improvement of zno and sno2 hydrogen gas sensors
DESCRIPTION
Ph. D. thesis in PDF format ...Physics Department - College of Science - Baghdad UniversityTRANSCRIPT
I
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
II
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:
Name: Abdulla M. Suhail
Title: Assist. Professor
Address: College of Science,
University of Baghdad
Date: November , 2011
Signature:
Name: Wasan R. Saleh
Title: Assist. Professor
Address: College of Science,
University of Baghdad
Date: November , 2011
In view of the available recommendation, I forward this thesis for
debate by the Examining Committee.
Signature:
Name: Dr. Raad M. S. Al-Haddad
Title: Professor
Address: Collage of Science, University of Baghdad
Date: November 29, 2011
Suhail Wasan
Raad
III
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: Title: Professor
Name: Dr. Raad M. S. Al-Haddad Date: November 29, 2011 Chairman
Signature:
Title: Professor Name: Dr. Izzat M. AL-Essa
Date: November , 2011 Member
Signature:
Title: Professor
Name: Dr. Emad Kh. Al-Shakarchi Date: November 27, 2011
Member
Signature:
Title: Assist. Professor
Name: Dr. Mayada Bedry Al-Quzweny
Date: November 28, 2011
Member
Signature: Title: Assist. Professor
Name: Abdulla M. Suhail
Date: November , 2011
Supervisor
Signature: Title: Assist. Professor
Name: Dr. Bassam Ghalib Rasheed
Date: November 29, 2011
Member
Signature: Title: Assist. Professor
Name: Wasan R. Saleh
Date: November , 2011
Co-Supervisor
Approved by the Dean of college of Science
Signature:
Title: Professor
Name: Dr. Saleh Mahdi Ali
The Dean of the College of Science
Date: December 5, 2011
Raad
Suhail Wasan
Bassam M. B. Q.
Emad
Salih M. Ali
IV
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.
V
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.
VI
Dedicated to
All Those Who Care…
Including…
Her
VII
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.
VIII
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.
:إال قال في غده إني رأيت أنه ال يكتب أحد كتابا في يومه
م هذا " ."لكان أفضل، ولو ترك هذا لكان أجمللو غير هذا لكان أحسن، ولو زيد هذا لكان يستحسن، ولو قد
.النقص على جملة البشر وهذا من أعظم العبر، وهو دليل على استيالء
العمـــاد األصفهانـــي
Assuming that he’s not dead, Qahtan can best be reached at his “lifetime” email address of:
Mobile No.: +009647702981421
IX
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
X
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
XI
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
XII
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.
XIII
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.
XIV
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
XV
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
XVI
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.
XVII
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
XVIII
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
XIX
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.
XX
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
1
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-
2
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
3
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-
4
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-
5
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].
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
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
Table 1.1: Examples of applications for gas sensors and electronic noses [17].
6
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-
7
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.
8
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
9
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
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.
Physisorption Chemisorption
Intermolecular (van der Waals) Force Covalent Bonding
(adsorbate & surface)
Low Temperature High Temperature
Low Activation Energy
(<< 0.5 eV)
High Activation Energy
(> 0.5 eV)
Low Enthalpy Change
(ΔH < 20kJ/mol)
High Enthalpy Change
(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
Table 2.1: Comparison of physisorption and chemisorption [20].
10
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 O
2- 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)
Temperature Range (°C) Adsorption Reaction(s)
Room Temp.< T < 175°C O2 (g) + e− → O2
− (ads) O2 (g) + 2e− → O2
2− (ads)
175°C < T < 500°C O2 (g) + 2e− → 2O− (ads)
T > 500°C O− (ads) + e− → O2− (ads) O2 (g) + 4e− → 2O2− (ads)
Table 2.2: Temperature ranges associated with molecular and dissociative oxygen
adsorption reactions [20].
11
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:
K(T) =[C]c[D]d
[A]a[B]b= e−
∆G
KT (2.3)
12
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:
ZnZn + OO → ZnZn + VO∙∙ + 2e′ +
1
2O2(g) (2.4)
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:
KR = [VO∙∙](pO2)
1
2 ⋅ n2 (2.5)
where pO2denotes 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:
[VO∙∙] = (
KR
4)
1
3
(pO2)−1
6 (2.7)
and
n = (2)1
3(KR)1
3(pO2)−1
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:
13
σ = q(2)1
3(KR)1
3(pO2)−
1
6μe (2.9)
where q is the electronic charge and 𝜇𝑒is 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:
nil = Zni⋅⋅ + 2e′ +
1
2O2(g) (2.10)
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 N2O on
ZnO. If the generation of two electrons proceeds as in (2.4), the decom-
position reaction was proposed as follows [2]:
2e− + N2O → N2 + O2−(ads. ) (2.11)
O2−(ads. ) + N2O → 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-
14
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:
ns = Ndexp (−eVs
kT) (2.13)
where Nd is the concentration of donors, k is Boltzmann’s constant, and T
is the temperature. Since conductance (or resistivity) is proportional to ns,
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.
15
SnO2
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].
EF
O-
O-
O-
O-
O-
O-
O-
O-
O-
O-
O-
O-
O-
Grain Grain
Thermionic
emission
EC
𝑒𝑉𝑆1 ;𝐺1 = 𝐺0exp(−𝑒𝑉𝑆1
𝐾𝑇)
1
2O2
− Electron – depleted
region
(a) In air
O-
O-
O
O-
O-
O-
O
O-
O-
O-
O
Grain Grain
𝑒𝑉𝑆2 ;𝐺2 = 𝐺0exp(−𝑒𝑉𝑆2
𝐾𝑇)
𝑠𝑒𝑛𝑠𝑖𝑡𝑖𝑣𝑖𝑡𝑦 ⇉ 𝑆 =𝐺2
𝐺1= exp (
𝑒 𝑉𝑆1−𝑉𝑆2
𝐾𝑇)
CO
CO2
CO CO CO2
CO2
(b) In the presence
of reducing gas
16
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 → 2H2O + 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 → O2− → 2O− → 2O2− (2.16)
of these forms, O2 is considered fairly inactive due to its non – dissocia-
tive state. With regards to O2− and O
−, electron spin resonance (ESR)
studies have shown that O− is far more reactive than O2− [26]. The nature
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
17
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-
Figure 2.2: The nature of oxygen species adsorbed on ZnO as reported by several
researchers [25].
18
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]:
= exp (∆H
T) (2.18)
C+D
A+B
Ener
gy
Ea
Ec ΔH
I
II
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 (Ea),
while the barrier to product formation is lowered (Ec) when a catalyst is used [27].
19
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
√2 T) exp (
∆H
T) (2.20)
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:
20
=
+ =
a
+ a (2.22)
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
1 + b (2.23)
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
1 + b + b =
b 1 + b + b
(2.24)
If bBPB >> bAPA, then the equations of (2.24) become:
=b
1 + b =
b 1 + b
(2.25)
in the case that bBPB is >> 1, the equations of (2.25) reduce to:
b
1 + b 1 (2.26)
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]:
a e = k (2.27)
21
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.
22
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
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]
Sensitive layer
Electrodes
Substrate
C B
A
Product Gas
Product
(a) Compact layer
Gas
(b) Porous layer
Gas
adsorp-
tion
Product desorption
23
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-
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].
(a)
Surface band
bending
Conducting
channel
Volume not accessible
to gases
Product Gas
Current flow
Z0
Zg
Energy
eVS Z
Z
X
Zg
Z
Energy
eVS
e∆VS
Eb
(b)
Figure 2.5: schematic of a compact layer with geom-
etry and energy band representation; Z0 is
the thickness of the depleted surface lay-
er; 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”)
[32].
24
q∆VS <KB T
Flat band condition
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].
X
Energy
Eb
q∆VS
Xg< λD
Current flow
Small grains
Extended surface influence
Gas
Xg
Xg> λD
Eb
2X0
Energy
Current
flow
qVS
X
X
Z
Product Large grains
25
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
eVS
Zn
Z
X
eVS
(b)
(a) eVS
Zn
Zo
Z Z
X
Current
flow Conduction
Channel
Energy
Zn- 2Z0
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].
26
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
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
Model
Geometric
Electronic
Band
Electrical equiva-
lent circuit
(Low current)
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].
Grain
boundary
Surface
bulk nano crystal
𝐿𝐷>12
O2−
X and
I
Z
Schottky contact
Metal
EF
Metal
LD
X
CO effect
EV
EC
EF
X
EC
EV Surface
LD O2
−
EF
bulk
Z
LD
EF
X
Surface
Bulk
27
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-
28
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
RG =
G
(2.28)
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|
R 100 =
RG −R
R 100 (2.29)
|∆ |
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
SG (2.31)
29
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:
D = | 1 −
1 − 0| (2.32)
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.
30
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.
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.
Conductance
Time
90% of maximum
conductance Maximum conductance
Recovered to within 10%
of original baseline
τres τrec
Figure 2.9: drawing showing how response and recovery times are calculated
from a plot of sensor conductance versus time [20].
31
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]:
ω = λD (eVS
kT)
12
(2.33)
λD = (εε0kT
e2n0)
12
(2.34)
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 λDis 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].
32
The conductance of 1-D metal oxide nanomaterial can be ex-
pressed as [45]
=1
=
A
ρl= n0eμ
π(D − 2ω)2
4l (2.35)
where R is the electrical resistance, ρ is the resistivity, no is the ini-
tial/nominal charge carriers concentration, e is the electron charge, μ is
the mobility of electrons, l is the length of the nanomaterial, D is the di-
ameter of the nanomaterial, and w is the width of surface charge region
that is related to the Debye length of the nanomaterial. Likewise, the
electrical conductance of ZnO nanofilms can be expressed as dependent
[24] upon the charge carriers’ concentration:
=1
=
A
ρl=
no|e|μA
l (2.36)
here A, l are the area and length of the nanofilm channel, respectively.
Therefore, the change in electrical conductance of the nanofilm exposed
to gas atmosphere is determined by the change in electrical charge carri-
ers’ concentration ∆n [12]:
∆ =∆no|e|μA
l (2.37)
The gas sensitivity S is given by [24]:
𝑆 =|∆ |
100 =
∆nS
no (2.38)
According to this expression, higher gas sensitivity could be obtained by
a larger modulation in the depletion region of the ZnO metal oxide nano-
film. The width of the depletion region is inversely proportional to the
square root of the free charge carrier concentration.
When the radius of ZnO nanostructure (grain) is of the order of or
less than Debye length/radius, the conductive channel is reduced substan-
tially. The modulation of the depletion region width can also be produced
33
by the control of electron density in the metal oxide ZnO nanostructure,
i.e. by means of surface defects.
Generally speaking, the response of the chemoresistors in ambient
environment can be defined as [13]:
= g − a
g 100 =
4
D ωa − ωg =
4
D√𝜀𝜀0𝑒𝑛0
(VSa
12 − V
Sg
12 ) (2.39)
where Gg and Ga are the conductance of ZnO nanostructure in H2 gas and
in air ambient, respectively, n0 is carrier concentration in air. VSa and VSg
are the adsorbance – induced band bending in air and in H2 gas, respec-
tively. According to this equation, enhancement of H2 gas sensitivity can
be realized by controlling the geometric factor (4/D), electronic character-
istics (εε0/en0), and adsorption induced band bending (VSa
12 − V
Sg
12 ) due
to adsorption on the ZnO nanostructure surface. This can be done by
doping, or by using modulation of operation temperature which is not
desirable for H2 gas sensors on single ZnO nanowire. Another way is to
make use of geometric parameters, of which the grain size is at the fore-
front of the parameters used for sensitivity enhancement.
2.7 Factors influencing the performance
2.7.1 Long term effects / Baseline Drift:
Baseline refers to the conductance of the sensor in clean air.
Changes over long operating times of both baseline and sensitivity are all
important in utilization of the sensors. These determine the frequency at
which the calibration checks should be carried out and the frequency at
which the sensors may have to be replaced. They can only be determined
over long periods of time and no method by which the process can be
accelerated is valid [18].
34
2.7.2 Sensor surface poisoning
The surface of ZnO and other oxides may become unstable because
of “poisons”. Sulfur (as H2S) is a potential poison that can block the cata-
lytic activity of Pd on the surface. Wagner et al found instability due to
the presence of H2S in commercial SnO2 based sensors. Another domi-
nant poison is chlorine gas. Thus it is important in the development of
sensors to be aware of the other reactive gases in the measurand environ-
ment [18].
2.8 Optimization of Sensor Performance
2.8.1 Use of Catalyst
Metal oxide gas sensors need a catalyst deposited on the surface of
the film to accelerate the reaction and to increase the sensitivity. A cata-
lyst is a material that increases the rate of chemical reactions without
itself getting changed. It does not change the free energy of the reaction
but lowers the activation energy. Catalysts are supposed to, and do, im-
part speed of response and selectivity to gas sensors [45].
The catalytic surface reaction used for gas sensing makes this field
close to that of heterogeneous catalysis, with the only difference that in
catalysis one is mainly interested in the products of the reaction whereas
in gas sensing one is interested in the reactants as shown in the figure
2.10. This approach is considered relatively standard in fields such as
heterogeneous catalysis but so far it has rarely been applied to solid-state
gas sensors.
The chosen catalyst influences the selectivity of sensor. Ideally, if
one wants to detect a particular gas in a mixture of gases, one will like a
catalyst combination that catalyzes the oxidation of the gas of interest and
35
does not catalyze the oxidation of any other gas. Unfortunately, such
ideal combinations are not easily found [46].
The widespread applicability of semi-conducting oxides such as
SnO2 or ZnO, as gas sensors is related both to the range of conductance
variability and to the fact that it responds to both oxidizing and reducing
gases.
Small amounts of noble metal additives, such as Pd or Pt are com-
monly dispersed on the semiconductor as activators or sensitizers to im-
prove the gas selectivity, sensitivity and to lower the operating tempera-
ture [47]. There are two ways in which the catalysts can affect the inter-
granular contact region and hence affect the film resistance. The first one
is the spill over mechanism and the other is Fermi energy control.
Catalytic theory proposed, as spillover and Fermi energy control,
have not led to a widely accepted catalyst mechanism that predicts or
explains sensor behavior in different environments [48]. In spite of all the
work reported, a deep analysis of the material – gas interaction and its
influence on the sensor electrical response is still lacking to completely
Figure 2.10: Illustration of catalyst effect. Nano – particles, having higher
surface area, act as catalysts. Here, R stands for reducing gas [46].
O-
O-
O-
O-
O-
O- O
-
O-
O-
O-
O-
O-
O-
O-
O-
O-
O-
O-
RH2+O2 RH2+H2O
36
understand the role played by the additives on the gas sensing mecha-
nism. A model for increase in sensitivity using nanoparticles has been
explained by activated charge carrier creation and tunneling through po-
tential barrier [48].
2.8.1.1 Spill over Mechanism
Spillover mechanism is a well-known effect in heterogeneous ca-
talysis and is probably most active with metal catalysts. This interaction
is a chemical reaction by which additives assist the redox process of met-
al oxides. The term spillover refers to the process, illustrated in figure
2.11, namely the process where the metal catalysts dissociates the mole-
cule, then the atom can ‘spillover’ onto the surface of the semiconductor
support. At appropriate temperatures, reactants are first adsorbed on to
the surface of additive particles and then migrate to the oxide surface to
react there with surface oxygen species, affecting the surface conductivi-
ty. For the above processes to dominate the film resistance, the spilled-
over species must be able to migrate to the inter-granular contact as
shown in the figure 2.12.
Figure 2.11: Mechanism of sensitization by metal or metal oxide additive [48].
O-
O-
O-
O-
H
H
H
H H2O H2O
Activation of gas followed
by spilling over
Change in surface
oxygen concentration
M
Acceptor of electrons
Change in redox state of additive
O-
O-
O-
O-
e
H2O H2
M
Electronic sensitization Chemical sensitization
37
Thus, for a catalyst to be effective there must be a good dispersion
of the catalysts, as shown in the figure 2.13, so that catalyst particles are
available near all inter – granular contacts. Only then can the catalysts
affect the important inter-granular contact resistance. The inverse effect
may also occur [49] when a nascent oxygen or gas atom is newly formed
from a reaction on a metal oxide site. The nascent atom may migrate to a
metal site and desorbs into gas molecule from there. This is called reverse
spillover or the porthole effect.
2.8.1.2 Fermi Energy Control
The second interaction is the electronic sensitization (figure
2.11); in which additives interact electronically with the metal – oxide as
a sort of electron donor or acceptor. For example, changes in the work
function of the additive due to the presence of a gas will cause a change
in the Schottky barrier between the metal and the oxide and thus, a
change in conductivity. This simply means that oxygen adsorption on the
O2
O-
O-
O-
O- O-
O-
O-
O-
O-
O-
O-
H2
Figure 2.12: Illustration of Spill Over caused by catalyst particles on the surface
of the grain of the polycrystalline particle [46].
38
catalyst removes electrons from the catalyst and the catalyst, in turn,
removes electrons from the supporting semiconductor. Figure 2.11 illus-
trates the situation with Fermi energy control.
Figure 2.13 demonstrates the catalyst, by Fermi energy control, to
dominate the depletion of electrons form the semiconductor surface, but
the poor catalyst dispersion prohibits any influence on the inter-granular
contact resistance. In other words, oxygen adsorbed on the catalyst re-
moves electrons from the catalyst and the catalyst, in turn, removes elec-
trons from the nearby surface region of semiconductor. But if only a few
catalyst particles are on each semiconductor particle, only a small portion
of the semiconductor surface will have a surface barrier controlled by the
catalyst. Then, the chances of a catalyst particle being near enough to the
inter – granular contact to control its surface barrier will be small [46].
Figure 2.13 (b) shows the more desired situation where one has a
good dispersion of the catalyst particles such that the depleted regions, at
the surface of a metal-oxide, overlap and the influence of the catalyst
extends to the inter-granular contact.
(b) Need adequate catalyst dispesion
(Particle separation < 500 A)
Electron
flow
O
2
O-
(a) Poor catalyst dispersion
Catalyst
Figure 2.13: An adequate dispersion of the catalysts is required in order to effective-
ly affect the grains of the semi-conducting material to serve the implied purpose of
increase in sensitivity [46].
39
2.8.2 Grain size effects
One of the most important factors which affect the sensing proper-
ty of semiconducting gas sensors is the microstructure of polycrystalline
element [48]. Each crystallite of semiconductor oxide in the element has
an electron depleted surface to a depth of L in air, where L is determined
the by Debye length LD (or λD) and the strength of chemisorptions. The
grain size effects are pictorially depicted in figure 2.14. If the diameter D
of the crystallite is comparable to 2L, the whole crystallite will be deplet-
ed of electrons and this will cause the gas sensitivity of the element to the
reducing gas to change with D. The crystallites in the gas sensing ele-
ments are connected to the neighboring crystallites either by grain bound-
ary contacts or by necks. In the case of grain boundary contacts the elec-
trons should move across potential barrier, the height of which changes
with surrounding atmosphere .The gas sensitivity in this case is inde-
pendent of the grain size. In the case of conduction through necks, elec-
trons move through the channel penetrating through each neck. The aper-
Core region
(Low re-
sistance)
Space – charge
region
(High re-
sistance)
𝐷 ≥ 2𝐿 (𝑛𝑒𝑐𝑘 𝑐𝑜𝑛𝑡𝑟𝑜𝑙)
𝐷 ≫ 2𝐿 (𝑔𝑟𝑎𝑖𝑛 𝑏𝑜𝑢𝑛𝑑𝑎𝑟𝑦 𝑐𝑜𝑛𝑡𝑟𝑜𝑙)
𝐷 < 2𝐿 (𝑔𝑟𝑎𝑖𝑛 𝑐𝑜𝑛𝑡𝑟𝑜𝑙)
Figure 2.14: Schematic models for grain – size effects [31]
40
ture of the channel is attenuated by the surface space charge layer. This
model is related to the grain size through the neck size.
It has been found out experimentally by Yamazoe et al, in 1991
[50] that the neck size X is proportional to D with a proportionality con-
stant of 0.8 ± 0.1. For D>>2L, conduction of electrons in the sensing
element is dominated by conduction through grain boundary contacts
(grain boundary control). For D≥2L, neck control forms the primary
mechanism of conductivity modulation (neck control). For D<2L, the
electrical resistance of the grains dominates whole resistance of the sen-
sor and thus sensitivity is controlled by grains themselves (grain control).
2.8.3 Thickness dependence
Thin and thick film sensing layers differ not only in their thickness
but also in their microstructures and can thus lead to rather different
transducer functions [48, 51]. The sensitivity of the layers depends
strongly on the layer thickness. In the case, that the thickness of the elec-
tron depleted surface thickness is about the size of a film, high gas sensi-
tivity can be expected. Thus, sensitivity of the metal oxide sensor is di-
rectly influenced by the size of the oxygen induced depletion layer at the
film surface relative to the thickness of the bulk semiconductor. In gen-
eral, when the depletion width equals the film thickness, more sensitivity
is expected [51].
Adsorption of the atmospheric oxygen on the surface of sensing
film, results in an increase of the resistance of sensing thin film. Upon
exposing to reducing gas, reduction in depletion layer depth occurs thus
decreasing the resistance of the film.
