistance meas-
ured when a gas is introduced as shown in equation 2.28 [39]:
R G
= (2.28)
RG
where R is the electrical resistance and G is the electrical conductance
and the subscript “AIR” indicates that background is the initial dry air
state and the subscript “GAS” indicates the analyte gas has been intro-
duced.
Another common approach to report S is shown in equations 2.29
and 2.30 [40]:
|∆R| RG −R
= 100 = 100 (2.29)
R R
|∆ |
100 100 (2.30)
The values calculated using the above equations scale from zero
while the values from equation 2.28 scale from 1. The relationship be-
tween the S values (using G as a metric) from 2.28 and equation 2.30 is
simply to add 1 to the reported value and they are equivalent. Both values
are acceptable and useful metrics for gas sensor response testing. In the
current research, the percentage conductance change has been selected as
the value of sensitivity as calculated in equation 2.30 (S = ΔG/Go) be-
cause it scales more intuitively from a value of zero.
2.5.2 Selectivity
Selectivity is defined as the ability to discriminate a particular gas
species from the background atmosphere. This is a task where metal ox-
ide based sensors face significant challenges and show poor discrimina-
tion between gas species. Selectivity has been defined as the ratio of sen-
sitivities to a particular gas as shown in equation 2.31.
SG
⁄ = (2.31)
SG
28

49. In this work only a single gas was introduced in a background of dry air
so this ratio does not describe the ability of the sensor to pick out a par-
ticular gas species from a complex mixture of gases. Selectivity has also
been applied to describe the ability of a sensor (or array) to detect and
distinguish a particular gas species in a mixture containing multiple ana-
lyte gases [41]. As with sensitivity there are many factors (e.g. tempera-
ture gas flow rate, device electrode material, etc.) that contribute to the
selectivity of a sensor [41]. A sensor array uses the combination of sig-
nals from multiple sensors to test and improve the selectivity of a system.
This is why analytically orthogonal (opposite) signals such as those from
n – type and p – type materials in a sensor array are valuable for data
analysis algorithms to enhance selectivity. Selectivity is not a focus of
this study. It will become a more important issue once other reducing
and/or oxidizing gases become available.
2.5.3 Stability
Stability measures the capability of a sensor to maintain sensitivity
over durations of time for a particular gas species. Stability is measured
in terms of baseline “drift” which is the change in baseline conductance
over some duration of time at a particular temperature. Here we define
drift as the change in baseline conductance relative to the initial conduct-
ance as written in equation 2.32 below:
1
−
D=| | (2.32)
1− 0
Drift is reported in units of Siemens per hour (S/hr). Elvin R.
Beach III [20] reported the drift over durations of 48 hours by individual-
ly introduceing reducing and oxidizing gases during testing to simulate
more realistic conditions the sensor would operate under.
29

50. 2.5.4 Response and Recovery Times
The response time (τres) of a gas sensor is defined as the time it
takes the sensor to reach 90% of maximum/minimum value of conduct-
ance upon introduction of the reducing/oxidizing gas [42]. Similarly, the
recovery time (τrec) is defined as the time required to recover to within
10% of the original baseline when the flow of reducing or oxidizing gas
is removed. Figure 2.9 shows how this is measured from sensor data
plotting the conductance as a function of time.
90% of maximum
Maximum conductance
conductance
Conductance
Recovered to within 10%
of original baseline
τres τrec
Time
Figure 2.9: drawing showing how response and recovery times are calculated
from a plot of sensor conductance versus time [20].
Xu et al., [43] prepared ultrathin Pd nanocluster film capable of detecting
2% H2 with a rapid response time down to tens of milliseconds (~ 70 ms)
and is sensitive to 25 ppm hydrogen, detectable by a 2% increase in con-
ductance due to the hydrogen – induced palladium lattice expansion. Four
years earlier that time, in 2001, Favier et al., [44] obtained a comparable
ultrafast response time of less than 75 ms towards H2 : N2 gas mixture
from 2 to 10% at room temperature using palladium mesowire arrays.
30

51. 2.6 Sensing Mechanism
Both thin – film and bulk semiconducting metal oxide materials
have been widely used for the detection of a wide range of chemicals
such as H2, CO, NO2, NH3, H2S, ethanol, acetone, human breath, and
humidity. The sensing mechanism of metal oxide gas sensors mainly
relies on the change of electrical conductivity contributed by interactions
between metal oxides and surrounding environment. The exchange of
electrons between the bulk of a metal oxide nanostructure grain and the
surface states takes place within a surface layer (charge depletion layer),
thus, contributing to the decrease of the net charge carrier density in the
nanomaterial conductance channel. This will also lead to band bending
near the surface of both conduction and valence bands. The thickness of
the surface layer is of the order of the Debye length/radius λD of the sens-
ing material which can be expressed as the following formula obtained in
the Schottky approximation [32]:
1
eVS 2
ω = λD ( ) (2.33)
kT
1
εε0 kT 2
λD = ( 2 ) (2.34)
e n0
where ω is the width of the surface charge region that is related to the
Debye length λD of the nanomaterial, ε0 is the absolute dielectric con-
stant, ε is the relative dielectric permittivity of the structure, k is the
Boltzmann’s constant, T is the temperature, e is the elementary charge, n0
is the charge carrier concentration and VS is the adsorbate – induced band
bending. The Debye length λD is a quantum value for the distribution of
the space charge region. It is defined as the distance to the surface at
which the band bending is decreased to the 1/e – th part of the surface
value [11].
31