When the depletion depth is more or less equal to the thickness of
sensing film, the resistance will be high and hence contributing for the
higher sensitivity. However, it has been pointed out that the columnar
41
growth of gas sensitive film leads to the thickness-independent gas sensi-
tivity of sensor.
It has also been shown that the thickness of the sensitive layer does
play a role in determining the sensitivity of the sensor for different gases
[50]. The thin SnO2 layer, (thickness 50-300 nm) mainly responds to the
oxidizing gases such as Ozone and NO2 whereas thick films (thickness
15-80 μm) respond to reducing gases like CO and CH4. However, upon
reducing the temperature of the sensor, the thick film showed a signifi-
cant response to oxidizing gases. This behavior can be explained with the
diffusion reaction model. . A model for the sensing mechanism in thick-
film has been presented in [50].
2.8.4 Temperature modulation
The temperature of the sensor surface is one of the forefront char-
acterization parameters. Firstly, adsorption and desorption are tempera-
ture activated processes, thus dynamic properties of the sensors viz. re-
sponse time, recovery etc. depends on the temperature. The surface cov-
erage, co-adsorption, chemical decomposition or other reactions are also
temperature dependent, resulting in different static characteristics at dif-
ferent temperatures. On the other hand, temperature has an effect on the
physical properties of the semiconductor sensor material such as charge
carrier concentration, Debye length, work function etc.
The optimum range of temperature for an effective sensor response
corresponds to that where the material is able to catalytically reduce or
oxidize the target gas, simultaneously changing the electrical properties
of the sensor material. The rate of reaction depends on the exact reducing
agent under study. It is found that, with a given reducing agent, there is
peak in the sensitivity: If the temperature is too low, the rate of reaction is
too slow to give a high sensitivity, whereas if the temperature is too high,
42
the overall oxidation reaction proceeds so rapidly that the concentration
of reducing agent [R] at the surface becomes diffusion limited and con-
centration seen by sensor approaches to zero [48]. At such temperatures,
the whole target gas concentration reaching the material surface could be
reduced/oxidized without producing a perceptible electrical change on the
metal-oxide material. The sensitivity again is low. However, on the one
hand, temperature should be high enough to allow gas reaction on the
material surface. The operating temperature is chosen empirically to pro-
vide the highest sensitivity to the determinate gases. So, a clear compre-
hension of the relation between the sensing material, catalytic properties
and the sensor electrical response is indispensable to understand the
whole gas sensing mechanism.
For each sensor-gas combination, an optimum temperature be-
tween these limits must be used. When higher degrees of selectivity are
needed, sensor arrays are used (sometimes termed “electronic noses”),
where the different response of different sensors is used for identifying
the gaseous species by pattern matching [52]. With such sensor array, the
lack of selectivity of the single metal-oxide gas sensor can also be over-
come by processing the signals of the same kind of sensor devices at
different operating temperatures or of the device using different materials
at the same temperature [53, 54].
2.8.5 Filters for selectivity
The use of filters forms another approach to improve the selectivity
of gas sensors. These filters either consume gases that one does not wish
to pass to the gas sensor or to permit the passage of selected gases to the
sensor. Their use is to a great extent empirical. For example, Ogawa et al
claim that ultra-fine SnO2 rejects methanol. Carbon cloth and low porosi-
ty materials are used to prevent highly reactive or large molecules from
43
reaching the sensor. Silica can be used to increase hydrogen sensitivity,
as hydrogen passes more freely through a silica surface layer. Similarly
Teflon is helpful in stopping H2O reaching the sensor and Zirconia can be
used at high temperature to pass oxygen [17].
2.8.6 AC and DC measurements
The sensors resistance change is the best-known sensor output sig-
nal and in most cases determined at constant operation temperature and
by DC measurement. The inherently noisy behavior of the resistor, 1/f
noise also known as flicker noise in the DC resistance measurements can
often approach the desired sensitivity threshold of the sensor. AC re-
sistance measurements are one way to overcome prohibitive 1/f noise, but
they incur more complex measurement electronics and calibration repro-
ducibility issues. AC measurements are more frequently used in imped-
ance spectroscopy at modeling level. Udo Weimar and Wolfgang Göpel
have reported [55] that sensitivity and selectivity of the gas sensors can
be improved by applying different conduction measurement methods viz.
DC and AC conduction measurement methods. They have shown that the
use of different contact arrangements and monitoring at different fre-
quencies make it possible to discriminate between different gases.
The equivalent circuit for the different contributions; intergranular
contact, bulk and electrode contact is illustrated in figure 2.15. Intergran-
ular contact: the ionosorption of oxygen at the grain surface results in the
creation of a potential barrier and the corresponding depletion layers at
the intergranular contacts. An intergranular contact can be represented
electronically by a resistor Rgb (due to the high resistive depletion layers)
and a capacitor Cgb (due to the sandwiching of high resistive depletion
layers between two high conductive ‘plates’ of bulk material) in parallel.
The electrode contact can also be represented by a (RC) element. The
44
values of the resistor RC and the capacitor CC are independent of the am-
bient gas atmosphere. The bulk contribution can be represented by a re-
sistor Rb, whose resistance value is hardly influenced by changes in the
ambient atmosphere.
Gas
Large grains
X
Z
Product
Xg
Xg> λD
Eb
2X0
Energy
Current
flow
X
… …
qVS ∆Φ
𝑅 𝐶 ~𝑛𝑏exp(∆Φ
𝑘𝐵𝑇)
𝐶 𝐶 ~(𝜀
∆Φ)0.5
𝑅 𝑔𝑏 ~𝑛𝑏exp(𝑞𝑉𝑆𝑘𝐵𝑇
)
𝐶 𝑔𝑏 ~(𝜀
𝑞𝑉𝑆)0.5
𝑅 𝑏 ~𝑛𝑏 𝑅 𝑏 ~𝑛𝑏
Figure 2.15: Equivalent circuit for the different contributions in a thin film gas sen-
sor; intergranular contact, bulk and electrode contact.
45
2.9 Zinc oxide
2.9.1 Properties of Zinc oxide
Zinc oxide, ZnO, is an interesting II – VI compound semiconductor
with a wide direct band gap of 3.4 eV at room temperature [56]. It is a
widely used material in various applications such as gas sensors, UV
resistive coatings, piezoelectric devices, varistors, surface acoustic wave
(SAW) devices and transparent conductive oxide electrodes [57]. In the
early 2000 ‘s, ZnO also attracted attention for its possible application in
short – wavelength light emitting diodes (LEDs) and laser diodes (LDs)
because the optical properties of ZnO are similar to those of GaN [56].
Figure 2.16 shows the phase diagram of the Zn – O binary system
[58]. The equilibrium solid phase of the condensed Zn – O system at 0.1
MPa hydrostatic pressure are the hexagonal closed packed (hcp) Zn with
a very narrow composition range, the hexagonal compound, ZnO (49.9 to
50.0 at % O), with a narrow but significant composition range, and a
cubic peroxide, ZnO2 (~66.7 a . O), with unknown composition range.
Even though the existence of ZnO2 has been reported, its nature and tem-
perature of formation are unknown. At elevated hydrostatic pressure, a
face centered cubic (fcc) modification of ZnO is stable. Also, it has been
reported that ZnO can exist metastably at room temperature in either of
two cubic modifications with structure of ZnO (sphalerite) and NaCl
(rock salt) types [58]. Table 2.3 summarizes data related to Zn – O crystal
structures.
ZnO crystals are composed of alternate layers of zinc and oxygen
atoms disposed in a wurtzite hexagonal close – packed structure with a
longitudinal axis (c – axis) as shown in figure 2.17 [59]. The oxygen
atoms (ions) are arranged in close hexagonal packing, with zinc ions
46
Stable phases at 0.1 MPa Other phases
Zn ZnO (I) ZnO2 ZnO (II)(a) ZnO (III)
Composition,
at. % O ~0 49.9 to 50.0 ~66.7 ~50 ~50
Pearson
Symbol hP2 hP4 cP12 cF8 cF8
Space group P63/mmc P63mc Pa3 Fm3(-)m F4(-)3m
Prototype Mg ZnO
(wurtzite)
FeS2
(pyrite) NaCl
ZnS
(sphalerite)
Figure 2.16: T – X diagram for condensed Zn- O system at 0.1 MPa [58].
O – Rich
Boundary
Unknown
Zn – Rich Boundary
49.999 Liq. ~0.005
Liq. ~ 7 10−7
T 0C
800
600
400
419.58 0
Liquid
0 10 20 30 40 50 60
At. %
ZnO2
ZnO
~50.00
O
70
~419.80
Z
n
(Zn)~O
(Zn)
M.P.
Table 2.3: Zn – O crystal structure data [58]
47
occupying half the tetrahedral interstitial positions with the same relative
arrangement as the oxygen ions. In the crystal structure, both zinc and
oxygen ions are coordinated with four ions of the opposite charge, and
the binding is strong ionic type. Owing to the marked difference in size,
these ions fill only 44% of the volume in a ZnO crystal leaving some
relatively large open spaces (0.095 nm). Typical properties of ZnO are
listed in Table 2.4 [58, 60] and Ellingham diagram including ZnO is
shown in figure 2.18 [61].
Pure zinc oxide, carefully prepared in a laboratory, is a good insu-
lator. However, its electrical conductivity can be increased many folds by
special heat treatments and by the introduction of specific impurities into
the crystal lattice. ZnO can even be made to exhibit metallic conductivity
as for transparent electrodes similar to ITO. In general, 0.5 – 1% addi-
Figure 2.17: Many properties of zinc oxide are dependent upon the wurtzite
hexagonal, close-packed arrangement of the Zn and O atoms, their cohesiveness
and void space [59].
O
Zn
Zn
Zn O
O
Zn
O
Zn
O
48
tions of trivalent cations (e.g. Al and Cr) decrease the resistivity of ZnO
by about 10 orders of magnitude. [58].
Property Value
Crystal structure Hexagonal, wurtzite
Molecular weight Zn:65.38, O:16 and ZnO:81.38
Lattice parameters at 300 K (nm) a0: 0.32495
c0: 0.52069
Density (g cm-3
) 5.606 or
4.21 x 1019
ZnO molecules/mm3
Stable phase at 300 K Wurtzite
Melting point (ºC) 1975
Thermal conductivity 0.6, 1-1.2
Linear thermal expansion coefficient a0: 6.5 10
-6
c0: 3.0 10-6
dielectric constant 𝜀0 = 8.75, 𝜀∞ = 3.75
Refractive index 2.008, 2.029
Energy band gap (eV) Direct, 3.37
Intrinsic carrier concentration (cm-3
)
<106
max n-type doping: n ~ 1020
max p-type doping: p ~ 1017
Debye temperature 370 K
Lattice energy 964 kcal/mole
Exciton binding energy (meV) 60
Pyroelectric constant 6.8 Amp./sec/cm2/K x 10
10
Piezoelectric coefficient D33 = 12 pC/N
Electron effective mass 𝑚𝑒∗ 0.24
Electron Hall mobility, n-type at 300 K (cm2V
-1s
-1) 200
Hole effective mass 𝑚ℎ∗ 0.59
Hole Hall mobility, p-type at 300 K (cm2V
-1s
-1) 5 – 50
Table 2.4: Typical properties of zinc oxide [58, 60].
49
2.9.2 Defects chemistry
Many properties of crystals, most particularly electrical, are deter-
mined by imperfections, e.g. defects. Point defects are defined as devia-
tions from the perfect atomic arrangement: missing ions, interstitial ions
and their associated valence electrons as shown in figure 2.19. A principal
difference between point defects in ionic solids and those in metals is that
in the former, all such defects can be electrically charged. Ionic defects
Figure 2.18: Ellingham diagram of oxides [61].
50
are point defects that occupy lattice atomic positions, including vacan-
cies, interstitial and substitutional solutes. Electronic defects are devia-
tions from the ground state electron orbital configuration of a crystal,
formed when valence electrons are excited into higher orbital energy
levels. Such an excitation may create an electron in the conduction band
and/or an electron hole in the valence band of the crystal. In terms of
spatial positioning, these defects may be localized near atom sites, in
which case they represent changes in the ionization state of an atom, or
may be delocalized and move freely through the crystal.
An equivalent way to view the formation of defects is as a chemi-
cal reaction, for which there is an equilibrium constant. Chemical reac-
tions for the formation of defects within a solid must obey mass, site and
charge balance. In this respect they differ somewhat from ordinary chem-
ical reactions, which must obey only mass and charge balance. Site bal-
ance means that the ratio of cation to anion sites of the crystal must be
preserved, although the total number of sites can be increased or de-
creased.
Vacancy
Vacancy Substitutional
impurity
Interstitial
impurity
Figure 2.19: various types of point defects in crystalline materials [62].
A A A A
B A A
A A A A
A A
A A A A
51
For example, the Schottky disorder for NaCl and Frenkel disorder
for ZnO, respectively, can be written using Kröger – Vink notation as:
null = VNa′ + VCl
⋅ (2.40)
and
ZnZnX = Zni
⋅⋅ + VZn′′ (2.41)
where null indicates the creation of defects from a perfect lattice. The
respective mass – action equilibrium constants are:
KScho ky = [VNa′ ] ⋅ [VCl
⋅ ] (2.42)
and
KF enkel = [Zni⋅⋅] ⋅ [VZn
′′ ] (2.43)
The brackets denote concentration, usually given in mole fraction. Writ-
ing the equilibrium constant as the product of concentrations implies that
the thermodynamic activity of each defect is equal to its concentration.
The free energies for these quasichemical reactions are simply the
Schottky or Frenkel formation energy, and the equilibrium constant is
given by:
KScho ky = KS∘ exp (−
∆HS
kT) (2.44)
and
KF enkel = KF∘ exp (−
∆HF
kT) (2.45)
The equilibrium constant is a function of temperature only and the prod-
uct of the cation and anion vacancy concentrations is a constant at fixed
52
temperature. Furthermore, when only intrinsic defects are present, the
concentration of anion and cation vacancies must be equal for charge
neutrality considerations,
[VNa′ ] = [VCl
⋅ ] = KS∘ 1 2⁄ exp (−
∆HS
2kT) (2.46)
and
[Zni⋅⋅] = [VZn
′′ ] = KF∘ 1 2⁄ exp (−
∆HF
2kT) (2.47)
In Kröger – Vink notation, free electrons and holes do not themselves
occupy lattice sites. The process of forming intrinsic electron – hole pairs
is excitation across the band gap, which can be written as the intrinsic
electronic reaction:
null = e′ + h⋅ (2.48)
The equilibrium constant for this reaction is:
Ke = n. p = [e′]. [h⋅] = NC. NVexp (− g
kT) (2.49)
where NC and NV are the density of state of conduction band and valence
band, respectively, and Eg is the energy band gap of the material.
When electrons and holes are tightly bound to an ion, or otherwise
localized at a lattice site, the whole is considered to be one ionic defect.
Thus, the valence state of defects such as vacancies and interstitials can
vary. For instance, an oxygen vacancy can in principle take on different
valence states (VO⋅⋅ VO
⋅ and VOX ), as can cation interstitials, e.g., Zni
⋅⋅, Zni⋅
and ZniX in the wurtzite structure compound ZnO.
53
Equilibration of ionic solids with an ambient gas plays an im-
portant role in determining defect structure.
For example, the reduction of ZnO can be written as the removal of
oxygen to the gas phase leaving behind doubly charged oxygen vacancies
or cation interstitials:
OOX =
1
2O2(g) + VO
⋅⋅ + 2e′ o OOX =
1
2O2(g) + Zni
⋅⋅ + 2e′ (2.50)
The equilibrium constant for the creation of double ionized oxygen va-
cancies is:
KR = n2. [VO⋅⋅].
O2
12 = KR
0 . exp (− R
kT) (2.51)
In ZnO, the electron is a major electronic charge carrier. Thus, the con-
ductivity of ZnO is:
σ ∝ n = 2 ⋅ [VO⋅⋅] = 2 ⋅ (2KR
0)1
3 ⋅ exp (− R
3kT) ⋅ (O2)
−16 (2.52)
With background acceptor,
[A′] = 2 ⋅ [VO⋅⋅] (2.53)
The conductivity of ZnO is:
σ ∝ n = (2KR0)
12 ⋅ exp (−
R
2kT) ⋅ (O2)
−14 ⋅ [A′]
−12 (2.54)
Also, the reduction of ZnO can be expressed by creating single charged
oxygen vacancies or cation interstitials:
OOX =
1
2O2(g) + VO
⋅ + e′ o OOX =
1
2O2(g) + Zni
⋅ + e′ (2.55)
54
The equilibrium constant for the creation of single ionized oxygen vacan-
cies is:
KR = n. [VO⋅ ].
O2
12 = KR
0 . exp (− R
kT) (2.56)
Thus, the conductivity of semiconducting ZnO in which single ionized
oxygen vacancies are dominant is:
σ ∝ n = [VO⋅ ] = (KR
0)1
2 ⋅ exp (− R
2kT) ⋅ (O2)
−14 (2.57)
With background acceptor,
[A′] = [VO⋅ ] (2.58)
The conductivity of ZnO is:
σ ∝ n = (KR0) ⋅ exp (−
R
kT) ⋅ (O2)
−12 ⋅ [A′] −1 (2.59)
Thus, investigating the conductivity of ZnO in reducing environments can
assist in determining the valence state of defects and the activation energy
for releasing electrons.
Substituted foreign atoms can also enhance the semiconducting
properties of ZnO. In the presence of selected metallic vapors at elevated
temperatures, the foreign metallic atom replaces a portion of the Zn at-
oms. The zinc atoms, on release from their lattice positions, diffuse to the
crystal surface where they are vaporized. This substitution process can
substantially alter the crystal properties, depending upon the nature, con-
centration and valence of the foreign atom. Optical and electrical proper-
ties are two of the several areas that can be readily modified.
55
2.9.3 Spray pyrolysis deposition technique
Chemical spray pyrolysis (CSP) is used for depositing a wide vari-
ety of thin films, which are used in devices like solar cells, sensors, solid
oxide fuel cells etc. It has evolved into an important thin film deposition
technique and is classified under chemical methods of deposition [63].
This method offers a number of advantages over other deposition pro-
cesses, the main ones being scalability of the process, cost – effectiveness
with regard to equipment costs and energy needs, easiness of doping,
operation at moderate temperatures (100 – 500 °C) which opens the pos-
sibility of wide variety of substrates, control of thickness, variation of
film composition along the thickness and possibility of multilayer deposi-
tion. Spray pyrolysis has been used for several decades in the glass indus-
try [64] and in solar cell production [65].
Typical spray pyrolysis equipment consists of an atomizer, precur-
sor solution, substrate heater, and temperature controller. The following
atomizers, table 2.5, are usually used in spray pyrolysis technique: air
blast (the liquid is exposed to a stream of air) [66], ultrasonic (ultrasonic
frequencies produce the short wavelengths necessary for fine atomiza-
tion) [67, 68] and electrostatic (the liquid is exposed to a high electric
field) [69, 70].
Key parameters of this process are the atomization technique, aero-
sol transport (carrier gas, pressure, distance, and reactor geometry), sub-
Atomizer Droplet size µm Atomization rate
cm3/min.
Droplet velocity
m/s
Pressure 10-100 3-no limit 5-20
Nebulizer 0.1-2 0.5-5 0.2-0.4
Ultrasonic 1-100 <2 0.2-0.4
Electrostatic 0.1-10
Table 2.5: Characteristics of atomizers commonly used for SPD [63].
56
strate temperature and material, the relative expansion coefficients of the
film and the substrate upon which it is deposited, and most importantly,
the chemical composition of the solution (both solvent and precursor salt
types). [71].
Many studies were made on CSP process since the pioneering work
by Hill and Chamberlin in 1964 on CdS films for solar cells [65]. Several
reviews on this technique have also been published.
Siu and Kwok made a detailed study of the properties of the Cu2S/CdS
thin film solar cells formed on chemically sprayed CdS films. Good and
reproducible films could be obtained using a spray-rate of 2.8 ml min-1
and a substrate temperature of around 340 0C. They demonstrated that
cells made on sprayed films could compete well with cells made on evap-
orated films, especially when cost is also considered [72].
Henry et al. reviewed CSP technique in which properties of specific films
of oxide superconductors in relation to deposition parameters and their
device applications were discussed in detail [73].
Brown and Bates discussed the preparation, properties and applications,
as solar cell, of spray – coated CulnSe2 thin films at 250 0C deposition
temperature [74].
Song et al., presented a preparation procedure of spray pyrolyzed un-
doped and aluminium doped zinc oxide thin films for solar cell. They
investigated the effects of the various deposition parameters and vacuum
– annealing of ZnO. ZnO:Al thin films with a transmittance at about 80%
and a resistivity as low as 3.5 x 10-3
Ω.cm were obtained using CSP dep-
osition route [75].
57
Next, in 1995, Roh et al., [76] did employ ultrasonic nozzle to deposit
CdS thin films on SAW devices intended for SO2 gas sensing.
Polycrystalline tin oxide SnO2 films with nano – size crystallites (8 – 20
nm) were prepared by Korotcenkov et al., [77]. The crystallites with ori-
entation (110) or (200) plane parallel to substrate, forming the surface of
the film, were predominant in the sprayed SnO2 films. The latter factor is
important in influencing the gas sensitivity characteristics similar to the
grain size effect.
Different atomization techniques and properties of metal oxide, chalco-
genide and superconducting films prepared using CSP were discussed by
Patil [78]. The results proved that film properties depended on the prepa-
ration conditions and could be easily tailored via optimizing spraying
parameters viz. substrate temperature, spray rate, precursor concentration
etc.
After that, Ebothe and El Hichou [79] examined the role of different
spraying flow rates deposition parameter f, between 1 and 8 mil min-1
, on
the surface irregularities evolution of a sprayed ZnO thin film of the same
thickness e evaluated at 1 mm by always adjusting the deposition time, t,
to the f value. This thickness has been confirmed from cross-sectional
images of the samples examined by scanning electron microscopy (SEM)
using a LEO 982 set. The substrate – nozzle distance d=44.5 cm is kept
constant and the optimal spraying temperature used is 450 0C. The XRD
results revealed that the variation of f has no effect on the material’s
structure as it remains hexagonal and has (002) preferred growth orienta-
tion which is normal to the substrate plane.
58
In 2005, Perednis and Gaukler gave an extensive review on the effect of
spray parameters on films as well as models for thin film deposition by
CSP [63].
Gümüs et al. [66] reported highly transparent ZnO thin films that had
successfully been prepared by pyrolytically spraying zinc acetate solution
on glass substrate at 400 °C using air as a carrier gas. The X-ray diffrac-
tion analysis shows that film is polycrystalline in nature and exhibits
excellent crystalline structure with (002) preferential orientation perpen-
dicular to substrate surface. The grain size is estimated to be 40 nm. Opti-
cal measurements show that the film possesses a high transmittance of
over 90 % in the visible region and a sharp absorption edge near 380 nm.
Envelope method is employed to calculate the refractive index and ex-
tinction coefficient as a function of wavelength. The film has a 3.27 eV
optical direct band gap which is close to the elsewhere – reported value
(3.25 – 3.27 eV) [69].
Recently, Sahay et al. [80] analyzed the optical and electrical properties
of a ZnO thin film obtained by spraying a 0.1 M zinc acetate precursor on
glass substrate held at 370 oC temperature. The optical energy gap for the
film of different thicknesses is estimated to be in the range 2.98 – 3.09
eV. The film exhibits thermally activated electronic conduction and the
activation energies depending on the film thickness. Moreover, the con-
ducted impedance spectra contained a single arc with a non – zero inter-
section with the real axis in the high frequency region.
Next, using simple, flexible and cost-effective ultrasonic spray pyrolysis
(USP) technique, Babu et al. prepared Al – doped ZnO (AZO) thin films
at substrate temperatures around 475 0C. Zinc acetate dehydrate (Zn
(CH3COO)2.2H2O) and Aluminum acetylacetonate (C15H21AlO6) were
59
used as precursors and the solvent was a mixture of de – ionized water,
methanol and acetic acid. The obtained films are polycrystalline with a
hexagonal wurtzite structure and are preferentially oriented in the (002)
crystallographic direction. Grain sizes varied from 21.3 to 25.3 nm based
on substrate temperature. An average transmission of 75% is observed
and the optical band gap of AZO films is varied from 3.26 to 3.29 eV
with the increase in substrate temperature [81].
In the early 2011, Gledhill et al. [82] prepared highly transparent, conduc-
tive ZnO films deposited by spray pyrolysis of zinc acetate – based solu-
tion. Quality films yielded as the spraying process is analogous to an
aerosol assisted chemical vapour deposition rather than a droplet deposi-
tion spray pyrolysis technique. Aluminum – doped zinc oxide (ZnO:Al)
films are grown with free charge carrier concentrations of more than
1020
cm−3
. The carrier density and mobility are measured by both Hall
probe and near infrared spectroscopy. Film growth and grain size, mor-
phology and orientation have been altered using an increased percentage
of ZnCl2 in the precursor, which resulted in a 10 – fold increase in charge
carrier mobility (~10 cm2V
−1 s
−1). An investigation is presented correlat-
ing the composition of the precursor solution with the chemical, structur-
al, electrical and optical properties of the grown films.
2.9.3.1 The deposition process and atomization models
CSP technique involves spraying a solution, usually aqueous, con-
taining soluble salts of the constituents of the desired compound onto a
heated substrate. Typical CSP equipment consists of an atomizer, a sub-
strate heater, temperature controller and a solution container. Additional
features like solution flow rate control, improvement of atomization by
electrostatic spray or ultrasonic nebulization can be incorporated into this
basic system to improve the quality of the films. To achieve uniform large
60
area deposition, moving arrangements are used where either nozzle or
substrate or both are moved. The schematic diagram of a typical spray
unit is depicted in figure 2.20.
Only crude models about the mechanism of spray deposition and
film formation have been developed. There are too many processes that
occur sequentially or simultaneously during the film formation by CSP.
These include atomization of precursor solution, droplet transport, evapo-
ration, spreading on the substrate, drying and decomposition. Understand-
ing these processes will help to improve film quality.
Deposition process in CSP has three main steps: atomization of
precursor solutions, transportation of the resultant aerosol and decompo-
sition of the precursor on the substrate. Atomization of liquids has been
Compressed
Air
Gas Regulator
Valve
Figure 2.20: Schematic diagram of chemical spray pyrolysis unit [83].
Spray
Solution
Power Supply
Heater
Substrate
Hot Plate
Temperature
Controller
Power Supply
Electronic
Controller
Mechanical
System
Thermocouple
Spray nozzle
61
investigated for years. It is important to know which type of atomizer is
best suited for each application and how the performance of the atomizer
is affected by variations in liquid properties and operation conditions. Air
blast, ultrasonic and electrostatic atomizers are normally used. Among
them, air blast atomization is the simplest. However this technique has
limitation in obtaining reproducible droplets of micrometer or submicron
size and in controlling their distribution [63, 83].
In ultrasonic nebulized atomization, precursor solutions are fogged
using an ultrasonic nebulizer [68]. The vapour generated is transported by
carrier gas to the heated substrate. Precursor solution is converted to
small droplets by ultrasonic waves and such droplets are very small with
narrow size distribution and have no inertia in their movement. Pyrolysis
of an aerosol produced by ultrasonic spraying is known as pyrosol pro-
cess [10, 75]. Advantage of this technique is that the gas flow rate is in-
dependent of aerosol flow rate, unlike the case of air blast spraying.
Electrostatic spray deposition technique has gained significance
only in recent years. Electrostatic atomization of liquids was first reported
by Zeleny [84]. Jaworek et al. published an article on this type of atomi-
zation [70]. A positive high voltage applied to the spray nozzle generated
a positively charged spray. Stainless steel discs acted as cathode and the
droplets under electrostatic force moved towards the hot substrate where
pyrolysis took place. In electrostatic spray, depending on the spray pa-
rameters, various spraying modes were obtained. They were classified as
cone – jet mode and multi – jet mode. Cone – jet mode split into multi –
jet mode with increase in electric field, where number of jets increased
with applied voltage.
62
The reaction process taking place in CSP is interesting. Many
models exist for the decomposition of precursor. Many simultaneous
processes occur when a droplet hits the substrate surface: evaporation of
residual solvent, spreading of droplet and salt decomposition. Vigue and
Spitz proposed that the following processes occur with increasing sub-
strate temperature [85]. Figure 2.21 given below illustrates the four pos-
sible processes that occur with increasing temperature.
In process A, droplet splashes on substrate, vaporizes and leaves a dry
precipitate in which decomposition occurs.
In process B, solvent evaporates before the droplet reaches the surface
and precipitate impinges on the surface where decomposition occurs.
In process C, solvent vaporizes as droplet approaches the substrate, then
solid melts and sublimes and vapour diffuses to substrate to undergo het-
erogeneous reaction there.
In process D, at highest temperature, the metallic compound vaporizes
before it reaches the substrate and chemical reaction takes place in vapour
phase.
Substrate
Finely divided solid
product
Vapour
Precipitate
D C B A
Figure 2.21: spray processes (A, B, C, and D) occurring with increase in substrate
temperature [85, 63].
63
Most of the spray pyrolysis deposition is of type A or B and our discus-
sion will be focused on these two.
2.9.3.2 Deposition parameters
Properties of film deposited depends on various deposition parame-
ters like substrate temperature, nature of spray and movement of spray
head, spray rate, type of carrier gas, nature of reactants and solvents used.
The effect of some important spray parameters are discussed here.
I Substrate temperature
Substrate temperature plays a major role in determining the proper-
ties of the films formed. It is generally observed that higher substrate
temperature results in the formation of better crystalline films [66, 80].
Grain size is primarily determined by initial nucleation density and re-
crystallization. Recrystallization into larger grains is enhanced at higher
temperature [86]. By increasing the substrate temperature, the film mor-
phology can be changed from cracked to dense and then to porous [87].
Variation of substrate temperature over different points results in non-
uniform films. Composition and thickness are affected by changes in
substrate temperature which consequently affect the properties of depos-
ited films. Though surface temperature is a critical factor, most investiga-
tors have not known the actual surface temperature of the substrate. Also,
maintenance of substrate temperature at the preset value and its uniformi-
ty over large area are challenging. Spraying in pulses or bursts also has
been used to assure that surface temperature is reasonably constant [86].
II Influence of precursors
The precursor used for spraying is very important and it extremely
affects the film properties. Solvent, type of salt, concentration and addi-
tives influence the physical and chemical properties of the films [71].
64
For ZnO thin films grown by spray pyrolysis, it was found that organic
salts (e.g., zinc acetate Zn(CH3COO)2.2H2O) are preferable over inorgan-
ic ones such as chlorides and nitrates. In the case of inorganic salts, un-
wanted etching processes, caused by acids formed as a result of the pre-
cursor decomposition, lead to degradation of the films performance.
Similarly, organic solvents are preferable over water due to a better drop-
let size distribution and, also, due to additional heat transfer toward the
sample surface by their burning. It was observed that transparency of as
deposited ZnO films increased when ethanol was used instead of water as
solvent for zinc acetate [71].
III Spray rate
Spray rate is yet another parameter influencing the properties of
films formed. It has been reported that properties like crystallinity, sur-
face morphology, resistivity and even thickness are affected by changes in
spray rate [88].
It is generally observed that smaller spray rate favours formation of
better crystalline films. Smaller spray rate requires higher deposition time
for obtaining films of the same thickness prepared at higher spray rate.
Also, the surface temperature of substrate may deviate to a lower value at
high spray rate. These two factors may contribute to the higher crystal-
linity at small spray rates. Decrease in crystallinity at higher spray rates is
observed in sprayed CuInS2 thin films [88]. Decrease in crystallinity
usually results in increased resistivity of the films.
Surface morphology of the films varies with spray rate. Higher
spray rate results in rough films. Also, it is reported that films deposited
at smaller spray rates are thinner due to the higher re-evaporation rate.
65
IV Other parameters
Parameters like height and angle of spray head, angle or span of
spray, type of scanning, pressure and nature of carrier gas etc., influence
the properties of deposited films. Different types of spray heads which
produce different spray patterns are commercially available. Relative
motion of the substrate holder and spray head should ensure maximum
uniformity and large area coverage.
2.9.4 Metal oxide gas sensors
The idea of using semiconductors as gas sensitive devices leads
back to 1952 when Brattain and Bardeen first reported that gas adsorption
on germanium semiconductor surface caused a variation in its electrical
conductivity [3]. The first realization of a working gas sensor was in
1962, when Seiyama et al. detailed the use of zinc oxide ZnO thin films
in the detection of gases as ethanol (C2H6O) and carbon dioxide (CO2)
[4]. Naoyoshi Taguchi, in that same year, published a patent for a tin
oxide (SnO2) – based gas sensor [5]. From then, the detection of hydro-
gen (H2), oxygen (O2) and hydrocarbon by means of surface conductivity
changes on various metal oxide crystals and thin films have been pro-
posed and demonstrated [89]. Although, many metal oxides have been
successfully demonstrated in gas sensing, SnO2 and ZnO have been in-
tensively investigated fundamentally and commercially due to the attrac-
tive structural, optical and electrical properties they possess and its easy
fabrication in thin film form with various methods [12, 13, 14, 41, 69, 92]
as well as, its improved sensor performance by addition of dopants [41,
47, 58, 93, 95].
The sensing performance (magnitude of gas response (sensitivity),
selectivity, sensing temperature, response /recovery time and so on) de-
66
pends on the electronic and structural properties of the sensor material
[36]. The sensing parameters can be promoted by the addition of metal
additives such as Al, Sn, Sb, Cu, Pt and Pd. These additives greatly en-
hance the gas sensor sensitivity, shorten the response time, and shift the
volcano – shaped correlations between gas response and temperature
toward the lower temperature side [48].
Nanto et al., [38] prepared a sensor with a high sensitivity and an excel-
lent selectivity for ammonia by using sputtered ZnO thin films. The sen-
sor exhibited a negative resistance change on exposure to oxidizing am-
monia gas (NH3) whereas the change became positive for exposure to
many other reducing inflammable and organic gases (H2, methane CH4,
butane C4H10, acetone C3H6O, ethanol C2H5OH). The resistance change
and the selectivity of the sensor were enhanced by doping group III metal
impurities such as AI, In, and Ga. On exposure to 200 ppm ammonia gas,
the resistance changed about three times as large as that in the undoped
ZnO sensors. The lower limit of the detection for ammonia gas was about
1 ppm at a working temperature of 350 0C. The sensing mechanism of
ammonia gas was related to the enhancement of adsorption of atmospher-
ic oxygen.
Five years later, in 1991, in a study to develop a cheap smell sensor capa-
ble of detecting the various gases, the freshness of sea foods and drinks,
and the fragrance from wine and coffee, Nanto et al., [90] investigated the
gas sensitivity of RF and DC magnetron – sputtered aluminum doped
zinc oxide (ZnO:Al) thin film on corning 17059 glass substrate with a
film thickness of about 300 nm. The sensitivity measurements were car-
ried out at an operating temperature of 200 – 350 °C in air. A testing gas
of 200 ppm was introduced by using an injector into the testing glass bell
– jar. The resistance of the sensor changed on exposure to odor from rot-
67
ten sea foods such as oyster, squid, sardine or fragrance from wine and
coffee. The high sensitivity of the sensor for the odor from rotten sea
foods was attributed to the high sensitivity of the sensor for trimethyla-
mine N(CH3)3 in the odor.
Hong et al., [91] successfully demonstrated the identification of methan-
ethiol CH3SH, trimethylamine (CH3)3N, ethanol C2H5OH and CO gases
in the 0.1 – 100 ppm concentration range by a gas recognition system
using a thin film oxide semiconductor micro gas sensor array and the
neural – network pattern recognition technique. The sensing materials of
1 wt. % Pd – doped SnO2, 6 wt. % A1203 – doped ZnO, WO3 and ZnO
were used for the gas sensor array whose power consumption was only 65
mW at 300 0C, and the back-propagation algorithm was applied as the
supervised learning rule. The recognition probability of the neural-
network was 100% for the discrimination of the gases and concentrations
used in the work.
Next, Mitra et al., [92] investigated the gas response of chemically depos-
ited ZnO films using a sodium zincate bath. The ZnO films prepared by
this method were highly resistive, signifying the presence of a large den-
sity of oxygen adsorbed acceptor – like trap states (O2− , O
-, etc.). Prelim-
inary studies of gas sensing characteristics, performed at 150 – 375 0C
temperature range, indicated that the Pd – sensitized ZnO films respond
strongly to 1 vol% H2, which is on the lower side of the hazardous explo-
sion range for H2 (4 –75 %). The ZnO film – based sensors exhibited
excellent sensitivity (more than 99%) at 200 0C and the response time
was reasonably rapid.
Nunes et al., [93] reported a maximum value of sensitivity for zinc oxide
(ZnO) thin films sensor of high electrical resistivity and low thickness
68
upon being exposed to methane (CH4), hydrogen (H2) and ethane (C2H6)
reductive gases. The sensitivity of the sensor increased with operating
temperature and its highest values were obtained at 200 ºC for methane
and hydrogen, while it occurred at 100 ºC for the test with ethane. More-
over, the sensitivity of the ZnO thin films changed linearly with the in-
crease of the gas concentration. The ZnO sensor demonstrated low selec-
tivity since it detects the presence of several gases. The increase of the
selectivity could be promoted by the use of an appropriated catalytic
metal such as Pd or Au.
Roy and Basu [94] explored the selectivity, towards dimethylamine
(DMA) (CH3)2NH and H2 test gases at different temperatures, of a good
quality undoped ZnO films deposited on glass and quartz by a novel CVD
technique using a 0.5 M zinc acetate as the starting solution. They ob-
tained faster response and higher sensitivity. The operating temperature
played a key role in the selectivity of such sensors, with the optimum
operating temperature being 300 0C.
B. Licznerski [19] observed that thick – film gas sensors based on semi-
conductive tin dioxide are suitable for detection of explosive and toxic
gases and vapours. Sensitivity, selectivity and stability of sensors working
in different temperature depend on the way the tin dioxide and additives
were prepared. A construction also plays an important role. He presented
an attitude towards the evaluation of transport of electrical charges in
semiconductive grain layer of SnO2, when dangerous gases appeared in
the surrounding atmosphere.
Bârlea et al., [8] characterized SnO2 metal oxide gas sensors deposited on
a ceramic cylinder heated to the functioning temperature (100 – 400 °C),
and exposed it to an atmosphere containing a reducing gas: carbon mon-
69
oxide (CO), liquid petroleum gas (LPG, mainly butane) and methane gas.
They investigated the influence of the supplied electric current to the
sensor sensitivity and the response time and recovery time. The electric
resistance of the semiconducting material dramatically modified, even at
very low gas concentrations. The response time was longer for lower
supply voltage (many seconds, even minutes) and became very short
(under 1 second) at the greatest voltages.
Gas sensing with fast response and recovery times is at the forefront of
gas sensing characteristic parameters particularly when it comes to
providing early alert of inflammable and toxic gas leaks. Several re-
searches are concerned on optimizing this parameter.
Xu et al., [43] demonstrated an ultrafast, ultrasensitive hydrogen gas
sensor based on self – assembled monolayer promoted 2 – dimensional
palladium film on glass substrate. Their enhanced sensor was sensitive
enough to detect hydrogen levels as low as 25 ppm with a fast response
time in tens of milliseconds (~70 ms) upon exposure to 2% H2, a con-
centration below the hydrogen explosion limit range of 4 – 75% for effec-
tive alarming.
Chou et al., [95] investigated the structural and sensing properties of
ZnO:Al films as an ethanol vapor gas sensor obtained by RF magnetron
sputtering on Si substrate using Pt as interdigitated electrodes. The struc-
tural characteristics revealed that flat and well – defined columnar films
with c – axis textured were formed. The film exhibited good sensitivity to
the ethanol vapors with quick response – recovery characteristics, and it
was found that the sensitivity for detecting 400 ppm ethanol vapor was
~20 at an operating temperature of 250 °C. The high sensitivity, fast re-
70
covery, and reliability implied that ZnO:Al seemed to be a promising
semiconducting material for the detection of ethanol vapor.
Patil et al., [41] explored the characterization and ethanol gas sensing
properties of pure and doped ZnO thick films prepared by the screen
printing technique. Pure zinc oxide was almost insensitive to ethanol.
Thick films of (1 wt%) Al2O3 – doped ZnO were observed to be highly
sensitive to ethanol vapours at 300 °C. Doping Al2O3 into zinc oxide
created surface misfits and since it is reported that the surface misfits,
calcination temperature and operating temperature could affect the micro-
structure and gas sensing performance of the sensor. The sensor showed
very rapid response and recovery to ethanol vapours. Moreover, it had
good selectivity to ethanol against LPG, NH3, CO2, Cl2 and H2 at 300 °C.
Mitra & Mukhopadhyay [96] studied the methane (CH4) sensitivity, re-
sponse time and recovery process of a chemically fabricated ZnO thin
film semiconducting layer and a Pd – catalyst layer coated on the surface
of the semiconducting ZnO. The sensor exhibited reasonable sensitivity
of about 86% to 1vol. % methane (CH4) in air at 200 0C optimum operat-
ing temperature.
The effect of film thickness on sensor performance was investigated by
Liewhiran and Phanichphant [42]. They fabricated ZnO thick films using
flame spray pyrolysis (FSP) on Al2O3 substrate interdigitated with Au
electrodes with various thicknesses (5, 10, 15 μm). The gas sensing char-
acteristics to ethanol (25 – 250 ppm) evaluated as a function of film
thickness at 400 °C in dry air, displayed the tendency of the sensitivity to
decrease with increasing film thickness and response time; with the thin-
nest sensing film (5 μm) showed the highest sensitivity and the fastest
response time (to 250 ppm, S=801, τres =5 s). They discussed the behavior
71
on the basis of diffusively and reactivity of the gases inside the oxide
films. The sensing characteristics were deteriorated evidently with in-
creasing film thicknesses. The recovery times were quite long within
minutes.
Aroutiounian et al., [97] reported hydrogen sensors working at and close
to room temperature and made of porous silicon covered by the TiO2-x or
ZnO<Al> thin films. The sensitivity of manufactured structures to 1000
– 5000 ppm H2 showed the possibility of realizing a durable, highly sen-
sitive and selective hydrogen sensor within its lower and upper explosion
limits of 4 – 75% by volume. The sensor had a relatively short time of
response and recovery (~20 s). Such sensors could also be a part of a
silicon integral circuit. The same research group grew Aluminum –
doped ZnO nano – size films on glass ceramic substrates by high – fre-
quency magnetron sputtering method [99]. Pt layer and gold interdigitat-
ed ohmic contacts were evaporated on the prepared films by the ion –
beam sputtering method. Sensitivity measurements in the temperature
range 40 −100 0C to different concentrations of hydrogen (1000 − 5000
ppm) in air was investigated. The glass ceramic/ZnO<Al>/Pt structure
showed sufficient sensitivity to hydrogen at the pre-heating of the work-
ing body already up to 40 0С.
The grain or crystallite size of the sensing element is one of the most
important factors affecting sensing properties, especially sensitivity. The
gas response to different gases is related to a great extent to the surface
state and morphology of the material [13]. The use of micro and
nanostructured films is advantageous due to the high surface to volume
ratio the nanostructure exhibits, offering faster response and higher gas
sensitivities to low concentration of the tested gases and eliminating the
need of operation at elevated temperatures. Some groups have explored
72
nano grain – sized ZnO structures and tested it for gas sensitivity. Mat-
thew Szeto [98] fabricated ZnO chemical gas sensors from prepared ZnO
nanoplatelets and their sensitivities to H2 gas were investigated under
conditions of varying concentration, sensor temperature, and intensity of
UV light. It was found that at room temperature and a source voltage of
5V that the ZnO sensor had the best sensitivity of greater than 85% to H2
gas at 50 ppm and was most sensitive in the absence of UV radiation. At
130 0C temperature, the ZnO sensor showed sensitivities near 100% to
~50 ppm of H2, where it both responded and recovered faster. Memory
effect previously observed was also non – existent at temperatures near
and above 100°C. The sensor was also observed to be both slower in
response and lower in sensitivity in the absence of UV light in a darkened
chamber.
Savu et al., [16] deposited a Porous nano and micro crystalline tin oxide
films by RF Magnetron Sputtering and doctor blade techniques, respec-
tively. Electrical resistance and impedance spectroscopy measurements,
as a function of temperature and atmosphere, were performed in order to
determine the influence of the microstructure and working conditions
over the electrical response of the sensors. The conductivity of all sam-
ples increases with the temperature and decreases in oxygen, as expected
for an n-type semiconducting material. The improved sensitivity and
response times at the 200 °C working temperature are due to the higher
rate of gas adsorption/desorption. The impedance plots indicate the exist-
ence of two time constants related to the grains and the grain boundaries.
The Nyquist diagrams at low frequencies reveal the changes that took
place in the grain boundary region, with the contribution of the grains
being indicated by the formation of a second semicircle at high frequen-
cies. The better sensing performance of the doctor bladed samples can be
73
explained by their lower initial resistance values, bigger grain sizes and
higher porosity.
Also, Korotcenkov et al., [36] reviewed the pioneering influence of mor-
phological and crystallographic structural parameters (i.e., film thickness,
grain size, agglomeration, porosity, faceting, grain network, surface ge-
ometry, and film texture) on the gas sensor main analytical characteristics
(absolute magnitude and selectivity of sensor response (S), response time
(τres), recovery time (τrec), and temporal stability). A comparison of stand-
ard polycrystalline sensors and sensors based on one – dimension struc-
tures was conducted. The structural parameters of metal oxides were
found to be the most important factors for controlling response parame-
ters of resistive type gas sensors. Thus, it was shown that the decreasing
of thickness, grain size and degree of texture was the best way to decrease
time constants of metal oxide sensors. However, it was concluded that
there is no universal decision for simultaneous optimization of all gas-
sensing characteristics. One have to search for a compromise between
various engineering approaches because adjusting one design feature may
improve one performance metric but considerably degrade another.
Kiriakidis et al., [100] exhibited highly porous ZnO films with character-
istic c-axis columnar growth structure deposited on glass substrates in a
home-made aerosol spray pyrolysis system at 350 0C. Their sensing re-
sponse to low ozone concentrations was evaluated at room temperature.
Films have shown to produce clear response signals to ozone concentra-
tions as low as 16 ppb with a response time of 1 min, demonstrating the
potential of applying these films as sensing elements in future metal ox-
ide gas sensing devices.
74
Al-hardan et al., [101] prepared ZnO thin films by thermal oxidation of
Zn metal at 400 0C for 30 and 60 min. The XRD results showed that the
Zn metal was completely converted to ZnO with a polycrystalline struc-
ture. The sensors had a maximum response to H2 at 400 0C and showed
stable behavior for detecting H2 gas in the range of 40 to 160 ppm. Film
oxidized for 60 min in oxygen flow exhibited higher response than that of
the 30 min oxidation which was approximately 4000 for 160 ppm H2 gas
concentration. The sensor with higher resistance yields higher response to
the gas under test. The sensing mechanism was modeled according to the
oxygen – vacancy model.
Tamaekong et al [102] investigated the gas sensing properties toward
hydrogen (H2) of ZnO nanoparticles doped with 0.2 – 2.0 at. % Pt which
were successfully produced in a single step by flame spray pyrolysis
(FSP) technique. ZnO nanoparticles paste was coated on Al2O3 substrate
interdigitated with gold electrodes to form thin films by spin coating
technique. The gas sensing properties toward hydrogen gas revealed that
the 0.2 at. % Pt/ZnO sensing film exhibited an optimum H2 sensitivity of
~164 at hydrogen concentration in air of 1 volume % at 300 0C and a low
hydrogen detection limit of 50 ppm at 300 0C operating temperature.
Al-hardan et al., [103] synthesized undoped and 1 at. % chromium (Cr) –
doped ZnO by RF reactive co-sputtering for oxygen gas sensing applica-
tions. The prepared films showed a highly c – oriented phase with a dom-
inant (002) peak at a Bragg angle of around 34.28 o. The Cr – doped ZnO
sensor has been shown to have a lower operating temperature of around
250 0C and enhanced sensitivity than previously reported. Good stability
and repeatability of the sensor were demonstrated when tested under
different concentration of oxygen atmosphere. The enhancement was
likely attributed to the higher oxidation state of the chromium.
75
Also, Al-hardan et al., [104] investigated the mechanism of hydrogen
(H2) gas sensing in the range of 200 – 1000 ppm of RF-sputtered ZnO
films. The I – V characteristics as a function of operating temperature
proved the ohmic behavior of the contacts to the sensor. The complex
impedance spectrum (IS) of the ZnO films exhibited a single semicircle
with shrinkage in the diameter as the temperature increased and as the
hydrogen concentration was increased in the range from 200 ppm to 1000
ppm.
One month later, Al-hardan et al. [12] studied the gas sensing properties
of RF reactively sputtered ZnO thin film towards volatile organic com-
pounds VOC in which the sensitivity of the sensor was the highest
( ~ 100 ) for 500 ppm acetone in comparison to that of isopropanol
and ethanol. An optimum operating temperature for maximum sensitivity
of 400 0C for the above vapors was obtained. The sensor showed a stable,
reversible and repeatable behavior in the acetone concentration of 15 up
to 1000 ppm. They explained the sensing mechanism in accordance with
the ionosoption model. The same ZnO based sensor also exhibited good
sensitivity for vinegar test in the concentration range of 4% to 9% and the
maximum sensitivity to vinegar test application was obtained at 400 0C.
The work revealed the validity of using ZnO gas sensor in estimating the
acid concentrations of the vinegars for food requirements [106].
Hussain et al., [105], grew ultra – fine thin films of pure and SnO doped
ZnO nanosensor on gold interdigitated ceramic substrate by ultrasonic
aerosol assisted chemical vapor deposition technique (UAACVD) at
around 450 °C temperature and under 5 Pa oxygen atmospheric pressure.
Both doped and undoped ZnO thin films sensing characteristic measure-
ments verified the nanosensor suitability for detecting ethanol vapor at a
temperature range of 60 – 150 °C. At room temperature (25 °C), the re-
76
sponse and recovery time of the sensor increased many orders of magni-
tude compared to 60 °C. Sensitivity of the ZnO sensor demonstrated
linear dependence with the increase of gas concentration. 1 % SnO dop-
ing of ZnO enhanced the sensitivity of the film drastically and thus im-
proved its detecting efficiency.
Later, Al-zaidi et al., [107] demonstrated spray – pyrolyzed palladium –
doped ZnO thin film deposited on glass substrate to be a fast hydrogen
gas sensor. The prepared ZnO films were doped by dipping in palladium
chloride PdCl2. Sensitivity dependence on the temperature and test gas
concentration was tested and the optimum operation temperature was
determined at around 280 oC. The response time of 2-3 s of the doped
ZnO film was so fast to detect flammable H2 leaks well below the lower
explosion limit (LEL) of 4%.
Chen et al., [118] successfully prepared tin dioxide SnO2 thin films with
interesting fractal features by pulsed laser deposition techniques under
different substrate temperatures. Tin oxide is a unique material of wide-
spread technological applications, particularly in the field of environ-
mental functional materials. New strategies of fractal assessment for tin
dioxide thin films formed at different substrate temperatures are of fun-
damental importance in the development of microdevices, such as gas
sensors for the detection of environmental pollutants. Fractal method
was applied to the evaluation of this material. The measurements of car-
bon monoxide gas sensitivity confirmed that the gas sensing behavior
was sensitively dependent on fractal dimensions, fractal densities, and
average sizes of the fractal clusters. The random tunneling junction net-
work mechanism was proposed to provide a rational explanation for this
gas sensing behavior. The formation process of tin dioxide nanocrystals
and fractal clusters could be reasonably described by a novel model.
77
In spite of the many researches on the as – deposited and doped
metal oxide ZnO and SnO2 – based gas sensors prepared via different
depositions techniques, there is still a need to obtain simple, cost – effec-
tive, sensitive and fast response gas sensor towards inflammable hydro-
gen reducing gas. Thus, in this work the fabrication and sensing charac-
teristics of undoped and Pd – doped ZnO and SnO2 semiconductor metal
oxides by chemical spray pyrolysis deposition are presented. The effect
of Pd doping on the sensitivity S, operating temperature and the response
time of both metal oxides gas sensing elements to hydrogen (H2) gas of
different concentrations is examined.
78
Chapter 3
Experimental Procedure
Introduction
The present chapter gives a detailed account of the work carried
out for the development of Zinc Oxide and tin oxide thin film based gas
sensors. The different steps followed in this regard are described. Chemi-
cal spray pyrolysis deposition on glass and silicon substrates, which is
used in the present work, is discussed in details. Following it is a discus-
sion on the characterization of the deposited gas sensitive ZnO and SnO2
materials. Details of the experimental set up made for testing and study-
ing the performance of the developed gas sensors is also presented in this
chapter. The results obtained and analyses of data on the performance of
the sensors fabricated are presented at the next chapter.
3.1 Gas Sensor Fabrication
The schematic conception of a typical simple metal oxide gas sen-
sor is illustrated in figure 3.1 below.
The different steps that have been followed for the realization of a
semiconductor metal oxide, ZnO or SnO2, gas sensor are outlined below:
Selection of substrate
Figure 3.1: Schematic of a typical gas sensor structure. Thicknesses are not to scale.
ZnO film (150 nm)
Electrode: Pt (200 nm)/ Ta (25 nm)
film
Insulation layer: SiO2 layer (1 μm)
Substrate: Si wafer
H2 H2
H2
H2
SiO2 layer
Pt electrode
Si wafer
ZnO film
Electrical Measurement
79
Substrate cleaning procedure
Deposition of gas sensitive thin film
Surface sensitization of the prepared thin film by palladium
noble metal catalyst.
Deposition of Al interdigitated electrodes IDE on the sensi-
tive film and attachment of leads for electrical measurement.
Fabrication of gas sensor testing system.
The substrate refers to the base on which the gas sensing material is de-
posited. A substrate used for gas sensing application should ideally be
[18]:
Good conductor of heat: The ability of a material to conduct heat is
quantified by either thermal conductivity or thermal diffusivity,
thereby, determining the power consumption of the sensor.
Electrically insulating: This is to ensure that the electrons generat-
ed due to gas – solid interaction are not being grounded by a con-
ducting substrate.
Rugged.
Stable and inert in measurand environment.
Inexpensive.
Capable of influencing the microstructure favorably (porous films
with granular microstructure).
Render itself suitably for cleaning.
The sensitivity of gas sensors depends mainly on the grain size of
the gas sensitive material [36]. The grain size of a given material on a
substrate is known to depend on the wettability of the substrate. Sub-
strates with lower surface tension are believed to result in smaller grain
size. Glass, having the lowest surface tension, resulted in the smallest
80
ZnO thin film grain size [18]. The large grain size of ZnO particles de-
posited on sintered alumina is probably due to of the rough topography
exhibited by the substrate itself.
3.2 Spray pyrolysis experimental set up
Chemical techniques for the preparation of thin films have been
studied extensively because such processes facilitate the designing of
materials on a molecular level. Spray pyrolysis, one of the chemical tech-
niques applied to form a variety of thin films, results in good productivity
from a simple apparatus. In the current research, zinc oxide thin films are
deposited on glass substrates employing locally – made spray pyrolysis
deposition chamber whose main components set up is illustrated in the
schematic diagram of figure 3.2. It is essentially made up of a precursor
solution, carrier gas assembly connected to a spray nozzle, and a tempera-
ture – controlled hot plate heater.
The atomizer, illustrated in the photo plate 3.1-B, has an adjustable
copper capillary tube nozzle of 0 - 0.8 mm inner diameter clamped to a
holder and supported by a metal tripod. The nozzle is driven by a com-
pressed atmospheric air. The prepared precursor solution is pumped
through the metal nozzle with a solution flow rate ranging from 1 to 2
mL/min. Due to the air pressure of the carrier gas; a vacuum is created at
the tip of the nozzle to suck the solution from the tube after which the
spray starts [63]. To regulate spraying time, a 16 – Bar Tork solenoid
valve controlled by an adjustable timer has been incorporated. The atom-
izer and the 1500 Watts hot plate heater are enclosed in a 1 1
1 𝑚3 ventilation hood, photo plate 3.1-A. A 220 V a.c. power was ap-
plied to the heater and temperature was measured using a type K (nickel-
chromium) thermocouple and precision digital temperature controller
(GEMO DT109 photo plate 3.1-C).
81
3.3 Precursor solution
A 0.2 M concentration precursor solution of zinc acetate dihydrate
Zn(CH3COO)2.2H2O (molecular weight 219.4954 g/mole) has been pre-
pared by dissolving a solute quantity of 4.389908 g of
Zn(CH3COO)2.2H2O (as weighed by a 10−4 g - precision balance) in 100
mL isopropyl alcohol C3H9O (the solvent). A magnetic stirrer is incorpo-
rated for this purpose for about 10 – 15 minutes to facilitate the complete
dissolution of the solute in the solvent. Furthermore, aqueous precursor of
Zinc chloride ZnCl2 (molecular weight 136.3146 g/mole) dissolved in
distilled water has also been employed in getting ZnO thin films. Organic
0 Substrate
Sprayer
Holder with
stand
Spray
cone
Air Nozzle
Substrate heater
Capillary Tube
Compressed
Air Tube
Thermocouple
Temperature
Controller
30 cm
Measuring
Cylinder
Ventilation Fan
Solenoid Valve
And Timer
0450°C
Air in
Precursor
Solution
Figure 3.2: Spray pyrolysis experimental set up
82
solvents are preferable over distilled water because the former enables the
attainment of homogeneous, highly – transparent, thin films of small
grain size [71].
Prior to depositing the films, the substrates, which are commercial
glass slides of 76×25×1 mm3 dimensions, are firstly cleaned by dipping
in distilled water to remove the dust and then are ultrasonically cleaned in
methanol for about 10 min. Finally they are soaked in distilled water,
dried, and polished with lens paper. The pretreatment of the substrates is
carried out to facilitate nucleation on the substrate surface. Presence of
contamination on the substrate surface is one of the reasons of the ap-
pearance of pinholes and film inhomogeneity [71].
Photo plate 3.1: A: experimental set up of the spray pyrolysis deposition SPD. B: Air atomiz-
er. C: Gemo DT109 temperature controller, and D: Digital balance with the magnetic stirrer.
B
Needle
in nozzle
Air exit
A
C D
83
The spray rate is usually in the range 2 – 3 mL min-1
. The optimum
carrier gas pressure for this rate of solution flow is around 5 kg cm-2
. At
lower pressures, the size of the solution droplets becomes large, which
results in the presence of recognized spots on the films and then reduction
of transparency. This situation increases the scattering of light from the
surface and then reduces the transmittance of the films.
The spray pyrolytic substrate temperature is maintained within 450
± 5 °C during the deposition. Film thickness is controlled by both the
precursor concentration and the number of sprays, or alternatively, spray-
ing time. Thus, a 4 – second spray time is maintained during the experi-
ment. The normalized distance between the spray nozzle and substrate
was fixed at 30 cm. Table 3.1 summarizes the optimized thermal pyroly-
sis deposition conditions for the preparation of ZnO thin films that were
employed in the current research.
3.4 The determination of film thickness
The thickness of the films is determined using a micro gravimet-
rical method. The films deposited on clean glass slides whose mass had
previously been determined. After the deposition, each substrate itself is
weighted again to determine the quantity of deposited ZnO. Measuring
Spray parameters Values
Concentration of precursor 0.2 M
Volume of precursor sprayed 100 mL
Solvent isopropyl alcohol
Substrate temperature 450 0C
Spray rate 2.3 mL/min.
Carrier gas pressure 1 bar
Nozzle-substrate distance 30 cm
Table 3.1.: Optimum thermal spray pyrolysis deposition conditions for the preparation
of ZnO thin films.
84
the surface area of the deposited film, taking account of ZnO specific
weight of the film, the thickness is determined using the relation:
=∆mZnO
A ∙ ρ (3.1)
where A is the actual area of the film in cm2, ∆mZnO
is the quantity of
deposited zinc oxide, and ρ is the specific weight of ZnO. Film thickness
was also confirmed and verified from cross sectional SEM image.
3.5 Surface modification of ZnO by palladium noble metal
Metal oxide gas sensors need a catalyst deposited on the surface of
the film to accelerate the reaction and to increase the sensitivity, impart
speed of response and selectivity [41].
Small amounts of noble metal additives, such as Pd or Pt are com-
monly dispersed on the semi conducting as activators or sensitizers to
improve the gas selectivity, sensitivity and to lower the operating temper-
ature [41, 48]. Many methods have already been tested for this purpose,
for example bulk doping during calcination, sol-gel technology, spray
pyrolysis deposition, thermal evaporation, CVD, laser ablation, magne-
tron sputtering, impregnation by salt solution. With the help of these
methods, it was possible to form on a surface of metal oxides surface
clusters of various components with sizes from 0.1 to 8 nm [16].
For the above reasons, the surface of the deposited ZnO thin films
were catalyzed using successive multiple dipping (or spraying) of the
prepared samples with a 50 ccm (0.0564 Molar) solution made up of
dissolving 1% by weight palladium chloride PdCl2 (Mwt.= 177.3256
g/mole) in ethanol alcohol C2H5OH. Each sample was successively
sprayed 10, 15, 20, 25, and 30 times of 4s spray interval and at 400 de-
grees hot plate heater temperature. Eventually, the sensitized samples
85
were heat – treated at the same temperature for a period of one hour in
atmospheric air. About 20 – time palladium spray or dipping was found to
be optimum for fast and sensitive zinc oxide H2 gas sensor.
3.6 Al Interdigitated Elecrtodes (IDE)
Figure 3.3 illustrates a schematic diagram of the thermal evapora-
tion system (Edward type E306A unit) which is used to thermally evapo-
rate the aluminum electrodes layer on the ZnO sample via the metal
Variable power
supply transformer
Boat
Valve
Pirani gauge
Valve
Rotary pump
Diffusion pump
High vacuum
Valve
Penning
gauge
Bell jar
Substrate holder
Bus bars
Cylindrical
shield
Ring shield
Shutter
Figure 3.3: Vacuum system for the vaporization from resistance – heated sources.
When replacing the transformer and heater with an electron gun, vaporization by
means of an electron beam occurs.
Glow ring Thickness
monitor
86
mask. Figure 3.4 illustrates two 8 and 10 – finger interdigitated electrode
IDE metal masks which were utilized in this work. The samples were
fixed in the evaporation system. The thickness t of the evaporated alumi-
num electrode was estimated using the following formula:
=m
2πR2ρ (3.2)
Where R is the separation distance of the tungsten boat to the substrate
holder, 𝜌 represents the density of aluminum (specific gravity of 2.7) and
22 mm
19 mm
1 mm
2 mm
3 mm
2 mm
14 mm
0.4 mm
0.4 mm
3 mm
3 mm
0.4 mm
13.6
mm
2 mm
3 mm
10 mm
15 mm
25 mm
2 mm
2 mm
Figure 3.4.: A schematic diagram of the IDE masks utilized in this work.
87
m is the mass of Al used during evaporation.
3.7 Gas sensor testing system
A schematic cross sectional view of the gas sensor testing system,
test chamber and photos of the mounted sensor and test chamber are illus-
trated schematically in figure 3.5 and in the photograph plate 3.2 respec-
tively. The unit consists of a vacuum – tight stainless steel cylindrical test
chamber of diameter 163 mm and of height 200 mm with the bottom base
made removable and of O – ring sealed. The effective volume of the
chamber is 4173.49 cc; it has an inlet for allowing the test gas to flow in
and an air admittance valve to allow atmospheric air after evacuation.
20 cm
16.3 cm
8 – pin feed through
Output to
vacuum
pump
Test gas in
Gas Manifold 2 cm
O –ring seal
V A
Ω
Needle Valve
Vacuum gage
3 mm
Auxiliary inlet
436
450
Gas
Flow meter
ZnO Sensor
PC – interfaced
DMM
Temp. Controller
Exhaust
USB
Cable
Air Flow
meter
Hydrogen Air
Relief
valve
Vacuum Pump
Figure 3.5: Gas sensor testing system
88
Another third port is provided for vacuum gauge connection.
A multi – pin feed through at the base of the chamber allows for
the electrical connections to be established to the heater assembly as well
as to the sensor electrodes via spring loaded pins [15].
The heater assembly consists of a hot plate and a k – type thermo-
couple inside the chamber in order to control the operating temperature of
the sensor. The thermocouple senses the temperature at the surface of the
film exposed to the analyte gas. The PC – interfaced multi meter, of type
UNI-T UT81B, is used to register the variation of the sensor conductance
(reciprocal of resistance) exposed to predetermined air – hydrogen gas
mixing ratio. The chamber can be evacuated using a rotary pump to a
rough vacuum of 1 10−3 ba . A gas mixing manifold is incorporated to
Photo plate 3.2.: A photo of the sensor testing system
Testing Chamber
H2 gas
Vacuum Pump
Air Supply
Temp Controller
DC Power
Supply
UNI-T81 DMM
Gas Flow
Meter
Gas Needle
Valve
Pressure Gauge
UNI-T81 DMM
89
control the mixing ratios of the test and carrier gases prior to being inject-
ed into the test chamber. The mixing gas manifold is fed by zero air and
test gas through a flow meter and needle valve arrangement. This ar-
rangement of mixing scheme is done to ensure that the gas mixture enter-
ing the test chamber is premixed thereby giving the real sensitivity.
3.8 Sensor testing protocol
The following is the protocol used in the operation of the test set-
up.
The test chamber is opened and the sensor placed on the heater.
The necessary electrical connections between the pin feed through
and the sensor spring loaded pins and the thermocouple are made.
Doing so, the test chamber is closed.
Then, the rotary pump is switched on to evacuate the test chamber
to approximately 1 10−3 ba . Setting the sensor desired operat-
ing temperature is done using the PID temperature controller.
After that, using the needle valves the flow rate of the carrier and
test gases flow meters is adjusted.
Next, the gas of known concentration in mixing chamber is al-
lowed to flow to the test chamber by opening the two-way valve.
Measurement of the current variation of the sensor for the known
concentration of test gas mixing ratio is observed by the PC – inter-
faced digital multimeter DMM.
After the measurement, the needle valve of the test gas is closed to
allow the sensor to recover to the base line current value I0.
90
The above measurements are repeated for the other required tem-
peratures and/or concentrations of the test gas.
The process of achieving a known concentration of test gas for
measurement is described below:
The flow meter connected to zero air cylinder is set to a known value (say
1000 sccm) using the needle valve. Then, the flow meter connected to the
test gas is set to the required value to achieve the desired concentration.
For example if 1000 ppm (0.1%) of test gas is required, the flow rate of
test gas is set to 1 sccm while keeping the flow rate of zero air at 1000
sccm.
A schematic diagram of the electrical circuit used for gas sensor
measurements is illustrated in figure 3.6. When the sensor is connected as
shown in the basic circuit, output across the load resistor (VRL) increas-
es/decreases as the sensor's resistance (RS) decreases/increases, depend-
ing on the analyte gas concentration and its type; i.e., whether it is a re-
ducing or oxidizing gas. A DC power supply feeds an adjustable bias
voltage Vb from 0 to 15 volts across the sensor resistance RS and the cor-
responding current of the circuit is measured via the digital multimeter
DMM whose signal is directly being interfaced to the PC for further anal-
ysis. The hot plate heater of the sensor is supplied by a 220 V A.C voltage
controlled by the GEMO DT109 PID temperature controller (not shown
in the figure) together with its k – type (nickel-chromium) thermocouple.
The same above electrical circuit was exploited to investigate the I – V
characteristic of the sensitive sensing material at various temperatures
both in pure air and in gas – containing atmosphere.
91
3.9 Crystalline structure of the prepared ZnO thin films
The crystalline structure is analyzed by a SHIMADZU 6000 X-ray
diffractometer (illustrated in photo plate 3.3) using Cu K𝛼 radiation
(1.5406 Å) in reflection geometry. A proportional counter with an operat-
ing voltage of 40 kV and a current of 30 mA is used. XRD patterns are
recorded at a scanning rate of 0.08333° s-1
in the 2𝜃 ranges from 20° to
60°.
PC – interfaced
DMM
Figure 3.6: A schematic diagram of the gas sensor basic measurement elec-
trical circuit.
RL
RS RH
A
220 V AC DC
Power
Supply
0 -15 V
Gas
Vb
92
3.10 Thin film surface topography
The surface topography is analyzed with Ultra 55 scanning electron
microscope SEM from ZEISS, with its photo plate 3.4 illustrated below.
Also, it is employed for thin film thickness measurement.
The morphological surface analysis is carried out employing an
atomic force microscope, AFM, (AA3000 Scanning Probe Microscope
SPM, tip NSC35/AIBS) shown in photo plate 3.5, from Angstrom Ad-
vance Inc.
3.11 Optical properties
The optical properties are examined via Optima sp-3000 plus UV-
Vis-NIR (Split-beam Optics, Dual detectors) spectrophotometer equipped
with a xenon lamp. Photo plate 3.6 illustrates this spectrophotometer.
Photo plate 3.3.: LabX XRD – 6000 Shimadzu diffractometer unit.
93
Photo plate 3.4.: Ultra 55 SEM unit from ZEISS.
Photo plate 3.5.: AA3000 Scanning Probe Microscope SPM, tip NSC35/AIBS), from Ang-
strom Advance Inc.
94
3.12 Tin oxide (SnO2) hydrogen gas sensors
In addition to the zinc oxide – based hydrogen gas sensor, and for
comparison purposes, undoped and palladium doped tin oxide thin films
have been prepared on glass substrates using the same deposition/doping
procedures described previously for ZnO thin films. The 0.2 – M concen-
tration spraying precursor is obtained by dissolving 4.5126 g stannous
chloride dihydrate SnCl2.2H2O (molecular weight 225.63 g/mole) solute
in 100 mL isopropyl alcohol C3H9O solvent. Likewise, the dissolution
process is facilitated by a magnetic stirrer for10 minutes. The structural,
optical and sensing properties of the prepared films are studied to reduc-
ing hydrogen gas environments at different operating temperatures and H2
gas mixing ratios.
Photo plate 3.6.: Optima sp-3000 plus UV-Vis-NIR spectrophotometer.
95
Chapter 4
Results and discussion
Introduction
In the preceding chapter, experimental setup and methods are de-
scribed. In this chapter, the detailed experimental results and sensing
performance characteristics of the ZnO and SnO2 thin films to hydrogen
gas will be presented.
Some important factors of sensor characteristics will be investi-
gated here. These include sensitivity, response and recovery time, the
optimum operating temperature, and gas concentration. Other characteris-
tics that we didn't measure such as selectivity, stability, repeatability,
resolution, and hysteresis, will not be covered.
We start with a comprehensive outline of ZnO thin film deposition,
its crystalline structure; optical and electrical properties. Surface mor-
phology characteristics will be examined as well. The sensitivity, transi-
ent response, and temperature effects will be analyzed in details for both
metal oxides sensing elements. At the end, results discussion of the sen-
sor performance will be outlined.
4.1 ZnO thin film deposition
Initially, zinc oxide thin films are obtained by spray pyrolytic de-
composition of 0.1 M zinc chloride ZnCl2 aqueous solution precursor.
Later on, 0.2 M precursor of zinc acetate dehydrate Zn(CH3COO)2.2H2O
dissolved in water, organic solvent, and mixture of both, are also used in
the realization of zinc oxide thin films. Spraying temperature, the crucial
parameter, is varied between 370 and 500 oC with the optimum spraying
temperature being around 450 oC. Table 4.1 outlines the optimum deposi-
tion conditions employed in the current research.
96
Figure 4.1: A photo of spray pyrolyzed ZnO thin film on glass samples
Zinc chloride aqueous precursor
Zinc acetate aqueous precursor
Figure 4.1 illustrates a zinc oxide thin film photos of 0.2 M zinc chloride
and zinc acetate dehydrate aqueous precursors sprayed on glass substrates
at 450 0C temperature.
The above ZnO thin films have a thickness of the order of 950 –
1200 nm as estimated by the weighing method and verified in figure 4.2
with cross sectional view of the scanning electron microscope SEM im-
age from which film thickness is estimated to be 1541 nm. Moreover,
film thickness is calculated using an interference method, a procedure
Spray parameters Values
Concentration of precursor 0.2 M
Volume of precursor sprayed ~100 mL
Solvent Isopropyl Alcohol
Substrate temperature 450 0C
Spray rate ~2.3 mL/min.
Carrier gas pressure 1 bar
Nozzle – substrate distance 30 cm
Table 4.1: spray pyrolysis deposition optimum parameters.
97
developed for calculating the thickness of thin films is given elsewhere
[108].
Two masks of different fingers spacing (1 mm and 0.4 mm) were
used to evaporate the Al interdigitated electrodes IDE and figure 4.3
Figure 4.3: 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
A B
Figure 4.2: Scanning Electron Micrograph photo of spray pyrolyzed ZnO thin
film on glass
98
shows two of these electrodes after being evaporated over the ZnO thin
film layer deposited on glass and silicon substrates.
4.2 Crystalline structural properties of the ZnO thin film
The structure and lattice parameters of undoped and Pd – doped
ZnO films are analyzed by a LabX XRD 6000 SHIMADZU XR – Dif-
fractometer with Cu Kα radiation (wavelength 1.54059 Å, voltage 30 kV,
current 15 mA, scanning speed = 4 °/min) as illustrated in figure 4.4 and
the effect of the Pd dopant on the structure of the film is displayed in
figure 4.5. Diffraction pattern spectra are obtained with 2𝜃 starting from
20 ° to 50 ° at 6 ° glancing angle. In both the as – deposited and Pd –
doped ZnO thin films, the X – ray diffraction spectra possess one sharp
and three small peaks. It means that the film is polycrystalline with crys-
tal planes (100), (002), (101) and (102). The film is crystallized in the
hexagonal wurtzite phase and presents a preferential orientation along the
c – axis indicated by the plane (002). The result is in a good agreement
with data mentioned in the literature (JCPDF card no 36-1451) [109]. The
strongest peak, observed at 2𝜃 = 34.3646 ° (d = 0.260 nm), can be at-
tributed to the (002) plane of the hexagonal ZnO. Another major orienta-
tion present is (101) observed at 2𝜃 = 36.1732 °. The other orientations
like (100) and (102) at 2𝜃 = 31.7013 ° and 47.4424 °, respectively are
also seen with comparatively lower intensities. Therefore, the crystallites
are highly oriented with their c – axes perpendicular to the plane of the
substrate. It is worth to mention here that a small intensity peak appears
in the Pd – doped film at 2𝜃 = 40.32 o which belongs to the plane (111) of
the palladium. The lattice constants: a = 3.24982 Å, c = 5.20661 Å. The c
– axis lattice constant of the ZnO thin film was calculated from XRD data
as 5.20 nm. This value is consistent with the one obtained by Gumu et al
[66]. The (002) peak full width at half maximum (FWHM) is 0.1958 0.
99
0.1958 °., while 2𝜃and d values are given in Tables 4.2 and 4.3, respec-
tively.
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50
I [
CP
S]
Theta - 2Theta [Degree]
(002)
(101)
(102) (100)
XRD 6000 SHIMADZU XR-Diffractometer
Figure 4.4: XRD crystal structure of as deposited ZnO thin film (thickness =1178 nm)
prepared from 0.1 M Zinc Chloride aqueous precursor on glass substrate.
0
200
400
600
800
1000
1200
1400
1600
1800
20 25 30 35 40 45 50
I [
CP
S]
Theta - 2Theta [Degree]
(101)
(002)
(100) (102)
XRD 6000 SHIMADZU XR-Diffractometer
Pd (111)
Figure 4.5: XRD crystal structure of Pd – doped ZnO thin film (thickness =1178 nm)
prepared from 0.1 M Zinc Chloride aqueous precursor on glass substrate.
100
4.3 Surface topography and morphology studies
Figure 4.6 (a) shows the surface micrograph of zinc oxide film
prepared at 400 0C which consists of a uniform distribution of spherical –
shaped nanostructure grains with a diameter of about 20 nm. This struc-
ture repeats throughout the materials with closely packed to each other
indicating good adhesiveness of film with the substrate. The grains size
seen is comparable with the value calculated from x-ray diffraction stud-
ies. Al-Hardan et al., had a uniform distribution of RF – sputtered ZnO
nanostructure grains with a similar average grain size diameter [12]. Film
prepared at 200 0C spraying temperature, displayed in the inset (b) of
Peak No. 2Theta
deg.
dExp.
dTheo I/I1 FWHM
deg.
Intensity
counts
Integrated
Int.
counts
1 31.6946 2.82084 2.857884 8 0.179 104 854
2 34.383 2.60618 2.65 100 0.1958 1355 8020
3 36.1701 2.48141 2.515484 13 0.2329 170 1287
4 47.4654 1.91393 1.943173 6 0.2588 82 578
Peak
No.
2Theta
deg. dTheo Å dExp. Å I/I1
FWHM
Deg.
Intensity
counts
Integrated
Int.
counts
1 31.7013 2.8578838 2.82026 5 0.179 56 379
2 34.3646 2.65 2.60754 100 0.1958 1166 6624
3 36.1732 2.5154837 2.4812 11 0.2329 131 938
4 47.4424 1.9431734 1.9148 7 0.2588 82 630
Table 4.2: Crystalline structure, Miller indices and d spacings of the as – deposited
ZnO crystal planes.
Table 4.3: Crystalline structure, Miller indices and d spacings of the Pd – doped ZnO
crystal planes.
101
figure 4.6, demonstrates a discontinuous nature. E. Arca et al. believe that
at low temperature the droplet splashes onto the substrate with lesser
decomposition which leads to porous and less adhesive film which could
sometimes be observed visually [71].
The surface morphology of the undoped ZnO films as observed
from the AFM micrograph (figures 4.7 and 4.8) confirms that the grains
are uniformly distributed within the scanning area (5 μm Χ 5 μm), with
individual columnar grains extending upwards. This surface characteristic
is important for applications such as gas sensors and catalysts [101]. It
was found that using isopropyl alcohol organic solvent other than water is
preferred. This is due to a better droplet size distribution and, also, due to
additional heat transfer toward the sample surface resulted from alcohol
burning [71]. The root mean square (rms) of the film surface roughness
deposited at 450 0C using precursor of zinc acetate dissolved in distilled
Figure 4.6: Scanning Electron Micrograph of ZnO film prepared at a) 400 0C and the
inset b) 200 0C
a
b
102
CSPM Imager Surface Roughness Analysis Image size: 20000.00*20000.00 nm Amplitude parameters: Sa(Roughness Average) 31.1 [nm] Sq(Root Mean Square) 39.2 [nm] Ssk(Surface Skewness) 0.101 Sku(Surface Kurtosis) 3.12 Sy(Peak – Peak) 304 [nm] Sz(Ten Point Height) 293 [nm]
CSPM Imager Surface Roughness Analysis Image size: 10000.00*10000.00 nm Amplitude parameters: Sa(Roughness Average) 31.4 [nm] Sq(Root Mean Square) 39.9 [nm] Ssk(Surface Skewness) 0.0412 Sku(Surface Kurtosis) 3.15 Sy(Peak – Peak) 283 [nm] Sz(Ten Point Height) 267 [nm]
CSPM Imager Surface Roughness Analysis
Image size: 2000.00*2000.00 nm Amplitude parameters: Sa(Roughness Average) 13.4 [nm] Sq(Root Mean Square) 17 [nm] Ssk(Surface Skewness) -0.366 Sku(Surface Kurtosis) 3.17 Sy(Peak – Peak) 108 [nm] Sz(Ten Point Height) 105 [nm]
Figure 4.7: Scanning Probe Microscope images of zinc oxide thin film spray pyrolysed on
glass substrate at 450 oC spraying temperature with the precursor of 0.2 M zinc acetate
dissolved in 100 mL distilled water.
103
CSPM Imager Surface Roughness Analysis
Image size: 20000.00*20000.00 nm Amplitude parameters: Sa(Roughness Average) 15 [nm] Sq(Root Mean Square) 25.8 [nm] Ssk(Surface Skewness) 0.766 Sku(Surface Kurtosis) 9.04 Sy(Peak – Peak) 223 [nm] Sz(Ten Point Height) 223 [nm]
CSPM Imager Surface Roughness Analysis
Image size: 5000.00*5000.00 nm Amplitude parameters: Sa(Roughness Average) 1.57 [nm] Sq(Root Mean Square) 2.21 [nm] Ssk(Surface Skewness) 1.35 Sku(Surface Kurtosis) 10.3 Sy(Peak – Peak) 34.7 [nm] Sz(Ten Point Height) 33.4 [nm]
CSPM Imager Surface Roughness Analysis
Image size: 2000.00*2000.00 nm Amplitude parameters: Sa(Roughness Average) 15 [nm] Sq(Root Mean Square) 25.8 [nm] Ssk(Surface Skewness) 0.766 Sku(Surface Kurtosis) 9.04 Sy(Peak – Peak) 223 [nm] Sz(Ten Point Height) 223 [nm]
Figure 4.8: Scanning Probe Microscope images of zinc oxide thin film spray pyrolysed on
glass substrate at 450 oC spraying temperature with the precursor of 0.2 M zinc acetate
dissolved in 100 mL isopropyl alcohol.
104
water is about 17 nm, indicating that the surface of the spray deposited
ZnO thin film is very smooth. This value increases to 25.8 nm and the
grain size decreases from 250 to 65 nm as zinc acetate is dissolved in
isopropyl alcohol organic solvent.
The higher nucleation with a lower growth rate, results in a fine
grains of the films. Further, the grain size of the film can also be deduced
from the AFM micrograph and the distribution of grain size value is ob-
served between a minimum of about 20 nm and a maximum of about 130
nm as illustrated in figure 4.9. The statistical mean grain size of the film
deposited at 450 °C is about 57.76 nm, which is a little bit higher than the
crystallite size calculated from the XRD profile (49.15211 nm).
0
20
40
60
80
100
120
0
20
40
60
80
100
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190
Cum
ula
tio
n
%
Per
centa
ge
%
Diameter nm
Granularity Cumulation Distribution Chart
Sample: ZnO_01 Code: 009 Line No.: lineno Grain No.:1072 Instrument: CSPM Date: 2011-03-29
Avg. Diameter: 57.76 nm <=10% Diameter: 20.00 nm <=50% Diameter: 50.00 nm <=90% Diameter: 100.00 nm
Figure 4.9: 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.
105
4.4 Optical properties
Figure 4.10 shows the optical transmittance spectra of the ZnO thin
films. Approximately, all the films demonstrate more than 60% transmit-
tance at wavelengths longer than 400 nm, which is comparable with the
values for the ZnO thin films deposited by Ju-Hyun Jeong [69] using
electrostatic spray deposition ESD method, P. P. Sahay [80] using SPD
method, B. J. Babu [81] by ultrasonic spraying USP scheme. Below 400
nm there is a sharp fall in the T% of the films, which is due to the strong
absorbance of the films in this region. It has been observed that the over-
all T% increases with the decrease in the film thickness. This happens
due to the overall decrease in the absorbance with the decrease in film
thickness [80]. The relationship between the morphology of the sample
and the solvent composition is straightforward. A smooth, homogeneous,
good quality layer can be obtained by just using organic solvents (isopro-
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
200 300 400 500 600 700 800 900
Tra
nsm
issi
on
Wavelength nm
613 nm
523 nm 279 nm
189 nm
Figure 4.10: Transmission spectra of ZnO thin films of different thicknesses sprayed
on – glass at 400 0C temperature. The precursor was 0.2 M zinc acetate dissolved in
distilled water except for the 189.34 – nm thick sample which was a 0.2 M dis-
solved in 3:1 volume ratio isopropyl alcohol and distilled water.
106
pyl alcohol or methanol) while the surface gets rougher with increasing
water content as it has experimentally been noticed during spray.
The morphology of the film has a direct influence on the optical
properties of the coating (figure 4.10). Increasing water content leads to a
significant decrease in the transparency of the film. It is also worth noting
that the growth rate (related to the film thickness) increases with the wa-
ter content. This has already been observed in the literature in the case of
SnO2 [110]. Nevertheless the above mentioned decrease in transmission
is not caused by the thickness of the sample, but it is a consequence of the
scattering losses at the rough surface [71].
The absorption spectra of the films are shown in figure 4.11. These
spectra reveal that films grown under the same parametric conditions
have low absorbance in the visible/ near infrared region while absorbance
is high in the ultraviolet region.
The absorption coefficient (α) was calculated using Lambert law as
0
0.5
1
1.5
2
2.5
200 300 400 500 600 700 800 900
Ab
sorb
ance
Wavelength nm
523 nm
613 nm
279 nm
189 nm
Figure 4.11: Absorption spectra of ZnO thin films of different thicknesses sprayed on –
glass at 400 0C temperature. The precursor was 0.2 M zinc acetate dissolved in distilled
water.
107
follows [111]:
log 0 = d log e → 2.30258 A = d (4.1)
where I0 and I are the intensities of the incident and transmitted light
respectively, A is the optical absorbance and d is the film thickness.
The absorption coefficient (α) was found to follow the relation
h = A (h − g )1 2⁄ (4.2)
where A is a constant and Eg is the optical energy gap. Plots of (αhυ)2
versus the photon energy (hυ) in the absorption region near the funda-
mental absorption edge indicate direct allowed transition in the film mate-
rial [66], as shown in figure 4.12. Extrapolating the straight line portion
of the plot (αhυ)2 versus (hυ) for zero absorption coefficient value gives
the optical band gap (Eg), and its dependence on film thickness t is illus-
trated in figure 4.13. The band gap of the films varied slightly between
3.21 eV to 3.224 eV as the film thickness is changed from 189 nm to 613
0
2
4
6
8
10
12
14
16
2 2.5 3 3.5 4
(αh
ν)2
cm-2
. eV
2
Χ
10
10
hν eV
Figure 4.12: Plots of (αhν)2 vs. photon energy hν for ZnO thin films of different
energy gaps Eg and thicknesses t.
Eg= 3.210 eV, t=613 nm
Eg= 3.216 eV, t=523 nm
Eg= 3.220 eV, t=279 nm
Eg= 3.224 eV, t=189 nm
108
nm. As it is obvious from the plot above, the variation of thickness t for
the sprayed ZnO thin films has no effect on the estimated values of Eg.
4.5 Electrical properties
4.5.1 Resistance – temperature characteristic
The film is initially tested to confirm its semi conducting behavior.
The sensor is placed on a heater and its resistance is measured as the
temperature is ramped up from 50 0C to 350
0C in the dry air atmosphere.
Figure 4.14 shows the variation of resistance of the spray – pyrolyzed
deposited zinc oxide films of 668 nm film thickness with temperature.
The variation of the resistance with the temperature reveals that resistance
of the film decreases as the temperature increases from room temperature
3.15
3.2
3.25
100 200 300 400 500 600 700
Ener
gy g
ap E
g
eV
Film thickness t nm
Figure 4.13: Relationship of the extrapolated energy gap Eg of sprayed ZnO thin
films at different film thicknesses.
109
to 200 0C showing a typical negative temperature coefficient of resistance
(NTCR) due to thermal excitation of the charge carriers in semiconductor
[104]. Above 240 0C, sensor film displays positive temperature coeffi-
cient of resistance (PTCR) as temperature increases further, which may
be due to the saturation of the conduction band with electrons promoted
from shallow donor levels caused by oxygen vacancies. At this point an
increase in temperature leads to a decrease in electron mobility and a
subsequent increase in resistance. Similar observations are made by other
research groups [104, 18]. According to Al-Hardan et. al., [12] below 150
0C temperature, oxygen adsorption at the surface is mainly in the form of
O2− , while above 150
0C, chemisorbed oxygen is present in the form of
O− or O
−2. Due to the conversion of O2
− into O− or O
−2, oxygen adsorbs
the additional electron from the zinc oxide, which is attributed to increase
in the resistance of the sensor film as temperature rises further. At tem-
perature higher than 275 0C up to 300
0C, the film resistance is not greatly
affected by the temperature variation, probably due to the equilibrium
0
100
200
300
400
500
600
700
800
900
1000
0 50 100 150 200 250 300 350 400
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
0.018
0.02
Res
ista
nce
kΩ
Temperature C
Co
nd
uct
ance
S
Figure 4.14: The variation of resistance of the spray – pyrolyzed deposited zinc
oxide film of 668 nm film thickness with temperature.
110
obtained between the two competing processes: thermal excitation of
electrons and the oxygen adsorption. Finally, at temperature higher than
300 0C the resistance decreases again, probably because of the dominant
excitation of electrons and desorption of electron species [12]. The tem-
perature range 200 – 240 0C is suitable for sensor operation due to the
small temperature dependence of the sensor [12].
4.5.2 I – V characteristic of the zinc oxide films
I – V characteristics of nanostructured ZnO films are shown in fig-
ure 4.15. Both dark current and current under illumination increase linear-
ly for both positive and negative applied bias voltages up to ±12 V. How-
ever, when observations are made in vacuum, the dark current increases
due to decrease in resistance [112], figure 4.16.
In air, the dark base line current decreases due to increase in re-
sistance. The resistance variation in air is attributed to the effect of oxy-
gen chemisorption.
-10
-8
-6
-4
-2
0
2
4
6
8
10
-14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14
Cu
rren
t μ
A
Bias Voltage V
UV - illuminated
Dark
Figure 4.15: The I–V characteristic at room temperature of undoped ZnO film of
1178 - nm thickness in dark and under UV illumination.
111
It is generally accepted that oxygen is chemiadsorbed at a surface
site such as oxygen vacancy in the form of an ionized oxygen atom or
molecule, i.e. O− or O2−, resulting in a reduced concentration of free elec-
trons at the surface and the observed reduction in the conductivity or dark
current [112, 113].
The effect of temperature on the I – V characteristics is depicted in
figure 4.17. It confirms the enhancement of the current with temperature
from room temperature up to 200 0C. As the temperature is increased,
more electrons have sufficient energy to surmount the barrier height be-
tween the grains.
It can be observed that there is a decrease in the measured current
as the temperature is further raised above 200 0C indicating an increase in
the film’s resistance. This effect is observed in the chemisorption region
at elevated temperatures (250 – 500 0C) [104] where the oxygen is ad-
sorbed at the surface of the metal oxide that enable an electron trapping.
Hence the charge carrier density is reduced which leads to an increase in
the resistance of the ZnO. This reaction can be expressed as follows:
10
15
20
25
30
35
40
0 50 100 150 200 250 300 350 400
Curr
ent
μ
A
Time s
maximum
vacuum
Atmospheric air Atmospheric
air
Vacuum
pump ON
Vacuum
pump OFF
Figure 4.16: The effect of vacuum on base line current of a ZnO thin film at 200 0C
and 10 v bias voltage.
112
1
2O2 + e− → O− (4.3)
where O2 is the adsorbed oxygen molecules, O- is the chemisorbed oxy-
gen and e- is the trapped electrons from the ZnO surface. In this region,
the film resistance shows weak dependence on the temperature, as the
equilibrium is achieved for the thermal excitation of electrons and oxygen
adsorption processes.
4.5.3 AC impedance spectroscopy
Figure 4.18 shows the Cole – Cole plot of the impedance spectrum
of ZnO thin film at room temperature. It was observed that as the temper-
ature of the films increases above room temperature (300 K), the imped-
ance spectra begin to distort and therefore, experimental observations
cannot be carried out above room temperature. The spectrum at room
temperature contains only a single arc, but the arc has a non-zero inter-
section with the real axis in the high frequency region. Also, the center of
-15
-10
-5
0
5
10
15
-12 -8 -4 0 4 8 12
Cu
rren
t
μA
Bias voltage v
36 0C
50 0C
100 0C
200 0C
300 0C300 0C
200 0C
100 0C
50 0C
36 0C
Figure 4.17: The I–V characterization of sprayed ZnO film in the temperature range
from RT to 300 0C.
113
each arc lies below the real axis at a particular angle of depression θ. This
indicates that in our films the relaxation time τ is not single – valued but
is distributed continuously or discretely around a mean τm = ωm-1
(the
ideal case). The angle θ is related to the width of the relaxation time dis-
tribution and as such is an important parameter [80].
The low frequency arc is interpreted as due to the grain boundary
effect and the high frequency arc is attributable to the grain effect, in
agreement with the conventional view. In this experiment, the arc is ob-
served in the low frequency region. This indicates that the electric
transport mechanism is associated with the grain boundaries.
A simple R – C equivalent circuit, shown in the inset of figure
4.18, is used to simulate the impedance spectrum. The values of real and
imaginary components for such a circuit are given by:
Z′ = S +
(1 + ω2C 2
2 ) (4.4)
Z′′ = −ωC
2 2
(1 + ω2C 2
2) (4.5)
0
10000
0 10000 20000 30000
Z''
Ω
Z' Ω
Figure 4.18: The Cole-Cole plot for the impedance spectrum of the films at room tempera-
ture. The inset is the R-C equivalent circuit of the simulation of the impedance spectrum.
RS
RP
CP
114
The values of RS, RP and CP of the circuit are estimated using the
experimental values. These values are fed back in equations 4.4 and 4.5
to evaluate Z’ and Z” for the spectrum. The simulated curve is shown in
the same figure 4.18 (dotted line). A close agreement between the two
shows that a simple circuit shown in the inset of figure 4.18 can be used
to analyze the Cole – Cole plot for impedance spectrum.
4.6 Gas sensing measurements
4.6.1 Sensing characteristics of pure ZnO towards hydrogen gas
The gas sensing characteristics of the as – sprayed ZnO film are
carried out for H2 reducing gas at different mixing ratios and operating
temperatures. A known amount of target gas is introduced after the ohmic
resistance of the sensor material gets stabilized. The recovery
characteristics (when the target gas is withdrawn) are also monitored as a
function of time. Figure 4.19 demonstrates sensor test at 6v bias voltage
and operating temperature of 210 0C. The hydrogen : air mixing ratio is
set at 3%, 2%, 1%, respectively. The ZnO sample is prepared from zinc
30
40
50
60
70
80
90
0 100 200 300 400 500 600
Cu
rren
t μ
A
Time s
3%
H 2%
H 1%
H
Figure 4.19: Sensing behavior of pure ZnO thin film at 6 v bias voltage and 210 0C tem-
perature to traces of H2 reducing gas mixing ratio in air of 3%, 2%, and 1% respectively.
115
acetate of 0.2 M precursor solution sprayed on the aluminum interdigitat-
ed electrodes (IDE) of 1– mm finger spacing and at 450 0C spraying tem-
perature.
The variation of sensor sensitivity S, as estimated using equation
2.30 with test gas mixing ratio C is illustrated in figure 4.20. The figure
displays that the sensitivity of the sensor is linear in the low gas concen-
tration region up to 2%, which benefits an actuator by enabling it to de-
tect different concentrations of combustible gases and organic vapors
[95], whereas, the sensitivity tends to saturate in the high gas concentra-
tion. This may be due to a saturation of adsorption of H2 atoms at the Al
electrode/ZnO nanofilm interface and lack of adsorbed oxygen ions at the
nanofilm surface to react with gas molecules [114]. [101] obtained a con-
sistent behavior on ZnO thin film prepared by thermal oxidation exposed
to H2 gas up to 120 ppm at 400 0C.
40
45
50
55
60
0 0.5 1 1.5 2 2.5 3 3.5
Sen
siti
vit
y
%
Hydrogen : air mixing ratio %
Figure 4.20: The sensitivity dependence of as – deposited ZnO sensor on hydrogen gas
mixing ratio
116
Figure 4.21 exhibits the transient response as a function of H2 gas
concentration for the as deposited, 668 – nm thick, ZnO sensing element
at 210 0C. The sensitivity of the ZnO gas sensor increases as the H2 gas
concentration is increased from 1% (10000 ppm) to 3% (30000 ppm) and
it drops relatively rapidly when the H2 gas is removed , indicating that the
gas sensor has a good response for different H2 concentrations. Besides, it
takes almost the same time for the sensor to reach the maximum sensitivi-
ty for different H2 concentrations. This result is consistent with the con-
clusion for the dominance of operation temperature for the response time
[115].
The response and recovery times of the undoped sensor as a func-
tion of testing gas mixing ratio is illustrated in figure 4.22. Both response
and recovery of the sensor have the same monotonicity behavior as the
hydrogen target gas concentration increases. They both decrease with
increasing hydrogen concentration up to 2% at which the lowest response
and recovery times of 21s and 15 s are observed. Figure 4.23 illustrates a
0
10
20
30
40
50
60
0 50 100 150 200
Sen
siti
vit
y
%
Time s
Figure 4.21: Transient responses of ZnO thin film (668 nm thick) at 210 0C testing
temperature upon exposure to hydrogen gas of mixing ratios of 1%, 2%, and 3%
respectively.
3%
2%
1%
117
comparison of the I – V characteristic curves of the undoped sensor from
0 up to 10 v of 2–v increment both in atmospheric air and in three H2 gas
containing air ambients.
A linear dependence is dominant for the measured maximum cur-
0
20
40
60
80
100
120
140
0
5
10
15
20
25
30
35
0 0.5 1 1.5 2 2.5 3 3.5
Rec
over
y t
ime
s
Res
po
nse
tim
e
s
Hydrogen : air mixing ratio %
Figure 4.22: 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.
0
1
2
3
4
5
6
7
8
9
0 2 4 6 8 10 12
Max
imu
m c
urr
ent
Imax
m
A
Bias Voltage v
Air
1% H2 3%H
5% H2
Figure 4.23: I - V characteristics of undoped ZnO gas sensor to 5%, 3%, and 1%
Hydrogen gas mixture in air and at 200 degrees temperature
118
rent Imax with its value almost doubled once the reducing gas being inside
the testing chamber.
4.6.2 Sensing characteristics of Pd – doped ZnO towards hydrogen gas
Figure 4.24 shows the switching behavior of the ZnO gas sensor
followed film surface modification with 20 palladium chloride PdCl2
layers. A drastic enhancement in sensor sensitivity is achieved when it is
exposed to 3% hydrogen gas traces in air.
The temperature at which the test is carried out was 200 0C with a
10 – v bias voltage. As can be seen from the figure 4.24, both response
and recovery times ( 𝑒 = | 0 − 10 |) are much faster as compared
to figure 4.19. The response and recovery time values at the level of 90%
and at 3% of H2, are about 3 s and 116 s, respectively.
For the above two gas traces, the sensitivity was 91.5528% and
Figure 4.24 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.
0
200
400
600
800
1000
1200
1400
1600
1800
0 100 200 300 400 500 600 700 800
Co
nd
uct
ance
μ
S
Time sec.
Rise time = 3 sec
H2 OFF
H2 OFF
H2 ON
Recovery time = 116 s
H2 ON
119
94.3787% respectively which are comparatively twice as that for the
undoped ZnO sensor. These drastic performance enhancements in the
ZnO based H2 gas sensor are believed to be due to the role of the palladi-
um noble metal surface promotion [43, 116] which lowers the reaction
activation energy and the low grain size ZnO crystallites of the prepared
sensing layer made possible by using organic solvent precursors [71].
4.7 Operation temperature of the sensor
One of the most important disadvantages of ZnO gas sensors is the
high temperature required for the sensor operation (200 – 500 ºC). For
that reason, the effect of the operation temperature on the thin films sensi-
tivity was studied with the aim of optimizing the operation temperature to
the lowest possible value.
The gas sensitivity tests performed at room temperature show no
variation on the film conductivity, even with the increase of the gas con-
centration. The increase in the operation temperature leads to an im-
provement of the films sensitivity. Figure 4.25 illustrates the results of
how the maximum conductance Gmax depends on the temperature T for
0
500
1000
1500
2000
2500
3000
3500
0 50 100 150 200 250 300 350 400
Max
imu
m C
on
du
ctan
ce G
max
μS
Temperature T 0C
Figure 4.25: 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.
120
hydrogen sensor based on surface – promoted, with Palladium, ZnO sens-
ing layer of about 1178 nm thickness. It is seen that the film maximum
conductance Gmax goes through a maximum on changing T, with the best
operating temperature at around 280 ºC. Roughly speaking, the increase
of Gmax (the left side of the maximum) results from an increase in the rate
of surface reaction of the target gas, while the decrease of Gmax (the right
side) results from a decrease in the utility of the gas sensing layer. At the
temperature of the maximum conductance (response), the target gas mol-
ecules have optimum penetration depth into the gas sensing grains (large
utility) i.e., optimum reactivity for the diffusion in the whole sensing
layer, as well as for exerting sufficiently large interaction with the surface
(large gas response coefficient). This explains qualitatively why the corre-
lations between Gmax and T take a volcano shape for semiconductor metal
oxide gas sensors [48].
It is evident from the figure that the ZnO film shows a negative
temperature coefficient of resistance (NTCR) up to ~ 270 oC, whereas
above 285 oC it shows a positive temperature coefficient (PTCR). Similar
behavior was obtained by other researchers [18, 101].
The sensitivity of the sensor was calculated after the response had
reached steady state condition as a function of the operating temperature
in the range of 150 to 350 oC with temperature increment of 50
oC. The
test was achieved by recording the change in the current I (and ultimately,
the conductance G) upon exposure to a specific concentration of the hy-
drogen target gas which was kept constant at 3% in atmospheric air. The
variation of sensitivity with the operating temperature is shown in figure
4.26. The sensitivity increased as the operating temperature increased,
reaching a maximum value (~ 97%) at 250 oC, and decreased thereafter
with further increase in the operating temperature. It is suggested that 250
121
oC is the optimum operating temperature for high sensitivity of the sen-
sor.
The sensitivity as well as response time is temperature dependent
since the chemical kinetics governing the solid-gas interface reaction is
temperature dependent [117].
Figure 4.27 illustrates the transient responses of palladium - pro-
moted ZnO thin film (245 nm thick) as exposed to 3% H2: air gas mixing
ratio and at three different testing temperatures of 250, 350, and 300 0C
successively. Here, the surface Pd promotion is achieved by spraying
(rather than dipping) of the palladium chloride PdCl2 solution. The tem-
perature at which PdCl2 solution is sprayed was 400 0C. After 20 sprays,
the Pd – promoted ZnO sample was heat treated for 1 hour in atmospheric
air. As it is obvious from the figure, maximum sensitivity S of 93.01228
% is obtained at a temperature around 300 0C with a comparatively fast
response time of ~ 4 s and a baseline recovery time of 72 s.
60
70
80
90
100
0 50 100 150 200 250 300 350 400
Sen
siti
vit
y
%
Temperature 0C
Figure 4.26: The variation of sensitivity with the operating temperature of the Pd –
doped ZnO gas sensor.
122
There seems to be no noticeable difference in sensing behavior of
the ZnO hydrogen gas sensor as its surface is promoted by either dipping
or spraying. However, it is worth to mention here that the palladium layer
over the ZnO surface, resulted by spraying, looks more homogeneous
than that obtained by dipping.
4.8 Tin oxide (SnO2) hydrogen gas sensor
4.8.1 Crystalline structure and morphology of undoped SnO2 thin film
It is known that tin dioxide SnO2 has a tetragonal rutile crystalline
structure (known in its mineral form as cassiterite) [118]. The unit cell
consists of two metal atoms and four oxygen atoms. Each metal atom is
situated amidst six oxygen atoms which approximately form the corners
of a regular octahedron. Oxygen atoms are surrounded by three tin atoms
which approximate the corners of an equilateral triangle. The lattice pa-
rameters are a= 4.7382 Å, and c= 3.1871 Å. Figure 4.28 shows the X-ray
diffraction (XRD) pattern of the SnO2 thin film prepared on glass sub-
0
20
40
60
80
100
0 50 100
Sen
siti
vit
y
%
Time s
Figure 4.27: Transient responses of Pd – sensitized ZnO thin film (245 nm thick) as
exposed to hydrogen gas of mixing ratio of 3% and at three different testing tempera-
tures of (1) 250, (2) 350, and (3) 300 0C successively.
1
2
3
123
strate at 450 °C spraying temperature. The major diffraction peaks of
some lattice planes can be indexed to the tetragonal unit cell structure of
SnO2 with lattice constants a= 4.738 Å and c= 3.187 Å, which are con-
sistent with the standard values for bulk SnO2 (JCPDS-041-1445) [119].
There are six peaks with 2θ values of 26.72418 0, 34.03094
0, 38.00002
0,
43.45998 0, 51.91576
0 and 54.85798
0 corresponding to SnO2 crystal
planes peaks of (110), (101), (200), (211), (220), and (002) respectively.
No characteristic peaks belonging to other tin oxide crystals or impurities
were detected. In our films, the XRD spectrum showed predomination of
the peaks, corresponding to reflection from the crystallographic (110),
(101) planes, parallel to the substrate. The intensity of other peaks is
small. It indicates that the current films are textured. At that, the degree of
the texturing depends on kind of sprayed solution we used, and increases
while using water solution instead of alcohol solution of SnCl2.
0
20
40
60
80
100
120
140
160
15 20 25 30 35 40 45 50 55 60 65
Inte
nsi
ty
I C
PS
Theta 2 -Theta degrees
(110)
(101)
(200)
(220)
SnO2
(211)
(002)
Figure 4.28: X-ray diffraction (XRD) pattern of SnO2 thin film spray pyrolyzed
on glass substrate at temperature of 450 oC.
124
The high intensity of these peaks suggests that these thin films
mainly consist of the crystalline phase.
The surface morphology of the undoped SnO2 thin films, as re-
vealed by the AFM image, is shown in figure 4.29 on a scanning area of
2000 nm x 2000 nm. The average roughness, Ra of the sample is of the
order of 1.26 nm, whereas the peak – to – valley roughness, RPV takes
value of up to 12.8 nm. This result indicates that the coating surface mor-
Figure 4.29: AFM image of undoped SnO2 thin film deposited at 450 oC on
glass substrate with the precursor being tin dichloride dehydrate dissolved in
isopropyl alcohol.
125
phology of SnO2 thin films is almost perfectly smooth with nanosize
grains. The estimated grain size of undoped films is in the range of 57.6 –
68.8 nm.
4.8.2 Optical properties of the undoped tin oxide SnO2 thin films
Figure 4.30 shows the transmittance spectra obtained at the wave-
length between 300 – 850 nm. The optical transmission depends on the
film thickness. The increase of the film thickness leads to higher absorp-
tion and thus reducing the transmittance. The average visible transmit-
tance calculated in the wavelength ranging 400 – 700 nm varied between
~58 and 80 %.
The calculated values of the direct optical energy gap varied be-
tween 3.49 and 3.79 eV for SnO2 thin films depending on film thickness
as obviously illustrated in figure 4.31. The variations of the optical ener-
gy gap could be attributed to changes in the film defect density.
The band gap decreases with the increase of the film thickness
from 145 nm to 466 nm. The decrease of band gap with the increase of
film thickness implies that SnO2 is an n-type semiconductor [80]. This
0.00%
20.00%
40.00%
60.00%
80.00%
100.00%
200 300 400 500 600 700 800 900
Tran
smis
sio
n
%
Wavelength nm
t=240.294 nm
t=145.633 nm
t=466.024 nm
t=466 nm
t=145 nm
t=240 nm
Figure 4.30: Transmission spectra of undoped SnO2 thin films of different
thicknesses deposited at 450 oC on glass substrates.
126
decrease of band gap may be attributed to the presence of unstructured
defects, which increase the density of localized states in the band gap and
consequently decrease the energy gap [80].
4.8.3 Sensing characteristics of pure SnO2 towards hydrogen gas
The percentage response, S, of the pure SnO2 towards hydrogen
gas of different mixing ratios has been explored. The successive tests
were performed at a bias voltage of 5.1v and a 210 0C operating tempera-
ture. The results are shown in figure 4.32.
As it is apparent from the figure, the sensor sensitivity to hydrogen
gas increases linearly with H2 test gas mixing ratio up to 3% H2 (S~83%)
after which the sensitivity tends to saturate with increasing the analyte
gas. A maximum sensitivity value of 88% has been registered when the
sensor is exposed to 4% H2 gas in air (figure 4.33). Moreover, it’s worth
to indicate here that both the response and recovery times of the undoped
SnO2 gas sensor decrease with increasing the hydrogen gas mixing ratio,
0
5
10
15
20
25
1.5 2 2.5 3 3.5 4 4.5
(αh
ν)2
eV
2 c
m-2
Χ1
01
0
hν eV
Sample 1 thickness t=240.294 nm , Eg=3.76 eV Sample 2 thickness t=145.633 nm , Eg=3.79 eV Sample 3 thickness t=466.024nm , Eg=3.49 eV
Figure 4.31: Absorption coefficient versus the photon energy for energy gap esti-
mation of undoped SnO2 thin films of different thicknesses deposited at 450 oC on
glass substrates.
127
with the shortest response and recovery times being about 29, 127 s re-
spectively at 4% H2:air gas mixing ratio. Both response and recovery
times were measured as 90% of the conductance change ∆ that the
sensor experiences upon the step introduction of the H2 reducing gas.
0
10
20
30
40
50
60
70
80
90
100
0 500 1000 1500
Sen
siti
vity
S
%
Time t s
1% H2
2% H2
3% H2
4% H2
Figure 4.32: Sensitivity behavior of undoped tin oxide SnO2 thin film to different hydro-
gen concentrations. The bias voltage was 5.1 v with the temperature set to 210 0C.
Figure 4.33: 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.
30
40
50
60
70
80
90
100
0% 1% 2% 3% 4%
Sen
siti
vity
S
%
H2:air mixing ratio C %
128
4.8.4 Sensing characteristics of Pd – doped SnO2 towards hydrogen gas
In a similar way to that done for the ZnO thin film sensing element,
and in order to enhance the sensing characteristics of the tin oxide SnO2 –
thin film based sensors, its surface is promoted by 20 palladium layers
applied by the same spraying method used to prepare the films. Figure
4.34 shows the representative real – time electrical responses of a 145 nm
– thick PdCl2 promoted SnO2sensing element to H2 gas concentrations up
to 4.5% in air. The test was performed at 210 0C sensing temperature and
10 v bias voltage.
A drastic enhancement in sensor sensitivity towards hydrogen gas
is achieved in which, the sensor maximum current increased upon being
exposed to successively incrementing hydrogen gas concentrations. Fig-
0
100
200
300
400
500
600
700
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
Cu
rren
t μ
A
Time s
3.3% H2 4.5% H2
2% H2
1% H2
0.5% H2
pulse due to H2 remaining in the
tubing of H2 when the manifold is cracked
open; NF is still closed
Current increased upon switching ON of
rotary - from atmosphere to
vacuum
Figure 4.34: 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.
129
ure 4.35 exhibits the transient sensitivity S variation with hydrogen gas
concentration.
The relationship of sensor response time versus hydrogen:air mix-
ing ratio is plotted in figure 4.36. The maximum sensitivity obtained is
95.744 % at 4.5% H2 gas concentration in air which is slightly less than
that reported by Mitra [117]. Also, the shortest response time of about 24
s was observed at 2% hydrogen gas concentration. This speed of response
is faster than that obtained by Sunita Mishra et al. [120]
Figure 4.37 shows the switching behavior to 4.5% H2 gas of the
SnO2 gas sensor followed film surface modification with 20 palladium
chloride PdCl2 layers. The response for three operating temperatures is
compared. It is evident from the figure that the sensor sensitivity increas-
es, and its response time decreases, with increasing the operating temper-
ature for the sensor.
Figure 4.35: Response transient of Pd – doped SnO2 gas sensor to different H2 : air
mixing ratios. The tests were performed at 210 degrees temperature and 10 v bias.
0
20
40
60
80
100
0 250 500 750 1000 1250 1500 1750 2000 2250
Sen
siti
vity
%
Time s
0.5% H2
1% H2
2% H2
3.3% H2 4.5% H2
130
Figure 4.37: Transient responses of SnO2 thin film of 248 nm thick at 150, 175, and
210 0C testing temperature upon exposure to 4.5% H2:air gas mixing ratio.
0
20
40
60
80
100
0 100 200 300 400 500
Sen
siti
vity
%
Time s
210 0C
150 0C
175 0C
Figure 4.36: 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.
0
20
40
60
80
100
0
10
20
30
40
50
60
70
80
0 1 2 3 4 5
Sen
siti
vity
%
Res
po
nse
tim
e s
H2 mixing ratio %
131
The optimum operating temperature for the palladium – doped tin
oxide hydrogen gas sensor was found to be around 210 0C. The response
versus temperature plot, figure 4.38, demonstrated a volcano - shape
relationship.
0
100
200
300
400
500
600
700
100 125 150 175 200 225 250 275 300
Max
imu
m c
urr
ent
Im
ax.
μ
A
Temperature T 0C
Figure 4.38: 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.
132
4.7 Conclusions and future work proposals
In this study, the influence of thin film processing conditions on
the properties and gas sensing performance of spray pyrolyzed Pd –
doped ZnO and SnO2 thin films have been investigated.
The spray pyrolysis deposition has proved to be a relatively simple
and reliable deposition technique in acquiring high quality thin films of
different types and at a wide range of deposition temperatures. This
method has the advantages of low cost, easy-to-use, safe and efficient
route to coat large surface areas in mass production.
The spray pyrolyzed ZnO thin films obtained from non – organic
solvent (distilled water) are observed to be comparatively of less trans-
parency, with thickness of about 1178 nm, as measured by weighing
difference method and confirmed via the SEM unit.
The XRD diffraction pattern proves that the prepared ZnO on glass
substrate is highly c – axis oriented, giving a peak at Bragg angle equal to
34.38 o, which belong to the (0 0 2) phase of the hexagonal wurtzite struc-
ture of the ZnO.
The nearly ohmic behavior of I–V characteristics reveals that the
prepared films contain high carrier concentrations.
AFM investigations reveal a porous morphology of spherical parti-
cles. 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.
Both spray pyrolyzed Pd – doped ZnO and SnO2 thin – film sen-
sors demonstrated high sensitivity, relatively fast, and excellent selectivi-
ty to hydrogen reducing gas. Thus, they exhibit an increase in the con-
ductance for exposure to hydrogen gas of different concentrations and
operating temperatures, showing excellent sensitivity. It is found that the
133
sensing of hydrogen gas in our metal oxide sensors is related to the en-
hancement of adsorption of atmospheric oxygen. The excellent selectivity
and the high sensitivity for hydrogen gas can be achieved by surface
modification of ZnO 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 leads 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. For tin oxide sensors, the optimum tem-
perature is 210 0C.
The response – recovery time of Pd:ZnO sensing element to hy-
drogen gas is very fast. ZnO thin films of 20 – time dipping (or spraying
with) 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 volume for
H2 gas). The current high sensitivity and the fast response time are com-
parable to that obtained by Mitra et al., [92].
There is a great effort to fabricate hand – held, low power con-
sumption gas sensing elements that possesses high sensitivity, fast re-
sponse and recovery characteristic, and can operate at room temperature.
The heating element in the current ZnO and SnO2 gas sensors is bulky
and requires a 220 v A.C current source, consuming much power. This
issue can be handled by applying a thin film heating element on the back
of the substrate. The thin film is composed of a spraying solution consist-
ed of 100 g SnCl4.5H2O, 4g SbCl3, 6 g ZnCl2, 50 cc H2O, and 10 cc HCl
equivalent to 93.2% SnO2, 5.5% Sb2O3 and 1.3% ZnO. Using this ap-
134
proach, Mochel [64] obtained such thin films of an electrical resistance of
42 ohms per square. This value increased to 56 ohms per square upon
passing an alternating current, equivalent to 300 Watts at 110 v, to the
deposited film. Doing so, the sample temperature increased to 825 0C in
just a few minutes. Seven heating and cooling cycles caused no substan-
tial change in the resistance and other properties.
135
References:
[1] J. R. Stetter, W. R. Penrose, and S. Yao, Sensors, chemical sensors, electro-
chemical sensors, and ecs, Journal of The Electrochemical Society, 150 (2)
S11-S16 (2003).
[2] C. Wagner, The mechanism of the decomposition of nitrous oxide on zinc
oxide as catalyst, J. Chem. Phys. Vol. 18, No.1, pp. 69 – 71, (1950).
[3] J. Bardeen, W. H. Brattain, Surface properties of germanium, Bell. Syst.
Techn. J. Vol. 32, No. 1, (1952).
[4] T. Seiyama, A. Kato, K. Fujushi, M. Nagatani, A new detector for gaseous
components using semiconductive thin films, Anal. Chem. 34, 1502 (1962).
[5] N. Taguchi, Japan Patent No. 45-38200 (1962).
[6] FIGARO products catalog, pag. 2, (1990).
[7] S. M. Kanan , O. M. El-Kadri, I. A. Abu-Yousef and M. C. Kanan, Semi-
conducting metal oxide based sensors for selective gas pollutant detection,
Sensors 9, pp. 8158-8196, (2009).
[8] N. –M. Bârlea, S. I. Bîrlea, Semiconductor gas sensor for detecting air pol-
lution, Studia Universitatis Babes-Bolyai, Physica, special issue, (2003).
[9] V. Brinzari, G. Korotcenkov, J. Schwanka, Y. Boris, Chemisorptional ap-
proach to kinetic analysis of SnO2:Pd-based thin film gas sensors, Journal of
Optoelectronics and Advanced Materials Vol. 4, No. 1, pp. 147 – 150,
(2002).
[10] E. Elangovan, K. Ramamurthi, Optoelectronic properties of spray deposited
SnO2:F thin films for window materials in solar cells, Journal of Optoelec-
tronics and Advanced Materials Vol. 5, No. 1, pp. 45 – 54, March 2003.
[11] J. Hildenbrand, Simulation and characterisation of a micromachined gas sen-
sor and preparation for model order reduction, Diploma Thesis, Albert Lud-
wig University, Freiburg, Germany, 2003.
[12] N. H. Al-Hardan, M. J. Abdullah, A. Abdul Aziz, H. Ahmad, L. Y. Low,
ZnO thin films for VOC sensing applications, Vacuum 85, 101 – 106,
(2010).
[13] O. Lupan, V. V. Ursaki, G. Chai, L. Chow, G.A. Emelchenko, I. M.
Tiginyanu, A. N. Gruzintsev, A. N. Redkin, Selective hydrogen gas na-
nosensor using individual ZnO nanowire with fast response at room temper-
ature, Sensors and Actuators B 144, pp. 56–66, (2010).
[14] R. Ferro, J. A. Rodríguez, I. Jiménez, A. Cirera, J. Cerdà, and J. R. Morante,
Gas-sensing properties of sprayed films of (CdO)x(ZnO)1- x mixed oxide,
IEEE Sensors Journal, Vol. 5, No. 1, (2005).
[15] R. K. Joshi, S. Krishnan, M. Yoshimura, A. Kumar, Pd nanoparticles and
thin films for room temperature hydrogen sensor, Nanoscale Res Lett 4, pp.
1191–1196, (2009).
[16] R. Savu, M. A. Ponce, E. Joanni, P. R. Bueno, M. Castro, M. Cilense, J. A.
Varela, E. Longo, Grain size effect on the electrical response of SnO2 thin
136
and thick film gas sensors, Materials Research, Vol. 12, No. 1, 83-87,
(2009).
[17] S. Capone, A. Forleo, L. Francioso, R. Rella, P. Siciliano, J. Spadavecchia,
D. S. Presicce, A. M. Taurino, Solid state gas sensors: state of the art and fu-
ture activities, Journal of Optoelectronics and Advanced Materials Vol. 5,
No. 5, pp. 1335 – 1348, (2003).
[18] N. Abhijith, Semiconducting metal oxide gas sensors: development and re-
lated instrumentation, M. Sc. Thesis, Indian Institute of Science, Bangalore,
India, 2006.
[19] B. Licznerski, Thick-film gas microsensors based on tin dioxide, Bulletin of
the polish academy of sciences technical sciences, Vol. 52, No. 1, (2004).
[20] E. R. Beach III, Picoliter drop deposition of oxide nanoparticles: a route to
high performance microsensor arrays, Ph. D. Dissertation, The Ohio State
University, 2009.
[21] N Barsan and U Weimar, Understanding the fundamental principles of metal
oxide based gas sensors; the example of CO sensing with SnO2 sensors in
the presence of humidity, J. Phys.: Condens. Matter 15, R813–R839, (2003).
[22] P. Reichel, Development of a Chemical Gas Sensor System, Ph. D. disserta-
tion, Faculty of Chemistry and Pharmacy, Universitat of Tubingen, Germa-
ny, 2005.
[23] R. Frank, Understanding Smart Sensors, 2nd Edition, Artech House, Inc.,
Norwood, MA 02062, 2000.
[24] O. Lupan, G. Chai, L. Chow, Novel hydrogen gas sensor based on single
ZnO nanorod, Microelectronic Engineering 85, pp. 2220 – 2225, (2008).
[25] M. Takata, D. Tsubone, H. Yanagida, Dependence of electrical conductivity
of ZnO on degree of sintering, J. Am. Ceram. Soc. 59/1-2, 4, (1976).
[26] J. H. Lunsford, Catalysis Reviews 8, 135, (1973).
[27] S. J. Gentry, T. A. Jones, The role of catalysis in solid – state gas sensors,
Sensors and Actuators 10, 141, (1986).
[28] G. A. Somorjai, Introduction to surface chemistry and catalysis, John Wiley
and Sons, Inc., New York, 1994.
[29] G. A. Somorjai, Principles of Surface Chemistry, Prentice-Hall, Inc, Eng-
lewood Cliffs, 1972.
[30] I. Langmuir, The adsorption of gases on plane surfaces of glass, mica and
platinum, J. Am. Chem. Soc. 40, 1361, (1918).
[31] S. M. Aygun, Gas sensors based on ceramic p-n heterocontacts, M. Sc. The-
sis, Iowa State University Arnes, Iowa, 2005.
[32] N. Barsan & U. Weimar, Conduction model of metal oxide gas sensors,
Journal of Electroceramics, 7, 143–167, (2001).
[33] M. Schweizer-Berberich, Gas Sensors Based on Stannic Oxide, Ph. D. The-
sis, University of Tubingen, 1998.
137
[34] Y. Jin, Toxic gas sensors using thin film transistor platform at low tempera-
ture, M. Sc. Thesis, Massachusetts Institute of Technology MIT, 2009.
[35] W. Göpel and K. D. Schierbaum, SnO2 sensors: current status and future
prospect, Sensors and Actuators B, 26-27, pp.1-12, (1995).
[36] G. Korotcenkov, The role of morphology and crystallographic structure of
metal oxides in response of conductometric-type gas sensors, Materials Sci-
ence & Engineering R 61, pp. 1-39, (2008).
[37] H. Huang, C. L. Chow, Y. C. Lee, C. K. Lim, O. K. Tan, Highly selective H2
gas sensors based on ZnO – modified SnO2 nanorod arrays, IEEE Sensors
Conference, pp. 124 – 126, (2009).
[38] H. Nanto, T. Minami, and S. Takata, Zinc-oxide thin – film ammonia gas
sensors with high sensitivity and excellent selectivity, J. Appl. Phys. 60 (2),
(1986).
[39] S. Bagga, Gas sensor-studies on sensor film deposition, ASIC design and
testing, M. Sc. Thesis, Indian institute of science, Bangalore - 560012, India,
2007.
[40] L. A. Patil, A. R. Bari, M. D. Shinde, V. V. Deo, and D. P. Amalnerkar,
Synthesis of ZnO nanocrystalline powder from ultrasonic atomization tech-
nique, characterization, and its application in gas sensing, IEEE Sensors
journal, Vol. 11, No. 4, pp. 939 – 946, (2011).
[41] D. R. Patil, L. A. Patil and D. P. Amalnerkar, Ethanol gas sensing properties
of Al2O3-doped ZnO thick film resistors, Bull. Mater. Sci., Vol. 30, No. 6,
pp. 553 – 559, (2007).
[42] C. Liewhiran and S. Phanichphant, Influence of thickness on ethanol sensing
characteristics of doctor – bladed thick film from flame-made ZnO nanopar-
ticles, Sensors, 7, pp.185 – 201, (2007).
[43] T. Xua and M. P. Zach, Self-assembled monolayer – enhanced hydrogen
sensing with ultrathin palladium films, Applied Physics Letters 86, 203104
(2005).
[44] F. Favier, E. C. Walter, M. P. Zach, T. Benter, R. M. Penner, , Hydrogen
sensors and switches from electrodeposited palladium mesowire arrays, Sci-
ence Vol. 293, 2227, (2001).
[45] V. V. Sysoev, B. K. Button, K. Wepsoec, S. Dmitriev, and A. Kolmakov,
Toward the nanoscopic “electronic nose”: Hydrogen vs. carbon monoxide
discrimination with an array of individual metal oxide nano- and mesowire
sensors, Nano Lett., Vol. 6, pp. 1584 – 1588, (2006).
[46] S. R. Morrison, Selectivity in semiconductor gas sensors, Sensors and Actu-
ators, B No. 12, pp. 425 – 440, (1987).
[47] C. Wang, L. Yin, L. Zhang, D. Xiang and R. Gao, Metal oxide gas sensors:
sensitivity and influencing factors, Sensors, 10, pp. 2088 – 2106, (2010).
[48] N. Yamazoe, G. Sakai, and K. Shimanoe, Oxide semiconductor gas sensors,
Catalysis Surveys from Asia Vol. 7, No. 1, (2003).
138
[49] A. Kolmakov, D. O. Klenov, Y. Lilach, S. Stemmer, and M. Moskovits, En-
hanced gas sensing by individual SnO2 nanowires and nanobelts functional-
ized with Pd catalyst particles, Nano Lett. 5 (4), pp. 667 – 673, (2005).
[50] C. Xu, J. Tamaki, N. Miura and N. Yamazoe, Grain size effects on gas sensi-
tivity of porous SnO2-based elements, Sensors and Actuators, Vol. B, No. 3
pp. 147 – 155 (1991).
[51] X. Du, S. M. George, Thickness dependence of sensor response for CO gas
sensing by tin oxide films grown using atomic layer deposition, Sensors and
Actuators B 135 152 – 160, (2008).
[52] K. Arshak, E. Moore, G. M. Lyons, J. Harris and S. Clifford, A review of
gas sensors employed in electronic nose applications, Sensor Review, Vol.
24, No. 2, pp. 181 – 198, (2004).
[53] C. Hagleitner, A. Hierlemann, D. Lange, A. Kummer, N. Kerness, O. Brand
& H. Baltes, Smart single-chip gas sensor microsystem, Nature 414, pp. 293
– 296, (2001).
[54] E. Llobet, R. Ionescu, S. Al-Khalifa, J. Brezmes, X. Vilanova, X. Correig,
N. Bârsan, and J. W. Gardner, Multicomponent gas mixture analysis using a
single tin oxide sensor and dynamic pattern recognition, IEEE Sensors Jour-
nal, Vol. 1, No. 3, (2001).
[55] U. Weimar and W. Goepel, AC measurements on tin oxide sensors to im-
prove selectivities and sensitivities, Sensors and Actuators, Vol. B, No. 26-
27 pp. 13 – 18 (1995).
[56] T. Gao, G. Meng, Y. Tian, S. Sun, X. Liu and L. Zhang, Photoluminescence
of ZnO nanoparticles loaded into porous anodic alumina hosts, J. Phys.:
Condens. Matter 14, 12651 – 12656, (2002).
[57] U. Ozgur, D. Hofstetter, and H. Morkoc, ZnO devices and applications: A
Review of Current Status and Future Prospects, Proceedings of the IEEE
Vol. 98, No. 7, (2010).
[58] Y. Min, Properties and Sensor Performance of Zinc Oxide Thin Films, Ph.
D. Thesis, Massachusetts Institute of Technology MIT, MA, USA, 2003.
[59] R. M. Balabai, P. V. Merzlikin, electronic properties of doped ZnO films: ab
init io calculations, Ukr. J. Phys. Vol. 55, No. 10, (2010).
[60] T. Nguyen H., Hartmann Andreas J. and Lamb Robert N., Structural Order
of Nanocrystalline ZnO Films, The Journal of Physical Chemistry B,
103(21), 4264 – 4268, (1999).
[61] H. J. T. Ellingham, The Ellingham diagram, J. Soc. Chem. Ind. (London) 63:
125 . (1944).
[62] K. Nordlund, J. Tarus, and J. Keinonen, Atomic structures in semiconduc-
tors: dynamics, defects, and distributions, in CSC Report on Scientific Com-
puting 1999-2000, edited by S. Kotila and J. Haataja, Center for Scientific
Computing, Finland, p. 72, 2001.
[63] D. Perednis & L. J. Gauckler, Thin film deposition using spray pyrolysis,
Journal of Electroceramics, 14, pp. 103 – 111, (2005).
139
[64] J. M. Mochel, Electrically conducting coatings on glass and other ceramic
bodies, US Patent 2,564,707 (1951).
[65] J. E. Hill and R. R. Chamberlin, Process for making conductive film, US pa-
tent 3148084, pp. 1 – 13, (1964).
[66] C. Gümü, O. M. Ozkendir, H. Kavak, Y. Ufuktepe, Structural and optical
properties of zinc oxide thin films prepared by spray pyrolysis method,
Journal of Optoelectronics and Advanced Materials Vol. 8, No. 1, pp. 299 –
303, (2006).
[67] Z. Jiao, M. Wu, Z. Qin, M. Lu, J. Gu, The NO2 sensing ITO thin films pre-
pared by ultrasonic spray pyrolysis, Sensors 3, pp. 285-289, (2003).
[68] K. C. Pingali, D. A. Rockstraw, and S. Deng, Silver Nanoparticles from Ul-
trasonic spray pyrolysis of aqueous silver nitrate, Aerosol Science and Tech-
nology, 39:1010 – 1014, (2005).
[69] Ju-Hyun Jeong, Young-Sun Jeon, Kyung-Ok Jeon, Kyu-Seog Hwang and
Byung-Hoon Kim, Preparation of zinc oxide films by an electrostatic spray
deposition process, Journal of Ceramic Processing Research, Vol. 7, No. 1,
pp. 70 – 74, (2006).
[70] A. Jaworek_, A. T. Sobczyk, A. Krupa, M. Lackowski, and T. Czech, Elec-
trostatic deposition of nanothin films on metal substrate, Bulletin of the
Polish Academy of Sciences Technical Sciences, Vol. 57, No. 1, (2009).
[71] E. Arca, K. Fleischer, and I. V. Shvets, Influence of the precursors and
chemical composition of the solution on the properties of zno thin films
grown by spray pyrolysis, J. Phys. Chem. C 113, 21074–21081 (2009).
[72] WC Siu and HL Kwok, CuxS/CdS thin-film solar cells using chemically
sprayed CdS films, J. Phys. D: Appl. Phys., Vol. 11, (1978).
[73] R. I. Henry, E. J. Cukauskas, S. B. Qadri, A. H. Singer, and G. G. Campisi,
Thin film growth of oxide superconductor materials, IEEE Transactions on
Magnetics, Vol. 25, No. 2, March (1989).
[74] B. J. Brown, C. W, Bates, Jr., The effect of substrate on the morphology of
CulnSe2 films prepared by chemical spray pyrolysis for CulnSe2/CdS solar
cells, J, Appl. Phys., Vol. 68 No. 5, 1 September (1990).
[75] J. Song, Woo-Yeong Cho, Koeng-Su Lim, Kyung-Hoon Yoon and Gwon-
Jong Yu, Properties of ZnO:Al thin films for solar cells by pyrosol method,
IEEE Xplore 1st WCPEC, pp. 630 - 633, (1994).
[76] Y. R. Roh, H. B. Kim, Y. I. Lee, H. M. Cho, J. S. Chung, S. Baik, Develop-
ment of SAW gas sensor for monitoring SOx gas, IEEE Ultrasonics Sympo-
sium, Vol. 1, pp. 473 – 476, (1995).
[77] G. Korotcenkov, J. Schwank, M. DiBattista, V. Brinzari, XRD study of gas
sensitive SnO2 thin films deposited by spray pyrolysis method, Semicon-
ductor Conference, CAS '99 Proceedings, IEEE Xplore, Vol. 1, pp. 319 –
322, (1999).
[78] P. S. Patil, Versatility of chemical spray pyrolysis technique, Materials
Chemistry and Physics Vol. 59, Issue: 3, pp. 185 – 198, (1999).
140
[79] J. Ebothe and A. El Hichou, Flow rate and interface roughness of zinc oxide
thin films deposited by spray pyrolysis technique, Journal of Applied Phys-
ics Vol. 93, No. 1, pp. 632 – 640, (2003).
[80] P. P. Sahay, S. Tewari, and R. K. Nath, Optical and electrical studies on
spray deposited ZnO thin films, Cryst. Res. Technol. 42, No. 7, 723 – 729
(2007).
[81] B. J. Babu, A. Maldonado and S. Velumani, Deposition and characterization
of ZnO:Al thin films by ultrasonic spray pyrolysis, 6th International Confer-
ence on Electrical Engineering, Computing Science and Automatic Control
(CCE), IEEE Xplore, pp. 1 – 5, (2009).
[82] S. Gledhill, A. Grimm, D. Greiner, W. Bohne, M. Lux-Steiner, Christian-
Herbert Fischer, Doping induced structural and compositional changes in
ZnO spray pyrolysed films and the effects on optical and electrical proper-
ties, Thin Solid Films, Volume 519, Issue 13, pp. 4293 – 4298, (2011).
[83] T. Sebastian, Automation of Chemical Spray Pyrolysis Unit and Fabrication
of Sprayed CuInS2/In2S3 Solar Cell, Ph. D. Thesis, Cochin University of
Science and Technology, Cochin – 682 022, Kerala, India, (2009).
[84] J. Zeleny, The electrical discharge from liquid points, and a hydrostatic
method of measuring the electric intensity at their surfaces, Phys. Rev. 3(2)
(1914) 69.
[85] J. C. Viguie, J. Spitz, Chemical vapor deposition at low temperature, J. Elec-
trochem. Soc: Solid State Science and Technology, 122(4), 585, (1975).
[86] D. Perednis, Thin film deposition by spray pyrolysis and the application in
solid oxide fuel cells, Ph.D. Thesis, Swiss Federal Institute of Technology,
Zurich, 2003.
[87] N. H. J. Stelzer, J. Schoonman, Synthesis of Terbia-Doped Yttria – Stabi-
lized Zirconia Thin Films by electrostatic spray deposition (ESD), J. Materi-
als Synthesis and Proceedings. 4(6), 429, (1996).
[88] T. Sebastian, R. Jayakrishnan, C. S. Kartha and K. P. Vijayakumar, Charac-
terization of spray pyrolyzed CuInS2 thin films, The Open Surface Science
Journal, 1, 1- 6, (2009).
[89] G. Heiland, Homogeneous semiconducting gas sensors, Sensors and Actua-
tors, Vol. 2, pp. 343-361, (1982).
[90] H. Nanto, H. Sokooshi And T. Usuda, Shell sensor using zinc oxide thin
films prepared by magnetron sputtering, IEEE Xplore, pp. 596 – 599,
(1991).
[91] Hyung-Ki Hong, H. W. Shin, H. S. Park, D. H. Yun, C. H. Kwon, K. Lee,
Sung-Tae Kim and T. Moriimmi, Gas identification using oxide semicon-
ductor gas sensor array and neural-network pattern recognition, The 8th In-
ternational Conference on Solid-state Sensors and Actuators, and Eurosen-
sors IX. Stockholm, Sweden, June 25-29, Vol. 1, pp. 687 – 690, (1995).
[92] P. Mitra, A. P. Chatterjee, H. S. Maiti, Chemical deposition of ZnO films for
gas sensors, Journal of Materials Science: Materials in Electronics 9, 441 -
445, (1998).
141
[93] P. Nunes, E. Fortunato, R. Martins, Thin film combustible gas sensors based
on zinc oxide, Mat. Res. Soc. Symp. Proc. Vol. 666 (2001).
[94] S. Roy and S. Basu, Improved zinc oxide film for gas sensor applications,
Bull. Mater. Sci., Vol. 25, No. 6, pp. 513 – 515, Nov. (2002).
[95] S. M. Chou, L. G. Teoh, W. H. Lai, Y. H. Su and M. H. Hon, ZnO:Al thin
film gas sensor for detection of ethanol vapor, Sensors, 6, 1420 – 1427,
(2006).
[96] P. Mitra and A. K. Mukhopadhyay, ZnO thin film as methane sensor, Bulle-
tin of the Polish Academy of Science, Technical Sciences Vol. 55, No. 3,
(2007).
[97] V. M. Aroutiounian, V. M. Arakelyan, V. E. Galstyan, Kh. S. Martirosyan,
and P. G. Soukiassian, Manufacture and investigation of hydrogen sensitive
TiO2-x or ZnO<Al> film – porous silicon devices, Armenian Journal of Phys-
ics, vol. 1, issue 3 pp. 219 – 226, (2008).
[98] M. Szeto, ZnO nanoparticle gas sensors, Department of Physics, Rochester
Institute of Technology, 84 Lomb Memorial Drive, Rochester, NY 14623 –
5603, USA, (2008).
[99] V. E. Galstyan, V. M. Aroutiounian, V. M. Arakelyan, and G. E. Shahnaz-
aryan, Investigation of hydrogen sensor made of ZnO<Al> thin film, Arme-
nian Journal of Physics, Vol. 1, issue 4, pp. 242 – 246, (2008).
[100] G. Kiriakidis, K. Moschovis, I. Kortidis, Gas sensors based on metal oxide
spray pyrolysis films, 1st International Conference on Nanostructured Mate-
rials and Nanocomposites (6 – 8 April, Kottayam, India), (2009).
[101] N. Al-Hardan, M. J. Abdullah, A. Abdul Aziz, The gas response enhance-
ment from ZnO film for H2 gas detection. Appl. Surf. Sci. 255:7794 – 7,
(2009).
[102] N. Tamaekong, C. Liewhiran, A. Wisitsoraat and S. Phanichphant, Sensing
characteristics of flame-spray-made Pt/ZnO thick films as H2 gas sensor,
Sensors 9, pp. 6652 – 6669, (2009).
[103] N. Al-Hardan, M. J. Abdullah, A. Abdul Aziz, H. Ahmad, Low operating
temperature of oxygen gas sensor based on undoped and Cr – doped ZnO
films, Applied Surface Science 256, 3468 – 3471, (2010).
[104] N. H. Al-Hardan, M. J. Abdullah, A. Abdul Aziz, Sensing mechanism of
hydrogen gas sensor based on RF – sputtered ZnO thin films, International
journal of hydrogen energy 35, 4428 – 4434, (2010).
[105] M. Hussain, M. Mazhar, T. Hussain and N.A. Khan, High efficiency ZnO
nano sensor, fabrication and characterization, J. Iran. Chem. Soc., Vol. 7,
Suppl., pp. S59 – S69, (2010).
[106] N. Al-Hardan, M. J. Abdullah, A. Abdul Aziz & H. Ahmad, ZnO gas sensor
for testing vinegar acid concentrations, Sains Malaysiana 40 (1) pp. 67 – 70,
(2011).
[107] Q. G. Al-zaidi, A. M. Suhail, W. R. Al-azawi, Palladium – doped ZnO thin
film hydrogen gas sensor, Applied Physics Research Vol. 3, No. 1, pp. 89 –
99, (2011).
142
[108] P. Hariharan, Basics of Interferometry, second edition, Elsevier Inc. Sydney,
Australia, 2007.
[109] Joint Committee on Powder Diffraction Standards (JCPDS), International
Center for Diffraction Data, Swarthmore, card no. 36 –1451, PA, 1980.
[110] [110] V. Vasu and A. Subrahmanyam, Reaction kinetics of the formation of
indium tin oxide films grown by spray pyrolysis, Thin Solid Films, Volumes
193-194, Part 2, 15, pp. 696 – 703, (1990).
[111] K. V. Gurav, V. G. Fulari, U. M. Patil, C. D. Lokhande, Oh-Shim Joo, Room
temperature soft chemical route for nanofibrous wurtzite ZnO thin film syn-
thesis, Applied Surface Science, 256, pp. 2680 – 2685, (2010).
[112] S. O’Brien, M. G. Nolan, M. Copuroglu, J. Hamilton, I. Povey, L. Pereira,
R. Martins, E. Fortunato, M. Pemble, Zinc oxide thin films: characterization
and potential applications, Thin Solid Films 518, 4515 – 4519, (2010).
[113] J. P. Borah and K. C. Sarma, Optical and optoelectronic properties of ZnS
nanostructured thin film, ACTA PHYSICA POLONICA A, Vol. 114, No. 4,
(2008).
[114] S. N. Das, J. P. Kar, Ji-Hyuk Choi, T. Il Lee, Kyeong-Ju Moon, and Jae-Min
Myoung, Fabrication and characterization of ZnO single nanowire-based
hydrogen sensor, J. Phys. Chem. C 114, pp. 1689–1693, (2010).
[115] Y. Abdul M. Ismail, M. N. Md Yusuf, W. N. W. Shamsuri , Y. Wahab and
Z. Othaman, The effect of thickness on ZnO thin film CO gas sensor, De-
partment of Physics , Faculty of Science , University of Technology Malay-
sia, The Proceedings of Regional Annual Fundamental Science Seminar
(RAFSS), pp. 89 – 94, (2006).
[116] S. Basu and P. K. Basu, Nanocrystalline metal oxides for methane sensors:
role of noble metals, Journal of Sensors, Vol. 2009, pp. 1 – 20, (2009).
[117] P. Mitra and S. Mondal, Hydrogen and LPG sensing properties of SnO2
films obtained by direct oxidation of SILAR deposited SnS, Bulletin of the
Polish Academy of Sciences Technical Sciences, Vol. 56, No. 3, (2008)
[118] Z. Chen, D. Pan, B. Zhao, G. Ding, Z. Jiao, M. Wu, Chan-Hung Shek, Law-
rence C. M. Wu, and Joseph K. L. Lai, Insight on fractal assessment strate-
gies for tin dioxide thin films, American Chemical Society, ACSNANO,
Vol. 4, No. 2, pp. 1202 – 1208, (2010).
[119] Joint Committee on Powder Diffraction Standards (JCPDS), International
Center for Diffraction Data, Swarthmore, card no. 36 – 1451, PA, 1980.
[120] S. Mishra, C. Ghanshyam, N. Ram, S. Singh, R. P. Bajpai And R. K. Bedi,
Alcohol sensing of tin oxide thin film prepared by sol – gel process, Bull.
Mater. Sci., Vol. 25, No. 3, pp. 231– 234, (2002).
I
راقــــــة العـــــجمهوري
يــــث العلمـــي والبحــــم العالــــوزارة التعلي
داد ــــــــــة بغـــــــــجامع
وم ـــــــــة العلــــــــــــكلي
الزنك وأوكسيد القصدير متحسس أوكسيد حسينت
لغـاز الهايدروجين
أطروحة مقدمة إلى
داد ـة بغـجامع –وم ـكلية العلمجلس
ة ــــوراه فلسفـة دكتـل درجـات نيـزء من متطلبـوهي ج
اء ـفي الفيزي
من قبل
ـالـع حيـقحطــان كـاطـ
9114بكالوريوس علوم في الفيزياء
9111ر علوم في الفيزياء ـــــماجستي
رافــــأش
صالـحوسـن رشيـــد د.
لــن سهيــدهللا محسـعبد.
ميالدي 1199 هجري9441
II
الخالصــــــه
المطعم بالباليديوم على قواعد وأوكسيد القصدير ZnOتم تحضير اغشيه نانويه ألوكسيد الزنك
الكيميائي الحراري وتم تفحصها كمتحسس سريع االستجابه لغاز لتحللزجاجيه باستخدام طريقه ا
450الكيميائي الحراري بدرجه حراره حوالي تحللالهايدروجين المختزل. أستخدمت تقنية ال
وكسيدوأ والهواء الجوي كغاز حامل لتحضير األغشيه المتحسسه ألوكسيد الزنك درجه مئويه
وأسيتات ZnCl2التي طبقت هي كل من أمالح كلوريد الزنك تذريه. كانت محاليل الالقصدير
SnCl2.2H2Oبينما أستخدمت ماده كلوريد القصدير Zn(CH3COO)22H2Oالزنك
بساطتها ومعوليتها في الكيميائي الحراري تحللأثبتت تقنيه ال. للحصول على أوكسيد القصدير
cوبأتجاه على امتداد المحور (002)ل على أغشيه رقيقه متعدده التبلور وذات طور والحص
لتركيب أوكسيد الزنك السداسي كما اظهرته تحليالت التركيب لحيود األشعه السينيه. أظهرت
األغشيه المحضره نفاذيه عاليه عند المدى المرئي للطيف الكهرومغناطيسي وبمعدل وصل الى
نانومترتقريبا. 481وعتبة قطع عند األشعه فوق البنفسجيه ذات الطول الموجي %95حوالي
فجوة الطاقه المباشره المحسوبه لألغشيه بنقصان سمك الغشاء الرقيق. ليس نفاذيه وازدادت ال
القوه الذريه ألغشيه جهرااللكتروني الماسح وم جهرأظهرت نتائج الدراسات السطحيه بالم
اوكسيد الزنك تشكل توزيع منتظم لحبيبات مساميه نانويه التركيب دائريه الشكل ذات اقطار
. بينت الخصائص الكهربائيه لألغشيه الرقيقه المحضره بهذه التقنيه مقاوميتها نانومتر 11بحدود
العاليه وان هذه الخصائص تغيرت تغيرا ملحوظا عند أجراء القياسات في الهواء اوفي الفراغ.
المطعم تحسسيه فائقه حيث أزدادت SnO2و ZnOسا اوكسيد المعدن متحسسكل من ابدى
لغاز الهايدروجين ذا تراكيز متنوعه وعند درجات حراريه مختلفه. المواصله له بتعرضه
لغاز الهيدروجين يتعلق بتحسن امتزاز االوكسجين معادنسيد الاأكمتحسس وجد أن تحسس
الجوي. االنتقائيه الممتازه والتحسسيه العاليه لغاز الهيدروجين يمكن تحقيقها بالمعامله السطحيه
. لقد كان مقدار التغير بالمواصله ألوكسيد الزنك المطعم /أوكسيد القصديرألغشيه اوكسيد الزنك
% حوالي مرتان بقدر مثيلتها للغشاء الغير 4 بالباليديوم بتعرضه لغاز الهايدروجين ذا تركيز
مطعم.ال
أدى تغيير درجة الحرارة التي يعمل عندها متحسس أوكسيد الزنك الى تغير ملحوظ في حساسيته
250يجة الحراره المثلى لألشتغال حوالكانت درحيث ± ت بعدها ضأنخف درجه مئويه 25
حساسية المتحسس. كان تغير التحسسيه لغشاء أوكسيد الزنك خطيا بزياده تركيز الغاز.
III
. نسبيارقصنه مفرط الكواألفاقه لمادة أوكسيد الزنك المطعم بالباليديوم ب –زمن األستجابه يتميز
% وزمن 11كلوريد الباليديوم اعلى تحسسيه قدرها في مره 11لقد اظهرت األغشيه المغطسه
ثانيه وهذا مالئم عمليا طالما أنه يعتبر من االمور الحاسمه الحصول 4استجابه مفرط القصر
على متحسس غازي سريع االستجابه والتحسس له القدره على كشف الغازات الملتهبه والسامه
4راكيز قليله دون الحد االدنى ألنفجار هذه الغازات )لغاز الهيدروجين يكون هذا الحد وعند ت
.)%
درجه مئويه 191ولعناصر التحسس ألوكسيد القصدير فقد كانت درجه حرارة االشتغال المثلى
%. 4.5 لغاز هيدروجين بتركيز قدره % 14.144وبنسبه مئويه للتحسس قدرها